ENCYCLOPEDIA OF

ENVIRONMENTAL
MICROBIOLOGY
VOLUMES 1 - 6
ENCYCLOPEDIA OF ENVIRONMENTAL MICROBIOLOGY

Editor-in-Chief James A. Nienow

Gabriel Bitton Valdosta State University
University of Florida Kate Scow
University of California

Editorial Board Robert J. Seviour

David L. Balkwill La Trobe University
Florida State University Linda D. Stetzenbach
Robert S. Burlage University of Nevada
Oak Ridge National Laboratory Mic H. Stewart
Douglas Capone Metro Water District of Southern California
University of Southern California David C. White
Thomas L. Crisman University of Tennessee
University of Florida
Scott E. Dowd
Editorial Staff
University of Arizona
Executive Publisher: Janet Bailey
Hans-Curt Flemming
University of Duisburg Publisher: Paula Kepos
Charles P. Gerba Executive Editor: Jacqueline I. Kroschwitz
University of Arizona
Senior Associate Managing Editor: Shirley Thomas
Mark W. Le Chevallier
American Water Works Service Company Associate Managing Editor: Kellsee Chu
Andrew Leis Assistant Managing Editor: Laurie Claret
University of Duisburg
Editorial Assistant: Surlan Murrell
Eugene Madsen
Cornell University
ENCYCLOPEDIA OF

ENVIRONMENTAL
MICROBIOLOGY
VOLUMES 1 - 6
Gabriel Bitton
University of Florida
Gainesville, Florida

A Wiley-Interscience Publication
John Wiley & Sons, Inc.
New York / Chichester / Weinheim / Brisbane / Singapore / Toronto
This book is printed on acid-free paper. ®
Copyright © 2002 by John Wiley & Sons, Inc., New York. All rights reserved.
Published simultaneously in Canada.
No part of this publication may be reproduced, stored in a retrieval system or transmitted in any
form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise,
except as permitted under Sections 107 or 108 of the 1976 United States Copyright Act, without
either the prior written permission of the Publisher, or authorization through payment of the
appropriate per-copy fee to the Copyright Clearance Center, 222 Rosewood Drive, Danvers,
MA 01923, (978) 750-8400, fax (978) 750-4744. Requests to the Publisher for permission should be
addressed to the Permissions Department, John Wiley & Sons, Inc., 605 Third Avenue, New York,
NY 10158-0012, (212) 850-6011, fax (212) 850-6008, E-Mail: [email protected].
For ordering and customer service, call 1-800-CALL-WILEY.
Library of Congress Cataloging in Publication Data:
Bitton, Gabriel.
Encyclopedia of environmental microbiology / Gabriel Bitton.
v.<I->.cm.
Includes bibliographical references and index.
ISBN 0-471-35450-3 (set: cloth : alk.paper) — ISBN 0-471-36046-5 (v. 1) —ISBN
0-471-36047-3 (v. 2) —ISBN 0-471-36048-1 (v. 3) —ISBN 0-471-36049-X (v. 4) —
ISBN 0-471-36050-3 (v. 5) —ISBN 0-471-36051-1 (v. 6)
1. Microbial ecology — Encyclopedias. 2. Sanitary microbiology—Encyclopedias. I.
Title.
QR100.B58 2001
579'.17'03 —dc21 2001026911
Printed in the United States of America.
10 9 8 7 6 5 4 3 2 1
 TABLE OF CONTENTS

A Algal Turf Scrubbing: Potential Use for Wastewater

Treatment
Alkaline Enzymes
Acetylene Reduction Assay
Alkaliphiles
Acid Mine Drainage
Alkaliphiles: Alkaline Enzymes and Their Applications
Acidophiles
Alkalithermophiles (Anaerobic)
Actinomycetes, Airborne
Allergens, Fungal
Actinomycetes in Activated Sludge Ammonification
Actinomycetes in Soils Amoeboid Protozoa
Actinomycetes: Role in Foaming of Activated Sludge Anaerobic Digestion of Biosolids
Actinorhizal Symbiosis Anaerobic Granules
Activated Carbon Anaerobic Granules and Granulation Processes
Activated Sludge, Filamentous Bacteria in Anammox
Activated Sludge — Foaming Anhydrobiosis
Activated Sludge, Methodology Anthrax
Activated Sludge — Microbiology of Nitrogen Removal Antibiotic Resistance
Activated Sludge Models: Microbiological Basis Anydrobiosis
Activated Sludge — Molecular Techniques for Aquifers, Atlantic Coastal Plain
Determining Community Composition Arbuscular Mycorrhizae
Activated Sludge — Sequencing Batch Reactors Archaea, Detection of
Activated Sludge—The Floe Archaea: Detection Methods
Activated Sludge —The "G-Bacteria" Archaea in Biotechnology
Activated Sludge — The Microbial Community Archaea in Marine Environments
Activated Sludge—The Process Archaea in Soil Habitats
Activated Sludge: Use of Molecular Tools Archaea in Soils
Activity and Carbon Transformations in Biofilms Archaea, Thermophilic
Adenoviruses Assessing Microbial Proteomes in the Environment
Adhesion, Immobilization and Retention of Assimilable Organic Carbon (AOC) in Drinking Water
Microorganisms on Solid Substrata Assimilable Organic Carbon (AOC) in Treated Water:
Adhesion (primary) of Microorganisms onto Surfaces Determination and Significance
Adhesion to Surfaces Astroviruses
Adsorption of Microorganisms to Surfaces Attachment of Microorganisms
Adsorption of Viruses to Surfaces
Aerobic Alkaliphiles B
Aerobic Endospores
Aerobic Respiration, Principles of Bacteria: Streams
Aggregates and Consortia, Microbial Bacterial Cell Structure
Agroterrorism Bacterial Contaminants in Residential Environments
Air Pollution Control Bacterial Pathogens in Waste Stabilization Ponds
Air Sampling Bacterial Phytostimulators in the Rhizosphere
Air-Water Interface Bacterioneuston
Airborne Bacterial Pathogens Bacteriophage as Indicators
Airborne Fungi Bacteriophage: Biology and Genetics
Airborne Microorganisms Bacteriophage Detection Methodologies
Airborne Toxigenic Molds Bacteriophage of Enteric Bacteria: Occurrence and
Algae Persistance in the Environment
Algae Biotechnology B acterioplankton
Algae, Blue-Green Benthic Algae
Algae: Eutrophication Benthic-Associated Primary Productivity
Algae, Polar Bioaerosol Sampling and Analysis
Algae: Streams Bioaerosols in Agricultural and Outdoor Settings
Algae: Toxicity Testing Bioaerosols in Industrial Settings
Algae, Use as Biological Indicators in Paleolimnology Bioaerosols: Modeling
Algae: Waste Stabilization Ponds Bioaerosols: Transport and Fate
Algal Blooms Bioaerosols, Virus
Algal Blooms — Impact on Treatment, Taste, and Odor Bioaerosols: Wastewater Treatment Plants
Problems Bioaugmentation
Algal Pigments as Indicators Bioaugmentation: Cold-Adapted Microbes
v
vi TABLE OF CONTENTS

Bioavailability of Organic Substrates Biodegradation: Wetlands

Biochip-Based Devices and Methods in Microbial Biodeterioration
Community Ribotyping Biodeterioration of Mineral Materials
Biocontrol, Microbial Agents in Soil Biodiversity in Soils: Use of Molecular Methods for its
Biocorrosion: Role of Sulfate Reducing Bacteria Characterization
Biodegradability: Methods for Assessing Biodegradability Bioemulsifiers
Under Laboratory and Field Condition Biofilm Detachment
Biodegradable Dissolved Organic Carbon (BDOC) Biofilms, Activity in
Biodegradable Dissolved Organic Carbon in Drinking Biofilms, Algal Biofilms
Water Biofilms: Bacterial-Fungal Biofilms
Biodegradation: Composts Biofilms, Carbon Transformations in
Biodegradation: Fuel Oxygenates Biofilms, Conditioning Films
Biodegradation, Halogenated Aromatics Biofilms: Extracellular Enzymes
Biodegradation: Landfills Biofilms, Extracellular Polymeric Substance
Biodegradation: Oxygenase Enzymes Biofilms, Formation Rate in Water Distribution Systems
Biodegradation: Reductive Dehalogenation and Biofilms, Growth Kinetics of Microbes in
Metabolism of Chlorinated Organics by Anaerobes Biofilms in Natural and Drinking Water Systems
 TABLE OF CONTENTS

B Biotechnological Applications of Biosurfactants and

Bioemulsifiers
Biotechnology, Fungi in
Biofilms in the Food Industry
Biotechnology, Methanotrophs in
Biofilms, Methodology
Bioterrorism
Biofilms, Modeling of
Biotoxins
Biofilms, Pathogens in
Biotrickling Filters for Air Pollution Control
Biofilms, Sorption Properties of
Blue-Green Algae
Biofilms: Suspended Biofilms Borreliosis, Lyme
Biofiltration Bottled Water
Biofiltration and Bioodors Bottled Water, Microbiology of
Biofouling: Chemical Control of Biofouling in Water Brown Rot Fungi
Systems Brown Tide
Biofouling in the Marine Environment Bulking Control
Biofouling of Industrial Systems Bulking in Activated Sludge
Biogenic Trace Gases Bulking of Activated Sludge
Biogeochemical Cycles
Biohydrometallurgy
Bioleaching C
Bioleaching of Metals
Biological Control, Use of Biosurfactants in Calici viruses
Biological Warfare Campylobacter jejuni and Other Enteric Campylobacter
Capillary Electrophoresis in Genetic Analysis and
Biology of Cryptosporidium
Ribotyping of Microbiota in the Environment
Bioluminescence, Methodology
Carbon Monoxide
Biomarkers, Lipid
Caves and Mines Microbiological Sampling
Biomass, Algal
Caves and Other Low-Light Environments: Aerophitic
Biomass, Bacterioplankton Photoautotrophic Microorganisms
Biomass: Soil Microbial Biomass Cell Viability, Methodology for
Biomineralization by Bacteria Chaperone and Chaperonins
Biomining Chemical Weapons, Biodegradation of
Biomining Using Archaea Chemolithoautotrophy
Bioodors Control in Waste Treatment Facilities Cholera
Biopesticides Chrysophytes
Bioplastics Ciliated Protozoa
Bioremediation: An Overview of How Microbiological Ciliates in Freshwater Ecosystems
Processes Can Be Applied to the Cleanup of Organic Cisterns
and Inorganic Environmental Pollutants Climate Change, Use of Chrysophytes as Indicators of
Bioremediation: Aquatic Ecosystems Cloning
Bioremediation, Aquifers Clostridium
Bioremediation, Biosurfactants Use in Coagulation in Water Treatment Plants: Pathogen
Bioremediation, Cold-Adapted Microorganisms in Removal
Bioremediation, Methanotrophs in Cold-Active Enzymes
Bioremediation of Hot Desert Soils Cold-Adapted Microorganisms: Adaptation Strategies
Bioremediation of Soils and Biotechnological Potential
Bioremediation, Pesticides in Soils Cold Deserts
Bioremediation, PLFA Analysis Cold-Induced Proteins
Bioreporters Cold Shock
Biosolids: Anaerobic Digestion of Cold-Shock Proteins
Biosolids, Bioaerosols from Coliform Bacteria as Indicators of Water Quality
Biosolids, Granules in Coliform Bacteria — Control in Drinking Water
Biosorbents for Metals Distribution Systems
Biostability of Drinking Water Coliform Regrowth in Water Distribution Systems
Biostimulation, Aquatic Ecosystems Coliphage: Detection Methodologies
Biostimulation, Aquifers Cometabolism
Biostimulation: Cold-Adapted Microbes Commercial Use of Microorganisms, Regulation of
Biostimulation, Hot Desert Soils Compost: Biodegradation of Toxic Organic Compounds
Biosurfactants: Types, Screening Methods, and Conditioning Films in Aquatic Environments
Applications Confocal Laser Scanning Microscopy (CLSM)
v
vi TABLE OF CONTENTS

Consortia, Microbial Disinfection, Effect on Viruses

Controlling the Microbial Quality of Drinking Water in Disinfection of Legionellae
Distribution Systems Disinfection of Protozoan Parasites
Corrosion Disinfection, Ultraviolet (UV)
Cosmetics, Use of Biosurfactants in Dna Chips
Crude Oil, Biodegradation of Domestic Microbiology
Crude Oil Spills Drinking Water Biofilms
Culturable Subsurface Microbial Communities Drinking Water, Cyanobacteria in
Cyanobacteria Drinking Water Distribution Systems, Invertebrates and
Cyanobacteria in Soils Protozoa in
Cyanobacteria-Toxins in Drinking Water Drinking Water Distribution Systems, Pathogens in
Cyanobacterial Mats Drinking Water, Nitrifying Bacteria in
Cyanobacterial Mats and Nitrogenen Fixation Drinking Water, Sulfur Bacteria in
Cyanophages Drinking Water, Viruses in
Cyclospora: Basic Biology, Occurrence Fate and
Methodologies
E
D

Data Analysis and Modeling Ecological Significance of Subsurface Microorganisms

Dehalogenation of Haloorganics Ecology of Marine Microbial Biofilms
Denaturing Gradient Gel Electrophoresis (DGGE) Ecology, Pathogenicity, and Systematics ofAeromonas in
Denitrification the Environment
Denitrification in Activated Sludge Ectomycorrhizal Fungi
Denitrification in the Marine Environment Electrophoresis
Desert Environments Emergency Water Supplies
Desert Environments: Biological Soil Crusts Endolithic Microorganisms in Arid Regions
Desert Environments — Soil Microbial Communities in Endosymbiosis in Ecology and Evolution
Cold Deserts Endosymbiosis Involving Amebas
Desiccation, Adaptation of Microorganisms to Endosymbiosis Involving Cyanobacteria
Desiccation by Exposure to Space Vacuum or Extremely Endotoxins in Bottled Water
Dry Deserts: Effect on Microorganisms Enhanced Biological Phosphorus Removal (EBPR) in
Desiccation of Microorganisms Activated Sludge
Desulfurization of Fossil Fuels Enhanced Detection of Airborne Microbial Contaminants
Detection of Airborne Microorganisms Entamoeba histolytica /Entamoeba dispar
Detection of Enteroviruses Enteric Adenoviruses
Diatoms in Biofilms Enterococci, Fecal
Diazotrophs Enterohemorrhagic E. Coli (EHEC)
Dinitrogenase Enteroinvasive E. Coli
Dinoflagellates, Toxic Enteropathogenic E. Coli (EPEC)
Dioxygenases Enterotoxigenic E. Coli (ETEC)
Disinfection: Chlorine, Monochloramine, and Chlorine Enteroviruses: basic Biology and Diseases
Dioxide Enteroviruses in Water: Concentration and Detection
 TABLE OF CONTENTS

E Food Contamination
Food Industry
Enteroviruses: Occurrence and Persistence in the Food Preservation
Environment Food Processing and Bioaerosols
Entomopathogens (Bacteria, Fungi, Prototozoa, Viruses) Fossil Fuels Desulfurization
Environmental Genomics Free-Living Amebas Present in the Environment Can
Environmental Release of Microorganisms Cause Meningoencephalitis in Humans and Other
Enzymatic Toxicity Tests Animals
Enzyme Inhibition Freeze-Drying
Enzyme Kinetics Freeze Drying: Preservations of Microorganisms by
Enzymes, Archaeal Freeze-Drying
Enzymes: Biotechnological Applications Fuel Additives, Biodegradation of
Enzymes, Cold-Active Fumonisins
Enzymes, Industrial Fungal Allergy and Allergens
Enzymes in Soils Fungal Biofilms
Enzymes: Oxygenases Fungal Contaminants
Epifluorescence Microscopy Fungi
Epilithic Microorganisms Fungi and Indoor Air
EPS Fungi and Pollutant Biodegradation
Eutrophication and Algae Fungi, Culture Media for
Evolution of Metabolic Pathways for Degradation of Fungi, for Biotechnology
Environmental Pollutants Fungi in Freshwater Ecosystems
Exobiology Fungi in Marine/Estuarine Waters
Extracellular Enzymes in Biofilms Fungi in Soils
Extracellular Polymeric Substances (EPS) Fungi in Streams
Extracellular Polymeric Substances (EPS): Structural,
Ecological and Technical Aspects
G
Extremophiles: Life in Extreme Environments
G-Bacteria in Activated Sludge
F Gallionella ferruginea: An Iron-Oxidizing and
Stalk-Forming Groundwater Bacterium
Fate and Microbial Degradation of Halogenated Gases, Trace
Aromatics Gene Chips
Fate of Viruses and Protozoan Parasites in Aquatic Gene Exchange in Biofilms
Sediments Gene Probes
Fecal Contamination, Sources of Gene Transfer in Biofilms
Fecal Streptococci/Enterococci in Aquatic Environments Genetic Ecology of Soil
Field Release of Genetically Engineered Microorganisms Genetic Manipulation of Fungi
(GEM) Genetically Engineered Microorganisms (GEMs)
Filamentous Bacteria in Activated Sludge: Current Genetically Engineered Microorganisms for
Taxonomic Status and Ecology Biodegradation of Recalcitrant Compounds
Filamentous Bulking in Activated Sludge, Control of Genetically Modified Microorganisms (GMM) in Soil
Filtration (High-Rate): Removal of Pathogenic Environments
Microorganisms Genomics, Environmental
Filtration: Occurrence of Protozoa in Spent Filter Geochemical and Geological Significance of Subsurface
Backwash Water Microbiology
Fingerprinting Techniques Geochemical Aspects of Subsurface Microbiology
Flagellated Protozoa Geostatistics for Determining the Distribution of Soil
Flocs, Microbial Microorganisms
Flooded Soils Giardia: Basic Biology, Genetics and Epidemiology
Fluorescence Microscopy Giardia: Detection and Occurrence of in the Environment
Fluorescent in situ Hybridization (Fish) Granular Activated Carbon, Bacteriology of
Fluorescent in situ Hybridization (Fish): Use in Activated Granular High-Rate Filtration: Removal of Pathogenic
Sludge Microorganisms
Fluorescent Probes for in situ Analyses of Microbial Granular Sludge
Communities Green Fluorescent Protein (GFP)
Fluorochromes Greenhouse Effect, Role of Microorganisms
Foaming in Activated Sludge Groundwater, Protistan Communities in
v
vi TABLE OF CONTENTS

Groundwater Sampling I
Groundwater, Sulfur Bacteria in
Groundwater, Viruses in Ice Environments
Growth, Diauxic Ice Microbial Communities
Growth, on Mixed Substrates Identification of Airborne Fungi
Igneous Rock Aquifers Microbial Communities
Image Analysis of Microorganisms
H
Imaging
Immobilization of Microorganisms on Solid Substrata
Haber-Bosch Process Indicators of Water Quality: Aerobic Spores
Halogenated Compounds, Bioremediation of Indicators of Water Quality: Bacteriophage
Halogenated Hydrocarbon (Gases) Indicators of Water Quality: Clostridium
Haloorganics Dehalogenation Indicators of Water Quality: Coliforms
Halophiles Indicators of Water Quality: Fecal
Halophiles: Aerobic Halophilic Microorganisms Streptococci/Enterococci
Halophiles: Anaerobic Prokaryotes from Hypersaline Indicators of Water Quality: Source Water Protection
Environments Indoor Exposure to Fungi
Halophilic Archaea Indoor Health
Halophilic Microorganisms Industrial Enzymes
Halotolerant Bacteria Industrial Settings, Bioaerosols in
Harmful Algal Blooms (HAB) Infectious Airborne Bacteria
Heat-Shock Genes Inorganic Nutrient Use by Marine Microorganisms
Heat-Shock Proteins Insecticides, Microbial
Heavy Metal Toxicity Invertebrate-Associated Microorganisms in Deep-Sea
Helicobacter pylori Hydrothermal Vents
Hepatitis Viruses (HAV-HEV) Invertebrates and Protozoa (Free-Living) in Drinking
Heterotrophic Bacteria as Indicators of Water Quality Water Distribution Systems
High Hydrostatic Pressure: Microbial Inactivation and Ion Exhange in Water Treatment Plants: Pathogen
Food Preservation Removal
Home Treatment Devices — Microbiology of Point of Use Ionizing Radiation, Effect on Microorganisms
and Point of Entry Devices Iron Cycling
Hot Desert Soil Microbial Communities Iron Oxidation
Hot Tubs Microbiology Iron-Reducing Microbes in Petroleum Reservoirs
Human Caliciviruses: Basic Virology and Epidemiology Isospora
Humic Substances in Aquatic Environments Isotopic Fractionation
Hydrocarbon Biodegradation in Cold Environments
Hydrocarbons, Biodegradation of
Hydrocarbons, Bioremediation of
K
Hydrodesulfurization
Hydrophobicity of Microorganisms: Methodology
Hydrothermal Vents: Biodiversity in Deep-Sea Kinetics (Microbial): Theory and Applications
Hydrothermal Vents Kinetics of Microbial Processes and Population Growth in
Hydrothermal Vents: Prokaryotes in Deep-Sea Soil
Hydrothermal Vents
Hyperthermophiles
 TABLE OF CONTENTS

L Methanotrophs
Methods for Flow Cytometry and Cell Sorting
Labelling of Microbial Cells (Enzymatic, Fluorescent, Methods for the Identification of Microbial Isolates
Immunological, Phylogenetic, and Physiological Methylation of Metals
Labels) Microarrays
Lakes, Periphyton in Microarrays: Applications in Environmental Microbiology
Landfilling of Municipal Solid Wastes: Microbiological Microbial Degradation of Explosives
Processes and Environmental Impacts Microbial Degradation of Fuel Oxygenates
Laser Scanning Microscopy in Combination With Microbial Enhanced Oil Recovery (MEOR), Use of and
Fluorescence Techniques for Biofilm Study Biosurfactants in
Leaf Decomposition By Fungi in Freshwater Ecosystems Microbial Flocs Suspended Biofilms
Legionellae Microbially-Influenced Corrosion (MIC)
Legionella in the Environment: Persistence, Evolution, Microbial Removal by Pretreatment, coagulation and Ion
and Pathogenicity Exchange
Legume Inoculation Microbial Starvation Survival in Subsurface
Leptospirosis Environments
Lipid Biomarkers in Environmental Microbiology Microbial Toxicity Tests
Lipids, Archaeal Microbiology of Atlantic Coastal Plain Aquifers and
Lithotrophic Microbial Ecosystems in the Subsurface Other Unconsolidated Subsurface Sediments
Low Light Environments Microbiology of Cretaceous Shales and Sandstones of the
Luciferase and Green Fluorescent Protein as Bioreporters Southwestern United States
in Microbial Systems Microbiology of Deep High Temperature Sedimentary
Lyme Borreliosis Environments
Microchips
M Microgravity Effects on Microorganisms
Microorganisms in Soil: Factors Influencing Their
Machining and Bioaerosols Activity
Macrophyte Colonization by Fungi Microspheres as Tracers in Groundwater
Magnetotactic Bacteria Microsporidia: Basic Biology
Manganese Cycling Microsporidia: Occurrence, Fate and Methodologies
Manganese Oxidation Mine Tailings
Marine Biotechnology Mines
Marine Methanogens Mixotrophs
Mats, Microbial Modeling of Biofilms
Membrane Filter Procedure for Heterotrophic Bacteria Modeling of Virus Transport and Removal in the
Mercury Cycling Subsurface
Meroplankton Modeling the Transport of Bioaerosols
Messenger RNA-Targeted Probes Modeling
Metabolic Pathways Models, Activated Sludge
Metabolism of Mixtures of Organic Pollutants Models: Role of Protozoa
Metals Molecular Methods for Soils
Metals, Bioremediation of Monooxygenases
Metals in Soils Mustard Gas, Biodegradation of
Metals: Interaction with Biofilms Mycobacterium avium Complex
Metals, Microbe Resistance to Mycorrhizae
Metals: Microbial Processes Affecting Metals Mycorrhizae: Arbuscular Mycorrhizae
Metal Stressed Environments, Bacteria in Mycorrhizae: Ectomycorrhizal Fungi
Metal Transformations Mycotoxins
Metal (U,Fe,Mn,Hg) Cycling
Methane N
Methanogenesis
Methanogenesis in the Marine Environment Natural Organic Matter (NOM) in Aquatic Environments;
Methanogens Neural Networks
Methanogens: Biotechnological Applications Neuston Microbiology: Life at the Air–Water Interface
Methanogens in Petroleum Reservoirs Nitrification
Methanogens in Soils Nitrification in Activated Sludge
Methanotrophic Bacteria Nitrification in Aquatic Systems
Methanotrophic Bacteria: Use in Bioremediation Nitrification in Soils
v
vi TABLE OF CONTENTS

Nitrification in The Marine Environment Oil Reservoirs

Nitrifiers Typing using Microchips Oligotrophic Bacteria
Nitrifying Bacteria in Drinking Water Opportunistic Pathogens
Nitogen Cycle Organic Matter in Water
Nitrogen Cycle in Aquatic Systems Organic Matter Removal in Water Treatment
Nitrogen Cycle in Soils Organic Substances: Interaction with Biofilms
Nitrogen Cycle in the Marine Environment Organophosphorus Nerve Agents, Biodegradation of
Nitogen Fixation Osmoprotection of Microorganisms
Nitrogen Fixation in Methanotrophs Oxygen: Effect on Marine Microbial Communities
Nitrogen Fixation in Soils — Free-Living Microbes Oxygenase Enzymes: Role in Biodegradation
Nitrogen Fixation in Soils (Symbiotic)
Nitrogen Fixation in the Marine Environment
Nitrogen Fixation Measurement P
Nitrogen Gases
Nitrogen in Marine Microorganisms Paleoecology
Nitrogen Removal in Activated Sludge Paleoenvironmental Reconstruction
Norwalk-Like Viruses: Detection Methodologies and Paleolimonology
Environmental Fate Paleolimnology: Subfossil Algae Other than Diatoms and
Norwalk-Like Viruses (NLVS) Chrysophytes
Nosocomial Infections Paleolimnology: Use of Algal Pigments as Indicators
Nuclear Waste Respository in Yucca Mountain: Paleolimnology: Use of Siliceous Structures of
Microbiological Aspects Chrysophytes as Biological Indicators in Freshwater
Nucleic Acids Extraction from Soils Systems
Nutrient Cycling: Role of Bacterioplankton Panspermia Thesis
Nutrients Use By Marine Microorganisms Paralytic Shellfish Poisoning
Parasitic Protozoa: Fate in Wastewater Treatment Plants
Pathogen, Bacterial
O Pathogenic Escherichia coli
Pathogens, Bacterial
Occurrence of Protozoa in Spent Filter Backwash Water Pathogens (Bacterial) in Waste Stabilization Ponds
Ochratoxins Pathogens in Environmental Biofilms
 TABLE OF CONTENTS

P Protozoa (Free-Living) in Drinking Water Distribution

Systems
Pathogen Survival in Aquatic Environmnents Protozoa in Activated Sludge
Periphyton Protozoa in Marine and Estuarine Waters
Permafrost Protozoa in Streams
Pesticide Degradation in Soils Protozoan Ciliates in Freshwater Ecosystems
Pesticides, Bioremediation of Protozoan Parasites
Pest Mangement Protozoan Parasites, Disinfection of
Petroleum and Other Hydrocarbons, Biodegradation of Protozoan Parasites in Food
Petroleum Reservoirs, Influence, Activity and Growth of Protozoan Parasites in Sediments
Subsurface Microflora in Pseudomonas
Petroleum Reservoirs, Microbial Diversity in Psychrophiles
Pfiesteria: The Toxic Pfiesteria Complex Psychrophilic Bacteria
Phage: Detection Methodologies Psychrophilic Bacteria: Isolation and Characterization
Phosphorus Cycling in Aquatic Environments: Role of Psychrophilic Microorganisms
Bacteria Psychrotolerant Microorganisms
Phosphorus in Marine Microorganisms Psychrotrophic Bacteria
Photosynthetic Bacteria Psychrotrophic Microorganisms
Photosynthetic Bacteria in Caves Pulp and Paper Industry: Microbiological Aspects of
Photosynthetic Bacteria in Drinking Water
Photosynthetic Bacteria in Soils
Q
Photosynthetic Bacteria in Waste Stabilization Ponds
Photosynthetic Pigments in Marine Algae and Bacteria
Quality Assurance/Quality Control in Subsurface
Phototrophic Purple and Green Bacteria in Marine and
Sampling and Processing
Hypersaline Environments
Quality, Soil Quality
Phycotoxins
Quantification of Microbial Biomass
Phyllosphere
Quorum Sensing in Biofilms
Phylogenetically Based Methods in Microbial Ecology
Phylogenetics, Archaea
Phylogeny R
Phylogeny of Aquifer Communities
Phytoneuston Radiation Effect on Bacteria
Phytoplankton Radiation Effects on Microorganisms
Phytoplankton, Polar Radioactive Waste Disposal, Geomicrobiology of
Phytostimulators, Bacterial Rainwater Roof Catchment Systems, Microbial Quality of
Pigments of Algae as Paleolimnological Indicators Red Tides and Other Harmful Algal Blooms
Planktonic Algae Reedbeds for Wastewater Treatment
Planktonic Algae in the Marine Environment Regrowth of Bacteria in Water Distribution Systems
Planktonic Microorganisms: Bacterioplankton Regulation of Bottled Water
Plant–Microbe Interactions in the Marine Environment Regulation of the Commercial Uses of Microorganisms
Plfa Analysis Regulations for Drinking Water
Point-of-Use Treatment Devices for Drinking Water Removal of Pathogenic Microbes by Granular High-Rate
Polar Marine Phytoplankton Filtration
Polymerase Chain Reaction (PCR) RESIDENTIAL ENVIRONMENTS Retention of
Polymers in Activated Sludge Microorganisms on Solid Substrata
Population Growth in Soils, Kinetics of Rhizosphere
Preservation of Algae for Toxicity Testing Rhizosphere, Archaea in
Pretreatment in Water Treatment Plants: Pathogen Rhizosphere Microbiology
Removal Rhizospheric Effect
Primary Productivity Ribosomal RNA-Targeted Probes.
Primary Productivity in the Marine Environment Ribotyping
Probes Ribotyping Methods for Assessment of in situ Microbial
Probes: Rrna-Targeted Probes for Activated Sludge Community Structure
Processing of Subsurface Samples Ribotyping of Microbial Communities
Prochlorococcus Rice Agriculture
Productivity, Bacterioplankton Rice Soils
Protection of Source Water Risk Assessment
Protistan Communities in Groundwater Risk Assessment of Environmental Exposure to Viruses
v
vi TABLE OF CONTENTS

Riverine Biofilms Siliceous Structures of Chrysophytes

Rivers, Periphyton in Silicon in Marine Microorganisms
Root Exudates Sludge
Root Secretions Sludge, Granules in
Rotaviruses Snow and Ice Environments
Rural Water Supplies Soil Archaea
Soil Bacteria
Soil Bioremediation (Pesticides)
S
Soilborne Diseases, Biocontrol of
Soil Crusts, Biological
Salinity Effects on the Physiology of Soil Microorganisms
Soil Distribution of Microorganisms
Salinity Tolerance
Soil Enzymes
Salmonella in Aquatic Environments
Soil Formation, Role of Microorganisms
Salt Production
Soil Fungi: Nature’s Nutritional Network
Salt Production, Involvement of Microorganisms in
Soil Genetic Ecology
Sampling in Subsurface Environments
Soil Microbial Biomass
Sampling of Bioaerosols
Soil and Soil Microorganisms
Sampling Techniques for Environmental Microbiology
Screening Chemical Toxicity in Soils Soil Nitrogen Cycle
Seagrasses Communities Soil Quality
Sea Ice Microorganisms Soil Quality: The Role of Microorganisms
Sediments Soil Sampling
Sediment Sampling Soil Slurry Sequencing Batch Reactor (Ss-Sbr)
Sediments and PLFA Analysis Soils, Biodegradation in
Sediments, Archaea in Soils, Biodiversity in
Sediments, Bioremediation of Soils, Distribution of Microorganisms in
Sediments: Nitrification in Soils, Kinetics of Microbial Processes
Sediments, Nitrogen Fixation in Soil Toxicity Tests
Sediments: Sulfate Reduction in Marine Sediments Soils, Virus Survival in
S Gases Solar Power: Disinfection of Cistern Water
S-Layer, Archaeal Solid Wastes
Septic Tank Systems Solid Wastes Microbiology
Sequencing Batch Biofilm Reactor (SBBR) Solvents, Effects on Microorganisms
Sequencing Batch Reactor (SBR) Technology Solvent-Tolerant Bacteria
Shelf Storage of Bottled Water Sorption Properties of Biofilms
Shellfish, Salmonella in Sorting of Microbial Cells.
Shigella Source Water Protection: Microbiology of Source Water
Siderophores Sources of Fecal Contamination
Siderophores in Marine Bacteria Spa and Hot Tub Microbiology
 TABLE OF CONTENTS

S Toxicity Testing in Soil, Use of Microbial and Enzymatic

Tests
Space Microbiology: Effects of Ionizing Radiation on Toxicity Testing in Wastewater Treatment Plants Using
Microorganisms in Space Microorganisms
Space Microbiology: Microgravity and Microorganisms Toxigenic Molds
Space: Survival in Space Toxins, Microbial
Space Vacuum Toxoplasma gondii
Spatial Distribution of Microorganisms in Soils Trace Elements in Marine Microorganisms
Spores As Indicators Trace Gases Soil
Stalk-Forming Bacteria Tracers in Groundwater: Use of Microorganisms and
Starvation of Microorganisms in Subsurface Sediments Microspheres
Statistical Analysis Transport of Bacteria Through Soils
Storage of Environmental Samples Transport of Pathogens in Surface Waters
Storage Polymers: Role in the Ecology of Activated Sludge Trophic State and Algal Community Structure
Stream Microbiology
Streams, Periphyton in
U
Streptococci, Fecal
Stress
Unbalanced Growth Conditions
Stress Genes
Uranium Cycling
Stress Proteins
Use of Cold-Adapted Microorganisms in Biotechnology
Stress Response in Archaea
Use of Microscopic Algae in Toxicity Testing
Stress Response in Bacteria
UV Disinfection — Theory to Practice
Stress Response in Bacteria: Heat Shock
Stromatolites
Subaerial Communities V
Subsurface Microbial Communities: Diversity of
Culturable Microorganisms Vadose Zone Microbiology
Subsurface Microbiology Viable But Not Culturable (Vbnc)
Subsurface Samples: Collection and Processing Viable But Not Culturable (VBNC) Microorganisms
Subsurface Sampling Viable Counts in Environmental Samples
Subsurface Sedimentary Environments Viral Aerosols
Subsurface Sediments Viral Diseases
Sulfate Reducing Bacteria Viral Disinfection
Sulfate-Reducing Bacteria: Environmental and Virus Survival in Soils
Technological Aspects Virus Transport in The Subsurface
Sulfate Reducing Bacteria in Petroleum Reservoirs Virus Transport Though Soils
Sulfate Reduction in Marine Sediments Viruses and Protozoan Parasites in Food, Including
Sulfur Bacteria in Drinking Water Methodology
Sulfur Cycle Viruses: Astroviruses
Sulfur Cycle in Soils Viruses: Caliciviruses
Sulfur Oxidizing Bacteria Viruses, Concentration and Detection
Surfaces Viruses: Enteroviruses
Suspended Biofilms Viruses: Hepatitis Viruses
Sustainable Agriculture Viruses in Drinking Water and Groundwater
Sustainable Agriculture: Role of Microorganisms Viruses in Food
Sustainable Solid Waste Management Viruses in the Marine Environment
Symbiotic Nitrogen Fixation Viruses in Waste Stabilization Ponds
Viruses: Rotaviruses
Volcanic Tuffs: Deep Subsurface Microbiology of
T

Textile Industry and Bioaerosols W

Thermophiles
Thermophiles: Anaerobic Alkalithermophiles Wastewater and Biosolids as Sources of Airborne
Thermophiles, Diversity of Microorganisms
Thermophilic Alkaliphiles Wastewater Stabilization Ponds
Tnt Biodegradation Wastewater Treatment and Bioaerosols
Toxicity of Metals to Microorganisms Wastewater Treatment and Giardia
Toxicity of Organic Solvents to Microorganisms Wastewater Treatment Microbiology
v
vi TABLE OF CONTENTS

Wastewater Treatment, Microbiology of Wetlands, Periphyton in

Wastewater Treatment-Septic Tank Systems White Rot Fungi
Water Activity White Rot Fungi in Bioremediation
Water Distribution Systems Wood Decomposition By Fungi
Water Fungi as Decomposers in Freshwater Ecosystems Wood Degrading Fungi
Water Sampling Wood Processing and Bioaerosols
Water Treatment Plants: Protozoa in Spent Filter
Backwash Water X
Weathering
Weathering, Microbiol
Xerotolerant Microorganisms
Weathering: Mineral Weathering and Microbial
Metabolism
Wetlands Z
Wetlands and Readbeds for Wastewater Treatment
Wetlands: Biodegradation of Organic Pollutants Zoonotic Transmission of Parasites
Wetlands, Fungal Activity in Zoonotic Transmission of Pathogens
 PREFACE

Environmental microbiology was born at the dawn despite their busy schedules. I am thankful and grateful
of the "environmental era" at the beginning of the to all of them for sharing their knowledge with anybody
1970s. Thirty years of maturation have led to an interested in this fascinating field. I am indebted to
exciting and vibrant field that has attracted countless my colleagues and co-editors on the editorial board who
numbers of productive and enthusiastic scientists and have patiently and expertly helped me in orchestrating
students at universities, research centers and government this major endeavor. They have helped tremendously in
agencies around the world. The wealth of environmental selecting suitable authors and in participating in the
microbiology literature has made it imperative to make a review of the manuscripts. Their names and affiliation
pause at this time and summarize our knowledge in an appear on a separate list in this encyclopedia. I am grateful
encyclopedia format. For the sake of organization, we have to my colleagues and my students at the Department of
identified 14 areas within environmental microbiology. Environmental Engineering Sciences for their support and
These areas are discussed in more details in the encouragement.
introductory chapter. The quality of manuscripts is greatly enhanced by
The rising tide of molecular biology has led to the use the participation of expert reviewers. I am indebted
and adaptation of modern molecular techniques along with to the hundreds of reviewers who offered many useful
sophisticated equipment to the study of microorganisms suggestions for improving the manuscripts. Their names
in their environments, especially extreme environments. do not appear in this encyclopedia because many of them
The study of extremophiles has increased our knowledge expressed the desire to remain anonymous. Thank you for
on the limits and origin of life on our planet. a job well done.
Over 420 authors from 25 countries contributed 320 This encyclopedia involved the participation and
entries to this encyclopedia. We have witnessed a small cooperation of several individuals at John Wiley and
reduction in the planned coverage of topics, due to the Sons. I would like to thank the team at the Encyclopedia
inability of some authors to deliver their manuscripts. Department who has worked tirelessly to see the
Despite this slight setback, some on-board authors have successful completion of this work. I thank Glenn Collins
courageously agreed to "plug the holes" by expanding who recruited me and convinced me that the job of
their own contribution to cover a missing topic or by editor-in-chief of the encyclopedia was a worthwhile
authoring another entry. In addition to the print version, endeavor. I extend special thanks to Laurie Claret, the
an expanded version of the encyclopedia will be available assistant managing editor, for her expert assistance as
in the near future on-line to cover those missing topics. she undertook this job without missing a beat, following
Cross-reference titles (orphan entries) or key words have the departure of Glenn Collins. I thank Surlan Murrel,
been included in the encyclopedia to help readers retrieve editorial assistant, for her patience and, along with
a given topic, and an author index is also provided.
Laurie, for shielding me from the massive amount of
The Encyclopedia of Environmental Microbiology will
correspondence and tedious record keeping. I thank
serve as a quick reference work to be used by professors,
them for graciously adapting to my academic lifestyle
undergraduate and graduate students, researchers in
the public and private sectors, research organizations, which sometimes consists of reviewing manuscripts in the
environmental and patent lawyers, and government Luxembourg Garden and sidewalk cafes in Paris.
officials for a quick introduction to a given topic in this I am grateful to Nancy, Julie and Natalie, and all
vast microbiology field. my family and friends, for their love, good wishes, moral
The preparation and completion of this encyclopedia support, and patience with me.
is a complex undertaking that involved the participation
and cooperation of several individuals. The authors are GABRIEL BITTON
the "soldiers" without whom this work would not have Gainesville, Florida
been possible. They contributed their expertise unselfishly October 2001

vii
 CONTRIBUTORS

Morteza Abbaszadegan, Arizona State University, Tempe, Arizona, Christian Blaise, Environment Canada, Montreal, Canada, Use of
Viruses in Drinking Water and Groundwater Microscopic Algae in Toxicity Testing
Maria Angeles Abril, Estacion Experimental del Zaidin, Granada, Adria A. Bodour, The University of Arizona, Tucson, Arizona, Biosurfac-
Spain, Genetically Engineered Microorganisms for Biodegradation of tants: Types, Screening Methods, and Applications
Recalcitrant Compounds Hinrich Bohn, University of Arizona, Tucson, Arizona, Biofiltration
Peter Adriaens, The University of Michigan at Ann Arbor, Ann, Michigan, Christophe J. P. Boonaert, Universite catholique de Louvain, Louvain-
Fate and Microbial Degradation of Halogenated Aromatics la-Neuve, Belgium, Adhesion (primary) of Microorganisms onto Surfaces
David B. Alexander, University of Portland, Portland, Oregon, Soil Juan J. Borrego, University of Malaga, Malaga, Spain, Fecal Strep-
Bacteria tococci/Enterococci in Aquatic Environments; Salmonella in Aquatic
David Allen, University of Washington, Seattle, Washington, Use of Cold- Environments
Adapted Microorganisms in Biotechnology Penelope J. Boston, University of New Mexico, Boulder, Colorado, Caves
Michael F. Allen, University of California, Riverside, California, Mycor- and Mines Microbiological Sampling
rhizae: Arbuscular Mycorrhizae John P. Bowman, University of Tasmania, Hobart, Tasmania, Australia,
Jonas S. Almeida, Medical University South Carolina, Charleston, South Psychrophilic Bacteria: Isolation and Characterization
Carolina, Data Analysis and Modeling S. N. Bradley, Pacific Northwest National Laboratory, Richland, Wash-
Harriet M. Ammann, Office of Environmental Health Assessment ington, Vadose Zone Microbiology
Services, Olympia, Washington, Fungi and Indoor Air Larry E. Brand, University of Miami, Miami, Florida, Planktonic Algae
James W. Ammerman, Rutgers University, New Brunswick, New Jersey, in the Marine Environment
Phosphorus Cycling in Aquatic Environments: Role of Bacteria Robin L. Brigmon, Savannah River Technical Center, Aiken, South
Penny S. Amy, University of Nevada, Las Vegas, Nevada, Methods for the Carolina, Methanotrophic Bacteria: Use in Bioremediation
Identification of Microbial Isolates; Nuclear Waste Respository in Yucca F. J. Brockman, Pacific Northwest National Laboratory, Richland,
Mountain: Microbiological Aspects; Volcanic Tuffs: Deep Subsurface Washington, Vadose Zone Microbiology
Microbiology of Volker S. Brozel, University of Pretoria, Pretoria, South Africa,
Garabed Antranikian, Technical University Hamburg-Harburg, Ham- Biofouling: Chemical Control of Biofouling in Water Systems
burg, Germany, Extremophiles: Life in Extreme Environments Lee A. Bulla Jr., Biological Targets, Tioga, Texas and The University of
David A. C. Aron, University of Southern California, Los Angeles, Texas at Richardson, Dallas, Texas, Insecticides, Microbial
California, Protozoa in Marine and Estuarine Waters Saul Burdman, The Hebrew University of Jerusalem, Rehovot, Israel,
Harish Arora, American Water Works Service Company, Inc., Voorhees, Bacterial Phytostimulators in the Rhizosphere
New Jersey, Occurrence of Protozoa in Spent Filter Backwash Water Joann M. Burkholder, North Carolina State University, Raleigh, North
Nicholas Ashbolt, School of Civil and Environmental Engineering, Carolina, Cyanobacteria; Pfiesteria: The Toxic Pfiesteria Complex
UNSW-Sydney, Australia, Methods for Flow Cytometry and Cell Sorting Robert S. Burlage, Oak Ridge National Laboratory, Oak, Tennessee,
Kwok-Keung Au, Greeley and Hansen, LLP, Chicago, Illinois, Removal of Green Fluorescent Protein (GFP)
Pathogenic Microbes by Granular High-Rate Filtration Gary A. Burlingame, Philadelphia Water Department, Philadelphia,
Steven D. Aust, Utah State University, Logan, Utah, Fungi, for Pennsylvania, Algal Blooms — Impact on Treatment, Taste, and Odor
Biotechnology Problems
Khin Saw Aye Myint, Armed Forces Research Institute of Medical Todd Burnes, University of Minnesota, St. Paul, Minnesota, Pulp and
Sciences (AFRIMS), Bangkok, Thailand, Hepatitis Viruses (HAV-HEV) Paper Industry: Microbiological Aspects of
K. H. Baker, Middletown, Pennsylvania, Bioremediation: Aquatic S. G. Burton, Department of Biochemistry and Microbiology, Graham-
Ecosystems stown, South Africa, Archaea in Biotechnology
David L. Balkwill, Florida State University, Tallahassee, Florida, Sub- Robert K. Bush, University of Wisconsin-Madison, William S. Middleton
surface Microbial Communities: Diversity of Culturable Microorganisms Memorial, Veterans Hospital, Madison, Wisconsin, Fungal Allergy and
Guy Baranton, Institut Pasteur, Paris, France, Leptospirosis; Lyme Allergens
Borreliosis Henk J. Busscher, University ofGroningen, Groningen, The Netherlands,
Andrei L. Barkovskii, Georgia College and State University, Milledgeville, Adhesion, Immobilization and Retention of Microorganisms on Solid
Georgia, Fate and Microbial Degradation of Halogenated Aromatics Substrata; Hydrophobicity of Microorganisms: Methodology
Larry L. Barton, University of New Mexico, Albuquerque, New Mexico, David Butler, Imperial College, London, United Kingdom, Septic Tank
Sulfate-Reducing Bacteria: Environmental and Technological Aspects Systems
Catherine Bass, University of Exeter, Devon, United Kingdom, Petroleum Mark P. Buttner, University of Nevada, Las Vegas, Nevada, Bioaerosol
Reservoirs, Influence, Activity and Growth of Subsurface Microflora in Sampling and Analysis
Christa Baumstark-Khan, German Aerospace Center DLR, Institute of D. K. Button, University of Alaska, Fairbanks, Alaska, Kinetics (Micro-
Aerospace Medicine, Cologne, Germany, Space Microbiology: Effects of bial): Theory and Applications
Ionizing Radiation on Microorganisms in Space Bruce A. Caldwell, PNW Research Station, Corvallis, Oregon, Mycor-
L. A. Beaudette, Tsinghua University, Beijing, China, Bioremediation of rhizae: Ectomycorrhizal Fungi
Soils Vitaliano A. Cam a, Lima, Peru, Cyclospora: Basic Biology, Occurrence
Iwona B. Beech, University of Portsmouth, Portsmouth, United Kingdom, Fate and Methodologies
Biocorrosion: Role of Sulfate Reducing Bacteria Michael D. Cameron, Utah State University, Logan, Utah, Fungi, for
Lawrence J. Bell-Perkins, University of Surrey, Guildford, United Biotechnology
Kingdom, Rhizosphere Microbiology James R. Campbell, U.S. Naval Medical Research Unit No. 2 (U.S.
Paul S. Berger, U.S. Environmental Protection Agency, Washington, DC, NAMRU-2), Jakarta, Indonesia, Hepatitis Viruses (HAV-HEV)
Source Water Protection: Microbiology of Source Water Lisa Campbell, Texas A&M University, College, Texas, Prochlorococcus
Costanzo Bertoldo, Technical University Hamburg-Harburg, Hamburg, Anne K. Camper, Montana State University, Bozeman, Montana,
Germany, Extremophiles: Life in Extreme Environments Granular Activated Carbon, Bacteriology of
Terry J. Beveridge, University of Guelph, Guelph, Ontario, Canada, Mehmet Candas, Biological Targets, Tioga, Texas and The University of
Biomineralization by Bacteria Texas at Richardson, Dallas, Texas, Insecticides, Microbial
Gabriel Bitton, Department of Environmental Engineering Sciences, Uni- Douglas G. Capone, University of Southern California, Los Angeles,
versity of Florida, Gainesville, Florida, Toxicity Testing in Wastewater California, Nitrogen Fixation in the Marine Environment; Nitrogen
Treatment Plants Using Microorganisms Cycle in the Marine Environment
ix
x CONTRIBUTORS

M. Carmen Ronchel, Estacion Experimental del Zaidin, Granada, Patricia Cruz, University of Nevada, Las Vegas, Nevada, Identification of
Spain, Genetically Engineered Microorganisms for Biodegradation of Airborne Fungi
Recalcitrant Compounds Carolyn Currin, NOAA National Ocean Service Center, Beaufort, North
M. B. Cassidy, University of Guelph, Guelph, Ontario, Canada, Bioreme- Carolina, Seagrasses Communities
diation of Soils Elisa D'Angelo, University of Kentucky, Lexington, Kentucky, Wetlands:
Michael A. Castellano, PNW Research Station, Corvallis, Oregon, Biodegradation of Organic Pollutants
Mycorrhizae: Ectomycorrhizal Fungi Holger Daims, Technische Universitdt Miinchen, Freising, Germany,
Dolores Castro, University of Malaga, Malaga, Spain, Fecal Strep- Activated Sludge—Molecular Techniques for Determining Community
tococci/Enterococci in Aquatic Environments; Salmonella in Aquatic Composition
Environments; Soil Genetic Ecology Greg A. Davis, Rockford, Tennessee, Sampling Techniques for Environ-
Antonella Cattaneo, Universite de Montreal, Montreal, Quebec, Periphy- mental Microbiology
ton Joseph S. Davis, The University of Florida, Gainesville, Florida, Salt
Christian Chauret, Indiana University, Kokomo, Indiana, Disinfection Production
of Protozoan Parasites Maricruz Alvarez De Mejia, American Embassy I Guatemala, Miami,
Wilfred Chen, University of California, Riverside, California, Chemical Florida, Entamoeba histolytica I Entamoeba dispar
Weapons, Biodegradation of Ana Maria P. De Merida, American Embassy I Guatemala, Miami,
Jing Cheng, Beijing National Biochip Research and Engineering Center, Florida, Entamoeba histolytica I Entamoeba dispar
Beijing, China, Biochip-Based Devices and Methods in Microbial Daniel Deere, Sydney Catchment Authority, Penrith, Australia, Methods
Community Ribotyping for Flow Cytometry and Cell Sorting
Leonid Chernin, The Hebrew University of Jerusalem, Rehovot, Israel, Mary F. Deflaun, Envirogen, Inc., Lawrenceville, New Jersey, Bioaugmen-
Biocontrol, Microbial Agents in Soil tation
Joel R. Cherry, Novozymes Biotech, Inc., Davis, California, Enzymes:
Joseph Defrank, United States Army Edgewood Chemical Biological
Biotechnological Applications
Center, Aberdeen Proving Ground, Maryland, Chemical Weapons,
T. Chesnot, UMR Universite-CNRS 7564, Nancy, France, Parasitic Biodegradation of
Protozoa: Fate in Wastewater Treatment Plants
Jody Deming, University of Washington, Seattle, Washington, Use of
Ilan Chet, The Hebrew University of Jerusalem, Rehovot, Israel, Biocontrol, Cold-Adapted Microorganisms in Biotechnology
Microbial Agents in Soil
Ming Qi Deng, University of Minnesota, St. Paul, Minnesota, Viruses and
Derek E. Chitwood, University of California, Riverside, California, Protozoan Parasites in Food, Including Methodology
Biofiltration and Bioodors
Marc A. Deshusses, University of California, Riverside, California,
Max Ciarlet, Baylor College of Medicine, Houston, Texas, Rotaviruses
Biotrickling Filters for Air Pollution Control
Dean O. Cliver, University of California, Davis, California, Viruses and
Joseph S. Devinny, University of Southern California, Los Angeles,
Protozoan Parasites in Food, Including Methodology
California, Biofiltration and Bioodors
Frederick S. Colwell, Idaho National Engineering and Environmental
Thomas J. Dichristina, Georgia Institute of Technology, Atlanta, Georgia,
Laboratory, Idaho, Idaho, Microbiology of Deep High Temperature
Metal (U,Fe,Mn,Hg) Cycling
Sedimentary Environments
Elizabeth S. Didier, Tulane Regional Primate Research Center, Coving-
Rita Colwell, University of Maryland Biotechnology Institute, Baltimore,
ton, Louisiana, Microsporidia: Basic Biology
Maryland, Cholera
Ralf Conrad, Max-Planck-Institut fiir Terrestrische Mikrobiologie, Mar- Giacomo Ditullio, University of Charleston, Charleston, South Carolina,
burg, Germany, Flooded Soils Photosynthetic Pigments in Marine Algae and Bacteria
Pamela Contag, Xenogen Corporation, Alameda, California, Biolumines- John Doran, USD A-Agricultural Research Service, University of
cence, Methodology Nebraska, Lincoln, Nebraska, Sustainable Agriculture: Role of Microor-
ganisms
Everly Conway De Macario, Wadsworth Center, New York State,
Department of Health, The University at Albany (SUNY), Albany, New John E. Dore, University of Hawaii, Honolulu, Hawaii, Marine Biotech-
York, Stress Response in Archaea; Stress Response in Bacteria: Heat nology
Shock Klaus Dose, Johannes Gutenberg University, Mainz, Germany, Desicca-
Keith E Cooksey, Montana State University, Bozeman, Montana, tion by Exposure to Space Vacuum or Extremely Dry Deserts: Effect on
Diatoms in Biofilms Microorganisms
E. Corre, Institut Universitaire Europeen de la Mer, Plouzane, France, Carlos G. Dosoretz, Technion-Israel Institute of Technology, Haifa, Israel,
Petroleum Reservoirs, Microbial Diversity in Compost: Biodegradation of Toxic Organic Compounds
Daniele Corsaro, Centre Hospitaller Universitaire de Nancy, Vandoeuvre- ScotE. Dowd, USDA-LIRU, Lubbock, Texas, Bacteriophage: Biology
les-Nancy, France, Endosymbiosis in Ecology and Evolution and Genetics; Microsporidia: Occurrence, Fate and Methodologies;
Andrew L. Corwin, U.S. Naval Medical Research Unit No. 2 (U.S. Wastewater and Biosolids as Sources of Airborne Microorganisms
NAMRU-2), Jakarta, Indonesia, Hepatitis Viruses (HAV-HEV) J. P. Dubey, USDA, Animal and Natural Resources Institute, Beltsville,
Peter D. Countway, University of Southern California, Los Angeles, Maryland, Toxoplasma gondii
California, Protozoa in Marine and Estuarine Waters Yves F. Dufrene, Universite catholique de Louvain, Louvain-la-Neuve,
D. A. Cowan, Department of Microbiology, Cape, South Africa, Archaea in Belgium, Adhesion (primary) of Microorganisms onto Surfaces
Biotechnology Peter F. Dunfield, Max-Planck-Institut fiir Terrestrische Mikrobiologie,
Huub H. J. Cox, University of California, Riverside, California, Biotrick- Marburg, Germany, Biodiversity in Soils: Use of Molecular Methods for
ling Filters for Air Pollution Control its Characterization
Rupert Craggs, National Institute of Water and Atmospheric Research, Robert S. Dungan, USDA-ARS, Riverside, California, Metals: Microbial
Hamilton, New Zealand, Algal Turf Scrubbing: Potential Use for Processes Affecting Metals
Wastewater Treatment Laurence Dupont, Universite de Nice-Sophia Antipolis, Nice, France,
Thomas L. Crisman, University of Florida, Gainesville, Florida, Pale- Salinity Effects on the Physiology of Soil Microorganisms
olimnology: Subfossil Algae Other than Diatoms and Chrysophytes; Estrella Duque, Estacion Experimental del Zaidin, Granada, Spain,
Protozoan Ciliates in Freshwater Ecosystems Genetically Engineered Microorganisms for Biodegradation of Recalci-
John H. Crowe, University of California, Davis, California, Freeze trant Compounds
Drying: Preservations of Microorganisms by Freeze-Drying Stephen C. Edberg, Yale University School of Medicine, New Haven,
Lois M. Crowe, University of California, Davis, California, Freeze Drying: Connecticut, Nosocomial Infections
Preservations of Microorganisms by Freeze-Drying Thomas Egli, Swiss Federal Institute for Environmental Science and
David E. Crowley, University of California, Riverside, California, Metals: Technology, Dubendorf, Switzerland, Metabolism of Mixtures of Organic
Microbial Processes Affecting Metals Pollutants
 CONTRIBUTORS xi

Dick Van Elsas, Plant Research International B.V., Wageningen, The Charles P. Gerba, University of Arizona, Tucson, Arizona, Enteroviruses:
Netherlands, Genetically Modified Microorganisms (GMM) in Soil basic Biology and Diseases; Fate of Viruses and Protozoan Parasites in
Environments Aquatic Sediments
L. England, University of Guelph, Guelph, Ontario, Canada, Bioremedia- Charles Gerday, University of Liege, Liege, Belgium, Cold-Adapted
tion of Soils Microorganisms: Adaptation Strategies and Biotechnological Potential
Carlos Enriquez, Clorox Services Company, Pleasanton, California, James J. Germida, University of Saskatchewan, Saskatoon, Canada,
Adenoviruses Sulfur Cycle in Soils
Mary K. Estes, Baylor College of Medicine, Houston, Texas, Rotaviruses Mark O. Gessner, Limnological Research Center, Kastanienbaum,
Abraham Esteve-Nunez, Estacion Experimental del Zaidin, Granada, Switzerland, Water Fungi as Decomposers in Freshwater Ecosystems
Spain, Genetically Engineered Microorganisms for Biodegradation of David Gilichinsky, Russian Academy of Science, Pushchino, Russia,
Recalcitrant Compounds Permafrost
T. Eugene Cloete, University of Pretoria, Pretoria, South Africa, Susan Glasauer, University of Guelph, Guelph, Ontario, Canada,
Biofouling: Chemical Control of Biofouling in Water Systems Biomineralization by Bacteria
Ian R. Falconer, University of Adelaide, Adelaide, Australia, Cyanobacteria- David J. Glass, D. Glass Associates, Inc., Needham, Massachusetts,
Toxins in Drinking Water Regulation of the Commercial Uses of Microorganisms
Per Falholt, NovozymesAIS, Bagsvaerd, Denmark, Enzymes: Biotechno- Alan Godfree, Public Health Section, Warrington, United Kingdom,
logical Applications Clostridium
Joseph O. Falkinham, Virginia Polytechnic Institute and State Univer- Robert M. Goodman, University of Wisconsin-Madison, Madison,
sity, Blacksburg, Virginia, Mycobacterium avium Complex Wisconsin, Archaea in Soil Habitats
Georges Feller, University of Liege, Liege, Belgium, Cold-Adapted Cristina Gomez-Suarez, University of Groningen, Groningen, The
Microorganisms: Adaptation Strategies and Biotechnological Potential Netherlands, Adhesion, Immobilization and Retention of Microorganisms
on Solid Substrata
James G. Ferry, The Pennsylvania State University, State College,
Pennsylvania, Methanogenesis in the Marine Environment SagarM. Goyal, University of Minnesota, St. Paul, Minnesota,
Enteroviruses in Water: Concentration and Detection
Barry S. Fields, Centers for Disease Control and Prevention, Atlanta,
Georgia, Legionellae Rolf Gradinger, University of Alaska, Fairbanks, Alaska, Sea Ice
Microorganisms
Ma Jose Figueras, University of Rovira I Virgili, Reus, Spain, Fecal
Streptococci/Enterococci in Aquatic Environments; Salmonella in David W. Graham, University of Kansas, Lawrence, Kansas, Methan-
Aquatic Environments otrophic Bacteria
Peter Graham, University of Minnesota, St. Paul, Minnesota, Nitrogen
Kai W. Finster, University of Aarhus, Aarhus, Denmark, Oxygen: Effect
Fixation in Soils (Symbiotic)
on Marine Microbial Communities
Linda C. Green, Tulane Regional Primate Research Center, Covington,
Hans-Curt Flemming, Gerhard Mercator Universitdt, Duisburg, Ger-
Louisiana, Microsporidia: Basic Biology
many, Activity and Carbon Transformations in Biofilms; Biofouling of
Industrial Systems; Extracellular Polymeric Substances (EPS): Struc- Ralf Grote, Technical University Hamburg-Harburg, Hamburg, Germany,
tural, Ecological and Technical Aspects; Sorption Properties of Biofilms Extremophiles: Life in Extreme Environments
Claire M. Fraser, The Institute for Genomic Research, Rockville, Piero Guilizzoni, CNR Istituto Italiano di Idrobiologia, Verbania, Italy,
Paleolimnology: Use of Algal Pigments as Indicators
Maryland, Genomics, Environmental
M. Habash, University of Guelph, Guelph, Ontario, Canada, Bioremedia-
James K. Fredrickson, Pacific Northwest National Laboratory, Richland,
tion of Soils
Washington, Ecological Significance of Subsurface Microorganisms
Yitzhak Hadar, Hebrew University, Rehovot, Israel, Compost: Biodegra-
Peter Frenzel, Max-Planck-Institut fur Terrestrische Mikrobiologie,
dation of Toxic Organic Compounds
Marburg, Germany, Flooded Soils
John D. Haddock, Southern Illinois University, Carbondale, Illinois,
E. I. Friedmann, NASA-Ames Research Center, Moffett, California,
Oxygenase Enzymes: Role in Biodegradation
Endolithic Microorganisms in Arid Regions
Annamaria Halasz, Biotechnology Research Institute, Quebec, Canada,
Christian H. Fritsen, Desert Research Institute, Reno, Nevada, Snow and Methods for the Identification of Microbial Isolates; Microbial Degra-
Ice Environments dation of Explosives; Nuclear Waste Respository in Yucca Mountain:
Jed Fuhrman, University of Southern California, Los Angeles, California, Microbiological Aspects; Volcanic Tuffs: Deep Subsurface Microbiology of
Viruses in the Marine Environment Lotta Hallbeck, Goteborg University, Goteborg, Sweden, Gallionella fer-
Roger S. Fujioka, University of Hawaii, Honolulu, Hawaii, Rainwater ruginea: An Iron-Oxidizing and Stalk-Forming Groundwater Bacterium
Roof Catchment Systems, Microbial Quality of Omar S. Harb, University of Pennsylvania, Philadelphia, Pennsylvania,
Yoshihiro Fujiwara, Japan Marine Science and Technology Center, Legionella in the Environment: Persistence, Evolution, and Pathogenic-
Yokosuka, Japan, Hydrothermal Vents: Biodiversity in Deep-Sea ity
Hydrothermal Vents Hauke Harms, Swiss Federal Institute of Technology, Lausanne,
C. Gantzer, Faculte de Pharmacie, Nancy, France, Enteroviruses: Switzerland, Tracers in Groundwater: Use of Microorganisms and
Occurrence and Persistence in the Environment Microspheres
Jun Gao, North Carolina State University, Raleigh, North Carolina, Ronald W. Harvey, U.S. Geological Survey, Boulder, Colorado, Tracers
Hyperthermophiles in Groundwater: Use of Microorganisms and Microspheres
Jean-Louis Garcia, Universites de Provence et de la Mediterranee, Majid Hassanizadeh, Delft University of Technology, Delft, The
Marseille, France, Halophiles: Anaerobic Prokaryotes from Hypersaline Netherlands, Modeling of Virus Transport and Removal in the Subsurface
Environments Jalal Hawari, Biotechnology Research Institute, Quebec, Canada,
Ferran Garcia-Pichel, Arizona State University, Tempe, Arizona, Desert Microbial Degradation of Explosives
Environments: Biological Soil Crusts John F. Heidelberg, The Institute for Genomic Research, Rockville,
Mark Gauci, Biotechfrontiers, North Ryde BC, Australia, Methods for Maryland, Genomics, Environmental
Flow Cytometry and Cell Sorting Hermann J. Heipieper, Centre for Environmental Research (UFZ),
Christine Gaylarde, UFRGS, Porto, Brazil, Biodeterioration of Mineral Leipzig, Germany, Toxicity of Organic Solvents to Microorganisms
Materials John E. Herrmann, University of Massachusetts Medical School,
Mark E. Geesey, University of Charleston, Charleston, South Carolina, Worcester, Massachusetts, Astroviruses
Photosynthetic Pigments in Marine Algae and Bacteria Mark E. Hines, University of Alaska-Anchorage, Anchorage, Alaska,
Tilman Gehrke, University of Hamburg, Hamburg, Germany, Bioleaching Sediments: Sulfate Reduction in Marine Sediments
Edwin E. Geldreich, U.S. Environmental Protection Agency, Washington Lucien Hoffmann, University of Liege, Liege, Belgium, Caves and Other
D.C., Coliform Bacteria as Indicators of Water Quality; Controlling Low-Light Environments: Aerophitic Photoautotrophic Microorganisms
the Microbial Quality of Drinking Water in Distribution Systems; Osmund Holm-Hansen, University of California-San Diego, La Jolla,
Heterotrophic Bacteria as Indicators of Water Quality California, Polar Marine Phytoplankton
xii CONTRIBUTORS

Wim Hoogenboezem, PWN Water Supply North-Holland, Bloemendaal, Ann Kennedy, USD A-Agricultural Research Service, Washington State
The Netherlands, Invertebrates and Protozoa (Free-Living) in Drinking University, Pullman, Washington, Sustainable Agriculture: Role of
Water Distribution Systems Microorganisms
Koki Horikoshi, Kanagawa and Toyo University, Kawagoe, Japan, Thomas L. Kieft, New Mexico Institute of Mining and Technology, Socorro,
Alkaliphiles: Alkaline Enzymes and Their Applications New Mexico, Hot Desert Soil Microbial Communities; Microbial Starva-
Gerda Horneck, German Aerospace Center DLR, Institute of Aerospace tion Survival in Subsurface Environments; Vadose Zone Microbiology
Medicine, Cologne, Germany, Space Microbiology: Effects of Ionizing Hyung J. Kim, University of Kansas, Lawrence, Kansas, Methanotrophic
Radiation on Microorganisms in Space Bacteria
Thomas Horton, PNW Research Station, Corvallis, Oregon, Mycorrhizae: Nancy E. Kinner, University of New Hampshire, Durham, New
Ectomycorrhizal Fungi Hampshire, Protistan Communities in Groundwater
William R. Horwath, University of California, Davis, California, David L. Kirchman, University of Delaware, Lewes, Delaware, Inorganic
Biomass: Soil Microbial Biomass Nutrient Use by Marine Microorganisms
Jonathan Hosier, University of Mississippi Medical Center, Jackson, J. L. Kirk, University ofGuelph, Guelph, Ontario, Canada, Bioremediation
Mississippi, Aerobic Respiration, Principles of of Soils
Debra E. Huffman, University of South Florida, St. Petersburg, Florida, David M. Klaus, University of Colorado, Boulder, Colorado, Space
Biology of Cryptosporidium Microbiology: Microgravity and Microorganisms
Anwar Huq, University of Maryland Biotechnology Institute, Baltimore, Robert E. Klein, American Embassy I Guatemala, Miami, Florida,
Maryland, Cholera Entamoeba histolytica I Entamoeba dispar
Adrienne Huston, University of Washington, Seattle, Washington, Use of Arthur L. Koch, Indiana University, Bloomington, Indiana, Viable But
Cold-Adapted Microorganisms in Biotechnology Not Culturable (VBNC) Microorganisms; Oligotrophic Bacteria
Geert Huys, University of Ghent, Ghent, Belgium, Ecology, Pathogenicity, Dick Van Der Kooij, Kiwa Water Research, Nieuwegein, The Netherlands,
and Systematics ofAeromonas in the Environment Assimilable Organic Carbon (AOC) in Treated Water: Determination and
Johannes F. Imhoff, Institut fur Meereskunde an der Universitdt Kiel, Significance; Invertebrates and Protozoa (Free-Living) in Drinking Water
Kiel, Germany, Phototrophic Purple and Green Bacteria in Marine and Distribution Systems
Hypersaline Environments Darren R. Korber, University of Saskatchewan, Saskatoon, Saskatchewan,
Masayori Inouye, Robert Wood Johnson Medical School, Piscataway, Canada, Image Analysis of Microorganisms
New Jersey, Cold Shock Anton Korenevsky, University of Guelph, Guelph, Ontario, Canada,
R. L. Irvine, University of Notre Dame, Notre, Indiana, Activated Biomineralization by Bacteria
Sludge — Sequencing Batch Reactors Marion Koster, Institut fur Okologie der Ernst-Moritz-Arndt-Universitat
Sonja Isken, Wageningen University, Wageningen, The Netherlands, Greifswald, Kloster I Hiddensee, Germany, Ecology of Marine Microbial
Toxicity of Organic Solvents to Microorganisms Biofilms
Masahiro Ito, Toyo University, Gunma, Japan, Aerobic Alkaliphiles George A. Kowalchuk, Netherlands Institute of Ecology, Heteren, The
Yoshikazu Izumi, Tottori University, Koyama-Minami, Japan, Desulfur- Netherlands, Ribotyping Methods for Assessment of in situ Microbial
ization of Fossil Fuels Community Structure
Karl-Erich Jaeger, Ruhr-Universitdt Bochum, Bochum, Germany, Lee R. Krumholz, The University of Oklahoma, Norman, Oklahoma,
Extracellular Enzymes in Biofilms Microbiology of Cretaceous Shales and Sandstones of the Southwestern
Bruce B. Jarvis, University of Maryland, College, Maryland, Airborne United States
Toxigenic Molds Jorma Kuparinen, Finnish Institute of Marine Research, Helsinki,
C. Jeanthon, Institut Universitaire Europeen de la Mer, Plouzane, France, Finland, Aggregates and Consortia, Microbial
Petroleum Reservoirs, Microbial Diversity in Yousef Abu Kwaik, University of Kentucky, Lexington, Kentucky,
Mike M. S. Jetten, University Nijmegen, Toernooiveld, The Netherlands, Legionella in the Environment: Persistence, Evolution, and Pathogenic-
Activated Sludge—Microbiology of Nitrogen Removal ity
Juan Jofre, University of Barcelona, Barcelona, Spain, Bacteriophage as Sarah E. Lambert, University of Southern California, Los Angeles,
Indicators California, Protozoa in Marine and Estuarine Waters
Eckardt Johanning, Mt. Sinai School of Medicine, New, Fungi and Indoor Andrea Lami, CNR Istituto Italiano di Idrobiologia, Verbania, Italy,
Air Paleolimnology: Use of Algal Pigments as Indicators
SamanthaB. Joye, The University of Georgia, Athens, Georgia, Hilary M. Lappin-Scott, University of Exeter, Exeter, United Kingdom,
Denitrification in the Marine Environment Biofilms: Bacterial-Fungal Biofilms; Petroleum Reservoirs, Influence,
Peter-Georg Jozsa, University of Hamburg, Hamburg, Germany, Activity and Growth of Subsurface Microflora in
Bioleaching; Weathering, Microbiol Chrysi S. Laspidou, Northwestern University, Evanston, Illinois, Biofilm
Edouard Jurkevitch, The Hebrew University of Jerusalem, Rehovot, Detachment
Israel, Bacterial Phytostimulators in the Rhizosphere John R. Lawrence, National Water Research Institute, Saskatoon,
Robert H. Kadlec, Wetland Management Services, Chelsea, Michigan, Canada, Biofilms in Natural and Drinking Water Systems; Image
Wetlands and Readbeds for Wastewater Treatment Analysis of Microorganisms; Laser Scanning Microscopy in Combination
Daniel Kadouri, The Hebrew University of Jerusalem, Rehovot, Israel, With Fluorescence Techniques for Biofilm Study
Bacterial Phytostimulators in the Rhizosphere MarkW. Lechevallier, American Water Works Service Company,
Masahiro Kamekura, Noda Institute for Scientific Research, Chiba, Voorhees, New Jersey, Occurrence of Protozoa in Spent Filter Backwash
Japan, Halophiles: Aerobic Halophilic Microorganisms Water; Microbial Removal by Pretreatment, Coagulation and Ion
Peter Kampfer, Justus-Liebig-Universitdt Giessen, Giessen, Germany, Exchange; Removal of Pathogenic Microbes by Granular High-Rate
Filamentous Bacteria in Activated Sludge: Current Taxonomic Status Filtration
and Ecology H. Lee, University ofGuelph, Guelph, Ontario, Canada, Bioremediation of
Richard Kanwal, NIOSH, Cincinnati, Ohio, Bioaerosols in Industrial Soils
Settings Lee H. Lee, Montclair State University, Upper Montclair, New Jersey,
Mohammad R. Karim, American Water Works Service Company, Algae Biotechnology
Belleville, Illinois, Fate of Viruses and Protozoan Parasites in Aquatic Laura G. Leff, Kent State University, Kent, Ohio, Stream Microbiology
Sediments Andrew Leis, Gerhard Mercator University, Duisburg, Germany, Activity
Robert M. Kelly, North Carolina State University, Raleigh, North and Carbon Transformations in Biofilms; Conditioning Films in Aquatic
Carolina, Hyperthermophiles Environments; Sorption Properties of Biofilms
M. M. Kendall, Portland State University, Portland, Oregon, Petroleum Estelle Levetin, The University of Tulsa, Tulsa, Oklahoma, Bioaerosols
Reservoirs, Microbial Diversity in in Agricultural and Outdoor Settings
 CONTRIBUTORS xiii

Werner Liesack, Max-Planck-Institut fiir Terrestrische Mikrobiologie, Harris W. Martin, ATC Associates, Burlington, New Jersey, Caves and
Marburg, Germany, Biodiversity in Soils: Use of Molecular Methods for Mines Microbiological Sampling
its Characterization Kenneth F. Martinez, NIOSH, Cincinnati, Ohio, Bioaerosols in Indus-
Nancy I. Lieu, Metropolitan Water District of Southern California, Los trial Settings
Angeles, California, Nitrifying Bacteria in Drinking Water A. M. Maszenan, Biotechnology Research Centre, La Trobe University,
Angela S. Lindner, University of Florida, Gainesville, Florida, Methan- Bendigo, Australia, Activated Sludge—The Microbial Community
otrophic Bacteria A. Matin, Stanford University School of Medicine, Stanford, California,
K. C. Lindrea, Biotechnology Research Centre, La Trobe University, Stress Response in Bacteria
Bendigo, Australia, Activated Sludge—The Process Suzanne M. Matsui, Stanford University School of Medicine, Stanford,
David S. Lindsay, Virginia Tech, Blacksburg, Virginia, Isospora California, Veterans Affairs Palo Alto Health Care System, Palo,
Erin Lipp, University of Maryland Biotechnology Institute, Baltimore, California, Astroviruses
Maryland, Cholera Tiina Mattila-Sandholm, VTT Biotechnology, Espoo, Finland, Biofilms
David A. Lipson, San Diego State University, San Diego, California, in the Food Industry
Kinetics of Microbial Processes and Population Growth in Soil M. Maux, UMR Universite-CNRS 7564, Nancy, France, Parasitic
Steven N. Liss, Ryerson University, Toronto, Ontario, Canada, Microbial Protozoa: Fate in Wastewater Treatment Plants
Floes Suspended Biofilms Ian G. Mckinley, NAGRA, Wettingen, Switzerland, Radioactive Waste
Philip E. Long, Battelle Pacific Northwest National Laboratory, Richland, Disposal, Geomicrobiology of
Washington, Subsurface Samples: Collection and Processing KristinaD. Mena, University of Texas-Houston, El, Texas, Risk
Beatriz Lopez, American Embassy / Guatemala, Miami, Florida, Enta- Assessment of Environmental Exposure to Viruses
moeba histolytica /Entamoeba dispar John S. Meschke, University of North Carolina, Chapel, North Carolina,
Human Caliciviruses: Basic Virology and Epidemiology; Norwalk-Like
Charles R. Lovell, University of South Carolina, Columbia, South
Viruses: Detection Methodologies and Environmental Fate
Carolina, Plant-Microbe Interactions in the Marine Environment
Lutz-Arend Meyer-Reil, Institut fur Okologie der Ernst-Moritz-Arndt-
Alexander Loy, Technische Universitdt Milnchen, Freising, Germany,
Universitdt Greifswald, Kloster /Hiddensee, Germany, Ecology of Marine
Activated Sludge—Molecular Techniques for Determining Community
Microbial Biofilms
Composition
Carmen Michan, Estacion Experimental del Zaidin, Granada, Spain,
Bonnie K. Lustigman, Montclair State University, Upper Montclair, New
Genetically Engineered Microorganisms for Biodegradation of Recalci-
Jersey, Algae Biotechnology
trant Compounds
James M. Lynch, University of Surrey, Guildford, United Kingdom,
Ralph Mitchell, Laboratory of Microbial Ecology Harvard University,
Rhizosphere Microbiology
Cambridge, Massachusetts, Biofouling in the Marine Environment
Xuemei Ma, Beijing National Biochip Research and Engineering Center, Keith R. Mitchelson, University of Queensland, Brisbane, Australia,
Beijing, China, Biochip-Based Devices and Methods in Microbial Capillary Electrophoresis in Genetic Analysis and Ribotyping of
Community Ribotyping Microbiota in the Environment
Alberto J. L. Macario, Wadsworth Center, New York State, Department Randy Molina, PNW Research Station, Corvallis, Oregon, Mycorrhizae:
of Health, The University at Albany (SUNY), Albany, New York, Stress Ectomycorrhizal Fungi
Response in Bacteria: Heat Shock Stress Response in Archaea
Charles M. Moore, Georgia Institute of Technology, Atlanta, Georgia,
Barbara J. Macgregor, Max Planck Institute for Marine Microbiology, Metal (U,Fe,Mn,Hg) Cycling
Bremen, Germany, Phylogenetically Based Methods in Microbial Ecology Thomas B. Moorman, USDA Agricultural Research Service, National
Janet M. Macher, California Department of Health Services, Berkeley, Soil Tilth Laboratory, Ames, Iowa, Soil Distribution of Microorganisms
California, Infectious Airborne Bacteria Glyn Morton, UCLAN, Preston, United Kingdom, Biodeterioration of
Sarah J. Macnaughton, AEA Technology, Oxfordshire, U.K., Mineral Materials
Paolo Madoni, Fluorescent Probes for in situ Analyses of Microbial Duane P. Moser, Princeton University, Princeton, New Jersey, Caves and
Communities, University of Parma, Parma, ItalyProtozoa in Activated Mines Microbiological Sampling
Sludge Patriciade Q. F. Mota, Yale University School of Medicine, New Haven,
Eugene L. Madsen, Cornell University, Ithaca, New York, Biodegradabil- Connecticut, Nosocomial Infections
ity: Methods for Assessing Biodegradability Under Laboratory and Field Ashok Mulchandani, University of California, Riverside, California,
Condition Chemical Weapons, Biodegradation of
Raina M. Maier, The University of Arizona, Tucson, Arizona, Biosurfac- Ellyn Murphy, Pacific Northwest National Laboratory, Richland,
tants: Types, Screening Methods, and Applications Washington, Microbiology of Deep High Temperature Sedimentary
Mauro Majone, University of Rome "La Sapienza", Rome, Italy, Storage Environments
Polymers: Role in the Ecology of Activated Sludge Alison E. Murray, Desert Research Institute, Reno, Nevada, Archaea in
James S. Maki, Marquette University, Milwaukee, Wisconsin, Biofouling Marine Environments
in the Marine Environment; Neuston Microbiology: Life at the Air-Water David D. Myrold, Oregon State University, Corvallis, Oregon, Soil
Interface Nitrogen Cycle
James P. Malley, Jr., University of New Hampshire, Durham, New Glenn E. Nedwin, Novozymes Biotech, Inc., Davis, California, Enzymes:
Hampshire, UV Disinfection—Theory to Practice Biotechnological Applications
Carol A. Mancuso Nichols, Antarctic CRC and School of Agricultural Karen E. Nelson, The Institute for Genomic Research, Rockville,
Science, University of Tasmania, Hobart, Australia, Archaea: Detection Maryland, Genomics, Environmental
Methods Thomas R. Neu, UFZ Centre for Environmental Research, Leipzig-Halle,
Raphi Mandelbaum, LDD Technologies, Petach, Israel, Compost: Germany, Laser Scanning Microscopy in Combination With Fluorescence
Biodegradation of Toxic Organic Compounds Techniques for Biofilm Study
Karine Mandon, Universite de Nice-Sophia Antipolis, Nice, France, Steven Y. Newell, University of Georgia, Sapelo, Georgia, Fungi in
Salinity Effects on the Physiology of Soil Microorganisms Marine/Estuarine Waters
Reiner Mansch, Institut fiir Allgemein Botani, Hamburg, Germany, Peter D. Nichols, CSIRO Marine Research and Antarctic CRC, Hobart,
Weathering, Microbiol Australia, Archaea: Detection Methods
Werner Manz, Technical University of Berlin, Berlin, Germany, Biofilms Per HalkjaeR Nielsen, Department of Environmental Engineering,
in Natural and Drinking Water Systems Aalborg University, Aalborg, Denmark, Activated Sludge — The Floe
Duncan Mara, University of Leeds, Leeds, United Kingdom, Wastewater Eva C. Nieminski, Utah Department of Environmental Quality, Salt Lake
Stabilization Ponds City, Utah, Aerobic Endospores
Rosa Margesin, University of Innsbruck, Innsbruck, Austria, Cold- James A. Nienow, Biology Department, Valdosta State University,
Adapted Microorganisms: Adaptation Strategies and Biotechnological Valdosta, Georgia, Endolithic Microorganisms in Arid Regions; Subaerial
Potential Communities
xiv CONTRIBUTORS

Nena Nwachuku, United States Environmental Protection Agency, Daniel Prieur, I.U.EM. Technopole Bresg-Iroisg, Plouzane, France,
Washington, D.C., Bacteriophage Detection Methodologies Hydrothermal Vents: Prokaryotes in Deep-Sea Hydrothermal Vents
R. Ocampo-Friedmann, NASA-Ames Research Center, Moffett, Califor- Roger C. Prince, ExxonMobil Research and Engineering Co., Annandale,
nia, Endolithic Microorganisms in Arid Regions New Jersey, Petroleum and Other Hydrocarbons, Biodegradation of;
Clifford A. Ochs, University of Mississippi, University, Mississippi, Bioremediation: An Overview of How Microbiological Processes Can
Planktonic Microorganisms: Bacterioplankton Be Applied to the Cleanup of Organic and Inorganic Environmental
Andrew Ogram, University of Florida, Gainesville, Florida, Soil Genetic Pollutants
Ecology A. B. Prisma, Lima, Peru, Cyclospora: Basic Biology, Occurrence Fate and
OladeleA. Ogunseitan, University of California, Irvine, California, Methodologies
Assessing Microbial Proteomes in the Environment Juan L. Ramos, Estacion Experimental del Zaidin, Granada, Spain,
Takashi Ohshiro, Tottori University, Koyama-Minami, Japan, Desulfur- Genetically Engineered Microorganisms for Biodegradation of Recal-
ization of Fossil Fuels citrant Compounds
Yaacov Okon, The Hebrew University of Jerusalem, Rehovot, Israel, Maria-Isabel Ramos-Gonzalez, Estacion Experimental del Zaidin,
Bacterial Phytostimulators in the Rhizosphere Granada, Spain, Genetically Engineered Microorganisms for Biodegra-
dation of Recalcitrant Compounds
Bernard Ollivier, Universites de Provence et de la Mediterranee,
Lutgarde Raskin, University of Illinois, Urbana, Illinois, Anaerobic
Marseille, France, Halophiles: Anaerobic Prokaryotes from Hypersaline
Granules and Granulation Processes
Environments
Bibek Ray, University of Wyoming, Laramie, Wyoming, High Hydrostatic
Tullis Onstott, Princeton University, Princeton, New Jersey, Geochemical
Pressure: Microbial Inactivation and Food Preservation
and Geological Significance of Subsurface Microbiology; Microbiology of
Deep High Temperature Sedimentary Environments N. Scott Reading, Utah State University, Logan, Utah, Fungi, for
Biotechnology
Ynes R. Ortega, University of Georgia, Griffin, Georgia, Cyclospora: Basic
Donald J. Reasoner, United States Environmental Protection Agency,
Biology, Occurrence Fate and Methodologies
Cincinnati, Ohio, Home Treatment Devices — Microbiology of Point of
Robin K. Oshiro, U.S. Environmental Protection Agency, Washington, Use and Point of Entry Devices
DC, Source Water Protection: Microbiology of Source Water
Guenther Reitz, German Aerospace Center DLR, Institute of Aerospace
Leo Van Overbeek, Plant Research International B.V., Wageningen, Medicine, Cologne, Germany, Space Microbiology: Effects of Ionizing
The Netherlands, Genetically Modified Microorganisms (GMM) in Soil Radiation on Microorganisms in Space
Environments
Niels Peter Revsbech, University ofAarhus, Aarhus, Denmark, Oxygen:
Hans W. Paerl, University of North Carolina at Chapel Hill, Morehead, Effect on Marine Microbial Communities
North Carolina, Aggregates and Consortia, Microbial
Anna-Louise Reysenbach, Portland State University, Portland, Oregon,
Robert J. Palmer, Jr., National Institutes of Health, National Institute of Petroleum Reservoirs, Microbial Diversity in; Thermophiles, Diversity of
Dental and Craniofacial Research, Bethesda, Maryland, Luciferase and Eugene W. Rice, U.S. Environmental Protection Agency, Cincinnati,
Green Fluorescent Protein as Bioreporters in Microbial Systems Ohio, Disinfection: Chlorine, Monochloramine, and Chlorine Dioxide;
Anthony V. Palumbo, Oak Ridge National Laboratory, Oak, Tennessee, Helicobacterpylori; Pathogenic Escherichia coli
Microbiology of Atlantic Coastal Plain Aquifers and Other Unconsoli- Steven Ripp, University of Tennessee, Knoxville, Tennessee, Field Release
dated Subsurface Sediments of Genetically Engineered Microorganisms (GEM)
Simon F. Park, University of Surrey, Guildford, United Kingdom, Margaret L. Rising, Portland State University, Portland, Oregon,
Campylobacter jejuni and Other Enteric Campylobacter Thermophiles, Diversity of
Salina Parveen, University of Florida, Gainesville, Florida, Delaware Bruce E. Rittmann, Northwestern University, Evanston, Illinois, Biofilm
State University, Dover, Delaware, Fecal Contamination, Sources of Detachment
Bharat K. C. Patel, Griffith University, Brisbane, Australia, Halophiles: Francisco F. Roberto, Idaho National Engineering and Environmental
Anaerobic Prokaryotes from Hypersaline Environments Laboratory, Idaho, Idaho, Bioleaching of Metals
Ian T. Paulsen, The Institute for Genomic Research, Rockville, Maryland, Sara K. Roberts, University of Exeter, Exeter, United Kingdom, Biofilms:
Genomics, Environmental Bacterial-Fungal Biofilms
Pierre Payment, Universite du Quebec, Laval, Canada, Clostridium Paul A. Rochelle, Metropolitan Water District of Southern California,
Karsten Pedersen, Goteborg University, Goteborg, Sweden, Gallionella La Verne, California, Giardia: Detection and Occurrence of in the
ferruginea: An Iron-Oxidizing and Stalk-Forming Groundwater Bac- Environment
terium; Igneous Rock Aquifers Microbial Communities Thore Rohwerder, Master, Oregon, Bioleaching
Mikael Pell, Swedish University of Agricultural Sciences, Uppsala, Julie M. Rose, University of Southern California, Los Angeles, California,
Sweden, Toxicity Testing in Soil, Use of Microbial and Enzymatic Tests Protozoa in Marine and Estuarine Waters
Jill Peloquin, College of William and Mary Science, Gloucester, Virginia, Fred A. Rosenberg, California Lutheran University, Thousand, Califor-
Primary Productivity in the Marine Environment nia, Bottled Water, Microbiology of
Susan M. Pfiffher, University of Tennessee, Knoxville, Tennessee, Micro- Jon Rosenberg, California Department of Health Services, Berkeley,
biology of Atlantic Coastal Plain Aquifers and Other Unconsolidated California, Infectious Airborne Bacteria
Subsurface Sediments Paul G. Rouxhet, Universite catholique de Louvain, Louvain-la-Neuve,
Sangita Phadtare, Robert Wood Johnson Medical School, Piscataway, Belgium, Adhesion (primary) of Microorganisms onto Surfaces
New Jersey, Cold Shock Daniel Le Rudulier, Universite de Nice-Sophia Antipolis, Nice, France,
Tommy J. Phelps, Oak Ridge National Laboratory, Oak, Tennessee, Salinity Effects on the Physiology of Soil Microorganisms
Microbiology of Atlantic Coastal Plain Aquifers and Other Uncon- Nick J. Russell, Imperial College at Wye, Kent, England, Cold-Adapted
solidated Subsurface Sediments; Subsurface Samples: Collection and Microorganisms: Adaptation Strategies and Biotechnological Potential
Processing Egil Sakshaug, University of Trondheim, Trondheim, Norway, Polar
Edward J. Phlips, University of Florida, Gainesville, Florida, Eutrophi- Marine Phytoplankton
cation and Algae Joseph P. Salanitro, Shell Oil Company, Houston, Texas, Microbial
Yvette M. Piceno, Rockford, Tennessee, Sampling Techniques for Degradation of Fuel Oxygenates
Environmental Microbiology Wolfgang Sand, Institut fur Allgemein Botani, Hamburg, Germany,
Suresh D. Pillai, Texas A&M University, College, Texas, Bacteriophage Bioleaching; Weathering, Microbiol
Detection Methodologies W. T. M. Sanders, Wageningen University, Wageningen, The Netherlands,
Richard M. Plunkett, University of New Mexico, Albuquerque, New Biosolids: Anaerobic Digestion of
Mexico, Sulfate-Reducing Bacteria: Environmental and Technological David Sartory, Severn Trent Water, Shrewsbury, United Kingdom,
Aspects Clostridium
Peter Pollard, Griffith University, Maroochydore, Australia, Wastewater Syed A. Sattar, University of Ottawa, Ottawa, Ontario, Canada, Viral
Treatment Microbiology Aerosols
 CONTRIBUTORS xv

Gary S. Sayler, University of Tennessee, Knoxville, Tennessee, Field Kevin R. Sowers, University of Maryland Biotechnology Institute,
Release of Genetically Engineered Microorganisms (GEM) Baltimore, Maryland, Methanogenesis in the Marine Environment
Rebecca A. Schaffher, University of Southern California, Los Angeles, A. J. M. Stams, Wageningen University, Wageningen, The Netherlands,
California, Protozoa in Marine and Estuarine Waters Biosolids: Anaerobic Digestion of
Jack F. Schij ven, National Institute of Public Health and the Environ- Robert J. Steffan, Envirogen, Inc., Lawrenceville, New Jersey, Bioaug-
ment, Bilthoven, The Netherlands, Modeling of Virus Transport and mentation
Removal in the Subsurface Karen A. Steidinger, Florida Marine Research Institute, St. Petersburg,
Joshua Schimel, University of California at Santa Barbara, Santa, Florida, Red Tides and Other Harmful Algal Blooms
California, Trace Gases Soil John R. Stephen, Horiculture Research International, Wellesbourne,
Claire L. Schleske, University of Florida, Gainesville, Florida, Mero- United Kingdom, Ribotyping Methods for Assessment of in situ Microbial
plankton Community Structure
Ingo Schmidt, Biotechnology Research Centre, La Trobe Univer- Claus Sternberg, Technical University of Denmark, Lyngby, Denmark,
sity, Bendigo, Australia, Activated Sludge — Microbiology of Nitrogen Luciferase and Green Fluorescent Protein as Bioreporters in Microbial
Removal Systems
Steven K. Schmidt, University of Colorado, Boulder, Colorado, Kinetics Linda D. Stetzenbach, University of Nevada, Las Vegas, Nevada,
of Microbial Processes and Population Growth in Soil Enhanced Detection of Airborne Microbial Contaminants
RenE P. Schneider, Universidade de Sao Paulo, Sao Paulo, Brasil,
Todd O. Stevens, Master, Oregon, Lithotrophic Microbial Ecosystems in
Conditioning Films in Aquatic Environments
the Subsurface
F. L. Schuster, State of California Department of Health Services,
Mic H. Stewart, Metropolitan Water District of Southern California, Los
Berkeley, California, Free-Living Amebas Present in the Environment
Angeles, California, Spa and Hot Tub Microbiology
Can Cause Meningoencephalitis in Humans and Other Animals
Frances L. Stites, Dugway Proving Ground, Dugway, Utah, Bioaerosols:
J. Schwartzbrod, UMR Universite-CNRS 7564, Nancy, France, Parasitic
Protozoa: Fate in Wastewater Treatment Plants Transport and Fate
L. Schwartzbrod, Faculte de Pharmacie, Nancy, France, Enteroviruses: Guenther Stotzky, New York University, New, New York, Microorganisms
Occurrence and Persistence in the Environment in Soil: Factors Influencing Their Activity
Elizabeth Scott, Newton, Massachusetts, Bacterial Contaminants in Marc Strous, University Nijmegen, Toernooiveld, The Netherlands,
Residential Environments Activated Sludge — Microbiology of Nitrogen Removal
Ana Segura, Estacion Experimental del Zaidin, Granada, Spain, Genet- Susan D. Sutton, Miami University, Oxford, Ohio, Quantification of
ically Engineered Microorganisms for Biodegradation of Recalcitrant Microbial Biomass
Compounds Jean Swings, University of Ghent, Ghent, Belgium, Ecology, Pathogenic-
Robert J. Seviour, La Trobe University, Bendigo, Australia, Activated ity, and Systematics ofAeromonas in the Environment
Sludge — Microbiology of Nitrogen Removal; Activated Sludge—The "G- Ulrich Szewzyk, Technical University of Berlin, Berlin, Germany, Biofilms
Bacteria"; Activated Sludge — The Microbial Community in Natural and Drinking Water Systems
James P. Shapleigh, Cornell University, Ithaca, New York, Aerobic M. Ali Tabatabai, Iowa State University, Ames, Iowa, Soil Enzymes
Respiration, Principles of Ken Takai, Japan Marine Science and Technology Center, Yokosuka,
Richard Sharp, South Bank University, London, U.K., Fluorescent Probes Japan, Hydrothermal Vents: Biodiversity in Deep-Sea Hydrothermal
for in situ Analyses of Microbial Communities Vents
Timothy J. Sheeran, Department of Defense, Washington, B.C., Bioter- Mino Takashi, The University of Tokyo, Tokyo, Japan, Activated Sludge
rorism Models: Microbiological Basis
Byron C. Shumate, University of Florida, Gainesville, Florida, Paleolim- Mark L. Tamplin, Water Examination Technologies Inc., Gainesville,
nology: Subfossil Algae Other than Diatoms and Chrysophytes Florida, Fecal Contamination, Sources of
Steve D. Siciliano, University of Saskatchewan, Saskatoon, Canada, Valter Tandoi, Water Research Institute, Italian National Research
Sulfur Cycle in Soils Council, Rome, Italy, Storage Polymers: Role in the Ecology of Activated
Holly M. Simon, University of Wisconsin-Madison, Madison, Wisconsin, Sludge
Archaea in Soil Habitats Shengce Tao, Tsinghua University, Beijing, China, Biochip-Based Devices
Peter A. Siver, Connecticut College, New Haven, Connecticut, Paleolimnol- and Methods in Microbial Community Ribotyping
ogy: Use of Siliceous Structures of Chrysophytes as Biological Indicators Dorothea Thompson, Oak Ridge National Laboratory, Oak, Tennessee,
in Freshwater Systems Microarrays: Applications in Environmental Microbiology
Frank Skraly, Metabolix, Inc., Cambridge, Massachusetts, Bioplastics R. Greg Thorn, University of Western Ontario, London, Canada, Soil
Darrell B. Smith, Regional Water Authority, New Haven, Connecticut, Fungi: Nature's Nutritional Network
Coliform Bacteria—Control in Drinking Water Distribution Systems
Jeanette Thurston-Enriquez, USDA—Agricultural Research Service,
Jane E. Smith, PNW Research Station, Corvallis, Oregon, Mycorrhizae: Lincoln, Nebraska, Viral Disinfection
Ectomycorrhizal Fungi
W. Timothy Griffin, Golder Associates Inc., Oak, Tennessee, Subsurface
Jeffrey L. Smith, USD A-Agricultural Research Service, Washington Samples: Collection and Processing
State University, Pullman, Washington, Soil Quality: The Role of
Gary A. Toranzos, University of Puerto Rico, San Juan, Puerto Rico,
Microorganisms
Pseudomonas
Stephen R. Smith, Imperial College, London, United Kingdom, Septic
Tank Systems Alfredo G. Torres, University of Maryland School of Medicine, Baltimore,
Maryland, Shigella
Walker O. Smith JR., College of William and Mary Science, Gloucester,
Virginia, Primary Productivity in the Marine Environment Lennart Torstensson, Swedish University of Agricultural Sciences,
Patricia A. Sobecky, Georgia Institute of Technology, Atlanta, Georgia, Uppsala, Sweden, Toxicity Testing in Soil, Use of Microbial and
Microbiology of Atlantic Coastal Plain Aquifers and Other Unconsoli- Enzymatic Tests
dated Subsurface Sediments J. T. Trevors, University of Guelph, Guelph, Ontario, Canada, Bioreme-
MarkD. Sobsey, University of North Carolina, Chapel, North Carolina, diation of Soils
Human Caliciviruses: Basic Virology and Epidemiology; Norwalk-Like Jean Charles Trinchant, Universite de Nice-Sophia Antipolis, Nice,
Viruses: Detection Methodologies and Environmental Fate France, Salinity Effects on the Physiology of Soil Microorganisms
Jacques Soddell, La Trobe University, Bendigo, Australia, Activated Douglas Trout, NIOSH, Cincinnati, Ohio, NIOSH, Cincinnati, Ohio-
Sludge—Foaming Bioaerosols in Industrial Settings
Guy Soulas, INRA-CMSE Microbiologie des Sols, Dijon-Cedex, France, Anders Tunlid, Lund University, Lund, Sweden, Lipid Biomarkers in
Pesticide Degradation in Soils Environmental Microbiology
Gordon Southam, University of Western Ontario, London, Ontario, Ron Turco, Purdue University, West, Indiana, Soil and Soil Microorgan-
Canada, Metal Stressed Environments, Bacteria in isms
xvi CONTRIBUTORS

William J. Ullman, University of Delaware, Lewes, Delaware, CSIRO Susan A. Welch, University of Wisconsin, Madison, Wisconsin, Weather-
Land and Water, Glen, Australia, Weathering: Mineral Weathering and ing: Mineral Weathering and Microbial Metabolism
Microbial Metabolism Lloyd Wells, University of Washington, Seattle, Washington, Use of Cold-
Richard F. Unz, The Pennsylvania State University, University Park, Adapted Microorganisms in Biotechnology
Pennsylvania, Sulfur Bacteria in Drinking Water Julia M. West, British Geological Survey, Nottingham, United Kingdom,
Jacqueline A. Upcroft, Queensland Institute of Medical Research, Radioactive Waste Disposal, Geomicrobiology of
Brisbane, Australia, Giardia: Basic Biology, Genetics and Epidemiology Juergen Wiegel, University of Georgia, Athens, Georgia, Thermophiles:
Peter Upcroft, Queensland Institute of Medical Research, Brisbane, Anaerobic Alkalithermophiles
Australia, Giardia: Basic Biology, Genetics and Epidemiology Barbara Wigglesworth-Cooksey, Montana State University, Bozeman,
Henny C. Van Der Mei, University of Groningen, Groningen, The Montana, Diatoms in Biofilms
Netherlands, Hydrophobicity of Microorganisms: Methodology; Adhesion, P. A. Wilderer, Technical University of Munich, Garching, Germany,
Immobilization and Retention of Microorganisms on Solid Substrata Activated Sludge — Sequencing Batch Reactors
Jan Roelof Van Der Meer, Swiss Federal Institute for Environmental C. William Keevil, University of Southampton, Southampton, United
Science and Technology (EAWAG), Dubendorf, Switzerland, Evolution of Kingdom, Pathogens in Environmental Biofilms
Metabolic Pathways for Degradation of Environmental Pollutants Jost Wingender, Gerhard-Mercator-Universitdt Duisburg, Duisburg,
J. Hein M. Van Lieverloo, Kiwa Water Research, Nieuwegein, The Germany, Extracellular Enzymes in Biofilms; Extracellular Polymeric
Netherlands, Invertebrates and Protozoa (Free-Living) in Drinking Substances (EPS): Structural, Ecological and Technical Aspects
Water Distribution Systems Gun Wirtanen, VTT Biotechnology, Espoo, Finland, Biofilms in the Food
Gunther Van Ryckegem, Ghent University, Gent, Belgium, Water Fungi Industry
as Decomposers in Freshwater Ecosystems Gideon M. Wolfaardt, University of Stellenbosch, Matieland, South
Carroll Vance, University of Minnesota, St. Paul, Minnesota, Nitrogen Africa, Image Analysis of Microorganisms
Fixation in Soils (Symbiotic) Roy L. Wolfe, Metropolitan Water District of Southern California, Los
Danielle Venditti, TREDI—Departement Recherche, Vandoeuvre-les- Angeles, California, Nitrifying Bacteria in Drinking Water
Nancy, France, Endosymbiosis in Ecology and Evolution Stefan Wuertz, Technical University of Munich, Garching, Germany,
Graham Vesey, Biotechfrontiers, North Ryde BC, Australia, Methods for Gene Exchange in Biofilms
Flow Cytometry and Cell Sorting Chin S. Yang, P & K Microbiology Services, Inc., Cherry, New Jersey,
Helen S. Vishniac, Oklahoma State University, Stillwater, Oklahoma, Fungal Contaminants
Desert Environments — Soil Microbial Communities in Cold Deserts Q.-J. Yao, Princeton University, Princeton, New Jersey, Microbiology of
G. S. Visvesvara, Centers for Disease Control and Prevention, Atlanta, Deep High Temperature Sedimentary Environments
Georgia, Free-Living Amebas Present in the Environment Can Cause MarylynnV. Yates, University of California, Riverside, California,
Meningoencephalitis in Humans and Other Animals Bacteriophage of Enteric Bacteria: Occurrence and Persistance in the
Christian J. Volk, Indiana-American Water Company, Inc., Muncie, Environment; Modeling the Transport of Bioaerosols; Virus Survival in
Indiana, Biodegradable Dissolved Organic Carbon in Drinking Water Soils
Michael Wagner, Technische Universitdt Munchen, Freising, Germany, G. Zeeman, Wageningen University, Wageningen, The Netherlands,
Activated Sludge — Molecular Techniques for Determining Community Biosolids: Anaerobic Digestion of
Composition; Filamentous Bacteria in Activated Sludge: Current Jonathan P. Zehr, University of California, Santa, California, Nitrogen
Taxonomic Status and Ecology Fixation in the Marine Environment
Jiri Wanner, Prague Institute of Chemical Technology, Prague, Czech Dandan Zheng, University of Illinois, Urbana, Illinois, Anaerobic
Republic, Filamentous Bulking in Activated Sludge, Control of Granules and Granulation Processes
Oskar Wanner, Swiss Federal Institute for Environmental Science and Jizhong Zhou, Oak Ridge National Laboratory, Oak, Tennessee,
Technology, Dubendorf, Switzerland, Modeling of Biofilms Microarrays: Applications in Environmental Microbiology
B. B. Ward, Princeton University, Princeton, New Jersey, Nitrification in Stephen H. Zinder, Cornell University, Ithaca, New York, Biodegrada^
Aquatic Systems tion: Reductive Dehalogenation and Metabolism of Chlorinated Organics
Donald F. Ward, North Carolina State University, Raleigh, North by Anaerobes
Carolina, Hyperthermophiles David A. Zuberer, Texas A&M University, College, Texas, Nitrogen
I. A. Watson-Craik, University of Strathclyde, Glasgow, United Kingdom, Fixation in Soils — Free-Living Microbes
Landfilling of Municipal Solid Wastes: Microbiological Processes and
Environmental Impacts
 ABBREVIATIONS USED IN AN OVERVIEW OF
IMAGING SCIENCE

Abs absorption LCD liquid crystal display

AgX silver halide MEG magnetoencephalography
ATC air traffic control MIS multispectral image segmentation
B magnetic MRA magnetic resonance angiography
ft beta particle (electron) MRI magnetic resonance imaging
/3+ positron MW microwave
Bio. biologists n neutron
BW bandwidth v frequency
CCD charge-coupled device NMR nuclear magnetic resonance
Chem chemical obj. objects
CID charge-injected device org. organic
CP charged particle PET positron emission tomography
CRT cathode-ray tube PMT photomultiplier tube
CT computed tomography Polar polarization
DIP digital image processing PIXIE proton-induced X-ray emission
E electric P proton
e~ electron RF radio frequency
EGG electrocardiogram Refl reflection
EE electrical engineers Scint. scintillator
EEG electroencephalogram SIMS secondary ion mass spectroscopy
ESR electron spin resonance SC soft copy
EMG electromyogram Scat. scattering
EMR electromagnetic radiation SQUID superconducting quantum interference device
fMRI functional magnetic resonance imaging SSDA solid-state detector array
FT Fourier transform Surf. surface
0 phase Syst. system
y gamma particle t time
HC hard copy tan tangent
9, t angle, time Tomo. tomographic
hv photon, radiation Tran. transmission
Impe. electrical impedance T2* inhomogeneous nuclear spin-spin relaxation
Intel. intelligent time
IR infrared UV ultraviolet
IS intelligent system x,y two perpendicular spatial dimensions
LC inductive capacitive 2-D two-dimensional

xvii
 CONVERSION FACTORS, ABBREVIATIONS, AND
UNIT SYMBOLS
SI UNITS (Adopted 1960)
The International System of Units (abbreviated SI), is being implemented throughout the world. This measurement
system is a modernized version of the MKSA (meter, kilogram, second, ampere) system, and its details are published and
controlled by an international treaty organization (The International Bureau of Weights and Measures) (1).
SI units are divided into three classes:

BASE UNITS

length meter1^ (m)

mass kilogram (kg)
time second (s)
electric current ampere (A)
thermodynamic temperature* kelvin (K)
amount of substance mole (mol)
luminous intensity candela (cd)

SUPPLEMENTARY UNITS

plane angle radian (rad)

solid angle steradian (sr)

DERIVED UNITS AND OTHER ACCEPTABLE UNITS

These units are formed by combining base units, supplementary units, and other derived units (2-4). Those derived units
having special names and symbols are marked with an asterisk in the list below.

Quantity Unit Symbol Acceptable equivalent

* absorbed dose gray Gy J/kg

acceleration meter per second squared m/s2
* activity (of a radionuclide) becquerel Bq 1/s
area square kilometer km2
square hectometer hm2 ha (hectare)
square meter m2
concentration (of amount of substance) mole per cubic meter mol/m3
current density ampere per square meter A/m2
density, mass density kilogram per cubic meter kg/m3 g/L; mg/cm3
dipole moment (quantity) coulomb meter C-m
*dose equivalent sievert Sv J/kg
* electric capacitance farad F C/V
* electric charge, quantity of electricity coulomb C A-s
electric charge density coulomb per cubic meter C/m3
* electric conductance Siemens S A/V
electric field strength volt per meter V/m
electric flux density coulomb per square meter C/m2
* electric potential, potential difference, volt V W/A
electromotive force
* electric resistance ohm £2 V/A
* energy, work, quantity of heat megajoule MJ

Quantity Unit Symbol Acceptable equivalent

^The spellings "metre" and "litre" are preferred by ASTM; however, "-er" is used in the Encyclopedia.
* Wide use is made of Celsius temperature (t) defined by
t = T - To
where T is the thermodynamic temperature, expressed in kelvin, and TO = 273.15 K by definition. A temperature interval may be
expressed in degrees Celsius as well as in kelvin.
xix
xx CONVERSION FACTORS, ABBREVIATIONS, AND UNIT SYMBOLS

kilojoule kJ
joule J N-m
electronvolt1^ eV1
kilowatt-hour1^ kW-h f
energy density joule per cubic meter J/m3
* force kilonewton kN
newton N kg-m/s2
* frequency megahertz MHz
hertz Hz 1/s
heat capacity, entropy joule per kelvin J/K
heat capacity (specific), specific joule per kilogram kelvin J/(kg-K)
entropy
heat-transfer coefficient watt per square meter W/(m 2 -K)
kelvin
* illuminance lux Ix lm/m2
* inductance henry H Wb/A
linear density kilogram per meter kg/m
luminance candela per square meter cd/m2
* luminous flux lumen 1m cd-sr
magnetic field strength ampere per meter A/m
* magnetic flux weber Wb V-s
*magnetic flux density tesla T Wb/m2
molar energy joule per mole J/mol
molar entropy, molar heat joule per mole kelvin J/(mol-K)
capacity
moment of force, torque newton meter N-m
momentum kilogram meter per kg-m/s
second
permeability henry per meter H/m
permittivity farad per meter F/m
* power, heat flow rate, radiant kilowatt kW
flux
watt W J/s
power density, heat flux watt per square meter W/m2
density, irradiance
*pressure, stress megapascal MPa
kilopascal kPa
pascal Pa N/m2
sound level decibel dB
specific energy joule per kilogram J/kg
specific volume cubic meter per kilogram m3/kg
surface tension newton per meter N/m
thermal conductivity watt per meter kelvin W/(m-K)
velocity meter per second m/s
kilometer per hour km/h
viscosity, dynamic pascal second Pa-s
millipascal second mPa-s
viscosity, kinematic square meter per second m2/s
square millimeter per mm 2 /s
second
volume cubic meter m3
cubic diameter dm3 L (liter) (5)
cubic centimeter cm3 mL
wave number 1 per meter m-1
1 per centimeter cm"1

^This non-Si unit is recognized by the CIPM as having to be retained because of practical importance or use in specialized fields (1).
 CONVERSION FACTORS, ABBREVIATIONS, AND UNIT SYMBOLS xxi

In addition, there are 16 prefixes used to indicate order of magnitude, as follows:

Multiplication Factor Prefix Symbol Note

IO18 exa E
1015 peta P
1012 tera T
109 giga G
106 mega M
103 kilo k
102 hecto ha "Although hecto, deka, deci, and centi are SI prefixes, their use should
10 1 deka daa be avoided except for SI unit-multiples for area and volume and
lo-2 deci da nontechnical use of centimeter, as for body and clothing measurement.
io-3 centi ca
io-6 milli m
io-9 micro M
io- nano n
io-12 pico P
io-15 femto f
io-18 atto a

For a complete description of SI and its use the reader is referred to ASTM E380 (4) and the article UNITS AND CONVERSION
FACTORS which appears in Vol. 24.
A representative list of conversion factors from non-Si to SI units is presented herewith. Factors are given to four
significant figures. Exact relationships are followed by a dagger. A more complete list is given in the latest editions of
ASTM E380 (4) and ANSI Z210.1 (6).

Conversion Factors to SI Units

To convert from To Multiply by

acre square meter (m2) 4.047 x IO3

angstrom meter (m) 1.0 x 10~lot
are square meter (m2) 1.0 x 102t
astronomical unit meter (m) 1.496 x IO11
atmosphere, standard pascal (Pa) 1.013 x IO5
bar pascal (Pa) 1.0 x 105t
barn square meter (m2) 1.0 x 10-28t
barrel (42 U.S. liquid gallons) cubic meter (m3) 0.1590
Bohr magneton OB) J/T 9.274 x IO-24
Btu (International Table) joule (J) 1.055 x IO3
Btu (mean) joule (J) 1.056 x IO3
Btu (thermochemical) joule (J) 1.054 x IO3
bushel cubic meter (m3) 3.524 x IO-2
calorie (International Table) joule (J) 4.187
calorie (mean) joule (J) 4.190
calorie (thermochemical) joule (J) 4.184t
centipoise pascal second (Pa-s) 1.0 x 10~3t
centistokes square millimeter per second (mm2/s) 1.0*
cfm (cubic foot per minute) cubic meter per second (m3/s) 4.72 x 10~4
cubic inch cubic meter (m3) 1.639 x IO-5
cubic foot cubic meter (m3) 2.832 x IO-2
cubic yard cubic meter (m3) 0.7646
curie becquerel (Bq) 3.70 x 10lot
debye coulomb meter (C-m) 3.336 x IO-30
degree (angle) radian (rad) 1.745 x IO-2
denier (international) kilogram per meter (kg/m) 1.111 x IO-7
tex* 0.1111
dram (apothecaries') kilogram (kg) 3.888 x IO-3
dram (avoirdupois) kilogram (kg) 1.772 x IO-3
f
Exact.
*See footnote on p. xxi.
xxii CONVERSION FACTORS, ABBREVIATIONS, AND UNIT SYMBOLS

To convert from To Multiply by

dram (U.S. fluid) cubic meter (m3) 3.697 x 10-6
dyne newton (N) 1.0 x 10~5t
dyne/cm newton per meter (N/m) 1.0 x 10~3t
electronvolt joule (J) 1.602 x 10~19
erg joule (J) 1.0 x 10-7t
fathom meter (m) 1.829
fluid ounce (U.S.) cubic meter (m3) 2.957 x 10~5
foot meter (m) 0.3048f
footcandle lux (Ix) 10.76
furlong meter (m) 2.012 x 10~2
gal meter per second squared (m/s2) 1.0 x 10~2t
gallon (U.S. dry) cubic meter (m3) 4.405 x 10~3
gallon (U.S. liquid) cubic meter (m3) 3.785 x 10-3
gallon per minute (gpm) cubic meter per second (m3/s) 6.309 x 10~5
cubic meter per hour (m3/h) 0.2271
gauss tesla (T) 1.0 x 10~4
gilbert ampere (A) 0.7958
gill (U.S.) cubic meter (m3) 1.183 x 10-4
grade radian, 1.571 x 10-2
grain kilogram (kg) 6.480 x 10-5
gram force per denier newton per tex (N/tex) 8.826 x 10-2
hectare square meter (m2) 1.0 x 104t
horsepower (550 ft-lbf/s) watt (W) 7.457 x 102
horsepower (boiler) watt (W) 9.810 x 103
horsepower (electric) watt (W) 7.46 x 102t
hundredweight (long) kilogram (kg) 50.80
hundredweight (short) kilogram (kg) 45.36
inch meter (m) 2.54 x 10-2t
inch of mercury (32°F) pascal (Pa) 3.386 x 103
inch of water (39.2°F) pascal (Pa) 2.491 x 102
kilogram-force newton (N) 9.807
kilowatt hour megajoule (MJ) 3.6f
kip newton (N) 4.448 x 103
knot (international) meter per second (m/S) 0.5144
lambert candela per square meter (cd/m3) 3.183 x 103
league (British nautical) meter (m) 5.559 x 103
league (statute) meter (m) 4.828 x 103
light year meter (m) 9.461 x 1015
liter (for fluids only) cubic meter (m3) 1.0 x 10-3t
maxwell weber (Wb) 1.0 x 10~8t
micron meter (m) 1.0 x 10-6t
mil meter (m) 2.54 x 10-5t
mile (statute) meter (m) 1.609 x 103
mile (U.S. nautical) meter (m) 1.852 x 103t
mile per hour meter per second (m/s) 0.4470
millibar pascal (Pa) 1.0 x 102
millimeter of mercury (0°C) pascal (Pa) 1.333 x 102t
minute (angular) radian 2.909 x 10~4
myriagram kilogram (kg) 10
myriameter kilometer (km) 10
oersted ampere per meter (A/m) 79.58
ounce (avoirdupois) kilogram (kg) 2.835 x 10~2
ounce (troy) kilogram (kg) 3.110 x 10-2
ounce (U.S. fluid) cubic meter (m3) 2.957 x 10-5
ounce-force newton (N) 0.2780
peck (U.S.) cubic meter (m3) 8.810 x 10-3
pennyweight kilogram (kg) 1.555 x 10-3
pint (U.S. dry) cubic meter (m3) 5.506 x 10~4
 CONVERSION FACTORS, ABBREVIATIONS, AND UNIT SYMBOLS xxiii

pint (U.S. liquid) cubic meter (m3) 4.732 x 10~4

poise (absolute viscosity) pascal second (Pa-s) O.IO1^
pound (avoirdupois) kilogram (kg) 0.4536
pound (troy) kilogram (kg) 0.3732
poundal newton (N) 0.1383
pound-force newton (N) 4.448
pound force per square inch (psi) pascal (Pa) 6.895 x 103
quart (U.S. dry) cubic meter (m3) 1.101 x 10-3
quart (U.S. liquid) cubic meter (m3) 9.464 x 10-4
quintal kilogram (kg) 1.0 x 102t
rad gray (Gy) 1.0 x 10-2t
rod meter (m) 5.029
roentgen coulomb per kilogram (C/kg) 2.58 x 10~4
second (angle) radian (rad) 4.848 x 10~6t
section square meter (m2) 2.590 x 106
slug kilogram (kg) 14.59
spherical candle power lumen (1m) 12.57
square inch square meter (m2) 6.452 x 10-4
square foot square meter (m2) 9.290 x 10-2
square mile square meter (m2) 2.590 x 106
square yard square meter (m2) 0.8361
stere cubic meter (m3) l.Ot
stokes (kinematic viscosity) square meter per second (m2/s) 1.0 x 10-4t
tex kilogram per meter (kg/m) 1.0 x 10-6t
ton (long, 2240 pounds) kilogram (kg) 1.016 x 103
ton (metric) (tonne) kilogram (kg) 1.0 x 103t
ton (short, 2000 pounds) kilogram (kg) 9.072 x 102
torr pascal (Pa) 1.333 x 102
unit pole weber (Wb) 1.257 x 10~7
yard meter (m) 0.9144t

i'Exact.

ABBREVIATIONS AND UNIT SYMBOLS

"When a new discipline such as imaging science evolves from several existing scientific and engineering disciplines, it
is difficult to establish an agreed upon list of abbreviations and unit symbols that will be used by all the disciplines.
Therefore, intead of presenting here a comprehensive list of abbreviations and unit symbols used in this encyclopedia,
each encyclopedia entry has a list of abbreviations and unit sybols used in the entry. Included in this section are a set of
rules for writing unit symbols."

Rules for Writing Unit Symbols (4):

1. Unit symbols are printed in upright letters (roman) regardless of the type style used in the surround text.
2. Unit symbols are unaltered in the plural.
3. Unit symbols are not followed by a period except when used at the end of a sentence.
4. Letter unit symbols are generally printed lower-case (for example, cd for candela) unless the unit name has been
derived from a proper name, in which case the first letter of the symbol is capitalized (W, Pa). Prefixes and unit
symbols retain their prescribed form regardless of the surrounding typography.
5. In the complete expression for a quantity, a space should be left between the numerical value and the unit symbol.
For example, write 2.37 1m, not 2.37 1m, and 35 mm, not 35 mm. When the quantity is used in an adjectival sense,
a hyphen is often used, for example, 35-mm film. Exception: No space is left between the numerical value and the
symbols of degree, minute, and second of plane angle, degree Celsius, and the percent sign.
6. No space is used between the prefix and unit symbol (for example, kg).
7. Symbols, not abbreviations, should be used for units. For example, use "A," not "amp," for ampere.
8. When multiplying unit symbols, use a raised dot:

N-m for newton meter

xxiv CONVERSION FACTORS, ABBREVIATIONS, AND UNIT SYMBOLS

In the case of W-h, the dot may be omitted, thus:

Wh

An exception to this practice is made for computer printouts, automatic typewrier work, etc, where the raised dot
is not possible, and a dot on the line may be used.
9. When dividing unit symbols, use one of the following forms:
m
/ or m-s -i or —
m/s
s

In no case should more than one slash be used in the same expression unless parentheses are inserted to avoid
ambiguity. For example, write:

J/(mol-k) or J-moH-KT 1 or (J/mol)/K

but not
J/mol/K

10. Do not mix symbols and unit names in the same expression. Write:

joules per kilogram or J/kg or J-kg""1

but not
joules/kilogram nor joules/kg nor joules-kg-i

BIBLIOGRAPHY

1. The International Bureau of Weights and Measures, BIPM (Pare de Saint-Cloud, France) is described in Appendix X2 of Ref. 4. This
bureau operates under the exclusive supervision of the International Committee for Weights and Measures (CIPM).
2. Metric Editorial Guide (ANMC-78-1), latest ed., American National Metric Council, 5410 Grosvenor Lane, Bethesda, Md. 20814,1981.
3. SI Units and Recommendations for the Use of Their Multiples and of Certain Other Units (ISO 1000-1981), American National
Standards Institute, 1430 Broadway, New York, 10018, 1981.
4. Based on ASTM E380-89a (Standard Practice for Use of the International System of Units (SI)), American Society for Testing and
Materials, 1916 Race Street, Philadelphia, Pa. 19103, 1989.
5. Fed. Reg., Dec. 10, 1976 (41 FR 36414).
6. For ANSI address, see Ref. 3.
R.P. LUKENS
ASTM Committee E-43 on SI Practice
 INTRODUCTION

Environmental Microbiology encompasses numerous lev- deal with the public health aspects (e.g., allergies)
els of organization ranging from microbial genes to of airborne fungi, especially in indoor environments.
microbial ecosystems. It is the study of the activity of Airborne microbes under agricultural and industrial
indigenous microorganisms in their habitats and their settings are also covered. Other entries deal with models
interactions with other microorganisms or higher organ- to predict the dispersion and fate of microorganisms
isms. It is also the study of the fate of microbial pathogens in the airborne state. A current preoccupation of
and microscopic parasites outside their hosts in natural public health officials around the world is bioterrorism
environments such water, sediments, soils, air, and engi- which concerns the intentional release of pathogenic
neered systems. Environmental microbiology also deals microorganisms in the air (e.g., Bacillus anthracis) for
with the applied aspects of microbiology as regards the harmful purposes.
environment, agriculture, food and water quality, resource The aquatic microbiology area covers various types
recovery, water and wastewater treatment, and human of aquatic environments, including freshwater, estuarine
and animal health. and marine waters, groundwater, wetlands, and aquatic
This Encyclopedia was designed to provide the most sediments. Some of these environments (freshwaters,
comprehensive coverage of the various fields within marine waters) have long been studied by microbiologists
Environmental Microbiology. We have identified 14 areas and limnologists while others (groundwater, wetlands)
within this discipline. These areas are illustrated in have been investigated only during the past 20 years.
Figure 1 (the names of associate editors for each area This vast area necessitated the enlisting of three associate
are indicated in the figure). editors to tackle the task.
Freshwater microbiology entries included some on
IDENTIFIED AREAS IN ENVIRONMENTAL MICROBIOLOGY bacteria, cyanobacteria, fungi, algae and protozoa and
their activities in lakes, rivers, streams and wetlands.
Aeromicrobiology is the study of microorganisms found Paleolimnology is also addressed through the use of
in the air and the public health implications of airborne paleolimnological indicators such as algal pigments and
microbes (fungi, bacteria, viruses), their metabolites remains (e.g., diatoms) to reconstruct past environmental
(e.g., mycotoxins), and cell components. Several entries changes in aquatic environments.

Figure 1.
xxv
XXVI INTRODUCTION

Marine waters represent approximately 97% of the diseases with the subsequent decrease in the yield of
aquatic environment on planet Earth. This explains the agricultural crops. Due to serious concerns for human
attention given to this environment by marine microbi- and animal health and for the environment, attempts are
ologists around the globe. Planktonic microorganisms in being made to use microbial control agents or biopesticides
the oceans and seas serve as food for higher organisms in lieu of chemical pesticides to control plant diseases,
and are the base of the food chain. The marine plankton nematodes or insects. The most successful biopesticide
includes algae, bacteria (eubacteria, cyanobacteria, photo- marketed for insect control is Bacillus thuringiensis
synthetic bacteria), fungi, viruses and protozoa. Algae are which produces a toxin that is lethal to insect pests.
responsible for primary productivity in the water column Unfortunately, soils are becoming the receptacles of a
while bacteria play a crucial role in nutrient (N, P, S) wide range of toxicants, pathogens and parasites as a
and metal cycling in both the water column and the sed- result of human activities which include agricultural
iments. This encyclopedia also covers chapters on archae practices, industrial and mining operations, and disposal
including methanogens, red tides and other harmful algae, of liquid and solid wastes. As regards soil contamination
plant-microorganisms interactions and seagrass micro- by hazardous wastes, current research is addressing
bial communities. The marine environment is the home soil bioremediation which consists of enhancing the
of a wide range of microorganisms, many of which can biodegradation capacity of indigenous soil microorganisms
be tapped for biotechnological applications which include or using genetically engineered microorganisms.
Pharmaceuticals, biomaterials, enzymes, nutritional sup- Advances in molecular genetics have helped launch the
plements and other useful products. field of soil genetic ecology with the goal of studying gene
The goal of subsurface microbiology is the study of exchange, microbial community fingerprinting, analysis
the microbiology of subsurface sediments and formations of functional genes that code for key enzymes, and the
as well as groundwater. Prior to the 1970s, research on this phylogenetic diversity of microorganisms in soils. It is
hidden resource was focused on groundwater supply. In the estimated that there are approximately 10,000 bacterial
1970s, researchers directed their attention to groundwater species in one gram of soils. Culture-based techniques yield
quality, leading to continuing studies on contamination only a very small fraction (<1%) of the total number of
of this precious resource by microbial pathogens and species uncovered by using molecular biology techniques
chemical contaminants. The 1980s saw the focus shifting such as amplified rDNA restriction analysis (ARDRA),
to subsurface sediments with studies on characterization polymerase chain reaction (PCR), denaturing gradient gel
of subsurface microorganisms, their concentrations, and electrophoresis (DGGE) or terminal restriction fragment
their activity. Cultivation techniques and molecular length polymorphism (T-RFLP). Soil genetic ecology may
biology tools showed the existence of a wide range benefit from the use of microarrays (gene chips). All
of bacterial types. The phylogenetic characteristics of of these efforts may lead to some potential industrial
culturable bacteria from diverse subsurface environments applications.
has been undertaken in the U.S., but only a small Dr. Rick Cavicchioli, from the university of New South
fraction of approximately 13,000 isolated strains has Wales, characterized the extremophiles as the "bungi
been characterized. We know now that microorganisms jumpers " of the microbial world. There are 36 entries
exist in subsurface environments extending to a depth covering this theme in the Encyclopedia. These microor-
of approximately two miles and possibly beyond this ganisms are able to survive and grow in extreme envi-
depth. Energy sources, both organic and inorganic, occur ronments with regard to environmental factors such
at relatively very low concentrations, enough to sustain as temperature (psychrophiles, hyperthermophiles), pH
the microorganisms in this oligotrophic environment. (acidophiles, alkaliphiles), salt concentration (halophiles),
Subsurface microorganisms are stressed, starved, display hydrostatic pressure (barophiles), water activity (xerotol-
very low metabolic activity and remain in a dormant state erant microorganisms), microgravity, ionizing radiation,
for very long periods of time (thousands to possibly millions light (e.g., growth under low-light in caves), interfaces
of years). such as the air-water interface, nutrient concentration
Soil microbiology is one of the oldest subdiscipline (oligotrophs), and in the presence of toxic chemicals (toxi-
within environmental microbiology. Soils play a key role tolerant microorganisms). This area is important to envi-
in the life support system on planet Earth. A combination ronmental microbiologists who want to gain knowledge on
of physical, chemical and biological factors play a role the limits and the origin of life on Planet Earth and poten-
in their formation. Soil microorganisms (bacteria, fungi, tially in Space. Some of these extreme environments (e.g.,
protozoa) interact with organic matter, mineral particles, permafrost) could serve as models for potential life in other
water and gases, the major components of this complex planets and for attempting to answer the question of life
ecosystem. They play a key role in nutrient cycling preservation on Earth. An active research area concerns
and affect soil quality by mineralizing plant nutrients, the biotechnological applications of these extremophiles
improving soil structure, processing plant residues into and their products.
organic matter, and degrading toxic compounds in Research on biodegradation in water, soil, sediments,
soils, particularly chemical pesticides used in modern wastewater treatment plants is now receiving increased
agriculture. Rhizosphere microbial communities enter into attention by microbiologists around the world. Microor-
symbiotic or mycorrhizal associations with plant roots, ganisms play a key role in biogeochemical cycles, par-
leading to promotion of plant growth. On the other hand, ticularly the carbon cycle. During the last few decades,
fungal and bacterial pathogens are the cause of plant an array of foreign compounds called xenobiotics (i.e.,
 INTRODUCTION xxvii

foreign to biological systems) was introduced into the Public health microbiologists, in collaboration with
environment. These compounds are generally resistant or environmental engineers, must also work towards reduc-
recalcitrant to degradation by microbial action. The topics ing the pathogen and parasite load in domestic wastewa-
addressed in this area include the biodegradation of halo- ter, biosolids, and drinking water.
genated aromatics, petroleum and other hydrocarbons, Drinking water microbiologists seek to understand
reductive dehalogenation and metabolism of haloorganics, the mechanisms behind the removal/inactivation of
or fuel oxygenates, as well as the methodology to assess the microbial pathogens and parasites, cyanobacterial toxins,
biodegradability of these recalcitrant compounds. Some algae, and microorganisms responsible for taste and
chemicals such as the halogenated organic compounds odor problems, by various processes involved in water
(halogenated hydrocarbons, halogenated aromatics, pes- treatment plants. These processes include coagulation-
ticides, PCBs; dioxin, polychlorinated dibenzofurans) are flocculation, filtration (sand, activated carbon), water
quite resistant to microbial action. Biotechnological appli- softening, disinfection (e.g., chlorination, ozonation, U.V
cations of this research include mainly the bioremediation irradiation) and, more recently, biological treatment
of soils and aquifers. of water. It is now realized that organic compounds,
Public health microbiologists deal with the fate present at very low concentrations ({ig/L levels) in water
(transport and survival) of pathogenic microorganisms distribution systems, are responsible for the formation of
and parasites in the environment, and with the develop- trihalomethanes microbial regrowth in distribution pipes,
ment of the appropriate methodology for their detection and chlorine demand with the subsequent reduction of free
in environmental matrices. This information is useful for available chlorine. Special topics such as home treatment
assessing the risk they pose to humans and animals, devices, cistern water and bottled water are also covered
and for indicating preventive measures for avoiding or in the encyclopedia.
reducing human exposure to these harmful agents. Dur- The objectives of wastewater treatment are the
ing the past three decades we have witnessed dramatic reduction of organic compounds (i.e., BOD reduction)
advances in the methodology for detection of pathogenic and suspended solids in wastewater, removal/reduction
microorganisms and parasites in environmental samples, of nutrients (N, P) and toxic trace organics and metals
to avoid or at least pollution of receiving waters,,
including wastewater, drinking water, soils, biosolids,
and removal/inactivation of pathogenic microorganisms
freshwaters, marine waters, sediments, and air. Three
and parasites. Microorganisms (bacteria, fungi, protozoa)
associate editors have been assigned to this vast subdis-
play a crucial role in the biological treatment of
cipline in environmental microbiology. One editor dealt
wastewater, particularly in BOD and nutrient removal.
with bacterial pathogens and the two others dealt with
Several chapters in this encyclopedia are devoted to the
viruses and protozoan parasites. Most of the "classical"
microbiology of activated sludge, trickling filters, waste
bacterial pathogens are included (e.g., Vibrio cholera,
stabilization ponds, biofiltration, biosolids treatment and
Salmonella, Shigella, E. coli, Pseudomonas, Mycobac-
disposal, and to the use of wetlands, reed beds, algal
terium avium Complex, Leptospira, Legionella, Campy-
turf scrubbers, septic tanks or landfills to treat domestic
lobacter) as well as emerging pathogens such as Heli- wastewater and municipal solid wastes. The activated
cobacter pylori for which there are indications that it may sludge process is often prone to solid separation problems
be transmitted via the waterborne or food-borne routes. such as filamentous bulking and foaming. The physiology,
The biology, methodology (concentration and assay), risk growth kinetics, and control of filamentous bacteria are
assessment, and fate in the environment of a wide range covered in a few chapters.
of pathogenic viruses are discussed in this encyclopedia. Biofilms develop on biological and non biological
The list include enteroviruses, hepatitis viruses, human surfaces immersed in water and wastewater. They
caliciviruses including the Norwark virus, rotaviruses, are ubiquitous in natural aquatic environments (e.g.,
astroviruses, and adenoviruses. In recent years, research stream beds) and in engineered systems (e.g., fixed-
has focused on the epidemiology, methodology, fate in the film bioreactors). While they are beneficial in fixed-
environment, including food, and risk assessment of proto- film bioreactors (biological treatment of drinking waster
zoan parasites which produce hardy cysts or oocysts which and wastewater) or in groundwater (biodegradation of
persist in the environment. Furthermore, they are not organic pollutants), their accumulation leads to problems
completely removed following passage though water and in water distribution pipes, biofouling which adversely
wastewater treatment plants and are generally resistant affect the performance of artificial structures, and to
to disinfection. The list of parasites covered in this work medical problems such as dental plaques or colonization
include Cryptosporidium, Giardia, Entamoeba histolytica, of artificial implants, leading to increased infection in
toxoplasma gondii, Isospora, Cyclospora, microsporidia, patients.
and free-living amebas. The detection of bacterial and The colonization process starts with chemical condi-
viral pathogens, and protozoan parasites discussed above tioning of the surface, followed by microbial colonization
requires costly and time-consuming methodology as well and finally by macroorganisms. Bacteria, the first col-
as well trained labor. The use of microbial indicators of onizing microorganisms, excrete extracellular polymeric
fecal contamination is thus required. These indicators substances that help "cement" the biofilm onto the sur-
include total coliforms, fecal coliforms, E. coli, fecal strep- face. Eukaryotes (algae, fungi, protozoa) are also part
tococci/enterococci, Clostridium perfringens and bacterial of the biofilm microbial community. The members of
phage. this community display increased exchange of nutrients
xxviii INTRODUCTION

and metabolites, facilitated gene exchange, protection desulfurization of coal, treatment of waste and air
from grazing, and increased resistance to toxic chemi- or bioremediation of contaminated soils and aquatic
cals. Members use quorum sensing to communicate with environments.
one another. Signaling chemicals consist of homoserine The achievements of researchers in all the areas
lactones in gram-negative bacteria. Various techniques of environmental microbiology discussed earlier would
have been developed for studying biofilm microorganisms. not have been possible without the development of
Molecular techniques (e.g., rRNA-based methods, finger- adequate methodology to tackle the various tasks.
printing techniques) and other techniques such as laser This area is extremely important to our understand-
scanning microscopy are now available for identification ing of the role of microorganisms in their natural
of biofilm microorganisms. Other methods are useful for settings, including those inhabiting extreme environ-
determining the metabolic activity of these microorgan- ments. We have witnessed significant advances in our
isms. Specific biofilms covered in the encyclopedia include ability to determine the numbers, biomass, activity,
diatom and bacterial-fungal biofilms, and biofouling in injury, and death of microorganisms in environmen-
industrial systems and its control. A few other impor- tal samples. The development of genetic probes, such
tant topics related to biofilms include biodeterioration as the 16S rRNA-targeted probes, has increased our
of mineral surfaces, bioleaching, biocorrosion, microbial knowledge of phylogenetics and microbial community
weathering, biofilms in natural and drinking water distri- structure in water, sediments, soils or biofilms, and
bution pipes, and biofilms in the food industry. has helped tremendously in the detection of micro-
Humans have exploited microbes and their products
bial pathogens and parasites in complex environmental
for thousands of years. Wine is the classical example
matrices. Methods covered detail in several entries of
of a biotechnological product, although the involvement
this encyclopedia, deal with the identification of micro-
of microscopic yeast cells in the process was not known
bial isolates, flow cytometry, biochip-based devices (i.e.,
until the pioneering work of Pasteur. Today, biotechnol-
microarrays) for microbial community ribotyping, image
ogists have developed a myriad of modern techniques
analysis, biophotonic imaging, bioreporters, fluorescent in
and processes that use microorganisms and their prod-
ucts (e.g., enzymes, biosurfactants, compatible solutes, situ rRNA probes, lipid biomarkers, phylogenetics, ribo-
ice nucleating proteins) for the benefit of humans, ani- typing, capillary electrophoresis, and methods for
mals and the environment. The microorganisms covered in specifically detecting microbes such as archaea or
this encyclopedia include archaea, bacteria (e.g., methan- psychrophiles.
otrophs, sulfate reducing bacteria), algae, and fungi. The It is hoped the information presented in this ency-
advances made in biotechnology are benefiting the fields clopedia will be the catalyst for more exciting research
of medicine, agriculture and environmental remediation. on the role of microorganisms in the environment, the
Much hope is now placed on marine microorganisms control of pathogenic and parasitic microorganisms by
and extremophiles (especially those belonging to the engineered treatment systems which serve as safeguards
archaea domain) for biotechnological breakthroughs. A against disease occurrence in humans and animals, a bet-
wide range of useful products are now derived from hyper- ter understanding of life at the edge, and the beneficial
thermophiles, psychrophiles, halophiles, alkaliphiles and use of microorganisms for restoring the environment, and
acidophiles, benefiting medicine, molecular biology, biore- improving our quality of life.
mediation, as well as the food, pharmaceutical, pulp and
paper or the oil industries. GABRIEL BITTON
The encyclopedia also covers the biotechnological Gainesville, Florida
aspects of processes such as bioleaching of metals, October 2001
 A
ACETYLENE REDUCTION ASSAY. See NITROGEN ACTIVATED SLUDGE-FOAMING
FIXATION IN SOILS — FREE-LIVING MICROBES
JACQUES SODDELL
La Trobe University
Bendigo, Australia

ACID MINE DRAINAGE. See BIOMINERALIZATION BY

BACTERIA One of the major solids separation problems in activated
sludge plants is the occurrence of a thick stable viscous
scum on the surface of aeration tanks or sedimentation
tanks (Fig. 1). This scum, or foam, a result of overgrowth of
certain microbes in the activated sludge, causes a number
ACIDOPHILES. See EXTREMOPHILES: LIFE IN EXTREME of operational problems (extra maintenance, reduction
ENVIRONMENTS in oxygen transfer, poor effluent, foam in anaerobic
digesters) and is a potential health hazard (spread of
pathogens) as well. It occurs often in activated sludge
plants [e.g., 50% of plants surveyed in Italy recently
ACTINOMYCETES, AIRBORNE. See INFECTIOUS exhibited foaming (1)]. Some consider that foaming has
AIRBORNE BACTERIA become the most serious operational problem in activated
sludge systems today (2). In this review, I will discuss
the organisms that cause the problem, factors affecting
foam formation, and methods currently used to control the
problem. There is more emphasis on recent issues. For
ACTINOMYCETES IN ACTIVATED SLUDGE. an examination of earlier studies on the microbiology and
See FILAMENTOUS BACTERIA IN ACTIVATED SLUDGE: CURRENT control of foaming, see References 3 and 4. Other types of
TAXONOMIC STATUS AND ECOLOGY foam are sometimes experienced due to nonbiodegradeable
detergents and foaming at plant start-up, but these foams
are not stable (5) and are not considered here.

ACTINOMYCETES IN SOILS. See SOIL BACTERIA

WHAT IS FOAMING?

Foam formation has been described as a flotation process

involving interaction among gas bubbles (produced by the
ACTINOMYCETES: ROLE IN FOAMING OF aeration system), surfactants (these can be produced by
ACTIVATED SLUDGE. See ACTIVATED many microbes in the plant and be present in the incoming
SLUDGE — FOAMING waste), and hydrophobic particles, principally the ‘‘foam-
forming’’ microorganisms (4). The foam is made stable
through the presence of surfactants and hydrophobic
particles. Nonfilamentous microbes may also contribute
to foam formation, particularly through their emulsifying
ACTINORHIZAL SYMBIOSIS. See NITROGEN
ability (6).
FIXATION IN SOILS (SYMBIOTIC)

WHICH MICROBES CAUSE FOAMING?

ACTIVATED CARBON. See GRANULAR ACTIVATED The principal method used for identifying the causative
CARBON, BACTERIOLOGY OF organisms in activated sludge mixed liquor and foam have
been the microscopic methods based on early work by
Eikelboom (7,8). These have been published in a number of
manuals (5,9,10). Although foam formation was originally
attributed to an actinomycete called ‘‘Nocardia,’’ surveys
ACTIVATED SLUDGE, FILAMENTOUS BACTERIA using these methods (11–14) have shown that the main
IN. See FILAMENTOUS BACTERIA IN ACTIVATED SLUDGE: foam-forming organisms belong to a number of different
CURRENT TAXONOMIC STATUS AND ECOLOGY; FILAMENTOUS morphological types, including branched and unbranched
BULKING IN ACTIVATED SLUDGE, CONTROL OF filaments and nonfilamentous forms.
1
2 ACTIVATED SLUDGE-FOAMING

(a) (b)

Figure 1. (a) Foam on surface of aeration tank, (b) major foam overflow. Photos: Wayne Murdoch, Coliban Water, Bendigo, Australia.

Microthrix parvicella (13,14), and South Africa (11,12), but less common in
published reports from the United States. It appears to be
Microthrix parvicella is a long coiled unbranched gram- more prominent in biological nutrient removal plants (12),
positive filamentous actinomycete (Fig. 2a). It has been and its appearance may be temperature related because
the major cause of foaming in Europe (1,2,15), Australia it appears in larger numbers in cooler months in Europe

(a) (b)

(c) (d)

Figure 2. (a) Long coiled unbranched filaments of M. parvicella, (b) GALO (Gordonia amarae-like
organism) or Nocardia showing typical right-angled branching, (c) Skermania piniformis, with
its tree-like branching, (d) Nostocoida limicola II. Photos: Beth Seviour, Biotechnology Research
Centre, Bendigo, Australia.
 ACTIVATED SLUDGE-FOAMING 3

(15,16) and is more prominent in cooler regions of Australia Recent advances in the taxonomy of these microorgan-
(13). It is also responsible for bulking - another solids isms have clarified that organisms with GALO morphol-
separation problem in activated sludge. ogy belong to a number of genera: Nocardia, Gordonia,
The taxonomy of M. parvicella has now been resolved Tsukamurella, Rhodococcus, and Dietzia (4). These are all
despite difficulty with growing the organism in the actinomycetes, which contain mycolic acids in their cell
laboratory (17) — it is a deep-branching member of high walls (together with S. piniformis, but not M. parvicella).
percentage molecular G + C group of actinomycetes closest The following have been isolated in pure culture from
to Acidimicrobium ferrooxidans. Organisms related to foam, illustrating the diversity of mycolata found in acti-
M. parvicella have been found in diverse environments, vated sludge: G. amarae, Gordonia terrae and Gordonia
for example, a peat bog in Germany, geothermally heated sputi, Nocardia asteroides Nocardia farcinica, Nocardia
soil in New Zealand, soil in Australia, and various otitidiscaviarum, Rhodococcus coprophilus, Rhodococcus
marine environments in Japan and Finland. However, equi, Rhodococcus erythropolis, Rhodococcus globerulus,
M. parvicella still has Candidatus status as it has not Rhodococcus rhodochrous, Rhodococcus ruber, Rhodococ-
yet been fully characterized due to poor growth on cus rubra, Tsukamurella paurometabola, Tsukamurella
conventional media (17). spumae, and Dietzia maris (4). In addition, there are
Growth of M. parvicella occurs when aeration is many new species belonging to most of these genera, and
intermittent, but is suppressed under continuous oxygen Mycobacterium (27–29) and maybe even new genera (29).
supply (18). This may be related to its requirement The mycolic acid containing actinomycetes are aerobic
for reduced sulfur and nitrogen compounds, which may organisms and are stimulated by increased dissolved
not be present in adequate quantities in fully aerated oxygen concentrations (30). Hence they can be controlled
systems (18). It also explains why the organism is by biological selectors, which reduce dissolved oxygen.
commonly found in biological nutrient plants (BNR), which Although these organisms have traditionally been thought
have both anaerobic and anoxic stages. to occur in plants with long sludge age, some GALOs have
Recent studies using microautoradiography (MAR) particularly fast growth rates and may not be washed out
confirmed earlier studies by Slijkhuis, which showed that by reducing sludge age.
M. parvicella will only grow on long-chain fatty acids (or Many wastes from food and chemical industries contain
their Tween esters) but not on simpler substrates (19). hydrophobic substrates such as vegetable oils, greases, and
However, another study contradicted these studies in hydrocarbons, and these may cause problems in activated
finding some growth on acetate (20). The MAR studies (19) sludge plants (31). The mycolic acid–containing actino-
showed that uptake of long-chain fatty acids occurred mycetes readily use such substrates, but with varying
under aerobic, anaerobic, and anoxic conditions, but there growth rates and affinities (29,32–34). The hydrophobic
was no proof that they were actually metabolized under nature of foam-formers with GALO morphology allows
anoxic and anaerobic conditions (they may only have been them to attach to hydrophobic substrates rather than to
incorporated as storage compounds). This rapid uptake grow in the aqueous phase. This may give them a means
of lipids probably explains the high lipid content of of competing with faster growing organisms present in the
M. parvicella, which can approach 35% dry weight (21). aqueous phase of activated sludge (34,35).
A property of the organism that has implications in The mycolic acid–containing actinomycetes grow at
foam control is its very high resistance to chlorination a range of temperatures (36). Those growing at lower
compared to other filamentous microbes found in activated temperatures (5 ° C) are principally rhodococci. Therefore,
sludge (22,23). the so-called Nocardia foams at low temperatures are more
Although its significance is not known, there appears to likely to be caused by the Rhodococcus species. Some foam
be a link between dominance of sludge by M. parvicella and isolates can grow at 40 ° C or higher, suggesting these are
high levels of uronic acids produced in the extracellular more likely to occur in plants treating warm wastewater,
polymers in the sludge (24). or possibly could grow in the foam itself when ambient air
temperatures are high (36).
Gordonia amarae-like Organisms (GALO)
Skermania piniformis
These bacteria are shorter branched gram-positive fil-
aments with branches at approximately 90 ° (Fig. 2b). Skermania piniformis is a branched filamentous actino-
They are commonly reported as Nocardia, but are some- mycete with a treelike appearance (Fig. 2c), and is the
times called NALO (Nocardia amarae-like organism) only Nocardia with morphology sufficiently different to
because their morphology is similar to N. amarae (25) and allow identification in its own right (37,38). However, most
because other mycolic acid–containing actinomycetes may surveys do not differentiate it from Nocardia or GALO.
also have similar morphology under certain conditions. Hence its true incidence outside Australia, where most
Since N. amarae was reclassified as a Gordonia (26), our studies have been carried out, is unknown. It was origi-
laboratory now uses the term G. amarae-like organisms or nally called PTLO (pine treelike organism) because of its
GALO to describe these morphotypes. treelike branching morphology, but was then classified
Although organisms such as M. parvicella and as Nocardia pinensis (39). More recently, these PTLOs
S. piniformis can be readily identified using microscopy, were reclassified as a new genus, with only one mem-
the morphotype described as GALO (Nocardia) represents ber, S. piniformis (40). Skermania piniformis is usually
many related organisms and hence cannot be identi- associated with plants operating at sludge ages greater
fied to genus level by microscopic morphology alone. than 20 days (14), and grows over a relatively narrow
4 ACTIVATED SLUDGE-FOAMING

range of temperatures (15–30 ° C), but this is still suffi- The Need to Identify?
cient to cause problems in many plants (36). Skermania
Although in some cases simple identification based on
piniformis grows better on hydrophobic substrates such
morphology is adequate for monitoring an activated sludge
as olive oil and Tween 80 than on glucose (34) as well as
plant, there are occasions when more precise identification
other hydrophobic substrates (29).
could be important: (a) To assist with foam control
of GALOs: Since GALOs have different growth rates,
Nostocoida limicola correct identification may indicate whether manipulation
Nostocoida limicola is an unusual filamentous bacterium of sludge age will be successful as a control measure. For
containing chains of rounded cells (Fig. 2d). Recent example, some rhodococci grow relatively fast (38) and
European surveys suggest it appears in foam more reduction of sludge age may not be sufficient to eliminate
commonly than previously thought (2,6,41). Although these, although it will eliminate slower growing rhodococci
there are three morphotypes, these surveys did not and other GALOs. (b) Some foam-formers are pathogens:
differentiate among N. limicola types I, II, and III, Although G. amarae, S. piniformis, and M. parvicella are
which are now known to belong to separate, unrelated not known to be pathogens, others organisms isolated
genera (42,43). The foam-former is probably N. limicola II, from foam (e.g., N. asteroides, N. farcinica, and R. equi)
which is now recognized as Candidatus N. limicola II (44), are potential pathogens. Since identification of pathogens
belonging to the Intrasporangiaceae. It is phylogenetically is not possible on morphological grounds, identification
closely related to and probably a member of the same using molecular probes is likely to become important.
genus as cocci in the genus Tetrasphaera, also isolated
from activated sludge (45). Nostocoida limicola II is a low USE OF MOLECULAR PROBES IN FOAM
F : M organism (5). Increasing the F:M ratio causes it to
lose its competitive advantage (30). While probes have made in situ identification possible
without the need to isolate an organism, the wide range
Others Filamentous Foam-Formers of mycolic acid producing organisms present in activated
sludge means that probes will probably still remain a
Other filamentous microbes have also been associated
research tool rather than a routine quality assurance tool
with foam formation. Eikelboom Type 1863 is one of
for a few years yet, as the number of different probes
major causes of foam (after Nocardia and Microthrix)
necessary to make a complete analysis is fairly large.
according to Jenkins and coworkers (5). Types 0675/0041
However, microarray/biochip probe technology (51) may
have a high incidence in some surveys, especially in eventually simplify such an analysis.
biological nutrient removal plants (6,46). Other organisms To date, probe studies have not explored the range
such as Eikelboom Types 0092, 0041, 0803, 0413, 1851, of mycolata likely to be present in activated sludge.
021N, 0914, 0581, 1701, Haliscomenobacter hydrossis, Most studies have concentrated on Gordonia, especially
Sphaerotilus sp, and cyanobacteria have been reported G. amarae (52–54) and Rhodococcus (47). It was con-
as the dominant organism in foam, but their incidence is cluded that G. amarae makes up only a small proportion
usually low (3,4). of gordoniae found in foam, strains previously identified as
G. amarae may represent two different species of Gordo-
Nonfilamentous Foam-Formers nia, and that a Rhodoccoccus probe detects both branching
Closely related taxonomically to the branching filaments organisms and short rods. The use of confocal laser scan-
are nonfilamentous forms of some of the mycolic ning microscopy is recommended to improve detection of
acid–producing actinomycetes, called ‘‘actinomycetes,’’ as organisms within flocs (47,55).
revealed in one survey (14) to differentiate them from A probe based on an Australia isolate of M. parvicella
GALO and S. piniformis. These ‘‘actinomycetes’’ were has also been designed and successfully detected
defined as gram-positive rods (often slightly branched M. parvicella in European studies (56). However, unless
or not at all) without the distinctive branching patterns the M. parvicella probe detects some unusual morphotypes
of GALO or S. piniformis. Other studies support their (not to date), it is less likely to be more useful than conven-
importance (6,47). These nonfilamentous foam-formers tional microscopy, which readily identifies M. parvicella.
have possibly been overlooked in earlier studies of foam
as they are less obvious than the branching organisms we MONITORING FOAM
call GALO or ‘‘Nocardia.’’
Cocci have also been reported in nonfilamentous foams Quantification is a prerequisite for rational investigation
in some plants. These include gram-positive cocci in of foaming in activated sludge (57) and various quantifi-
Australian plants (13,48), which may also be mycolata, cation techniques for monitoring the level of organisms in
as some of these undergo a coccus stage in the so-called foam and mixed liquor have been reported. These include
rod-coccus life cycle. The gram-negative coccobacillus measuring the abundance of filaments (5), immunofluo-
Acinetobacter sp. have also been described as the dominant rescence (58,59), use of a selective medium containing
organism in the foam of a nutrient removal plant (49), but n-octodecane and nalidixic acid (60) and quantitative in
its more important contribution to foaming is probably situ hybridization (57).
through the emulsifiers it produces (6), hence encouraging However, many of these methods have problems, and
the growth of organisms such as G. amarae (50). choosing a meaningful method is difficult. Levels of
 ACTIVATED SLUDGE-FOAMING 5

foam-formers in the mixed liquor are only meaningful Manipulation of Loading Rate
if foaming has not occurred. Once foaming occurs, it
Washout of foam-forming organisms by reduction of sludge
separates to varying degrees. The foam-forming organisms
age or increase in F:M (Food:Microorganisms) ratio is
from the mixed liquor and the level in the mixed liquor
commonly practiced. Reduction of sludge age should wash
is not necessarily then meaningful. Another problem
out slowly growing organisms but this does not always
with quantification is that the mycolic acid–containing
work since different foam-formers have different growth
actinomycetes may be present as short rods or cocci, and rates (4). Microthrix parvicella and S. piniformis are slow-
branching filaments, and some methods do not address this growers and can be controlled this way, but this may be at
problem, counting only filaments. Lack of specificity can the expense of nitrification, which requires a long sludge
also be a problem with the immunofluorescent techniques age. However, organisms with the morphology of GALO
described. may represent both fast- and slow-growing actinomycetes,
Foaming potential of a biomass can also be measured and therefore reduction of sludge age may not successfully
through foam production by aeration under laboratory control the foam.
conditions (61) and hydrophobicity measurements.
Use of Selectors

HYDROPHOBICITY AND FOAM Selectors, also known as contact zones (67), are separate
reaction vessels added to the system to allow kinetic and/or
Mixed liquor biomass is more hydrophobic in foaming metabolic selection of certain groups of microbes before
plants than in nonfoaming plants, and onset of foaming the waste enters the main treatment basin. They can be
incidents is often correlated with increase in cell surface aerobic, anoxic, and anaerobic (4). However, these are less
hydrophobicity (CSH) (62,63). Therefore, there is interest successful for foaming control than for bulking control (2).
in the hydrophobic properties of foam-forming organisms Anoxic selectors favor growth of floc formers at the expense
and their role in foam formation, particularly the of Nocardia, but are not useful for M. parvicella foams
hydrophobic mycolic acids present in the cell walls of because M. parvicella grows well anoxically.
GALOs and S. piniformis. There has recently been interest in two-stage selectors.
Pure culture studies with foam isolates of mycolic One method (68) involves a first stage, which is aerated,
acid–containing actinomycetes show that the CSH varies but the second stage is capable of variable aeration. When
with culture age, nutrient sources, and temperature (64), the second stage was subjected to an aeration period of
between 9 and 12 hours, the population of M. parvicella
but variations in the mycolic acid composition have
was reduced sufficiently to prevent stable foam formation
little influence on their CSH or foaming ability (Stratton
in the aeration tank, but had no effect on N. limicola.
and coworkers, 1999). However, using cell-water contact
Another approach is to have separate ‘‘feasting’’ and
angles as a measure of CSH, studies of Corynebacterium,
‘‘fasting’’ units (69), which capitalize on the sensitivity
Rhodococcus, Gordonia, and Mycobacterium revealed a
of filamentous bacteria to F : M ratio, with the combined
tendency for CSH to increase with mycolic acid size. An
effect of both units resulting in the suppression of the
exception was Mycobacterium, which was less hydrophobic
overgrowth of filamentous bacteria without adversely
than Gordonia (65). Possibly other cell wall or excreted
affecting the organic treatment efficiency of the modified
components neutralize the hydrophobic nature of the
process.
mycolic acids (66). Thus, although foam-formers are
hydrophobic, the relationship between CSH and mycolic Physical Methods
acids is complex.
There is also some evidence that M. parvicella, Some feel that physical methods rather than biological
which does not contain mycolic acids, is also strongly methods are more successful in treating foaming prob-
hydrophobic (21). lems (2), and emphasize the proper design and construc-
tion of secondary clarifiers to avoid the escape of floating
biomass to the final effluent. A number of different physi-
FOAM CONTROL cal approaches have been used. Water sprays are used to
physically break down the foam, but with mixed results,
There are many inconsistencies and disagreement as to probably because of the stability of the foam (3). Physical
which methods are the best, and many are based on removal of scum as it accumulates is also practised. A vari-
anecdotal data. The best methods to use may depend on ation on this is to use a so-called classifying selector (70)
which organisms are causing the foam, but this is not to encourage flotation so that foam-formers rise to surface
possible if microscopic examination is not carried out. more rapidly, and then remove them. This reduced Nocar-
Microscopic examination will easily identify M. parvicella, dia levels, particularly when nonionic detergent was added
S. piniformis, and N. limicola II, and action can readily to the system. It is important not to recycle any trapped
be instigated based on our knowledge of the physiology of foam back into the activated sludge biomass — this recy-
these organisms. However, if the foam-former is identified cles the organisms that cause the problem.
as a GALO, identification to species is more difficult.
Chemical Additives
Consequently, choosing the right course of action may be
more difficult (4). The most common approaches to foam The addition of chemicals to control foaming is always
control are summarized as follows. a consideration, but is in most cases an emergency
6 ACTIVATED SLUDGE-FOAMING

control method rather than a fundamental solution (23). A foams (81). This is equivalent to the situation in vessels
number of approaches have been described, with various with foam-trapping, which tend to have dispersed growth
degrees of success. of ‘‘Nocardia’’ (79). It is now possible to design plants to
minimize scum formation by eliminating any chance of
Chlorination. This kills filaments protruding from flocs surface trapping of scum in a reactor. There should be
but does not kill flocs unless too much chlorine is applied. a free-surface flow in the reactor, with the whole sur-
Chlorination is useful, but care must be taken not to face layer moving through the reactor with essentially the
over chlorinate. Microthrix is possibly more resistant to same retention time as the bulk of the mixed liquor. This
chlorine than ‘‘Nocardia’’ (22,23). However, G. amarae is ensures that solids trapped on the surface do not have
more sensitive to chlorine than nonfoaming organisms retention periods in excess of the liquid retention period,
such as Pseudomonas aeruginosa and Escherichia coli in particularly the sludge age (82).
pure culture and in a laboratory scale plant (71).
FOAMING IN ANAEROBIC DIGESTERS
Ozonation. Like chlorine, ozone is a strong oxidizing
agent. Recent studies suggest that it can be useful Recent investigations into foaming in anaerobic digesters
in suppressing scum without affecting the rest of the suggest that this problem is related to that in activated
biomass (72,73) added it to aeration tank at slightly sludge plants. Digesters fed with sludge containing
higher levels and found that it eliminated foaming in the GALO or M. parvicella end up with the same organisms
aeration tank. Ozone also improves sludge settleability dominating the digester foam (52,53,83–87).
and accelerates nitrification (72). The operating strategy to prevent foam in the anaerobic
digesters at these plants is to control the growth of
Antifoam. The use of antifoams is not often success- M. parvicella in the activated sludge tanks by increasing
ful (46,74), probably because the foams they were devel- the sludge load. Top installed stirrers and the addition of
oped against are usually much less stable than those polyaluminum salt have also been used to prevent foam
experienced in activated sludge plants (75). They are also formation (86).
very expensive. Pilot scale studies show that heat pretreatment
of excess activated sludge with a high content of
Bioaugmentation. These are commercial mixtures of M. parvicella (87) and ‘‘Nocardia’’ (84) reduces its foam
microbes, sometimes supplemented with enzymes. They potential, but chlorination increases its foaming poten-
are expensive and often do not work, or need to keep tial (85) when placed in an anaerobic digester.
adding for continued effect (therefore, expensive). In both
lab-scale and full-scale experiments such mixtures do not
BIBLIOGRAPHY
appear to affect foaming, although in some cases, they
resulted in a change of bacteria involved (from ‘‘Nocardia’’
1. P. Madoni, D. Davoli, and G. Gibin, Water Res. 34, 1767–1772
to M. parvicella) (76). (2000).
2. J. Wanner, I. Ruzickova, O. Krhutkova, and M. Pribyl, Water
Iron Salts. The dosing of mixed liquor with iron salts has Sci. Technol. 41, 217–225 (2000).
also been successful in some instances, but the response
3. J. A. Soddell, and R. J. Seviour, J. Appl. Bacteriol. 69,
appears to be variable (77). Iron salts have been found to 145–176 (1990).
be useful for a Rhodococcus induced scum, but laboratory
4. J. A. Soddell, in R. J. Seviour, and L. L. Blackall, eds., The
trials with different cultures of mycolic acid containing Microbiology of Activated Sludge, Kluwer, Dordrecht, The
actinomycetes showed a variable response to iron (78). Netherlands, 1999, pp. 161–202.
5. Jenkins, M. G. Richard, and G. T. Daigger, Manual on the
Addition of Polymers. Addition of cationic polymer to the Causes and Control of Activated Sludge Bulking and
clarifier to combat poor settling activated sludge resulted Foaming, 2nd ed., Lewis Publishers, Boca Raton, Fla., 1993.
in the elimination of ‘‘Nocardia’’ foam after three days (79), 6. H. Lemmer et al., Acta Hydrochim. Hydrobiol. 28, 34–40
probably by neutralizing and precipitating the foam- (2000).
stabilizing anionic surfactants present in the wastewater. 7. D. H. Eikelboom, Water Res. 9, 365–388 (1975).
Successful suppression of foaming by M. parvicella 8. D. H. Eikelboom, Water Res. 8, 153–162 (1977).
through the addition of quaternary ammonium-based
9. D. H. Eikelboom, and H. J. J. van Buiysen, Microscopic
antifilament polymer has also been reported (23). Sludge Investigation Manual, 2nd ed., TNO Research Insti-
tute of Environmental Hygiene, Delft, The Netherlands, 1983.
Manipulation of Cell Surface Chemistry. Addition of clay 10. R. J. Seviour, and L. L. Blackall, The Microbiology of Acti-
compounds such as bentonite, talc, zeolite, and montmo- vated Sludge, Kluwer, Dordrecht, The Netherlands, 1999.
rillinite can remove foam-forming organisms (63,80). 11. J. R. Blackbeard, G. A. Ekama, and G. V. R. Marais, J. Water
Pollut. Control Fed. 85, 90–100 (1986).
Appropriate Design 12. J. R. Blackbeard, D. M. D. Gabb, G. A. Ekama, and G. V. R.
It is thought that subsurface culture withdrawal encour- Marais, Water SA 14, 1–18 (1988).
aged the growth of dispersed (i.e., free floating) G. amarae 13. E. M. Seviour et al., Water Res. 24, 493–498 (1989).
filaments rather than filaments within flocs; and these 14. E. M. Seviour et al., Water Res. 28, 2335–2342 (1994).
dispersed filaments are more likely to produce stable 15. S. Knoop, and S. Kunst, Water Sci. Technol. 37, 27–35 (1998).
 ACTIVATED SLUDGE-FOAMING 7

16. D. H. Eikelboom, Water Sci. Technol. 29, 271–279 (1994). 49. M. A. Hart, Water SA 11, 171–178 (1984).
17. L. L. Blackall et al., Int. J. Syst. Bacteriol. 46, 344–346 50. H. Lemmer, and M. Baumann, Water Res. 22, 765–767
(1996). (1988).
18. H. Slijkhuis, and M. H. Dienema, Water Res. 22, 825–828 51. M. Schena, DNA Microarrays - A Practical Approach. Oxford
(1988). University Press, Oxford, U.K., 1999.
19. K. Andreasen, and P. H. Nielsen, Water Res. 5, 1599–1569 52. F. L. Delosreyes, W. Ritter, and L. Raskin, Appl. Environ.
(2000). Microbiol. 63, 1107–1117 (1997).
20. V. Tandoi, S. Rosseti, L. Blackall, and M. Majone, Water Sci. 53. M. F. Delosreyes, F. L. Delosreyes, M. Hernandez, and
Technol. 37, 1–8 (1998). L. Raskin, Appl. Environ. Microbiol. 64, 2503–2512 (1998).
21. H. Slijkhuis, Appl. Environ. Microbiol. 46, 832–839 (1983). 54. D. B. Oerther, F. L. de los Reyes, M. Hernandez, and
22. J. B. Neethling, K. M. Hohnson, and D. Jenkins, J. Water L. Raskin, FEMS Microbiol. Ecol. 29, 129–136 (1999).
Pollut. Control Fed. 57, 890–894 (1985). 55. M. Schuppler, M. Wagner, G. Schon, and U. B. Gobel, Micro-
23. Y. W. Hwang, and T. Tanaka, Water Res. 32, 1678–1686 biology 144, 249–259 (1998).
(1998). 56. R. Erhart et al., Syst. Appl. Microbiol. 20, 310–318 (1997).
24. V. S. Baxter-Plant, E. Hayes, and C. F. Forster, J. Inst. Water 57. R. J. Davenport et al., Appl. Environ. Microbiol. 66,
Environ. Man. 13, 131–136 (1999). 1158–1166 (2000).
25. M. P. Lechevalier, and H. A. Lechevalier, Int. J. Syst. Bacte- 58. M. Hernandez, D. Jenkins, and B. L. Beaman, Water Sci.
riol. 24, 278–288 (1974). Technol. 29, 249–259 (1994).
26. S. Klatte, F. A. Rainey, and R. M. Kroppenstedt, Int. J. Syst. 59. B. N. Jacobsen, Water Res. 29, 279–283 (1995).
Bacteriol. 44, 769–773 (1994). 60. M. Fujita, K. Iwahori, and H. Taki, J. Ferment. Bioeng. 77,
27. M. Goodfellow, R. Davenport, F. M. Stainsby, and T. P. 674–678 (1994).
Curtis, J. Ind. Microbiol. 17, 268–280 (1996).
61. H. Chua, P. H. F. Yu, S. N. Sin, and K. N. Tan, Appl.
28. M. Goodfellow et al., Water Sci. Technol. 37, 511–519 (1998). Biochem. Biotechnol. 84-6, 1127–1135 (2000).
29. J. A. Soddell, R. J. Seviour, L. L. Blackall, and P. Hugen- 62. H. M. Stratton, R. J. Seviour, and T. G. Flapper, AWWA 16th
holtz, Water Sci. Technol. 37, 495–502 (1998). Federal Convention, 1996, pp. 69–75.
30. C. E. Scruggs, and C. W. Randall, Water Sci. Technol. 37, 63. H. M. Stratton, B. Seviour, and P. Brooks, Water Sci. Technol.
263–270 (1998). 37, 503–509 (1998).
31. C. F. Forster, J. Chem. Technol. Biotechnol. 55, 402–404 64. H. M. Stratton, R. J. Seviour, and J. A. Soddell, Actino-
(1992). mycetes 4, 40–46 (1993).
32. K. Iwahori, M. Wang, H. Taki, and M. Fujita, J. Ferment. 65. B. Bendinger, H. Rijnaarts, K. Altendorf, and A. J. B.
Bioeng. 79, 186–189 (1995). Zehnder, Appl. Environ. Microbiol. 59, 3973–3977 (1993).
33. H. Chua, K. N. Tan, and M. W. L. Appl. Biochem. Biotechnol. 66. M. Sunairi et al., J. Appl. Microbiol. 82, 204–210 (1997).
57-8, 851–856 (1996).
67. R. Pujol, and J. P. Canler, Water Sci. Technol. 26, 209–212
34. J. Soddell, and R. Seviour, Water Sci. Technol. 34, 113–118 (1994).
(1996).
68. C. F. Forster, and R. J. Foot, Environ. Technol. 18, 237–241
35. T. Bendt, A. Rahman, C.-H. Rolfs, and A. Kelle-Embden, (1997).
Korrespondez Abwasser 36, 152–156 (1989).
69. H. Chua et al., Water Sci. Technol. 42, 95–102 (2000).
36. J. A. Soddell, and R. J Seviour, Water Res. 29, 1555–1558
70. K. R. Pagilla, D. Jenkins, and W. H. Kido, Water Environ.
(1995).
Res. 68, 235–239 (1996).
37. J. A. Soddell, and R. J Seviour, Water Res. 28, 2343–2351
71. P. K. Wong, and W. K. Chung, J. Environ. Sci. Health Part
(1994).
A-Environ. Sci. Eng. 28, 1615–1628 (1993).
38. J. A. Soddell, and R. J Seviour, J. Appl. Microbiol. 84,
272–284 (1998). 72. M. Goi et al., Water Sci. Technol. 30, 231–234 (1994).
39. L. L. Blackall et al., J. Gen. Microbiol. 135, 1547–1558 73. T. Okouchi et al., Water Sci. Technol. 34, 283–290 (1996).
(1989). 74. P. Pitt, and D. Jenkins, J. Water Pollut. Control Fed. 62,
40. J. Chun et al., Int. J. Syst. Bacteriol. 47, 127–131 (1997). 143–150 (1990).
41. J. Wanner et al., Water Sci. Technol. 37, 271–279 (1998). 75. J. Wanner, Activated Sludge Bulking and Foaming Control,
Technomic Publishing, Lancaster, Calif., 1994.
42. J. R. Liu et al., Syst. Appl. Microbiol. 23, 528–534 (2000).
76. A. Franz, and N. Matsche, Water Sci. Technol. 29, 281–284
43. J. R. Liu et al., Int. J. Syst. Evol. Microbiol. 51, 195–202
(1994).
(2001).
77. P. Duchene, Water Sci. Technol. 29, 239–247 (1994).
44. L. L. Blackall et al., Int. J. Syst. Evol. Microbiol. 50, 703–709
(2000). 78. H. Lemmer, and R. M. Kroppenstedt, Syst. Appl. Microbiol.
5, 124–135 (1984).
45. A. M. Maszenan et al., Int. J. Syst. Evol. Microbiol. 50,
59–603 (2000). 79. Y. J. Shao et al., Water Environ. Res. 69, 25–27 (1997).
46. R. Pujol, P. Duchene, S. Schetrite, and J. P. Canler, Water 80. L. L. Blackall, and K. C. Marshall, J. Ind. Microbiol. 4,
Res. 25, 1399–1404 (1991). 181–188 (1989).
47. R. J. Davenport, J. N. Elliott, T. P. Curtis, and J. Upton, 81. L. L. Blackall, V. Tandoi, and D. Jenkins, J. Water Pollut.
Antonie van Leeuwenhoek Int. J. Gen. Mol. Microbiol. 74, Control. Fed. 63, 44–50 (1991).
41–48 (1998). 82. A. R. Pitman, Water Sci. Technol. 34, 291–298 (1996).
48. H. M. Stratton et al., Lett. Appl. Microbiol. 22, 341–346 83. M. Hernandez, and D. Jenkins, Water Environ. Res. 66,
(1996). 828–835 (1994).
8 ACTIVATED SLUDGE — MICROBIOLOGY OF NITROGEN REMOVAL

84. K. R. Pagilla, K. C. Craney, and W. H. Kido, Water Sci. Table 1. Oxidation States of Nitrogen
Technol. 36, 463–470 (1997).
Compound Formula Oxidation State
85. K. R. Pagilla, D. Jenkins, and W. Kido, Water Sci. Technol.
38, 49–54 (1998). Ammonium NH4 + −III
86. A. D. Westlund, E. Hagland, and M. Rothman, Water Sci. Hydrazine N2 H4 −II
Technol. 37, 51–55 (1998). Hydroxylamine NH2 OH −I
87. A. D. Westlund, E. Hagland, and M. Rothman, Water Sci. Dinitrogen N2 0
Technol. 38, 29–34 (1998). Nitrous oxide N2 O +I
Nitric oxide (Nitrogen monoxide) NO +II
Nitrite NO−2 +III
Nitrate NO−3 +V
ACTIVATED SLUDGE, METHODOLOGY.
See ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR
DETERMINING COMMUNITY COMPOSITION
N2

Nit
r
fixa ogen
tion
ACTIVATED SLUDGE — MICROBIOLOGY OF
NITROGEN REMOVAL N2O NH4+
Anammox
INGO SCHMIDT

n ory
MARC STROUS Denitrification

tio lat
MIKE M. S. JETTEN

uc mi
ed ssi
University Nijmegen

e r Di
Toernooiveld, The Netherlands

trit y,
Ni ator
il

n
sim

io
The elimination of inorganic nitrogen compounds plays

at
As

fic
an important role in wastewater treatment, because these

tri
NO NH2OH

Ni
compounds contribute significantly to the eutrophication
of aquatic environments and are toxic to fish. Ammo-
nium occurs as the main nitrogen component of untreated
wastewater. Organically bound nitrogen is also released
NO2−
as ammonium. Generally, ammonium nitrogen is removed
by biological treatment such as activated sludge systems
or trickling filters. Removal is also effected by several
chemical and physical methods such as precipitation,
stripping, or ion exchange. The first step of biological
ammonium elimination is nitrification. The oxidation of
ammonium to nitrite is catalyzed by the ammonia oxidiz- NO3−
ing bacteria. The subsequent oxidation of the nitrite to
Figure 1. Biological nitrogen cycle.
nitrate is catalyzed by the nitrite oxidizing bacteria. In
the course of denitrification, the second step of biologi-
cal nitrogen removal, nitrite and nitrate are reduced to
The biological nitrogen cycle is shown in Figure 1.
gaseous N-compounds such as nitric oxide, nitrous oxide,
After fixation, nitrogen is available to other organisms
or dinitrogen.
in the form of ammonia/ammonium (NH3 /NH4 + ). It can
subsequently be assimilated into biomass, or it can be
THE BIOLOGICAL NITROGEN CYCLE used for energy generation by specialized bacteria in the
process of nitrification.
Nitrogen (N) is an element with oxidation states from −III Two groups of microorganisms are responsible for
to +V (Table 1). nitrification. Ammonia oxidizers catalyze the first step,
Nitrogen is an important element in proteins, genetic the oxidation of ammonia to nitrite, and nitrite oxidizers
material, and further essential organic molecules. In the catalyze the second step, the oxidation of nitrite to nitrate.
atmosphere, nitrogen mainly exists as dinitrogen gas (N2 ), Nitrate can be reduced to ammonia via nitrite. Further,
where it constitutes about 78% of the atmosphere. nitrite and nitrate can be used as terminal electron
Nitrogen is fixed by bacteria in primary production and is acceptors during denitrification (anaerobic respiration) to
liberated as ammonium in the biological food chain (N- form N2 . N2 O and NO may occur as intermediates in the
cycle). Part of the ammonium is returned to primary process.
production and the bacterial processes of nitrification From a microbiological point of view, the nitrogen cycle
and denitrification recycle the surplus to the atmosphere (Table 2) is made up of four catabolic reactions (oxida-
as N2 . tion of ammonia, oxidation of nitrite, denitrification, and
 ACTIVATED SLUDGE — MICROBIOLOGY OF NITROGEN REMOVAL 9

dissimilatory nitrite reduction) and three anabolic reac-

tions (ammonium assimilation, assimilatory nitrite reduc-
tion, and nitrogen fixation). Oxidation of ammonia can
be divided into aerobic (O2 -dependent), anaerobic (NO2 - C
dependent), and anaerobic (nitrite-dependent, Anammox)
oxidation of ammonia. I

R
OXIDATION OF AMMONIA

Physiology of Ammonia Oxidation

The gram-negative ammonia oxidizers, for example, mem-
bers of the genera Nitrosomonas and Nitrosospira (1) are
lithoautotrophic organisms using carbon dioxide as the
main carbon source. Species of the genera Nitrosomonas
reveal extensive intracytoplasmic membrane (ICM) sys- Figure 2. Electron micrograph of ultrathin sections of N. eutro-
tems (Fig. 2). In Nitrosococcus oceanus, ICMs are arranged pha. C: carboxysome, Cy: cytoplasm with ribosomes (R), I:
as a centrally located stack of parallel, flattered vesicles, intracytoplasmic membranes. Bar: 0.5 µm.
and in other species, as flattened vesicles in the periph-
eral cytoplasm (e.g., Nitrosomonas eutropha). Recently,
molecular tools that infer the presence of ammonia oxi- several apolar compounds such as methane, carbon
dizing bacteria in the environment have been supple- monoxide, or some aliphatic and aromatic hydrocar-
mented by PCR primers for specific amplification of the
bons. These compounds can act as competitive inhibitors
ammonia monooxygenase structural genes amoA. Envi-
of ammonia oxidation. The second step (Eq. 2) in this
ronmental 16S rRNA and amoA libraries extend the
metabolism is performed by the enzyme hydroxylamine
knowledge on the natural diversity of ammonia oxidiz-
oxidoreductase (HAO). This enzyme oxidizes hydroxy-
ing bacteria. Comparative 16S rRNA sequence analyses
lamine to nitrite. Two of the four electrons released
revealed that members of this physiological group are
from this reaction are required for the AMO-reaction
confined to two monophyletic lineages within the Pro-
and the remaining are used for the generation of
teobacteria. Nitrosococcus oceanus is affiliated with the
a proton motive force in order to regenerate ATP
gamma-subclass of the Proteobacteria, whereas members
and NADH.
of the genera Nitrosomonas and Nitrosospira form a closely
related grouping within the beta-subclass of Proteobacte-
ria (2). NH3 + O2 + 2H+ + 2e− → NH2 OH + H2 O (1)
Ammonia oxidizers are widespread in nature as well as
in artificial ecosystems. Strains of different species have NH2 OH + H2 O → HNO2 + 4H+ + 4e− (2)
been isolated from fresh, brackish, and salt waters and
soils. Some ammonia oxidizers are adapted to special, NH3 + O2 → HNO2 + 2H+ + 2e− (3)
sometimes extreme environments. For example, strains of
Nitrosospira have been isolated from acid tea soil at pH
values of 4.6, Nitrosomonas strains from marine sediments ANOXIC METABOLISM (ANOXIC AMMONIA
with low temperatures, and from alkaline soda lakes with OXIDATION)
pH values above 10.0 (3).
Ammonia is oxidized in two steps. First, ammonia is oxi- Recently, an anoxic metabolism (denitrification and oxida-
dized to hydroxylamine by the ammonia monooxygenase tion of ammonia) in ammonia oxidizers (Nitrosomonas)
(Eq. 1). This enzyme is unspecific and also oxidizes was discovered (4,5). Furthermore, a new autotrophic

Table 2. Catabolic Reactions of the Nitrogen Cycle

Reaction Equation

Ammonia oxidation O2 -dependent NH3 + O2 → HNO2 + 2H+ + 2e−

NO2 -dependent NH3 + N2 O4 → HNO2 + 2NO + 2H+ + 2e−
Anammox NH3 + 1.3HNO2 → 0.3HNO3 + 1N2 + 1.7H2 O + 0.6H+ + 0.6e−
Nitrite oxidation HNO2 + H2 O → HNO3 + 2H+ + 2e−
Denitrification HNO3 + 2H+ + 2e− → HNO2 + H2 O
HNO2 + H+ + e− → NO + H2 O
2NO + 2H+ + 2e− → N2 O + H2 O
N2 O + 2H+ + 2e− → N2 + H2 O
Dissimilatory nitrite HNO2 + 6H+ + 6e− → NH3 + 2H2 O
reduction
10 ACTIVATED SLUDGE — MICROBIOLOGY OF NITROGEN REMOVAL

member of the order Planctomycetales was identified, oxi- partly used as an electron acceptor, leading to the
dizing ammonia under anoxic conditions to N2 with nitrite formation of dinitrogen (Eq. 7).
as electron acceptor via the intermediate hydrazine (6).
HNO2 + 3H+ + 3e− → N2 (7)
Denitrification by Ammonia Oxidizers
Equation 1 to 6 indicate that there are only a
During oxic oxidation of ammonia by ammonia oxidizers, few differences between the anaerobic and the aerobic
small amounts of nitric and nitrous oxide are released. oxidation of ammonia by N. eutropha. Instead of O2 in
Both gases are also produced in the course of aero- the course of aerobic oxidation of ammonia, N2 O4 is used
bic denitrification by ammonia oxidizing bacteria. Addi- as electron acceptor and NO, an additional product, is
tionally, the formation of dinitrogen is observed (5,7). released.
In the absence of dissolved oxygen, Nitrosomonas is
capable of anoxic denitrification (Fig. 3) with molec- Effect of Nitrogenous Oxides on Oxidation of Ammonia
ular hydrogen (4) or simple organic compounds serv-
ing as the electron donors, whereas nitrite is used as Ammonia oxidizing bacteria of the genus Nitrosomonas
the electron acceptor. However, high aerobic denitri- are inhibited when gaseous nitric oxide is removed from
fication activities were only obtained when the cells laboratory-scale cultures by means of intensive aeration.
were grown under extremely oxygen-limited conditions. Nitrification starts again when nitric oxide is added to
But under these conditions ammonia oxidation rates the gas inlet of the culture vessels (8). Using a laboratory-
are low. scale fermenter with complete biomass retention, it could
be shown that nitrogenous oxides such as nitric oxide and
Anaerobic NO2 -Dependent Oxidation of Ammonia by especially nitrogen dioxide have a significant promoting
Nitrosomonas effect on pure cultures of N. eutropha (9). Compared
with cultures grown without these externally added
In the absence of dissolved oxygen, oxidation of ammonia
nitrogenous oxides their addition resulted in a pronounced
has recently been observed in cultures of N. eutropha (5).
increase in nitrification rate, specific activity of ammonia
These cells were able to replace molecular oxygen by
oxidation, growth rate, maximum cell density, and aerobic
nitrogen dioxide or dinitrogen tetroxide, respectively,
denitrification capacity. Maximum cell numbers amounted
in the course of ammonia monooxygenase reaction
to 2 · 1010 Nitrosomonas cells per ml. Further, about 50%
(Eqs. 4–6).
of the nitrite produced was aerobically denitrified to
dinitrogen when nitrogen dioxide was present.
NH3 + N2 O4 + 2H+ + 2e− → NH2 OH + H2 O + 2NO (4) Recently, a new hypothetical model for oxidation of
NH2 OH + H2 O → HNO2 + 4H + 4e + −
(5) ammonia by N. eutropha was developed (10). A new
complex role for nitrogen oxides in the metabolism of
these organisms was proposed. Anaerobic oxidation of
NH3 + N2 O4 → HNO2 + 2NO + 2H+ + 2e− ammonia (Eqs. 4–6) is dependent on the presence of the
(6) oxidizing agent N2 O4 . NO is produced in stoichiometric
amounts and released into the atmosphere. The situation
Hydroxylamine and nitric oxide were formed in is more complex under oxic conditions. In the presence of
this reaction. Although nitric oxide was not further O2 , the produced NO can be oxidized to NO2 . According
metabolized, hydroxylamine was oxidized to nitrite, as to the new model (Fig. 4), N2 O4 is the oxidizing agent
shown, under oxic conditions. The nitrite produced was under oxic conditions. During the oxidation of ammonia,

Ammonia, Nitrite (mM), Nitrous oxide (µM) Cell number (cells ml−1)
1,E + 09

1,E + 08

1
1,E + 07

Figure 3. Denitrification and cell growth of

N. eutropha in a hydrogen atmosphere with 20%
carbon dioxide. Hydrogen, not ammonia, was used
as electron donor indicated by decreasing the H2 1,E + 06
0
concentration (not shown), whereas the ammonia
0 20 40 60 80 100
concentration remained unchanged. (
), ammonia;
(), nitrite; (), N2 O; (−), cell number. Time (days)
 ACTIVATED SLUDGE — MICROBIOLOGY OF NITROGEN REMOVAL 11

hydroxylamine and NO are produced as intermediates. used to survey the presence of B. anammoxidans and
Although hydroxylamine is further oxidized to nitrite, NO related bacteria in several wastewater treatment systems
is (re)oxidized to NO2 (N2 O4 ). with a very high nitrogen load and limited air sup-
The new hypothetical model is in agreement with ply. Indeed ‘‘Candidatus Brocadia anammoxidans’’ and
the reaction described earlier for aerobic oxidation the closely related ‘‘Candidatus Kuenenia stuttgartiensis’’
of ammonia. Although the total consumption rates could be detected in many of these systems throughout the
(ammonia, oxygen) and production rates (hydroxylamine world (12).
as intermediate) remain unchanged, the mechanism of the For the application of the Anammox process, it
reaction is different. is important to know how B. anammoxidans coped
with oxygen. Batch experiments showed that oxygen
ANAEROBIC AMMONIUM OXIDATION (ANAMMOX) as low as 2 µM completely, but reversibly, inhibited
the anammox activity. The obligately anaerobic nature
Engelbert Broda (1977) predicted the existence of of B. anammoxidans is in sharp contrast with the
chemolithoautotrophic bacteria capable of anaerobic more versatile aerobic Nitrosomonas-like ammonium
ammonium oxidation (Anammox). oxidizers.
As a result of this inhibition by oxygen, it seems
NH4 + + NO2 − → N2 + 2H2 O (8) likely that the aerobic ammonium oxidizing bacteria and
B. anammoxidans are only able to coexist under oxygen
limiting conditions. The ammonia oxidizers would oxidize
Only recently the experimental confirmation of Broda’s
ammonia to nitrite and keep the oxygen concentration low,
prediction was described (6). The biological nature of the
and B. anammoxidans would convert the toxic nitrite and
process was verified and nitrite was shown to be the pre-
ferred electron acceptor. Hydroxylamine and hydrazine the remaining ammonium to nitrogen gas. By gradually
supplying more and more air into an anammox sequencing
were identified as intermediates. Bacteria were enriched
in a mineral medium containing ammonia and nitrite, batch reactor, it was indeed possible to establish such a
and bicarbonate was the only carbon source. The growth cooperation. In the reactor, Nitrosomonas-like bacteria
rate of the cultures was extremely low (doubling time consumed the oxygen effectively, so that the actual oxygen
three weeks), and reactor systems with very efficient concentration remained below the detection limit of 2 µM.
biomass retention had to be used. The enrichments were Nitrite concentrations never exceeded 1 mM, indicating
dominated by a bacterium with a conspicuous distinc- that B. anammoxidans was active as well. After five
tive morphology. The bacterium was physically purified months wastewater with an ammonia concentration of
from the enrichment by density gradient centrifuga- 30 mM was converted into dinitrogen gas and some
tion (11). The purified cell suspension had high anam- nitrate, according to Equation 9 (13).
mox activity and fixed carbon dioxide. The 16S rDNA
gene of the bacterium was shown to branch very deep 2.5NH4 + + 2.1O2 → 0.2NO3 − + 1.15N2 + 3.6H2 O + 2.8H+
within the planctomycete lineage of descent. The anaer- (9)
obic ammonium oxidizing planctomycete-like bacterium The microbial composition of the biomass in
was named ‘‘Candidatus Brocadia anammoxidans.’’ The this reactor was analyzed with FISH, using
16S rDNA sequence information was used to design B. anammoxidans (Amx820) and Nitrosomonas (Neu653)
specific oligonucleotide probes for application in fluores- specific probes. Initially B. anammoxidans domi-
cence in situ hybridization (FISH), and these probes were nated (70%) but over time more and more Nitrosomonas-
like bacteria were detected. Aerobic nitrite oxidizers
(Nitrobacter or Nitrospira) were never detected, consis-
NH2OH + H2O
tent with the absence of nitrite oxidizing activity in oxic
batch tests.
The symbiosis of aerobic and anaerobic ammonium
oxidizing bacteria is relevant for wastewater treatment.
NO
The experiments showed that ammonium can be removed
NO O2 in a simple, single oxygen-limited step. The process was
NH3
named CANON (completely autotrophic N-removal over
nitrite). Figure 5 shows images of CANON biomass after
N2O4 hybridization with labeled probes.
The introduction of anammox to N-removal would
lead to a substantial reduction of operational costs. The
NO2 process would be suitable to treat waters that contain
NO2 high ammonium and little organic COD. The Anammox
process would replace the conventional denitrification step
Figure 4. NOx -cycle: New hypothetical model of ammonia completely and would also save half of the nitrification
oxidation by Nitrosomonas. According to this model, N2 O4 is the aeration costs.
oxidant for oxidation of ammonia. Under oxic conditions oxygen The feasibility of anammox to treat sludge liquor was
will be used to reoxidize NO to NO2 (N2 O4 ). Hydroxylamine will investigated in combination with the SHARON (single
be oxidized to nitrite. reactor system for high ammonium removal over nitrite)
12 ACTIVATED SLUDGE — MICROBIOLOGY OF NITROGEN REMOVAL

(a) (b)

(c) (d)

Figure 5. Phase contrast (a,c) and epifluorescence (b,d) images of biofilm aggregates from a
CANON reactor after hybridization with Cy-3 labeled probes PLA46 (b, specific for planctomycetes)
and BET42 (d, specific for beta proteobacteria, mainly nitrifiers in this case). The figure shows that
both aerobic and anaerobic ammonium oxidizers are present in this reactor system for nitrogen
removal from wastewater. Bar equals 5 micrometer. See color insert.

process. SHARON was developed recently for the removal The feasibility of SHARON for the production of
of ammonium via the so-called nitrite route (14). It was ammonium and nitrite (1 : 1) was demonstrated in a 20-
tested for two years in the laboratory and successfully L laboratory system (16). The ammonium was oxidized for
scaled-up to full scale (1,800 m3 ) (15). Anammox needs 53% to nitrite at 1.2 kg N m−3 per day without pH control.
ammonium and nitrite in a ratio of about one to one. The effluent of the SHARON reactor was fed directly
This ratio can be achieved without control using the to an anammox reactor system. This reactor removed
SHARON process as a prenitrification step. When half all nitrite, and left some ammonium. During the test
of the ammonium is converted, the alkalinity of the sludge period the nitrogen load was 0.75 kg N m−3 per day. The
liquor is depleted, leading to a pH drop and preventing specific activity of the anammox biomass was very high:
further nitrification (Eq. 10). 0.8 kg N (kg dry weight)−1 per day.
The combined SHARON-Anammox process is presently
evaluated for full-scale implementation. On the basis of
NH4 + + HCO3 − + 0.75O2 → 0.5NH4 + + 0.5NO2 −
the design of the combined SHARON-Anammox process, a
+ CO2 + 1.5H2 O (10) cost estimate of 0.75 Euro kg−1 N was calculated. This is
 ACTIVATED SLUDGE — MICROBIOLOGY OF NITROGEN REMOVAL 13

very low compared with the 2 to 5 Euro kg−1 N estimated Oxidation of nitrite occurs under obligately oxic
for other processes that have been tested on a pilot plant conditions. The involved organisms are much more
scale for N-removal from sludge liquors. sensitive to oxygen limitation than ammonia oxidizers are.
At dissolved oxygen concentrations of about 0.5 mg l−1 ,
oxidation of nitrite is inhibited. Additionally, Nitrobacter is
HETEROTROPHIC AMMONIA OXIDATION
inhibited at high oxygen concentrations. Thus, the oxygen
content of a nitrite oxidizing nitrification tank has to
The oxidation of ammonia, hydroxylamine, or organic
be maintained carefully to avoid accumulation of nitrite.
nitrogen compounds, for example, oximes, to nitrite by
With sufficient oxygen supply oxidation of nitrite proceeds
various chemoorganotrophic microorganisms is called
at a faster rate than conversion of ammonia to nitrite.
heterotrophic nitrification. The latter is a cometabolism,
Therefore, high nitrite concentrations are rarely found in
which is assumed not to be coupled to energy conservation.
either natural environments or in wastewater treatment
Heterotrophic nitrifiers are found among algae, fungi,
plants.
and bacteria. Compared with those of autotrophic
nitrifiers, nitrification rates of heterotrophic nitrifiers are
Denitrification by Nitrite Oxidizers
low. Therefore, heterotrophic nitrification was thought
to occur preferentially under conditions that are not Some strains of Nitrobacter have been shown to be
favorable for autotrophic nitrification, for example, acidic denitrifying organisms as well. Nitrate can be used
environments. However, recent research has revealed that as an acceptor for electrons derived from organic
heterotrophic nitrification only contributes to a minor compounds to promote anoxic growth. Because the
extent to overall nitrate production, even in acid soils. oxidation of nitrite is a reversible process, the nitrite
Recently, heterotrophic nitrification has received some oxidoreductase can reduce nitrate to nitrite in the
attention, because it frequently occurs while accompanied absence of oxygen. Further, the nitrite oxidoreductase
by aerobic denitrification. In principle, this combination was purified together with a nitrite reductase, which
could lead to the elimination of dissolved nitrogen reduces nitrite to nitric oxide. Evidence is given that
compounds in a single-step reactor system. the subsequent oxidation of nitric oxide to nitrite is
involved in NAD-reduction in Nitrobacter hamburgensis.
However, because denitrifying cells of Nitrobacter grow
OXIDATION OF NITRITE
very slowly, it seems improbable that denitrification
by nitrite oxidizers might have any significance in the
Physiology of Nitrite Oxidation
treatment of wastewater.
The second step in nitrification, the oxidation of nitrite
to nitrate, is performed by nitrite oxidizing bacteria, for
DENITRIFICATION
example, members of the genera Nitrobacter, Nitrospira,
or Nitrococcus. Several strains of Nitrobacter and one
General Aspects of Denitrification
strain of Nitrospira are the only nitrite oxidizers that are
not restricted to marine environments (17). Denitrification is the reduction of oxidized nitrogen com-
The key enzyme of nitrite oxidizing bacteria is the pounds such as nitrite or nitrate to gaseous nitro-
membrane bound nitrite oxidoreductase, which oxidizes gen compounds. This process is performed by various
nitrite with water as the source of oxygen to form nitrate. chemoorganotrophic, lithoautotrophic, and phototrophic
The electrons released from this reaction are transferred bacteria and some fungi, especially under oxygen-reduced
via a- and c-type cytochromes to a cytochrome oxidase or anoxic conditions. Denitrification can be described as
of the aa3 -type. However, the mechanism for energy anoxic respiration. Electrons originated from, for example,
conservation in nitrite oxidizers is still unclear. No electron organic matter, reduced sulfur compounds, or molecular
transport chain linked to proton translocation could be hydrogen are transferred to reduce nitrogen compounds
found, a factor that is necessary to maintain a proton instead of oxygen in order to establish a proton motive
motive force for ATP regeneration. Thus, NADH is thought force usable for ATP regeneration. Enzymes involved are
to be produced as the first step of energy conservation. nitrate reductase, nitrite reductase, nitric oxide reduc-
The overall process deserves further research for complete tase, and finally nitrous oxide reductase. Dinitrogen is
elucidation. the main end product of denitrification, whereas the
nitrogenous gases nitric oxide and nitrous oxide can
Growth Characteristics of Nitrite Oxidizers occur as intermediates at low concentrations (18). How-
ever, these gases are also released as end products when
Nitrite oxidizers are generally lithoautotrophic organisms.
denitrification enzymes are not completely expressed, for
Carbon dioxide is fixed as the main carbon source
example, when the concentration of dissolved oxygen is
using RuBisCO, which is in part carboxysome-bound.
too high.
Higher growth rates are obtained when the cells are
maintained mixotrophically. In contrast to ammonia
Substrate Requirements of Denitrifiers
oxidizing bacteria, several strains of Nitrobacter are
capable of heterotrophic growth under oxic as well as The common denitrifiers in municipal wastewater
anoxic conditions. Heterotrophic growth is significantly treatment plants are chemoorganotrophic bacteria, for
slower than lithoautotrophic, although, 10- to 50-fold example, Paracoccus denitrificans or Alcaligenes eutro-
higher cell densities are obtained. phus, which require an electron donor of an organic
14 ACTIVATED SLUDGE MODELS: MICROBIOLOGICAL BASIS

nature. Hence, complete denitrification requires a suffi- 13. M. S. M. Jetten et al., Curr. Opin. Biotechnol. 12, 283–288
cient supply of organic matter. In detail, a specific C : N (2001).
ratio is required to provide an adequate amount of elec- 14. C. Hellinga et al., Water Sci. Technol. 37, 135–142 (1998).
tron donors for the reduction of a distinct amount of 15. J. W. Mulder, M. C. M. van Loosdrecht, C. Hellinga, and
nitrogen oxides. Wastewater with an inadequate COD R. van Kempen, Proceedings of the First IWA Conference,
loading should therefore be supplemented with an exter- IWAQ, London, U.K., 2000, pp. 267–274.
nal organic electron donor–like acetate or methanol 16. U. van Dongen, M. S. M. Jetten, and M. C. M. van Loos-
to avoid exceeding outlet concentrations of nitrate or drecht, Water Sci. Technol. 44, 153–160 (2001).
nitrite. Recently, denitrification activity could be demon- 17. E. Bock, H.-P. Koops, H. Harms, and B. Ahlers, Variations of
strated in pure cultures of ammonia oxidizers such Autotrophic Life, Academic Press, London, U.K., 1991.
as N. eutropha (4). Under anoxic conditions N. eutropha 18. W. G. Zumft, in A. Balows et al., eds., The Prokaryotes, 2nd
is able to denitrify with molecular hydrogen as elec- ed., Springer, New York, 1992, pp. 554–582.
tron donor and nitrite as the only electron acceptor in 19. L. A. Robertson and J. G. Kuenen, Antonie van Leeuwenhoek
a sulfide-reduced medium producing nitrous oxide and 57, 139–152 (1990).
dinitrogen. Cell growth is directly coupled to nitrite 20. W. G. Zumft and H. Körner, Antonie van Leeuwenhoek 71,
reduction. 43–58 (1997).

Aerobic Denitrification
Denitrification also occurs in the presence of oxygen. ACTIVATED SLUDGE MODELS:
The range of oxygen concentration permitting aerobic MICROBIOLOGICAL BASIS
denitrification is broad and differs from one organism
to another (19). The onset of aerobic denitrification MINO TAKASHI
does not depend on the oxygen sensitivity of the The University of Tokyo
corresponding enzymes but rather on regulation of oxygen- Tokyo, Japan
or redox-sensing factors involved in the regulation on
a transcriptional level. This is further indicated by The activated sludge (AS) process is the most common
denitrifying enzymes expressed under anoxic conditions wastewater treatment process. It was primarily designed
that remain active in the presence of oxygen (20). for the removal of organic material or carbon (C) (see
Generally, the ability to denitrify under oxic conditions ACTIVATED SLUDGE — THE PROCESS), but has been modified
seems to be the rule rather than the exception among to remove nitrogen (N) and phosphorus (P) as well.
denitrifiers. However, in the treatment of wastewater From an engineering point of view, it is necessary
denitrification commonly proceeds under anoxic and not to model the behavior of the process mathematically
under oxic conditions because in the presence of oxygen and to predict, for example, the effluent water quality
large amounts of organic matter would be ‘‘wasted ‘‘on or the amount of excess sludge produced. The main
respiration and biomass production. Hence, denitrification mechanism for the substrate removal is biological.
would cease or run incompletely as a result of a lack of Therefore, mechanistic and structured mathematical
electron donors. models are needed for describing the major biological
events occurring within the system or the behavior of
the microorganisms responsible for the treatment. Among
BIBLIOGRAPHY
the mechanistic mathematical models available, the IWA
(International Water Association) AS models (1) have been
1. S. W. Watson et al., in J. T. Staley, M. P. Bryant, N. Pfennig,
and J. G. Holt, eds., Bergey’s Manual of Systematic Bac- recognized as the worldwide standard. In the present
teriology, Williams & Wilkins, Baltimore, Md., 1989, article, the fundamental concepts and the underlying
pp. 1822–1834. microbiological bases of the IWA models are critically
2. U. Purkhold et al., Appl. Environ. Microbiol. 66, 5368–5382 discussed.
(2000).
3. D. Sorokin et al., Arch. Microbiol. 176, 170–177 (2001).
HISTORY OF IWA AS MODELS
4. E. Bock, I. Schmidt, R. Stüven, and D. Zart, Arch. Microbiol.
163, 16–20 (1995).
A pioneer work on AS modeling was carried out in the
5. I. Schmidt and E. Bock, Arch. Microbiol. 167, 106–111 (1997). late 1970s to early 1980s by the group of G. v. R. Marais
6. M. S. M. Jetten et al., FEMS Microbiol. Rev. 22, 421–437 from University of Cape Town, South Africa. They
(1999). proposed basic ideas for general structures of AS
7. M. Poth, Appl. Environ. Microbiol. 52, 957–959 (1986). models (2), chemical oxygen demand (COD) fractionation
8. D. Zart, I. Schmidt, and E. Bock, Antonie van Leeuwenhoek for modeling purpose (2), application of oxygen uptake rate
77, 49–55 (2000). measurements for determination of model parameters (3),
9. D. Zart and E. Bock, Arch. Microbiol. 169, 282–286 (1998). and so on. Their work has been the basis for the
10. I. Schmidt, E. Bock, and M. S. M. Jetten, Microbiology 147, development of the IWA AS models.
2247–2253 (2001). In 1983, a task group was organized within the
11. M. Strous et al., Nature 400, 446–449 (1999). International Association on Water Pollution Research
12. M. Schmid et al., Syst. Appl. Microbiol. 23, 93–106 (2000). and Control (IAWPRC, later renamed as the International
 ACTIVATED SLUDGE MODELS: MICROBIOLOGICAL BASIS 15

Association on Water Quality [IAWQ] and then as in the system; and (4) prediction of the behavior of
the IWA) to formulate general mathematical tools for concerned microbes in the system. Recently, the impor-
describing the behavior of the AS process. In 1987 they tance of mechanistic mathematical models is increasing
published a mechanistic mathematical model called the significantly because of the progress of computer tech-
IAWPRC Activated Sludge Model No. 1 (ASM 1) that nology. Huge computations can be performed within
incorporated carbon and nitrogen removal (4). This model a short period of time, which allows these AS mod-
became very popular and was adopted as the standard els to be complex, without limitations of computation
AS model worldwide. In the present article, ASM 1 time, and to be applicable not only for academic or
will not be discussed in detail because newer models educational purpose but also for more practical uses
have since been developed. In 1991, the IAWQ task such as design, process control, and operation sup-
group was reorganized, and they attempted to formulate port. Increasing demand for nitrogen and phosphorus
an updated model that included enhanced biological removal (see ACTIVATED SLUDGE — THE PROCESS, ACTIVATED
phosphorus removal (EBPR) as well as removal of carbon SLUDGE — MICROBIOLOGY OF NITROGEN REMOVAL, ACTIVATED
and nitrogen. Thus, ASM 2, describing carbon, nitrogen, SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY
and phosphorus removal, was published in 1995 (5). In COMPOSITION) has facilitated the increased application of
1999, two additional models were proposed: ASM 2d (6) such mechanistic AS models in process control. Precise pro-
was a partial extension of ASM 2, and ASM 3 (7) was cess control is not necessarily needed for carbon removal,
developed to overcome many defects in ASM 1, 2, and 2d but it is extremely important for nitrogen and phosphorus
and was designed as the core model that formed the basis removal.
for further extension. As IAWPRC or IAWQ has been now The AS process is a biological process in which carbon
changed to IWA, all these models presented by the task removal is carried out by heterotrophic organisms, nitro-
group are hereafter called the ‘‘IWA’’ models. The main gen removal is achieved through biological nitrification
characteristics of these IWA AS models are summarized and denitrification, and EBPR is made possible because
in Table 1. a group of bacteria are able to accumulate intracellular
Apart from the IWA AS models, a new approach was polyphosphate (poly-P) (13). The more these biological pro-
proposed by the group of van Loosdrecht and Heijnen from cesses have been understood, the more complex, but more
Delft University of Technology, the Netherlands (8–10) generally applicable, the models developed have become.
for modeling the EBPR processes. They developed a The use of AS mathematical models allows wastewa-
structured metabolic model for the EBPR processes based ter engineers to test many different possible designs or
on bioenergetics and stoichiometry of known metabolic operational strategies by simulation, without carrying
reactions. All relevant metabolic reactions underlying the out expensive and time-consuming pilot plant studies.
EBPR metabolism were described, and their reaction rates Therefore, once conceptual understanding of the AS pro-
were correlated with each other by applying stoichiometry- cess had expanded sufficiently, attempts were made to
based linear relations between these reaction rates. This incorporate microbiological and biochemical information
approach allowed for significant reduction in the number of relevant biological processes into AS models. Mechanis-
of kinetic or stoichiometric parameters to be determined: tic and structured mathematical models have thus been
it was possible to reasonably describe the dynamic formulated that can describe in general terms the major
behavior of EBPR processes with a set of 14 parameters. events occurring within the system or the behavior of
On the other hand, the apparent complexity of the the microorganisms responsible for the treatment. The
model increased and much biochemical knowledge is IWA models are the most commonly accepted models of
required to understand the behaviors of the model. This such kind.
Delft model is too complicated to be used for practical
purposes. However, as this Delft model describes more
Processes and Components
detailed biochemical structures of the EBPR metabolism
in a scientific way, it is considered to be a promising In AS systems, different groups of microorganisms per-
research tool. form various biological reactions, such as carbon oxidation,
Currently, commercial software for running AS models nitrification, denitrification, and poly-P accumulation.
are available in the market, and many attempts have been These biological reactions actually determine the concen-
made to use AS models for either practical or scientific trations of different system components. In AS modeling,
purposes (11,12). In such modeling practices, the IWA AS the term ‘‘process’’ is used to define an event that occurs
models are recognized as benchmark models. independently in the system (4) and is concerned with
the transformation or conversion of one or more system
components such as the microorganisms, COD, and ammo-
BASIC CONCEPTS OF IWA AS MODELS
nia. So, the identification of important processes affecting
the system behavior and the selection of key components
Mechanistic Models
relevant to the identified processes are two major tasks
There are several essential objectives of AS modeling. required for AS modeling. From a modeling point of view,
These include: (1) prediction of effluent water quality microorganisms are classified into a few groups, each of
in terms of key water quality indexes, such as COD, which should reasonably represent a functional microbial
nitrogen, and phosphorus species; (2) estimation of excess group and be responsible for at least one process (4). The
sludge production; (3) estimation of oxygen consumption system components and the processes in the model must
Table 1. Characteristics of IWA AS Models
No. of
Components
Year Target No. of Microbial Groups Characteristics, Advantages, and
Abbreviation Published Pollutant Processes Soluble Particulate Described Constraints of Model

ASM 1 1986 C, N 8 7 6 Heterotrophs, The first model proposed by the

nitrifiers IAWPRC Task Group, which became
a standard AS model and the
common basis for further research
A structured/mechanistic model simu-
lating carbon and nitrogen removal
Peterson’s matrix format is adopted
for the presentation of the model,
offering the best opportunity to the
users to trace all the interactions of
the system components
Heterotrophs are assumed to perform
not only the aerobic growth with
the oxidation of organic substrates,
but also denitrification
It is assumed that hydrolysis of
high-molecular compounds is done
only by heterotrophs
As a general rule, the Monod type
kinetic expression is used
ASM 2 1995 C, N, P 19 9 10 Heterotrophs, Biological phosphorus removal is
nitrifiers, modeled in addition to carbon and
PAOs nitrogen removal, thus PAOs are
introduces as bacteria responsible
for EBPR
Cell internal storage components,
PHA, and polyphosphate, are intro-
duced to describe the behavior of
PAOs
The COD fraction is assumed to carry
a defined amount of nitrogen and
phosphorus. The ammonification of
organic nitrogen, which is adopted
in ASM 1, is replaced by hydrolysis
of nitrogen-containing COD
The hydrolysis process eventually
involves not only true hydrolysis of
complex molecules but many other
processes such as degradation of
stored substrates and predation by
higher animals
ASM2d 1998 C, N, P 21 9 10 Heterotrophs, An extension of ASM 2: denitrifying
nitrifiers, capability is given to PAOs so that
PAOs anoxic phosphorus uptake can be
realized in the model
Basic structure is the same as that of
ASM 2
ASM 3 1998 C, N 12 7 6 Heterotrophs, An extension of ASM 1: phosphorus
nitrifiers removal not included, and only
carbon and nitrogen removal
modeled
Developed by revising some defects in
ASM 1 and 2, and designed as the
core model for further extension
The concept of biomass decay in
ASM 1 and 2 is replaced by the
endogenous respiration concept
All heterotrophs are assumed to be
able to store cell internal storage
compounds

Note: PAO–polyphosphate accumulating organisms, EBPR–enhanced biological phosphate removal.

16
 ACTIVATED SLUDGE MODELS: MICROBIOLOGICAL BASIS 17

be defined as simple as possible, but should be complex processes and microbial groups adopted in ASM 2, 2d, and
enough to describe the system behavior in such a way 3 are listed in Table 2.
as to fit the model user’s purpose (4). The components
in the model do not always represent something identi- Stoichiometry and Kinetics
fiable in reality. In the IWA models, there are several
Then, the stoichiometry (quantitative relationship bet-
variables clearly defined on the basis of certain modeling ween components in a chemical reaction) and the kinetics
concepts that are neither measured by chemical analyses (dependency of process rate on the concentrations of
nor separated by physical means. Similarly, the processes relevant components) should be defined for each process.
in the model may not always reflect the corresponding When the stoichiometry of a process is determined, the
reactions in reality. For example, ‘‘hydrolysis’’ in the IWA continuity of the process must be strictly maintained.
models should not only mean true hydrolytic biochemical Continuity is a term that carries the mathematical
reactions but may also include other processes such as equivalence to the principle that in chemical reactions,
predation by higher animals and utilization of intracel- elements, electrons (or COD), and net electrical charges
lular storage polymers (see details a later section). The may neither be created nor be destroyed (4,5). For the

Table 2. Biological Processes in ASM 2, 2d, and 3

Model Microorganisms Responsible for the Process Biological Processes

ASM 2 Heterotrophs Aerobic hydrolysis

Anoxic hydrolysis
Anaerobic hydrolysis
Aerobic growth of heterotrophs on fermentation
products
Aerobic growth of heterotrophs on fermentable
readily biodegradable organic substrate
Anoxic growth of heterotrophs (denitrification)
on fermentation products
Anoxic growth of heterotrophs (denitrification)
on fermentable readily biodegradable organic
substrate
Fermentation
Lysis of heterotrophs
PAOs Storage of PHA
Storage of poly-P
Aerobic growth of PAOs
Lysis of PAOs
Lysis of poly-P
Lysis of PHA
Autotrophs (nitrifiers) Aerobic growth of nitrifiers (nitrification)
Lysis of nitrifiers
ASM 2d Heterotrophs Same as ASM 2
PAOs Storage of PHA
Aerobic storage of poly-P
Anoxic storage of poly-P
Aerobic growth of PAOs
Anoxic growth of PAOs
Lysis of poly-P
Lysis of PHA
Autotrophs (nitrifiers) Same as ASM 2
ASM 3 Heterotrophs Hydrolysis
Aerobic storage of cell internal product
Anoxic storage of cell internal product
Aerobic growth of heterotrophs
Anoxic growth of heterotrophs (denitrification)
Aerobic endogenous respiration of heterotrophs
Anoxic endogenous respiration of heterotrophs
Aerobic endogenous respiration of cell internal
storage product
Anoxic endogenous respiration of cell internal
storage product
Autotrophs (nitrifiers) Aerobic growth of nitrifiers
Aerobic endogenous respiration of nitrifiers
Anoxic endogenous respiration of nitrifiers
18 ACTIVATED SLUDGE MODELS: MICROBIOLOGICAL BASIS

kinetics, the Monod model (14) is generally adopted as the ammonia to nitrite and nitrate), and phosphorus accu-
rate expression for the biological growth process in the mulating organisms (PAOs, microorganisms that have
IWA models: a capability to accumulate poly-P and are responsi-
ble for EBPR). Functions of these three biomass frac-
vXS tions in ASM 2 are schematically shown in Figures 1 to
Process Rate =
(S + K) 3.
Heterotrophs are modeled as ‘‘almighty’’ microorgan-
where v is the rate coefficient for the process, X the isms that can grow aerobically, denitrify, hydrolyze com-
concentration of biomass that performs the process, plex organic compounds and carry out fermentations.
S the limiting substrate concentration, and K the Nitrifiers are autotrophs and do not contribute to organic
saturation coefficient. carbon removal. They are involved in the oxidative part
The Monod model is generally used, not because it may of biological nitrogen removal. PAOs are a group of het-
either best fit the experimental data or have a definite erotrophs, but they are considered as a separate group
microbiological basis. Generally speaking, the growth rate when EBPR is modeled. The present classification of
of an organism is dependent on the limiting substrate microorganisms is too simple, in some cases, to describe
concentration in relatively low substrate concentration the complicated and poorly understood behavior of AS pro-
ranges, whereas its growth rate is not affected by substrate cesses. However, it is sufficient to simulate major events
concentration at higher concentrations. The Monod model
that actually affect fundamental system design or control
can express this general dependency of the process rate
strategy of the AS process.
on the substrate concentration reasonably well in simple
terms. Only for a few specific processes other types of
kinetic expressions are adopted, because of several specific (Anoxic) (Aerobic)
reasons; for example, a kinetic model for a surface- SNO3 SO2
limited reaction is used for hydrolysis processes (4) and
a saturation equation is used for poly-P storage (5). XS SF SA
After both the stoichiometry and the kinetics of the
concerned system are defined, the mass balance for each Growth on SA
Growth on SF
component may be created for a given system boundary Hydrolysis
Fermentation
(e.g., a completely mixed AS reactor). Various processes
can affect the concentration of a specific component, XH : Heterotrophs
and all the processes affecting the concentration of that
particular component are taken into account in the mass Lysis
balance. From the mass balance, the concentration of
each system component can be calculated and the system XI
behavior can then be simulated (4). In order to simulate
carbon oxidation, nitrogen, and phosphorus removal, COD Figure 1. Transformation of organic substrates by heterotrophs
(measured using potassium dichromate as the oxidizing in ASM 2. SNO3 : nitrate and nitrite nitrogen, SO2 : dissolved
reagent), nitrogen and phosphorus are selected as the oxygen, SF : readily fermentable biodegradable substrates, SA :
basis for mass balance in the IWA models (4,5). Instead of fermentation products, XS : slowly biodegradable substrates, XH :
COD, carbon itself can be used for the mass balance for heterotrophic biomass, XI : inert organics.
organic substances (7). The advantage of using COD (or
actually the oxygen equivalent in the oxidation-reduction
reaction) for the mass balance is that the oxygen balance N fraction in XS S N2
can be calculated and thus used to predict the oxygen
requirement in the system (4). Because the AS process S F or SA
Hydrolysis
is an oxygen-dependent system, it is extremely useful, in Anoxic growth of XH
XH (denitrification)
practice, that the simulation model can predict reasonably S O2
the oxygen concentration in the system and its oxygen XH
requirement. S NH4 S NO3
Growth of XAUT
(nitrification)
ESSENTIAL PROCESSES IN IWA MODELS
XS XAUT : Autotrophs
General
Lysis
Processes mainly adopted in ASM 2 are discussed in this
section. For modeling purposes, microorganisms in the XI
AS are usually classified into several functional groups.
Figure 2. Function of autotrophs (nitrifiers) in biological nitro-
In ASM 2, only three groups of microorganisms are rec- gen removal in ASM 2. SNH4 : ammonia nitrogen, SN2 : nitrogen
ognized and represent a broad range of different micro- gas, SNO3 : nitrate and nitrite nitrogen, SO2 : dissolved oxygen, SF :
bial populations. These three groups are: heterotrophs readily fermentable biodegradable substrates, SA : fermentation
(microorganisms that live on organic substrates as carbon products, XS : slowly biodegradable substrates, XH : heterotrophic
and energy sources), nitrifiers (autotrophs that oxidize biomass, XAUT : autotrophs (nitrifiers), XI : inert organics.
 ACTIVATED SLUDGE MODELS: MICROBIOLOGICAL BASIS 19

Glycogen that RB-COD represents the low-molecular or soluble

[H+] fraction (4,5).
Succinate-
From a modeling point of view, SB-COD is defined as
Glycolytic the COD fraction that is hydrolyzed to smaller molecules
propionate
pathway
pathway before being utilized, and RB-COD is defined as the COD
External
[H+] [H+] substrates fraction that can be directly utilized by heterotrophs
(Ex. axetate) with no prior hydrolysis (4). In other words, SB-COD
Propionyl-CoA Acetyl-CoA need to be first hydrolyzed to RB-COD and only then
is utilized for heterotrophic growth. The conversion of SB-
COD to RB-COD is referred to as ‘‘hydrolysis’’ in the IWA
PHA
synthesis [H+] models (5). The two observed rates in the profile can be
interpreted as the hydrolysis rate and the aerobic growth
3HA Unit (PHA) rate of heterotrophs. However, the distinction between
Figure 3. Anaerobic carbon metabolism of PAOs for enhanced
SB- and RB-COD is not always done clearly in practice.
biological phosphate removal. Glycogen functions as regulator of Some soluble COD fractions may contain rather complex
redox balance in the cell. molecules that cannot be utilized directly by heterotrophs
and should be classified as SB-COD even though they
are soluble (5). The present COD fractionation is simple
Aerobic Growth of Heterotrophs and Carbon Removal enough for practical use and complex enough to reasonably
The major mechanism of carbon removal in the AS process describe the behavior of AS systems under most situations.
is by the growth of heterotrophs on organic substrates. In ASM 2, the RB-COD faction is further divided
These heterotrophs include diverse phylogenetic, taxo- into two fractions, ‘‘fermentation products (SA )’’ and
nomic, and functional groups (15), but all, except for PAOs ‘‘fermentable readily biodegradable substrate (SF ).’’ Thus,
as discussed later, are assumed to be one functional group
of microorganisms in ASM 2. Their common nature is SS = SF + SA
that they all utilize organic substrates as the carbon and
energy sources for growth. The process of conversion from SA to SF is defined as
In heterotrophic growth, organic substrates are used fermentation. This fractionation as well as the process of
in two ways: one for the energy-generating metabolism fermentation are necessary when EBPR is modeled, as
(adenosine triphophosphate [ATP] production) and the discussed later. The transformation of organic substrates
other for assimilative metabolism (biosynthesis of the cell). by heterotrophs, including the fermentation step, is shown
Some substrates in the wastewater are oxidized completely graphically in Figure 1.
to carbon dioxide and utilized for energy (ATP) generation. In the real world, many wastewaters do not show
This is an oxidation process and requires a terminal the typical two-stage response in the OUR test. For
electron acceptor. The AS reactors are usually aerated and example, OUR profiles of some municipal wastewaters
kept aerobic so that oxygen serves as the electron acceptor. show a gradual decreasing trend along with time when
The rest of the substrates are used for the synthesis of organic substrates are removed in a batch test (17).
biomass and withdrawn from the AS system as excess Such results imply that a wastewater may contain many
sludge. The ratio of this carbon distribution is represented compounds with different utilization rates. However, it is
by the yield (Y, the ratio of mass of produced biomass not recommended to increase the number of COD fractions,
to mass of the substrate utilized) in the AS models (16). because determination of each COD fraction as well as
This yield is assumed to be a constant, an assumption the parameters associated with it becomes increasingly
that is generally acceptable for wastewaters with similar difficult as the numbers increase. On the other hand,
compositions. the composition of industrial wastewaters depends totally
Wastewaters may contain a large variety of organic on the type of industry and the industrial processes used,
compounds that are utilized by microorganisms at and a few major organic components can be often identified
different rates. Therefore, an important question to be from the raw materials or from the processes. Then, each
asked when heterotrophic growth is modeled is: how major organic component can represent a COD fraction
many fractions of organic substrates should be defined that has a single process rate. Specialist microorganisms
in the model to satisfactorily describe the substrate can also be defined, which thrive on certain particular
removal profiles in the AS system? In the IWA models, components. When the wastewater composition is defined
the organic substrates are principally divided into two clearly enough and the process rates for the identified
fractions, namely, slowly biodegradable substrate (SB- components are measurable, it is possible to assign a
COD, denoted as XS ) and readily biodegradable substrate process rate to each relevant component and, in more
(RB-COD, denoted as SS ). This is because the oxygen limited cases, a specialist microorganism that utilizes it.
uptake rate (OUR) during aerobic substrate removal from
a municipal wastewater showed (3) a two-stage response,
Denitrification and Nitrification
which implies that organic substrates present in municipal
wastewaters can be divided into two fractions with two Biological nitrogen removal is achieved by the combination
different OURs. It is instructive to think that SB-COD of autotrophic nitrification and heterotrophic denitrifi-
represents the high-molecular or particulate fraction and cation (see ACTIVATED SLUDGE — MICROBIOLOGY OF NITROGEN
20 ACTIVATED SLUDGE MODELS: MICROBIOLOGICAL BASIS

REMOVAL). The nitrogen conversion processes modeled in through nitrification has been studied in detail, but there
ASM 2 are shown graphically in Figure 2. is very little kinetic information available about the nitrite
Under anoxic conditions (i.e., where oxygen is not production through denitrification (5). This means that the
available but nitrate or nitrite is), nitrate or nitrite can be nitrite concentration is not satisfactorily predictable.
used as electron acceptors in heterotrophic growth (15). Nitrous oxide (N2 O) is a nitrogenous species that can
This anoxic growth is referred to as denitrification. be produced both in nitrification and denitrification and
Denitrifying heterotrophs that grow anoxically usually should not be neglected in constructing mass balance
grow aerobically utilizing oxygen as electron acceptor. of nitrogen (22). N2 O is of concern these days from an
Thus, denitrifiers should be a part of aerobic heterotrophs. environmental point of view, because it is a greenhouse gas
In ASM 2, heterotrophs are defined as one biomass fraction that may contribute to global warming (23). However, N2 O
and modeled in such a way that all heterotrophs can production is not taken into account in the IWA models.
denitrify. But, this is not true in reality. The mechanism and the conditions for N2 O production are
Aerobic and the anoxic growth of heterotrophs are still poorly understood. In practice, it is almost impossible
very similar to each other in terms of stoichiometry and to determine satisfactorily all the kinetic parameters
kinetics, but differ in that the observed specific growth involved in these complex processes affecting its behavior.
rates of heterotrophs are usually lower under anoxic
conditions than under aerobic conditions (18). A factor (the EBPR
reduction factor for denitrification) is introduced to allow
for this reduced growth rate under anoxic conditions (4,5). Phosphorus is an essential element for biomass growth,
This factor is defined as the ratio of the specific growth rate and the sludge produced as excess sludge will contain
under anoxic conditions to that under aerobic conditions, a certain amount of phosphorus. Thus, phosphorus can
which actually implies the portion of denitrifiers within be removed from wastewater by withdrawing the excess
the total heterotrophic biomass (5). An advantage of the sludge from the AS system. However, municipal wastewa-
introduction of this factor is that most actual situations ters usually contain more phosphorus than is required for
can then be simulated without increasing the number of the biomass growth, so phosphorus in the influent may not
processes. However, this factor may not be a universal be completely removed only by normal microbial growth
constant but a parameter that should be calibrated for equivalent. On the other hand, the phosphorus content of
each specific system or situation, because the ratio of the sludge may increase up to 5 to 6% or even more than
denitrifiers may vary according to different environmental 10% in extreme cases, whereas the phosphorus content of
conditions. It is of course possible to create a model a typical AS may be only about 1.5 to 2% of phosphorus
that involves two heterotrophic biomass fractions, one (on dry weight basis) (13). Such high phosphorus con-
with and the other without denitrification capability. tent is caused by an excess accumulation of poly-P in the
The dual-heterotroph model may describe denitrification sludge and may lead to much higher phosphorus removal
behavior of systems more sensitively than the IWA models. efficiency than usual (see ACTIVATED SLUDGE — MOLECULAR
However, the complexity of the model will increase and TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION). This
the identification of stoichiometric or kinetic parameters high phosphorus removal by biological means is referred
will become more difficult. to as EBPR. The bacteria that can achieve the high phos-
Nitrification (19) is a process in which ammonia phorus content through the accumulation of poly-P and
is biologically oxidized to nitrite and nitrate in the are thus responsible for EBPR are called PAOs. It is nec-
presence of oxygen as the terminal electron acceptor. essary to describe the behavior of PAOs in mathematical
Nitrifiers are autotrophic bacteria and use inorganic terms to incorporate EBPR into the AS model.
carbon (carbon dioxide) as the carbon source. Nitrification A high phosphorus content of the sludge from EBPR
involves two steps: oxidation of ammonia to nitrite, and can be achieved by introducing an anaerobic phase
further oxidation of nitrite to nitrate (see ACTIVATED into the influent end of the AS process, followed by
SLUDGE — MICROBIOLOGY OF NITROGEN REMOVAL). These two conventional aerobic phase (see ACTIVATED SLUDGE — THE
steps are carried out by phylogenetically different PROCESS; 24). In such EBPR processes, the AS passes
groups of bacteria. The ammonia oxidizers include through anaerobic and aerobic conditions alternately,
Nitrosomonas and Nitrosospira and nitrite oxidizers, and the influent wastewater (or actually the organic
including Nitrobacter and Nitrospira (20). However, the substrates) is introduced into the anaerobic phase.
IWA models do not distinguish between these two groups In practice, this anaerobic-aerobic alternation can be
of nitrifiers, but only one biomass fraction is used for achieved either by spatial configuration of anaerobic
all bacteria that carry out complete nitrification. There and aerobic zones in series with sludge recycle in
are two major reasons for this. First, the growth rate of continuous flow reactors or by temporal arrangement of
nitrite oxidizers is usually higher than that of ammonia anaerobic and aerobic periods in sequence batch reactors
oxidizers. Therefore, nitrite does not accumulate in the (see ACTIVATED SLUDGE — SEQUENCING BATCH REACTORS). Such
system, but is consumed by nitrite oxidizers immediately EBPR processes are called anaerobic-aerobic or anaerobic-
after its production by ammonia oxidizers (21). So, in oxic processes. By exposing the sludge to these alternative
most cases, nitrification takes place as if it were a single anaerobic and aerobic conditions, PAOs are selected and
process. Secondly, nitrite can be produced not only as grow to become dominant in the process. To achieve high
an intermediate product in nitrification but sometimes as and stable EBPR performance, it is essential to maintain
a by-product of denitrification (22). Production of nitrite PAOs in the system (25,26).
 ACTIVATED SLUDGE MODELS: MICROBIOLOGICAL BASIS 21

The mechanism of proliferation of PAOs in the EBPR SO2

process has been described as follows (25–27; also see
another contribution of this Encyclopedia, ACTIVATED Poly-P Storage PAO's Cell
SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY
COMPOSITION). Because the AS is mixed with the influent SPO4 XPP : Polyphosphate
wastewater in the anaerobic phase, organisms that can
utilize organic substrates more rapidly in an anaerobic Lysis
environment have an advantage. PAOs are the organisms PHA Storage
that utilize the energy generated by the hydrolysis of
polyphosphate for the uptake of the carbon substrates SA XPHA : Polyhydroxyalkanoate
in the anaerobic phase of the EBPR process and
can proliferate over other microorganisms under such Lysis
conditions. The carbon substrates taken up by PAOs in the SO2 Phosphorus Accumulating
XPHO :
anaerobic phase are converted to polyhydroxyalkanoates XS Organisms (PAOs)
(PHA) and stored inside the cell (28). Because PHA is
a reduced polymer, its synthesis requires a source of
reducing power, which is obtained by partial oxidation of Lysis
intracellularly stored glycogen to carbon dioxide (29). The XI
glycogen is partially converted via acetyl coenzyme A (CoA)
Figure 4. Metabolism of PAOs in ASM 2. SPO4 : orthophos-
to polyhydroxybutyrate (PHB). Thus, uptake of organic
phate, SO2 : dissolved oxygen, SA : fermentation products, XS :
substrates and concomitant PHA synthesis by the sludge,
slowly biodegradable substrates, XPAO : phosphorus accumulat-
degradation of stored poly-P and the consequent release ing organisms, XPP : polyphosphate associated with XPAO, XPHA :
of orthophosphate to the bulk solution, and utilization polyhydroxyalkaonates associated with XPAO, XI : inert organics.
of stored glycogen are typical metabolic characteristics
of the anaerobic phase of EBPR processes (25). In the
subsequent aerobic phase, PAOs grow aerobically, recover
component, XPHA (cell internal storage product of PAOs)
the glycogen level, and assimilate orthophosphate to
resynthesize poly-P, by using the stored PHA as their is defined as carbon storage material in the model.
carbon and energy sources. Such typical EBPR metabolism This simplification is made in ASM 2 because the actual
is shown graphically in Figure 3. metabolism of the PAO is so unfamiliar to practitioners
Anaerobic metabolism of PAOs may involve glycogen that they may have difficulty in understanding the model
being metabolized via propionyl-CoA to 3-hydroxyvale- behavior fully. Glycogen and poly-P are the two key
rate-rich PHA (the propionate-succinate pathway). This storage polymers that enable PAOs to take up organic
conversion does not lead to carbon dioxide production substrates under anaerobic conditions. The former either
and consumes reducing power. This means that glycogen supplies or consumes reducing power to maintain the
can either produce reducing power if metabolized through redox balance in the cell and the latter generates energy
acetyl CoA, or consume reducing power if metabolized necessary for the uptake of organic substrates and for their
through propionyl-CoA. Glycogen stored in the cell should storage in the form of PHA under anaerobic conditions.
function as a regulator of the redox balance in the Both are essential factors for PAOs to survive in the
cell (25). The generation of reducing power should occur anaerobic-aerobic process. If glycogen is excluded from the
when more oxidative substrates are available, whereas model, situations in which glycogen, as the limiting factor,
its consumption should occur when more reductive determines the system behavior may not be simulated
substrates are utilized (30). It is essential for PAOs to properly. However, the relative size of the glycogen pool
take up carbon substrates at a faster rate than the other is usually almost the same as that of the poly-P pool (31).
bacteria in the anaerobic stage, no matter what kinds This means that most of the real situations in the operation
of substrates are available in the influent wastewater. of the EBPR processes, in practice, can be described with
The mechanism described earlier can provide PAOs with a only one of the two parameters: either glycogen or poly-
capability to assimilate different kinds of both reduced and P. In ASM 2, poly-P is selected as a model component,
oxidized organic substrates in the anaerobic phase without and glycogen, which is not familiar to practitioners, is not
disturbing the redox balance in the cell. Glycogen storage adopted.
appears to be the key strategy for PAOs to maintain the From the viewpoint of reaction stoichiometry, the
redox balance in the anaerobic uptake of various organic component XPHA in the model should represent all
substrates, and hence to emerge dominant in the face of carbon storage materials in PAOs, including PHA and
the selective pressures existing in EBPR processes (25). glycogen. Therefore, the model implies that the substrates
This EBPR mechanism is partially adopted in ASM 2. anaerobically taken up by PAOs are converted to PHA
Conceptual structures of the PAO metabolism adopted in without any loss or supply of carbon. However, a part
ASM 2 are schematically shown in Figure 4. A prominent of PHA is actually synthesized from glycogen that has
difference between this model structure and the generally already accumulated in PAOs (32,33). This argument
accepted scientific mechanism is that glycogen is not should lead to an important warning: XPHA in the
introduced in the model, whereas glycogen probably model does not necessarily represent the analytically
plays an essential role in the real world. Only one measured PHA.
22 ACTIVATED SLUDGE MODELS: MICROBIOLOGICAL BASIS

It is believed that some organic substrates are supplied through hydrolysis and too much denitrification
selectively utilized by PAOs. As short-chain fatty acid (under anoxic conditions) or phosphate release (anaerobic
such as acetate are favored as the carbon source for conditions) emerges from the model compared with that
EBPR (34), the fraction of RB-COD that can be utilized by actually observed in the AS systems. Differences in
PAOs is defined as SA in ASM 2. The fraction of RB- the hydrolysis rates under different electron acceptor
COD that can be utilized by heterotrophs other than conditions are allowed for by introducing anoxic and
PAOs but not by PAOs is defined as SF . For the growth anaerobic hydrolysis reduction factors. As true hydrolysis
of heterotrophs, SF and SA have no difference: both are is not an oxidative or reductive process, the availability
utilized by heterotrophs exactly in the same way in terms of electron acceptors should not affect the process rate.
of stoichiometry and kinetics. In the modeling context the However, several possible factors may contribute to
term ‘‘fermentation’’ is defined as the process in which the apparent reduced hydrolysis rates under anoxic or
SF is transformed to SA and is introduced to describe anaerobic conditions in the model as follows:
the possible selective preference of PAOs for certain
substrates. Fermentation in ASM 2 does not exactly imply 1. The activity of predators (protozoa and higher
its microbiological meaning, as discussed later. animals) may be less under anaerobic and anoxic
ASM 2d (6) is an extended version of ASM 2. It conditions than under aerobic conditions.
has been shown that some PAOs have denitrification 2. Carbon storage may take place under any of these
capability (35). In other words, some PAOs can utilize conditions, but the utilization of the storage products
nitrate or nitrite as the electron acceptor, and so they should be more under aerobic and anoxic conditions
take up orthophosphate with a consumption of nitrate than under anaerobic conditions.
or nitrite under oxygen-free conditions. To describe this
anoxic uptake of orthophosphate, this denitrification These may lead to an underestimation of the hydrolysis
capability is attributed to PAOs in ASM 2d (‘‘d’’ stands for rate in the model. The hydrolysis rate in ASM 2 is one of
‘‘denitrification’’) (6). PAOs with denitrification capability the most difficult parameters to estimate. This problem
(DN-PAOs) are very common in EBPR systems (35,36), has now partly been solved by the introduction of the
so without the introduction of DN-PAOs, the EBPR storage process in ASM 3, where this anoxic or anaerobic
processes incorporating recycling of nitrified liquid cannot hydrolysis reduction factor is no longer used (7).
be modeled properly. When nitrification is significant Municipal wastewaters usually contain organic nitro-
in the system, ASM 2d not ASM 2 should be used for gen compounds. After their hydrolysis, ammonia nitrogen
its simulation. An anoxic reduction factor (the factor is released. In the ASM models, it is assumed that SB-
expressing the ratio of the anoxic growth rate to that of COD should include a nitrogen fraction and that ammonia
the aerobic) is again introduced to account for the reduced is released when SB-COD is hydrolyzed to RB-COD. This
growth rate under the anoxic conditions (6). approach is extended to other COD fractions. Once the
nitrogen content of each COD fraction is determined exper-
Hydrolysis imentally, the model describes the behavior of organic
nitrogen compounds quite well. The same philosophy is
Hydrolysis in the IWA models is defined as the
also applied to describe the behavior of phosphorus bound
process in which complex molecules not metabolized
to organic matters, namely, each COD fraction include a
by microorganisms (SB-COD) are degraded to relatively
phosphorus fraction.
simple molecules that can then be assimilated and
metabolized directly by heterotrophic bacteria (RB-
Fermentation
COD) (4,5). Hydrolysis is not an oxidation-reduction
reaction and theoretically should not involve any electron The process ‘‘fermentation’’ was introduced in ASM 2
transfer. However, the hydrolysis process in the model to describe EBPR metabolism, and is different from
may include not only the real ‘‘biochemical’’ hydrolysis fermentation in the microbiological sense. In ASM 2, a
caused by hydrolytic exoenzymes, but also processes fraction of RB-COD utilized by PAOs (SA ) is distinguished
such as predation by higher animals or any other from the rest of RB-COD that cannot be directly taken
processes contributing to breakdown of polymers to low- up by PAOs (SF ). The conversion of the latter to the
molecular compounds. In addition, carbon storage may former is defined as fermentation. The process is called
be strongly correlated with the hydrolysis process. As fermentation, because compounds that are likely to be
modeled in ASM 3, external substrates can be polymerized classified as SA include fermentation products such as
and stored inside the cell when they are utilized by acetate and other short-chain fatty acids.
microorganisms. Subsequent breakdown of these internal In a microbiological context, fermentation is a growth
storage products to smaller metabolic intermediates generating process. However, fermentation in ASM 2 is
cannot be distinguished from the extracellular hydrolysis not modeled as such, but it is assumed to be a conversion
of SB-COD in the oxygen balance–based models, because process in which SF is simply converted to SA without any
neither upsets the oxygen balance. production of biomass. There are two major reasons for
In ASM 2, the hydrolysis rate is considered to be this assumption. First, the biomass yield in fermentative
greatest under aerobic conditions, medium under anoxic growth processes is small. Second, two yield values for the
(denitrifying) conditions, and lowest under anaerobic same heterotrophs (XH ), one for aerobic and anoxic growth
conditions (5). If the same hydrolysis rate is applied under processes and the other for fermentation processes, are
anaerobic, anoxic, and aerobic conditions, too much SS is confusing in the modeling context.
 ACTIVATED SLUDGE MODELS: MICROBIOLOGICAL BASIS 23

Decay, Lysis, and Endogenous Respiration Carbon Storage

When microorganisms are put into an environment in It has been pointed out that the carbon storage
which no substrates are available, a gradual decrease in capability of microorganisms plays a key role in the
biomass level will occur. Several possible mechanisms can ecological selection in biological wastewater treatment
be proposed to account for this biomass reduction. They (38). Microorganisms in AS are exposed to carbon-rich
include decay or lysis and endogenous respiration (37). and -deficient conditions alternatively. For example, either
Conceptual schemes of decay and endogenous respiration spatial or temporal gradients of substrate concentration
are given in Figure 5. are formed in many AS reactors, such as sequence
The term ‘‘decay’’ may be used to describe degradation batch reactors, plug flow reactors, systems with selector,
of biomass in general, in which the biomass is degraded and contact-stabilization processes (39). Also, carbon
and converted either to substrates that are then reutilized loadings from influent wastewater tend to fluctuate
by heterotrophic biomass or to inert organic substances dynamically from time to time. These dynamic conditions,
that are biologically inactive. Decay may occur partly in terms of the availability of carbon sources, will create
because of activities of lytic enzymes produced by the highly competitive environments, in which substrate
biomass and partly because of external factors such uptake rates during the carbon-rich periods will actually
as mechanical or physical stress, chemical attack, or determine the survival and proliferation of certain
predation activities (37). The indigenous enzymatic decay microorganisms (39,40). So, some microorganisms adopt a
is often called lysis. When the biomass reduction is strategy to store carbon sources in the cell under carbon-
modeled allowing for this decay, the decay rate does rich conditions by achieving a high substrate uptake rate.
not correspond to the biomass reduction rate, because They use this stored carbon sources under carbon-deficient
some biomass is regenerated by the secondary substrates conditions. In fact, PAOs have the capability to store
produced by the decay process itself. The first two IWA PHA under anaerobic conditions, and this explains their
models (ASM 1 and 2) adopt this approach. However, the dominance in anaerobic-aerobic EBPR systems (25). This
complex nature of this decay-regeneration model structure strategy for carbon storage seems to be adopted by aerobic
often leads to difficulties in evaluating relevant kinetic or anoxic heterotrophs as well, because storage seems to be
parameters. Therefore, ASM 3 adopts a different approach an essential capability of microorganisms to outcompete
by incorporating endogenous respiration. In this approach, others for substrates under dynamic conditions (39).
it is assumed that the biomass is continuously oxidized Therefore, the process of carbon storage is introduced
to carbon dioxide under aerobic or anoxic conditions, in ASM 3 (7).
thus generating the necessary energy for endogenous In ASM 3, a single heterotrophic biomass (XH ) is defined
metabolism or for cell maintenance. The essential as being responsible for carbon and nitrogen removal,
advantage of this endogenous respiration approach is its and this XH stores RB-COD (SS ) in the form of cell
simplicity. No regeneration of substrate is considered and internal storage products (XSTO ). It is assumed in the
the endogenous respiration rate then directly expresses stoichiometry of ASM 3 that XSTO is exclusively PHB.
the biomass reduction rate. On the other hand, biomass PHA, including PHB, are also likely to be major storage
reduction under anaerobic conditions cannot be described products, and possibly other polymers, such as lipids,
by this approach because no inorganic electron acceptors glycogen, and polypeptides, exist (16). At present, little
are available under anaerobic conditions. In reality, the information is available about the presence, importance,
decay rate seems to be very small and can be usually and characteristics of these carbon storage products in
neglected under anaerobic conditions. However, hydrolytic AS. The storage process is modeled as an energy-requiring
decay can theoretically occur even under anaerobic process, and the ratio of utilized SS to stored XSTO is defined
as the yield for the storage process. This yield should be
conditions and its significance in actual wastewater
much higher than yields of growth processes, because the
treatment processes still remains unclear.
energy requirement for storage is much smaller than that
for growth (41). Storage occurs both under aerobic and
anoxic conditions (36), and anoxic storage is identical to
(a) Decay
aerobic storage, but denitrification rather than aerobic
Growth Biomass
respiration provides the energy required. As more energy
is produced in aerobic respiration than in denitrification,
Substrate a higher yield is applied to aerobic storage than to anoxic
Decay storage.
An assumption that may not be true in the real world
(b) Endogenous but is incorporated into the models, is that all SS taken
respiration up by XH is first converted to and stored as XSTO , which
Growth Biomass is then utilized by XH as the substrate for growth. This
means that XH accumulate and utilize XSTO at the same
Substrate time during their growth. Heterotrophs utilize external
Endogenous substrates not necessarily via storage products but directly
respiration for growth. However, it is not wise, during modeling, to
Figure 5. Conceptual scheme of decay and endogenous respi- provide two heterotrophic growth processes (one on SS
ration. and the other on XSTO ), because more kinetic parameters
24 ACTIVATED SLUDGE MODELS: MICROBIOLOGICAL BASIS

must then be identified. To avoid such a complexity in problem in the operation of the AS process, because
the model structure, the foregoing assumption is made in solutions to it have not yet been established in practice.
ASM 3. From a practical point of view, there is a high demand
for models to describe the bulking phenomena, but widely
accepted bulking models have not been established yet.
LIMITATIONS AND FUTURE PERSPECTIVES OF AS MODELS
The main reason is that the mechanism of bulking
is not clearly understood. Many different filamentous
It is important to note that microbiological information
organisms can cause bulking and each of them may
is not always incorporated into these AS models. There
have different microbiological and kinetic natures (46).
are many processes that are known to microbiologists
It is also very difficult to quantify filamentous biomass
but that have not yet been introduced to IWA models
reliably. The conditions responsible for bulking are not
for several reasons. It is assumed by the AS models that
always known. All these difficulties lead to lack of definite
processes with engineering significance should be modeled.
information about the mechanism of bulking, and thus to
In other words, only those processes that affect the
concentrations of system components relevant to practical lack of acceptable bulking models. It is essential to obtain
objectives should be considered. Thus, the activities of more detailed microbiological and kinetic information on
absolute anaerobic bacteria, for example, are neglected different causative filaments to establish reliable bulking
in the IWA models, despite the fact that various strict models.
anaerobic bacteria, such as sulfate-reducing bacteria, Predation by protozoa or higher animals commonly
can be detected even in fully aerobic AS processes (42). occurs in the AS process (see PROTOZOA IN ACTIVATED SLUDGE).
However, their activities are usually considered to be Predation is not included in the IWA models but may
negligible in terms of the COD balance and they are not contribute to decay and hydrolysis in the modeling context.
introduced in the AS models. As their metabolisms are Several different types of predators are present in the AS,
well understood, it would be relatively simple to model and some of their characteristics have been studied. They
them if necessary. can be modeled relatively easily from a structural point of
As far as nitrogen removal is concerned, the model view. However, predator populations cannot be quantified
structures are far behind the microbiological information, easily, and thus identification of related parameters may
and so the following are not considered in ASM 2 or other be difficult.
IWA models, as discussed previously. Soluble microbial products (SMP) are often considered
as important COD fraction in wastewater treatment
1. Nitrite production, either through nitrification or systems (47). When decay or lysis of the whole cell or
denitrification, is not taken into account. Nitrite a part of the cell occurs, secondary substrates will be
accumulation and its inhibitory effects on microbial released into the bulk solution. This process is modeled
activities are not predicted. as decay or lysis process in ASM 1 and 2. It is assumed
2. N2 O production is not considered. It is known that in ASM 1 and 2 that the biomass is degraded to SB-COD
under certain conditions part of nitrogen can be and inert COD. This means that the secondary substrates
released to the air as N2 O in the courses of both produced through decay or lysis are exactly the same as
nitrification and denitrification (22). This leads to other SB-COD fractions from a modeling perspective. This
additional loss of nitrogen from the system, and to is not likely: secondary substrates produced by the biomass
an incomplete mass balance of nitrogen, unless it is will almost certainly have different kinetic characteristics.
taken into account. Therefore, SMP should be defined as a separate component
representing the secondary substrates produced by the
3. There are several other nitrogen-related metabolic
lytic activities of the cells. Contribution of the SMP to the
reactions, such as those involved in anaerobic ammo-
COD balance is more significant in biofilm systems than
nia oxidation known as ANAMMOX (43), in which
in AS systems, and SMP is sometimes introduced in the
ammonia is oxidized by nitrite to produce nitrogen
gas, autotrophic denitrification (44) (denitrification modeling of the biofilm process (47).
by inorganic compounds as electron donor), and so It is easy to allow a model to become more complex based
on. They will disturb the nitrogen balance, but are on microbiological information. For example, we have only
not considered at all in the IWA models, because it one heterotrophic biomass in ASM 1 and 3, but this
is considered unlikely that these metabolisms are can be divided into many different heterotrophic groups,
significantly active in the AS systems. including those with and without denitrifying capability or
with and without storage capability. One may think that
Nitrite production during nitrification is more common such a multiple heterotroph model should have a higher
than that in denitrification, and more microbiological and capability to describe more comprehensive situations, but
kinetic information is available about the former than the this is not always true. Parameter identification becomes
latter. Therefore, nitrite production during nitrification more difficult as the complexity of the model increases,
can be modeled satisfactorily and such models may be and the predictability of models may decrease. As stated
applied to situations in which nitrite production through before, it is evident that cell internal glycogen plays an
denitrification is not significant. essential role in the EBPR metabolism, but glycogen is
Bulking is a situation in which AS does not settle not incorporated into ASM 2 so that the structure of the
in the secondary clarifier because of excessive growth of model may not be too complicated for practitioners. In the
filamentous organism (45). This is a serious and critical practical application of the AS model, most of the PAO
 ACTIVATED SLUDGE MODELS: MICROBIOLOGICAL BASIS 25

behavior can be described with only one of the two storage 7. W. Gujer et al., Water Sci. Technol. 39, 183–193 (1999).
polymers: poly-P or glycogen. It is reasonable, therefore, 8. G. J. F. Smolders et al., Biotechnol. Bioeng. 43, 461–470
that poly-P, which can be easily understood by wastewater (1994).
engineers, is adopted in ASM 2 to reduce its complexity. 9. G. J. F. Smolders et al., Water Sci. Technol. 31, 9–93 (1995).
However, glycogen in PAOs is sometimes essential and 10. G. J. F. Smolders et al., Biotechnol. Bioeng. 47, 227–287
critical if the AS models are to be used for scientific (1995).
purposes. In fact, several AS models with glycogen have 11. T. Ohtsuki et al., Water Sci. Technol. 37, 77–85 (1998).
been proposed (48).
12. G. Olsson and B. Newell, Wastewater Treatment Sys-
The concentration of hydrogen ion (H+ ) or pH is not
tems — Modeling, Diagnosis and Control, IWA Publishing,
considered as a state variable in the IWA models. This is London, U.K., 1999.
simply because pH is not exactly predictable. To describe
13. T. Mino et al., Water Sci. Technol. 17, 93–106 (1984).
the behavior of pH (H+ ), it is necessary to be able to
14. D. Herbert, in G. Tunevall, ed., Recent Progress in Micro-
model the acid-base equilibrium. However, many chemical
biology, Almquist and Wiksell, Stockholm, Sweden, 1958,
species can affect this in wastewater or in the AS systems
pp. 381–396.
in general. It is impossible to include all these species into
the model. On the other hand, the acid-base equilibrium is 15. R. S. Stanier et al., Microbial World, 5th ed., Prentice-Hall,
Englewood Cliffs, N.J., 1986.
well understood theoretically and it is easy to create a pH
prediction model once the species that are involved in the 16. G. Tchobanoglous and F. L. Burton, Wastewater Engineer-
this equilibrium are identified. For example, the carbonate ing — Treatment Disposal and Reuse, 3rd ed., McGraw-Hill,
New York, 1991.
equilibrium can be introduced into the AS model to predict
pH. In oligotrophic water environments, the carbonate 17. E. W. Randall et al., Water SA 17, 11–18 (1991).
equilibrium actually determines the pH in the system. 18. T. Kuba et al., Biotechnol. Bioeng. 52, 685–695 (1996).
So, a carbonate-based pH prediction module of the AS 19. J. I. Prosser, Adv. Microb. Physiol. 30, 125–181 (1989).
model can be developed as a first step. For this purpose, 20. A. Teske et al., J. Bacteriol. 176, 6623–6630 (1994).
ASM 3c (49), an adapted version of ASM 3, where organic 21. G. Keen and J. Prosser, Arch. Microbiol. 147, 73 (1987).
components are expressed in terms of organic carbon 22. H. Itokawa et al., Water Res. 35, 657–664 (2000).
rather than COD, may be useful. If carbon is used as the
23. K. Hanaki et al., Water Sci. Technol. 26(5–6), 1027–1036
basis for mass balance, carbon dioxide should be accounted
(1992).
for in this balance and the pH prediction model based on
24. J. L. Barnard, Water Res. 9, 485–490 (1975).
the carbonate equilibrium will become realistic. Because
pH is a very important factor affecting microbiological 25. T. Mino et al., Water Res. 32, 3193–3207 (1998).
activities, the development of AS models containing the 26. T. Mino, Biochemistry (Moscow) 165, 341–348 (2000).
effect of pH will be useful both from a scientific and a 27. M. C. Wentzel et al., Water Sci. Technol. 23, 567–576 (1991).
practical point of view. 28. H. Satoh et al., Water Sci. Technol. 23, 899–905 (1992).
29. T. Mino et al., in R. Ramadori, ed., Biological Phosphate
CONCLUSION Removal from Wastewater, Advances in Water Pollution
Control, Pergamon Press, Oxford, U.K., 1987, pp. 27–38.
In the present article, various microbiological issues asso- 30. T. Mino et al., Water Sci. Technol. 29, 67–70 (1994).
ciated with the IWA AS models are discussed in terms of 31. D. Brdanovic et al., Water Sci. Technol. 37, 541–547 (1998).
modeling. The IWA models have been developed on the 32. H. Pereira et al., Water Res. 30,2128–2138 (1996).
basis of evidence accepted by the majority of microbiolog- 33. M. Maurer et al., Water Res. 31, 907–917 (1997).
ically oriented wastewater engineers. Therefore, the IWA 34. M. C. Wentzel et al., Water Sci. Technol. 17(11/12), 57–71
models are not the most advanced AS models, but rather (1985).
give a fundamental and common basis for the further
35. J. P. Kerrn-Jespersen and M. Henze, Water Res. 27, 617–624
development of revised, detailed, or innovative models. (1993).
Extension and improvement of the present models will
36. T. Kuba et al., Water Res. 31, 777–786 (1997).
surely be made according to practical demands in the
future. 37. M. C. M. van Loosdrecht and M. Henze, Water Sci. Technol.
39(1), 107–118 (1999).
38. M. A. van Aalst-van Leewen et al., Bioeng. Biotech. 55,
BIBLIOGRAPHY
773–782 (1997).

1. M. Henze et al., Activated Sludge Models ASM 1, ASM 2, 39. M. Majone et al., Water Sci. Technol. 39(1), 651–74 (1999).
ASM2d and ASM 3, IWA Publishing, London, U.K., 2000. 40. M. C. M. van Loosdrecht et al., Water Sci. Technol. 35(1),
2. P. L. Dold et al., Prog. Water Technol.12, 47–77 (1980). 44–52 (1997).
3. G. A. Ekama et al., Water Sci. Technol. 18, 91–114 (1986). 41. G. J. F. Smolders et al., Biotech. Bioeng. 47, 277–287 (1995).
4. M. Henze et al., Activated Sludge Model No. 1, International 42. R. Yamamot-Ikemoto et al., Water Sci. Tecnol. 34, 119–128
Association on Water Pollution Control and Research, (1996).
London, U.K., 1987. 43. M. S. M. Jetten et al., Antonie Van Leeuwenhoek 71, 75–93
5. M. Henze et al., Activated Sludge Model No. 2, International (1997).
Association on Water Quality, London, U.K., 1995. 44. B. Batchelor and B. W. Lawrence, J. Water Pollut. Control
6. M. Henze et al., Water Sci. Technol. 39, 165–182 (1999). Fed. 5, 1986–2001 (1978).
26 ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION

45. D. Jenkins et al., Manual on the Causes and Control of some guidance on how this knowledge could be extended
Activated Sludge Bulking and Foaming, 2nd ed., Lewis by implementing newly developed methods and be used
Publishers, Boca Raton, Fla., 1993. for future improvement in wastewater treatment.
46. D. Jenkins et al., Manual on the Causes and Control of
Activated Sludge Bulking and Foaming, Ridgeline Press,
Lafayette, Calif., 1986. METHODS FOR MICROBIAL DIVERSITY ANALYSIS IN
47. H. Furumai and B. Rittmann, Water Sci. Technol. 26, WASTEWATER TREATMENT PLANTS (WWTPS)
493–502 (1992).
48. T. Mino et al., Water Sci. Technol. 31(2), 25–34 (1995). This article briefly summarizes established methods for
49. W. Gujer et al., Water Sci. Technol. 39(1), 183–193 (1999). microbial community analysis. It is not the goal to
provide an encompassing overview on technical details
of each method but rather to discuss the advantages and
ACTIVATED SLUDGE — MOLECULAR limitations of each approach for its use in wastewater
TECHNIQUES FOR DETERMINING treatment microbiology.
COMMUNITY COMPOSITION
Light Microscopy and Cultivation
ALEXANDER LOY
HOLGER DAIMS Traditionally, two approaches to investigate the microor-
MICHAEL WAGNER ganisms in wwtps were applied. Most frequently, obtained
Technische Universität samples were analyzed by standard light microscopy to
München obtain an overview of the abundance of floc-forming and fil-
Freising, Germany amentous bacteria. Because of the importance of the latter
group of organisms for sludge bulking and foaming (Acti-
Wastewater treatment is one of the most important vated sludge — bulking; Activated sludge — foaming), keys
biotechnological processes that is used worldwide to were developed for a provisional identification of filamen-
treat polluted sewage and to ameliorate anthropogenically tous bacteria using (1) their reaction to gram- and Neisser-
induced damage to the environment. Depending on the staining and (2) morphological characteristics (1,2). How-
treatment goals, different types of sewage treatment ever, recent molecular approaches revealed that microor-
plants are used (see ACTIVATED SLUDGE — THE PROCESS). ganisms affiliated with different bacterial domains are
This article focuses on the microbiology of the activated identified as the same filament type using these keys
sludge process, which is most commonly used and in which (see FILAMENTOUS BACTERIA IN ACTIVATED SLUDGE: CURRENT
the microbial biomass is aerated and kept in suspension TAXONOMIC STATUS AND ECOLOGY). Furthermore, polymor-
during the treatment process. In addition, we also cover phism of certain filaments has been described (3), which
activated sludge and biofilm nutrient removal plants in further complicates morphology-based identification.
which anaerobic and aerobic treatments are combined to
The second approach is based on cultivation and
allow for complete nitrogen and/or biologically enhanced
isolation of bacteria from wwtps. The number of
phosphorus removal (EBPR). It is common knowledge that
active bacteria can be estimated by most probable
in all types of wastewater treatment plants (wwtps),
number (MPN) techniques and/or total plate counts. After
prokaryotic microorganisms dominate and represent the
isolation, bacteria are identified using either physiological
‘‘causative agent’’ responsible for the observed conversions.
parameters or chemotaxonomic markers such as cellular
On the other hand, certain microorganisms cause the
fatty acids or respiratory quinone profiles. The latter two
most frequently encountered problems in wastewater
biomarkers can also be used directly for profiling activated
treatment (see FILAMENTOUS BACTERIA IN ACTIVATED SLUDGE:
CURRENT TAXONOMIC STATUS AND ECOLOGY). Thus, the sludge microbial communities (4–6) but provide only
efficiency and robustness of a wastewater treatment plant relatively low-resolution information. Numerous bacteria
mainly depend on the composition and activity of the were isolated and identified using cultivation-based
microbial communities present in its different stages (see techniques, leading to the perception that, for example,
ACTIVATED SLUDGE — THE MICROBIAL COMMUNITY). Although pseudomonads, enterobacteriaceae, and acinetobacters
biological wastewater treatment has been intentionally are key components of the microbiota of these systems
used for more than a century, because of methodological (see ACTIVATED SLUDGE — THE MICROBIAL COMMUNITY). After
limitations, knowledge on the microbiology of this process the introduction of molecular techniques for community
was scarce until a decade ago. Consequently, these analyses it became obvious that only less than 16%
microbial communities were considered as a ‘‘black box,’’ of the microorganisms can be isolated from activated
and progress in the design and control of plants was sludge by standard cultivation and that those bacteria
derived mainly from empirical research in civil and process forming colonies on the respective media are generally
engineering. Only after the introduction of molecular of minor numerical importance in situ (7–10). Therefore,
techniques in microbial ecology, it has become possible cultivation offers a very biased and incomplete view of
to determine the composition and dynamics of microbial the bacterial diversity there. On the other hand, directed
communities in these systems and to identify the microbial cultivation of in situ important microorganisms is still
key players for the different process types. It is the aim a prerequisite for their encompassing physiological and
of this article to review these new insights and to provide genetical analysis.
 ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION 27

Immunofluorescence purpose, the clones should be grouped into operational

taxonomic units (OTUs) according to their 16S rDNA
A more direct identification and quantification of bacteria
similarities with each other. We recommend to use a
is offered by the use of fluorescent antibodies (FA). How-
97% similarity threshold for OTU assignment, because
ever, the production of these antibodies, generally, still
it has been demonstrated that two organisms with
requires the prior isolation of the target organism, restrict-
a lower 16S rDNA similarity also have DNA-DNA
ing the method to culturable bacteria. Once available, the
similarities below 70% and thus represent different
FA technique can be used to identify and quantify defined
genospecies (20,21). However, organisms with highly
microorganisms within activated sludge or biofilms. For
similar or even identical 16S rDNA sequence might
example, Sphaerotilus natans, Acinetobacter sp. (11), and
nevertheless represent two different species (20), and
Thiothrix sp. have been detected by this technique (12).
thus the OTU concept leads to underestimation of the
The specificity of antibodies cannot be adjusted and is
actual species richness. Once the clones have been
usually below the species level. This is an advantage if,
assigned to OTUs, rarefaction analyses (22,23) or coverage
for example, the fate of specific strains should be moni-
estimates (24) should be performed to determine whether
tored. On the other hand, the high serological diversity
the analyzed number of clones represents a sufficient
of many bacterial genera and even species (13,14) would
sample size. Figure 1 shows rarefaction analyses of
require the simultaneous use of a huge number of FA
16S rDNA surveys of wwtps, the respective coverage
for their detection, thus rendering the method cumber-
estimates are depicted in Table 1.
some. Furthermore, FA are relatively large molecules and
The rarefaction analyses and coverage estimates show
thus do not easily penetrate through dense extracellular
that most studies did not adequately harvest the diversity
polymeric substances of flocs and biofilms, although this
of their 16S rDNA libraries. This is understandable,
problem can be avoided if the FA technique is applied
because the number of identical or highly similar clone
to cryosections. Another difficulty of the FA technique is
sequences increases with the number of sequenced
that it often suffers from a high background fluorescence
clones and thus extensive clone sequencing creates
caused by unspecific binding of the FA to, for example,
highly redundant information. Therefore, screening of
detritus (14,15) and filament sheaths. In addition, the
the 16S rDNA clones with fingerprinting techniques
expression of the detected epitopes on the cell surface can
(discussed later) for selection of different clones for
vary with changing environmental conditions, a fact that
sequence analyses is recommended.
can cause false-negative results. Despite these limitations,
However, it is important to realize that even if the
the FA technique has the unique advantage that detection
complete diversity of an environmental 16S rDNA clone
does not require the killing of the cells. Thus, antibodies
library is harvested, the obtained species inventory might
can be used to enrich target cells, for example, via flow
not represent the naturally occurring diversity. In other
cytometry, coated microtiter plates, or magnetic beads for
words, not all, and sometimes even not all, numerically
subsequent cultivation.
important bacterial populations of an environmental
sample will be found in a respective 16S rDNA library.
The 16S rDNA Approach
This failure might be caused by (1) inefficient DNA
The comparative sequence analysis of environmen- extraction (27) (e.g., certain gram-positive bacteria are
tally retrieved 16S rDNA sequences has become the difficult to lyse), (2) inadequate coverage of the selected
gold standard for cultivation-independent assessment PCR primers, (3) kinetic and stochastic biases introduced
of bacterial diversity in natural and engineered sys- by the PCR amplification (28–30), and (4) cloning biases.
tems (16). Current 16S rDNA databases contain more For example, the high coverage but low overall OTU
than 20,000 entries (17) and thus provide a high- number obtained in the 16S rDNA library analysis of
resolution framework for the assignment of those activated sludge from a municipal high-load wastewater
sequences obtained in 16S rDNA libraries from environ- treatment basin (31) does not necessarily reflect low
mental diversity surveys (18). The approach consists of species richness but was probably caused by the omission
DNA extraction, subsequent PCR amplification of (a frag- of a dedicated DNA-extraction protocol before PCR
ment of) the 16S rDNA gene using primers targeting amplification.
regions conserved in the bacterial domain, cloning, and For establishing a most representative 16S rDNA
sequencing. The obtained sequences are analyzed together library, one might consider to (1) use different DNA-
with adequate reference sequences to infer their phy- extraction techniques and to combine the isolated nucleic
logenetic affiliation. A meaningful phylogenetic analysis acids before PCR amplification, (2) to use more than
requires the use of almost full-length 16S rDNA sequences one primer set for amplification, and (3) to use different
and of different treeing methods applied to different data vectors for cloning. Regarding the PCR conditions best
sets (19). suited for a general diversity survey, extremely stringent
As a result of the development of premanufactured annealing conditions should be avoided to allow binding
kits (e.g., for cloning) and automated sequencers, the of the primers to organisms with nucleotide exchanges
analyses of relatively high clone numbers per library in the target position. This procedure might lead to the
has become possible. It is however important, but amplification of undesired PCR products and the necessity
rarely done in published studies, to determine whether to purify the expected PCR amplificate by agarose
the number of analyzed clones does sufficiently well gel electrophoresis and band excision before cloning.
represent the diversity in the established library. For this Furthermore, the PCR cycle number has a profound
28 ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION

80
Nitrifying sequencing batch biofilm reactor 1; this study

70 Nitrifying/ denitrifying industrial wwtp; Juretschko et al. (submitted)

High-load aeration basin of a full-scale municipal wwtp; Snaidr et al. (1997)

60 EBPR lab-scale continuous flow reactor; Christensson et al. (1998)

EBPR lab-scale SBR1; Bond et al. (1995)

50 EBPR lab-scale SBR 2; Bond et al. (1995)

Number of OTUs

40

30

20

10

0
0 10 20 30 40 50 60 70 80 90
Number of clones
Figure 1. Rarefaction analyses of 16S rDNA inventories of different wwtps. A 97% similarity
threshold was used for assignment to an operational taxonomic unit (OTU). The studies of Liu
and coworkers (25) and Dabert and coworkers (26) were not considered because only one sequence
per OTU has been deposited in public databases.

influence on the composition of the PCR amplificate if composition can only be obtained if it is combined with
complex template mixtures are used (29). If high cycle quantitative dot blot or in situ hybridization techniques,
numbers are used, highly abundant and less abundant which are described later.
organisms are more likely to be represented in comparable
numbers in the PCR product, an effect caused by template 16S rDNA-Based Fingerprinting Techniques
reannealing at high concentrations (29). In contrast,
the 16S rDNA composition of PCR products generated 16S rDNA-based fingerprinting techniques nicely amend
with low cycle numbers more accurately represents the the 16S rDNA approach described earlier. The com-
composition of the 16S rDNA sequences in the isolated mon principle of these methods is to separate PCR
nucleic acids. products of the same length but different sequence
Additionally, the 16S rDNA approach can also indi- to visualize the diversity within the amplificate by
cate or even create artificial diversity. Small sequence a banding pattern. Many of these techniques were
differences among multiple rRNA operons of a single bac- invented to analyze mutations in medical research
terium, the so-called microheterogeneities (32), might be and were later adapted to environmental microbiology.
interpreted after amplification and cloning as microdiver- The most frequently applied fingerprinting technique
sity, which, however, does not exist at the organism level. in wastewater microbiology is the denaturing gradient
Furthermore, the formation of chimeric sequences via in gel electrophoresis (DGGE) (6,38–40), but also terminal
vitro recombinations (33,34) and the introduction of point restriction fragment length polymorphism (T-RFLP) (41),
mutations via the thermostable DNA polymerase (35,36) gelretardation (42), and single strand conformation poly-
leads to the retrieval of sequences that differ from their morphism (SSCP) (43,44) have been used. The main
natural counterparts. Although chimeric sequences com- advantage of the fingerprinting techniques is that high
posed of distantly related sequence fragments can be recog- sample numbers can be processed in a relatively short
nized easily by standard software tools (37) or comparative time to gain an overview of the dynamics of complex
phylogenetic analyses of different sequence regions (27), microbial communities in wwtps. DGGE and gelretarda-
it is virtually impossible to identify chimeras among tion bands of interest can be excised, cloned, and sequenced
closely related organisms. Polymerase-induced mutations for subsequent identification. DGGE offers a higher res-
can only be detected with a certain probability in highly olution compared with gelretardation but is, in contrast
conserved sequence regions by database comparison. to the latter method, limited to fragments smaller than
In essence, the 16S rDNA approach described earlier is 500 nucleotides that are not well-suited for phylogenetic
an essential tool for microbial diversity analyses in wwtps. analyses. The organism represented by a T-RFLP or SSCP
However, because of the numerous biases inherent to this band can only be identified if a 16S rDNA clone library is
approach quantitative data on the microbial community also established, and if the obtained clones are sequenced
 ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION 29

Table 1. Summary of 16S rRNA-Based Diversity Surveys of Wastewater Treatment Plants and Reactors. EBPR = Enhanced
Biological Phosphorous Removal. SBR = Sequencing Batch Reactor. SBBR = Sequencing Biofilm Batch Reactor. Clones
Were Assigned to Different Operational Taxonomic Units (OTUs) If They Shared Less Than 97% 16S rDNA Sequence
Similarity with Each Other

for identification and are used also as references in the This evaluation requires different procedures for the dot
respective fingerprinting protocol. In general, fingerprint- blot (45) and in situ format (46,47). Today a considerable
ing techniques are affected by the same DNA-extraction number of ready-to-use domain-, division-, genus-, and
and PCR-amplification biases as the 16S rDNA approach species-specific probes are available for identification of
and thus cannot provide quantitative data. the respective target microorganisms within their natural
habitat (for a review see Amann and coworkers (16,48)).
Hybridization Techniques Using rRNA-Targeted Probes For high-resolution analyses of a certain environmental
Dot blot and in situ hybridization techniques using rRNA- sample, additional probes must be designed against
targeted oligonucleotide probes allow us, in contrast cloned sequences obtained with the 16S rDNA approach.
to all other methods discussed in this article, to Application of these probes in the dot blot or in
quantitatively determine the composition of complex situ format allows the detection and quantification of
microbial communities. The specificity of the probes can the corresponding microorganisms within the habitat
be adjusted to different phylogenetic levels. Specialized for which the 16S rDNA library was established. The
software tools have been developed to assist in probe complete procedure including the 16S rDNA gene library
design and evaluation (17). After design, for each probe analysis, clone-specific probe design, and quantitative
optimal hybridization conditions have to be carefully hybridization experiments is referred to as the full-cycle
determined using target and nontarget microorganisms. rRNA approach (16).
30 ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION

In the dot blot format, total RNA is extracted from and the use of different probe labels. Therefore, sets of
the environment and is immobilized on a membrane probes with hierarchical or identical specificity can be
together with dilution series of RNA of reference species. used to increase the reliability of the identification.
Reference rRNA of uncultured microorganisms can be For many years, quantitative FISH data on the
obtained by in vitro transcription of the respective cloned bacterial community structure of activated sludge or
16S rDNA sequences. Subsequently, the membrane is biofilms required microscopic counting of stained cells, a
hybridized with a radioactively labeled probe, and after time-consuming procedure that is relatively inaccurate in
a stringent wash step the amount of target rRNA is samples containing densely clustered cells. Generally, the
quantified with a Phospho-Imager. Results are expressed abundance of a certain population of interest is expressed
as ng of target rRNA per weight or volume of the as the percentage of all cells that can be stained with
environmental sample. Alternatively, the membrane can either bacterial probes (7,47) or a DNA-binding dye such
be rehybridized with a bacterial or universal probe, and as DAPI (61). The combination of FISH and confocal
the amount of population-specific rRNA detected with laser scanning microscopy (62) not only dramatically
the specific probe is expressed as fraction of the total improved the image quality by exclusion of blurring out-
(bacterial) rRNA. For details of the method the reader of-focus fluorescence, but also allowed the development of
is referred to Raskin and coworkers (49). Quantitative semiautomatic quantification protocols by use of digital
dot blot hybridization measures the abundance of the image analyses (42,63). These methods measure the
target population via its rRNA content. Because the rRNA specifically stained biovolume of a target population and
content of a microbial cell varies with its activity and refer it to the volume of those microorganisms stained by
physiological history (50), and prevailing environmental the bacterial probe set or a DNA-binding dye. Although
conditions, the obtained data cannot be translated to using the manual technique only a few hundred or
cell numbers. Furthermore, the method might be biased thousand cells are counted, the semiautomated protocols
by a varying efficiency of RNA-extraction from different detect more than 100,000 cells per measurement and thus
microbial populations and degradation of the rRNA (which are much more reliable and reproducible.
might affect different probe target sites in a different However, it is important to realize that numbers
way) during the procedure. Quantitative dot blot analyses obtained by this technique or its manual counterpart
have been used in wastewater microbiology (45,51–53) cannot be used directly to compare the abundance of
but are more widely applied on environments such as detected bacterial populations among different samples,
soil and microbial mats (54,55), which because of their because differences in biomass among the samples are
high background fluorescence are not well-suited for not considered. Thus, for intersample comparison, the
fluorescence in situ hybridization (FISH). respective numbers need to be normalized by taking
In situ hybridization with fluorescently labeled, rRNA- into account biomass differences estimated, for example,
targeted oligonucleotide probes is perfectly suited to via volatile suspended solids (VSS) determinations (64) or
identify the target organisms within activated sludge total DAPI counts per sample volume on a membrane
flocs or biofilms by specific staining. For this purpose, filter (8).
samples are pretreated with chemical fixatives to kill the Confocal laser scanning microscopic analyses of FISH
cells, to preserve their morphology, and to make them results may also be used to investigate the spatial
accessible to oligonucleotide probes. For gram-negative distribution of microorganisms in activated sludge flocs
bacteria, a paraformaldehyde-based fixative fulfills these and biofilms and to study colocalization of microbial
requirements and allows storage of the samples at populations (27,62,65). However, such analysis can only
−20 ° C for several years (16). However, a fixation protocol be performed with high accuracy, if the architecture of the
permeabilizing all gram-positive bacteria still has to be flocs or biofilms is preserved by embedding (for embedding
developed. Many, but not all, gram-positive bacteria can protocols see (66) or cryosectioning (65,67)) and is thus not
be permeabilized efficiently for FISH by addition of 50% destroyed during fixation and hybridization.
EtOH (56). Others, for example, the mycolata containing One potential disadvantage of FISH is its relatively
mycolic acids in their cell walls, require additional or high detection limit of approximately 103 to 104 target
alternative pretreatments including the use of cell wall cells ml−1 . Furthermore, only cells with a ribosome
lytic enzymes (57–59). Fixed samples are hybridized with content of more than about 1,000 copies can be detected
probes labeled with fluorescent dyes. For environmental by probes without additional signal amplification (68).
studies, the dyes FLUOS 5 (6)-carboxyfluorescein-N- Because microorganisms with a low cellular ribosome
hydroxysuccimide ester: green fluorescence), Cy3 (orange content are probably not active, this requirement does
fluorescence), or Cy5 (infrared fluorescence) are well- not hamper the in situ analysis of physiologically active
suited and permit the specific detection of the target cells cells in wwtps. This is also reflected by the fact that
with up to three simultaneously applied probes (60). If about 90% of all DAPI-positive cells can be detected
more than one probe is used in a hybridization experiment with the bacterial probe set in a typical plant (47). On
and if the probes require different conditions to ensure the other hand, it is frequently assumed that a high
specificity, subsequent hybridization reactions from high cellular ribosome content of a microbial cell leading to
to low stringency have to be performed (3). In contrast to a bright FISH signal is an indicator of its physiological
dot blot hybridization, simultaneous binding of multiple activity at the time of sampling. This interpretation can
probes to the same target cells can be proven by FISH be misleading because nitrifying bacteria, for example,
through microscopic observation of the hybridized cells maintain high ribosome levels even after complete
 ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION 31

inhibition for several days (14) or starvation for a activated sludge plant connected to a rendering plant (78).
month (69). More direct in situ analyses of the activity may It is obvious that the number of plants for which
be achieved by the use of oligonucleotide probes targeting 16S rDNA libraries have been established is too low
the 16S/23S intergenic spacer region as demonstrated to provide an encompassing overview of the entire
for Acinetobacter sp. (70) and the anaerobic ammonium- microbial diversity of such systems. Although several
oxidizer Candidatus ‘‘Kuenenia stuttgartiensis’’ (71). SBRs with enhanced biological phosphorus removal have
In summary, a variety of molecular methods is avail- been investigated, for the other plant types only single
able for investigating microbial diversity in wastewa- surveys were performed. Nevertheless, some general
ter treatment plants. In particular, the full-cycle rRNA conclusions can be drawn already. Of the 36 divisions
approach allows to precisely measure the composition recognized for the bacterial domain (18), members of
and dynamics of microbial communities in these systems. 13 divisions were detected in the surveys, indicating
The following section provides an overview on the cur- considerable microbial diversity in wastewater treatment
rent status of knowledge that has been accumulated by plants. Consistent with the FISH results mentioned
application of these research approaches in wastewater earlier, Proteobacteria are abundant in each library
microbiology. and represent more than 50% of the clones in five of
the eight surveys. With one exception, β-Proteobacteria
GENERAL MICROBIAL DIVERSITY IN WASTEWATER are the most frequently retrieved members of this
TREATMENT PLANTS division. Apart from the Proteobacteria, the Cytophagales,
the Green-nonsulfur bacteria and the Planctomycetes
FISH with Group-Specific Probes were detected in significant numbers. One library of
a plant designed for enhanced biological phosphorus
Group-specific rRNA-targeted probes (Table 2) have been removal in a continuous flow system is dominated by
widely applied in the so-called top-to-bottom approach high-G+C gram-positive bacteria, consistent with the
for the investigation of activated sludge to obtain a proposed importance of this group for phosphorus removal
fast but relatively low-resolution analysis of the respec- (discussed later) (8).
tive microbial communities (7,8,10,25,72). These studies
showed the dominance of Proteobacteria and revealed Fluorescence In Situ Hybridization and the Full-Cycle rRNA
that the β-subclass is the most abundant subclass in Approach
these systems. α- and γ -subclass Proteobacteria are
So far the microbial community structures of activated
found in lower but generally still considerable num-
sludges from only two wastewater treatment plants have
bers. Subsequently, the widespread distribution of mem-
been investigated using the full-cycle rRNA approach.
bers of the Cytophagales (8,73), the Planctomycetales (74),
Snaidr and coworkers analyzed a high-load aeration basin
and the high-G+C gram-positive bacteria (8,25) in acti-
of a large municipal wastewater treatment plant (31,60),
vated sludge was demonstrated. In addition, applica-
whereas Juretschko and coworkers studied an intermit-
tion of group- and genus-specific probes showed that
tently aerated industrial wastewater treatment plant
the numerically important bacterial groups in activated
designed for simultaneous nitrification and denitrifica-
sludge are dramatically underrepresented after cultiva-
tion (27,78). The results from the respective 16S rDNA
tion and that certain bacterial groups with a low in
clone libraries are depicted in Table 1. Snaidr and
situ abundance are best adapted to laboratory cultiva-
coworkers designed probes for a few selected clones and
tion and thus are isolated most frequently. For example,
showed that a high microdiversity of bacteria of the
γ -subclass Proteobacteria generally occur in relatively
beta-1 group of the β-subclass of Proteobacteria was
low numbers in activated sludge but clearly dominate
present in the municipal activated sludge. Furthermore,
the heterotrophic flora obtainable on nutrient rich agar
Sphingomonas- and Arcobacter-related populations were
plates (7–9).
detected with a relative abundance of 3 and 4% of all cells,
respectively.
16S rDNA-Based Diversity Surveys
The composition of the microbial community in the
The most detailed knowledge on species richness of industrial plant was investigated in more detail using
microbial communities in wastewater treatment plants semiautomatic quantitative FISH. Hybridization with
can be obtained by phylogenetic inventories via 16S rDNA group-specific probes demonstrated that the β-subclass
clone libraries established using bacterial or universal of Proteobacteria made up almost half of the total
primers. However, probably because of the tediousness biovolume of those bacteria detectable with the bacterial
of the method, such surveys have only been published probe set (Fig. 2a). Other in situ important groups were
for six treatment systems, including five laboratory the α-subclass of Proteobacteria, the Nitrospira-phylum,
scale reactors (25,26,75,76), and one high-load basin of the Planctomycetes, and the green nonsulfur bacteria.
a full-scale municipal wastewater treatment plant (31). Figure 2a also depicts differences between the in situ
Results of these surveys are summarized in Table 1. abundance of the groups and their representation in the
In addition to published results, we have included in clone library.
this table two studies from our laboratory dealing with The composition of the β-subclass of Proteobacteria, the
a pilot-scale nitrifying, sequencing batch biofilm reactor numerically most important group within this system,
(partially published by Daims and coworkers (77)) and an was further analyzed using OTU-specific probes for
intermittently aerated nitrifying/denitrifying industrial quantitative FISH (Fig. 2b). Bacteria related to Zoogloea
32 ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION

Table 2. Domain and Group-Specific rRNA-Targeted Probes Suitable for FISH Coverage of the Main
Prokaryotic Lines of Descent by Group-Specific, rRNA-Targeted Oligonucleotide Probes. Sensitivity
and Nontarget Hits of the Probes Were Determined by Using the Arb ‘‘Probe Match’’ Module on
Current Arb Databases of Almost Complete 16S and 23S rRNA Sequences

Target Group and

Probe Sensitivitya (%) Nontarget Hitsb Reference Total Coverage (%)c

ALF1b 37.6 1,105 46 α-Proteobacteria 84.7

ALF968 76.6 132 79
BET42ad 92.6 22 46 β-Proteobacteria 92.6
GAM42ad 90.8 45 46 γ -Proteobacteria 90.8

LGC A 41.1 9 59 Low-G+C gram-positive bacteria 52.1

LGC B 29.1 7 59
LGC C 10.9 0 59

DHP1006 100 0 80 Synergistes 100

TM7905 100 2 81 TM7 100
GSB532 75 0 82 Green sulfur bacteria 75

CFB286 49.3 0 83 Cytophagales 90.5

CFB563 27.2 0 83
CFB719 31.1 4 83
CFB972 27.2 96 83
CFB1082 39.2 2 83
CF319 41.9 28 73
BAC303 33.1 0 73

Fibro 100 0 84 Fibrobacter 100

Pla46 92.6 6 85 Planctomycetales 95.1

Pla886d 82.7 2,561 85
EUB338-IId,e 75.3 1 47

EUB338-IIId,e 70 31 47 Verrucomicrobiales 70

Ntspa712d 85 0 77 Nitrospira 85

IRog1 38.4 1 86 Acidobacterium/Holophaga 44.2

IRog2 44.2 0 86

HGC 82.5 10 56 High-G+C gram-positive bacteria 82.5

EUB338 90.4 0 87 domain Bacteria 91.8

EUB338-II 0.8 0 47
EUB338-III 0.6 0 47

CREN499 30.6 0 88 Crenarchaeota 30.6

EURY498 50 2 88 Euryarchaeota 50
Arch915 88.3 0 48 Archaea 88.3
domain Archaea 89.8
a
The fraction of the sequences within the target group that have not more than 0.4 weighted mismatches to the probe sequence.
b
The number of nontarget sequences that have up to 0.4 weighted mismatches to the probe sequence.
c
Total coverage of the target group by a combination of all listed probes that are specific for this group.
d
The probe specificity is improved by a competitor oligonucleotide as detailed in the publication cited.
e
The probe targets parts of the specified group and additional bacteria, because it belongs to the Bacteria-specific probe set.

ramigera and the Azoarcus/Thauera cluster were the most detectable in situ could not be assigned to a specific
abundant members of this subclass in situ, and accounted genus.
for 36% and 34% of the biovolume of the β-Proteobacteria.
In addition, significant numbers of the ammonia-oxidizer FUNCTIONAL GROUPS OF MICROORGANISMS IN
Nitrosococcus mobilis (which was not present in the clone TREATMENT PLANTS
library), Alcaligenes latus-like bacteria, and Brachymonas
denitrificans-related microorganisms were recorded. In From an applied perspective, identification and char-
total, only 6.5% of the bacteria of the β-Proteobacteria acterization of those bacteria responsible for specific
 ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION 33

(a) Unidentified bacteria 11% (8%∗)

Holophaga/Acidobacterium 1% (5%)

β-Proteobacteria 47% (31%)

Green nonsulfur bacteria 5% (16%)

Planctomycetales 6% (12%)

Nitrospira 12% (2%)

γ-Proteobacteria 2% (0%)

α-Proteobacteria 16% (26%)

(b) Unidentified β-Proteobacteria 6,5% (35.5%)

Brachymonas denitrificans
-related bacteria 2% (3%)
Zoogloea ramigera-related bacteria
Alcaligenes latus 36% (31%)
-related bacteria 6,5% (6.5%)

Nitrosococcus mobilis 15% (0%)

Azoarcus/ Thauera 34% (24%)

Figure 2. Microbial community structure of an industrial nitrifying/denitrifying activated sludge

analyzed by FISH using group-specific (panel a) or β-subclass OTU-specific (panel b) probes. The
biovolume of the specifically stained cells was determined using confocal laser scanning microscopy
and digital image analyses and expressed as percentage of the biovolume of those cells detectable
by in situ hybridization with the bacterial probe set. For panel B the biovolume of the β-subclass
Proteobacteria was set to 100% and the other groups shown in this panel were scaled accordingly.
Numbers in brackets show the abundance of the respective target organisms in the 16S rDNA
clone library established for the same sample. ∗sequences not affiliated with the depicted groups.

transformations of sewage compounds are of primary been implicated with enhanced biological phosphorus
importance. In particular, the organisms responsible for removal (EBPR) are closely related to the phototrophic
nitrogen and phosphorus removal are of interest. However, β-subclass proteobacterium Rhodocyclus purpureus ( (89);
16S rDNA sequence-based identification of a microorgan- mentioned later also), which is not capable of performing
ism generally does not allow to infer its functional prop- the required phosphorus transformations. On the other
erties. Phylogenetically closely related microorganisms hand, several physiological traits important for wwtps, for
may possess different metabolic potentials. For example, example, denitrification, are found dispersed in many dif-
certain obviously nonphotosynthetic bacteria that have ferent phylogenetic lineages. Only for a few physiological
34 ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION

groups does an unambiguous linkage between phylogeny are not available for all important substances (e.g., a
and physiology exist. One example is the lithoautotrophic microelectrode suitable for phosphate determinations has
ammonia oxidizers that form two monophyletic clusters not yet been developed), and those that exist do not func-
in the β- and γ -subclasses of Proteobacteria, respec- tion under all environmental conditions. Microelectrode
tively (90). All known organisms affiliated with these measurements are one-dimensional and thus are heavily
clusters gain their energy by oxidation of ammonia to influenced by spatial heterogeneity of the sample. This is
nitrite. particularly important because it is currently not possible
One way to identify microorganisms capable of to detect the injection channel of the microelectrode in
performing a certain function in a wwtp is to use a subsequent FISH analysis in order to investigate the
lab-scale reactors inoculated with activated sludge or microbial communities surrounding this site. Therefore,
biofilm material, which select for the respective functional FISH-microelectrode measurements are much easier to
microbial group. Using this approach, enrichments of yet interpret in biofilms with a layered distribution of microor-
uncultured nitrite oxidizers (91), anaerobic ammonium ganisms than in heterogeneous and structurally dynamic
oxidizers (92), and bacteria catalyzing EBPR (89,93) were activated sludge flocs. Furthermore, the spatial resolu-
obtained. Although these studies provided important tion of microelectrodes of 10 to 50 µm is clearly above the
new insights, enrichment in reactors might favor the single-cell level and does not allow to directly investigate
growth of organisms not important in a full-scale wwtp the function of single microbial cells in an environment.
by inducing biases comparable to those caused by The in situ physiology of microorganisms in wwtps can
cultivation approaches. This effect is probably pronounced, be investigated using a combination of FISH and microau-
in particular, if the reactor is fed with artificial sewage or toradiography (101,102). A small volume of the native
if the sewage is amended with nutrients such as acetate. sample is incubated with radioactively labeled substrates
Therefore, the identification of organisms catalyzing a under suitable environmental conditions. After incubation
certain function in a reactor enrichment always has the samples are fixed, and substrate not incorporated by
to be complemented with investigations on the in situ the cells is removed by a series of wash steps. Subse-
abundance and function of these bacteria in full-scale quently, samples are cryosectioned and transferred to a
wwtps. cover slip for FISH analyses. Finally, the samples are cov-
A major current research challenge is the development ered with an autoradiographic emulsion, developed, and
and validation of techniques for in situ analyses of the viewed with an inverse confocal laser scanning microscope.
function of microorganisms within their environment. For Cells are identified by their probe-conferred fluorescence,
example, selected genes coding for key enzymes of certain and silver grain formation on top of the cells (in the emul-
metabolic pathways (often referred to as functional genes) sion) indicates physiological activity and substrate uptake
can be used both as phylogenetic and functional markers under the selected environmental conditions. If combined
for the respective organisms. Using this approach, specific with an appropriate study design, most functionally impor-
primers are used for PCR amplification of a fragment of tant groups of microorganisms can be identified and
the respective gene from a wwtp sample. After cloning quantified using the combination of FISH and microau-
and sequencing, the affiliation of the environmental toradiography (101). It should be stressed that radioactive
gene fragments is determined by comparative sequence labeling of cells in the FISH-microautoradiography pro-
analyses with reference cultures. Although this approach cedure is unlikely to be caused by simple substrate
also suffers from DNA extraction, PCR- and cloning- uptake because the cells are permeabilized for oligonu-
biases (discussed earlier), it combines, in contrast to cleotide penetration and therefore unbound substrate is
the 16S rDNA approach, identification with assignment removed during the multiple wash steps. Consequently,
of a metabolic potential. In wastewater microbiology, the microautoradiographic detection requires incorporation of
dsrAB genes coding for the dissimilatory sulfite reductase the substrate into a macromolecule by the enzymatic
of sulfate-reducers (94) and the amoA gene encoding the apparatus of the cell. The major disadvantages of FISH-
active-site subunit of the ammonia-monooxygenase of microautoradiography are that it is relatively expensive
ammonia-oxidizers (95,96) have been used to investigate and time-consuming, and that radioactive derivatives of
the diversity of both functional groups (90,97). However, many substances are not easily available. Furthermore,
the mere detection of a functional gene does not prove that the spatial resolution of microautoradiography depends on
this gene was transcribed, translated, and functionally the isotope used for substrate labeling. Although tritium-
active in the respective organism. labeled substrates often allow single cell resolution in
Another way to link in situ identification of microorgan- cryosections, lower resolution on a cell cluster level is
isms in wwtps with a specific function is to combine FISH achieved using 14 C- or 33 P-labeled substrates.
In the following sections, the current knowledge on the
and microelectrodes (98). For example, zones of nitrifica-
diversity of important physiological groups of bacteria in
tion or denitrification in a biofilm can be determined using
wwtps is reviewed.
the appropriate microelectrodes. Equipped with this infor-
mation, the bacteria living in the respective zones can be
Nitrifying Bacteria
identified using FISH (65,67,99,100). If quantitative FISH
data are available, this sophisticated approach can even The nitrifying bacteria encompass two groups of microor-
be used to estimate substrate conversion rates per cell and ganisms, the ammonia- and the nitrite-oxidizing bacteria,
KS values (65). Nevertheless, the combination of FISH and which catalyze the oxidation of ammonia to nitrite and
microelectrodes has several limitations. Microelectrodes of nitrite to nitrate, respectively. Because most of the
 ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION 35

nitrogen in the influent of a wwtp is present either in a surprise (106). Using the full-cycle rRNA approach,
form of urea (which is hydrolyzed to ammonia) or ammo- the occurrence of novel, yet uncultured, Nitrospira-
nium/ammonia, the nitrifying bacteria play a central role like nitrite-oxidizing bacteria in nitrifying wwtps could
in nitrogen removal in wwtps. It is important to lower be demonstrated (27,77,100,107,108). The importance of
the ammonia concentrations in the effluent of wwtps these microorganisms for nitrite oxidation in wwtps was
because this compound is toxic to aquatic life and pro- also confirmed by reactor enrichment studies (91). Today
motes eutrophication in the receiving water. Nitrifying four different phylogenetic lineages, two of them con-
bacteria are extremely slow-growing microorganisms and taining 16S rDNA clones of wwtps, within the genus
are recalcitrant to cultivation attempts. Because of the Nitrospira have been recognized (Fig. 4) and phylum-
sensitivity of nitrifying bacteria to disturbances such as and genus-specific probes suitable for FISH are avail-
pH and temperature shifts, breakdown of the nitrifica- able (109). Combination of FISH and microautoradiogra-
tion process is frequently reported from municipal and phy showed that the Nitrospira-like nitrite-oxidizers in
especially industrial wwtps. activated sludge fix carbon dioxide and can also grow
According to microbiology and civil engineering text- mixotrophically using pyruvate but not acetate, butyrate,
books, the model ammonia-oxidizer is Nitrosomonas and propionate (109).
europaea. However, FISH analyses in nitrifying acti- It has been postulated that the predominance of
vated sludge and biofilms showed that other ammonia- Nitrospira-like bacteria over Nitrobacter in most wwtps
oxidizers are more important. In an industrial nitrify- is a reflection of their different survival strategies.
ing/denitrifying plant, the dominant ammonia-oxidizer Although Nitrospira-like nitrite oxidizers are, according
was N. mobilis, a bacterium that was previously consid- to data extracted from microelectrode-FISH analyses, K-
ered to occur in brackish water only (27). Subsequently, strategists and thus may possess a low µmax , but they
N. mobilis was also detected in significant numbers in a are well-adapted to low nitrite and oxygen concentrations;
nitrifying sequencing batch biofilm reactor (77). In con- Nitrobacter was postulated to be a relatively fast-growing
trast, Nitrosospira-related ammonia-oxidizers were found r-strategist with low affinities to nitrite and oxygen (65).
to be dominant in situ in a laboratory scale-fluidized bed Because nitrite concentrations in most reactors from
reactor (103). Although Nitrosospira was also reported wwtps are low, Nitrospiras will outcompete Nitrobacter
in a PCR-based study as important ammonia-oxidizer in these systems. In plants with temporally or spatially
genus in wwtps (104), this finding could not be con- elevated nitrite concentrations, for example, in nitrifying
firmed by FISH analyses of various wwtps and by a sequencing batch reactors, both nitrite-oxidizers should
large amoA-based ammonia-oxidizer diversity survey in be able to coexist. Consistent with this hypothesis, co-
wwtps (90). Today it is generally accepted that nitro- occurrence of Nitrobacter and Nitrospira-like bacteria has
somonads (including N. mobilis), and not nitrosospiras been observed by FISH in a nitrifying sequencing batch
(encompassing the genera Nitrosospira, Nitrosolobus, and biofilm reactor (109).
Nitrosovibrio), are important for ammonia oxidation in
wwtps. This perception is also reflected in Figure 3 that Denitrifying Bacteria
shows the affiliation of 178 amoA clones retrieved from Denitrification, the anaerobic respiration with nitrite or
various nitrifying wwtps. Only 10 amoA clones from nitrate as electron acceptor, is used in wastewater to
these systems cluster with the nitrosospiras, whereas convert the product(s) of nitrification into gaseous nitrogen
the remaining 168 clones are affiliated with the nitro- compounds (mainly dinitrogen), and thus to remove them
somonads. Figure 3 also shows that almost all recognized from the sewage. Most attempts to identify and enumerate
lineages of β-subclass ammonia-oxidizers can be found denitrifiers in activated sludge were based on cultivation-
in wwtps. Numerically, the N. europaea/Nitrosomonas dependent approaches. Members of the genera Alcaligenes,
eutropha-lineage, the N. mobilis-lineage, and the Nitro- Pseudomonas, Methylobacterium, Bacillus, Paracoccus,
somonas marina/Nitrosomonas oligotropha/Nitrosomonas and Hyphomicrobium were isolated as part of the den-
urea cluster are most frequently detected. itrifying microbial flora from wwtps (110–114). The latter
In conclusion, wwtps harbor a diversity of ammonia genus was also detected microscopically by its typical
oxidizers of the β-subclass of Proteobacteria, which was cell morphology in denitrifying activated sludge (115,116).
enormously underestimated previously. Most of these However, little is known about whether the aforemen-
ammonia-oxidizers are, on the basis of comparative tioned listed bacterial genera are representative for the
amoA sequence analyses, relatively close relatives of in situ active denitrifiers in wwtps. Neef and coworkers,
described ammonia-oxidizer species. Interestingly, quan- using specific FISH probes, detected significant numbers
titative FISH results indicate that some nitrifying wwtps of Paracoccus spp. and Hyphomicrobium spp. in a denitri-
are dominated by a single ammonia-oxidizer species (27), fying sand filter supplemented with methanol as reduced
whereas other plants harbor at least five different coex- carbon compound for nitrate reduction. But both gen-
isting ammonia-oxidizer populations that are present in era were present in numbers below 0.1% of the total cell
significant numbers (77). counts in a nondenitrifying sand filter run in parallel with-
Traditionally, members of the genera Nitrobacter out addition of methanol, indirectly suggesting an active
were considered as the most important nitrite-oxidizers participation of both genera in the denitrification pro-
in wwtps (105). Therefore, the finding that Nitrobacter cess (74). Molecular studies of the community composition
could not be detected by FISH with specific 16S rRNA- of denitrifying bacteria are difficult to perform because
targeted probes in various nitrifying wwtps came as the denitrifying phenotype cannot be inferred from the
 Isolates/ No. of wwtps
described Isolates Wwtp in which the respective clones
species obtained from clones were detected

84

N. europaea/ eutropha lineage 5 Wwtp, soil 79 10

36
Nc. mobilis lineage1 4 Wwtp, seawater 32 5

19
Nc. mobilis lineage2 0 - 19 4

2 N. halophila lineage 2 Seawater, alkaline lake 0 0

5 N. communis lineage 1 Soil 4 2

2 Nitrosomonas sp. 33 lineage 1 Soil 1 1

2 N. nitrosa lineage 1 Wwtp 1 1

4 environmental lineage 1 0 - 4 2

36
3 environmental lineage 2 0 - 3 1
Nitrosomonas sp. 41 1 Soil 0 0
To outgroups
31
N. marina cluster Soil, sandstone,
N. oligotropha lineage 6 seawater, 25 7
N. ureae lineage brackish water

Nitrosomonas cryotolerans cluster 1 Arctic seawater 0 0

1 Nitrosospira lineage Np22 1 n. d. 0 0

3 Nitrosospira cluster 2 2 Soil 1 1

14 Nitrosospira cluster 3 5 n. d. 9 3

Σ 30 178

Figure 3. Diversity of amoA clones retrieved from different nitrifying wwtps. A schematic
AmoA–based phylogenetic classification of the β-subclass ammonia oxidizing bacteria is shown.
Multifurcations connect branches for which a relative order could not be unambiguously
determined by applying different treeing methods. The cluster designations were adopted from
those of Purkhold and coworkers (90). The height of each tetragon indicates the number of
sequences in the cluster. γ -subclass ammonia oxidizers are not included because they currently
have not been detected by FISH nor by amoA-analyses in nitrifying wwtps.
 ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION 37

Six SBBR 1 clones

Activated sludge clone A1-11 (AF033558)
Activated sludge clone A1-4 (AF033559)
SBR clone RC73 (Y14641)
SBR clone RC90 (Y14642)
SBR clone RC11 (Y14636)
0.10 SBR clone RC99 (Y14643)
SBR clone RC7 (Y14640) I
SBR clone RC14 (Y14637)
SBR clone SBR2016 (X84560)
SBR clone RC25 (Y14639)
SBR clone RC19 (Y14638)
Fluidized bed reactor clone b30 (AJ224041)
Fluidized bed reactor clone b2 (AJ224038)
Fluidized bed reactor clone b21 (AJ224043)
Fluidized bed reactor clone b9 (AJ224045)
Fluidized bed reactor clone g6 (AJ224039)
Fluidized bed reactor clone o14 (AJ224044)
Fluidized bed reactor clone o9 (AJ224042)
SBR clone SBR1015 (X84460)
SBR clone SBR2046 (X84584)
SBR clone SBR1024 (X84468)
Activated sludge clone GC86 (Y14644)
Soil clone (AF010059)
SBR clone SBR2046 (AF155155)
Mesotrophic lake clone LCo23 (AF337199)
Lake Baikal clone 1405-19 (AJ007652)
Rhizosphere clone (AJ005878) II
Rhizosphere clone (AJ232865)
Rhizosphere clone RSC-II-52 (AJ252684)
Freshwater clone MNC2 (AF293010)
Nitrospira moscoviensis (X82558)
Clone C26-12 (AF332345)
Aquarium clone (AF035813)
Aquifer clone WJGRT-122 (AF175639)
SBR clone SBR1065 (X84499)
Soil clone (AF010096)
Nullarbor caves clone wb1_C02 (AF317758)
Clone T26-17 (AF332306)
Clone C26-14 (AF332330)
Soil clone C059 (AF128729)
Nullarbor caves clone wb1_N07 (AF317779)
Nullarbor caves clone wb1_K02 (AF317775)
Nullarbor caves clone wb1_K22 (AF317777)
Nullarbor caves clone wb1_F07 (AF317764) III
Clone T26-16 (AF332296)
Nullarbor caves clone wb1_C17 (AF317762)
Deltaic mud clone n4r (AF194185)
Nitrospira marina (X82559)
IV
Deep sea clone BD3-8 (AB015550)
Sponge symbiont WS65 (AF186451)
Sponge symbiont WS69 (AF186454)
Sponge symbiont WS58 (AF186446)
Nullarbor caves clone wb1_B21 (AF317752)
Figure 4. Phylogenetic tree of the genus Nitrospira based on comparative analysis of 16S rRNA
sequences. The basic tree topology was determined by maximum-likelihood analysis of all
sequences longer than 1,300 nucleotides. Shorter sequences were successively added without
changing the overall tree topology. Branches leading to sequences shorter than 1,000 nucleotides
are dotted to point out that the exact affiliation of these sequences cannot be determined.
Black spots on tree nodes symbolize high parsimony bootstrap support, above 90%, based on
100 iterations. The scale bar indicates 0.1 estimated changes per nucleotide. Clones from wwtps
and reactors and sequences that belong to isolated strains are depicted in bold. The four sublineages
of the genus Nitrospira are boxed in gray and marked by the numbers I to IV. Figure modified
from Daims and coworkers (109).

phylogeny of a microorganism. The capability to denitrify with radioactively labeled substrates that are typically
is found in many bacterial divisions and is not confined used as electron donors for denitrification. In the second
to defined phylogenetic clusters within these divisions. experiment, the sample is incubated with the same labeled
However, the combination of FISH and microautoradiog- substrates under anaerobic conditions but in the presence
raphy (101) allows to identify denitrifiers in situ. For this of nitrate or nitrite. Bacterial species, identified by FISH,
purpose, two types of experiments need to be performed. that take up substrate under anaerobic conditions exclu-
In the first experiment, the wwtp sample is incubated sively in the presence of nitrate or nitrite are most likely
under anaerobic condition in absence of nitrate or nitrite denitrifiers. The use of this technique in combination with
38 ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION

the full-cycle rRNA approach revealed that novel, still the bacteria responsible for EBPR are supposed to gain
uncultured Azoarcus-related bacteria are important deni- energy from polyphosphate hydrolysis accompanied by
trifiers in an industrial nitrifying/denitrifying wwtp (117). subsequent Pi release for uptake of short-chain fatty acids
and their storage in the form of polyhydroxyalkanoates
Anaerobic Ammonium Oxidizing Bacteria (PHA). Two different models were postulated for the
Unexpected nitrogen losses are frequently observed in production of the reducing equivalents for this anaerobic
wastewater treatment plants. In these plants, ammonia metabolism (125,126). In the subsequent aerobic stage, the
was obviously converted to gaseous nitrogen compounds polyphosphate accumulating organisms (PAOs) possess a
in the absence of organic electron donors excluding selective advantage compared with other microorganisms
the activity of conventional denitrifiers. Using the that were not able to take up fatty acids under the
full-cycle rRNA approach, we could demonstrate that preceding anaerobic conditions by using the stored PHA
novel planctomycetes distantly related to the recently in an otherwise carbon-poor environment. In parallel,
described anaerobic ammonium oxidizer Candidatus PAOs restore their polyphosphate pools by aerobic
‘‘Brocadia anammoxidans’’ (92,118) did occur in high uptake of available phosphate from the wastewater. After
numbers in these systems (42,118) and were thus probably sedimentation in the secondary clarifier, a part of the
responsible for the nitrogen removal. Consequently, the biomass is recycled to the anerobic stage and mixed with
new candidatus genus ‘‘Kuenenia stuttgartiensis’’ was new wastewater, whereas the excess sludge containing the
proposed (42). Using specific 16S and 23S rRNA-targeting intracellular polyphosphates is removed from the system.
probes for both anaerobic ammonium oxidizers (42,71), a In contrast to chemical precipitation, EBPR plants
predominance of Candidatus ‘‘Kuenenia stuttgartiensis’’ have been frequently reported to fail. This raised interest
over Candidatus ‘‘Brocadia anammoxidans’’ in wwtps was in the microbiology of the process. Traditionally, on the
observed. basis of cultivation experiments γ -subclass Proteobacteria
The anaerobic ammonium oxidizers are a perfect of the genus Acinetobacter were believed to be the only
example that completely unexpected organisms are hid- PAOs (127–129). However, today it has become apparent
den in wwtps. Both anaerobic ammonium oxidizers form a that Acinetobacter can accumulate polyphosphate but does
monophyletic deep-branching lineage within the Planc- not possess the earlier-described PAO metabolism (122).
tomycetales (Fig. 5) and oxidize ammonium anaerobi- Furthermore, cultivation-independent methods such as
cally with nitrite to dinitrogen gas (42,92,119). Anaerobic fluorescent antibody staining (11), respiratory quinone
ammonium oxidation in contrast to classical nitrification profiles (4), and FISH with a genus-specific probe (8,9)
and denitrification allows to remove ammonium in a one- demonstrated that the relative abundance of Acinetobacter
step process saving space, energy, and money because no in EBPR systems was dramatically overestimated because
aeration or addition of organic carbon for denitrification of cultivation biases, further confirming that Acinetobacter
are required. Anaerobic ammonium oxidation is a very is not of importance for EBPR.
promising method for the treatment of sewage with a Several other bacteria isolated from EBPR reactors
high ammonium and low organic carbon content. How- have been suggested as PAO candidates. Microlunatus
ever, this process needs a mixture of ammonium and phosphovorus, a high-G+C gram-positive bacterium accu-
nitrite and thus must be coupled to partial nitrification mulates aerobically polyphosphate and consumes it under
for nitrite production (42,118,120). Furthermore, estab- anaerobic conditions, but fails to take up acetate or accu-
lishment of anaerobic ammonium oxidizing activity in a mulate PHA under anaerobic conditions (130,131). FISH,
wwtp takes a very long time, and once established the with a probe specific for M. phosphovorus, demonstrated
process is sensitive to perturbation. the presence of this organism in an EBPR plant (2.7%
of the total bacteria) (132), but no direct indications for
Bacteria Important for Enhanced Biological Phosphorus the importance of this bacterium for EBPR are available.
Removal Furthermore, Lampropedia spp. were shown to possess
Phosphorus removal from wastewater is important to the basic metabolic features of a PAO (133), but their
prevent eutrophication, and is therefore an integral acetate-uptake phosphate-release ratio was much lower
part of modern municipal and industrial nutrient than what EBPR models predict, and no additional data
removal wwtps. Phosphorus can be precipitated by the regarding the abundance or activity of these morpholog-
addition of iron or aluminum salts and subsequently be ically conspicuous bacteria in EBPR systems have been
removed with the excess sludge. Chemical precipitation published.
is a very reliable method for phosphorus removal Compared with these cultivation-based attempts, the
but increases significantly the sludge production and hunt for PAOs was more successful using molecular tools
thus creates additional costs. Furthermore, the use for analyses of EBPR systems. β-subclass Proteobacte-
of chemical precipitants often introduces heavy metal ria and high-G+C gram-positive bacteria (actinobacteria)
contamination into the sewage and increases the salt increased in number after addition of acetate to the
concentration of the effluent. Alternatively, phosphorus raw sewage of a EBPR full-scale wwtp suggesting that
removal can be achieved by microbiological mechanisms these groups benefit from the enhanced availability of
in a process called enhanced biological phosphorus short-chain fatty acids in the anaerobic basin and thus
removal (EBPR) (121–124). This process is characterized represent candidates for PAOs (8). Additional support
by cycling the activated sludge through alternating for the importance of both groups for EBPR stems from
anaerobic and aerobic conditions. In the anaerobic stage, FISH experiments in an efficient EBPR laboratory-scale
 Rotating-biological-contactor-clone,
(AJ 224943)
Antarctic-clone
Sequencing batch (AF146242)
reactor clone methanogenic
(AF280847) sludge-clone Antarctic-clone Marine-clone (U70712)
(AB011309) (AF146252)
Marine-clone (L10943)
Deep sea sediment clone (AB015552)
Antartic clone (AF142804)
Antarctic clone (AF142940)
Candidatus “Brocadia anammoxidans”
Antarctic clone
(AF146239)
Antarctic clone (AF146246)
Anammox-enrichment-clone (AJ250882)
Marine sediment clone Candidatus “Kuenenia stuttgartiensis”
(AJ241009)

Antarctic-clone (AF146262)
Soil-clone (AJ390448)

Soil-clone (AJ390447)

39
Isosphaera

Gemmata
Soil-clone (AJ390442)

To outgroups

Planctomyces

0.10

Pirellula
Figure 5. Phylogenetic 16S rDNA tree showing the deep-branching of the anaerobic ammonium
oxidizers within the Planctomycetales. In addition to the described anaerobic ammonium oxidizers
Candidatus ‘‘Brocadia anammoxidans’’ and Candidatus ‘‘Kuenenia stuttgartiensis,’’ several
environmental clone sequences, including sequences from wwtps depicted in bold, are unaffiliated
with cultured members of the division. The physiology of the organisms represented by these
sequences still needs to be clarified. Figure modified from M. Schmid et al. (71).
40 ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION

sequencing batch reactor (72) and respiratory quinone pro- increased effort on the development of suitable cultivation
filing in a lab-scale EBPR system (6). Recently, high-G+C strategies for these bacteria is needed. In parallel, the
gram-positive bacteria related to Terrabacter tumescens, use of techniques referred to as environmental genomics
Tetrasphaera japonica, and Tetrasphaera austrialensis should allow to investigate the genome composition of
were reported to be abundant in EBPR systems and these bacteria without the need of cultivation (138).
thus might be important for EBPR (25,76,93). Although Although researchers interested in microbial ecology
the function of actinobacteria as PAOs still has to be will probably find the results summarized in this article
proven, evidence is available that β2-subclass Proteobac- interesting and important, engineers might ask what
teria related to Rhodocyclus and Propionibacter are impor- practical value can be extracted from the accumulated
tant PAOs in many EBPR systems. These bacteria, for knowledge. Regarding application, the most obvious
which the name Candidatus ‘‘Accumulibacter phosphatis’’ benefit of the progress described is that it provides a
was suggested (89), were detected in high numbers in basis for a more knowledge-driven treatment of wwtp
EBPR systems (26,75,89,93,134). Furthermore, phospho- failures. One strategy to improve a particular aspect
rus accumulation by these bacteria in the aerobic phase of process performance in a wwtp, for example, during
was demonstrated by sequential FISH and polyphosphate start-up or after its breakdown, is the addition of
staining (25,93). In addition, acetate uptake in the anaero- specialized microorganisms or activated sludge from
bic phase and phosphorus uptake under aerobic conditions another wwtp (139). This operational tool that is called
could be demonstrated for Candidatus ‘‘Accumulibacter bioaugmentation does however frequently fail ( (63) and
phosphatis’’ using FISH and microautoradiography (134). references therein). Such failure is typically caused by
A potential reason for the failure of EBBR plants addition of the ‘‘wrong’’ microorganisms, for example, the
is the presence of bacteria that use previously stored model organisms for nitrification, which cannot compete
compounds such as glycogen (also referred to as glyco- successfully with the autochthonous bacteria in the plant
gen accumulating organisms (GAOs)) to compete with and are thus eliminated or washed-out. Therefore, it is
the PAOs for substrate uptake under anaerobic condi- important to identify those microorganisms responsible
tions (41,135,136). Cech and Hartmann (135) described for nitrogen or phosphorus removal in a functioning wwtp.
gram-negative cocci in clumps and packages of tetrades If problems arise, these results can be used as guidance
in activated sludge and called these morphotypes G- to select the appropriate bacterial additive (for culturable
bacteria, because their numbers increased after glucose microorganisms) or a well-suited activated sludge from a
addition. Mino and coworkers (123) suggested that bac- neighboring wwtp containing a comparable microbial flora.
teria with a similar morphology are GAOs. However, Once the appropriate bacteria have been selected, they
recent molecular and physiological analyses of various need to be protected, for example, by polymer embedding
G-bacteria isolated from wwtps revealed a high phyloge- from grazing by protozoa (63).
netic diversity of this morphotype and provided no support Compared with curing failure of a certain process in
for their role as GAOs (Fig. 6; 137). Molecular community a wwtp by bioaugmentation, protecting the plant from
analyses of deteriorated EBPR reactors revealed the pre- process deterioration is a more sustainable strategy. For
dominance of a novel bacterial group within the γ -subclass this purpose we need to understand the links between
of Proteobacteria (40). The phylogenetic affiliation of this the diversity of a functionally important bacterial group
deeply branching group cannot unambiguously be resolved and the stability of the catalyzed process. Preliminary
(Fig. 6). These bacteria are good GAO candidates because observations indicate that plants with low functional
they occur in deteriorated EBPR systems, probably accu- redundancy caused by the presence of a low diversity
mulate PHA, and store little or no polyphosphates (25,40). of bacteria of a certain functional group are more
sensitive to failure of the respective process than plants
harboring a high diversity of the same bacterial group.
CONCLUSION If increase in diversity can indeed be proven to cause
process stabilization, then it will be important to learn
During the last decade, the application of molecular tools how plant design and process parameter control can
in wastewater microbiology has revolutionized our view on be optimized to increase the diversity of functionally
the microbial ecology of these systems. Different groups of important bacterial groups. To answer these ecologically
still not culturable bacteria were identified and shown to and economically important questions it will be necessary
be responsible for nitrite oxidation, denitrification, and to determine the microbial community composition of a
enhanced biological phosphorus removal. Surprisingly, large number of different samples obtained from tightly
the model organisms described in text books for these controlled reactor studies and full-scale wwtps. Because
processes and for ammonia oxidation were shown to be of the tediousness of many of the established molecular
generally not of importance for wastewater treatment. methods, this kind of research will greatly benefit from the
Furthermore, a new process, anaerobic ammonium implementation of modern high-throughput techniques
oxidation, was detected in wwtps and the responsible such as DNA microarrays for measuring microbial
(still not cultured) microorganisms were identified. It community composition in complex samples (140).
is important to note that significant diversity exists
in each of these functional groups of bacteria. In the Acknowledgments
next research phase in wastewater microbiology, more The authors acknowledge Justyna Adamczyk, Matthias Horn,
detailed knowledge on the biology of the aforementioned Stefan Juretschko, Natuschka Lee, Angelika Lehner, Ulrike
noncultured bacteria needs to be gained. Therefore, an Purkhold, Markus Schmid, and Stefan Schmitz-Esser from
 ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION 41

EBPR reactor clone, AF255641

EBPR reactor clone, AF255631
P++ sludge clone SBRP112, AF204248
P+++ sludge clone SBRA245A, AF204245
P+++ sludge clone SBRA220, AF204244
Candidatus "Accumulibacter phosphatis", AJ224937 PAOs
uncultured bacterium PHOS-HC05, AF314417
SBR2 clone, X84620
P+++ sludge clone SBRB34, AF204247
EBPR reactor clone, AJ225358
EBPR reactor clone, AJ225350
0.10 Propionibacter pelophilus, AF016690
Propionivibrio dicarboxylicus, Y17601
EBPR reactor clone, AJ225382
Rhodocyclus tenuis, D16208
Rhodocyclus purpureus, M34132
Azospira oryzae, AF011347
Dechlorimonas agitatus, AF047462
Azonexus fungiphilus, AF011350
Thauera terpenica, AJ005817
Thauera selenatis, Y17591
Zoogloea ramigera, X74913 β-Proteobacteria
Nitrosomonas europaea
uncultured bacterium PHOS-HE54, AF314424
DBPR sludge DGGE band 6, AF093781
uncultured bacterium Px2-DGGE band d, AF124656
uncultured bacterium Px12-DGGE band e, AF124652 Putative GAOs
DBPR sludge DGGE band 4, AF093778
DBPR sludge DGGE band 3, AF093777
DBPR sludge DGGE band 5, AF093780
Nitrosococcus oceani
Acinetobacter lwoffii, U10875
Acinetobacter johnsonii, X81663
Acinetobacter calcoaceticus, M34139
Acinetobacter baumannii, X81660
Vibrio cholerae, X76337
γ-Proteobacteria
Amaricoccus macauensis, U88042
Amaricoccus veronensis, U88043
Amaricoccus tamworthensis, U88044
G-Bacteria
Amaricoccus kaplicensis, U88041
Agrobacterium tumefaciens, D01256 D14500 α-Proteobacteria
Azospirillum parooense, X90760
18
EBPR sludge clones
Tetrasphaera australiensis, AF125090
Nostocoida limicola II, X85212 G-Bact er ia Putative PAOs
Tetrasphaera japonica, AF125092
uncultured High G+C Gram-positive bacterium lpha5, AF109792
uncultured High G+C Gram-positive bacterium lpha7, AF109793
uncultured bacterium Px2-DGGE band c, AF124655
uncultured bacterium Px12-DGGE band d, AF124650
Dermatophilus congolensis, M59057
Micrococcus luteus, M38242
Micrococcus lylae, X80750
Janibacter limosus, Y08539
Janibacter thuringensis, Y08540
Terrabacter tumescens, X83812
Friedmanniella antarctica, Z78206
High-G+C gram-positive bacteria
Microlunatus phosphovorus 3, Z78207

Figure 6. Phylogenetic 16S rDNA tree showing the affiliation of putative polyphosphate
accumulating organisms (PAOs), glycogen accumulating organisms (GAOs) and G-bacteria.
Environmental sequences of putative PAOs and GAOs derived from wwtps are depicted in bold.
The tree was calculated using the maximum likelihood method with a 50% bacterial conservation
filter. The bar corresponds to 10% estimated sequence divergence.

the ‘‘Microbial Ecology’’ group at the Technische Universität 3. M. Wagner et al., Syst. Appl. Microbiol. 17, 405–417 (1994).
München for their contributions and enthusiasm as well as Sibylle 4. A. Hiraishi et al., Appl. Environ. Microbiol. 55, 897–901
Schadhauser, and Jutta Elgner for excellent technical assistance. (1989).
Support by the Deutsche Forschungsgemeinschaft (SFB 411) is
5. A. Hiraishi et al., Appl. Environ. Microbiol. 64, 992–998
gratefully acknowledged.
(1998).
6. W. T. Liu et al., Microbiology 146, 1099–1107 (2000).
BIBLIOGRAPHY
7. M. Wagner et al., Appl. Environ. Microbiol. 59, 1520–1525
1. D. H. Eikelboom, Water Res. 9, 365–388 (1975). (1993).
2. D. Jenkins et al., Manual on the Causes and Control of 8. M. Wagner et al., Appl. Environ. Microbiol. 60, 792–800
Activated Sludge Bulking and Foaming, Lewis Publishers, (1994).
Boca Raton, Fla., 1993. 9. P. Kämpfer et al., Microb. Ecol. 32, 101–121 (1996).
42 ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION

10. W. Manz et al., Water Res. 28, 1715–1723 (1994). 46. W. Manz et al., Syst. Appl. Microbiol. 15, 593–600 (1992).
11. T. E. Cloete and P. L. Steyn, in R. Ramadori, eds., Biological 47. H. Daims et al., Syst. Appl. Microbiol. 22, 434–444 (1999).
Phosphate Removal from Wastewaters, Pergamon Press, 48. D. A. Stahl and R. Amann, in E. Stackebrandt and M. Good-
Oxford, U.K., 1987, pp. 335–338. fellow, eds., Nucleic Acid Techniques in Bacterial Systemat-
12. R. L. Brigmon et al., Appl. Environ. Microbiol. 61, 13–20 ics, John Wiley & Sons, Chichester, U.K., 1991, pp. 205–248.
(1995). 49. L. Raskin et al., Appl. Environ. Microbiol. 60, 1241–1248
13. L. W. Belser, Annu. Rev. Microbiol. 33, 309–333 (1979). (1994).
14. M. Wagner et al., Syst. Appl. Microbiol. 18, 251–264 (1995). 50. K. Flardh et al., J. Bacteriol. 174, 6780–6788 (1992).
15. H. Szwerinski et al., Appl. Microbiol. Biotechnol. 21, 51. R. A. Rosselló-Mora et al., Appl. Environ. Microbiol. 61,
125–128 (1985). 702–707 (1995).
16. R. I. Amann et al., Microbiol. Rev. 59, 143–169 (1995). 52. B. K. Mobarry et al., Appl. Environ. Microbiol. 62, 2156–
17. O. Strunk and W. Ludwig, http://www.arb-home.de/ (2001). 2162 (1996).
18. P. Hugenholtz et al., J. Bacteriol. 180, 4765–4774 (1998). 53. M. F. de los Reyes et al., Appl. Environ. Microbiol. 64,
2503–2512 (1998).
19. W. Ludwig et al., Electrophoresis 19, 554–568 (1998).
54. J. B. Risatti et al., Proc. Natl. Acad. Sci. 91, 10173–10177
20. E. Stackebrandt and B. M. Goebel, Int. J. Syst. Bacteriol.
44, 846–849 (1994). (1994).
55. D. Minz et al., Appl. Environ. Microbiol. 65, 4659–4665
21. R. Rossello-Mora and R. Amann, FEMS Microbiol. Rev. 25,
39–67 (2001). (1999).
56. C. Roller et al., Microbiology 140, 2849–2858 (1994).
22. J. C. Tipper, Paleobiology 5, 423–434, (1979).
23. K. L. Heck et al., Ecology 56, 1459–1461 (1975). 57. R. Erhart et al., Syst. Appl. Microbiol. 20, 310–318 (1997).

24. S. J. Giovannoni et al., in J. Joint, eds., Molecular Ecology 58. R. J. Davenport et al., Appl. Environ. Microbiol. 66,
of Aquatic Microbes, Springer-Verlag, Berlin, Heidelberg, 1158–1166 (2000).
1995. 59. H. Meier et al., Syst. Appl. Microbiol. 22, 186–196 (1999).
25. W. T. Liu et al., Environ. Microbiol. 3, 110–122 (2001). 60. R. Amann et al., J. Bacteriol. 178, 3496–3500 (1996).
26. P. Dabert et al., Appl. Microbiol. Biotech. 55, 500–509 61. R. E. Hicks et al., Appl. Environ. Microbiol. 58, 2158–2163
(2001). (1992).
27. S. Juretschko et al., Appl. Environ. Microbiol. 64, 3042– 62. M. Wagner et al., J. Microsc. 176, 181–187 (1994).
3051 (1998). 63. T. Bouchez et al., Environ. Microbiol. 2, 179–190 (2000).
28. A. Meyerhans et al., Nucleic Acids Res. 18, 1687–1691 64. R. Nogueira et al., Water Res. (in press).
(1990). 65. A. Schramm et al., Appl. Environ. Microbiol. 65, 3690–
29. M. T. Suzuki and S. J. Giovannoni, Appl. Environ. Micro- 3696 (1999).
biol. 62, 625–630 (1996). 66. A. Moter et al., Microbiology 144, 2459–2467 (1998).
30. M. Suzuki et al., Appl. Environ. Microbiol. 64, 4522–4529 67. A. Schramm et al., Appl. Environ. Microbiol. 62, 4641–
(1998). 4647 (1996).
31. J. Snaidr et al., Appl. Environ. Microbiol. 63, 2884–2896 68. W. Schönhuber et al., Appl. Environ. Microbiol. 65,
(1997). 1259–1267 (1999).
32. S. C. Dryden and S. Kaplan, Nucleic Acids Res. 18, 69. E. Morgenroth et al., Water Sci. Technol. 41, 105–113
7267–7277 (1990). (2000).
33. E. D. Kopczynski et al., Appl. Environ. Microbiol. 60, 70. D. B. Oerther et al., Appl. Environ. Microbiol. 66, 2154–
746–748 (1994). 2165 (2000).
34. W. Liesack et al., Microb. Ecol. 21, 191–198 (1991). 71. M. Schmid et al., Environ. Microbiol. 3, 450–459 (2001).
35. N. Arnheim and H. Erlich, Annu. Rev. Biochem. 61, 131–156 72. P. L. Bond et al., Appl. Environ. Microbiol. 65, 4077–4084
(1992). (1999).
36. A. G. Speksnijder et al., Appl. Environ. Microbiol. 67, 73. W. Manz et al., Microbiology 142, 1097–1106 (1996).
469–472 (2001). 74. A. Neef et al., Appl. Environ. Microbiol. 62, 4329–4339
37. N. Larsen, http://rdp.cme.msu.edu/cgis/chimera.cgi?su= (1996).
SSU. 75. P. L. Bond et al., Appl. Environ. Microbiol. 61, 1910–1916
38. G. Muyzer et al., Appl. Environ. Microbiol. 59, 695–700 (1995).
(1993). 76. M. Christensson et al., Appl. Microbiol. Biotechnol. 49,
39. C. Wawer and G. Muyzer, Appl. Environ. Microbiol. 61, 226–234 (1998).
2203–2210 (1995).
77. H. Daims et al., Water Sci. Technol. 43, 9–18 (2001).
40. A. T. Nielsen et al., Appl. Environ. Microbiol. 65, 1251– 78. S. Juretschko et al., Syst. Appl. Microbiol. (submitted).
1258 (1999).
79. A. Neef, Anwendung der in situ-Einzelzell-Identifizierung
41. W. T. Liu et al., Appl. Environ. Microbiol. 63, 4516–4522 von Bakterien zur Populationsanalyse in komplexen mikro-
(1997).
biellen Biozönosen, Ph.D. thesis, Technische Universität
42. M. Schmid et al., Syst. Appl. Microbiol. 23, 93–106 (2000). München, 1997.
43. M. Leclerc et al., FEMS Microbiol. Ecol. 34, 213–220 (2001). 80. C. S. McSweeney et al., Appl. Environ. Microbiol. 59,
44. P. Dabert et al., Water Sci. Technol. 43, 1–8 (2001). 1607–1612 (1993).
45. F. L. de los Reyes et al., Appl. Environ. Microbiol. 63, 81. P. Hugenholtz et al., Appl. Environ. Microbiol. 67, 411–419
1107–1117 (1997). (2001).
 ACTIVATED SLUDGE — SEQUENCING BATCH REACTORS 43

82. C. Tuschak et al., Arch. Microbiol. 171, 265–272 (1999). 118. M. S. M. Jetten et al., Curr. Opin. Biotechnol. 12, 283–288,
83. R. Weller et al., Syst. Appl. Microbiol. 23, 107–114 (2000). (2001).
84. C. Lin and D. A. Stahl, Appl. Environ. Microbiol. 61, 119. K. Egli et al., Arch. Microbiol. 175, 198–207 (2001).
1348–1351 (1995). 120. C. Hellinga et al., Water Sci. Technol. 37, 135–142 (1998).
85. A. Neef et al., Microbiology 144, 3257–3266 (1998). 121. D. Jenkins and V. Tandoi, Water Res. 25, 1471–1478 (1991).
86. W. Ludwig et al., FEMS Microbiol. Lett. 153, 181–190 122. M. C. van Loosdrecht et al., Antonie Van Leeuwenhoek 71,
(1997). 109–116 (1997).
87. R. I. Amann et al., Appl. Environ. Microbiol. 56, 1919– 123. T. Mino et al., Water Res. 32, 3193–3207 (1998).
1925 (1990). 124. T. Mino, Biochemistry (Mosc) 65, 341–348 (2000).
88. S. Burggraf et al., Appl. Environ. Microbiol. 60, 3112–3119 125. Y. Comeau et al., Water Res. 20, 1511–1521 (1986).
(1994).
126. T. Mino et al., in R. Ramodori, eds., Biological Phosphate
89. R. P. X. Hesselmann et al., Syst. Appl. Microbiol. 22, Removal from Wastewaters, Pergamon Presss, Oxford, U.K.,
454–465 (1999). 1987, pp. 27–38.
90. U. Purkhold et al., Appl. Environ. Microbiol. 66, 5368– 127. G. W. Fuhs and M. Chen, Microbiol. Ecol. 2, 119–138 (1975).
5382 (2000).
128. L. H. Lötter and M. Murphy, Water SA 11, 179–184 (1985).
91. P. C. Burrell et al., Appl. Environ. Microbiol. 64, 1878–
129. R. C. Bayly et al., in N. H. Pilkington and R. C. Bayly, eds.,
1883 (1998).
AWWA 13th Federal Convention, Canberra, Australia, 1989,
92. M. Strous et al., Nature 400, 446–449 (1999). pp. 162–166.
93. G. R. Crocetti et al., Appl. Environ. Microbiol. 66, 1175– 130. K. Nakamura et al., J. Ferment. Bioeng. 71, 258–263 (1991).
1182 (2000).
131. K. Nakamura et al., Int. J. Syst. Bacteriol. 45, 17–22 (1995).
94. M. Wagner et al., J. Bacteriol. 180, 2975–2982 (1998).
132. M. Kawaharasaki et al., Water Sci. Technol. 37, 481–484
95. J. H. Rotthauwe et al., FEMS Microbiol. Lett. 133, 131–135 (1998).
(1995).
133. L. Stante et al., Water Res. 31, 1317–1324 (1979).
96. J. H. Rotthauwe et al., Appl. Environ. Microbiol. 63,
134. N. Lee et al., Appl. Environ. Microbiol. (submitted).
4704–4712 (1997).
135. J. S. Cech and P. Hartman, Water Res. 27, 1219–1225
97. C. M. Santegoeds et al., Appl. Environ. Microbiol. 65,
(1993).
4618–4629 (1999).
136. H. Satoh et al., Water Sci. Technol. 26, 933–942 (1992).
98. R. Amann and M. Kühl, Curr. Opin. Microbiol. 1, 352–358
(1998). 137. R. J. Seviour et al., Environ. Microbiol. 2, 581–593 (2000).
99. A. Schramm et al., Environ. Microbiol. 2, 680–686 (2000). 138. C. Schleper et al., J. Bacteriol. 180, 5003–5009 (1998).
100. S. Okabe et al., Appl. Environ. Microbiol. 65, 3182–3191 139. B. E. Rittmann and R. Whiteman, Biotechnology 1, 12–16
(1999). (1994).
101. N. Lee et al., Appl. Environ. Microbiol. 65, 1289–1297 140. D. Y. Guschin et al., Appl. Environ. Microbiol. 63, 2397–
(1999). 2402 (1997).
102. P. H. Nielsen et al., Environ. Microbiol. 2, 389–398 (2000).
103. A. Schramm et al., Appl. Environ. Microbiol. 64, 3480–
3485 (1998). ACTIVATED SLUDGE — SEQUENCING BATCH
104. W. D. Hiorns et al., Microbiology 141, 2793–2800 (1995). REACTORS
105. M. Henze et al., Wastewater Treatment, Springer-Verlag,
Berlin, Germany, 1997. R. L. IRVINE
106. M. Wagner et al., Water Sci. Technol. 34, 237–244 (1996). University of Notre Dame
107. H. Daims et al., Water Sci. Technol. 41, 85–90 (2000). Notre Dame, Indiana
108. A. Gieseke et al., Appl. Environ. Microbiol. 67, 1351–1362 P. A. WILDERER
(2001). Technical University of Munich
109. H. Daims et al., Appl. Environ. Microbiol. (in press). Garching, Germany
110. I. Y. Vedenina and N. I. Govorukhina, Mikrobiologiya 57,
320–328 (1988). The sequencing batch reactor (SBR) belongs to a broad
111. R. Knowles, Microbiol. Rev. 46, 43–70 (1982). class of treatment systems that use a flocculated suspen-
112. G. T. Sperl and D. S. Hoare, J. Bacteriol. 108, 733–736 sion of microorganisms referred to as activated sludge.
(1971). This microorganism suspension is mainly composed of
113. M. A. Attwood and W. Harder, Antonie Leeuwenhoek 38, many different species of bacteria (i.e., a mixed rather
369–378 (1972). than a pure culture that is spatially organized in the
114. F. Schmider and J. C. G. Ottow, Arch. Hydrobiol. 106, form of clusters and microcolonies) that destroy, trans-
497–516 (1986). form, or simply remove unwanted contaminants present
115. U. Nyberg et al., Water Sci. Technol. 26, 1077–1086 (1992). in liquid streams. Within this frame of reference, the objec-
116. P. Timmermans and A. van Haute, Water Res. 17, tive of the SBR and virtually all other biological reactors
1249–1255 (1983). is to select and enrich a mixed culture of microorganisms
117. S. Juretschko, Mikrobielle Populationsstruktur und-dyna- that convert biodegradable contaminants into new cellular
mik in einer nitrifizierenden/denitrifizierenden Belebtschla- material, carbon dioxide, water, and other stable innocu-
mmanlage, Ph.D. Thesis, Technische Universität München, ous organic and/or inorganic products. Although they are
2000. sustained by the consumption of readily biodegradable
44 ACTIVATED SLUDGE — SEQUENCING BATCH REACTORS

energy-rich compounds called contaminants, the resul- zones of the treatment system); others use nitrite-
ting microbial community also removes nonbiodegradable nitrogen or nitrate-nitrogen (i.e., in anoxic zones);
and some slowly biodegradable contaminants by physi- and some use none at all (i.e., in anaerobic zones). It
cal, chemical, or physiochemical means. In such instances, is noted that strict aerobes only grow in the presence
insoluble contaminants and soluble contaminants that of oxygen; strict anaerobes only grow in the absence
precipitate or sorb become part of a flocculated mass of of oxygen; and facultative anaerobes grow in the
microorganisms (i.e., the ‘‘flocs’’) and are thus removed presence or absence of oxygen.
from the waste stream that is being treated. The SBR 2. The ability of bacteria to grow attached to a surface
differs from conventional continuous-flow activated sludge as a biofilm or unattached and in suspension
systems because it is a periodically operated, time-oriented as dispersed organisms or in clusters or flocs.
system, with flow, energy input, and tank volume vary- The most common treatment systems distinguish
ing according to a predetermined operating strategy. As organisms that grow attached as fixed films (e.g.,
a result, the SBR is a time-oriented, periodic process in trickling filters or rotating biological contactors)
that readily and easily develops controlled unsteady state from those that grow in suspension as ‘‘activated
conditions that are required to meet discharge limits dic- sludge.’’ In both these systems, the mixed cultures
tated by the regulators. These distinguishing features and of microorganisms convert the contaminants to
others are described in this article. end products that depend on the nature of the
contaminants and the organism distribution that
BACKGROUND is present.
3. The mixing pattern established in the reactor. Two
Contaminants mixing patterns are most commonly employed:
completely mixed or plug flow. Intermediate mixing
Unwanted contaminants are present in liquids (e.g.,
intensities are often employed but are difficult to
municipal sewage, industrial wastewater, landfill lea-
model effectively.
chate, and contaminated groundwater), solids (e.g., soils
after a gasoline spill), and gases (e.g., odors from a 4. The type of reactor used. These include the com-
composting plant). They are organic or inorganic, more pletely mixed batch reactor (CMBR), the completely
or less dense than water, and either soluble or insoluble mixed flow reactor (CMFR), the plug flow reactor
in water. Some are emulsions in water while others (PFR), the plug flow reactor with dispersion (PFDR),
are not; some are volatile and some nonvolatile. Those and the sequencing batch reactor (SBR). Microor-
contaminants that are sparingly soluble in water are ganisms in these systems are either in suspension or
termed hydrophobic or nonaqueous phase liquids (NAPLs) attached to a bed of support medium (e.g., packing
and those that are not are termed as hydrophilic. material such as rocks, sand, ceramic, styrofoam,
Contaminants that are biodegradable serve as electron etc.) that is fixed, fluidized, or expanded.
donors for microorganisms that catalyze a plethora of 5. The control strategy used to operate the reactor
conversions that range from the minor modification of (i.e., periodically operated or used under continuous
a contaminant’s structure (e.g., through the addition or flow conditions). Continuous-flow reactors normally
removal of a hydroxy group) to the complete mineralization recycle settled biomass but can also be operated with
or oxidation of the contaminant to carbon dioxide, water, no-recycle.
and other inorganic compounds.
The primary conversions that take place in biological A typical process schematic of a continuous-flow
reactors involve soluble organic compounds that provide activated sludge system is shown in Figure 1. The
carbon and energy for the growth of a mixed culture of schematic represents one of the typical ‘‘controlled
heterotrophic microorganisms present in the biomass. As a unsteady state systems’’ that are in use today. As shown,
result, the success of a biological treatment system is often the biological wastewater treatment system is depicted
judged first by its ability to remove soluble biodegradable as a cascade of three CMFRs with biomass (or sludge)
organic compounds. When combined with powdered returned (i.e., recyled) from the third CMFR in the
activated carbon (PAC), the ‘‘partitioning power’’ of the cascade to the first CMFR after settling in the clarifier,
floc (i.e., the flocculated mass of microorganisms, PAC, which is sketched in Figure 1 as a cone-shaped vessel.
and other associated materials) is appreciably enhanced In this system, the microorganisms circulating in the
because soluble nonbiodegradable organic compounds are system are periodically exposed to different substrate
removed more aggressively. Biological systems are also concentrations with the highest concentration in the first
often operated to minimize the discharge of the nutrients, CMFR and the lowest in the last CMFR. As a result,
namely, nitrogen and phosphorus. The success of such biomass settled and recycled to the first tank is repeatedly
treatment systems is then judged by their ability to subjected to high and low growth-rate environments as
remove soluble biodegradable organic compounds and the it moves from reactor to reactor and back again into the
nutrients of interest. cascade.
Biological treatment systems are categorized in a In order for the sludge to concentrate well in the
number of ways: clarifier, the rate at which flocs settle must be suffi-
cient. Unfortunately, the rate of sedimentation is slowed
1. The microbial community’s ability to use exogenous when the relative abundance of microorganisms that
electron acceptors. Some use oxygen (i.e., in aerobic grow head to tail within long sheaths as filaments is
 ACTIVATED SLUDGE — SEQUENCING BATCH REACTORS 45

Influent Effluent THE CONCEPT

1 2 3 The SBR technology is used to gain control over the

structure and functions of the microbial community in a
multipurpose bioreactor that is exposed to varying influent
Return sludge conditions. An SBR system differs from a continuous-flow
activated sludge systems in the following manner
C in
1. The influent and effluent streams are uncoupled.
2. Biomass separation occurs in the biological reactor
C1 and not in a separate clarifier, as was shown in
C2
Figure 1.
C3 3. The unit operations and unit processes that take
place in a reactor follow each other in a time sequence
Figure 1. Process schematic of a continuous-flow activated that is progressively repeated in a ‘‘periodic’’ or
sludge system with the aeration basin designed as a reactor
‘‘cyclic’’ manner and not from tank to tank as is
cascade (top) and a profile of the substrate concentration from
the influent port of the cascade to the discharge from the cascade
done in space-oriented systems.
(bottom). Microorganisms circulated in the system are repeatedly 4. A portion of the treated water is periodically
exposed to high and low substrate concentration. discharged from each tank to make room for a new
batch of wastewater (see Fig. 2).

Because of these features, SBRs are referred to as

greater than some minimum acceptable level. When this
periodic processes, single-tank systems, fill and draw
happens, poor sedimentation results and the microor-
reactors, or as variable volume reactors.
ganism (or activated sludge) shown in the shaded zone
As represented in Figure 2, the SBR process is
of the clarifier in Figure 1 extends upward toward the characterized by a series of process phases (e.g., fill,
overflow weir causing the ‘‘clarified’’ effluent to contain react, settle, draw, and idle), each lasting for a defined
unacceptable quantities of suspended solids (i.e., the period. During the fill phase, the SBR behaves as a
flocs). Various researchers (2,3,10) demonstrated experi- semicontinuous CMFR (if the reactor is well mixed and/or
mentally that the repeated exposure of activated sludge aerated during fill) that receives flow but has no discharge.
to feast (e.g., in the first CMFR in Fig. 1) and then As is described later, the fill phase may be unmixed, mixed,
famine (e.g., in the third CMFR) conditions is a very or mixed and aerated. During the react phase, the SBR
effective means of controlling the excess growth of fil- behaves as a CMBR. Sludge wasting as a means to keep
amentous organisms. The frequent shifting of activated the biomass concentration in the reactor at a certain level
sludge between aerobic, anoxic, and anaerobic zones normally takes place after the settle phase but may take
establishes the conditions needed to enrich for microbial place near the end of react phase or during the settle phase
communities that carry out nitrification, denitrification, and may take place weekly, daily, or during each cycle.
and enhanced biological phosphate removal. Frequent As originally defined, the idle phase was included as
shifts of certain process conditions and the resulting con- a ‘‘buffer’’ phase in the SBR (4,5) (Fig. 2). No wastewater
trolled unsteady state conditions have significant and very enters the reactor during the idle phase and, unless sludge
important long-term effects on microorganism selection is wasted during this phase, there is no discharge. It is
and enrichment and on the physiological state of the usually neither mixed nor mixed and aerated. As such, the
microorganisms enriched. Accordingly, these short-term name ‘‘idle’’ is suitable. However, the time in the idle phase
unsteady state conditions, if properly selected and con- has an important function in the SBR as originally defined.
trolled, are an effective tool to maintaining long-term Idle phase provides the excess capacity in the system so
quasi–steady state conditions that control the structure
and function of the microbial communities on which the
performance of activated sludge systems is so greatly Fill React Settle Draw Idle
dependent.
In practice, the factors known to be effective in
controlling the structure and function of microbial ∆V f ∆Vd
aggregates (e.g., activated sludge or fixed films) are
difficult to maintain in continuous-flow systems. The Vo
frequency and amplitude of the changes needed to control
Effluent
variations in rate and time (e.g., from feast to famine)
Sludge
are not easily applicable to continuous-flow systems wasting
as they are to periodically operated systems such as
the SBR. Because of inflow rate variations, biological Cycle
treatment systems are often subjected to suboptimal Figure 2. Schematic of a typical SBR cycle. The Idle phase is
control conditions even when they are designed to operate optional if an alternative means for handling excess inflow is
as controlled unsteady state systems. provided. A cycle may include aerobic, anoxic, or anaerobic phases.
46 ACTIVATED SLUDGE — SEQUENCING BATCH REACTORS

that permit limits are met during periods of high flow. The feast conditions are established during static fill for the
time for idle phase seamlessly reduces toward zero when following reasons
high volumetric flow rates are observed and it increases
when there are low volumetric flow rates. The idle phase 1. The reaction rates are reduced in systems with little
can only be eliminated when an equalization tank or or no mixing. The only energy supplied to the system
holding tank is available or when there is some other during this static fill comes from the introduction of
means for handling the excess inflow (e.g., continuous the wastewater. The highest rates of reactions occur
influent as in the intermittent cyclic extended aeration in well-mixed systems.
system depicted in Table 1). 2. Because of the lack of mixing, contact between the
In some of the variable-volume activated sludge microorganisms and contaminant is minimal.
systems, there is no distinct react phase, and the 3. The extent of oxidation is limited by the avail-
settle and draw phases occur while the influent enters ability of exogenous electron acceptors. Accord-
the system. These differences are briefly described in ingly, although some electron donors (i.e., contam-
Table 1 for selected variable-volume activated sludge inants) and electron acceptors are utilized, and
systems. despite some fermentation reactions occurring in
In addition to the descriptions given earlier, the fill- the absence of exogenous electron acceptors, anoxic
and-react phases can have several subphases based on the and/or anaerobic reactions take place at a much
energy input to the system that results in various aeration greater rate during the mixed fill than during the
and mixing operating strategies. These subphases have static fill.
been labeled in various ways as follows
Similarly, initial aerobic reactions take place during
aerated fill. In general, the extent of the aerobic reactions
• static fill no energy input to the system allows
depends on the mass flow rate of the readily biodegradable
the accumulation of unreacted
substrates and the specific fill strategy employed. For
substrate.
example, anoxic reactions also occur during the aerated fill
• mixed fill mixing without forced aeration
if oxidized forms of nitrogen are present, the concentration
typically allows either anoxic or
of dissolved oxygen is sufficiently low, and an adequate
anaerobic reactions.
supply of electron donors (i.e., oxidized forms of nitrogen)
• aerated fill mixing with forced aeration typically
is available.
allows aerobic reactions.
The reactions initiated during the fill phase are
• mixed react mixing without forced aeration allows
completed during the react phase by initiating the proper
anoxic and possibly anaerobic reactions.
mixing and/or aeration strategy. The settle (separation of
• aerated react mixing with forced aeration allows
the biomass from the treated waters), draw (withdrawal of
aerobic reactions.
the treated water, i.e., the supernatant that forms during
the sedimentation), and idle (time between draw and fill
Contaminants in the wastewater enter the reactor that reflects the excess hydraulic capacity of the system)
during the fill phase. Unless a temporary influent storage phases complete the cycle.
volume is provided, more than one SBR is required The generic cycle arbitrarily commences with the fill
to handle a continuous inflow of wastewater. When a phase and terminates at the end of the idle phase.
holding tank for the influent is used, the time for the The volume of wastewater introduced into the reactor
fill phase is often only a small fraction of the total is Vf (Fig. 2). It is added to the volume of water and
cycle time and the wastewater constituents accumulate, sludge that remains in the reactor at the end of the
thus providing the conditions required for ‘‘feast’’ and past cycle (Vo ). At the end of the fill phase, the reactor
for the control of filaments. Without a holding tank, contains VR = Vo + Vf . Once the react phase has been

Table 1. Distinguishing Features of Selected Variable-Volume Activated Sludge Systems (6)

Distinct Distinct Distinct Idle

Named Systems Influent React Phase Settle Phase Effluent Phase

Cyclic Activated Sludge Periodic No Yes Periodic No

Technologya
Intermittent Cyclic Continuous No No Periodic No
Extended Aeration
Systemb
Intermittent Pasveer Ditchc Continuous No No Periodic No
SBR With Equalization Tank Periodic Yes Yes Periodic No
SBR without equalization Periodic Yes Yes Periodic Yes
tank
a
CAST Cyclic Activated Sludge Technology (7)
b
ICEAS Intermittent Cycle Extended Aeration System (8)
c
Pasveer ditch carousel ditch with no external clarifier (9)
 ACTIVATED SLUDGE — SEQUENCING BATCH REACTORS 47

Primary Influent SBRs Effluent Each of these generic groups is described in detail in
treatment holding tank buffer tank the following section (Fig. 4).
(optional) (optional)
1
Group A. The SBR as originally defined (4–6) is
shown in Figure 4a. Two or more tanks are normally
2 used. The influent is diverted from the tank being
filled when its water level reaches a predetermined
maximum value. The fill time ratio (FTR) for each
3 tank is less than one. The time for the fill phase is
defined by
tc
tf =
( ... ) n
where tf is the fill time, tc is the total cycle
n time, and n is the number of SBR tanks avail-
able.
Group B. SBR systems with a holding tank in front
Figure 3. Schematic of an SBR plant. (Fig. 4b) are often used in Germany (3). The influent
enters a holding tank before being distributed into
the SBR tanks. Two or more tanks are normally
completed and the mixing energy has been dissipated, used. The configuration of tanks allows the system
the activated sludge starts coagulating and settling. After to be operated with a variable fill strategy including
wasting excess sludge ( Vw ) and discharging the treated instantaneous fill to control filaments or extended
supernatant ( Vd ), the reactor is available to receive a fill strategies to avoid the accumulation of inhibitors
new supply of wastewater. (e.g., as could be present in industrial wastewater).
Thus, a SBR process is characterized by the following Therefore, an idle phase is not required to control the
set of parameters: system during variable flow conditions. The holding
tank in Group B provides the same function as the
ti Time for the ith phase idle phase in group A.
tc Total time of one cycle (tc =
ti )
Group C. Figure 4c describes SBR systems with an
FTR fill time ratio, tf /tc , where tf is the time for
internal or external constant volume selector basin.
fill
In these systems, fill is interrupted either during the
VER Volumetric exchange ratio, Vf /VR
draw phase (e.g., as in the Cyclic Activated Sludge
HRT or τ Hydraulic residence time, n • VR Q−1 ,
System) or during the settle and draw phases (e.g.,
where n is the number of tanks and Q is
as in the Cyclic Activated Sludge Technology). All of
the volumetric flow rate of the influent to
these systems have a recirculation line that returns
the treatment plant
mixed liquor from the main reactor to the selector.
The Cyclic Activated Sludge System has a chamber
In addition, process parameters that are typical for
that follows the selector to buffer the flow during
activated sludge or biofilm systems apply. For instance, the
design and operation of an activated sludge SBR usually settle.
includes considerations of key factors such as HRT, sludge Group D. Figure 4d describes SBR systems that
age, and sludge loading. have a continuous inflow of wastewater that is
not interrupted during the settle and draw phases.
These SBRs can be operated as single tanks or
DESIGN AND OPERATION OF SBR PLANTS as sets of tanks operated in parallel. Once the
water level in a tank reaches a predetermined
SBR plants typically consist of a number (n) of reactors. maximum value, the aerators and mixers are
Pre- and posttreatment facilities may be added. A general switched off, the activated sludge is allowed to
layout of an SBR plant is shown in Figure 3. Four generic settle, and the supernatant is withdrawn from
groups of SBRs can be distinguished (6). The specific fill the reactor. To minimize cross-contamination of
strategy used and/or the inclusion of a react or idle phase
the treated effluent with the influent wastewater
characterize each of them. The generic groups are as
during the draw phase, the reactor is divided
follows:
into two zones separated by a baffle. The first
of the two zones limits cross-contamination and
(A) Systems with a periodic influent, a react phase,
serves as a selector. Normally, SBRs such as the
and an idle phase
Intermittent Cyclic Aeration System operate with a
(B) Systems with a periodic influent, a react phase, completely aerated fill. In contrast to the operation
and no idle phase of Groups A and B, static fill occurs at the end
(C) Systems with interrupted influent, a selector, and of the cycle and is termed settle when there is no
no react and idle phases discharge, and draw when the treated effluent is
(D) Systems with a continuous influent. discharged.
48 ACTIVATED SLUDGE — SEQUENCING BATCH REACTORS

Influent Influent
Holding tank

Intermittent effluent
Static fill Intermittent effluent
Mixed fill
Aerated fill Mixed fill
Aerated react Aerated react
Settle Settle
Draw
Draw
Idle
Cycle
Cycle
SBR system: Group A SBR system: Group B

Interrupted influent Continuous influent

Selector
Effluent Intermittent effluent

Return flow
Fill
Fill Aerated fill
Static fill Settle (+ static fill)
Aerated fill Draw (+ static fill)
Settle
Draw
Return flow Cycle
Cycle

SBR system: Group C SBR system: Group D

Figure 4. Four different groups of SBR systems currently applied in practice (6).

Each of these modifications provides specific advan- phases) should be based on results obtained from
tages and disadvantages. Selection of a specific SBR treatability studies conducted at either bench or pilot
type must be made on a case-to-case basis after tak- scale. Treatability studies are especially recommended
ing into account the conditions under which the SBR when the wastewater to be treated is unique with
will be operated and the treatment goals that must be respect to composition, concentration, and variability as
achieved (6). Care should be taken, however, to select is the case for virtually all industrial wastewaters and
a system that maximizes the difference between the landfill leachates. In these cases, design is based on a
‘‘intensity’’ of the feast and the ‘‘depth’’ of the famine. mathematical description of the facility that integrates
Doing so will provide protection against bulking sludge projections of future flow and loading conditions with
while maximizing the potential for biological nutrient the data collected during the treatability studies (e.g.,
removal. stoichiometric, kinetic, settling, oxygen transfer, nutrient
requirements, etc.).
SIZING OF SBRS The design of SBRs for municipal wastewater treatment
facilities is often done without the aid of treatability stud-
Whenever possible, design calculations for an SBR plant ies because it is assumed, sometimes incorrectly, that
for tank size and cycle program (i.e., the operating municipal wastewater varies little from location to loca-
strategy that defines the sequence and duration of tion. When the characteristics of municipal wastewater
 ACTIVATED SLUDGE — SEQUENCING BATCH REACTORS 49

is ‘‘typical,’’ the design and operating conditions that are on the basis of the mass of influent COD removed.
selected are based on experience using methods that sim- The water volume needed in the reactor after the draw
ply set the hydraulic residence time (HRT) and organic phase, V0 , is calculated by determining the volume
loading to size the system. In other cases, especially for occupied by the biomass after settling for 30 minutes.
the design of large facilities and/or when nutrient removal The value is multiplied with a safety factor (SFV ) to
is required, engineers often combine experience with one cover any uncertainties with respect to the ability of the
of the many commercially available versions of computer sludge to settle and growth of the respective groups of
models based on International Water Association (IWA, microorganisms.
formerly IAWQ) (10) to test and refine possible design. The exchanged volume ( Vf ) depends on the selected
These simulation programs use default values for stoichio- cycle time. The total volume of the SBRs (VR ) is obtained
metric and kinetic coefficients. An intermediate approach by adding V0 and Vf . Further information on using
to these two extremes is used in Germany where the mass balances for the design of activated sludge systems
Waste and Wastewater Association (ATV) issued guide- can be found in Orhon and Artan (14) and Artan and
lines (11) for the design of SBR plants treating municipal coworkers (15).
wastewater. This guideline is based on established design
procedures for continuous-flow systems. Table 2 describes EXPERIENCES
the various assumptions and calculations that must be
made when using these guidelines. Unlike conventional technology, SBRs tend to be designed
It is assumed that the mass and the activity of the and marketed by equipment suppliers. To date there
activated sludge in the SBR will be the same as in a is little well-documented information on how existing
continuous-flow activated sludge system with plug flow SBRs are performing, including their cost-effectiveness,
characteristics (i.e., a cascade of three CMFRs). It is reliability, optimum design and operation, cost, and
further assumed that the order and the duration of the performance associated with different SBR configurations
SBR process phases are proportional to the design setting and equipment suppliers. The growth and acceptance of
and to the mean volumetric retention time in the various SBRs appear to be based on their ability to meet permit
zones of the continuous-flow systems (e.g., anoxic zone and limits and at their competitive capital and operating
aerobic zone). costs.
One of the several possible design procedures directly During the past three years, several surveys were
based on mass balances is presented in Table 3. The conducted in various countries to study the performance
example chosen is based on COD as a measure for the of SBR facilities (6). Survey results revealed that designer
organic fraction of the wastewater constituents. and operators were often not aware of the full range of
In determining an appropriate θX needed for nitrogen control that is provided by SBR technology. Despite these
removal, nitrification capacity (NOX ) and denitrification problems, effluent values were well below permit limits.
potential (NDP ) are two important factors that were In rural areas, the combination of a high-technology
first defined for recirculation-type continuous-flow pro- system (i.e., an SBR) and a natural system (e.g., effluent
cesses (12,13). polishing ponds) was very effective with respect to
After selection of the relevant sludge age (θX ), the chemical oxygen demand (COD) and five-day biological
total biomass present in the SBR reactors is calculated oxygen demand (BOD5 ) removal. Only small variations

Table 2. Design of an Activated Sludge SBR Plant Based on the German ATV Guidelines (11)

1. Input: Design information from continuous flow

system, operating strategy for the SBR
2. Calculation of the time (ta ) during which biological ta = tc − ts − td − tidle − tstaticfill
reactions may take place
tc
3. Effective mass of sludge in the SBR equals the mass MX,SBR = MX,Conti. ·
tc − ts − td
of sludge in the continuous-flow system
MX,SBR
4. Select MLSS in SBR after fill and, thus, n · VR n · VR =
CX
5. Select V0 depending on the volume of the settled min(n · V0 ) = MX,SBR · SVI · SFV
sludge or VER = 1 − CX · SVI × SFV
6. Select exchange ratio or cycle time
tc · Q
7. Select number of SBRs (n) and calculate missing < VER
n · VR
value (volumetric exchange ratio (VER) or cycle
time)
tD VD
8. For a system with nitrogen removal, select mixed =
tD + tN VD + VN
and aerated process phases according to mixed and
aerated volumes of the continuous-flow system
9. Check ts and td based on settling velocity and
maximal draw flow rate, respectively

Note: Table prepared by E. Morgenroth, University of Illinois at Urbana-Champaigne, U.S.A.

50 ACTIVATED SLUDGE — SEQUENCING BATCH REACTORS

Table 3. Design Procedure Based on Mass Balances

1. Input: operating strategy for SBR

2. Select solids retention time (θX )
tc − ts − td
3. Estimate effective fraction of tc from ts and td effective fraction =
tc
tc
4. Calculate mass of sludge in all reactors MX,SBR = YH,net · Q · CS · θX ·
tc − ts − td
5. Calculate volume of settled sludge in all reactors n · V0 = MX,SBR · SVI × SFV
6. Select exchange ratio or cycle time
n · V0 · VER
7. If exchange ratio was selected, then calculate cycle tc =
Q · (1 − VER)
time from
8. Calculate exchange volume of all reactors n · V = Q · tc
9. Calculate total volume of the SBRs n · VR = n · V0 + n · Vf
10. Select number of SBRs (n)
min θX,a
11. Calculate minimum aerated fraction of reaction min .aerated fraction = · SFX
θX
time on the basis of minimal aerobic solids
retention time, for example, nitrification
12. Choose number of fill phases and distribution of
the influent on the fill phases depending on the
wastewater composition (e.g., TKN/COD ratio)
13. For a system with nitrogen removal, check Nnit = TKN0 -NH4,e -Norg,e -NSL
nitrification capacity and denitrification potential 1
NDP = (1 − YH,net ) · CS,in · · fDN · η
depending on cycle design 2.86
NDP · n · Vf
14. Check react time for nitrogen removal using tD =
rDN,X · MX,SBR
denitrification rates
15. Check ts and td based on settling velocity and
maximum draw flow rate, respectively

Note: Table prepared by E. Morgenroth, University of Illinois at Urbana–Champaign.

are found in the effluent despite the wide range of influent The responses received during the survey were used
concentrations. Even very low effluent standards such as to generate a list of concerns, which, if addressed,
40 mg L−1 COD were safely met. would likely improve the performance of these SBR-like
A study funded by Ontario Ministry of the Environ- installations. The top ten of these are summarized in the
ment, Environment Canada, and the Water Environment following text:
Association of Ontario evaluated 75 SBR-like installations
in Ontario and the United States Great Lakes Region (16). 1. Operators have no formal training on SBR
A telephone and mail survey was conducted and 10 SBR operation or process control.
facilities were visited in 1998 to define possible design and 2. Mechanical equipment located outdoors may freeze
operational problems. The average daily design flow for the or malfunction (e.g., air valves, solenoid valves,
67 of the 75 SBR plants providing such data was approx- decanter arms, and level monitoring floats).
imately 3,600 m3 d−1 and ranged from a low of 7 m3 d−1 to 3. Decanters were unable to meet some specific
a high of 22,710 m3 d−1 . Results obtained from the 75 SBR treatment requirements (e.g., the discharge of
plants were prioritized according to operating costs, plant floatables).
capacity, effluent quality, and frequency of occurrence. 4. Discontinuous SBR effluent flow have an impact on
No effort was made to subcharacterize the findings in downstream treatment processes (e.g., ultraviolet
terms of the four generic groups of SBRs described ear- (UV) disinfection and the backwashing of filters).
lier.
5. Lack of online DO monitoring and control instru-
The average daily flow for the 49 of the 75 SBR plants mentation eliminated the possibility of increased
reporting actual flow data was nearly 2,600 m3 d−1 . Forty- energy savings.
six of the plants reported effluent readings for BOD5 ; 58 for
6. No specific control strategy was provided for the
SS; 52 for ammonia-nitrogen; and 33 for total phosphorus
solids residence time.
(as P). The average results for those reporting were less
than 6.5 g m−3 , 9 g m−3 , 1.5 g m−3 , and 1.4 g m−3 for BOD5 , 7. The pretreatment systems (e.g., bar screens and
SS, NH3 , and phosphorous, respectively. These results are comminutors) were inadequately designed.
quite impressive for facilities located in either warm or 8. Lack of automation for the selection of the time
cold regions. In addition, a preliminary cost comparison needed for sludge wasting.
between SBRs and continuous-flow activated sludge plants 9. Potential secondary phosphorus release in aerobic
indicated that for similar effluent requirements, SBRs- digesters.
like facilities were more economical than continuous-flow 10. Partial failure of the SBR control program during
activated sludge plants. peak flows.
 ACTIVATED SLUDGE — SEQUENCING BATCH REACTORS 51

Most of the concerns listed earlier could easily be Water from

corrected if the operators were given a formal training filter washing
program. Unfortunately, there is not a simple solution
Effluent
because the term SBR is used to include many different
types of periodically operated reactor configurations, even
more than the four generic types described earlier, and Influent
the specific training for each type would be different. The Recirculation
SBR is also a new technology that is being implemented by
some who have little basic understanding of how to make Influent
it work best. holding Storage of
A few of the concerns listed previously are specific tank washing
water
to the equipment used for SBR facilities. These are
difficulties that can be overcome by using the equipment
specified by qualified engineers and manufactured and Figure 5. Schematic of a SBBR plant (6).
installed by reputable vendors. The remaining concerns
were independent of the type of treatment facility installed
and are just as likely to cause difficulties at a conventional Influent
continuous-flow activated sludge plant as they are at Fill
an SBR system. It is worth noting that the effluent
Packing
data compiled showed that the effluent criteria were
consistently met, and in most cases, exceeded.
Blower
The evaluations indicate that SBRs are a cost-effective
wastewater treatment technology (17,18). Unfortunately,
limited historical data have been compiled comparing the Drain
cost of SBRs with other types of activated sludge treatment
systems. Clearly, the lack of need for an external secondary
Cycle
clarifier and its associated hardware and the return sludge
pumping system offers potential savings in construction
costs. In addition, primary clarification is not normally
employed (none of the 75 plants evaluated had primary Aeration
Effluent Carbon source
clarifiers).

NOVEL APPLICATIONS FOR BIOFILM AND SLURRY

SYSTEMS

The concept of implementing short-term unsteady state

conditions to control the long-term performance of
biological reactors is applicable to either aerobic or
anaerobic systems, or to suspended growth or biofilm, or Denitrification
fluidized bed systems that treat wastewater, contaminated Figure 6. Schematic representation of the SBBR process.
soil, or gases. Over the past years, a wide variety
of innovative batch treatment technologies have been
described in literature. The common basis of these novel Breakthrough events can be successfully avoided by
approaches is application of well-defined sequences (i.e., temporary prolongation of the duration of the react phase.
operating strategies) that alter process conditions in a Jaar and coworkers (22), Kolb (23), Chozick and
controlled manner. The common goal is control of the Irvine (24), and Irvine and coworkers (25) packed the
composition and activity of microbial communities by reactor with granular activated carbon. By means
targeting the microorganisms that should be selected, of adsorption processes during the fill phase, the
enriching these organisms to the levels needed, and concentration of critical pollutants (e.g., volatile organic
then adjusting their physiological state so that permit compounds or inhibitory organic substances) were reduced
limits are achieved. Two of those systems, the Sequencing in the bulk liquid. During the react phase, biodegradation
Batch Biofilm Reactor (SBBR) and the Soil Slurry- and desorption became the dominant processes. The
Sequencing Batch Reactor (SS-SBR), are described wastewater was effectively treated and the activated
briefly here. carbon was regenerated biologically.
Gonzales and Wilderer (19) and Wilderer (20) proposed Irvine and Montemagno (26) proposed the sequencing
the sequencing batch operation of a biofilm reactor. The batch operation of a soil slurry reactor in 1989 for the reme-
schematic of a SBBR plant is presented in Figure 5. The diation of trinitrotoluene-contaminated soils. In 1993,
sequence of process phases is sketched in Figure 6. Irvine and coworkers (27,28) reported on how the SS-SBR
Kaballo and coworkers (21) compared the response can be used to treat soil contaminated with the plasticizer
of continuous-flow and SBBR reactors to peak loading. bis-(2-ethylhexyl) phthalate and petroleum products.
52 ACTIVATED SLUDGE — SEQUENCING BATCH REACTORS

The schematic of an SS-SBR facility is presented in the impact of varying influent conditions is minimized. The
Figure 7. SBR concept can be applied to the treatment of wastewater
The SS-SBR is operated on a fill and draw basis in in both activated sludge and biofilm systems and to the
a manner similar to that described for the SBR. Each treatment of contaminated soils and gases.
tank is filled during a discrete period (fill phase), and then In contrast to the continuous-flow activated sludge
operated as a batch reactor during the react phase. Just as systems, biomass separation occurs in the biological
for the SBBR, the cycle for the SS-SBR does not include the reactor and not in a separate clarifier. The unit operations
settle phase (Fig.7). After the desired contaminant levels and unit processes that take place in each reactor follow
have been reached, a fraction of the slurry is removed each other in a time sequence that is progressively
from the SS-SBR (draw phase). A certain fraction of the repeated in a ‘‘periodic’’ or ‘‘cyclic’’ manner and not from
treated slurry remains in the reactor to provide acclimated tank to tank as they do in space-oriented systems. A
microorganisms for the next batch of untreated slurry. portion of the treated water is periodically discharged from
The optimum recycle fraction in the SS-SBR must be each tank to make room for a new batch of wastewater.
determined for each contaminated soil (27–29). Because of these features, SBRs are referred to either
As can be seen in the bioslurry system illustrated as periodic processes, single-tank systems, fill-and-draw
in Figure 7, the soil is first screened to remove rocks, reactors, or as variable volume reactors.
branches, and other debris. Water is added to the The SBR process is characterized by a series of process
screened soil to make a slurry in a small mixing tank. phases (e.g., fill, react, settle, draw, and idle) each
The concentration of solids in a bioslurry reactor can lasting for a defined period. Four generic groups of SBRs
range from 5 to 50% (weight/volume) depending on the were distinguished, each characterized by the specific
contaminant concentration and the capacity of the mixing fill strategy utilized, or the inclusion of a react or idle
and aeration equipment. Nutrients and other possible phase. Each generic group provides specific advantages
amendments (e.g., microbes and surfactants) are also and disadvantages. Selection of a specific SBR type
added to the mixing tank if needed. The slurry is then must be made on a case-to-case basis after taking into
transferred to the bioslurry reactor via a slurry pump. account the conditions under which the SBR will be
Compressed air diffusers or surface agitators can be operated and the treatment goals that must be achieved.
used to provide oxygen. Bioslurry reactors are easily Care should be taken, however, to select a system that
covered and equipped with an emissions recovery system. maximizes the difference between the ‘‘intensity’’ of the
Fugitive emissions may require a separate treatment feast and the ‘‘depth’’ of the famine; this will protect
before discharge or they can be recycled back into the against bulking sludge while maximizing the potential for
slurry reactor for biodegradation. After the soil is treated, biological nutrient removal. The technical report on SBR
solids are separated from the slurry water using standard technology issued by the International Water Association
dewatering techniques. (IWA Publishing) was written to provide a comprehensive
The SBR has also been shown to be a cost-effective description of the information needed by designers and
and energy-efficient means of removing hazardous organic operators (6).
compounds found in industrial wastes (30) and leachates
from landfills (31–33). Periodically operated biofilters
have recently been proposed for the treatment of air NOMENCLATURE
contaminants (33,34–36).
BOD5 Five-day biological oxygen g m−3
CONCLUSION demand
COD Chemical oxygen demand g m−3
SBR technology defines a controlled unsteady process that CS Substrate concentration g m−3
imposes selective pressures and manipulates both the Cx Biomass concentration g m−3
community structure and expression of a heterogeneous FTR Fill time ratio, tf /tc , where tf –
microbial consortium in a multipurpose bioreactor so that is the time for fill

Influent air with oxygen

Effluent air with VOCs

Bioslurry
reactor
Effluent slurry
Soil screening and
nutrient addition
Blower and
air scrubber

Figure 7. Schematic representation of a SS-SBR.

 ACTIVATED SLUDGE — SEQUENCING BATCH REACTORS 53

fDN Fraction of electron – BIBLIOGRAPHY

acceptors potentially
associated with anoxic 1. J. Chudoba, P. Grau, and V. Ottova, Water Res. 7, 1389–
reactions 1406 (1973).
HRT Hydraulic residence time, h 2. S. C. Chiesa and R. L. Irvine, Water Res. 19, 471–479 (1985).
n · VR Q−1 3. P. A. Wilderer and E. D. Schroeder, Anwendung des Sequenc-
M or Mx Mass of sludge g ing Batch Reactor (SBR)-Verfahrens zur biologischen
MLSS Mixed liquor suspended g m−3 Abwasserreinigung (Application of SBR Technology for the
Biological Treatment of Wastewater), vol. 4, Hamburger
solids
Berichte zur Siedlungswasserwirtschaft, Germany, 1986.
N Concentration of nitrogen g m−3
compounds 4. R. L.Irvine and W. B.Davis, Use of Sequencing Batch Reac-
−3 tors for Waste Treatment — CPC International, 26th Annual
NDP Denitrification potential gm
Purdue Industrial Waste Conference, Purdue University,
Nnit Nitrification potential g m−3 West Lafayette, Ind., 1971, pp. 450–462.
NSL Nitrogen concentration of g m−3
5. R. L. Irvine and A. W. Busch, J. Water Pollut. Control Fed.
the waste sludge
51, 235–243 (1979).
N Number of tanks or reactors –
6. P. A. Wilderer, R. L. Irvine, and M. C. Goronszy, Sequencing
PAC Powdered activated carbon –
Batch Reactor Technology, No. 10, IWA Technical Report
Q Flow rate (plant influent) m3 h−1 Series, IWA Publishing, London, U.K., 2000.
rDN,X Denitrification rate g g−1 h−1
7. M. C. Goronszy, Proceedings 7th International Conference on
SVI Sludge volume index ml g−1 Design and Operation of Large Wastewater Treatment Plants,
tc Total cycle time (tc =
ti ) h ‘‘Large scale application of cyclically operated activated
td Time of the decant phase h sludge technology’’, Aug.–Sept. 1995, Vienna, Austria, 1995.
tD Time of the denitrification h 8. M. C. Goronszy, J. Water Pollut. Control Fed. 51, 274–287
phase (1979).
ti Time of the ith phase h 9. A. Pasveer, Tech. Eau (Brussels) 15(171), 17–28 (1961).
tidle Time of the idle phase h 10. M. Henze et al., Activated Sludge Model No. 1, IAWPRC
tf Time of the fill phase h Scientific and Technical Reports, No. 1, IAWQ, London, U.K.,
tN Time of the nitrification h 1987.
phase 11. ATV, Belebungsanlagen im Aufstaubetrieb (Activated sludge
ts Time of the sedimentation h variable volume reactors), Merkblatt ATV-M 210, Abwasser-
phase technische Vereinigung, Hennef, Germany, 1997.
tstaticfill Time of the h 12. A. C. van Haandel, P. L. Dold, and G. v. R. Marais, Water
no-mix/no-aeration fill Sci.Technol. 14, 443–461 (1982).
phase 13. G. A. Ekama, I. P. Siebritz, and G. v. R. Marais, Water Sci.
SFv Safety factor – Technol. 15, 283–318 (1983).
TKN Total Kjeldahl nitrogen mg L−1 14. D. Orhon and N. Artan, Modeling of Activated Sludge
VER Volumetric exchange ratio, – Systems, Technomics, Lancaster, Calif., 1994.
Vf / VR 15. N. Artan et al., Water Sci. Technol. 43(3), 53–60 (2001).
V0 Liquid volume at low water m3 16. D. Nolasco et al., Environ. Sci. Eng. 11(4), 42–49 (1998).
lever
3
17. USEPA, An Emerging Technology — Sequencing Batch Reac-
Vf Fill volume m tors — A Project Assessment, Washington, D.C., U.S. Envi-
VR Total liquid volume at high m3 ronmental Protection Agency, 1983.
water level 18. USEPA, Sequencing Batch Reactors for Nitrification and
VD Volume of the denitrification m3 Nutrient Removal, Washington, D.C., U.S. Environmental
zone (continuous-flow Protection Agency, 832-R-92-002, 1992.
plant) 19. S. Gonzalez-Martinez and P. A. Wilderer, Water Sci. Tech.
VN Volume of the nitrification m3 23(7–9), 1405–1415 (1991).
zone (continuous-flow 20. P. A. Wilderer, Sequencing Batch Biofilm Reactor Technol-
plant) ogy, in M. R. Ladisch and A. Bose, eds., Harnessing Biotech-
Vw Volume of the wasted sludge m3 nology for the 21st Century, American Chemical Society, 1992,
Vd Volume of the decanted m3 pp. 475–479.
effluent 21. H. -P. Kaballo, Y. Zhao, and P. A. Wilderer, Water Sci.
Y Cell yield g g−1 Technol. 31(1), 51–60 (1995).
θX Sludge age (HRT) d 22. M. A. A. Jaar and P. A. Wilderer, Water Sci. Technol. 26(5,6),
θX,a Aerobic sludge age d 1195–1203 (1992).
η Correction factor for oxic – 23. F. R. Kolb and P. A. Wilderer, Water Sci. Technol. 31(1),
respiration 205–213 (1995).
τ Hydraulic retention time h 24. R. Chozick and R. L. Irvine, Environ. Prog. 10, 282–289
(HRT) (1991).
25. Granular Activated Carbon — Sequencing Batch Biofilm
Reactor (GAC-SBBR), U.S. Patent No. 5, 126, 050 (June
54 ACTIVATED SLUDGE — THE FLOC

30, 1992), R. L. Irvine and L. H. Ketchum Jr., P. A.Wilderer, Typical problems are often encountered because of poor
and C. Montemagno. flocculation or poor settling of the sludge flocs (1,3,4).
26. R. L. Irvine, and C. D. Montemagno, TNT Slurry Reactor, The dewatering and further treatment of the surplus
presented by R. L. Irvine at the Workshop on Research sludge is also an important step in the solid-liquid
and Development Efforts in Composting of Explosives separation (5). This step is usually costly, and special
Contaminated Soils, U.S. Army Toxic and Hazardous care must be taken to handle sludge with a high content of
Materials Agency, New Orleans, La., Sept., 1989.
pathogenic bacteria, heavy metals, micropollutants, and
27. R. L. Irvine, J. P. Earley, G. J. Kehrberger, and B. T. Dela-
other unwanted substances. The dewatering process is
ney, Environ. Prog. 12, 39–44 (1993).
strongly affected by the floc properties, and problems with
28. R. L. Irvine, P. S. Yocum, J. P. Earley, and R. Chozick, Water
the dewatering process caused by poor floc properties are
Sci. Technol. 27, 97–104 (1993).
often encountered (5).
29. D. P. Cassidy and R. L. Irvine, Water Sci. Technol. 35(1)
Activated sludge flocs, like other biological aggregates
185–192 (1997).
such as biofilms, are complex heterogeneous structures in
30. R. L. Irvine and L. H. Ketchum, Jr., CRC 18(4), 255–294
which bacteria and other chapter provides an overview
(1988).
of the physical, chemical, and microbial composition
31. R. L. Irvine, R. L. Sojka, and J. F. Colaruotolo, Haz. Waste 1,
of the flocs. Furthermore, the principles of flocculation
123–135 (1984).
and possible correlations between floc composition,
32. P. A. Herzbrun, R. L. Irvine, and K. C. Malinowski, J. Water
floc properties, and functional sludge properties are
Pollut. Control Fed. 57, 1163–1167 (1985).
described.
33. R. G. Smith, and P. A.Wilderer, Treatment of Hazardous
Landfill Leachate Using Sequencing Batch Reactors with
Silicone Membrane Oxygenation, 41st Industrial Waste PHYSICAL AND CHEMICAL STRUCTURE OF FLOCS
Conference, Purdue University, Lewis Publisher, Chelsea,
Mich., 1987, 272–282. An activated sludge floc consists of many different
34. R. L. Irvine and W. M. Moe, Water Sci. Technol. 141(4,5), components as shown in Figure 2. The components are
441–444 (2000). bacterial cells, various types of extracellular polymeric
35. W. M. Moe and R. L. Irvine, J. Environ. Eng. 126(9), 815–825 substances (EPS), adsorbed organic matter, organic fibers,
(2000). and inorganic compounds. The bacteria can be single,
36. W. M. Moe and R. L. Irvine, J. Environ. Eng. 126(9), 826–832 growing in microcolonies (sometimes termed clusters or
(2000). microflocs) or growing as filaments. This basic composition
is common to all flocs, but the relative proportion of the
components and the exact types of chemical compounds or
ACTIVATED SLUDGE — THE FLOC types of microorganisms vary from plant to plant. In plants
treating domestic wastewater, the cell biomass typically
PER HALKJÆR NIELSEN constitutes 10 to 20% of the organic matter, whereas the
Department of Environmental size of other fractions is not certain. Typical values from
Engineering various sources are shown in Figure 3 (6–8).
Aalborg University The size of activated sludge flocs ranges from very
Aalborg, Denmark small aggregates of only a few cells (few µm) to large
flocs of more than 1 mm. In most activated sludges, the
Use of activated sludge is the most common process world- flocs are typically 40 to 100 µm in diameter, relatively
wide for treatment of numerous types of wastewaters. strong, and not easy to break apart (3,4,9). Usually, this
The main process is the biological conversion of organic fraction also constitutes the largest part of the sludge as
matter, nitrogen, phosphorus, and other compounds into measured by volume, whereas very small flocs that are
biomass (sludge) and gaseous compounds (1,2). Another numerous and single cells are less important in terms of
key process is the separation of the biological sludge flocs mass or volume. An example of size, surface area, mass,
from the treated water, the solid-liquid separation. This and volume distribution of activated sludge flocs for five
process relies entirely on the formation of good, activated conventional treatment plants is shown in Table 1 (10).
sludge flocs to ensure a proper effluent quality and a good Under certain circumstances, very dense flocs or granules
dewaterability of the sludge (1,3,4).
The solid-liquid separation involves several steps
(Fig. 1): floc formation in the process tanks and floccu- Flocculation
lation, clarification, and settling in the final clarifier to Clarification
retain the majority of the sludge. A large percentage of the Settling
Effluent
retained sludge (typically 80–97%) is recycled in order to Wastewater Biol. processes
water
keep a sufficiently high sludge concentration in the pro- Floc formation
cess tanks. In a typical activated sludge treatment plant,
Sludge
the amount of suspended solids in the process tanks, the Return sludge
mixed liquid suspended solids (MLSS), is typically 3 to 5 g Thickening Dewatering
suspended solids/L (gSS/L). To meet effluent standards of and storage
less than 20 to 30 mgSS/L, it is obvious that the entire Figure 1. Floc formation and solid-liquid separation in an
floc formation and the separation process must work well. activated sludge treatment plant.
 ACTIVATED SLUDGE — THE FLOC 55

Adsorbed matter high content of mixed liquor suspended solids. In the clar-
ifier, the floc-aggregates have sizes up to several mm in
diameter. Although these may have a lower density than
Microcolony smaller flocs, they settle faster on account of their greater
size and lower friction (Fig. 4). The very loose structure
Filamentous of large floc-aggregates appears to allow a flow of water
bacterium
through the porous structure during the movements in the
process tank and during settling (12).
Morphology of individual flocs is highly variable,
Organic fiber ranging from dense, strong flocs to very irregular,
Extracellular loose structures (Fig. 5). The organization of the various
polymeric substances structures, microcolonies, bacteria and the like is very
(EPS) and single heterogeneous (8,13). It is possible to quantify the floc
bacteria shape by its fractal dimension. Typical values of 1.9 to
Figure 2. Schematic illustration of the constituents of activated 2.5 (14) have been reported, showing the loose structure
sludge flocs. and high porosity of flocs. The floc size, shape, and
other properties can be described qualitatively (3,15)
according to simple microscopic investigations and related
to common operational problems in activated sludge
Protein (40−60%) plants (3):
Extracellular polymeric
substances (EPS) Humics (20−30%) 1. Dispersed growth, where hardly any true flocs are
(40−60%) Polysaccharides (10−20%) formed and cells are freely suspended in liquid;
Nucleic acids (2−5%)
Fibers, other organic 2. Pinpoint floc, where the flocs are small and fragile
Other, lipids, .., (5−10%)
material (10−30%) without a proper macrostructure or backbone from
Cell biomass (10−20%) EPS filamentous bacteria;
Total organic matter 3. The ideal floc, where the filaments are largely
Figure 3. Composition of the organic part of sludge (EPS: restricted to the floc interior with limited extension
Extracellular polymeric substances). into bulk water;
4. Filamentous bulking sludge, with extensive growth
of filamentous bacteria inside and between the flocs.
with a diameter of 300 to 500 µm can be formed in aerobic
The Chemical Composition of Flocs
activated sludge systems. So far, however, aerobic granule
formation has only been reported from laboratory-scale On the basis of weight, the organic matter (the volatile
reactors (11). suspended solids, VSS) is usually the largest fraction,
The individual flocs can undergo three main types typically 60 to 80% of the dry matter weight (16). The
of transformation when they are subjected to some inorganic fraction consists of various ions, some within
stirring (shear): erosion, fragmentation, and flocculation the bacteria and some (mainly Mg2+ , Ca2+ , Fe3+ ) adsorbed
(Fig. 4). Fragmentation, during which flocs are split into in the EPS matrix, and attached minerals, such as clay.
a few large new flocs, is not considered important for Most of the organic matter in activated sludge is protein
activated sludge flocs (4). Erosion takes place when small
particles, typically bacteria or EPS fragments are sheared
off the main floc because of several processes that are (a)
described later.
Flocculation of the individual flocs into loosely con- Erosion
nected, voluminous floc-aggregates takes place in the
process tanks and in the final clarifier because of the

(b)
Fragmentation
Table 1. Examples of Distribution in Activated Sludge
Flocs in Size, Surface Area, Volume, and Mass (10)

Distributions (c)
Size (µm) By Number By Area By Volume By Mass Flocculation

<2 30–80% <2% Negligible 2–8%

2–16 17–70% 1–8% Negligible 5–28%
16–128 1–4% 20–70% 10–80% 50–80%
128–256 <1% 5–50% 10–70% 5–25%
>256 Negligible 0–50% 0–60% 0–30%
Figure 4. Flocculation behavior of activated sludge flocs.
56 ACTIVATED SLUDGE — THE FLOC

(a) very important for the floc properties (see the following
section), several attempts have been made to quantify
and characterize this fraction in greater detail. However,
today there is no single universal extraction method
that effectively extracts all EPS components without cell
lysis (17). The most efficient extraction method at present
is based on a combination of shear and ion exchange (17)
and shows that protein is the predominant compound
(Fig. 3). The total amount of EPS has been estimated to
constitute 40 to 60% of the organic matter in activated
sludge flocs (17). The exopolymers may be very different
from one microcolony-forming species to another, and
because part of the EPS components seems to be loosely
attached to the flocs as a ‘‘cloud’’ (18), it is difficult to
extract EPS that is representative of the entire floc.
(b) It is important to note that polysaccharides, which are
assumed to be an important part of bacterial exopolymers,
are not present in large amounts in typical sludge
flocs (6,17). Instead, protein seems to act as the most
important ‘‘glue’’ component. The exact function of the
large protein pool is not well understood, but exoenzymatic
activity is present (19,20) and lectin-like proteins may
have a structural function (21). Humic substances can be
a large fraction in systems in which these substances are
present in the wastewater and in which the sludge age
is long.
As described in the next section, the EPS fraction
is important for the formation of a three-dimensional
gel-like network that keeps all the cells and other
constituents together. The surface charge of the EPS-
(c)
covered entities is generally assumed to be important
for the gel properties and can be characterized by the
zeta potential (or electrophoretic mobility), pH titration,
and colloid titration. Typical values for the surface
charge of activated sludge floc components are −0.5 to
−1 meq/g SS (22,18).

Floc Microbiology
Most bacteria are known to grow in aggregates (flocs
or biofilms) in natural and engineered systems, which
provide a number of advantages for the bacteria when
compared to suspended growth. In particular, the pres-
ence of EPS components ensures a well-buffered local
Figure 5. Examples of activated sludge flocs. (a) Large strong chemical environment that provides substrate and impor-
flocs form a nutrient removal plant, (b) Typical activated sludge tant ions and protection against predators and toxic
from nutrient removal plant, (c) Small, dispersed flocs. compounds. Furthermore, close proximity to other cells
improves interspecies substrate and gene transfer. Con-
siderable effort has been made in recent years to gain
(40–60%), and polysaccharide material, humic substances knowledge about the most important mechanisms con-
and other compounds (e.g., lipids, nucleic acids) each trolling floc and biofilm formation. This has been sup-
constitute 10 to 20% (6). Only a minor part of these ported by the recent development of tools, such as light
compounds represents the living cell biomass. Therefore, microscopy, epifluorescence microscopy, and confocal laser
these compounds arise mainly from the EPS fraction, the scanning microscopy (CLSM) (23), for in situ studies. Use
adsorbed organic debris, and the fibers. of microsensors has provided valuable knowledge about
The EPS matrix consists of various macromolecules, chemical gradients (24); fluorescence in situ hybridization
such as proteins, polysaccharides, nucleic acids, humic (FISH) with gene probes and other molecular methods
substances, various heteropolymers, lipids, and so on. have revealed information about the identity and phy-
(Fig. 3). The macromolecules are partly exopolymers logeny of important microorganisms (25,26, FILAMENTOUS
produced by bacterial activity and lysis and hydrolysis BACTERIA IN ACTIVATED SLUDGE: CURRENT TAXONOMIC STATUS
products, but they are also adsorbed from the wastewater. AND ECOLOGY), and microautoradiography (MAR) provides
Because the EPS matrix is widely acknowledged to be information about their ecophysiology (27,28).
 ACTIVATED SLUDGE — THE FLOC 57

The bacteria within the flocs are present as single total bacterial population. Less is known about the other
cells, as microcolonies, or as filamentous bacteria (Fig. 2). groups such as Fe(III)-reducers and sulfate- reducers, but
The filamentous bacteria are, in general, well known they are always present (35).
and well described in municipal treatment plants (3,15,
ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING Mechanisms of Floc Formation
COMMUNITY COMPOSITION), but much less is known about Formation of activated sludge flocs (bioflocculation)
filamentous bacteria in industrial plants. Little is known is controlled by several processes as illustrated in
about single cells in the flocs. Some are probably attached Figure 6. Formation is partly determined by adsorption
and incorporated into the floc matrix from the incoming or flocculation of various components (e.g., of bacteria
wastewater and may not be active. However, several from the wastewater) and partly by growth and other
species, for example, Spirochaetes and Spirillum, can be biological processes taking place within the floc. Among
observed by microscopy to be moving around in some the dominant physicochemical processes are adsorption-
sludge types (15), showing that not all active bacteria desorption of bacteria, EPS-components, and other organic
grow in microcolonies. and inorganic compounds. The desorption process, that is,
Microcolony formation is described among several when the components are detaching, is usually termed
bacterial groups in activated sludge, including Zoogloea deflocculation, disintegration, or erosion of the flocs (4).
spp., often forming very typical finger-like colonies easily It is also important to note, as illustrated in Figure 6,
recognizable by microscopy (15,29). Zoogloea spp. were that bacteria that do not properly adhere to the flocs
originally believed to be the most important species will not be retained in the treatment plant and will
forming flocs in activated sludge, but it appears that leave with the treated effluent from the treatment plant.
they are mainly present in large amounts in high-loaded Thus, an accumulation and selection of bacteria (and other
conventional systems and hardly present in advanced compounds) only takes place if they adhere well to the flocs
treatment plants with N- and P-removal (30,31). They or if they grow within the flocs.
are also often present in treatment plants with plug flow
or selectors to control filamentous bulking (1). Bacteria Physicochemical Factors. The different constituents of
belonging to the genus Thauera can form zoogloeal the floc are embedded in an EPS matrix (or gel) kept
clusters in industrial wastewater treatment plants (31). together by various intermolecular forces. The forces are
Little is known about other important microcolony- the DLVO interactions (van der Waals and electrostatic
forming heterotrophic bacteria with regard to both their forces) and non-DLVO interactions (hydrogen bonds,
phylogeny and physiology. Several species have been hydrophobic forces, steric forces, and bridging of EPS
isolated from flocs (2), but their importance remains by means of multivalent cations) (36,37). Furthermore,
unknown. In activated sludge from treatment plants physical entanglement of the long EPS macromolecules
performing biological phosphorus removal, microcolonies may be important in the formation of the network
of phosphorus accumulating organisms (PAOs) are present structure of the floc. In particular, electrostatic forces
and seem to form strong colonies of importance to the and bridging are believed to be important. Most floc
floc properties. However, these bacteria have not yet constituents are negatively charged overall because of
been properly identified (ACTIVATED SLUDGE — MOLECULAR the presence of proteins, polysaccharides, and humic
TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION). In substances with net negative charges at neutral pH
many plants the so-called ‘‘G-bacteria’’ are also present, (e.g., carboxylic groups). These are largely cross-linked
and they form small, distinct microcolonies (ACTIVATED by divalent (Mg2+ , Ca2+ ) and trivalent cations (Fe3+ ,
SLUDGE — THE ‘‘G-BACTERIA’’). Al3+ ). The hydrophobic forces are probably also important,
Autotrophic nitrifying bacteria can form large micro- but less understood. It is known, however, that when
colonies within the flocs in nitrifying treatment plants. relatively hydrophobic bacteria are added to activated
Several studies using FISH have clearly shown that sludge, they adhere better to sludge flocs than more
ammonium oxidizers (mainly Nitrosomonas, Nitrosococ- hydrophilic bacteria (38).
cus, and Nitrosospira) as well as nitrifiers (mainly Nitro- Figure 6 deals with adsorption-desorption phenomena
spira) grow in colonies, often in close proximity to each and consequently with various equilibria. These are
other (32: ACTIVATED SLUDGE — MICROBIOLOGY OF NITROGEN determined by the turbulence in the system (e.g., aeration,
REMOVAL). These colonies appear to be strong and difficult stirring, and pumping) and by the interactive forces of
to break apart. the entities, which are largely determined by the ionic
The total number of bacteria in activated sludge, strength, ionic composition, and pH mentioned earlier.
as determined by DAPI-staining using epifluorescence As described in the following section, changes in these
microscopy, is reported to be approximately 1012 per g parameters will affect the floc composition and the floc
VSS (6). Living or active cells can be identified using strength.
fluorescence in situ hybridization (FISH) with gene
probes, by reduction of the redox dye CTC, or by Biological Mechanisms. A main process in floc formation
microautoradiography (26,33,34). This fraction typically in the activated sludge is growth and activity of the
constitutes 80 to 90% of all cells in activated sludge bacteria within the floc. In addition, bacteria and
although CTC active cells are often fewer (33). Most of the particles from the incoming wastewater are adsorbed
bacteria seem to be heterotrophic (34). In nutrient removal by existing flocs. Most bacteria entering the treatment
plants, the nitrifiers typically constitute 1 to 10% of the plant are already aggregated in the sewer system or
58 ACTIVATED SLUDGE — THE FLOC

Dissolved EPS components

FLOC

Single bacteria Microbial growth

Polymer production
Inorganic ions Lysis
(e.g. Mg 2+, Ca 2+, Fe3+ ) Hydrolysis
Respiratoric activity

Adsorbed matter

Figure 6. Formation of activated slud-

ge floc. Adsorption - desorption Microbial activity

present as detached sewer biofilm components. Therefore,

the aggregation mechanisms known from pure culture,
including specific adhesion and aggregation of a number of
single cells (39), rarely occur in activated sludge systems.
Biological activity within the floc encompasses cell
growth, exopolymer production, hydrolysis, death, and

Turbidity
lysis (Fig. 6), as is also described for biofilm systems (40).
Cell growth depends on the availability of soluble
substrates, which is controlled by their diffusion into the
floc, and by hydrolysis of the particles adsorbed from
the wastewater. Often most of the organic substrate
in wastewater is present as particles and colloids, Time
and hydrolysis is required before it is available for
Figure 7. Measurement of the floc strength of activated sludge
consumption. This is reflected by the high activity of flocs. The sludge is stirred under defined conditions for two to
extracellular enzymes in the floc matrix (19,41). Very three hours, and the increase in small particles in suspension
little is known about the controlling factors for general (measured as turbidity after centrifugation) reflects the strength,
exopolymer production within the flocs. From pure culture that is, a high turbidity reflects a weak floc (44).
studies it is known that a number of factors, such as
bacterial species, their growth stage, type, concentration
of substrate, and limitation of N, P, or micronutrients, defined conditions (44,45). The floc weakness or the shear
and so on (40), affect the production of exopolymers sensitivity can be quantified using the initial slope (45) or
(mainly proteins and polysaccharides). In activated sludge, the ratio between the developed turbidity and total mass
bacteria are generally starved and exposed to dynamic in the system after equilibrium (44). A more fundamental
conditions (1). These conditions seem to promote floc approach and more comprehensive modeling can be
formation, perhaps as a result of an increased exopolymer applied where the interaction energies and enthalphies
production compared to cell production. for flocculation-deflocculation of activated sludge particles
Flocs and other aggregates are characterized by high can be determined (44). The model by Parker describes the
cell densities, which induce inter- and intraspecific inter- production of small particles by floc-breakup (4).
actions between the cells. An example is the production Strong flocs are believed to be formed from strong
of low molecular weight, diffusible signaling molecules microcolonies, a large amount of exopolymers with a
(autoinducers). These compounds can be involved in the high charge density, many cations that effectively cross-
induction of various genes that are responsible for aggre- link the matrix, and the presence of a macrostructure
gation behavior, EPS-production, disaggregation, and so (filamentous bacteria or fibers). Changes in floc strength
on. Thus, cell-to-cell communication may be of fundamen- may be due to changes in the electrostatic interactions
tal importance to the dynamics of aggregation in flocs and among exopolymers and particles in the floc, which can
biofilms and needs more attention in the future (42,43). be induced by changes in environmental parameters such
as ionic composition or pH, or to microbial activity, either
Stability of Activated Sludge Flocs. If activated sludge directly or indirectly (Table 2).
in a specific treatment plant is exposed to changes Changes typically observed are a decreased floc
in certain environmental parameters, changes in the strength when pH increases and when ionic strength is
stability of the flocs also take place. A change typically lowered, both a result of increased electrostatic repulsion.
causes a weakening of the floc and thus an undesired Changes can also be induced indirectly by microbial
deflocculation, in particular if it is exposed to shear forces. activity, for example, by denitrifying activity (which
This is illustrated in Figure 7, where the floc strength increases the local pH) or by the activity of Fe(III)-reducing
is characterized by the development in suspension bacteria, in which Fe(III) is reduced to a poorer flocculant,
turbidity with time when sludge is stirred under Fe(II) (46,47). The presence of sulfide, either produced by
 ACTIVATED SLUDGE — THE FLOC 59

Table 2. Examples of Changes in Floc Strength Due about the composition of the microbial community, (2) lack
to Short-Term Changes in Environmental Factors or of reproducible and comprehensive measurements of floc
Microbial Activity and sludge properties, and (3) difficulties in comparing
Increased Deflocculation the large diversity of wastewater types, process designs,
(decreased floc strength) and plant operations. Some of the generally accepted
correlations are presented below.

Environmental Condition
pH Increased pH (>8–9) Flocculation and Clarification. If the floc formation is
Ionic concentration Change from high to low concentration working well in the process tanks and in the inlet to
Ionic composition Reduced ratio: (trivalent + divalent)/ the clarifier, relatively large flocs are formed with an
monovalent ions irregular shape that can collect smaller flocs during the
Detergents Increased concentrations (>10–20 mg/l) initial flocculation and formation of floc-aggregates. The
flocs must be strong to prevent breakup and production of
Microbial Activity
Lack of oxygen Lack of EPS production? many small particles. A number of factors are known to be
Denitrification Increased pH (>8–9) important for flocculation (4). Among the physical factors
Fe(III)-reduction Removal of the floc-forming Fe(III) are the level of turbulence (shear rate) and the design of
Sulfide production Precipitation of Fe(III) to black FeS the tanks. Important biological parameters include the
F/M ratio (food to microorganisms, kgCOD/kgSS day)
or sludge age (defined as total sludge mass divided by
daily sludge production). At high F/M ratio or low sludge
sulfate-reducing bacteria or present in the wastewater,
age (5–10 days), production of large but weak flocs has
can seriously damage the flocs because of precipitation of
been reported, whereas smaller and stronger flocs are
the Fe(III) as ferrous sulfide (48). Furthermore, the active
produced at lower organic loading and longer sludge
aerobic metabolism of the bacteria is essential to maintain
ages (20–30 days). However, conflicting observations are
strong aggregates on a short-term basis so that any lack
reported in the literature, undoubtedly because other
of oxygen may induce some deflocculation of activated
factors are important as well. For instance, large amounts
sludge flocs (49). The floc strength has been shown to have
of humic substances in the incoming wastewater or large
a seasonal variation in a certain treatment plant, with the
variations in salinity can be expected to affect the floc
strongest floc present during the summer period (50). The
composition and the floc properties.
reason for this is yet to be ascertained.
Settling. A high settling velocity is observed if the flocs
Influence of Floc Structure on Functional Sludge Properties
are spherical and large with a relatively high density and
The important functional sludge properties in the without too many filamentous bacteria extruding from
activated sludge process are flocculation and clarification, the flocs (1,3). A high density is obtained if the inorganic
settling, and dewatering. Possible parameters that are fraction is high and if the porosity of the flocs is low. It is
believed to lead to floc formation with certain sludge also reported that often flocs from plants with biological
properties are shown in Figure 8. Different environmental removal of phosphorus settle well, despite the presence of
factors determine the microbial population as well a high number of filamentous bacteria, probably because
as other components present in the floc. All these of high-density microcolonies of phosphorus-accumulating
components, with their inherent properties, build up the bacteria or precipitates. Low floc density often appears
flocs, determining size distribution, shapes, density, and due to formation of conglomerates of flocs and filaments
strength. Ultimately, these floc properties determine the or due to extensive amounts of water-binding EPS, either
functional sludge properties that largely determine the adsorbed from wastewater or produced by certain bacteria
operation of the plant. (e.g., Zoogloea spp., (51)). Lack of inorganic cations has
Today, only little is known about the cause-effect been reported to cause deflocculation and production of
relationship between the external factors that determine flocs with low density (52).
the various floc properties, whereas slightly more is known
about which floc properties determine certain sludge Dewatering. Good sludge dewatering properties are
properties. This mainly is due to (1) lack of knowledge defined by low consumption of conditioning agents (organic

Microorganisms Size
Wastewater and
treatment plant Extracellular polymeric Flocculation
Shape
- Substrates Substances (EPS) Settling
- pH Organic debris Density Dewatering
- Temperature Inorganic particles
- Ionic composition Ions Strength
- ...
Figure 8. Overview of factors leading to floc
Environmental Floc composition Floc Sludge formation with certain functional sludge prop-
factors and structure properties properties erties.
60 ACTIVATED SLUDGE — THE FLOC

polymers or iron/lime), a high dewatering rate, and a (2) floc structure and floc properties, and (3) floc proper-
high dry matter content in the dewatered sludge cake (5). ties and sludge properties. Such knowledge can be used
Conditioning agents are used to flocculate small particles, for troubleshooting and optimization of the solid-liquid
often by neutralizing the negative surface charges of separation in existing treatment plants and to facilitate
the particles. Therefore, it is important to minimize the the development of new concepts, including active control
number of small particles (and thus the surface area) and, of important floc properties.
if possible, only have large flocs with a narrow floc size
distribution. It is also important that the flocs are strong BIBLIOGRAPHY
and do not break apart during dewatering, when a high
shear is applied. This comment also applies to the rate of 1. J. Wanner, Activated Sludge Bulking and Foaming Control,
the dewatering, which is affected by many small particles Technomic Publishing Company, Inc., 1994.
because of filter clogging. Examples of poor dewatering are 2. N. Y. Osée Muyima, M. N. B. Momba, and T. E. Cloete,
often observed when the floc size distribution is broad with Biological Methods for Treatment for Wastewaters, in
many small particles resulting from low floc strength (53). T. E. Cloete and N. Y. O. Muyma, eds., Microbial Commu-
A high dry matter content requires sludge with a low nity Analysis, IAWQ Scientific and Technical Report No. 5,
water-binding capacity, that is, a relatively low content 1997, pp. 1–24.
of highly charged EPS (54). The presence of an extensive 3. D. Jenkins, M. G. Richard, and G. T. Daigger, Manual on
amount of water-binding EPS in zoogloeal bacteria can Causes and Control of Activated Sludge Bulking and
deteriorate the dewatering (31). Foaming, Lewis Publisher, Boca Raton, Fla., 1993.
4. G. A. Ekama et al., Secondary settling tanks: Theory, Model-
The Ideal Floc. It is not simple to define the ‘‘ideal’’ ing, Design and Operation, IAWQ Scientific and Technical
floc, one that is optimum for clarification, settling, Report No. 6, 1997.
and dewatering. For instance, strong flocs are good 5. L. Spinosa and P. A. Vesilind, Sludge into Biosolids, Process-
for flocculation and clarification but not necessarily ing, Disposal, Utilization ISBN 1-900222-08-6, 2001.
for settling and dewatering. Regarding dewatering, 6. B. Frølund, R. Palmgren, K. Keiding, and P. H. Nielsen,
the high water-binding capacity of most strong flocs Water Res. 30, 1749–1758 (1996).
(caused by many charged EPS components and cations) 7. V. Urbain, J. C. Block, and J. Manem, Water Res. 5, 829–838
reduces the obtainable sludge dry matter content during (1993).
dewatering (54). Furthermore, this high water-binding 8. F. Jorand et al., Water Res. 29, 1639–1647 (1995).
capacity may reduce the density of the flocs and thus, 9. D. H. Li and J. J. Ganczarczyk, CRC Crit. Rev. Environ.
potentially, the settling velocity. The best compromise Control 17, 53–87 (1986).
for floc properties that ensure good sludge properties 10. D. H. Li and J. J. Ganczarczyk, Res. J. Water Pollut. Control
for the entire solid-liquid separation process is probably Fed. 63, 806–814 (1991).
for it to be relatively large, irregularly shaped with 11. J. J. Beun, M. C. M. van Loosdrecht, and P. A. Wilderer,
some filaments present, not too strong, and high density. Water Res. 33, 2283–2290 (1999).
However, it is obvious that an ideal floc with respect to 12. D. H. Li and J. J. Ganczarczyk, Water Res. 22, 789–792
solid-liquid separation is different from the ideal structure (1988).
for biological transformations because for the latter, small 13. F. Zartarian et al., Water Sci. Tech. 30(11), 243–250 (1994).
flocs with low diffusional resistance would be desirable. 14. T. D. Waite, J. Guan, and R. Amal, in H. H. Hahn et al.,
eds., Chemical Water and Wastewater Treatment V, Springer
Publishing, 1998, pp. 269–283.
CONCLUSION 15. D. H. Eikelboom and H. J. J. van Buijsen, Microscopic Sludge
Investigation Manual, TNO research institute for environ-
The activated sludge floc is a heterogeneous, complex mental hygiene, 1983.
structure consisting of a multitude of bacteria, extra- 16. M. Henze, P. Harremoes, J. la Cour Jansen, and E. Arvin,
cellular polymeric substances, and other organic and Wastewater Treatment–Biological and Chemical Processes,
inorganic components. Floc formation is determined partly 2nd ed., Springer-Verlag,1995, pp. 383.
by biological activity (growth-related processes, lysis, and 17. P. H. Nielsen and A. Jahn, in T. R. Neu, H. -C. Flemming,
hydrolysis) and partly by physicochemical, adsorption- and J. Wingender, eds., Microbial Extracellular Polymeric
desorption processes. Little is presently known about the Substances, Springer-Verlag, Berlin, 1999, pp. 49–72.
identity of the dominating bacteria involved in these pro- 18. K. Keiding and P. H. Nielsen, Water Res. 31, 1665–1672
cesses and about what controls their activity, whereas (1997).
more is known about the dominating physicochemical 19. B. Frølund, T. Griebe, and P. H. Nielsen, Appl. Microbiol.
factors. The most significant sludge properties for a well- Biotechnol. 43, 755–761 (1975).
working treatment plant (flocculation, settling, and dewat- 20. S. Wuertz et al., Water Sci. Technol. 37(4–5), 379–384 (1998).
erability) are determined by the floc properties (mainly floc 21. M. J. Higgins and J. T. Novak, Journ. Environ. Eng. 123,
size, shape, density, water-binding capacity, and strength). 479–485 (1997).
Future investigations of both microbiological and physico- 22. L. H. Mikkelsen et al., Water Sci. Tech. 34(3–4), 449–457
chemical aspects of the flocs, for example, by using novel (1996).
in situ methods, will improve our knowledge about cause- 23. T. R. Neu and J. R. Lawrence, in J. Wingender, T. R. Neu,
effect relationships between (1) wastewater characteris- and H. -C. Flemming, eds., Microbial Extracellular Polymeric
tics, plant operation, and floc composition or structure, Substances, Springer-Verlag, Berlin, 1999, pp. 21–48.
 ACTIVATED SLUDGE — THE ‘‘G-BACTERIA’’ 61

24. D. De Beer, J. C. van den Heuvel, and P. S. Stewart, Appl. ACTIVATED SLUDGE — THE ‘‘G-BACTERIA’’
Environ. Microbiol. 59, 573–579 (1993).
25. R. Amann, W. Ludwig, and K. H. Schleifer, FEMS Microbiol. ROBERT J. SEVIOUR
Lett. 59, 143–169 (1995). La Trobe University
26. M. Wagner and R. Amann, in T. E. Cloete and N. Y. O. Bendigo, Australia
Muyma, eds., Microbial Community Analysis, IAWQ Scien-
tific and Technical Report No. 5, 1997, pp. 61–72.
Many enhanced biological phosphorus removal (EBPR)–
27. K. Andreasen and P. H. Nielsen, Appl. Env. Microbiol. 63, activated sludge systems (ACTIVATED SLUDGE — MOLECULAR
3662–3668 (1997).
TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION;
28. N. Lee et al., Appl. Environ. Microbiol. 65, 1289–1297 (1999). ACTIVATED SLUDGE — THE PROCESS), where the biomass is
29. R. J. Seviour and L. L. Blackall, The Microbiology of Activated recycled alternatively through anaerobic and aerobic
Sludge, Kluwer Academic Publisher, 1999. regimes, contain large numbers of bacterial cells with a
30. R. Rossello-Mora et al., Water Environ. Res. 72, 56–64 (2001). distinctive morphology. In the absence of any detailed
31. C. A. Lajoie et al., Water Environ. Res. 72, 56–4 (2001). physiological or biochemical information about them,
32. M. Wagner et al., Syst. Appl. Microbiol. 18, 251–264 (1995). they are defined here as a morphotype, appearing as
33. W. Manz, M. Eisenbrecher, T. R. Neu, and U. Szewzyk, cocci in tetrads or clusters, often of different sizes
FEMS Microbiol. Ecol. 25, 43–61 (1998). and usually closely associated with the flocs (ACTIVATED
34. J. L. Nielsen and P. H. Nielsen, Water Res. (2001, in press). SLUDGE — THE FLOC), as shown in Figure 1. They are
35. P. H. Nielsen, B. Frølund, S. Spring, and F. Caccavo Jr., also seen in large numbers in sequencing batch reactors
System. Appl. Microbiol. 20, 645–651 (1997). (ACTIVATED SLUDGE — SEQUENCING BATCH REACTORS) operating
36. J. Gregory, CRC Critical Rev. Environ. Control. 19, 185–230 under aerobic–anaerobic conditions. An early report (1)
(1989). suggested that they occurred more commonly in systems-
37. M. Hermansson, Coll. Surf. B: Biosurfaces 14, 105–119 fed glucose and not acetate, and so they were referred to
(1999). as G-Bacteria (glucose-bacteria). Bacteria with the same
38. A. Zita and M. Hermansson, Appl. Environ. Microbiol. 63, appearance were reported earlier by Takii (2), and have
1168–1170 (1997). now been recorded in plants all over the world (3).
39. P. Bossier and W. Verstrate, Appl. Microbiol. Biotechnol. 45, Unfortunately, most of these have probably not been
1–6 (1996). obtained in pure culture as they are very difficult
40. P. H. Nielsen, A. Jahn, and R. Palmgren, Water Sci. Tech. to grow (3), although now enough have been cultured
36(1), 11–19 (1997). to demonstrate that the term G-Bacteria describes a
41. M. Hoffman and A. W. Decho, in J. Wingender, T. R. Neu, group of phylogenetically quite different organisms, all
and H. C. Flemming, eds., Microbial Extracellular Polymeric sharing the same microscopic morphological appearance.
Substances, Springer-Verlag, Berlin, 1999, pp. 217–230. Because their taxonomy is so different, it is likely
42. D. G. Davies, in J. Wingender, T. R. Neu, and H. C. Flem- that their physiologies and ecologies also vary. Hence
ming, eds., Microbial Extracellular Polymeric Substances, their significance in activated sludge systems will
Springer-Verlag, Berlin, 1999, pp. 93–118. probably differ and interpretations of their impact on
43. A. W. Decho, in J. Wingender, T. R. Neu, and H. -C. Flem- plant performance need to be made with caution.
ming, eds., Microbial Extracellular Polymeric Substances, This article briefly discusses the systematics of these
Springer-Verlag, Berlin, 1999, pp. 155–169. ‘‘G-Bacteria,’’ what is known and would like to be
44. L. H. Mikkelsen and K. Keiding, Adv. Coll. Interface. Sci. 80, known about their biology, and how such information
151–182 (1999). might help in understanding activated sludge systems
45. L. Eriksson, I. Steen, and M. Tendaj, Water Sci. Tech. 25, better.
251–265 (1992).
46. F. Caccavo Jr. et al., Appl. Env. Microbiol. 62, 1487–1490
TAXONOMY OF THE ‘‘G-BACTERIA’’
(1996).
47. P. H. Nielsen, Water Sci. Tech. 34(5–6), 129–136 (1996).
Taxonomy of the Gram-Negative ‘‘G-Bacteria’’
48. P. H. Nielsen and K. Keiding, Water Res. 32, 313–320 (1998).
49. B. M. Wilén, J. L. Nielsen, K. Keiding, and P. H. Nielsen, It is now clear that the term G-Bacteria embraces both
Coll. Surf. Biointerfaces 18, 145–156 (2000). gram-negative and gram-positive bacteria, all capable of
50. B. M. Wilén, K. Keiding, and P. H. Nielsen, Water Res. 34, growing in this distinctive manner. The gram-negative
3933–3942 (2000). ‘‘G-Bacteria’’ named by (1) were initially wrongly ‘‘identi-
51. L. Novák, L. Larrea, J. Wanner, and J. L. Garcia-Heras, fied’’ on their morphology only as members of the genus
Water Res. 27, 1339–1346 (1993). Methanosarcina, an Archaea. It is now known from 16S
52. M. Higgins and J. T. Novak, Water Env. Res. 69, 225–232 rRNA gene sequence data obtained from pure cultures,
(1997). most of which possess the same distinctive morphology
53. P. R. Karr and T. M. Keinath, J. Water Poll. Cont. Fed. 50, as seen in plant biomass, that these belong to a novel
1911–1928 (1978). genus, Amaricoccus in the α-Proteobacteria (4). Some (e.g.,
54. L. H. Mikkelsen, A Colloidal Chemical Approach to the Amaricoccus kaplicensis) often appear pleomorphic and
Floc Strength Concept–With Dewatering Implications, Ph.D. produce large amounts of extracellular capsular material
Thesis, Environmental Engineering Laboratory, Aalborg (Fig. 2a), which on drying appears to join the cells together
University, Aalborg, Denmark, 1999. in a microfibrillar network (Fig. 2b). Other Amaricoccus
62 ACTIVATED SLUDGE — THE ‘‘G-BACTERIA’’

(a) (b)

(c)

Figure 1. (a) Light microscope view of biomass showing ‘‘G-Bacteria’’ as distinctive cocci in tetrads
and clusters associated with the floc. (b) ‘‘G-Bacteria’’ showing distinctive staining reaction with
Neisser stain for polyphosphate, where cell contents do not stain but cell walls do. (c) Size variation
in ‘‘G-Bacteria’’ from same biomass sample, where small and large cocci are visible (arrowed).

isolates are more regular in appearance in axenic cul- Once the 16S rRNA sequence data are available, it is
ture (Fig. 2c). It also seems that isolates from different then possible to design oligonucleotide probes that are spe-
parts of the world represent several different species cific for organisms at different taxonomic levels, that is,
of Amaricoccus, and these have been named after their genus, species, and so on for fluorescent in situ hybridiza-
places of isolation, that is, A. kaplicensis, A. veronensis, tion or FISH (ACTIVATED SLUDGE — MOLECULAR TECHNIQUES
A. tamworthensis, and A. macauensis. Some of the same FOR DETERMINING COMMUNITY COMPOSITION). A genus-specific
isolates characterized by (4), that is, A. kaplicensis were probe for Amaricoccus has been designed (AMAR 839) and
also sequenced in a parallel study by (5) who named used to survey the occurrence of this organism in sev-
them Tetracoccus cechii. Under the rules of the Code eral plants (Fig. 3). There was no clear evidence from this
of Bacterial Nomenclature, the genus name Amaric- survey that it was seen more commonly in EBPR plants
occus has precedence, although the name Tetracoccus with poor P-removal performance, although high numbers
still appears in the literature. Amaricoccus kaplicensis were detected, but not exclusively, in aerobic–anaerobic
and A. tamworthensis have been isolated from plants in SBR systems and plants treating carbohydrate-rich wastes
countries other than where they were initially grown (6). These bacteria were also seen in many conventional
(Maszenan, 2000). systems, and in some plant biomass samples, the probe
 ACTIVATED SLUDGE — THE ‘‘G-BACTERIA’’ 63

(a) (b)

(c)

Figure 2. Light microscope of pure cultures of Amaricoccus, the gram-negative ‘‘G-Bacteria’’

(a) isolated originally by (1), showing the cells associated with large amounts of extracellular
capsular material. (b) Dried capsular material in gram stained pure cultures of the isolate from
Tamworth, NSW appearing as an interconnecting microfibrillar matrix between cells. (c) The
strain from Macau showing no capsular material associated with the tetrads.

AMAR 839 was able to confirm that the ‘‘G-Bacteria’’ been cultured and so whether they do accumulate PolyP
visible under the microscope but never successfully recov- or another storage polymer is not known. Two other gram-
ered into pure culture, were in fact Amaricoccus spp. (see negative ‘‘G-Bacteria’’ have been grown in an laboratory,
following section). which after 16S rRNA gene sequence analysis appear as
It is now clear that Amaricoccus is not the novel genera. One, an unusual α-Proteobacteria has been
only gram-negative coccus associated with activated placed in the genus Defluvicoccus (8), whereas the other,
sludge systems. Thus, molecular techniques using strain Ben 117, a β-Proteobacteria, has been allocated to
16S rDNA fragments separated by DGGE (ACTIVATED the new genus Cardococcus (9). These data clearly demon-
SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY strate that these gram-negative coccal ‘‘G-Bacteria’’ are
phylogenetically diverse (Fig. 4), and it is highly probable
COMPOSITION) detected bacterial sequences in a labora-
that many more await description.
tory reactor showing poor P-removal, which emerged on
sequencing as phylogenetically novel α-Proteobacteria and
Taxonomy of the Gram-Positive ‘‘G-Bacteria’’
β-Proteobacteria (7). After FISH probing based on only
their partial sequences, they appeared in biomass sam- It is now clear that the gram-positive coccal ‘‘G-Bacteria’’
ples as large cocci (i.e., ‘‘G-Bacteria’’). Neither of these has are also taxonomically diverse, but until more have been
64 ACTIVATED SLUDGE — THE ‘‘G-BACTERIA’’

• Micropruina glycogenica (20) closely related to Fried-

maniella and M. phosphovorus and obtained from an
EBPR system.

The phylogenetic relationships based on 16S rRNA

sequence analysis between these and the other gram-
positive ‘‘G-Bacteria’’ are shown in Figure 5. The questions
of how widely occurring these gram-positive ‘‘G-Bacteria’’
are and what operating conditions might favor their pres-
ence in activated sludge systems await the development
of 16S rRNA-targeted probes for FISH of each of these.
Probes have been described for Microlunatus (21) and Tes-
saracoccus (Maszenan, unpublished) so their population
dynamics should now become clearer. Clone library data
Figure 3. In situ identification of Amaricoccus by FISH using have shown that organisms very similar to Tetrasphaera
the fluorescein labeled AMAR 839 probe, showing the distinctive occur elsewhere (22), but surprisingly no reports of the
cell arrangement of tetrads and clusters in a sample of biomass.
presence of any of the other ‘‘G-Bacteria’’ in activated
See color insert.
sludge biomass have appeared when molecular techniques
grown in pure culture and fully characterized, the true of community analyses have been used.
level of their biodiversity remains unknown. Those that
have been grown axenically in the laboratory, and for
which 16S rRNA sequence data (Fig. 5) have been obtained PHYSIOLOGY OF ‘‘G-BACTERIA’’
include
Evidence (largely anecdotal) suggests that these
• possibly novel Micrococcus strains (10), ‘‘G-Bacteria’’ are favored in activated sludge and SBR
• two new species of Friedmanniella, F. spumicola, and systems operating with alternative anaerobic–aerobic
F. capsulata (11), which grow in clusters of tetrads in regimes. Thus they possibly share the same ecological
a mucilaginous mass (Fig. 6), niches as some of the so-called polyphosphate accu-
mulating bacteria (PAB) (ACTIVATED SLUDGE — MOLECULAR
• two novel genera, Tessaracoccus (as T. bendigoensis)
TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION), but
(12) and
there is insufficient information on their physiology so far
• Tetrasphaera (as T. australiensis and T. japonica) to really explain why. The physiology and ecology of the
(13). few pure cultures available have not been examined in
The taxonomies of Friedmanniella and Tetrasphaera detail, and so the information comes largely from using
are particularly interesting. The only other Fried- mixed liquor communities enriched with these organ-
manniella species currently known, the slow growing isms. The taxonomic status of the bacteria present in
F. antarctica was isolated, not from activated sludge, these is usually not known, and given the earlier discus-
but from antarctic sandstone, a nutritionally stressed sions on the phylogeny of these ‘‘G-Bacteria,’’ they may
habitat (14). Tessaracoccus japonica is the coccus iso- in fact represent additional novel organisms. The reader
lated from activated sludge after prolonged starvation is directed to (ACTIVATED SLUDGE — MOLECULAR TECHNIQUES
by (15) and incorrectly ‘‘identified’’ then as a Micrococ- FOR DETERMINING COMMUNITY COMPOSITION) for a critical dis-

cus. Totally unexpectedly, it also emerges as a member cussion on current understanding of how these EBPR
of the same genus as the bulking activated sludge fila- plants are thought to work. The roles of polyphosphate
mentous bacterium ‘‘Nostocoida limicola’’ II (FILAMENTOUS (PolyP), poly b-hydroxybutyrate (PHB) and glycogen in
BACTERIA IN ACTIVATED SLUDGE: CURRENT TAXONOMIC STATUS aerobic–anaerobic systems in providing the PAB with a
AND ECOLOGY), which is known to grow as cocci in clus- selective advantage in EBPR, and the crucial importance
ters (16), even though Tetrasphaera has never been seen of anaerobic substrate uptake by them are also discussed.
to grow as filaments in pure culture. Such an outcome It has been suggested that the ‘‘G-Bacteria’ are
serves to emphasize the dangers inherent in relying solely undesirable in EBPR systems because of their ability
on morphology to ‘‘identify’’ bacteria (17), especially those to synthesize intracellular storage carbohydrates such
in activated sludge. as glycogen aerobically and to use these to provide
These strains are in addition to those gram-positive the energy to assimilate readily utilizable substrates
cocci so far isolated only from Japanese activated sludge anaerobically (1). Thus, the PAB were considered in
plants and not described originally as ‘‘G-Bacteria.’’ These danger of being outcompeted in the anaerobic stage, and
are also in the high mol% G + C gram-positive bacteria instead of PolyP accumulation occurring aerobically, cells
and include the novel genera store polysaccharide material instead.
This attractive idea has never received any direct sup-
• Microlunatus (as M. phosphovorus) (18) with a dis- port from pure culture work (23), and although none
tinctive cell morphology of the Amaricoccus isolates or Defluvicoccus that were
• Microsphaera (as M. multipartita) obtained from a examined store PolyP in pure culture (4), Cardococ-
plant treating a sugar-rich synthetic waste (19). cus does. Whether these non-PolyP accumulators can
 10%

Ruegeria algicola, (ATCC51442)

98 Roseobacter denitricans, (strain OCh 114)
Octadecabacter arcticus, (strain 238)
Antarctobacter heliothermus, (DSM 11440)
Rhodobacter veldkampii, (ATCC33303)
Paracoccus denitrificans
100 Rhodovulum robiginosum, (strain N2)

100
Amaricoccus group

100 Caulobacter variabilis, (ATCC15255)

Phenylobacterium immobile, (DSM1986)
100
95 Erythrobacter longus, (strain OCh 101)
α-Proteobacteria
Sphingomonas capsulata
100 Sandaracinobacter sibiricus, (strain RB 16-17)
99 Agrobacterium vitis, (strain K309)
Rhizobium leguminosarum, (strain RCR 3644)
Rhodomicrobium vannielii, (ATCC 51194)
Hyphomicrobium zavarzinii
Methylosinus trichosporium, (strain OB3b)
97
Blastobacter denitrificans
Methylobacterium rhodinum
Azospirillum halopraeferens, (DSM3675)
Magnetospirillum gryphiswaldense, (DSM6361)
Rhodospirillum sodomense, (ATCC51195)
Defluvicoccus vanus, (strain Ben 114T)
100 Roseococcus thiosulfatophilus, (DSM8511)
Acidiphilium multivorum
Gluconobacter asaii
Rhodovibrio salinarium
Hydrogenophaga palleronii, (strain Stanier)
98
Aquabacterium parvum, (strain B6)
Ideonella dechloratans
100 100
Roseateles depolymerans, (strain 61B2)
100 Acidovorax facilis
95 Hydrogenophilus thermoluteolus, (strain TH-1)
96 Cardococcus australiensis (strain Ben 117T)
Dechlorimonas agitatus, (strain CKB)
Rhodocyclus purpureus, (DSM 168)
100
Azoarcus denitrificians, (strain Td-21)
Thauera aromatica (DSM 6984)
β-Proteobacteria Nitrosospira multiformis (ATCC25196)
Gallionella ferruginea, (Stock Johan)
Neisseria canis, (ATCC14687)
Chromobacterium violaceum
Sutterella wadsworthensis, (strain WAL9054)
Iodobacter fluviate
Burkholder mallei, (ATCC 23344)
Ralstonia solanacearum, (strain H1)
Oxalobacter formigenes, (ATCC 35274)
Janthinobacterium lividum, (DSM 1522)
Figure 4. Phylogenetic tree (6) based on 16S rDNA sequence analysis of the cultured
gram-negative ‘‘G-Bacteria’’ (in bold) among the α-Proteobacteria and β-Proteobacteria. Bootstrap
values are given for the major clades and the bar marker indicates 10 nucleotide substitutions per
1,000 nucleotides.

65
66 ACTIVATED SLUDGE — THE ‘‘G-BACTERIA’’

2%

100 Tetrasphaera australiensis (strain Ben 109)

99 Tetrasphaera australiensis (strain Ben 110)
100 Tetrasphaera australiensis (strain Ben 111)
97 Tetrasphaera japonica (strain T1-X7)
Intrasporangium calvum (IFO 129989)
Janibacter limosus (DSM 11140)
Terracoccus luteus (DSM 44267)
Terrabacter tumescens (ATCC 6947)
Dermatophilus congolensis (ATCC 14637)
Brevibacterium casei (DSM 20657)
100 Genus sanguibacter

Clavibacter michiganensis (DSM 7483)

100 Micrococcus luteus (strain Ben 112)
100 Micrococcus luteus (ATCC 381)
100 Micrococcus lylae (strain Ben 113)
Micrococcus lylae (DSM 20315)
Mycobacterium neoaurum (ATCC 25795)
Corynebacterium bovis (ATCC 7715)
Micropruina glycogenica (strain Lg2)
99 Friedmanniella antarctica (DSM 11053)
Friedmanniella spumicola (strain Ben 107)
Microlunatus phosphovorus (strain NM-1)
Friedmanniella capsulata (strain Ben 108)
100
Genus nocardioides
Propioniferax innocua (DSM 8251)
Luteococcus japonicus (IFO 12422)
Tessaracoccus bendigoensis (strain Ben 106)

Family propionibacteriaceae

Streptomyces bikiniensis (DSM 40581)

Sporichthya polymorpha (DSM 46113)
Figure 5. Phylogenetic tree (6) based on 16S rDNA sequence analysis of the cultured
gram-positive ‘‘G-Bacteria’’ (in bold) among the high G + C gram-positive bacteria. Bootstrap
values are given for the major clades and the bar mark represents 10 nucleotide substitutions per
1,000 nucleotides.

synthesize glycogen (either aerobically or anaerobically)

instead is known only for Amaricoccus, where glycogen in
high amounts (about 20% of cell dry weight) is stored
under aerobic conditions from glucose (Kong, unpub-
lished). PHB is also synthesized from acetate and sev-
eral other carbon sources (23). Most of the gram-positive
‘‘G-Bacteria’’ that have been observed can accumulate
PolyP in pure culture, although with the exception of
M. phosphovorus (18,24), nothing is known about which
conditions might influence its synthesis. This bacterium
was described as behaving in pure culture in a man-
ner close to that expected for the PAB, except that it
could not assimilate acetate anaerobically. In fact, NMR
studies (25) have shown that this organism fits little of
the expected behavior of a PAB. Unable to grow anaer-
obically, it could still ferment glucose to accumulate
anaerobically intracellular reserves of free acetate without Figure 6. Scanning electron micrograph of Friedmanniella
synthesis of any PHB. Under aerobic conditions, glu- spumicola showing the coccal tetrads embedded in mucilage
cose was respired without intracellular storage polymer (bar marker = 2 µm).
(e.g., glycogen) production. However, it could assimilate
and release phosphate at high rates in pure culture,
similar to those measured in EBPR activated sludge sam- ‘‘G-Bacteria’’ are likely to be rewarding in the efforts to
ples. Similar experimental protocols applied to the other understand how these organisms function.
 ACTIVATED SLUDGE — THE ‘‘G-BACTERIA’’ 67

MIXED CULTURE STUDIES WITH THE ‘‘G-BACTERIA’’ and not using them to increase what appears from
pure culture evidence to be very slow growth rates.
Much of what is thought to be known about how these It could be argued from the organisms’ perspective
‘‘G-Bacteria’’ behave in activated sludge comes from that this makes sound ecological sense because an ‘‘r-
studies using mixed culture communities of unknown strategy’’ based on a high growth rate in activated sludge
taxonomic status. Often the biomass in these experiments is certain to fail. Most of the ‘G-Bacteria’’ that were
where conditions for very poor phosphorus removal have observed appear to produce often large amounts in pure
been established, has been dominated by morphotypes cultures of polysaccharide capsular material. Probably as
of the ‘‘G-Bacteria,’’ although they were not always a consequence of this material imparting hydrophobicity
described as such, instead being referred to as the to their cell surfaces, the tetrads are usually seen closely
glycogen-accumulating bacteria, or GAO (25). As these associated with the activated sludge flocs. This may serve
bacteria have not yet been cultured, it is difficult to to
interpret the chemical transformations reported in terms
of the organisms actually carrying them out. However, • protect them from possible predation from the
these experiments did attempt to explain why the PAB protozoa feeding on suspended bacteria (PROTOZOA IN
seemingly were replaced by GAO in EBPR systems- ACTIVATED SLUDGE)
fed glucose and carbohydrate-rich substrates, and not • provide protection against potentially harmful abiotic
acetate (1). The main proposition is that the GAO under influences in the biomass
these conditions could assimilate glucose anaerobically • ensure that they stay in the system in the return
at a faster rate than the PAB could (who were thus activated sludge and not leave the plants suspended
disadvantaged), to synthesize PHB using glycogen as in the bulk liquid
the source of reducing power and energy. Aerobically,
glycogen was then replenished in the biomass from Their frequency of occurrence and predominance in
the PHB, and the PAB in the absence of PHB were biomass samples containing them together make them
no longer advantaged and able to accumulate PolyP. worthy of study.
An acetate feed favored the PAB according to the
models for EBPR (ACTIVATED SLUDGE — THE PROCESS). It
CONCLUSION
was suggested (26) that the energy demand for acetate
transport into the cells is high and cells need an energy
• The cocci referred to as the ‘‘G-Bacteria’’ and often
source such as PolyP to support it, whereas glucose
seen in activated sludge systems in large numbers are
transport has a comparatively low energy demand that
phylogenetically diverse and contain mainly novel
can be met from glycogen catabolism. However, no other
gram-positive and gram-negative bacteria
convincing evidence has been produced to support such
claims. Unfortunately, as we do not know yet which • This finding reinforces the dangers in relying on
organisms are the GAO in these studies, and FISH-based morphology to ‘‘identify’’ bacteria and the risks in
identifications would help us here, it can only be explained using this ‘‘identification’’ to interpret their possible
in simplistic terms what might be going on. However, role in communities such as activated sludge
anaerobic glucose assimilation and aerobic glycogen • Generally, the literature seems to agree that these
synthesis may be a physiological adaptation of many ‘‘G-Bacteria’’ appear to favor activated sludge sys-
of these ‘‘G-Bacteria,’’ although some biomass samples tems where the levels and durations of periods of
dominated by Amaricoccus in fact remove phosphate very nutrient stress are particularly high, that is, SBR
well (27). and EBPR processes
• These ‘‘G-Bacteria’’ appear to be found in niches
shared by the PAB, although it is known how widely
ECOLOGY OF ‘‘G-BACTERIA’’
distributed each ‘‘G-Bacterium’’ is or the influences
that promote or discourage their growth
Can the little that is known about these slow growing
bacteria help understand why and how they survive • Although these organisms are difficult to obtain in
in such highly competitive habitats as activated sludge pure culture, the molecular methods now available
systems, especially those removing phosphorus? They (e.g., FISH) provide the tools for studying their
seem to possess the physiology to tolerate the wildly ecology and population dynamics
fluctuating availability of nutrients and the inevitable • The general understanding of their physiology is poor
periods of nutrient stress and starvation encountered in and so their role in EBPR plants is still not certain. All
such systems. An ability to synthesize energy storage seem to possess an ability to synthesize intracellular
polymers such as PolyP, PHB and/or glycogen appears storage polymers but the conditions affecting their
to be a common attribute of these, and there is some production await clarification
evidence that they are able to withstand prolonged • However, all seem to possess many of the required
periods of starvation better than other activated sludge traits for a slow growing organism to survive in an
bacteria (15). It has also been shown (23) (albeit not activated sludge system that is able to cope with
for all these ‘‘G-Bacteria’’) that they can access any feast or famine conditions and remain in the system
substrates when they do become available, immobilizing indefinitely. Therefore, with some confidence they
them almost exclusively into these storage compounds might be described as true ‘‘K-strategists’’
68 ACTIVATED SLUDGE — THE MICROBIAL COMMUNITY

• It is likely that there are many more ‘‘G-Bacteria’’ in ACTIVATED SLUDGE — THE MICROBIAL
activated sludge systems and probably other stressful COMMUNITY
habitats (i.e., Friedmanniella antarctica) waiting to
be described, and so views on what these bacteria ROBERT J. SEVIOUR
might be doing will probably need to be revised. A. M. MASZENAN
Biotechnology Research Centre
La Trobe University
Acknowledgments Bendigo, Australia
I wish to thank all the people who provided data for this
article, especially A.M.(Mas) Maszenan whose Ph.D. work this The activated sludge process (see ACTIVATED SLUDGE — THE
contribution largely uses. PROCESS), the most popular process for treating both domes-
tic and industrial wastes has been used around the world
for almost a century, yet it is still very much a ‘‘black
BIBLIOGRAPHY box.’’ Some of the major chemical changes that are tak-
ing place can be readily measured (Fig. 1), but still,
1. J. S. Cech and P. Hartman, Water Res. 27(7), 1219–1225 little is known about the microbes that are responsible
(1993).
for these and how their activities are affected by vari-
2. S. Takii, Water Res. 11, 85–89 (1977). ables such as influent composition, process configuration,
3. R. J. Seviour et al., Environ. Microbiol. 2, 581–593. and process operation. The reasons for such ignorance
4. A. M. Maszenan et al., Int. J. Syst. Bacteriol. 47, 727–734 are probably many, but until quite recently the meth-
(1997). ods available to gather such information from systems
5. L. L. Blackall et al., Lett. Appl. Microbiol. 25, 63–69 (1997). designed and run largely by engineers were not avail-
6. A. M. Maszenan et al., J. Appl. Microbiol. 88, 826–835. able. This has now begun to change. The application of
7. A. T. Nielsen et al., Appl. Environ. Microbiol. 65, 1251–1258 molecular methods to the activated sludge ecosystem (see
(1999). ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING
8. A. M. Maszenan et al., Int. J. Syst. Evol. Microbiol. (in the COMMUNITY COMPOSITION) has not only revealed a previ-
press). ously unsuspected level of microbial diversity, but has
9. A. M. Maszenan, R. J. Seviour, and B. K. C Patel, (unpub- also questioned many earlier ideas as to which bac-
lished). teria were responsible for important processes such as
10. A. M. Maszenan, R. J. Seviour, and B. K. C. Patel (unpub- phosphorus (see ACTIVATED SLUDGE — MOLECULAR TECHNIQUES
lished). FOR DETERMINING COMMUNITY COMPOSITION) and nitrogen

11. A. M. Maszenan et al., Int. J. Syst. Bacteriol. 49, 1667–1680 (see ACTIVATED SLUDGE — MICROBIOLOGY OF NITROGEN REMOVAL)
(1999). removal. However, it could be argued that such important
12. A. M. Maszenan et al., Int. J. Syst. Bacteriol. 49, 459–468 advances in the understanding of these systems are yet
(1999). to help operators run them better or engineers to achieve
13. A. M. Maszenan et al., Int. J. Syst. Evol. Microbiol. (in the improved designs. This will take some time to happen,
press). but it is clear that it will not occur until the microbiol-
14. P. Schumann et al., Int. J. Syst. Bacteriol. 47, 278–283 (1997). ogy of activated sludge is far better understood than it
15. N. Kataoka et al., Appl. Microbiol. Biotechnol. 45, 771–777 is now.
(1996).
16. L. L. Blackall et al., Int J. Syst. Evol. Microbiol. 50, 703–709. WHY THIS IGNORANCE?
17. C. R. Woese, Microbiol. Rev. 51, 221–271 (1987).
18. K. Nakamura et al., Int. J. Syst. Bacteriol. 45, 17–22 (1995). Applying traditional or classical microbiological tech-
19. Y. Yoshimi, A. Hiraishi, and K. Nakamura, Int. J. Syst. niques to activated sludge is beset with difficulties (1,2).
Bacteriol. 46, 519–525 (1996). These are often a consequence of the biomass being orga-
20. T. Shintani et al., Int. J. Syst. Evol. Microbiol. 50, 201–207 nized into discrete complex aggregates called flocs (see
(2000). ACTIVATED SLUDGE — THE FLOC). Their chemical and phys-
21. M. Kawaharasaki, T. Kanagawa, H. Tanaka, and K. Naka- ical structures and how they are formed are still not
mura, Water Sci. Technol. 37, 481–484 (1998). well understood, but unless these flocs settle rapidly in
22. M. Christensson, L. L. Blackall and T. Welander, Appl. the clarifiers, the treatment process will ultimately fail.
Microbiol. Biotechnol. 49, 226–239 (1998). Activated sludge is also a highly complex ecosystem, so
23. A. Falvo et al., J. Appl. Microbiol. (in press). microbiologists with a preference for working with pure
24. M. M. Santos, P. C. Lemos, M. A. M. Reis and H. Santos, cultures have not been attracted readily to its study.
Appl. Environ. Microbiol. 65, 3920–3928 (1999). The impact of applying molecular methods to overcom-
25. T. Mino, M. C. M. van Loostrecht, and J. J. Heinen, Water ing some, if not all, of the problems associated with
Res. 32, 3193–3207 (1998). culture-dependent methods has been considerable. These
26. M. Maurer, W. Gujer, R. Hany, and S. Bachmann, Water Res. include
31, 907–914 (1997).
27. A. Carucci, M. Majone, R. Ramadori, and S. Rossetti, Water • Sampling. Floc structure and organization mean
Sci. Technol. 30, 237–246 (1994). that representative samples are difficult to obtain,
 ACTIVATED SLUDGE — THE MICROBIAL COMMUNITY 69

Degradation or catabolism

Organic Mineralised CO2 + H2O + SO42−

substrates products PO43− + NO3−
{C,H,O,N,P,S} Energy
+ (ATP)
Oxygen Reducing
power
+ Precursors
Biomass

Biomass

Biosynthesis or anabolism
Figure 1. Summary of main chemical transformations
Waste
thought to occur in the activated sludge process (2).

since the floc-associated microbes have to be sus- • Measuring what they do. One particularly difficult
pended first. Various physical and chemical methods question to answer is what role these microbes are
have been developed for this (2,3), but none are performing in the treatment process from the data
without problems, and all give serious underesti- using 16S rRNA targeted probes and FISH, it seems
mations of the bacterial numbers in quantitative likely that most of the bacteria present are metaboli-
studies (1) as well as the true level of bacterial cally active and not dead or moribund (10) as once
biodiversity present in this system. Furthermore, thought and thus probably contribute to the mea-
it is not surprising that considerable heterogene- surable chemical changes occurring. Only recently
ity can exist within individual flocs, where anoxic have attempts been made to find the specific micro-
zones with different active bacterial populations have bial populations that are responsible for them in
been detected in flocs in fully aerobic systems (4). the hope that clearer structure–function relation-
Individual cells with a distinctive morphology (e.g., ships emerge. FISH probing directed at locating
filamentous bacteria) can be recovered by microma- specific functional genes within particular popula-
nipulation (2), a technique with considerable advan- tions is one approach. For example, the nitrate
tages over other culture dependent methods, since reductase gene in denitrifying bacteria (11,12), the
the organism(s) of interest can first be ‘‘identified’’ sulfate reductase gene for dissimilatory sulfate
under the microscope and can be selectively and reducing bacteria (4), and the ammonia monooxy-
physically separated from the rest. Flow cytometry genase gene for the nitrifiers (13) have all been
used for this purpose. Respirometry via oxygen
allows for bacteria recovery but not for the growth
uptake rates (OUR) and substrate utilization and
of specific populations (5,6) that have been distin-
product formation rates for nitrifying and denitrify-
guished with fluorescently tagged (FISH) probes.
ing bacteria (see ACTIVATED SLUDGE — MICROBIOLOGY OF
(see ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR
NITROGEN REMOVAL) are popular general methods (2),
DETERMINING COMMUNITY COMPOSITION).
especially for modelling purposes (see ACTIVATED
• Culturing and quantifying isolates. Even the most SLUDGE — THE PROCESS). Dye reduction enzyme assays
nonselective microbiological media are unlikely to with CTC (14) and INT (15) for detecting respira-
support the growth of more than a small subsample tory active (i.e., viable) cells microscopically have
(<1%) of the bacteria present in activated sludge. been applied to activated sludge systems. How-
Various molecular methods, which themselves are ever, when results obtained with all these differ-
not without criticism (1,2,7,8), have all strikingly ent macrotechniques are compared, the data do not
exposed the biased nature of culture-dependent tech- always correlate well (2). Determining carbon utiliza-
niques and the confusion that the data obtained from tion patterns with the Biolog system to fingerprint
them have created (see ACTIVATED SLUDGE — MOLECULAR heterotrophic metabolic activity, especially for sys-
TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION). tems treating industrial wastes, has also provided
It is the view of the authors that there is still promising data, but these are not always easily inter-
an important role for pure cultures in any study preted (16).
of activated sludge microbiology, especially if phys- The fate of specific populations with metabolic
iological and kinetic data are sought from indi- activities of interest has been followed using
vidual populations (9). Unfortunately, without prior green fluorescent protein tagging (17), which seems
knowledge of the nutritional requirements or tax- to be a technique of promise for the elucida-
onomy of the bacteria sought (e.g., the bulk- tion and assessment of their contributions to
ing and foaming filamentous bacteria), it is not particular chemical events (18). Recent applica-
easy to design media rationally for their cul- tion (19,20) of microautoradiography (MAR) in com-
ture (2). bination with FISH probing to activated sludge
70 ACTIVATED SLUDGE — THE MICROBIAL COMMUNITY

(see ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR tools, and it is not hard to predict that their appli-
DETERMINING COMMUNITY COMPOSITION) has already pro- cation to activated sludge systems will continue to
vided invaluable information on the in situ phys- expand.
iology of bacterial populations of interest. A good
example of how valuable this technique can be is SO WHAT IS KNOWN SO FAR ABOUT THE MICROBIAL
seen in Microthrix parvicella (20). This bulking and COMMUNITIES IN ACTIVATED SLUDGE PLANTS?
foaming filament (FILAMENTOUS BULKING IN ACTIVATED
SLUDGE, CONTROL OF; ACTIVATED SLUDGE — FOAMING) It is quite clear from recent studies applying these
was shown to be selectively favored in aerobic : rRNA-based methods that perceptions of what were pre-
anaerobic systems such as enhanced biological viously considered as more important (physiologically and
nutrient removal (EBNR) plants (see ACTIVATED numerically) populations in activated sludge systems are
SLUDGE — THE PROCESS) because of its ability to largely wrong. Organisms that are revealed using culture-
assimilate hydrophobic substrates, such as oleate, dependent methods are a reflection more of these methods
anaerobically and anoxically as well as under aero- than the populations actually present. Several indepen-
bic conditions. Uptake of trioleate by M. parvicella dent studies have shown that while the γ -Proteobacteria
was in fact best under anaerobic regimes and dominate cultured gram-negative bacterial populations,
could be maintained under the starvation condi- those revealed by using culture-independent methods are
tions inevitably encountered in such systems (see mainly the β-Proteobacteria (24–27). The relative impor-
STORAGE POLYMERS: ROLE IN THE ECOLOGY OF ACTIVATED tance of the high mol% G+C Bacteria differs between
SLUDGE). As more bacterial populations are studied studies, which may reflect different operational condi-
in this way, our current understanding of acti- tions of the systems studied or variations in efficiencies of
vated sludge microbiology will change dramatically, the individual methods of DNA extraction used, which
and such information should allow plant function are usually not checked for possible biases (28). Such
to be related meaningfully to community struc- observations also emphasize the well-known risks asso-
ture. ciated with trying to assess data quantitatively from these
rRNA-based molecular approaches (7). The information
• Identifying the microbes present. The classical
available on which organisms are present in activated
approach with activated sludge has been to use phe-
sludge is inevitably skewed. More effort has been directed
notypic characters to identify those isolates that are at understanding those considered responsible for par-
obtainable in pure culture. For those not culturable ticular changes such as those involved in nitrogen and
(e.g., many of the filamentous bacteria and proto- phosphate removal (see ACTIVATED SLUDGE — MICROBIOLOGY
zoa), ‘‘identification,’’ by necessity, is based on their OF NITROGEN REMOVAL) or the bulking and foaming filamen-
diagnostic microscopic appearance in biomass sam- tous bacteria (see FILAMENTOUS BULKING IN ACTIVATED SLUDGE,
ples. Neither approach is appropriate if the aim is CONTROL OF; ACTIVATED SLUDGE — FOAMING). Even so, these
to understand which microbes are really present in populations are still not always well understood, and often
activated sludge. Even if they can be cultured, most the picture is much more complex now than once thought
microbes, as is the case of many filamentous bac- likely. A brief overview of some recent additions to the
teria (see FILAMENTOUS BACTERIA IN ACTIVATED SLUDGE: view of the microbial composition of the activated sludge
CURRENT TAXONOMIC STATUS AND ECOLOGY), are novel community follows. This community is likely to change
organisms (21,22). Therefore, conventional identifi- in response to many factors in ways that are not yet
cation methods are difficult, and morphology is a understood, and so any generalization made here is risky.
notoriously unreliable phylogenetic indicator with
bacteria. Again, molecular RNA-based methods have • Viruses. As might be expected, activated sludge
provided the means for the in situ identification of systems contain many viruses of animal and human
many bacterial groups in the absence of any need origin, but because of problems in their isolation and
to culture them, independently of their morphol- characterization, their diversity is poorly understood,
ogy, using FISH (see ACTIVATED SLUDGE — MOLECULAR and many more different types may be present
TECHNIQUES FOR DETERMINING COMMUNITY COMPOSITION). about which little is known. Molecular methods (e.g.,
The number of FISH probes available for detect- probes) may eventually provide this information.
ing an increasingly diverse range of bacteria in These viruses are known to include the HIV virus,
activated sludge continues to grow rapidly and which is unlikely to survive the process, and more
their advantages over both fluorescent and mono- robust viruses including the rotaviruses, many of
clonal antibody techniques, which have also been which will survive (29). Bacteriophages are readily
employed for the same purpose, are well recog- isolated. Although their possible roles in bacterial
nized (10). Again, FISH techniques are not above population density control have not been fully
criticism (7,23), but they have already supplied vital assessed, it seems that some may have broad host
information about the role of the bacterial pop- ranges making such a function likely.
ulations in processes such as nitrogen and phos- • Bacteria. Unquestionably, applying a battery of
phate removal (see ACTIVATED SLUDGE — MICROBIOLOGY molecular rRNA-based techniques has changed views
OF NITROGEN REMOVAL). In combination with MAR, on the true extent of biodiversity among activated
they appear to be even more powerful experimental sludge bacterial populations, and many populations
 ACTIVATED SLUDGE — THE MICROBIAL COMMUNITY 71

previously unknown or ignored, or considered to Molecular approaches have also forced us to reap-
represent unimportant members of this community, praise anaerobic respiration in activated sludge
now emerge as numerically significant. Often, the systems. For example, the ANAMMOX process
populations detected in large numbers have no known illustrates how much is still to be learned about
function, and most of these are yet to be cultured in the diversity of denitrifying bacteria (39), and
the laboratory and it is usually not known why they their ability to reduce nitrate under aerobic con-
predominate. Activated sludge systems will certainly ditions now seems to be a common one (12,40)
contain pathogenic bacteria of human fecal origin (see ACTIVATED SLUDGE — MICROBIOLOGY OF NITROGEN
such as Salmonella spp., and some will survive the REMOVAL). Which bacteria are the more active
process. These will need to be removed by disinfection denitrifiers is not clear although unexpected can-
later because it is important to recognize that these didates such as Hyphomicrobium may dominate
processes are designed primarily to remove nutrients in some systems (41). Bacteria that are capable
and not bacteria. of reducing Fe3+ during anaerobic respiration are
readily detected in activated sludge systems in
a. Chemoheterotrophs. The process relies on aerobi-
which their activity may lead to floc destabiliza-
cally respiring bacteria to degrade and mineralize
tion (42) and one isolated strain is closely related
organic compounds present in raw sewage that
to Geobacter sulfurreducens (43). Sulfate reduc-
are often complex and chemically diverse. Yet,
tion may also occur in anoxic microniches in flocs,
little is known about the main bacteria that are
even in fully aerated systems (4).
responsible or how the communities of these might
The microbes responsible for phosphate removal,
change with different operating conditions. Fin-
often appearing in large clusters, have proven elu-
gerprinting techniques such as denaturing gradi-
sive despite a great deal of directed effort, with
ent gel electrophoresis (DGGE) (30) and terminal
the early favorite candidate Acinetobacter increas-
restriction length polymorphism (T-RFLP) (31)
ingly dismissed from molecular evidence (44)
(see ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR
(see ACTIVATED SLUDGE — MOLECULAR TECHNIQUES FOR
DETERMINING COMMUNITY COMPOSITION), which allow
DETERMINING COMMUNITY COMPOSITION). The puzzle is
changes in community composition to be followed
still not satisfactorily resolved in the absence of
with time, may help provide answers. Substantial pure cultures that behave as the models require
progress has been made in properly identifying but the β-Proteobacteria seem to be more likely
some of the distinctive bacterial morphotypes participants (45) than yeast spores (46). Many dif-
readily seen under the microscope. These include ferent populations, each possibly achieving impor-
the bulking and foaming bacteria (see FILAMENTOUS tance under different operational conditions, may
BACTERIA IN ACTIVATED SLUDGE: CURRENT TAXONOMIC be involved in phosphate removal, which will only
STATUS AND ECOLOGY) and the ‘‘G-Bacteria’’ (see complicate the story.
ACTIVATED SLUDGE — THE ‘‘G-BACTERIA’’). Both con-
b. Chemoautotrophs. Although it was thought that
tain bacteria that were not described previously, the microbiology of nitrification in activated sludge
and emphasize how collections of bacteria indistin- is quite well understood, molecular data have
guishable morphologically can be phylogenetically shown this not to be the case, and doubts are now
quite unrelated to each other. Spiral bacteria, raised about a role in activated sludge for the gen-
often recognized in biomass samples because of era that was once considered to be the major bacte-
their characteristic corkscrew pattern of move- ria involved in NH3 oxidation (13,47). What FISH
ment, are not spirochaetes as they have often been has elegantly shown is that the Nitrosobacteria
referred (2). Instead, on the basis of molecular and the Nitrobacteria are closely and physically
characteristics, they were previously undescribed arranged in distinctive clusters in the biomass,
members of the Cytophaga-Flavobacterium divi- allowing syntrophism to occur readily between
sion (6), a group long known to exist in activated the NO2 -producing and NO2 -consuming organ-
sludge systems (2). Novel isolates readily detected isms (48) (see ACTIVATED SLUDGE — MICROBIOLOGY OF
among populations of Aeromonas, Acinetobac- NITROGEN REMOVAL). Nitrifying bacteria are not
ter, Acidovorax, Sphingomonas, and Hyphomicro- the only autotrophic bacteria present in acti-
bium (32–36) confirm activated sludge as a rel- vated sludge. Thus, Fe2+ oxidizers have been
atively unexploited natural source of bacteria. In detected (42) and some bulking filamentous bac-
addition, some bacteria that were never considered teria such as Thiothrix may obtain energy from
previously as important members of this commu- H2 S oxidation. However, whether these are strict
nity now appear to be present in significant num- autotrophs or mixotrophs that are unable to use
bers. These include the Planctomycetes (25,37,38) carbon dioxide as a carbon source is uncertain (2).
and oligotrophs such as Hyphomicrobium (36). • Fungi. The importance of fungi in activated sludge
What the reason or reasons for their presence is not well understood, although their role as
might be are not understood, but such data imply predators in consuming nematodes (49) deserves
that appreciation of the total spectrum of the further study. Most of the fungi that have been
metabolic changes occurring in activated sludge cultured appear to be the Deuteromycotina that is
must also be simplistic. commonly seen as part of the air microflora or in soil
72 ACTIVATED SLUDGE — THE MICROBIAL COMMUNITY

samples (50), but experiences with activated sludge

Substrate Activated
bacteria might suggest that many more that are nutrients sludge
currently nonculturable may be present. Sludge floc
• Protozoa. Because of their occurrence in such bacteria
large numbers, the protozoa have attracted much
Flagellate
interest, and it is quite clear that they play a Sewage protozoa Attached Attached
vital role as predators in reducing the levels of bacteria & crawling carnivorous
suspended particles (e.g., bacterial cells) in the bulk ciliate ciliate
liquid, thus reducing its turbidity (see PROTOZOA Protozoa Protozoa
IN ACTIVATED SLUDGE). Most studies have looked at
the ciliates, especially their ecology and value as
Free-swimming
indicators of plant performance. Less is known about Free-swimming
carnivorous
the roles of the flagellates and ameboid protozoa, ciliate protozoa
ciliate
and it may turn out that their importance has Protozoa
been underestimated so far. Current interest in Figure 2. Proposed food chain for the activated sludge commu-
pathogenic protozoa such as Giardia lamblia and nity (2).
Cryptosporidium parvum is high, and many sensitive
molecular and other methods are now available for
WHICH FACTORS DECIDE THE FATE OF THESE MICROBES
their detection and enumeration at low levels in water
IN ACTIVATED SLUDGE?
and wastewater samples. Activated sludge seems
to be more effective at removing cysts of Giardia, An activated sludge process in simple terms is a continuous
which could be detected in influents at high levels, culture system with biomass recycling (56). Therefore, the
than Cryptosporidium oocysts (51). How selective microbial community there is subjected to considerable
the predatory bacteriovorous protozoa are in their selective pressures. These will (at least partly) decide
feeding habits is uncertain, but an increasing body of whether they remain in the system, die, or leave in the
evidence suggests that feeding is not random. Thus, treated effluent. Some of the more influential selective
cell size (52), hydrophobicity (53), and floc-forming factors are assumed to be the following:
abilities of the bacteria have all been suggested to
directly determine their fate in activated sludge. Yet, • Growth rate of a population will decide how
some ciliated protozoa appear to prefer filamentous successfully an organism can compete for limiting
bacteria to unicells (54). resources, and the chemostat theory states that an
• Metazoa. Under the microscope, activated sludge organism with a higher µmax and a lower ks will out-
compete one in which the µmax is lower and the ks is
biomass samples from plants operating well contain
higher for a growth-limiting nutrient (57). Whether
nematodes, rotifers, and oligochaete worms (55). All
this happens in an activated sludge is doubtful.
probably act as predators contributing to the removal
Little is known about how fast different populations
of suspended bacterial cells, although how rapidly can grow in activated sludge, and with such a
they feed and the selectivity or otherwise of their diverse range of available substrates that are possibly
feeding are less well understood than with those of utilized preferentially by different populations, any
the protozoa. kinetic analysis of activated sludge becomes a
daunting task. Furthermore, the opportunity for
populations to interact both negatively and positively
is considerable. Such interactions will certainly
THE ACTIVATED SLUDGE FOOD CHAIN?
influence their fate, but what these are and what
effect they might have on population dynamics is still
Despite a poor understanding of how individual popula-
completely unknown.
tions might interact, models based on simple food chains
• Ability to tolerate the prevailing abiotic factors will
have been proposed, and these often fit experimental
affect an organism’s µ, and, possibly, its ability
data quite well. Most of these are simplified and based
to survive. These factors include pH, pO2 , redox
on predator–prey relationships. They recognize several
potential, and temperature, as well as toxic chemicals
trophic levels, with the activated sludge bacteria (both
including heavy metals (2). There is little information
freely suspended and associated with the flocs) represent- on how any of these might influence community
ing the base level. These bacteria are the main degraders composition and activity, and because of the physical
of the soluble and colloidal organic substrates present and heterogeneity of flocs, many microbes will experience
are then fed on by the flagellate and free-swimming, crawl- conditions quite different from those that can be
ing and sessile, ciliated protozoa, which then serve as prey measured in the plant biomass as a whole (2).
to the carnivorous ciliates (Fig. 2). The population dynam- However, temperature may affect the communities
ics of such a complex process is not easy to detail until of foaming bacteria (see ACTIVATED SLUDGE — FOAMING),
the understanding of the basic microbiology of activated and a low pH may favor fungi in bulking episodes (see
sludge is improved. FILAMENTOUS BULKING IN ACTIVATED SLUDGE, CONTROL OF).
 ACTIVATED SLUDGE — THE MICROBIAL COMMUNITY 73

The composition of the protozoan community may used in their studies. However, much is still to be learned.
also be affected by heavy metals (58). Obtaining this information will be slow and fragmentary
• Becoming part of the floc means that a population and will tax the experimental skills of the microbiologist.
will be recycled in the return activated sludge (RAS) The real challenge then is to translate this basic microbi-
and thus survive longer. This ability probably is ological information into improved engineering outcomes
more important than a high µ, since fast-growing in both plant design and operation.
freely suspended cells will still leave in the bulk
liquid. How cells might achieve this is simple BIBLIOGRAPHY
to detect microscopically. Many apparently slow-
growing bacteria (see ACTIVATED SLUDGE — THE ‘‘G- 1. R. J. Seviour and L. L. Blackall, Microbiology of Activated
BACTERIA’’) often produce extensive capsular material Sludge, Kluwer Academic Publishers, Dordrecht, The Nether-
enabling them to become integrated into the floc, lands, 1999.
and probably at the same time resisting predation by 2. M. Wagner and R. Amann, in T. E. Cloete and N. Y. O. Muy-
the floc-grazing protozoa (2). The sessile protozoa, ima, eds., Microbial Community Analysis, IAWQ, London,
also probably slow growers, attach to flocs with U.K., 1996, pp. 61–72.
holdfasts, and many of the filamentous bacteria that 3. E. B. Pike, Ecological Aspects of Used Water Treatment, vol. 1,
certainly grow slowly in the laboratory are part of the Academic Press, New York, 1975, pp. 1–64.
floc matrix. 4. A. Schramm et al., Appl. Environ. Microbiol. 65, 3690–3696
• Accumulate storage compounds such as glycogen, (1999).
polyphosphate, and poly β-hydroxyalkanoates to 5. G. Wallner et al., Appl. Environ. Microbiol. 63, 4223–4231
provide them with the biochemical equipment to (1997).
survive the inevitable periods of famine, which will 6. J. Snaidr et al., Environ. Microbiol. 1, 125–135 (1999).
be encountered in an activated sludge system. These 7. I. M. Head et al., Microb. Ecol. 35, 1–21 (1998).
are especially acute in anaerobic : aerobic systems 8. J. Theron and T. E. Cloete, Crit. Rev. Microbiol. 26, 37–57
such as EBNR plants in which only cells that are (2000).
able to assimilate substrates into these polymers in 9. M. C. Tomei et al., Water Sci. Technol. 39, 21–29 (1999).
the anaerobic stage can then grow in the subsequent 10. R. Amann et al., Microbiol. Rev. 59, 143–169 (1995).
aerobic stage in which available nutrients are no 11. S. Hallin and P. -E. Lindgren, Appl. Environ. Microbiol. 65,
longer present. Many slow-growing bacterial isolates 1652–1657 (1999).
possess this ability, and the evidence suggests that 12. C. McDevitt et al., FEMS Microbiol. Lett. 184, 113–118
cells provided with readily metabolizable substrates (2000).
use them very rapidly not to maintain an increased 13. S. Juretschko et al., Appl. Environ. Microbiol. 64, 3042–
µ but to synthesize one or more of these storage 3051 (1998).
polymers. 14. G. G. Rodriguez et al., Appl. Environ. Microbiol. 58, 1801–
1808 (1992).
CAN OUR COMMUNITIES BE MANIPULATED TO 15. C. Maurines-Carboneill et al., Water Res. 32, 1213–1221
IMPROVE PLANT PERFORMANCE? (1998).
16. L. Vittorio et al., Water Res. 30, 1077–1086 (1996).
The idea of adding populations of microbes with spe- 17. L. Eberl et al., FEMS Microbiol. Lett. 149, 77–83 (1997).
cific, desirable, metabolic, or other properties to activated 18. D. Errampalli et al., J. Microbiol. Methods 35, 187–199
sludge is not a new one (59). Yet, when this has been (1999).
attempted, the results have generally been unimpres- 19. N. Lee et al., Appl. Environ. Microbiol. 65, 1289–1297 (1999).
sive (60). It is not too surprising from an ecological
20. K. Andreasen and P. -H. Nielsen, Water Res. 34, 1559–1569
perspective that most of the niches in activated sludge (2000).
would already be occupied, and any introduced strains
21. L. L. Blackall et al., Syst. Appl. Microbiol. 17, 513–518
would struggle to overcome the prevailing selective factors. (1994).
However, some success has been achieved in laboratory
22. L. L. Blackall et al., Int. J. Syst. Evolut. Microbiol. 50,
studies with genetically engineered bacterial strains car- 703–709 (2000).
rying plasmids, which by conjugation may spread through
23. A. Moter and U. B. Göbel, J. Microbiol. Methods 41, 85–112
the activated sludge community (61). It may be that we (2000).
are trying to add the wrong organisms to the system, and
24. M. Wagner et al., Appl. Environ. Microbiol. 60, 792–800
once we understand better what populations do, the more
(1984).
success will be achieved with this aproach.
25. P. L. Bond et al., Appl. Environ. Microbiol. 61, 1910–1916
(1995).
CONCLUSION 26. P. Kämpfer et al., Microb Ecol. 32, 101–121 (1996).
27. J. Snaidr et al., Appl. Environ. Microbiol. 63, 2884–2896
Understanding the composition of the microbial commu- (1997).
nities of activated sludge systems has been revolutionized 28. Z. Yu and W. W. Mohn, Can. J. Microbiol. 45, 269–272 (1999).
by the increase in the interest shown in them by molecu- 29. W. J. Lodder et al., Appl. Environ. Microbiol. 65, 5624–5627
lar microbial ecologists and by the techniques they have (1999).
74 ACTIVATED SLUDGE — THE PROCESS

30. C. A. Eichner et al., Appl. Environ. Microbiol. 65, 102–109 ACTIVATED SLUDGE — THE PROCESS
(1999).
31. W. -T. Liu et al., Appl. Environ. Microbiol. 63, 4512–4522 K. C. LINDREA
(1997). ROBERT J. SEVIOUR
32. E. Nsabimana et al., Water Res. 34, 1696–1704 (1999). Biotechnology Research Centre
33. E. Carr et al., J. Appl. Microbiol. 90, 309–319 (2001). La Trobe University
Bendigo, Australia
34. R. Schultze et al., Syst. Appl. Microbiol. 22, 205–214 (1999).
35. A. Neef et al., J. Appl. Microbiol. Biotechnol. 23, 261–267
(1999).
Although recorded history and surviving archaeological
sites show that the Babylonians and Assyrians, and
36. A. C. Layton et al., Appl. Environ. Microbiol. 66, 1167–1174
later Romans, used quite sophisticated technology to
(2000).
collect and dispose off their wastes, this environmental
37. A. Neef et al., Microbiology (UK) 144, 3257–3266 (1998).
awareness was not adopted by later societies, until the
38. J. Liu et al., Int. J. Syst. Evolut. Microbiol. 51, 195–202 Industrial Revolution in Britain in the 19th century forced
(2001). political intervention (1–3). Poor quality, densely packed
39. M. Strous et al., Nature 400, 446–449 (1999). housing with no provision for proper sewage disposal led
40. L. Frette et al., FEMS Microbiol. Ecol. 24, 363–370 (1997). to overbearing odor problems, polluted river systems,
41. C. G. Gliesche and A. Fesefeldt, Syst. Appl. Microbiol. 21, and inevitably epidemics of fatal diseases caused by
315–320 (1998). pathogenic bacteria of faecal origin such as Vibrio cholerae
42. F. Caccavo et al., Appl. Environ. Microbiol. 62, 1487–1490 and Salmonella spp. dispersed by them. Eventually,
(1996). authorities were forced to take action and after several
43. P. -H. Nielsen et al., Syst. Appl. Microbiol. 21, 315–320 royal commissions, the Rivers Pollution Prevention Act
(1998). was brought into effect in 1876. In fact, this act was
44. P. L. Bond and G. N. Rees, in R. J. Seviour and L. L. Blackall,
fairly ineffectual, and achieved little except that it led
eds., Microbiology of Activated Sludge, Kluwer Academic
to another Royal Commission on Sewage Disposal in
Publishers, Dordrecht, The Netherlands, pp. 227–256. 1898, with the function of coordinating evaluation of
new treatment systems. Those that had been developed
45. G. Crocetti et al., Appl. Environ. Microbiol. 66, 1175–1182
following the disease epidemics in Britain such as
(2000).
biological trickling filters, land effluent percolation, and
46. H. Melasniema and A. Hernesmaa, Microbiology (UK) 146,
septic tanks (see SEPTIC TANK SYSTEMS) suffered from
701 (2000).
two major disadvantages. They were slow and inefficient
47. P. C. Burrell et al., Appl. Environ. Microbiol. 64, 1878–1883 and/or operated on a relatively small scale.
(1998).
One important outcome of the Commission was the
48. B. K. Mobarry et al., Appl. Environ. Microbiol. 62, 2156– stipulation of standards for treated wastes, and in 1908, it
2162 (1996). recommended that treated effluent should meet the 30:20
49. N. F. Gray, Ann. Appl. Biol. 104, 143–149 (1984). standard (for BOD5 and Suspended Solids, respectively),
50. T. G. Tomlinson and I. L. Williams, Ecological Aspects of which was adopted in Britain in 1912. This requirement
Used Water Treatment, vol. 1, Academic Press, New York, provided further impetus for developing more satisfactory
1975, pp. 93–112. treatment systems that could meet this standard, and
51. L. J. Robinson et al., Water Res. 34, 2310–2322 (2000). undoubtedly played an influential role in the development
52. M. W. Hahn and M G. Höfle, Appl. Environ. Microbiol. 65, of the activated sludge process in 1914.
4863–4872 (1999). In fact, this process grew from a trip Dr. Gilbert
53. K. R. Gurijala and M. Alexander, Appl. Environ. Microbiol. Fowler of the University of Manchester made in 1912
56, 1631–1635 (1990). on behalf of the Manchester Corporation to the Lawrence
54. Y. Inamori et al., Water Sci. Technol. 23, 963–971 (1991).
Experimental Station in Massachusetts to look at different
systems of treating sewage, including aerated processes
55. C. H. Ratsak et al., Water Res. 27, 739–747 (1993).
under study there. On his return to Manchester, he passed
56. G. Hamer, in C. F. Forster and D. A. Joh-Wase, Environmen- on his experiences to Edward Arden and William Lockett
tal Biotechnology, Ellis Horwood, Chichester, U.K., 1987,
at the Manchester-Davy Hulme wastewater treatment
pp. 318–346.
plant. They both understood the importance of the
57. J. C. Gottschal, Methods in Microbiology, vol. 22, Academic American work, and were inspired to carry out laboratory
Press, New York, 1992, pp. 87–124. scale research during 1913 and 1914, which led to the
58. P. Madoni et al., Water Res. 30, 135–141 (1996). activated sludge process. Their brilliantly original idea
59. D. Stephenson and T. Stephenson Biotechnol. Adv. 10, was to retain the sedimenting biomass (sludge) produced
549–559 (1992). from the waste and reuse it to treat further waste material,
60. H. van Limbergen et al., Appl. Microbiol. Biotechnol. 50, instead of discarding it as had been the standard practice
16–23 (1998). until then. By adopting this strategy, they showed that the
61. J. C. Fry et al., in E. M. H. Wellington and J. D. van Elsas, aeration times needed to degrade the organic compounds
eds., Genetic Interactions Among Microorganisms in the in the sewage and achieve complete nitrification were
Natural Environment, Pergamon Press, Oxford, UK, 1992, spectacularly shortened over those in existing wastewater
pp. 194–215. treatment systems.
 ACTIVATED SLUDGE — THE PROCESS 75

The solids were perceptively called ‘‘activated sludge’’ which the success of the process ultimately depends,
by Arden and Lockett, the results of their work can occur efficiently and rapidly in the clarifier.
were published (4,5) and their process, very similar in • In many plants this separation is not achieved
configuration to current conventional activated sludge because of the operational problems of bulking
systems, was tested on a pilot plant scale in 1914. (FILAMENTOUS BULKING IN ACTIVATED SLUDGE, CONTROL
Both continuous flow and fill-and-draw (the Sequencing OF) and foaming (ACTIVATED SLUDGE — FOAMING).
Batch Reactors Technology, which is discussed elsewhere)
• Most of the settled biomass, the activated sludge,
configurations were so successfully tested that a full-scale
is recycled as ‘‘return activated sludge’’ or RAS
continuous plant was installed in Worcester, England, in
to inoculate the incoming raw sewage, with the
1915, the first one of its kind in the world. It used a
understanding that it will contain a selected
mechanical aeration system (which is still preserved and
community of microbes well adapted to treating that
functioning at a plant in Stockport, England) since the
particular raw sewage.
submerged diffusers used in the pilot plant system became
rapidly clogged, a problem still encountered in many plants • Some of this sludge is discarded or wasted, and
today and one which led to development around the world although rich in nutrients, valuable minerals, and
of a variety of aeration systems. The attractions of the a possible food source, the sludge is viewed as a
activated sludge system soon spread outside of Britain, liability that is disposed of, often at great expense.
and by the beginning of World War II, plants had been built
in many European and Commonwealth countries (2,3). Because of all these uncontrollable operational vari-
Since then the activated sludge process, which was ables, it seems remarkable that activated sludge systems
developed originally mainly to remove carbonaceous work as well as they do. However, because they are so
materials and produce an effluent meeting the 30:50 versatile and adaptable, it is very likely that they will
standard, has undergone many modifications to its design continue to represent the most widely used methods for
and operation to increase its efficiency and flexibility so aerobic waste treatment into the foreseeable future and
that many plants built now will also effectively remove evolve as they have in the past to meet additional require-
both nitrogen and phosphorus (6,7). ments and process stringencies (7,11).

FEATURES OF THE PROCESS CONFIGURATIONS FOR ACTIVATED SLUDGE SYSTEMS

In terms of the volume of material handled, the activated Many activated sludge configurations are in use around
sludge process probably ranks as the largest biotechnology the world, differing in their design features and their
process in the world (8). Its features (Fig. 1) include (9–11) performance criteria. It is convenient here to categorize
the following: them on their abilities to remove different organic and
inorganic chemical components in the influent, a feature
• It uses a complex community of mixed populations that reflects their design and operation into:
of bacteria and protozoa (ACTIVATED SLUDGE — THE
MICROBIAL COMMUNITY). • conventional (those plants designed to remove
• This community has to cope with an uncontrollable primarily organic carbonaceous compounds) and
diverse range of organic and inorganic compounds, • nutrient removal (where the plants remove nitrogen
some of which will be toxic to these microbes. and/or phosphorus either chemically or microbiologi-
• The microbes are organized into discrete aggregates cally) (7,12).
called flocs (ACTIVATED SLUDGE — THE FLOC), which are
maintained in suspension in the aeration tank by CONVENTIONAL ACTIVATED SLUDGE SYSTEMS
mechanical agitation/aeration or by the mixing action
of air bubbles from submerged aeration systems. These were the early activated sludge systems introduced.
• The flocs must have good settling properties so that Some of the first ran as semicontinuous ‘‘fill-and-draw’’
separation of solids (biomass) and liquid phases, upon systems, the basis for the increasingly popular sequencing
batch reactor configurations (7), while most conventional
systems now run continuously and operate either as
Aeration distributed uniformly
‘‘plug flow’’ or ‘‘completely mixed systems’’, although the
Raw distinction between them when operating is not always
sewage clear (2,3,13).

Waste Plug Flow Systems

sludge
Clarifier These are systems in which the incoming sewage is
Return activated sludge recycle thought to move as a discrete plug through the reactor(s),
although some mixing will inevitably accompany aeration,
either from submerged diffusers or surface agitators. The
Effluent
general view is that these plug flow systems are less prone
Figure 1. The activated sludge process. to filamentous bacterial bulking (FILAMENTOUS BULKING
76 ACTIVATED SLUDGE — THE PROCESS

IN ACTIVATED SLUDGE, CONTROL OF), although they also

Aeration
operate relatively inefficiently because of their hydraulic
characteristics that generate an uneven load distribution Raw
along the reactors in terms of their oxygen demand (7,11). sewage
Engineering approaches to this problem include adjusting
the aeration systems to redistribute the oxygen to meet
these different requirements, through modifications such Waste
as tapered aeration (Fig. 2), step aeration (Fig. 3), or sludge
altering the feed regime as in the step feed (Fig. 4; 6,7,11). Clarifier
Return activated
sludge recycle
Completely Mixed Systems
Such systems evolved to overcome the problems with Effluent
‘‘plug flow’’ systems, in which all the components, including
Figure 5. The complete mix activated sludge process.
the RAS, are rapidly mixed (Fig. 5). There is a price in
terms of loss of nitrification efficiency if underaeration
or plant overloading occurs, short circuiting may be a
with ‘‘plug flow’’ systems (2,3). Completely mixed systems
problem, and the sludge usually settles less well than
were found to be less susceptible to toxic materials in the
feed due to the rapid dilution of the incoming waste.
Aeration distributed according
to perceived process demand Extended Aeration Systems
As the name suggests these processes operate best with
Raw low loadings, and long sludge ages (2,3). These systems
sewage were designed as inexpensive processes for small rural
communities and were called Pasveer ditches after the
Waste Dutch Engineer who introduced them. These small plants
sludge were operated intermittently (‘‘fill-and-draw’’) but the
Clarifier later-introduced Carrousel plants, which borrowed and
Return activated sludge recycle
developed the original concept, operated continuously and
used separate clarifiers (Fig. 6). Surface agitation provides
Effluent aeration and circular mixing, in which high dissolved
oxygen levels close to the agitators encourage nitrification
Figure 2. The activated sludge process with tapered aeration.
and denitrification then follows in the anoxic regions,
which develop as the liquid moves further away from them
(ACTIVATED SLUDGE — MICROBIOLOGY OF NITROGEN REMOVAL).
Aeration distributed stepwise
Less sludge is produced because of their long sludge ages
and they need little maintenance as long as the mixing is
Raw sufficient to keep the solids/sludge in suspension.
sewage
Stimuli for Improving Performances of Conventional Plants
Waste
sludge Most attempts at improving performances of these plants
Clarifier have been directed at increasing their operational rates,
Return activated sludge recycle
especially with potentially toxic industrial wastes and
reducing the volumes of sludge produced. Modifications
Effluent include ‘‘contact stabilization,’’ high rate systems including
processes such as the Deep Shaft process in which pure O2
Figure 3. The activated sludge process with step aeration.

Aeration and mixing energy

Uniformly distributed aeration provided by surface aerator

Waste
sludge Return activated
Raw
sludge recycle
Raw Clarifier sewage Clarifier
sewage Return activated sludge recycle Waste
sludge

Effluent Effluent

Figure 4. The activated sludge process with step feed. Figure 6. The Carrousel activated sludge process.
 ACTIVATED SLUDGE — THE PROCESS 77

is used instead of air. Gray (2) discusses these processes simpler in configuration and enabling a reduction in
in detail. Also, conventional systems fail to remove the size of the anoxic zone;
sufficient levels of inorganic nutrients such as nitrogen • an organic carbon and energy source to allow the
and phosphorus, both of which are serious environmental denitrifying microbial community in the anoxic zone
hazards, to prevent eutrophication and formation of to reduce the nitrate to nitrogen gas. Deliberate
‘‘algal’’ or blue-green bacterial (Cyanobacterial) blooms
addition of chemicals such as methanol for this
in receiving rivers and lakes. So, many of the changes
purpose was once common, but as their costs
to conventional activated sludge processes have been
increased, plants were designed to use substrates
in developing systems to remove these nutrients. Only
already present for this anaerobic respiration. Many
microbiology-based activated sludge systems are discussed
plants, for reasons not yet understood, probably only
here, and the reader is directed to the reviews of
References 6 and 14 for discussions on chemical nutrient partially reduce the nitrate to nitric oxide and nitrous
removal systems, which pose their own environmental oxide, both serious long-term air pollutants, in the
problems (15). Some plants using microbes often have anoxic zone (16);
facilities for chemical polishing when they fail to perform • in plants in which the aerobic zone precedes the
adequately to satisfy discharge license requirements (6). anoxic zone, as seen in the early Wuhrmann (18)
design (Fig. 7), these organic substrates originate
ACTIVATED SLUDGE SYSTEMS DESIGNED TO REMOVE from lysis of the biomass as most readily biodegrad-
NITROGEN able organic substrates would already have been
utilized in the aerobic zone. Consequently, denitri-
Some nitrogen (N) removal will always occur in con- fication rates are low;
ventional plants since cells in the microbial community • later designs (19) placed the anoxic zone in front of
require nitrogen to grow, and some of these will be and in partial contact with the aerobic zone (Fig. 8).
removed in sludge wasting. Furthermore, nitrification Variable plant performance with this configuration
and denitrification leading to N removal will occur as and a better comprehension of the process require-
long as conditions for these processes are met (ACTIVATED ments led to the suggestion by Barnard to modify
SLUDGE — MICROBIOLOGY OF NITROGEN REMOVAL). Because the it by physically separating the anoxic and aero-
amount of N removed by conventional processes is often bic zones and providing a recycle between the two
small, some plants are deliberately designed and con- into a two-stage process (Fig. 9). However, this also
figured to achieve much higher rates and levels for has an inherent weakness since nitrate removal will
its removal. Several conditions need to be met and all
always be incomplete because the liquor recycled to
plants deliberately designed to remove N incorporate
the anoxic zone and the clarifier feed are derived from
these features. These (ACTIVATED SLUDGE — MICROBIOLOGY
the same source (the aeration basin);
OF NITROGEN REMOVAL) and their operational requirements
are (10,16,17): • Barnard (20) recognized these faults and came up
with the four-stage Bardenpho process (Fig. 10) by
• an aerobic zone in which nitrification by the incorporating a secondary anoxic reactor after the
chemoautotrophic nitrifying bacteria can occur. This aerobic zone, with the dual advantage of maintaining
step is essential to generate nitrate from the oxidation a higher denitrification capacity and a nearly nitrate-
of ammonia as the levels of nitrate in domestic free effluent. It also included a secondary aerobic
sewage from prior nitrification are generally too reactor (reaeration reactor) to allow nitrification of
low to encourage microbial activity and permit the any ammonia produced in the secondary anoxic zone;
chemical transformations needed for N removal; • this configuration should be regarded as the direct
• an anoxic zone in the reactor, with high levels ancestor of most of the nutrient removal systems that
of nitrate from the nitrifying bacteria, which have followed, and required only minor modification
acts in the absence of dissolved oxygen as the to successfully remove P (6,7,11,20,21).
terminal electron acceptor for the anaerobically
respiring chemoheterotrophic denitrifying bacteria.
Many taxonomically diverse bacteria can denitrify,
although it is still unclear as to which ones are Uniformly distributed aeration
important in activated sludge systems. It is known
that many of these can also denitrify under aerobic Raw
conditions, although the impact of these aerobic sewage
denitrifiers on nitrogen removal in activated sludge Waste
systems is again unclear (16); sludge
• most plants designed to remove N use the same Clarifier
biomass for both the nitrification and denitrification Return activated sludge recycle
steps. In the aerobic zones of many such plants, simul-
Nonaerated Aerated
taneous nitrification/denitrification (SND) is thought zone zone Effluent
to occur (16), raising the tantalizing possibility of
manipulating SND to make future N removal plants Figure 7. The Wuhrmann process.
78 ACTIVATED SLUDGE — THE PROCESS

Uniformly distributed aeration Aeration

Raw Raw
sewage sewage
Back-mixing
Waste
sludge Waste
Clarifier sludge
Return activated sludge recycle Clarifier
Return activated sludge recycle
Nonaerated Aerated
zone zone Effluent Aerobic Anaerobic
zone zone
Figure 8. The Ludzack-Ettinger process.
Figure 11. The two-stage or high rate Phoredox process.

Aeration
Aeration
Raw
sewage Raw
sewage

‘a’- recycle Waste Waste

sludge ‘a’- recycle sludge
Clarifier Clarifier
Return activated sludge recycle Return activated sludge recycle

Aerobic Anoxic Anaerobic 1° anoxic

zone zone zone zone Effluent

Figure 9. The modified Ludzack-Ettinger process (MLE). Aerobic 2° anoxic

zone zone

Figure 12. The five-stage Bardenpho or Phoredox process.

Effluent
Aeration

Raw use have evolved from there. Barnard’s pioneering work

sewage in the 1970s helped to strictly define the operational
requirements for EBPR, although this phenomenon
Waste of EBPR was reported much earlier (23). He showed
‘a’- recycle sludge
empirically that adding an anaerobic zone to the head
Clarifier of a conventional plant gave a process (the high rate
Return activated sludge recycle
Phoredox system, Fig. 11) that would remove P when the
Aerobic 1° anoxic 2° anoxic system was operated at a low enough sludge age to avoid
zone zone zone Effluent nitrification. This process, in combination with the four-
Figure 10. The Bardenpho process. stage Bardenpho (Fig. 10) configuration, gave rise to what
is known as the five-stage Bardenpho or Phoredox system
(Fig. 12), a plant configuration capable of removing both
ACTIVATED SLUDGE SYSTEMS DESIGNED TO REMOVE N and P.
PHOSPHORUS His and other studies elucidated the main design
features needed for successful P removal. These include
Conventional activated sludge systems remove very little
of the phosphorus that enters in the raw sewage, except • an initial anaerobic zone or reactor, which is an
that required by the biomass to grow. Thus, their essential component of all EBPR systems. This zone
effluents are usually rich in P and they are recognized appears to selectively favor bacteria [including the P
as one of the major point sources for environmental accumulating bacteria (PAB) responsible for EBPR]
P pollution (22). P-removing activated sludge systems that are capable of anaerobically assimilating the
work because the microbes assimilate more P than is organic substrates present there (mainly short-chain
needed for their growth, and so are also called enhanced volatile fatty acids, SCVFA, such as acetate). Cells
biological phosphorus (EBPR) systems (7,20,21). Their store these substrates as intracellular reserves of poly
detailed microbiology is discussed elsewhere (ACTIVATED β-hydroxyalkanoates (PHA) to be used as sources of
SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING COMMUNITY energy and carbon for their growth in the aerobic zone
COMPOSITION). Most agree that current EBPR systems, if (see later), where now other metabolizable substrates
properly designed, should yield a final effluent P level of are no longer available.
less than 1 mg l−1 (21). • the stored polyP that is used as the energy source
Much of the development of these processes occurred for PHA synthesis and the phosphate produced is
in South Africa and most of the plants currently in released into the bulk liquid. Strictly anaerobic
 ACTIVATED SLUDGE — THE PROCESS 79

conditions are essential here to eliminate any aerobic Aeration

respiration of these organic substrates, reducing their
availability to the PAB. Raw
sewage Waste
• a prefermenter using an anaerobic digestion of settled sludge
sludge that is installed in many EBPR plants before
the anaerobic reactor to guarantee a supply of high ‘a’- recycle
levels of SCVFA to cells in the anaerobic zone (24). In
some designs, acetate itself is added to the anaerobic Clarifier
Return activated sludge recycle
reactor. Claims are that these modifications both
improve and stabilize EBPR in systems in which
Anaerobic 1° anoxic Aerobic
they have been introduced, although odor can become zone zone zone Effluent
problematic. However, the size of the anaerobic
reactor can now be reduced. Figure 13. The three-stage Phoredox process.
• an anoxic zone where denitrification is encouraged
(in plants that nitrify) so that the levels of nitrate
entering the anaerobic zone are insufficient to permit Aeration
any anaerobic respiration of these organic substrates,
Raw
reducing their availability to the PAB. Consequently, sewage Waste
most of the later designed plants removing P sludge
also remove N, although the presence if not the
importance, of denitrifying PAB in activated sludge ‘a’- recycle
systems has been well documented (25) and these
may make a contribution to denitrification in the Clarifier
‘s’- recycle
anoxic zone.
• an aerobic zone, where the nitrifying bacteria Anaerobic 1° anoxic
function and those with stores of PHA produced zone zone Effluent
in the anaerobic zone can grow and use the
Aerobic 2° anoxic
energy made available from them to assimilate zone zone
phosphorus from the bulk liquid and store it as polyP
granules. As more P is assimilated aerobically than Figure 14. The Johannesberg process.
released anaerobically, net P removal occurs, and this
assimilated P immobilized in the biomass is wasted
with it. ‘r’- recycle Aeration
• alternating the biomass between the anaerobic and
aerobic zones in the plants, although this was not a Waste
deliberate design feature of some of the earlier plants. sludge
Raw
Stimuli for Improving Performances of EBPR Systems sewage ‘a’- recycle
Many of these improvements have been based on a better
Clarifier
understanding of the microbiology of these systems, which ‘s’- recycle
is a trend that is likely to continue. These were introduced
in attempts to (26): Anaerobic 1° anoxic Aerobic
zone zone zone Effluent
• simplify plant configuration, reduce the size of
Figure 15. The University of Cape Town (UCT) process.
the reactors (see the preceding text) making them
inexpensive to build and run
• cope with high sudden flow rates, which would dilute This design has in turn produced several progeny with
SCVFA levels below those capable of supporting characteristics to cope with difficulties encountered over
good EBPR time with their use. A few examples are given here. The
• overcome the perceived operational problems in many Johannesburg process (Fig. 14) incorporated a secondary
of the earlier design changes caused by returning anoxic zone in the RAS line to the anaerobic reactor
nitrate back to the anaerobic zone, a stimulus that to ensure complete nitrate removal, but not without
led to many new plant designs, some of which operational consequences, which were addressed by the
are operating successfully around the world. Thus, University of Cape Town (UCT) process that recycles
the three-stage Phoredox process evolved from the the RAS back into the anoxic rather than the anaerobic
five-stage system by increasing the size of the zone first before another recycle then transfers it from
primary anoxic zone to increase denitrification levels the anoxic to the anaerobic reactor (Fig. 15). Additional
and reduce recycled nitrate, so that the protecting improvement was achieved when the anoxic zone was split
secondary anoxic zone and reaeration reactor were into two separate reactors to provide further control, as
no longer needed (Fig. 13). seen with the Modified UCT (MUCT) process (Fig. 16). The
80 ACTIVATED SLUDGE: USE OF MOLECULAR TOOLS

‘r’- recycle Aeration case, the costs of building the systems and running them
will increase and additional demands will be placed on the
operational staff who will need to be highly skilled.
Waste
sludge
Raw
sewage ‘a’- recycle BIBLIOGRAPHY
Clarifier
‘s’- recycle 1. H. A. Hawkes, Ecological Aspects of Used-Water Treatment,
vol. 1, Academic Press, New York, 1983, pp. 77–162.
Anaerobic 1° anoxic Aerobic 2. N. F. Gray, Biology of Wastewater Treatment, Oxford Univer-
zone zone zone Effluent
sity Press, Oxford, U.K., 1989.
Figure 16. The modified UCT process. 3. N. F. Gray, Activated Sludge: Theory and Practice, Oxford
University Press, Oxford, U.K., 1990.
4. E. Ardern and W. T. Lockett, J. Soc. Chem. Ind. 33, 112
MUCT process sacrifices some of its overall denitrification (1914).
capacity to insure that nitrate is completely removed from 5. E. Ardern and W. T. Lockett, J. Soc. Chem. Ind. 33, 523–539
the recycle to the anaerobic basin, thus protecting this (1914).
zone and optimizing the P removal process. 6. M. T. J. Meganck and G. M. Faup, Biotreatment Syst. 3,
111–204 (1988).
• prevent P release (secondary release) from the settled 7. R. J. Seviour et al., in The Microbiology of Activated Sludge,
sludge in the clarifier and the RAS by avoiding Kluwer Academic Publishers, Dordrecht, The Netherlands,
establishment of anaerobic conditions anywhere in 1999, pp. 44–75.
the sludge handling processes (e.g., by avoiding 8. P. Grau, Activated Sludge Process Design and Control: Theory
prolonged sludge storage). and Practice, Technomic Publishing Co., Lancaster, Pa., 1992,
pp. 1–36.
THE FUTURE FOR ACTIVATED SLUDGE PROCESSES? 9. A. L. Downing, Activated Sludge Design and Control: Theory
and Practice, Technomic Publishing Co., Lancaster, Pa., 1992,
What activated sludge plants might look like and how they pp. 69–126.
might operate in 50 or 100 years from now is difficult to 10. M. Henze et al., Wastewater Treatment: Biological and
predict, but a few speculations are allowed. Chemical Processes, Springer-Verlag, Berlin, Germany, 1995.
11. D. F. Toerien et al., Adv. Microb. Ecol. 11, 173–230 (1990).
• Methods and instrumentation for automatic monitor- 12. J. Wanner, Water Sci. Technol. 20, 1–8 (1994).
ing and control of these systems will rapidly continue
13. J. Wanner, Activated Sludge Bulking and Foaming Control,
to increase.
Technomic Publishing Co., Lancaster, Pa., 1994.
• Single ‘‘add-on’’ units containing communities of 14. D. Jenkins and S. W. Hermanowicz, Phosphorus and Nitro-
bacterial populations with specific metabolic traits gen Removal from Municipal Wastewater, Lewis Publishers,
to treat particular wastes (e.g., toxic and industrial) Boca Raton, Fla., 1991, pp. 167–202.
will become common.
15. G. J. J. Kortstee et al., FEMS Microbiol. Rev. 15, 137–155
• Advances in membrane technology will revolutionize (1994).
solids separation so that clarifiers will become 16. L. A. Robertson and J. G. Kuenen, Microbial Control of
archaeological oddities and problems such as bulking Pollution, Cambridge University Press, Cambridge, U.K.,
will become simple to handle. 1992, pp. 227–267.
• Membranes will be used increasingly in plant design 17. M. C. Wentzal et al., Water Sci. Technol. 23, 555–565 (1992).
so that most will use them to polish treated effluent 18. K. Wuhrmann, Schweiz, Z. Hydrol. 19, 409–417 (1957).
water for reuse as potable water.
19. F. J. Ludzack and M. B. Ettinger, J. Water Pollut. Control
• Our knowledge of the bacteria responsible for N Fed. 34, 920–931 (1962).
and P removal and the factors that affect these
20. J. L. Barnard, Water SA. 2, 136–144 (1976).
processes in these organisms will undoubtedly
21. J. L. Barnard, Proc. 2nd Aust. Conf. Biol. Nut. Remov., 1–14
increase significantly, and this information will lead
(1994).
to the design of novel configurations with input
from both engineers and microbiologists, ensuring 22. Y. Argaman, in Biological Degradation of Wastes, Elsevier
Applied Science, Newfoundland, Canada, 1991, pp. 85–111.
that these populations and their pertinent metabolic
properties are particularly encouraged. 23. P. I. Bond and G. N. Rees, The Microbiology of Activated
Sludge, Kluwer Academic Publishers, Dordrecht, The Nether-
• Sludge reuse will also become more sophisticated and
lands, 1999, pp. 227–256.
less costly.
24. J. L. Barnard, Water Sci. Technol. 15, 1–13 (1983).
It is possible that few or none of these changes 25. T. Mino et al., Water Res. 32, 3192–3207 (1998).
will occur, and future developments in the treatment of 26. J. L. Barnard, Advanced Wastewater and Sludge Treatment
wastewater will use totally different technology. In either Conference, IBC Conferences, Sydney, Australia, 1999.
 ACTIVITY AND CARBON TRANSFORMATIONS IN BIOFILMS 81

ACTIVATED SLUDGE: USE OF MOLECULAR which photosynthetic organisms synthesize organic carbon
TOOLS. See ACTIVATED SLUDGE — MOLECULAR TECHNIQUES compounds from carbon dioxide. Secondary producers con-
FOR DETERMINING COMMUNITY COMPOSITION vert the products of primary production into biomass (2).
By convention, metabolism of carbon substrates is
divided into two closely linked processes.

— Dissimilatory metabolism involves catabolic (degra-

ACTIVITY AND CARBON TRANSFORMATIONS dative) activities performed to provide energy for
IN BIOFILMS cellular respiration.
— Assimilatory metabolism describes anabolic reac-
ANDREW LEIS tions, in which energy is expended for the assimila-
HANS-CURT FLEMMING tion of low-molecular-weight substrates into macro-
Gerhard Mercator Universität molecules, which have structural or physiological
Duisburg, Germany functions.

Biofilms are an important sink of the carbon-associated Most of the energy obtained from catabolic reactions is
biomass in aquatic and terrestrial ecosystems. Much of this used for growth, although biofilm cells spend a dispropor-
biomass comprises eubacteria, although archaebacteria, tionate part of their energy budget on maintenance and
cyanobacteria, fungi, and various forms of algae can the production of the biofilm hydrogel matrix (extracellu-
be quantitatively important or perhaps, dominant in lar polymeric substances, EPS). Planktonic bacteria seem
biofilms under favorable circumstances. The superior to have adopted a strategy for maximum substrate utiliza-
growth rates of prokaryotes enables a rapid turnover of tion, rather than efficiency (3). The major transformation
usable substrates generated by primary producers, thus carried out by microorganisms in biofilms can be summa-
establishing the critical role of these organisms in the rized as the metabolism of inorganic and organic electron
cycling of organic matter. It is an apparent paradox that donors and acceptors to produce soluble by-products, EPS,
biofilms form in natural waters, which are considered to carbon dioxide, and/or methane, and water (4). The mech-
be nutrient-depleted or ‘‘oligotrophic.’’ While a reduction anisms used to achieve these goals are considerably more
in the supply of nutrients in technical systems below diverse, and all major metabolic classes of microorganism
a certain threshold level can minimize the formation are represented in biofilms. In fact, it is often the biofilm
of biofilms (1), this principle cannot be extended to the that creates the potential for a metabolic group of organ-
fundamental complexities of natural environments. A isms to become established. Recent studies employing
complete description must take into account the diversity either microelectrodes or fluorescent in situ hybridiza-
of substrates available for use, the fluxes of substrates tion (FISH) in combination with confocal laser scanning
to and from biofilms, the taxonomic diversity of biofilm microscopy (CLSM) have provided direct evidence for the
organisms, and the numerous metabolic pathways, which existence of different metabolic groups within discrete
have evolved to enable the transformation of structurally areas of the biofilm matrix (5). In general, classifications
dissimilar and dilute carbon sources into potentially of biofilms according to metabolic pathways include the
utilizable forms. The majority of abiotic carbon compounds designations aerobic, anaerobic (sulfate-reducing bacte-
in soils, sediments, and natural waters are decay ria, methanogens), anoxic (denitrifying), and fermentative,
products known as humic substances, which bear only depending on the type and concentration of the elec-
a superficial resemblance to their molecular precursors. tron donor and acceptor (6). Microorganisms can also be
The quantitative dominance of humic substances in described in terms of carbon utilization:
the dissolved phase is a consequence of the relative
recalcitrance of these compounds and the fact that most — Autotrophs assimilate single-carbon (C-1) com-
assimilable carbon compounds are immobilized as cellular pounds, including CO2 , CH4 , and methanol.
biomass in biofilms. The use of humic substances as a — Photoautrophs harness electromagnetic energy to
source of carbon and energy is not necessarily restricted generate ATP and the reducing power required for
by an absence of suitable degradation pathways; at the assimilation of carbon dioxide.
least, part of the explanation for the dominance of these — Chemoautotrophs (lithotrophs, lithos Gr. for rock)
compounds in natural systems is related to the inefficiency obtain cellular carbon by assimilative carbon dioxide
of microbial degradation pathways. Microorganisms have reduction or from reduced C-1 compounds and energy
evolved diverse and efficient pathways for the utilization from oxidation of inorganic compounds such as NH4 + .
of those carbon substrates that warrant a metabolic — Heterotrophs (chemoorganotrophs) obtain energy
investment. from the oxidation of complex organic carbon sources.

DIVERSITY OF STRATEGIES FOR CARBON METABOLISM The terms autotrophy and heterotrophy can apply
BY BIOFILM ORGANISMS to both dissimilatory and assimilatory metabolism.
Autotrophs may require additional carbon in the form
Environmental biofilms contain both eukaryotic and of growth factors or vitamins (6), but they assimilate
prokaryotic organisms, which act as primary and sec- all necessary elements as inorganic forms. The energy
ondary producers. Primary production is the process by expenditure required for synthesizing macromolecules
82 ACTIVITY AND CARBON TRANSFORMATIONS IN BIOFILMS

partly accounts for the lower growth rates of strict within flocs (see MICROBIAL FLOCS SUSPENDED BIOFILMS, this
autotrophs compared to most biofilm bacteria, which Encyclopedia).
are (chemo)heterotrophic — they depend on organic com-
pounds to generate energy and the precursors used to Respiration
synthesize cellular biomass. A low-molecular-weight sub-
The primary goal of respiration is the production of
strate may serve as a source of carbon for building
ATP (phosphorylation) by any combination of organic or
macromolecules and as a source of cellular energy, pro-
inorganic electron donor and electron acceptor. Aerobic
ducing carbon dioxide as a by-product (6). Phototrophic
respiration of an organic substrate is described most
microorganisms use different catabolic and anabolic sub-
simply as:
strates. The complete metabolism of carbon substrates to
(CH2 O) + O2 → CO2 + H2 O (1)
CO2 or CH4 is called ‘‘mineralization.’’ Confusingly, some
of the inorganic carbon in biofilms is converted to min-
This process yields the most energy of all catabolic
eral forms such as calcium carbonate, partly by abiotic
processes, so it is preferred by facultative microorganisms
processes and partly as a result of microbially-induced
when oxygen is available. Many microorganisms are
increases in pH (7). The biotic mechanisms are known as
able to perform oxidative phosphorylation of organic
‘‘biomineralization.’’ Together with various geologic pro-
substrates. Most respiratory organisms use the citric
cesses, these reactions form sedimentary deposits known
acid (or tricarboxylic, TCA) cycle for the oxidation of
as microbialites, which constitute fossilized biofilms. More
acetate to carbon dioxide, whereas phototrophic green
specifically, stromatolites and thrombolites are formed by
sulfur bacteria and some archaea use the reverse of
a consortia of bacteria and filamentous cyanobacteria, and
this cycle to assimilate carbon dioxide. Purple nonsulfur
bacteria and unicellular cyanobacteria, respectively (7).
bacteria use the same electron transport system for
Energy-yielding processes (catabolism) are further
both respiration and photophosphorylation. Anaerobic
categorized as fermentation, respiration, phototrophy, and
respiration is characterized by the use of exogenously
methanogenesis.
derived terminal electron acceptors other than oxygen.
Fermentation Energy yields are always lower than those obtained
by oxidative phosphorylation. Ecologically important
Fermentation is often used in a general sense to anaerobic respiration processes include the reduction
describe the utilization of a substrate by a microbial of sulfate, nitrate, iron, and manganese, although
culture (heterotrophic fermentation). In a metabolic sense, some low-molecular-weight organic compounds may also
fermentations are anaerobic processes in which the serve as terminal electron acceptors. Nitrate or nitrite
substrates are sequentially transformed by reduction- reduction is performed by many facultative anaerobes,
oxidation (redox) processes in the absence of an external including members of the ubiquitous genera Bacillus and
electron acceptor. The redox levels of the substrate and Pseudomonas. In these organisms, O2 inhibits synthesis
metabolite(s) remain the same, and the energy yields are of the key enzyme, nitrate reductase (6).
low. Fermenting bacteria may be: Sulfate-reducing organisms are common in biofilms due
to the depletion of oxygen by respiring organisms and
— strict (obligate, O2 -sensitive) anaerobes, for example, subsequent diffusion limitation within the EPS matrix.
Clostridium spp.; Sulfate reduction is an important contribution to anaerobic
— facultative anaerobes, which are capable of oxidative metabolism in both freshwater and marine sediments.
phosphorylation (production of adenosine triphos- The principal carbon substrates for this process are
phate, ATP) in the presence of O2 , for example, low-molecular-weight organic acids (acetate, butyrate,
Pseudomonas spp.; or propionate, lactate), alcohols, and H2 . Sulfate reducers
— aerotolerant anaerobes, for example, Lactobacillus can be subdivided into two further groups on the basis of
spp., some Streptococcus spp. carbon utilization. Organisms such as Desulfovibrio spp.
do not use a citric acid cycle. They can use H2 or lactate
Facultative anaerobes are especially suited to the as substrates. The use of lactate is inefficient in that the
gradients, which develop suddenly within biofilms as carbon product is acetate (rather than methane), but this
a consequence of the rapid depletion of oxygen in the substrate is readily assimilated by heterotrophs within
vicinity of immobilized cells. Anaerobic fermentation of the biofilm matrix. Other organisms such as Desulfobacter
simple sugars yields little energy compared with aerobic spp. make use of a citric acid cycle to oxidize acetate
metabolism. This type of fermentation is exploited in completely. Members of the genus Desulfuromonas use
industry, but it is rare in environmental biofilms because elemental sulfur rather than sulfate for the complete
the heterotrophic microorganisms usually ensure that oxidation of acetate, but also produce CO2 and HS− .
high concentrations of sugar substrates do not accumulate. The methanotrophs in soil and aquatic environments
Amino acids can be fermented via the Stickland reaction, are gram-negative bacteria, which oxidize methane or
whereby one kind of amino acid is used to oxidize another other C-1 compounds, such as methanol or methylamine.
to acetate and ammonia (8). This process is important They act as biofilters for methane produced in anaerobic
when high concentrations of proteins are available environments. Methanotrophs are classified into three
for anaerobic degradation. Planktonic microorganisms groups, partly according to their ability to assimilate the
can encounter relatively high-protein concentrations on key intermediate, formaldehyde (10). Anaerobic oxidation
suspended particles (9), where anaerobic zones form of CH4 takes place in sediments and soils; when oxygen
 ACTIVITY AND CARBON TRANSFORMATIONS IN BIOFILMS 83

is present, soil methanotrophs can oxidize atmospheric enabling the operation of two separate reaction systems,
methane. Some methanotrophs can synthesize a soluble Photosystem I and Photosystem II. The so-called purple
methane monooxygenase capable of oxidizing priority and green bacteria carry out photosynthesis, but do not
pollutants such as trichloroethylene (10). produce oxygen. This form of photosynthesis involves the
production of ATP by cyclic phosphorylation. Anoxygenic
Methanogenesis and Acetogenesis photosynthesis occurs in illuminated sediments, microbial
mats, and anoxic zones in (other) biofilms, in which H2 S
Methanogens belong to the archaea domain and reduce
acts as proton and electron donor, and elemental sulfur
carbon substrates to methane, using H2 as electron
is produced. The exposure of oxygenic, sediment-dwelling
donor. The carbon substrate is usually carbon dioxide,
cyanobacteria to sulfide inhibits Photosystem II; however,
which also constitutes autotrophy. Carbon dioxide, carbon
the active Photosystem I allows anoxygenic photosynthe-
monoxide, and formate comprise one group of potential
sis with HS− as the electron donor. In dark and anaer-
substrates for methanogens. A second group of substrates
obic conditions, some cyanobacteria seem to be capable
comprises compounds with methyl groups, which can
of fermentation or sulfate reduction (6). Four, unrelated
act as both electron donor and acceptor. The cleavage
groups of eubacteria can perform anoxygenic photosynthe-
of acetate (acetotrophy) is a third mechanism, producing
sis, and different groups have different combinations of
methane and carbon dioxide. Methanogenesis is important
carotenoid pigments. The purple bacteria contain bacteri-
in anaerobic wastewater treatment, during which soluble
ochlorophyll a or bacteriochlorophyll b. The purple sulfur
organics are degraded to fatty acids and subsequently
bacteria (e.g., Thiocapsa, Chromatium) use reduced sul-
converted to methane and carbonate. The fermentation
fur as an electron donor, and also store elemental sulfur.
of acetate is responsible for the bulk of the methane
The ‘‘purple nonsulfur bacteria’’ (e.g., Rhodospirillum) do
formation (11). This process is performed by acetogens
not store elemental sulfur, and they are more sensitive to
and homoacetogens. high sulfide concentrations. The designation is a misnomer
because these organisms are also capable of using HS− as
Phototrophy and Photoautotrophy electron donor. Purple nonsulfur bacteria are versatile in
Phototrophy and photoautotrophy are terms to describe that they can use some carbon compounds (and Fe2+ ) as
carbon dioxide fixation, which is mediated by photosyn- electron donors, and they can assimilate low-molecular-
thesis: weight compounds such as acetate. The green sulfur
bacteria (e.g., Chlorobium) have bacteriochlorophylls c
nCO2 + 2nH2 O → n(CH2 O) + O2 + nH2 O (2) or d. They are autotrophic, strictly anaerobic organisms,
which use sulfide as an electron donor, but they only
Light energy is converted to chemical energy (ATP oxidize it to S0 , which is excreted. The green nonsulfur
and the reduced form of nicotinamide adenine dinu- bacteria (e.g., Chloroflexus) are also erroneously named,
cleotide phosphate, NADPH) in the ‘‘light’’ reactions, and being able to use HS− . These are filamentous gliding forms
subsequently used to reduce carbon dioxide to organic found in microbial mats, and they can live as photoau-
totrophs using HS− and CO2 or as aerobic heterotrophs.
compounds in the ‘‘dark’’ reactions. Phototrophs and some
The heliobacteria are strict phototrophic anaerobes, which
other microorganisms contain various forms of chloro-
have been isolated from anoxic soils. Halobacteria are
phyll and accessory pigments, including carotenoids and
a group of archaebacteria, which usually act as aerobic
phycobilins. Carotenoid pigments can transfer energy,
heterotrophs, but under conditions of anoxia and light,
but not directly through photophosphorylation. Phycobilo-
they revert to phototrophy. A ‘‘purple’’ membrane con-
proteins are the principal light-harvesting pigments of
sisting of a photosensitive pigment, rhodopsin, facilitates
cyanobacteria and red algae. They are arranged as high-
proton export from the cell. The proton motive force is
molecular-weight aggregates called phycobilisomes. The
then harnessed by ATPase in the cell membrane for ATP
energy transfer from the biliprotein complex to chloro-
synthesis.
phyll a is very efficient, enabling growth under low-light
conditions. The phycobilisome content correlates with light
intensity. SYNTROPHY AND ANTAGONISM IN BIOFILMS
The two modes of phototrophy are oxygenic pho-
tosynthesis and anoxygenic photosynthesis. Oxygenic Ultimately, the establishment of different organisms
photosynthesis can be performed by organisms that con- within a biofilm is dependent on competition, and many
tain chlorophyll a, including various microalgae. Light interactions may be antagonistic (14). There is no convinc-
energy is used to synthesize ATP and NADPH by non- ing, direct evidence for antagonistic interactions within
cyclic photophosphorylation, whereby O2 is produced from a biofilm, as can be observed frequently in the case of
water. Cyanobacteria are the only prokaryotes capable inhibition zones surrounding colonies of microbial soil
of oxygenic photosynthesis (12). Filamentous cyanobac- isolates on laboratory media. It can be inferred, how-
teria dominate shallow sediments and microbial mats, ever, that sulfate reducers and methanogens will compete
which are essentially thick, photosynthetic biofilms. The for carbon dioxide. Also, methanogens and homoaceto-
unicellular cyanobacteria (e.g., Synechococcus) are poorly gens use the same precursors, CO2 and H2 , to produce
represented in microbial mats, but they are important as methane and acetate, respectively. Unless carbon diox-
planktonic primary producers (13). Oxygenic phototrophs ide is subject to considerable diffusion limitation within
have two, spectrally distinct forms of chlorophyll a, a biofilm, its availability is far better than the relatively
84 ACTIVITY AND CARBON TRANSFORMATIONS IN BIOFILMS

insoluble O2 , which is depleted rapidly by respiring organ- Lipids

isms. Competition is especially relevant for the limited Cutin
(1.3%)
supply of low-molecular-weight carbon substrates such Proteins
(1.3%) (1.3%)
as monosaccharides and amino acids, which constitute
a ‘‘superlabile’’ carbon pool. The proximity of metaboli- Hemicellulose and Cellulose
cally diverse neighbors and selection pressures, however, other polysaccharides (42%)
implies that neutral or even favorable interactions will (25%)
allow particular metabolic groups to establish within the
biofilm matrix. This is especially important for photosyn-
thetic biofilms, in which the carbon and oxygen cycles are
intimately related. Algal exudates in epilithic biofilms
provide heterotrophic bacteria with significant quanti-
ties of organic and inorganic substrates (15,16). Syntrophy
Lignin
occurs when two or more organisms catabolize substrates (29%)
in a mutually-dependent manner. In freshwater systems,
Figure 1. Schematic diagram showing the syntrophic metabolic
purple nonsulfur bacteria commonly form syntrophic asso-
interactions in a mixed phototrophic/heterotrophic biofilm.
ciations with heterotrophic sulfur reducers. Methanogens
Note the concentration gradients within the biofilm, and the
depend on other organisms for a supply of the few com- dependence of O2 and CO2 gradients on diurnal and nocturnal
pounds, which can act as precursors for methane pro- cycles. Carbon dioxide is buffered to hydrogen carbonate and is
duction, including some products of respiration. Respiring approximately 5 times more soluble in water than oxygen.
organisms can hydrolyze polysaccharides, proteins, and
other macromolecular organic substrates. Low-molecular-
weight organic acids produced by fermenting bacteria can reducing agent is often supplied by organisms performing
be used by sulfate reducers, accounting for the presence of oxygenic photosynthesis and subsequently degraded under
these organisms in consortia (6). Fermentation of organic anaerobic conditions. When referring to the use of HS−
substrates proceeds only as far as acetate + H2 , but the as a substrate, the designation of autotrophy is only
acetate can be respired by other organisms. Thus, com- justified if the HS− is provided abiotically, for example,
plete mineralization depends on the supply of substrates geothermal vents or acid mine drainage. In the water
by actively fermenting bacteria. In the case of phototrophs, column of a deep lake, temperature gradients induce
energy derived from the oxidation of a noncarbon sub- ‘‘thermal stratification,’’ in which the lower strata are
strate, such as sulfide, requires the assimilation of some colder and oxygen-depleted (reducing conditions). This
other carbon source, such as carbon dioxide produced by prevents admixing among the lower, hypolimnion, and
fermenting organisms. The expected carbon transforma- other layers. Photosynthesis is limited to the upper
tions in a mixed phototrophic/heterotrophic biofilm are (euphotic) zone, which is defined by the penetration depth
depicted in Figure 1. The fermentation of glucose to acetate of solar radiation, and commonly measured by the Secchi
and H2 by obligate acetate-producing bacteria requires the disk method. Particulate matter absorbs much of the
recycling of hydrogen by reoxidation of NADH. In thermo- radiation, such that when POC levels in natural waters are
dynamic terms, this process is only possible when the high, the euphotic zone is extremely small. Within biofilms,
hydrogen tension is low (less than 10−4 atm). The close absorption or other attenuation of incoming radiation
proximity of H2 -consuming biofilm bacteria such as sulfate confines some types of photosynthetic activity to the upper
reducers or methanogens facilitates ‘‘interspecies hydro- strata (17). Significant physicochemical gradients also
gen transfer’’ (6), and the equilibrium hydrogen tension exist in sediments and soil microcosms. Thus, substrate
largely determines the dominant organisms. and physicochemical gradients become established within
The diversity of microbial metabolism is exemplified other gradients. Biofilm organisms are ultimately affected
by the fact that some classes of microorganisms use by the trophic status of the bulk fluid compartment.
more than one metabolic strategy. The most obvious and
important example is the switch between autotrophy and SOURCES OF NATURAL ORGANIC CARBON SUBSTRATES
heterotrophy under different conditions. Purple nonsulfur
bacteria are strictly autotrophic during the day, when they Carbon substrates belong to the dissolved or particulate
use H2 or HS− as reductants for photosynthesis and for organic carbon pools (DOC and POC, respectively).
assimilating carbon dioxide. At night, they can also respire Colloid scientists use the thermodynamic concept of
oxygen by using low-molecular-weight organic substrates, speciation, where the term ‘‘dissolved’’ refers to chemical
such as acetate. Some chemoautotrophs can assimilate potential (18). In this sense, DOC is the fraction of greatest
organic substrates, and under anaerobic conditions, some significance to surface phenomena because it adsorbs
sulfur bacteria can either reduce sulfur or perform onto surfaces to form conditioning films (see CONDITIONING
heterotrophic fermentations. Autotrophy is an ambiguous FILMS, this Encyclopedia). This molecular diffusion is
term for biofilm organisms because the energy cycle considerably more rapid than the adsorption of colloids,
is intimately related to other cycles occurring in close including bacteria, as described by Fick’s First Law of
proximity. For example, purple sulfur bacteria carry molecular diffusion. Thus, a continuous flow of (albeit
out photosynthesis using HS – as an electron donor, limited) substrate replenishes nutrients within a porous
thereby assimilating carbon as carbon dioxide, but the biofilm matrix. These nutrients can be in the form of
 ACTIVITY AND CARBON TRANSFORMATIONS IN BIOFILMS 85

trapped particles or macromolecules, which are degraded carbon compounds, such as lignin, appear to degrade
if the appropriate exoenzymes are available. Sorption more readily in the ocean than would be expected from
of kaolin particles to a Pseudomonas fluorescens biofilm laboratory-based studies. Some of the most important
caused an increase in microbial activity compared to findings were derived from the use of specific biomarker
biofilms grown in the absence of clay particles, as explained compounds, which can be traced with some level of cer-
by the greater biomass in the former (19). Nutrient tainty; however, only a small fraction of the organic matter
sources are described as allochthonous or autochthonous, dissolved in seawater and preserved in marine sediments
depending on their origins. Allochthonous carbon sources appears to be terrestrial in origin, despite the fact that
are produced in locations remote from the environment of annual exports from rivers are more than sufficient to
interest and reach the location by some physical process. account for deposition in sediments (24). This anomaly is
The leaves of vascular plants, for example, provide an the most elusive aspect of the global carbon cycle. Similar
exogenous source of lignin to lakes and rivers. This may difficulties apply to soil ecosytems, in which percolation
be a direct event, such as in the case of the leaf falling of water-soluble organic carbon through the soil matrix is
into a stream, or a consequence of a more indirect route, relatively small in comparison to inputs from leaf mate-
such as the leaching of soluble components into the water rial, roots, and root exudates. It is possible, however,
from an adjacent soil matrix, in which case the microbial to estimate the above-ground organic matter prior to its
decomposition processes are initiated prior to reaching the incorporation into the soil matrix (Fig. 2). Microorganisms
water. Anthropogenic carbon sources (i.e., derived from in some groundwater systems live in conditions that are
human activity) are mostly allochthonous. Autochthonous completely remote from aquatic or soil inputs. The sub-
carbon sources, as the name suggests, are produced in strates in these environments are the products of microbial
situ. Examples include cell lysis products and exudates metabolism and lysis and kerogen. These two different
from planktonic algae. The availability of autochthonous scenarios emphasize the ability of microorganisms to live
substrates is determined by the rates of incorporation into under conditions of ‘‘feast and famine,’’ in which the latter
biomass, but also by the remobilization of carbon from predominates.
particles. This is especially important when sediments
are mixed by advection. Most of the DOC in oceans is Sources and Classes of Proteins, Carbohydrates, and Lipid
autochthonous, especially in coastal regions and estuaries, Substrates in Natural Environments
due to the photosynthetic activity of phytoplankton (20).
Hydrolysis of proteins and protein conjugates from
The broad range of potential substrate compositions
seawater produces both dissolved free amino acids (DFAA)
(Table 1) is dependent on the extreme variability of spatial
and dissolved combined amino acids (DCAA). DCAA
and temporal parameters.
derived from marine organisms constitute the largest,
The low DOC of groundwater may be due to
well-defined form of all DOC; however, studies of the
any combination of limited leaching from aquifer
amino acid composition of DCAA have yielded little
solids (kerogen), complete mineralization of carbon sub-
information because proteins from living organisms have
strates by heterotrophic microorganisms, and adsorption-
similar amino acid compositions (25). Carbohydrates are
immobilization (22). Like humic substances, kerogen is
quantitatively dominant over proteins in soils, freshwater,
thought to be derived from the decay of plants and microor-
and marine environments, accounting for up to 20%
ganisms (23). Ultimately, the major source of terrestrial
of seawater DOC. Most of this is associated with
organic matter is vascular land plants. Some of these
photosynthetic activity in irradiated surface waters. All
major classes of carbohydrates occur in marine DOC,
Table 1. Concentrations of Dissolved Organic including amino sugars, uronic acids, and aldoses (26).
Carbon in the Major Types of Natural Waters Like amino acids, carbohydrates in natural waters exist
both as free monomers and as polymers. Polysaccharides
Type of Water [Dissolved Organic Carbon] (mg/L)
can account for 80% or more of the total dissolved
carbohydrates in seawater. Fibrillar polysaccharides may
groundwater 0.2–15 (0.7† )
marine 0.5–3.0
account for up to 25% of the entire pool of natural
lake 2–10 organic matter (NOM) in fresh waters (27) and therefore
— oligotrophic 1–3 constitute the second largest pool of aquatic DOC after
— eutrophic 2–5 humic substances. The major polysaccharides in soils
— dystrophic‡ 20–50 are cellulose and hemicellulose. The most abundant
rivers 5–60§ protein in photosynthetic organisms is Rubisco (ribulose-
1,5-bisphosphate carboxylase). The turnover time of
Source: Data from E. M. Thurman, Organic Geochemistry this enzyme is thought to be rapid (28), although
of Natural Waters, Martinus Nijhoff/Dr W. Junk Publish-
incubation with the nocturnal Rubisco inhibitor 2’-
ers, Dordrecht, Netherlands, 1985 and R. L. Malcolm, in
A. J. Beck, K. C. Jones, M. H. B. Hayes and U. Mingelgrin, carboxy-D-arabitinol-1-phosphate was shown to confer
eds., Organic Substances in Soil and Water: Natural Con- protection against proteolytic enzymes (29).
stituents and their Influences on Contaminant Behaviour, Lipids constitute a small fraction of the total dis-
Royal Society of Chemistry, Cambridge, U.K., 1993, solved organic carbon. The major classes of lipids in
pp. 19–30.
† mean concentration. DOC are hydrocarbons, chlorins, and their esters (i.e., pig-
‡ Dystrophic lakes have high levels of fulvic acids (21). ments), alcohols, sterols, triacylglycerols, free fatty acids,
§ ‘‘uncolored’’ water. and phospholipids (30), which can usually be regarded as
86 ACTIVITY AND CARBON TRANSFORMATIONS IN BIOFILMS

Air O2
Water CO2
Light
gradient

O2 CO2 CO2

O2, organic compounds

Diffusion
Alga Bacterium gradients
CO2

EPS
Figure 2. Carbon inventory of aboveground,
terrestrial plant matter (data from A. M.
Romani and S. Sabater, Freshwater Biol. 41(4), Substratum
729–736 (1999)).

autochthonous. The aqueous solubilities of many lipids are polyhydroxycarboxylates. These compounds account for 50
in the range of milligrams per liter (31); however, because to 75% of aquatic dissolved organic carbon (34) and differ
they are relatively hydrophobic, most lipids tend to parti- from other polymers in that they are the products of ran-
tion to air-water and solid-water interfaces, including par- dom chemical synthesis rather than intentional biological
ticles. Alternatively, the lipids may be released as particles processes. The long-standing assumptions of soil humic
or micelles. Lipids represent 10 to 25% of POC in marine chemistry were changed to reflect the fact that aliphatic,
surface waters. Interestingly, a large proportion (approx. rather than aromatic, compounds form the major car-
70–90%) can be found in the dissolved fraction (less than bon reservoir (35). Aquatic humic substances have always
10 kDa) (31). Free fatty acids are thought to be liber- been considered predominantly aliphatic and compara-
ated by microbial degradation of particulate substrates, tively low in molecular weight, as inferred from solubility
whereas the phospholipids are considered to be derived characteristics. Much recent information on the aliphatic
almost exclusively from cell membranes (32). Some lipids and aromatic nature of humic substances has come from
are more refractory than others, and migrate through the Nuclear Magnetic Resonance spectroscopy (NMR). NMR
water column relatively intact, where they are converted spectroscopy of humic substances usually involves sequen-
into alkanes at the sediment-water interface (30). A pre- tial extractions with sodium pyrophosphate. Pyrolysis
dominance of odd-carbon-number n-alkanes greater than in conjunction with mass spectrometry yields molecular
C23 H48 has been used as a marker for inputs of terrestrial fragments, which must then be reconstructed into a coher-
plant waxes to sediments. The uncommon predominance ent network. It is difficult to make specific statements
of even-carbon alkanes is considered to be an indication about the microbial degradation of humic compounds if
of specific microbial inputs, degradation of algal detritus, their chemical identities cannot be determined with any
and the reduction of alkanoic acids or other lipids (33). degree of certainty, or if they cannot be separated from
The degradation of lipids and triacylglycerols from algae biochemical compounds without altering their original
and terrestrial plants produces long-chain (nonvolatile) properties. Far from being inert and irrelevant to microbial
fatty acids, such as palmitic acid and stearic acid. These metabolism, humic substances can hinder the hydrolysis
compounds are poorly soluble in water and thus poorly of bound compounds and may act as electron donors for
bioavailable. Long-chain fatty acids predominate over some metabolic reactions (36).
short-chain forms in surface waters (oxidizing conditions),
especially in the hydrophobic air-water slick. Conversely, Anthropogenic Carbon Sources
microbial oxidation of POC and DOC produces volatile Increasing amounts of environmental carbon compounds
fatty acids, which accumulate only under reducing con- are misappropriated by human activity. These include
ditions. These conditions establish significant gradients vast quantities of untreated or partially treated sewage
within the biofilm matrix, but also more generally, such effluent, point inputs of pesticides and fertilizers, and
as in hypolimnetic waters and sediment cores. (other) xenobiotic carbon compounds. Not all of these
compounds are unique — for example, phthalate esters
Humic Substances
and halogenated organic compounds are often described
The bulk of DOC compounds are described as as undesirable anthropogenic compounds (37), yet natural
humic substances, which are complex and variable sources of these compounds can be significant. Phthalate
 ACTIVITY AND CARBON TRANSFORMATIONS IN BIOFILMS 87

derivatives are abundant in natural humic substances, general terms, ‘‘bioavailability’’ describes the accessibil-
as are organohalogens in emissions from volcanoes and ity or otherwise of a substrate to living microorganisms,
hydrothermal vents (38). Methane is produced in vast and can therefore be independent of the physiological
quantities from the fermentation of cellulose by the capability of the organism to degrade the substrate. Phys-
intestinal flora of termites and ruminant animals. Only ical accessibility is governed by parameters such as pH,
very preliminary estimates are available to predict the ionic strength, temperature, and water activity. These
total impact of anthropogenic carbon on total microbial parameters have a pronounced effect on factors such as
activity in terms of the global carbon cycle. It should solubility and partitioning. For example, an increase in
be noted, however, that the oceans act as a major temperature enhances the solubility and hydrolysis of
buffer system for carbon as soluble carbon dioxide, macromolecular substrates, thereby increasing their avail-
and that biofilms act as efficient buffer matrices for ability (44). Even for common macromolecular substrates
(nonxenobiotic) point disturbances. This property is such as proteins, the presence of appropriate enzyme sys-
exploited in wastewater treatment and bioremediation. tems is not a trivial issue. Some bacteria are apparently
unable to utilize high-molecular-weight substrates under
ASSIMILABLE ORGANIC CARBON AND BIOAVAILABILITY any conditions, suggesting the name ‘‘dissipotrophs’’ (45).
CONCEPTS Especially in aquatic environments, dissolved organic
matter can become more bioavailable by the action of
Descriptions of metabolic pathways give the impression solar radiation. Photochemical degradation of NOM pro-
that all heterotrophic activities are dependent on a duces carbonyl compounds, including carboxylic acids
limitless supply of low-molecular-weight precursors, such and aldehydes, and probably generates assimilable amino
as glucose. In aquatic environments, low-molecular-weight acids and carbohydrates (46,47). Ozonation has a similar
substrates have very low turnover times in the range effect on high-molecular-weight organic compounds dur-
of minutes to hours, and are therefore exceedingly rare ing the treatment of drinking water. The activity levels
(Table 2; 34). The major part of the dissolved organic of humic bound enzymes are also partially restored by
carbon pool comprises high-molecular-weight compounds UV photolysis (48). If particulate organic carbon in the
from primary production and decay processes. The euphotic zone of natural water contains chromophores,
utilization of high-molecular-weight organic compounds which strongly absorb solar radiation, then photolysis will
depends on the ability of an organism to degrade the have a major influence on bioavailability. A dual effect
polymer into its constituent oligomers or monomers. This, is observed when extensive absorption by humic com-
in turn, depends on the proximity of degradable substrate, pounds is sufficient to significantly raise the temperature
the production of appropriate exoenzymes, and the in shallow, brown-colored lakes and surface soils. Chem-
optimum size for passive transport through the cell wall. ical accessibility for a particular compound must also be
Proteins and polynucleotides are easily hydrolyzed, and considered. Different isomers or functionalized forms of
the constituent monomers are mineralized fully to CO2 or molecules can be comparatively labile or nonlabile. For
CH4 . Simple sugars are also easily hydrolyzed. Hydrolytic example, chitin is generally less bioavailable than the
activity is mostly confined to the biofilm matrix (39). deacetylated form, chitosan.
Structural polysaccharides constitute a considerably Particles are less easily degraded than individ-
larger pool of potential substrates, which is directly ual macromolecules; physical disintegration of particles
related to the resistance of these compounds to hydrolysis. increases the sites available for enzyme action. Adsorbed
The β-1,4 polysaccharide linkage is especially resistant macromolecules must be hydrolyzed before assimilation by
to degradation. Interestingly, this linkage may be the attached organisms. Surface immobilization may promote
key to the cohesiveness in some biofilms (40), and is an access of bacterial enzymes to the substrata, thereby facil-
intrinsic part of the holdfast structure of at least one, itating utilization. In some circumstances, immobilization
stalked bacterium (41). Suggestions that microorganisms can have the opposite effect. High-molecular-weight sub-
produce humic substances independently of lignin-type strates with extended conformations can have multiple
mechanisms are not convincing, since infrared spectra of binding sites that hinder hydrolysis, and the activities
aged extracts from laboratory cultures (42) resemble the of bacterial enzymes can be reduced or even inhibited
spectra of bacterial isolates (43). if they are immobilized by adsorption (49). Nucleic acids
The use of a compound as a metabolic substrate is and enzymes can be preserved through interactions with
linked intrinsically to the concept of bioavailability. In humic substances, organic sediment, or clay minerals (50).

Table 2. Concentrations and Turnover Times for Dissolved Free Amino Acids (DFAA),
Dissolved Free Carbohydrates (DFCHO) and Recalcitrant, ‘‘Humic’’ Substances in
Natural Waters (34)

Marine Freshwater
Substrate Concentration (nM) Turnover Time Concentration (nM) Turnover Time

DFAA 3–1,400 2.6–4,124

1.4 hrs–948 days 2 hrs–51 days
DFCHO 0.4–5,000 14–1,111
recalcitrant 70–90% of compounds, with turnover time of 2,000–6,000 years
88 ACTIVITY AND CARBON TRANSFORMATIONS IN BIOFILMS

Some inorganic particles, such as three-layered clay min- Table 3. Classes of Poorly Hydrolyzable, Aliphatic
erals can reduce the availability of citrate and glucose (51). Compounds and Their Origins
Similarly, bacterial degradation of herbicides such as (2,4- Compound(s) Origin
dichlorophenoxy)acetic acid (2,4-D) or insecticides such as
diquat can be inhibited by adsorption onto particles. In Algaenans Freshwater and marine algae
other words, some substrates must desorb before they can Cutans Cuticles of higher plants
be assimilated, a phenomenon described as ‘‘desorption Suberins Periderm tissue of Higher plants
limitation’’ (52). Ogram and coworkers (53) showed that Tegmens Inner seed-coats of Freshwater plants
in one system, the degradation of 2,4-D could only take Sporopollenins Spores and pollen grains
Polycadinenes Resins
place if the substrate was in solution. Of special relevance
to biofilm organisms is the fact that solid substrata may
also act as a metabolic substrate. Some common examples
are the bacterial degradation of agar, cellulose, wood, and of aliphatic products produced by the degradation of
poly (vinyl) chloride (PVC) in drinking water distribution phytoplankton. These nonhydrolyzable and insoluble
systems. Alternatively, the substrate may leach from the ‘‘algaenans’’ are components of the cell walls of many
solid phase and percolate through, for example, a soil col- algae. Algaenan-like compounds reportedly present in
umn. The water-soluble organic matter from leaf litter is cyanobacteria, and some bacteria are probably artifacts
a major source of biodegradable carbon. Removal of leaf resulting from the hydrolysis procedure used in their
litter at a test site yielded soil respiration rates one-third isolation (58). Thus, algaenans have so far only been
lower than at control sites (54). Cycling of interstitial car- reported from a relatively few species of photosynthetic
bon by biofilm organisms undoubtedly contributes much microorganisms. The comparative abundance of algaenan-
of the remainder, but the relationship between POC and containing organisms has not been investigated, so
DOC in the soil matrix is not clear. the cumulative effect of this class of compounds on
The major concept for bioavailability is the recalcitrance biogeochemical cycles is mostly unknown. It seems
of the bulk pool of NOM. Humic substances, according to likely, however, that aliphatic products from planktonic
classical definitions, are refractory to both chemical and algae contribute much of the humiclike substances in
biological degradation. This definition is problematic in marine environments. Table 3 shows a summary of these
that some humic compounds may be artifacts of the harsh waxy compounds and their origins. It is postulated
extraction procedures. Structurally related, but immature that algaenans comprise polymethylenic (n-alkyl) chains
compounds, such as lignin and cutin have been shown to cross-linked via ether bridges (23). The persistence of
degrade extensively during incubation experiments (55), these compounds in aquatic sediments has led to
and a measurable percentage of the operationally defined the suggestion of a humification route known as the
humic substances are readily biodegradable. Studies ‘‘selective preservation’’ pathway (59). Although, it might
that show ‘‘surprisingly’’ high levels of humic substance appear from these discoveries that the humic substances
degradation by biofilms seem to contradict the classical in marine environments are exclusively aliphatic, and
definitions. At least in freshwater environments, this therefore somewhat unrelated to freshwater or soil
discrepancy can be explained by the strong affinities of systems, many recalcitrant aromatic compounds are
carbohydrates and amino acids for the truly recalcitrant, present in these environments, and the origins of these
humic compounds (56), resulting in humic fractions, compounds have not been established fully. Lignin phenols
which contain large amounts of relatively assimilable are easily detected in estuaries, but quickly become
substrates. Unfortunately, it is difficult to abandon diluted in the marine environment (60). Other peculiar
classical definitions without comprehensive information but common forms of carbon are the condensed residues
on chemical structure, which is clearly unfeasible from incomplete combustion processes, known as ‘‘carbon
given the demands on analytical techniques, and the black’’ or ‘‘black carbon.’’ Highly aromatic compounds can
extensive variability of humic substances in different be produced from oxidized black carbon (61), suggesting
environments. More useful terms such as ‘‘refractory that humic substances in some environments can be
organic carbon’’ are becoming more common, although produced from charred biomass. Black carbon is extremely
it would be desirable if these terms discriminated resistant to biodegradation (62), which partly explains its
between biotic and abiotic processes. Measurements accumulation in many soil environments, and freshwater
of assimilable organic carbon (AOC) and biodegradable and marine sediments.
organic carbon (BDOC) are based on bioassays (57), which
therefore give operationally useful data concerning the MICROBIAL TRANSFORMATIONS OF DIVERSE AND
substrate. Conversely, these parameters are of limited DILUTE SUBSTRATES
benefit to the understanding of natural environments
if the microbial inoculum is not representative of the The biofilm mode of existence demonstrates a dependence
system (AOC), or if the physical and chemical parameters on transfer processes, or flux, to and from the biofilm.
represent ‘‘ideal’’ conditions for degradation, such as an The ecology of biofilm organisms is unique in terms
absence of adsorptive mineral phases. of the proximity of the attachment substratum, the
It has, only recently, been established that many comparatively high cell density, and the long residence
of the recalcitrant compounds in freshwater sediments time of the cells (63). There are several means by which
and possibly marine sediments are a distinct class biofilm cells can obtain nutrients: from the attachment
 ACTIVITY AND CARBON TRANSFORMATIONS IN BIOFILMS 89

substratum, which can be degradable, but always contains secondary production. Primary production is measured as
an adsorbed conditioning film (but see Schneider and Leis, reactants such as CO2 or H+ (pH), while the products are
CONDITIONING FILMS, this Encyclopedia); from sorption of usually measured as O2 and pH. Direct measurements
compounds (DOC or POC) onto biofilm cells and EPS; of carbon dioxide production are useful for estimates
or from degradation of adjacent biofilm components. of metabolism in whole systems, but do not provide
Geesey (64) suggested that an organism will not degrade any insight into transformations or partitioning of key
its own EPS. Biofilms treated with dissolved organic components. The use of radiotracers such as 14 CO2 is more
carbon derived from decomposing macrophytes contained specific. Radiolabeled, low-molecular-weight compounds
10 to 57% less EPS than control biofilms (65), suggesting are certainly not representative of the bulk of NOM,
that the amount of EPS, which is produced by biofilm and the use of labeled, high-molecular-weight compounds
organisms is closely related to the type and amount of depends on their commercial availability and their ability
DOC. Classical microbiology has centered on the concept to disperse within a system (69). Clearly, the use of any
of a ‘‘limiting’’ nutrient, and the ability of microorganisms standards for high-molecular-weight NOM is subjective.
to use only one carbon source at any one instant, as Microautoradiography can be used for the analysis of
exemplified by the diauxic growth of Entamoeba coli on labeled carbon dioxide assimilated by photosynthesis (9),
glucose and lactose. In natural systems, assimilable, low- but the visualization of biofilm cells in photographic
molecular-weight substrates are rarely present in high emulsions requires considerable technical effort to avoid
concentrations. Under these conditions, the formation artifacts.
of a biofilm would not take place if all cells were The fundamental prerequisite for studies of biofilm
specialized for the utilization of a single substrate. activity is the ability to distinguish between the metabolic
Furthermore, biofilm organisms cannot translocate to processes of biofilm and planktonic environments, which
nutrient sources as readily as planktonic organisms. in turn depends on the development of appropriate tech-
Carbon starvation or slow growth induces the expression niques for in situ measurements. The inadequacies of
of many catabolic systems, despite the absence of the culture-based methods for measuring growth processes
appropriate carbon source. This ‘‘catabolite repression’’ have been appreciated for some time, especially for
gives microorganisms an almost instantaneous ability biofilms. The purposeful detachment of microorganisms
to utilize a substrate should it become available, and from particles is a selective process, and the extrapola-
in general, serves to optimize an organism’s growth tion from total cell number to biomass is problematic
rate. In most bacteria, the enzymes involved in sugar because it is necessary to assume a universal conver-
transport and phosphorylation are important for signal sion factor to account for the variations in cell volume
transduction, although the regulation mechanisms vary among species, and among nutrient-limited phenotypes
widely between organisms (66). In one study, it was of the same species (32). Estimates of primary produc-
found that a strain of Pseudomonas aeruginosa was ers by measurement of chlorophyll a requires the same
able to form a biofilm in tap water with a mixture of assumptions for average biovolume. Measurements of
45 carbon compounds, each present at just 1 µg C L−1 , specific biochemical components is one strategy to avoid
whereas any one of these compounds in isolation could not culture-based techniques and microscopy, but there are a
produce growth at this concentration (67). The sorption number of assumptions inherent to these methods. Uni-
of particulate organic matter by biofilms represents a versal indicators of metabolism, such as ATP, can persist
major input of nutrients. These abiotic particles and after cell death (70), and the turnover times of excreted
flocs contain a relatively small proportion of easily compounds, which become localized in the biofilm matrix
biodegradable matter, but newly attached cells, and their can be totally dissimilar to the turnover time of these
metabolites and lysis products contribute to the cycling compounds in the bulk phase. Measurements of extra-
of organic matter. The turnover time for total organic cellular enzymes are complicated by the fact that the
carbon (TOC) in marine ‘‘snow’’ is approximately eight enzymes often bind to the EPS matrix, and therefore do
to nine days, suggesting that flocs are ‘‘hotspots’’ for not necessarily represent the presence of active cells (71).
microbial respiration (68). The biofilm environment is Humic compounds complex with microbial phosphatases,
therefore, a dynamic exchange interface, even when the thereby reducing enzyme activity (48). The same applies
activity levels of the biofilm organisms are observed to for the coenzyme, F420, which is found in nonviable
be low. cells of archaea and some eubacteria, as well as extra-
cellularly (72). Conversely, polar lipids such as the cell
MEASUREMENT OF TRANSFORMATIONS AND ACTIVITY wall phospholipids are degraded rapidly after cell death,
IN BIOFILMS with the added advantage of diversity profiles obtained
from fatty acid analysis (32). High levels of activity do
The ability of microorganisms to rapidly alter levels not necessarily imply an actively growing population.
of metabolic processes explains their ability to resume Cells in immature biofilm populations can be outnum-
activity after extended stages of dormancy. Microbial bered by planktonic organisms, yet the biofilm carbon
activity can be measured in terms of individual metabolic demand and hydrolytic enzyme activity can account for
reactions, or as a cumulative effect, as in the case close to 100% of the system’s total (73). Protease activ-
of biomass measurements. Approaches to determine ity can be measured with the protein substrate azocoll,
microbial activity include measurements of substrates, which yields a blue product on hydrolysis. The incorpo-
metabolic intermediates, or products of primary and ration of radiolabeled nucleotides into DNA allows an
90 ACTIVITY AND CARBON TRANSFORMATIONS IN BIOFILMS

estimate of DNA synthesis, and therefore an estimate that a lack of oxygen restricted limited active protein
of the rate of cell division. This is commonly measured synthesis to the top 30 µm of a biofilm (84). Other stud-
by the incorporation of 3 H-thymidine (74), although mea- ies have demonstrated stratifications in nitrifying and
surement of protein synthesis from the incorporation sulfate-reducing biofilms (85). Planar optodes are supe-
of radiolabeled amino acids such as 3 H-leucine is use- rior to conventional O2 microelectrodes because the lat-
ful as an independent validation, and is probably more ter create hydrodynamic disturbances, which can affect
sensitive because of the large amounts assimilated by an area of several millimeters surrounding the mea-
growing cells (2,75). Both methods are specific for het- surement site (86). Reporter gene techniques allow the
erotrophic bacteria, and contain a number of implicit identification of genetic changes, which are invoked on
assumptions. Measurements of the rates of nucleic acid recognition of an attachment substratum (87,88). These
synthesis from the incorporation of adenine, and perhaps changes include the production of exopolysaccharides (89),
also 35 SO4 into bacteria and algae provide a measure of and presumably other components of the EPS. Observa-
the entire microbial community. An alternative to radioac- tions relating to cellular morphology are less discriminate
tive isotopes is an immunoassay in which the uptake because they depend on an ability to recognize a partic-
of 5-bromo-2’deoxyuridine is measured by chemilumines- ular species within a complex biofilm community. This
cence (76). is only feasible with biofilms containing one or a few
Measurements of transformations and microbial activ- organisms, preferably with simple, distinguishing char-
ity within biofilms constitute a unique challenge because acteristics. The best examples are given by the initial
of the microscale heterogeneity within the biofilm matrix. attachment and subsequent morphological changes of the
Total cell counts of biofilm populations are hindered by the prosthecate (stalked) bacteria belonging to the genera
presence of aggregates, although confocal microscopy is an Hyphomicrobium, Flexibacter, Caulobacter, Prosthecobac-
indispensable tool in this respect. With the aid of suitable ter, and Hyphomonas (41). In studies of an Acinetobacter
fluorescent dyes, it is possible to observe relatively undis- sp. attaching to solid substrata, it was found that a
turbed biofilm populations as three-dimensional recon- nutrient-rich medium induced a transition from coccoid
structions, and to probe for features such as membrane morphology to a bacillary form, and the converse occurred
integrity and redox activity. Membrane integrity is con- when cells were supplied with a nutrient-depleted (star-
veniently tested with the Live/DeadTM kit (Molecular vation) medium (90). Figure 3 shows the variability in
Probes, Oregon). A potential complication of this tech-
morphology of a single species of marine bacterium, which
nique is that dye uptake could vary depending on the
size and charge of the dye molecules, and the uptake
systems of particular organisms. Further indications of
actively metabolizing cells are provided by the cleavage
of fluorogenic substrates such as fluorescein diacetate to
fluorescein, providing an indication of total esterase activ-
ity (77), and the reduction of tetrazolium salts by cellular
redox mechanisms, generating an insoluble, colored, and
perhaps fluorescent product. Changes in intracellular com-
ponents such as organic phosphorus and carbohydrate, or
NADH, can be measured by NMR spectroscopy and fluores-
cence spectrophotometry, respectively (14,78). Measure-
ments of small amounts of heat from metabolic activity
are possible using microcalorimetry (79). Measurements
of changes in the isotope compositions of carbon, nitro-
gen, and sulfur are possible using an isotope ratio mass
spectrometer (IRMS), since microorganisms discriminate
against heavy isotopes during metabolic processes (80).
This approach is especially useful for determining the
sources and fates of groundwater carbon, and its partition-
ing between biotic and abiotic phases. Isotope enrichment
is especially important for methanogenesis, acetogene-
sis, and methanotrophy. Microbially produced methane
represents the most 13 C-depleted carbon compound in
nature (80).
Microelectrodes allow the measurement of metabo-
lites, diffusion coefficients (81) and flow velocity (82) in
situ. Specific microelectrodes have been developed for
measurements of several important compounds, includ- Figure 3. Differences in the morphologies of dividing bacteria
ing nitrate, nitrite, ammonium, carbon dioxide, hydrogen (Psychrobacter sp. SW8) from a colony grown on minimal medium
sulfide, and methane, and fiber-optical sensors are avail- with acetate as the sole source of carbon. In continuous culture
able for measurements of O2 , pH and temperature (83). with carbon as the limiting nutrient, all cells of this isolate have
In P. aeruginosa biofilms, microelectrodes have shown a similar shape and indistinguishable surface morphology.
 ACTIVITY AND CARBON TRANSFORMATIONS IN BIOFILMS 91

was isolated for its ability to attach to hydrophobic sub- 12. H. Paerl, in C. J. Hurst et al., eds., Manual of Environmental
strata, thus implicating it as a primary colonizer of Microbiology, American Society for Microbiology Press,
immersed surfaces. Washington, D.C., 1997, pp. 252–262.
13. P. H. Burkill, R. J. G. Leakey, N. J. P. Owens, and R. F. C.
Mantoura, Deep-Sea Res. II 40(3), 773–782 (1993).
CONCLUSION 14. M. Fletcher, in L. F. Melo, T. R. Bott, M. Fletcher and
B. Capdeville, eds., Biofilms–Science and Technology, Kluwer
Academic, Dordrecht, Netherlands, 1992, pp. 113–124.
Sustainable, high levels of microbial productivity in low-
nutrient environments are testimony to the diversity and 15. R. K. Neely and R. G. Wetzel, Microb. Ecol. 30(3), 227–237
(1995).
adaptability of environmental microorganisms. Most of
this growth is associated with biofilms. Microorganisms 16. A. M. Romani and S. Sabater, Freshwater Biol. 41(4),
729–736 (1999).
have evolved many mechanisms to utilize available
17. O. Pringault, M. Kihl, R. Dewit, and P. Caumette, Microbiol-
substrates, only some of which are efficient. Biofilms
ogy 144(4), 1051–1061 (1998).
can sometimes account for nearly all of the enzyme
18. W. Stumm, Colloids Surf., A: Physicochem. Eng. Aspects 73,
activity in a natural environment. The bioavailable
1–18 (1993).
natural organic matter is rapidly utilized, leaving a
19. M. J. Vieira and L. F. Melo, Water Sci. Technol. 32(8), 45–52
recalcitrant macromolecular skeleton. While much is
(1995).
known about the mechanisms of substrate utilization,
20. T. I. Eglinton, J. J. Boon, E. C. Minor, and R. J. Olson, Mar.
and to a lesser extent, the genetic mechanisms regulating
Chem. 52, 27–54 (1996).
microbial activity, the largest knowledge gaps are the
21. E. M. Thurman, Organic Geochemistry of Natural Waters,
extent to which the poorly characterized pool of natural
Martinus Nijhoff/Dr W. Junk Publishers, Dordrecht, Nether-
organic matter is available for microbial growth, and lands, 1985.
the mechanisms controlling this bioavailability. Further 22. R. L. Malcolm, in A. J. Beck, K. C. Jones, M. H. B. Hayes and
knowledge of the bioavailable organic carbon must U. Mingelgrin, eds., Organic Substances in Soil and Water:
consider the flux of particulate substrate to biofilms, Natural Constituents and their Influences on Contaminant
the identification of biodegradable substrates using Behaviour, Royal Society of Chemistry, Cambridge, U.K.,
nondestructive techniques, and the measurement of 1993, pp. 19–30.
metabolic activity in situ. Such fundamental knowledge 23. I. M. Höld, S. Schouten, H. M. E. van Kaam-Peters, and
requires the use of complex biofilms and substrates, J. S. Sinninghe-Damsté, Org. Geochem. 28(3–4), 179–194
monospecies biofilms and fractionated organic matter, and (1998).
model systems capable of simulating key aspects of natural 24. J. I. Hedges, R. G. Keil, and R. Benner, Org. Geochem.
environments. 27(5–6), 195–212 (1997).
25. E. Tanoue, Mar. Chem. 51, 239–252 (1995).
26. M. McCarthy, J. Hedges, and R. Benner, Mar. Chem. 55(3–4),
BIBLIOGRAPHY 281–297 (1996).
27. K. J. Wilkinson, A. Joz-roland, and J. Buffle, Limnol.
1. J. C. Block in L. F. Melo, T. R. Bott, M. Fletcher, and Oceanogr. 42(8), 1714–1724 (1997).
B. Capdeville, eds., Biofilms–Science and Technology, Kluwer 28. J. Moreno, L. Penarrubia, and C. Garciaferris, Plant Physiol.
Academic, Dordrecht, Netherlands, 1992, pp. 469–485. Biochem. 33(1), 121–127 (1995).
2. G. Chin-Leo, in C. J. Hurst et al., eds., Manual of Envi- 29. S. Khan, P. J. Andralojc, P. J. Lea, and M. A. J. Parry, Eur.
ronmental Microbiology, American Society for Microbiology J. Biochem. 266(3), 840–847 (1999).
Press, Washington, D.C., 1997, pp. 263–271.
30. J. Laureillard, L. Pinturier, J. Fillaux, and A. Saliot, Deep-
3. P. A. del Giorgio and J. J. Cole, Annu. Rev. Ecol. Systemat. Sea Res. II 44(5), 1085–1108 (1997).
29, 503–541 (1998).
31. Q. Liu, C. C. Parrish, and R. Helleur, Mar. Chem. 60,
4. J. D. Bryers and W. G. Characklis in W. G. Characklis and 177–188 (1998).
K. C. Marshall, eds., Biofilms, John Wiley & Sons, New York,
32. D. C. White, H. C. Pinkart and D. B. Ringelberg, in C. J.
1990a, pp. 671–696.
Hurst et al., eds., Manual of Environmental Microbiology,
5. W. Manz et al., Microb. Ecol. 37(4), 225–237 (1999). American Society for Microbiology Press, Washington, D.C.,
6. T. Fenchel, G. M. King and T. H. Blackburn, Bacterial Bio- 1997, pp. 91–101.
geochemistry: The Ecophysiology of Mineral Cycling, 2nd ed., 33. V. O. Elias, B. R. T. Simoneit, and J. N. Cardoso, Naturwis-
Academic Press, San Diego, Calif., 1998. senschaften 84, 415–420 (1997).
7. S. Douglas and T. J. Beveridge, FEMS Microbiol. Ecol. 26, 34. U. Münster, Antonie van Leeuwenhoek 63, 243–274 (1993).
79–88 (1998). 35. M. H. B. Hayes, P. MacCarthy, R. M. Malcolm, and R. S.
8. A. Schneeberger, J. Frings, and B. Schink, FEMS Microbiol. Swift, in M. H. B. Hayes, P. MacCarthy, R. M. Malcolm, and
Lett. 177(1), 1–6 (1999). R. S. Swift, eds., Humic Substances II: In Search of Structure,
9. R. A. Long and F. Azam, Aquat. Microb. Ecol. 10(3), 213–221, John Wiley & Sons, Chichester, U.K., 1989, pp. 689–733.
(1996). 36. D. R. Lovley, J. L. Fraga, J. D. Coates, and E. L. Blunt-
10. R. S. Hanson and T. E. Hanson, Microbiol. Rev. 60(2), 439 Harris, Environ. Microbiol. 1(1), 89–98 (1999).
(1996). 37. C. A. Staples, D. R. Peterson, T. F. Parkerton, and W. J.
11. H. Weber, K. Kulbe, H. D. Chmiel, and W. Trosch, Appl. Adams, Chemosphere 35(4), 667–749 (1997).
Microbiol. Biotechnol. 19, 224–228, (1984). 38. G. W. Gribble, J. Nat. Prod. 55(10), 1353–1395 (1992).
92 ADENOVIRUSES

39. D. R. Confer and B. E. Logan, Water Res. 32(1), 31–38 (1998). 71. B. Frølund, T. Griebe, and P. H. Nielsen, Appl. Microbiol.
40. D. Mack et al., J. Bacteriol. 178(1), 175–183 (1996). Biotechnol. 43(4), 755–761, (1995).
41. S. E. Langille and R. M. Weiner, Appl. Environ. Microbiol. 72. M. A. Barlaz, in C. J. Hurst et al., eds., Manual of Envi-
64(8), 2906–2913 (1998). ronmental Microbiology, American Society for Microbiology
42. H. Claus, G. Gleixner, and Z. Filip, Acta Hydrochim. Hydro- Press, Washington, D.C., 1997, pp. 541–557.
biol. 27, 200–207 (1999). 73. D. C. Smith, G. F. Steward, R. A. Long, and F. Azam, Deep-
43. J. Schmitt and H.-C. Flemming, Int. Biodeter. Biodegr. 41, Sea Res. II 42(1), 75–97 (1995).
1–11 (1998). 74. C. Freeman and M. A. Lock, Sci. Total Environ. 138(1–3),
44. K. Kovárová-Kovar and T. Egli, Microbiol. Mol. Biol. Rev. 161–167 (1993).
62(3), 646–666 (1998). 75. S. M. Thomas and R. G. Wetzel, Microb. Ecol. 29(1), 63–70
45. L. V. Vasileva and G. A. Zavarzin, Microbiology 64(2), (1995).
198–202 (1995). 76. G. F. Steward and F. Azam, Aquat. Microb. Ecol. 19(1), 57–66
46. N. O. G. Jørgensen et al., FEMS Microbiol. Ecol. 25, 217–227 (1999).
(1998). 77. T. J. Battin, Sci. Total Environ. 198(1), 51–60 (1997).
47. J. Dahlen, S. Bertilsson, and C. Pettersson, Environ. Int. 78. C. Freeman, and M. A. Lock, Water Res. 29(3), 789–792
22(5), 501–506 (1996). (1995).
48. M.-J. Boavida and R. G. Wetzel, Freshwater Biol. 40(2), 79. M. A. Lock and T. E. Ford, in F. Megusar and M. Gantar,
285–293 (1998). eds., Perspectives in Microbial Ecology, Slovene Society for
49. H. Quiquampoix, P. Chassin, and R. G. Ratcliffe, Prog. Col- Microbiology, Ljubljana, Solvenia, 1986, pp. 367–374.
loid Polym. Sci. 79, 1–5 (1989). 80. E. L. Grossman, in C. J. Hurst et al., eds., Manual of Envi-
50. M. G. Lorenz and W. Wackernagel, Appl. Environ. Microbiol. ronmental Microbiology, American Society for Microbiology
53, 2948–2952 (1987). Press, Washington, D.C., 1997, pp. 565–576.
51. D. L. Jones and A. C. Edwards, Soil Biol. Biochem. 30(14), 81. E. E. Beuling, J. C. van den Heuvel, and S. P. P. Ottengraf,
1895–1902 (1998). Biotechnol. Bioeng. 67(1), 53–60 (2000).
52. M. C. M. van Loosdrecht, J. Lyklema, W. Norde, and A. J. B. 82. F. H. Xia, H. Beyenal, and Z. Lewandowski, Water Res.
Zehnder, Microbiol. Rev. 54(1), 75–87 (1990). 32(12), 3631–3636 (1998).
53. A. V. Ogram, R. E. Jessup, L. T. Ou, and P. S. C. Rao, Appl. 83. C. M. Santegoeds, A. Schramm, and D. DeBeer, Biodegrada-
Environ. Microbiol. 49, 582–587 (1985). tion 9(3–4), 159–167 (1998).
54. R. Jandl and P. Sollins, Biol. Fertil. Soils 25, 196–201 (1997). 84. K. D. Xu et al., Appl. Environ. Microbiol. 64(10), 4035–4039
55. S. Opsahl and R. Benner, Geochim. Cosmochim. Acta 59(23), (1998).
4889–4904 (1995). 85. P. L. Bishop and T. Yu, Water Sci. Technol. 39(7), 179–185
56. C. J. Volk, C. B. Volk, and L. A. Kaplan, Limnol. Oceanogr. (1999).
42(1), 39–44 (1997). 86. R. N. Glud et al., Aquat. Microb. Ecol. 14(3), 223–233 (1998).
57. P. M. Huck, J. AWWA, July, 78–86 (1990). 87. H. M. Dalton and P. E. March, Curr. Opin. Biotechnol. 9(3),
58. B. Allard, J. Templier, and C. Largeau, Org. Geochem. 26, 252–255 (1998).
691–703 (1997). 88. Sternberg et al., Appl. Environ. Microbiol. 65(9), 4108–4117
59. L. D. Stasiuk, Org. Geochem. 30, 1021–1026, (1999). (1999).
60. G. Kattner et al., Mar. Chem. 65(1–2), 25–39 (1999). 89. D. G. Davies and G. G. Geesey, Appl. Environ. Microbiol. 61,
860–867 (1995).
61. E. D. Goldberg, Black Carbon in the Environment, John Wiley
& Sons, New York, 1985. 90. G. A. James, D. R. Korber, D. E. Caldwell, and J. W. Coster-
62. L. Haumaier, and W. Zech, Org. Geochem. 23(3), 191–196 ton, J. Bacteriol. 177(4), 907–915 (1995).
(1995).
63. W. G. Characklis, M. H. Turakhia, and N. Zelver, in W. G.
Characklis, and K. C. Marshall, eds., Biofilms, John Wiley & ADENOVIRUSES
Sons, New York, 1990, pp. 265–340.
64. G. G. Geesey, Am. Soc. Microbiol. News 48, 9–14 (1982). CARLOS ENRIQUEZ
65. R. G. Wetzel, A. K. Ward, and M. Stock, Arch. Hydrobiol. Clorox Services Company
139(3), 289–299 (1997). Pleasanton, California
66. J. Stulke and W. Hillen, Curr. Opin. Microbiol. 2(2), 195–201
(1999).
As early as 1926, several outbreaks of swimming pool con-
67. D. van der Kooij, J. P. Oranje, and W. A. M. Hijnen, Appl.
Environ. Microbiol. 44, 1086–1095 (1982). junctivitis had been documented in the United States (1).
It is likely that some outbreaks were caused by aden-
68. H. Ploug, H. P. Grossart, F. Azam, and B. B. Jørgensen, Mar.
Ecol.: Prog. Ser. 179, 1–11 (1999). oviruses, as no bacterial pathogens could be identified,
and the description of clinical manifestations matched
69. G. M. King, in C. J. Hurst et al., eds., Manual of Environmen-
tal Microbiology, American Society for Microbiology Press, that of pharyngoconjunctival fever caused by aden-
Washington, D.C., 1997, pp. 317–323. oviruses (2–6). Strong epidemiological evidence of the
70. D. M. Karl, in P. F. Kemp, B. F. Sherr, E. B. Sherr, and waterborne transmission of these agents was obtained
J. J. Cole, eds., Handbook of Methods in Aquatic Micro- in 1955, when adenoviruses were isolated from patients
bial Ecology, Lewis Publishers, Boca Raton, Fla., 1993, suffering form conjunctivitis during an outbreak that orig-
pp. 359–368. inated in a swimming pool (3). Numerous waterborne
 ADENOVIRUSES 93

outbreaks of pharyngoconjunctival fever, caused by ade- GENETICS

noviruses, have been documented since then (5,7–12).
The first detection of adenoviruses during a waterborne Evolution and diversity of adenoviruses result mostly from
outbreak of gastroenteritis was recorded in Finland in recombination among members of the same subgroup;
April 1994 (13), when approximately 25 to 50% of the however, recombination among adenoviruses of different
population (1,500 to 3,000) of a municipality developed subgroups can also occur. In fact, adenovirus type 4
acute gastroenteritis. This outbreak was associated with (the only member of group E) originated, apparently,
contaminated well water. In addition to adenoviruses, from recombination between viruses from groups B and
a Norwalk-like agent, small round viruses (SRV), and C (16).
group A and C rotaviruses were identified as the eti-
ological agents, Norwalk virus being the most preva-
BIOLOGY
lent of the implicated viruses. In addition to gastroen-
teritis (adenovirus types 31,40, and 41) and pharyngo-
Adsorption and internalization in adenoviruses is a two-
conjunctival fever (adenovirus types 3,7,8,11,14,19, and
step process. Attachment of adenovirus to susceptible
37), adenovirus infections may cause other clinical con-
host cells is mediated by the fiber protein; however,
ditions such as acute respiratory disease (adenovirus
for internalization to occur, the penton base protein
types 1–7, 14, and 21), acute hemorrhagic cystitis (ade-
needs to attach also to a cell receptor of the family of
novirus types 11 and 21), and acute respiratory dis-
‘‘integrins.’’ The penton–integrin interaction precipitates
ease (ARD) of military recruits (types 3, 4, 7, 14, and
a receptor-mediated endocytosis. Once the virion enters
21) (14).
the cell, it is disassembled by a systematic elimination
of structural proteins, leading to the transport of the
CLASSIFICATION viral DNA into the cell nucleus (16). The adenovirus
genome replicates by an unusual mechanism in which
Adenoviruses were first identified in 1953 as agents single-stranded (ss) DNA intermediates are formed, and a
causing spontaneous degeneration of primary cultures protein, but not short RNA molecules serves as a primer
of human adenoid tissue (15). In 1956, the name ade- to initiate replication (17). Adenovirus DNA replication
novirus was given to these agents denoting the tis- can start at one or both ends of the molecule. A similar
sue in which these viruses were discovered. Aden- general organization has been found in the genomes of
oviruses belong to the Adenoviridae family, which is different adenoviruses. The adenovirus ds-DNA molecule
represented by the Mastadenovirus and the Aviaden- possesses two identical origins for replication, which
ovirus genera (16). The Mastadenovirus genus contains is found in each of the virus short terminal repeats.
human (49 serotypes) (14), simian (27 serotypes), bovine Adenovirus mRNA synthesis takes place in the cell
(10 serotypes), equine (1 serotype), porcine (4 serotypes), nucleus (16) adenovirus genes are expressed in two
ovine (1 serotype), and canine (3 serotypes) viruses. The stages, early transcription (before viral DNA synthesis),
Aviadenovirus genus (11 serotypes) includes only bird and late transcription (after viral DNA synthesis);
viruses (4). The 49 human serotypes have been fur- however, the functional distinction of the two stages
ther classified into six subgroups (A through F) is frequently obscure. A cellular RNA polymerase II
based on their ability to agglutinate red blood transcribes each adenovirus gene into multiple mRNAs
cells, capacity to cause tumors in rodents, relat- that are differentiated by alternative splicing (18). The
edness of tumor antigens, electrophoretic mobility accumulation of large quantities of viral proteins and
of viral proteins, genome homology, cross hybridiza- DNA induces virus assembly. The formation of empty
tion, and DNA digestion with restriction endonucle- capsids seems to be followed by the internalization of
ases (16). the DNA. Virus release is mediated by disruption of
cytoskeleton intermediate filaments, which renders the
STRUCTURE cell more susceptible to lysis (16).

Adenoviruses are nonenveloped icosahedric viruses of

EPIDEMIOLOGY AND CLINICAL DISEASE
approximately 70 to 100 nm in diameter, composed of DNA
(13% of mass), protein (87% of mass), and trace amounts of
Clinical Symptoms
carbohydrates present in the virion fibers (16). Members
of the Adenoviridae family possess a linear double- Most adenovirus human diseases are associated only
stranded (ds) DNA molecule with about 30,000 base with one-third of adenovirus types. Although many
pairs (17) and molecular weight of 1.5–1.8 × 108 (14). adenovirus infections are subclinical, these viruses
Virions consist of a protein shell (capsid) surrounding may cause acute respiratory disease (types 1–7,14,
a DNA-containing core. This structure is composed by and 21), conjunctivitis (types 3,7,8,11,14,19, and 37),
252 subunits (capsomeres), 240 hexons, and 12 pentons. acute hemorrhagic cystitis (11 and 21), acute respiratory
Each penton contains a projecting fiber that varies in disease (ARD) of military recruits (types 3,4,7,14, and 21),
length in different adenovirus serotypes. All human and gastroenteritis (types 31,40, and 41) (Table 1; 14).
adenoviruses seem to possess a single fiber protein, with Enteric adenoviruses 40 and 41 have been recognized
the exception of the enteric adenoviruses 40 (Eda 40) and as important etiological agents of gastroenteritis in
41 (Ead 41) that possess two fiber proteins (16). children (2,19–25), second in importance only to rotavirus.
94 ADENOVIRUSES

Table 1. Disease Caused by Human Adenoviruses Adenovirus in AIDS patients

Subgenus Serotype Human Illness Adenovirus may cause opportunistic infections in immuno-
compromised patients, including individuals suffering
A 12 Meningoencephalitis from AIDS (29). Although uncommon, these infections may
18,31 Gastroenteritis
result in a high mortality (30). Several clinical conditions
B 3 Acute febrile pharyngitis; associated with adenovirus infections in immunocompro-
adenopharyngo- mised patients have been described. These conditions
conjunctival fever;
include encephalitis (31,32), gastroenteritis (33,34), liver
pneumonia; follicular
conjunctivitis; fatal
necrosis (35), nephritis (36), parotitis (29), hemorrhagic
infection of neonates cystitis (37,38), and fatal pneumonia (39–41).
7 Acute febrile pharyngitis;
adenopharyngo- Modes of Transmission
conjunctival fever; acute
respiratory disease with Adenoviruses can replicate in the respiratory tract, the
pneumonia; fatal eye mucosa, the intestinal tract, the urinary bladder, and
infection of neonates; the liver. Adenoviruses can gain access into susceptible
meningoencephalitis individuals through the mouth, the nasopharynx, or the
11 Follicular conjunctivitis; conjunctiva. Although initial infection may occur through
hemorrhagic cystitis in the respiratory route, the fecal-oral transmission accounts
children for most adenovirus infections in young children due
21 Hemorrhagic cystitis in to prolonged shedding of large numbers of viruses in
children; fatal infection
feces (14). This may explain the high rate (close to 50%) of
in neonates
14,16 Acute respiratory disease adenovirus secondary transmission in households (42). In
with pneumonia addition to the type of infection, the duration of adenovirus
34,35 Acute and chronic infection shedding depends on the age, and immune status of the
in patients with infected individual. Adenovirus can be isolated for one to
immunosuppression and three days from the throat of adults suffering from the
AIDS common cold, and for three to five days from nose, throat,
C 1,2,6 Acute febrile pharyngitis; stool, and eye of adults with pharyngoconjunctival fever.
pneumonia in children At a younger age, adenovirus excretion in throat or stool is
5 Acute febrile pharyngitis; more prolonged than in adults. In children suffering from
pertussis-like syndrome; respiratory or generalized infection, adenoviruses can be
acute and chronic recovered from throat or stool for three to six weeks, and
infection in patients with
in immunocompromised patients for up to 12 months or
immunosuppression and
AIDS
longer from urine, stool, throat, or organ biopsies. In some
individuals, intermittent shedding of adenoviruses may
D 8,19,37,9,10,13, Epidemic
occur for months or even years (14).
15,17,42,19, keratoconjunctivitis
20,22–29
Most, if not all, waterborne outbreaks of viral
30 Fatal infection in neonates conjunctivitis are caused by adenoviruses. Documented
32,33,36,38 Asymptomatic infections waterborne outbreaks of conjunctivitis, by adenovirus
39,42–47 Acute and chronic infection type 3 (3,8,43) and type 4 (5) have been reported; however,
in patients with adenovirus types 1,7, and 14 have also been associated
immunosuppression and with family outbreaks of pharyngoconjunctival fever (44).
AIDS According to the Centers for Disease Control (45), the
E 4 Respiratory infection only viral waterborne outbreak between 1991 and
F 40,41 Diarrhea 1992 in the United States, associated with recreational
waters, was caused by adenovirus type 3. In this
Adapted from Maier, Pepper, and Gerba (107) outbreak, conjunctivitis, pharyngitis, and fever were the
most common clinical manifestations among 595 affected
individuals (8).
Children infected with enteric adenoviruses can shed up
The leading routes of transmission of adenoviruses are
to 1011 particles/g in feces (26). direct contact via host-to-host, and indirect contact via
An electron microscopy survey of stool samples in fecally contaminated water, food, and fomites (46). Intact
London, from patients suffering from viral gastroenteritis, conjunctival mucosa seem to be resistant to infection by
showed that enteric adenoviruses were found in 14% of adenoviruses.
the examined specimens (27). Enteric adenoviruses are The enteric adenoviruses 40 and 41 have been recog-
ubiquitous, and antibodies to these viruses have been nized as the second most important etiological agents of
found throughout the world (28). An investigation in gastroenteritis in children (21). These viruses, in contrast
Guatemala (25) showed that Ead 40 and Ead 41 were to other adenoviruses, are not shed in respiratory secre-
associated with diarrheal episodes in ambulatory children tions (47,48), thus their transmission is likely to be limited
three times more often than rotaviruses. to the oral-fecal route. It has been suggested that among
 ADENOVIRUSES 95

the adenoviruses, Ead 40 and 41 are the most likely to be room temperature was longer than that observed for a
transmitted by water (49), probably through recreational laboratory strain (61).When the same study compared the
or drinking water (50). This idea was supported by find- survival of adenovirus type 5 with that of poliovirus 1 in
ings by researchers from the United Kingdom (51), who tap water at 4 and 22 ° C, it was found that poliovirus 1
reported a gastroenteritis outbreak in which a common was inactivated faster at both temperatures.
unknown source of enteric adenoviruses affected several Adenoviruses seem to be more stable than enteroviruses
children from a long-stay ward at the same time. under different relative humidity conditions. A compara-
Not only are the enteric adenoviruses a significant tive study on the survival of several viruses at different
cause of gastroenteritis among the enteroviruses, aden- temperatures and relative humidity values, reported that
ovirus type 31 has been increasingly detected during the adenovirus type 2 survived longer than poliovirus 2, vac-
last few years as an important cause of infant gastroen- cinia virus, coxsackievirus B3, and herpesvirus, at relative
teritis (52). humidity values ranging from 3 to 96%, at either 25
Contaminated inanimate surfaces may also play a or 37 ° C (53). In a study that compared the stability of
significant role in adenovirus transmission because of their enterovirus 70 and human adenovirus type 19 at room
resistance to drying. At room temperature, adenovirus 2 temperature (63), it was observed that the drying of
survives for 8 and 12 weeks under low (7%) and high (96%) these agents on ophthalmic instruments resulted in a
relative humidity, respectively. This virus was more stable 5-log10 reduction of the titer of enterovirus 70, whereas
under these conditions than poliovirus 2, vaccinia virus, the titer of adenovirus 19 decreased by less than 1 log10
coxsackievirus B3, and herpesvirus (53). after 11 days. A study on the survival of hepatitis A virus
(HAV), human rotavirus (HRV), polio 1, and Ead 40 on
Reservoirs porous and nonporous materials showed that when dried
on these materials, HAV and HRV were more stable than
Although adenoviruses are common in many types
either polio 1 or Ead 40; however, Ead 40 survived longer
of animals, including simian species (14), adenoviruses
in porous surfaces such as paper or cotton cloth than in
infecting humans seem naturally limited to humans (46).
nonporous surfaces such as latex, aluminum, china, and
glazed tiles (64). The presence of fecal matter affected the
OCCURRENCE AND PERSISTENCE IN THE ENVIRONMENT survival of Ead 40, depending on the type of material.
On nonporous surfaces, this agent persisted longer in the
A greater number of adenoviruses than enteroviruses presence of fecal matter, whereas in porous materials the
has been consistently found in raw sewage around the presence of feces resulted in a diminished infectivity of
world (50,54–58). Results of a comparative study of this agent (64).
cytopathogenicity, immunofluorescence and in situ DNA When the survival of the enteric adenoviruses 40 and
hybridization as methods for the detection of adenoviruses 41 was compared with the survival of poliovirus type 1
from water have suggested that 80% of infectious aden- (polio 1) and HAV (in tap water at 4 ° C, and at room
oviruses in raw sewage may be enteric adenoviruses (50). temperature, and with the survival of polio 1 in primary
Adenoviruses are also commonly found in primary sewage and secondary wastewater at 4 and 15 ° C, and with
sludge, where they have been found in concentrations ten seawater at 15 ° C), it was observed that the survival of
times greater than that of the enteroviruses (59). In con- Ead 40 and Ead 41 in primary and secondary wastewater
trast to these reports, a five-year study of river water was slightly greater than that of polio 1. However, in tap
in Japan (60) indicated that adenoviruses were present and seawater, the enteric adenoviruses were substantially
throughout the duration of the study, but in lower number more stable than either polio 1 or HAV (65).These results
than enteroviruses or reoviruses. The peak concentration are in agreement with previous research in which it was
of reoviruses and enteroviruses was observed in winter reported that adenovirus type 5 survived longer than both
and summer, respectively, whereas the number of aden- poliovirus 1 and echovirus 7 in tap water, at either 4 or
oviruses remained constant throughout the year. 18 ° C (61), and that HAV in groundwater was more stable
Adenoviruses seem to be relatively stable in the aquatic than polio 1 at different temperatures (66).
environment. Adenovirus type 5 survives longer in tap Additional evidence of the stability of the enteric
water, at either 4 or 18 ° C, than either poliovirus 1, adenoviruses at higher temperatures was obtained in an
or echovirus 7 (61). In contrast to the survival patterns experiment in which both Ead 40 and polio 1 were held
observed with enteroviruses, the enteric adenoviruses at 50,65, and 80 ° C in phosphate buffered saline. After
do not survive significantly longer in wastewater. The incubation at 50 ° C for six minutes, polio 1 lost 0.88 log10 ,
increased survivability of the enteric adenoviruses in tap whereas Ead 40 lost less than 0.2 log10 . Similarly, at
and seawater and their faster inactivation in sewage may 65 ° C, polio 1 lost more than 2.5 log10 in 30 seconds,
indicate that these viruses are inactivated by mechanisms whereas Ead 40 lost only 1 log10 after the same incubation
that are different from those affecting the enteroviruses. time. When the thermal inactivation was conducted at
The enteric adenoviruses are more thermally stable than 80 ° C polio 1 lost more than 4 log10 in 30 seconds, in
poliovirus type 1 (polio 1), which is inactivated faster at contrast, adenovirus 40 lost only 1.8 log10 during the same
temperatures above 50 ° C (62). period (62).
Differences in survival among members of the same In addition to thermal stability, Ead 40 shows some
adenovirus type have been reported. The survival of two resistance to extreme pH levels. No reduction of its
clinical isolates of adenovirus type 5 in tap water at infectivity in 0.1 M glycine buffer was observed after
96 ADENOVIRUSES

Table 2. UV Doses for 90% and 99.99% A 99% reduction of human adenoviruses 3 and 7 was
Inactivation of Test Viruses achieved with a concentration of free chlorine of 0.69 and
Inactivation Dose 0.89 mg/L, respectively. In contrast, a similar disinfection
(mW sec/cm2 ) level of the simian adenoviruses M2,M3, and M4 was only
Relative
obtained with 0.93 mg/L of residual chlorine. Disinfection
Virus Type 90% 99.99% Sensitivity
experiments with HAV, human rotavirus, poliovirus 1,
Ead 40 30 124 1.0 and adenovirus 5 in tap water by copper and silver ions
Ead 41 23.6 111.8 1.16 and reduced levels of free chlorine (1. 0.5 or 0.2 mg/L)
MS-2 14 65.2 1.8 showed that adenovirus type 5 was more resistant to
PRD-1 8.7 31.6 4.08 all disinfection treatments than poliovirus 1 but less
Polio 1 4.1 21.7 5.6 resistant than either HAV or human rotavirus (72).
Similar results were reported in a study that estimated the
Adapted from Meng and Gerba (69)
time required to eliminate adenovirus 40 from artificially
contaminated mussels in a continuous flow of ozonated
45 minutes at either pH 3.5, 9.5, or 10. However, after marine water. This investigation showed that it took much
45 minutes at pH 10.5, Ead 40 underwent a titer reduction longer to eliminate Ead 40 than poliovirus 1. However,
of 2.7 log10 . These results are in contrast to those reported it was depurated faster than either HAV or human
earlier (67) in which the infectivity of adenovirus type 5 rotaviruses (73).
decreased steadily at pH higher than 9.0, with rapid
inactivation at pH 10.
ENVIRONMENTAL STABILITY OF ADENOVIRUSES

ULTRAVIOLET DISINFECTION The increased resistance showed by enteric adenoviruses

when compared with other enteric viruses may be
Differences in the stability to inactivating factors among associated with the double-stranded nature of their DNA,
adenovirus types were first reported in 1962 (68). Research which if damaged, may be repaired by the host cell DNA-
on the inactivation of several adenoviruses by ultraviolet repair mechanisms, which in human cells, in addition
irradiation showed that the adenoviruses represent a to repairing pyrimidine dimers can also repair a wide
very diverse group. Adenovirus 4 and 20, for instance, range of DNA damages (74). In addition, both adenovirus
were more resistant to inactivation by ultraviolet (UV) DNA strands may serve as template for replication (75).
light than adenovirus 1. Adenovirus 4 and 20 had a Thus, if one strand is damaged by environmental factors,
titer reduction of approximately 2.5 log10 . In contrast, the other may still serve as template for replication of
adenovirus type 1 lost almost 3 log10 during the same progeny DNA. This mechanism would not be effective
period. In another study, when the enterovirus 70 on ssRNA-genome viruses such as polio 1 or HAV. It has
and human adenovirus type 19 were exposed to UV been suggested that nucleic acids from DNA viruses would
irradiation, it took less than 10 mW/cm2 /min to cause persist longer in the aquatic environment (76) because
a 4.5 log10 reduction of the enterovirus 70 titer, whereas indigenous DNAases require cofactors to be active and are
a UV dose of 60 mW/cm2 /min was necessary to reach denatured by temperature more readily than RNAases.
similar levels of inactivation of adenovirus 19 (63). It could be speculated then, that the longer survival of
These results agree with a previous investigation that the enteric adenoviruses in tap and seawater might have
compared the resistance of enteric adenoviruses to UV been associated with DNA damage that could have been
disinfection with poliovirus type 1 and coliphages MS-2 repaired by the host cell, whereas the faster inactivation in
and PRD-1 (Table 2; 69). Purified stocks of the viruses sewage might have resulted from protein capsid damage,
in deionized-sterile water were exposed to collimated rendering the virions unable to attach and enter the
UV irradiation in a stirred reactor for a total dose host cell.
of up to 140 mW sec/cm2 . The doses of UV to achieve
a 90% inactivation of adenovirus 40, adenovirus 41,
coliphages MS-2 and PRD-1, and poliovirus type 1 CONCENTRATION AND RECOVERY
were 30,23.6,14,8.7, and 4.1 mW sec/cm2 , respectively.
Although MS-2 is considered to be much more resistant to Although Ead 40 and 41 are important human
UV disinfection than enteroviruses and enteric bacteria, pathogens (19), methods for their concentration in water
and has been suggested as a good model for enteric viruses are very limited (77). However, some research has been
disinfection (70), adenovirus 40 was significantly more conducted on the concentration of adenoviruses (67,78–82)
resistant to UV irradiation than this coliphage. Therefore, and enteric adenoviruses (77).
adenoviruses appear to be among the most resistant The earliest attempts to concentrate adenoviruses from
waterborne enteric viruses to UV light disinfection (69). water were conducted in 1965 (78). Sanitary napkins
were suspended for two to three days in sewage or
sewage effluents. In the laboratory, the napkins were
CHEMICAL DISINFECTANTS wrung, the liquid centrifuged, and the resuspended pellet
inoculated into different combinations of primary cultures
Differences in the susceptibility to chlorine among of African green monkey kidney cells (MK), human
members of the adenovirus group have been reported (71). amnion cells (HA), and a continuous green monkey kidney
 ADENOVIRUSES 97

cell line (BCS-1). Using this procedure, 96 adenovirus different even among different serotypes of the same
isolates were obtained from 661 samples positive for virus (86).
viruses. In a study that compared two aqueous-polymer-phase
The concentration of adenovirus was improved by a separation methods, as a second step of adenovirus recon-
protamine-sulfate precipitation procedure (79).Using this centration, it was demonstrated that a dextran-methyl
method, a 50- to 400-fold concentration of adenovirus was cellulose-polyethylene glycol phase separation method
achieved. It was observed that adenoviruses were pref- was superior to a sodium dextran sulfate-polyethylene
erentially precipitated when mixed with enteroviruses, glycol sodium chloride procedure (67). Furthermore, the
and it was pointed out that the protamine-sulfate pre- dextran-methyl cellulose-polyethylene glycol phase sepa-
cipitation method could be used to prevent overgrowth ration method selectively concentrated adenovirus 5 when
of adenoviruses by faster replicating enteroviruses. This mixed with poliovirus 2 (16 : 2 ratio).
methodology was used successfully to detect indigenous As an alternative to the organic flocculation method,
adenoviruses from raw sewage in Greece (55). polyethylene glycol (PEG) precipitation has been used
Electronegative epoxy-fiberglass filters and fiberglass to concentrate Ead 40 from beet extract (BE). The
textile filters have been used as primary adenovirus- reconcentration efficiency by this method (40%) was
adsorbing filters. Elution of adenovirus 5 from fiberglass similar to that obtained by organic flocculation (38%).
filters was more efficient with 3% beef extract pH 9.0 PEG precipitation may offer some advantages over the
than with 0.05 M glycine pH 11.5, which resulted in organic flocculation procedure to obtain high-quality RNA
inactivation of this agent (67). An adenovirus high-pH
for RT (reverse transcriptase)-PCR analysis, especially
sensitivity was observed in another study in which the
when the number of viruses in the sample is low (87).
simian adenovirus SV-11 infectivity was greatly reduced
at pH 10 or above (83).
Elution of viruses from adsorbing materials is often
achieved by the use of high pH solutions (28). However, METHODS OF DETECTION
when adenoviruses are exposed to an eluting solution of
pH 11.5, more than 90% of the viruses are inactivated Several methods have been used to detect adenoviruses.
in less than one minute. Enteroviruses, on the other Clinical specimens for detection may include feces (88),
hand, can be reconcentrated with an overall efficiency of nasal washes, and pharyngeal, anal, and conjuncti-
15.4% (81). This suggests that the enteroviruses may not val swabs (42); whereas environmental samples consist
be good surrogates for adenovirus recovery from water, and mostly of water (58) and wastewater (58,59). Adenoviruses
that failure to recover adenoviruses from water samples in these types of samples are relatively stable and will
may be due, in part, to virus inactivation during elution of withstand freezing and storage at −70 ° C (14). Although
filters with high pH solutions. adenovirus isolation by cell culture is the procedure
In a study to evaluate the ability of microporous filters used by most diagnostic and reference laboratories, it
to recover Ead 40 from different types of water, it was requires highly trained personnel and specialized facil-
found that with electronegative filters the recovery of ities. Therefore, the enzyme immunoassay (EIA) is the
this virus from drinking water was of 36%, whereas with method of choice for the detection of adenoviral sol-
electropositive filters, the average recovery efficiency was uble antigens in feces or respiratory secretions (42).
26.5% (77).These results suggest that adsorption Ead 40 Other methods used for the detection of adenoviruses
to electropositive filter surfaces at neutral pH may be include direct electron microscopy (89), immune electron
less affected by specific electrostatic characteristics of the microscopy (90), immunofluorescence (21), enzyme-linked
virions than their attachment to electronegative surfaces immunosorbent assay (ELISA) (91), gel electrophoresis,
at pH 3.5, and that the higher recovery of this virus with and restriction analysis of viral DNA (92), nucleic
electronegative filters may have been associated with a acid hybridization (93,94), and polymerase chain reac-
more efficient elution of the virus from the electronegative tion (PCR) (9,57,58,88). With the exception of PCR, detec-
filtering medium (77). tion by these techniques is only possible if a relatively
large number of viral particles are present in the sam-
Reconcentration of Adenoviruses ple.
Although the filter-adsorption-elution procedure allows for Enteric adenoviruses propagate poorly, or not at
the processing of large samples of water (84), the volume all, in conventional human cell lines in which other
of the eluate still needs to be reduced to be assayed adenoviruses can be propagated effectively (95). The
in a practical way. Among the methods for second-step enteric adenoviruses 40 and 41 are even difficult to grow
reconcentration, the organic flocculation procedure (85) in fetal intestinal organ cultures (96). However, in 1992
continues to be the most efficient (86). The organic the successful propagation of both Ead 40 and Ead 41
flocculation method (85) is widely used to reconcentrate in PLC/PRF/5 human liver cell line was reported (97).
viruses from environmental water samples. It has been Later, these agents were also propagated in a cell
used efficiently for the recovery of Simian adenovirus SV- line derived from a human colon carcinoma (CaCo-2
11, with recoveries of 50 to 60% (80). A limitation of this cells) (98). The availability of these cell lines has provided
technique, in addition to extreme pH changes, is that it is a practical method for the propagation and detection of
not equally effective with all viruses. The reconcentration Ead 40 and Ead 41 to study these viruses in the aquatic
efficiency by the organic flocculation method, may be environment (62,65).
98 ADENOVIRUSES

Purification and Separation samples of healthy patients. It has been suggested

that PCR using these primers may be detecting latent
Several continuous cell lines have been used in the
adenoviruses of different types because EIB primers are
propagation of many types of human adenoviruses. Some
usually negative in normal individuals (14).
of these include Graham 293, HeLa, Hep-2, KB, and A549
cells (99). Propagation of the enteric adenoviruses 40 and
41 has been accomplished in PLC/PRF/5 human liver Oligoprobe Hybridization and Restriction Endonuclease
cell line (97) and CaCo-2 cells (98). Adenovirus cytopathic Digestion. Adenoviral DNA can be characterized by oligo-
effect (CPE) is distinguished by large pycnotic cells that probe hybridization (59,94,104) and restriction endonucle-
eventually aggregate into sheets that detached from the ase digestion (52,95). In contrast to the relatively small
culture vessel (99). regions of the genome analyzed by either hemaggluti-
Adenoviruses replicate in the nucleus of infected nation inhibition (HI) or serum neutralization (SN) tests,
cells, and most virions remain cell-associated; there- restriction endonucleases recognize and cleave DNA at
fore, the release of viral particles can be achieved by specific sequences throughout the viral genome. The
disruption of the nuclear membrane by freezing and sensitivity of restriction endonuclease digestion can be
thawing, sonication, or by extraction with an organic increased by Southern blotting of the DNA restriction
solvent such as Freon (100). Because of environmen- digests (105). Depending on the hybridizing probe, DNA
tal concerns resulting from the use of Freon, other from adenovirus isolates can be identified and classified
organic solvents have been proposed as an alternative by group or type (14). Hybridization of adenoviral DNA
for the extraction of nonenveloped viruses (101). After from fecal extracts by the direct-spot method has been
removing cell debris, virions can be purified by density successfully used to identify enteric adenoviruses from
gradient centrifugation (100). Depending on the aden- clinical specimens from patients suffering from gastroen-
ovirus type, titration of can be accomplished by the teritis (106).
plaque method or by an end point CPE method such
as the tissue-culture-infectious dose fifty (TCID50 ) (65).
CONCLUSION
Although the plaque method is more accurate, it requires
lengthy incubation periods and a very stable cell mono-
Adenoviruses are important human pathogens causing an
layer (100). Furthermore, not all adenovirus types respond
array of clinical conditions ranging from asymptomatic
to plaquing (97). In contrast, the end point CPE method
infections to fatal illness. These viruses are ubiquitous in
is less sophisticated than the plaque method, but
sewage-contaminated water and are relatively resistant to
still sensitive and accurate enough for most applica-
tions (100). environmental stresses. Although their impact on human
health, from the environmental point of view in particular
Assay as a waterborne organism, has not been fully assessed,
evidence suggests that these agents may represent a
Immune Assays. The enzyme immunoassay (EIA) is significant hazard to public health. Further research
the method most commonly used for the detection of on detection methods and disinfection procedures could
adenoviral soluble antigens in feces, or in respiratory provide the tools to assess the risk that adenoviruses
secretions (42,91). Monoclonal antibodies (Mab) directed pose for the population more objectively and also to
to a family-shared hexon epitope can be used in EIA implement actions to reduce their threat to public
to identify any member of the Adenoviridae family. If health.
needed, a type-specific Mab can be utilized to further
determine the precise adenovirus type (42). All known
mastoadenoviruses share a family antigen that can BIBLIOGRAPHY
be detected by the Complement-Fixation (CF) (102) and
1. C. A. Bahn, New Orleans Med. Surg. J. 79, 586–592 (1927).
immunofluorescence (IF) tests (14). The hemagglutination
inhibition (HAI) and serum neutralization (SN) tests can 2. M. J. Albert, Arch. Virol. 88, 1–17 (1986).
be used to further differentiate specific members of the 3. J. A. Bell et al., J. Am. Med. Assoc. 157, 1083–1092 (1955).
Adenoviridae family (14,21). 4. C. H. Calisher et al., Stedman’s/ICTV Virus Words, Will-
iams & Wilkins, Baltimore, Md., 1992, pp. xxvi, 271.
Molecular Methods 5. L. J. D’Angelo et al., J. Infect. Dis. 140, 42–47 (1979).
6. J. J. Eady, J. H. Peacock, and T. J. McMillan, Br. J. Cancer
PCR. The polymerase chain reaction (PCR) has been
66, 113–118 (1992).
used to identify adenovirus DNA in tissue specimens
7. H. M. Foy, M. K. Cooney, and J. B. Hatlen, Arch. Environ.
and body fluids (14), diluted stool (52,88), air filters (103),
Health 17, 795–802 (1968).
swimming pool water (9), sewage, and polluted river
8. CDC, Infect. Control Hosp. Epidemiol. 13, 499–500 (1992).
water (58). Most primers for PCR have been selected from
the hexon coding region because there is an extensive 9. M. Papapetropoulou and A. C. Vantarakis, J. Infect. 36,
101–103 (1998).
homology among adenovirus serotypes; whereas sequences
from the early coding region E1B have been used for 10. M. Turner et al., South. Med. J. 80, 712–715 (1987).
the detection of specific adenoviruses (88). Although PCR 11. S. Darougar et al., Br. J. Ophthalmol. 67, 1–7 (1983).
using hexon primers is more sensitive than culture 12. G. G. Caldwell et al., Am. J. Epidemiol. 99, 230–234 (1974).
methods, it may give false-positive results from stool 13. M. Kukkula et al., Scand. J. Infect. Dis. 29, 415–418 (1997).
 ADENOVIRUSES 99

14. M. S. Horwitz, Adenoviruses, Lippincott-Raven Publishers, 53. M. C. Mahl and C. Sadler, Can. J. Microbiol. 21, 819–823
Philadelphia, Pa., 1996. (1975).
15. W. P. Rowe et al., Proc. Soc. Exp. Biol. Med. 84, 570–573 54. L. G. Irving and F. A. Smith, Appl. Environ. Microbiol. 41,
(1953). 51–59 (1981).
16. T. Shenk, Adenoviridae: The Viruses and Their Replication, 55. V. Krikelis et al., Can. J. Microbiol. 31, 24–25 (1985).
Lippincott-Raven Publishers, Philadelphia, Pa., 1996. 56. V. Krikelis et al., Water Sci. Technol. 17, 159–164 (1985).
17. E. K. Wagner and M. Hewlett, Basic Virology, Blackwell 57. R. Girones et al., Water Sci. Technol. 27, 235–241 (1993).
Science, Malden, Mass., 1999, pp. xxv, 466.
58. M. Puig et al., Appl. Environ. Microbiol. 60, 2963–2970
18. N. J. Dimmock and S. B. Primrose, Introduction to Modern (1994).
Virology, Blackwell Science, Boston, Mass., 1994, pp. viii,
59. F. P. Williams and C. J. Hurst, Water Res. 22, 847–851
384.
(1988).
19. I. Uhnoo, E. Olding-Stenkvist, and A. Kreuger, Arch. Dis.
60. N. Tani et al., Microbiol. Immunol. 39, 577–580 (1995).
Child. 61, 732–738 (1986).
61. G. A. Bagdasar’Yan and R. M. Abieva, Hyg. Sanit. 36,
20. I. Uhnoo et al., J. Clin. Microbiol. 20, 365–372 (1984).
333–337 (1971).
21. I. Uhnoo et al., Dev. Biol. Stand. 53, 311–318 (1983).
62. C. E. Enriquez, J. Sandoval-Garzon, and C. P. Gerba, Water
22. C. D. Brandt et al., J. Infect. Dis. 151, 437–443 (1985).
Quality Technology Conference, Part II, American Water
23. C. D. Brandt et al., J. Clin. Microbiol. 18, 71–78 (1983). Works Association, New Orleans, La., 1995.
24. J. P. Leite et al., J. Med. Virol. 15, 203–209 (1985). 63. J. Hara et al., Jpn. J. Ophthalmol. 34, 421–427 (1990).
25. J. R. Cruz et al., J. Clin. Microbiol. 28, 1780–1784 (1990). 64. F. X. Abad, R. M. Pinto, and A. Bosch, Appl. Environ.
26. G. Wadell, Curr. Top. Microbiol. Immunol. 110, 191–220 Microbiol. 60, 3704–3710 (1994).
(1984). 65. C. E. Enriquez, C. J. Hurst, and C. P. Gerba, Water Res. 29,
27. W. D. Cubitt, CIBA Found. Symp. 128, 126–143 (1987). 2548–2553 (1995).
28. V. C. Rao and J. L. Melnick, Environmental Virology, Amer- 66. M. D. Sobsey et al., Water Sci. Technol. 18, 97–106 (1986).
ican Society for Microbiology, Washington, D.C., 1986. 67. H. A. Fields and T. G. Metcalf, Water Res. 9, 357–364 (1975).
29. M. S. Gelfand et al., Clin. Infect. Dis. 19, 1045–1048 (1994). 68. F. E. Wassermann, Virology 17, 335–341 (1962).
30. C. P. Gerba, J. B. Rose, and C. N. Haas, Int. J. Food 69. Q. S. Meng and C. P. Gerba, Water Res. 30, 2665–2668
Microbiol. 30, 113–123 (1996). (1996).
31. K. H. Anders et al., Pediatr. Neurosurg. 16, 316–320 (1990). 70. B. Wilson, P. Roessler, and E. V. Dellen, Water Quality
32. D. Schnurr et al., J. Med. Virol. 47, 168–171 (1995). Technology Conference, American Water Works Association,
33. D. Dionisio et al., Scand. J. Infect. Dis. 29, 305–307 (1997). Toronto, Canada, 1992.
34. E. S. Yi and H. C. Powell, Ultrastruct. Pathol. 18, 549–551 71. A. P. Il’Nitskii, Hyg.Sanit. 31, 162–166 (1966).
(1994). 72. F. X. Abad et al., Appl. Environ. Microbiol. 60, 2377–2383
35. F. Dombrowski et al., Virchows Arch. 431, 469–472 (1997). (1994).
36. M. Shintaku, K. Nasu, and M. Ito, Histopathology 23, 73. A. Bosch, R. M. Pinto, and F. X. Abad, IAWQ 17th Biennial
588–590 (1993). International Conference, Budapest, Hungary, July 1994.
37. D. Ghez et al., Am. J. Hematol. 63, 32–34 (2000). 74. C. Bernstein and H. Bernstein, Aging, Sex, and DNA repair,
38. S. A. Hedderwick et al., Clin. Infect. Dis. 26, 997–999 (1998). Academic Press, San Diego, Calif., 1991, pp. xii, 382.
39. J. Koopmann et al., Infection 28, 323–325 (2000). 75. T. J. Kelly, in H. S. Ginsberg, ed., The Adenoviruses, Plenum
40. C. Maslo et al., Am. J. Respir. Crit. Care Med. 156, Press, New York, 1984, pp. xvii, 605.
1263–1264 (1997). 76. C. E. Enriquez et al., Water Res. 27, 1113–1119 (1993).
41. G. T. Valainis et al., J. Infect. Dis. 160, 349–351 (1989). 77. C. E. Enriquez and C. P. Gerba, Water Res. 29, 2554–2560
42. D. O. White and F. Fenner, Medical Virology, Academic (1995).
Press, San Diego, Calif., 1994, pp. xvii, 603. 78. B. England et al., in G. Berg, ed., Transmission of Viruses
43. W. J. Martone et al., Am. J. Epidemiol. 111, 229–237 (1980). by the Water Routes, John Wiley & Sons, New York, 1965.
44. H. M. Foy, Adenoviruses, Plenum Medical Book Company, 79. B. England, Appl. Microbiol. 24, 510–512 (1972).
New York, 1997. 80. M. D. Sobsey et al., J. Am. Water Works Assoc. 72, 350–355
45. A. C. Moore et al., MMWR CDC Surveill. Summ. 42, 1–22 (1980).
(1993). 81. M. D. Sobsey et al., J. Am. Water Works Assoc. 72, 292–299
46. C. J. Hurst and N. J. Adcock, in C. J. Hurst, ed., Viral (1980).
Ecology, Academic Press, San Diego, Calif., 2000, pp. 41–62. 82. M. D. Sobsey and J. S. Glass, Appl. Environ. Microbiol. 47,
47. M. Petric et al., Lancet 1, 1074–1075 (1982). 956–960 (1984).
48. N. R. Blacklow and H. B. Greenberg, N. Engl. J. Med. 325, 83. R. R. Jacobs and M. D. Sobsey, 78th Annual Meeting of the
252–264 (1991). American Society for Microbiology, ASM, Las Vegas, Nev.,
49. A. H. Havelaar, in G. F. Craun, ed., Safety of Water 1978, p. 199.
Disinfection: Balancing Chemical and Microbial Risks, ILSI 84. C. P. Gerba, in G. Berg, ed., Methods for Recovering Viruses
Press, Washington, D.C., 1993, pp. xxvi, 690. from the Environment, CRC Press, Boca Raton, Fla., 1987.
50. C. J. Hurst, K. A. McClellan, and W. H. Benton, Water Res. 85. E. Katzenelson, B. Fattal, and T. Hostovesky, Appl. Envi-
22, 1547–1552 (1988). ron. Microbiol. 32, 638–639 (1976).
51. C. A. Morris et al., Lancet 1, 4–5 (1975). 86. R. Morris and W. M. Waite, Water Res. 14, 791–793 (1980).
52. P. A. Thorner et al., Arch. Virol. 133, 397–405 (1993). 87. X. Jiang et al., J. Clin. Microbiol. 30, 2529–2534 (1992).
100 ADHESION, IMMOBILIZATION, AND RETENTION OF MICROORGANISMS ON SOLID SUBSTRATA

88. A. Allard, B. Albinsson, and G. Wadell, J. Med. Virol. 37, 40

149–157 (1992).
E. coli 2239
89. C. D. Brandt et al., J. Clin. Microbiol. 20, 1008–1009 (1984). S. epidermidis HBH 171
S. mitis 398
90. S. Chiba et al., Lancet 2, 954–957 (1983). 20 L. acidophilus ATCC 9224

Zeta potential (mV)

91. M. E. Johansson et al., J. Med. Virol. 17, 19–27 (1985).
92. M. Brown, M. Petric, and P. J. Middleton, J. Clin. Microbiol.
20, 334–338 (1984). 0
93. B. Genthe et al., Water Quality International, IAWQ,
Budapest, Hungary, 1994.
94. T. H. Scott-Taylor et al., J. Med. Virol. 41, 328–337 (1993). −20
95. J. C. de Jong et al., J. Med. Virol. 11, 215–231 (1983).
96. C. T. Tiemessen, M. Ujfalusi, and A. H. Kidd, Arch. Virol.
132, 193–200 (1993). −40
97. W. O. Grabow, D. L. Puttergill, and A. Bosch, J. Virol. 0 2 4 6 8 10
Methods 37, 201–207 (1992). pH
98. R. M. Pinto, J. M. Diez, and A. Bosch, J. Med. Virol. 44, Figure 1. Zeta potentials as a function of pH for E. coli 2,239,
310–315 (1994). Staphylococcus epidermidis HBH 276, Streptococcus mitis BA and
99. D. L. Wiedbrauk and S. L. G. Johnston, Manual of Clinical Lactobacillus acidophilus ATCC 9,224 as measured in 10 mM
Virology, Raven Press, New York, 1993, pp. ix, 275. potassium phosphate, with pH adjusted by addition of HCl
100. B. Precious and C. Russell, in B. W. J. Mahy, ed., Virology: or KOH. Data from H. C. van der Mei, R. Bos, H. J. Busscher,
A Practical Approach, IRL Press, Washington, D.C., 1985. and P. S. Handley, in A. Baszkin and W. Norde, eds., Physical
101. I. I. Mendez et al., J. Virol. Methods 90, 59–67 (2000). Chemistry of Biological Interfaces, Marcel Dekker, New York,
2000, pp. 431–458.
102. G. D. Hsiung and C. K. Y. Fong, Diagnostic Virology: Illus-
trated by Light and Electron Microscopy, Yale University
Press, New Haven, Conn., 1982, pp. xxii, 276.
103. M. Echavarria, et al., J. Clin. Microbiol. 38, 2982–2984 four selected microbial strains are shown in Figure 1.
(2000). Charge reversal upon association of acidic (carboxyl or
104. A. E. van Loon et al., Virology 140, 197–200 (1985). phosphate) groups at low pH values is indicative of the
105. N. Suzuki et al., Microbiol. Immunol. 25, 1291–1301 (1981). presence of charged basic (amino) groups on the cell sur-
106. G. Hammond et al., J. Clin. Microbiol. 25, 1881–1885 face. Microorganisms also exhibit variation in surface-free
(1987). energies, with some strains being hydrophilic and oth-
107. R. M. Maier, I. L. Pepper, and C. P. Gerba, Environmental ers being relatively hydrophobic (5). In contrast to inert
Microbiology, Academic Press, New York, 2000. colloidal particles, however, microorganisms are living
microorganisms capable of metabolism, growth, and, in
some instances, independent motion. An expression of the
ADHESION, IMMOBILIZATION, AND metabolism of the microorganisms is the production of var-
RETENTION OF MICROORGANISMS ON SOLID ious amounts of extracellular polymers (6) that together
SUBSTRATA with flagella, pili, fimbriae, and fibrils (4,7) (Fig. 2) are
important in the interaction between microorganisms and
CRISTINA GÓMEZ-SUÁREZ solid substrata. Microorganisms can interact with inor-
HENNY C. VAN DER MEI ganic, living, or dead materials such as plants, animals,
HENK J. BUSSCHER artificial structures, rocks, clays, sand grains, and organic
University of Groningen remains of living microorganisms in various stages of
Groningen, The Netherlands decomposition.
Once microorganisms are attached to a surface, a mul-
Microorganisms occur in nearly all natural fluid envi- tistep process starts leading to the formation of a complex,
ronments on earth and have a strong tendency to adhering microbial community termed biofilm. A biofilm
become associated with surfaces (1). The interaction mech- can be defined as a microecosystem in which different
anism between microorganisms and substratum surfaces strains and species efficiently cooperate to protect them-
involves the physicochemical properties of the interacting selves against environmental stresses and to facilitate
surfaces (2) and, in fact, microorganisms can be consid- more efficient nutrient uptake (8). Biofilms can be benefi-
ered as living colloidal particles with a density slightly cial, for instance, in sewage treatment (9), bioreactors (10),
greater than that of water, and size of about 1 µm in and animal microflora (11). However, biofilms are most
length or diameter (3). Stable suspensions of microorgan- frequently unwanted as in food processing (12,13), on
isms in dilute electrolytes result, in part, from a mutual ship hulls (14), in drinking water (15) and dairy man-
electrostatic repulsion between like charges on micro- ufacturing plants (16), in pipelines (17), on biomaterial
bial cell surfaces. Virtually all microorganisms known implants (18), or in the oral cavity (19). Although the
at present possess a net-negative surface charge at the function and appearance of biofilms in various environ-
pH range found in most natural habitats (4), as revealed ments may be different, all biofilms form from the same
by zeta potential measurements. The pH dependence of basic sequence of events (20), as depicted in Figure 3.
 ADHESION, IMMOBILIZATION, AND RETENTION OF MICROORGANISMS ON SOLID SUBSTRATA 101

(a) are transported toward the solid surface through diffusion,

convection, sedimentation, or by intrinsic motility. Alter-
natively, planktonic (suspended) microorganisms may be
transported toward each other to form microbial coag-
gregates (22) and subsequently adhere, or a few sessile
(already adhering) microorganisms may stimulate coad-
hesion of planktonic microorganisms (23). In later stages,
microbial anchoring occurs through extrapolymer (i.e.,
polysaccharides) production (6,24) and eventually, adher-
ing microorganisms start to grow, which is the main
contributing factor for the accumulation of microorgan-
isms in a biofilm.
The thermodynamic theory (25,26) considers the
surface-free energies of the solid substratum, the micro-
bial cell surface, and the suspending medium yielding the
interfacial free energies between the interacting surfaces,
as schematically illustrated in Figure 4. Accordingly, this
comparison is expressed in the so-called free energy of
adhesion
Gadh = γsm − γsl − γml (1)

where γsm , γsl , and γml , are the interfacial free energies,
(b) where ‘‘s’’ denotes the substratum, ‘‘m’’ the microbial
cell surface and ‘‘l’’ the liquid surface. Adhesion is
thermodynamically favorable if Gadh is negative, because
systems tend to minimize their free energy, whereas
adhesion is energetically unfavorable when Gadh is
greater than zero.
A three-point hypothesis of microbial interaction mech-
anisms (27) related to the distance between the microbial
cell and substratum surfaces is schematically illustrated
in Figure 5. At large separation distances (>50 nm), only
Lifshitz-Van der Waals forces are operative. Upon closer
approach, both Lifshitz-Van der Waals and electrostatic
forces (i.e., nonspecific interactions) occur for separation
distances between 10 to 20 nm. Finally, at near contact,
Lifshitz-Van der Waals forces, electrostatic and acid-base
interactions, and so-called specific interactions dictate
adhesion. Specific interactions may be defined as the inter-
actions between stereochemically complementary surface
components and act over extremely short distances allow-
ing specific ionic, hydrogen, and possibly chemical bonds
to form.
Figure 2. Electron micrographs of (a) E. coli with rigid, peritric- In this paper, a distinction is made among adhesion,
hous fimbriae, approximately 7 nm in diameter and up to 1.0 µm immobilization, and retention of microorganisms on sub-
length. Cells are shadowed with palladium. Bar marker repre-
stratum surfaces, and the importance to make such a
sents 0.5 µm; (b) Streptococcus salivarius showing peritrichous
clumped fibrils with a length of 191 nm. Cells are negatively
distinction is exemplified. A microorganism adheres to a
stained with 1% methylamine tungstate. Bar marker represents substratum surface when it maintains the same separa-
100 nm. Micrographs taken with permission from H. C. van der tion distance from the substratum surface as a function
Mei, R. Bos, H. J. Busscher, and P. S. Handley, in A. Baszkin of time through a balance of attractive and repulsive
and W. Norde, eds., Physical Chemistry of Biological Interfaces, interactions homogeneously acting over an entire sub-
Marcel Dekker, New York, 2000, pp. 431–458. stratum surface. Consequently, adhering microorganisms
are still free to move parallel to the substratum as
opposed to immobilization. Immobilized microorganisms
When organic matter is present in an aqueous environ- adhere to substratum surface but are kept at the same
ment (e.g., blood, saliva, tear fluid, seawater, milk, or position by immobilization forces acting parallel to the
urine), a so-called conditioning film of adsorbed compo- surface. Often, both in natural habitats and in laboratory
nents (i.e., proteins and other organic molecules) is formed experiments, adhering microorganisms are challenged by
on the substratum surface before adhesion of microor- external forces yielding detachment of all or subpopula-
ganisms, because molecules diffuse much faster than tions of adhering organisms. Retention will be explained
microorganisms (21). As a second step, microorganisms to as the capacity of adhering microorganisms to remain
102 ADHESION, IMMOBILIZATION, AND RETENTION OF MICROORGANISMS ON SOLID SUBSTRATA

(a) (b)
Conditioning film Transport

(c) (d)
Initial adhesion Co-adhesion

(e) (f)
Anchoring Growth

Yeasts Bacteria Polymers Conditioning film Surface structures

Figure 3. Sequential steps in the formation of a biofilm on a solid substratum. (a) Adsorption of
conditioning film components. (b) Microbial transport and coaggregation. (c) Adhesion of single
organisms and of microbial coaggregates. (d) Coadhesion between microbial pairs. (e) Anchoring
yielding firm, irreversible adhesion through exopolymer production and surface structures.
(f) Microbial growth and ingrowth.

adhering after having been submitted to an external force Waals, electrostatic, and acid-base components. In the
other than their mutual interaction. original theory by Derjaguin, Landau, Verwey, and Over-
beek (the classical DLVO theory), microbial adhesion
ADHESION AND THE DLVO THEORY is described as a balance between attractive Lifshitz-
The perpendicular interaction between microorganisms Van der Waals and repulsive or attractive electrostatic
and a substratum surface consists of Lifshitz-Van der forces. Accordingly, the interaction energies between
 ADHESION, IMMOBILIZATION, AND RETENTION OF MICROORGANISMS ON SOLID SUBSTRATA 103

Liquid Liquid

glm

Microorganism
glm

Microorganism

Figure 4. Schematic presentation of the

gsl gsm gsl interfacial free energies (γij ) involved in
the adhesion of a microorganism to a
Substratum Substratum solid substratum surface from a liquid sus-
pension. The solid-microorganism (γsm ),
the solid-liquid (γsl ), and the microorgan-
isms-liquid (γml ) interfacial free energies
DGadh > 0 DGadh < 0
are indicated.

microorganisms and substratum surfaces as a function of introducing an electrolyte into the liquid phase, yield-
their separation distance (d) can be expressed as (28–30) ing changes in the electrostatic double layer at a sur-
face (33,34). Electrostatic double-layer attraction or repul-
GTOT (d) = GLW (d) + GEL (d) (2) sion arises from the interpenetration of interacting double
layers, and assuming constant potentials during interac-
tions it is given by
where GTOT , GLW , and GEL denote the total, the Lifshitz-
Van der Waals, and the electrostatic interaction energies, 
2ζm ζs 1 + exp(−κd)
respectively (31,32). GEL (d) = π εa(ζm2 + ζs2 ) ln
ζm2 + ζs2 1 − exp(−κd)

+ ln[1 − exp(−2κd)] (5)
Lifshitz-Van der Waals Interactions
The Lifshitz-Van der Waals interaction between a where ε denotes the permittivity of the medium, ζm and
microorganism and a solid surface at separation distances ζs the zeta potential of microorganism and substratum,
small enough to ignore retardation effects is given by respectively, and κ −1 is the double layer thickness,

  and constant potentials are considered. κ −1 depends on
A a a d the total ion concentration in the solution and can be
G LW
(d) = − + + ln (3)
6 d d + 2a d + 2a calculated from
 (1/2)
where A denotes the Hamaker constant (i.e., an indication e2 
κ= zi ni (6)
of the polarizability of the interacting surfaces) and a is εkT
i
the radius of the microorganism. The Hamaker constant
depends on the nature of the interacting media. For the where e denotes the electron charge, k the Boltzmann
interaction of a microorganism (m) with a substratum (s) constant, T the absolute temperature, zi is the valence of
in a liquid medium (l), A is (29) the ions present and ni is the number of ions per unit
volume. For a symmetrical 1-1 electrolyte this equation
(1/2) (1/2)
Amsl = (A(1/2)
ss − All mm − All
)(A(1/2) ) (4) reduces to
κ = 0.328 × 1010 (z2i Mi )(1/2) (7)
In most cases, A is positive (attractive Lifshitz-Van der
Waals forces), because in aqueous systems Ass is greater where Mi is the molarity (mol/L) of the ions.
than All and Amm is greater than All , and consequently Figure 6 illustrates the decay with the distance of the
GLW is less than zero. Lifshitz-Van der Waals and electrostatic energies between
a charged microorganism and a substratum surface for
repulsive electrostatic interactions (i.e., like charges) at
Electrostatic Interactions
different ionic strengths. It can be seen that at low
In aqueous media, electrostatic interactions are present ionic strengths (Fig. 6a), a high-energy barrier has to
in addition to Lifshitz-Van der Waals forces. These forces be overcome if the microorganisms are to come in close
arise from the electrical surface charge of microorgan- contact with the surface. At intermediate ionic strengths
isms and substratum surfaces, and can be influenced by (Fig. 6b), the energy barrier, although still present,
104 ADHESION, IMMOBILIZATION, AND RETENTION OF MICROORGANISMS ON SOLID SUBSTRATA

Liquid

Bacterium
Lifshitz-Van der Waals
interactions

> 50 nm

Substratum

− −
Additional electrostatic − +
interactions − −
+ −
− 10-20 nm
− − − − − − − −

Additional specific
interactions

< 1.5 nm

+ +
− −
Figure 5. Three-point hypothesis of bacterial adhesion mech-
anisms related to the distance of the bacterium from
the substratum surface. Adapted from H. J. Busscher and
A. H. Weerkamp, FEMS Microbiol. Rev. 46, 165–173 (1987).

is lower and microorganisms can become reversibly both the range and the magnitude of the electrostatic
captured in a so-called secondary minimum. At high ionic interactions decrease with an increasing ionic strength.
strengths (Fig. 6c), the energy barrier disappears and
a strong net attraction exists between microorganisms The Extended DLVO Theory
and the substratum surface at all separation distances.
The classical DLVO theory has been extended by Van Oss
Microorganisms held in the deep primary interaction
and coworkers (35) and recently related to the origin of
minimum are frequently considered to adhere irreversibly.
hydrophobic interactions in microbial adhesion (36). The
Alternatively, Figure 7 depicts the interaction energies
extended DLVO theory considers four fundamental, nonco-
as a function of separation distance at different ionic
valent interactions: Lifshitz-Van der Waals, electrostatic,
strengths for unlike microbial and substratum surface
Lewis acid-base, and Brownian motion forces.
charges. The net effect is a strong attraction regardless of
separation distance, which decreases with the increasing
Acid-Base Interactions
ionic strength.
In summary, the Lifshitz-Van der Waals interaction Acid-base interactions originate from electron-donating
between suspended microorganisms and a surface is not and electron-accepting interactions between polar moi-
influenced by the ionic strength of the suspension, but eties in aqueous solutions. The decay with separation
 ADHESION, IMMOBILIZATION, AND RETENTION OF MICROORGANISMS ON SOLID SUBSTRATA 105

(a) (b)

30 30

G EL
20 20

10 G TOT 10

GTOT(kT)
GTOT(kT)

0 0

−10 −10

G LW
−20 −20

Secondary minimum

−30 −30
Primary minimum

0 5 10 15 20 25 0 5 10 15 20 25
Distance (nm) Distance (nm)
Figure 6. Interaction energy curves (classical DLVO theory) of a microorganism with a solid
surface of like charge versus their separation distance for (a) low (i.e., one mM), (b) intermediate
(i.e., 10 mM) or (c) high (i.e., 100 mM) ionic strength of the medium.


 

distance of the acid-base interaction energy (GAB ) between
− + −
γmv γlv+ − −
γmv γlv−
a microorganism and a substratum surface can be
expressed as 




 − + −
γsv γlv+ − −
γsv γlv− (10)
d0 − d
G AB
(d) = 2π aλGAB
adh exp (8)
λ where γ − and γ + denote the electron-donating and
electron-accepting parameters of the interfacial free
where GAB adh is the acid-base interfacial free energy of energy, respectively.
adhesion, d0 is the distance of closest approach between Figure 8 illustrates the decay with the distance of the
two surfaces 1.57 Å according to (37), and λ denotes the interaction energies in the extended DLVO theory between
correlation length of molecules in a liquid medium and a microorganism and a substratum surface for apolar,
varies with the ionic strength and with the sign of GAB adh . monopolar, and bipolar surfaces under conditions of elec-
For GAB adh less than zero (‘hydrophobic attraction’), λ trostatic repulsion. It can be seen that the influence of
may be as large as 13 nm. For situations of hydrophilic acid-base interactions is quite strong compared with elec-
repulsion (GAB adh > 0) λ equals 0.6 nm (38). The acid-base trostatic and Lifshitz-Van der Waals interactions. How-
free energy of adhesion can be obtained from the following ever, the acid-base interactions are only relatively short-
expressions ranged, and a close approach between interacting surfaces


 

 is required for these forces to become operative, which may
GAB
adh = −2 AB
γmv − γlvAB AB
γsv − γlvAB (9) well be impeded in microbial adhesion because of the dif-
ferent structural features on microbial cell surfaces (39).

where γmv AB
, γlvAB , and γsvAB
are, respectively, the IMMOBILIZATION: PERPENDICULAR AND LATERAL
microorganism-vapor, liquid-vapor, and substratum-vapor INTERACTIONS
acid-base interfacial free energies, or from
    In the case of adhesion, the main body of a microbial
adh = +2
GAB + −
γmv +
γsv − −
γmv −
γsv cell is held at a small, but finite, distance from the
106 ADHESION, IMMOBILIZATION, AND RETENTION OF MICROORGANISMS ON SOLID SUBSTRATA

(c) because the substratum surface cannot exert any lateral

force. A lateral interaction minimum between microor-
ganisms occurs at large separation distances (500 nm)
30 and is extremely shallow (about 0.1 kT) as compared with
the deep (i.e., 12 kT) perpendicular interaction minimum
occurring at heights of around 6 nm above the substra-
20 tum surface (see Fig. 10). Consequently, regardless of the
strength of the perpendicular interactions, microbial adhe-
sion is always mobile on ideally homogeneous and smooth
10 surfaces.

Chemically Heterogeneous Substrata

GTOT(kT)

0
Virtually all substratum surfaces contain chemical
heterogeneities at a submicron level, which can be
−10 demonstrated by contact-angle hysteresis (41). Their
importance in particle adhesion has recently been
demonstrated by Wit and Busscher (42), showing that
−20 colloidal particles have a strong preference to deposit at
the same locations on a substratum surface in repeated
deposition experiments after detachment of the adhering
−30 particles. Different chemistries along a substratum surface
yield different DLVO interaction energies with adhering
microorganisms, as shown in Figure 11a, and secondary
interaction minima with different depths are found for
hydrophilic and hydrophobic regions on the surface, from
which a lateral interaction force develops, driving adhering
0 5 10 15 20 25
organisms to a location of minimum interaction energy,
Distance (nm)
that is, the site with the deepest interaction minimum is
Figure 6. (Continued) occupied. The lateral interaction energy dG = (G2 − G1 )
for a chemical heterogeneity in the example amounts to
about −5 kT (see Fig. 11b), which is more than the lateral
substratum surface by a balance between interaction
interaction energy on an ideally homogeneous and smooth
forces acting perpendicular to the surfaces. Regardless
surface (0.1 kT, compare with Fig. 10b).
of whether this position above the substratum surface
is occupied in a reversible or irreversible fashion, the
balance of forces acting perpendicular to the surface does Substratum Roughness
not ensure immobilization of adhering organisms along Because of rugosities on a substratum surface, macro-
the surface, which requires lateral interaction forces. As scopically lateral interaction forces can arise from per-
a consequence, mobile adhesion is not solely a property pendicular interactions and stimulate immobilization.
of the so-called mobile bacteria, but of all colloidal Microorganisms adhering near structural heterogeneities
particles adhering on a surface in the absence of lateral on a substratum surface have an energetically less favor-
interaction forces. The distinction between adhesion and able situation than when adhering on an overall smooth
immobilization is best illustrated by comparing the ease
surface (43). Mobility on a rough surface is driven by a lat-
with which a magnet adhering on a smooth metal surface
eral interaction energy, which depends on the size of the
can be slid over the surface with the force required to pull
rugosities and can be estimated to amount to several kTs
the magnet off the surface (see Fig. 9).
in magnitude. In Figure 11b, it can be seen that lateral
Mobility of adhering microorganisms is due to forces
mobility of a microorganism adhering on a rough surface
on the adhering microorganisms acting laterally along
is inevitably associated with a change in the perpendicu-
the substratum surface and not due to perpendicular
lar position with respect to the overall substratum surface
interactions between an adhering microorganism and a
from which the majority of the interaction forces occur (44).
substratum surface (40). Accordingly, adhering microor-
It is important to make a distinction between
ganisms can be sheared, they can diffuse, or they can
micrometer roughness and nanometer roughness, because
actively swim freely, laterally until they encounter a sticky
new developed techniques (i.e., atomic force microscopy)
patch on the surface (chemical heterogeneity) or become
permit the measurement of such values (45). A bacterial
mechanically trapped in ‘‘rugosities.’’
cell (size of about a micrometer) will interact laterally with
a micrometer roughness and stipulate its mobility over the
Ideally Homogeneous and Smooth Substrata
substratum. On the other hand, if the substratum surface
On ideally homogeneous and smooth surfaces (see Fig. 10), has a roughness at a nanometer scale, the bacterial cell
the only lateral interactions operative result from a multi- will not interact laterally with such rugosity and might
body DLVO interaction between all adhering bacteria, even be repelled by it (46).
(a) (b)

30 30

20 20

10 10

GTOT(kT)
GTOT(kT)

0 0

−10 −10
G LW

−20 −20

−30 −30

G EL

G TOT

0 5 10 15 20 25 0 5 10 15 20 25
Distance (nm) Distance (nm)

(c)

30

20

10
GTOT(kT)

−10

−20

−30

0 5 10 15 20 25
Distance (nm)

Figure 7. Interaction energy curves (classical DLVO theory) of a microorganism with a solid
surface of unlike charge versus their separation distance for (a) low (i.e., 1 mM), (b) intermediate
(i.e., 10 mM), or (c) high (i.e., 100 mM) ionic strength of the medium.
107
(a) (b)

30 30
TOT
GClassical G TOT
Classical

20 20

10 10

G AB G AB
GTOT(kT)

GTOT(kT)
0 0

−10 −10

−20 −20

−30 −30

TOT
GExtended GTOT
Extended

0 5 10 15 20 25 0 5 10 15 20 25
Distance (nm) Distance (nm)

(c)

30

20

10

G AB
GTOT(kT)

0
TOT
GExtended

−10

−20

GTOT
Classical
−30

0 5 10 15 20 25
Distance (nm)

Figure 8. A comparison of the interaction energies according to the classical and extended DLVO
theory between a microorganism and a solid substratum as a function of the separation distance
for (a) apolar, (b) monopolar, and (c) bipolar surfaces.

108
 ADHESION, IMMOBILIZATION, AND RETENTION OF MICROORGANISMS ON SOLID SUBSTRATA 109

Ideally homogeneous and smooth surfaces Nonideal surfaces

(a) Mobile adhesion (c) Immobilization due to surface
heterogeneities
wood

iron

(b) Immobilization due to lateral (d) Immobilization due to surface

interactions with neighbors roughness

Figure 9. Motility of bacteria is, from a physicochemical point of view comparable with the
interaction of a magnet with a metal plate. On an ideally homogeneous and smooth surface:
(a) when a magnet can be easily slid over the surface of a metal plate, one speaks of adhesion
without immobilization, (b) when a magnet is not free to move over the surface because of lateral
interactions with neighbors, there is immobilization. On a nonideal surface: (c) Immobilization
may result from lateral interaction forces as a result of heterogeneities on the metal surface, (d)
surface roughness can provoke immobilization as a result of lateral interactions with the magnet.

(a) (b)

40 0.8

30 0.6
G(r)
G(d)
20 0.4
G (kT)

G (kT)

10 0.2

0 0

−10 −0.2

0 5 10 15 20 0 1 2 3 4 5
Distance (nm) Distance (µm)
Figure 10. (a) The interaction energy G(d) according to the extended DLVO theory as a
function of separation distance between a microorganism and an ideally homogeneous and
smooth, hydrophilic, and negatively charged model surface. (b) The lateral interaction energy
G(r) as a function of separation distance among microorganisms (r) is calculated from the
radial pair distribution of the adhering microorganisms. With permission from H. J. Busscher,
A. T. Poortinga, and R. Bos, Curr. Microbiol. 37, 319–323 (1998).

Polymer and Other Short-Range Interactions extracellular polysaccharides, whereas on gram-positive

The surfaces of bacteria are covered with a variety of cells, lipoteichoic, teichoic, and teichuronic acids, mycolic
polymers, which differ from one species to another (47). acids, capsules of extracellular polysaccharides, and
The surfaces of gram-negative bacteria are mainly covered proteins can be present (48). Polymer interactions can
by lipopolysaccharides, surface proteins, or capsules of be attractive or repulsive (49) and arise when bacterial
110 ADHESION, IMMOBILIZATION, AND RETENTION OF MICROORGANISMS ON SOLID SUBSTRATA

(a) (b)

40 40

30 30

G1(d) G2(d)
d
20 20
G (kT)

G (kT)
10 G1 10
G2 d

0 0

−10 dG −10 dG

0 5 10 15 20 0 5 10 15 20
Distance (nm) Distance (nm)
Figure 11. Interaction energies G(d) of a microorganism, according to the extended DLVO
theory, as a function of separation distance with different regions on a heterogeneous substratum
surface. (a) Hydrophobic patches on a hydrophilic surface exert different interaction energies on an
approaching microorganism, yielding a lateral interaction between the adhering microorganisms.
(b) Microorganisms adhering on structural heterogeneities on a substratum experience lateral
interaction forces on a macroscopic scale. NB. Organisms and heterogeneities are not drawn
to scale. With permission from H. J. Busscher, A. T. Poortinga, and R. Bos, Curr. Microbiol. 37,
319–323 (1998).

surface polymers have a high affinity for the solid surface (a)
and are long enough to reach and adsorb to the surface (see Bacterial cell
Fig. 12a). Such polymers may bridge the distance between
a substratum surface and the microbial cell body, the
latter of which is kept at a larger distance by electrostatic
repulsion (30). If polymer interactions are site-specific on
the substratum surface, they do not only assist in adhesion
but also in immobilization. Excessive polymer production
by bacteria may serve to effectively disperse bacteria in
suspension and prevent them from adhering to a surface, Substratum
as shown in Fig. 12b (50,51).
Short-range interactions arise upon direct contact
between bacteria and solid substrata. Often these site-
(b) Bacterial cell
specific interactions involve covalent bonds, requiring 40
to 200 kT (52) to break, which is considerably more than,
for instance, hydrogen bonds that can be disrupted at the
expense of only 4 to 16 kT (53).

Motile Bacteria
Motility of adhering bacteria has been described so Substratum
far from a physicochemical point of view, but motile Figure 12. (a) Attractive polymer interaction occurs when bac-
bacteria constitute a special class of microorganisms terial polymers have affinity for the solid surface and are long
in this respect. Depending on their orientation when enough to reach and adsorb to the surface. As a consequence,
approaching a substratum surface, bacteria (i.e., some these polymers bridge the distance separating surface and bacte-
strains of Escherichia coli and Pseudomonas aeruginosa) rial cell. (b) On the other hand, when the bacterial cell produces
may exhibit a continuous gyratory motion around the an excessive amount of polymers, a repulsive polymer interac-
pole of the cell or a propeller-like motion (54). Some motile tion occurs, which tends to keep the bacterial cell far from the
substratum surface.
bacteria near or adhering to a surface are able to overcome
 ADHESION, IMMOBILIZATION, AND RETENTION OF MICROORGANISMS ON SOLID SUBSTRATA 111

the attractive forces at the secondary minimum and swim exposed to extremely large detachment forces provoked
away (55,56), often becoming reversibly attached at other by the surface tension at the passing liquid-air interface
sites. Gliding and near-surface swimming bacteria are (see Fig. 13). A significant detachment of adhering
a mobile form of microbial adhesion, as discussed earlier, microorganisms can result from the passage of a liquid-
with the major difference that the lateral interaction forces air interface over adhering microorganisms depending on
for mobile behavior now originate from the organism itself. the velocity of the passing liquid-air interface, the surface
tension at the liquid-air interface, the number of passing
interfaces, hydrophobicity, the charge of the interacting
RETENTION
surfaces, and the size of the microorganisms (60–62).
As most of the methods summarized in Table 1 actually
The balance between repulsive and attractive interactions
involve the passage of a liquid-air interface, retention
in microbial adhesion is a delicate one and can be easily
of adhering microorganisms is measured in these assays
disturbed by dynamic factors in the natural microbial
rather than adhesion.
habitat or by experimental conditions in the laboratory.
External forces other than the interaction forces between
adhering microorganisms and a substratum surface CONCLUSION
frequently cause detachment of all or subpopulations of
adhering organisms. Therefore, in natural habitats and The interaction between microorganisms and substratum
experimental systems, microbial retention more readily surfaces is mediated by different types of interaction
occurs as a dominant factor in biofilm formation than forces. A distinction should be made with regard to the
microbial adhesion. Unfortunately, the distinction is not type of interaction according to:
well made in microbial adhesion research (see section on
Materials and methods, which mentions that ‘‘surfaces 1. adhesion, mobile or immobile, of microorganisms
with adhering microorganisms were slightly rinsed to on substratum surfaces as a result of a balance
remove loosely adhering organisms.’’) Instead, a definition between Lifshitz-Van der Waals, electrostatic, Lewis
of ‘‘loosely adhering’’ should be given, along with a acid-base, and Brownian motion forces;
quantitative indication of the rinsing forces, although 2. immobilization of adhering microorganisms on a
finally, such papers should be written in terms of microbial substratum surface due to the lateral interaction
retention rather than adhesion. The reluctance of authors forces, polymer bridging, or chemical bonds that keep
to do so greatly contributes to the inability of researchers an adhering microorganism at a certain location on
in the field to compare microbial adhesion data, whereas, a substratum surface;
in fact, in many natural habitats, most notably the oral 3. retention of adhering microorganisms, denoting
cavity, retention is more important than adhesion (57). microorganisms that remain adhering on a substra-
Methods to study microbial adhesive interactions tum surface after application of an external force.
(58,59) are summarized in Table 1, together with a critical
evaluation of whether adhesion or retention is actually It is important to distinguish between adhesion, immo-
studied. In all the methods involving dipping, rinsing, bilization, and retention for a proper interpretation of
and washing steps (60), adhering microorganisms are experimental results. Real-life biofilm formation often

Table 1. Currently Employed Methods to Study Microbial Adhesive Interactions with Solid
Substrata
Retention (R) or
Method Relevant Details Adhesion (A) References

Filtration — High shears applied R Ofek and Doyle (58), Ofek

— Risk of bacterial and coworkers (59), Ellen
clumping and Gibson (63)
Differential — Shear forces depend on R Lowry and coworkers (64),
centrifugation number of washings and Beachy and Ofek (65),
centrifugations Graham (66)
Beads — No centrifugation needed R Clark and coworkers (67),
— Washing steps included Wassal and coworkers (68)
Packed beads — No washing steps A Lytle and coworkers (69),
included Elimelech (70)
— Controlled mass
transport
Immobilization of — Washing steps included R Ofek and coworkers (71)
macromolecules — Nonspecific adhesion
Slide methods — Bacterial enumeration R Liu (72), Wirtanen and
after washing steps coworkers (73)
Critical forces tests — Fluid-shear forces R Taylor (74), Mohandas, and
exerted coworkers (75)
112 ADHESION, IMMOBILIZATION, AND RETENTION OF MICROORGANISMS ON SOLID SUBSTRATA

Idealized, undeformed
interface Three-phase
contact
Air γ lv γ lv

Liquid θ −φ h0
φ
Figure 13. Schematic presentation of the three-phase θ
contact between a liquid-air interface and a micron-sized
particle. The surface-tension force can oppose the adhesion R
force between particle and solid substratum, depending p
on the three-phase boundary line as expressed by
H
Fγ = 2π Rp γlv sin[φ(t)] sin[θ − φ(t)]. With permission from
A. F. M. Leenaars and S. B. G. O’Brien, Philips J. Res. 44, Substratum
183–209 (1989).

involves adhesion, immobilization, or retention as a dom- 20. A. Escher and W. G. Characklis, in W. G. Characklis and
inant contributing factor that should be identified before K. C. Marshall, eds., Biofilms, John Wiley & Sons, New York,
meaningful laboratory experiments can be set up. U.S.A., 1990, pp. 445–486.
21. H. C. van der Mei, J. M. Meinders, and H. J. Busscher,
Microbiology 140, 3413–3419 (1994).
BIBLIOGRAPHY
22. P. E. Kolenbrander and J. London, in K. C. Marshall, eds.,
1. M. C. M. van Loosdrecht, J. Lyklema, W. Norde, and A. J. B. Advances in Microbial Ecology, vol. 12, Plenum Press, New
Zehnder, Microbiol. Rev. 54, 75–87 (1990). York, 1992, pp. 183–217.
2. R. Bos, H. C. van der Mei, and H. J. Busscher, FEMS 23. R. Bos, H. C. van der Mei, and H. J. Busscher, J. Microbiol.
Microbiol. Rev. 23, 179–230 (1999). Methods 23, 169–182 (1995).
3. S. Frieberg, Swed. Dent. J. 1, 207–214 (1977). 24. I. W. Sutherland, in I. W. Sutherland, eds., Surface Carbohy-
drates of the Prokaryotic Cell, Academic Press, New York,
4. H. C. van der Mei, R. Bos, H. J. Busscher, and P. S. Handley,
1977, pp. 27–96.
in A. Baszkinand W. Norde, eds., Physical Chemistry of
Biological Interfaces, Marcel Dekker, New York, 2000, 25. D. R. Absolom et al., Appl. Environ. Microbiol. 46, 90–97
pp. 431–458. (1983).
5. H. C. van der Mei, R. Bos, and H. J. Busscher, Colloid Surf., 26. H. J. Busscher et al., Appl. Environ. Microbiol. 48, 980–983
B: Biointerfaces 11, 213–221 (1998). (1984).
6. T. R. Neu and K. C. Marshall, J. Biomater. Appl. 5, 107–133 27. H. J. Busscher and A. H. Weerkamp, FEMS Microbiol. Rev.
(1990). 46, 165–173 (1987).
7. S. M. Hammond, P. A. Lambert, and A. N. Rycroft, The Bac- 28. P. R. Rutter and B. Vincent, in R. C. W. Berkeley et al., eds.,
terial Cell Surface, Croom Helm Ltd., Kent, U.K., 1984, Microbial Adhesion to Surfaces, Ellis Horwood, Chichester,
pp. 119–146. U.K., 1980, pp. 79–92.
8. B. Gottenbos, H. C. van der Mei, and H. J. Busscher, Methods 29. J. N. Israelachvili, in J. N. Israelachvili, eds., Intermolecular
in Enzymology in R. J. Doyle, eds., Biofilms, vol. 310, and Surface Forces, Academic Press, San Diego, Calif., 1992,
Academic Press, New York, 1999, pp. 523–534. pp. 139–337.
9. G. M. Wolfaardt et al., Appl. Environ. Microbiol. 60, 434–446 30. K. C. Marshall, in D. C. Savage and M. Fletcher, eds., Bac-
(1994). terial Adhesion. Mechanisms and Physiological Significance,
10. R. M. M. Diks and S. P. P. Ottengraf, Bioprocess Eng. 6, Plenum Press, New York, 1985, pp. 133–145.
93–100 (1991). 31. Th. F. Tadros, in R. C. W. Berkeley et al., eds., Microbial
11. G. Reid et al., Clin. Microbiol. Rev. 3, 335–344 (1990). Adhesion to Surfaces, Ellis Horwood, Chichester, U.K., 1980,
pp. 93–113.
12. S. Notermans, J. A. M. A. Dormans, and G. C. Mead, Biofoul-
ing 5, 21–36, (1991). 32. M. C. M. van Loosdrecht, J. Lyklema, W. Norde, and A. J. B.
13. J. P. G. Piette and E. S. Idziak, Appl. Environ. Microbiol. 55, Zehnder, Microb. Ecol. 17, 1–15 (1989).
1531–1536 (1989). 33. B. V. Derjaguin and L. D. Landau, Acta PhysicaChimica,
14. K. E. Cooksey and B. Wigglesworth-Cooksey, Aquat. Microb. U.S.S.R. 14, 633–652 (1941).
Ecol. 9, 87–96 (1995). 34. E. J. W. Verwey and J. Th. G. Overbeek, Theory of Stability
15. S. L. Percival, J. S. Knapp, D. S. Wales, and R. G. J. Edy- of Colloids, Elsevier, Amsterdam, The Netherlands, 1948.
vean, Biofouling 13, 259–277 (1999). 35. C. J. van Oss, R. J. Good, and M. K. Chaudhury, J. Colloid
16. S. H. Flint, P. J. Bremer, and J. D. Brooks, Biofouling 11, Interface Sci. 111, 378–390 (1986).
81–97 (1997). 36. C. J. van Oss, Colloids Surf., B: Biointerfaces 5, 91–110
17. J. T. Walker, K. Hanson, D. Cadwell, and C. W. Keevil, Bio- (1995).
fouling 12, 333–344 (1998). 37. C. J. van Oss, J. Mol. Recognit. 3, 128–136 (1990).
18. J. Dankert, A. H. Hogt, and J. Feijen, Crit. Rev. Biocompat. 38. C. J. van Oss, Interfacial Forces in Aqueous Media, Marcel
2, 219–301 (1986). Dekker, New York, 1994, pp. 18–46.
19. P. D. Marsh and M. V. Martin, Oral Microbiology, 3rd ed., 39. R. Bos and H. J. Busscher, Colloids Surf., B: Biointerfaces 14,
Chapman and Hall, London, U.K., 1992. 169–177 (1999).
 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES 113

40. H. J. Busscher, A. T. Poortinga, and R. Bos, Curr. Microbiol. 70. M. Elimelech, in M. Elimelech, J. Gregory, X. Jia, and
37, 319–323 (1998). R. Williams, eds., Particle Deposition and Aggregation.
41. J. Meinders, H. C. van der Mei, and H. J. Busscher, J. Colloid Measurement, Modelling and Simulation, Hartnolls, Bodmin,
Interface Sci. 164, 355–363 (1994). U.K., 1995, pp. 290–309.
42. P. J. Wit and H. J. Busscher, J. Colloid Interface Sci. 208, 71. I. Ofek, H. S. Courtney, D. M. Schifferli, and E. H. Beachy, J.
351–352 (1998). Clin. Microbiol. 24, 512–516 (1986).
43. M. Kostoglou and K. J. Karabelas, J. Colloid Interface Sci. 72. Y. Liu, Colloids Surf., B: Biointerfaces 5, 213–219 (1995).
171, 187–199 (1995). 73. G. Wirtanen, T. Alanko, and T. Mattila-Sandholm, Colloids
44. J. Czarnecki and P. Warszynski, Colloids Surf. 22, 207–214 Surf., B: Biointerfaces 5, 319–326 (1996).
(1987). 74. A. C. Taylor, in R. S. Manly, ed., Adhesion in Biological
45. D. M. Schaefer et al., J. Adhes. Sci. Technol. 9, 1049–1062 Systems, Microbial Cell Surface Analysis, Academic Press,
(1995). New York, 1970, pp. 51–71.
46. M. Soltani, G. Ahmadi, R. G. Bayer, and M. A. Gaynes, J. 75. N. Mohandas, R. M. Hochmuthand, E. E. Spaeth, J. Biomed.
Adhes. Sci. Technol. 9, 453–473 (1995). Mater. Res. 8, 119–136 (1974).
47. B. A. Jucker, H. Harms, S. J. Hug, and A. J. B. Zehnder, 76. H. J. Busscher, B. van de Belt-Gritter, and H. C. van der Mei,
Colloids Surf., B: Biointerfaces 9, 331–343 (1997). Colloids Surf., B: Biointerfaces 5, 111–116 (1995).
48. I. C. Hancock, in N. Mozes, P. S. Handley, H. J. Busscher, 77. H. C. van der Mei, B. van de Belt-Gritter, and H. J. Busscher,
and P. G. Rouxhet, eds., Microbial Cell Surface Analysis, VCH Colloids Surf., B: Biointerfaces 5, 116–126 (1995).
Weinheim, New York, 1991, pp. 21–59. 78. A. Abbott, R. C. W. Berkeley, and P. R. Rutter, in R. C. W.
49. H. H. M. Rijnaarts et al., Colloids Surf., B: Biointerfaces 4, Berkeley et al., eds., Microbial Adhesion to Surfaces, Ellis
5–22 (1995). Horwood, Chichester, U.K., 1980, pp. 117–142.
50. C. J. van Oss, J. Dispersion Sci. Technol. 12, 201–219 (1991). 79. T. Dabros and T. G. M. van de Ven, Phys. Chem. Hydrodyn.
8, 161–172 (1987).
51. H. H. M. Rijnaarts, W. Norde, J. Lyklema, and A. J. B. Zehn-
der, Environ. Sci. Technol. 30, 2877–2883 (1996). 80. R. B. Dickinson et al., Infect. Immun. 63, 3143–3150 (1995).
52. J. N. Israelachvili, Intermolecular and Surface Forces, Aca- 81. J. Sjollema, H. J. Busscher, and A. H. Weerkamp, J. Micro-
demic Press, London, U.K., 1987, pp. 31–47. biol. Methods 9, 73–78 (1989).
53. Y. Pradip, A. Attia, and D. W. Fürstenau, Colloid Polym. Sci. 82. J. Yang et al., Langmuir 15, 4671–4677 (1999).
258, 1343–1353 (1980). 83. A. F. M. Leenaars and S. B. G. O’Brien, Philips J. Res. 44,
54. G. W. Jones, L. A. Richardson, and D. Ulhman, J. Gen. 183–209 (1989).
Microbiol. 127, 351–360 (1981).
55. H. A. S. Vigeant and R. M. Ford, Appl. Environ. Microbiol.
63, 3474–3479 (1997). ADHESION (PRIMARY) OF MICROORGANISMS
56. Z. Liu and K. D. Papadopoulos, Appl. Environ. Microbiol. 61, ONTO SURFACES
3567–3572 (1995).
57. H. J. Busscher, M. M. Cowan, and H. C. van der Mei, FEMS CHRISTOPHE J. P. BOONAERT
Microbiol. Rev. 88, 199–210 (1992). YVES F. DUFRÊNE
58. I. Ofek and R. J. Doyle, in I. Ofek, eds., Bacterial Adhesion PAUL G. ROUXHET
to Cells and Tissues, Chapman & Hall, New York, 1994, Université catholique de
pp. 17–40. Louvain
59. I. Ofek, D. Mirelman, and N. Sharon, Nature 265, 923–925 1348 Louvain-la-Neuve,
(1977). Belgium
60. J. Noordmans, P. J. Wit, H. C. van der Mei, and H. J. Buss-
cher, J. Adhes. Sci. Technol. 11, 957–969 (1997). Microbial adhesion and biofilm formation are widespread
61. C. Gómez-Suárez, J. Noordmans, H. C. van der Mei, and phenomena that play a crucial role in natural environ-
H. J. Busscher, Langmuir 15, 5123–5127 (1999). ments as well as in industrial processes. Natural surfaces
62. C. Gómez-Suárez, J. Noordmans, H. C. van der Mei, and on which microbial adhesion occurs vary from inanimate
H. J. Busscher, Phys. Chem. 1, 4423–4427 (1999). materials to living tissues. In aquatic habitats, microor-
63. R. P. Ellen and R. J. Gibson, Infect. Immun. 5, 826–830 ganisms accumulate on other living organisms, suspended
(1972). particles, rocks, and sediments. Soil particles, often coated
64. O. H. Lowry, N. J. Rosebrough, A. L. Farr, and R. J. Randall, by clay minerals, metal oxides, and organic matter,
J. Biol. Chem. 193, 265–275 (1951). constitute another complex microbial habitat. Biofilm
65. E. H. Beachey and I. Ofek, J. Exp. Med. 143, 759–771 (1976). formation has a deleterious effect in many situations
66. J. M. Graham, in G. D. Birnie and D. Rickwood, eds., Cen- (commonly called biofouling): decrease of heat transfer
trifugal Separations in Molecular and Cell Biology, Butter- on heat exchangers, increase of pressure drop in pipelines,
worths, London, U.K., 1978. corrosion enhancement, and source of bacterial contam-
67. W. B. Clark, L. L. Bamman, and G. J. Gibson, Infect. Immun. ination (1). Development of biofilms on medical devices
19, 846–853 (1978). often requires their surgical replacement (2). Removal
68. M. A. Wassal, G. Embery, and J. Bagg, Colloids Surf., B: of biofilms using biocides (3) or antibiotics (2) is often a
Biointerfaces 5, 143–152 (1995). difficult task due to the protective role of extracellular
69. M. S. Lytle, J. C. Adams, D. G. Dickman, and W. R. Bressler, substances. On the other hand, biofilm formation may be
Appl. Environ. Microbiol. 55, 29–32 (1989). beneficial in water and wastewater treatment (4–6), in
114 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES

bioremediation processes (7,8), in the production of fer- METHODS TO STUDY MICROBIAL ADHESION
mented food (9,10), or in establishment of symbiosis on
plant roots (11). Overview
The formation of microbial biofilms at the interface
Generally speaking, two types of approaches are used in
between a bare solid surface and a cell suspension
experimental microbial cell attachment studies (23). In
may be seen as a succession of various steps, as
the first type, adhesion is evaluated in environments of
illustrated in Figure 1 (2,12,13). The solid substratum is
practical significance. To this end, test surfaces are placed
first conditioned by adsorption of organic (macro)molecules
in the environment of interest (water delivery system,
(proteins, polysaccharides, lipids; humic substances,
cooling system, soil, etc.), left for a period of time, retrieved,
natural organic matter,. . .) which diffuse faster than cells
and the number of attached bacteria is evaluated. Another
and are likely to modify the substratum physicochemical
approach, typical of laboratory studies (24), is more
properties. Transport of the microorganisms to the
appropriate for investigating mechanisms or establishing
substratum surface will occur by Brownian motion
guidelines. The aim is typically to determine (1) the
(i.e., diffusion), sedimentation, convection, and possibly
ability of one or several microorganisms to adhere,
also intrinsic motility. During this step, aggregation
(2) the influence of particular environmental conditions
occurring between planktonic microorganisms (14,15)
(ionic strength, pH, substratum nature) on adhesion,
may considerably affect the ‘‘simple vision’’ of one
and (3) the adhesion force between the microorganisms
microorganism approaching a surface. Attachment may
and the substratum. Usually, model materials are used;
occur between single or aggregated microorganisms and
they may be chosen according to their hydrophobicity
the conditioned substratum, which is referred to as
(hydrophilic such as glass, metal oxides, minerals;
the ‘‘primary adhesion’’ step. It is worth noting that
hydrophobic such as nonoxidized polymers), surface
coadhesion between planktonic and adhering cells may
electrical properties, roughness, or mode of conditioning.
also be encountered (15). Subsequently, the production
The results obtained may be strongly influenced by
of extracellular substances may allow the anchorage of
the hydrodynamic conditions (turbulent, laminar, resting
attached bacteria while bacterial growth leads to biofilm
liquid) and the rinsing step.
development. Later, the detachment of biofilm portions
Two limiting cases may be distinguished when
from the substratum may occur as a result of flow shear
considering the kinetics of particle attachment. On the
or chemical treatment (see BIOFILM DETACHMENT). It may
one hand, in the bulk transport controlled regime, the
also be induced by the enzymatic activity of entrapped
surface interactions create no resistance and the kinetics
cells (16,17).
is determined by convection, diffusion, and/or an external
The aim of this chapter is to provide an overview of
force field (sedimentation). On the other hand, in the
(1) the various methods available for studying the primary
surface barrier limited regimen, the bulk transport is
adhesion of microbial cells, and (2) the physicochemical
fast and the resistance is caused by physicochemical
approaches, which help to understand and control this
interactions between the two surfaces (25). The former
process. The importance of considering microbial cells
case will be mainly considered in this section; the latter
as complex colloidal particles is emphasized. However,
case will be mainly treated in the next section.
the role of specific (key-lock) interactions between
The experimental systems will be presented here
the microorganism and a biological substratum is not
according to the hydrodynamic conditions. Therefore,
considered in detail; more comprehensive details on this
it is worth recalling the types of flow which may be
subject may be found in (18–20).
distinguished on the basis of the Reynolds number, Re,

TERMINOLOGY vo
Re = (1)
ν
It has been proposed to use the term ‘‘deposition’’ to
designate the attachment of microbial cells to a surface, where vo is the fluid velocity, ν is the kinematic viscosity
and the term ‘‘adhesion’’ to refer to the force necessary to (Table 1) and  is a reference length scale (26,27). Regard-
separate the adherents, that is, to express the strength of ing the flow around colloidal particles (size ≤ 1 µm), Re is
the adhesive force (21). On the other hand, ‘‘adsorption’’ usually much lower than 1. In the absence of particular
was proposed to designate immobilization of an organism force fields (gravity, concentration gradients, interfacial
on a collector surface, and ‘‘attachment’’ to designate interactions), particles with sufficient symmetry follow
the consolidation of the interface between an organism the fluid streamlines. For higher Re, a laminar flow is
and a collector (22); however, these distinctions are not obtained, that is, the fluid streamlines are parallel to each
of universal use. In this article, in order to avoid the other. This is the case of the Poiseuille flow through a
confusion between the denomination of a phenomenon circular tube or a parallel-plate channel; the fluid velocity
(attachment, deposition, adhesion) and the name of a is 0 at the wall, while the velocity gradient, also referred
physical quantity characterizing the phenomenon or its to as the shear rate, is maximal at the wall and 0 at
result (rate, amount, strength); the terms deposition, the center. For Re above a critical value, typically 2,100
attachment, and adhesion are considered as synonymous. in pipes, unsteady and chaotic motion occurs, giving a
Adsorption is sometimes used as a synonym of attachment turbulent flow. However, near the wall, no random fluc-
or deposition; in this chapter, however, adsorption will be tuation occurs, thereby providing the laminar or viscous
reserved for the retention of molecules at a surface. sublayer. The concept of boundary layer finds its origin
 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES 115

(a) (b)

(c) (d)

(e) (f)

Figure 1. Schematic representation of the various steps involved in the formation of biofilms:
(a), initial situation; (b), substratum conditioning; (c), cell transport; (d), primary adhesion; (e),
anchorage; (f), growth.

in approximations, which involve a distinction between a cells present in a 1 mm layer of the suspension is needed
fluid behavior near a wall and far from a wall. to make a dense monolayer. The ratio of transport by
sedimentation with respect to diffusional transport may
Static Systems be evaluated by the Peclet number, Pe, also defined in
In static tests, the microbial suspension remains station- Table 1. For bacteria, Pe is about 300 (Table 1), indicat-
ary with respect to the substratum. Consider a quiescent ing that transport caused by Brownian motion may be
liquid containing dispersed particles; the transport of neglected compared to gravitational transport. The latter
which is governed by sedimentation and Brownian motion will control the transport to the upper face of a horizontal
(diffusion). Table 1 presents the physical quantities, which substratum immersed in a suspension of bacteria, but the
influence the rate of transport according to the two mecha- cells will not contact the lower face.
nisms (26,27). It gives also indicative values for nonmotile A bacterial suspension may also contain macro-
bacteria. The characteristic distance L, that is, the dis- molecules, which tend to be adsorbed onto the substratum.
tance along which the major concentration change takes A deposition test performed using a closed cell with a
place, may be evaluated as follows: a dense monolayer rectangular section showed that A. brasilense cells were
of bacteria represents about 5 × 107 cells/cm2 ; with a sus- attached on the lower wall and not on the upper wall;
pension concentration of 5 × 108 cells/cm3 , the amount of however a surface analysis showed that the amount of
116 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES

Table 1. Physical Quantities, with Indicative Values, Which Are Relevant to the
Transport of Bacteria by Sedimentation and Diffusion (26,27)

Physical Quantity Symbol and Definition Indicative Values

Bacteria equivalent radius a 0.8 µm

Acceleration of gravity g 9.81 m/s2
Viscosity coefficient η 1.0 × 10−3 Pa s, for water
Kinematic viscosity ν = η/ρliquid 1.0 × 10−6 m2 /s, for water
Sedimentation velocity vsed = 2 g a2 ρ/9η 0.8 cm/daya , 10−7 m/s
Difference of specific gravity ρ = ρcell − ρliquid 0.07 g/cm3b
Diffusion coefficientc D = kT/6π ηa 3 × 10−13 m2 /sd
Characteristic distance L 1 mm
Rate of transport by diffusion vdiff = D/L 3 × 10−10 m/sd
Particle fluxe J = vCo sediment.: 5 × 103 cells/cm2 s
diffusion: 15 cells/cm2 s
Peclet number Pe = vsed L/D 300d
a
close to the progression rate of the front of sedimentation measured in water for Azospirillum
brasilense and Pseudomonas strains (non published).
b
computed from vsed
c
k = Boltzmann constant, T = temperature (K)
d
computed using the above data
e
Co = initial bacterial concentration; indicative value of 5 × 108 cells/cm3

adsorbed proteins was similar for the lower wall after cell In practice, the cells are left to settle on the substratum
detachment, and for the upper wall. This demonstrated and the nonattached cells are removed by rinsing. These
that proteins were released by the cells into the solution tests are simple and quick, allowing rapid screening to
and adsorbed by the substratum, and were not exclusively be performed. However, the definition of adhering cells is
brought to the substratum by the cells (29,30). strongly dependent on the conditions used for the washing
A characteristic distance of 1 mm is still relevant for procedure. Several rinsing procedures are used to remove
the migration of a macromolecule to a surface: consider microorganisms that are ‘‘unattached’’ or ‘‘loosely bound’’:
a macromolecule with a molecular weight of 50 kDa, a successive dipping in the relevant fluid, mild shaking of
specific gravity of 1 g/cm3 and thus an equivalent radius the container with progressive replacement of the liquid,
of 2.7 nm; an adsorbed monomolecular layer represents or washing under flow. All of these procedures may yield
about 6 × 10−8 mole/m2 or 3 mg/m2 ; this is equivalent to nonreproducible results due to the lack of control over the
the amount present in a 1 mm thick layer of a solution at a removal force (33). Progressive dilution and mild shaking
concentration of 3 µg/cm3 . The diffusion coefficient of these avoid the passage of an air-water interface, and the
macromolecules is about 8 × 10−11 m2 /s. It turns out that possible influence of a water meniscus on the deposited
the ratio of the rate of sedimentation of bacteria (10−7 m/s) cells. The hydrodynamic conditions may be controlled
over the rate of diffusion of the macromolecules (8 × better by placing the test plates in a device, which ensures
10−8 m/s) is on the order of 1. This simple computation a laminar flow (31,34). The parallel-plate flow chambers
shows that, depending on the respective concentrations described later can also be used to investigate bacterial
and properties of macromolecules and bacteria, different adhesion under static conditions, while controlling the
situations may occur: most of the bacteria may get in hydrodynamic conditions used for rinsing (29).
contact with the native substratum surface or with a
surface fully modified by adsorption; the substratum Evaluation of the Strength of Adhesion
surface may also be progressively modified by adsorption An overview of earlier methods dealing with the
during the sedimentation process, which means that the determination of the strength of adhesion of small particles
substratum-cell interface will be different for bacteria can be found in (35). Among the direct methods, a force is
contacting the substratum surface at different times. applied in a direction opposite to the adhesive force and
By varying the cell concentration, the duration of the the value required to cause cell detachment is measured.
experiment and the height of the suspension, static tests The centrifugal method is attractive due to its apparent
allow the amount of cells brought in contact with the simplicity; the applied force is given by
support to vary from a small fraction of a monolayer
to a thick sediment. The thickness of the sedimented FC = Vp (ρp − ρm )ω2 r (2)
layer was found to influence the density of adhering yeast
cells under conditions of low cell-support affinity; this where Vp is the particle volume, ρp and ρm are the
was attributed to an increased pressure exerted on the densities of the particles and medium respectively, ω is the
bottom layer of cells or to a reorganization leading to the rotational speed and r is the radius of gyration. However,
formation of a dense layer of cells in close contact with the in the case of microbial cells, this method is strongly
substratum (31,32). limited as a result of the small difference of density
 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES 117

between the cells and the usual types of surrounding the disk and its velocity is vz . Within the boundary layer,
liquid medium. Experiments performed with yeast cells the radial and tangential velocity components are not 0.
adhering to glass in an aqueous medium showed that an If the suspended particles are irreversibly captured when
acceleration up to 5,700 g did not cause cell detachment they arrive close enough to the collector surface (‘‘perfect
(unpublished). Centrifugation in air of a sample tilted sink’’ model), the flux of particles to the disk surface
with respect to the rotation axis allowed a cell detachment J may be considered as being controlled by diffusion
threshold to be determined. However, the interpretation through a diffusion boundary layer. The thickness of
of the data was not clear, because of possible interferences the latter, δ, is constant over the entire disk surface,
from capillary forces created by the water meniscus at the neglecting edge effects, and is much smaller than δo .
cell-substratum interface and of drying. Table 2 presents relevant physical quantities for two disk
Indirect methods consist of measuring the force rotation frequencies. It gives also the apparent rate vp
required to provoke cell detachment exerted by a liquid at which the particles are transported to the surface.
flow parallel to the surface (35,36); the force is thus of This is larger than the rate of diffusional transport in
hydrodynamic origin. The drag force FH exerted by a static conditions (Table 1), because δ is smaller than
laminar liquid flow on a spherical particle of radius a the characteristic distance used in the latter case. At a
adhering to a wall is given by rotation frequency of 1 cycle/s, it becomes close to the
sedimentation velocity that should no longer be neglected;
FH = 32a2 τ (3) note also that δ becomes close to the cell radius. The
influence of particle size and density, and of Re is further
where the wall shear stress τ is given by discussed in (25). The rotating disk method has been
used to study deposition of microorganisms associated
dv with dental caries, namely, Streptococcus mutans (40),
τ =η (4) and with infections of biomaterials after blood contact,
dz
namely, Staphylococcus epidermidis (41).
where v is the fluid velocity and z the direction The radial-flow chamber (36) mentioned earlier has
perpendicular to the surface. Equations have been been used under conditions of microbial growth, by
developed for the following systems: (i) flow between pumping the culture fluid at a constant volumetric rate.
rotating concentric cylinders (21,35); (ii) flow created The linear fluid velocity, and hence the surface shear stress
between two parallel discs with a narrow spacing, by decreases radially across the disc. With a suitable flow
injection of the fluid at the center of one disc (radial- rate, a clear zone is found around the inlet, with attached
flow chamber) (36,37); and (iii) flow in a rectangular pipe cells beyond a certain distance from the inlet. This
having a small thickness (38,39). In the latter case, ‘‘critical radius’’ gives the critical shear stress below which
provided a laminar flow is ensured in the pipe, attachment takes place. The radial-flow chamber was
combined with automated videomicroscopy, which allowed
a2 the kinetics of bacterial attachment and detachment to be
FH = 96νvlo (5) determined (24,42,43).
b
Deposition of colloidal particles has also been studied
where vlo is the average linear flow rate and b is half the using the impinging jet method (25,44,45). The dispersion
pipe thickness. However, the drag force required to detach flows through a circular hole and the fluid impinges
an adhering cell has a clear meaning only when the cells on the substratum. If the latter is transparent, the
are isolated and distributed with a low density on the deposition process can be observed under the microscope.
substratum. A stagnation point flow is created near the substratum
surface facing the inlet hole. Mass transfer equations were
Dynamic Tests Under Controlled Hydrodynamic Conditions solved for the stagnation point region, taking into account
hydrodynamic, gravitational, and interfacial forces. The
Hydrodynamic conditions influence deposition tests in two variation S of the density of deposited particles, as a
ways: (i) the liquid flow governs the rate at which the function of time, can be approximated by
suspended cells make contact with the substratum surface
and, consequently, cell deposition; (ii) it also influences J0
the shear stress near the substratum surface, thus the S= [1 − exp{−(βm + βe )t}] (6)
βm + βe
residence time of nonattached cells at a given spot of the
surface and the drag force exerted on attached cells. where J0 is the initial flux of particles towards the collector
The transport of dissolved compounds or suspended surface, βm is a coefficient reflecting the surface blocked
particles in a moving liquid is governed by two per particle, and βe is an escape coefficient depending on
mechanisms, convection and diffusion, the combination the transport conditions near the collector surface and on
of which is called convective diffusion (26). The equations the particle-support interactions. J0 can be determined
describing the convective diffusion have been solved when from the initial slope of S versus t; J0 , βm , and βe
the collector is a smooth rotating disc immersed in an can be analyzed in relation to the characteristics of
otherwise non agitated liquid (25–27). A fluid flow of the system investigated, according to mass transport
well-controlled intensity is created by the rotation. At a equations.
distance from the disk larger than δo , the thickness of the A theoretical analysis was made for the particle
hydrodynamic boundary layer, the liquid moves toward deposition kinetics onto walls of parallel-plate and
118 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES

Table 2. Physical Quantities, with Indicative Values, Which are Relevant to the
Transport of Bacteria to a Rotating Disc of Radius r = 10 cm at Steady State (26)a

Physical quantity Symbol and definition Indicative values Units

Rotation frequency — 10−2 1 cycle/s

Angular velocity ω 6.3 × 10−2 6.3 radius/s
√
Rate of the liquid to the vz ≈ 0.89 νω 2.2 × 10−4 2.2 × 10−3 m/s
surface at a large
distance from the disc
√
Thickness of the δo ≈ 3.6 ν/ω 1.4 × 10−2 1.4 × 10−3 m
hydrodynamic
boundary layer
Thickness of the diffusion δ ≈ 0.5(D/ν)(1/3) δo 4.7 × 10−5 4.7 × 10−6 m
boundary layer
Rate of transport of the vp = J/Co = D/δ 6.4 × 10−9 6.4 × 10−8 m/s
particle to the surface
a
ν taken for water, D and Co taken for bacteria, as given in Table 1.

cylindrical channels through which a dilute suspension of Batch Systems and Packed Beds
spherical particles moves in laminar flow (46). Transport
With batch systems, the amount of adhering cells is often
equations were formulated by taking into account surface determined by the difference between the initial and final
forces as well as external forces. The parallel-plate flow concentration of microorganisms in the liquid phase. A
chamber has been broadly used to investigate microbial direct quantification of adhering bacteria can be achieved
adhesion (12,47). Therefore, the substratum of interest, by using cell labelling techniques, with fluorescent or
which must be transparent, constitutes the main chamber radioactive probes (24,53,54).
walls and the cell deposition is observed directly under the In packed-bed systems, which are relevant to soil or to
optical microscope. The adhesion kinetics may be analyzed industrial bioreactors (55), the retention of bacterial cells,
by using relations similar to Eq. (6) (48). assumed to be caused only by attachment, is deduced from
A flexible and frequently used device to study the concentration of cells remaining in the effluent and
microbial adhesion is the modified Robbins device (49–51). recording breakthrough curves (56–59). A disadvantage
Aseptically removable studs contain a dish of the is that direct observation and localization of adhering
substratum to be examined, which is exposed to the fluid microorganisms is not possible. In particular, microbial
passing through a pipe. This is a convenient method to cells may be held in the bed by sedimentation in zones
compare different materials in the same conditions. of low liquid velocity or by a mechanical retention of
aggregates, which may erroneously be considered as
Table 3 shows a comparison of different systems regard-
attachment.
ing the possibility of examining a large substratum area,
Rijnaarts and coworkers (60) compared bacterial
the possibility of direct observation, which facilitates the
attachment performed on glass and teflon surfaces, using
quantification of deposition kinetics, and the possibility packed-beds and batch systems, with a large collection
of comparing different materials in the same run. Other of Pseudomonas and coryneform bacteria selected accord-
systems consist of exposing several substratum samples ing to their environmental and biotechnological relevance.
in a flow chamber or on a rotating cylinder. They allow They showed that the transport of microbial cells from
a comparison of different substratum samples to be made the bulk liquid to the substratum surface was faster in
in the same conditions, but provide a less accurate control packed-bed columns (transport ensured by diffusion and
of the hydrodynamic conditions. The relation to biofilm convection) compared with batch tests (transport domi-
reactors can be made using reference (52). nated by diffusion). The two methods were consistent for a

Table 3. Comparison of Different Methods to Study Microbial Deposition

Simultaneous study of
Method Large area Direct observation∗ different samples

Rotating disc yes no no

Radial-flow chamber yes yes no
Impinging jet no yes no
Parallel-plate flow cell yes yes no
Modified Robbins device no no yes
∗
provided the substratum is transparent.
 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES 119

certain number of strain-substratum combinations. Devi- that two regions can be identified near a charged
ations observed for other combinations could be attributed surface of potential ψ0 , as illustrated by Figure 2 (65).
to method-dependent effects resulting from specific cell In the first region, called the Stern layer, the ions,
characteristics such as ability to aggregate, presence of chemically adsorbed or retained by localized electrostatic
capsular polymers, large cell sizes or detachment by an attraction, are in close contact with the surface and the
air-liquid meniscus. potential is considered to fall linearly as a function of
the distance from the surface. In the second region,
PHYSICOCHEMICAL ASPECTS OF PRIMARY ADHESION called the diffuse double layer, the potential falls
exponentially as a function of the distance, caused
In the course of the approach of a substratum by a by the balance between electrical interactions of ions
microbial cell, physicochemical aspects become important with the surface and molecular motion. The electrical
when the separation distance is smaller than the cell size. potential at the inner limit of the diffuse double layer
Depending on the nature of the surfaces involved, they is called ψd . The zeta potential, ζ , is the potential
may be the factor controlling the rate of attachment and at the plane of shear between the surface and the
even determining the possibility of attachment. Former solution.
reviews can be found in (21,61–63). Considering a sphere (cell) and a plate (substratum),
GE is expressed as follows for small values of the relevant
DLVO (Derjaguin, Landau, Verwey, Overbeek) Theory electric potential (ψ < 50 mV) (62,66,67)
Theoretical Frame. According to DLVO theory (28,64),
 
2ψM ψS 1 + exp(−κh)
the free enthalpy of interaction, GT , when two surfaces GE = π εa(ψM
2
+ ψS2 ) ln
approach each other, may be accounted for by the ψM2
+ ψS2 1 − exp(−κh)

sum of two contributions, corresponding respectively to + ln(1 − exp(−2κh)) (8)
electrostatic and van der Waals interactions
where ε is the permittivity of the suspending fluid; a is the
GT = GE + GV (7)
cell radius; ψM and ψS are the values of ψd near the cell and
the substratum surfaces, respectively; κ is the reciprocal
The electrostatic contribution, GE , is caused by the
of the Debye length; h is the separation distance between
overlap of electrical diffuse double layers. Remember
the interacting surfaces (28,68). For computing GE , the
value taken for ψd is usually the zeta potential (68).
Stern layer κ determines the rate of the variation of the electrical
potential as a function of the distance within the diffuse
double layer
Diffuse layer
κ = [(2,000Ie2 N)/(εkT)](1/2) = 3.29 × 109 (I)(1/2) (at 25 ° C)
(9)
Ψ0 Surface of shear
where I is the ionic strength of the bulk solution, e is
the electron charge, N is the Avogadro number, k is the
Boltzmann constant, and T is the temperature. At ionic
strengths of 10−1 , 10−3 , and 10−5 M, κ −1 is close to 1, 10
Ψd
and 100 nm, respectively. Note that the ionic strength
ζ
is commonly in the range of 10−2 to 10−3 M for natural
freshwater and near 10−1 M for most microbial culture
media. Strictly speaking, h should represent the distance
between the inner limits of the two diffuse layers facing
each other. Considering that it is the distance between the
Potential

0
two surfaces neglects the thickness of the Stern layer. For
Distance
this reason and because of the uncertainties concerning
the definition of ψd , computing GE for distances shorter
than 1 nm is meaningless.
The contribution of van der Waals interactions varies
with the separation distance according to
ζ
 
A 2a(h + a) h + 2a
Ψd GV = − − ln (10)
6 h(h + 2a) h

Figure 2. Potential decay curves for an electrical double layer which may be simplified for very small separation distance
associated with a surface of potential 0 . The upper and lower (h a), giving
curves arise, respectively, from weak and strong adsorption of Aa
counterions in the Stern layer. GV = − (11)
6h
120 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES

where A is the Hamaker constant (62,66,68). This favor microbial attachment by lowering the magnitude
is mainly determined by dispersion forces (surface- of the potential energy barrier. Such variation of surface
surface and surface-liquid interactions) and thus by the potential can be encountered when comparing different
polarizability of the surface constituents. For bacterial bacterial strains, when comparing different substrata or
cells suspended in water, A is estimated to be about 0.035, when changing the pH of the surrounding solution.
1.54 and 3.58 kT for Teflon, glass, and iron oxide substrata, Experimental data obtained during the last decades are
respectively (kT = 4.1 × 10−21 J/mol) (58,64,69). in good agreement with trends predicted by DLVO theory.
Figure 3 presents the computed GT vs h curves for A more negative zeta potential of biotechnologically
microbial cells, considered as spherical particles of 1 µm relevant microorganisms (70) and of soil bacteria (71), both
radius, A = 1.54 kT, ψd of both cells and substrata = hydrophilic, was shown to be unfavorable to attachment.
−20 mV, and κ −1 = 10 nm; the respective influence of Electrostatic repulsion may be reduced in different ways:
variation of A, ψd , and κ −1 is also illustrated. The (i) by increasing the ionic strength, (ii) by bringing the pH
attractive contribution of van der Waals forces is shown to a value closer to the cell isoelectric point, or (iii) by
in Figure 3 (center; d = 0 mV). At physiological pH, most modifying the surfaces.
of biological and substratum surfaces have a negatively
charged surface. As a consequence, an overall repulsion Surface Modification. The role of electrostatic interac-
appears at a distance that may vary from 10 to hundreds tions is illustrated by the influence of substratum mod-
of nanometers, depending on ionic strength (Fig. 3). The ifications that confer a positive character to the surface.
repulsion energy must be put in perspective, considering Figure 4 (left) presents a micrograph of a glass plate which
the kinetic energy associated with Brownian motion, had been partially immersed for one hour in a solution of
equal to 1.5 kT. Figure 3 (left, center) shows that the 1.8 mM Fe(NO3 )3 brought at pH 4, washed with distilled
chemical composition of the surfaces involved acts much water and used as substratum for a static adhesion test
more through the ionized functions, which determine the of Saccharomyces cerevisiae suspended in water (unpub-
electrical potentials, than via the van der Waals forces. An lished). The cells did not attach to untreated glass but
increase of ionic strength from 10−5 to 10−1 M (reducing formed a denser layer on the modified substratum. Hydrol-
the Debye length from 100 to 1 nm) reduces considerably ysis of ferric ions produces ferric hydroxide particles of
the magnitude of the potential energy barrier (Fig. 3, about 10 nm size, which adsorb to negatively charged sur-
right). At sufficiently high ionic strength, the variations faces; this provokes a charge reversal that allows adhesion
of GE and GV as a function of the distance bring about of a dense layer of yeast cells (72).
a secondary minimum, which may keep the cell near the Figure 4 (right) presents data from static adhesion
substratum, although not in close contact. As illustrated tests performed with S. cerevisiae and a glass substratum
in Figure 3 (center), a decrease of surface potential may modified by particles of hydrous alumina with a 0.25 µm

1500
κ −1 = 10 nm κ −1 = 10 nm Ψd = −20 mV
Ψd = −20 mV A = 1.54 kT A = 1.54 kT

1000
−30 mV
0.035 kT
500 100 nm
GT( kT)

10 nm
1.54 kT
−20 mV
0
1 nm

−500 0 mV
3.58 kT

−1000
1 10 100 1 10 100 1 10 100
h (nm) h (nm) h (nm)
Figure 3. Variation of the free enthalpy of interaction as a function of the distance (h) between
a sphere of 1 µm radius and a plane substratum, according to DLVO theory. Middle curve:
Hamaker constant A = 1.54 kT; potential of the diffuse double layer d = −20 mV; Debye
length κ −1 = 10 nm. Left: d = −20 mV, κ −1 = 10 nm, A varying from 0.035 to 3.58 kT; center:
A = 1.54 kT, κ −1 = 10 nm, d varying from 0 to −30 mV; right: A = 1.54 kT, d = −20 mV, κ −1
varying from 1 to 100 nm.
 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES 121

Amount adhering × 10−6 (cells / cm2)

3

Figure 4. Left, micrograph of a

1 glass slide partially immersed in a
Fe(NO3 )3 solution and subsequently
submitted to a static adhesion test of
S. cerevisiae. Right, variation of the
0 amount of S. cerevisiae attached to
a glass substratum as a function of
0 20 40 60 80 100
the degree of coverage with hydrous
Degree of coverage by alumina (%) alumina particles.

diameter and an isoelectric point at pH 9.3 (unpublished). sand particles (56,86). In similar experiments performed
Different surface coverages of the substratum surface with negatively charged glass and Teflon beads, the
with alumina particles were obtained by controlling retention of several negatively charged Pseudomonas and
the alumina concentration and were quantified by coryneform bacteria was enhanced by increasing ionic
redissolution of the coating and chemical analysis. Below strength (57,59). Similar effects were reported for the
20% coverage, the yeast cells were distributed in patches attachment of these microorganisms to glass and Teflon
on the glass support. The amount of adhering cells plates (87), of several food-borne pathogens to stainless
increased with the degree of coverage by alumina up to steel (88,89), of biomaterial pathogens to glass and to
65% coverage, in which case it reached the amount typical siliconized glass (50). The opposite was observed with a
of a densely packed cell layer. positively charged bacterium (57).
Table 4 presents an overview of various systems The effect of electrolytes may sometimes be more
in which treatments have been applied either to the complex. Deposition of a Pseudomonas strain studied in
cells or the substrata to promote primary microbial sand columns was more efficient in solutions containing
adhesion by decreasing electrostatic repulsion or achieving Mg2+ as the dominant cation compared with solutions
electrostatic attraction. These treatments include the use of identical ionic strengths but containing Na+ . The
of positively charged colloidal particles (hematite, Fe2 O3 ; authors showed, by electrophoretic measurements, that
hydrous alumina, Al(OH)3 ), of cations (Fe3+ , Al3+ ) and of this could be explained by adsorption of Mg2+ within the
polycations either adsorbed (chitosan, diethyl aminoethyl- Stern layer, reducing the zeta potential (90). Williams
dextran, polyethyleneimine) or grafted (aminosilane); and Fletcher (91) showed that, although the attachment
further details may be found in (73). One may argue that of Pseudomonas fluorescens to surface-treated polystyrene
these treatments may affect cell physiology. However, for (water contact angle of 66° ) was increased with a rise
various microorganisms, it was shown that the adhering of ionic strength, the inverse effect was observed on
cells retained their metabolic activity (34,74,75) or their native polystyrene (water contact angle of 90° ). This
germinating and growing capabilities (76). apparent contradiction may be explained by the respective
It must be kept in mind that, when immersed in a influence of electrostatic interactions (cfr. DLVO theory),
solution, the positively charged substrata may tend to which would dominate in the case of the more hydrophilic
adsorb multivalent or organic anions, which will reduce or surface-treated polystyrene, and of other interactions (see
eliminate the effect of pretreatment. This can be examined below). The attachment of Aureobasidium pullulans, a
by measuring the electrokinetic properties of the cells and fungus associated with the deterioration of polyvinyl
of the substratum (34,65,72,85). chloride and painted wood surfaces was maximum
at intermediate ionic strength (92). The explanation
Influence of pH and Ionic Strength. The effect of ionic proposed by the authors was that the electrolyte inhibited
strength on adhesion may be easily understood from attachment, either by screening short-range electrostatic
Figure 3 (right): the decrease of Debye length with attraction between oppositely charged groups, or by
increasing ionic strength causes a lowering of the potential modifying the conformation of cell surface molecules.
barrier, which has to be passed by the cells in order In many instances, the isoelectric points of both cells
to attach. In studies of the transport of soil bacteria and substratum are low (in the range of 1 to 4) and the
through sand columns, an increase of ionic strength zeta potential will generally become less negative when
decreased the recovery of cells in the effluent, which decreasing the pH, for a given ionic strength. Accordingly,
may be attributed to an increase of cell attachment on the potential barrier preventing cell attachment will
122 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES

Table 4. Immobilization of Microorganisms by Adhesion Through Modification of the Surface

Charge of the Cells or of the Substratum

Treatment of∗
Microorganism Substratum Cells Carrier References

Saccharomyces cerevisiae glass, polycarbonate Al3+ , Fe3+ — (34,74)

Fe2 O3 , Al(OH)3 (74)
glass, metals, organic — Fe3+ (70,72)
polymers
Kluyveromyces lactis, glass — chitosan (77,78)
Klebsiella oxytoca
Arthrobacter simplex glass Al3+ — (75)
— Al (OH)3 (75)
Acetobacter aceti glass, mica, organic — Fe3+ (70,79)
polymers
Corynebacterium glass — DEAE-dextran, (80)
glutamicum aminosilane
Xanthomonas campestris glass Al3+ — c

— Al3+ , Al (OH)3 c

Escherichia coli glass Al3+ Al3+ c

Bacillus licheniformis glass, organic polymer — Fe3+ , chitosan, c

DEAE-dextran,
aminosilane
Syntrophomonas wolfei glass — Fe3+ (81)
Acetobacter cotton cloth PEIa PEI (82)
Lactobacillus casei glass — PEI (83)
Phanerochaete glass, organic polymer — Fe3+ , PEI, PAAb (76)
chrysosporium
Baker’s yeast glass PEI PEI (84)
∗
Al3+ , Fe3+ , chitosan, PEI, PAA, and DEAE-dextran refer to treatment by solutions of the indicated substance; Fe2 O3 and
Al (OH)3 refer to coating by colloidal particles; aminosilane designates a polymer coating.
a
polyethyleneimine.
b
polyacrylamide.
c
unpublished.

decrease or will be suppressed (Fig. 3, center). The density energies (γ )

of attached cells has indeed been shown to increase when Gadh = γMS − γML − γSL (12)
the pH of the suspending medium was brought near
the isoelectric point. This was observed for Moniliella The interfacial free energy is the sum of several terms,
pollinis (a yeast — like fungus) (70), A. pullulans (92), and each of which corresponds to different interactions: tran-
A. aceti (79). sient dipole-induced dipole interactions or dispersion
Besides cell-substratum interactions, cell–cell interac- forces (London), permanent dipole-permanent dipole inter-
tions may be affected by the pH and the ionic strength of actions (Keesom), permanent dipole-induced dipole inter-
the cell suspension. In this context, an increase of ionic actions (Debye), hydrogen bonding and ionic bonding (93).
strength was shown to become unfavorable to attachment Like all systems in nature, the system made by these
because of the formation of aggregates of P. chrysosporium interacting surfaces will strive to reach a state of mini-
conidiospores, the latter presenting higher susceptibility mal free energy. Therefore, microbial attachment will be
with respect to shear during rinsing (76). Also, flocculation favored when Gadh is negative.
of M. pollinis was found to compete with adhesion when Surface free energies are deduced by using the Young
decreasing the pH towards the cell isoelectric point (70). equation
γLV cos θ = γXV − γXL (13)
Balance of Interfacial Free Energies
where X stands for solid (either substratum, X = S, or
Conceptual Frame. This approach considers the adhe- microorganism, X = M) and V for the saturated vapor
sion process as the creation of a new microorganism- phase; θ is the contact angle measured on a lawn of
substratum (MS) interface, and the destruction of two microorganisms or on the substratum; γLV is the surface
preexisting interfaces, that is, the microorganism-liquid tension of liquid L. In this relation, two parameters can
(ML) and substratum-liquid (SL) interfaces. Thus, this be experimentally assessed (θ and γLV ). The other two
involves a molecular contact between the two adhering parameters can be tentatively deduced using equations
surfaces. The tendency of the surfaces to associate is which relate γXL to γXV and γLV . Similar equations relate
expressed as a balance of the relevant interfacial free γMS to γML and γSL , allowing Gadh to be computed.
 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES 123

Several approaches have been used to establish these surface tension (γMV ) lower than that of water (γLV ),
equations (32,94–96). One key point among the diverging was studied on different solid substrata presenting
hypotheses is whether γXV may be approximated by γX , the a large range of surface tensions (γSV ): as expected
surface free energy of X under vacuum. from the Equation of State approach, the adhesion
One approach assumes an Equation of State to relate density decreased with increasing γSV . Moreover, when
the three interfacial free energies involved in the Young dimethyl sulfoxide was added to water, the inverse
equation (97). It allows γX to be calculated from contact correlation was observed when the added amount was
angles measured with one liquid. According to Spelt sufficient to decrease the value of γLV below that of
and coworkers (98), apolar and polar liquids of the same γBV (100). Similar results were reported by Fletcher
surface tension should give the same θ value for a given and Pringle for P. fluorescens (105). More recently, Wang
solid; however, conflicting results are reported regarding and coworkers (41) found that interfacial free energies,
this aspect (99). According to this approach, if γLV is assessed by the Equation of State, are strong driving
lower than γMV , Gadh decreases with increasing γSV , forces for adhesion of S. epidermidis to a large range of
predicting increasing bacterial adhesion with increasing blood-contacting biomedical materials.
surface tension of substrata. On the other hand, when γLV The Owens & Wendt approach was used to pre-
is higher than γMV , the opposite trend is observed (100). dict the adhesion of three bacteria encountered in
In another set of approaches one assumes the additive implant infections (E. coli, Pseudomonas aeruginosa and
contribution of dispersion and polar surface free energy S. epidermidis) to a large selection of orthopedic implant
components (γ d and γ p , respectively) in the total surface polymers (106). The comparison between computed Gadh
free energy (γ = γ d + γ p ) (101). While γ d is determined and experimental results revealed the following features:
by London dispersion forces, γ p comprises hydrogen (1) as predicted by the negative values of Gadh , significant
bonding, Keesom, and Debye interactions. Moreover, it adhesion was always detected whatever the microorgan-
is considered that the dispersion and polar contributions ism and the substratum; (2) the least negative value of
of the interfacial energy γXY (where X and  Ypstandforp M,p S, Gadh obtained for S. epidermidis was related to a lower
or L) combine pairwise [γXY d
= f γXd , γYd ; γXY = f γX , γY ]. adhesion density; (3) when comparing the polymers, an
In this frame, the methods differ according to whether the inverse correlation between attached cell density and
pairwise combination involves a harmonic- or a geometric- Gadh was expected for the three bacteria, whereas it
mean equation (102), and whether the vapor adsorption was observed only for P. aeruginosa.
by the solids is neglected or not (103). The most frequently Predictions based on the thermodynamic approach
used approach, the Owens-Wendt method, is based on are not always in agreement with experimental data.
the geometric-mean equation and neglects the influence of According to the Owens-Wendt approach, a positive
vapor adsorption; it involves the use of two liquids, usually correlation between attached cell density and γS indicates
α-bromonaphthalene (apolar, γL = γLd ) and water (highly that the polar component of the bacterial surface free
p
polar, γL = γLd + γL ). The γLd value of water is determined p
energy M must be higher than γL (104). Nevertheless, the
from the water contact angle on a solid allowing only attachment of the teeth-colonizing Streptococcus sanguis
p
London dispersion forces (polyolefin). The dispersion bacteria (γM = 52 mJ/m2 ) in phosphate buffered saline
p
component does not vary much with the nature of the (γL = 48 mJ/m2 ) was decreased with increasing γS (107).
p
substratum considered. As a consequence, if γM is lower Adhesion of L. monocytogenes to relevant industrial
p p
than γL , adhesion will be favored if γS decreases (104). This surfaces (polypropylene, rubber, glass, and stainless steel)
fits the common understanding of hydrophobic bonding. could not be correlated to the predictions when using
A variant considers that γXV is different from γX , the the van Oss approach (108). Actually, the degree of
difference being due to vapor adsorption by the solid (103). success of the prediction of the adhesion of Streptococcus
Another type of approach proposed by van Oss (95) thermophilus and Leuconostoc mesenteroides, both at the
takes into account not only the polar character but also its origin of contamination in dairy processing, depended
electron-donor or electron-acceptor nature. It emphasizes greatly on which thermodynamic approach was used
that nonpolar interactions result from all three types of to convert measured contact angles into surface free
van der Waals interactions, lumped together in Lifshitz energies (94).
theory, giving γ LW (LW for Lifshitz-van der Waals). A more global but simpler approach is to consider
Further, the polar interactions are designated as Lewis the influence of surface hydrophobicity, as assessed
acid-base (γ AB ), resulting from the combination of electron- from water contact angle measurements. The direct
donor (γ − ) and electron-acceptor (γ + ) contributions. This correlation between hydrophobicity, determined either on
method involves the use of at least three different liquids, cell or substratum surface, and microbial adhesion was
such as water and formamide (two polar liquids), and demonstrated in several studies. A study of eight strains
α-bromonaphthalene or diiodomethane (assumed to be of Coryneform and four strains of Pseudomonas showed
apolar). denser attachment and stronger adhesion with increasing
surface hydrophobicity of both bacteria and substrata,
Experimental Data. The validity of approaches on the the latter being glass, polystyrene, and Teflon (58).
basis of a balance of interfacial energies for predicting The deteriogenic fungus A. pullulans adhered in higher
microbial adhesion has been tested. The adhesion of five amount on substrata of higher hydrophobicity (92). The
bacteria (two strains of E. coli, Staphylococcus aureus, incorporation of a biocide into polyvinylchloride provoked a
S. epidermidis, and Listeria monocytogenes), having a reduced attachment of P. fluorescens, isolated from biofilm
124 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES

developed on a swimming pool liner; this was partially would indicate the electron-donor and electron-acceptor
attributed to a reduction of polymer hydrophobicity (109). character of the bacterial surface, respectively. Recently,
The attachment of soil bacteria to sulfonated polystyrene, the adhesion behavior of L. monocytogenes to stainless
a hydrophobic substratum, was directly correlated with steel was examined in the light of this method (89). The
their water contact angle values, which were between parameters γ LW , γ + , γ − were assessed by the MATS
20 and 70° depending on the microorganism. It may be method for bacteria and by the Van Oss approach for
noted that, while electrostatic interactions were always stainless steel: both surfaces had strong electron-donor
repulsive, the effect of cell surface charge was visible and weak electron-acceptor characteristics. Nevertheless,
only for more hydrophilic strains. Furthermore, the a correlation between adhesion results to stainless steel
attachment of the same microorganisms to glass, a and cell affinity for ethyl acetate was observed, indicating
hydrophilic substratum, was shown to be controlled by that Lewis acid-base interactions play an important role
electrostatic interactions, the cell surface hydrophobicity in the adhesion mechanism. It may be noted that this
having only a small effect (71,110,111). Correspondingly, approach is equally open to criticism. Indeed, MATS is
the attachment of hydrophilic microorganisms was found very sensitive to the surface area developed by the solvent
to be governed by electrostatic interactions (69). While droplets created during mixing of the two liquid phases,
these results demonstrate that water contact angles which can depend on the mixing conditions (temperature,
point to the role of hydrophobicity in the attachment type of mixing vessel . . .). Moreover, an emulsion is
mechanisms, they also illustrate that attachment results frequently produced and the cells tend to accumulate at
from a combination of hydrophobic and electrostatic the interface between the two solvents. As a consequence
interactions. there is no simple partition of the cells between two
defined liquid phases and the measured decrease of the
Limitations of Classical Physicochemical Approaches cell concentration in one liquid phase may be misleading.
Moreover, the cells may have an amphiphilic character.
Although the DLVO theory and the approach based on a
Finally, the usage of organic solvents was reported to affect
balance of interfacial free energies have provided valuable
the bacterial integrity (113).
insight into the mechanism of microbial adhesion, as
illustrated earlier, they suffer from several limitations.
In the DLVO theory, (i) the interacting surfaces are COMPLEXITY OF MICROORGANISM — SUBSTRATUM
assumed to be chemically homogeneous, smooth, rigid, and INTERFACES
nonpenetrable by ions; (ii) only long-range interactions,
that is, electrostatic and van der Waals interactions, This section is devoted to particular interfacial aspects,
are considered, neglecting short-range interactions like which may influence markedly the primary bacterial
hydrogen bonding. In the approach based on the balance adhesion. Some features have already been pointed out as
of interfacial free energies, (i) the surfaces are considered limitations for the classical physicochemical approaches.
as homogeneous and in molecular contact; (ii) the system
Surface Roughness
is considered to be at equilibrium; (iii) the interfacial
free energy computations rely on theoretical frames Contrary to frequent assumptions made to model the cell-
that lack accuracy; (iv) electrostatic interactions between substratum interactions, cell and substratum surfaces are
approaching surfaces are not taken into account because not atomically smooth. On the one hand, the presence
the surface energies are determined from contact angles of morphological heterogeneities on the substratum
of liquids. (roughness) may influence the biofilm formation. With
The extended DLVO theory includes an additional regard to initial adhesion, a few reports have indicated
term GAB , compared to classical DLVO theory, in order that roughness has a minor effect. The adhesion of
to account for short-range ‘‘acid-base’’ interactions (95). S. thermophilus and L. monocytogenes to stainless steel
According to van Oss, these acid-base interactions form samples covering a range of finishes, representative
the basis of the hydrophobic or hydrophilic interactions. of food industry equipment, did not vary significantly
It is important to note that these acid-base interactions according to substratum roughness (88,114). In the
are short-range, that is, they become significant at small same way, the density of P. aeruginosa, E. coli, and
separation distance (5 to 10 nm) (12). S. epidermidis attached on cast biomedical films was
While examining the relevance of the extended DLVO similar to extruded films, the latter being rougher
theory, Bellon-Fontaine and coworkers (112) developed a than the former (106). However, surface roughness
partitioning method, called microbial adhesion to solvents at the micrometer scale appears to be favorable by
(MATS), which is based on comparing microbial cell protecting the cells against shear stress or antimicrobial
affinity to a polar solvent (electron acceptor or electron substances (115,116). Moreover, in conditions where the
donor) and to a nonpolar solvent. The following pairs of adsorption of a conditioning film plays an important role in
solvents were used: (1) chloroform (electron acceptor) and the adhesion process, one may anticipate that substratum
hexadecane (nonpolar); (2) ethyl acetate (electron donor) roughness at the nanometer scale will affect cell adhesion
and decane (nonpolar). Due to the similar Lifshitz-van via changes in the organization of the adsorbed layer. It
der Waals components of the surface tension in each has indeed been shown that substratum roughness may
pair of solvents, differences between the results obtained affect the adsorption of macromolecules, such as proteins,
with chloroform and hexadecane, on the one hand, and and in particular, the supramolecular organization of the
between ethyl acetate and decane, on the other hand, adsorbed layer (117).
 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES 125

On the other hand, microbial cells may bear sur-

face appendages such as flagella, fimbriae, and fib-
rils (67). A relationship between the presence of flagella
and cell behavior was pointed out for attachment of
P. aeruginosa (colonizing contact lenses and catheters) to
polyvinylchloride (118) and attachment of P. fluorescens
in sand columns (119); this was made by comparing the
behavior of a flagellated wild-strain and a nonflagellated
mutant. However, interpretation of the data is delicate
since the presence of flagella may affect both the transport
step and the physicochemical interactions. To separate
these effects, the attachment of the food-borne pathogen
L. monocytogenes to stainless steel was studied by com-
paring the behavior of the wild-type strain possessing
an inactive flagellum and a nonflagellated mutant: a
higher adhesion was observed when the flagellum was Figure 5. Schematic illustration of the penetration of surface
appendages, allowing attachment of a microbial cell to a
present (120). The presence of a particular protein on
plane substratum, through a potential energy barrier (location
fimbriae was shown to be necessary for the adhesion of represented by dashed line).
the pathogen S. epidermidis, of concern in intravascular
devices, to polystyrene spheres (53). Finally, the presence
of fibrillar structures on the surfaces of Streptococcus recent years, atomic force microscopy (AFM) has made
strains has been linked to their ability to adhere to saliva- it possible to probe cell surface softness in a direct
coated hydroxyapatite (121), to host surfaces (122), and to way (125–127). This is illustrated in Figure 6, which
fibronectin-conditioned microwells (123). shows AFM force-distance curves recorded under water
The modes of action of cell surface appendages to upon approach between a silicon nitride probe and the
allow bacterial attachment are still unclear. Firstly, cell surface of fungal spores (Fig. 6a) and bacteria (Fig. 6b).
surface appendages may influence bacterial adhesion For P. chrysosporium spores, no significant deviation from
through hydrophobic interactions. Water contact angles linearity was seen in the contact region, indicating that
were significantly higher on fibrillated parent strains the sample was not deformed by the probe (128,129).
of Streptococcus salivarius and S. sanguis compared to Conversely, the curvature observed for both strains of
their mutants devoid of fibrils (67). Similar observations S. salivarius may be attributed to surface softness (126).
were made for Serratia marcescens, a pathogen that is The difference in softness observed between the two
found in water, soil, sewage, and foodstuffs. Secondly, the strains was related to the presence or the absence of
interactions occurring between the cell and the substratum fibrils for S. salivarius HB and HBC12, respectively.
upon approach could be controlled by the physicochemical Direct observation of bacterial surface deformability was
properties of surface appendages different from those of provided after three-dimensional reconstruction of height
the overall cell surface. Handley and coworkers (124) used AFM images of Lactococcus lactis, which revealed grooves
cationized ferritin to probe negatively charged sites, and induced by the scanning AFM probe (Fig. 7). While these
colloidal gold to probe positive and hydrophobic sites on grooves represent an alteration of the cell surface by the
S. sanguis strains bearing fibrils. Their results suggested probe, they were informative about its nanomechanical
that these structures may have electrical and hydrophobic properties (127).
properties, which are different from the neighboring bald
part of the cell surface (67,124). Thirdly, the radius of Macromolecules at Interfaces
curvature ruling the DLVO interactions may be much
The presence of solvated macromolecules, either consti-
smaller than the overall cell radius. In particular, it
tutive of surfaces or adsorbed, may generate interactions
is conceivable that cells become bound via appendages
that may be either attractive or repulsive, depending on
through an electrostatic potential barrier, which results
the properties of the macromolecules, the nature of the
from average surface properties, as illustrated by Figure 5.
solvent and the degree of coverage (64). When two sur-
The distinct role of the flagellum in bacterial adhesion
faces carrying nonionic (or uncharged) polymers approach
is illustrated by the fierce rotation movements observed
each other, the segments of these macromolecules over-
with attached Pseudomonas putida cells at low ionic
lap, which leads to a repulsive force (Fig. 8a). The latter
strength, that is, when the overall electrical potential
is caused by the decrease of entropy associated with a
barrier is high (87). Finally, it must also be noted that
closer confinement of the polymer chains between the two
appendages may be responsible for specific interactions
surfaces and is known as steric repulsion. The extent of
with a biological substratum or with an inert material
this repulsive force will be influenced by the nature of
coated by substances of biological nature.
the solvent: in a good solvent, that is, in a solvent in
which the polymer segments have a higher affinity for the
Surface Deformability
solvent than for one another, the repulsion will occur at
Although some microbial cells may be considered as a large separation distance (Fig. 9). On the other hand,
rigid spheres, as assumed in DLVO theory, most of attractive forces may also be mediated by adsorbed macro-
them should be viewed as soft and deformable. In molecules (Fig. 9). In a poor solvent, segment attraction
126 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES

(a) 1 (a)

0.5
Force (nN)

−0.5

−1
0 50 100 150 200 (b)
Separation distance (nm)

(b) 2
Force (nN)

(c)
−1

0 50 100 150 200

Separation distance (nm)

Figure 6. Representative AFM force-distance curves recorded,

under water, upon approach between the silicon nitride probe
and the surface of P. chrysosporium spores (a), S. salivarius
HB (B; continuous line) or S. salivarius HBC12 (b; interrupted
line) (126,128,129). Figure 8. Illustration of steric forces generated by polymers
present at an interface. Adapted from (64).

High coverage

Steric repulsion

Low
coverage Poor Good

Good
Bridging

Poor
Segment attraction, bridging

Figure 9. Variation of interaction free enthalpy as a function of

the separation distance between two surfaces as a result of macro-
molecular interactions; influence of the polymer concentration at
the surface (low or high coverage), and solvent quality (good or
Figure 7. 3D reconstructed AFM height image (475 × 475 nm2 ) poor). Adapted from (64).
obtained under water on L. lactis at an imaging force of
about 7.5 nN; the groove direction was parallel to the scanning
direction (127).
The influence of macromolecules at the interface will
depend on the pH and on the ionic strength, which
between molecules bound to opposite surfaces leads to determine the electrostatic repulsion between the different
overall attraction between the surfaces until a balance partners. A microbial cell may be bound to a substratum
with steric repulsion is reached (Fig. 8b). When the degree if bridging macromolecules extend far enough from the
of coverage is weak and the polymer is strongly adsorbed, surface to span the distance at which the electrical
the macromolecules may form bridges between the two repulsion operates, as illustrated by Figure 10. On the
surfaces (Fig. 8c). left-hand side of the figure, the molecular mass or the
 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES 127

the effect of conditioning film on attachment may

be positive or negative, it appears from a recent
compilation (12) that the force exerted by the liquid-
air interface tends to detach a higher percentage of
microorganisms attached via a conditioning film as
compared with microorganisms attached to a bare
substratum.
Figure 10. Illustration of the effect of ionic strength or surface
potential on polymer bridging; the dashed lines represent the Microorganisms as Living Colloidal Particles
distance at which the electrostatic repulsion operates. The
The study of microbial adhesion to living tissue (human
increase of ionic strength or decrease of surface potential from
or vegetal) indicated the role played by specific interac-
left to right allows polymer bridging to occur. Adapted from (130).
tions, on the basis of highly stereochemically selective
(key-lock) interactions between the microorganism and
ionic strength are too low to allow bridging. On the right- the substratum (18–20,135). The attachment of Lacto-
hand side, the ionic strength is higher or the surface bacillus plantarum to a human colonic cell line was
potential is weaker, the electrical repulsion operates shown to occur via a mannose-specific interaction (136).
at short distances compared with the thickness of the Another strain of Lactobacillus, that is, L. crispatus, had
adsorbed macromolecules, and bridging occurs (130,131). the capacity to attach to the collagen of human intesti-
nal extracellular matrix via its S-layer protein (137).
Conditioning Film Specific interaction may also effect adhesion of micro-
bial cells on vegetable surfaces. The attachment of
In most environments of practical relevance (natural the basidiomycetous yeast Rhodosporidium toruloides to
waters, soil solution, biological fluids, industrial fluids, . . .), barley leaf surfaces was mediated by the interaction
the substratum surface is rapidly conditioned by dissolved between the mannose residues of yeast surface and a
substances, particularly by dissolved macromolecules lectin (138).
(proteins, polysaccharides, lipids). The formation of these Microbial cells may influence the adhesion mecha-
conditioning films will affect both the physicochemical nisms by modification of their surface properties or
properties of the substratum surface and the adhesion by releasing extracellular compounds into the sur-
of microorganisms. Nevertheless, explaining the change rounding liquid. The surface chemical composition of
in microbial attachment is not straightforward (see the soil bacterium A. brasilense was shown to become
CONDITIONING FILMS FROM ENVIRONMENTAL WATERS, this richer in protein and hydrocarbon-like compounds dur-
Encyclopedia). ing its growth. These variations were related to an
The changes of hydrophobicity and electrical properties increase of both hydrophobicity and adhesion to glass
of silicone rubber, used in voice prostheses, could not and polystyrene (139). The chemical composition and/or
be related to the systematic inhibitory effect of a the physicochemical properties of the surface of other
salivary film on bacterial and yeast adhesion (132). The bacteria were also shown to be influenced by the
attachment (assessed after drying) of bacteria to natural growth and environmental conditions (140–142). On the
(shale, sandstone, andesite) and industrial (stainless steel, other hand, compounds released by microbial cells may
polypropylene) materials was decreased or increased influence in a positive or negative way their behav-
after the formation of a conditioning layer by dissolved ior at interfaces. Extracellular compounds released by
compounds; however, no relation could be established L. lactis may affect its attachment by increasing the
with water contact angles determined on the conditioned ionic strength locally, near the cell-substratum inter-
substratum (133). The study of the influence of adsorbed face (143). Extracellular proteins released by A. brasilense
milk-related products (skim milk, sodium caseinate, adsorb on the substratum, allowing attachment to
whey protein isolate and whey permeate) on L. lactis occur (29,30). In the latter case, the influence of cell
attachment revealed a beneficial effect of skim milk physiology was demonstrated by the influence of tem-
and whey permeate, which could not be attributed perature and of tetracycline (an inhibitor of pro-
to the overall chemical composition, hydrophobicity, or tein synthesis) on both protein release and attach-
electrical properties of the conditioning film. Force- ment (144).
distance curves recorded by AFM, combined with a study Biosurfactants released by microorganisms have a dis-
of dynamic wetting indicated that the mobility of adsorbed tinct tendency to accumulate at hydrophobic-hydrophilic
macromolecules play a key role in favoring attachment interfaces and to lower the interfacial tension (145) and
of L. lactis to skim milk-conditioned substratum (134). On were shown to inhibit microbial adhesion. Streptococcus
the other hand, conditioning stainless steel with skim thermophilus, regularly encountered on heat exchanger
milk was found to reduce the adhesion of five food- plates in dairy industry, was shown to be able to
borne pathogens; cross-linking adsorbed proteins partially produce a biosurfactant which caused its own detach-
reversed the inhibition of bacterial attachment (88). These ment (146). A small amount of attached oral Streptococcus
two studies illustrate that adsorbed macromolecules of mitis (1 to 4% of surface coverage) was able to pre-
similar nature may condition substratum surfaces in vent attachment of S. mutans, pointing to the potential
a way that is favorable or unfavorable to attachment, protective role of S. mitis in the oral cavity against
depending on the partner surfaces involved. While the cariogenic S. mutans (147). It may be noted that
128 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES

conditioning the substratum alone with the same bio- the surface properties of these microorganisms. Neverthe-
surfactant was less efficient, indicating that inhibition less, the exact nature of the interactions (hydrophobic,
occurred by conditioning both substratum and cell sur- electrostatic, steric) leading to the attachment of these
faces. Attachment of S. thermophilus to voice prosthe- microspheres (both negatively charged and hydrophobic)
sis materials (silicone rubber) prevented yeast coloniza- could not be specified (160). The meaning of the micro-
tion by biosurfactant production, allowing the prosthesis sphere adhesion test must be therefore approached with
indwelling lifetime to increase (148). Further, biosurfac- caution.
tants may be used to form a conditioning film that In a similar way, the adhesion of gold colloidal
discourages microbial attachment, as for the uropathogen particles (both hydrophobic and negatively charged) to
Streptococcus faecalis (149); such application presents a a tufted streptococcal strain showed an asymmetry of
high potential to inhibit adhesion on catheter materi- surface properties (124). Electron microscopy revealed
als. that colloidal gold did not attach to the bald part
Are the microorganisms able to react upon contact of tufted S. sanguis strains over the pH range 3.7 to
with a solid surface by inducing a specific activity that 9.0 because of the hydrophilic nature of this portion
leads to their attachment? Comparison of gene expres- of the cell surface. Adhesion of colloidal gold to the
sion for E. coli cells, either planktonic or from a biofilm, long tuft fibrils occurred irrespective of pH, confirming
showed that this microorganism is able both to repress conclusions from contact angle measurements that
the flagellum formation, which might destabilize the these fibrils convey hydrophobicity to the cell surface.
biofilm, as well as to activate the production of exopolysac- Adhesion of colloidal gold to short fibrils occurred at
charide (colanic acid), which will reinforce the biofilm pH 3.7 but not at elevated pH, because of increased
structure (150). However, this study does not answer electrostatic repulsion between the gold and the negatively
the question. By comparing E. coli strains proficient or charged short fibrils with increasing pH. Accordingly, the
deficient in colanic acid production, Danese and cowork- short and long fibrils can be associated with carrying
ers (151) showed that exopolysaccharide production is not negative charges and hydrophobicity, respectively (67).
required for attachment, in contrast with biofilm devel- It may be noted that, whether using polystyrene
opment. The study of attachment of P. aeruginosa under microspheres or gold particles, this approach does not
account for the possible role played by macromolecular
flowing conditions revealed that the situation may be
interactions.
rather complex: (1) the biofilm development was related
to the expression of the gene responsible for alginate
Atomic Force Microscopy. Regarding the assessment of
synthesis, allowing bacteria to be retained by a glass
the heterogeneity of substratum and cell surfaces, exciting
substratum; (2) this expression was not a prerequisite possibilities are now offered by atomic force microscopy
for attachment, indicating that alginate synthesis was (AFM) to investigate the molecular interactions and
not necessary for attachment to glass; (3) in the absence properties of surfaces on the nanometer scale. AFM,
of such expression, the bacteria were detached from invented in 1986 (162) consists of measuring the forces
the substratum as time passed (152). In the same way, acting between a sharp probe and the surface of a sample.
synthesis of proteins was shown to be essential to the It allows the nanoscale mapping of surface topography, as
anchorage of A. brasilense to polystyrene (29,30). The well as the direct measurement of mechanical properties,
production of quorum-sensing molecules (153) has been friction, and interaction forces (electrostatic, van der
shown to be necessary to the differentiation step of Waals, solvation, steric interactions). As opposed to more
Pseudomonas biofilm formation, but not to the primary conventional surface analysis techniques, AFM can be
adhesion step (154). operated in aqueous solutions, which makes it possible to
study living cells under physiological conditions (125,163).
Heterogeneity of Cell and Substratum Surfaces A very challenging goal for future research is to directly
map the surface properties of living cells in aqueous
The hydrophobicity, electrical properties, and chemical
conditions.
composition of microbial surfaces, can be characterized
AFM offers new perspectives concerning the nanoscale
by water contact angle measurement (155,156), elec-
mapping of physicochemical properties (hydrophobicity,
trophoretic mobility measurement (65) and X-ray photo-
electrical properties, macromolecular interactions). Ger-
electron spectroscopy (157,158), respectively. While rel-
minating spores of P. chrysosporium were shown to
evant information may thereby be obtained regarding present a heterogeneous ultrastructure made of gran-
cell behavior at interfaces, most of these techniques do ular and smooth zones (128). Figure 11a shows a map
not probe the cell surface in its native state. More- of adhesion forces recorded on an area presenting both
over, all give information on overall surface properties zones. A force-distance curve was recorded at each pixel
and do not provide information at high lateral resolu- of this map; lighter levels correspond to larger adhesion
tion. forces. The comparison of this map with the height image
recorded on the same area (Fig. 11b) demonstrated the
Microsphere Attachment. Polystyrene latex micro- relevance of AFM to assess the heterogeneous character
spheres were shown to attach preferentially to germi- of microbial cell surfaces with regard to physicochemical
nating tubes and hyphae compared with the parent cell, interactions.
for yeast (159) and other fungal cells (160,161). The micro- Recently, Dufrêne (164) described a new method
spheres thus give information about the heterogeneity of for characterizing the physicochemical properties of
 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES 129

(a) CONCLUSION

The methods used to study microbial adhesion differ

according to two major characteristics that will influence
the significance of the results: (i) the rate of transport of
microorganisms to the surface; and (ii) the shear exerted
near the substratum surface. The experimental conditions
may also influence the physicochemical properties of the
substratum surface, which may be affected by adsorption
of substances present in the liquid or released by the cells.
The choice of the methodology will differ depending on
whether the interest is focused on the rate of attachment,
on the strength of adhesion, on the study of the cell-
substratum surface, or on the comparison of different
substratum materials.
The role of electrostatic interactions and hydrophobic
interactions in controlling microbial adhesion is well
established. However, approaches attempting to quantify
cell-substratum interactions and to predict cell attachment
or adhesion strength are subject to severe limitations. The
(b)
approach based on the balance of interfacial energies is
limited by the impossibility to determine accurate values
of the surface energy from contact angles of liquids or
alternative methods.
This approach and the DLVO theory rely on hypotheses,
which may be crude simplifications, particularly regarding
the microbial cell surfaces. Of concern are the surface
roughness, deformability, and heterogeneity. Solvated
macromolecules or biosurfactants, whether present at the
cell surface or adsorbed by the substratum, may also play
a major role. Moreover, specific interactions may come into
the play with biological substrata or substrata conditioned
with compounds of biological origin. AFM offers promising
perspectives regarding the possibility to probe directly
interfacial interactions involving microbial cells and to
map the physicochemical properties and adhesion forces
over a cell surface.

Figure 11. AFM adhesion force map (a) and height image (b) Acknowledgments
recorded with a silicon nitride probe on the same area (2 × 2 µm2 ) The authors thank R.A. Jacob, F. Bouvy, and L. Durinckx for the
of a germinating spore of P. chrysosporium. Each pixel of the map preparation of the manuscript, and Prof. P. Grange for the use
corresponds to a force-distance curve, the gray level corresponding of the atomic force microscope. The support of the Foundation
to the adhesion force value (125). Z-range = 10 nN (a) and for Training in Industrial and Agricultural Research (FRIA), the
100 nm (b). National Foundation for Scientific Research (FNRS), the Research
Department of Communauté Française de Belgique (Concerted
Research Action), the Federal Office for Scientific, Technical, and
Cultural Affairs (Interuniversity Poles of Attraction Programme),
native microbial cells by using AFM with chemically and the European Union (Research Project N° ENV4-CT97-
functionalized probes. Functionalization was shown to 0634; COST Action 520 Biofouling and Materials) is gratefully
make the probe very sensitive to surface hydrophobic- acknowledged.
ity, the magnitude of the adhesion forces measured
between functionalized probes and substrata decreas-
BIBLIOGRAPHY
ing in the order CH3 /CH3 > CH3 /OH > OH/OH. The
surface of P. chrysosporium spores was homogeneously 1. H. C. Flemming and G. Schaule, in E. Heitz, H. C. Flem-
hydrophilic, as revealed by the lack of adhesion forces ming, and W. Sand, eds., Microbially influenced corrosion
measured among the spore using CH3 - and OH- of materials: scientific and engineering aspect, Springer-
terminated probes. An AFM probe can also be func- Verlag, Berlin, Germany, 1996, pp. 39–54.
tionalized to assess the specific interaction between a 2. J. W. Costerton, P. S. Stewart, and E. P. Greenberg, Science
ligand and a receptor, as performed for streptavidin- 284, 1318–1322 (1999).
biotin (165) or concanavalin-mannan (lectin-like) (166) 3. D. De Beer, R. Srinivasan, and P. S. Stewart, Appl. Environ.
interactions. Microbiol. 60, 4339–4344 (1994).
130 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES

4. J. D. Bryers and W. G. Characklis, in W. G. Characklis and 29. Y. F. Dufrêne, C. J. P. Boonaert, and P. G. Rouxhet, Col-
K. C. Marshall, eds., Biofilms, John Wiley & Sons, New loids Surf., B: Biointerfaces, 7, 113–128 (1996).
York, 1990, pp. 671–696. 30. Y. F. Dufrêne, C. J. P. Boonaert, and P. G. Rouxhet, in
5. L. F. Melo, T. R. Bott, M. Fletcher, and B. Capdeville, J. Berthelin, P. M. Huang, J. M. Bollag, and F. Andreux,
Biofilms-science and technology, Kluwer Academic Publish- eds., Effect of Mineral-Organic-Microorganism Interactions
ers, Dordrecht, The Netherlands, 1992, pp. 409–466. on Soil and Freshwater Environments, Kluwer Academic,
6. R. Cam and K. J. Kennedy, in H. M. Lappin-Scott and Plenum Publishers, New York, 1999, pp. 261–274.
J. W. Costerton, eds., Microbial Biofilms, Cambridge Uni- 31. J. L. Van Haecht, M. De Bremaeker, and P. G. Rouxhet,
versity Press, Cambridge, Mass., 1995, pp. 183–195. Enzyme Microb. Technol. 6, 221–227 (1984).
7. D. F. Day, M. L. Marceau-Day, and B. M. Day, in M.-D. 32. P. G. Rouxhet and N. Mozes, Water Sci. Technol. 22, 1–16
Legoy and D. Thomas, eds., Annals of the New York (1990).
Academy of Sciences: Enzyme Engineering XII, The New
33. H. J. Busscher and H. C. van der Mei, Methods Enzymol.
York Academy of Sciences, New York, 1995, pp. 432–434.
253, 455–476 (1995).
8. M. Martin et al., Appl. Environ. Microbiol. 66, 1190–1194
(2000). 34. J. L. Van Haecht, M. Bolipombo, and P. G. Rouxhet, Biotech-
nol. Bioeng. 27, 217–224 (1985).
9. E. A. Zottola, Food Technol. 48, 107–115 (1994).
35. J. Visser, in E. Matijevic, ed., Surface and Colloid Science,
10. T. Tosa and T. Shibatani, in M.-D. Legoy and D. Thomas,
John Wiley & Sons, New York, 1976, pp. 3–84.
eds., Annals of the New York Academy of Sciences: Enzyme
Engineering XII, The New York Academy of Sciences, New 36. H. W. Fowler and A. J. McKay, in R. C. W. Berkeley et al.,
York, 1995, pp. 364–375. eds, Microbial Adhesion to Surfaces, Ellis Horwood,
Chichester, U.K., 1980, pp. 143–161.
11. D. Pearce, M. J. Bazin, and J. M. Lynch, in H. M. Lappin-
Scott and J. W. Costerton, eds., Microbial biofilms, Cam- 37. C. Cozens-Roberts, J. A. Quinn, and D. A. Lauffenburger,
bridge University Press, Cambridge, Mass., 1995, pp. 207– Biophys. J. 58, 107–125 (1990).
220. 38. A. J. Goldman, R. G. Cox, and H. Brenner, Chem. Eng. Sci.
12. R. Bos, H. C. van der Mei, and H. J. Busscher, FEMS 22, 653 (1967).
Microbiol. Rev. 23, 179–230 (1999). 39. M. E. O’Neill, Chem. Eng. Sci. 23, 1293 (1968).
13. P. Watnick and R. Kolter, J. Bacteriol. 182, 2675–2679 40. A. Abbott, R. C. W. Berkeley, and P. R. Rutter, in R. C. W.
(2000). Berkeley et al., eds., Microbial Adhesion to Surfaces, Ellis
14. P. E. Kolenbrander, Crit. Rev. Microbiol. 17, 137–155 Horwood, Chichester, U.K., 1980, pp. 117–142.
(1989). 41. I. Wang, J. M. Anderson, M. R. Jacobs, and R. E. Marchant,
15. K. W. Millsap, H. C. van der Mei, R. Bos, and H. J. Buss- J. Biomed. Mater. Res. 29, 485–493 (1995).
cher, FEMS Microbiol. Rev. 21, 321–336 (1998). 42. R. B. Dickinson and S. L. Cooper, AIChE J. 41, 2160–2174
16. A. Boyd and A. M. Chakrabarty, Appl. Environ. Microbiol. (1995).
60, 2355–2359 (1994). 43. R. B. Dickinson et al., Infect. Immun. 63, 3143–3150 (1995).
17. D. G. Allison et al., FEMS Microbiol. Lett. 167, 179–184 44. T. Dabros and T. G. M. van de Ven, Colloid Polym. Sci. 261,
(1998).
694–707 (1983).
18. R. J. Doyle, A. M. Calderon de La Barca, and G. Buck, in
45. T. Dabros and T. G. M. van de Ven, Phys. Chem. Hydrodyn.
H. J. Busscher and L. V. Evans, eds., Oral Biofilms Plaque
8, 161–172 (1987).
Control, Harwood, Amsterdam, The Netherlands, 1998,
pp. 111–124. 46. Z. Adamczyk and T. G. M. van de Ven, J. Colloid Interface
Sci. 80, 340–356 (1981).
19. P. Klemm and M. A. Schembri, Int. J. Med. Microbiol. 290,
27–35 (2000). 47. K. M. Wiencek and M. Fletcher, in L. F. Melo, T. R. Bott,
20. D. C. Savage and M. Fletcher, Bacterial Adhesion — Mecha- M. Fletcher, and B. Capdeville, eds., Biofilms — Science and
nisms and Physiological Significance, Plenum Press, New Technology, Kluwer Academic Publishers, Dordrecht, The
York, 1985. Netherlands, 1992, pp. 99–104.
21. T. F. Tadros, in R. C. W. Berkeley et al., eds., Microbial 48. J. Sjollema, H. J. Busscher, and A. H. Weerkamp, J. Micro-
Adhesion to Surfaces, Ellis Harwood Limited, Chichester, biol. Methods 9, 79–90 (1989).
U.K., 1980, pp. 93–116. 49. J. W. Costerton, J. C. Nickel, and T. I. Ladd, in J. S. Poin-
22. M. Fletcher, M. J. Latham, J. M. Lynch, and P. R. Rutter, dexter and E. R. Leadbetter, eds., Bacteria in Nature — Vol-
in R. C. W. Berkeley et al., eds., Microbial Adhesion to ume 2: Methods and Special Applications in Bacterial
Surfaces, Ellis Harwood, Chichester, U.K., 1980, pp. 67–78. Ecology, Plenum Publishing, New York, 1986, pp. 49–84.
23. M. Fletcher, Methods Microbiol. 22, 251–283 (1990). 50. C. J. Linton, A. Sherriff, and M. R. Millar, J. Appl. Micro-
24. M. Fletcher, in L. F. Melo, T. R. Bott, M. Fletcher, and biol. 86, 194–202 (1999).
B. Capdeville, eds., Biofilms — Science and Technology, 51. E. P. J. M. Everaert et al., J. Mater. Sci.: Mater. Med. 9,
Kluwer Academic Publishers, Dordrecht, The Netherlands, 147–157 (1998).
1992, pp. 603–614. 52. J. D. Bryers and W. G. Characklis, in L. F. Melo, T. R. Bott,
25. Z. Adamczyk, Colloids Surf. 35, 283–308 (1989). M. Fletcher, and B. Capdeville, eds., Biofilms — Science and
26. V. G. Levich, Physicochemical Hydrodynamics, Prentice- Technology, Kluwer Academic Publishers, Dordrecht, The
Hall, Englewood Cliffs, N.J., 1962. Netherlands, 1992, pp. 615–629.
27. T. G. M. van de Ven, Colloidal Hydrodynamics, Academic 53. G. J. C. Veenstra, F. F. M. Cremers, H. Van Dijk, and
Press, London, U.K., 1989. A. Fleer, J. Bacteriol. 178, 537–541 (1996).
28. P. C. Hiemenz and R. Rajagopalan, Principles of Colloid and 54. G. Fleminger and Y. Shabtai, Appl. Environ. Microbiol. 61,
Surface Chemistry, 3rd ed., Marcel Dekker, New York, 1997. 4357–4361 (1995).
 ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES 131

55. A. B. Cunningham, E. J. Bouwer, and W. G. Characklis, in 83. D. Guoqiang, R. Kaul, and B. Mattiasson, Appl. Microbiol.
W. G. Characklis and K. C. Marshall, eds., Biofilms, John Biotechnol. 37, 305–310 (1992).
Wiley & Sons, New York, 1990, pp. 697–732. 84. S. F. D’Souza, J. S. Melo, A. Deshpande, and G. B. Nad-
56. J. Gannon, Y. Tan, P. Baveye, and M. Alexander, Appl. karni, Biotechnol. Lett. 8, 643–648 (1986).
Environ. Microbiol. 57, 2497–2501 (1991). 85. P. G. Rouxhet, A. Doren, J. L. Dewez, and O. Heuschling,
57. B. A. Jucker, H. Harms, and A. J. B. Zehnder, J. Bacteriol. Prog. Org. Coatings 22, 327–344 (1993).
178, 5472–5479 (1996). 86. Q. Li and B. E. Logan, Water Res. 33, 1090–1100 (1999).
58. H. H. M. Rijnaarts et al., Colloids Surf., B: Biointerfaces, 4, 87. H. H. M. Rijnaarts, W. Norde, J. Lyklema, and A. J. B.
5–22 (1995). Zehnder, Colloids Surf., B: Biointerfaces, 14, 179–195
59. H. H. M. Rijnaarts et al., Environ. Sci. Technol. 20, (1999).
2877–2883 (1996). 88. L.-M. Barnes, M. F. Lo, M. R. Adams, and A. H. L. Cham-
60. H. H. M. Rijnaarts et al., Appl. Environ. Microbiol. 59, berlain, Appl. Environ. Microbiol. 65, 4543–4548 (1999).
3255–3265 (1993). 89. R. Briandet, T. Meylheuc, C. Maher, and M.-N. Bellon-
61. K. C. Marshall, in D. C. Savage and M. Fletcher, eds., Bacte- Fontaine, Appl. Environ. Microbiol. 65, 5328–5333 (1999).
rial Adhesion — Mechanisms and Physiological Significance, 90. S. F. Simoni, T. N. P. Bosma, H. Harms, and A. J. B. Zehn-
Plenum Press, New York, 1985, pp. 133–161. der, Environ. Sci. Technol. 34, 1011–1017 (2000).
62. P. R. Rutter and B. Vincent, in R. C. W. Berkeley et al., eds, 91. V. Williams and M. Fletcher, Appl. Environ. Microbiol. 62,
Microbial Adhesion to Surfaces, Ellis Horwood, Chichester, 100–104 (1996).
U.K., 1980, pp. 79–92.
92. J. S. Webb et al., Appl. Environ. Microbiol. 65, 3575–3581
63. L. F. Melo, T. R. Bott, M. Fletcher, and B. Capdeville, (1999).
Biofilms — Science and Technology, Kluwer Academic Pub-
93. R. J. Good, in R. J. Good and R. R. Stromberg, eds., Surface
lishers, Dordrecht, The Netherlands, 1992.
and Colloid Science, Plenum Press, New York, 1979,
64. J. Israelachvili, Intermolecular and Surface Forces, 2nd ed., pp. 1–29.
Academic Press, London, U.K., 1992.
94. M.-N. Bellon-Fontaine et al., Cell. Biophys. 17, 93–105
65. A. M. James, in N. Mozes, P. S. Handley, H. J. Busscher, (1990).
and P. G. Rouxhet, eds., Microbial Cell Surface Analysis:
95. C. J. van Oss, Interfacial Forces in Aqueous Media, Marcel
Structural and Physico-chemical Methods, VCH Publishers,
Dekker, New York, 1994.
New York, 1991, pp. 221–262.
96. M. Morra and C. Cassinelli, J. Biomater. Sci.: Polym. Edn.
66. W. Norde and J. Lyklema, Colloids Surf. 38, 1–13 (1989).
9, 55–74 (1997).
67. H. J. Busscher, R. Bos, H. C. van der Mei, and P. S. Hand-
97. A. W. Neumann, R. J. Good, C. J. Hope, and M. Sejpal, J.
ley, in A. Baszkin and W. Norde, eds., Physical Chemistry
of Biological Interfaces, Marcel Dekker, New York, 2000, Colloid Interface Sci. 49, 291–304 (1974).
pp. 431–458. 98. J. K. Spelt, D. R. Absolom, and A. W. Neumann, Langmuir
68. J. Lyklema, Fundamentals of Interface and Colloid Science, 2, 620–625 (1986).
vol. 2, Academic Press, London, U.K., 1995. 99. C. J. van Oss, R. J. Good, and M. K. Chaudhury, Langmuir
69. S. Nir, Prog. Surf. Sci. 8, 1–58 (1976). 4, 884–891 (1988).
70. N. Mozes et al., Biotechnol. Bioeng. 30, 439–450 (1987). 100. D. R. Absolom et al., Appl. Environ. Microbiol. 46, 90–97
(1983).
71. M. C. M. van Loosdrecht et al., Appl. Environ. Microbiol.,
53, 1898–1901 (1987). 101. D. K. Owens and R. C. Wendt, J. Appl. Polym. Sci. 13,
1741–1747 (1969).
72. C. Changui et al., J. Chim. Phys. 84, 275–281 (1987).
102. S. Wu, J. Adhes. 5, 39–55 (1973).
73. N. Mozes and P. G. Rouxhet, in L. F. Melo, T. R. Bott,
M. Fletcher, and B. Capdeville, eds., Biofilms — Science and 103. H. J. Busscher, A. W. van Pelt, H. P. de Jong, and
Technology, Kluwer Academic Publishers, Dordrecht, The J. Ahrends, J. Colloid Interface Sci. 95, 23–27 (1983).
Netherlands, 1992, pp. 69–85. 104. H. J. Busscher et al., Appl. Environ. Microbiol. 48, 980–983
74. N. Mozes and P. G. Rouxhet, Appl. Microbiol. Biotechnol. (1984).
22, 92–97 (1985). 105. M. Fletcher and J. H. Pringle, J. Colloid Interface Sci. 104,
75. N. Mozes and P. G. Rouxhet, Enzyme Microb. Technol. 6, 5–14 (1985).
497–502 (1984). 106. A. J. Barton, R. D. Sagers, and W. G. Pitt, J. Biomed. Mater.
76. P. A. Gerin, M. N. Bellon-Fontaine, M.Asther, and P. G. Res. 30, 403–410 (1996).
Rouxhet, Biotechnol. Bioeng. 47, 677–687 (1995). 107. A. W. J. van Pelt et al., Appl. Environ. Microbiol. 49,
77. B. Champluvier, B. Francart, and P. G. Rouxhet, Biotech- 1270–1275 (1985).
nol. Bioeng. 34, 844–853 (1989). 108. A. A. Mafu, D. Roy, J. Goulet, and L. Savoie, Appl. Environ.
78. B. Champluvier, F. Marchal, and P. G. Rouxhet, Enzyme Microbiol. 57, 1969–1973 (1991).
Microb. Technol. 11, 422–430 (1989). 109. C. R. Jones et al., J. Appl. Bacteriol. 81, 553–560 (1996).
79. M. P. Hermesse, C. Dereppe, Y. Bartholome, and P. G. 110. M. C. M. van Loosdrecht et al., Appl. Environ. Microbiol. 53,
Rouxhet, Can. J. Microbiol. 34, 638–644 (1988). 1893–1897 (1987).
80. J. Büchs, N. Mozes, C. Wandrey, and P. G. Rouxhet, Appl. 111. M. C. M. van Loosdrecht, W. Norde, J. Lyklema, and
Microbiol. Biotechnol. 29, 119–128 (1988). A. J. B. Zehnder, Aquat. Sci. 52, 103–114 (1990).
81. P. M. Van Schie and M. Fletcher, Appl. Environ. Microbiol. 112. M.-N. Bellon-Fontaine, J. Rault, and C. J. van Oss, Colloids
65, 5082–5088 (1999). Surf., B: Biointerfaces 7, 47–53 (1996).
82. N. Kamath and S. F. D’Souza, Enzyme Microb. Technol. 13, 113. R. S. Pembrey, K. C. Marshall, and R. P. Schneider, Appl.
935–938 (1991). Environ. Microbiol. 65, 2877–2894 (1999).
132 ADHESION TO SURFACES

114. L. Boulangé-Petermann, J. Rault, and M.-N. Bellon-Fon- 147. C. G. van Hoogmoed, M. van der Kuijl-Booij, H. C. van der
taine, Biofouling 11, 201–216 (1997). Mei, and H. J. Busscher, Appl. Environ. Microbiol. 66,
115. M. Quirynen and C. M. L. Bollen, J. Clin. Periodontol. 22, 659–663 (2000).
1–14 (1995). 148. H. J. Busscher et al., Appl. Environ. Microbiol. 63, 3810–
116. D. R. Korber et al., Appl. Environ. Microbiol. 63, 3352– 3817 (1997).
3358 (1997). 149. M. M. C. Velraeds, H. C. van der Mei, G. Reid, and H. J.
117. Y. F. Dufrêne, T. G. Marchal, and P. G. Rouxhet, Langmuir Busscher, Appl. Environ. Microbiol. 62, 1958–1963 (1996).
15, 2871–2878 (1999). 150. C. Prigent-Combaret, O. Vidal, C. Dorel, and P. Lejeune, J.
118. G. A. O’Toole and R. Kolter, Mol. Microbiol. 30, 295–304 Bacteriol. 181, 5993–6002 (1999).
(1998). 151. P. N. Danese, L. A. Pratt, and R. Kolter, J. Bacteriol. 182,
119. M. F. DeFlaun et al., Appl. Environ. Microbiol. 56, 112–119 3593–3596 (2000).
(1990). 152. D. G. Davies and G. G. Geesey, Appl. Environ. Microbiol. 61,
120. S. Vatanyoopaisarn et al., Appl. Environ. Microbiol. 66, 860–867 (1995).
860–863 (2000). 153. B. L. Bassler, Curr. Opin. Microbiol. 2, 582–587 (1999).
121. E. J. Morris, N. Ganeshkumar, M. Song, and B. C. McBride, 154. D. G. Davies et al., Science 280, 295–298 (1998).
J. Bacteriol. 169, 164–171 (1987). 155. H. C. van der Mei, M. Rosenberg, and H. J. Buscher, in
122. A. H. Weerkamp, P. S. Handley, A. Baars, and J. W. Slot, J. N. Mozes, P. S. Handley, H. J. Busscher, and P. G. Rouxhet,
Bacteriol. 165, 746–755 (1986). eds., Microbial Cell Surface Analysis: Structural and
Physico-Chemical Methods, VCH Publishers, New York,
123. R. McNab et al., J. Bacteriol. 181, 3087–3095 (1999).
1991, pp. 263–287.
124. P. S. Handley, L. M. Hesketh, and R. A. Moumena, Biofoul-
156. N. Mozes and P. G. Rouxhet, J. Microbiol. Methods 6,
ing 4, 105–111 (1991).
99–112 (1987).
125. Y. F. Dufrêne, Micron 32, 153–165 (2001).
157. P. G. Rouxhet and M. J. Genet, in N. Mozes, P. S. Handley,
126. H. C. van der Mei et al., Biophys. J. 78, 2668–2674 (2000). H. J. Busscher, and P. G. Rouxhet, eds., Microbial Cell Sur-
127. C. J. P. Boonaert et al., Colloids Surf., B: Biointerfaces in face Analysis: Structural and Physico-Chemical Methods,
press (2001). VCH Publishers, New York, 1991, pp. 173–220.
128. Y. F. Dufrêne et al., J. Bacteriol. 181, 5350–5354 (1999). 158. P. G. Rouxhet et al., Colloids Surf., B: Biointerfaces 2,
129. C. J. P. Boonaert, P. G. Rouxhet, and Y. F. Dufrêne, Surf. 347–369 (1994).
Interface Anal. 30, 32–35 (2000). 159. K. C. Hazen and B. W. Hazen, J. Microbiol. Methods 6,
130. F. Lafuma, in J. C. Roberts, ed., Paper Chemistry, 2nd ed., 289–299 (1987).
Chapman & Hall, London, U.K., 1995, pp. 44–63. 160. P. A. Gerin et al., J. Bacteriol. 175, 5135–5144 (1993).
131. B. Vincent, in T. F. Tadros, ed., Solid/Liquid Dispersions, 161. M. C. Penalver, M. Casanova, J. P. Martinez, and M. L. Gil,
Academic Press, London, U.K., 1987, pp. 149–162. Microbiology 142, 1597–1604 (1996).
132. H. J. Busscher, G. I. Geertsema-Doornbusch, and H. C. van 162. G. Binnig, C. F. Quate, and C. Gerber, Phys. Rev. Lett. 56,
der Mei, J. Biomed. Mater. Res. 34, 201–210 (1997). 930–933 (1986).
133. R. P. Schneider et al., FEMS Microbiol. Ecol. 14, 243–254 163. M. Radmacher, R. W. Tillmann, M. Fritz, and H. E. Gaub,
(1994). Science 257, 1900–1905 (1992).
134. C. J. P. Boonaert, S. R. Derclaye, and P. G. Rouxhet, Bio- 164. Y. F. Dufrêne, Biophys. J. 78, 3286–3291 (2000).
fouling, to be published. 165. G. U. Lee, L. A. Chrisey, and R. J. Colton, Science 266,
135. H. J. Busscher, M. M. Cowan, and H. C. van der Mei, FEMS 771–773 (1994).
Microbiol. Rev. 88, 199–210 (1992). 166. M. Gad, A. Itoh, and A. Ikai, Cell. Biol. Int. 21, 697–706
136. I. Alderberth et al., Appl. Environ. Microbiol. 62, 2244– (1997).
2251 (1996).
137. T. Toba et al., Appl. Environ. Microbiol. 61, 2467–2471
(1995).
ADHESION TO SURFACES. See CONDITIONING FILMS
138. J. W. Buck and J. H. Andrews, Appl. Environ. Microbiol. 65, IN AQUATIC ENVIRONMENTS
465–471 (1999).
139. Y. F. Dufrêne and P. G. Rouxhet, Can. J. Microbiol. 42,
548–556 (1996).
140. C. F. Brush, M. F. Walter, L. J. Anguish, and W. C. Ghiorse,
Appl. Environ. Microbiol. 64, 4439–4445 (1998).
141. K. W. Millsap, H. C. van der Mei, G. Reid, and H. J. Buss- ADSORPTION OF MICROORGANISMS TO
cher, J. Microbiol. Methods 27, 239–242 (1996). SURFACES. See ADHESION (PRIMARY) OF MICROORGANISMS
142. T. K. Jana, A. K. Srivastava, K. Csery, and D. K. Arora, ONTO SURFACES
Can. J. Microbiol. 46, 28–37 (2000).
143. C. J. P. Boonaert, Y. F. Dufrêne, S. R. Derclaye, and P. G.
Rouxhet, Colloids Surf., B: Biointerfaces 22, 171–182 (2001).
144. Y. F. Dufrêne, H. Vermeiren, J. Vanderleyden, and P. G.
Rouxhet, Microbiology 142, 855–865 (1996). ADSORPTION OF VIRUSES TO SURFACES.
145. T. R. Neu, Microbiol. Rev. 60, 151–166 (1996). See MODELING OF VIRUS TRANSPORT AND REMOVAL IN THE
146. H. J. Busscher et al., Biofouling 2, 55–63 (1990). SUBSURFACE
 AEROBIC ALKALIPHILES 133

AEROBIC ALKALIPHILES interesting from both the viewpoints of physiology and

industrial applications.
MASAHIRO ITO In this article, I present an introduction to alkaliphilic
Toyo University microorganisms that includes (1) historical background,
Gunma, Japan (2) diversity, and (3) physiology.

Recently, it has become clear that life exists in Milieux that HISTORICAL BACKGROUND
were earlier thought to be incompatible with growth, for
example, extremely high temperatures, very low tempera- Although alkaliphilic microorganisms are now well
tures, abyssobenthic conditions, strongly alkaline pH, and recognized, their existence was little noted until the 1970s.
strongly acid pH. The ‘‘extremophiles’’ that grow in such Koki Horikoshi, a leading investigator of alkaliphiles,
conditions are the subject of current active research. Alka- was the first to initiate thorough studies of these
liphilic microorganisms are extremophiles that actively bacteria, starting in the late 1960s. Horikoshi has
grow in extremely alkaline Milieux, for example, pH 10, commented that when he began these investigations, he
and generally require sodium ions for growth. Alkaliphilic found only 16 scientific papers referring to alkaliphilic
microorganisms are classified as either facultative or obli- microorganisms (1). Two of the earliest papers dealing
gate alkaliphiles, depending on their capacity to grow at with the growth of bacteria in alkaline environments were
neutral pH (Fig. 1). Obligate alkaliphiles are restricted those of Meek and Lipman in 1922 (2) and Downie and
to alkaline pH, whereas facultative alkaliphiles can also Cruickshank in 1928 (3). However, both these reported on
grow at neutral pH. Microorganisms called alkalitolerant the growth of non-alkaliphiles when exposed to an alkaline
typically have an upper limit of pH 9 for robust growth; environment. The earliest papers on true alkaliphiles
they grow poorly in the pH ranges of true alkaliphilic bac- were those of Gibson (4) on B. pasteurii growing at pH
teria. Alkaliphilic microorganisms are distributed widely 11 and of Vedder (5) on B. alcalophilus growing well
in nature and include actinomycetes, eubacteria, archaea, at pH 8.6 to 11. Early in the 1960s, Takahara and
and fungi. There are many interesting and unresolved coworkers (6) demonstrated the industrial importance of
issues with respect to how alkaliphilic microorganisms alkaliphiles when he improved the process of indigo
adapt to their extremely alkaline Milieux. The mechanism fermentation through the addition of the alkaliphilic strain
of this adaptation has been most extensively studied in Bacillus S-8. This was the first industrial application
the B. species. Data have been presented for the roles of of alkaliphilic bacteria. Since then, more than 1,000
Na+ /H+ antiporters, which are present in the cell mem- papers on alkaliphiles and their adaptations have been
brane, and of a barrier of charged cell wall–associated published (1).
macromolecules in the accommodation of the bacteria to
the alkaline Milieux. In addition to mechanistic stud- DISTRIBUTION
ies, the applications of alkaliphile biology to industry
are numerous. For example, extracellular enzymes pro- Distribution of Alkaliphiles in the Extremely Alkaline Milieux
duced by alkaliphiles are used in laundry detergents (e.g., in Nature
alkaline cellulase and alkaline protease) and are used Research into the distribution of alkaliphiles in the
in the production of cyclodextrin, which can encapsulate extremely alkaline environments was developed by Grant
large compound molecules (e.g., cyclomaltodextrin glucan- and his colleagues (7). Natural soda lakes and soda deserts
otransferase). Thus, alkaliphilic microorganisms are very exist worldwide (e.g., Lake Magadi in Kenya (pH 10.9),
Wadi Natrum in Egypt (pH 11), and Lake Sambhar in
India (pH 9.0)), and they are summarized in Table 1.
Obligative alkaliphile Some of them have NaCl concentrations of about 5% w/v,
Bacillus firmus RAB whereas others are hypersaline, with NaCl concentrations
(pH 8.5~11.5) greater than 15% w/v. These lakes are strongly inhabited
by prokaryotic and archaeal alkaliphiles, which prosper
Facultative alkaliphile
in these selective environments. Because Ca2+ and Mg2+
Bacillus halodurans C-125
(pH 6.8 ~11) are precipitated as carbonates, the ionic concentrations of
these important divalent cations are very low in these
Neutrophile lakes. The major ions of the lake are Na+ , Cl− and
Bacillus subtilis bicarbonate/carbonate. The alkaliphilic isolates from these
(pH 5.5~8.5) lakes generally require Na+ for growth. Most of them are
haloalkaliphiles (7).
4 6 8 10 12 Artificial alkaline environments are probably selected
pH for particular alkaliphiles from adjoining soils and waters.
Isolates from such environments include Bacillus sp.
Figure 1. The relationship of pH to the growth of a typical
neutrophile and two categories of alkaliphiles. The typical pH No. S-8 from an indigo ball of indigo fermentation (6),
dependency of the growth of neutrophile, B. subtilis, and two gram-positive Exguobacterium auratiacum from potato
categories of alkaliphiles, obligative alkaliphile, B. firmus RAB processing waste (8), and a gram-positive facultatively
and facultative alkaliphile, B. halodurans C-125 is shown by alkaliphilic Bacillus sp. from the alkaline wash waters
two-way arrows. derived from the preparation of edible olives (9).
134 AEROBIC ALKALIPHILES

Table 1. Worldwide Distribution of Soda Lakes and Soda Deserts

North America
Canada Manito
United States Alkali Valley, Albert Lake, Lake Lenore, Soap Lake, Big Soda Lake, Mono Lake,
Searles Lake, Deep Springs, Rhodes, Marsh, Harney Lake, Summer Lake,
Surprise Valley, Pyramid Lake, Walker Lake
Central America
Mexico Texcoco
South America
Venezuela Langunilla Valley
Chile Antofagasta
Europe
Hungary Lake Feher
Yugoslavia Pecena Slatina
Russia Kulunda Steppe, Tanatar lakes, Karakul, Araxes plain, Chita, Barnaul,
Slavgerod
Asia
Turkey Van
India Lake Looner, Lake Sambhar
China Qinhgai Hu, Sui-Yuan, Heilungkiang, Kirin, Jehol, Chahar, Shansi, Shensi,
kansu
Africa
Libya Lake Fezzan
Egypt Wadi Natrun
Ethiopia Lake Aranguadi, Lake Kilotes, Lake Abiata, Lake Shala, Lake Chilu, Lake
Hertale, Lake Metahara
Sudan Dariba Lakes
Kenya Lake Bogoria, Lake Nakuru, Lake Elmentieta, Lake Magadi, Lake Simbi, Lake
Sonachi
Tanzania Lake Natron, Lake Embagi, Lake Megad, Lake Manyara, Balangida, Basotu
Crater Lakes, Lake Kusare, Lake Tulusia, El Kekhooito, Momela Lakes, Lake
Lekandiro, Lake Reshitani, Lake Lgarya, Lake Ndutu, Lake Rukwa North
Uganda Lake Katwe, Lake Mahega, Lake Kikorongo, Lake Nyamunuka, Lake
Munyanyange, Lake Murumuli, Lake Nunyampaka
Chad Lake Bodu, Lake Rombou, Lake Dijikare, Lake Momboio, Lake Yoan
Australia Lake Corangamite, Red Rock Lake, Lake Werowrap, Lake Chidnup

The Distribution of Alkaliphiles in Nonalkaline Milieux in reproduced from a review by Jones and coworkers (11).
Nature The enormous taxonomic diversity of extreme alkaliphiles
is evident, and it is further reflected in the diversity
As for the broader distributions of alkaliphilic microor-
of characteristics. Alkaliphilic cyanobacteria are among
ganisms, alkaliphiles are surprisingly distributed almost
the primary photosynthetic organisms that produce oxy-
everywhere. Alkaliphiles are isolated from the soil of deep-
gen; such organisms include Spirulina, Cyanospira, Syne-
sea trenches at 11,000-m depth (10) and from diverse soil
chococcus, and Chroococcus. Some of the soda lakes harbor
and water samples from more conventional sites (1). Alka-
many characteristic alkaliphilic phototrophic bacteria,
liphiles can be found in soil at pH 4, but the highest
for example, Ectothiorhodospira together with halophilic
frequency was observed in alkaline soil, especially above
archaea (12). Zhilina and coworkers (13) isolated and
pH 8.3. The number of alkaliphiles found in soil was about
characterized Desulfonatronovibrio hydorogenovorans, an
1/10 to 1/100 of that of neutrophilic bacteria. The number
of organisms was to a certain extent correlated with the alkaliphilic, sulfate-reducing bacterium, from a soda-
pH of the soil sample (Fig. 2). depositing lake, Lake Magadi, Kenya.

Alkaliphiles Are Not Restricted to Any Taxonomic Group Diversity of Alkaliphilic Bacillus Strains
Alkaliphiles are thought to change the pH of their Milieux Since Horikoshi and his colleagues began the screening
to suit their optimum growth. In addition to prokaryotic and research of alkaliphilic microorganisms in 1968,
alkaliphiles, various types of alkaliphilic microorganisms, many alkaliphilic microorganisms have been isolated.
including actinomycetes, fungi, yeast, and phages, have Most of them are gram-positive, spore-forming, motile,
been isolated (1) and some of them are presented at and aerobic rods, and so are classified in the genus
Table 2. Bacillus. Bacillus species are among the most commonly
A summary of the taxonomic group of prokaryotes iso- found aerobic and eubacterial alkaliphiles, both in alkaline
lated from soda lakes, an alkaline Milieux in nature, is and in nonalkaline Milieux. A very large number of
 AEROBIC ALKALIPHILES 135

Table 2. Taxonomic Group Containing Prokaryotes Iso-

106
lated from Soda Lakes (Boldface Type). Reproduced from
Jones, et al., (1998) with Permission from the Publisher

Eubacteria
105
Cyanobacteria
Counts of alkaliphilic bacteria /g of soil

Chronococcales
Oscillatoriales
104 Spirulina
Firmicutes (gram-positive bacteria)
Actinobacteria (high G + C gram-positive bacteria)
Actinomycetales
103 Micrococcaceae
Nocardiform actinomycetes
Streptomycetaceae
Streptomyces
102
Bacillus/Clostridium group (Low G + C
gram-positive bacteria)
Bacillaceae
10 Clostridiaceae
Haloanaerobiales
Proteobacteria(purple nonsulfur bacteria)
Beta subdivision
Delta subdivision
1 2 3 4 5 6 7 8 9 10 11 12
Gamma subdivision
pH of soil Ectothiorhodospiraceae (purple sulfur bacteria)
Figure 2. Distribution of alkaliphilic bacteria in soil (Repro- Ectothiorhodosprira
duced with permission from Horikoshi and Akiba, Alkalophilic Halomonadaceae
Microorganisms: A New World, p11, Japan Scientific Societies Pseudomonadaceae
Press, Tokyo (1982)). Pseudomonas
Spirochaetales
Spirochaetaceae
alkaliphilic bacilli have been isolated because of interest in Spirochaeta
the biotechnological application of useful alkali-resistant Thermotogales
extracellular enzymes. Thermopallium
Archaea
In the 1990s, Fritze and coworkers (14) and Nielsen
Euryarchaeota
and coworkers (15) classified alkaliphilic and alkali-
Halobacteriales
tolerant Bacillus. They reported phylogenetic diversity Halobacteriaceae
of alkaliphilic Bacillus strains isolated from various Halorubrum
sources and proposed that they were classified into Natrialba
13 taxa and a number of unassigned single-membered Natronobacterium
clusters by using DNA-DNA hybridization, numerous Natronococcus
physiological and biochemical characteristics, DNA base Natronomonas
composition, and 16S rRNA analyses. Two taxa were Natronorubrum
equated with B. cohnii and B. alcalophilus and nine of Methanomicrobiales
Methanosarcinaceae
the remainder were proposed as new species with the
Methanohalophilus
following names: B. agaradhaerens, B. clarkii, B. clausii,
B. gibsonii, B. halmapalus, B. halodurans, B. horikoshi,
B. pseudoalcalophilus, and B. pseudofirmus. Two taxa
were insufficiently distinct to allow confident identification and Natronococcus are alkaliphilic archaea that grow only
and these have therefore not been proposed as new species. in specialized Milieux with the combination of extremely
In a separate study, Agnew and coworkers (16) reported high salt concentrations (2.5–5.2 M NaCl) and high pH
the isolation and characterization of novel obligate values (8.5–11.0). Such haloalkaliphiles have been iso-
alkaliphiles from bauxite-processing waste. They are lated from soda lakes and soda deserts. The amount of
currently assigned to the new species B. vdderi. The new glycolipid in the membrane hardly varied in haloalka-
isolates appear to be members of the genus Bacillus but liphilic archaea when compared with that of haloneu-
16S rRNA sequence characters and other physiological
trophilic archaea. A novel osmolyte, 2-sulfotrehalose,
and biochemical analysis indicate that the new isolates
was discovered in several Natronobacterium species of
are not members of any validly identified Bacillus species.
haloalkaliphilic archaea (17). The concentration of sul-
fotrehalose increases with elevating concentrations of
Haloalkaliphiles
external NaCl, a behavior consistent with its identity
Oren (12) classified the Halobacteriaceae into six gen- as an osmolyte. Other common osmolytes (glycine betaine,
era, Halobacterium, Haloarcula, Haloferax, Halococcus, glutamate, and proline) were neither accumulated nor
Natronobacterium, and Natronococcus. Natronobacterium used for osmotic balance in place of the sulfotrehalose.
136 AEROBIC ALKALIPHILES

Kamekura and coworkers (18) studied the diversity of Line 1a Line 1b Line 2
haloalkaliphiles on the basis of phylogenetic tree recon- H.jadinii (10.5) H.ciferrii (10.5)
H.beijerinckii (10.0)
structions, signature bases specific to individual gen- H.saturnus var. (C.utilis (10.5) H.anomala var.
era, and sequences of space regions between 16 and subsufficiens (10.0) schneggii (9.0)
23S rRNA genes. They proposed the following changes: H.saturnus var. H.anomala var.
anomala (10.0)
Natronobacterium pharaonis be transferred to a new Line 5 saturnus (10.0) H.petersonii (10.5)
H.mrakii (10.0) H.subpelliculosa (8.5)
genus, Natronomonas as Natronomonas pharaonis and
Natronobacterium magadii to be transferred to a new H.polymorpha (8.5)
H.dimennae (10.5)
species of the genus Natrialba as Natrialba magadii. H.glucozyma (7.5)
(P.pinus (7.5)) H.holstii (8.5)
Xu and coworkers (19) isolated and characterized novel H.californica (10.5)
haloalkaliphilic archaea from a soda lake in Tibet, China. H.minuta (7.5) Line 4
Line 3 H.silvicola (8.5)
They proposed to classify their strains in a new genus
Natronorubrum. At present, microorganisms that belong H.beckii (7.5) H.wingei (9.0)
H.capsulata (7.5)
to Halobacteriaceae have 16 genera registered in the taxon- H.canadensis (8.0)
omy section of the National Center for Biotechnology Infor- (C.mellinii (8.5))
H.bimundalis (7.5)
mation (NCBI) (http://www.ncbi.nlm.nih.gov/htbin-post H.henricii (7.5)
/Taxonomyhome.html/). These six genera include Haloal- H.nonfermentans (7.5) H.wickerhamii (8.5)
calophilium in Halobacteriaceae, belonging to haloalka-
liphilic archaea.
Romano and coworkers (20) isolated and characterized
a haloalkaliphilic microorganism from hard sand of Lake Ancestor species
Venere on Pantelleria Island, Italy. This haloalkaliphile is Figure 3. Alkali-tolerant yeasts on a phylogenetic tree of the
a gram-negative pleomorphic rod, strictly aerobic, capable genus Hansenula. The number in parentheses shows the upper
of accumulating polyhydroxybutyrate, and growing opti- pH limit for growth of a given strain. Species related to Hansenula
mally at pH 9.0 in the presence of 10% NaCl at 33–35 ° C. are shown in parentheses. (Reproduced with permission from
They classified this strain in a new species of the genus Aono, Syst. Appl. Microbiol. 15, 589 (1992)).
Halomonas.

Alkaliphilic Cyanobacteria
Yeast and Fungi Gerasimenko and coworkers (26) reported the discovery
Most species of yeasts and yeast-like fungi can grow of a wide variety of alkaliphilic cyanobacteria (16 genera
well in the acidic to neutral pH range. It has long and 34 species were found). Buck and Smith (27) reported
been assumed that yeasts cannot grow at alkaline pH an Na+ /H+ electrogenic antiporter in alkaliphilic Syne-
except for unusual strains. A few investigators have chocystis sp. Singh (28) partially purified a urease from an
reported that certain yeasts could grow in initially alkaline alkaliphilic diazotrophic cyanobacterium, Nostoc calcicola.
media (21,22). However, these investigations were not Strain Z7935 (T) is an obligatory sodium-dependent alka-
carried out under appropriate conditions. The media used liphile, which grows in a sodium carbonate medium and
were poorly buffered and the change in pH was not does not grow at pH 7; the maximum pH for growth is more
followed. than pH 10, and the optimum pH is 9.5 to 9.7. The opti-
Goto and coworkers (23) examined their microbial mum NaCl concentration for growth is only 3% w/v (13).
properties and proposed a new species, Exophiala
alcalophila Goto et Sugiyama, with an accompanying Thermoalkaliphiles
new yeast morph Phaeococcomyces alcalophilus Goto et Early work on alkaliphiles suggested that microorganisms
Sugiyama. These strains grow between pH 5.4 and 10.4 adapted to the two extreme environmental conditions of
on YM medium (5 g peptone, 3 g malt extract, 3 g yeast high temperature and high alkaline pH did not exist.
extract, 15 g glucose, and 15 g agar in 1,000 ml distilled However, many kinds of thermoalkaliphiles have been
water). Aono (24) studied the taxonomic distribution isolated from different habitats since the 1980s. These
of alkali-tolerant yeasts in his laboratory. Yeasts and include an obligate alkaliphilic Clostridium species iso-
yeast-like fungi (433 strains, 296 species, 53 genera) were lated from sewage (29), an asporogenous, gram-positive
examined to determine the upper limit of pH for growth. ammonifying anaerobe from a soda lake deposit, Tindallia
Among these, 135 strains of 86 species were found to mamadii (30), a xylan-degrading anaerobic thermoalka-
be capable of growth at a pH above 10. These alkali- liphile designated as strain LB3A (31), and an actino-
tolerant species belonged to 27 genera of yeasts, of mycete, Thermoactinomyces sp. HS682 (32). Stetter and
which 10 genera contained only alkali-tolerant species. colleagues (33) have even described a hyperthermophilic,
Furthermore, Aono (25) measured upper pH limits for alkaliphilic archaea, Thermococcus alcaliphilus, which
growth of yeasts belonging to the genus Hansenula and grows on polysulfide at temperatures between 56 ° C and
some related strains. As shown in Figure 3, alkali-tolerant 90 ° C and with an optimum temperature around 85 ° C.
species occupied particular positions of a phylogenetic tree The pH range for growth was 6.5 to 10.5, with an optimum
proposed for the genus. around 9.0.
 AEROBIC ALKALIPHILES 137

Other Alkaliphiles walls groups 1 and 2 have specific acidic polymers called
teichuronopeptide (TUP); TUP is a polymer in which an
In contrast to thermoalkaliphiles, Kimura and Horikoshi
acidic polypeptide binds covalently to polyglucuronic acid.
(34) isolated a Micrococcus that was alkalipsychrotrophic
Cell wall concentrations of TUP increase with respect to
in nature. This microorganism showed the highest growth
peptidoglycan as the culture pH is elevated (39). While
rate at 0 ° C and produced an amylase whose properties
the mechanism is not clear, it is thought that the acidic
might be useful in food-processing circumstances.
fraction of the cell walls may act as an obstacle to the
Several alkaliphilic spirochetes were also isolated from
high extracellular concentrations of hydroxide ions or as a
Lake Magadi and from Lake Khatyn, Central Asia,
reservoir of hydrogen ions (39).
by Zhilina and coworkers (35). Analysis of the genes
In other alkaliphilic Bacillus species, for example,
encoding 16S rRNA indicated a possible fanning out of
B. pseudofirmus OF4, the uronic acid polymers are not
the phylogenetic tree of spirochetes.
found (40). Recent studies have shown that B. pseudofir-
mus OF4 has an acidic S-layer polymer produced from
PHYSIOLOGY a gene that has strong homology with genes from neu-
trophilic B. anthracis and B. licheniformis. This protein
The Feature of Cell Surfaces in the Alkaliphilic Bacillus was identified in studies of the 2D gel electrophoresis
Species patterns of membrane-associated proteins of pH 10.5- and
Cell Walls. Aono and Horikoshi classified the alka- 7.5-grown cells of B. pseudofirmus OF4. It is a homoge-
liphilic Bacillus species into three groups on the basis neous, apparently processed, protein that is present in
of cell wall composition and other physiological char- greater amounts at high pH. On cloning, sequencing, and
acteristics (36). As described in Table 3, the alkaliphilic disruption of the gene, the S-layer was found not to be
Bacillus strains of group 1 contain large amounts of glu- required for alkaliphily in B. pseudofirmus OF4, but to
curonic acid and hexosamine in the cell walls and cannot confer an advantage for growth at pH 10.5. Cytoplasmic
grow at a neutral pH. The cell walls of group-2 strains pH homeostasis, on a sudden shift from pH 8.5 to 10.5,
contain large amounts of acidic amino acids and uronic is somewhat better in the S-layer-containing wild type
acids. Na+ is essential for their growth. The quantities than in mutant strains lacking the S-layer (41). Bacillus
of acidic compounds in the cell walls of both groups 1 pseudofirmus OF4 may also have an acidic capsular layer
and 2 are increased when cells are grown at elevated pH. partial sequence for genes that are likely to encode the
In contrast, the cell walls of group-3 organisms contain synthetic enzymes for a polyglutamate capsule that were
phosphorus and neutral sugars in large amounts and are characterized in this species (42). It will be of interest to
essentially similar to the walls of the neutrophilic Bacillus complete the characterization of the sequence and the role
subtilis in chemical composition. Growth was observed in of this locus.
the presence of Na+ or K+ .
Auxotrophy, Antibiotic Resistance
The Role of the Cell Walls in Growth at Alkaline The majority of alkaliphilic microorganisms isolated from
pH. Protoplasts prepared from alkaliphilic Bacillus are nature are apparently auxotrophic. Such alkaliphiles
unstable at alkaline pH and regenerate only at neutral cannot grow in completely synthetic media (e.g., the M9
pH but not at alkaline pH (37). This suggests that some culture media in which Escherichia coli and B. subtilis can
mechanisms of adaptation to high pH, ‘‘alkaliphily,’’ grow). However, some alkaliphilic Bacillus strains can;
might involve the cell walls, the part of the cell that is these are used for molecular biological manipulations. It
directly exposed to the extracellular environment. The is possible, in these strains, to introduce useful markers
cell walls of alkaliphilic Bacillus consist of peptidoglycan by chemical mutagenesis (1). Transformants are selected
and non-peptidoglycan components. The peptidoglycan with the antibiotic resistance when plasmid DNA is
from all three groups is of the A1γ type, identical to incorporated into alkaliphilic microorganisms. In this case,
that in the cell walls of neutrophilic B. subtilis (38). Cell common antibiotics such as kanamycin and tetracycline
cannot be used because they are very unstable under
alkaline pH (1). Chloramphenicol and erythromycin are
Table 3. Classification of Alkaliphilic Bacillus Strains
stable at alkaline pH and can be used.
Major Components Ion
Group Growth pH of Cell Wall Requirement Lipid Composition
1 No growth at pH 7 Glucuronic acid Na+ (essential) Clejan and coworkers (43,44) have reported that obligately
Hexosamine and facultatively alkaliphilic bacteria exhibit significant
2 Capable of growth Aspartic acid Na+ (essential) differences in lipid constituents. They compared the
at pH 7 Glutamic acid lipid composition of several obligately and facultatively
Galacturonic alkaliphilic Bacilli. High levels of cardiolipin were
acid reported, that is, 13% and 25% of the polar lipids
Glucuronic acid
in B. pseudofirmus RAB and OF4, respectively. The
3 Capable of growth Phosphoric acid Na+ or K+
at pH 7 Glycerol
obligately alkaliphilic B. pseudofirmus RAB had a much
Neutral sugar higher neutral/polar lipid ratio than the facultatively
alkaliphilic B. pseudofirmus OF4. The B. pseudofirmus
138 AEROBIC ALKALIPHILES

RAB had 90% branched chain fatty acids as opposed to in alkaliphily is associated with the cell surface and its
72% in the B. pseudofirmus OF4. The lipids of obligate transporters, which must contribute to an intracellular
alkaliphiles contain a significant amount of unsaturated neutral Milieux that is separate and distinct from the
fatty acids, but these fatty acids are hardly present in the extracellular alkaline Milieux.
lipids of facultative alkaliphiles. Although the relationship
of unsaturated fatty acid content to obligate alkaliphily Na+ /H+ Antiporters Involved in the Intracellular
has not been demonstrated, it may be hypothesized to Acidification Relative to the Extracellular Milieux
involve membrane fluidity. Since alkaliphiles live in an extremely alkaline Milieux,
their most challenging problem is keeping a cytoplasmic
Flagella and Sodium Ion pH more acid than the outside pH, often by more
Flagellar movement of neutrophilic, nonmarine bacteria is than 1.5 pH units. In other words, they must actively
not caused by ATP but by H+ -driven motors (45). However, bring protons into the cell. In contrast to research with
flagellar movement of alkaliphiles and marine bacteria neutrophiles, an especially strong case can be made for
is accomplished by Na+ -driven motors. These Na+ -driven the acidification of the cytoplasm of alkaliphiles by Na+ /H+
flagellar motors are inhibited by amiloride (46). It has been antiporters (48). When alkaliphiles are placed in a medium
reported that even on a significant shift to alkaline pH, without Na+ at pH 10, the intercellular pH quickly rises to
Na+ -driven flagellar motors do not rotate without sodium the value of the outside pH. However, when Na+ is present
ions (1). The structural difference of a flagellum motor of in the external medium, the intercellular pH does not
the H+ type and the Na+ type is not seen by observation at rise on shifting to the more basic medium. Furthermore,
the electron microscope level. Even if there is a difference mutants of alkaliphiles that cannot grow at pH values
in the coupling ion between H+ - and Na+ -coupled flagella, above 9 are defective in Na+ /H+ antiporter activity. The
the basal structure and the mechanism of the rotation are antiporter is electrogenic (H+ >Na+ ) and driven by the
thought to be similar (1). membrane potential (), which is generated by the
primary proton pumps of respiratory chains. The sodium
Membrane Transport ion circuit is completed when sodium ion enters the cell
via Na+ /solute symporters that are also driven by the .
In the growing cell, nutrients can be taken up into the cell The use of Na+ /solute symporters has the advantage that
and metabolized, followed by secretion of the products. solute transport is driven by the sodium potential rather
Uptake of the nutrient and secretion of the products than the proton potential, the latter being low because of
involve passage through the cytoplasmic membrane. The the inverted pH (pHin -pHout ).
cytoplasmic membrane also plays an active role in pH
homeostasis (see section on Antiporters). Alkaliphiles have Respiration-Dependent ATP Synthesis
numerous transport systems that can take up nutrients
efficiently from the highly alkaline environment (47). In Alkaliphiles always maintain their cytoplasmic pH lower
bacteria, amino acids are usually taken up against a than the outside pH, the opposite of most bacteria. This
concentration gradient by active transport. Usually, the raises a serious conundrum for energy conservation. To
energy required for active transport is provided either
by ATP or by the proton motive force. Under the high Table 4. Intracellular pH Values in Alkaliphilic
alkaline Milieux, the proton motive force decreases, and Bacillus Strains at Different External pH Values
alkaliphiles tend not to utilize proton as a coupling ion.
Alkaliphiles have Na+ -coupled uptake systems instead of Microorganism External pH Internal pH
proton as a coupling ion. Other monovalent cations such as
B. alcalophilus 8.0 8.0
K+ , Li+ , and NH4 + could not be substituted for Na+ , and
9.0 7.6
different species of counteranions for Na+ did not affect 10.0 8.6
the uptake. The uptake of amino acids was stimulated by 11.0 9.2
the addition of Na+ ; this was especially true for glycine, B. pseudofirmus OF4 7.0 7.7
L-alanine, L-serine, and L-asparagine (47). 9.0 8.0
10.8 8.3
Intracellular pH 11.2 8.9
11.4 9.6
Most alkaliphiles have an optimum growth pH at around Bacillus strain YN-2000 7.5 7.5
pH 10, which is significantly different from that toler- 8.5 7.9
ated by well-investigated neutrophilic microorganisms. 9.5 8.1
Therefore, the question arises as to how these alka- 10.2 8.4
liphiles can grow in such an extreme Milieux. Are there B. halodulans C-125 7.0 7.3
any fundamental differences in the physiology of alka- 7.5 7.4
liphilic and neutrophilic microorganisms? If intracellular 8.0 7.6
8.5 7.8
pH is estimated by measuring the extra- and intra-cellular
9.0 7.9
distribution of weak bases, which cells do not actively
9.5 8.1
transport, the intracellular pH is found to be maintained 10.0 8.2
at around 8, despite a high extracellular pH of 8 to 11, 10.5 8.4
as shown in Table 4 (1). Therefore, one of the key features
 AEROBIC ALKALIPHILES 139

survive in an alkaline Milieux, alkaliphiles must adjust OF4, B. halodulans C-125, and Micrococcus sp. Y-
their strategies of energy conservation. One possibility is 1 (50–52). The first whole genome sequence project
to change the coupling ion. For example, some anaer- for an alkaliphilic Bacillus, B. halodurans C-125, was
obic alkaliphiles use Na+ -coupled ATP synthases (49). completed in 2000 (53) and the postgenome project,
However, this is surprisingly not adopted by the alka- especially directed toward clarifying the mechanisms of
liphilic aerobic Bacillus species. These microorganisms alkaliphily, has begun. The vastly increasing database on
have respiratory chains that pump H+ ion outward and sequences of alkaliphilic proteins will provide insights
H+ -coupled ATP synthases, apparently like those of neu- into global adaptations in cytoplasmic proteins or in
trophilic microorganisms. For the present, the mecha- functional cytoplasmic assemblies, namely, ribosome,
nism of ATP synthesis in these microorganisms remains secretory particles, and so forth.
obscure. Krulwich (47) presented several possible models
of the proton translocated outward by respiratory chains
CONCLUSION
during ATP synthesis. One possibility is that the protons
in electron transfer reactions are not released into free
Alkaliphiles are unique microorganisms with great
solution but extruded directly by the respiratory chains
potential for physiology and biotechnology. The aspects
to the ATP synthase. A diagram of these elements is
that have received most attention in the late 1990s
shown in Figure 4. An Na+ /H+ antiporter, which is the include (1) extracellular enzymes and genetic analysis
major mechanism for extruding Na+ and which also func- of their production, (2) mechanisms of ‘‘alkaliphily,’’ and
tions to bring protons into the cell for pH homeostasis (3) taxonomy of alkaliphiles. Taxonomy of haloalkaliphiles
in alkaliphilic Bacillus species, is shown. The Na+ /H+ and anaerobic alkaliphiles is a very active field, and
antiporter creates the sodium potential necessary for the the number of microorganisms keeps increasing. So far,
Na+ /solute symporters because a primary Na+ pump is studies of alkaliphiles belonging to the genus Bacillus
not present. The antiporter uses the proton motive force have led to the discovery of many alkaline enzymes and
(p) as an energy source. Also shown are Na+ /solute sym- mechanisms of ‘‘alkaliphily’’ under extremely alkaline
porters that use the sodium electrochemical potential to pH. However, new findings and an even more diverse
accumulate solutes. The flagellar motor is drawn to turns group of applications of alkaliphiles from genome analysis,
at the expense of the sodium electrochemical potential. proteome studies, and novel ecological, molecular, and
The respiratory chain represents a hypothesis in which biophysical approaches will soon provide greatly expanded
protons are transferred directly from this complex to the developments.
proton coupled F1 Fo-ATP synthase in a protein–protein
interaction. The cell wall–associated layer reflects the
finding that, in B. halodurans C-125 and B. pseudofirmus BIBLIOGRAPHY
OF4, respectively, different negatively charged polymers
1. K. Horikoshi, Alkaliphiles, Kodansha, Tokyo, Japan, 1999.
play at least some role in pH homeostasis (47).
2. H. W. Johnson, Iowa State Coll. Eng. Exp. Sta. Bull. 76,
307–344 (1923).
Genomics
3. A. W. Downie and J. Cruickshank, Br. J. Exp. Pathol. 9,
Numerous alkaliphile genes have been sequenced and 171–173 (1928).
physical maps have been reported from B. pseudofirmus 4. T. Gibson, J. Bacteriol. 28, 313–322 (1934).

Acidic polymer pHout = 10.5

Negative charged Cell wall
Respiratory chain H+-ATPase
S-layer protein
capsule H+ H+ H+
Cytoplasmic
Interaction membrane
?

Figure 4. A model depicting the energy coupling

+ + mechanisms of alkaliphilic Bacillus species. The active
Na /H
pH homeostasis mechanism is shown as including
antiporter
+ respiration, Na+ /H+ antiporter, and Na+ re-entry
H+ H H+ H+
ADP, Pi ATP pathway. The respiration chain extrudes H+ (electron
+
H+
+
Na Na + donors/acceptor not shown). The Na+ /H+ antiporter
S Na
is probably more Na+ -specific than those involved in
Flagellar pHin = 8.5 pH homeostasis in neutralophilic bacteria. The Na+
motor re-entry pathways are shown as a channel associated
+
with the flagellum and Na+ /solute symporters (47).
Na Other participants that have been suggested to
function in the establishment and/or maintenance of
an acidified cytoplasm relative to the medium are
Flagellar
S Na+ indicated and are discussed in the text. See color
Na+/solute symporter insert.
140 AEROBIC ENDOSPORES

5. A. Vedder, Ant. V. Leeuenhoek 1, 141–147 (1934). 52. H. Takami et al., Extremophiles 3, 21–28 (1999).
6. Y. Takahara and O. Tanabe, J. Ferment. Technol. 38, 53. H. Takami et al., Nucleic Acids Res. 28, 4317–4331 (2001).
329–331 (1960).
7. W. D. Grant et al., FEMS Microbiol. Rev. 75, 235–270
(1990).
8. M. D. Collins et al., J. Gen. Microbiol. 129, 2037–2042 (1983).
AEROBIC ENDOSPORES
9. S. Ntougias and N. J. Russell, Extremophiles 4, 201–208
EVA C. NIEMINSKI
(2000).
Utah Department of
10. H. Takami et al., FEMS Microbiol. Lett. 152, 279–285 (1997).
Environmental Quality
11. B. E. Jones et al., Extremophiles 2, 191–200 (1998). Salt Lake City, Utah
12. A. Oren, FEMS Microbiol. Rev. 13, 415–439 (1994).
13. T. N. Zhilina et al., Int. J. Syst. Bacteriol. 47, 144–149 (1997). Aerobic spores are not pathogenic to humans, which makes
14. D. Fritze et al., Int. J. Syst. Bacteriol. 40, 92–97 (1990). them useful as indicators of treatment effectiveness for
15. P. Nielsen et al., Microbiology 141, 1745–1761 (1995). drinking water. Because direct monitoring for waterborne
pathogens such as protozoa and viruses is very inaccurate,
16. M. D. Agnew et al., Syst. Appl. Microbiol. 18, 221–230 (1995).
time-consuming, and expensive, monitoring for aerobic
17. D. Desmarais et al., J. Bacteriol. 179, 3146–3153 (1997).
spores can become a viable surrogate for the removal
18. M. Kamekura et al., Int. J. Syst. Bacteriol. 47, 853–857 effectiveness of pathogens and parasites. Moreover,
(1997).
because most pathogenic organisms, even those found
19. Y. Xu et al., Int. J. Syst. Bacteriol. 49, 261–266 (1999). abundant in water sources, seldom are detected in treated
20. I. Romano et al., Syst. Appl. Microbiol. 19, 326–333 (1996). drinking water, calculation of their removal through
21. M. E. DiMema, J. Gen. Microbiol. 20, 13–23 (1959). treatment is not possible. In contrast, aerobic spores are
22. B. Norkrans, Arch. Mikrobiol. 54, 374–392 (1966). present in high concentration in most surface waters, and
23. S. Goto et al., Trans. Mycol. Soc. Japan 22, 429–439 (1981). their presence can also be detected in treated water, thus
24. R. Aono, Syst. Appl. Microbiol. 13, 394–397 (1990). allowing for an evaluation of treatment effectiveness.
25. R. Aono, Syst. Appl. Microbiol. 15, 587–589 (1992). The effectiveness of disinfection, especially the ultravio-
26. L. M. Gerasimenko et al., Microbiology 65, 736–740 (1996).
let (UV) light treatment of both water and wastewater, can
also be evaluated using aerobic spores as indicators. Being
27. D. P. Buck and G. D. Smith, FEMS Microbiol. Lett. 128,
resistant to chemical disinfection and UV irradiation, aer-
315–320 (1995).
obic spores can be used in development of disinfectant
28. S. Singh, Folia Microbiol. Prague 40, 529–533 (1995).
dose-response curve, thus allowing for a calibration and
29. J. Wiegel, Extremophiles 2, 257–267 (1998).
measurement of disinfection efficiency.
30. V. V. Kevbrin et al., Curr. Microbiol. 37, 94–100 (1998).
31. A. Sunna et al., FEMS Microbiol. Lett. 148, 209–216 (1997).
BACKGROUND
32. K. Tsuchiya et al., Biosci. Biotechnol. Biochem. 56, 246–250
(1992).
As we enter the new millennium, the drinking water
33. M. Keller et al., Arch. Microbiol. 164, 390–395 (1995).
industry is undergoing a major paradigm shift. This
34. T. Kimura and K. Horikoshi, Agric. Biol. Chem. 53, is because of increasing concerns over the potential
2019–2020 (1989).
presence of pathogens in potable waters and the large
35. T. N. Zhilina et al., Int. J. Syst. Bacteriol. 46, 305–312 (1996). number of people with compromised immune systems. It
36. R. Aono and K. Horikoshi, J. Gen. Microbiol. 129, 1083– has become clear that continuous process optimization
1087 (1983). based only on the current practice of turbidity monitoring
37. R. Aono et al., Biochem. J. 285, 99–103 (1992). (and to some extent particle counting) of treated water
38. R. Aono et al., J. Bacteriol. 157, 688–689 (1984). needs to be reevaluated (1). Although monitoring finished
39. R. Aono et al., Microbiology 141, 2955–2964 (1995). water quality is an essential step in the production
40. A. A. Guffanti and T. A. Krulwich, J. Biol. Chem. 269, of safe drinking water, traditional linkages between
21576–21582 (1994). process monitoring and plant operation fail to provide
41. R. Gilmour et al., J. Bacteriol. 182, 5969–5981 (2000). an adequate margin of safety to enable operators to
42. M. Ito et al., Extremophiles 1, 22–28 (1997). respond to temporal changes in source water quality (2).
43. S. Clejan et al., J. Bacteriol. 168, 334–340 (1986). This monitoring paradox manifests itself periodically
when cases of waterborne disease outbreaks occur even
44. S. Clejan and T. A. Krulwich, Biochim. Biophys. Acta 946,
40–48 (1988). when finished water is in full compliance with regulatory
requirements (3–5).
45. D. F. Blair and H. C. Berg, Cell 60, 439–449 (1990).
Water treatment regulations are becoming increasingly
46. T. Atsumi et al., J. Bacteriol. 172, 1634–1639 (1990).
more stringent. However, emerging regulatory require-
47. T. A. Krulwich et al., Biochim. Biophys. Acta 1,505, 158–168 ments do not provide utilities with a framework for
(2001).
optimizing treatment operations and processes to achieve
48. N. Kaieda et al., FEMS Microbiol. Lett. 167, 57–61 (1998). finished water quality objectives. Alternative approaches
49. T. A. Krulwich, Mol. Microbiol. 403, 27365–27371 (1995). for accommodating variability in source water quality are
50. A. Gronstad et al., Extremophiles 2, 447–453 (1998). needed to provide plant operators with a sound basis for
51. J. H. Park et al., Microbiology 140, 2247–2250 (1994). ensuring reliable plant performance.
 AEROBIC ENDOSPORES 141

Monitoring for microbial pathogenic contaminants inadequate. Besides, the reliability and reproducibility of
presents a special challenge. Sampling and analytical the existing methods being an issue, there is a serious
methods that are available to detect Giardia and time lag between sample collection and data availability.
Cryptosporidium have limited sensitivity and accuracy (6). Therefore, it is impossible to use even these flawed
The time required for analyses does not allow for the methods as a means to optimize treatment processes.
practical use of the results in process control. Lack of Despite the water industry and the U.S. Environmental
‘‘real-time’’ measurements of pathogens in water prevents Protection Agency’s (EPA’s) attempts to improve this
utilities from being able to react to source water quality analytical method, the problems with its use remain.
changes and provide proactive treatment optimization. In Although research into innovative and more rapid methods
the absence of quick and reliable analytical methods for using such techniques as polymerase chain reaction
pathogens, pathogen indicators should be used. and flow cytometry is promising, additional research is
The need for some measure that utility operators can required before practical monitoring tools are available to
use to determine how well their plants are optimized the water industry.
for the removal of pathogens is paramount if they are
to achieve maximum treatment effectiveness. It has SURROGATES FOR PROTOZOA IN DRINKING WATER
become clear that turbidity, as currently regulated, is an
inadequate measure for this purpose. Recent waterborne In recent years, a substantial effort has been focused
outbreaks in the United States have been reported on investigating the use of surrogates for optimizing
even though turbidity was within regulatory limits. treatment of drinking water and pathogen removal.
Moreover, research has shown the presence of Giardia Generally, it is agreed that an ideal surrogate would be
and Cryptosporidium even at turbidities of less than abundant in water but not multiplying in treatment plant
0.1 NTU (4,7,8). A further complication to the turbidity basins, resistant to treatment, and easy to detect and
issue is that optical turbidimeters are not accurate at enumerate with cost-effective methods. The search for
readings less than 0.1 NTU (9). the surrogate includes both microbial and nonmicrobial
In recent years, the use of on-line measurements by parameters.
utilities to supplement routine sampling and analysis A research project on biological particle surrogates
has increased substantially. Currently, the major types for filtration performance, performed at Colorado State
of on-line monitors in use at water treatment facilities University in 1995–1998 (13), has addressed the need for
include turbidity, pH, and ion-specific electrodes. Other indicators that would serve as tools in plant optimization
options include particle counting and streaming current and was designed to evaluate a wide spectrum of
monitors. Relationships between on-line measurements microorganisms and their potential to serve as surrogates
and other monitoring parameters have not been developed. for monitoring of pathogen removal through filtration.
Correlation between suspended solids, particle counting, This effort indicated that some bacteria, bacterial phages,
turbidity, and microbial content are highly source- and algae could represent Giardia and Cryptosporidium
specific (7), and most likely depend on the fraction of in removal mechanisms. In pilot tests, the use of
organic matter contained in the colloidal and suspended various microbial surrogates was tested through seeding
solids (10). experiments. Significant correlation was established
Common indicators are needed to establish a platform between the removal of Giardia and Cryptosporidium and
for comparison of water quality of various sources some of the surrogates tested. However, because of a lack
and their treatability, as well as to evaluate the of supportive data, the significance of these correlations
efficiency of treatment achieved at various plants. may be suspect. One suggested conclusion was that the
Some indicators can be selected from the commonly strength of the correlations was dependent on the quality
used water quality parameters, such as turbidity, of measurements or enumeration of the surrogates. In fact,
particle counts, and coliform bacteria (10–12). A critical the strongest correlation was found with those surrogates
assessment of microbiological parameters that could serve that had the smallest deviation within the data set. A
as surrogates for pathogenic microorganisms is needed conclusion was also suggested that the deviation observed
to identify readily measurable analytes that can be was due to error in measurement — the smaller the
quickly incorporated into the decisions made about water error, the more apparent the ‘‘true’’ relationship between
treatment plant operation. pathogen and surrogate removal. The study demonstrated
The need for a surrogate has become increasingly the difficulty in finding a reliable surrogate when there
apparent in the water industry in light of well-documented is so much uncertainty in the analytical methods for
cases of disease outbreaks traced to drinking water as the enumerating the pathogens.
route of transmission. Although the causative agent for a Optical particle counting and sizing (‘‘particle count-
number of these outbreaks was identified as Giardia or ing’’) has been considered a good candidate for a surrogate
Cryptosporidium, the affected water treatment plants did measure for pathogen occurrence in watersheds and treat-
not have an effective way to monitor for the presence ment processes. Importantly, many utility operators and
of the pathogen or to judge the treatment efficiency consulting engineers have begun to use this analytical
in eliminating the organisms from the finished water. technique in water production. A project evaluating data
Since these outbreaks, it has become even more apparent from 100 drinking water plants across the United States
that the current methods for analyzing for Giardia and was initiated to assess particle removals by filtration (8).
Cryptosporidium in watersheds and in finished water are Specific objectives of this study included: distribution of
142 AEROBIC ENDOSPORES

particle counts, factors that affect finished water parti- at various laboratories, making data comparison diffi-
cle counts, evaluation of the relationship between particle cult, the resulting database concerning removal of these
counts and pathogen concentrations in source and filtered pathogens could be a valuable source of information
water. Some relevant findings of this project indicated concerning pathogen occurrence. The Giardia and Cryp-
that particle counting can be used by utilities to opti- tosporidium occurrence data, collected during the ICR,
mize treatment for particulate removal, but a relationship show an overwhelming majority of sample results falling
between the two parameters is not straightforward. In below the analytical detection limit, making the evalu-
full-scale sampling, a correlation between the occurrence ation of national occurrence of these pathogens and an
of pathogens and the number and distribution of particles assessment of their treatment effectiveness more difficult
in watersheds was not substantial. Particle count was not and stressing a need for their surrogate more than before.
found to be a good indicator of pathogen occurrence in
watersheds. An exception to this observation was noted
AEROBIC SPORES AS SURROGATES FOR REMOVAL OF
during storm and run-off events that resulted in higher lev-
PROTOZOAN PARASITES
els of particles and in pathogens entering the source water
intake. These data indicated that particle-counting mea-
A research effort, sponsored in 1995 by the U.S. EPA
surement could be used as an alarm for unusual events,
in Cincinnati, Ohio, focused on comparison between
but not as a reliable indicator of the presence or absence of
occurrence and removal of indigenous aerobic bacterial
pathogens in a watershed. The uncertainty in the method spores and particle counts (15). The project findings
used in enumerating the pathogen was indicated as a fac- indicated that spores of spore-forming aerobic bacteria
tor preventing the accurate comparison of particles and such as Bacillus could serve as a good surrogate of
pathogen concentrations. pathogen removal. Naturally occurring Bacillus spores
One of the important caveats when looking for a can be easily and quickly detected in both the source and
surrogate, however, is that the concept of log removal as the treated water. The spores did not propagate in the
a measure of filtration effectiveness is somewhat limiting various treatment processes. Removal of Bacillus spores
in that the maximum log removal is governed by the correlated highly with removal of particles. The working
detection limit of the analytical methods. Therefore, the relationship between the project team at EPA and the team
maximum log removal that could have been obtained when involved in this proposed project allowed for an exchange
considering the pathogen data is governed by the source of information and a continuation of efforts undertaken by
water pathogen levels. Much higher levels of particles other researchers.
occurring in source water allowed calculation of higher In July 1994, the State of Utah and five major
removal effectiveness. It is almost numerically impossible Utah water utilities organized the Utah Water Quality
to obtain similar log removals for particles and pathogens. Alliance and initiated a Water Quality Enhancement
However, because pathogens are ‘‘particles,’’ it stands to Study to evaluate treatment effectiveness based on an
reason that measuring particulate concentrations through increasing need to provide protection against pathogenic
a treatment process and optimizing the performance of microorganisms (16). In this study, 10 water treatment
that process to reduce particles will result in lower levels plants were sampled for Giardia, Cryptosporidium, total
of pathogens. Particle counting is a good measure of these aerobic spores, Bacillus spores, turbidity, and particle
particulate changes. Measurements of finished water count. The project tasks included source and finished
concentrations of particles or any microbial indicator of water quality data collection, development of water quality
water quality could provide the assurance of consistency standards, search for surrogate analytical tools, and
of plant operations. Water treatment systems should focus treatment optimization. A major conclusion of the study
on the consistency of providing very good water quality, was that conventional monitoring practices do not provide
even during changes in source water quality. Consistently sufficient information on source water quality variations
low concentration of particulates (and therefore microbial to provide plant personnel with adequate response time
contaminants) in finished water may present a higher to adjust plant operations. Thus, the ability to ensure
confidence in high water quality than the assurance a consistently reliable treated water quality, free of
of a specified log removal of these particulates. The pathogenic organisms, is handicapped by inadequate
direct monitoring of finished water quality may therefore linkages between monitoring and operating practices. A
be in conflict with the concept of log removal because key issue that resulted from the study is the need for
when source water concentrations increase, so could the improved integration of source water quality parameters
log removal increase even if the finished water quality into strategies for operation of treatment unit processes.
deteriorates. To evaluate potential pathogen surrogates for treat-
The database of the U.S. EPA’s Information Collection ment plant performance optimization, log removals for
Rule (ICR), (14), is an additional source of information all measured parameters (Giardia, Cryptosporidium, total
about presence and removal of pathogens in treatment aerobic spores, Bacillus, particle count, and turbidity) were
plants across the United States. The ICR data were compared. The best correlation was seen between either
collected and made available as a result of monthly mon- total aerobic spores or Bacillus and Cryptosporidium fol-
itoring for Giardia, Cryptosporidium, and enteric viruses lowed by Giardia, particle counts, and turbidity. It was
conducted for 18 months (July 1997–January 1988) at concluded that with lack of detectable amounts of Giardia
more than 500 plants. Although samples were analyzed and Cryptosporidium in filtered water, either total aerobic
 AEROBIC ENDOSPORES 143

spores or Bacillus spores proved to be a reliable measure pathogens were evaluated. Finally, the effects of treatment
for assessing plant performance effectiveness. type, filter type, disinfectant type and application point,
In 1996, the American Water Works Association filtered water pH, as well as source water quality (turbidity
Research Foundation sponsored a study to evaluate and pathogen concentration) on the effectiveness of using
the applicability of pathogen surrogates to be used by surrogates to evaluate treatment plant performance were
treatment plant operators as a tool in continuous plant assessed.
performance optimization (17,18). The target pathogens
Aerobic Spores Versus Other Surrogate Parameters for
were Giardia, Cryptosporidium, and enteric viruses.
Protozoa in Drinking Water
Pathogen surrogates were selected among indigenous,
nonpathogenic microorganisms that represent the target Among possible microbial indicators of the removal of pro-
pathogens in removal through treatment. In undertaking tozoa in drinking water treatment plants, aerobic spores
this project, it was assumed that the detection of the could serve best because they meet most of the require-
surrogate should be quick, simple, and inexpensive. ments of a surrogate measure. Surrogate parameters used
The selected surrogate could then become a tool to in evaluating and improving water treatment plant perfor-
integrate source and finished water quality information mance should meet the following requirements: First, the
with treatment plant operations. The implementation of a organisms should be indigenous, nonpathogenic, and ubiq-
representative surrogate as a monitoring and operational uitous in an aquatic environment, especially in surface
tool should improve overall treatment process reliability waters. They should be occurring naturally at concentra-
and provide a sound basis for reducing the likelihood of tions greater than but corresponding to those of target
microbial pathogen breakthrough. The project findings pathogens. Next, their removal through treatment pro-
provided information on the source water occurrence of cesses should be at levels comparable with removal of
Giardia, Cryptosporidium, enteric viruses, total aerobic target pathogens. The surrogate parameters should be
bacterial spores, Bacillus spores, total anaerobic bacterial easy to analyze through techniques that are simple, quick,
spores, Clostridium spores, heterotrophic plate-count and economically feasible.
bacteria (HPC), Escherichia coli, fecal and total coliforms, Table 1 contains a list of candidate surrogates and their
bacterial phage, turbidity, and total particle count. The capability to meet these criteria (19–26).
corresponding source water concentrations of Giardia,
Cryptosporidium, and viruses were compared to these ANALYSIS OF WATER SAMPLES FOR AEROBIC SPORES
of other organisms to evaluate the potential of using
surrogates to estimate the pathogen occurrence. Removal The twentieth edition of the Standard Methods for the
of the potential surrogates through treatment was Examination of Water and Wastewater (27) does not
examined, and compared to the removal of Giardia, contain methodology for analysis of aerobic spores in
Cryptosporidium, and viruses to evaluate the potential of water. Analysis of total aerobic spores and their major
using surrogates to estimate treatment plant performance. representative, Bacillus, is based on methods adopted by
The effects of variables such as water source type, source Rice and coworkers (14) from the dairy industry (28). The
water quality, and seasonal changes on the effectiveness method has been further modified (18) to facilitate the
of using surrogates to assess source water occurrence of adoption of the method by drinking water laboratories. On

Table 1. Candidate Surrogate Parameters for Giardia and Cryptosporidium in Water

Natural Removal Analysis Turn-Around

Occurrence in Similar to Time, Simplicity, and Candidate
Potential Surrogate Surface Waters Pathogens Costs Surrogate

Turbidity Yes Site- specific Quick, simple, inexpensive Yes

Particles >2 µm Yes Yes? Quick, simple, relatively Yes
inexpensive
Streaming current Yes Treatment Quick, simple, relatively Used only as
potential specific inexpensive baseline
Total and fecal Yes No Quick, simple, inexpensive Used only as
coliform, and baseline
Escherichia coli
Heterotrophic Yes Treatment Quick, simple, inexpensive Yes
plate count specific
Micrococcus luteus Yes, very low ? Quick, simple, inexpensive No
Anaerobic bacterial Yes, very low ? Quick, complicated, inexpensive Yes
spores and
Clostridium
perfringens
Aerobic bacterial Yes Yes Quick, simple, inexpensive Yes
spores and
Bacillus subtilis
144 AEROBIC ENDOSPORES

the basis of the results of a study on methods comparison, An example of the distribution of pathogens and their
the incubation temperature and time needed to kill all potential surrogates in source waters, based on the number
vegetative cells and to induce the spores to germinate of samples given in Table 2, is presented in Table 3.
were set at 60 ° C for 15 minutes in a shaking water Typically, Giardia and Cryptosporidium were detected
bath (150 rpm), (following pretreatment in a water bath in surface water samples at concentrations ranging
at 37 ° C for 30 minutes). The germinated spores in water from 1 to 10 per liter. In this study, only one sample
samples should be filtered through 47-mm diameter, 0.45- revealed concentrations of Giardia in excess of 10 per liter.
µm pore-size membrane filters and plated on a nonselective Anaerobic spores were usually detected at concentrations
nutrient agar also containing soluble starch (0.1%) and ranging from 1 to 10,000 per liter (0.1 to 1,000/100 mL).
trypan blue dye (0.01 µg/L) and incubated for 21 hours Aerobic spores were found to be more abundant, with
at 35 ° C in a humid environment. The total number of typical concentrations ranging from 100 to 1,000,000 per
colonies on that agar represented the number of aerobic liter (10 to 100,000/100 mL). Heterotrophic bacteria were
spores present in the sample. The number of colonies that found in a wide range of high concentrations as well,
have cleared zones around them (indicating the presence that is, from 1,000 to 10,000,000 per liter (100 to
of amylase, an enzyme produced by Bacillus spores that 1,000,000/100 mL). The physical surrogates (turbidity
degrades starch) was a presumptive representation of the and particle count) vary significantly in various source
number of Bacillus spores present in the sample. waters.
It is very difficult, if not impossible, to establish a
relationship between the Giardia and Cryptosporidium
OCCURRENCE OF PATHOGENS, AEROBIC SPORES, AND measurements and any of their potential surrogates,
OTHER SURROGATES IN SURFACE WATERS including aerobic spores. Figure 1 presents an example
of the lack of correlation found between aerobic spores
An example of the relative concentration of Giardia and and the pathogens (17). Correlation coefficients (R2 ) for
Cryptosporidium and their microbial indicators, including occurrence among these pathogens to any of the intended
aerobic spores, as well as some nonmicrobial indicators, surrogates in the source waters were less than 0.2 and
is presented in Table 2, which was developed, based showed no statistical significance. These results support
on the source water analytical results collected over a recommendation that the search for a Giardia and
one year of sampling at 24 water treatment utilities Cryptosporidium surrogate in natural waters should
in the United States and Canada (17,18). Typically, be put on hold pending the development of more
the microbial content of surface source waters varies sensitive and accurate analytical methods for these
dramatically. Results of microbial analysis revealed pathogens.
that Cryptosporidium and Giardia are usually found in
the least number of samples, whereas aerobic spore- Factors Influencing Occurrence of Aerobic Spores in Surface
forming bacteria were found in all source water samples Waters
tested. With the frequently observed exception of coliform Several factors can influence the concentration of aerobic
bacteria and anaerobic spores, all the potential surrogate spores found in waters. Because aerobic spores are soil-
organisms occur more consistently in source waters than based organisms, their presence in water is associated
the pathogenic organisms, and at orders of magnitude with the presence of solids. Major factors responsible for
greater concentration than those of the pathogenic the presence of spores in water are water quality (turbidity
organisms. and particle concentrations), source water type (river

Table 2. Example of the Surface Water Microbial Quality Results (18)

Number of Percent Detection Median Twenty-Fifth Ninety-Fifth

Pathogen or Surrogate Samples in All Samples Concentr. Percentile Conc. Percentile Conc.

Giardia (/100L) 137 79 21 5 224

Cryptosporidium (/100L) 137 59 5 2 66
Enteric viruses (/100 L) 12 75 5 1 465
Aerobic spores (/100 mL) 242 100 1,820 528 25,280
Bacillus spores (/100 mL) 243 93 130 33 3,330
Anaerobic 160 94 10 5 120
spores (/100 mL)
Clostridium (/100 mL) 160 92 6 2 78
HPC (/100 mL) 211 99 15,000 3,800 462,200
E. coli (/100 mL) 143 64 4 1 129
T. coliform (/100 mL) 176 85 39 6 4,496
F. coliform (/100 mL) 149 65 4 1 214
PhiX174 phage (/100 mL) 62 71 8 1 2,414
MS2 phage (/100 mL) 62 56 3 1 689
Turbidity (NTU) 195 100 3 1 33
Particles >2 µm (/mL) 68 100 8,937 5,675 360,633
 AEROBIC ENDOSPORES 145

Table 3. Distribution of Pathogens and Their Surrogates in Source Waters (18)

Number of Observations
Range of Observed Anaerobic Aerobic
Number of Organisms per Liter Cryptosporidium Giardia Spores Spores HPC

1–10 79 105 14 1 0
10–100 0 1 52 8 0
100–1,000 0 0 73 19 1
1,000–10,000 0 0 10 58 16
10,000–100,000 0 0 0 116 65
100,000–1,000,000 0 0 0 38 72
1,000,000–10,000,000 0 0 0 0 45
10,000,000–100,000,000 0 0 0 0 5

a presence of particles that are soil-based, and not

an indicator of water contamination that would pose
a human health concern. They were found in sufficient
concentrations in many of the source waters to be used
as a direct indicator of treatment performance. Even for
those waters with lower concentrations, aerobic spores can
be used as relative indicators of treatment performance.

USING AEROBIC SPORES IN THE ASSESSMENT OF

TREATMENT PLANT EFFECTIVENESS IN REMOVING
Figure 1. Correlation between occurrence of aerobic spores and
PARTICULATES AND MICROBIAL CONTAMINANTS
parasitic protozoa in surface waters (18). See color insert.
The calculation of removal of microbial contaminants
through treatment, and thus the ability to directly assess
or flowing stream and lake or reservoir), and seasonal treatment plant performance, is controlled (or constrained)
changes. by the following factors:
A relationship between increasing aerobic spore
concentrations and turbidity is presented in Figure 2. 1. The source water concentrations of microorgan-
isms, or
PRESENCE OF AEROBIC SPORES IN TREATED DRINKING 2. The detection limits for microorganisms in finished
WATER water.

Aerobic spores can be found in treated drinking water It has been documented that a well-operated water
samples. The median value, calculated after sampling at treatment plant can reliably remove more than 3 log of
24 drinking water plant effluents for 12 months (17,18) microorganisms and particles 2 µm in size (1,8–10,16).
was 1,820 spores per 100 mL (the ninety-fifth percentile Therefore, it is necessary for the source water concentra-
was 25,280 spores per 100 mL). Aerobic spores were tion and the filter water detection limit to be separated
detected in 84% of the filtered water samples. Because by at least 3 orders of magnitude to accurately assess a
aerobic spores are not human pathogens, their presence 3-log removal. Since most treatment plants are capable
in treated drinking water samples is an indicator of of greater than 3-log removal, even this range in source
water occurrence and filtered water detection level limits
an accurate assessment of performance.
To illustrate the relationship between source and
finished water concentrations of Giardia, Cryptosporidium
and aerobic spores, data in Figure 3 are plotted to show
the frequency distribution of the organisms.
It can be noted that the range of concentrations of
organisms in source water is significantly wider than the
range of their observed concentration in filtered water.
It can also be observed that the range between the
median concentration of Giardia and Cryptosporidium
in filtered water and the source water spans between
0.001 per liter and 0.1 per liter. The respective difference of
median concentrations of aerobic spores in filtered and in
Figure 2. Comparison of aerobic and Bacillus spore occurrence source water ranges from 10 per liter to 100,000 per liter.
to source water turbidity (18). See color insert. This observation leads to an estimation of ‘‘a theoretically
146 AEROBIC ENDOSPORES

water was found to date due to the lack of accuracy

in defining occurrence of these pathogens. The best
relationship was found between turbidity and the
pathogens. Even this relationship is only indicative of
trends and not predictive in nature. All other comparisons
had no perceptible relationship.
No correlation between surrogates and pathogens for
removal could be derived in treated drinking water
samples because it is not possible to calculate true
pathogen removal. However, it is possible to use aerobic
spores and particles to evaluate treatment performance
based on direct removal calculations or on relative
measurements of filtered water concentrations.
Figure 3. Distribution of aerobic spores and protozoan cysts in
source and filtered water (18). See color insert. Benefits of Aerobic Spore Analyses to Water Treatment Plant
Operations

probable removal capacity’’ that could be documented at a Aerobic spores are sufficiently plentiful in the majority
water treatment plant. Using naturally occurring aerobic of source waters to allow for removal calculations that
spores detected in both the source and filtered water, will demonstrate from 3- to 4-log treatment efficiency.
up to 4-log removal can be documented at a majority of In contrast, in most source waters only 1- to 3-log
treatment plants. The removal of aerobic spores has been pathogen removal can be demonstrated. Future method
found to result from their physical removal through water development suggested that the sensitivity of the aerobic
treatment rather than inactivation (18). Inactivation spore analysis can be improved, and its use as a monitoring
of these organisms contributes little to their overall technique could be extended.
reduction in treated waters. In contrast, monitoring In addition to calculating removal, evaluating finished
of Giardia and Cryptosporidium would only allow for water microbial quality could be performed by aerobic
documentation of 2-log removal through treatment in a spore analysis. This type of microbial monitoring com-
typical plant. plements the physical measurements of turbidity and
Because of the limitations of the analytical procedures particles.
for Giardia and Cryptosporidium, no reasonable method Efficiency of drinking water treatment will continue
has been found to date to develop a surrogate to predict to be evaluated based on surrogates until a full range
treatment performance characterizing the removal of of pathogen analyses are developed that can easily
these pathogens. Thus, the search for a surrogate that test for all of the waterborne pathogens of concern.
correlates directly to pathogen treatment should be put on Even though an ideal pathogen surrogate has not been
hold pending the development of sensitive and accurate found, surrogates can and will continue to be needed to
analytical methods for the target organisms. assist in assessing treatment efficiency. Several studies
and drinking water utility practices have identified
Practical Application of Aerobic Spores in Drinking Water a number of factors that should be considered when
Treatment selecting surrogates as predictors of treatment efficiency
for pathogen removal.
Aerobic spores and particle counts can be used as
The most important issue when selecting a surrogate
surrogate measures to evaluate or improve treatment
is an understanding of the site-specific source water
plant performance. The spores present a more sensitive
characteristics and treatment impacts on the surrogates.
tool than particle counting as they maintain their size and
The surrogate parameter must be present in sufficient
shape, remaining intact through the treatment process.
concentration to allow for an assessment of treatment
In contrast, particle counting does not reflect changes
efficiency. The concentration of the surrogate will dictate
in the size of particles as they enter the plant, undergo
whether it can be used as a ‘‘direct indicator’’ of
coagulation into larger particles, and break apart when
performance (e.g., removal of >3-log aerobic spores) or
passing through the filters.
a ‘‘relative indicator’’ of performance (e.g., absolute value
Among known indicators, only aerobic spores and
such as a turbidity value of 0.1 NTU).
particles are found in sufficient quantities in the source
Finally, understanding the impact of treatment on
waters to provide a basis for assessing treatment plant
the surrogate concentration also is very important. For
performance, (i.e., removal through water treatment plant
example, concentration of HPC bacteria may increase in
processes). HPC bacteria can also be found in abundance
a filter and obscure the actual removal efficiency of the
but cannot normally be used as a surrogate to directly
process.
evaluate treatment efficiency because they multiply in
filters.
Using Aerobic Spores in Monitoring for Minimum Required
Removal of Pathogens. To be used as a direct indicator
Limitations to Aerobic Spores as Microbial Surrogates
of performance, the concentration of a surrogate in the
No ‘‘ideal’’ surrogate for predicting occurrence of protozoan source water must be greater than the detection limit in
pathogens (i.e., Giardia and Cryptosporidium) in source the filtered water by a factor equivalent to the removal to
 AEROBIC ENDOSPORES 147

be demonstrated and a safety factor. When this occurs, it a number of surrogates in the filtered water (both
is possible to monitor directly for the minimum required physical and biological), it is reasonable to conclude
log removal by the treatment process. that a specified level of treatment is being achieved.
For example, if aerobic spores are to be used as a For example, a turbidity of less than 0.1 NTU 95% of
direct indicator of performance for 3-log removal and the the time, particles greater than 2 µm below 50 particles
filtered water detection limit is 0.1 spores per 100 mL, the per milliliter, and aerobic spores below 2 per 100 mL may
source water concentration must be at least 100 spores reliably indicate absence of pathogens. The actual level
per 100 mL. A testing safety factor should be applied of treatment must be determined based on site-specific
to assure reasonable accuracy, typically a factor of 2 criteria and studies. Additional research in this area may
(or 200 spores/100 mL should be satisfactory). Of the lead to more universally applied treated water quality
24 utilities studied in the United States and Canada (18), parameters.
more than 75% can use aerobic spores as a direct indicator
of performance for 3-log removal. Over 50% of these Using Aerobic Spores in Challenge Studies. Challenge
utilities can use aerobic spores to demonstrate 4-log studies can be used to relate surrogate removal efficiency
removal. or a filtered water quality value to pathogen removal
Particles greater than 2 µm are present in sufficient performance in a treatment facility. This is the basic
quantities in a number of source waters to allow for approach that was used to produce the background
direct log removal determinations. More than 75% of data to support the establishment of a 0.3 NTU filter
the observed values were greater than 5,000 particles per effluent quality being equivalent to a minimum of a 2-log
milliliter. This allows for a 3-log removal calculation. Cryptosporidium removal.
Probably the most important issue associated with Challenge studies can be conducted at pilot-,
monitoring to demonstrate minimum required assurance demonstration-, or full-scale. The studies can use viable
of log removal is an understanding that log removal is not pathogens, inactivated pathogens, bacterial spores, or
appropriately assessed if the source water concentration is other proven nonmicrobial pathogen surrogates such as
insufficient to calculate the desired log removal. As greater microspheres. The selection of the appropriate constituent
physical removal is achieved during treatment and there to use for the challenge study is site-specific.
is a need to demonstrate better removal capabilities, it The premise for a challenge study is that the
will be more difficult to demonstrate direct removal. surrogate and challenge constituent will be present in
the source water at a concentration sufficient to measure
Monitoring of Treated Water Quality Through Monitoring the removal efficiency desired. Thus, seeding of the
of Aerobic Spores. Often, a surrogate is present in source challenge constituent at a concentration sufficient to
water at a concentration below that necessary to allow demonstrate the desired removal, and then comparing
for a direct determination of removal at a level that this to the surrogate treatment efficiency, can lead to the
actually is achievable by the treatment process. A less establishment of a treated water quality goal based on the
than adequate source water concentration of the intended surrogate measurement.
surrogate should not preclude its use as a parameter to When there is a sufficient concentration of the intended
assist in determining treatment plant performance. If this surrogate in the source water supply, the challenge study
occurs, the surrogate might be used as a relative indicator can directly relate the removal efficiency of the surrogate
of performance. to that of the pathogen. The treatment plant performance
For example, source water may have variable particle can then be determined based on the removal efficiency of
concentrations from 1,000 to greater than 10,000 particles the surrogate.
per milliliter. If the filtered water quality remains at 1 Where the concentration of the surrogate(s) in the
to 2 particles per milliliter, the actual log removal will source water is not sufficient, it will be necessary to
be calculated at approximately 3 to 4 log. However, both relate the pathogen treatment efficiency to a filter water
calculations, and especially the 3-log removal calculation, quality value, for example, aerobic spores. The result can
are constrained by the source water concentration or be a set performance value that will relate to the level
the filtered water detection limit. It is not reasonable of pathogen performance desired. For example, it could
to imply that the calculated log removals are the true be demonstrated that an aerobic spore concentration in
treatment efficiency of the facility. It is more reasonable the filtered water of less than 2 per 100 mL will assure a
to assign a given treatment efficiency for a given filtered 3-log removal of Cryptosporidium oocysts. Turbidity and
water quality. This approach is being used to assure a particle counts can be set in much the same way. The end
2-log Cryptosporidium removal in the U.S. EPA’s Interim result is the establishment of filtered water quality values
Enhanced Surface Water Treatment Rule. A 2-log removal for the surrogate(s) that relate to a desired pathogen
will be credited to any facility achieving a filtered water removal.
turbidity of 0.3 NTU 95% of the time. Pilot-scale or For those source waters with highly variable surrogate
demonstration-scale studies can be used to show that concentrations, it may be desirable to establish both a
an absolute value for a surrogate in the filtered water filtered water value and a removal value. When the
indicates a predetermined level of treatment efficiency for concentration in the source water is sufficient, a calculated
pathogen removal. removal can provide the desired treatment information.
A single surrogate measurement should not be As the source water concentration drops below the value
considered sufficient to assess treatment. By measuring necessary to calculate the removal directly, a finished
148 AEROBIC ENDOSPORES

water value may be sufficient to verify appropriate 3. K. R. Fox and D. A. Lytle, J. AWWA 88(9), 87–94 (1996).
treatment. In most cases, there will be a maximum set for 4. P. A. Roefer, J. T. Monscvitz, and D. J. Rexing, J. AWWA
the finished water value, regardless of the source water 88(9), 95–106 (1996).
concentration. 5. H. Solo-Gabriele and S. Neumeister, J. AWWA 88(9), 76–86
Each utility needs to determine which surrogates to (1996).
develop and how the surrogates will be evaluated to verify 6. J. L. Clancy, W. D. Gollnitz, and Z. Tabib, J. AWWA 86(5),
the treatment efficiency of the treatment plant. The goal 89–97 (1994).
is to establish a number of surrogate measures that will 7. N. E. McTigue, Proceedings of the AWWA Annual Conference,
assure appropriate treatment. Selection of the correct Anaheim, Calif., 1995.
surrogates and verification of the surrogate measures 8. N. E. McTigue, M. W. LeChevallier, and J. L. Clancy, Nation-
as indicators of pathogen treatment will accomplish al Assessment of Particle Removal by Filtration, AWWARF
this goal. Research Project Report, AWWARF, Denver, Colo., AWWARF/
AWWA, 1998.
9. G. A. Burlingame, M. J. Pickel, and J. T. Roman, J. AWWA
90(8), 57–69 (1998).
CONCLUSION
10. M. W. LeChevallier and W. D. Norton, J. AWWA 84(12),
54–60 (1992).
Because removal calculations and the ability to directly
11. D. W. Hendricks et al., Proceedings of the AWWA Water
assess water treatment plant performance is controlled
Quality Technology Conference, Denver, Colo., 1984.
(or constrained) by the concentration of pathogens
12. J. B. Rose, H. Darbin, and C. P. Gerba, Water Sci. Technol.
in the source water and the detection limits for
20(11,12), 271–276 (1988).
pathogens in finished water, it is difficult to accurately
13. D. W. Hendricks et al., Biological Particle Surrogates for
determine the rate of removal of parasitic protozoa
Filtration Performance Evaluation, AWWARF Research
such as Giardia and Cryptosporidium using the current
Project Report, AWWARF, Denver, Colo., AWWARF/AWWA,
analytical methods. Although sampling and processing 1998.
significantly larger sample volumes results in lowering
14. U.S. Environmental Protection Agency, Fed. Reg. 59(28),
method detection limits and allows for calculation 6332–6444 (1994).
of higher log removals, it is very expensive and
15. E. W. Rice et al., J. AWWA 88(9), 122–129 (1996).
time consuming and not practical from the analytical
16. E. C. Nieminski et al., J. AWWA 88(8), 70–80 (1996).
standpoint. Aerobic spores are sufficiently plentiful in
17. E. C. Nieminski, W. D. Bellamy, and L. R. Moss, J. AWWA
the majority of source waters to allow for removal
92(3), 67–78 (2000).
calculations that would demonstrate from 3- to 5-log
18. E. C. Nieminski and W. D. Bellamy, Application of Surro-
treatment efficiency. In contrast, only 1- to 3-log pathogen
gate Measures to Improve Treatment Plant Performance,
removal can be demonstrated based on natural occurrence
AWWARF Research Project Report, AWWARF, Denver, Colo.,
of pathogens and analytical sensitivity. Aerobic spores AWWARF/AWWA, 2000.
can be used as effective surrogate measures to improve
19. C. Chauret et al., J. AWWA 87(11), 76–84 (1995).
treatment plant performance. They cannot be considered,
20. G. F. Craun, P. S. Berger, and R. L. Calderon, J. AWWA
however, as direct surrogate measures of pathogen
89(3), 97–104 (1997).
removal.
21. W. A. M. Hijnen, F. A. P. Houtepen, W. M. H. van der Speld,
Because aerobic pores are nonpathogenic to humans
and D. van der Kooij, Proceedings of the 1997 International
and are not associated with fecal contamination, they Symposium on Waterborne Cryptosporidium, Newport Beach,
could be used in challenge studies, being applied in a slug Calif., 1997.
or at a step dose to water treatment plant unit processes 22. E. C. Nieminski, Proceedings of the AWWA Annual Confer-
and measured in the effluent to establish log removal. ence, Atlanta, Georgia, 1997.
The analysis of aerobic spores is simple (using plate-count 23. E. C. Nieminski, Proceedings 1997 International Symposium
agar), quick (less than one day), and inexpensive ($25 per on Waterborne Cryptosporidium, Newport Beach, Calif., 1997.
sample). 24. E. W. Rice et al., Proceedings AWWA Water Quality Technol-
Evaluating finished water microbial quality can be per- ogy Conference, San Francisco, Calif., 1994.
formed by aerobic spore analysis. This type of microbial 25. K. N. Scott et al., Proceedings of the 1997 International
monitoring complements the physical measurements of Symposium on Waterborne Cryptosporidium, Newport Beach,
turbidity and particles. The filtered water spore mon- Calif., 1997.
itoring could be used as plant performance verifica- 26. S. M. Teefy, Proceedings of the AWWA Annual Conference,
tion. Cincinnati, Ohio., 1991.
27. American Public Health Association, American Water Works
Association, and Water Environment Federation, Standard
BIBLIOGRAPHY Methods for the Examination of Water and Wastewater, 20th
ed., Washington, D.C., 1999.
1. W. D. Bellamy, J. L. Cleasby, G. S. Logsdon, and M. J. Allen, 28. G. H. Richardson, Standard Methods for the Examination
J. AWWA 85(12), 81 (1993). of Dairy Products, Tests for Groups of Microorganisms,
2. J. B. Rose, C. P. Gerba, and W. Jakubowski, Environ. Sci. American Public Health Association, Washington, D.C., 1986,
Technol. 25, 1393 (1991). pp. 194–198.
 AEROBIC RESPIRATION, PRINCIPLES OF 149

AEROBIC RESPIRATION, PRINCIPLES OF of the free energy available from the hydrolysis of ATP
is approximately −7.7 kcal/mol. (Standard free energy
JONATHAN HOSLER values (Go ) are calculated by setting the concentrations of
University of Mississippi both the products and the reactants at 1 M, and, according
Medical Center to convention, the larger the negative value of the Gibbs
Jackson, Mississippi energy change the greater the energy available.) Because
actively growing cells maintain a high concentration of
JAMES P. SHAPLEIGH
ATP relative to that of ADP, the actual free energy
Cornell University
(G) of ATP hydrolysis is greater (more negative) than
Ithaca, New York
the standard free energy value. For example, it may be
−10 kcal/mol. The biosynthesis of ATP requires the input
Aerobic respiration is by definition the transfer of electrons
from a donor compound through a series of carriers of an amount of energy equal to that of the actual free
to molecular oxygen (O2 ), which is reduced to water. energy of hydrolysis, G. Respiration can provide this
Molecular oxygen serves as the terminal acceptor of energy because the flow of electrons through the carriers
electrons or terminal oxidant in this process. An oxidant that make up the respiratory chain to the terminal oxidant
is a compound that brings about the oxidation of other is energetically favorable. One form of the Nernst equation
compounds and is therefore an acceptor of electrons. (Eq. 1) relates the amount of energy in electrons flowing
Likewise, a reductant is a compound that brings about the through the electron transport chain to the amount of
reduction of other compounds and is therefore an electron Gibbs free energy available to do work. In this equation,
donor. Reactions involving the transfer of electrons are G is the Gibbs energy change, n is the number of electrons
frequently termed redox reactions because they involve transferred, F is the Faraday constant (23.06 kcal/V/mol)
both oxidation and reduction reactions. An oxidation is and Em is the difference between the redox potentials of
the loss of electrons by an atom or molecule, whereas a the oxidation and reduction half-reactions of the reaction
reduction is the gain of electrons by an atom or molecule. of interest.
The use of O2 as a terminal electron acceptor differentiates G = −nFEm (1)
aerobic respiration from ‘‘anaerobic’’ respiration reactions
that use alternative oxidants such as nitrate, sulfate, or It can be seen from Equation (1) that, as long as
oxidized metals. Aerobic respiration is also distinct from mechanisms are available, cells can maximize energy
fermentation in that carbohydrates serve as the terminal production by transferring electrons from donors of low
electron acceptors during fermentation. This article will potential to terminal acceptors of high potential. The
focus on the mechanisms organisms have evolved to use O2 larger the overall voltage drop the greater the energy
as a terminal oxidant. A basic description of the structure available, similar to a battery. The redox potential of the
and function of the components involved in aerobic reduction of O2 to H2 O is +0.82 V, which is one of the
respiration will be included. The organization of these highest redox potentials of any biologically useful oxidant.
complexes into functional respiratory chains will then be This is one of the principal reasons O2 is preferred as a
discussed. Examples of aerobic respiratory chains from terminal oxidant. This high midpoint potential also makes
well-studied model systems, along with a brief description it possible to use a wide range of compounds as electron
of some of the mechanisms used by organisms to regulate donors while maintaining an energetically favorable Em .
the make-up of these chains, will also be provided. The The most common electron donor is the enzyme cofactor
primary focus will be on aerobic respiration in bacteria. NADH, which is produced by the reduction of NAD+ during
However, the information discussed here is applicable to many reactions in the cell involved in the oxidation of
respiration by higher organisms because the mechanisms reduced carbon compounds. The redox potential of the
used by all organisms to couple respiration with energy NAD+ /NADH couple is −0.32 V. Using equation (1), the
production are essentially the same. Aerobic bacteria use a transfer of two electrons from NADH to O2 yields a G of
much broader range of respiratory strategies than higher −53 kcal/mol, which is sufficient free energy to drive the
organisms, so it is useful to discuss these strategies to synthesis of several molecules of ATP.
help understand how bacteria can adapt to many different Peter Mitchell first elucidated the mechanism, by
physiological niches. which energetically favorable electron flow is coupled
with energy requiring biosynthesis of ATP; the Mitchell
theory of energy production and conservation is termed
BASIC FUNCTIONS OF AEROBIC RESPIRATION
chemiosmosis (1,2). In the now proven chemiosmotic
mechanism, the energy of electron transfer to O2 , via
Conservation of Energy
the respiratory electron transfer chain, is used to move a
The most important function of respiration is energy few negative and positive charges to opposite surfaces of
conservation via the formation of ATP. This is termed a membrane, either by pumping protons to the positive
oxidative phosphorylation because the oxidation of reduced surface or electrons to the negative surface (Fig. 1).
substrates leads to the phosphorylation of ADP to produce The protein components of the respiratory chain are
ATP. The hydrolysis of phosphate from ATP, which is part oriented with specific topology in the membrane to drive
of numerous reactions in the cell, releases a relatively such charge separation events (Fig. 2). Lipid bilayers are
large amount of Gibbs free energy (G), a measure of relatively impermeable to charged species; this allows the
the energy available for work. The ‘‘standard’’ value (Go ) transmembrane charge separation to persist as long as
150 AEROBIC RESPIRATION, PRINCIPLES OF

Electrical electron transfer continues. The charge separations create

H+
energy a transmembrane voltage gradient and this gradient
+ + attracts positively charged ions to the negative surface of
+ the membrane. In fact, it is the flow of protons through the
− − ATP synthase in the bacterial membrane, from the positive
e− −
H+ O2 surface to the negative surface, which provides the energy
+ − ADP + Pi for ATP synthesis (this is discussed in more detail later).
− + Because only protons move through the ATP synthase,
H+ H+
Chemical the transmembrane voltage gradient is also termed the
ATP proton motive force (PMF). A more detailed explanation of
energy
−
Electrical to − chemiosmosis can be found in (3).
mechanical to − +
Those components of the respiratory electron trans-
chemical bond fer chain that use some of the energy of electron flow to
energy + + generate PMF via charge separations are termed electro-
Figure 1. Schematic representation of the chemiosmotic circuit genic protein complexes. The respiratory electron transfer
in a bacterial cell. The membrane bilayer is shown with the chain in all mitochondria contains three electrogenic
positive (+) and negative (−) charges that are closely associated complexes (4). The situation in bacteria, however, is con-
with each face of the membrane during aerobic respiration. The siderably more complicated. Some bacteria modulate the
redox energy of electron flow to oxygen within a terminal oxidase number of electrogenic complexes in their electron trans-
is used to pump protons through the oxidase to the outer face of port chain depending on the environmental conditions as
the cytoplasmic membrane, thus establishing the separation of well as their physiological requirements. Moreover, many
charges that is termed the transmembrane voltage gradient, or
bacteria synthesize fewer electrogenic complexes than do
.  provides the driving force for proton flow through the
mitochondria, although they use the same reductants and
F0 portion of the F1 F0 -type ATP synthase, where the proton flow
initiates a carousel-like rotation of a ring of proteins within the terminal oxidants.
membrane. This rotary motion leads to the rotation of a protein Aerobic bacteria can also use a much wider variety of
spindle within the F1 portion of the ATP synthase. Rotation of this compounds as electron donors than higher organisms. In
spindle forces the active sites in F1 that bind ADP and phosphate addition to carbon containing compounds, many inorganic
(Pi) to change shape, with the result being the formation of compounds can serve as sources of electrons. Bacteria
a chemical bond between ADP and Pi and the release of the that can use inorganic compounds as sources of energy
product, ATP.

2H2O
2H2O
H+ H+ Cytochrome c H+ H+
H+
Fe-S 00 P
QH2 QH2
0 Q QH2 0
Q Q Q
9 Fe-S
3 Fe-S
N
FM

FAD N
H+
O2 H+ H+ O2 H+
H+
Succinate Fumarate
ADP + Pi ATP
NADH NAD+ + H+

IV I III II IV
Heme-copper NADH bc1 complex Succinate Cytochrome c F1Fo
quinol oxidase dehydrogenase dehydrogenase oxidase ATP synthase
Figure 2. Representation of some of the protein complexes involved in aerobic respiration and the
reactions they catalyze. The name of each complex is indicated below the complex along with its
shorthand designation as a Roman numeral. Solid ovals represent heme centers and open circles
represent copper. P and N represent the positive and negative surfaces of the membrane. Fe−S
indicates complexes containing iron-sulfur centers and FAD and FMN represent flavin moieties.
Dashed arrows within the membrane indicate the redox cycling of the quinone pool that occurs
as NADH dehydrogenase and succinate dehydrogenase reduce quinone (Q), while the bc1 complex
and quinol oxidases oxidize quinol (QH2 ). Redox cycling of the cytochrome c pool in the periplasmic
space of the cell also occurs as cytochrome c is reduced by the bc1 complex and oxidized by
cytochrome c oxidases. The ability of four of the five electron transfer complexes shown to function
as proton pumps is indicated by the entry and exit of H+ from each complex.
 AEROBIC RESPIRATION, PRINCIPLES OF 151

are known as lithotrophs. For example, there are bacteria Consume Reducing Equivalents
termed nitrifiers that use nitrite as their energy source
Another possible use for electron transfer to O2 is to
by oxidizing it to nitrate. The relatively high midpoint
replenish the supply of NAD+ . Bacteria growing on highly
potential of this reaction (+0.41 V) decreases the Eh of reduced carbon sources, such as fatty acids or aliphatic
electron transfer to O2 , as compared with the oxidation hydrocarbons, produce NADH at such a rate that the
of NADH, thus reducing the amount of the Gibbs free supply of NAD+ may become limiting for the metabolism
energy that can be recovered (Eq. 1). Consequently, these of the cell. Under such conditions, a cell might modify
bacteria must oxidize more reductant than is oxidized its respiratory chain to ensure that NADH oxidation can
by bacteria that use NADH to produce an equivalent keep pace with NAD+ reduction. Some bacteria have been
amount of ATP. The benefit of this strategy is that the shown to reduce nitrate to nitrite in order to oxidize
development of an aerobic respiratory chain that oxidizes the NADH pool (7). Although never directly shown, it
inorganic compounds allows lithotrophs to occupy a unique seems likely that O2 also serves as a sink for excess
environmental niche. reducing equivalents under certain conditions and may
The high concentration of O2 in the atmosphere, along help explain the diversity of aerobic respiratory chains in
with the energetic efficiency of the aerobic respiratory many aerobes.
chain, makes aerobic respiration the preferred mode of
energy conservation for those organisms that can switch COMPONENTS INVOLVED IN AEROBIC RESPIRATION
between aerobic and anaerobic modes of growth. Other AND THEIR FUNCTION
modes of energy conservation, such as fermentation, are
not as efficient in the production of ATP. In addition, Electron Transfer and Charge Separation
fermentation and anaerobic respiration may result in
The process of oxidative phosphorylation can be viewed as
the buildup of products that can be toxic at modest a series of energy transformations (Fig. 1). The function
concentrations. Aerobic respiration by most organisms of the components of the electron transport chain is to
generates CO2 and H2 O as its end products, neither of take high-energy electrons (those with a more negative
which is particularly cytotoxic. Accordingly, the regulatory redox potential) and transfer them stepwise to O2 , which
circuitry of aerobic bacteria is designed to preferentially is reduced to water. Proper arrangement of the respiratory
express those genes required for O2 respiration at the components allows some of the redox energy released in
expense of alternative modes of energy conservation the stepwise transfer of electrons to O2 to drive charge
whenever O2 is present (5). separation events across the cytoplasmic membrane
(Figs. 1 and 2). Charge separation is accomplished by
moving a positive charge (a proton) to the outer, positively
Detoxification charged surface of the membrane or by moving a negative
charge (an electron) to the cytoplasmic, negatively charged
Aerobic respiration can fulfill other physiological functions
surface of the membrane. Thus, a transmembrane voltage
besides ATP production. One of these alternative functions
gradient is created and maintained, which is another
is to protect O2 -sensitive components of the cell. For
form of electrical energy (Fig. 1). The voltage gradient
example, the bacterium Azotobacter vinelandii uses the
is used to pull protons through the membrane-bound
O2 -sensitive enzyme nitrogenase to fix nitrogen even ATP synthase, which is a nanomachine that converts the
under aerobic conditions. Azotobacter minimizes O2 energy of the voltage gradient into mechanical energy. As
damage to nitrogenase by maintaining high rates of the ATP synthase couples phosphate to ADP and forms
aerobic respiration (6). This results in low intracellular ATP, mechanical energy is transformed into chemical bond
concentrations of O2 . Because these bacteria want to energy. A good deal of Gibb’s free energy is made available
maximize O2 consumption, they must also maximize upon the hydrolysis of ATP to ADP and phosphate; for
electron flow through the respiratory chain. This can this reason the hydrolysis of ATP is incorporated into the
be accomplished by modifying the components of the mechanisms of numerous reactions in the cell to make
respiratory chain. By definition, the processes of electron them thermodynamically favorable.
transfer and charge separation are linked, or coupled, in
the electrogenic complexes of respiratory electron transfer. Description of Basic Components of the Respiratory Chain
As a result of this situation, it becomes energetically
For aerobes that use carbon compounds as electron donors,
more difficult to pump additional charges (protons or the flow of electrons during respiration centers on a pool
electrons) against the transmembrane voltage gradient of lipid-soluble quinone (Q); each quinone molecule is a
as the magnitude of the gradient increases. This results two-electron (and two-proton) carrier. In these bacteria,
in a slowing of electron transport. However, bacteria can all of the possible electron transfer pathways of aerobic
adjust the components of the respiratory chain to include respiration can be described by defining which membrane-
complexes that are less efficient at coupling electron flow bound electron transfer proteins are adding electrons to
to the generation of a voltage gradient. This makes it the Q-pool and which protein complexes are removing
possible to maintain higher rates of electron transfer to O2 . electrons from the Q-pool (Fig. 2). The chemical structure
As discussed later, Azotobacter uses a particular terminal of Q varies somewhat throughout the eubacteria and
oxidase that is less efficient at coupling electron flow to the archea, but all contain a simple quinone headgroup
the generation of PMF. coupled to a long hydrocarbon tail that restricts the
152 AEROBIC RESPIRATION, PRINCIPLES OF

(a) O OH Eectrons enter the Q-pool in two ways: via the oxidation
of NADH or via the direct oxidation of reduced metabolic
CH3O CH3 2e− + 2H+ CH3O CH3 intermediates, such as succinate. NAD+ is a small, water-
CH3O CH3O soluble two-electron carrier that is reduced to NADH in
numerous cellular oxidations of reduced carbon substrates
O OH CH3
CH3 n n (see Table 1 for listing of respiratory complexes). When
NADH is the source of electrons for respiration, it
(b) O OH is oxidized by a NADH dehydrogenase. Many aerobic
bacteria use a NADH dehydrogenase that is very similar to
CH3 2e− + 2H+ CH3 the mitochondrial enzyme (4). This enzyme is the largest
of the membrane-bound electron transfer complexes of
aerobic respiration, being composed of 13 to 14 separate
O OH
CH3 n CH3 n
protein subunits (8–11). NADH binds at a site on the
internal extramembrane domain of the complex, where
Figure 3. Structure of common quinones found in E. coli in its electrons are transferred to a flavin prosthetic group
their reduced and oxidized forms. Ubiquinol is shown in (a) (Fig. 2). Flavins are two-electron, two-proton carriers
and menaquinone is shown in (b). The length of the hydrophobic derived from riboflavin; they are always bound within
side chain (n) is variable but is typically 8 in E. coli.
proteins. The electrons then travel through a series of
electron carriers known as iron-sulfur centers ([Fe−S])
quinone to the lipid bilayer (Fig. 3). Fully oxidized quinone spaced throughout the protein complex, and finally to
and fully reduced quinol diffuse rapidly in the bacterial a quinone that is transiently bound at a site accessible
membrane to shuttle electrons between the membrane- to the lipid bilayer. [Fe−S] centers are complexes of
bound electron transfer complexes. During the actual non-heme iron and acid-labile sulfur bound to particular
reduction or oxidation of each quinone, its headgroup binds residues in the protein complex (Fig. 4). These [Fe−S]
to a specific site on the electron transfer complex. In this clusters are of variable iron and sulfur stoichiometry and
manner, the one-electron reduced quinone intermediates, variable redox potential but are all one-electron carriers.
which may be negatively charged, are sequestered from The quinone reduced by the NADH dehydrogenase enters
the hydrophobic lipid bilayer. the lipid-soluble Q-pool. NADH dehydrogenase is one of

Table 1. Occurrence and Molecular Characteristics of the Protein Complexes of Aerobic Respiration in Representative
Bacteria and Mitochondria

Number Presence in:

Respiratory of Complex Prosthetic Escherichia Paracoccus Rhodobacter
Complex Types Subunits Size (kDa) Groups coli denitrificans sphaeroides Mitochondria

NADH NDH I; nuo 13–14 ∼500–600 FMN, [Fe−S],Q + + + +a

dehydrogenase
(Complex I)
NDH II; ndh 1 ∼50 FAD + ? + −
Succinate 4 ∼115 FAD, [Fe−S], + + + +
dehydrogenase heme B
(Complex II)
bc1 complex 4 ∼110 hemes B and C, − + + +a
(Complex III) [Fe−S]
Heme-copper cbb3 -type 3 ∼120 hemes B and C, − + + −
terminal cytochrome c Cu
oxidases oxidase
(Complex IV)
aa3 -type 4 ∼120 hemes A, O, − + + +a
cytochrome c ±Cb , Cu
oxidase
quinol oxidase 4 ∼120 hemes A, B, and + + + −
Ob
bd-type quinol bd-type 2 ∼100 hemes B and D + ? − −
oxidases
(Complex IV)
bb-type 2 ∼100 heme B − ? + −
F1 F0 ATP synthase 20–23 ∼500 none + + + +a
a
The mitochondrial enzyme has a greater number of subunits.
b
Different hemes are present in various versions of these oxidases.
 AEROBIC RESPIRATION, PRINCIPLES OF 153

(a) type of terminal oxidase that uses the electrons donated

Cys
by successive cytochrome c molecules to reduce O2 to
S water.
Cys [Heme is composed of a conjugated ring of four pyrrole
Fe
Fe groups, termed porphyrin, with a central iron atom bound
Cys by one nitrogen from each pyrrole (Fig. 5). Hemes can be
Cys S used as one-electron carriers; proteins that bind redox-
active heme are termed cytochromes. Bacteria modify
substituent groups of the porphyrin ring to produce at
(b) least four chemically distinct forms of heme, hemes B, O,
S Cys
Fe
A, or D. More than one type of heme may be present in
Cys Fe multi-heme cytochromes.]
Fe Electron transfer from reduced quinol to cytochrome c
S
S in the bc1 complex occurs through a complicated dance that
has been termed the Q-loop (18,20,21). The bc1 complex is
S Cys
Fe composed of three to four subunits (Table 1). One subunit
contains two quinone-binding sites and two b-type hemes
that are oriented perpendicular to the plane of the bilayer
Cys and span the membrane. A second subunit contains a
c-type heme, and a third contains an [Fe−S] center in
an extramembrane domain that moves back and forth
Figure 4. Examples of a [2Fe−2S] cluster (a) and a [4Fe−4S]
on a flexible tether between the quinol oxidation site
cluster (b). The S represents acid-labile sulfur. One of the sulfur
and the c-type heme (19,21). Some of the redox energy
ligands for each iron (Fe) comes from a cysteine residue in the
protein. released by electron transfer from quinol to cytochrome
c is used to push electrons through the b-type hemes
from the positive surface of the cytoplasmic membrane
the three-electrogenic complexes that are used by aerobes, (outside) to the negative surface (inside). Thus, electron
meaning that the enzyme contributes to the maintenance transfer through the bc1 complex is electrogenic, that is, it
of the transmembrane voltage gradient. It does this by contributes to the transmembrane voltage gradient.
using a portion of the redox energy released during The third electrogenic component, present in all of
electron transfer from NADH to Q to pump protons from the possible electron transfer pathways to O2 , is the
the negative surface of the cytoplasmic membrane to the terminal oxidase (4). Although the terminal oxidases
positive surface. The mechanism of this proton pumping of bacteria show the greatest level of diversity of
process is not fully understood (12,13). any of the respiratory components (discussed later),
Succinate dehydrogenase is a membrane-bound protein basic structural and mechanistic similarities are shared
complex of three to four subunits that also transfers by all of these enzymes (Fig. 6). All of the terminal
electrons from succinate to the Q-pool (Table 1, Fig. 2). oxidases are composed of two to four integral membrane
This protein can be an important part of the respiratory proteins (16,22) (Table 1). Electrons from quinol or from
chain of some aerobes if the cultures contain succinate reduced cytochrome c are passed by various means to a
or significant concentrations of tricarboxylic acid cycle six-coordinate heme in the largest subunit of the complex.
intermediates. Like NADH dehydrogenase, the electrons This heme transfers electrons, one at a time, to a nearby
flow through a flavin cofactor and a series of [Fe−S] centers five-coordinate heme in the same subunit. Being five-
to reach a lipid-accessible quinone-binding site (4,14,15). coordinate, as opposed to the fully occupied six-coordinate
The energy released during electron transfer from geometry, allows one of the ligand binding sites of this
succinate to Q falls short of that required to drive a charge heme’s iron to serve as the O2 binding site (23,24). Once
separation event across the cytoplasmic membrane. Thus, O2 binds, it is reduced by four successive electrons to two
the reaction catalyzed by succinate dehydrogenase is not water molecules (25,26). The O2 reduction site, or active
electrogenic. site, of the oxidases is buried within the transmembrane
Electrons flow out of the Q-pool via two paths, both of region of the large subunit. The four electrons required
which terminate with the reduction of O2 to water through to reduce O2 follow a path from the outside to the buried
the activity of a terminal oxidase (Fig. 2). The first path active site (27,28); this path carries the electrons part of
requires a single membrane-protein complex that oxidizes the way across the membrane toward its negative surface.
quinol and reduces O2 ; this type of terminal oxidase is Simultaneously, the four protons required to convert O2
referred to as a quinol oxidase (16,17). The second path to 2H2 O are transferred from the cytoplasm to the buried
from Q to O2 , commonly referred to as the cytochrome c active site; this path carries the protons part of the way
pathway, involves two electron transfer complexes in the toward the positive surface of the membrane (29–32).
bacterial membrane. In the two-step pathway, quinol is Because these electrons and protons combine at the active
oxidized by the bc1 complex (18,19), which then reduces site to form H2 O, a neutral molecule, the partial electron
a mobile cytochrome c, a small electron transfer protein and proton transfers add up to one charge separation
containing a single c-type heme (Fig. 2). Cytochrome c then event (i.e., the movement of one charge completely across
shuttles its electron to a cytochrome c oxidase, another the membrane) for each electron transferred to O2 . All
154 AEROBIC RESPIRATION, PRINCIPLES OF

COO− −OOC

COO− −OOC
CH2 CH2
CH2 CH2 CH2 CH2
CH2 CH2
CH3 CH3
CH3 CH3 N N
N N Fe
Fe CH2 N N
CH2 N N CH CH3
CH CH3 OH
CH3 CH CH3
CH3 CH CH2 CH2 CH2 CH C CH3 H
3

heme B heme O

COO− −OOC COO− −OOC

CH2 CH2 CH2 CH2

CH2 CH2 OH CH2 CH2
O
CH3 C CH3 CH3
N N H N N
OH
Fe Fe
CH2 CH2
N N N N
CH CH3 CH CH3
OH
CH3 CH CH3 CH3 CH CH2
CH2 CH2 CH2 CH C CH3 H
3

heme A heme D

Figure 5. The different structures of the hemes found in common cytochromes of aerobic bacteria.

terminal oxidases are electrogenic in this fashion. The In addition to the proton-pumping NADH dehydroge-
redox energy available in the oxidation of either quinol nase, numerous eubacteria synthesize a simpler enzyme
or cytochrome c by O2 is substantial, however, and most that oxidizes NADH and reduces Q (36). This simpler
terminal oxidases have evolved a second mechanism to dehydrogenase consists of a single integral membrane pro-
convert more of this redox energy into the energy of the tein containing a flavin prosthetic group; it does not use
transmembrane voltage gradient. This is accomplished the energy of its electron transfer reaction to contribute to
by pumping one proton, through the protein, from the the transmembrane voltage gradient (Table 1).
negative to the positive surface of the membrane, for each The greatest degree of structural diversity is found
electron transferred to O2 (29). This proton is in addition in the terminal oxidases. Terminal oxidases are first
to the charge separation achieved through the chemistry of divided into two groups on the basis of the structures
O2 reduction. It is not clear how the proton is transferred of their active sites, in which O2 is reduced to water.
through the protein, but its transfer is obligatorily linked The first of these groups is composed of a family of
to electron flow through the oxidase complex (33,34). quinol oxidases, the prototype of which is the bd-type
quinol oxidase (Fig. 6). The bd-designation derives from
the types of hemes associated with the complex. The bd-
Structural Diversity of the Essential Respiratory Complexes
type oxidases are composed of two integral membrane
The degree of structural diversity, as far as is currently subunits, one of which contains three hemes (26) (Table 1).
known, varies considerably for each of the principal protein Electrons from quinol are first transferred to a heme
complexes. The structures and mechanisms of the bc1 b, and then to the active site, which consists of a
complex, succinate dehydrogenase, and the ATP synthase second b-type heme and a closely associated d-type
appear to be largely similar throughout the eubacteria and heme. This two-heme center forms the active site. The
the archaea. Perhaps the greatest diversion known to date bd-type oxidases have a high affinity for O2 and thus
is the ability of certain marine bacteria, for example, Vibrio allow respiration to continue under low ambient O2
alginolyticus, to synthesize a NADH dehydrogenase, a concentrations (37). This high affinity for O2 allows some
terminal oxidase, and an ATP synthase that pump and bacteria to use this oxidase for detoxification. For example,
use Na+ rather than protons (35). Azotobacter species use rapid electron flow through a
 AEROBIC RESPIRATION, PRINCIPLES OF 155

H2O A wide degree of structural variation is found within the

H 2O H 2O heme-copper oxidase superfamily (39,40). Some of these
enzymes oxidize quinol, and thus bypass the bc1 complex,
c 00 Q
cc
whereas others oxidize cytochrome c. The hemes that are
present vary considerably. In different enzymes, heme B,
0 0 0 O, or A (Fig. 5) may be present in the heme-CuB active site,
b b aa a, a, whereas heme B or A may be present as the six-coordinate
b o heme that transfers electrons to the active site. There
are no obvious mechanistic differences associated with the
O2 O2 O2 type of heme found in different heme-copper oxidases.
The oxidases of the heme-copper superfamily can be
cbb3-type aa3-type ba3-, bo3, and aa3-type
cytochrome c oxidase cytochrome c oxidase quinol oxidase
further divided into two groups. One group contains
three core subunits that have obvious structural homology
H2O H2O between species. In this group, electrons from quinol or
cytochrome c enter through a protein designated subunit
00 II; as in all of the heme-copper oxidases the six-coordinate
c
heme and the heme-CuB active site are located in the
Q subunit designated subunit I (39). Those enzymes of this
0
group that oxidize cytochrome c contain an additional
a d b b
a copper center in subunit II, termed CuA , close to the
cytochrome c binding site. The quinol oxidases of this group
appear to have evolved from the cytochrome c oxidases by
O2 O2 losing the residues that bind CuA and cytochrome c and
caa3-, cao-type bd-type by gaining a quinone-binding site (40). One subset of this
cytochrome c oxidase quinol oxidase first group, the heme aa3 -type cytochrome c oxidases of the
Figure 6. Schematic representation of the structural organiza- α subgroup of the proteobacteria (including Rhodobacter
tion of the different type of bacterial terminal oxidases. Each sp. and Paracoccus sp.), are the progenitors of the aa3 -type
complex is drawn such that the top of the figure would face the cytochrome c oxidase of mitochondria (40).
periplasmic side of the cytoplasmic membrane and the bottom The second group of the heme-copper oxidases consists
would face the cytoplasmic side of the membrane. Closed ovals of the more ‘‘primitive’’ cbb3 -type (or FixN) cytochrome
represent hemes and open circles represent copper. Q indicates c oxidases (41) (Table 1). Only the largest subunit of
the likely quinone-binding site in the quinol oxidases. The letters this enzyme, containing the active site, is structurally
below the hemes indicate the possible hemes found in a particular
homologous to the first group of heme-copper oxidases.
type of oxidase. The subscript 3 is used to indicate those hemes
Although this enzyme oxidizes cytochrome c, it does
that bind oxygen. This form of designation was developed to dif-
ferentiate between the two chemically identical, but functionally not contain CuA (42). The cbb3 -type oxidase has a
distinct, a hemes in the aa3 type cytochrome c oxidase. As addi- higher affinity for O2 than the other heme-copper
tional variants of the heme- copper oxidase have been found the oxidases, possibly because this enzyme arose when the
subscript 3 designation has been used primarily in those com- O2 concentration in earth’s atmosphere was low (41,43).
plexes that have chemically identical, but functionally distinct
hemes. The catalytic site is indicated by the arrows indicating
Diversity of Respiratory Chains
where O2 is bound and water is produced.
Not surprisingly, given the diversity of respiratory
enzymes, aerobes also exhibit a diversity of respiratory
bd-type oxidase to maintain the low intracellular O2 chains. Not only is there diversity between species, but
concentrations required for nitrogenase activity (38). As because most aerobes are capable of synthesizing multiple
discussed earlier. It appears that the bd-type oxidases do terminal oxidases, the components of the respiratory
not pump protons, although their O2 reduction mechanism chain of a single bacterium often changes in response
is electrogenic. to environmental conditions. Only a few well-studied
The second, and more diverse, group of terminal oxi- examples of aerobic respiratory chains will be discussed.
dases is that of the heme-copper oxidase superfamily (39) It should be emphasized, however, that the diversity of
(Table 1). In these enzymes, electrons are transferred from aerobic respiratory chains is a hallmark of bacteria. The
a six-coordinate heme to an active site composed of a presence of multiple respiratory complexes having similar
five-coordinate heme, where O2 binds, and a closely asso- functions allows electrons that are being transferred from
ciated copper atom, termed CuB (Fig. 6). It appears that reductant to oxidant to potentially flow through more
all of the heme-copper oxidases are capable of pumping than one path (Fig. 7). This ‘‘branched chain’’ structure of
protons. Because the heme-copper oxidases consume one respiratory chains is common in bacteria. There are few
cytoplasmic proton per electron as part of the O2 reduction examples of branched chain respiratory chains in higher
mechanism, as do all terminal oxidases, their additional organisms.
ability to pump one proton per electron (as described ear- In some sense, the Escherichia coli aerobic respiratory
lier) makes them twice as efficient at energy conservation chain is relatively simple (Table 1, Fig. 7). Escherichia coli
as the bd-type oxidases. encodes both a mitochondrial-type NADH dehydrogenase
156 AEROBIC RESPIRATION, PRINCIPLES OF

and a single subunit, nonelectrogenic NADH dehydroge- there is a quinol oxidase of the heme-copper family that
nase (44,45). Electrons from the dehydrogenase flow to can directly oxidize quinol. This oxidase is designated the
one of the three types of quinones, ubiquinone, or either ba3 -type oxidase because of its heme content (50). Unlike
of two napthoquinones, menaquinone, or demethylme- E. coli, electrons can also flow from reduced quinol to the
naquinone (46). The quinones transfer electrons to either a bc1 complex. Electrons then flow from the bc1 complex
bo-type heme-copper oxidase or a bd-type oxidase (16). The to one of the several c-type cytochromes that can either
regulation of the expression of the various components of be located in the periplasm or are membrane-bound (49).
the E. coli chain is complex and not completely understood. Paracoccus denitrificans has two terminal oxidases that
In general, ubiquinone and the bo-type oxidase are prefer- use reduced cytochrome c as a source of electrons. One of
entially expressed if O2 levels are high (47). Menaquinone these oxidases is the aa3 -type oxidase, which has signifi-
and the bd-type oxidase are more highly expressed under cant sequence similarity with the mitochondrial aa3 -type
limiting O2 concentrations (47). Regulation of the state- oxidase, consistent with the proposal that the bacterial
ment of the NADH dehydrogenases is not so clear-cut (48). progenitor of the mitochondria was closely related to
The single subunit enzyme is apparently expressed aero- P. denitrificans (23,51). The other oxidase is the cbb3 -
bically and repressed during fermentation. Expression of type oxidase (52). The aa3 -type oxidase is preferentially
the electrogenic enzyme is apparently stimulated by O2 expressed when O2 concentrations are high. The cbb3 -type
and nitrate. oxidase, consistent with it being a high-affinity oxidase,
The aerobic respiratory chain of Paracoccus den- is more highly expressed when O2 concentrations are
itrificans is similar to the mitochondrial respiratory limiting (53). The expression of the ba3 -type quinol oxi-
chain and both more complicated than the E. coli chain dase is more complex and its role in the physiology of
(Table 1, Fig. 7). Paracoccus denitrificans apparently P. denitrificans is unclear (54).
only encodes the mitochondrial-type NADH dehydroge- In addition to multiple terminal oxidases, some
nase (49). Ubiquinone is the major quinone. As in E. coli, bacteria also have the capacity to use electrons from

(a) H+

NADHdh H+

bo -type
H+ oxidase
NADHdh H+
Q

bd -type
Succ.dh oxidase

(b) H+
H+
bo3-type H+
NADHdh oxidase
aa3-type
H+ oxidase
H+ Q
H+ H+
H+
Succ.dh
bc1 cyt.c
complex
cbb3-type
H+ oxidase

H+

(c) H+

Nitrite cyt.c aa3-type

oxidase oxidase

H+

Figure 7. Representation of the aerobic respiratory chain of (a) E. coli, (b) P. denitrificans and
(c) Nitrobacter species. Abbreviations are: Q, quinone pool, cyt. c, soluble cytochrome c, NADHdh,
NADH dehydrogenase, Succ.dh, succinate dehydrogenase. Arrows indicate the direction of electron
flow, and H+ entering and exiting the boxes indicate electron transfer complexes that pump
protons.
 AEROBIC RESPIRATION, PRINCIPLES OF 157

multiple substrates during aerobic growth. For example, R. sphaeroides encounters conditions in the environment
some, including P. denitrificans, can use thiosulfate as in which the heme-copper quinol oxidase is functional,
an energy source (55). Others use the single carbon although it is not yet obvious what these conditions would
compounds methanol and methylamine as both a source be. Another example is provided by E. coli, which con-
of carbon and high-energy electrons (49,56). Oxidation of tains a set of genes that encode an alternative bd-type
thiosulfate involves a periplasmic thiosufate-cytochrome c oxidase (61). The role of this oxidase is not clear, but it
oxidoreductase, a periplasmic cytochrome c, and the aa3 - has been shown to be functional (62). It is likely that
type terminal oxidase. Similarly, the oxidation of methanol as more genomic sequences become available they will
or methylamine requires periplasmic dehydrogenases that show that the true diversity of aerobic respiratory chains
feed electrons to the aa3 -type oxidase via periplasmic has been underestimated because standard laboratory cul-
cytochrome c’s. In these pathways, the only electrogenic ture conditions do not induce expression of the full set of
component contributing to the PMF is the terminal respiratory enzymes.
oxidase, because electron flow through the NADH
dehydrogenase, the Q-pool and the bc1 complex is The F1 FO ATP Synthase
bypassed. However, the oxidation of formaldehyde, one of
The protein complex that allows the electrochemical
the end products of methanol and methylamine oxidation,
gradient to be coupled to ATP synthesis is termed the
produces NADH (49,56). Because the oxidation of this
F1 FO ATP synthase. Recent structural and mutagenesis
NADH involves the entire respiratory chain, cells growing
studies have shown that the ATP synthase of the bacterial
on methanol and methylamine still synthesize the NADH
membrane is actually a molecular motor (63–66). The
dehydrogenase and the bc1 complex.
ATP synthase complex is composed of two domains: a
The lithotroph Nitrobacter provides another example
transmembrane domain that contains a ring of small
of the versatility of bacterial aerobic respiration (Fig. 7).
protein subunits and an extramembrane domain that
Nitrobacter obtains electrons from the oxidation of nitrite
is shaped like a mushroom (Table 1, Fig. 8). The stalk
to nitrate; the redox potential of the nitrite/nitrate couple
of the extramembrane domain consists of one elongated
is about +0.4 V. This is too high a potential for the direct
subunit (γ ), which extends through the cap of the complex.
reduction of cytochrome c, as occurs in the examples given
The cap is a ring of six large subunits (3α and 3β); an active
earlier. Thus, Nitrobacter places a nitrite-cytochrome
site where ATP is synthesized from ADP and phosphate
c oxidoreductase within the cytoplasmic membrane in
is present in each of the three β-subunits. The proposed
which it can take advantage of the PMF to help drive
mechanism of this machine is as follows. Protons are
the transfer of electrons to the positive surface of the
pulled through the transmembrane domain, toward the
membrane in order to reduce a periplasmic cytochrome
inner, negatively charged surface of the membrane, using
c. The reduced cytochrome c then transfers its electrons
to an aa3 -type terminal oxidase (57). The entire PMF is
generated through the electrogenic activity of the aa3 -
type oxidase (3). The situation is further complicated by γ
the fact that Nitrobacter, as an obligate autotroph, must
generate NADH for CO2 reduction. Using more of the ADP+ Pi
PMF to drive electron transfer in reverse, such that
ATP
the NADH dehydrogenase oxidizes quinol and reduces
β α β
NAD+ to NADH, does this. This strategy allows the cell to
generate high-energy (low potential) electrons in the form
of NADH from lower energy (higher potential) electrons
from nitrite. To accomplish this, the cell must oxidize H+
several nitrite molecules for each molecule of NADH γ
produced.
−−−N
As shown by the examples discussed earlier, aerobes
can use many compounds as reductants and, by expressing
a variety of oxidases, can grow over a wide range of O2 con-
a
centrations. However, genomic sequence information sug- c c c
gests that some aerobes have an even larger respiratory
repertoire than currently available biochemical evidence
demonstrates. For example, it is known that the bac-
terium Rhodobacter sphaeroides expresses three terminal +++P
oxidases. These include an aa3 -type cytochrome c oxidase,
a cbb3 -type cytochrome c oxidase, and an uncharacterized H+
quinol oxidase (58–60) (Table 1). However, recent genomic
Figure 8. Structural organization of the F1 Fo -type ATPase. Not
analyses indicate that R. sphaeroides encodes at least five all subunits making up the complete complex are present. The Fo
terminal oxidases: a bd-type quinol oxidase, a heme-copper part of the complex resides in the membrane and is indicated by
quinol oxidase, and the three-cytochrome c oxidases. The the a and c subunits. The F1 part lies on the cytoplasmic side of
quinol oxidase that is expressed is likely to be the bd-type the membrane. The dashed arrow indicates direction of rotation
quinol oxidase, in a modified form (58). It seems likely that of the Fo complex during ATP synthesis.
158 AEROBIC RESPIRATION, PRINCIPLES OF

the energy of the transmembrane voltage gradient or PMF.

Through an electrostatic mechanism, the flow of protons apo apo
causes the ring of small subunits in the membrane to
rotate in a clockwise direction (64,67–69). This ring is
connected to the γ -subunit, which also rotates, and the
extramembrane ring of α- and β-subunits is prevented Fe, S Fe, S
from rotating by additional subunits that act as a stator.
As the γ -subunit rotates within the cap structure it
forms different associations with each β-subunit, much
↓O2 ↑O2
the same way as a camshaft in a multicylinder engine
forms different associations with each cylinder at any Concentration Concentration
given second. As γ rotates, the active site on each β-
subunit moves successively through three conformations:
one that loosely binds ADP and phosphate, one that binds [Fe-S] [Fe-S]
these substrates tightly and catalyzes the condensation
reaction to form ATP, and a third conformation that opens
the active site to release ATP. Thus, the overall result Binding to target sequence
of continued proton transfer through the transmembrane
domain of the synthase is constant consumption of ADP
and phosphate by the extramembrane headgroup and the 5'----TTGATNNNNATCAA----3'
constant release of ATP. 3'----AACTANNNNTAGTT----5'
Figure 9. Model of how the transcriptional regulator Fnr
responds to changes in O2 concentration in the cell. [Fe−S]
REGULATION OF AEROBIC METABOLISM represent O2 labile iron-sulfur centers found in Fnr. The
consensus Fnr target sequence is given below the active
As stated earlier, O2 is the preferred oxidant for aerobic [Fe−S]-containing Fnr.
bacteria, and when it is available, aerobes will adjust
their physiology to preferentially use this oxidant over all
others. Bacteria have developed complicated regulatory DNA-binding domain at the C-terminus and an effector-
circuitry to assess O2 concentrations. A few of these have binding domain at the N-terminus. There is little sequence
been studied in some detail and will be briefly described. similarity among the N-terminal domains among the
These proteins are found in many aerobic bacteria; various members of this family of proteins. However, most
however, it should not be assumed that every protein seem to be designed to use small molecules as effectors.
would have the same function in every aerobe. Each For example, there are members of the family that detect
species of bacteria has its own regulatory circuitry and the gas carbon monoxide and other members that respond
physiological set points making it difficult to extrapolate to changes in cyclic AMP (73,74).
from one organism to the next. One common theme among In E. coli, Fnr has been shown to repress the
the regulatory proteins studied so far is that they are expression of the genes encoding the cytochrome bd-
active when O2 is absent. That is, they are used to activate type oxidase under anaerobic conditions (75). A similar
genes whose products are required when O2 is limiting. role for Fnr occurs in A. vinelandii, with Fnr repressing
Regulatory proteins that are activated by increases in O2 synthesis expression of the bd-type oxidase used to protect
have not been studied in bacteria. nitrogenase from O2 (76). In Pseudomonas aeruginosa,
inactivation of the gene encoding Fnr caused increased
Oxygen Sensing Regulatory Proteins expression of an oxidase with high affinity for O2 ,
probably a cbb3 -type oxidase, along with a cyanide-
The prototype of this type of regulatory protein is Fnr.
insensitive oxidase that is probably a bd-type enzyme (77).
Fnr apparently senses O2 directly and binds to DNA
Inactivation of a gene encoding an Fnr homologue in
when O2 concentrations are low (Fig. 9). This is because
P. denitrificans also caused increased expression of a
Fnr contains an O2 labile [Fe−S] center (70). When
quinol oxidase, but repressed expression of the cbb3 -type
O2 concentrations are high the [Fe−S] center is not
oxidase (53). Expression of the genes encoding cbb3 -type
completely assembled. Consequently, the protein does not
oxidase in R. sphaeroides also requires Fnr (78).
dimerize and this prevents it from binding to its target
sequence (71). As O2 concentrations decline, conditions
Regulators That Indirectly Sense Changes in O2
become more favorable for assembly of a stable [Fe−S]
Concentration
center. Once the [Fe−S] is formed, Fnr can dimerize and
interact with its target sequence. In most cases, Fnr is Although Fnr appears to interact directly with O2 , other
a positive activator of gene expression. This protein is proteins appear to sense changes in O2 concentration by
found in a wide range of bacteria and is responsible indirect mechanisms. One of these systems is termed
for the expression of many genes. It is now evident the Arc system, which has been extensively studied in
that Fnr is member of a large family of transcriptional E. coli and its relatives. The Arc system consists of two
regulators termed the Fnr/Crp family (72). All of these proteins, ArcB, which is a membrane-associated histidine
regulators have similar structural organization, with a kinase, and ArcA, which is its associated response
 AEROBIC RESPIRATION, PRINCIPLES OF 159

regulator (79). Together, these two proteins form a two- 4. M. Saraste, Science 283, 1488–1493 (1999).
component regulatory system. ArcB apparently senses 5. G. Unden and J. Bongaerts, Biochim. Biophys. Acta 1,320,
some aspect of respiration, perhaps the redox state of 217–234 (1997).
an intermediate in the respiratory chain. There is no 6. M. J. Kelly, R. K. Poole, M. G. Yates, and C. Kennedy, J.
evidence that ArcB directly senses O2 . When conditions Bacteriol. 72, 6010–6019 (1990).
limit electron flow through the aerobic respiratory chain, 7. H. J. Sears et al., Microbiology 146, 2977–2985 (2000).
ArcB phosphorylates ArcA (80). ArcA then acts as either 8. N. Grigorieff, Curr. Opin. Struct. Biol. 9, 476–483 (1999).
an activator or a repressor of its target genes. The activity 9. T. Yagi, T. Yano, S. DiBernardo, and A. Matsuno-Yagi,
of ArcB also seems to be influenced by the presence Biochim. Biophys. Acta 1,364, 125–133 (1998).
of metabolic intermediates that indicate O2 is limiting, 10. A. Dupuis et al., Biochim. Biophys. Acta 1,364, 147–165
such as pyruvate and D-lactate (81). The ArcAB system (1998).
functions as a positive activator of expression of the genes 11. T. Friedrich and D. Scheide, FEBS Lett. 479, 1–5 (2000).
encoding the bd-type oxidase in E. coli (82). 12. T. Yano, S. Magnitsky, and T. Ohnishi, Biochim. Biophys.
A second example of a regulatory system that regulates Acta 1,459, 299–304 (2000).
anaerobically expressed genes, but does not directly sense 13. U. Brandt, Biochim. Biophys. Acta 1,318, 79–91 (1997).
O2 , is the PrrA/PrrB system (also termed RegA/RegB) 14. B. A. Ackrell, FEBS Lett. 466, 1–5 (2000).
in Rhodobacter species and related bacteria. Like the
15. C. R. D. Lancaster and A. Kroger, Biochim. Biophys. Acta
ArcBA system, the PrrBA system is a two-component
1,459, 422–431 (2000).
sensor-regulator system, with PrrB being the membrane-
16. Y. Anraku and R. B. Gennis, Trends Biochem. Sci. 12,
associated histidine kinase and PrrA the associated
262–266 (1987).
response regulator (79). Recent work in R. sphaeroides
17. R. B. Gennis, Biochim. Biophys. Acta 1,058, 21–24 (1991).
indicates that PrrBA interacts with the cbb3 -type
18. B. L. Trumpower and R. B. Gennis, in C. C. Richardson, ed.,
oxidase (83). When O2 is present, the cbb3 -type oxidase
vol. 63, Ann. Rev of Biochem, Annual Reviews Inc., Palo Alto,
is active and PrrB does not phosphorylate PrrA. However,
Calif., 1994, pp. 675–716.
when electron flow through the oxidase decreases because
19. E. Darrouzet, M. Valkova-Valchanova, T. Ohnishi, and
of O2 limitation, PrrB phosphorylates PrrA, causing it to
F. Daldal, J. Bioenerg. Biomembr. 31, 275–288 (1999).
activate transcription of a number of genes. In this case,
20. C. H. Snyder, T. Merbitz-Zahradnik, T. A. Link, and B. L.
the activity of the oxidase is the gauge the regulator uses
Trumpower, J. Bioenerg. Biomembr. 31, 235–242 (1999).
to determine ambient O2 levels.
21. A. R. Crofts and E. A. Berry, Curr. Opin. Struct. Biol. 8,
501–509 (1998).
TAXIS RESPONSE TO O2 22. J. van der Oost, et al., FEMS Microbiol. Lett. 121, 1–9 (1994).
23. S. Iwata, C. Ostermeier, B. Ludwig, and H. Michel, Nature
In addition to using their regulatory systems to monitor 376, 660–669 (1995).
environmental conditions, cells can use taxis to move 24. T. Tsukihara, et al., Science 272, 1136–1144 (1996).
themselves to positions that optimize aerobic respiration. 25. S. Ferguson-Miller and G. T. Babcock, Chem. Rev. 96,
Taxis toward terminal oxidants such as O2 has been 2889–2907 (1996).
observed for many years. This taxis response to O2 , 26. J. P. Osborne and R. B. Gennis, Biochim. Biophys. Acta 1,410,
termed aerotaxis, has been extensively studied in E. coli 32–50 (1999).
and its relatives (84). Taxis toward O2 is dependent on 27. B. C. Hill, J. Biol. Chem. 266, 2219–2226 (1991).
the ability of the cells to respire O2 , which suggests that 28. L. M. Geren, et al., J. Biol. Chem. 270, 2466–2472 (1995).
O2 is not directly sensed by cellular receptors. Rather, 29. R B. Gennis and S. Ferguson-Miller, Curr. Biol. 6, 36–38
the cell appears to respond to electron flow to a particular (1996).
oxidant. In E. coli, a membrane-bound flavoprotein termed 30. D. A. Mills and S. Ferguson-Miller, Biochim. Biophys. Acta
Aer has been shown to be involved in the aerobic taxis 1,365, 46–52 (1998).
response (85). It is likely that the redox status of the 31. P. Brzezinski and P. Adelroth, J. Bioenerg. Biomembr. 30,
flavin allows Aer to monitor electron flow to O2 (86). In the 99–107 (1998).
bacteria in which aerotaxis has been studied, O2 is always 32. D. Zaslavsky and R. B. Gennis, Biochim. Biophys. Acta 1,458,
the preferred oxidant and its presence will prevent taxis 164–179 (2000).
to alternative oxidants. 33. H. Michel, Biochemistry 38, 15129–15140 (1999).
Acknowledgments 34. M. Wikstrom, Biochemistry 39, 3515–3519 (2000).
We thank Sara Shapleigh for assistance with the figures. NIH 35. P. Dimroth, Biochim. Biophys. Acta 1,318, 11–51 (1997).
(GMS56824) supports work in the JH lab and the Department of 36. S. J. Kerscher, Biochim. Biophys. Acta 1,459, 274–283 (2000).
Energy (95ER20206) supports work in the JS lab. 37. R. D’mello, S. Hill, and R. K. Poole, Microbiology. 142,
755–763 (1996).
BIBLIOGRAPHY 38. R. K. Poole and S. Hill, Biosci. Rep. 17, 303–317 (1997).
39. M. W. Calhoun, J. W. Thomas, and R. B. Gennis, Trends
1. P. Mitchell, Nature 191, 144–148 (1961). Biochem. Sci. 19, 325–330 (1994).
2. P. Mitchell, Biol. Rev. Cam. Philos. Soc. 41, 445–502 (1966). 40. J. Castresana, M. Lubben, M. Saraste, and D. G. Higgins,
3. D. G. Nichols and S. J. Ferguson, Bioenergetics, vol. 2, Aca- EMBO J. 13, 2516–2525 (1994).
demic Press, London, U.K., 1992. 41. M. Saraste and J. Castresana, FEBS Lett. 341, 1–4 (1994).
160 AGGREGATES AND CONSORTIA, MICROBIAL

42. J. A. Garcia-Horsman et al., Biochemistry 33, 3113–3119 78. N. J. Mouncey and S. Kaplan, J. Bacteriol. 180, 2228–2231
(1993). (1998).
43. O. Preisig, R. Zufferey, L. Thony-Meyer, C. A. Appleby, and 79. A. S. Lynch and E. C. Lin, in Escherichia Coli and Salmo-
H. Hennecke, J. Bacteriol. 178, 1532–1538 (1996). nella, Cellular and Molecular Biology, vol. 1, ASM Press,
44. A. Jaworowski et al., Biochemistry 20, 3621–3628 (1981). Washington, D.C., 1996, pp. 1526–1538.
45. H. Leif et al., Biochem. Soc. Trans. 21, 998–1001 (1993). 80. O. Kwon, D. Georgellis, and E. C. Lin, J. Bacteriol. 182,
3858–3862 (2000).
46. R. B. Gennis and V. Stewart, in F. C. Neidhardt, ed.,
81. S. Iuchi et al., J. Bacteriol. 176, 1695–1701 (1994).
Escherichia Coli and Salmonella, Cellular and Molecular
Biology, ASM Press, Washington, D.C., 1996 pp. 217–261. 82. P. A. Cotter, S. B. Melville, J. Albrecht, and R. P. Gunsalus,
Mol. Microbiol. 25, 605–615 (1997).
47. C. P. Tseng, J. Albrecht, and R. Gunsalus, J. Bacteriol. 178,
1094–1098 (1996). 83. J. I. Oh and S. Kaplan, Biochemistry 38, 2688–2696 (1999).
48. M. W. Calhoun et al., J. Bacteriol. 175, 3020–3025 (1993). 84. B. L. Taylor, I. B. Zhulin, and M. S. Johnson, Annu. Rev.
Microbiol. 53, 103–128 (1999).
49. S. C. Baker et al., Microbiol. Mol. Biol. Rev. 62, 1046–1078
85. S. I. Bibikov, R. Biran, K. E. Rudd, and J. S. Parkinson, J.
(1998).
Bacteriol. 28, 683–690 (1997).
50. I. Zickermann et al., Biochim. Biophys. Acta 1,277, 93–102
86. A. Rebbapragada et al., Proc. Natl. Acad. Sci. U.S.A. 94,
(1996).
10541–10546 (1997).
51. D. Yang et al., Proc. Natl. Acad. Sci. U.S.A. 82, 4443–4447
(1985).
52. J. W. de Gier et al., Mol. Microbiol. 13, 183–196 (1994).
53. R. J. M. van Spanning et al., Mol. Microbiol. 23, 893–907 AGGREGATES AND CONSORTIA, MICROBIAL
(1997).
54. I. Zickermann et al., Eur. J. Biochem. 246, 618–624 (1997). HANS W. PAERL
55. T. Yamanaka, Plant Cell Physiol. 37, 569–574 (1996). University of North Carolina at
Chapel Hill
56. S. J. Ferguson, Periplasmic Electron Transport Reactions,
Morehead City, North Carolina
Academic Press, New York, 1998.
57. T. Nomoto, Y. Fukumori, and T. Yamanaka, J. Bacteriol. 175, JORMA KUPARINEN
4400–4404 (1993). Finnish Institute of Marine
58. N J. Mouncey et al., FEMS Microbiol. Lett. 192, 205–210 Research
(2000). Helsinki, Finland
59. J. P. Shapleigh and R. B. Gennis, Mol. Microbiol. 6, 635–642
Nearly a century of microbiological research indicates that
(1992).
structurally and functionally complex microbial communi-
60. J. H. Zeilstra-Ryalls and S. Kaplan, J. Bacteriol. 177,
ties commonly inhabit surface-associated environments in
6422–6431 (1995).
aquatic ecosystems. Surficial microbial communities often
61. J. Dassa et al., Mol. Gen. Genet. 229, 341–352 (1991). show higher metabolic activities and growth rates than
62. M. G. Sturr, T. A. Krulwich, and D. B. Hicks, J. Bacteriol. their freely suspended counterparts. This has been linked
178, 1742–1749 (1996). to surfaces facilitating close metabolic coupling among
63. R. H. Fillinghame, Science 286, 1687–1688 (1999). community members, leading to enhancement of growth
64. R. H. Fillinghame, W. Jiang, and O. Y. Dmitriev, J. Exp. Biol. and nutrient cycling that exceeds what is achievable at
203, 9–17 (2000). the single species level. These synergistic, mutually ben-
65. D. Stock, A. G. W. Leslie, and J. E. Walker, Science 286, eficial associations are termed consortia. Consortia are
1700–1705 (1999). common to aggregate and other surficial communities
66. P. D. Boyer, Annu. Rev. Biochem. 66, 717–749 (1997). broadly distributed along geographic and trophic gradi-
67. R. H. Fillinghame, W. Jiang, O. Y. Dmitriev, and P. C. Jones, ents, ranging from the tropics to the poles. Consortial
Biochim. Biophys. Acta 1,458, 387–403 (2000). activity, growth, and diversity are enhanced by biogeo-
68. Y. Wada, Y. Sambongi, and M. Futai, Biochim. Biophys. Acta chemical (i.e., oxygen, Eh , pH, nutrients) gradients in
1,459, 499–505 (2000). surficial microenvironments, or microzones. In part, con-
69. Y. Sambongi et al., Science 286, 1722–1724 (1999). sortia themselves promote biogeochemical zonation and
microbial diversification through localized production and
70. P. J. Kiley and H. Beinert, FEMS Microbiol. Rev. 341–352
(1998). mineralization of organic matter. This positive interac-
tion is a requisite for biodiversification, which in turn
71. D. M. Bates et al., J. Biol. Chem. 275, 6234–6240 (2000).
enhances microzonal productivity and nutrient cycling.
72. S. Spiro, Antonie von Leeuwenhoek 66, 23–36 (1994).
For these reasons, consortia are viewed as the functional
73. D. Shelver, R. L. Kerby, Y. He, and G. P. Roberts, Proc. Natl. units of biocomplexity in aquatic ecosystems.
Acad. Sci. U.S.A. 94, 11216–11220 (1997). The consortial ‘‘lifestyle’’ also appears to be a highly
74. S. Busby and R. H. Ebright, J. Mol. Biol. 293, 199–213 (1999). effective strategy for colonizing and exploiting nutrient
75. F. Govantes, J. A. Albrecht, and R. P. Gunsalus, Mol. Micro- and energy resources under environmentally extreme
biol. 37, 1456–1469 (2000). conditions at ‘‘the edge of life.’’ Examples from oligotrophic
76. G. Wu et al., Microbiology 143, 2197–2207 (1997). planktonic, hypersaline, and lithifying benthic, Baltic Sea
77. A. Ray and H. D. Williams, FEMS Microbiol. Lett. 156, and Antarctic lake ice communities are used to exemplify
227–232 (1997). the important roles aggregate and surface microbial
 AGGREGATES AND CONSORTIA, MICROBIAL 161

communities play in ecosystem, regional and global of processes are often required to fully metabolize and
biogeochemical cycling and production dynamics. Because derive energy from organic and inorganic compounds. For
they are highly sensitive, exhibit rapid community growth, example, cyanobacteria utilize oxygenic photosynthesis to
and compositional responses to environmental change obtain energy and carbon compounds to support growth.
and extremes, microbial consortia are excellent indicators However, cyanobacteria also rely on highly reduced pro-
of climatic and geochemical changes, as well as human cesses, such as nitrogen (N2 ) fixation to obtain biologically
pollutants and other perturbations in aquatic ecosystems. available nitrogen in nitrogen-deplete waters. The oxygen
evolved from photosynthesis is inhibitory to N2 fixation; it
SUSPENDED AND SOLID SURFACES AS can also exert a strong negative feedback control on photo-
MICROENVIRONMENTS synthesis itself (i.e., the Warburg effect, see Refs. 11,12).
Some filamentous cyanobacteria are able to separate these
Microorganisms are the dominant producers and cyclers cross-inhibitory processes by cellular specialization (e.g.,
of organic and inorganic compounds in the biosphere. In the formation of heterocysts to which N2 fixation is con-
aquatic environments, at least two-thirds of primary pro- fined) (12,13). Such specialization requires multicellular
duction is mediated by microorganisms, including bacteria compartmentalization, a luxury that may not be practical
(cyanobacteria and photosynthetic bacteria) and microal- within the confines of micron-scale surficial habitats.
gae (1). A vast percentage of aquatic production is either To circumvent the ‘‘packaging problem,’’ physiologically
sedimented or recycled via microbial mineralization (2,3). distinct prokaryotes, eukaryotic microalgae, protozoans,
Therefore, environmental controls of microbial production, and fungi participate in metabolically coupled associa-
mineralization, and associated nutrient transformations tions (13–16), enabling them to colonize, optimize growth,
are key determinants of aquatic ecosystem structure, establish redox gradients, and maintain effective nutrient
function, and material flux. A vast proportion of the cycling within the confines of a microenvironment. Under
Earth’s marine and freshwater ecosystems is nutrient- and these conditions, smallness, metabolic specialization, and
energy-poor, and as such production and mineralization close proximity along micron-scale biogeochemical gra-
are tightly controlled by nutrient and energy availability. dients are advantageous for (1) survival, (2) obtaining
In natural waters, nutrient resources are heterogeneously nutrient and energy sources, (3) establishing high bio-
distributed in time and space (4), with relatively high diversity, (4) facilitating nutrient and genetic exchange,
concentrations associated with particles and submersed and (5) exploiting environmentally ‘‘extreme’’ habitats not
surfaces (2,5), which can act as sources or concentra- colonizable by single species acting alone (14,17).
tion sites of nutrients. Microbial biomass and activities Microorganisms participating in the ‘‘lifestyle’’ descri-
are generally higher on surfaces than in the free-floating bed earlier operate as consortia. That is, populations
state (2,5,6). As a result, microbial growth and nutrient have evolved to functionally specialize and diversify along
transformation are often stimulated by surfaces (5,7). This
micron-scale biogeochemical (redox, nutrients, organic
has been attributed to the nutrient-rich ‘‘oasis’’ that sur-
matter supply, etc.) gradients, or ‘‘microzones,’’ where
ficial microenvironments provide in an otherwise dilute
effective exchange of metabolites leads to ‘‘one organism’s
‘‘desert’’ macroenvironment (2,8).
trash being another’s treasure.’’ Consortia are defined
Relatively high rates of primary productivity and
as the operational units of biocomplexity in nature,
respiration associated with surfaces lead to localized
as several species or populations that function in a
regions of elevated organic-matter enrichment and con-
complementary fashion such that production and growth
sumption, accompanied by the development of oxygen
are enhanced above that which a single species or
gradients (2,9,10), which provide conditions for species
loosely-knit collection of populations (i.e., assemblage)
enrichment along the resultant biogeochemical contin-
uum. These are the essential ingredients for intensification can achieve independently (14,17). On regional and global
and diversification of surficial biogeochemical cycling. scales, consortial associations in aggregates, biofilms, and
Owing to their rapid growth rates and small size microbial mats play basic roles in production and nutrient
(i.e., ∼µm), microorganisms are excellent exploiters of cycling dynamics. The ‘‘microbial loop’’ in trophodynamics,
nutrient-enriched surficial microenvironments. Small cell energy flow, and biogeochemical cycling highlights the
size enables large numbers of cells to colonize these central role consortia play in ecosystem processes (3).
microenvironments. Small cell size also provides high Clarification and quantification of consortial components
surface to volume ratios, facilitating nutrient uptake at and their exchange rates are essential for developing
low concentrations. The metabolic activities (i.e., oxy- process-based models linking biocomplexity to function
gen production and consumption, pH changes) associated in aquatic and terrestrial ecosystems.
with metabolically active microbial communities further Consortia develop within, and as a consequence of, an
enhance biogeochemical zonation, promoting microbial interactive matrix of geological, physical, and chemical
diversification. However, there is a price to pay for parameters. Over time, consortia develop by interactions
being small. The small cell size and prokaryotic (lack of at some quasi ‘‘steady-state’’ whose determinants arise
organelles and nucleus) nature of bacteria restricts cellu- from the prevailing interplay of system parameters
lar biochemical specialization. This metabolic ‘‘packaging (Fig. 1). We will explore this operational definition of
problem’’ is particularly problematic when considering consortial life in aggregates and surficial microzones and
the need for harboring aerobic (oxidized) and anaero- its role in production and biogeochemical cycling dynamics
bic (reduced) processes on the cellular level. Both types in planktonic and benthic habitats.
162 AGGREGATES AND CONSORTIA, MICROBIAL

Physical factors

T l S pH
Chemical
modulators

N, P
Aggregate/surface
microbial consortia
Transport
Organic Recycling
matter Uptake
processes

Fe &trace Autotroph Heterotroph

Grazing
elements
Production
Toxic
substances

Diffusion&
shear Higher
effects Vertical
trophic
mixing Transport
levels
&
Resuspension

Sediment environment
Figure 1. Conceptual diagram depicting influence of natural and anthropogenic environmental
factors, and physical modulators and biotic interactions on aggregate/surface microbial consortial
assemblages. Physical modulators include temperature (T), irradiance (I), salinity (S), and pH
(pH).

FORMATION OF AGGREGATES AND CONSORTIA aggregation (27). Aging microaggregates collide with each
other and form visible macroaggregates, such as ‘‘marine
Suspended Aggregates snow’’ and ‘‘lake snow.’’
Formation of visible, suspended macroscopic aggregates A two-state coagulation theory, in which coagulation is
composed of microscopic particles is a common phe- insignificant at low, but significant at high algal con-
nomenon in aquatic ecosystems (18–23). Microscopic par- centrations with a rapid transition between the two
ticulate matter in surface waters consists of living organ- states has been suggested for diatom blooms (28–30).
isms, inorganic particles, and nonliving organic matter, The critical cell concentration is dependent on cell
or detritus. Aggregation of these ‘‘particles’’ is a complex size, cell surface stickiness, and turbulence, and may
process involving physical, chemical, and biological inter- be reached rapidly, with the entire sequence of aggre-
actions on different spatial and temporal scales (Fig. 1). gation and sedimentation completed in 24 hours (31).
Physical processes, such as shear (difference in fluid In the Baltic Sea, such rapid transition from diatom
movements on small scales), Brownian motion (thermal growth to aggregation and sedimentation has been
movement of one particle toward another), and differen- recorded during spring blooms (32–35). Nutrient deple-
tial settlement (one particle sinking faster than the other, tion at the end of a bloom has been shown to increase
overtaking and colliding with it) bring particles together diatom cell stickiness, enhancing aggregation (30,36,37).
(Fig. 2) (24). The number of particles per unit volume of Despite its episodic nature, aggregation is signifi-
water is crucial for encountering other particles and form- cant and widespread enough to control ecosystem-level
ing aggregates. Frequently, physical discontinuities in the annual energy flow and nutrient dynamics owing to
water column, such as the pycnocline, act as sites for the transport of organic matter to the benthos by
enhanced aggregation. sedimentation.
Chemical and biological properties of colliding parti- Physical and biological processes may also decrease
cles (‘‘stickiness’’) determine whether they adhere and aggregate size (Fig. 2). Turbulence may tear particles
aggregate. The mucus sheaths and other organic sur- apart or bring them together (38,39). Biological disaggre-
face materials, especially polysaccharide particles known gation by microorganisms or consumption by zooplankton
as transparent exopolymer particles surrounding organ- can, at times, be an important factor regulating aggregate
isms, act as adsorption surfaces for additional parti- size (25,40). Grazing of aggregates by fish and zooplank-
cles (19,25,26). Attachment to surfaces stimulates bacte- ton is an important shortcut in the food web, transferring
rial exopolysaccharide synthesis, which further enhances organic carbon to higher trophic levels more efficiently
 Aggregate size
A4, H4

Growth rate
A3, H3

dm A2, H2

A3, H3
A1, H1
A2, H2
cm

A1, H1
A2, H2

mm
A1, H1

A1, H1
µm

Physical forcing
Brownian Brownian motion, Brownian motion,
motion shear, shear,
differential sedimentation differential sedimentation,
mesoscale hydrodynamics:
pycnocline, fronts, eddies,

Heterotrophs

Autotrophs
Mesozoopl. >200
µm

A3 Microplankton Microzoopl. 20-200 µm H3

A2 Nanoplankton Nanozoopl. 2-20 µm H2

A1 Picoplankton Bacteria 0,2-2 µm H1

Virus

DOM

Figure 2. Aggregate size development in relation to physical forcing and time. A1-H3 and H1-H3
represent autotrophic (A) and heterotrophic (H) plankton in the classical size structured pelagial
food web. Dashed line represents the artificial partitioning between DOM and particulate organic
matter. Brownian motion is a relevant collision mechanism with particles smaller than 1 µm
(A1, H1 particles) (38). Shear contact from laminar or turbulent processes is a common collision
mechanism for particles larger than 1 µm (A2, H2, and above). The third collision mechanism
is differential sedimentation (A2, H2, and above). The critical cell concentration for aggregation
depends on cell size, cell surface properties, and environmental shear. Spherical particles have
upper limits to their size as large particles become vulnerable to break up by the same shear that
brings them together. Large aggregates (>cm A4, H4 particles) are formed when organisms have
appendages, hairy structures, or grow in filaments. The physiological status of the cell adds to the
‘‘stickiness’’ of the particle.

163
164 AGGREGATES AND CONSORTIA, MICROBIAL

than via the microbial loop (23,41). The relative impor- physically modify the attachment substrates. Examples
tance of these mechanisms in aggregate dynamics depends include, complex biofilms on submersed surfaces (54,58),
on the origin and age of the aggregates (28,42,43). adhesive mats that consolidate sediments (55), beach
sands and lagoonal environments (55), providing protec-
Sea and Lake Ice Aggregates tion against erosion from storm surges, and calcifying mats
Seasonal formation of sea ice traps organisms in that form permanent lithified, layered deposits called stro-
organic-matter rich aggregates. Ice-associated microbial matolites, the oldest known evidence (2 billion years ago)
assemblages in the brackish Baltic Sea resemble those for microbial consortial life on Earth (59,60).
found in Antarctic and Arctic sea ice (44–46). The biotic
assemblages in the Baltic Sea ice consist of diatoms, photo- Microbial Modification of Aggregates and Biofilms
and heterotrophic flagellates, cyanobacteria, heterotrophic
Once attached, microbial communities can physically and
bacteria, and metazoa (44–48). Diatoms and various
chemically modify the substrate. Recently divided ‘‘young’’
autotrophic flagellates are the most important algal
groups; hetero- and mixotrophic organisms (flagellates and phytoplankton cells usually show the least amount of bac-
ciliates) compose 2 to 10% of carbon-based biomass (49). terial colonization (61,62). As the cells age, they are more
The accumulation of inorganic nutrients and abundant readily colonized by microorganisms. Bacterial coloniza-
heterotrophic organisms seasonally found in the Baltic Sea tion may be quite intense along algal filaments surrounded
is indicative of active carbon mineralization in the interior by mucilage (23,62,63) and cyanobacterial heterocysts (61)
parts of the ice. The interior layers of 2- to 3-month- (Fig. 3). These bacterial epibionts represent a key food
old ice are also active sites for nitrogen transformations. source for protozoan grazers, leading to complex microbial
The enrichment of denitrifying organisms in the middle food webs (19,21,39,62,64–66). Among aging aggregates,
layers of thick ice, together with elevated nitrite senescent algae provide degradable organic matter for
concentrations, indicate active nitrate reduction and large numbers of attached bacteria. The role of the recy-
potentially denitrification, that is, microbial reduction of cling of organic matter via microzooplankton back to
NO3 − to N2 gas (50). Despite the low rates of N2 O and N2 bacteria is enhanced as the aggregates mature (39,41,62).
production from seasonal ice (50), the large aerial expanse Cyanobacterial aggregates that avoid sedimentation
of seasonal ice on the global scale renders sea ice microbial and grazing can maintain ‘‘steady-state’’ aggregation con-
processes of high significance. ditions for long periods owing to recycling of organic
Microbial aggregates may also be embedded in the ice matter and mineral nutrients by attached microflora
cover of lakes and marine systems for many years. The within the phycosphere (62,67,68). Under these condi-
permanent ice cover of the glacial lakes in the McMurdo tions, ‘‘host’’ cyanobacteria can reveal high growth rates,
Dry Valley region of Antarctica is one example (51–53). frequently exceeding rates obtainable under axenic con-
Here, aggregates originate as windblown soil particles ditions (10). In fact, some bloom-forming cyanobacterial
(originating in the surrounding exposed desert soils) species are only culturable when bacterial epibionts are
harboring desiccated phototrophic (cyanobacteria) and present (13). Among other cyanobacterial genera (e.g.,
heterotrophic microflora that are deposited on the ice Oscillatoria, Lyngbya, Nodularia), bacterially colonized
cover. Solar heating of the soil particles enables them to strains revealed higher growth rates and were easier
melt into the ice cover, where they remain for several to maintain in culture than axenic strains. Gibson and
years. The ecophysiology of these ‘‘entombed’’ aggregates Smith (69) reported that both axenic Oscillatoria redekei
is discussed in detail later. As with sea ice, lake ice primary and Oscillatoria agardhii isolates ‘‘always appeared to
production and nutrient cycling rates are extremely low; grow better in the presence of contaminant heterotrophic
however, on the ecosystem-scale, the aggregates are a bacteria.’’ Similar findings were reported by Meffert (70),
major source of ‘‘new’’ production and nutrient input to Herbst and Overbeck (71), Caldwell (72), and Lehtimäki
these lakes (51). and coworkers (73), suggesting consortial interactions.
Cyanobacteria excrete organic compounds, including
Surficial and Benthic Habitats organic and amino acids, peptides, alkaloids, carbohy-
The bottom (benthos) and other surficial attached environ- drates, and lipopolysaccharides (74–76). Diverse excre-
ments provide diverse colonization sites for metabolically tion products chemotactically attract and support the
complex microbial consortia. These include, biofilms on growth of phycosphere-associated bacteria (77,78). In
submersed inorganic and organic nonliving substrates, N2 fixation and fate experiments, some of the 15 N-
including rocks, metals, glass, wood, plastics, and other labeled N2 fixed by host A. oscillarioides was rapidly
synthetic materials (5,54). Consortia also inhabit the sur- transferred to heterocyst-associated Pseudomonas aerug-
faces of living plants (epiphytic) and animals (epizoic). inosa (79). Axenic isolates of A. oscillarioides exhibited
Benthic habitats supporting microbial consortia include optimal growth and N2 fixation rates when reinocu-
submersed and intertidal mud- and sandflats, hard (rock) lated with P. aeruginosa (16). When released in axenic
and soft (mud-silt) bottoms, marshes, reef, and shelf A. oscillarioides cultures, P. aeruginosa reestablished the
environments (55–57). Consortia inhabiting these envi- heterocyst-specific association. These findings suggest
ronments range from ephemeral, loosely organized, floccu- close metabolic coupling and mutually beneficial relation-
lent, accumulations of microalgae, and associated bacteria, ships in cyanobacterial-bacterial aggregates.
fungi, and metazoans, to well-developed laminated micro- Aggregation into large flocculates by filamentous
bial mat communities (55). Microbial communities may cyanobacteria may be of major importance for prolongation
 AGGREGATES AND CONSORTIA, MICROBIAL 165

(a) pH pO2
Surface

Eh

Day
0.5 to 1.0 cm

(b) Eh pH pO2
Surface

Night
0.5 to 1.0 cm

Magnitude
Figure 4. Idealized set of vertical profiles of redox potential
(Eh ), pH, and percent saturation of dissolved oxygen (pO2 ) in a
microbial mat. The differences among relative values for each
parameter are shown for daytime (upper frame) and nighttime.
(c) Typically, the biogeochemical gradients show strongest vertical
zonation in the upper 0.5 to 1.0 cm of the mat where light can
potentially penetrate and photosynthesis takes place.

sands, muds, rocks, metals, glass, concrete, wood, even

plastics, and other synthetic substances. Mats are mor-
phologically complex microecosystems that possess a high
degree of phylogenetic and functional (metabolic) diver-
sity compressed into only a few millimeters (55,83). Mats
are widely distributed, with their range including polar
deserts, glacial lakes and streams, lakes and marine
ecosystems, the deep-sea, polar to tropical shallow seas,
reefs, mud- and sandflats, saltmarshes, hypersaline, and
Figure 3. Microbial associations with bloom-forming cyanobac- hyperthermal ecosystems. In many of these systems, mats
teria: (a) bacteria inside Anabaeana spiroides coils; (b) bacteria are among the most productive and diverse biotic commu-
around Gomphosphaeria colonies; (c) bacteria on heterocyst of nities. Mats are composed of microbial phototrophs (anoxy-
Anabaeana oscillarioides. See color insert. genic and cyanobacterial), heterotrophs, and chemoau-
totrophs (55,84–87). Characteristically, mats exhibit ver-
tical biogeochemical zonation (10,88,89), which usually
of bloom events. Further, the interior of these large floc- includes an upper oxic and lower anoxic zone (Fig. 4).
culates can become anaerobic (80) like the microbial mats, In aphotic regions (i.e., deep sea), or under anaerobic
support denitrification (81) and sulfate reduction (82), bio-
(i.e., sulfidic hot springs) conditions, anaerobic processes
geochemical transformations not predicted by models
prevail. In euphotic, oxygenated environments, photosyn-
that assume random distribution of organic matter and
thetic production is usually conducted by a phylogenet-
microbes in the pelagic environment.
ically diverse assemblage of cyanobacteria and diatoms
that dominate the upper oxic layer. Lower regions of
Microbial Mats
well-developed mats are often O2 subsaturated or anoxic,
Microbial mats are laminated microbial communities asso- leading to sulfide accumulation generated by heterotrophic
ciated with a variety of benthic substrates, including sulfate-reducing bacteria (SRB). These mats also have
166 AGGREGATES AND CONSORTIA, MICROBIAL

steep vertical redox (Eh ) and pH gradients. Gradients cre- can be formed even in the upper layers of mats. In addi-
ate a range of microzonal (micron to millimeter scale) tion, facultative anaerobic and aerobic bacteria associated
habitats, in which metabolically diverse microbial con- with anaerobic bacteria provide removal of potentially
sortia are located according to energetic, nutrient, and toxic O2 (79). Microscopic observations of mats treated
ecological needs and limitations. Consortia closely inter- with redox (tetrazolium) dyes reveal a highly heteroge-
act by exchange of metabolites, enabling major biochemical neous matrix supporting anoxic microzones even within
processes with different environmental requirements (e.g., largely oxic regions (determined by microelectrodes) (103).
oxygenic photosynthesis, N2 fixation, denitrification, nitri- This heterogeneity promotes high genetic and physiolog-
fication, sulfate reduction, methanogenesis) to function ical diversity, needed to fully optimize production and
contiguously and contemporaneously (9,10,90–92). nutrient cycling processes requiring a range of oxygen
Depending on environmental conditions, critical energy conditions. The production of mucilaginous exopolysac-
and nutrient transformation processes oriented along charides (EPS) by resident cyanobacteria and associated
these gradients may be cross-inhibitory. For example, bacteria promotes sediment stabilization and cohesion of
during sunny days, rates of oxygenic photosynthesis the mat, acts as diffusional barrier, provides surfaces and
may be high, leading to a buildup of oxygen-saturated substrates for growth (104), and selectively binds potential
conditions in the mat matrix. This could prove inhibitory toxins such as heavy metals (105).
to anaerobic carbon, nitrogen, and sulfur transformation
processes, especially near the mat surface. For example, ECOLOGICAL AND BIOGEOCHEMICAL ROLES OF
daytime O2 supersaturation presents a problem for AGGREGATES AND CONSORTIA
N2 fixation, denitrification, sulfate reduction, and other
anaerobic processes (e.g., methanogenesis) that are reliant By virtue of ensuring quasi-stable, small-scale biogeo-
on photosynthetically produced organic matter as an chemical gradients, planktonic aggregates and surficial
energy source, yet are inhibited by even low ambient microzonal habitats provide a protective set of niches
O2 concentrations (91–93). Sulfate reducers account for for complex, highly interactive microbial production and
a large fraction of the heterotrophs in some mats (90), nutrient cycling processes. These microzones harbor
and are also suspected of being significant contributors of metabolically diverse microbes even in the face of severe
fixed nitrogen (94). Although N2 fixation can be observed environmental constraints. As a result, life in some of
during daytime, rates often increase at night or in the the most extreme environments on Earth largely oper-
early morning when pO2 is low. It is suspected that sulfate ates in a consortial mode. We will explore the consortial
reducers may contribute to N2 fixation at this time. ‘‘players,’’ their metabolic and structural strategies for
Denitrification can be an important process mediating survival and growth in a range of environments, including
nitrogen availability in microbial mats (95). Denitrify- extreme ones. We will also examine the implications for
ing bacteria include taxonomically diverse heterotrophs aquatic production and biogeochemical cycling in physi-
that are either facultatively anaerobic (Pseudomonas and cally, geochemically, and ecologically diverse planktonic
Thiobacillus) or obligate anaerobic (Bacillus and Spiril- and benthic habitats. These include
lum) genera. Molecular O2 strongly inhibits denitrification
because O2 can outcompete NO3 − as a respiratory electron 1. Cyanobacterial aggregates in lakes, oceans, and in
acceptor. Oxygen can also be directly toxic to some den- the brackish Baltic Sea
itrifiers. Therefore, denitrifiers are often most effective
2. Nitrogen-fixing cyanobacterial (Trichodesmium)
in anoxic zones near NO3 − sources. Organically enriched
aggregates in the open ocean
surface sediments and mats harboring anoxic microzones
overlain by NO3 − rich bottom water are optimal conditions 3. Microbial mats in temperate and tropical ecosystems
for denitrifiers (96,97). 4. Microbial mats in hypersaline waters
Even though sulfate reduction, N2 fixation, and den- 5. Aggregates embedded in the permanent ice cover of
itrification are all inhibited by even low pO2 , they can Antarctic lakes.
co-occur during peak photosynthetic (i.e., high O2 evolving)
periods (56,90,97,98). Consortial syntrophic associations CYANOBACTERIAL AGGREGATES IN LAKES, OCEANS,
between presumed metabolically incompatible organisms, AND IN THE BRACKISH BALTIC SEA
such as oxygenic cyanobacteria and obligate anaerobic pur-
ple sulfur bacteria, sulfate reducers and facultative aer- Planktonic diazotrophic cyanobacteria are common in
obes (99,100), and sulfate reducers and methanogens (101) eutrophic lakes, rare in oceans, and nearly absent in
have been demonstrated. These results can be reconciled estuaries and coastal seas (106). Exceptions are the Baltic
with the observation of small-scale (i.e., 1 to 100 µm) anoxic Sea and the Peel-Harvey estuary in Australia (107).
microzones in an oxygenated matrix (i.e., mats) as deter- Cyanobacteria are responsible for most of the planktonic
mined by localized reduction of tetrazolium salts (79,102) N2 fixation, but rates are high only when these organisms
and the presence of active photosynthetic bacteria (103). make up a majority of the biomass (107). In eutrophic
In mats, anoxic microzones are promoted and main- lakes, planktonic N2 fixation may account for 6 to 82%
tained by abundant supplies of organic matter (via sedi- of the nitrogen input into the system. In the surface
mentation and high rates of primary production), which waters of the world’s oceans and estuaries, including
support high rates of respiration. If respiratory O2 con- eutrophic estuaries, N2 fixation provides a far lower
sumption exceeds inward O2 diffusion, anoxic microzones percentage of nitrogen input (107). In the Baltic Sea,
 AGGREGATES AND CONSORTIA, MICROBIAL 167

however, planktonic N2 fixation provides at least 15% (a)

of the nitrogen input (108).
The mass appearance of N2 fixing cyanobacteria in
eutrophic lakes has been linked to the ratio of nitrogen
to phosphorus loading. An N/P ratio near or below the
Redfield ratio of 16 : 1 provides a selective advantage
for N2 fixing phytoplankton species. However, planktonic
N2 fixation is low in many estuaries despite low N/P
input ratios. One explanation for this conspicuous absence
is iron or possibly trace element limitation (109,110).
Another explanation is that persistent high turbulence
negatively impacts N2 fixing cyanobacterial growth and
hence reduces their dominance (111). Lastly, organic
matter enrichment may be a prerequisite for bloom
formation (112,113). The cyanobacterial blooms in the
Baltic Sea and Peel-Harvey estuary have been linked to
high concentrations of dissolved organic matter, which
increase the availability of iron and possibly other
metals through chelation (114,115). With the exception
of iron (discussed later), none of these factors appear
to explain large blooms of the pelagic marine N2 fixer
Trichodesmium, which inhabit some of the most nutrient-
deprived waters on Earth (116).
Nodularia spumigena is the dominant cyanobacterium
in the Peel-Harvey estuary, Australia (117). In the Baltic
Sea, N. spumigena is also abundant but other heterocys-
tous cyanobacteria, such as Aphanizomenon flos-aquae
and Anabena spp. coexist in blooms (118–121). Nodularia
spumigena and A. flos-aquae blooms may be initiated in
the open Baltic following deep-water phosphorus intru- (b)
sions to the warm surface layer, providing a competitive
advantage for N2 fixers (121).
In the open sea, contrary to many enclosed coastal
bays, hydrodynamically induced nutrient additions to the
phytoplankton community are pulsed, episodic events. As
shown by Rothhaupt and Güde (122) and Thingstad and
coworkers (123), even minor amounts of nutrients may
cause a shift from picoplanktonic (less than 2 µm) to larger
organisms. The cyanobacterial blooms occurring in the
open Baltic often start with A. flos-aquae, which, owing
to high phosphorus storage ability, is favored by pulsed
phosphorus inputs (124). Once the blooms are established,
N. spumigena may cover large areas of the open Baltic Sea
(Fig. 5) (121).
During their initiation, N. spumigena blooms start as
straight filaments. During this early stage, phosphorus
intrusions from deeper water can occur (121), enabling
Nodularia to experience unrestricted growth conditions
when it fixes N2 (125,126). If phosphorus inputs to the Figure 5. Observations of a bloom of the N2 fixing cyanobac-
planktonic community are intermittent, the development teria Nodularia spp. Left-hand side: a Nodularia-dominated
and persistence of a bloom must rely on regeneration cyanobacterial bloom observed in the Gulf of Finland, Baltic Sea
of the stored phosphorus. In later growth stages, when (photo courtesy of Pia Moisander). Upper right-hand side: pho-
Nodularia dominates over Aphanizomenon, formation of tomicrograph (400x) of several Nodularia sp. filaments, showing
large aggregates of coiled Nodularia filaments has been the nitrogen fixing heterocysts (round, light green cells), neigh-
observed (Fig. 5). boring heavily (green) pigmented vegetative (photosynthetically
active) cells, as well as associated bacteria. Lower right: coiled
As a proposed survival strategy, coiling provides a
N. spumigena aggregates with associated bacteria and protozoans
means to engulf and support the development of an
present inside the coils (400x). See color insert.
active microbial food web promoting phosphorus recy-
cling within the confines of the aggregate. In contrast,
A. flos-aquae forms tight flake-like aggregates that are not the outside of these aggregates (127,128). In the absence
penetrated by other microbes. Rather, bacteria attach to of additional phosphorus pulses, A. flos-aquae loses its
168 AGGREGATES AND CONSORTIA, MICROBIAL

(C) in aggregates, complicating their roles in the nitrogen

budget.

NITROGEN-FIXING CYANOBACTERIAL
(TRICHODESMIUM SPP.) AGGREGATES IN THE
SUBTROPICAL AND TROPICAL OCEAN

Among planktonic cyanobacterial aggregate-formers, the

filamentous nonheterocystous N2 fixing genus Tri-
chodesmium is the most widespread and biogeochem-
ically important contributor of ‘‘new’’ nitrogen to
nitrogen-depleted subtropical and tropical oligotrophic
oceans (138,139). Trichodesmium occurs as buoyant
macroscopic (a few millimeters in length) radial (puff)
or fusiform (tuft) aggregates (Fig. 6), which exhibit high
rates of primary production and N2 fixation, as well as
associated microbial (bacterial, microalgal, and protozoan)
and crustacean grazer populations (140,141). When calm
Figure 5. (Continued) seas prevail, buoyant aggregates accumulate as large
(sometimes greater than 100 km2 ) yellow-brown slicks
covering the ocean’s surface (139,142). Trichodesmium
dominance owing to its reliance on this phosphorous sup- is one of the few planktonic nonheterocystous genera
ply mechanism. Under these circumstances, N. spumigena that can fix N2 and CO2 (and hence evolve O2 ) simul-
can persist because it is able to maintain effective phos- taneously, exhibiting parallel maxima of these processes
phorus regeneration mediated by the microbial consortia during daytime. Capone and coworkers (143) showed that
in its coiled matrix. Thus, coiling-based aggregate forma- even when supplied with light during nighttime, N2 fix-
tion provides a competitive advantage to N. spumigena, in ation (but not CO2 fixation) ceased, suggesting strong
part owing to its ability to ‘‘win the battle’’ against dif- daytime dependence and a diel ‘‘rhythm’’ in its dia-
fusive loss of nutrients. In addition, N. spumigena has a zotrophic behavior. Molecular studies indicate that the
higher overall affinity for phosphorus than A. flos-aquae, nitrogenase enzyme complex of Trichodesmium is struc-
which helps explain the successional sequence of the two turally (i.e., amino acid sequence) conserved and similar
species (129). to that found in other diazotrophs (144,145). As with
What is the biogeochemical contribution of cyanobac- all other diazotrophs, Trichodesmium’s nitrogenase is O2
terial blooms to the Baltic ecosystem? Apart from the sensitive. However, unlike heterocystous cyanobacteria
toxicity aspect of the cyanobacterial blooms, little is (e.g., Nodularia, Aphanizomenon), in which nitrogenase
known about the contribution of blooms to, for example, is protected in O2 -devoid heterocysts, Trichodesmium
carbon flux. Although visually impressive, the cyanobac- reveals no obvious cellular differentiation. A great deal
teria may contribute only little to the pelagic food web of research and speculation has been directed toward clar-
in terms of carbon as the two dominant cyanobacterial ifying how this enigmatic diazotroph manages to optimize
species N. spumigena and A. flos-aquae are poorly eaten these potentially cross-inhibitory processes in the open
(only a few percentage of the biomass) by mesozooplank- ocean. Trichodesmium fixes N2 at highest cellular rates
ton (130,131). Little or no sedimentation of the summer when present as surface-dwelling aggregates as opposed
plankton (including cyanobacteria) has been recorded to single filaments (116,142,146,147). When viewed in
in various basins (132–134). During the cyanobacterial situ, aggregates reveal highly-compacted, dense, dark-
bloom period, less than 1% of the suspended concen- pigmented central cores, while their peripheral regions are
trations settled daily (133), which led the authors to often more diffuse and less compacted (148,149) (Fig. 6).
conclude that the cyanobacterial bloom had disintegrated Fogg (13) hypothesized that CO2 and N2 fixation were
into dissolved and particulate organic pools incorpo- spatially segregated, with CO2 fixation confined to high
rated into the food web via the microbial loop. Nev- irradiance external regions and N2 fixation confined to
ertheless, in N. spumigena-dominated blooms, a highly optically dense internal regions. Carpenter and Price (150)
active microbial food web may develop within aggre- and Paerl (149) provided support of Fogg’s hypothesis by
gates and thus contribute to pelagic planktonic produc- microautoradiographically showing that external regions
tion (135). In terms of the Baltic Sea’s nitrogen budget, of individual trichomes were photosynthetically much
cyanobacterial aggregates appear to be important nitro- more active than internal regions. Furthermore, numer-
gen transformation sites. N2 fixation by cyanobacterial ous studies (148–151) have shown a positive relationship
bloom-formers is a significant source of externally sup- between calm sea state (facilitating aggregation) and rates
plied or ‘‘new’’ nitrogen (108,136). It is suspected that of N2 fixation in natural Caribbean Sea, Indian, and
aggregates may also be sites of denitrification (137); how- Pacific Ocean Trichodesmium populations. Bryceson and
ever, rates and overall contributions to nitrogen losses Fay (148) also showed a strong direct correlation between
via this process have not been adequately assessed. It is aggregate shape, size, and N2 fixation potentials in Tri-
possible that N2 fixation and denitrification may co-occur chodesmium sampled off the east African coast.
 AGGREGATES AND CONSORTIA, MICROBIAL 169

(a) external regions. One explanation for this partitioning

is self-shading, imposed by the dense aggregates. It
is also possible that respiration is heterogeneously
distributed in aggregates. Immunological studies did
not show similar partitioning for nitrogenase, with
this enzyme being scattered, possibly as groups of
cells, throughout aggregates (140,153). These studies
indicate a widespread potential for N2 fixation throughout
aggregates. Thus, while photosynthesis appears to be
strongly partitioned between inner and outer regions,
nitrogenase is not. The implications are that when inner
regions of aggregates are photosynthetically inactive
(and hence O2 subsaturated), N2 fixation can proceed
there. Energy needed for supporting N2 fixation would
be supplied by photosynthetically active cells near
(b) the outer regions transporting reductant to the inner
regions along the filaments, which was shown by
autoradiography (149). These studies provide evidence for
spatial partitioning of these processes, which is promoted
by aggregation. Interestingly, Paerl (149) demonstrated
that even single Trichodesmium filaments were able to
partition photosynthesis along their length, with central
regions having lowest rates. However, N2 fixation and
photosynthesis only co-occurred at very low irradiance
levels (less than 100 µE m−2 second) in single filaments,
while aggregates were able to simultaneously conduct
these processes at irradiances exceeding 500 µE m−2
second. Aggregation therefore appears advantageous by
enabling this genus to optimize these essential processes
in radiant energy-rich, oligotrophic surface waters.
Figure 6. Surface-dwelling aggregates of the filamentous marine As observed among Nodularia aggregates in the Baltic
planktonic N2 fixing cyanobacteria Trichodesmium spp. Upper Sea, Trichodesmium aggregates host diverse bacterial,
frame: radial, or ‘‘puff’’-shaped aggregates, sampled during a microalgal, and protozoan epibionts (154,155). Bacteria
bloom in the Gulf Stream, W. Atlantic Ocean, off the coast of associated with Trichodesmium are actively engaged
North Carolina. Lower frame: fusiform ‘‘tuft’’-shaped aggregates in assimilating organic compounds, including amino
sampled in the W. Atlantic Ocean near San Salvador Island, acids and sugars known to be excreted by the ‘‘host’’
Bahamas. For more details on the N2 fixing characteristics, and
cyanobactria (139,155). Bacteria isolated from naturally
microbial ecology of these aggregates see Paerl (149). See color
insert. occurring Trichodesmium aggregates are mostly aerobic
heterotrophs. Some reveal chemotactic behavior, being
attracted to nitrogen-containing amino acids (155), reflect-
Paerl and Bland (102) demonstrated, using tetrazolium ing Trichodesmium excretion products (139). Culturing
dye reduction, that highly reduced (O2 depleted) regions experiments show high Trichodesmium growth and N2
were present in actively N2 fixing Trichodesmium fixation rates in the presence of these bacteria. These find-
aggregates obtained from the western Atlantic Ocean (Gulf ings suggest mutualistic consortial associations, in which
Stream). Using O2 microelectrodes, Paerl and Bebout (147) organic excretion products from Trichodesmium support
directly measured small (µm) O2 -deplete internal regions the growth of associated bacteria, while these bacteria
in Trichodesmium aggregates. Collectively, these studies may promote Trichodesmium growth. Factors potentially
showed that photosynthetically active Trichodesmium involved in stimulating ‘‘host’’ Trichodesmium growth
aggregates could simultaneously exhibit high and low pO2 include localized respiratory O2 consumption, nutrient
microzones. (phosphorus, CO2 recycling), and exchange of beneficial
Recently, Fredricksen and coworkers (152) showed bioactive compounds, such as vitamins and chelators.
partitioning of nitrogenase and photosynthetic activities An additional functional role of surface-dwelling Tri-
within single filaments of both naturally occurring chodesmium aggregates is their possible ability to inter-
and cultured Trichodesmium populations, suggesting cept atmospheric nutrient inputs. Iron (Fe) is a nutrient
spatial segregation of these processes without the aid required as a cofactor in enzymes and energy transfer-
of heterocysts. Microautoradiographic examinations of ring molecular mediating the synthesis and activities of
14 the photosynthetic and nitrogen fixing machinery. This
C−CO2 fixation under both natural and artificial
irradiance conditions (149) showed internal regions of metal is known to be in short supply (i.e., limiting) in
aggregated filaments to be far less photosynthetically- the tropical open oceans (156). Field and laboratory stud-
active (and hence having low O2 evolution rates) than the ies (116) indicated that iron supply controlled N2 fixation
170 AGGREGATES AND CONSORTIA, MICROBIAL

and growth in Trichodesmium blooms off the North Car- (a)

olina coast (Gulf Stream). The dominant source of new
iron in these as well as open ocean waters is atmospheric,
either via dry or wet deposition (157,158). Rueter and
coworkers (159) hypothesized that, by forming surface-
dwelling aggregates, Trichodesmium enhances its ability
to intercept iron associated with small particles (dust,
aerosols) deposited on the ocean surface. Indeed, there
appears to be geographic overlap between relatively high
frequencies and densities of Trichodesmium blooms and
areas of iron deposition downwind of iron-enriched African
Saharan dust storms the Atlantic Ocean between 20° N and
the equator (160). Alternatively, the concentrated blooms
may simply arise from lengthy periods of calm seas state
(i.e., ‘‘horse latitudes’’) periodically encountered in this
(b)
region (160).

CYANOBACTERIAL MATS IN A TEMPERATE MARINE

ECOSYSTEM: BIRD SHOAL, NORTH CAROLINA

Microbial mats are common features of shallow water

marine environments. Coastal (Atlantic Ocean) North
Carolina is blessed with expansive lagoonal, intertidal
mud- and sandflat habitats supporting mats. Bird Shoal
(∼220 ha), located near Beaufort, NC (34° 40 N, 74° 42 W)
forms the southern edge of the Outer Banks barrier
island system ringing the major estuaries and Pamlico
Sound, the second largest estuarine complex in the United
States. Microbial mats on Bird Shoal are mainly located
in the high intertidal on a protected, siliclastic sandflat. Figure 7. Laminated microbial mat obtained from the Bird
They are inundated twice daily with coastal Atlantic Shoal, coastal North Carolina. As is the case with most marine
Ocean water. The climate is temperate to subtropical, intertidal mats, microalgal biomass and production are domi-
with air temperatures ranging from an extreme low of nated by cyanobacteria. The cyanobacteria are concentrated in
−5 ° C in the winter to temperatures often exceeding the upper, illuminated segment of the mat. The mat thickness
30 ° C in the summer. Water temperature ranges from is approximately 0.5 cm. Upper frame: cross-section of the mat,
12 to 30 ° C. Irradiance can reach up to 2,000 µE m−2 s−1 showing distinct microbial laminations. Phototrophic cyanobac-
teria dominate the upper, euphotic zone of the mat. Below
during summer, when extended periods of low tide
that is a layer composed of photosynthetic bacteria, cyanobac-
occur at midday, and evaporation greatly elevates the
teria, and heterotrophic bacteria. The lowest layer contains
porewater salinity. Owing to this evaporation, mats obligate anaerobic heterotrophs, including sulfate-reducing bac-
may also experience desiccation. Mats in the supratidal teria. Methanogenic bacteria can also be found in this layer. The
may become desiccated for extended periods (days to mat is situated on highly porous sands, which allow for rapid
weeks). The mats go through a dramatic diel cycle of exchange of sweater and oxygen. Hence, the lighter oxygenated
O2 supersaturation during the day to anoxia at night. appearance of the lower sand layer. The tip (∼100 µm diameter)
Ambient waters are nitrogen-limited year-round (161). of an oxygen microelectrode is shown for size reference. Lower
The impact of major storms such as hurricanes and frame: photomicrograph of the mat matrix. Shown are nonhete-
nor’easters is unknown, but is likely to have a strong rocystous filamentous cyanobacteria (Microcoleus chthonoplastes)
and a variety of bacteria, including microcolonies of purple pho-
structuring effect.
tosynthetic bacteria (Chromatium sp.). See color insert.
The Bird Shoal mats are highly laminated, with resul-
tant strong vertical biogeochemical (i.e., pO2 , Eh ) zona-
tion (91–93) (Figs. 4 and 7). The upper layers of the
a (163,164). Ambient waters are nitrogen-deplete, leading
mat are dominated by non-heterocystous, filamentous
to year-round nitrogen-limitation of planktonic and ben-
cyanobacterial genera, including Microcoleus, Lyngbya,
thic primary production (93,161). Therefore, mats must
and Oscillatoria, and coccoid forms (e.g., Synechocys-
tis, Synechococcus) (91,93). Other filamentous, nonhete- rely on N2 fixation to satisfy microbial nitrogen require-
rocystous cyanobacterial genera, including Phormidium, ments (91,165). Recent estimates of nitrogen inputs by
Arthrospira, and unicellular morphotypes are present microbial N2 fixation indicate that this process meets at
to a lesser extent. Diatoms are present in the surface least half of the phototrophic community’s demand for
layer, but they comprise less than 15% of total microal- nitrogen, the remaining nitrogen being supplied through
gal biomass (162). Purple sulfur bacteria are occasionally efficient nitrogen recycling within the mat (166). NifH phy-
observed below the green cyanobacterial layer and pig- logenetic analysis of genomic DNA extracted directly from
ment analyses indicate the presence of bacteriochlorophyll the mats and from cultured mat isolates has demonstrated
 AGGREGATES AND CONSORTIA, MICROBIAL 171

a diverse assemblage of diazotrophs (86,167). These organ- fixation, indicative of both taxonomic and metabolic
isms include filamentous nonheterocystous cyanobacteria, diversity (87,93).
microaerophilic beta/gamma proteobacteria, and anaer- The observed rates and patterns of N2 fixation (nitro-
obic bacteria similar to delta SRB. The diazotrophic genase activity) appear to be dependent on complex
cyanobacteria genera include Lyngbya and Oscillatoria. interactions between seasonality, mat development, and
Interestingly, the numerically dominant filamentous community structure. In the winter, NA rates are low, but
cyanobacterium M. chthonoplastes does not appear to exhibit maximum activity during the day. In the summer,
possess the genetic (nifH) ability to fix N2 (87). Instead, it average rates increase greatly. In less developed mats dur-
supports bacterial N2 fixing epibionts that most likely ing the summer, substantial daytime NA may be observed,
utilize organic matter excreted by this cyanobacterial with maximum rates observed just after sunrise (93,166).
‘‘host’’. In addition to being a major source of organic In more developed mats during the summer, daytime NA
matter supporting consortial N2 fixation, Microcoleus is greatly repressed and maximum rates occur sometime
forms the ‘‘fabric’’ that consolidates the mats and during the night (Fig. 8). Results from the experiments
maintains structural integrity, while providing numerous designed to assess the inhibition of total photosynthesis,
microaerophilic and anaerobic niches for metabolically oxygenic photosynthesis, and protein synthesis suggest
diverse bacteria involved in nutrient cycling. In this that different physiological strategies translate into con-
manner, the presence of a highly productive non-N2 fixer trasting diel NA patterns. Additionally, the results infer
provides adequate organic carbon to support consortial that multiple physiological strategies are utilized to sus-
N2 fixation and complete nutrient cycling (regeneration of tain daytime NA during the summer. The inhibitory effect
existing nitrogen and other nutrients), thereby enabling of O2 on NA occurs very early. Evidence suggests that
the mat to remain highly self-sufficient in a nutrient de novo protein synthesis is required to sustain day-
deplete, periodically hostile and environmentally extreme time NA. Reverse transcriptase-polymerase chain reaction
environment. The major N2 fixing taxa in these mats (RT-PCR) results demonstrate that organisms other than
can be found in other temperate shallow-water intertidal cyanobacteria are at least actively transcribing the func-
marine habitats having mats (Tomales Bay, California; tional genes for N2 fixation (166). They further imply that
Sippewisett Marsh, Massachusetts) (167). In coastal transcription of nitrogenase for different organisms may
North Carolina, mats are present year-around, but reveal occur as environmental conditions within the mat change.
distinct seasonal patterns in community composition, Mats are an important source of organic matter
photosynthetic performance, and diel patterns of N2 (i.e., primary production) and fixed nitrogen in what

100 Apr 16, 1997

Jul 10, 1997
Jun 30, 1998
80
nitrogenase activity
Percent maximum

60

40

20

0
06:00 09:00 12:00 15:00 18:00 21:00 00:00 03:00 06:00
Time of day

60 350
Max. activity
Total NA 300
50
(nmol C2H2 red cm−2 h−1)
Nitrogenase activity (NA)

250
(C2H2 red cm−2)
Total daily NA

40 Figure 8. Observed diel variation in N2 fixation (nitroge-

200 nase activity) in the Bird shoal microbial mats, Beaufort,
30 North Carolina. The diel variation observed on April 16,
150 1997 depicts typical late wintertime observations. Highest
specific nitrogenase activity is generally observed during
20
100 midday. Diel variations observed on July 10, 1997 and
June 30, 1998 depict diel nitrogenase activity variations
10 50 observed from late spring through fall. Highest specific
nitrogenase activity is generally observed during very low
0 0 light periods near sunrise. Values are shown as percentage
Apr 16, 1997 Jul 10, 1997 Jun 30, 1998
of maximum nitrogenase activity that was observed during
Date diel experiments. (Data from 166)
172 AGGREGATES AND CONSORTIA, MICROBIAL

are often very nutrient-deprived (oligotrophic) coastal (a)

regions (91,113,165). They represent an important nutri-
tional source for micro- and macroinvertebrates that are
critical food sources for a variety of fish species that uti-
lize lagoonal marsh environments as nurseries and refugia
from predators. The full biogeochemical and trophic impor-
tance of microbial mats is only now beginning to be
understood.

STROMATOLITIC MATS OF HIGHBORNE CAY, EXUMA

SOUND, BAHAMAS

Stromatolites are highly laminated, lithified (CaCO3 ) rock

formations that are at least partially formed by microbially
mediated CaCO3 precipitation (60). Fossil stromatolites,
(b)
dating back 2.5 billion years, are among the oldest
evidence for organized life on Earth (59). The stromatolites
of Highborne Cay are subtidal along 2.5 km of an
eastward-facing beach facing Exuma Sound, Bahamas.
They form as ridges and columnar heads up to a half
a meter height. The ambient waters range in salinity
from 35 to 37 psu and are saturated with respect to
calcite and aragonite (60). Surface irradiance can exceed
2,200 µE m−2 s−1 . Their depth at low tide is approximately
one and a half meters. Water temperature ranges from 21
to 30 ° C. The stromatolites are protected by a fringing reef
complex, but are still subjected to substantial wave energy.
Ambient nutrient levels (nitrogen, phosphorus, trace
metals) are extremely very low. Laminated cyanobacteria-
(c)
dominated mats cover the surface of the stromatolites
(Fig. 9) and their metabolic activity is thought to be
responsible for the formation of lithified laminae and
subsequent buildup of the stromatolites (60). These mats
constitute the dominant source of ‘‘new’’ production and
nitrogen input to this nutrient deprived ecosystem. The
internal microenvironment is characterized by vertical
O2 concentration gradients that are typical of microbial
mats (168,169).
The phototrophic community is dominated by non-
heterocystous, filamentous cyanobacteria Schizothrix spp.,
but the biomass is very low (Pinckney, unpublished).
Other prevalent nonheterocystous cyanobacterial genera
include Phormidium, Lyngbya, and Microcoleus. Unicel-
lular cyanobacteria, such as Solentia spp. are also com- Figure 9. Stromatolitic mat communities on the Exuma Islands,
monly observed. Diatoms are present, but not abundant. Bahamas. Upper frame: view of the subtidal stromatolites
NifH phylogenetic analysis has yielded sequences most situated off a beach at Highborne Cay, Exumas. Middle
frame: cross-sectional view of a laminated (CaCO3 ) stromatolite,
similar to unicellular cyanobacteria, filamentous, non-
showing the cyanobacteria-dominated mat residing on the
heterocystous cyanobacteria, including Lyngbya sp. and surface of the stromatolite. Lower frame: photomicrograph,
Phormidium sp., alpha proteobacteria, and anaerobic bac- showing the aggregated filamentous cyanobacteria (Schizothrix
teria similar to SRB. sp.) comprising a bulk of the surface mat (from 104). See color
Rates of N2 fixation show seasonality, but diel patterns insert.
do not. In both late winter and summer, most N2 fixation
is confined to nighttime, but summertime rates are much
higher. However, the ratio of N2 fixed to net primary different effects on N2 fixation rates depending on the
production (CO2 fixation) is relatively low. Net production season, suggesting contrasting physiological controls on
values in Highborne Cay stromatolites for comparable this process at different times of the year.
times of the year range from one-third to a little over There are close associations between cyanobacteria and
half the values for temperate Bird Shoal mats. Maximum bacterial epibionts that appear to play an important role
N2 fixation values are an order of a magnitude less in the layered deposition of CaCO3 that characterizes
than what has been observed in the Bird Shoal mats. stromatolites. Microscopic and microautoradiographic
Inhibition of protein synthesis by chloramphenicol had observations indicate that the dominant cyanobacterial
 AGGREGATES AND CONSORTIA, MICROBIAL 173

taxa (most notably Schizothrix and Lyngbya) produce (a)

copious amounts of sheath and exopolysaccharide (EPS)
materials (60,105). These materials can strongly bind
calcium ions and by doing so prevent CaCO3 precipitation
and lithification (60). Bacteria that are embedded in
the sheath and EPS decompose this material, thereby
liberating Ca++ ions into solution. These Ca++ ions are
then rapidly precipitated, allowing lithification to proceed.
Microscopic investigation reveals bacteria encapsulated in
CaCO3 (aragonite), while the filamentous cyanobacteria
remain free of precipitates as long as they are covered
by sheaths and EPS (104). Thus, rather than being
directly mediated by cyanobacterial precipitation, the
lithification of stromatolitic mats is indirectly controlled
by cyanobacteria, which produce organic matter that is
mineralized by bacteria mediating precipitation (104). The (b)
ecological and biogeochemical significance (and underlying
physiological mechanisms) of this consortially mediated
lithification process require further clarification.

MICROBIAL MATS IN HYPERSALINE LAKES: CONSORTIA

‘‘AT THE EDGE OF LIFE’’

The Bahamian hypersaline lakes are among the most

extreme aquatic environments on Earth. These lagoonal
ponds are isolated from the nearby oligotrophic oceans
(Atlantic and Caribbean) by narrow dunes and carbon-
ate ridges. They are exposed to hot (sometimes exceed-
ing 40 ° C), arid and extremely high irradiance (greater
Figure 10. Upper frame: hypersaline (100 to >160 psu) Salt
than 2,200 µE m−2 s−1 ) conditions throughout much of the
Pond, located on San Salvador Island, Bahamas. The entire
year. Annually, evaporative water loss greatly exceeds bottom of this lake is dominated by laminated mats dominated
rainfall input, leading to persistent hypersaline condi- by filamentous, nonheterocystous cyanobacteria (Microcoleus sp.,
tions exceeding 150 psu (seawater salinity is ∼35 psu). Lyngbya sp., and Oscillatoria sp.). Lower frame: close-up side
Because rainfall is extremely low and restricted to trop- view pieces of laminated mat excised from the pond. See color
ical storms of oceanic origin and surrounding carbonate insert.
soils are highly leached, nutrient (nitrogen, phosphorus,
iron, and trace metal) inputs to these lakes are very
small and restricted to aeolian sources (i.e., dust). As a
Chl a cm−2 , which is very high by benthic microalgal
result, the lakes exhibit a broad spectrum of environ-
standards (164,170). Biomass is highly compacted in the
mental extremes and limitations. Despite these hostile
surface layer of the mat, causing rapid light extinction
conditions, the lakes support an abundant microbial
with depth. Photosynthetically active radiation (400 to
community, largely confined to microbial mats lining
800 nm) decreases from greater than 2,200 µE m−2 s−1 at
the lake bottoms (164). The nitrogen requirements of
these mats are largely met by N2 fixation and nitro- the surface to undetectable levels at 4 to 5 mm depth.
gen regeneration in the mat matrix (166). San Sal- This leads to a highly compressed euphotic zone with
vador Island, Bahamas (24° 00 N, 72° 05 W) has numer- strong vertical biogeochemical (O2 , Eh ) zonation; typically,
ous hypersaline lagoonal systems, including: (1) Storr’s during daylight the upper 1 to 2 mm of the mat is O2
lake (45 → 90 psu), which supports calcifying, stroma- supersaturated, while the lower 3 to 10 mm is permanently
tolitic mats dominated by the heterocystous cyanobacterial anoxic and sulfidic. A layer of anoxic photosynthetic
genus Scytonema, nonheterocystous Schizothrix, and coc- bacteria (Chromatium, Rhodopseudomonas) is present at
coid forms Synechocystis, Synechococcus; and (2) Salt Pond the oxic–anoxic interface. The O2 gradients spanning
(45 → 100 psu), which contain a highly laminated, non- oxic and anoxic zones are dynamic and light attenuation-
calcified cyanobacterial mat dominated by nonheterocys- dependent, varying with time of day and season.
tous filamentous Microcoleus, Lyngbya, and Oscillatoria Both light- and dark-mediated N2 fixation occur in
species (Fig. 10). these mats. Rates are generally high (relative to pho-
The hypersalinity is inhibitory to both photosynthesis tosynthesis (162,166,170). Both oxygenic (cyanobacterial)
and N2 fixation (164,170), and may play a role in and anoxygenic (photosynthetic bacteria) photosynthesis
selecting for cyanobacterial species, because eukaryotic contribute significantly to mat primary production, and
microalgae are confined to only a few diatom species are a dominant source of ‘‘new’’ carbon supporting the
that constitute less than 10% of the total microalgal lake’s unusual (extensive microbial loop and invertebrate
biomass. Cyanobacterial biomass often exceeds 50 nmol grazing, but no fish) food web (170).
174 AGGREGATES AND CONSORTIA, MICROBIAL

AGGREGATES AND MATS IN THE PERMANENT ICE COVER (a)

OF LAKE BONNEY, MCMURDO DRY VALLEYS,
ANTARCTICA

The McMurdo dry valleys of Southern Victorialand,

Antarctica (160–164° E, 76° 30
–78° 20
S) contain a series
of glacial lakes having permanent ice covers varying in
thickness from approximately 3 to over 6 m. The dry
valleys are among the highest, driest, coldest, and most
nutrient-deprived habitats on Earth, with wintertime air
temperatures ranging from −40 to less than −60 ° C,
and summer temperatures ranging from −20 ° C to
near 0 ° C. Because the dry valleys are surrounded by
greater than 3,000 m mountains, they are in a ‘‘snow
shadow,’’ resulting in annual snowfall of only 2 to
(b)
4 cm. Under these conditions, evaporation (sublimation)
greatly exceeds precipitation, leaving the soils of the
valley floor exposed (51). As a result, the dry valley area
forms the largest expanse of ice free soils (∼4,000 km2 )
on Antarctica. Liquid water is confined to a short
summer period (December to early February), when
air temperatures and solar heating are high enough to
generate meltwater from permanent glaciers that reach
the valley floor. Meltwater enters the lakes via narrow
streams, which flow for only about 2 to 3 months.
However, much of the rest of the valley floor is frozen,
desiccated desert soil. Under these extremely cold, water-
stressed conditions, prokaryotic life largely exists as
cyanobacterial-bacterial aggregates and mats associated
with the ice cover surface and embedded in the ice
(c)
matrix (51–53).
Microbial consortia associated with the ice cover
include desiccated mats around the edge of the ice
cover, and aggregates associated with desert soil particles
that are periodically dispersed over the ice surface
by strong adiabatic and catabatic winds sweeping
this region (42,43). Soil particles are very dark and
absorb enough radiant energy to warm and melt their
way down into the ice matrix during the austral
summer (Fig. 11). These particles aggregate and tend
to accumulate in distinct layers in the ice cover. The
depths of the layers depend on the extent to which
solar heating can maintain molten conditions around the
particles. During relatively clear periods, more extensive
melting and deeper layers result. A cyanobacterial
community is associated with the soil particles (Fig. 11). Figure 11. Soil-based microbial aggregates in the permanent ice
cover of Lake Bonney, McMurdo Dry Valley, Antarctica. Upper
This includes nonheterocystous genera (Phormidium,
frame: surface view of the ice cover, showing the cracks in which
Oscillatoria), at least one heterocystous genus (Nostoc)
windblown soil accumulates. Middle frame: close-up view of the ice
and coccoid forms (Synechococcus spp.) (52,171). During surface, showing localized melting of ice by dark soil aggregates.
at least nine months of the year (fall through late Eventually, this melting activity causes soil aggregates to settle
spring), ice-embedded aggregates exist in a desiccated, into the ice matrix. Lower left frame: a core taken from the
‘‘freeze dried’’ state. Rehydration occurs during the ice cover, showing soil aggregates embedded in the ice matrix.
summer period (late November to February) when Lower right frame: microautoradiograph of 14 CO2 uptake by
localized solar heating of aggregates leads to localized microorganisms associated with a soil aggregate sampled from
melting, followed by activation of metabolic processes the ice matrix. The darkened filaments are photosynthetically
and growth (51,52). When desiccated aggregates are active cyanobacteria attached to a soil particle (from 52 and 103).
See color insert.
experimentally rehydrated, photosynthesis, heterotrophic
activity, and N2 fixation can be detected within a few
hours.
The cyanobacterial-bacterial consortia that are associ- aggregate microenvironment. Following rehydration, exu-
ated with aggregates physically and chemically modify the dation of organic matter takes place and redox gradients
 AGGREGATES AND CONSORTIA, MICROBIAL 175

are established in response to microbial photosynthetic compounds) at individual microorganism and biogeochem-
O2 production and heterotrophic O2 consumption along ical gradient levels (57). Complementary techniques, such
microscale nutrient, light, and organic matter gradients. as process-specific (i.e., CO2 and N2 fixation, nitrifica-
Ice aggregates represent structural and functional biocom- tion, denitrification) immunofluorescence assays can be
plexity in the form of interactive production and nutrient used to visualize those cells actively expressing the struc-
cycling consortia ensuring microbial self-sustenance at the tural genes (nitrogenase; NifH, ribulose 1,5-diphosphate
‘‘edge of life’’ on Earth. Molecular studies of aggregate con- carboxylase; RubisCo, nitrate reductase; Nir) and hence
sortia (171,172) indicate a diverse assemblage, including capacities for enzyme function (176–178).
some of the most ancient (i.e., Precambrian) known forms Experimental approaches to examining the in situ
of life on Earth. impacts of O2 tension on aggregate and mat production and
nutrient cycling processes include the use of inhibitors of
CHARACTERIZING AGGREGATE AND MAT CONSORTIAL oxygenic photosynthesis (photosystem II). Among these, 3-
STRUCTURE AND FUNCTION (3,4-dichlorophenyl)-1,1-dimethylurea (DCMU) at greater
than 10−5 M effectively blocks PS II activity, leading
Ecophysiological Approaches to a cessation of O2 evolution (and CO2 fixation) with-
The past two decades have seen significant develop- out altering respiration (i.e., O2 consumption) (91,114).
ments and improvements in aggregate and mat consortial Studies (57,58,92,93) have shown that when mats are pho-
rate measurement techniques that are both appropriately tosynthetically active (i.e., high rates of O2 evolution),
scaled and sensitive. There have been breakthroughs in DCMU additions arrest O2 production and CO2 fixa-
the areas of microsensing (i.e., microelectrodes), micro- tion (93), with concurrent stimulation of N2 fixation rates,
chemical, and molecular techniques that have greatly indicating that, within the aggregate and mat matrix,
expanded the ‘‘toolbox’’ microbial ecologists, chemists, in situ O2 evolution rates are high enough to negatively
hydrobiologists, geologists, and molecular biologists can impact O2 -sensitive N2 fixation. However, when photo-
deploy for examining aggregate and consortial structure synthetic rates are low (i.e., wintertime), or at nighttime
and function. Consortial structure and function are most when photosynthesis ceases, DCMU has little impact on
commonly analyzed in terms of (1) production and trans- diazotrophy. These results indicate N2 fixation may be
formation rates (and their interactions), (2) community restricted by increased O2 levels accompanying periods of
composition and structure, and (3) biogeochemical gra- active photosynthesis.
dient and localization studies. Establishing diurnal and To compensate for this metabolic incompatibility, dia-
seasonal rate patterns of key production and nutrient zotrophs have evolved various morphological and physi-
cycling processes, such as CO2 and N2 fixation, denitrifi- ological adaptations, including (1) temporally separating
cation, and sulfate reduction reveal how carbon, nitrogen, photosynthesis from N2 fixation, by confining the former to
sulfur, and other essential elements are cycled and allo- daytime and the latter to nighttime; (2) spatial separation
cated in aggregates and mats in response to long- and of the two processes, with photosynthesis concentrated
short-term physicochemical influences. Microscale sensing near the illuminated surface of aggregates and mats
and chemical characterization techniques are detailed by and diazotrophy taking place in deeper, aphotic layers;
Jørgensen (9), Revsbech and coworkers (168), Canfield and (3) formation of localized O2 deplete microzones through-
Des Marais (90), Rysgaard coworkers (173), and Stal and
out the matrix, in which R exceeds P; and (4) confinement
Caumette (55), while molecular applications have recently
of the N2 fixation to O2 devoid cells (heterocysts) or nonpho-
been reviewed by Amann (174), Nielsen and cowork-
tosynthetic regions of filaments and colonies. Despite the
ers (175), Zehr and Paerl (145), and Paerl and Zehr (113).
obvious advantage of forming heterocysts, many aggregate
Aggregates and mats reflect broad taxonomic and
and mat systems in nitrogen deplete waters are devoid of
physiological diversity and plasticity, which interacts
heterocystous cyanobacteria (e.g., Trichodesmium aggre-
with biogeochemical gradients. Gradients in large part
gates, many intertidal mat systems) (55,113). This ironic
reflect the combined product of O2 production (P) versus
consumption via respiration (R) in time and space (10). situation is a puzzling aspect of cyanobacterial evolu-
Gradients are controlled by the concentrations of dissolved tion, that is, why do microbial mats, biofilms, aggregates,
O2 and salinity in water surrounding aggregates and mats and other surficial communities in nitrogen limited oxic
and the internally generated P/R ratios. O2 gradients can waters and soils not contain more heterocystous genera?
be experimentally manipulated, allowing for examinations The resolution to this dilemma should yield key insights
of the impacts of pO2 on aggregate and mat structure and into ecophysiological controls of microbial N2 fixing con-
function. sortia.
Oxygen, pH, and H2 S gradient fluctuations can occur Consortial composition has recently been assessed
on temporal scales of seconds to minutes and spatial scales by bulk community analysis (DNA and photopigment)
of microns to millimeters; spatial scales reflecting consor- and specific gene expression (mRNA) (discussed later).
tial dimensions. Microelectrode analyses and tetrazolium Chemosystematic carotenoids and chlorophylls, quanti-
redox salt assays have been used to determine chem- fied using high-performance liquid chromatography have
ical gradients on these scales (89,92,102). In addition, been used as indicators of relative abundance of major pho-
microautoradiography allows visualization of those organ- totrophic groups (162,179). 16S rRNA or rDNA sequences
isms responsible for primary production (CO2 fixation) and are considered accurate measures of microbial diver-
dissolved organic matter (DOM) uptake (of radiolabeled sity (180). Genes encoding for biogeochemically important
176 AGGREGATES AND CONSORTIA, MICROBIAL

and evolutionarily conserved processes, including N2 fixa- small scales and sample sizes needed to adequately assess
tion, nitrification, and denitrification are receiving increas- microzonal consortial processes.
ing scrutiny for phylogenetic and physiological studies. Small subunit rRNA (16S) sequence analysis is
Their application is discussed later. In the case of primary routinely used for the identification and phylogenetic
production (CO2 fixation), the ribulose-1,5-carboxylase placement of microorganisms without the need for cul-
gene (RubisCo) is responsible for virtually all primary tivation. Members of the bacterioplankton that have not
production in the world and the large subunit, rbcL, been previously described have been identified (186,187),
is evolutionarily well conserved (181). Examining rbcL, and their diversity and distribution have been inves-
nifH, amoA, and nirSK expression and then correlating tigated (188–190). New members of the Archaea have
the expression with patterns of CO2 fixation, N2 fixation, also been identified in the water column (191,192) and
denitrification, and nitrification will allow us to identify in sediments (193,194), and their spatiotemporal variabil-
consortial members responsible for these processes, envi- ity has also been assessed (190,195). Functional genes
ronmental factors controlling genetic expression of key encoding for enzymes directly involved in the transfor-
functional genes, and how they are manifested in terms mation processes have also been used to characterize
of community compositional or physiological responses natural microbial assemblages. Target genes involved
(Fig. 12). in denitrification (nosZ) (185), nitrification (amoA) (196),
nitrogen fixation (nifH) (197,198), and the oxidation of
Molecular Approaches methane (pmoA) (199) have been successfully used. More-
over, the sequences for the dissimilatory sulfate reductase
A serious limitation to an improved understanding of the genes (dsr) have been recently obtained from different
‘‘players’’ involved in microbial consortia has been the lack organisms (200,201), and dsr-specific PCR primers have
of availability of quantitative identification and characteri- been developed to track sulfate-reducing microorganisms,
zation techniques. Standard microbiological approaches to which play a central role in the global sulfur cycle. Detec-
the identification of individual populations within micro- tion of target genes such as 16S, amoA, nosZ, and nifH can
bial consortia, such as selective culturing, require the be coupled with high throughput techniques, such as ter-
extensive knowledge of microzonal environmental condi- minal restriction length polymorphism analysis (TRFLP)
tions. This information is not easily determined along a to characterize consortia in a timely, quantitative, and
natural gradient. Furthermore, environmental conditions unequivocal manner.
are often transient, depending on the balance between The utility of highly conserved functional genes
oxygen production and consumption, nutrient uptake and for microbial consortial characterization studies can be
release, and physical control processes, such as turbu- readily demonstrated for nifH. Because studies of nifH-
lence and diffusion, all of which operate on highly variable based diversity and phylogeny have been underway
temporal and spatial scales. As a consequence, only a for at least a decade, a substantial database on
small portion (less than 1%) of microorganisms from the taxa-specific sequence regions of interest has been
environment can be grown using routine culturing tech- established (86,87,145,167,172,197,198). This database
niques (182–185). can then be used to establish phylogenetic trees,
Determination of functions, including rates of basic which serve as references against which established
metabolic processes, such as photosynthesis, respi- as well as novel sequences derived from N2 fixing
ration, utilization of organic matter, and nutrient consortia can be compared and characterized according
uptake/assimilation have traditionally been limited by the to specific functional groups (i.e., heterotrophs, such as

Sample

Rate analysis Community analysis Consortial characterization

DNA isolation Pigment extraction RNA isolation Microautoradiography:

Primary production:
H 14 CO − Uptake H14 CO3−
3
O2 flux
labeled DOM
Labeled DOM uptake PCR HPLC Dot blot RT-PCR In Situ:
Diazotrophy: Nitrogenase (NifH)
Acetylene reduction immunofluorescence

16S rDNA RubisCo RubisCo

+ + +
nifH nifH nifH
Figure 12. Schematic diagram showing complementation of rate measurements, diagnostic
indicators of microalgal community composition with nucleic acid determinations of functional
(i.e., nifH), phylogenically important (16s rRNA) genes for determining the structure and function
of microbial consortia in natural samples.
 AGGREGATES AND CONSORTIA, MICROBIAL 177

nifH phylogenetic tree

reducers
Sulfate
Desulfovibrio gigas
Desulfonema limicola

anaerobes
Desulfobacter curvatus
Bird Shoals microbial mat
Bird Shoals microbial mat
Clostridium pasteurianum
Clostridium sp.
Clostridium cellobioparum
Antarctic isolate

proteobacteria
Rhodobacter capsulatus
Rhodospirillum rubrum

alpha
Rhizobium sp.
Rhizobium meliloti
Bradyrhizobium sp.

aerophiles
Highborne Cay isolate

Micro
Frankia alni
Frankia sp.
Microcoleus chthnonplastes
Trichodesmium thiebautii
Fisherella sp.
Crescent L., FL Heterocystous
Neuse River, NC
Anabaena oscillariodes
Anabaena azollae
Nostoc commune
Phormidium

cyanobacteria
Highborne Cay isolate
Crescent Lake, FL
Cyanothece
Dermocarpa sp.
Xenococcus sp.
Myxosarcina sp. Non
Plectonema boryanum Heterocystous
Calothrix sp.
Bird Shoals microbial mat
Neuse River, NC
Lyngbya lagerheimii
Synechococcus sp. Figure 13. Phylogenetic tree of nitrogen
Gloeothece sp. fixing microbial groups, many of which
Klebsiella pneumoniae participate in consortial associations. The
methano- beta and gamma

tree was based on analysis of a 326

proteobacteria

Vibrio diazotrophicus
Azotobacter chroococcum base pair nucleotide fragment of the
Azotobacter vinelandii nifH gene, which encodes for the highly
Alcaligenes faecalis conserved dinitrogenase reductase subunit
Chromatium purpuratum of nitrogenase, the N2 fixing enzyme
Bird Shoals microbial mat complex. The tree was constructed by
Bird Shoals microbial mat the neighbor-joining method and bootstrap
Methanobrevibacter arboriphillicus
gens

values greater than 50% are given above or

Methanococcus janneschii beside the corresponding nodes. See color
insert.

sulfate reducers, chemolithotrophs, coccoid, filamentous- nutrient transformations using TRFLP analysis of clonal
heterocystous, heterocystous cyanobacteria) (Fig. 13). libraries to minimize the time necessary to identify spe-
While molecular characterization of target genes, such cific target gene clones. For example, TRFLP analyses
as nifH and 16S rRNA, is a routine procedure in many have been used to screen for denitrifying bacteria in sedi-
microbial ecology laboratories, the procedure is slow, ments from salt marshes, fresh water marshes, and coastal
laborious, and potentially prone to biases. These biases sediments (207).
are being overcome by developing highly reproducible Specific target genes (and the nutrient transformations
extraction procedures (202) and quantitative nucleic acid they catalyze) are available for characterizing aquatic
characterization methods (202–204). Recently, rapid ‘‘fin- microbial consortia. These biomarkers can be used to
gerprinting’’ methods, utilizing fluorescent end-labeling assess the numbers and activity of various consortial
of PCR product (target genes) and screening by ter- members, using TRFLP analysis. For example, one way to
minal restriction length polymorphism (TRFLP; see ascertain whether bacteria involved in denitrification are
Refs. 205,206) have greatly streamlined and improved indeed active in an environment involves determining
the accuracy of these methods. Highly reproducible fin- if a functional gene (like nosZ) is being transcribed
gerprints can be obtained from complex samples. It is (i.e., mRNA synthesis) (207). Rate measurements can
now possible to rapidly identify target genes for specific then accompany transcript information to elucidate
178 AGGREGATES AND CONSORTIA, MICROBIAL

consortia members that are actively growing and hence descriptive approaches (i.e., inventories of biotic diver-
participating in nutrient transformations of interest (in sity at higher organismal levels) is multifold. These
this case denitrification). microbial parameters: (1) simultaneously allow for mea-
sures of community structure (biodiversity) and function
(rates of relevant processes, including primary produc-
CONCLUSION tion, mineralization, specific nutrient transformations,
etc.); (2) can be coupled to large-scale survey and remote
Aggregates and other surficial microzones are ecophysio- sensing tools, including aircraft and satellite imagery (of
logical, environmental, and evolutionary units of microbial ocean color, turbidity, photopigments diagnostic of certain
biodiversity (i.e., community structure) and biocomplexity microalgal groups); (3) are amenable to microchip and
(i.e., community function) in aquatic and terrestrial habi- other molecular-based, multiparameter indices of carbon,
tats. There are commonalities and redundancies embodied nitrogen, and other nutrient cycling processes, (4) have
in microzonal structure and function among the myriad a role in the detection and characterization of micro-
aquatic and soil habitats. Using a variety of examples bial pathogens, harmful bloom algae, and other indices
from geographically and geochemically distinct environ- of environmental degradation of ecosystem and human
ments, we have shown that prokaryotic consortia, able health.
to develop and colonize microscale biogeochemical gra- Now, at the beginning of the third millennium, we are
dients, are a common feature of these environments. facing unprecedented changes in Earth’s biotic resources.
Aggregate, mat, biofilm systems can be used as ‘‘mod- One of the greatest and most important challenges is being
els’’ to identify and functionally understand a basic set able to detect, quantify, and differentiate human impacts
of process-level requirements or ‘‘ground rules’’ for micro- form natural climatic and geological and geochemical
bial life to establish, maintain, and proliferate itself on a changes facing our planet. For better or worse, microbially
community organizational level. mediated changes are among the most fundamental and
Research on diverse marine, estuarine, and freshwa- important in the biosphere. We have the opportunity and
ter habitats points to aggregates and surficial microbial
responsibility to judiciously apply the tools of microbiology,
communities as biogeochemically and ecologically impor-
molecular biology, and ecology to gauge anthropogenic
tant nutrient transformation and trophic transfer sites.
and natural change, using microbial consortia as key
These sites are particularly significant in nutrient-poor
indicators.
and otherwise extreme environments, in which microbial
production and nutrient cycling are severely restricted Acknowledgments
by nutrient and energy availability, as well as other Research and logistic support were provided by the National
environmental constraints (i.e., extremely low or high Science Foundation (Projects OCE 94-15985, DEB 9815495,
temperatures, desiccation, hypersalinity, high UV light). OPP 94-19423 and LexEn Program, Project 98-08959), the
Changes in these conditions are likely to have a cas- U.S. Department of Agriculture (Project NCR-9600509), the
cading effect on the structure, function, and overall U.S. Environmental Protection Agency (Project R825243-01-0),
production and nutrient cycling characteristics of these the National Oceanographic and Atmospheric Administration
microzonal communities, with ramifications for biogeo- (NOAA/North Carolina Sea Grant Program; RMER 30), and the
Finnish Academy of Science. We appreciate the technical assis-
chemical cycling on larger ecosystem, regional and global
tance, informational input, and critical reviews from C. Buzzelli,
scales. In particular, the biocomplexity and productivity L. Kerkhof, J. Pinckney, J. Zehr, M. Go, P. Moisander, J. Dyble,
of oligotrophic open ocean waters are likely to be affected. and B. Peierls.
Contemporary issues of global change, including warming,
altered amounts and patterns of rainfall and (conversely)
drought, resultant changes in runoff and riverine dis- BIBLIOGRAPHY
charge of nutrients and sediments, coupled to human
land-use changes and associated nutrient and other pol- 1. W. Schlesinger, Biogeochemistry: An Analysis of Global
lutant loading patterns, will have profound impacts on Change, Academic Press, New York, 1997, p. 288.
the structure and function of aggregate/surficial microbial 2. T. Fenchel and T. H. Blackburn, Bacteria and Mineral
consortia. Cycling, Academic Press, New York, 1979.
Microzonal communities as described here are sen- 3. F. Azam et al., Mar. Ecol.: Progr. Ser. 10, 257–263 (1983).
sitive, metabolically active, and hence responsive indi- 4. G. E. Hutchinson, Am. Naturalist 95, 137–146 (1961).
cators of environmental change, in essence serving as 5. C. E. Zobell and D. Q. Anderson, Biol. Bull. 712, 324–342
the microbial equivalent of the ‘‘canary in the mine.’’ (1936).
As such, microzonal communities may serve as excellent 6. M. Fletcher and K. C. Marshall, in K. C. Marshall, ed.,
indices of environmental change spanning the size spec- Advances in Microbial Ecology, Plenum Press, New York,
trum from specific habitat, ecosystem, regional-coastal 1977, pp. 190–230.
to oceanic scales. Standard microbiological, molecular, 7. T. E. Ford, ed., Aquatic Microbiology: An Ecological
physiological, and analytic assessments of aggregate and Approach, Blackwell Science Publications, Oxford, U.K.,
surficial genetic composition, function, and rates of rel- 1993.
evant processes serve as the parameters of change. 8. H. W. Paerl, in T. E. Ford, ed., Aquatic Microbiology:
The advantage of using this suite of parameters for An Ecological Approach, Blackwell Science Publications,
gauging environmental change over more conventional Oxford, U.K., 1993, pp. 343–382.
 AGGREGATES AND CONSORTIA, MICROBIAL 179

9. B. B. Jørgensen, in B. Bolin and R. Cook, eds., The Major 45. J. Ikävalko, Walter Andrée de Nottbeck Found. Sci. Rep. 14,
Biogeochemical Cycles and their Interactions, John Wiley & 1–24 (1997).
Sons, New York, 1983, pp. 477–509. 46. T. Mock et al., Mar. Ecol. Prog. Ser. 158, 23–40 (1997).
10. N. P. Revsbech et al., Adv. Microb. Ecol. 9, 273–352 (1986). 47. M. Huttunen and Å. Niemi, Mem. Soc. Fauna Flora Fenn.
11. W. D. P. Stewart, Algal Physiology and Biochemistry, 62, 58–62 (1986).
Blackwell, Oxford, U.K., 1974. 48. M. Laamanen, Algol. Stud. 83, 423–433 (1996).
12. G. E. Fogg, in W. D. P. Stewart, ed., Algal Physiology and 49. P. Haecky et al., Polar Biol. 20, 1–8 (1998).
Biochemistry, Blackwell, Oxford, U.K., 1974, pp. 560–582. 50. H. Kaartokallio, Polar Biol. 24, 21–28 (2001).
13. G. E. Fogg, in N. G. Carr and B. A. Whitton, eds., The 51. J. C. Priscu et al., Science 280, 2095–2098 (1998).
biology of Cyanobacteria, Blackwell Science Publications, 52. H. W. Paerl and J. C. Priscu, Microb. Ecol. 36, 221–230
Oxford, U.K., pp. 491–513. (1998).
14. B. A. Whitton, in Fogg et al., eds., The Blue-green Algae, 53. J. L. Pinckney and H. W. Paerl, Antarct. J. US Review 1,996,
Blackwell, Oxford, U.K., 1973, pp. 415–443. 215–216 (1998).
15. H. W. Paerl, in W. Reisser, ed., Algal symbioses, Biopress 54. J. W. Costerton et al., Ann. Rev. Microbiol. 49, 711–745
Bristol, U.K., 1992, pp. 537–566. (1995).
16. H. W. Paerl in N. G. Carr and B. A. Whitton, eds., The 55. L. J. Stal and P. Caumette, NATO ASI Series G: Ecological
Biology of Cyanobacteria, Blackwell Scientific, Oxford, U.K., Sciences, vol. 35, Springer Verlag, Berlin, Germany, 1994.
1982, pp. 441–461. 56. L. J. Stal and W. E. Krumbein, Arch. Microbiol. 143, 67–71
17. H. W. Paerl and J. L. Pinckney, Microb. Ecol. 31, 225–247 (1985).
(1996). 57. H. W. Paerl et al., in Y. Cohen and E. Rosenberg, eds.,
18. N. Suzuki and K. Kato, Bull. Facul. Fish. Hokkaido Univ. 4, Microbial Mats: Physiological Ecology of Benthic Microbial
132–135 (1953). Communities, American Society for Microbiology Press,
19. A. L. Alldredge and M. W. Silver, Prog. Oceanogr. 20, 41–82 Washington, D.C., 1989, pp. 326–341.
(1988). 58. H. W. Paerl and R. G. Carlton, Nature 332, 260–262 (1988).
20. P. Weiss et al., Appl. Environ. Microbiol. 62, 1998–2005 59. J. W. Schopf and M. R. Walter, in N. G. Carr and B. A. Whit-
(1996). ton, eds., The Biology of Cyanobacteria, Blackwell Scientific,
Oxford, U.K., 1982, pp. 543–564.
21. H. Zimmerman and K. Kausch, Adv. Limnol. 48, 85–92
(1996). 60. R. P. Reid et al., Nature 406, 989–992 (2000).
22. H. Zimmermann, Aquat. Microb. Ecol. 13, 37–46 (1997). 61. H. W. Paerl, J. Phycol. 12, 432–435 (1976).
23. H.-P. Grossart et al., Aquat. Microb. Ecol. 14, 59–67 (1998). 62. H.-G. Hoppe, Kieler Meeresforsch., Sonderh. 5, 291–303
(1981).
24. G. Jackson, Deep Sea Res. 41, 429–437 (1994).
63. D. E. Caldwell and S. J. Caldwell, Can. J. Microbiol. 24,
25. A. L. Alldredge et al., Limnol. Oceanogr. 35, 1415–1428 922–931 (1978).
(1990).
64. M. W. Silver et al., Science 201, 371–373 (1978).
26. U. Passow et al., Deep Sea Res. 41, 335–357 (1994).
65. B. A. Biddanda, Mar. Ecol.: Prog. Ser. 20, 241–251 (1985).
27. P. Vandevivere and D. L. Kirchman, Appl. Environ. Micro-
66. A. L. Alldredge and C. C. Gotschalk, Cont. Shelf. Res. 10,
biol. 59, 3280–3286 (1993). 41–58 (1990).
28. U. Riebesell, Mar. Ecol. Prog. Ser. 69, 281–291 (1991b). 67. D. C. Smith et al., Nature 359, 139–142 (1992).
29. G. Jackson, Deep Sea Res. 37, 1197–1211 (1990). 68. A. L. Shanks and J. D. Trent, Limnol. Oceanogr. 24,
30. T. Kiørboe et al., J. Mar. Res. 52, 297–323 (1994). 850–854 (1979).
31. A. L. Alldredge and C. C. Gotschalk, Deep Sea Res. 36, 69. C. E. Gibson and R. V. Smith, in N. G. Carr and B. A. Whit-
159–171 (1989). ton, eds., The Biology of Cyanobacteria, Blackwell Science,
32. J.-M. Leppänen, Finnish Mar. Res. 255, 97–118 (1988). Oxford, 1982, pp. 463–490.
33. J.-M. Leppänen and K. Kononen, Finnish Mar. Res. 255, 70. M. E. Meffert, Arch. Hydrobiol. Suppl. 41, 235–24l (1972).
37–54 (1988). 71. V. Herbst and J. Overbeck, Naturwiss 65, 598–599 (1978).
34. J. Kuparinen et al., AMBIO Special rep. 8, 13–21 (1996). 72. D. E. Caldwell, in J. M. Nichols, ed., Abstracts of the
35. H. T. Kankaanpää et al., Boreal Env. Res. 2, 257–274 (1997). 3rd International Symposium Photosynthesis Prokaryotes,
Oxford University Press, Liverpool, U.K., 1979.
36. V. S. Smetacek, Mar. Biol. 84, 239–251 (1985).
73. J. Lehtimäki, Appl. Environ. Microbiol. 63, 1647–1656
37. B. E. Logan and A. L. Alldredge, Mar. Biol. 101, 443–450 (1997).
(1989).
74. J. A. Hellebust, in G. H. Lauff, ed., Estuaries, AAAS Special
38. I. N. MaCave, Deep Sea Res. 31, 329–352 (1984). Publications, Washington, D.C., 1967, pp. 361–366.
39. U. Riebesell, Mar. Ecol. Prog. Ser. 69, 273–280 (1991a). 75. A. E. Walsby, Br. Phycol. J. 9, 371–381 (1974).
40. G. T. Taylor et al., Mar. Ecol. Prog. Ser. 29, 141–155 (1986). 76. C. Nalewajko, in J. A. Hellebust and J. S. Craigie, eds.,
41. R. S. Lampitt et al., Mar. Biol. 116, 689–702 (1993). Handbook of Phycological Methods, Cambridge University
42. B. A. Biddanda and L. R. Pomeroy, Mar. Ecol. Prog. Ser. 42, Press, Cambridge, Mass., 1978, pp. 389–398.
79–88 (1988). 77. K. K. Gallucci et al., Appl. Environ. Microbiol. 45, 557–561
43. C. M. Turley, Arch. Hydrobiol. Beih. Ergebn. Limnol. 37, (1983).
155–163 (1992). 78. H. W. Paerl and K. K. Galluci, Science 227, 647–649 (1985).
44. B. Norrman and A. Andersson, Polar Biol. 14, 531–537 79. H. W. Paerl, J. Phycol. 20, 600–608 (1984).
(1994). 80. A. L. Alldredge and Y. Cohen, Science 235, 689–691 (1987).
180 AGGREGATES AND CONSORTIA, MICROBIAL

81. D. M. Karl et al., Nature 309, 54–56 (1984). 116. H. W. Paerl et al., Appl. Environ. Microbiol. 60, 1044–1047
81. A. L. Shanks and M. L. Reeder, Mar. Ecol. Prog. Ser. 96, (1994).
43–47 (1993). 117. A. L. Huber, Hydrobiologia 133, 193–203 (1986).
82. Y. Cohen and E. Rosenberg, Microbial Mats: Physiological 118. K. Kononen, Kieler Meeresforch. Sonderh. 6, 281–294
Ecology of Benthic Microbial Communities, American (1988).
Society for Microbiology, Washington, D.C., 1989. 119. K. Kononen, Finn. Mar. Res. 261, 3–36 (1992).
83. N. Ramsing et al., Appl. Environ. Microbiol. 59, 3840–3849 120. G. Lindahl and K. Wallström, Arch. Hydrobiol. 104,
(1993). 193–204 (1985).
84. A. Teske et al., Appl. Environ. Microbial. 62, 4210–4215 121. K. Kononen et al., Limnol. Oceanogr. 41, 98–112 (1996).
(1996).
121. Å. Niemi, Acta. Bot. Fenn. 110, 57–61 (1979).
85. B. Risatti et al., Proc. Natl. Acad. Sci. 91, 10173–10177
122. K. O. Rothhaupt and H. Güde, Limnol. Oceanogr. 37,
(1994).
739–749 (1992).
86. J. P. Zehr et al., Appl. Environ. Microbiol. 61, 2527–2532
123. T. F. Thingstad et al., Mar. Ecol. Prog. Ser. 99, 239–259
(1995).
(1993).
87. T. F. Steppe et al., FEMS Microbiol. Ecol. 21, 149–154
124. U. Sommer, Limnol. Oceanogr. 30, 335–346 (1985).
(1996).
125. A. J. Horne and D. L. Galat, Limnol. Oceanogr. 30,
88. B. B. Jørgensen and N. P. Revsbech, Ophelia 31, 51–72
1229–1239 (1985).
(1989).
126. F. Sörensson and E. Sahlsten, Mar. Ecol. Prog. Ser. 37,
89. M. Kuhl and B. Jørgensen, Appl. Environ. Microbiol. 58,
277–284 (1987).
1164–1174 (1992).
90. D. Canfield and D. J. Des Marais, Geochim. Cosmochim. 127. J. W. Ammerman and F. Azam, Science 227, 1338–1340
Acta. 57, 3971–3984 (1993). (1985).
91. B. M. Bebout et al., Appl. Environ. Microbiol. 59, 1495– 128. A. Heinänen et al., Mar. Ecol. Prog. Ser. 116, 233–245
1503 (1993). (1995).
92. B. M. Bebout et al., Appl. Environ. Microbiol. 53, 2353– 129. K. Wallström et al., Acta. Phytogeogr. Suec. 78, 25–31
2362 (1987). (1992).
93. H. W. Paerl et al., Limnol. Oceanogr. 41, 419–427 (1996). 130. J. F. Haney, N. Z. J. Mar. Freshw. Res. 21, 467–475 (1987).
94. T. F. Steppe et al., Microb. Ecol. 41, 36–44 (2001). 131. K. G. Sellner et al., Hydrobiologia 292/293, 249–254 (1994).
95. S. B. Joye and H. W. Paerl, Mar. Biol. 119, 285–295 (1994). 132. S. Blomqvist and U. Larsson, Limnol. Oceanogr. 39,
880–896 (1994).
96. K. Henriksen and W. M. Kemp, in T. H. Blackburn and
J. Sprensen, eds., Nitrogen Cycling in Coastal Marine 133. A.-S. Heiskanen and K. Kononen, Arch. Hydrobiol. 131,
Environments SCOPE 33, John Wiley & Sons, New York, 175–198 (1994).
1988, pp. 207–249. 134. J. Kuparinen et al., Rapp. Réun. Cons. int Explor. Mer. 183,
97. S. B. Joye and H. W. Paerl, Mar. Ecol. Progr. Ser. 94, 180–192 (1984).
267–274 (1993). 135. T. Tamminen et al., in J. S. Gray and M. E. Christiansen,
98. C. Fründ and Y. Cohen, Appl. Environ. Microbiol. 58, 70–77 eds., Marine Biology of Polar Regions and Effects of Stress
(1992). on Marine Organisms, John Wiley & Sons, New York, 1985,
pp. 363–378.
99. A. N. Teske et al., Appl. Environ. Microbiol. 64, 2943–2951
(1998). 136. R. Elmgren, AMBIO 30, 222–231 (2001).
100. F. P. Van den Ende et al., FEMS Microbiol. Ecol. 23, 65–80 137. V. Michotey and P. Bonin, Mar. Ecol. Prog. Ser. 160, 47–56
(1997). (1997).
101. T. M. Hoehler et al., Global Biogeochem. Cycles 8, 451–463 138. E. J. Carpenter and K. Romans, Science 254, 1356–1358
(1994). (1991).
102. H. W. Paerl and P. T. Bland, Appl. Environ. Microbiol. 43, 139. D. G. Capone et al., Science 276, 1221–1229 (1997).
218–226 (1982). 140. H. W. Paerl et al., Appl. Environ. Microbiol. 55, 2965–2975
103. H. W. Paerl et al., Environ. Microbiol. 2, 11–26 (2000). (1989).
104. H. W. Paerl et al., Environ. Microbiol. 3, 123–130 (2001). 141. J. M. O’Neil, in L. Charpy, ed., Marine Cyanobacteria,
105. A. Decho, Oceanogr. Mar. Biol. Ann. Rev. 28, 73–154 (1990). Bulletin de Institut Oceanographique, Monaco, France,
1999, pp. 293–318.
106. H. W. Paerl, Limnol. Oceanogr. 33, 823–847 (1988).
142. E. J. Carpenter and J. McCarthy, Limnol. Oceanogr. 20,
107. R. W. Howarth et al., Limnol. Oceanogr. 33, 669–687 (1988).
389–401 (1975).
108. U. Larsson et al., Ambio 14, 9–14 (1985).
143. D. G. Capone et al., Appl. Environ. Microbiol. 56, 3532–
109. D. G. Capone and E. J. Carpenter, Limnol. Oceanogr. 39, 3536 (1990).
1140–1142 (1982).
144. J. P. Zehr and D. G. Capone, Microb. Ecol. 32, 263–281
110. E. F. DeLong et al., Science 243, 1360–1363 (1989). (1996).
111. H. W. Paerl et al., Limnol. Oceanogr. 40, 634–638 (1995). 145. J. P. Zehr and H. W. Paerl, in K. E. Cooksey, ed., Molecular
112. G. E. Fogg, Proc. R. Soc. London, Ser. B 173, 175–189 (1969). Approaches to the Study of the Ocean, Chapman & Hall,
113. H. W. Paerl and J. P. Zehr, in D. Kirchman, ed., Microbial London, 1998, pp. 285–302.
Ecology of the Oceans, Academic Press, New York, 2000, 146. E. J. Carpenter, Deep Sea Res. 20, 285–288 (1973).
pp. 387–426. 146. E. J. Carpenter, in E. J. Carpenter and D. G. Capone, eds.,
114. H. W. Paerl, Adv. Microb. Ecol. 11, 305–344 (1990). Nitrogen in the Marine Environment, Academic Press, New
115. H. W. Paerl et al., Limnol. Oceanogr. 32, 525–536 (1987). York, 1983, pp. 65–104.
 AIRBORNE BACTERIAL PATHOGENS 181

147. H. W. Paerl and B. M. Bebout, Science 241, 442–445 (1988). 185. D. J. Scala and L. J. Kerkhof, Appl. Environ. Microbiol.,
148. I. Bryceson and P. Fay, Mar. Biol. 61, 159–166 (1981). 65(4), 1681–1687 (1999).
149. H. W. Paerl, J. Phycol. 30, 790–799 (1994). 186. S. J. Giovannoni et al., Nature (London) 345, 148–149
150. E. J. Carpenter and C. C. Price, Science 191, 1278–1280 (1990).
(1976). 187. J. A. Fuhrman et al., Appl. Environ. Microbiol. 59, 1294–
151. T. H. Mague, Mar. Biol. 41, 213–227 (1977). 1302 (1993).
152. C. Fredriksson and B. Bergman, Protoplasma 197, 76–82 188. K. G. Field et al., Appl. Environ. Microbiol. 63, 63–70
(1997). (1997).
153. B. Bergman and E. J. Carpenter, J. Phycol. 27, 158–165 189. M. S. Rappe et al., Appl. Environ. Microbiol. 64, 294–303
(1991). (1998).
154. G. A. Borstad and L. E. Bordstad, in H. B. Stewart Jr., ed., 190. A. E. Murray et al., Appl. Environ. Microbiol. 62, 2676–
Cooperative Investigations of the Caribbean and Adjacent 2680 (1996).
Regions, FAO Fish Report 200, 1977, pp. 51–577. 191. J. A. Fuhrman et al., Nature (London) 356, 148–149 (1992).
155. H. W. Paerl et al., J. Phycol. 25, 773–784 (1989). 192. E. F. DeLong et al., Limnol. Oceanogr. 38, 924–934 (1993).
156. J. H. Martin, Nature 28, 159–182 (1994). 193. M. A. Munson et al., Appl. Environ. Microbiol. 63, 4729–
157. T. M. Church et al., Atmos. Environ. 18, 2657–2664 (1984). 4733 (1997).
158. R. Duce et al., Cycles 5, 193–259 (1991). 194. C. Vetriani et al., FEMS Microbiol. Letters 161, 83–88
(1998).
159. J. D. Rueter et al., in E. Carpenter, D. Capone, and
J. Rueter, eds., Marine Pelagic Cyanobacteria: Trichodes- 195. R. Massana, Limnol. Oceanogr. 43, 607–617 (1998).
mium and Other Diazotrophs, Kluwer Academic Press, 196. R. C. Hastings et al., Appl. Environ. Microbiol. 64, 3674–
Dordrecht, Netherlands, 1992, pp. 289–306. 3682 (1998).
160. D. G. Capone and E. J. Carpenter, in L. Charpy, ed., Marine 197. J. D. Kirshtein et al., Appl. Environ. Microbiol. 57, 2645–
Cyanobacteria, Monaco Musee Oceanographique, Monaco, 2650 (1991).
France, 1999, pp. 235–256. 198. J. B. Olson et al., Aquat. Microb. Ecol. 19, 26–36 (1999).
161. G. Thayer, Oecologia 14, 75–92 (1974). 199. I. R. Mcdonald and J. C. Murrell, FEMS Microbiol. Letters
162. J. L. Pinckney et al., J. Exp. Mar. Biol. Ecol. 187, 223–237 156, 205–210 (1997).
(1995b). 200. R. R. Karkhoffschweizer et al., Appl. Environ. Microbiol.
163. J. L. Pinckney et al., Mar. Ecol. Progr. Ser. 123, 207–216 61(1), 290–296 (1995).
(1995a). 201. M. Wagner et al., J. Bacteriol. 180, 2975–2982 (1998).
164. J. L. Pinckney et al., Microb. Ecol. 29, 19–37 (1995). 202. L. Kerkhof and B. B Ward, Appl. Environ. Microbiol. 59,
165. C. A. Currin et al., Mar. Ecol. Progr. Ser. 121, 99–116 1303–1309 (1993).
(1955). 203. L. Kerkhof, Anal. Biochem. 205, 359–364 (1992).
166. T. F. Steppe, Ph.D. Dissertation, University of North Car- 204. L. Kerkhof and P. Kemp, FEMS Microb. Ecol. 30, 253–260
olina, Chapel Hill, N.C. 2001, p. 130. (1999).
167. J. B. Olson et al., Aquat. Microb. Ecol. 19, 29–36 (1999). 205. W. T. Liu, Appl. Environ. Microbiol. 63(11), 4516–4522
168. N. P. Revsbech et al., Limnol. Oceanogr. 28, 1062–1074 (1997).
(1983). 206. C. Phelps et al., FEMS Micro. Ecol. 27, 269–279 (1998).
169. P. T. Visscher et al., FEMS Microbiol. Ecol. 86, 383–294 207. D. Scala and L. Kerkhof, Appl. Environ. Microbiol. 66,
(1992). 1980–1986 (2001).
170. J. Pinckney and H. W. Paerl, Appl. Environ. Microbiol. 63,
420–426 (1997).
171. D. Gordon et al., Antarctic J. U.S. 33, 222–224 (1998).
172. J. B. Olson et al., Microb. Ecol. 36, 231–238 (1998). AGROTERRORISM. See BIOTERRORISM
173. S. Rysgaard et al., Appl. Environ. Microbiol. 59, 2093–
2098 (1993).
174. R. I. Amann et al., J. Bacteriol. 172, 762–770 (1990).
175. A. T. Nielsen et al., Environ. Microbiol. 2, 59–68 (2000).
176. C. A. Currin et al., Limnol. Oceanogr. 35, 59–71 (1991). AIR POLLUTION CONTROL. See VIRAL AEROSOLS
177. B. B. Ward and A. F. Carlucci, Appl. Environ. Microbiol. 50,
194–201 (1985).
178. B. B Ward et al., J. Gen. Microbiol. 139, 2285–2293 (1993).
179. D. F. Millie et al., Can. J. Fish. Aquat. Sci. 50, 2513–2527
(1993). AIR SAMPLING. See SAMPLING TECHNIQUES FOR
180. C. R. Woese, Microbiol. Rev. 51, 221–271 (1987). ENVIRONMENTAL MICROBIOLOGY
181. H. H. Xu and F. R. Tabita, Appl. Environ. Microbiol. 62,
1913–1921 (1996).
182. I. M. Head et al., Microb. Ecol. 35, 1–21 (1998).
183. P. Hugenholtz et al., Limnol. Oceanogr. 33, 99–113 (1998).
184. M. T. Suzuki et al., Appl. Environ. Microbiol. 63(3), 983–989 AIR-WATER INTERFACE. See NEUSTON MICROBIOLOGY:
(1997). LIFE AT THE AIR –WATER INTERFACE
182 AIRBORNE TOXIGENIC MOLDS

AIRBORNE BACTERIAL PATHOGENS. good agricultural practices. But even in more developed
See INFECTIOUS AIRBORNE BACTERIA countries, contamination of food products results in
significant economic losses, prompting governments to
monitor levels of certain mycotoxins (e.g., aflatoxins) in
food products to ensure that people do not consume
unacceptable levels of these fungal toxins.
AIRBORNE FUNGI. See IDENTIFICATION OF AIRBORNE Mycotoxins are low molecular weight (generally
FUNGI <1, 000 Daltons) and nonvolatile compounds; the odors
associated with moldy buildings are due to microbial
volatile organic compounds (mVOCs) that are typically
low molecular weight alcohols, ketones, esters, and hydro-
carbons of low inherent toxicity (11). Importantly, most
AIRBORNE MICROORGANISMS. See BIOAEROSOLS: mycotoxins are fairly stable compounds that can survive
TRANSPORT AND FATE; WASTEWATER AND BIOSOLIDS AS SOURCES unchanged for long periods of time after the fungi are no
OF AIRBORNE MICROORGANISMS
longer viable. Thus, information obtained from culturing
air samples (e.g., CFUs/m3 ) needs to be evaluated with
this in mind because low levels of viable (culturable) fungi
do not necessarily reflect low levels of toxigenic fungi in the
air. Furthermore, some airborne toxigenic fungal spores
AIRBORNE TOXIGENIC MOLDS
(e.g., Stachybotrys) have much shorter lifetimes in the dry
state than do others, complicating the interpretation of
BRUCE B. JARVIS
CFU data from air sampling (12).
University of Maryland
There is a great deal of information available on
College Park, Maryland
mycotoxins (4,13,14). In brief, mycotoxins are fungal
secondary metabolites (15) that pose a risk to both
In the past 20 years, there is increasing recognition that
human and animals that come into contact with these
an important factor in the health of people in indoor
natural products, principally through ingestion of food
environments is the dampness of the buildings in which
and feed. They span an enormous range of chemical
they live and work (1). Furthermore, it is now appreciated
structures (Fig. 1), with diverse biological activities and
that the principal nonpathogenic biologics responsible for
are produced by a wide variety of fungal genera (16).
the health problems in such buildings are fungi rather
Mycotoxins, like other secondary metabolites in general,
than bacteria or viruses (2,3). Although traditionally fungi
tend to be idiosyncratic in that specific mycotoxins are
in this context have been viewed as allergens (and in
often restricted in their production to a small number
some circumstances, pathogens), data have accumulated
of fungal genera or even single species. Although some
to show that the adverse health effects resulting from
classes of mycotoxins may be restricted to a small number
inhalation of fungal spores are due to multiple factors. One
of species in a single genus (e.g., the aflatoxins to
factor associated with certain fungi is the small molecular
Aspergillus), some fungi are notorious for producing many
weight toxin (mycotoxin) produced by these fungi (1).
different classes of mycotoxins within their genus (e.g.,
Mycotoxins are held to be important in human and animal
Penicillium). Additionally, there may be great variation in
health because of their production by toxigenic fungi
toxigenic potential within a single species. Most isolates
associated with food and feed (4). However, mycotoxins
of Aspergillus flavus are aflatoxin producers (although
tend to concentrate in fungal spores (5,6) and thus present
the percentage varies between areas of the country) (17),
a potential hazard to those inhaling airborne spores. There
but only one-third of Stachybotrys chartarum isolates
are well documented cases of mycotoxicosis resulting from
produce the satratoxins, highly toxic members of the
inhalation of toxigenic spores by agricultural workers
trichothecene class of mycotoxins (18). Furthermore, there
handling moldy farm material (7,8), but until recently
is a general lack of knowledge about the long-term
there have been few reports of such toxicoses in an urban
chronic effects of most mycotoxins, especially in humans.
setting (9).
With all this diversity in the biology, chemistry, and
toxicology of mycotoxins, it is not surprising that there
MYCOTOXINS are no general methods available for mycotoxin analysis.
The only useful analyses for mycotoxins are those for
The most potent toxin known is that produced by specific ones for which the analytical chemistry has been
Clostridium botulinum, the bacterium responsible for established.
botulism. Fungi also produce toxins that, like the From an agricultural view, the three most important
botulism toxin, can enter our food and livestock feed. genera of toxigenic fungi are Fusarium, Aspergillus, and
These fungal toxins are called mycotoxins, and they Penicillium. The principal classes of mycotoxins of concern
present both medical and economic problems to the from ingestion are (fungal genera): the trichothecenes
community. Exposure to mycotoxin-producing (toxigenic) and fumonisins (Fusarium), aflatoxins (Aspergillus), and
fungi has a long history and is a continuing threat to the ochratoxins (Aspergillus and Penicillium). Also, there
human and animal health (10), especially in developing are numerous other mycotoxins that intermittently cause
countries where people do not often have the luxury of problems. Some trichothecenes are acutely toxic (e.g., the
 AIRBORNE TOXIGENIC MOLDS 183

O
O

COOH O OH O
O
H N O O
OCH3 O
H
CH3 OH
O O O HO
Cl CH2OH
H

Aflatoxin B1 Ochratoxin A Deoxynivalenol

N S S N CH34
OH
O CH2OH COO−

H3C Gliotoxin COCH2CHCH2COO−

CH3
H
O CH3 O CH3 OH OH OH

N HN CH3 CH3(CH2)3CHCHCHCH2CHCH2CH(CH2)4CHCH2CHCHCH3
H
+
OO CH3 OCOCH2CHCH2COO− NH3
OH
O COO−

Chaetoglobosin A Fumonisin B1

Figure 1. Selected mycotoxins.

satratoxins), whereas others are considerably less acutely former species grow well on building material with water
toxic but are immunosuppressant or cause feed refusal in activities (aw ) in the range of 0.8; whereas S. chartarum
livestock (e.g., deoxynivalenol). Fumonisins, and especially requires aw of 0.9 and higher and greater than 0.96 on
some of the aflatoxins (e.g., B1 ), are carcinogenic (10). building material to grow well (22). Generally, moisture
There is now overwhelming epidemiological evidence levels this high are the result of local leakage (e.g.,
that aflatoxin B1 in the diet contributes significantly to from a broken pipe) and are confined to a small
the high incidence of liver cancer in many developing area, but with leaking roofs, water intrusion can be
countries (10). Ochratoxin A is nephrotoxic and a possible quite extensive. Like all fungi, the growth of toxigenic
cause of urinary tract tumors and Balkan-endemic fungi depends on substrate, temperature, and water
nephropathy (10). activity (23).
In an indoor environment, the toxigenic molds of most
concern are Penicillium, Aspergillus Chaetomium, and
Stachybotrys. However, the specific species of Penicillium AIRBORNE FUNGAL SPORES
and Aspergillus in indoor environments differ from
those of an agricultural concern. For agricultural-based The mycotoxins found in indoor air are most likely confined
mycotoxicoses, P. verrucosum, P. cyclopium, A. flavus, and to the aerosolized spores (or more correctly referred to
A. parasiticus (the latter two are aflatoxin producers) are as conidia) of the toxigenic fungi (and building material
of most concern; whereas, indoors, P. aurantiogriseum, on which these fungi grow) (24), with levels unlikely to
P. brevicompactum, P. chrysogenum, A. versicolor, and the approach those encountered in animal feed intoxications.
A. nidulans group are the most important toxigenic However, mycotoxins are considerably more toxic when
species (19). Stachybotrys chartarum (Ehrenb. ex Link) inhaled than when ingested (25,26). Furthermore, the
Hughes (S. atra Corda) has been focused on in recent effects of these toxins are multiple, and in particular,
years because of its association with acute bleeding in the immunosuppressant effects of these toxins on alveolar
the lungs of infants (idiopathic pulmonary hemosiderosis, macrophage function is very pronounced (27,28). Studies
IPH), several of whom have died (20,21). However, of the effects of fungal spores on rat pulmonary
the Penicillium spp. and A. versicolor are far more alveolar macrophage cells have shown variability probably
commonly encountered in water-damaged buildings than mediated by spore-borne toxins (29). Spores of five
is S. chartarum. The basic reason for this is that the Aspergillus, Penicillium spinulosum, and Cladosporium
184 AIRBORNE TOXIGENIC MOLDS

cladosporioides caused the release of leukotriene B4 and MYCOTOXINS IN THE AIR

increased superoxide anion production and activation
of complement. Several species of Aspergillus inhibited Although a great deal is known about the spectrum
LPS-stimulated IL-1; whereas the other species had no of mycotoxins produced by the relevant toxigenic fungi
effect (30). Data are available to support the role of found indoors (20,21,47), there are few data available
the immunosuppressant mycotoxin, gliotoxin (Fig. 1), in on the presence of the mycotoxins themselves in mold-
the pathogenicity of A. fumigatus in animal lungs (31). contaminated buildings. In those few cases where
Jakab and coworkers (32) demonstrated that inhalation mycotoxins have been found in mold-contaminated
exposure of aflatoxin in rats led to persistent reductions buildings (47–55), they have been found in bulk samples
in phagocytosis, among other effects. In rats, inhalation rather than in air samples. Following from this, there
exposure to aflatoxin-containing spores of Aspergillus remains the critical question of the exposure of people in
has been demonstrated to be an effective route of these buildings to the airborne toxigenic fungal spores, a
exposure (33). Damage to clearance mechanisms would problem that merits further discussion.
affect processing of antigen and lead to accumulation As noted earlier, many studies report the presence
of material in granulomatous matter, leading to an of toxigenic fungi in mold-contaminated buildings, but
immunotoxic effect (34–36). Low exposure to spores of few actually determine the presence of the mycotoxins
toxigenic S. chartarum affects lung surfactant production in the buildings. Nielsen and coworkers (47) showed that
in laboratory animals (37). bulk samples from mold-contaminated building materials
Toxigenic spores that strongly affect alveolar macro- from water-damaged domestic residence contained both
phage function pose a threat to workers handling toxigenic fungi and the specific mycotoxins produced
mycotoxin-contaminated material. Occupational expo- by these fungi. Although their data are complicated
sures to spores of A. flavus, a producer of the potent by these natural cultures being mixtures of two or
carcinogenic aflatoxins, have been shown to correlate more fungal species growing together (mixed cultures
to increased risk of liver cancer (7,8). A female agri- are the rule rather than the exception under natural
cultural worker suffered acute renal failure following conditions), the mycotoxins found are those expected from
exposure to grain dust in an enclosed granary, whose the fungi observed. Thus, the sterigmatocystin (an IARC
wheat was shown to be contaminated by A. ochraceous and 2A carcinogen i.e., a probable human carcinogen) and 5-
its mycotoxin metabolite, ochratoxin A (Fig. 1). Although methoxysterigmatocystin found in the bulk samples most
ochratoxin A was not demonstrated to be in air samples certainly came from A. versicolor (which in fact produces
from the granary, caged experimental animals (guinea these mycotoxins when grown on building materials) (54),
pigs and rabbits) experienced acute renal failure when and the chaetoglobosins from the Chaetomium sp.
exposed for eight hours to aerosols generated by their P. chrysogenum were commonly observed in these bulk
movement on moldy wheat (38). In this regard, a recent samples, but no mycotoxins were found that could be
report of high concentrations (1,500 ppb) of ochratoxin ascribed to this fungus. In a similar study of mold-
A in dust from a home air-handling system in which contaminated homes in Finland, workers showed that
occupants had complained of recurring health problems 34 of 79 bulk dust samples contained one or more of
is noteworthy (39). Other reports have indicated that the 17 mycotoxins for which analyses were conducted.
S. chartarum (40–42), A. versicolor, and several toxigenic Of the 79 samples, 24% contained sterigmatocystin
species of Penicillium are potentially hazardous, espe- and 19% contained one or more of the trichothecene
cially when the air-handling systems have become heavily mycotoxins (55).
contaminated (28,43).
Epidemiological studies have connected the increased STACHYBOTRYS CHARTARUM — A MODEL?
incidence of premature births in Norwegian farmers
to their exposure to grain-borne toxigenic fungi (44). As noted earlier, S. chartarum is less commonly encoun-
Researchers have shown that Finnish farmers are com- tered indoors than other toxigenic fungi (e.g., A. versicolor)
monly exposed to relatively high levels of airborne but has nonetheless garnered more attention than the
spores (103 –106 CFU/m3 ), and that low to moderate others because of its association with acute bleeding in
(0.00–11 ppm) levels of Fusarium toxins (e.g., deoxyni- the lungs of infants in Cleveland (20,21). This fungus was
valenol, see Fig. 1) are commonly found in the grain, held responsible for serious health problems of a fam-
although less commonly in air samples (45). ily living in a water-damaged home in Chicago (40) and
The literature on the growth characteristics of fungi in has been implicated in several cases of building-related
damp buildings is extensive (1–3,23,28). However, with illness (39–41,56,57). A cluster of cases of acute bleeding
respect to the production of organic compounds by these in the lungs of infants was reported in Cleveland, Ohio
fungi, most of the literature has focused on the microbial- where over 30 infants from homes that suffered flood dam-
generated mVOCs (11). Although some of the mVOCs age became sick (12 deaths) with the illness beginning in
might be of health concern from a chronic perspective (e.g., January 1993 (21).
styrene), mVOCs are not considered an acute toxic hazard Stachybotrys chartarum was first described more than
(although the associated odors can present problems) (11). 150 years ago by Corda in 1837, who isolated the mold
However, confounding factors such as the co-occurrence of from damp wallpaper in a home in Prague. Although
ozone with mVOCs may result in a significant increase in S. chartarum-related animal intoxications have no doubt
volatile irritants (46). existed for some time, it was not until 1931 that this
 AIRBORNE TOXIGENIC MOLDS 185

toxicosis was described and not until the late 1930s was agent(s) found in several S. chartarum isolates from the
the condition recognized as a mycotoxicosis: stachybotryo- Cleveland study (72). Stachybotrys chartarum can produce
toxicosis (58). Horses are particularly sensitive to this a diverse array of compounds, and the variation in
mold that is a common contaminant of damp hay individual and overall metabolite levels from one isolate
and straw (58). However, there are intermittent reports to the next is remarkable — and troublesome to any
where major portions of a livestock herd have been who wish to draw toxicological inferences from levels
killed following ingestion of S. chartarum–contaminated of airborne S. chartarum spores. Similarly, the toxicity
hay (58,59). of two isolates of S. chartarum recovered from almost
Stachybotryotoxicosis in humans is rare and has identical environmental conditions, and then grown in the
been reported most commonly in workers who han- laboratory, may exhibit widely different biological activity
dle moldy straw and hay (60). Workers handling profiles (73).
S. chartarum–contaminated hay experienced bloody nose It is instructive to see how chemically complex the
and nasal secretions, dyspnea, sore throats, and inflam- profile of toxin production by this fungus is. Figure 2
mation of the skin (61). More recently, severe dermatitis illustrates members of the classes of mycotoxins reported
appeared on the hands of those workers who were han- to be produced by isolates S. chartarum and Memnon-
dling decomposable flower pots made of recycled paper iella echinata, obtained from Cleveland homes involved
that had a heavy growth of S. chartarum; however, no res- in the IPH study (20,21,73–75). The latter fungus is
piratory distress was noted, although viable spore levels closely related taxonomically to Stachybotrys (76) and is
of S. chartarum up to 7,500 CFUs were found during the often found growing on the same substrates indoors with
handling of the pots (62). Until 1986, there were no reports S. chartarum. The spectrum of compounds produced by
in the literature of stachybotryotoxicosis in North America. S. chartarum in the laboratory depends not only on the
The first report of a toxicosis in North America attributed particular isolate but also on the substrate. There are more
to S. chartarum occurred in a Chicago household where than 10 each of the trichothecene (71) and atranone (77,78)
the home had suffered significant water intrusion from classes of compounds produced by S. chartarum and they
a leaky roof over a period of several years (40). The are the minor secondary metabolites. The major myco-
occupants of the home had suffered recurring maladies, toxins produced by S. chartarum and M. echinata are the
including cold and flu symptoms, diarrhea, headaches, benzodrimanes (Fig. 2: stachybotrylactone, stachybotry-
fatigue, dermatitis, sore throats, intermittent loss of hair, dialdehyde, and memnobotrin A), which are the ‘‘signa-
and general malaise. Examination of the home revealed ture’’ compounds of this fungus (65,73). There are more
extensive growth of S. chartarum on the ceiling of a bed- than 20 members of this class, and unless steps are taken
room and, more significantly, in the air-handling system. to separate these compounds from the others before chem-
Ethanol extracts of the contaminated material were lethal ical analysis by high performance liquid chromatography
(injected per os) to mice and rats within 24 hours. From (HPLC), it is very difficult to even observe the presence
these extracts were isolated a series of potent cytotoxic tri- of the other classes of mycotoxins (79). In addition, the
chothecenes (40). After the home was thoroughly cleaned more highly immunosuppressant dialdehydes (80), such
(protective clothing was necessary to prevent severe der- as stachybotrydialdehyde (Fig. 2), are unstable and tend
matitis in the workers), the residents returned, and their to readily rearrange to the corresponding significantly less
symptoms abated and within a few weeks, their health active (80) stachybotrylactones (79,81).
returned to normal. As noted earlier, isolates of toxigenic fungi may
The rare occurrence of S. chartarum–induced toxicosis differ significantly in their toxigenic potential. In the
in humans stems from the specific conditions this fungus laboratory, isolates (approximately 40) of S. chartarum,
requires for significant growth (e.g., this fungus is never grown on rice, from the Cleveland study (73), display
found growing on food products). In damp buildings, high cytotoxicities of their crude extracts that vary by five
cellulose and fibrous surfaces are favored for growth if orders of magnitude. The most toxic isolates (about 1/3)
the moisture level is sufficient (aw > 0.9). For example, if are those that produce the macrocyclic trichothecene
sufficiently moist, S. chartarum readily grows on gypsum mycotoxins (e.g., satratoxin H). Of the remaining two-
board (63,64), wood fiberboard, wall paper, and dust- thirds of the isolates, some produce the considerably
lined air-conditioning ducting (40,43,45–49,54). In the less toxic trichothecenes, trichodermol, and trichodermin;
laboratory, S. chartarum is commonly grown on rice and most isolates yield the considerably less biologically
although toxin production is observed when the fungus active atranones (Fig. 2) (77,78). Although we have yet
is grown on building material as well (48,54,63,64). to find any significant biological activity associated with
Chemical investigation of S. chartarum has provided the atranones, these compounds are related in structure
many highly toxic and novel compounds (65). Stachybotrys to the diterpenoid dolabellanes (two of whom have been
chartarum is known to produce the very potent cytotoxic isolated from S. chartarum cultures) (78), several of which
macrocyclic trichothecenes (e.g., satratoxin H) along with do have appreciable biological activity (82). Others also
a variety of immunosuppressants (65–67) and endothelin have noted great variation in various biological activities
receptor antagonists (68–70). In fact, some of the most of S. chartarum isolated from indoor environments (83),
cytotoxic fungal metabolites ever discovered are products and thus this is no doubt a general characteristic of
of S. chartarum fermentation (71), and by no means S. chartarum.
have all the active constituents been isolated and The cluster of cases of acute lung bleeding (idiopathic
identified — there remains an uncharacterized hemolytic pulmonary hemosiderosis, IPH) in infants (two weeks to
186 AIRBORNE TOXIGENIC MOLDS

O
O
O

O
O
O HO HO CHO
O O
HO HO
O CHO

O O
HO HO HO
OH

Satratoxin H
(A Trichothecene) Stachybotrylactone Stachybotrydialdehyde

O
CH3 HO
CH3
CH3 NH
H
O
CH3 O O
H O O
H O
HO OR
H CH3
AcO CH3
O CH3
O Trichodermol (R = H)
Atranone A Memnobotrin A Trichodermin (R = Ac)

OCH3 O OCH3
O
HO
O O O
H3CO O

OH CH3

Memnoconone Dechlorogriseofulvin

Figure 2. Mycotoxins isolated from S. chartarum and M. echinata.

six months of age) from Cleveland, Ohio has garnered a care in pointing out that other environmental factors (e.g.,
good deal of public attention. Between 1993 and 1998, second-hand cigarette smoke) also may play important
there were 37 cases of IPH reported in the Cleveland area, roles in the etiology of IPH. Nonetheless, others (86–90)
a normally rare syndrome (20,21). Twelve of the cases have emphasized that the epidemiology of IPH vis a
resulted in deaths, including seven that had originally vis Stachybotrys is tenuous, and that reports such as
been diagnosed as sudden infant death syndrome. The the Cleveland studies may raise undue concern in the
researchers reported that all the case homes in the public mind. It is worth noting that it took 30 years
study had varying degrees of water damage and that (with hundreds of time-consuming and expensive studies)
S. chartarum was significantly more commonly found and before the medical community accepted as fact the
at higher levels than in control homes (20,21). However, causal relationship between the consumption of aflatoxin-
a strange observation in these cases is that occupants contaminated food and an increased incidence of liver
of the IPH homes older than six months of age appear cancer in humans.
not to suffer from any undue adverse health effects,
although the investigators have offered a rationale for the
hypersensitivity of infants to the effects of S. chartarum THE FUTURE
spores. In addition to the Cleveland IPH studies, there
have been several recent reports of individual cases of IPH There are presently a number of serious limitations
apparently associated with the presence of S. chartarum in our ability to assess the role played by mycotoxins
in homes (51,84), including one where culturable spores in the health of those exposed to toxigenic fungi in
of S. chartarum were isolated from the bronchoalveolar water-damaged buildings. Toxicologists prefer to study
lavage fluid of a child with pulmonary hemorrhage (85). pure toxins, but mycotoxicoses always involve mixtures
However, ‘‘association’’ is not the same as cause and of chemicals and usually mixed fungal cultures as
effect, and workers in this area (21) have taken some well. Furthermore, the spores appear to have noticeable
 AIRBORNE TOXIGENIC MOLDS 187

effects themselves apart from their mycotoxins. Intranasal 2. J. D. Miller, Atmos. Environ. 26, 2163–2172 (1992).
exposure of mice to spores of a highly toxigenic strain of 3. J. D. Miller, in Approaches to Assessment of the Microbial
S. chartarum resulted in a remarkable deterioration of Flora of Buildings, ASHRAE IAQ 92, Environments for
the lung tissue of the animals (91,92). In the case where Healthy People, ASHRAE, Atlanta, Ga., 1993, pp. 147–159.
animals inhaled pure trichothecene mycotoxins (25,26), 4. V. Betina, Mycotoxins: Chemical, Biological, and Environ-
little or no damage was observed in their lung tissue, even mental Aspects, Elsevier, New York, 1989.
though the trichothecene toxin (T-2 toxin) administered 5. W. G. Sorenson, D. G. Frazer, and B. B. Jarvis, Appl. Envi-
in this manner was more than an order of magnitude ron. Microbiol. 53, 1370–1375 (1987).
more toxic relative to iv or ip injection. But spores in 6. T. O. Larsen and J. C. Frisvad, in R. Samson, B. Flannigan,
which toxins have been extracted are considerably less and S. Graveson, eds., Health Implication of Fungi in
deleterious to lung tissue than are the spores prior to Indoor Environments, Elsevier Biomedical, Amsterdam, The
extraction of their toxins (93). Thus, it appears that the Netherlands, 1994, pp. 251–279.
spores themselves play an important role in the responses 7. J. L. Autrup, J. Schmidt, T. Seremet, and H. Autrup, Scand.
observed in the lung tissue of animals who inhale the J. Work Environ. Health 17, 436–440 (1991).
spores, strongly reinforcing data on the disregulation of 8. J. L. Autrup, J. Schmidt, T. Seremet, and H. Autrup, Scand.
pulmonary macrophages caused by toxigenic spores. Environ. Health Persp. 99, 195–197 (1993).
The critical question to be answered is that of the level 9. B. B. Jarvis, in AST STP 1071, P. R. Morey, J. C. Feely, Sr.,
of exposure. Without a reliable assessment of exposure, and J. A. Otten, eds., Biological Contaminants in Indoor
Environments, ASTM, Philadelphia, 1990, pp. 201–214.
the risks to human health cannot be properly ascertained
in individual cases. Although the epidemiological data 10. M. Peraica, B. Radic, A. Lucic, and M. Pavlovic, Bull. WHO
77, 754–766 (1999).
clearly show that heightened exposure to indoor molds
compromises the general health of those exposed (1), what 11. T. Godish, Sick Buildings: Definition, Diagnosis, and Mitiga-
tion, Lewis Publishers, Boca Raton, Fla., 1996.
cannot be gleaned from these data is the role played by
mycotoxins. At present, few good methods for measuring 12. J. D. Miller, in F. Flannigan, J. D. Miller, and R. Samson,
eds., Methods for Investigating Fungal Contamination of
exposure of individuals to indoor molds are available.
Buildings, Harwood Academic Publishers, Amsterdam, The
The measurements of airborne mold levels indoors is Netherlands, 2001.
problematic for a variety of reasons (12) and tend to be
13. Y. Ueno, Crit. Rev. Toxicol. 14, 99–132 (1985).
meaningful only when careful comparisons are made to
14. A. Pittet, Rev. Med. Vet. 49, 79–92 (1998).
the concurrent levels of molds in outdoor air. Indoor
air sampling may be useful to demonstrate unacceptable 15. C. Pearce, Adv. Appl. Microbiol. 44, 1–80 (1997).
amplification of indoor molds, but beyond that, these data 16. R. A. Samson, J. Med. Vet. Mycol. 30, 9–18 (1992).
are usually not statistically reliable (12). 17. B. W. Horn and J. W. Dorner, Appl. Environ. Microbiol. 65,
Except for a small number of allergenic molds, no 1444–1449 (1999).
direct way of determining whether an individual has 18. B. B. Jarvis and S. Hinkley, in E. Johanning and C. S. Yang,
been exposed to a particular mold is currently available. eds., Bioaerosols, Fungi, and Mycotoxins: Health Effects,
In fact, the biochemical markers for general exposure Assessment, Prevention, and Control, Boyd Printing, Albany,
New York, 1999, pp. 232–239.
to heightened levels of fungal spores are presently only
of limited use, and certainly are such that one cannot 19. J. A. Scott, N. A. Straus, and B. Wong, in E. Johanning and
C. S. Yang, eds., Bioaerosols, Fungi, and Mycotoxins: Health
distinguish exposure to toxigenic molds from exposure
Effects, Assessment, Prevention, and Control, Boyd Printing,
to nontoxigenic molds (12), let alone exposure to specific
Albany, New York, 1999, pp. 335–342.
mycotoxins. What are needed are biological markers for
20. R. A. Etzel et al., Arch. Pediatr. Adolesc. Med. 152, 757–762
exposure to both specific fungi and for their specific toxins.
(1998).
A few mycotoxins (e.g., ochratoxin A) are metabolically
21. D. G. Dearborn et al., Environ. Health Persp. 107(Suppl. 3),
stable enough to give sustainable and measurable levels
495–499 (1999).
in blood (94), but most metabolize readily and few give
22. S. Gravesen, P. A. Nielsen, R. Iversen, and K. F. Nielsen,
metabolic products that have been characterized. An
Environ. Health Persp. 107(Suppl. 3), 505–508 (1999).
exception to this is aflatoxin B1 that conjugates to genomic
23. R. S. Samson, in E. Johanning and C. S. Yang, eds., Bio-
DNA and can be assayed as the guanine adduct, giving a
aerosols, Fungi, and Mycotoxins: Health Effects, Assessment,
measure of exposure to the mycotoxin (10). Prevention, and Control, Boyd Printing, Albany, New York,
Future advances in analytical chemistry coupled 1999, pp. 33–37.
with the identification of biological markers for specific 24. W. G. Sorenson, Environ. Health Persp. 107(Suppl. 3),
mycotoxins will be a great aid in addressing the 469–472 (1999).
fundamental questions of exposure to airborne toxigenic 25. D. A. Creasia et al., Fundam. Appl. Toxicol. 8, 230–235
molds and their risk to people living and working in (1987).
mold-contaminated buildings. 26. D. A. Creasia, J. D. Thurman, R. W. Wannemacher, and
D. L. Bunner, Fundam. Appl. Toxicol. 14, 54–59 (1990).
BIBLIOGRAPHY 27. W. G. Sorenson, Dev. Ind. Microbiol. 31, 205–211 (1990).
28. B. Flannigan and J. D. Miller, in R. Samson, B. Flannigan,
1. J. D. Miller, in L. Morawska, ed., Indoor Air Quality — An M. Flannigan, and S. Graveson, eds., Health Implication
Integrated Approach, Elsevier Biomedical, Amsterdam, 1995, of Fungi in Indoor Environments, Elsevier, Biomedical,
pp. 159–168. Amsterdam, The Netherlands, 1994, pp. 3–28.
188 AIRBORNE TOXIGENIC MOLDS

29. W. G. Sorenson and D. M. Lewis, in D. H. Howard and 60. E.-L. Hintikka, in T. D. Wyllie and L. G. Morehouse, eds.,
J. D. Miller, eds., The Mycota, Vol. VI, Human and Animal Mycotoxic Fungi Mycotoxins Mycotoxicoses, vol. 3, Marcel
Relationships, Springer Verlag, Berlin, Germany, 1996, Dekker, N.Y., 1978, pp. 87–89.
pp. 159–172. 61. K. Andrassy, I. Horvath, T. Lakos, and Z. Toke, Mykosen 23,
30. W. G. Sorenson, T. A. Shahan, and D. M. Lewis, in R. Sam- 130–133 (1979).
son, B. Flannigan, M. Flannigan, and S. Graveson, eds., 62. D. Dill, C. Trautmann, and R. Szwzyk, Mycoses 40, 110–114
Health Implication of Fungi in Indoor Environments, (1997).
Elsevier, Biomedical, Amsterdam, The Netherlands, 1994,
63. M. Nikulin, A.-L. Pasanen, S. Berg, and E.-L. Hintikka, Appl.
pp. 325–343.
Environ. Microbiol. 60, 3421–3424 (1994).
31. J. L. Richard, T. J. Dvorak, and P. F. Ross, Mycopathologia
134, 167–170 (1996). 64. K. F. Nielsen and S. Gravesen, in E. Johanning and
C. S. Yang, eds., Bioaerosols, Fungi, and Mycotoxins: Health
32. G. J. Jakab et al., Toxicol. Appl. Pharmacol. 125, 198–205
Effects, Assessment, Prevention, and Control, Boyd Printing,
(1994).
Albany, N.Y., 1999, pp. 423–431.
33. A. Zarba et al., Carcinogenesis 13, 1031–1033 (1992).
65. B. B. Jarvis, J. Salemme, and A. Morais, Nat. Toxins 3, 10–16
34. K. Donaldson and D. M. Brown, Respir. Med. 84, 432–434 (1995).
(1990).
66. K. Sakamoto et al., J. Antibiot. 46, 1788–1798 (1993).
35. K. E. Driscoll, J. K. Maurer, and L. L. Crosby, J. Aerosol Med.
3, 83–91 (1990). 67. R. Kaneto et al., J. Antibiot. 47, 727–730 (1994).
36. P. G. Holt, Am. J. Indust. Med. 47, 47–54 (1990). 68. K. Ogawa et al., J. Antibiot. 48, 1396–1400 (1995).
37. C. D. Mason et al., Nat. Toxins 6, 27–33 (1998). 69. B. E. Roggo et al., J. Antibiot. 49, 13–19 (1996).
38. N. Di Paolo et al., Nephron 64, 621–625 (1993). 70. B. E. Roggo et al., J. Antibiot. 49, 374–379 (1996).
39. J. L. Richard, R. D. Plattner, J. May, and S. L. Liska, Myco- 71. B. B. Jarvis, in R. P. Sharma and D. K. Salunkhe, eds.,
pathologia 146, 99–103. Mycotoxins and Phytoalexins in Human and Animal Health,
CRC Press, Boca Raton, Fla., 1991, pp. 361–421.
40. W. A. Croft, B. B. Jarvis, and C. S. Yatawara, Atmos. Envi-
ron. 20, 549–552 (1986). 72. S. J. Vesper et al., Appl. Environ. Microbiol. 65, 3175–3181
41. Anonymous, in E. Johnanning and C. S. Yang, eds., Fungi (1999).
and Bacteria in Indoor Air Environments, Eastern New 73. B. B. Jarvis et al., Appl. Environ. Microbiol. 64, 3620–3625
York Occupational Health Program, Latham, N.Y., 1995, (1998).
pp. 201–207. 74. B. B. Jarvis et al., J. Nat. Prod. 59, 553–554 (1996).
42. E. Johanning et al., Inter. Arch. Occup. Environ. Health 68, 75. S. F. Hinkley, J. C. Fettinger, K. Dudley, and B. B. Jarvis,
207–218 (1996). J. Antibiot. 52, 988–997 (1999).
43. J. E. Smith, J. G. Anderson, C. W. Lewis, and Y. M. Murad, 76. S. C. Jong and D. C. Davis, Mycotaxon 3, 409–485 (1976).
FEMS Microbiol. Lett. 100, 337–344 (1992).
77. S. F. Hinkley, J. Jiang, E. P. Mazzola, and B. B. Jarvis, Tetra.
44. P. Kristensen et al., Am. J. Epidemiol. 146, 329–338 (1997). Lett. 40, 2725–2728 (1999).
45. S. Lappalainen et al., Atmos. Environ. 30, 3059–3065 (1996). 78. S. F. Hinkley et al., Phytochemistry 55, 663–673 (2000).
46. P. Wolkoff et al., Atmos. Environ. 33, 693–698 (1999). 79. S. Hinkley and B. B. Jarvis, in M. W. Trucksess and
47. K. F. Nielsen et al., Mycopathologia 145, 43–56 (1999). A. E. Pohland, eds., Methods in Molecular Biology: The Myco-
48. K. F. Nielsen, M. O. Hansen, T. O. Larsen, and U. Thrane, toxin Protocols, vol. 157, Humana Press, Totowa, NJ, 2000,
Int. Biodeterior. Biodegrad. 42, 1–7 (1998). pp. 173–194.
49. M. A. Andersson et al., Appl. Environ. Microbiol. 63, 387–393 80. H. Kaise et al., J.C.S. Chem. Comm. 726–727 (1979).
(1997). 81. W. A. Ayer and S. Miao, Can. J. Chem. 71, 487–493 (1993).
50. E. Johanning et al., Mycotox. Res. 14, 60–73 (1998). 82. A. D. Rodrı́guez, E. González, and C. Ramı́rez, Tetrahedron
51. S. M. Flappan, J. Portnoy, P. Jones, and C. Barnes, Environ. 54, 11683–11729 (1998).
Health Persp. 107, 927–930 (1999). 83. M. Ruotsalainen et al., Environ. Toxicol. Pharmacol. 6,
52. E. Johanning et al., Int. Arch. Occup. Environ. Health 68, 193–199 (1998).
207–218 (1996). 84. M. A. Hegenbarth, Pediatr. Emer. Care 15, 288–293 (1999).
53. M. J. Hodgson et al., J. Occup. Environ. Med. 40, 241–249
85. O. Elidemir, G. N. Colasurdo, S. N. Rossmann, and L. L. Fan,
(1998).
Pediatrics 104, 964–966 (1999).
54. K. F. Nielsen et al., Int. Biodeterior. Biodegrad. 42, 9–16
86. F. Fung, R. Clark, and S. Williams, J. Toxicol. Clin. Toxicol.
(1998).
36, 79–86 (1998).
55. T. Tuomi et al., Appl. Environ. Microbiol. 66, 1899–1904
87. H. A. Burge, in E. Johanning and C. S. Yang, eds.,
(2000).
Bioaerosols, Fungi, and Mycotoxins: Health Effects,
56. E. Johanning, P. R. Morey, and B. B. Jarvis, Indoor Air’93 1, Assessment, Prevention, and Control, Boyd Printing, Albany,
225–230 (1993).
N.Y., 1999, pp. 306–312.
57. H. B. Schiefer, in E. Johnanning and C. S. Yang, eds., Fungi
88. D. L. Sudakin, Meds. Gen. Med. February 29 (2000).
and Bacteria in Indoor Air Environments, Eastern New
York Occupational Health Program, Latham, N.Y., 1995, 89. Anonymous MMWR Weekly 49, 180–184 (2000).
pp. 69–72. 90. C. A. Robbins et al., Appl. Occup. Environ. Hyg. 15, 773–778
58. J. Forgacs, in S. Kadis, A. Ceigler, and S. J. Ajl, eds., (2000).
Microbial Toxins, vol. VIII, Academic Press, New York, 1972, 91. M. Nikulin, K. E. Reijula, B. B. Jarvis, and E.-L. Hintikka,
pp. 95–128. Int. J. Exp. Pathol. 77, 213–218 (1996).
59. D. J. Schneider et al., J. S. Afr. Vet. Assoc. 50, 73–81 (1979). 92. M. Nikulin et al., Fundam. Appl. Toxicol. 35, 182–188 (1997).
 ALGAE BIOTECHNOLOGY 189

93. C. Y. Rao, J. D. Brain, and H. A. Burge, Appl. Environ. the health and pharmaceutical industries, and for the
Microbiol. 66, 2817–2821 (2000). manufacturers of animal feed.
94. A. Breitholtz, M. Olsen, Å. Dahlback, and K. Hult, Food Spirulina is the most widely cultivated cyanobacteria
Addit. Contamin. 8, 183–192 (1991). (Fig. 1). It is a prokaryotic organism containing blue
phycocyanin in addition to chlorophyll. It is found in lakes
that are rich in salts in Central and South America and
in Africa. It is characterized by forming helical strands
ALGAE. See MEROPLANKTON; PLANKTONIC ALGAE IN THE of cells in aqueous environments and grows in alkaline
MARINE ENVIRONMENT; PRIMARY PRODUCTIVITY IN THE MARINE media, surviving pH of 11. It can grow in temperatures
ENVIRONMENT as low as 18 ° C, but the optimum is 35 to 37 ° C. It
grows well in mass cultivation in open, shallow racetrack
ponds, or in closed tube systems. The media for growth
was developed by Zarouk (2) but was modified to reduce
the costs by substituting cheaper substances, such as
ALGAE BIOTECHNOLOGY ammonia or urea for nitrate. Spirulina is grown for its
high protein content (71%) and is widely marketed as a
BONNIE K. LUSTIGMAN health food product and for animal feed. In addition to its
LEE H. LEE high protein content, it has essential fatty acids, minerals,
Montclair State University and vitamins. Most health benefits to humans claimed for
Upper Montclair, New Jersey supplementation come from anecdotes and not scientific
research. In vitro and animal tests have demonstrated a
In recent years there has been increased interest in wide range of healthful benefits including anticancer (3),
the use of algae for biotechnology. Algal biotechnology lipid lowering (4), and blood-iron status enhancing (5).
represents an area of current and future development Chlorella vulgaris was the first microalgae to be
as new sources for products are being explored. Algae isolated. It is a unicellular green algae and does not have
require minimal substances for growth and provide a flagella (Fig. 2). All strains have a cell wall, and asexual
rich source for bioproducts. Traditionally, seaweeds were reproduction occurs within the parent cell wall producing
harvested and used as a source of complex carbohydrates four to 16 new cells. The cell wall then breaks to release
such as agar and carrageenan. People in several countries, the offspring. Chlorella has high tolerance for acidic pH
particularly in Asia, have used algae as a food source for ranging from 2.0 to 6.0. Its salinity tolerance ranges from
centuries. Yet, of more than 30,000 species of algae known, 0 to 5%. Optimum temperature varies depending on the
only a small number have been exploited for commercial strain, ranging from 25 to 38 ° C (2). Maximum growth
purposes. Spirulina has been cultivated for several years rate can be induced by the use of mixotrophic cultures
as a health food product. Biotechnology for valuable (supplementation with acetic acid), rather than strict
biochemicals from algae is currently a reality, and growth reliance on autotrophic growth (6). Chlorella is grown
of algae through aquaculture has developed into an commercially for its chlorophyll and protein content. It
important industry. Several countries produce carotenoids is used as a health-food supplement and for animal feed.
by growing Dunaliella and Haematococcus and harvesting Chlorella can be used for bioremediation of heavy metals
the pigments. With the search for new natural products in wastewater (Fig. 3).
for use in pharmaceuticals, food materials, cosmetics, and Dunaliella are unicellular, flagellated, green algae that
other uses increasing, algae represent a likely resource to lack a cell wall (Fig. 4). There are halotolerant and
be explored for future development.
Algae, both micro- and macro-, are increasingly being
sought as a means of bioremediation that can compete
with other technologies in terms of efficiency, cost,
and reliability. The application for bioremediation and
for natural products derived from algae is even being
combined, with algae being grown on wastewater and their
proteins harvested for industrial use. Algal biotechnology
represents an important area for current and future
research.

COMMERCIALLY IMPORTANT ALGAE

Among the thousands of marine, freshwater, and soil

microalgae, about 100 species have been shown to
produce valuable compounds (1), a few of which have
gained considerable significance commercially. Spirulina,
Chlorella, Dunaliella, Haematococcus, and Porphyridium
are five genera of microalgae with great economic Figure 1. Cyanobacteria-Spirulina sp. http://vis-pc.plantbio.
relevance. Cultivation of these species is important for ohiou.edu/algaeimage/pages/spirulina.html See color insert.
190 ALGAE BIOTECHNOLOGY

Figure 2. Chlorella vulgaris, http://www.ifa.au.dk/∼jabraham

/algae.htm See color insert.

halophilic strains; the species grows in a range from

brackish to hypersaline waters. Optimum conditions for
growth vary depending on the strain involved, and the
interaction between the physical and chemical growth Figure 3. Bioremediation of heavy metals in wastewater,
conditions. Several different media have been developed http://zobell.biol.tsukuba.ac.jp/∼inouye/
for growth, with commercially important strains adapted ino/g/chl/chl picl.html See color insert.
to high salt concentrations. These strains were isolated
from environments such as the Dead Sea. Temperature
requirements vary depending on the light intensity and
salinity. With high salt concentrations, high temperatures,
and intense light, halophilic varieties of Dunaliella
produce significant quantities of beta-carotene. This is
harvested for use in health-food supplements and as
coloration in foods and cosmetics. In response to increasing
osmotic pressure, Dunaliella produces glycerol to maintain
osmotic balance, and this is also harvested for commercial
distribution (2).
Haematococcus pluvialis is a unicellular green algae
that has two anterior flagella (Fig. 5). It grows in
coastal rock pools, water holes, and other similar small
bodies of water. Its optimum temperature is between 15
and 25 ° C. Haematococcus pluvialis is able to respond
to environmental stress by changing from the motile,
vegetative green cell form to red stationary aplanospores.
This occurs when nutrients become scarce or under
hypersaline conditions, high temperature, or intense light.
In the cyst phase, it produces high amounts of carotenoids
and astaxanthin, 4 to 6% or more of their dry weight (7).
Astaxanthin is a red pigment that is valued commercially
for cosmetics and as an antioxidant. It is also used as
a color additive in feed for salmon to enhance the pink Figure 4. Dunaliella, http://zobell.biol.tsukuba.ac.jp/∼inouye/
color (Food and Drug Administration 21 CFR Part 73 ino/g/chl/chl\picl.html See color insert.
 ALGAE BIOTECHNOLOGY 191

[Docket No. 98C-0212]). These characteristics have led Cultivation of algae can occur in areas that are not
to increased interest in developing culture methods for suitable for conventional agriculture, but have high light
commercial exploitation. availability and warm temperatures, making this suitable
Porphyridium is a unicellular red algae (Fig. 6). to thirdworld nations. Many species can be grown in saline
Porphyridium has been isolated from a wide range of water or in marine environments. All systems for mass
environments, freshwater streams, brackish and marine production must include supply systems, culture systems,
water, and damp soils. It is characterized by a single harvesters, and processing equipment (Fig. 7). Special
large stellate chloroplast. It lacks a true cell wall, but technology is needed for each of the phases of the growing
is surrounded by a gel polysaccharide layer that causes process: cultivation, enrichment of the required product,
clumping of the cells so that they form loose colonies. harvesting, and extraction of the products. The phases are
Its pigment, phycoerythrin, has commercial significance dependent on each other and the process varies depending
for cosmetics, medical diagnostics, and food coloring. on the algae grown and the product to be harvested.
Porphyridium is grown in large-scale open ponds and The standard method used for large-scale production
in closed tube systems in marine media. is shallow ponds, 10 to 20 cm deep that are open or
covered. The ponds are circular or oval and are stirred
by paddle wheels to provide an even suspension and
CULTURE TECHNOLOGY prevent the algae from settling out. The most common
commercial structure is an open raceway, which is
To produce sufficient quantities of algae for industrial stirred by paddle wheels. There are several problems
purposes, methodologies for mass cultivation and harvest with large-scale cultivation, mainly concerned with
have been developed. Considerable technological research low productivity and contamination by microorganisms
has been undertaken to determine the optimum conditions and chemicals (8). Guterman and Ben-Yaakov proposed
for growth and to remove problems with contamination. application of modified, simple, mathematical algorithms
to computerized control systems to provide for optimum
operating conditions by automatic systems in which the
parameters of light intensity, optical density, pH, and
temperature are included in the formula to calculate the
conditions needed to increase yield (9).
Closed, long, polyethylene sleeves, first proposed
by Trotta in 1980 (10), have been used to limit the
problems found with open ponds. There are various
types of horizontal or vertical closed systems including
horizontally arranged glass or polyethylene tubes (11).
The tube photobioreactors are arranged as vertical closed
bioreactors that have limited contact with the external
environment, especially compared to open ponds. These
systems reduce the problems of contamination and
increase control over temperature, evaporation, and light
availability (12). Comparison of near horizontal straight
tubular reactors and near horizontal flat panels in outdoor
Figure 5. Haematococcus, http://www.microscopy-uk.org.uk/ cultivation of Spirulina showed that the photosynthetic
mag/wimsmall/extra/haema.html See color insert. efficiency in the tubular systems was significantly higher
because the curved surface reduced the light saturation

Figure 6. Rhodophyta-Porphyridium purpureum http://vis-pc.

plantbio.ohiou.edu/algaeimage/pages/Porphyridium.html Figure 7. http://members.nbci.com/ XMCM/rekel/ushn/helo/
See color insert. photo daphnia.htm See color insert.
192 ALGAE BIOTECHNOLOGY

effect (13). Comparison of curved versus straight tubes except for the sun (12). The advantage of the closed
indicates that curved tubes are superior to straight system is to limit contamination and fluctuations in
ones (14) (Fig. 8). environmental conditions. In other localities, where the
Water is added by a system of tubes that connect growth of algae is linked to remediation of agricultural
the sleeves to a control center where the temperature wastewater or similar environments, other technologies
and the densities of the materials are controlled and may be applicable (19,20). It is easier to commercially
harvesting begins. Models have been developed to estimate grow hardy organisms, such as Chlorella, Spirulina, and
the biomass productivity depending on the solar irradiance Dunaliella than Haematococcus, which cannot compete
on the culture surface (15). The determination of the final with contaminant organisms in open outdoor systems.
product is subject to the environment present (16). Thus, Haematococcus, which produces the pigment astaxanthin
changes in the composition of the medium, light intensity, in the cyst phase, is grown in a combination of enclosed
and gas flow will affect the productivity of the algal outdoor photobioreactors and pond-culture systems (1)
cultures (17). Low-dose ultrasonic treatment can affect as well as other culture methods intended to increase
the growth rate and biomass yield of algae depending on yield (21).
the species of algae cultivated; the ultrasonic treatment The yield and type of product can vary depending on
either increasing or decreasing growth rates (18). the nutrients and other growth conditions supplied to
Microalgae can be grown by various technological the algae. Nitrogen deprivation can increase the yield of
means that are dependent on the needs and economics polysaccharides of Porphyridium (22), but higher nutrient
of the environment. Algae exist primarily in freshwater availability will increase lipid content of Dunaliella
and marine habitats. They acquire the substances they tertiolecta (23). High biomass protein production can be
need for growth and metabolism (water, carbon dioxide, achieved by growth with high nitrogen content, such as
and nutrients) directly from this environment. In Israel, swine wastewater (19). Temperature can be an important
Dunaliella and Porphyridium are grown in vertical sleeves factor affecting the productivity and biomass in Spirulina
in which the growth conditions are closely monitored grown outdoors (24). The growth stage can be followed
and controlled. In these rows of vertical plastic sleeves, by an enrichment stage aimed at the concentration of
they are separated from the external environment, the desired product (25,26). The algae must be harvested
from the culture medium, concentrated, and dried. The
harvesting technique employed depends on the algae. For
microalgae, harvesting occurs by centrifugation, filtration,
flocculation and settling, or flotation, depending on the
organism involved. Degradation of the product, as with
beta-carotene, can occur because of exposure to light and
air (27). Irradiation can change the proportions of the
stereoisomers of beta-carotene in Dunaliella (28).
With the high costs involved in the cultivation of
microalgae and harvesting the bioproducts, it is probable
that in the future the identification of the genes
responsible for the production of these products will
take place. Genetic manipulation will then be used
for enhancement of these products in organisms, not
necessarily algae. Identification of some of the genes
responsible for production of astaxanthin (29,30) and their
transfer into tobacco plants has taken place (31).

NATURAL PRODUCTS FROM ALGAE

It is in the area of natural products that algae have

become increasingly important as sources for the health
and pharmaceutical industry and for use in animal feed.
Algae have long been used as a source of commercial
products, particularly for the food and pharmaceutical
industry. These products such as agar, sodium alginate,
and carrageenan are complex polysaccharides whose usage
is well established. They are used as thickening agents,
emulsifiers, and gelling agents in food products and paints,
photographic films, and drug formulations. With increased
consumer interest in natural products as opposed to
synthetic ones, attention has been focused on algae as
a source of these substances. Marine organisms including
Figure 8. Polyethylene culture sleeves from Ben–Gurion Uni- algae have been an important source of natural products
versity. Provided by Dr. Lustigman and Dr. Lee. See color insert. with biomedical potential (32).
 ALGAE BIOTECHNOLOGY 193

Protein show that high supplementary doses leads to precancer-

ous lesions in ferrets exposed to tobacco smoke, whereas
Although other algae also have high protein content,
low levels were antiproliferative (51).
Spirulina is probably the most widely known algae grown
The pathway for biosynthesis of beta-carotene in
for its natural products. Because its protein is complete,
Dunaliella has shown that several intermediates are
Spirulina is raised for use in protein supplements and for
produced, terminating as two stereoisomers (all-trans and
animal feed. In addition to protein, it produces essential
9-cis) found in equal concentration. The synthetic variety
fatty acids, minerals, and vitamins.
consists of the all-trans stereoisomer. In view of the
question as to the possible greater value of natural versus
Pigments synthetic beta-carotene, studies have been performed
comparing serum and tissue concentrations of the all-trans
Carotenoids. One of the most widely used products
and 9-cis isomers. The steroisometric mixture of beta-
derived from algae are the pigments that are harvested
carotene in Dunaliella has been shown to be preferentially
from mass cultures. Microalgae are frequently grown for absorbed in animal tissues (34), but much of the 9-
their pigments, particularly the lipid-soluble carotenoids. cis molecules are converted to all-trans (52,53). Serum
These are used for food coloration, cosmetics, and medic- analysis in humans shows that 9-cis-beta-carotene is more
inal purposes. Carotenoids are found in a large number efficient as an in vivo lipophilic antioxidant than all-
of organisms including bacteria, algae, higher plants, and trans-beta-carotene, indicating that natural sources may
animals. Carotenoids are lipid-soluble pigments made of be superior to synthetic ones (54). Comparative studies
isoprene units with a yellow-orange-red coloration. Beta- of natural and synthetic beta-carotene using in vitro tests
carotene is the most abundant of the carotenoids making indicate that natural, or a mixture of natural and synthetic
up 10% of the dry weight of Dunaliella bardawil and beta-carotene, was more effective in antigenotoxicity than
D. salina (33). synthetic, and may be of value as a supplementary
In plants, carotenoids are secondary photosynthetic treatment in cancer prevention (55). Other carotenoids
pigments that serve to absorb additional light energy for used commercially are xanthophylls, which are yellow-
the photosynthetic process. They increase in concentration orange in color, used as food dyes for poultry skin, egg
when the plant is undergoing stress such as high yolk, and salmon (56).
light intensity or high osmotic stress. This is true
particularly for microalgae such as Dunaliella, which has Astaxanthin. Another carotenoid is the red pigment
been cultivated commercially for several years. Extensive astaxanthin produced almost exclusively by Haematococ-
cultivation of Dunaliella occurs in Israel, Australia, United cus pluvialis cysts (Fig. 9). It has commercial value as a
States, Spain, and China (34–36). The amount of beta- red colorant in cosmetics and foods, and enhances the pink
carotene produced is directly related to the amount of color of salmon (57). The gene CrtO, responsible for the
light absorbed by the algae. Therefore this is a mechanism production of astaxanthin has been transferred to tobacco
to protect against high light intensity (37). Increased plants imparting red color in the manipulated plants (31).
salinity also increased the concentration of carotenoids, Astaxanthin present in cyst cells appears to function as
particularly beta-carotene (38). an antioxidizing agent against oxidative stress (58). It
Industrially, carotenoids have been used for food col- increases in oxidative stress along with superoxide dismu-
oration, cosmetics, and as a health-food product (39). tase, an antioxidative enzyme (59). It is being evaluated
Beta-carotene is a vitamin A precursor and acts as an for presumed cancer prevention (60). The development
antioxidant (40). Exposure of algal cells to various envi- of methods for increased production and purification of
ronmental stresses causes an increase in the presence of astaxanthin is currently a very active area of commercial
antioxidative enzymes and beta-carotene production (41). interest in several locations including Iceland, Israel, and
Beta-carotene has been shown to scavenge reactive oxygen Hawaii (1,61).
species generated in vitro (42) and to cause slight lowering
of spontaneous and X-ray-induced chromosomal damage in Phycobiliproteins. The water-soluble phycobiliproteins,
bone marrow cells (43). Beta-carotene has long been con- produced by the cyanobacteria and red algae, are other
sidered as an anticarcinogen (44,45). Because of its antiox- important pigments. These pigments create the blue-green
idant properties, it has been studied as a means of chemo- and deep red color, respectively. They have potential as
prevention to prevent development of gastric carcinoma food dyes and application in the cosmetic industry (62).
in high-risk populations with precancerous lesions (46). Phycocyanin, which is the blue-green pigment obtained
However, two large-scale studies, the Physicians’ Health from cyanobacteria, has been patented and is commercially
Study (PHS) and the Beta-Carotene and Retinol Efficacy available as Linablue (Dainippon Ink and Chemicals Inc.
Trial (CARET), question the benefit against heart disease 1980. Jap. Patent 8077890). It is used in the food indus-
or cancer (47). In fact, an increase in the risk of lung can- try for coloring dairy products, candy, and soft drinks.
cer among smokers who took beta-carotene supplements Spirulina platensis has been suggested for phycocyanin
was reported in the alpha tocopherol, beta-carotene cancer production (63). It has also been shown to have hep-
prevention study (48), and among smokers and asbestos atoprotective effects against toxic substances (64). The
workers in CARET (49), but not among male physicians red fluorescent phycobiliprotein, phycoerythrin, gives the
in the Physicians’ Health Study (50). Studies using ferrets red algae, Rhodophyta, their name and coloration. The
supplemented with high and low doses of beta-carotene phycobiliprotein from the red alga, Rhodella reticulata
194 ALGAE BIOTECHNOLOGY

Figure 10. Rhodella reticulata, http://www.cibnor.org/malgas

/ifotrhr.html See color insert.

actually produced by the marine algae that the fish con-

sume (74). Algae and cyanobacteria are rich in ω-3 fatty
acid homologs (75), and if the fatty acids are harvested
directly from the algae, the undesirable components, like
Figure 9. Astaxanthin, http://www.bioprocess.is/product.html
See color insert. the fishy smell and cholesterol, will not be present (76).
These fatty acids are suggested to be of value in preven-
tion of heart disease by reducing levels of plasma lipids
and lipoproteins. They act to decrease serum cholesterol
(Fig. 10), have been purified and could be considered as
a potential source of these substances for commercial and triglyceride levels (77). High quantities of essential
purposes (65). They are also found in several microal- fatty acids such as arachidonic and docosahexaenoic acids
gae including Porphyridium (66) and Synechococcus. The are found in marine algae. Greater variation exists in
purification of phycoerythrin from Synechococcus sp. DC-2 the quantities of fatty acids in different algae than in
has shown that two forms of phycoerythrin are pro- the class of fatty acid produced, although within the
duced (67). Heat-resistant C-phycocyanin has been puri- class unique metabolites may exist (78). The Phaeophyta
fied from the red alga Cyanidium caldarium, and has and Rhodophyta generally produce more fatty acids than
potential for use in industrial processes because it does the Chlorophyta, including eicosatetraenoic and eicos-
not denature up to 60 ° C (68). apentaenoic, and less linolenic acids (79). Arachidonic
These pigments are used for diagnostic purposes in acid (5,8,11,14-eicosatetraenoic acid) is a natural precur-
multiple-color-flow cytofluorimetric analysis. The phy- sor of a large family of structurally related C-20 compounds
cobiliproteins and molecules with specific biological including prostaglandins, leukotrienes, and prostacyclins.
activity bind to target molecules forming fluorescent Arachidonic acid and docosahexaenoic acid are impor-
conjugates. The phycoerythrin conjugates are solu- tant for neurological development in newborns (80). The
ble in water, have a long shelf life, and are not tropical red marine alga, Murrayella periclados has been
affected by most natural molecules (69). Glazer and shown to produce high quantities of eicosanoids, including
Stryer developed markers, phycoprobes, made of fluo- a leukotriene, usually found in animal sources (81). Other
rescent phycobiliproteins covalently bound to biologically red algae including Porphyridium, are also rich in high
active molecules (70). These include phycoerythrin- unsaturated fatty acids, which is considered as a potential
immunoglobulins, phycoerythrin-biotin, phycoerythrin- source of arachidonic acid (82,83). The green microalgae,
avidin, and phycoerythrin-protein (71). The use of Chlorella has been suggested for the commercial produc-
phycoerythrin-antibody conjugates allow for early detec- tion of gamma-linolenic acid (84) as has Spirulina (85).
tion of leukemia and other blood disorders (72). Commercial fish feed enriched with fatty acids derived
from algae produces better growth of fish larvae (86,87)
and juvenile scallops (88). There is considerable difference
Fatty Acids
in the success of the juvenile scallops, depending on the
Algae are also used commercially to obtain unsaturated species of algae fed. There was greater growth when those
fatty acids (73). Although fish oils are considered to be species having high levels of ω-3 and ω-6 polyunsaturated
the source of unsaturated fatty acids, these acids are fatty acids and carbohydrates were used (89).
 ALGAE BIOTECHNOLOGY 195

Polysaccharides
The Rhodophyta are the traditional source of complex
polysaccharides used commercially as gelling agents,
stabilizers, thickeners, and emulsifiers for dairy products
and other foods. They are also used in paints, photographic
films, and pharmaceuticals. The most widely used are
agar-agar and carregeenan. The insoluble carrageenan
fraction extracted from Stenogramme interrupta has also
shown antiviral and anticoagulant properties (90).
Polysaccharides derived from seaweeds are frequently
used as fiber for nutritional supplements. Seaweeds
such as Ulva, an edible green algae, have been raised
in aquaculture for its polysaccharides and used as
dietary fiber (91). For several years there has been
an increase in the development of additional marine
algal polysaccharides for pharmaceutical purposes (92).
The polysaccharide derivative, calcium spirulan, from
Spirulina has been shown to limit tumor metastasis (3).

Antimicrobials
Antibacterial. Because of the increase in the number
of strains of bacteria that are resistant to antibiotics, Figure 11. Chlorococcum oleofaciens, http://www.cibnor.mx/
there has been increased interest in the development malgas/efotcho.html See color insert.
of antimicrobial substances produced by algae. Most
of the antibiotics currently used clinically are derived
from substances produced by fungi or actinomycetes. Screening of 89 seaweeds from British Columbia, Canada,
However, the production of antibiotic substances from and Korea for substances with antiviral activity has
marine algae has long been known, dating back to shown that 37% of the species tested were able to inhibit
Pratt in 1942 (93). Since the 1960s, many studies have viruses (111).
been undertaken on the growth-inhibitory substances Natural substances derived from algae that were
produced by marine algae (94,95). Several reviews have active against HIV include terpenoids, xanthones, alka-
been published (32,96–98) on this topic. loids, flavenoids, polyphenols, and polysaccharides and
Among the seaweeds, the Chlorophyta, Phaeophyta, appear to be active against reverse transcriptase, pro-
and Rhodophyta have been surveyed for their activity. tease, and integrase (112). In particular, the sulfated
It has been shown that activity may be present, but polysaccharides, such as dextran sulfate, are known
it is species specific and varies depending on the sea- to interfere with the adsorption and penetration of
son (99–102). Cyanobacteria have been found to produce retro- and other viruses (113,114). Such substances
antibiotic substances and other potentially pharmacolog- have been extracted from Spirulina and Cochlodinium
ically active products (103,104). Green algae including polykrikoides (115–117). Two mechanisms have been
Chlorococcum (Fig. 11; 95) and Dunaliella (105,106), have proposed: (1) Inhibition of retroviral reverse transcrip-
also been shown to produce antibacterial substances. tase (118) and (2) binding to CD4 receptors (119). The
Many of the substances produced by marine microor- sulfated polysaccharide of Porphyridium (120) and that of
ganisms show a high degree of toxicity. The toxic nature of Spriulina (116) have shown activity against Herpes sim-
these substances, which include acrylic acid, phenols, ter- plex virus types 1 and 2 and Varicella zoster virus. The
penoids, and halogenated compounds, has made them unfit mechanism for action is the inhibition of production of
for, or have limited use in, commercial exploitation (107). new viral particles within host cells and/or prevention of
However, the need to develop new antibiotics because of adsorption of the virus into host cells (120).
increasing resistance by clinical bacteria to traditional The potential for the identification and development of
ones has led to renewed interest in the search for antibi- antimicrobial substances from algae is an area of great
otics from algal sources (108,109). A survey of 100 strains potential for future development. With the increasing
of marine microalgae from Japan determined that sev- proliferation of resistant bacteria and the need for
eral strains showed activity, with a species of Chlorella antiviral drugs, the exploitation of substances from algae
displaying the greatest results, producing a light-induced is sure to increase.
substance (84). In a similar survey of 84 marine algae, Additional uses for algal products are the proposal
Dunaliella primolecta showed the highest antibiotic activ- of development for use as natural oral deodorants
ity (106). Nine extracts made from 16 marine algae from by members of the brown algae belonging to the
the Atlantic coast of Brittany, France, showed activity Laminariales (121). Algae and other marine organisms
against isolates of marine fungi, bacteria, and yeasts (110). have been suggested as a source of substances that are
environmentally safe to control barnacles that live on
Antiviral. Development of substances that show activity ships’ hulls, oil platforms, and pipelines (122). Methods,
against viruses is another active area of research. such as new assay gels containing extracts from these
196 ALGAE BIOTECHNOLOGY

organisms, are being developed to assist in the screening Because of the environment they inhabit, sediment
of such substances (123). microorganisms exhibit a higher level of tolerance to
high heavy-metal concentrations. Aquatic microorganisms
ALGAE AS FOOD SOURCES display the lowest tolerance to heavy-metal concentrations
because the heavy-metal suspension in water is generally
As we search for new, safe, and less costly sources of very low (132).
feed for animals and humans, and for supplementary A microorganism’s ability to tolerate specific heavy-
dietary products like protein, vitamins, and minerals, metal toxicities is because of the organism’s exposure to
algae provide an important source for these substances. that substance. Thus, many bacterial strains show high
Algae have long been used as human food sources in tolerance to zinc (Zn) and copper (Cu), and few to silver
cultures with access to the marine environment, especially and arsenic because of the greater percentage of Zn and Cu
in Asian cultures. in the environment compared to silver and arsenic (133).
There is increased awareness of the potential for Collard and Matagne (134) showed that most toler-
use of agar by consumers in countries outside of ance appears to be physiological rather than genetic, but
Japan (124). Aquaculture has become an important genetically caused tolerance to heavy metals was present
source of food for humans and animals. Technological in Euglena gracilis (Fig. 12), Stigioclonium (Fig. 13),
advances have provided the ability to raise and harvest Chlorella vulgaris (Fig. 2), and Chlamydomonas rein-
algae grown in controlled environments, with specialized hardtii (Fig. 14). Cd tolerance in Chlamydomonas is
media formulated to increase yield. The blue-green alga, achieved by alteration in the metabolic pathways asso-
Spirulina, has been effectively marketed as a healthfood ciated with the chloroplast, causing reduced chlorophyll
supplement because of its high protein content (125). content and not by increased efficiency of a detoxifica-
Although other species, including Chlorella, Dunaliella, tion system (135). Pretreatment provides for increased
and Scenedesmus, are also suggested to be used for their tolerance. Preexposure of Euglena gracilis to low con-
protein content (126), Spirulina, in particular, has been centrations of Hg protected cells to concentrations up to
proposed for use in poultry feed, fish meal, and as a 5 µM. Hg pretreatment also conferred tolerance against
supplement in other animal feed (127–129). Since the cost Cd exposure and vice versa. These algae were able to
of traditional protein sources such as fish, groundnuts, tolerate higher levels of Pb as well (136). Some algal
and soybean meals is expensive, especially for developing species display cotolerance in which they can tolerate
nations, the search for less expensive alternative sources one or more additional metals to which they were not
such as algae, has increased (130,131). previously exposed (137). Exposure of Dunaliella minuta
Algae provide a source of high protein, comparable to Cu or Cd led to the acquisition of tolerance toward both
to those from conventional sources. Dunaliella, grown Cu and Cd (138).
under optimum conditions, can produce up to 70% Anacystis nidulans, (Fig. 15) a unicellular cyanobacte-
protein content (23). However, production costs are high ria, and Chlamydomonas reinhardtii, a unicellular green
and this limits use as animal feed, although many algae, have served as indicator species for a long-term
third world nations are exploring the cultivation of project that was undertaken to determine the inhibitory
microalgae for protein and other biochemicals. This, in concentrations of the heavy metals identified by the EPA.
conjunction with wastewater treatment (128), or to clean In these studies, Anacystis with A. nidulans was treated
liquid manure (129), provides dual benefits in terms of with the heavy metals in media with and without the
production and is ecologically sound. chelator, ethylene diamine tetraacetic acid (EDTA) (139).
The results of this survey indicate a wide range of
TOXICOLOGY inhibitory concentrations of the heavy metals. These
studies show the following order of inhibitory concen-
Algae, as simple microorganisms, can be used as trations: Hg > Tl > Al = Cu > Ni > Co > Se = Zn > Cd >
indicator species for environmental contamination and Mn > Ba = Pb with EDTA. A similar, but not identical pat-
ecotoxicology (130). Large quantities of heavy metals are tern was seen when EDTA was not used in the medium.
released into the environment mainly through industrial Without EDTA the results were: Hg > Cu > Tl > Ni >
discharges. Agricultural runoff and sewage treatment Co > Se = Zn > Cd > Mn > Ba. In this series, Pb was not
are also sources of environmental contamination. The studied and the results with manganese show a long lag
effect of these substances on algae will reflect in the period at lower concentrations before the onset of log
entire ecosystem because they are the world’s largest phase (140–148).
group of primary producers. Heavy metals exert their There does not appear to be consistency based
toxic effects by competing with essential metals for active on toxicity. Matulova, working with the green alga
enzyme or membrane protein sites, and by reacting with Scenedesmus quadricauda (Fig. 16), and using lethality
biologically active groups, thereby disrupting normal cell as the measurement found that the order was: Ni > Cr =
processing. The Environmental Protection Agency (EPA) Zn Pb (149). Rosko and Rachlin using Chlorella vulgaris
has listed major toxic target contaminants including cultures with EDTA found toxicity to heavy metals in
13 heavy metals. Of these contaminants, the ones that the following order: Cd > Cu > Hg > Zn Pb (150) and
have received the most attention are mercury(Hg), Raichlin and colleagues using Chlorella saccharophila
lead(Pb), and cadmium(Cd), although several others are found that the order was Cd > Cu Zn Pb (151).
of important environmental significance. However, Raichlin and colleagues using the freshwater
 ALGAE BIOTECHNOLOGY 197

Figure 12. Euglenophyta-Euglena gracilis http://vis-pc.plant-

bio.ohiou.edu/algaeimage/pages/Euglena.html See color insert.

Figure 13. Chlorophyta-Stigeoclonium sp. http://vis-pc.plant-

diatom Navicula incerta (Fig. 17) found the toxic hierarchy bio.ohiou.edu/algaeimage/pages/Stigeoclonium.html See color
insert.
to be Cd > Pb Zn > Cu (152). Comparison in terms of
toxicity to thallium shows a wide difference between
Anacysitis nidulans and Chlamydomonas reinhardtii, On the other hand, phytoplankton appear to be
with inhibitory concentrations at 10 mg/L and 0.25 mg/L, highly sensitive to Hg; most organisms display greatest
respectively (148). sensitivity to this metallic ion (153–156). Hg is one
These results contrast to environmental studies that of the most lethal heavy metals studied, with no
contain a mixed phytoplankton population. Studies of the known biological function. There are three forms of Hg,
Saanich inlet showed the order of toxicity to be: Hg elemental, inorganic, and organic compounds. Inorganic
Cu = Pb > Cd > Ni = Cr in two experiments (153). Krock Hg is far less dangerous than organic methyl Hg.
and Mason found the order of toxicity based on inhibition Because of limited solubility, Hg compounds are deposited
of photosynthesis in phytoplankton communities of San in bottom mud of rivers, lakes, and so on. Through
Francisco Bay to be: Hg > Cu Cd = Cr = Ni = Pb (154). biological processes Hg compounds are converted into
It appears that the results depend largely on the type of dimethyl and methyl Hg and thereby enter the food
algae used, and this constitutes an important factor in chain (157).
determining the effects of heavy metal toxicity. Cd has widespread industrial usage and is capa-
Pb is the most abundant and widely used heavy metal, ble of bioaccumulation (158). It is released as a
although recent use has declined with lead-free gasoline. result of electroplating and alloy preparation. Plants
It exerts its toxicity more in terms of chlorophyll a absorb Cd more readily than Pb. Cd binds to organic
content than on cell division (150) and binds strongly to molecules through sulfur and nitrogen (159), thereby
cell membranes (155). It appears to be the most widely inactivating proteins. The degree of toxicity to Cd
tolerated of the major heavy-metal contaminants for appears to vary depending on the algae used as a
phytoplankton (140–151). model (150–154).
198 ALGAE BIOTECHNOLOGY

Figure 16. Chlorophyta-Scenedesmus sp. http://vis-pc.plantbio.

ohiou.edu/algaeimage/pages/Scenedesmus.html See color insert.

Figure 14. Chlorophyta-Chlamydomonas sp. http://vis-pc.plan-

tbio.ohiou.edu / algaeimage / pages / Chlamydomonas.html See
color insert.

Figure 17. Bacillariophyceae-Navicula http://vis-pc.plantbio.

ohiou.edu/algaeimage/pages/navicula.html See color insert.

of heavy metals. These compounds act by binding cys-

teine and lysine to heavy metals (165) thereby providing
a mechanism by which the ions are incapable of adversely
interacting with essential metabolic enzymes. Another
detoxification mechanism is conformational changes in
Figure 15. Anacystis nidulans, provided by Dr. Lee H. Lee. See the cell wall that will trap the heavy-metal ions and
color insert. prevent them from entering the cell (133). Most met-
als, when absorbed into the cytosol are concentrated in
the vacuole and excluded from vital portions of the cell,
Resistance of algae to toxicants has significance for such as the nucleus and chloroplast (166). Nassiri and col-
bioremediation. Elucidation of the mechanisms for resis- leagues showed that microalgae could develop a tolerance
tance may produce more efficient means to remediate toward metallic pollutants by exclusion or by internal trap-
contaminated waters or reclaim the metals involved. ping (159). Skeletonema costatum, when exposed to Cd,
Algal resistance to heavy metals may be due to sev- Hg, or Zn forms electron-dense inclusions, multivesicu-
eral mechanisms. These include production of extra- late bodies, and cytoplasmic tubules. Several trace metals
cellular chelating factors (160), the presence of heavy including Cu, Pb, Hg, and Zn retard the flow of elec-
metal–resistant genes present in plasmids, or as part trons in electron-transfer reactions of plant mitochondria
of the genome (134,161,162). Specific enzymes have been and chloroplasts (167). It is currently possible to make
identified that are capable of reducing toxic heavy metals use of some of these resistance mechanisms in products
to forms that are nontoxic (163,164). Metallothionein- derived from algae. Thus dried biomass produced from
like compounds have been identified in several types of Chlorella has been proposed for bioremediation (168) and
microorganisms that function in reducing the toxicity polysaccharides from Ascophyllum can be used to remove
 ALGAE BIOTECHNOLOGY 199

heavy metals (169). Similarly, genetic engineering may be

used to produce strains of microalgae with greater resis-
tance and ability to adsorb pollutants than those currently
in use.

REMEDIATION

In view of increasing environmental protection control,

the use of microorganisms has received attention in
recent years as a means of removing soluble metal-
lic ions from wastewaters. Existing treatment technolo-
gies may have problems with removal of the ions;
therefore, alternative treatment techniques have been
explored (170). Absorption of heavy metals by microbial
cells has become recognized as an alternative to exist-
ing methods such as precipitation methods or use of
synthetic ion exchangers. Several microorganisms have
been studied, including bacteria (171), fungi (172), and Figure 18. Oscillatoria http://vis-pc.plantbio.ohiou.edu/algae-
algae. image/pages/oscillatoria.html See color insert.
Algae, like other aquatic microorganisms, have the abil-
ity to take up dissolved metals and other contaminants
from the environment. Biooxidation by Cryptophyceae in groups are considered to be the major metal-sequestering
lake water showed processes that can detoxify organic sites (104).
contaminants such as resorcinol (173). Polycyclic aromatic pH plays an important role in the removal of metallic
amines are promutagenic/procarcinogenic. Because they ions. The pH for optimum adsorption depends on the
can accumulate and metabolize promutagenic pollutants, physiochemical conditions of the metal ions and the
algae, in combination with other microorganisms, can be algae (168,180). The pH will affect the bond formation
used in aquatic environments for the detection and con- of the algal cell wall, particularly the amino groups,
version of environmental promutagens (174). Microalgae to the metallic ions. It has been found that higher
such as Botryococcus-braunii, can be used for hydrocar- pH provides for greater adsorption of Pb and (Ni)
bon recovery, but results in lower cell viability (175,176). with Chlorella (168), and with the brown algae, Fucus,
Bacteria and microalgae act symbiotically in wastewater Ascophyllum, and Sargassum (183). However, at pH
treatment to provide for removal/metabolism of materi- values above 5.0 for Pb and Cu, and 6.7 for Ni, the ions
als (177). The process, microalgae use the end-products precipitate out of solution. pH may also affect the ionic
of bacterial metabolism (carbon dioxide, ammonia, etc.,) form of the ions and therefore plays a role with regard to
and in turn, supply bacteria with oxygen for total degra- toxicity.
dation of organic compounds. The algae serve for reox- Many algae are known to be able to absorb and concen-
idation and mineralization, and contribute to the food trate metal ions from dilute solutions and to accumulate
chain (178). Filamentous Oscillatoria has been proposed them within the structure of the cells. The accumu-
as an agent to remove nutrients from secondary acti- lation of trace metals can be of particular importance
vated sludge effluent, and the proteins produced could from several perspectives. Although accumulation of trace
be utilized commercially (179). Spirulina maxima grown metals may not be of sufficient toxicity to the algae, mag-
in seawater has been suggested as a means of control- nification can occur up the food chain, disrupting the
ling pollution from anaerobic effluents of animal wastes metabolism of higher organisms, thereby causing biomag-
and for harvest for high protein (19). Cultures of algae nification. Similarly, absorption and accumulation can
and Daphnia are raised in Holland on sewage effluent be used for metal reclamation with gold (Au), cobalt,
(Fig. 18). and other ions (184–186) and remediation of industrial
Algal cell walls contain polysaccharides and pro- wastewaters (187,188). Use of algae for the biosorption
teins having amino, carboxyl, phosphate, and sul- of radioactive Strontium 90 and Yttrium 90 allows for
phate groups. The amino and carboxyl groups, and concentrating and removal of these radionuclides (189).
the nitrogen and oxygen of the peptide bond could be Acidophilic microorganisms including Euglena sp., have
used for coordination bonding with metallic ions (180). been proposed for the treatment of highly acidic mine-
Algae need not be living to act for biosorption. Het- tailing waste to remove Ni–Cu sulphide minerals and
erosigma akashiwo (hada) Hada, which was inactivated other toxic heavy metals (190). Use of microorganisms
by steam sterilization, showed biosorption of bivalent including algae for the removal of Hg is less satisfactory
metal ions owing to binding to carboxylic- and phos- because whereas the cells absorb the Hg, volatilization of
phatic sites (181). Modification of carboxyl and amino significant quantities occurs (191).
surface functional groups has been shown to decrease Metal uptake by intact algal cells has been found to con-
the adsorption capacity of Cd (II) and Zn (II) binding sist of two processes: (1) A rapid passive adsorption to the
with Chlorella (182) indicating the role these groups cell surface and (2) a slower metabolism-dependent uptake
play with regard to metallic ion absorption. Carboxyl into the cytosol. Dead cells accumulate heavy-metal ions
200 ALGAE BIOTECHNOLOGY

to the same or greater extent as living cells. The pas- produced from a wide range of algal and nonalgal sources,
sive process of adsorption is called biosorption (192). The including dried biomass of Chlorella (168). Similarly,
sorption phenomenon is most frequently expressed by the screening of 191 strains of marine microalgae showed
Freundlich adsorption isotherm and is used for calculat- that 24 of these were able to remove Cd from their
ing the residual or adsorbed metal-ion concentration with growth media. Chlorella showed the highest rate of
application in wastewater treatment (180). The Langmuir removal. Six strains out of 19 green algae and one
isotherm is used to quantitatively determine sorption per- out of five cyanobacteria removed more than 10%
formance and reflects the attraction between the sorbent of total Cd from the medium. Although 12 days of
and the sorbate (168). The biopolymers involved include incubation showed that most of the Cd was accumulated
Na/Ca alginate, agarose, and cellulose-acetate. These intracellularly, dried cells adsorb the ions more quickly
biopolymers are nontoxic, selective, efficient, and inexpen- than living cells (203). Ascophyllum nodosum produces
sive (193). When heavy metals are accumulated intracel- a polysaccharide that provides it with the ability
lularly, toxicity can occur that is not present in cell surface to remove and sequester high levels of Cd/g dry
adsorption (194). Wehrheim and Wettern (195) compared weight (169).
dry weight of whole cells and isolated cell walls rather than Biosorption of metals is not based on only one
definite surface area and found that isolated cell walls mechanism. It consists of several mechanisms that
have 10-fold higher accumulation rates than whole cells vary according to the species used, the origin of the
for Cd, Cu, and Pb. Similar rates have been observed with biomass, and its processing (197). The mechanisms involve
Vaucheria (Fig. 19) (196), fungi, and bacteria (197–199). chelation, adsorption by physical forces, and binding to cell
However, the amount of uptake varies depending on micro- wall polysaccharides and amino, phosphate, sulfhydryl,
bial biomass employed and the type of metal ion (200). and carboxyl groups in proteins. Within cells, heavy
As a result, biosorption processes, which are the passive metals adsorbed strongly to biological membranes or to
accumulation of metals by biomass, have been proposed cytoplasmic polyphosphate bodies (152,204).
as an efficient and cost-effective way of removing toxic Combinations of ions can have a variety of effects in
heavy metals from industrial effluents (201). For cost- biosorption. Combination of Cr VI and Fe III indicate that
effectiveness, they can be manipulated for better efficiency these ions can act antagonistically with regard to levels of
and multiple reuse (202). Biosorption processes can be adsorption of Cr and Fe in Chorella vulgaris (205). Studies
with reinforced Ascophyllum nodosum using two and three
metal sorption systems containing Cd, Cu, and Zn have
shown that Zn and Cu solutions showed an antagonistic
effect on the uptake of Cd (197). Combinations of Cd,
Cu, and Zn with formaldehyde cross-linked Ascophyllum
showed that the total metal sorption increases at low
concentrations, but each metal can inhibit the sorption
of the others. At high metal concentrations the total
metal sorption uptake is either constant or slightly
decreases (206). However, there was little effect on uptake
of Au by Sargassum biomass by other metal ions (207).
Chlorella has been found to be effective in adsorbing
metallic ions (203). Dried cells showed high capacity for Pb
adsorption and reached saturation within 10–15 minutes,
with pH 5.0 being optimum (168). Comparison of whole
cells and isolated cell walls indicated that whole cells
accumulated more metal ions than cell walls using Cd,
Cu, and Pb (195). Similarly, Chlorella was shown to have a
stronger affinity for Cd II than Zn in uptake studies (182).

CONCLUSION

It is expected that future trends in algal biotechnology

will involve the commercial development of additional
products derived from algae. With improvements in
culture methodology and harvesting of natural products,
the cost for production of bioproducts from algae should
make them competitive compared to production from
synthetic sources. Similarly, the need for cheap, safe,
animal feed will bring about increased aquaculture of
algae for feed. Fish farming is increasing and therefore the
Figure 19. Xanthophyceae-Vaucheria http://vis-pc.plantbio. need for high-quality feed. The search for antimicrobials,
ohiou.edu/algaeimage/pages/Vaucheria.html See color insert. especially those that have antiviral properties or new
 ALGAE BIOTECHNOLOGY 201

antibiotics to fight resistant bacterial strains represents 28. S. C. Orset and A. J. Young, Plant Physiol. (Rockville) 122,
an important area for future development. 609–617 (2000).
Genetic engineering provides for a new area of 29. G. Hershkovits, Z. Dubinsky, and D. J. Katcoff, Mol. Gen.
biotechnology of algae. As genes are isolated and their Genet. 254, 345–350 (1997).
functions determined, transfer of valuable genes from 30. H. Linden, Biochim. Biophys. Acta 1,446, 203–212 (1999).
algae to other species will provide another avenue for 31. V. Mann, M. Harker, I. Pecker, and J. Hirschberg, Nat.
bioproducts to be produced for commercial exploitation. Biotechnol. 18, 888–892 (2000).
Algal biotechnology is a field that has only begun to be 32. B. Carte, Bioscience 46, 271–286 (1996).
developed and one with much promise for the future. 33. A. Ben Amotz, A. Katz, and M. Avron, J. Phycol. 18,
529–537 (1982).
34. A. B. Amotz, A. Shaish, and M. Avron, Bioresour. Tech. 38,
BIBLIOGRAPHY 233–235 (1991).
35. E. S. Miliko, Microbiology 32, 299–307 (1963).
1. M. A. Borowitzka, J. Appl. Phycol. 7, 3–15 (1995).
36. E. S. Miliko, Microbiology 32, 590–597 (1963).
2. A. Richmond, ed., CRC Handbook of Microalgal Mass
37. A. B. Amotz, J. Gressel, and M. Avron, J. Phycol. 23,
Culture, CRC Press, Boca Raton, Fla., 1986.
176–181 (1987).
3. T. Mishima, J. Murata, and M. Toyoshima, Clin. Exp.
38. M. A. Borowitzka, J. Borowitzkal, and D. Kessly, J. Appl.
Metastasis, 16, 541–550 (1998).
Phycol. 2, 111–120 (1990).
4. C. G. de Rivera, R. Miranda-Zammora, and J. C. Diaz-
39. H. Klani and J. C. Bauernfeind, in J. C. Bauernfeind, ed.,
Zagoya, Life Sci. 53, 57–61 (1993).
Carotenoids as Colorant and Vitamin A Precursors,
5. R. Kapoor and U. Mehta, Plant Foods Hum. Nutr. 52, Academic Press, New York, 1981, pp. 147–317.
315–324 (1998).
40. M. A. Azuine, U. C. Goswami, J. J. Kaya, and S. V. Bhicle,
6. L. Lin and T. Chen, J. Chin. Agric. Chem. Soc., 32, 91–102 Nutr. Cancer 17, 287–295 (1992).
(1994).
41. N. Mallick and F. H. Mohn, J. Plant Physiol. 157, 183–193
7. P. Z. Margalith, Appl. Microbiol. Biotechnol., 51, 431–438 (2000).
(1999). 42. M. V. R. Kumari, T. Yoneda, and M. Hiramatsu, Biochem.
8. A. Richmond, CRC Crit. Rev. Biotechnol. 4, 369–438 (1986). Mol. Biol. Intl. 38, 1163–1170 (1996).
9. H. Guterman and S. Ben-Yaakov, Biotechnol. Bioeng. 35, 43. K. Umegaki, N. Takeuchi, S. Ikegami, and T. Ichikawa,
417–426 (1990). Nutr. Cancer 22, 277–284 (1994).
10. A. Trotta, Aquaculture 22, 383–387 (1989). 44. S. T. Mayne and R. S. Parker, Nutr. Cancer 12, 225–236
11. D. Chaumont, C. Thepenier, C. Gudin, and C. Junjas, in (1989).
T. Stadler, ed., Algal Biotechnology, Elsevier Applied 45. R. G. Ziegler, J. Nutr. 119, 116–122 (1989).
Science, London, 1988, pp. 199–208. 46. P. Correa et al., J. Natl. Cancer Inst. 92, 1881–1888 (2000).
12. E. Cohen and S. Arad, Biomass 18, 59–67 (1989). 47. HYPERLINK ‘‘http://www.heartinfo.com/betacrt.html’’
13. M. R. Tredici and G. C. Zittelli, Biotechnol. Bioeng. 57, 48. D. Albanes et al., J. Natl. Cancer Inst. 88, 1560–1570 (1996).
187–197 (1998).
49. G. S. Omenn et al., J. Natl. Cancer Inst. 88, 1550–1559
14. P. Carlozzi and G. Torzillo, Appl. Microbiol. Biotech. 45, (1996).
18–23 (1996). 50. C. H. Hennekens et al., N. Engl. J. Med. 334, 1145–1149
15. A. Fernandez et al., Biotechnol. Bioeng. 58, 605–616 (1998). (1996).
16. R. K. Mandalam and B. Palsson, Biotechnol. Bioeng. 59, 51. X. D. Wang et al., J. Natl. Cancer Inst. 91, 60–66 (1999).
605–611 (1998). 52. C. S. You, R. S. Parker, and K. J. Good, Am. J. Clin. Nutr.
17. A. Contreras, F. Garcia, E. Molina, and J. C. Merchuk, 64, 177–183 (1996).
Biotechnol. Bioeng. 60, 317–325 (1998). 53. J. W. Erdman Jr. et al., J. Nutr. 128, 2009–2013 (1998).
18. B. J. Thomas et al., Biotechnol. Techniq. 3, 389–392 (1989). 54. A. B. Amotz and Y. Levy, Am. J. Clin. Nutr. 63, 729–734
19. E. J. Olgiun et al., World J. Microbiol. Biotech. 10, 576–578 (1996).
(1994). 55. K. Xue et al., Mutat. Res. 418, 73–78 (1998).
20. R. D. Canizares et al., Biotechnol. Lett. 15, 321–326 (1993). 56. M. Grung, F. M. L. D’Souza, and M. Borowitzka, Appl.
21. R. T. Lorenz and G. R. Cysewski, Trends Biotechnol. 18, Phycol. 4, 165–171 (1992).
160–167 (2000). 57. O. J. Torrissen and R. Christiansen, J. Appl. Ichthyol. 11,
22. S. Arad, O. Friedman, and A. Rotem, Appl. Environ. Micro- 225–230 (1995).
biol. 54, 2411–2414 (1998). 58. M. Kobayashi, Appl. Microbiol. Biotechnol. 54, 550–555
23. J. Fabregas et al., Appl. Microbiol. Biotechnol. 44, 287–292 (2000).
(1995). 59. M. Kobayashi et al., Appl. Microbiol. Biotechnol. 48,
24. G. Torzillo, A. Sacchi, and R. Materassi, Bioresour. Technol. 351–356 (1997).
38, 95–100 (1991). 60. H. Nishino, Mutat. Res. 402, 159–163 (1998).
25. A. B. Amotz and M. Avron, Ann. Rev. Microbiol. 37, 95–119 61. J. P. Yuan and F. Chen, J. Agric. Food Chem. 47, 31–35
(1983). (1999).
26. A. B. Amotz and M. Avron, Plant Physiol. 72, 593–597 62. A. Yaron, E. Cohen, and S. Arad, J. Agric. Food Chem. 40,
(1983). 1316–1320 (1992).
27. S. Orset, G. C. Leach, R. M. Morais, and A. Young, J. Agric. 63. F. Chen, Y. Zhang, and S. Guo, Biotechnol. Lett. 18, 603–608
Food Chem. 47, 4782–4790 (1999). (1996).
202 ALGAE BIOTECHNOLOGY

64. B. B. Vadiraja, N. W. Gaikwad, and K. M. Madyastha, 101. B. Lustigman and C. Brown, Bull. Environ. Contam. Toxicol.
Biochem. Biophys. Res. Commun. 249, 428–431 (1998). 46, 329–335 (1990).
65. S. G. Mihova, D. Georgiev, and K. M. Minkova, J. Biotech- 102. B. Lustigman, L. H. Lee, N. Thees, and J. Masucci, Bull.
nol. 48, 251–257 (1996). Environ. Contam. Toxicol. 49, 743–749 (1992).
66. S. Arad and A. Yaron, Trends Food Sci. Technol. 3, 92–96 103. S. Bloor and R. R. England, Enzyme. Microb. Technol. 13,
(1992). 76–81 (1991).
67. G. W. Niven, S. J. Smith, and A. T. Andres, Biotechnol. 104. V. Majidi, D. A. Laude, and J. A. Holcombe, Environ. Sci.
Tech. 4, 373–378 (1990). Technol. 24, 1300–1312 (1990).
68. L. Eisele et al., Biochim. Biophys. Acta 1,456, 99–107 105. B. Lustigman, Bull. Environ. Contam. Toxicol. 43, 342–349
(2000). (1988).
69. V. T. Ol, A. N. Glazer, and L. Stryer, J. Cell Biol. 93, 106. T. Chang et al., Bioresour. Technol. 44, 149–153 (1993).
981–986 (1982). 107. P. R. Burkholder and G. M. Sharma, Lloydia, 32, 466–483
70. A. N. Glazer and L. Stryer, Trends Biochem. Sci. 9, 423–427 (1969).
(1984). 108. I. Anderson, New Scientist 145, 5 (1995).
71. I. Schneider, Genetic Eng. News 18, 26,32,47 (1984). 109. I. Mahasneh, M. Jamal, M. Kashashneh, and M. Zibdeh,
Microbios 83, 23–26 (1995).
72. D. R. Ryan et al., J. Immunol. Methods 7, 115–128 (1984).
110. C. Hellio et al., Appl. Microbiol. Biotechnol. 54, 543–549
73. M. A. Borowitzka, Microbiol. Sci. 3, 372 (1986).
(2000).
74. M. Mansour et al., Phytochemistry 50, 541–548 (1999).
111. J. Kim et al., Can. J. Bot. 75, 1656–1660 (1997).
75. G. Klyachko-Gurvich et al., Physiol. Plant 107, 240–249
112. T. B. Ng, B. Huang, W. Fong, and H. W. Yeung, Life Sci. 61,
(1999).
933–949 (1997).
76. A. Klausner, Biotechnology 4, 947–953 (1986).
113. O. Bagasa and H. W. Lischner, J. Infect. Dis. 158,
77. A. Yaron et al., in G. Charalambous, ed., Development 1094–1087 (1988).
of Food Science, Elsevier Science, Amsterdam, 1995, 114. H. Sudo et al., Curr. Microbiol. 30, 219–222 (1995).
pp. 665–674.
115. S. Ayehunie, A. Belay, T. W. Baba, and R. M. Ruprecht, J.
78. Z. D. Jiang, S. O. Ketchum, and W. H. Gerwick, Phytochem- Acquir. Immun. Defic. Syndr. Hum. Retrovirol. 18, 7–12
istry 53, 129–133 (2000). (1998).
79. R. H. Al-Hasan, F. M. Hantash, and S. S. Radawan, Appl. 116. K. Hayashi, T. Hayashi, and I. Kojima, AIDS Res. Hum.
Microbiol. Biotechnol. 35, 530–535 (1991). Retroviruses 12, 1463–1471 (1996).
80. J. Hodgson, Nat. Biotechnol. 14, 700 (1996). 117. M. Hasui, M. Matsuda, K. Okutani, and S. Shigeta, Int. J.
81. M. W. Bernart and W. H. Gerwick, Phytochemistry 36, Biol. Macromol. 17, 293–297 (1995).
1233–1240 (1994). 118. H. Nakasima et al., J. Cancer Res. Clin. Oncol. 113,413–416
82. T. J. Ahern, S. Katoh, and E. Sada, Biotechnol. Bioeng. 25, (1987).
1057–1070 (1983). 119. G. Lynch, L. Low, and S. Li, J. Leukocyte Biol. 56, 266–277
83. E. Sada, S. Katoh, A. Kheirolomoon, and H. Yokoi, J. (1994).
Ferment. Bioeng. 67, 135–137 (1989). 120. M. Huheihel, V. I. Shanu, J. Tal, and S. Arad, Pres. Israel
84. Y. Miura et al., FEMS Microbiol. Lett. 107, 163–168 (1993). Soc. Microbiol. 15, 103 (1995).
85. G. Mahajan and M. Kamat, Appl. Microbiol. Biotechnol. 43, 121. N. Kita et al., J. Appl. Phycol. 2, 155–162 (1990).
466–469 (1995). 122. A. Clare, New Scientist 145, 38–41 (1995).
86. R. J. Shields et al., J. Nutr. 129, 1186–1194 (1999). 123. A. Henrikson and J. R. Pawlik, J. Exp. Mar. Biol. Ecol. 194,
87. E. Von Elert and P. Stampfl, Freshwater Biol. 45, 189–200 157–165 (1995).
(2000). 124. K. J. Becker and K. W. G. Rotmann, J. Appl. Phycol. 2,
88. C. C. Parrish, J. S. Wells, Z. Yang, and P. Dabinett, Mar. 105–110 (1990).
Biol. (Berlin) 133, 461–471 (1999). 125. R. A. Kay, Crit. Rev. Food Sci. Nutr. 30, 555–573 (1991).
89. B. Klein et al., Mar. Biol. (Berlin) 135, 191–198 (1999). 126. B. Lipstein and S. Hurwitz, Brit. Poult. Sci. 21, 9–21 (1980).
90. P. J. Caceres et al., Phytochemistrys 53, 81–86 (2000). 127. E. Ross and W. Dominy, Poult. Sci. 69, 794–800 (1990).
91. M. Lahaye, J. L. G. Phinchetti, M. J. de Rio, and G. Garcia 128. R. O. Canizares-Vilanueva, A. R. Dominguez, M. S. Cruz,
Reina, J. Sci. Food Agric. 68, 99–104 (1995). and E. Rios-Leal, Bioresour. Technol. 51, 111–116 (1995).
92. J. K. Baird, P. A. Sandford, and I. W. Cottrell, Biotechnology 129. M. A. Olvera-Novoa, L. J. Dominguez cen, and L. Olivera-
1, 778–783 (1983). Castillo, Aquacul. Res. 29, 709–715 (1998).
93. R. H. Pratt, Am. J. Bot. 29, 142–148 (1942). 130. L. V. Venkataraman, T. Somasekaran, and E. W. Becker,
Brit. Poult. Sci. 35, 373–381 (1994).
94. Y. Miura, K. Sode, V. Narasaki, and T. Matsunaga, J. Mar.
131. A. Anaga and G. Abu, Bioresour. Technol. 58, 93–95 (1996).
Biotechnol. 1, 143–146 (1993).
132. D. Dean-Ross and A. L. Mills, Appl. Environ. Microbiol. 55,
95. S. Ohta et al., Bull. Environ. Contam. Toxicol. 50, 171–178
2002–2009 (1989).
(1993).
133. W. Zhang and V. Majidi, Environ. Sci. Technol. 28,
96. J. McN. Sieburth, Dev. Ind. Microbiol. 5, 124–134 (1964).
1577–1581 (1994).
97. W. Fenical, Science 215, 923–928 (1982).
134. J. M. Collard and R. Matagne, Appl. Environ. Microbiol. 56,
98. G. M. Konig et al., Planta Med. 60, 532–538 (1997). 2051–2055 (1990).
99. I. S. Hornsey and D. Hide, Brit. Phycol. J. 9, 353–361 (1974). 135. K. Nagel and J. Voight, Appl. Environ. Microbiol. 55,
100. I. S. Hornsey and D. Hide, Brit. Phycol. J. 11, 63–67 (1976). 526–528 (1989).
 ALGAE BIOTECHNOLOGY 203

136. S. Devars, R. Hernandez, and R. Moreno-Sanchez, Arch. 167. J. Vymazal, Acta Hydrochim. Hydrobiol. 18, 513–535
Environ. Contam. Toxicol. 34, 128–135 (1998). (1990).
137. M. R. Twiss, P. M. Welbourn, E. Schwartzel, and J. Cana- 168. Z. Aksu and T. Kutsal, J. Chem. Tech. Biotechnol. 52,
dian, J. Bot. 71, 333–338 (1992). 109–118 (1991).
138. I. Visviki and J. Rachlin, Arch. Environ. Contam. Toxicol. 169. Z. R. Holan, B. Volesky, and I. Prasetyo, Biotechnol. Bioeng.
20, 271–275 (1998). 41, 819–825 (1993).
139. L. H. Lee, B. Lustigman, I. Chu, and H. Jou, Bull. Environ. 170. C. Williams and R. G. J. Edyvean, Biotechnol. Prog. 13,
Contam. Toxicol. 46, 720–726 (1991). 424–428 (1997).
140. L. H. Lee, B. Lustigman, I. Chu, and S. Hsu, Bull. Environ. 171. C. L. Brierley, in H. L. Ehrlich and C. L. Brierley, eds.,
Contam. Toxicol. 48, 230–236 (1992). Microbial Mineral Recovery, McGraw Hill, New York, 1990,
141. L. H. Lee et al., Environ. Contam. Toxicol. 49, 272–278 pp. 302–324.
(1992). 172. M. D. Mueller, D. Wolf, T. J. Beveridge, and G. W. Bailey,
142. L. H. Lee, B. Lustigman, and J. Maccari, Bull. Environ. Soil Biol. Biochem. 24, 129–135 (1992).
Contam. Toxicol. 50, 600–607 (1993). 173. R. Suetfeld, Phytochemistry 49, 451–459 (1998).
143. L. H. Lee, B. Lustigman, and D. Dandorf, Bull. Environ. 174. E. Miadokova et al., Environ. Mol. Mutagen. 31, 383–389
Contam. Toxicol. 53, 158–165 (1994). (1998).
144. B. Lustigman, L. H. Lee, and C. Weiss-Magasic, Bull. Envi- 175. J. Frenz et al., Biotechnol. Bioeng. 34, 754–762 (1989).
ron. Contam. Toxicol. 55, 65–72 (1995). 176. J. Frenz, C. Largeau, and E. Casadevall, Enzyme Microb.
145. B. Lustigman, L. H. Lee, and A. Khhalil, Bull. Environ. Technol. 11, 717–724 (1989).
Contam. Toxicol. 55, 73–80 (1995). 177. C. J. Soeder, in A. Richmond, ed., Handbook of Microalgal
146. L. H. Lee and B. Lustigman, Bull. Environ. Contam. Toxicol. Mass Culture, CRC Press, Boca Raton, Fla., 1986, pp. 25–41.
56, 985–992 (1996). 178. W. J. Oswald, in M. A. Borowitzka and L. J. Borowitzka,
147. L. H. Lee, B. Lustigman, S. Murray, and S. Koepp, Bull. eds., Microalgal Biotechnology, Cambridge University Press,
Environ. Contam. Toxicol. 62, 591–599 (1999). New York, 1987, pp. 305–328 and pp. 357–394.
148. B. Lustigman, L. H. Lee, J. Morata, and F. Khan, Bull. 179. S. Hashimoto and K. Furukawa, J. Ferment. Bioeng. 67,
Environ. Contam. Toxicol. 64, 565–573 (2000). 62–69 (1989).
149. D. Malutova, Water Manage. B29, 148–154 (1979). 180. Z. Aksu and T. Kutsal, Process Biochem. 33, 7–13 (1998).
150. J. J. Rosko and J. W. Rachlin, Bull. Torrey Bot. Club 104, 181. H. Seki, A. Suzuk, Y. Iburi, and J. Colloid, Interface Sci. 229,
226–233 (1977). 196–198 (2000).
151. J. W. Raichlin, T. E. Jensen, and B. Warkentine, in D. D. 182. D. Y. Cho, S. T. Lee, S. Park, and A. S. Chung, J. Environ.
Hemphill, ed., Trace Substances in Environmental Health- Sci. Health A29, 389–409 (1994).
XVI, University of Missouri, Columbia, 1982, pp. 145–154. 183. Z. R. Holan and B. Volesky, Biotechnol. Bioeng. 43,
152. J. W. Raichlin, T. E. Jensen, and B. Warkentine, Bull. 1001–1009 (1994).
Torrey Bot. Club 110, 217–223 (1983). 184. D. Darnall et al., Environ. Sci. Technol. 20, 206–228 (1986).
153. J. T. Hollibaugh, D. L. R. Seibert, and W. H. Thomas, Sci- 185. D. Darnall, Technol. Sr. Biol. Proc. 3, 65–72 (1991).
ence 10, 93–105 (1980). 186. N. Kuyucak and B. Volesky, Biotechnol. Bioeng. 33, 823–831
154. H. J. Krock and D. T. Mason Berkley, Sanit. Eng. Res. Lab. (1989).
Univ. Calif. Rept. No. 71–8 6, 1–32 (1971). 187. P. Harris and G. Ramelow, Environ. Sci. Technol. 24,
155. J. Bittell, R. J. Miller, and D. E. Koeppe, Physiol. Plant 30, 220–228 (1990).
226–230 (1974). 188. R. Crist et al., Environ. Sci. Technol. 22, 755–760 (1988).
156. C. Weiss-Magasic, B. Lustigman, and L. H. Lee, Bull. Envi- 189. H. H. Liu, W. L. Chen, and Y. Wu, Sci. Total Environ. 91,
ron. Contam. Toxicol. 59, 823–833 (1997). 275–282 (1990).
157. J. M. Wood, Science 183, 1049–1052 (1974). 190. H. Mann, W. S. Fyfe, R. Kerrich, and M. Wiseman, Biore-
158. S. E. Fayed and H. I. Abd-El-Shafy, Environ Int. 11, 77–87 covery 1, 155–172 (1989).
(1985). 191. S. C. Wilkinson, K. H. Goulding, and P. K. Robinson, Bio-
159. Y. Nassiri, J. L. Mansot, and J. Wery, Arch. Environ. technol. Lett. 11, 861–864 (1989).
Contam. Toxicol. 33, 147–155 (1997). 192. N. Friis and P. Myers-Keith, Biotechnol. Bioeng. 28, 21–28
160. L. C. Rai and M. Raizada, Ecotoxicol. Environ. Saf. 15, (1986).
195–205 (1998). 193. Z. Aksu, G. Egretli, and T. Kutsal, Process Biochem. 33,
161. C. Rai, T. Gaur, and J. Kumar, Environ. Res. 25, 250–259 393–400 (1998).
(1981). 194. G. M. Gadd, Experentia 46, 38–48 (1990).
162. J. Trevors, G. Stratton, and G. Gadd, Can. J. Microbiol. 32, 195. B. Wehrheim and M. Wettern, Appl. Microbiol. Biotechnol.
447–464 (1986). 41, 725–728 (1994).
163. A. Abbas and C. Edwards, Appl. Environ. Microbiol. 55, 196. R. H. Crist, K. Oberholser, N. Shank, and M. Nguyen, Env-
2030–2035 (1989). iron. Sci. Technol. 15, 1212–1217 (1980).
164. Y. Tsai and B. H. Olson, Appl. Environ. Microbiol. 56, 197. B. Volesky and Z. R. Holan, Biotechnol. Prog. 11, 235–250
3266–3272 (1990). (1995).
165. D. H. Hamer, D. J. Theile, and J. E. Lemont, Science 228, 198. G. Venkateswerlu and G. Stotzky, Appl. Microbiol. Biotech-
685–690 (1985). nol. 31, 619–625 (1989).
166. Y. P. Ting, F. Lawson, and I. G. Prince, Biotechnol. Bioeng. 199. S. G. Walker et al., Appl. Environ. Microbiol. 55, 2976–
34, 990–999 (1989). 2984 (1989).
204 ALGAL BLOOMS

200. A. A. Hamdy, Curr. Micrbiol. 41, 232–238 (2000). ALGAL BLOOMS. See RED TIDES AND OTHER HARMFUL
201. S. Schiewer and M. H. Wong, Chemosphere 41, 271–282 ALGAL BLOOMS
(2000).
202. R. H. Vieira and B. Volesky, Int. Microbiol. 1, 17–24 (2000).
203. T. Matsunaga, H. Takeyama, T. Nakao, and A. Yamazaw, J.
Biotechnol. 70, 33–38 (1999).
204. I. Visviki and J. W. Rachlin, Arch. Environ. Contam. Toxicol. ALGAL BLOOMS — IMPACT ON TREATMENT,
20, 271–275 (1991). TASTE, AND ODOR PROBLEMS
205. Y. Sag, U. Aickel, A. Zaksu, and T. Kutsal, Process Biochem.
33, 273–284 (1998). GARY A. BURLINGAME
Philadelphia Water Department
206. N. Kuyucak and B. Volesky, Biorecovery 1, 189–204 (1989).
Philadelphia, Pennsylvania
207. R. P. de Carvalho, K. H. Chong, and B. Volesky, Biotechnol.
Prog. 11, 39–44 (1995). Algae are not regulated in drinking water. However, algae
can cause water quality to exceed secondary maximum
contaminant levels (SMCLs) such as color, taste, and odor.
Algae can increase the pH above the SMCL limit of 8.5
ALGAE, BLUE-GREEN. See CYANOBACTERIA and they can increase the threshold odor number above
the limit of three. SMCLs are nonenforceable guidelines
under the Safe Drinking Water Act as set by the U.S.
EPA; however, in some states they have been adopted as
enforceable standards. In most cases, algae are a nuisance.
They affect the water’s aesthetics or the water’s usability.
ALGAE: EUTROPHICATION. See EUTROPHICATION Algae are also known for their interference with the
AND ALGAE treatment of water. More recently, concern has been raised
over the production of toxins by cyanobacteria (blue-green
algae).
Although algae themselves do not typically pass into
drinking water, their by-products often do. These by-
ALGAE, POLAR. See POLAR MARINE PHYTOPLANKTON products can be (1) chemicals released upon death or cell
lysis, (2) chemicals formed by treatment, that is, chemical
reactions with algae materials, or (3) chemicals released
by the active algae before treatment.

WHAT ARE ALGAE?

ALGAE: STREAMS. See STREAM MICROBIOLOGY
The terminology is alga for the singular form and algae
for the plural form of the word. The study of algae is
called algology or phycology. Algae are related to plants
because they contain pigments that cause photosynthesis
to take place. They are eukaryotic organisms (their cells
ALGAE: TOXICITY TESTING. See USE OF have a true nucleus as do plant and animal cells). But
MICROSCOPIC ALGAE IN TOXICITY TESTING
algae have no roots or leaves like plants. (There is an
exception with the blue-green algae or cyanobacteria, and
they will be mentioned in more detail later.) The cell wall
of most algae is made of cellulose. The diatoms have an
outer protective shell made of silica. When dead diatoms
ALGAE, USE AS BIOLOGICAL INDICATORS IN accumulate they form deposits called diatomaceous earth.
PALEOLIMNOLOGY. See PALEOLIMNOLOGY: SUBFOSSIL Algae are typically photoautotrophic: they only require
ALGAE OTHER THAN DIATOMS AND CHRYSOPHYTES; light, water, and carbon dioxide to produce their own food
PALEOLIMNOLOGY: USE OF SILICEOUS STRUCTURES OF source. Photosynthesis is the process in which light is
CHRYSOPHYTES AS BIOLOGICAL INDICATORS IN FRESHWATER absorbed by chlorophylls (green photosynthetic pigments)
SYSTEMS and converted to chemical energy (sugars). Chlorophyll a
and b and c are common pigments, giving algae their green
color. Other carotenoids are also present, such as those,
which give rise to the color of the golden-brown algae.
Algae in water can be found in a variety of places:
benthic algae are found in sediments, planktonic are free-
ALGAE: WASTE STABILIZATION PONDS. floating, and periphyton are attached to aquatic plants,
See WASTEWATER STABILIZATION PONDS rocks, submerged debris, and walls. They can be found in
 ALGAL BLOOMS — IMPACT ON TREATMENT, TASTE, AND ODOR PROBLEMS 205

various waters, from salt water to freshwater and from

swamps to hot springs. Planktonic algae are most often
associated with taste and odor and filter-clogging problems
at water treatment plants.
Algae also vary in physical structure, which makes
them relatively easy to identify and differentiate. Algae
vary in size from microscopic forms to aquatic seaweeds
or kelp:

• Picoplankton are of the size 0.2 to 2.0 µm.

• Nanoplankton are of size 2 to 20 µm.
• Microplankton are of the size 20 to 200 µm and are the
types involved directly in water-treatment problems.

Finally, algae differ in the ways their cells associate

with each other. For example, they can be unicellular Figure 3. An example of a colonial diatom (Asterionella sp.). See
(Fig. 1), filamentous (Fig. 2), or colonial (Fig. 3). color insert.
In clear or slightly polluted water, diatoms are often
found as planktonic algae. They are an important source
of food for fish and wildlife. Increasing the nutrient algae to filamentous green algae, colonial green algae,
pollution load (nitrogen and phosphorus) to the water unicellular green algae, and blue-green algae. There are
causes a shift in the predominant forms of planktonic many variations and many other secondary factors also
affect algal growth and predominance. Blue-green algae
are often indicators of organic pollution or excessive
nutrient loading. They can form blooms that, upon death
of the cells, cause a depletion in dissolved oxygen, which
in turn can result in fish mortality.

COMMON ALGAE THAT AFFECT DRINKING WATER

Waters in which algae are present are either surface

waters (such as rivers, lakes, and reservoirs) or ground-
waters under the influence of surface waters. These waters
are required, in most cases, to be filtered to remove particu-
late matter such as algae. If treatment is insufficient, algae
can pass into finished drinking water and have been found
in water distribution systems. In general, water never
contains a single alga specie, but always consists of some
Figure 1. An example of a unicellular green alga (Closterium combination of algae species. Sometimes the alga of great-
sp.). See color insert. est dominance in water is not the alga causing the problem.
Palmer’s Guide to Algae (1) provided an empirical classifi-
cation of algae according to their impacts on water supplies
and treatment. Table 1 provides an overview of Palmer’s
guide. The following divisions of algae are of common
concern for drinking water and its treatment (2).

Green Algae
Chlorophyta or the green algae are the largest and most
diverse division of algae, which range from microscopic
single-celled algae to colonial and filamentous algae
(examples include the genera Chlorella, Scenesdesmus,
Spirogyra, Cladophora, and Ankistrodesmus).

Dinoflagellates
Pyrrhophyta or the dinoflagellates are important com-
Figure 2. An example of a filamentous blue-green alga (Oscilla- ponents of plankton. They have flagella and, therefore,
toria sp.) surrounding a single strand of a green alga (Spirogyra are motile and are typically single-celled in occurrence
sp.). See color insert. (examples include the genera Peridinium and Ceratium).
206 ALGAL BLOOMS — IMPACT ON TREATMENT, TASTE, AND ODOR PROBLEMS

Table 1. Algae Associated with Certain Source Water and Treatment Conditions
Based on Palmer’s Guide (1)

Taste & Odor Clogging of Filters in Nutrient Pollution

Association: Problems in Tap Water Water Treatment of Natural Waters

Most often Blue-green algae Diatoms Blue-green algae

Common Pigmented flagellates Blue-green algae Pigmented flagellates
Infrequent Diatoms Green algae Green algae
Least often Green algae Diatoms

Yellow-Green or Golden-Brown Algae

Chrysophyta or golden-brown algae include two classes of
algae, which are important in drinking water treatment.
The Chrysophyceae are pigmented flagellates (examples
include the genera Volvox, Pandorina, Synura, Peri-
dinium, Chlamydomonas). The Bacillariophyceae are the
diatoms (examples include the genera Cyclotella, Aster-
ionella, Navicula, Melosira Diatoma, Fragillaria, and
Synedra).

Blue-Green Algae
Cyanophyta or blue-green algae are also known as
cyanobacteria (examples include the genera Oscillatoria,
Microcystis, Anabaena, Aphanizomenon). These algae are
more closely related to bacteria than algae. They can
be single-celled, colonial, or filamentous in occurrence,
and can appear as floating scum or benthic mats, in
eutrophic waters or stratified reservoirs. Blue-green algae
are not plantlike in their cellular structure as are the
other algae. They exhibit a gliding motility, have resting
spores or endospores, have gas vacuoles to regulate their
buoyancy, and can have a sheath surrounding the cells.
They are diverse in their metabolism and are oxygenic
photosynthetic bacteria. Some fix nitrogen and develop
heterocysts for nitrogen storage. Some are found in
symbiotic relationships in lichens.

Figure 4. Example of visible abundance of a blue-green alga as

WHAT IS A BLOOM AND WHERE DOES IT OCCUR? a surface scum on water. See color insert.

A bloom of an alga is commonly considered a visible

abundance of plankton at the surface of water (Fig. 4). since carbon is derived by algae through photosynthesis
Algal blooms can also appear as carpetlike mats of algae from dissolved carbon dioxide. Phosphorus is most often
attached to the bottom of a water body or to submerged the limiting nutrient. Sediments can contain insoluble
matter (Fig. 5). Perhaps more accurately, in the context of phosphorus bound to iron, calcium, or aluminum. Under
water treatment, a bloom is a significant increase in the anaerobic or elevated pH conditions the bound phosphor-
density of an alga to the point where the use of the water is us can become soluble and fuel an algal bloom. When
negatively affected. For some algae a bloom could translate phosphorus becomes the limiting nutrient, any addition of
into a density of 100,000 cells per milliliter, whereas for phosphorus can lead to green algae blooms. When nitrogen
others it could mean only 10,000 cells per milliliter of is limiting, this can select for blue-green algae, which can
water or less. fix nitrogen from other sources. When a bloom occurs, an
An accepted cause of an algal bloom is the excess avail- alga can deplete the nutrient that fueled the bloom and
ability of nutrients such as phosphorus and nitrogen (3). can, in turn, limit the bloom.
Nutrients are considered very important in determining Although algal blooms occur most often during the
the types of algae that will grow. Nutrients can come from warmer months of the year, clear lakes, which are enriched
sources external to the body of water (farm runoff and with nitrogen, during the late winter to early spring, can
wastewater discharges) or within the body of water (recy- experience blooms of pigmented flagellates, diatoms, and
cling from the sediment, aquatic vegetation). The three green algae.
primary nutrients are carbon, nitrogen, and phosphorus. Depending on the characteristics of the lake or
In most cases, there is no limit to the availability of carbon reservoir, the distribution both vertically and horizontally
 ALGAL BLOOMS — IMPACT ON TREATMENT, TASTE, AND ODOR PROBLEMS 207

as by chemical treatment or stress, they release intra-

cellular organic matter, which can affect water-treatment
processes.
Perhaps the most dramatic impact caused by algal
cellular products are caused by toxins (4,5). Freshwater
blue-green algae or cyanobacteria can produce two
important human and animal toxins. Hepatotoxins inhibit
enzyme activity and can cause acute liver damage. Animal
poisoning has resulted in animal mortality in one hour
because of liver failure. They are produced by species of
Microcystis (M. aeruginosa and M. flos-aquae), Anabaena,
Nostoc, Oscillatoria. Microcystis aeruginosa is a commonly
implicated cyanobacterium. Neurotoxins are nerve poisons
and can cause death by respiratory distress. They are
produced by such algae as Anabaena flos-aquae and
Aphanizomenon flos-aquae. Microcystis aeruginosa can
also produce a neurotoxin.
Toxins can have effects in two ways. Direct effects
are where algae release their toxins into water, which is
ingested, or the toxin is released after ingestion of the
algal cells. Mediated effects are where the algae or toxin
accumulate in a vector organism, such as shellfish, which is
unaffected by the poison but causes illness when the vector
is consumed. Human contact, rather than ingestion, can
cause skin and eye irritation, whereas the ingestion of very
low levels of toxin can cause gastrointestinal illness such
as vomitting, cramps, and fever. (See also CYANOBACTERIA-
TOXINS IN DRINKING WATER.)
Figure 5. Example of how filamentous blue-green algae appear
attached to bottom sediment. See color insert. OVERVIEW OF TREATMENT PROCESSES AFFECTED BY
ALGAE

of nutrients, temperature, pH, light, and oxygen can result The following discussion explains the treatment stages of a
in different types of algae in different concentrations conventional drinking water treatment system (6) (Fig. 6):
at different depths. As a result, the depth at which
a treatment plant withdraws water from a reservoir Presedimentation Basin
can affect the types of algae that must be treated. It
Some treatment works contain a presedimentation basin
is important that a treatment plant manager develop
where treatment is achieved by gravity settling in a large
an understanding of the source water. The ability to
basin at the head of the treatment plant. The primary
predict algal blooms to provide an early warning to
purpose is for the removal of silt and sand and debris,
water treatment operators depends on sanitary surveys
which can damage pumps and processes. This basin
of the watershed and measurements of nutrients such as
also provides for additional emergency storage. Generally,
total phosphate, nitrate, ammonia, total organic carbon,
presedimentation basins are ineffective at removing algal
dissolved silica, dissolved oxygen, pH, water temperature,
cells and may serve to increase algal production because
available light, and water turbidity.
of their long detention times.

OVERVIEW OF IMPACTS FROM ALGAL BLOOMS Preoxidation

In the presedimentation basin or in the conduits that
There are two categories of impacts that algae can have deliver water to the plant, an oxidant such as potassium
on water treatment (2): direct impacts caused by presence permanganate, chlorine dioxide, or chlorine might be used
of algae cells, and indirect impacts caused by cellular to precondition the water or satisfy the oxidant demand.
by-products. The direct impacts of algal blooms include The demand can be caused by organic matter (algae,
filter clogging, surface scums, and change in color of the decaying plant matter, suspended natural organic matter)
water. Indirect impacts include the production of odors, and inorganic matter (ammonia, iron, manganese). The
consumption of oxygen in the water when algae decompose, achieving of this oxidant demand helps the plant operators
and change in the pH of the water. achieve disinfection and produce stable water before
Algae convert inorganic matter (carbon dioxide, nitro- it leaves the treatment plant. Aeration has at times
gen, phosphorus) to organic matter through photosyn- been used to oxidize dissolved metals such as iron and
thesis. This organic matter makes up the algal cells. It manganese. Preoxidation can be effective at inactivating
can contribute to extracellular organic matter, which is algal cells and at improving their removal in subsequent
excreted outside the cells. When algae cells are lysed, treatment processes.
208 ALGAL BLOOMS — IMPACT ON TREATMENT, TASTE, AND ODOR PROBLEMS

Water source

Intake and
pre-sedimentation

Oxidant
addition

Flocculation Coagulation

Disinfection

Filtration

Sedimentation

Post-treatment

Figure 6. Stages of a conventional water Consumers

treatment plant.

Coagulation/Flocculation/Sedimentation are removed from the water. This sequence of coagulant

addition, rapid mixing, flocculation, and sedimentation
Coagulation is the addition into water of coagulant
has been commonly used in the treatment of surface
chemicals to neutralize the surface charges of the
waters and is effective at removing a large percentage
remaining suspended matter. Aluminum salts (alum), and
of algal cells.
ferric salts such as ferric chloride and ferric sulfate, when
added to water form hydroxide complexes in water. These
positively charged complexes neutralize the negative
Filtration
surface charges of particles in the water, allowing them
to stick together. This is followed by flocculation, which Filtration by sand media, or a combined bed of sand
is the gentle mixing of water to cause the floc particles and coal (dual media) has been commonly applied for the
to collide and form larger particles that will settle out removal of suspended microscopic particulates and floc,
by gravity (sedimentation). Coagulation and flocculation which did not settle. Typical turbidity levels in well-run
are often followed by sedimentation and filtration. The filter effluents can range from 0.1 to 0.3 nephelometric
combination of these processes constitutes a treatment turbidity units or less. Turbidity is a measure of the
train by which dissolved inorganic and organic matter, light-scattering properties of water and, thus, is a relative
biological particulates, and suspended inorganic matter measure of the suspended particulate matter in water. At
 ALGAL BLOOMS — IMPACT ON TREATMENT, TASTE, AND ODOR PROBLEMS 209

low turbidity levels, such as these, very few algal cells will Performance of Settling and Clarification
have penetrated the filters.
Although algae can be effectively incorporated into the floc,
the presence of the algae and their by-products such as
Disinfection extracellular organic matter and cellular oils may impair
Disinfection using chlorine, chloramines, or ozone is the settling of the floc. One effect is the carryover of
common. Disinfection renders the water free of micro- algal cells and floc to the filters. The size and shape of
bial pathogens. Once the water has been adequately the algae are two important factors in determining the
pretreated, the addition of a disinfectant to the water extent of these effects to occur. It is generally believed
can effectively remove or inactivate most disease-causing that coagulation and flocculation should be preceded by
organisms without interference from other organic and oxidation to inactivate the algae.
inorganic substances. Disinfection should also ensure that
no viable algal cells are transmitted to the distribution Performance of Filtration
system.
Algae that pass through the treatment processes and make
their way onto the filters can cause the clogging of the
Posttreatment filters. This requires more frequent filter backwashing,
Posttreatment can include fluoride addition, pH adjust- which increases turbidity and the quantity of wasted
ment, corrosion control, and the addition of a disinfectant water. For example, rapid sand filters that are normally
residual to make the water acceptable for distribution and backwashed every forty hours into operation might have
storage in reservoirs. Posttreatment has little or no effect to be backwashed every twelve hours. Along with this
on the control of algal cells. effect on the filters comes a reduction in water quality.
Increased backwashing of the filters and the penetration
of algal cells through the filters can increase the turbidity
SUMMARY OF NEGATIVE IMPACTS FROM ALGAE ON of the finished water, but the largest effect of clogging is
TREATMENT to limit the amount of water that the treatment plant can
produce.
There are sources of information on the negative impacts of
algae on water treatment processes (1,2). The size, shape, Disinfectant Demand and By-Product Formation
motility of algae, release of extracellular organic matter
(EOM), and aggregation of algae affect the treatability of Algal cell products such as extracellular organic matter or
the algae. For example, algae that associate as dense rafts products released upon cell lysis may constitute sufficient
will more readily clog filters. Smaller, singly occurring organic matter to react with oxidants such as chlorine
algal cells can more readily break through the filters. to form disinfection by-products. The oxidation of algae
Algae that have cell structures that prevent the proper cells causes cell lysis, which releases organic matter that
formation of floc will interfere with the flocculation and itself will get oxidized and exert an oxidant demand.
settling processes. Motile algae can stay near the surface Chlorination by-products (trihalomethanes and haloacetic
of settling basins and avoid getting caught in the floc that acids) are regulated under the Safe Drinking Water Act.
is settling out. Aldehydes are one other group of by-products. Sufficient
levels of aldehydes can alter the odor quality of the finished
water. Algae and their extracellular organic matter impart
Control of Coagulant Demand
a chlorine demand, thus requiring increased use of chlorine
The amount of coagulant that must be added to water in the treatment process, which increases disinfection by-
to achieve the process goal for coagulation, flocculation, product formation in general.
and sedimentation is termed the coagulant demand.
The process goal could be charge neutralization for the Overall Control of the Treatment Process
suspended particulates or it could be a reduction in total
organic carbon or turbidity. Algae or their extracellular The treatment plant operator and manager must adjust
organic matter can induce an additional coagulant demand chemical dosages and other process controls according to
and interfere with the efficiency of the process. changes in water quality (such as during a rainstorm
that washes excessive sediment into the source water and
decreases alkalinity). Algal blooms that cause changes
Maintenance of Coagulation pH
in chlorine demand, pH, coagulant demand, and filter
The pH of the water is important for determining the operations require the plant operator to adjust treatment
actual mechanism by which particles settle out and the processes. Such changes impact other processes such as
efficiency of the coagulation and flocculation processes. sludge management, storage of treatment chemicals as
Algal blooms can increase the pH of the water above 9. more chemical might have to be used, or the recycling
Optimum coagulation pH is usually from 5.0 to 8.5 units. of filter backwash water. In addition, adjustments to
Treatment plants may not be able to counteract the pH treatment to offset the direct and indirect effects of
increase. Water with an elevated pH then enters the the algae can in turn increase turbidity, disinfection by-
coagulation and flocculation processes, thereby affecting product formation, and change the chemical quality of the
the efficiency of the processes. water. All of these issues must be carefully balanced.
210 ALGAL BLOOMS — IMPACT ON TREATMENT, TASTE, AND ODOR PROBLEMS

Quality of Water Passed into Distribution and onto 2-MIB is an abbreviated name for 2-methylisoborneol.
Customers This compound has been known since the 1960s. It is
naturally produced and blue-green algae are a common
Changes in treatment or the inability of treatment to
source though not the only source. The odor of 2-MIB is
control the impacts of algal blooms can result in increased
likened to earth, damp musty soil, and peat. It can be
turbidity, lower chlorine residuals, increased disinfection
noticed as an odor at levels around 5 nanograms per liter.
by-products, and tastes and odors. These can negatively
Analytical detection methods exist for detection of such
affect compliance with regulatory standards along with
low levels in water. Typical levels in water are around 5
customer relations.
to 100 nanograms per liter.
Cyanobacteria that have been reported to produce
TASTE AND ODOR OF DRINKING WATER geosmin include Anabaena, Aphanizomenon, Lyngbya,
Oscillatoria, Phormidium, Schizothrix, Symploca, and
One goal of the treatment process is to produce a drinking Fischerella. Cyanobacteria reported to produce 2-MIB
water that has minimal taste and odor (or flavor). One include Oscillatoria, Pseudanabaena, Synechococcus, and
consequence is that customers become accustomed to the Phormidium (8).
flavor of their tap water. Customers will complain if The decay of algae can produce sulfurous odors such
they detect a change in tap water flavor irrespective of as from dimethyldi- and dimethyltrisulfides (decaying
whether that change is an improvement or not. Therefore, vegetation, septic and fishy odors). The sulfides are
the flavor of water must be consistent. In the United decomposition products of proteins and amino acids. Other
States, the Safe Drinking Water Act requires that tap decay products include isovaleric and butyric acids, which
water contain a detectable residual of chlorine to control give off rancid, sour milk, and dirty sock odors.
Grassy odors have been noticed during algal blooms
microbiological quality. Thus, the minimal flavor of tap
and have been associated with cis-3-hexen-1-ol and cis-
water is a chlorinous flavor.
3-hexenyl acetate. The grassy odors sometimes are more
There are only four basic tastes (sweet, sour, salty,
noticeable after oxidation of the algae. Algae can produce
bitter). There are innumerable odor qualities (such
linolenic acid, which has a watermelon type of odor, and
as musty, earthy, fishy, grassy, rotten eggs, swampy,
this can degrade microbially to the grassy odorants cis-3-
chlorinous) and their names are usually representative
hexenal and t-2-hexenal.
of associations or experiences (smells like fresh cut grass).
A cucumber odor has been found to be produced by alga
Feeling sensations, produced at free nerve endings in
such as Synura. The responsible compound is trans,2-
the nasal and oral regions by chemical reactions, include
cis,6-nonadienal. This chemical is naturally produced and
drying, slick, metallic, and astringent. Changes in the
is also found in cucumbers and melons. The cucumber
inorganic and organic chemistry of drinking water have
odorant can be detected as an odor down to levels under
caused such off-tastes, off-odors and off-feeling-factor
10 nanograms per liter.
sensations.
Fishy odors have been noticed during algal blooms.
Fishy odors have been related to the compounds n-hexanal,
IMPACTS ALGAE HAVE ON TASTE AND ODOR n-heptanal, decadienal, and heptadienal. Some of these
same compounds also produce fishy odor problems in fish
There has been much research into the taste and odor oils and related products.
of drinking water and the impacts that algae have A hay-woody-tobacco-like or fruity-fragrant odor can be
on that quality (7–9). Findings of this research are produced by other alga such as Microcystis. The odor comes
summarized in the following section. Odors that have from the chemical beta-cyclocitral. It is a degradation
occurred in source waters include grassy, fishy, earthy, product of carotenes (which are cellular pigments).
musty, cucumber, geranium-flowery, rancid, sulfide-like, The diatoms (Asterionella, Tabellaria, Synedra,
decaying vegetation, aquarium, rotting hay, tobacco, and Melosira) that bloom during the spring and fall, have been
septic odors. The most common odors that affect drinking associated with grassy and fishy odors. The yellow-brown
waters and cause customer complaints come from two algae or pigmented flagellates that bloom sometimes in
naturally produced, compounds known as geosmin and winter (Dinobryon, Mallomonas, and Synura) have been
2-MIB. They impart earthy or musty odors to the associated with cucumber and fishy odors depending on
water. the stage of the bloom such as during the exponential
Geosmin (‘‘earth odor’’) is an abbreviated name for stage of growth or during the death of the bloom. The
trans-1,10-dimethyl-trans,9-decalol. This compound has green algae that bloom more typically during the summer
been known since the 1960s. It is naturally produced, have also been associated with grassy and fishy odors.
and blue-green algae are a common source, although
not the only source. The odor of geosmin has been IMPACTS ON TAP WATER CONSUMERS
likened to earth, dirt, corn silk, and beets. It can be
smelled at levels as low as about 5 nanograms per liter. Consumers of tap water rely on their senses of taste
Analytical detection methods exist for detection of this and smell and sight to detect changes in water quality.
compound in water at these low levels. Typical levels Consumers can choose to install point-of-use treatment
in source waters are around 5 to 200 nanograms per systems (such as faucet filters) or purchase bottled water
liter. to avoid consumption of water they do not trust or find
 ALGAL BLOOMS — IMPACT ON TREATMENT, TASTE, AND ODOR PROBLEMS 211

acceptable. Consumers can choose to register complaints to aluminum hydroxides. The floc lays a blanket over the
with the water provider or with the local health agency sediment and prevents recycling for a period of time.
and regulatory authority. Depending on the response they Hypolimnetic aeration is the process of adding oxygen
receive from the water provider and the attention given to the hypolimnion (colder bottom layer of water) without
to the problem by the local media, their trust of the water upsetting the stratification of the reservoir. This aeration
provider can be greatly challenged. is used in deep reservoirs that develop anoxic bottom
layers. The aeration maintains the water density layers in
EVIDENCE OF THE OCCURRENCE OF TASTE AND ODOR the reservoir and helps prevent the recycling of phosphorus
PROBLEMS and the release of other nutrients, such as iron and
manganese, from the sediment into the photic areas where
It is an unacceptable, but often unavoidable practice algae thrive.
to wait until customers complain to know that an Artificial destratification is used in the late spring or
algae bloom has affected the taste and odor quality early summer when reservoirs stratify. It attempts to
of the drinking water. There do exist some means mix the layers of water to prevent the bottom layer from
by which to better predict the occurrence of algal becoming anoxic. It can be accomplished by hydraulically
blooms and better prepare for changes in treatment. For mixing the waters and by injecting air across the bottom.
example, algal photosynthesis requires pigments such as The objective is to keep the water uniform in temperature
chlorophyll. There are analytical techniques for measuring throughout the whole body of water. Mixing can also stir
the chlorophyll content of water. On-line analyzers are together the various types of algae growing at different
available to help operators monitor trends in algal levels, and thereby cause a shift to more favorable algae
growths before the water turns green in color. Other types. Other reservoir control options include removing
predictors of algal blooms include water temperature, sediment; harvesting aquatic vegetation; draining and
light transmittance, light availability, orthophosphate, cleaning the reservoir.
ammonium nitrogen, nitrate nitrogen, dissolved silica, and Algicides are chemical agents, which are toxic to algae.
pH. Finally, the odorous chemicals produced by algae can Copper sulfate has been the most common reservoir control
be monitored for increasing trends over time. option since about 1905. It is less effective for lakes, which
The actual source of a taste-and-odor problem may not have a high pH or high alkalinity or water temperatures
be easy to find. The algal growth that is releasing an below 15 ° C. Algae vary in sensitivity and some can be
odorous compound could be in one of many coves of a quite resistant to copper sulfate such that its use can
reservoir, only in a certain bottom area of a body of water, even encourage the growth of more problematic algae.
only at a certain depth of a reservoir, or many miles up Copper sulfate-type algicides can be fed to a reservoir as
river from a treatment plant’s intake. small crystals such as in a burlap bag filled with crystals
and pulled behind a boat to dissolve the chemical into
the water. Helicopters and boats can also apply larger
OVERVIEW OF CONTROL OPTIONS crystals, which dissolve slowly and sink to reach the algae
at lower depths or on the bottom of the reservoir.
There are sufficient sources of information from treatment Potassium permanganate has been used to control algae
chemical manufacturers to treatment handbooks, which in the water source as well as algae coming into the
cover the options for control of algae and their associated treatment plant. However, it seems to work best as a pre-
problems (3,5,6–9). The following is a summary of these oxidant in the treatment plant. Potassium permanganate,
options. in all cases, must be applied at a sufficient dosage and
given sufficient time of contact.
Source Controls Another control option is watershed management. This
includes land use management, control of runoff and land
Controls include chemical oxidation or inactivation, development, use of agricultural practices that capture
nutrient reduction, alteration of the ecological balance (in runoff, waterway preservation, and control of the use of
smaller reservoirs), and aeration or mixing. The average chemicals in the watershed. Included is the reduction of
cost per year for algae control by treatment works ranged nutrients by upgrading wastewater treatment plants to
from 1,000 to 25,000 dollars during the 1980s (9). Controls achieve better nitrogen and phosphorus removals.
are often initiated according to time of year, start of a A final option is avoidance of the source of water
taste or odor problem, occurrence of certain algae, and the containing the nuisance algae. This includes providing
temperature of the water. multiple intakes in a reservoir, which allow the draw off
If phosphorus is the limiting nutrient for a reservoir of water from different depths, where algal blooms are not
or slow moving river, it is possible that when the found. This also includes changing source waters.
sediment becomes anoxic the bound phosphorus that
had precipitated out redissolves into the water. With
Inplant Treatment Controls
mixing or reservoir turnover this bottom concentration
of dissolved phosphorus then becomes available to the Treatment can either physically remove or kill the algal
algae. Attempts have been made to precipitate out the cells. It is preferable to capture and settle out the algae,
phosphorus with aluminum sulfate, aluminum silicate, avoiding cell lysis and the release of intracellular toxins
and ferric chloride because phosphorus is tightly bound or organic matter, which could increase taste and odor,
212 ALGAL BLOOMS — IMPACT ON TREATMENT, TASTE, AND ODOR PROBLEMS

chlorine demand, coagulant demand, or disinfection by- Adsorption of By-Products

products. Chemical oxidation can kill the algal cells, or at
Although treatment can effectively remove algae from the
lower dosages, immobilize motile algae so that they can be
water, algal intracellular and extracellular chemicals can
more readily trapped in floc. Oxidation also changes the
still pass through into the treated water. Carbon has been
surface of the cells, which can enhance or impede charge
used to adsorb algal by-products from the water. Carbon
neutralization. This may aid or interfere with the removal
can be added to water in an activated powdered form (as a
of algal cells during flocculation.
slurry that settles or filters out) or a filter can be made of
Conventional particle removal treatment (coagulation,
granular activated carbon to adsorb the organic materials
sedimentation, rapid sand filtration) is ineffective for
from the water. Since many algal blooms are infrequent
soluble algal toxins. Adsorption onto activated carbon
or episodic in nature, powdered activated carbon has been
or oxidation can be effective in controlling the toxins.
the most commonly used form. Geosmin and 2-MIB have
Lysed cells can release high levels of toxins, which may
been removed up to 40 to 90 percentage of their initial
require the use of large chemical dosages. Therefore, it
levels. However, different types of carbon are more or less
is preferable for treatment to remove the cells before cell
effective on different odorous compounds depending on the
lysis. chemical quality of the water being treated.
Jar testing is often used by treatment plant operators
to determine the optimum dosages of treatment chemicals.
The amount of coagulant added, the mixing energy, the CONCLUSION
settling time, the water temperature and the pH of the
water all affect the efficiency of settling of the floc. Algae in source waters can affect drinking water quality.
Treatment plant operators can measure charge neu- Typically, they affect the aesthetic quality of water (color,
tralization with a streaming current detector, which gives taste, odor) but some algae can impart toxins that have
a relative measure of the surface charge on particles in the human health impacts.
water. One would, in the lab, add increasing amounts Algae in source waters can affect drinking water
of coagulant until charge neutralization is achieved. treatment processes in a treatment plant. They can
Cationic, anionic, and nonionic polymers (flocculant aids) exert an oxidant demand, interfere with the removal of
have been used to improve floc formation and settleability particulate matter, and clog filters.
to get charge neutralization. Microsand enhanced floccu- The treatment of drinking water can be optimized to
lation and dissolved air flotation are other options for reduce the effects of algae on the treatment processes and
improving the physical removal of algal cells. This is the finished water quality. However, at times, source water
important in preventing the loading of the filters with treatments, watershed controls, selection of alternative
algae. The algal cells must be inactivated or immobilized water supplies, and advanced treatments have to be used
and physically captured in floc and removal by sedimenta- because the problems caused by algae cannot be mitigated
tion. The goal is to accomplish 90 to 99 percentage removal using existing treatment.
of algae before filtration without producing other negative Much is known about algae, their impacts on drinking
side effects. water quality and its treatment, and their control. Algal
The predominant algae species in the surface water impacts can be documented and, at times, even predicted
supply might not be the same species that cause filter by water quality testing and algae monitoring. However,
clogging or that break through filtration into the drinking the solutions to problems caused by algae remain site-
water. Algae vary in their propensity to resist different specific. Each water quality manager and treatment plant
stages of treatment and even low counts of certain species engineer must effectively apply existing experience and
can cause treatment or water quality problems. knowledge to secure a high quality drinking water for
Chemical oxidation attempts to accomplish a variety their consumers.
of tasks for algae control: inactivate the algae; oxidize
taste and odor compounds; and prevent the biological
formation of other odor compounds. Chlorine has been BIBLIOGRAPHY
commonly used at dosages from 1 to 5 milligrams per liter
and is effective for most algae and many algae odorants 1. C. M. Palmer, Algae in Water Supplies, U.S. Public Health
Service Publication No. 657, Washington, D.C., 1959.
but causes formation of disinfection by-products. Chlorine
dioxide has been used at dosages from 1 to 3 milligrams 2. American Water Works Association, Problem Organisms in
Water: Identification and Treatment, 2nd ed., Manual of Water
per liter in place of chlorine to reduce the formation of
Supply Practices, M7, American Water Works Association,
disinfection by-products. Ozone has been used at dosages
Denver, Colo., 1995.
from 1 to 3 milligrams per liter and has been effective for
3. American Water Works Association Research Foundation,
the control of geosmin or 2-MIB, which are not affected
Current Methodology for the Control of Algae in Surface
by other oxidants. Potassium permanganate at dosages Reservoirs, American Water Works Association Research
from 1 to 5 milligrams per liter, though a relatively weak Foundation, Denver, Colo., 1987.
disinfectant, has been found to be effective for certain 4. R. S. Yoo, W. W. Carmichael, R. C. Hoehn, and S. E. Hrudey,
algae and certain odorants. For example, it works more Cyanobacterial (Blue-Green Algal) Toxins: A Resource Guide,
effectively for the cucumber odor and does not form American Water Works Association Research Foundation,
disinfection by-products. Denver, Colo., 1995.
 ALGAL TURF SCRUBBING: POTENTIAL USE FOR WASTEWATER TREATMENT 213

5. American Water Works Association, Waterborne Pathogens, DEVELOPMENT

1st ed., Manual of Water Supply Practices, M48, American
Water Works Association, Denver, Colo., 1999. ATS was developed by Adey and coworkers at the
6. American Water Works Association, R. D. Letterman, ed., Smithsonian Institution, Washington, D.C., during their
Water Quality and Treatment: A Handbook of Community research on coral reef ecosystems (1–4). Adey found that
Water Supplies, 5th ed., McGraw-Hill, New York, 1999. the algal turfs growing on coral reefs have very high
7. American Water Works Association Research Foundation growth rates despite the low nutrient concentrations
and Lyonnaise des Eaux, J. Mallevialle and I. H. Suffet, (mg m−3 ) in the surrounding seawater (3). The high
eds., Identification and Treatment of Tastes and Odors in growth rates were attributed to two main factors.
Drinking Water, American Water Works Association Research First, the constant surging of seawater across the turf,
Foundation, Denver, Colo., 1987. which by turbulent mixing breaks down boundary layers
8. American Water Works Association Research Foundation and provides a constant supply of nutrients to the
and Lyonnaise des Eaux, I. H. Suffet, J. Mallevialle and algae. Second, the efficient grazing of the algal turf by
E. Kawczynski, eds., Advances in Taste-and-Odor Treatment herbivorous invertebrates and fish maintains the turf
and Control, American Water Works Association Research
species growth in exponential phase.
Foundation, Denver, Colo., 1995.
By using a screen to mimic the reef surface, a dump
9. D. M. C. Rashash et al., Identification and Control of Odor- bucket to provide a wave of turbulent mixing, artificial
ous Algal Metabolites, American Water Works Association
lighting, and periodic manual scrapping of the screen to
Research Foundation, Denver, Colo., 1996.
remove accumulated turf biomass, the algal turf scrubber
can simulate these natural conditions (Fig. 1). Adey and
Hackney (3) found that ATS provides a much more natural
ALGAL PIGMENTS AS INDICATORS. means of maintaining water quality and community
See PALEOLIMNOLOGY: USE OF ALGAL PIGMENTS AS INDICATORS structure in model coral reef ecosystems than bacterial
filters and protein skimmers that are traditionally used in
aquarium systems. Bacterial filters and protein skimmers
are not suitable for use in ecosystem models as they filter
out the zooplankton and larval stages of many aquatic
ALGAL TURF SCRUBBING: POTENTIAL USE FOR organisms.
WASTEWATER TREATMENT Following the success of the ATS system as part of
the coral reef ecosystem model, ATS was used in several
RUPERT CRAGGS other ecosystem models at the Smithsonian Institution,
National Institute of Water and including a salt marsh estuary, a northern temperate
Atmospheric Research littoral system, and a tropical mangrove (2,5,6). ATS was
Hamilton, New Zealand used as an integral part of the 3,400 m−3 mesocosm of
the Space Biosphere 2 project in Arizona (4,7). Algal
turf scrubbers have been used to remove nutrients and
Algal turf scrubbing (ATS) is a novel algal technology that maintain water quality in the aquarium systems of
has been designed and engineered to promote natural several cities in the United States, including Indianapolis,
wastewater treatment processes. Algal turf scrubbing Pittsburgh, and St. Louis, and in the 2,500 m3 Great
improves water quality by passing a shallow stream of Barrier Reef Aquarium in Townsville, Australia (4).
water over the surface of a gently sloped floway in a The use of ATS systems to treat aquaculture wastew-
series of pulses. The floway is colonized by a natural ater at pilot-scale has been demonstrated (3,4) and a
heterogeneous assemblage of periphyton consisting of full-scale Tilapia aquaculture facility is now operating
filamentous algae and symbiotic aerobic bacteria and in San Antonio, Texas. The system uses 1.6 hectares of
fungi. Algal photosynthesis provides oxygen for aerobic parallel algal turf scrubber floways to treat the aquacul-
breakdown of wastewater by heterotrophic bacteria. ture wastewater, and the harvested algae biomass is used
Pollutants are extracted from the wastewater by several as a feed supplement for the fish (8).
processes including assimilation, adsorption, filtration, In terms of wastewater treatment, small-scale labo-
and precipitation. The algal turf is periodically harvested ratory experiments have indicated the potential of ATS
to remove the accumulated periphyton biomass and systems to remove both refractory organic compounds and
associated pollutants from the system. heavy metals from polluted water (9). The capability of

400 – 100 watt

Dump trough metal halide
lights

Algal retention screen Scrubber tray

Figure 1. Schematic diagram of the component of ATS.

214 ALGAL TURF SCRUBBING: POTENTIAL USE FOR WASTEWATER TREATMENT

ATS to remove nutrients from agricultural runoff has for treatment of cow manure. The systems consist of 200-m
also been demonstrated at pilot-scale using a 15.0-m long, long, 1-m wide floways. Their ability to remove nitrogen
0.75-m wide algal turf scrubber floway (10). and phosphorus from raw or anaerobically digested cow
The first large-scale ATS system for the treatment manure diluted with water is being determined. Two pairs
of domestic wastewater was constructed in Patterson, of 50-m by 1-m floways (one pair at a 1% slope and the
California (Plate 1). The capability of ATS to polish other at a 2% slope) are also under construction and will
secondary treated wastewater at several influent flow be used to study seasonal variation in algal growth and
rates (ranging from 108.9 m3 d−1 to 1,336 m3 d−1 ) was nutrient uptake rates (8).
evaluated over a three-year period (1994–1996) (11,12).
The system consisted of a single floway that was 152.4 m
ATS WASTEWATER TREATMENT SYSTEMS
long, 6.5 m wide, and had a surface area of 1,012 m2 .
A second trial to assess the potential of an ATS floway
Algal turf scrubbers are low-cost treatment systems,
to polish domestic wastewater was conducted at Fruitland,
which are simple in design and construction (Fig. 2).
Maryland. The system consisted of 10 parallel, 91.4-m long The floway is formed from graded (0.25% to 2.0% slope),
floways, which were connected so that the effluent from compacted earth covered with an impervious liner and
the first floway was pumped to the top of the second and so an overlying biomass retention screen between two
on, until the wastewater was treated by all 10 floways. The raised sidewalls. The liner may be constructed from any
system was evaluated over a one-year period (1998–1999) impervious material including concrete, asphalt, high-
for its ability to treat secondarily treated wastewater at a density polyethylene, polyvinyl chloride, or sprayed on
flow rate of 284 m3 d−1 (8). asphalt rubber surfaces. Geomembranes with low thermal
At the time of writing, the United States Department of expansion are particularly suitable as they do not wrinkle
Agriculture (USDA) in Beltsville, Maryland in association or retain manufacturing creases.
with the University of Florida, is evaluating ATS systems
Pulsed Influent
The influent is normally delivered to the ATS by a surging
device, which releases the wastewater in a series of waves.
Various techniques have been tested to produce pulsed
wastewater application including pulsed pumping, counter
weighted dump troughs, and displacement troughs. Pulsed
application was found to be important in maintaining
the treatment efficiency of marine ATS systems (3,4,13).
However, research at Patterson showed that surged flow
may not be necessary for freshwater and wastewater
treatment systems (12). Freshwater algae may be better
adapted to the continuous turbulent flow that naturally
occurs in streams and rivers.

Water Depth
Algal growth on ATS systems declines with increased
water depth (3). Vymazal (14) found that periphyton
growth was much higher at shallow depth of 6 cm than
at 64 cm. Therefore, the water depth of ATS systems
is maintained between 2 and 4 cm to allow optimum
exposure of the algal turf to sunlight.

Residence Time
One factor influencing the treatment efficiency of ATS
Plate 1. The ATS pilot system at Patterson, California. See color systems is the time that the wastewater is in contact with
insert. the algal turf (the hydraulic residence time). A simple

Wastewater inflow
Algae retention screen
Effluent collection sump Dump trough
Graded (0.5 – 2.0% slope), compacted earth
Impervious liner

Figure 2. Schematic diagram of an ATS for the treatment of wastewater.

 ALGAL TURF SCRUBBING: POTENTIAL USE FOR WASTEWATER TREATMENT 215

means of optimizing treatment is by controlling residence

time, by changing the hydraulic loading velocity, or by
passing the wastewater down a floway of variable length.

Hydraulic Loading Velocity. Reducing the hydraulic

loading velocity of the Patterson ATS system from
1.36 m d−1 to 0.44 m d−1 increased removal efficiency
and produced an effluent with lower concentrations of
nutrients, total suspended solids (TSS), and biochemical
oxygen demand (BOD) (11).

Floway Length. For a particular hydraulic loading

velocity, a minimum length of floway is required for
efficient treatment to be achieved. Some parameters (e.g.,
dissolved oxygen (DO), temperature and concentrations Plate 2. ATS filamentous algae canopy. See color insert.
of all forms of nitrogen) have a linear correlation
with floway length. Other parameters (particularly, pH,
soluble reactive phosphorus concentration, hardness, and provided a source of algal spores in the mesocosm and
conductivity) change more rapidly at or after a particular ecosystem model systems (4).
floway length.
Seeding and Recirculation. Algal turf development and
Algal Species species diversity are much improved by seeding the floway
with periphyton from nearby streams and rivers. This can
The algal species that typically occur on algal turf
be done either by placing the biomass retention screens
scrubbers can be divided into three ecological types: the
in natural waterways and transferring them to the top
mat-forming species consisting primarily of Cyanophyta,
of the floway once a turf has established or by collecting
the green filamentous periphyton that grow through the
algal turf, breaking it up, and evenly distributing it over
mat to form a canopy (Plate 2), and the diatomaceous
the floway surface. The latter method can bring a floway
epiphytes that grow on the surface of the periphyton or
into full productivity after only a few weeks (12). Species
are embedded within the mat (Fig. 3). The population
diversity may also be maintained by recirculating a portion
density and algal species diversity of algal turfs in
of the algal turf scrubber effluent to introduce algal spores
wastewater treatment systems are much lower than
with the influent.
in systems treating agricultural drainage water and
seawater (3,4,10,12). The dominant algal species is also
Invertebrates
strongly influenced by season. For much of the year
they are cyanobacteria (specifically, Oscillatoria sp.) Several invertebrate species are found in wastewater algal
and diatoms (including Navicula sp., Nitzschia sp., and turfs. The most abundant are amphipods and chironomid
Cyclotella sp.). It is only in the summer that the green midge larvae. Chironomid infestation can be a significant
filamentous periphyton Ulothrix sp. and Stigeoclonium problem because the larvae, which live in cocoons on
sp. are dominant with Cladophora sp., Spyrogyra sp., the floway surface, dislodge the surrounding algal turf
Tribonema sp., Oedogonium sp., and Rhizoclonium sp. causing a reduction in algal turf standing crop. Chironomid
present to a lesser extent. problems have been observed in several other periphyton
The low population density and species diversity of growth systems (15–17). A simple method of controlling
algal turfs in wastewater treatment systems probably is chironomid populations on algal turf scrubber floways is
due to the absence of algal spores in secondary treated by drying out the floway following harvest of the turf,
wastewater compared to agricultural drainage water to especially as the productivity of the turf can be quickly
seed the system (10) and the lack of recirculation, which restored by seeding.

(b)

(a)
(c)

Biomass retention screen Floway liner

Figure 3. Schematic drawing of algal turf species growing on an ATS screen: (a) cyanobacterial
mat, (b) green filamentous algae, (c) epiphytic diatoms.
216 ALGAL TURF SCRUBBING: POTENTIAL USE FOR WASTEWATER TREATMENT

Algal Standing Crop

Self-shading by accumulated turf biomass is one major
factor limiting the growth rate and productivity of
the turf algae. This can be prevented by frequent
harvest, although a sufficient standing crop of algal
biomass is required to maintain regrowth of the turf and
effective treatment immediately following harvest. Several
operational parameters influence the standing crop of ATS
systems, and these are harvest technique and interval, and
growth surface texture.

Harvest Technique. Harvesting the floway has several

roles. It physically removes all the pollutants accumulated
by the algal turf from the system before they naturally
slough off. It simulates heavy grazing of the plant
Plate 3. Unharvested (left) and harvested turf biomass (right)
community, which has been shown to stimulate algal on an ATS floway without a biomass retention screen. See color
growth and nutrient removal (4,17,18). Harvesting also insert.
controls the population densities of most invertebrate
grazers so that pollutants remain entrapped in the turf
community and are not reintroduced into the water. productivity (total solids dry weight) averaged for the
Harvest is achieved by initially stopping the flow of whole floway can attain values of up to 61 g m−2 d−1 ,
wastewater and allowing the water to drain from the although the annual mean productivity (24 g m−2 d−1 ) is
floway. Several harvest methods have been successfully similar to that previously reported for periphyton water
employed, including both manual scraping or vacuum treatment systems (10,12,17). Sedimentation, filtration,
suction of the screen surface on the floway as well as and precipitation of particulates on the floway surface
removal and scrapping of a screen that is rolled up off the contribute to the accumulation of algal turf solids,
floway. Vacuum harvest, which was found to reduce the such that nearly 50% of turf dry weight is nonvolatile
population density of filamentous algae, is not as suitable and contains high concentrations of calcium (Ca) and
as harvest by scraping (12). magnesium (Mg).

Harvest Interval. Because the growth rate of algal WASTEWATER TREATMENT CAPABILITIES
turf varies with season, treatment performance can be
improved by varying the harvest interval over the year to Algal turf scrubber wastewater treatment systems have
maintain sufficient biomass on the ATS floway. A harvest been demonstrated to improve water quality in many
interval of less than one week may be appropriate when the ways. They increase dissolved oxygen concentrations to
highest productivity is achieved during summer, whereas between 100% and 300% saturation during the day and
intervals of up to a month may be required during winter reduce alkalinity, conductivity, and hardness (11).
when productivity is at its lowest. BOD, TSS, and turbidity of wastewater can be removed
by ATS treatment. Removal is due to filtration of
Growth Surface Texture. The texture of the floway liner particulates by filamentous algae and bacterial oxidation.
seems to be of particular importance for the maintenance Particulate filtration is increased when the predominant
of algal species diversity and the promotion of turf turf algae are filamentous rather than the mat-forming
productivity. Particularly, the addition of a biomass cyanobacteria and diatoms (12). Bacterial oxidation of
retention screen to the floway surface improves treatment BOD varies diurnally, correlating with dissolved oxygen
performance by increasing the surface area available for concentration. Mean annual concentrations in Patterson
turf colonization and retaining a larger standing crop of secondary treated wastewater, BOD (10 g m−3 ), chemical
filamentous algae beneath the screen following harvest oxygen demand (COD) (45 g m−3 ), and total organic
(Plate 3). Biomass retention screens were used in much of carbon(TOC) (10 g m−3 ) were minimally affected and are
the original research conducted on ATS (3,4,9,13). The probably close to background levels for ATS systems.
mesh becomes impregnated with algal holdfasts and Removal of BOD and COD by ATS systems was shown
rhizoids, which are not removed during harvest (4,19). by Adey and coworkers (9) although the operational
Algal productivity has been found to vary with screen parameters of the recirculating mesocosm system used
mesh size (optimum mesh size appears to be between in that study are very different from those of a single
1 mm and 5 mm) and color (3). The lower turf productivity pass wastewater treatment floway. The capability of
of black screens is possibly due to its greater absorbance of ATS floways to significantly remove BOD remains to be
heat than white screens, which may inhibit algal growth. evaluated.

Algal Turf Productivity Pathogen and Indicator Removal by ATS

There is a gradual decline in the productivity of the Because of the shallow water depth and large surface area
algal turf down the length of the floway. Algal turf of ATS systems, wastewater is exposed to high amounts of
 ALGAL TURF SCRUBBING: POTENTIAL USE FOR WASTEWATER TREATMENT 217

solar-ultraviolet (UV) radiation, which can result in a high nitrate tend to dominate the turf. Annual mean inorganic
removal rate of the fecal indicator bacteria Escherichia nitrogen levels in the Patterson influent and effluent
coli (11). However, because of the very short residence were 9.2 g m−3 and 6.1 g m−3 , respectively. The longer
time (typically 20 to 40 minutes) of wastewater on the Fruitland ATS system reduced inorganic nitrogen levels
ATS floway, there is little (<1 log) overall reduction in from 20 g m−3 to 4 g m−3 and could probably be enhanced
fecal coliform bacteria concentrations by this system. if additional carbon was added.

Nutrient Removal Phosphorus Removal

ATS systems have been shown to reduce nutrient (see also ACTIVATED SLUDGE — MICROBIOLOGY OF NITROGEN
concentrations to low levels in agricultural runoff and REMOVAL.)
marine systems. In coral reef mesocosms, nitrate can be Mean annual total phosphorus concentrations in the
maintained at levels between 7 and 14 mg m−3 (4). This influent and effluent of Patterson ATS systems were
level is limited by the nitrogen fixation of cyanobacteria 3.1 g m−3 and 1.5 g m−3 , respectively. Periphyton can
present in the system. Phosphorus can be reduced to levels remove phosphorus from wastewater by a combination
below limits of detection. of filtration, adsorption, assimilation (including luxury
ATS wastewater treatment systems are effective at uptake), and precipitation (20). Most of the phosphorus
removing nitrogen and phosphorus from secondary treated removal in ATS wastewater treatment systems is by
wastewater, with greater removal achieved at higher assimilation into algal biomass or by precipitation of
influent concentrations. Nutrient removal is a result of soluble reactive phosphorus (SRP), which is the main form
several microbiological and physical processes (Fig. 4) of phosphorus in secondary treated effluents. Precipitation
that are affected by a number of parameters, including of SRP with cations such as Ca2+ , Mg2+ , and Al3+ is
hydraulic loading, the seasonal change in turf growth and known to occur at pHs between 8.9 and 9.5, depending
species, and attainment of a pH greater than 9.0 in the on the buffering capacity of the water (21). SRP removal
floway effluent. Reduction of nutrient concentrations to therefore is dependent on the wastewater pH reached
mg m−3 levels is limited in wastewater treatment systems while the wastewater is on the floway. The increase in
as there is not enough available organic carbon to enable the pH of the ATS effluent is a result of carbon limitation
the nutrients to be assimilated into algal turf biomass. of the turf algae and their subsequent use of bicarbonate
Addition of a carbon source to the wastewater on the as a carbon source for photosynthesis (22). At Patterson,
floway when carbon is limiting may enhance nutrient a threshold pH of approximately 9.0 was required before
removal by the system. SRP removal occurred to any great degree.
Reducing the hydraulic loading velocity of the ATS
Nitrogen Removal
system from 1.4 to 0.11 m d−1 increases the residence time
(see also ACTIVATED SLUDGE — MICROBIOLOGY OF NITROGEN of the wastewater on the floway and the pH attained in
REMOVAL.) the floway effluent, hence SRP removal by precipitation
All forms of nitrogen can be removed by ATS is improved. SRP precipitation on the algal turf scrubber
wastewater treatment systems. Particulate nitrogen floway is indicated by the reduction of alkalinity, hardness,
removal is related to the dominance of filamentous and conductivity that occurs as the pH of the effluent
species on the floway. The simultaneous removal of both increases. Maintenance of the pH of the ATS system
ammoniacal-nitrogen and nitrate is due to the mixed effluent above the level at which phosphate precipitates
assemblage of algal species including both cyanobacteria may provide an effective means of controlling SRP removal
that generally prefer nitrate and filamentous green algae by ATS systems. The threshold pH may be easily achieved
that prefer ammonium as a nitrogen source. However, by controlling the length of time the wastewater is in
as nitrate is the principal form of inorganic nitrogen in contact with the algal turf. This may be done either by
secondary treated wastewater, algal species that prefer altering the hydraulic loading rate of the floway or by

Particulate filtration NH3 N2

Ammonia volatilisation
at high pH
Assimilation

NH4+

Nitrification

NO32−
Denitrification

SRP precipitation at high pH

Figure 4. Schematic drawing of nutrient removal processes that occur on an ATS.

218 ALGAL TURF SCRUBBING: POTENTIAL USE FOR WASTEWATER TREATMENT

passing the wastewater down different lengths of the Fertilizer: Dried or composted algal turf biomass is very
floway. However, other factors need to be considered in suitable for use as a high-quality soil amendment
determining the residence time of the floway, such as and slow-release fertilizer.
the temperature attained in the ATS effluent, which may Energy: The turf biomass could also be used to
restrict algal growth (12). produce energy through fermentation to alcohol or
Nighttime decline of the ATS system effluent pH to methane gas.
below the threshold pH results in the dissolution of some Feed: Turf algae have a high protein content and thus
of this precipitate and the release of SRP back to the ATS could be used as a feed additive for ruminants,
effluent. Discharge of this released SRP can be prevented poultry, and fish.
either by recirculating the ATS effluent at night or by
greatly reducing the flow at night such that little or no
water is discharged until the effluent pH has risen above CONCLUSION
the threshold again.
ATS systems could be designed to enhance phosphorus Algal turf scrubbers are capable of removing nitrogen and
removal from wastewater by a two-stage process (11). phosphorus from secondary wastewater without added
The first floways would be only operated during the day chemicals and with minimum energy expenditure. When
and have low hydraulic loading to promote pH-mediated covered with filamentous algae, ATS systems are capable
precipitation with cations. This would be followed by of filtering significant amounts of suspended solids and
the second floways with high hydraulic loading and the BOD, TSS, nitrogen, and phosphorus associated with
perhaps recirculation to promote rapid algal growth these suspended solids. Inorganic nitrogen removal is due
and assimilation of phosphorus into the algal biomass. to assimilation and some denitrification, whereas soluble
The efficiency of treatment under both of these modes reactive phosphorus removal is a result of assimilation and
of operation is affected by the amount of biomass on precipitation. Maintenance of the pH of the ATS system
the floway (11). Frequent harvesting would maintain an effluent above the threshold at which precipitation occurs
exponential growth of the biomass in the second floways, could provide a simple means of controlling phosphorus
whereas a longer harvest interval would maintain a large removal by ATS. Control of phosphorus removal by
standing crop to raise pH in the first floways. ATS systems may be achieved by altering the length of
time the wastewater is in contact with the algal turf,
either by reducing the hydraulic loading velocity of the
Algal Turf Nutrient Content
floway or by passing the wastewater down a longer
The mean annual percentages of nitrogen and phosphorus floway. Stopping overnight flow until the threshold pH
in the accumulated solids of the ATS floway at Patterson is regained the following day also prevents resolution
were 3.96% and 1.83%, respectively (11). The values for of precipitated phosphorus at night. A comparison of
nitrogen are typical of algal biomass. However, those the hydraulic loading velocities used for the Patterson
for phosphorus are higher than that (<1%) normally ATS wastewater treatment system with those for pond
associated with periphyton biomass (17,23–25). Adey and wetland treatment systems (26–28) indicates that a
and coworkers (10) found a phosphorus content of 0.4% hydraulic loading velocity of as low as 0.25 m3 m−2 d−1
in periphyton grown on agricultural runoff. The high would remain economical (Table 1).
phosphorus content of the algal turf of ATS wastewater ATS treatment systems have good potential for the
treatment systems further indicates the role of pH- removal of nutrients and other contaminants from
mediated SRP precipitation in phosphorus removal. wastewaters. The simplicity of ATS treatment systems
On the basis of percentages of nitrogen and phosphorus and the ease with which configuration and operational
in the accumulated solids and the mean annual turf parameters such as hydraulic loading, floway length, and
solids accumulation rate (24 g m−2 d−1 ), the mean removal harvest period can be changed should enable process
of nitrogen was 0.95 ± 0.48 g m−2 d−1 and the mean
removal of phosphorus was 0.44 ± 0.22 g m−2 d−1 . The
pilot ATS system treating agricultural runoff achieved Table 1. Comparison of the Typical Hydraulic
phosphorus removal rates of 0.12 g m−2 d−1 (10). Mass Loading Rate of ATS Floway Wastewater Treat-
balance calculations of the wastewater treatment system ment Systems to Those of Other Wastewater Treat-
ment Methods
show that more nitrogen is removed than is accumulated in
the algal turf and indicate that denitrification does occur, Hydraulic Loading Rate
perhaps in anoxic micro environments of the algal turf. Treatment Method (m3 m−2 d−1 )

Activated sludge 10–50

Algal Turf Biomass
Oxidation ditch 1.0–1.5
The solid content of harvested turf biomass usually Algal turf scrubber 0.25–0.5
does not exceed 5%. However, the fibrous material may Stabilization pond 0.1–0.2
simply be dried using drying racks or sand beds and is Rotating biological contactor 0.08–0.2
Rapid infiltration 0.02–0.3
otherwise suitable for dewatering by filter press or drum
Overland flow 0.02–0.2
and spray dryers. Finding a use for the turf biomass Wetland 0.01–0.05
remains unresolved, although there are several potential Irrigation 0.002–0.02
applications (11):
 ALKALIPHILES: ALKALINE ENZYMES AND THEIR APPLICATIONS 219

control and optimization of the system for treatment of 22. C. J. Soeder and E. Hegewald, in M. A. Borowitzka and
different wastewaters. Appropriate use of ATS technology L. J. Borowitzka, eds., Micro-algal Biotechnology, Cambridge
may be for the remediation of eutrophic reservoirs and University Press, Cambridge, U.K., 1988, pp. 59–84.
lakes, for polishing secondary effluents, and for algal 23. D. Swift, Technical Publication 81-5. South Florida Water
biomass production. Management District, West Palm Beach, Fla., 1981.
24. M. T. Auer and R. P. Canale, J. Great Lakes Res. 8, 93–99
Acknowledgments (1982).
Funding for the Patterson study was provided by Aquatic Bio 25. D. H. Kesler, J. Freshwater Ecol. 1, 507–514 (1982).
Enhancement Systems Inc. Texas. The author wishes to thank 26. W. J. Oswald, in M. A. Borowitzka and L. J. Borowitzka,
Professor William Oswald and all other members of the Applied eds., Cambridge University Press, Cambridge, U.K., 1988,
Algae Group at the University of California at Berkeley for their pp. 305–328.
assistance during the evaluation of the Patterson system. ATS is a 27. J. T. Watson et al., in D. Hammer, ed., Constructed Wetlands
patented environmental control system developed by W. Adey and for Wastewater Treatment, Lewis Publishers, Chelsea,
held by the Smithsonian Institution. Aquatic Bio Enhancement Michigan, 1989.
Systems Inc. holds an exclusive license for the technology.
28. C. J. Richardson et al., Annual report. Duke Wetland Centre,
Durham, North Carolina, 1991, 91–98.
BIBLIOGRAPHY

1. W. H. Adey, Coral Reefs 1, 193–201 (1983).

ALKALINE ENZYMES. See ALKALIPHILES: ALKALINE
ENZYMES AND THEIR APPLICATIONS
2. W. H. Adey, in W. Jordan, J. Aber, and M. Gilpin, eds.,
Restoration Ecology: Progress toward a Science and Art of
Ecological Healing, Blackwell Science, Malden, Mass., 1987.
3. W. H. Adey and J. M. Hackney, in W. H. Adey, ed., The
Biology, Ecology, and Mariculture of Mithrax Spinosissimus
ALKALIPHILES. See EXTREMOPHILES: LIFE IN EXTREME
ENVIRONMENTS
Utilizing Cultural Algal Turfs, The Mariculture Institute,
Washington, D.C., 1989.
4. W. H. Adey and K. Loveland, Dynamic Aquaria: Building
Living Ecosystems, Academic Press, New York, 1998.
5. S. L. Williams and W. H. Adey, Aquat. Bot. 16, 181–188 ALKALIPHILES: ALKALINE ENZYMES AND THEIR
(1983). APPLICATIONS
6. L. Tangley, Bioscience 35, 618–619 (1985).
7. W. H. Adey, in Encyclopedia of Environmental Biology, KOKI HORIKOSHI
Academic Press, New York, 1995. Kanagawa and Toyo University
8. L. Bridgewater, Research note. Water Environ. Technol. 11, Kawagoe, Japan
59 (1999).
9. W. H. Adey, C. Luckett, and M. Smith, Ecolog. Eng. J. There are no precise definitions of what characterizes
Ecotechnol. 7, 191–212 (1996). an alkaliphilic or an alkali-tolerant organism. Several
10. W. H. Adey, C. Luckett, and K. Jensen, Restor. Ecol. March, microorganisms exhibit more than one pH optimum for
29–39 (1993). growth, depending on the growth conditions, which are
11. R. J. Craggs et al., Water Sci. Technol. 33(7), 191–198 influenced by nutrients, metal ions, and temperature.
(1996a). Therefore, in this article, the term alkaliphile is used
12. R. J. Craggs, W. H. Adey, B. K. Jessup, and W. J. Oswald, for microorganisms that grow optimally or very well at pH
Ecol. Eng. 6, 149–169 (1996). values above 9, often between 10 and 12, but cannot grow
13. W. H. Adey and T. Goertemiller, Phycologia 26, 374–386 or grow only slowly at the near-neutral pH value of 6.5.
(1987). The discovery of alkaliphiles occurred fairly recently.
14. J. Vymazal, J. Hydrobiologia. 166, 225–237 (1988). Only 16 scientific papers on alkaliphiles could be found
15. M. A. Lock et al., Oikos 42, 10–22 (1984). when the author started experiments on alkaliphilic
16. P. E. Cook, J. P. Hoffmann, J. W. Morris, and L. S. Stuart, bacteria in 1968 (1). Since then, the author and his
Project 84-06 VWRRC, University of Vermont, Nebraska, colleagues have isolated a great number of alkaliphilic
1986. microorganisms using a variety of different media.
17. L. S. Davis, J. P. Hoffmann, and P. W. Cook, J. Phycol. 26, Alkaliphiles are a diverse group, including representatives
617–623 (1990). from the three major domains of life; many of these have
18. A. Sladeckova, P. Marvan, and J. Vymazal, in R. G. Wetzel, been proposed as new taxa. They can be isolated from
ed., Periphyton of Freshwater Ecosystems, Dr. W. Junk nominal environments, but viable counts of alkaliphiles
Publishers, The Hague, The Netherlands, 1983, pp. 299–303. are higher in alkaline environments. The particular
19. S. H. Brawley and W. H. Adey, Mar. Biol. 61, 167 (1981). adaptations of alkaliphiles to their environment are not
20. D. Swift and R. Nicholas, Technical Publication 87-2, South completely understood, but it is thought that the cell
Florida Water Management District, West Palm Beach, Fla., surface plays a key role in maintaining the intracellular
1987. pH between 7 and 8.5.
21. D. K. Belsare and S. D. Belsare, Arch. Hydrobiol. Beih. The industrial use of alkaliphilic microorganisms has
Ergebn. Limnol. 28, 123–128 (1987). a long history in Japan. From ancient times, indigo
220 ALKALIPHILES: ALKALINE ENZYMES AND THEIR APPLICATIONS

from indigo leaves has been reduced by the particular GENETIC MAPS OF CHROMOSOMAL DNAS OF
bacteria that grow under the highly alkaline conditions of ALKALIPHILIC BACILLUS STRAINS
the traditional process called ‘‘indigo fermentation.’’ The
most important factor in this process is the control of How alkaliphiles have adapted to their alkaline envi-
the pH value. Formerly, this could only be accomplished ronment is one of the most interesting and challenging
by the skill of the craftsmen. It was not until we topics facing microbiologists. In addition, the genes of alka-
rediscovered alkaliphiles that the process could be studied liphiles are potentially a valuable source of information
and understood from a microbiological point of view. Even waiting to be explored and exploited by biotechnologists.
then, alkaliphiles remained little more than interesting Physical maps of the chromosomes of two alkaliphilic
biological curiosities, and no further industrial application Bacillus strains, Bacillus pseudofirmus OF4 (3) and Bacil-
was attempted or even contemplated. The situation lus halodurans C-125 (4), have been published; the com-
has changed since then. The first paper concerning plete sequence of the genome of B. halodurans C-125
an alkaline protease was published in 1971 (2). Now, has also been determined (5). The physical map of the
biological detergents contain enzymes, notably proteases chromosome in B. pseudofirmus OF4 is consistent with
and cellulases derived from alkaline bacteria; 60% of the a circular chromosome of approximately 4 Mb, with an
total world production of enzymes is destined for the extrachromosomal element of 110 kb (3,6). Although the
laundry detergent market. The industrial production of analysis is still in progress, several open reading frames
cyclodextrin using cyclomaltodextrin glucanotransferase for Na+ /H+ antiporters that may be involved with pH
is another important application of alkaliphile-derived homeostasis have been detected. The physical map of the
enzymes. This enzyme reduced the cost of production chromosome in B. halodurans C-125 is consistent with a
and paved the way for the use of cyclodextrin in large circular chromosome of approximately 4.20 Mb. As can be
quantities in foodstuffs, chemicals, and pharmaceuticals. seen in Figure 1, many open reading frames show signif-
It has also been reported that alkali-treated woodpulp icant similarities to those of other microorganisms, such
could be biologically bleached by xylanases produced by as a fragment of an intracellular serine protease from
alkaliphiles. Other applications are also contemplated. Thermoactinomyces sp. HS682 or a G6-amylase from the

0° secY, rpoA
pel 4-b 3.94M 0.1M
H-167
tupG oriC 0.44M
3.776M
tupA
galT 3.476M
3.719 groEL 0.5M
M

galT
pel 4-a
0.75M

bgl 3.35M AH 101

0.928M

Bacillus halodurans C-125

270° 90°
4.20 Mb

Alkaliphily ?
2.94M

sigA
pALK 1.4M

resB 1.6M
rnh
Thermoactinomyces
HS682 2.30M xy 1 (A) 180°
2.2M

Figure 1. Genetic map of the chromosome of B. halodurans c-125.

 ALKALIPHILES: ALKALINE ENZYMES AND THEIR APPLICATIONS 221

alkaliphilic Bacillus clausii H-167. These DNA fragments Independently, Aono and coworkers isolated an alkali-
do not code for functional enzymes because they are trun- hypersensitive mutant AS-350 from B. halodurans C-125
cated, but they do suggest that there must be some degree as a host for protoplast transformation. A gene that
of horizontal gene transfer between B. halodurans and restored the alkaliphilic growth of the mutant was cloned
other microorganisms (5,7–11). from the parent strain (23). The mutation in the AS-350
was complemented with a 1.0-kb fragment expressed as
‘‘Alkaliphily’’ in Figure 1. However, role of this fragment
MECHANISMS OF CYTOPLASMIC pH REGULATION
is not yet clear.
Krulwich and her coworkers have focused their studies
The cells of B. halodurans C-125 have two barriers to
on the B. pseudofirmus OF4, which is routinely grown
reduce pH values from 10.5 to 8.
on malate-containing medium either at pH 7.5 or at
The plasma membrane must be kept below pH 9
pH 10.5. Current work is directed toward clarification of
because it is very unstable at alkaline pH values much
the characteristics and energetics of membrane-associated
below the pH optimum for growth (12). It has been
proteins that must catalyze inward proton movements (see
suggested that the cell wall may play a role in protecting
the previous section by M. Ito).
the cell from alkaline environments. In addition to
peptidoglycan, alkaliphilic B. halodurans C-125 contains
certain acidic polymers, such as teichuronopeptide, ALKALINE ENZYMES
composed of polyglucuronic acid and a polypeptide of
acidic amino acids. The negative charges on the acidic Studies of alkaliphiles have led to the discovery of many
non-peptidoglycan components may act as a barrier to types of enzymes that exhibit interesting properties. The
sodium and hydronium ions and repulse hydroxide ions, first report concerning an alkaline enzyme published in
and consequently, may assist cells to grow in alkaline 1971 described an alkaline protease produced by Bacillus
environments (13,14). Aono and coworkers reported that sp. 221 (2). More than 100 new enzymes have been isolated
teichuronopeptide (TUP) is one of the major structural and purified in many laboratories (24). Some of these are
components of the cell wall of B. halodurans C-125. summarized in Table 1.
A mutant defective in TUP synthesis grows slowly at
alkaline pH. An upper limit of pH for growth of the mutant Alkaline Proteases
was 9.4, while that of the parental strain C-125 was 11.0.
A gene tupA, directing synthesis of TUP, was cloned from In 1971, Horikoshi (2) reported the production of an
the parental C-125 chromosomal DNA. Introduction of the extracellular alkaline serine protease from alkaliphilic
tupA gene into the TUP-defective mutant complemented B. clausii No. 221. This strain, isolated from soil, produced
the mutation responsible for the pleiotropic phenotypes large amounts of alkaline protease that differed from the
of the mutant, leading to simultaneous disappearance of subtilisin group. The optimum pH of the purified enzyme
the defect in TUP synthesis, the diminished ability for was 11.5 with 75% of the activity maintained at pH 13.0.
cytoplasmic pH homeostasis, and the low tolerance for The enzyme was completely inhibited by diisopropylflu-
alkaline conditions. These results demonstrate that the orophosphate or 6 M urea but not by ethylenediamine
acidic polymer TUP in the cell wall plays a role in pH tetraacetic acid or p-chloromercuribenzoate. The addition
homeostasis in this alkaliphile (15). of a 5 mM solution of calcium ions was reflected in a
Plasma membranes may also maintain pH homeostasis 70% increase in activity at the optimum temperature
by using Na+ /H+ antiporter system (-dependent and (60 ° C). The gene encoding this enzyme was cloned in
pH-dependent), K+ /H+ antiporter and ATPase-driven E. coli and expressed in Bacillus subtilis. An open reading
H+ expulsion. Recent works in several laboratories on frame of 1,140 bases, identified as the protease gene was
the critical antiporters have begun to clarify the number preceded by a putative SD sequence (AGGAGG) with a
and characteristics of the porters that support active spacing of 7 bases. The deduced amino acid sequence had
mechanisms of pH homeostasis (see AEROBIC ALKALIPHILES). a peptide of 111 residues followed by the mature protease
The author’s group isolated a non-alkaliphilic mutant comprising 269 residues (25). Subsequently, two Bacillus
strain from B. halodurans C-125 as the host for cloning species, AB42 and PB12, which also produced an alka-
genes related to alkaliphily. A 3.7-kb parental DNA line protease were reported (61). These strains exhibited
fragment (pALK fragment) from the parental strain a broad pH range of pH 9.0 to 12.0, with a temperature
restored the growth of an alkaline-sensitive mutant 38,154 optimum of 60 ° C for AB42 and 50 ° C for PB12. Since
at alkaline pH. The transformant was able to maintain an these reports, many alkaline proteases have been isolated
intracellular pH that was lower than the external pH and from alkaliphilic microorganisms (62–66). Fujiwara and
contained an electrogenic Na+ /H+ antiporter driven only coworkers (67) purified a thermostable alkaline protease
by  membrane potential, interior negative. Membrane from a thermophilic alkaliphilic Bacillus sp. B18. The opti-
vesicles prepared from the mutant 38,154 did not mum pH and temperature for the hydrolysis of casein were
show membrane potential -driven Na+ /H+ antiporter pH 12 to 13 and 85 ° C, both of which are higher than those
activity. These results indicate that the mutant 38,154 of alkaline proteases. Han and Damodaran (68) reported
affects, either directly or indirectly, electrogenic Na+ /H+ the purification and characterization of an extracellular
antiporter activity. This was the first report of a DNA endopeptidase from a strain of B. pumilus displaying high
fragment responsible for a Na+ /H+ antiporter system in stability in 10% (w/v) sodium dodecyl sulfate and 8 M
the mechanism of alkaliphily (16–22). urea. Following these studies, many alkaline proteases for
222 ALKALIPHILES: ALKALINE ENZYMES AND THEIR APPLICATIONS

Table 1. Major Extracellular Alkaline Enzymes Produced by Alkaliphiles

Optimum Stable
pH∗ pH Applications References

Alkaline Proteases
B. clausii No. 221 11.5–12.0 4–11 Detergent additives (2,25)
B. halodurans AH-101 12–13 5–13 (26)
Bacillus sp. D-6 10.0–11.0 4–12 (27)

Alkaline Amylase
Bacillus sp. A-40-2 10.5 6–9 α-amylase type (28)
B. clausii H-167 10–11 6–11 G6 forming∗∗ (10,29–31)
Bacillus sp. 38-2 4.5–9 6–9 CD forming∗∗∗ (32–35)

Alkaline Pectinase
Bacillus sp. P-4-N 10.0 5–9 (36,37)

Alkaline Pullulanase
B. halodurans No. 202-1 9.0 6–10 (38)
Micrococcus sp. 207 7.0–8.0 Food processing (39,40)

Alkaline Cellulase
Bacillus sp. N-4 6–11 5–11 Detergent additives (41,42)
Bacillus sp. No. 1,139 9 5–11 (42–47)

Alkaline Lipase
Pseudomonas sp. 9.5 Detergent additives (48)
Bacillus sp. 9.5 9.5 (49)

Xylanses
B. halodurans C-59-2 5.5–9 5–9 Biobleaching (50)
B. halodurans C-125 6–10 4–12 (51)
Bacillus sp. TAR-1 5.0–9.5 9 (52)

Alkaline Chitinase
Bacillus sp. BG-11 7.5–9.0 6.0–9.0 (53–55)

Alkaline Mannanase
Bacillus sp. AM001 9.0–9.5 8–9 Food processing (56–58)

β-1,3-gucanase
Bacillus sp. K-12-5 6–9 6–8 (60)
Bacillus sp. AG-430 9.0–10.0 4–10
∗
Optimum pH for enzyme action.
∗∗
Maltohexaose forming.
∗∗∗
Cyclodextrin forming.

laundry detergent additives were isolated from various was stable from pH 5 to 12. The optimum reaction pHs
alkaliphilic Bacillus strains, although their properties are for feather powder and casein were 8.5 and 10.5 to 11.5,
almost the same as those reported previously (69–71). respectively. Zaghloul and coworkers also reported isola-
Some of the enzymes are now commercially available tion, identification, and keratinolytic activity of several
as detergent additives. feather-degrading bacteria isolated from an Egyptian soil.
Takami and coworkers (26) isolated a new alkaline pro- These isolates could degrade chicken feathers (78).
tease from alkaliphilic B. halodurans No. AH-101. The
enzyme was most active toward casein at pH 12 to 13 and Other Proteases
stable under 10-minute incubation at 60 ° C and pH 5 to Several alkaline proteases produced by alkaliphilic acti-
13. The optimum temperature was about 80 ° C in the pres- nomycetes have been reported. Tsuchiya and cowork-
ence of 5 mM calcium ion. The alkaline protease showed ers (11,79,80) isolated thermostable alkaline protease
a higher hydrolyzing activity against insoluble fibrous from alkaliphilic Thermoactinomyces sp. HS682. The
natural proteins such as elastin and keratin in compar- strain grew in alkaline media between pH 7.5 to 11.5.
ison with subtilisins and proteinase K (72–76). Cheng Maltose gave the highest productivity of protease at a
and coworkers (77) reported a keratinase of a feather- concentration of 10 g/L. The protease had the maximum
degrading Bacillus licheniformis PWD-1. This enzyme proteolytic activity around pH 11.0 and at 70 ° C. In the
 ALKALIPHILES: ALKALINE ENZYMES AND THEIR APPLICATIONS 223

presence of Ca2+ ions, maximum activity was observed were observed. No alkaline amylases produced by neu-
at 80 ° C. An intracellular alkaline serine protease gene trophilic microorganisms have so far been reported. Recent
of alkaliphilic Thermoactinomyces sp. HS682 was cloned studies revealed that the starch-degrading enzymes α-
and expressed in Escherichia coli. Yum and coworkers (81) amylase and cyclodextrin glycosyltransferase (CGTase)
purified an extracellular alkaline serine protease produced are functionally and structurally closely related.
by Streptomyces sp. YSA-130. The optimum temperature
and pH for the enzyme activity were 60 ° C and 11.5, respec- a-Amylases of Alkaliphilic Bacillus Strains
tively. The enzyme was stable at 50 ° C, and between pH 4
The production of the alkaline amylase was first achieved
and 12. A keratin-degrading serine protease of Strepto-
in an alkaliphilic Bacillus species strain No. A-40-2
myces pactum DSM 40,530 was purified by casein agarose
(ATCC21592) that was selected from about 300 colonies of
affinity chromatography by Bockle and coworkers (82).
bacteria grown in Horikoshi-II medium (28). The enzyme
The proteinase was optimally active in the pH range from
is most active at pH 10.0 to 10.5 and retains 50% of
7 to 10 and at temperatures from 40 to 75 ° C. After incu-
its activity between pH 9.0 and 11.5. The enzyme is not
bation with the purified proteinase, a rapid disintegration
inhibited by 10 mM EDTA at 30 ° C and is completely
of whole feathers was observed. Extracellular proteolytic
inactivated by 8 M urea. The enzyme can hydrolyze
activity was also detected in the haloalkaliphilic archaeon
70% of starch to yield glucose, maltose, and maltotriose
Natronococcus occultus as the culture reached the sta-
and is a type of saccharifying α-amylase. Boyer and
tionary growth phase (83). Proteolytic activity was pre-
Ingle (88,89) reported an alkaline amylase in the strain
cipitated with ethanol and subjected to a preliminary
NRRL B-3881, which was the second report of an
characterization. Optimum conditions for activity were
alkaline amylase. The B-3881 amylase had its optimum
attained at 60 ° C and 1 to 2 M NaCl or KCl. Gelatin
pH at 9.2. The enzyme yields maltose, maltotriose, and
zymography in the presence of 4 M betaine revealed a
a small amount of glucose and maltotetraose, all of
complex pattern of active species with apparent molec-
which have a β-configuration. Considerable diversity of
ular masses ranging from 50 to 120 kDa. Subsequently,
α-amylases has been reported. Kim and coworkers (90)
several haloalkaliphilic proteases have been isolated from
reported that an alkaliphilic Bacillus sp. Strain GM8901
haloalkaliphilic archaea. These enzymes were dependent
produced five alkaline amylases in a culture broth.
on high NaCl concentrations in order to express protease
McTigue and coworkers studied the alkaline amylases
activity and stability (84,85).
of three alkaliphilic Bacillus strains (91,92). Bacillus
halodurans A-59 (ATCC 21591), Bacillus sp. NCIB 11203,
Industrial Applications of Alkaline Proteases and Bacillus sp. IMD370 produce alkaline α-amylases
with maxima for activity at pH 10.0. One of the strains
1. Detergent Additives. isolated, Bacillus sp. No. H-167, produced α-amylases that
The main industrial application of alkaliphilic yielded maltohexaose and the main product from starch.
enzymes is in the detergent industry, and detergent The nucleotide sequence of the G6-amylase gene from
enzymes account for approximately 30% of the total alkaliphilic Bacillus sp. H-167 was determined. The open
worldwide enzyme production. Not all of these reading frame of the gene consisted of 2,865 bp encoding
are produced by alkaliphilic bacteria. However, 955 amino acids. The DNA sequence and the deduced
almost all alkaline proteases have been produced by amino acid sequence of the G6-amylase gene showed
alkaliphilic Bacillus strains and are commercially no homology with those of other bacterial α-amylases,
available. although the consensus amino acid sequences of the active
2. Dehairing. center were well conserved (10,29–31). It is of interest that
this nucleotide sequence, which is fragmented, was well
Alkaline enzymes have been used in the hide-
conserved in the DNA of B. halodurans C-125, although
dehairing process, in which dehairing is carried out
no enzymatic activity was expressed.
at pH values between 8 and 10. These enzymes are
Kelly and coworkers (93) found that the alkaline
commercially available from several companies.
amylase of Bacillus sp. IMD 370 could hydrolyze raw
3. Others. starch. The enzyme digested raw cornstarch to glucose,
An interesting application of alkaline protease was maltose, maltotriose, and maltotetraose. The maximum
developed by Fujiwara and coworkers (66,67,86,87). pH for raw starch hydrolysis was pH 8.0 compared with
They reported that an alkaline protease was used pH 10.0 for soluble starch hydrolysis. It is of interest
to decompose the gelatinous coating of X-ray films, that degradation of raw starch was stimulated sixfold
from which silver was recovered. Protease B18 had in the presence of β-cyclodextrin. Lin and coworkers (94)
a higher optimum pH and temperature, around 13.0 have recently reported the production and properties of
and 85 ° C, respectively. The enzyme was most active a raw-starch-degrading amylase from the thermophilic
toward gelatin on film at pH 10. and alkaliphilic Bacillus sp. TS-23. Activity staining
revealed that two amylases with molecular masses of
STARCH-DEGRADING ENZYMES 150 and 42 kDa were produced. The 42-kDa amylase only
hydrolyzed raw starch. Igarashi and coworkers isolated a
The first alkaline amylase was produced in Horikoshi-II novel liquefying α-amylase (LAMY) from cultures of an
medium by cultivating alkaliphilic Bacillus sp. No. A-40- alkaliphilic Bacillus isolate, KSM-1378 (95). The enzyme
2 (28). Several types of alkaline starch-degrading enzymes had a pH optimum of 8.0 to 8.5 and displayed maximum
224 ALKALIPHILES: ALKALINE ENZYMES AND THEIR APPLICATIONS

activity at 55 ° C. The structural gene for the amylase 400 soil bacteria (108). An enzyme of isolate No. 76 was
contained a single open reading frame 1,548 bp in length, partially purified by starch adsorption and some of its
corresponding to 516 amino acids that included a signal properties were investigated. The enzyme properties seem
peptide of 31 amino acids. The four conserved regions were to be quite similar to those of Bacillus sp. 38-2 enzyme (33).
found in the deduced amino acid sequence. Essentially, Chung and coworkers isolated a thermostable CGTase
the sequence of LAMY was consistent with the tertiary from alkaliphilic Bacillus stearothermophilus ET1 (109).
structures of reported amylolytic enzymes, which are The optimum pH for the enzyme-catalyzed reaction was
composed of domains A, B, and C. Furthermore, Igarashi pH 6.0, and the optimum temperature was observed at
and coworkers (96) improved the thermostability of the 80 ° C. A 13% (w/v) cornstarch solution was liquefied and
amylase by deleting an arginine-glycine residue in that converted to CDs solely using this enzyme. The cornstarch
molecule. Then they reported that thermostability of the conversion was 44%, and α-, β-, and γ -CDs were produced
enzyme was also improved by proline substitution (Arg124 in the ratio of 4.2 : 5.9 : 1. Terada and coworkers studied
to Pro124) (97). the initial reaction of CGTase from an alkaliphilic Bacillus
Another modification to improve stabilization was sp. A2-5a on amylose (110). Cyclic α-1,4-glucans with a
reported by Villalonga, and coworkers (98). Carboxy- degree of polymerization ranging from 9 to more than 60,
methylcellulose activated by periodate oxidation was cova- in addition to α-, β-, and γ -cyclodextrin, were detected at
lently linked to α-amylase. The thermostability and pH an early stage of enzymatic reactions. Subsequently, large
stability were improved for α-amylase by this modifica- cyclic α-1,4 glucans were converted into smaller cyclic
tion. The conjugate was also more resistant to the action α-1,4-glucans. The final major product was β-CD. From
of denaturant agents such as urea and sodium dodecyl- the industrial point of view, γ -CD is of practical interest
sulfate. They concluded that modification of enzymes by because it is rare and can encapsulate larger compound
the anionic polysaccharide carboxymethylcellulose might molecules. Recently, Parsiegla and coworkers reported
be a useful method for improving enzyme stability under that mutation of Bacillus circulans Strain No. 8 was
various denaturing conditions. effective in increasing the γ -cyclodextrin production (111).
Yamane’s group has extensively investigated molecular
Other a-Amylases
structure of the CGTase of alkaliphilic Bacillus sp.
Kimura and Horikoshi (39,40,99–102) isolated a number 1,011. Kimura and coworkers isolated the enzyme,
of starch-degrading alkaliphilic psychrophilic microorgan- purified, and cloned its gene (112,113). The enzyme,
isms from the environments. Recently, a gene coding for consisting of 686 amino acid residues, was crystallized
a new amylolytic enzyme from Pseudomonas sp. KFCC and subjected to X-ray analysis. The molecule consists
10818 was cloned, and its nucleotide sequence was deter- of five domains, designated A to E, and its backbone
mined (103). A deduced amino acid sequence contained structure was similar to the structure of other bacterial
four highly conserved regions of α-amylases. A haloal- CGTases. The molecule had two calcium binding sites
kaliphilic Natronococcus sp. strain Ah-36 produced an where calcium ions were coordinated by seven ligands,
extracellular maltotriose-forming amylase (104). The amy- forming a distorted pentagonal bipyramid. Three histidine
lase exhibited maximal activity at pH 8.7 and 55 ° C in the residues in the active center of CGTase participate
presence of 2.5 M NaCl. Kobayashi and coworkers (104) in the stabilization of the transition state. His-327 is
have cloned this α-amylase and expressed it in Haloferax especially important for catalysis over the alkaline pH
volcanii. range (114). Three-dimensional structures of cyclodextrin
glucanotransferases (CGTases) revealed four aromatic
Cyclomaltodextrin Glucanotransferases (CGTase) residues, which are highly conserved among CGTases but
Nakamura and Horikoshi discovered several alkaliphilic not in α-amylases, are located in the active center (115).
Bacillus strains producing CGTases. A crude enzyme of Ishii and coworkers analyzed the crystal structure of
Bacillus sp. No. 38-2 was a mixture of three enzymes: asparagine-233-replaced enzyme. The enzyme that was
acid CGTase having optimum pH for enzyme action a site-directed mutation of histidine-233 to asparagine
at 4.6, neutral CGTase at 7.0, and alkaline CGTase changed the nature of the enzyme such that it no longer
at 8.5 (32–35). In 1975, Matsuzawa and coworkers produced α-cyclodextrin. The neighborhood of asparagine-
established the industrial production of cyclodextrin 233, maintaining the architecture of the active site cleft,
by using the crude CGTase of Bacillus sp. 38-2 (105). seems to be responsible for the change in molecular
Since then, several alkaliphilic microorganisms producing recognition of both substrate and product of the mutant
CGTases have been reported (35,105). Georganta and CGTase (116,117).
coworkers (106) isolated CGTase producing alkaliphilic
psychrophilic bacteria from samples of deep-sea bottom Industrial Production of Cyclodextrins
mud. The isolate No. 3-22 grew at 4 ° C and showed activity
in both broad temperature and pH 5 to 9 ranges. The In 1969, Corn Products International Co. in the United
CGTase produced predominantly β-cyclodextrin (β-CD), States began producing β-CD using B. macerans CGTase.
with minor amounts of α- and γ -CDs. The formation of Teijin Ltd. in Japan also produced β-CD using the
various cyclodextrins was observed after accumulation macerans enzyme in a pilot plant. However, there were
of maltooligosaccharides with degrees of polymerization several serious problems in both the production processes:
more than seven (107). Salva and coworkers isolated sixty- (1) The yield of CD from starch was not high, and (2) toxic
eight CGTases from alkaliphilic bacteria from among organic solvents such as trichloroethylene, bromobenzene,
 ALKALIPHILES: ALKALINE ENZYMES AND THEIR APPLICATIONS 225

and toluene were used to precipitate CD because of the bacteria were alkaliphilic and moderately thermoactive;
low conversion rate. optimum activity was detected at pH 8.0 to 10.0 and
The use of CGTase of alkaliphilic Bacillus sp. No. 38- between 50 ° C and 60 ° C.
2 overcame all these weak points and led to the mass In screening alkaline cellulases for detergent additives,
production of crystalline α-, β-, γ -CD at low cost without Ito’s group isolated a novel alkaline pullulanase from
using any organic solvents. The yield of CD was 85 to alkaliphilic Bacillus sp. KSM-1876, which was identified
90% from amylose and 70 to 80% from potato starch on as a relative of B. circulans (125,126). The enzyme had
a laboratory scale. Owing to the high conversion rate, an optimum pH for enzyme action of around 10.0 to
CDs could be directly crystallized from the hydrolyzate 10.5. This enzyme is a good candidate for use as an
of starch without the addition of organic solvents (105). additive to dishwashing detergents. Furthermore, another
These simple methods reduced the cost of β-CD from alkaline amylopullulanase from alkaliphilic Bacillus sp
200,000 yen to 1,000 yen/kg, and that of α-CD to within KSM-1378 was also found. This enzyme efficiently
3,000 yen/kg. This has paved the way for its use in large hydrolyzed α-1,4 and the α-1,6 linkages of amylose,
quantities in foodstuffs, chemicals, and pharmaceuticals. amylopectin, and glycogen at alkaline pH values. The
Several other processes to produce CDs have been kinetic studies revealed two independent active sites
reported after Matsuzawa’s original paper (105). Kim for the α-1,4 and α-1,6 hydrolytic reactions. Incubation
and coworkers reported an enzymatic production of of the enzyme at 40 ° C and pH 9.0 caused complete
cyclodextrins from milled cornstarch in an ultrafiltration
inactivation of the amylase activity within 4 days,
membrane bioreactor to reduce product inhibition (118).
but the pullulanase activity remained at the original
As compared with an operation without ultrafiltration, the
level under the same conditions (126). This alkaline
conversion yield was increased by 57% in a batch operation
amylopullulanase can therefore be considered to be
with ultrafiltration. Okada and coworkers also developed
a ‘‘two-headed’’ enzyme molecule. Limited proteolysis
a bioreactor system with the enzyme immobilized on
with papain also revealed that the α-1,6 and α-1,4
a capillary membrane (119). The percentage of CDs to
hydrolytic activities were associated with two different
total sugar obtained was slightly more than 60% under
most operating conditions. Abraham (120) immobilized active sites (127). Furthermore, amino acid sequence and
CGTase for the continuous production of cyclodextrins. molecular structure analysis of the enzyme showed two
The immobilized enzyme had a pH optimum shifted to different active sites (128). The enzyme was observed by
the alkaline side (from 6.5 to 7.5) and had a reduced transmission electron microscopy; it appeared to be a bent
temperature optimum (from 60 ° C to 50–55 ° C). Lima and dumbbell-like molecule with a diameter of approximately
coworkers developed a unique cyclodextrin production 25 nm.
process (121). A 10% (w/v) solution of cassava starch Takagi and coworkers reported diversity in size
liquefied with α-amylase was incubated with CGTase and alkaliphily of thermostable α-amylase pullulanases
using only the enzyme, added ethanol (from 1 to 5%) produced by recombinant E. coli, B. subtilis, and the wild-
followed by the addition of yeast, Saccharomyces cerevisiae type Bacillus sp. XAL601 (129). It was revealed that
(12% w/v), plus nutrients. However, production costs were the noncatalytic C-terminal region may be responsible
not reported. for the high optimum pH of the enzyme activity. These
observations were also reported in CGTase (130,131) and
Pullulanases alkaline CMCase (132).
Lin and coworkers purified a thermostable pullulanase
In 1975, Nakamura and coworkers discovered an alkaline
from thermophilic alkaliphilic Bacillus sp. strain TS-23.
pullulanase of Bacillus sp. No. 202-1 (38). The enzyme
This purified enzyme had both pullulanase and amylase
had an optimum pH for enzyme action at 8.5 to 9.0 and
activities. The temperature and pH optima for both
was stable for 24 hours at pH 6.5 to 11.0 at 4 ° C. The
pullulanase and amylase activities were 70 ° C and pH 8
enzyme was most active at 55 ° C and was stable up to
50 ° C for 15 minutes in the absence of a substrate. Kelly to 9, respectively. The enzyme remained more than 96%
and coworkers (122) found that alkaliphilic B. halodurans active at temperatures below 65 ° C, and both activities
No. A-59 (ATCC21591) produced three enzymes, namely, were retained at temperatures up to 90 ° C in the presence
α-amylase, pullulanase, and α-glucosidase, in culture of 5% SDS (133).
broth. These three enzymes were separately produced, In some food industries, enzymes showing activity at
and the levels of α-glucosidase and pullulanase reached lower temperatures have been requested for food pro-
maxima after 24-hour cultivation at the initial pH 9.7. cessing. Psychrophilic bacteria are thought to be potential
Although this pullulanase was not purified, the indicated producers of these enzymes. Kimura and Horikoshi (39,40)
pH optimum was at 7.0. reported that an alkalipsychrophilic strain, Micrococcus
Two highly alkaliphilic pullulanase-producing bacteria sp. 207, produced amylase and pullulanase extracellularly.
were isolated from Korean soils (123,124). The two isolates The pullulanase of Micrococcus sp. 207 was purified to
were extremely alkaliphilic since bacterial growth and an electrophoretically homogeneous state by conventional
enzyme production occurred at pH values ranging from ways. The purified enzyme was free of α-amylase activity.
pH 6.0 to 12.0 for Micrococcus sp. Y-1 and pH 6.0 The enzyme had a pH optimum at 7.5 to 8.0 and was
to 10.0 for Bacillus sp. S-1. The enzyme displayed a relatively thermostable (stable up to 45 ° C). The enzyme
temperature optimum of around 60 ° C and a pH optimum could hydrolyze the α-1,6-linkages of amylopectins, glyco-
of around pH 9.0. The extracellular enzymes of both gens, and pullulan. Although many alkaline pullulanases
226 ALKALIPHILES: ALKALINE ENZYMES AND THEIR APPLICATIONS

have been reported as described earlier, no industrial Bacillus sp. No. KSM-635 from a soil and succeeded in
application has been developed yet. producing an alkaline cellulase as a laundry detergent
additive on an industrial scale.
Besides the KSM-635 enzyme, Shikata and coworkers
CELLULASES
isolated three strains, alkaliphilic Bacillus KSM-19, KSM-
64, and KSM-520, producing alkaline cellulases for
Alkaline Cellulases of Alkaliphilic Bacillus Strains laundry detergents (139). Their activities (pH optima:
Commercially available cellulases display optimum activ- 8.5 to 9.5) were not inhibited at all by metal ions
ity over a pH range from 4 to 6. No enzyme with or various components of laundry products, such as
an alkaline optimum pH for activity (pH 10 or higher) surfactants, chelating agents, and proteinases. With a
had been reported before the rediscovery of alkaliphiles. view to increasing industrial production, Sumitomo and
Horikoshi and coworkers found bacterial isolates (Bacillus coworkers (140,141) overexpressed alkaline cellulase of
sp. No. N4 and No. 1,139) producing extracellular alkaline alkaliphilic Bacillus sp. KSM-64 by using B. subtilis
carboxymethylcellulases (CMCases) (41,43). One of these harboring their vector pHSP64. By this process, they
alkaliphilic Bacillus sp. No. N-4 (ATCC21833) produced produced 30 g of alkaline cellulase in one liter. After
multiple CMCases that were active over a broad pH range the discovery of the industrial application of alkaline
(pH 5 to 10). Sashihara and coworkers cloned the cellu- cellulase as a detergent additive, many microbiologists
lase genes of Bacillus sp. No. N-4 in E. coli HB101 (42). have extensively studied alkaline cellulases (142–147).
Another bacterium, Bacillus sp. No. 1,139, produced one Further details are reviewed by Ito (148).
CMCase, which was purified and shown to have optimum
pH for activity at pH 9.0. The enzyme was stable over the Alkaline Lipases
range of pH 6 to 11 (24 hours at 4 ° C and up to 40 ° C for Although the initial motivation for studying alkaline
10 minutes). The CMCase gene of Bacillus sp. No. 1,139 lipases was its application to detergents, many alkaline
was also cloned in E. coli (43–47). Nakamura and cowork- lipases were significantly inhibited in the presence of
ers (132) and Park and coworkers (134) constructed many either alkylbenzene sulfate or dodecyl benzene sulfonate.
chimeric cellulases from B. subtilis and Bacillus sp. N-4 Watanabe and coworkers (48) conducted an extensive
enzyme genes in order to understand the alkaliphily of N- screening for alkaline lipase-producing microorganisms
4 enzymes. Despite the genes having high homology, the from soil and water samples. Two bacterial strains
pH activity profiles of the two enzymes are quite different; were selected as potent producers of alkaline lipases.
B. subtilis (BSC) has its optimum pH at 6 to 6.5, whereas These were identified as Pseudomonas nitroreducens var.
Bacillus sp. N-4 enzyme (NK1) is active over a broad pH thermotolerans and Ps. fragi. The optimum pH of the two
range from 6 to 10.5. The chimeric cellulases showed var- lipases was 9.5. Both enzymes were inhibited by bile salts
ious chromatographic behaviors, reflecting the origins of such as sodium cholate, sodium deoxycholate, and sodium
their C-terminal regions. The pH activity profiles of the taurocholate at a concentration of 0.25 %.
chimeric enzymes in the alkaline range could be classified A thermophilic lipase-producing bacterium was isolated
into either the BSC or the NK1 type, mainly depending on from a hot spring area of Yellowstone National Park (49).
the origins of the fifth C-terminal regions. The organism characterized as Bacillus sp. grew optimally
at 60 to 65 ° C and in the pH range of 6 to 9. The partially
Alkaline Cellulases from Other Alkaliphiles purified lipase preparation had an optimum temperature
Park and coworkers and Damude and coworkers (135,136) of 60 ° C, at an optimum pH of 9.5. It retained 100%
studied a semi-alkaline cellulase produced by alkaliphilic of the original activity after being heated at 75 ° C for
Streptomyces strain KSM-9. Dasilva and coworkers (137) half an hour. The enzyme was active on triglycerides
reported two alkaliphilic microorganisms, Bacillus sp. containing fatty acids with a carbon chain length of 16 to
B38-2 and Streptomyces sp. S36-2. The optimum pH and 22 as well as on natural fats and oils. Bhushan (149,150)
temperature of the crude enzyme activities ranged from found a lipase produced from an alkaliphilic Candida
6.0 to 7.0 at 55 ° C for the Streptomyces and 7.0 to 8.0 at species in a solid-state fermentation. The lipase from this
60 ° C for the Bacillus sp. B38-2. However, their results microorganism had temperature and pH optima of 40 ° C
indicated that the properties of these enzymes were not and 8.5, respectively, and was stable at 45 ° C for 4 hours.
sufficient for industrial purposes. Enzyme activity was stimulated by Ni2+ and Ca2+ ions
whereas Fe2+ and Fe3+ ions inhibited the activity.
Large-scale industrial application such as laundry
Cellulases as Laundry Detergent Additives
detergent additives has not been reported yet.
The discovery of alkaline cellulases created a new
industrial application of cellulase as a laundry detergent
XYLANASES
additive. Ito (personal communication) mixed alkaline
cellulases with laundry detergents and studied the
Xylanases of Alkaliphilic Bacillus Strains
washing effect on cotton underwear. The best results were
obtained by one of the alkaline cellulases produced by an The first paper describing a xylanase from alkaliphilic
alkaliphilic Bacillus strain. However, the yield of enzyme bacteria was published in 1973 by Horikoshi and
was not sufficient for industrial purposes. Consequently, Atsukawa (50). The purified enzyme of Bacillus sp. No. C-
Yoshimatsu and coworkers (138) isolated an alkaliphilic 59-2 exhibited a broad optimum pH ranging from 6.0
 ALKALIPHILES: ALKALINE ENZYMES AND THEIR APPLICATIONS 227

to 8.0. In the culture broth of B. halodurans No. C-125, compared with known pectate lyases reported so far. The
two xylanases were found (51). Xylanase A had molecular enzyme activity and stability were stimulated by the
weight of 43,000 and xylanase N had a molecular weight addition of NaCl at an optimum of 100 mM. Fogarty and
of 16,000. Xylanase N was most active at pH 6 to Kelly (171,172) then reported that Bacillus sp. No. RK9
7 and xylanase A was most active at a pH range produced endo-polygalacturonate lyase. The optimum pH
of 6 to 10 and had some activity at pH 12. The for the enzyme activity toward acid-soluble pectic acid
xylanase A gene was cloned, sequenced, and expressed in was 10.0. Ito and his colleagues isolated two excellent
E. coli (151–155). Four thermophilic alkaliphilic Bacillus pectate lyase producers, Bacillus sp. strains KSM-P7,
strains (W1 (JCM2888), W2 (JCM2889), W3 and W4) KSM-P15 (173–175). This high-alkaline pectate lyase,
produced xylanases (156,157). The pH optima for enzyme Pel-7 from strain KSM-P7 was purified to homogeneity
action of strains W1 and W3 was 6.0 and that for strains and its molecular weight was about 33,000. The isoelectric
W2 and W4 was between 6 to 7. The enzymes were stable point was close to or higher than pH 10.5. It exhibited
between pH 4.5 and 10.5 at 45 ° C for 1 hour. The optimum optimum activity at pH 10.5 and around 60 to 65 ° C in
temperatures of xylanases of W1 and W3 were 65 ° C and glycine-NaOH buffer. The gene was cloned and sequenced,
those of W2 and W4 were 70 ° C. The degree of hydrolysis and the deduced amino acid sequence of mature Pel-7 (302
of xylan was about 70 % after 24-hour incubation. amino acids) was obtained. The Pel-7 is basically grouped
After the demonstration that alkali-treated woodpulp into the Pel super family, although the enzymatic and
could be biologically bleached by xylanases instead of molecular properties are different. Another high molecular
the usual environmentally damaging chemical process weight enzyme, Pel-15W from strain KSM-P15, had a
using chlorine, the search for thermostable alkaline molecular weight of 70,000, the pI was around pH 4.6.
xylanases has been extensive. Dey (158) isolated an Pel-15 HHrandomly trans-eliminated polygalacturonate
alkaliphilic thermophilic Bacillus sp. (NCIM 59) that in the presence of Ca2+ ions, and the maximum activity
produced two types of cellulase-free xylanase at pH 10 was observed at pH 11.5 and at 55 ° C in glycine-NaOH
and 50 ° C. Khasin and coworkers reported that alkaliphilic buffer. The gene for Pel-15 HHwas cloned and sequenced,
B. stearothermophilus T-6 produced an extracellular and the structural gene contained a 2,031-bp open reading
xylanase that was shown to optimally bleach pulp frame that encoded 677 amino acids including a possible
at pH 9 and 65 ° C (159). Nakamura and coworkers 28-amino acid signal sequence.
also reported that an alkaliphilic Bacillus sp. strain, Several papers on potential applications of alkaline
41M-1, isolated from soil produced multiple xylanases pectinase have been published. The first application
extracellularly (160–162). One of the enzymes, xylanase J, of alkaline pectinase-producing bacteria in the retting
was most active at pH 9.0. The optimum temperature for of Mitsumata bast (Edgeworthia papyrifera, plant for
the activity at pH 9.0 was around 50 ° C. Then, another Japanese paper making) was reported by Yoshihara
alkaliphilic thermophilic Bacillus sp. strain TAR-1 was and Kobayashi (176). The pectic lyase (pH optimum 9.5)
isolated from soil (52). The xylanase of Bacillus sp. produced by an alkaliphilic Bacillus sp. No. GIR 277
strain TAR-1 was most active over a pH range of 5.0 has been used in improving the production of a type
to 9.5 at 50 ° C. Optimum temperatures of the crude of Japanese paper. A new retting process produced a
xylanase preparation were 75 ° C at pH 7.0 and 70 ° C high-quality, nonwoody paper that was stronger than
at pH 9.0. Many thermostable alkaline xylanases have the paper produced by the conventional method. Tanabe
been also produced from various alkaliphiles isolated from and coworkers (177,178) tried to develop a new waste
geothermal areas (163–165). treatment by using an alkaliphilic Bacillus sp. No. GIR
Recently, biobleaching by xylanases of Streptomyces 621-7. Cao and coworkers (179) isolated four alkaliphilic
thermoviolaceus and Staphylococcus sp. SG-13 have been bacteria, NT-2, NT-6, NT-33, and NT-82, producing
reported besides Bacillus sp. Xylanase (166–170). These polygalacturonase and xylanase. They selected high-
xylanases did not act on cellulose, indicating a possible producers of polygalacturonase from mutants resistant
application of the enzyme in biological debleaching to rifampin in alkaliphilic Bacillus sp NT-33. One strain,
processes. NT-33, had an excellent capacity for degumming ramie
fibers.
PECTINASES
CHITINASES
The first study on alkaline endo-polygalacturonate lyase
produced by alkaliphilic Bacillus sp. No. P-4-N was Tsujibo and coworkers (53) isolated chitinases from an
published in 1972 (36). The optimum pH for enzyme alkaliphilic Nocardiopsis albus subsp. prasina OPC-131.
action was 10.0 for pectic acid. Recently, the gene The isolate produced two types of chitinases. The optimum
for this alkaline pectate lyase, Pel-4A of Bacillus sp. pH of chi-A was pH 5.0 and that of chi-B was pH 7.0.
No. P-4-N was cloned, sequenced, and overexpressed in Suresh and Chandrasekaran (180) isolated a chitinolytic
B. subtilis cells (37). The deduced amino acid sequence of fungus, Beauveria bassiana, from a marine sediment.
the mature enzyme (318 amino acids, 34,805 Da) showed The organism was strongly alkaliphilic and produced
moderate homology to those of known pectate lyase in the maximum chitinase at pH 9.20. The NaCl and colloidal
polysaccharide lyase family 1. The purified recombinant chitin requirements varied with the type of moistening
enzyme had an isoelectric point of pH 9.7 and a molecular medium used. The addition of phosphate and yeast extract
mass of 34 kDa and exhibited very high specific activity resulted in the enhancement of chitinase yield. This is the
228 ALKALIPHILES: ALKALINE ENZYMES AND THEIR APPLICATIONS

first report of the production of chitinase from a marine with samples from a number of alkaline environmental
fungus. sources yielded 10 isolates. From this group, selections
Recently, Bhushan and colleagues isolated an alka- were made on the basis of growth at high pH and the
liphilic, chitinase-producing Bacillus sp. BG-11 (54,55). gallium-binding capacity of the siderophores. It was found
The purified chitinase exhibited a broad pH and temper- that some isolates grew well and high concentrations of
ature optima of 7.5 to 9.0 and 45 ° C to 55 ° C, respectively. siderophore were detected, whereas others grew well in
It was stable between pH 6.0 to 9.0 and 50 ° C for more the presence of much lower concentrations of siderophore.
than 2 hours. The pH and thermostability of immobi- The effect of iron, gallium, and aluminum on growth and
lized chitinase were enhanced significantly. The chitinase siderophore production batch culture was investigated for
immobilized on chitosan was stable between pH 5.0 and six isolates. The presence of iron greatly decreased the
10.0, and the half-life of chitosan-immobilized enzyme at siderophore concentration in these cultures, whereas the
70, 80, and 90 ° C was 90, 70, and 60 minutes, respectively. response to added gallium or aluminum was dependent on
No industrial application has been investigated. the isolate.

Cholic Acid Derivatives

METABOLITES PRODUCED BY ALKALIPHILES
Kimura and coworkers isolated an alkaliphilic Bacillus
Antibiotics strain from soil that grew well in media containing cholic
acid (CA) at 5% (w/v) or higher concentrations (188).
The first report was published in 1980 by Sato and cowork-
The 7- and 12-hydroxyl groups of CA were efficiently
ers (181), in which they recorded the isolation of Pae-
converted into keto groups, with the conversion rate
cilomyces lilacinus No. 1,907 from soil using an alkaline
for both hydroxyl groups reaching 100% after 72 hours
medium with a pH of 10.5. New antibiotics were produced
of cultivation. The strain also converted a 3α-hydroxyl
only under alkaline conditions (pH 9 to 10.5). Paecilomyces
group to a keto group, but the conversion rate was about
lilacinus No. 1,907 produced one major product, 1907-II,
5% after 72 hours. The strain neither affected any other
and a minor product, 1907-VIII, which had antibacterial
part of the CA molecule nor oxidized 7β-or 12β-hydroxyl
and antifungal activities. Subsequently, an alkaliphilic
groups. By nitrosoguanidine (NTG) mutagenesis, Kimura
Nocardiopsis dassonvillei that produced phenazine antibi-
and coworkers isolated five mutants that selectively
otics under alkaline culture conditions was isolated (182).
produced the following compounds from CA at high
Bahn and coworkers isolated a novel aldose reductase
yield (closed 100%): Strains M-4 and M-5 produced
inhibitor from alkaliphilic Corynebacterium sp. YUA25-1.
7,12-diketolithocholic acid; strain No. 250 produced 7-
Compound YUA001 was purified from the supernatant of
ketodeoxycholic acid; strains No. 124 and 336 yielded
culture broth by successive silica gel column chromatogra-
ketochenodeoxycholic acid from CA. Furthermore, strains
phy. The compound had no antimicrobial activity against
M-4, M-5, and No. 250 produced only 7-ketolithocholic acid
some gram-positive and gram-negative bacteria, fungi,
from chenodeoxycholic acid.
and yeasts (183).
Since alkaliphiles have been rediscovered, many
Organic Acids
Japanese pharmaceutical companies have tried using
alkaline media to discover new microorganisms producing During the cultivation of alkaliphiles, pH values of the
new antibiotics. Although several have been found and culture media often decrease sharply owing to the produc-
reported, none are as yet commercially available. tion of organic acids, which are produced by growth on
carbohydrates. Paavilainen and coworkers (189) reported
2-Phenylamine comparative studies of organic acids produced by alka-
liphilic bacilli. Four bacilli, Bacillus sp. 38-2 (ATCC21783),
Hamasaki and coworkers found that a large amount of 2-
B. alkalophilus subsp.halodurans (ATCC27557), Bacil-
phenylethylamine was synthesized by cells of alkaliphilic
lus alcalophilus (ATCC27648) and Bacillus sp. 17-1
Bacillus sp. strain YN-2000. This amine was secreted in
(ATCC31007) were cultured in the presence of various 1%
the medium during the cell growth (184).
(w/v) sugars and related compounds such as sugar alco-
hols. All these alkaliphiles produced acetic acid (4.5–5 g/L
Carotenoid of Alkaliphilic Bacillus Strains
at the maximum) while formic acid was produced by only
Aono and Horikoshi reported that alkaliphilic Bacillus one of the strains. In contrast to neutrophilic bacilli, ace-
sp. No. A-40-2, No. 2B-2, No. 8-1, and No. 57-1 produce toin, butanediol, or ethanol were not detected. Moderate
yellow pigments in the cells (185). These are triterpenoid amounts of isobutyric, isovaleric, α-oxoisovaleric, α-oxo-
carotenoids. However, A-59, M-29, and Y-25 white strains β-methylvaleric, α-oxoisocaproic, and phenylacetic acids
when grown in Horikoshi-II medium did not produce were generated with three of the alkaliphiles.
carotenoids. The physiological role of the yellow pigments
was not reported.
CONCLUSION
Siderophores
Since the rediscovery of alkaliphilic bacteria, more than
Gascoyne and coworkers isolated a siderophore-producing 1,200 studies have been published on many aspects of
alkaliphilic bacterium that accumulated iron, gallium, alkaliphiles and alkaliphily. The alkaliphiles are unique
and aluminum (186,187). Enrichment cultures initiated microorganisms with great potential for microbiology
 ALKALIPHILES: ALKALINE ENZYMES AND THEIR APPLICATIONS 229

and biotechnological exploitation. The aspects that have 27. K. Saeki et al., Biochem. Biophys. Res. Commun. 279,
received the most attention in recent years include 313–319 (2000).
(1) extracellular enzymes and their genetic analysis; 28. K. Horikoshi, Agric. Biol. Chem. 35, 1783–1791 (1971).
(2) mechanisms of membrane transport and pH regulation; 29. T. Hayashi, T. Akiba, and K. Horikoshi, Agric. Biol. Chem.
and (3) the taxonomy of alkaliphilic microorganisms. 52, 443–448 (1988).
It is unclear what the next development will be, but 30. T. Hayashi, T. Akiba, and K. Horikoshi, Appl. Microbiol.
it may be the wider application of enzymes. Alkaline Biotechnol. 28, 281–285 (1988).
enzymes should find additional uses in various fields of 31. O. Shirokizawa, T. Akiba, and K. Horikoshi, Agric. Biol.
industry, such as chiral synthesis, biological wood pulping, Chem. 53, 491–495 (1989).
and more sophisticated enzyme detergents. Furthermore, 32. N. Nakamura and K. Horikoshi, Agric. Biol. Chem. 40,
alkaliphiles may be very good general genetic resources 753–757 (1976).
for such applications as signal peptides for secretion and 33. N. Nakamura and K. Horikoshi, Agric. Biol. Chem. 40,
promoters for hyperproduction of enzymes. 935–941 (1976).
34. N. Nakamura and K. Horikoshi, Agric. Biol. Chem. 40,
BIBLIOGRAPHY 1647–1648 (1976).
35. N. Nakamura and K. Horikoshi, Agric. Biol. Chem. 40,
1. K. Horikoshi, T. Akiba, Alkalophilic Microorganisms: A New 1785–1791 (1976).
Microbial World, Springer-Verlag, Gakkai-shuppan center, 36. K. Horikoshi, Agric. Biol. Chem. 362, 85–293 (1972).
Heidelberg, Tokyo, 1982. 37. T. Kobayashi et al., Extremophiles 4, 377–383 (2000).
2. K. Horikoshi, Agric. Biol. Chem. 36, 1407–1414 (1971). 38. N. Nakamura, K. Watanabe, and K. Horikoshi, Biochim.
3. A. Gronstad et al., Extremophiles 2, 447–453 (1998). Biophys. Acta 397, 188–193 (1975).
4. H. Takami et al., Extremophiles 3, 227–233 (1999). 39. T. Kimura and K. Horikoshi, Agric. Biol. Chem. 53,
5. H. Takami et al., Nucleic Acids Res. 28, 4317–4331 (2000). 2963–2968 (1989).
6. M. G. Sturr, T. A. Krulwich, and A. B. Kolsto, Extremophiles 40. T. Kimura and K. Horikoshi, Appl. Microbiol. Biotechnol.
2, 447–453 (1998). 34, 52–56 (1990).
7. H. Takami et al., Extremophiles 3, 29–34 (1999). 41. K. Horikoshi, M. Nakao, Y. Kurono, and N. Saschihara,
8. H. Takami et al., Biosci. Biotechnol. Biochem. 63, 452–455 Can. J. Microbiol. 30, 774–779 (1984).
(1999). 42. N. Sashihara, T. Kudo, and K. Horikoshi, J. Bacteriol. 158,
9. H. Takami, N. Masui, K. Nakasone, and K. Horikoshi, 503–506 (1984).
Biosci. Biotechnol. Biochem. 63, 1134–1137 (1999). 43. F. Fukumori, T. Kudo, and K. Horikoshi, J. Gen. Microbiol.
10. O. Shirokizawa, T. Akiba, and K. Horikoshi, FEMS Micro- 131, 3339–3345 (1985).
biol. Lett. 70, 131–136 (1990). 44. F. Fukumori, T. Kudo, Y. Narahashi, and K. Horikoshi, J.
11. K. Tsuchiya, I. Ikeda, T. Tsuchiya, and T. Kimura, Biosci. Gen. Microbiol. 132, 2329–2335 (1986).
Biotechnol. Biochem. 61, 298–303 (1997). 45. F. Fukumori, T. Kudo, and K. Horikoshi, FEMS Microbiol.
12. R. Aono, M. Ito, and K. Horikoshi, Biochem. J. 285, 99–103 Lett. 40, 311–314 (1987).
(1992). 46. F. Fukumori, K. Ohishi, T. Kudo, and K. Horikoshi, FEMS
13. R. Aono, M. Ito, and K. Horikoshi, J. Gen. Microbiol. 139, Microbiol. Lett. 48, 65–68 (1987).
2739–2744 (1993). 47. F. Fukumori et al., Gene 76, 289–298 (1989).
14. R. Aono, M. Ito, K. N. Joblin, and K. Horikoshi, Microbiology 48. N. Watanabe, Y. Ota, Y. Minoda, and K. Yamada, Agric.
141, 2955–2964 (1995). Biol. Chem. 41, 1353–1358 (1977).
15. R. Aono, M. Ito, and T. Machida, J. Bacteriol. 181, 49. Y. X. Wang, K. C. Srivastava, G. J. Shen, and H. Y. Wang,
6600–6606 (1999). J. Ferment. Bioeng. 79, 433–438 (1995).
16. T. Kudo, M. Hino, M. Kitada, and K. Horikoshi, J. Bacteriol. 50. K. Horikoshi and Y. Atsukawa, Agric. Biol. Chem. 37,
172, 7282–7283 (1990). 2097–2103 (1973).
17. T. Hamamoto et al., Mol. Microbiol. 14, 939–946, (1994). 51. H. Honda, T. Kudo, Y. Ikura, and K. Horikoshi, Can. J.
18. M. Hashimoto et al., Biosci. Biotechnol. Biochem. 58, Microbiol. 31, 538–542 (1985).
2090–2092 (1994). 52. H. Takahashi, R. Nakai, and S. Nakamura, Biosci. Biotech-
19. M. Kitada, M. Hashimoto, T. Kudo, and K. Horikoshi, J. nol. Biochem. 64, 887–890 (2000).
Bacteriol. 176, 6464–6469, (1994). 53. H. Tsujibo et al., Biosci. Biotechnol. Biochem. 56, 1304–
20. Y. Seto et al., Biosci. Biotechnol. Biochem. 59, 1364–1366, 1305 (1992).
(1995). 54. B. Bhushan and G. S. Hoondal, Biotechnol. Lett. 20,
21. S. Kosono, S. Morotomi, M. Kitada, and T. Kudo, Biochim. 157–159 (1998).
Biophys. Acta Bioenerg. 1,409, 171–175 (1999). 55. B. Bhushan, J. Appl. Microbiol. 88, 800–808 (2000).
22. M. Kitada, S. Kosono, and T. Kudo, Extremophiles 4, 56. T. Akino, N. Nakamura, and K. Horikoshi, Appl. Microbiol.
253–258 (2000). Biotechnol. 26, 323–327 (1987).
23. R. Aono et al., Nucleic Acids Symp. Ser. 29, 139–140 (1993). 57. T. Akino, N. Nakamura, and K. Horikoshi, Agric. Biol.
24. K. Horikoshi, Microbiol. Mol. Biol. Rev. 63, 735–750 (1999). Chem. 52, 773–779 (1988).
25. H. Takami et al., Biosci. Biotechnol. Biochem. 56, 1455– 58. T. Akino, C. Kato, and K. Horikoshi, Arch. Microbiol. 152,
1460 (1992). 10–15 (1989).
26. H. Takami, T. Akiba, and K. Horikoshi, Appl. Microbiol. 59. K. Horikoshi and Y. Atsukawa, Agric. Biol. Chem. 37,
Biotechnol. 30, 120–124 (1989). 1449–1456 (1973).
230 ALKALIPHILES: ALKALINE ENZYMES AND THEIR APPLICATIONS

60. Y. Nogi and K. Horikoshi, Appl. Microbiol. Biotechnol. 34, 92. M. A. McTigue, C. T. Kelly, E. M. Doyle, and W. M. Fogarty,
704–707 (1990). Enzyme Microb. Technol. 17, 570–573 (1995).
61. K. Aunstrup, H. Ottrup, O. Andresen, and C. Dambmann, 93. C. T. Kelly, M. A. McTigue, E. M. Doyle, and W. M. Fogarty,
eds., 4th International Fermentation Symposium, Society of J. Ind. Microbiol. 15, 446–448 (1995).
Fermentation Technology, Kyoto, Japan, 1972, pp. 299–305. 94. L. L. Lin, C. C. Chyau, and W. H. Hsu, Appl. Biochem.
62. M. Nomoto et al., Agric. Biol. Chem. 48, 1627–1628, (1984). Biotechnol. 28, 61–68 (1998).
63. Y. Tsai, M. Yamasaki, and G. Tamura, Biochem. Int. 8, 95. K. Igarashi et al., Appl. Environ. Microbiol. 64, 3282–3289
283–288, (1984). (1998).
64. Y. Tsai et al., Biochim. Biophys. Acta 883, 439–447 (1986). 96. K. Igarashi et al., Biochem. Biophys. Res. Commun. 248,
65. O. Tsuchida et al., Curr. Microbiol. 14, 7–12 (1986). 372–377 (1998).
66. N. Fujiwara and K. Yamamoto, J. Ferment. Technol. 65, 97. K. Igarashi et al., Biosci. Biotechnol. Biochem. 63, 1535–
531–534 (1987). 1540 (1999).
67. N. Fujiwara, A. Masui, and T. Imanaka, J. Biotechnol. 30, 98. R. Villalonga, L. Gomez, H. L. Ramirez, and M. L. Villa-
245–256 (1993). longa, J. Chem. Technol. Biotechnol. 74, 635–638 (1999).
68. X. Q. Han and S. Damodaran, J. Agric. Food Chem. 46, 99. T. Kimura and K. Horikoshi, Agric. Biol. Chem. 53,
3596–3603 (1998). 2019–2020 (1989).
69. C. G. Kumar, M. P. Tiwari, and K. D. Jany, Process Bio- 100. T. Kimura and K. Horikoshi, FEMS Microbiol. Lett. 71,
chem. 34, 441–449 (1999). 35–42 (1990).
70. J. Singh, R. M. Vohra, and D. K. Sahoo, Biotechnol. Lett. 21, 101. T. Kimura and K. Horikoshi, Starch/Starke 42, 403–407
921–924 (1999). (1990).
71. P. L. Jorgensen et al., Appl. Environ. Microbiol. 66, 825–827 102. T. Kimura and K. Horikoshi, J. Ferment. Bioeng. 70,
(2000). 134–135 (1990).
72. H. Takami, T. Akiba, and K. Horikoshi, Appl. Microbiol. 103. H. K. Na et al., Mol. Cells 6, 203–208 (1996).
Biotechnol. 33, 519–523 (1990). 104. T. Kobayashi et al., J. Bacteriol. 174, 3439–3444 (1992).
73. H. Takami, T. Akiba, and K. Horiakoshi, Biosci. Biotechnol. 105. M. Matsuzawa, M. Kawano, N. Nakamura, and K. Hori-
Biochem. 56, 333–334 (1992). koshi, Die Staerke 27, 410–413 (1975).
74. H. Takami, T. Akiba, and K. Horiakoshi, Biosci. Biotechnol. 106. G. Georganta et al., Starch 45, 95–99 (1993).
Biochem. 56, 510–511 (1992).
107. V. A. Abelyan, T. Yamamoto, and E. G. Afrikyan, Bio-
75. H. Takami, S. Nakamura, R. Aono, and K. Horikoshi, Biosci. chemistry-Engl. Tr. 59, 573–579 (1994).
Biotechnol. Biochem. 56, 1667–1669 (1992).
108. T. D. G. Salva, V. B. deLima, and A. P. Pagan, Rev. Micro-
76. H. Takami, T. Kobayashi, R. Aono, and K. Horikoshi, Appl.
biol. 28, 157–164 (1997).
Microbiol. Biotechnol. 38, 101–108 (1992).
109. H. J. Chung et al., J. Agric. Food Chem. 46, 952–959 (1998).
77. S. W. Cheng et al., Biosci. Biotechnol. Biochem. 59,
2239–2243 (1995). 110. Y. Terada et al., J. Biol. Chem. 272, 15729–15733 (1997).
78. T. I. Zaghloul, M. AlBahra, and H. AlAzmeh, Appl. Biochem. 111. G. Parsiegla, A. K. Schmidt, and G. E. Schulz, Eur. J.
Biotechnol. 70–2, 207–213 (1998). Biochem. 255, 710–717 (1998).
79. K. Tsuchiya, H. Sakashita, Y. Nakamura, and T. Kimura, 112. K. Kimura, T. Takano, and K. Yamane, Appl. Microbiol.
Agric. Biol. Chem. 55, 3125–3127 (1991). Biotechnol. 26, 147–153 (1987).
80. K. Tsuchiya, Y. Nakamura, H. Sakashita, and T. Kimura, 113. K. Kimura et al., Appl. Microbiol. Biotechnol. 27, 372–377
Biosci. Biotechnol. Biochem. 56, 246–250 (1992). (1988).
81. D. Y. Yum et al., Biosci. Biotechnol. Biochem. 58, 470–474 114. A. Nakamura, K. Haga, and K. Yamane, Biochemistry 32,
(1994). 6624–6631 (1993).
82. B. Bockle, B. Galunsky, and R. Muller, Appl. Environ. 115. A. Nakamura, K. Haga, and K. Yamane, Biochemistry 33,
Microbiol. 61, 3705–3710 (1995). 9929–9936 (1994).
83. C. A. Studdert, R. E. DeCastro, K. H. Seitz, and J. J. San- 116. N. Ishii, K. Haga, K. Yamane, and K. Harata, J. Biochem.
chez, Arch. Microbiol. 168, 532–535 (1997). (Tokyo) 127, 383–391 (2000).
84. H. StanLotter et al., Extremophiles 3, 153–161 (1999). 117. N. Ishii, K. Haga, K. Yamane, and K. Harata, J. Mol.
85. M. I. Gimenez, C. A. Studdert, J. J. Sanchez, and R. E. De- Recognit. 13, 35–43 (2000).
Castro, Extremophiles 4, 181–188 (2000). 118. T. J. Kim, Y. D. Lee, and H. S. Kim, Biotechnol. Bioeng. 41,
86. N. Fujiwara, K. Yamamoto, and A. Masui, J. Ferment. 88–94 (1993).
Bioeng. 72, 306–308 (1991). 119. T. Okada, M. Ito, and K. Hibino, J. Ferment. Bioeng. 77,
87. A. Masui, N. Fujiwara, M. Takagi, and T. Imanaka, Biotech- 259–263 (1994).
nol. Tech. 13, 813–815 (1999). 120. T. E. Abraham, Biocatal. Biotransform. 12, 137–146 (1995).
88. E. W. Boyer and M. B. Ingle, J. Bacteriol. 110, 992–1000 121. H. O. S. Lima, F. F. DeMoraes, and M. Zanin, Appl. Bio-
(1972). chem. Biotechnol. 70–2, 789–804 (1998).
89. E. W. Boyer, M. B. Ingle, and G. D. Mercer, Int. J. Syst. 122. C. T. Kelly, F. O’ Reilly, and W. M. Fogarty, FEMS Micro-
Bacteriol. 23, 238–242 (1973). biol. Lett. 20, 55–59 (1983).
90. T. U. Kim et al., Appl. Environ. Microbiol. 61, 3105–3112 123. C. H. Kim, H. I. Choi, and D. S. Lee, Biosci. Biotechnol.
(1995). Biochem. 57, 1632–1637 (1993).
91. M. A. McTigue, C. T. Kelly, W. M. Fogarty, and E. M. Doyle, 124. C. H. Kim, H. I. Choi, and D. S. Lee, J. Ind. Microbiol. 12,
Biotechnol. Lett. 16, 569–574 (1994). 48–57 (1993).
 ALKALIPHILES: ALKALINE ENZYMES AND THEIR APPLICATIONS 231

125. K. Igarashi et al., Biosci. Biotechnol. Biochem. 56, 514–516 159. A. Khasin, I. Alchanati, and Y. Shoham, Appl. Environ.
(1992). Microbiol. 59, 1725–1730 (1993).
126. K. Ara, K. Igarashi, K. Saeki, and S. Ito, Biosci. Biotechnol. 160. S. Nakamura et al., Appl. Environ. Microbiol. 59, 2311–
Biochem. 59, 662–666 (1995). 2316 (1993).
127. K. Ara et al., Biosci. Biotechnol. Biochem. 60, 634–639 161. S. Nakamura et al., World J. Microbiol. Biotechnol. 9,
(1996). 221–224 (1993).
128. Y. Hatada et al., J. Biol. Chem. 271, 24075–24083 (1996). 162. S. Nakamura et al., Nucleic Acids Symp. Ser. 34, 99–100
129. M. Takagi, S. P. Lee, and T. Imanaka, J. Ferment. Bioeng. (1995).
81, 557–559 (1996). 163. P. L. Dimitrov, M. S. Kambourova, R. D. Mandeva, and
130. T. Kaneko et al., J. Gen. Microbiol. 135, 3447–3457 (1989). E. I. Emanuilova, FEMS Microbiol. Lett. 157, 27–30 (1997).
131. K. Kaneko, T. Kudo, G. Georganta, and K. Horikoshi, in 164. A. Sunna, S. G. Prowe, T. Stoffregen, and G. Antranikian,
T. Hoshi, ed., The Fifth International Symposium of FEMS Microbiol. Lett. 148, 209–216 (1997).
Cyclodextrins, Minutes, Paris, France, 1990, p. 19. 165. C. Lopez, A. Blanco, and F. I. J. Pastor, Biotechnol. Lett. 20,
132. A. Nakamura et al., J. Biol. Chem. 266, 1579–1583 (1991). 243–246 (1998).
133. L. L. Lin, M. R. Tsau, and W. S. Chu, Appl. Microbiol. 166. A. P. Garg, J. C. Roberts, and A. J. McCarthy, Enzyme
Biotechnol. 42, 51–56 (1994). Microb. Technol. 22, 594–598 (1998).
134. J. -S. Park, J. Hitomi, S. Horinouchi, and T. Beppu, Protein 167. S. Gupta, B. Bhushan, and G. S. Hoondal, World J. Micro-
Eng. 6, 921–926 (1993). biol. Biotechnol. 15, 511–512 (1999).
135. J. -S. Park, S. Horinouchi, and T. Beppu, Agric. Biol. Chem. 168. S. Gupta, B. Bhushan, and G. S. Hoondal, J. Appl. Micro-
55, 1745–1750 (1991). biol. 88, 325–334 (2000).
136. H. G. Damude et al., Gene 123, 105–107 (1993). 169. S. Gupta, R. C. Kuhad, B. Bhushan, and G. S. Hoondal,
137. R. Dasilva, D. K. Yim, E. R. Asquieri, and Y. K. Park, Rev. World J. Microbiol. Biotechnol. 17, 5–8 (2001).
Microbiol. 24, 269–274 (1993). 170. L. S. Zheng, Y. M. Du, and J. Y. Zhang, Bioresour. Technol.
138. T. Yoshimatsu et al., J. Gen. Microbiol. 136, 1973–1979 78, 89–94 (2001).
(1990).
171. W. M. Forgarty and P. O. Ward, in D. J. D. Hockenhull, ed.,
139. S. Shikata et al., Agric. Biol. Chem. 54, 91–96 (1990). Progress in Industrial Microbiology, Churchill Livingstone,
140. N. Sumitomo, K. Ozaki, S. Kawai, and S. Ito, Biosci. Biotech- Edinburgh, London, 1977, pp. 59–119.
nol. Biochem. 56, 872–877 (1992). 172. C. T. Kelly, W. M. Fogarty, Can. J. Microbiol. 24, 1164–
141. N. Sumitomo et al., Biosci. Biotechnol. Biochem. 59, 1172 (1978).
2172–2175 (1995). 173. T. Kobayashi et al., Biosci. Biotechnol. Biochem. 63, 65–72
142. Y. Hakamada et al., Extremophiles 1, 151–156 (1997). (1999).
143. K. Hayashi et al., Seibutsu-Kogaku Kaishi 74, A7–A10 174. T. Kobayashi et al., Biochim. Biophys. Acta Gen. Subjects
(1996). 1,427, 145–154 (1999).
144. S. Ito, Extremophiles 1, 61–66 (1997). 175. A. Ogawa et al., Biosci. Biotechnol. Biochem. 64, 1133–
145. H. Khyamihorani, World J. Microbiol. Biotechnol. 12, 1141 (2000).
525–529 (1996). 176. K. Yoshihara and Y. Kobayashi, Agric. Biol. Chem. 46,
146. M. Miyatake and K. Imada, Biosci. Biotechnol. Biochem. 61, 109–117 (1982).
362–364 (1997).
177. H. Tanabe et al., J. Ferment. Technol. 65, 243–246 (1987).
147. J. Sancheztorres, P. Perez, and R. I. Santamaria, Appl.
178. H. Tanabe, Y. Kobayashi, and I. Akamatsu, Agric. Biol.
Microbiol. Biotechnol. 46, 149–155 (1996).
Chem. 52, 1855–1856 (1988).
148. S. Ito et al., Extremophiles 2, 185–190 (1998).
179. J. W. Cao, W. H. Sun, Y. Pan, and S. Y. Chen, Enzyme
149. B. Bhushan, N. S. Dosanjh, K. Kumar, and G. S. Hoondal, Microb. Technol. 27, 545–548 (2000).
Biotechnol. Lett. 16, 841–842 (1994).
180. P. V. Suresh and M. Chandrasekaran, Process Biochem. 34,
150. B. Bhushan and G. S. Hoondal, Biotechnol. Lett. 16, 257–267 (1999).
837–840 (1994).
181. M. Sato, T. Beppu, and K. Arima, Agric. Biol. Chem. 44,
151. H. Honda, T. Kudo, and K. Horikoshi, Agric. Biol. Chem. 49,
3037–3040 (1980).
3165–3169 (1985).
182. H. Tsujibo et al., Agric. Biol. Chem. 52, 301–306 (1988).
152. H. Honda, T. Kudo, and K. Horikoshi, Agric. Biol. Chem. 49,
3011–3015 (1985). 183. Y. S. Bahn et al., J. Antibiot. 51, 902–907 (1998).
153. H. Honda, T. Kudo, and K. Horikoshi, J. Bacteriol. 161, 184. N. Hamasaki et al., Appl. Environ. Microbiol. 59, 2720–
784–785 (1985). 2722 (1993).
154. H. Honda, T. Kudo, and K. Horikoshi, Syst. Appl. Microbiol. 185. R. Aono and K. Horikoshi, Agric. Biol. Chem. 55, 2643–
8, 152–157 (1986). 2645 (1991).
155. H. Honda, T. Kudo, and K. Horikoshi, J. Ferment. Technol. 186. D. J. Gascoyne, J. A. Connor, and A. T. Bull, Appl. Micro-
64, 373–377 (1986). biol. Biotechnol. 36, 130–135 (1991).
156. W. Okazaki, T. Akiba, K. Horikoshi, and R. Akahoshi, Appl. 187. D. J. Gascoyne, J. A. Connor, and A. T. Bull, Appl. Micro-
Microb. Biotechnol. 19, 335–340 (1984). biol. Biotechnol. 36, 136–141 (1991).
157. W. Okazaki, T. Akiba, K. Horikoshi, and R. Akahoshi, Agric. 188. H. Kimura, A. Okamura, and H. Kawaide, Biosci. Biotech-
Biol. Chem. 49, 2033–2039 (1985). nol. Biochem. 58, 1002–1006 (1994).
158. D. Dey, J. Hinge, A. Shendye, and M. Rao, Can. J. Microbiol. 189. S. Paavilainen, P. Helisto, and T. Korpela, J. Ferment.
38, 436–442 (1992). Bioeng. 78, 217–222 (1994).
232 ANAEROBIC GRANULES AND GRANULATION PROCESSES

ALKALITHERMOPHILES (ANAEROBIC). to allow operation at high loading rates. Recently, the

See THERMOPHILES: ANAEROBIC ALKALITHERMOPHILES formation of granular sludge has also been observed in a
few other types of reactors, in particular in the anaerobic
migrating bed reactor (AMBR) (6) and the anaerobic
sequencing bed reactor (ASBR) (7).
Various aspects of granular sludge characteristics and
ALLERGENS, FUNGAL. See FUNGAL ALLERGY AND granulation processes are reviewed in this article. We start
ALLERGENS with a review of studies that determined the abundance of
various microbial populations in anaerobic granules. The
research reviewed in this first section used techniques
that did not allow preservation of granular structure.
In the second part, we summarize the results of studies
AMMONIFICATION. See SOIL NITROGEN CYCLE on techniques that preserved granular structure. In a
third section, we review research that studied granulation
processes with the general objective to minimize the start-
up period for granule formation. Finally, the recommended
start-up procedures for UASB reactors are summarized.
AMOEBOID PROTOZOA. See PROTOZOA IN MARINE
AND ESTUARINE WATERS
MICROBIAL COMPOSITION OF ANAEROBIC GRANULES

Various anaerobic microorganisms exhibit wide-ranging

abilities to utilize different substrates and have vari-
ANAEROBIC DIGESTION OF BIOSOLIDS. able affinities for similar substrates. Furthermore, growth
See BIOSOLIDS: ANAEROBIC DIGESTION OF and death rates, and requirements for nutrients and
microenvironments vary greatly for different anaerobes.
Therefore, it can be expected that wastewater character-
istics, reactor configuration, and reactor operating con-
ditions determine the microbial composition of anaerobic
ANAEROBIC GRANULES. See ANAEROBIC GRANULES granules. However, as discussed later, the microbial com-
AND GRANULATION PROCESSES position of granules obtained from widely varying systems
is often quite similar, suggesting key roles for certain
populations present in granules.
Wu and coworkers (8) used enrichment methods and
microscopy to identify microorganisms in granules from
ANAEROBIC GRANULES AND GRANULATION a UASB reactor treating brewery wastewater (with
PROCESSES 0.6 to 1.2 mM sulfate present). It was found that
Methanobacterium-like cells were the prevalent H2 -CO2
DANDAN ZHENG and formate utilizers, whereas Methanospirillum-like cells
LUTGARDE RASKIN were occasionally found. Methanosaeta-like cells were
University of Illinois the most abundant aceticlastic methanogens. In addition,
Urbana, Illinois
Desulfovibrio-like and Desulfobulbus-like cells were found
in enrichments prepared with ethanol and propionate
In the 1970s, Lettinga and his coworkers developed (with or without sulfate), respectively. In a subsequent
the upflow anaerobic sludge blanket (UASB) reactor study, enrichments were obtained from granules fed a
to allow anaerobic treatment of a variety of high- medium containing acetate, propionate, and butyrate
strength wastewaters at relative high loading rates (1). (with 0.15 mM sulfate) (9). Using most probable number
An important feature of the UASB system is the ability (MPN) estimates combined with microscopy, the authors
to retain a high biomass concentration, which makes found that fermentative Clostridium-like spore-forming
it possible to operate at high loading rates. Since the rods were present at 109 cells/g suspended solids (SS), that
first development of the UASB system, a variety of syntrophic propionate-degrading bacteria were detected
modifications of this process have been developed, but all at 1010 to 1011 cells/g SS, and that syntrophic butyrate-
consist of an upflow-type reactor. Examples of modified degrading bacteria were found to be present at 1011 cells/g
UASB systems include the expanded granular sludge SS. The observed hydrogenotrophic methanogens included
blanket (EGSB) reactor (2), compartmentalized UASB Methanobacterium formicicum-like cells (1012 cells/g SS)
reactors (3), and upflow sludge bed filters (USBF) (4). The and Methanospirillum-like cells (109 cells/g SS). The
biomass in UASB and related reactors typically develops aceticlastic methanogens Methanosarcina mazei-like and
into separate spherical entities of a few millimeters in Methanosaeta-like cells were observed at levels of
diameter, which are called anaerobic granules. Therefore, 108 cells/g SS and 1012 cells/g SS, respectively.
the biomass as a whole is referred to as granular sludge. These two studies demonstrated that the most
Although the formation of granules is not necessary for the abundant aceticlastic and hydrogenotrophic methanogens
operation of a UASB reactor (5), granulation is desirable were similar in granules from UASB reactors fed different
 ANAEROBIC GRANULES AND GRANULATION PROCESSES 233

wastewaters. Methanosaeta spp. and Methanobacterium decreased to lower levels, which persisted throughout the
spp. dominated granules from both systems. It was also observed period. The populations of methanogens related
observed that Methanobacterium formicicum-like cells to M. concilii Opfikon and M. arboriphilus AZ increased,
were dominant in the MPN test using acetate (9). This and reached a stable level after 21 and 76 days, respec-
might indicate that a significant amount of acetate was tively. The population of Methanobacterium thermoau-
first converted to CO2 and H2 by acetogens, which totrophicum
H related organisms continued to increase
was then converted to methane by M. formicicum-like over the observed period.
cells. So far, only two strains have been isolated that Several research groups have used oligonucleotide
can convert acetate to CO2 and H2 , the thermophilic hybridization probes targeting ribosomal RNA (rRNA)
strain AOR (acetate oxidizing rod) and Clostridium to determine the microbial composition of anaerobic
ultunense (10). Other researchers also found that the granules. Zheng (16) applied 16S rRNA targeted oligonu-
indirect conversion of acetate to methane by acetogens cleotide probes to analyze the microbial composition of
and hydrogenotrophic methanogens can be significant in granules sampled at different heights from a UASB reac-
granular sludge (11–13). tor treating wastewater from a corn processing plant,
Several methods were used to reveal the microbial which produced ethanol as its major product. The lev-
composition of granules obtained from a UASB reactor els of archaeal (methanogenic) rRNA ranged from 18.4 to
treating sugar refinery wastewater (original granules) and 30.6%, with 4 to 5.9% of Methanosaeta spp., 3.4 to 4.3%
after they were adapted to ethanol and propionate for 6 Methanomicrobiales, and 10.5 to 13.6% Methanobacteri-
and 36 months, respectively (14). Direct counting based aceae. The predominant sulfate-reducing bacteria (SRB)
on morphology and autofluorescence showed that 20% of were Desulfovibrionaceae (28.6 to 53.3%), but Desulfobul-
the total cells in the original granules were Methanosaeta- bus spp. (2.7 to 6.2%), and Desulfobacter spp. (1.3 to 2.6%)
like organisms, 10% were Methanosarcina-like, and 15% were also present.
were hydrogenotrophic methanogens. After six months The same method was used to show the population
of adaptation to ethanol, Methanosaeta-like cells made changes after granules formed in potato-processing
up less than 1% of the community, Methanosarcina-like industry wastewater adapted to different substrates (17).
cells increased to 20%, and hydrogenotrophic methanogens The inoculum had very low levels of Desulfobulbus spp.
decreased to around 10%. On the other hand, granules (around 2% of total 16S rRNA) and of the syntrophic
adapted to propionate for 36 months contained 30% of propionate degrader, strain SYN7 (less than 1% of total
Methanosaeta-like cells, less than 1% Methanosarcina- 16S rRNA). After adaptation, the Desulfobulbus spp.
like cells, and 40% hydrogenotrophic methanogens. increased significantly (up to 35%) in the granules fed
MPN methods were used to estimate the abundance propionate and sulfate, whereas SYN7 increased to around
of populations in these three sets of granules using 10% in the granules fed propionate only.
various enrichment media and antibodies were used to Granules from two laboratory-scale UASB reactors also
identify the cells present in the positive highest dilution were characterized using 16S rRNA targeted oligonu-
tubes of the MPN test. Both Methanosarcina barkeri cleotide probes (18,19). The feed for one of the UASB reac-
and Methanosaeta concilii were detected in the original tors contained glucose (Reactor NP), whereas the second
granules, but only M. barkeri was present in the ethanol- reactor was fed glucose and propionate (Reactor P). Despite
adapted granules and M. concilii in the propionate- these differences in feed composition, the methanogenic
adapted granules. Similarly, both Methanospirillum composition of granules from the two reactors were very
hungatei JF1 and Methanobrevibacter arboriphilus AZ similar. Granules from both reactors contained around
were found in the original granules, but only M. hungatei 40% archaeal (methanogenic) rRNA. Methanosaeta con-
JF1 was present in the ethanol-adapted granules and cilii was the predominant methanogen in both reactors.
M. arboriphilus AZ in the propionate-adapted granules. Species within the family Methanobacteriaceae were the
Visser and coworkers (15) used monoclonal antibodies dominant hydrogenotrophic methanogens in Reactor NP,
to show changes in the microbial composition of gran- whereas species belonging to the Methanobacteriaceae and
ules, which were obtained from a mesophilic UASB reactor Methanomicrobiales were equally abundant in Reactor P.
treating a potato-processing wastewater and were exposed Sekiguchi and coworkers (20) evaluated the diver-
for 20 days to a medium containing acetate, propionate, sity of mesophilic and thermophilic granular sludge
and butyrate. The authors found no changes in population that had been maintained in UASB reactors treating a
patterns (in terms of the degree of antigenic relatedness sucrose/propionate/acetate-based artificial wastewater by
to different methanogens) after 20 days of adaptation to constructing 16S ribosomal DNA (rDNA) clone libraries
the VFA-based medium. However, the total number of using prokaryotic primers. Partial sequencing of the clones
methanogens detected dropped from 12% in the origi- showed that about 20% of the clones were Archaea, all
nal granules to 5%. The authors documented changes in of which were close relatives of known methanogens
the population profiles for different groups of immunolog- and most of them were affiliated with M. concilii and
ically related methanogens after the reactor was oper- M. thermophila for the mesophilic and thermophilic gran-
ated at thermophilic conditions for four months. The ules, respectively. An important difference between the
abundance of Methanobrevibacter smithii PS, M. smithii two clone libraries was that a major group in the
ALI, M. hungatei JF1, Methanogenium cariaci JR1, and mesophilic clone library (27% of clones) belonged to the
Methanosarcina thermophila TM1 related cells increased delta subclass of the Proteobacteria and contained close rel-
immediately after the rise in temperature, and gradually atives of SRB (Desulfovibrio spp. and Desulfobulbus spp.)
234 ANAEROBIC GRANULES AND GRANULATION PROCESSES

and syntrophic bacteria (strain SYN7 and Syntrophobac- potential to reveal microbial diversity, but can be biased
ter fumaroxidans [strain MPOB]). In contrast, none of the when DNA is not uniformly retrieved.
clones in the thermophilic clone library was affiliated with
the Proteobacteria, but clones that were closely related STRUCTURE OF GRANULAR SLUDGE
to gram-positive syntrophic bacteria (Thermosyntropha
lipolytica and propionate-oxidizing spore formers A and Layered Structure
B) were identified. Using fluorescence in situ hybridiza-
tions (FISH), they further determined that approximately To study the structure of granules, MacLeod and
40% and 45% of total cells (as determined using the inter- coworkers (21) used mesophilic granular sludge that had
calating dye 4’,6-diamidino-2-phenylindole [DAPI]) were been fed a sucrose-based wastewater for one month (after
labeled with archaeal and bacterial oligonucleotide probes, transfer from a UASB reactor treating cheese whey
respectively. This discrepancy between FISH and cloning wastewater). Transmission electron microscopy (TEM)
results lead Sekiguchi and coworkers (20) to believe that and scanning electron microscopy (SEM) determined that
their DNA extraction method did not result in sufficient the granules consisted of three layers:
lysis of all cells, which may have limited their estimate of
• Organisms in the external layer were of diverse
diversity.
morphology: chain-forming coccoids, long thin fila-
In summary, the studies reviewed earlier have shown
ments (Methanospirillum-like), small rods and cocci,
that the microbial composition of granules depends on clustering coccoids (Methanococcales-like), and chain-
the type of substrate, and can change substantially forming rods (Methanosaeta-like). Some of the large
when substrates or environmental conditions are changed. coccoids were surrounded by extracellular poly-
However, there are similarities, especially with regard mers (ECP).
to methanogen representation. This suggests that the
• The middle layer was tightly packed with ECP and
presence of certain methanogens is critical for the
rod-shaped bacteria similar to Methanobrevibacter
anaerobic degradation of organic compounds in general
spp. and Syntrophobacter spp.
and for systems depending on granulation in particular.
• Methanosaeta-like organisms were dominant in the
Because different microbial characterization techniques
center. In addition, the center contained many
were used in the various studies, results need to be
cavities, which might help to release produced biogas.
interpreted with care. For example, cultivation-based
methods, such as MPN estimates, may underestimate the On the basis of these observations Guiot and cowork-
abundance of Methanosaeta spp. and syntrophic bacteria ers (22) proposed a model for anaerobic granules formed
as pointed out by Grotenhuis and coworkers (14). When on carbohydrate-based substrates (Fig. 1), consisting of:
using oligonucleotide probes for population quantification, • an external layer, in which fermenters are predomi-
only those organisms that are targeted by the probes nant, but which also contains some hydrogenotrophic
are included in the quantification. Furthermore, rRNA microorganisms;
targeted oligonucleotide probe hybridizations provide • a middle layer that contains syntrophic consortia;
estimates for rRNA levels, and not for biomass levels • a core that contains mainly aceticlastic methanogens,
or cell numbers. Techniques based on cloning have the usually Methanosaeta spp.

CARB.
Hydrogenic acidogens
Sulfate-reducers VFA H2
Methanosarcina spp.
H2-using methanogens
CH3COOH
Hydrogenic acetogens
H2-using methanogens
CH4
CO2
Methanosaeta spp.
transport

Bulk liquid
Mass

Granule
zone

G P G: Active glucose use

Hla Hha P: Active propionate use
Ala Ala Aha Hha: Active H2 use (high affinity)
Glucose
Hla: Active H2 use (low affinity)
Concentration

Aha: Active acetate use (high affinity)

Acetate Ala: Active acetate use (low affinity)
Propionate
Figure 1. Proposed structure of arrange-
H2
ment of populations in glucose-fed gran-
ules. Reproduced with permission from.
S. R. Guiot et al., Water Sci. Technol. 25,
1–10 (1992). Distance
 ANAEROBIC GRANULES AND GRANULATION PROCESSES 235

As discussed by Guiot and coworkers (22), the layered relatively high for all three fractions and the specific
structure develops because of diffusion limitations. activity on acetate was similar for all three fractions.
Because the substrates in the wastewater can only diffuse Thus, each layer was not exclusive of other trophic groups.
for a certain distance into the granules, fermenters stay in Similar layered granular structures with Methanosaeta
the external layer. Volatile fatty acids (VFAs) produced by spp. dominant in the center were also reported for
the fermenters diffuse further into the granules and are mesophilic granules treating sucrose (23), paper factory
utilized by syntrophs. Acetate produced here diffuses into wastewater (24), paper-mill wastewater (25), and sugar
the core and is used by aceticlastic methanogens. Because refinery wastewater (25). These observations were made
H2 is produced in the external layer and the middle using SEM, TEM, phase contrast microscopy, or epifluo-
layer, hydrogenotrophic microorganisms are present in rescence microscopy on sectioned granules.
both layers. Using FISH with rRNA targeted oligonucleotide probes,
The model was supported also by results from metabolic Harmsen and coworkers (17,26) were able to show the
activity tests for granules formed on carbohydrate-based microbial structure of granules in more detail. They
substrates (22). First, the metabolic activity on formate, observed three layers in granules that were originally
acetate, and propionate was greater for larger granules. obtained from a system treating sugar beet wastewater
This observation was explained by the low surface-to- and were adapted to sucrose for six months (Fig. 2; 26):
volume ratio of larger granules and by assuming that
glucose can only diffuse for a certain distance into the • The external layer contained mainly bacterial cells.
granules. Second, metabolic activity tests were performed • The middle layer consisted of syntrophic micro-
on the three fractions that resulted from abrasions of colonies that contained propionate degraders and
anaerobic granules. The external layer had the highest Methanobrevibacter spp. (as suggested by positive
specific activity on glucose, but the lowest specific activity hybridization signals with a probe for the family
on propionate, whereas the core exhibited the opposite Methanobacteriaceae and morphological similarity to
effect. However, the specific activity on glucose was Methanobrevibacter spp.). This layer also contained

(a) (b)

Figure 2. FISH of a sucrose-fed granule viewed by epifluorescence microscopy at a 200-fold

magnification. (a) A fluorescein-labeled bacterial probe detects bacterial cells mainly located in the
outer layer of the granule (P, peripheric location) and the syntrophic microcolonies more toward
the inside of the granule (S, syntrophic microcolony). (b) A rhodamine-labeled probe specific for a
syntrophic population detects only the microcolonies inward of the granule. X, autofluorescence.
Bar, 50 µm. Reproduced with permission from H. J. M. Harmsen et al., Appl. Environ. Microbiol.
62, 1,656–1,663 (1996). See color insert.
236 ANAEROBIC GRANULES AND GRANULATION PROCESSES

microcolonies of Methanosaeta spp. (as revealed by propionate degradation caused by the unfavorable thermo-
both TEM and FISH), which were usually located dynamics of propionate conversion. Methanosaeta spp. and
close to the syntrophic microcolonies. Methanosarcina spp., present in the thick layer, removed
• The core contained inorganic material with large cav- the acetate before it reached the syntrophic microcolonies.
ities and some methanogens (positive hybridization Because the feed contained only butyrate, propionate, and
signal with an Archaea-specific probe, but not with a acetate, the authors suggested that the Bacteria in the
Methanosaeta-specific probe). external layer were mainly butyrate degraders.
A similar structure was observed also in granules from
Thus, in contrast with the observation made by
a thermophilic UASB reactor treating alcohol distillery
MacLeod and coworkers (21), Harmsen and coworkers
wastewater using SEM (27). This reactor was inoculated
(26) found that the core did not contain large numbers
with mesophilic granules treating fruit-juice wastewater.
of Methanosaeta spp., but that Methanosaeta spp. formed
The surface of the granules was covered with filamentous
microcolonies near the syntrophic microcolonies in the
middle layer. Granules that were originally obtained from bacteria, presumably fermenters. The second layer was
a system treating sugar beet wastewater and adapted dominated by Methanosaeta spp., whereas the core
for six months to a mixture of butyrate, propionate, and contained Methanobacterium-like cells.
acetate, had an additional thick layer rich in microcolonies The same mesophilic and thermophilic granules that
of Methanosaeta spp. and Methanosarcina spp. between were characterized in terms of diversity by constructing
the external layer and the middle layer (as detected by clone libraries by Sekiguchi and coworkers (20) (see pre-
FISH and based on the morphology of Methanosarcina vious discussion) were recently analyzed using FISH (28).
spp.) (26). The authors explained the presence of this Granules from both reactors consisted of outer and inner
extra layer by the high acetate concentration in the feed. layers dominated by Bacteria and Archaea, respectively
High concentrations of acetate are inhibitory to syntrophic (Fig. 3). The center did not show hybridization signals with

(a) (b)

(c) (d)

Figure 3. FISH of sections from mesophilic (a and b) and thermophilic (c and d) granules
viewed by confocal laser scanning microscopy. The sections were simultaneously hybridized with
a Cy-5-labeled bacterial probe (green) and a rhodamine-labeled archaeal probe (red). Reproduced
with permission from Y. Sekiguchi et al., Appl. Environ. Microbiol. 65, 1,280–1,288 (1999). See
color insert.
 ANAEROBIC GRANULES AND GRANULATION PROCESSES 237

a bacterial probe or with an archaeal probe. Methanosaeta 5 ° C, but not at 15 ° C and 25 ° C (7). As discussed by the
spp. were the predominant Archaea in both types of authors, the nonlayered structure at 15 ° C and 25 ° C was
granules. In the mesophilic granules, Methanobacteri- probably related to the low hydrolysis rate of the substrate.
aceae cells juxtaposed with Bacteria and some Metha- At 5 ° C, the diffusion rate was also low, possibly allowing
nomicrobiales cells were distributed in the granules. the fermenting bacteria to degrade the substrate before it
In the thermophilic granules, Methanobacteriaceae and diffused into the granules.
Methanosarcina cells were detected. The authors also tried A nonlayered structure also can develop when the
to locate the microorganisms that were found to be signif- substrate can be fully degraded by a single population. For
icant in their earlier study using clone libraries (20,28). example, granules obtained from a methanol-fed UASB
For mesophilic granules, Desulfobulbus cells were found reactor were dominated by Methanosarcina with small
in the outer layer of the granules. Cells that were closely numbers of rods and filaments in the center (31).
related to Syntrophobacter were shown to form micro-
colonies together with Methanobacteriaceae cells in the Microcolonies
mesophilic granules. FISH with a probe that targets clones
Microcolonies often have been observed in granules. Wu
associated with green nonsulfur bacteria revealed filamen-
and coworkers (8) observed microcolonies in mesophilic
tous cells on the surface of the thermophilic granules.
granules from a full-scale UASB reactor treating brewery
In another recent study, FISH revealed a layered struc-
wastewater. The wastewater contained mainly ethanol,
ture for granules used to treat a methanol-containing
some propionate and acetate, sulfate, and trace amounts
wastewater (29). Methanosarcina barkeri were found in
of glucose, butyrate, lactate, and succinate. Three types of
the outer layer, and M. concilii cells were located in
microcolonies were found:
the inner layer and core. Layered structures were also
observed using FISH for granules from two lab-scale • type 1: Methanosaeta-like cells mixed with small
UASB reactors treating glucose- and glucose/propionate- amount of Methanobacterium-like cells;
containing wastewaters (18,19). In both reactors, granules
• type 2: Methanobacterium-like cells juxtaposed with
contained large cores that consisted almost exclusively of
Desulfobulbus-like cells;
M. concilii. Granules from both reactors exhibited a thin
outer layer of Bacteria, most of which had a filamentous • type 3: Methanobacterium-like cells combined with
morphology. For granules treating glucose/propionate- Desulfovibrio-like cells.
containing wastewater, a layer surrounding the core con-
Type 2 and type 3 were believed to be syntrophic
sisted of both Bacteria andArchaea. Finally, Santegoeds
microcolonies, which were responsible for syntrophic
and coworkers (29) used molecular techniques (denaturing
degradation of propionate and ethanol, respectively.
gradient gel electrophoresis and FISH) and microsen-
Two types of microcolonies were found in propionate-fed
sors for hydrogen sulfide and methane to study anaerobic
granules (14). One was solely made up by M. concilii-
granules originally obtained from three full-scale UASB
related organisms, and the other one contained
reactors. They observed a layered structure and found that
Methanobrevibacter arboriphilicus AZ and a possible pro-
the localization of sulfidogenic, methanogenic, and syn-
pionate degrader, which was not closely related to Syntro-
trophic populations determined by FISH and microsensor
phobacter wolinii.
data corresponded well.
Similarly, Harmsen and coworkers (17) found two types
of microcolonies in propionate-fed granules using FISH.
Nonlayered Structure
One microcolony was dominated by Methanosaeta spp.;
Nonlayered granules were observed under a number of the other one was a syntrophic microcolony that contained
conditions. A random distribution of an ethanol-degrader the syntrophic propionate degrader, strain SYN7, and
(strain EE121), M. hungatei JF1, and M. mazei-related a methanogen from the order Methanomicrobiales, mor-
organisms were found in granules adapted to ethanol (14). phologically similar to Methanospirillum spp. Harmsen
The same study also demonstrated a nonlayered structure and coworkers (26) also found two types of microcolonies
for granules fed propionate. These granules were dom- in both sucrose-fed granules and VFA-fed granules (dis-
inated by M. arboriphilus and M. concilii together with cussed earlier), that is, Methanosaeta dominated micro-
a bacterium, presumably a propionate oxidizer. Recently, colonies and syntrophic microcolonies containing syn-
Rocheleau and coworkers (29) found that granules treating trophic propionate degraders and Methanobrevibacter spp.
protein-rich wastewater also exhibited nonlayered struc- Three types of syntrophic microcolonies were
tures in which M. concilii were randomly distributed. found in granules treating carbohydrate-containing
On the basis of observations that granules fed wastewater (32). Two types consisted of acetogen
propionate, peptone, or glutamate exhibited nonlayered cells juxtaposed with hydrogenotrophic methanogens
structures, Fang (30) proposed that the initial degradation (Methanobrevibacter-like), whereas the other type
rate of the substrate determines the granular structure. consisted of clusters of acetogens in close proximity to
If the initial step for degradation of a substrate is the clusters of Methanobrevibacter-like cells.
rate-limiting step, some of the substrate can diffuse into Using wax-embedded sections and histogram stains,
the granules before it is degraded by the microorganisms Morgan and coworkers (25) observed that the diameters of
at the surface, resulting in nonlayered granules. microcolonies increased from the outer layer to the center.
Temperature can also play a role. A layered structure To explain this observation, they suggested that inner
was only observed for granules treating low-fat milk at microcolonies were older than those at the outside.
238 ANAEROBIC GRANULES AND GRANULATION PROCESSES

In summary, studies on granular structure generally In addition, lots of empty sheaths of Methanosaeta, which
have found two types of microcolonies. One type consists are rich in proteins, were observed. This might explain
of syntrophic consortia, and the other type contains why other researchers have found proteins to be dominant
mainly Methanosaeta spp. The mechanism of microcolony in the center of granules (see following section).
formation seems to be obvious. Once a microorganism Quarmby and Forster (37) used gram staining, as well
colonizes, it starts to grow and its offspring stays in the as staining for carbohydrates, proteins, and lipids to show
same area. This results in the formation of a single-species the structure of granules taken from UASB reactors
microcolony. The juxtaposition of syntrophic partners is treating different industrial wastewaters. The surface
probably the result of internal mixing of the offspring (10). contained mainly gram-positive cells with some colonies of
gram-negative cells, whereas the center mainly consisted
Other Studies Related to Granular Structure of gram-negative cells. Carbohydrates were dominant in
the outer layer and decreased inwards, whereas proteins
Distribution Patterns of Methanogens. Macario and
were present at relatively low levels in the outer layer
coworkers (33) observed that thermophilic VFA-fed gran-
and increased inwards. Lipids were evenly distributed
ules consisted of a two-layered structure with an outer
throughout the granules. The structures of granules
layer and a core. Both the outer layer and the core
treating the same types of wastewater were observed to be
had a thin, dense outer layer. In some granules, this
different, demonstrating that other factors such as organic
dense layer separated the core into several parts. Using
loading rate, pH, and the strength of the wastewater may
antibody probes, the authors showed the distribution pat-
also affect the structures.
tern of different groups of methanogens (methanogens
immunologically related to M. thermoautotrophicum
H, In summary, the structure of anaerobic granules
M. thermophila TM1, M. concilii Opfikon, M. arboriphilus can be either layered or nonlayered, depending on
AZ, and M. smithii ALI) located in a spongy matrix formed substrate characteristics. In addition, diffusion factors and
by other organisms or by intercellular material. degradation rates play important roles in determining the
The immobilization patterns of Methanosarcina and structure. Microcolonies are observed in both layered and
Methanosaeta cells in autoclaved granules were studied nonlayered granules. They fall into two groups, one group
using polyclonal antibodies (34). Methanosaeta was immo- consists of juxtaposed syntrophs and hydrogenotrophic
bilized on the surface and in the center of the granules, methanogens; the other one contains mainly Methanosaeta
and all immobilized Methanosaeta cells were single short spp. Methanobrevibacter spp. also are observed frequently
rod-shaped. Methanosarcina, on the other hand, stayed in syntrophic microcolonies.
on the surface or in the cavities. They grew as single
cells, small clusters, or large clusters (especially when the
acetate concentration was high). GRANULATION PROCESSES

Metabolic Activity Zones. Lens and coworkers (35) Because it is generally accepted that granules are
observed metabolic activity zones in granules using pH merely spherical biofilms, Schmidt and Ahring (38) used
and glucose microelectrodes. For example, in granules a modification of biofilm formation theory to explain
(diameter 0.5 to 2.0 mm) from a full-scale UASB reactor that the granulation process consists of four steps:
treating potato starch wastewater that were acclimated (1) transport of cells to the surface of uncolonized inert
to glucose or diluted molasses for about 10 days, they material or other cells (nucleus), (2) initial reversible
identified an outer, glucose-converting zone of 200 to adsorption to the nucleus by physicochemical forces,
300 µm and an inner methanogenic activity zone. The (3) irreversible adhesion of the cells to the nucleus by
authors were not able to correlate these two activity zones microbial appendages and/or polymers attaching the cells
to the granular structure. Methaonsaeta-like cells were to the substratum, and (4) cell growth and development
found throughout the granules, even in the surface layer. of granules. Thus, the granulation process starts with the
In addition, coenzyme F420 was found to be distributed colonization of cells onto a nucleus (inert material or other
homogeneously in the granules. cells). The cells then grow into granules of visible size.

Importance of Nuclei
Hydrophobicity. Three layers of different hydropho-
bicities were observed in gray colored granules fed Nuclei can be particles consisting of inert material and/or
sucrose (36). The outer layer was hydrophilic, the second cells present in the original inoculum, or they can be
layer was very hydrophobic, and the core was moder- particles added to a reactor together with inoculum to
ately hydrophobic. In general, fermenting bacteria are enhance the granulation process. Precipitates, such as
hydrophilic, whereas acetogens and methanogens are iron sulfide can also act as nuclei (39).
hydrophobic. This finding again supported the structured Hulshoff Pol (40) demonstrated that particles with
model proposed by Guiot and coworkers (22). dimensions between 40 and 100 µm are important for
granulation. He showed that granulation still happened
Extracellular Polymers (ECP). ECP were found around after particles larger than 100 µm had been removed by
some clusters of microorganisms in the surface layers of sieving of the inoculum. However, when the inoculum was
granules from two full-scale UASB reactors treating paper sieved using a 40-µm sieve, no granulation was observed.
mill and sugar refinery wastewater (25). There was very Several studies demonstrated that the addition of
little ECP in the area where Methanosaeta was dominant. particles resulted in a higher rate of granulation.
 ANAEROBIC GRANULES AND GRANULATION PROCESSES 239

Addition of 100-µm hydroanthracite particles (with Fermenting Bacteria. On the basis of observations that
good adhesion properties) reduced start-up time and granulation was promoted by wastewater containing car-
increased granulation rate (40). Using a water-absorbing bohydrates but not by preacidified wastewater, Vander-
polymer (WAP, made mainly of an acrylic compound haegen and coworkers (47) suggested that fermenting
with particles between 100 and 200 µm), Imai and bacteria that produce ECP act as ‘‘nucleation centers’’ and
coworkers (41) were able to enhance the start-up process that other microorganisms subsequently associate with
of UASB reactors using anaerobic digester sludge as this nucleus. Filamentous fermenters also were believed
the inoculum. Laboratory-scale reactors fed with glucose to be responsible for promoting granulation in UASB and
showed that the granulation time reduced from 60 days anaerobic fixed film reactors treating partially acidified
to 40 days when WAP was added. For a laboratory- wastewater (48).
scale reactor fed with acetate, propionate, and butyrate,
granulation was observed after 70 days when WAP was Methanosaeta. Wiegant (49) proposed the ‘‘spaghetti
added, and no granulation had taken place in the theory’’ by analyzing available data from other re-
control reactor after four months of operation. In pilot- searchers. This theory states that as a result of the upflow
scale studies with acetate, propionate, and butyrate as velocity in a UASB reactor, Methanosaeta cells are forced
substrates, however, granulation enhancement was less to grow as filaments and/or attached to inert material
significant (65 days with WAP addition versus 75 days in order to be retained in the system. These filaments
without). SEM showed that WAP particles appeared serve as precursors of granulation for ‘‘healthy’’ UASB
in the center of the granules and were gradually reactors. Because granulation usually is much faster
decomposed. than the maximum growth rate of Methanosaeta spp.,
Not all inert particles can enhance granulation. Four Wiegant suggested that aggregation of Methanosaeta with
different carriers, powdered activated carbon (PAC), or without inert material is very important during the
granular activated carbon (GAC), gamet sand, and silica initial stage of start-up. Once the nuclei are formed,
sand were tested for the enhancement of granulation in they grow out to granules by individual growth of
ASBRs treating synthetic wastewater containing mainly trapped microorganisms and entrapment of nonattached
microorganisms.
sucrose (42). All average particle diameters were 0.5 mm.
On the basis of their observation of layered, sucrose-fed
Only the addition of GAC and PAC showed significant
granules, MacLeod and coworkers (21) proposed that the
enhancement of the granulation process, suggesting that
granulation process started with Methanosaeta aggrega-
GAC and PAC provided sites to which microorganisms
tion. Then, acetogens (syntrophs) and hydrogenotrophic
could adhere and thus be retained.
methanogens colonized outside the Methanosaeta aggre-
The addition of 30- to 100-µm gravel particles to
gates. Finally, fermenting bacteria and hydrogenotrophic
digested sewage sludge did not show significant enhance-
methanogens colonized the surface. A similar process was
ment of granulation for a thermophilic UASB reactor fed
also proposed by Morgan and coworkers (21) after they
acetate as the major substrate (43). Methanosaeta was
observed layered granules treating paper mill and sugar
believed to be more important as a nucleus than the inert
refinery wastewater. They proposed that the granula-
material. Attachment of Methanosaeta filaments to the tion process started with a nucleus consisting of aggre-
inert particles was considered a requirement for granula- gates of Methanosaeta spp. and other microorganisms.
tion and the lack of significant enhancement of granulation As Methanosaeta filaments grow, they form bundles. The
in this case was believed to be caused by the low numbers bundle sizes increase and eventually grow into each other
of Methanosaeta in the inoculum. and exclude other microorganisms from the center.
These results are consistent with observation reported
by Yoda and coworkers (44). Powdered zeolite (50 to Methanosarcina. During a granulation experiment with
100 µm) was used in an anaerobic expanded microcarrier defined species (including a butyrate-degrading strain BH,
bed reactor. A thin biofilm developed on the zeolite powder, a propionate-degrading strain PT, Methanosaeta sp. strain
which was later transformed into granules. Methanosaeta M7, M. formicicum T1N, and M. mazei T18), M. mazei
spp. were the predominant organisms in the thin film. T18 was found to be in the center of the granules (50).
Beeftink and Staugaard (45) demonstrated that sand Obviously, the aggregation of M. mazei T18 served as
grains were important during the start-up of an anaerobic the nucleus for the granulation. However, the authors
gas-lift reactor. Later, granules were found to have no sand were not convinced that Methanosarcina was essential for
center and the sand was gradually lost in the system. This granulation and suggested that inert particles may be able
indicates that once granules have been formed, there is no to function in a similar manner.
need for further addition of inert particles to maintain the Some Methanosarcina spp. (usually isolated from
granules. nonmarine environments) can form aggregates up to
The nucleus can also consist of small aggregates several millimeters during their life cycle (51). A pure
of microorganisms. For example, by recycling small culture of M. barkeri formed granules in a UASB
microbial aggregates back to a UASB reactor, Thiele reactor with recycling (52). It is possible that small
and coworkers (46) found that granule bed growth was aggregates of Methanosarcina form and serve as nuclei
stimulated. Microorganisms or groups of microorganisms for granulation because the acetate concentration can
believed to play important roles in the granulation process be sufficiently high (>200 mg/liter) to favor the growth
are discussed below. of Methanosarcina during start-up (10). Later, when
240 ANAEROBIC GRANULES AND GRANULATION PROCESSES

the acetate concentration decreases, the growth of the mean sizes were 2.4 and 2.7 mm, respectively. In the
Methanosarcina aggregates will stop. reactors seeded with fermenter nuclei, granules were still
Although it is possible to use Methanosarcina aggre- in the lag phase and had a mean diameter of 1.1 mm.
gates as nuclei, Methanosarcina dominated granules are The authors suggested that the following two factors led
usually of less practical significance because this type of to the fast start-up of the reactors seeded with the syn-
granules are easily washed out of UASB reactors (43,49). trophic, Methanosarcina, and Methanosaeta nuclei. First,
syntrophic bacteria and methanogens have relatively low
Unidentified Filaments. A UASB reactor fed with growth rates. The presence of these organisms in the
methanol was inoculated with sludge (fine floc) from nuclei secured a fast start-up. Second, the nuclei with syn-
another UASB reactor treating methanolic waste (31). trophs and aceticlastic methanogens effectively promoted
Electron micrographs showed that granules contained rod- the attachment of fermenters and prevented flocculant
type and long filaments in the core and Methanosarcina growth.
at the surface. The rod and filament cells likely served as In another study, the potential for M. concilii and
nuclei for the granulation. Thin filaments (possibly SRB) propionate-degrading syntrophic consortia to serve as
also may have served as nuclei for granulation during nuclei for granulation was tested by monitoring granula-
start-up of a UASB reactor treating wastewater high in tion in two laboratory-scale UASB reactors inoculated with
sulfate (5 g/L) (53). nongranular sludge and treating synthetic wastewaters
containing glucose (Reactor NP) and glucose and pro-
Syntrophic Consortia. In a study by Grotenhuis and pionate (Reactor P) (19). The effluent acetate concentra-
coworkers (52), granulation of a coculture consisting of tions were controlled to be below 200 mg/L. Quantitative
a propionate degrader and Methanosaeta proceeded very membrane hybridizations and FISH with oligonucleotide
fast (11 days) in a UASB reactor with recycling. A few other probes targeting 16S rRNA were used to monitor changes
studies also compared the effectiveness of syntrophic con- in microbial communities and to study cell aggregate struc-
sortia as nuclei with other possible nuclei. For example, tures during granulation. Methanosaeta concilii cells were
El-Mamouni and coworkers (54–56) compared the influ- present at high levels and demonstrated good settling
ence on granulation of nuclei that contained mainly ability. Increases in their abundance were associated with
Methanosaeta spp., Methanosarcina spp., syntrophic con- significant increases in cell aggregate sizes at the early
sortia, or fermenters. The characteristics of four types stage of granulation. Methanosaeta concilii cells were
of nuclei enriched in four different USBF reactors are found to serve as backbones in small cell aggregates with
summarized in Table 1. The fermenter nuclei had the other archaeal cells and bacterial cells attached to them,
smallest size, lowest settling velocity, and lowest ash con- and they remained dominant in large cell aggregates and
tent, whereas they had the highest ECP content and were mature granules. These observations support the hypothe-
rich in carbohydrates. The Methanosaeta nuclei had the sis that Methanosaeta cells serve as nuclei for granulation.
highest ash content, which might be due to the precipita- On the other hand, syntrophic propionate-oxidizing bacte-
tion of calcium carbonate, magnesium carbonate, and iron ria, namely Syntrophobacter spp., exhibited poor settling
sulfide. Together with suspensions of municipal anaerobic ability and were easily washed out from the system. Their
sludge, the four types of nuclei were used to inoculate four contribution to granulation was probably minimal.
USBF reactors treating sucrose (54). The mass ratios of In summary, studies have shown that nuclei can be
municipal anaerobic sludge to nuclei were 1 : 4 g VSS/g inert particles or aggregates of microorganisms. Particles
VSS. At the end of the experiment (on day 120), the with diameters ranging from 40 to 100 µm are important
granules in the reactor seeded with syntrophic nuclei for granulation and the addition of certain types of inert
reached their plateau of stabilization and had the largest particles can increase the granulation rate. Also, several
size (mean diameter of 3.1 mm). The granules in reac- studies pointed to the importance of Methanosaeta, which
tors seeded with Methanosarcina or Methanosaeta nuclei is believed to attach to inert particles first after which
were still in the exponential phase at this point, and granulation is started. Filamentous microorganisms (e.g.,

Table 1. Characteristics of Four Types of Microbial Nuclei for Granulation (55)

Nucleus Type Syntrophic Consortia Methanosaeta spp. Methanosarcina spp. Fermenters

Enrichment medium Ethanol/Acetate Acetatea Methanol Sucroseb

Settling velocity (m/h) 10.5 11.3 8.7 3.2
Size (mm) up to 1.2 up to 1.2 up to 1.2 <0.6
Ash content (%) 30 60 40 16
Extracellular polymers (mg/g 35.2 42.2 26.6 216.5
VSS)
Carbohydrates: ECP (%)c 12 19 21 84
Proteins: ECP (%)d 41 43 64 7
a
Effluent acetate concentrations were controlled to be below 150 mg/Liter;
b
Reactor was overloaded (3 to 5 g COD/g VSS·day) and pH was 5;
c
Percentage of carbohydrates in ECP.
d
Percentage of proteins in ECP.
 ANAEROBIC GRANULES AND GRANULATION PROCESSES 241

Methanosaeta, fermentings filaments) or small aggregates of two identical bacteria is determined by liquid surface
(e.g., syntrophic consortia and Methanosarcina) can serve tension (γLV ) and the surface tension of the bacteria
also as nuclei. Finally, syntrophic consortia, at least in (γBV ) (61). Calculations demonstrated that the adhesion
some studies (54,55,56), appear to be better nuclei than of hydrophilic cells was favored for γLV less than 50 mN/m,
Methanosaeta, Methanosarcina, and fermenters. and adhesion of hydrophobic cells was favored for γLV
greater than 55 mN/m. When γLV was between 50 and
Colonization 55 mN/m, neither process was favored. Contact angle
measurement of pure cultures showed that, in general,
Once the nuclei are formed, microorganisms need to attach
fermenters are more hydrophilic than acetogens and
to them in order to be retained in the system, which results
methanogens (62). Granules treating sucrose wastewater
in the formation of granules. The following factors can
consisted of three layers, an outer hydrophilic layer, a
affect the adhesion process.
middle hydrophobic layer, and a core with an intermediate
level of hydrophobicity. When the surfactant linear
Divalent Cations and Cation Polymers. Divalent cations
alkylbenzenesulfonate was added to a UASB reactor fed
such as Ca2+ and Mg2+ can promote granulation in two
with sucrose to lower the γLV , the outer layer became more
ways. First, they condense the diffusive double-layers
hydrophilic and the middle layer more hydrophobic (62).
resulting in a stronger effect of van der Waals attractive
These granules had the advantage that biogas bubbles
forces. Second, they may directly neutralize the negatively
did not adhere strongly to their more hydrophilic outer
charged cell surface. Mahoney and coworkers. (57) con-
firmed this experimentally. They shortened the start-up layers and thus were less susceptible to washout than
time of a process treating a synthetic wastewater con- granules with less hydrophilic outer layers, resulting
sisting of acetate, propionate, and butyrate through the in higher reactor stability when the volumetric loading
addition of 2.5 mM Ca2+ . Compared to a system without rate was high (36,61). It appeared that microorganisms
Ca2+ addition, the granules formed in the system with in the granules were able to change their position
Ca2+ addition were larger and exhibited a settling ability according to the γLV . When the feed contained higher
three to four times higher. Similarly, addition of 0.2 mM levels of VFAs, the surface of the granules became
Mg2+ , 0.34 mM Ca2+ , and trace elements enhanced the more hydrophobic because of the presence of fewer
granulation in a mesophilic UASB treating sucrose (58). fermenters.
However, too much Ca2+ can deteriorate the granulation In addition, protein degraders are very hydrophilic
process. For example, Thiele and coworkers (46) demon- compared to other fermenters (62). Their high γBV might
strated that the addition of 25 mM Ca2+ caused washout explain why it is difficult to form granules in protein-rich
of the biomass. wastewater. Daffonchio and coworkers (62) demonstrated
Schmidt and Ahring (59) studied the importance of that the addition of a polycation, which reduces the γBV of
Mg2+ concentration (0 to 100 mM) for the formation of the cells, to a wastewater consisting of 20% molasses
stable granules (using acetate as the substrate at 55 ° C). and 80% peptone with an intermediate liquid surface
At very low Mg2+ concentrations (0 and 0.5 mM), the tension (γLV = 55 mN/m) increased the yield of granules
biomass was washed out at high rates (50% and 20%, significantly (60 to 75%).
respectively). On the other hand, high Mg2+ concentrations Modification of hydrophobicity can help in certain
(100 mM) resulted in breakage of Methanosarcina cell reaction systems, but does not promote granulation per
aggregates and adversely affected the granulation process. se because granules should contain both hydrophilic
In addition, the presence of high Mg2+ concentrations fermenters and hydrophobic acetogens and methanogens.
(30 and 100 mM) resulted in biofilm formation on the Changing the level of surface tension will only favor one
reactor wall rather than enhancement of granulation. The type of cell. Two lab-scale UASB reactors were started
authors recommended that the Mg2+ concentration should with sludge from a full-scale anaerobic digester in a
be higher than 0.5 mM but lower than 30 mM to promote yeast production plant (63). Linear alkylbenzenesulfonate
granulation. was added to one of the reactors to reduce the liquid
Analyses of ECP indicated that Co2+ and Fe2+ were surface tension to 48 mN/m. No granulation was observed
found in extracted ECP (60), in which they might bind to in either of the reactors 50 days after start-up. When
polymers and play important roles in linking ECP, which the sugar level in the wastewater increased from
is usually negatively charged. Furthermore, calcium and 7% to 27%, granules formed in both reactors at the
phosphorus were found to be the dominant elements in same time.
ashed ECP (39). They might also contribute to the stability
of granules. Species-Specific Interactions. Some species prefer to
Adding cationic polymers during the start-up of an form aggregates with certain species rather than with
ASBR reduced the granulation time from four months to others. This type of aggregation may be determined by the
one month (42). The cationic polymers were believed to production and consumption of intermediates, for example,
enhance granulation through bridging negatively charged in the case of syntrophic interactions.
bacterial cells and through the formation of large biomass In batch experiments with pure cultures, butyrate-
aggregates via sweep-floc mechanisms. degrading bacteria (strain BH) formed strong aggregates
with M. formicicum, but not with M. hungatei (50). In
Hydrophobicity. From the point of view of surface addition, propionate-degrading bacteria (strain PT) only
thermodynamics, the free energy change during adhesion formed strong aggregates with M. formicicum when a
242 ANAEROBIC GRANULES AND GRANULATION PROCESSES

Methanosaeta sp. was present. Wu and coworkers (50) produce polysaccharides (38). For granules grown on ace-
demonstrated that M. formicicum played a very impor- togenic or methanogenic substrates, proteins may play an
tant role during granulation when a number of defined important role in granulation, and lipids might be com-
species (strain BH, strain PT, Methanosaeta sp. strain pensating for the lower amounts of polysaccharides and
M7, M. formicicum T1N, and M. mazei T18) were used. proteins. The authors also pointed out that high ECP
Methanobacterium formicicum covered the surface of levels are not necessary for the production of active gran-
Methanosarcina aggregates and served as a link between ules.
the nuclei and the other cells. This study also demon- Thermophilic granules were found to have a lower total
strated the importance of a Methanosaeta sp., which ECP yield as well as a lower content of polysaccharides
encouraged the aggregation between M. formicicum and and proteins, but higher lipid levels (66). Similar obser-
strain PT and thus helped with syntrophic degrada- vations for thermophilic granules were reported by other
tion. researchers (67,68). This indicates that the production of
polysaccharides and proteins may be limited, or that they
Extracellular Polymers (ECP). ECP is an alterna-
are degraded faster under thermophilic conditions. It was
tive name for glycocalyx, which is defined as the
also found that thermophilic granules were not as strong
polysaccharide-containing structures from bacterial (or
archaeal) origin located outside the cell wall (64). It con- as mesophilic ones (68), which might be due to the lower
tains polymers of saccharides, proteins, lipids, phenols, total ECP in thermophilic granules.
and nucleic acids (38). ECP can be present in forms of slime In general, fermenters are believed to play an important
layers or capsules and appears outside of almost all types role in producing ECP and enhance the granulation
of prokaryote cells (64). ECP is sometimes also referred process. Using vinasse as the feed for a UASB reactor,
to as extracellular polymeric substances (EPS). However, Vanderhaegen and coworkers (47) demonstrated that
EPS also stands for extracellular polysaccharides. Here, preacidification of the feed resulted in a growth yield of
we use ECP to represent extracellular polymers, and EPS granules ten times lower than that fed with fresh vinasse.
to designate extracellular polysaccharides. The authors believed that the presence of high-energy
As discussed previously, ECP play an important role carbohydrates favors granulation by enhancing the growth
in granulation. They help to form strong bonds between of populations of fermenters, which produce ECP. First,
nuclei and cells. Also, they help to trap the offspring VFAs produced through the degradation of high-energy
of attached cells to form microcolonies. This theory was carbohydrates provide selective conditions for fermenters,
supported by MacLeod and coworkers (21) who found ECP which are insensitive to VFAs. The insensitiveness is
in all three layers of sucrose-fed granules. acquired by secreting ECP. Second, because of the
The ECP content in granules ranges from 0.6 to 20% of production of hydrogen, fermenters that do not produce
VSS (38), and is much lower than in activated sludge (39). hydrogen are favored, such as Propionibacterium, which
The main components of the ECP extracted from granules can effectively produce ECP. In another study (discussed
are proteins and polysaccharides, and a typical ratio earlier), granulation only happened when the sugar
between these two components is 2 : 1 to 6 : 1 (w:w). The content in the feed increased to 27% (in terms of COD) (63).
lipid contents are usually very low, only 0.02 to 0.05% Yoda and coworkers (44) found that when a reactor was fed
of VSS. VFAs (acetate, propionate, and butyrate), only 10 to 40%
The production of ECP depends on the growth of the biomass changed to granular sludge, whereas in the
conditions and characteristics of wastewater. At high reactor fed with glucose or molasses, all the sludge became
substrate concentrations and thus, high growth rates, granular. The authors believed that the presence of fast-
the ECP content of anaerobic sludge is high, and vice growing fermenters produced large amounts of ECP, which
versa (65). Excessive production of ECP can have a
was beneficial for granulation.
negative effect on granulation because the negative surface
However, the addition of fermenters to a UASB reactor
charge increases with increasing ECP yields (39) in a
resulted in flotation of granules (69), suggesting the need
linear relationship (65). When the substrate is propionate
of the right types of fermenters (i.e., fermenters with
or butyrate, the increase of negative surface charge is
suitable surface characteristics). In another study by Brito
linearly correlated with the increase of protein content
and coworkers (70), granulation only started in a UASB
of ECP. However, when the substrate is glucose, the
negative surface charge increases with both protein and reactor inoculated with digested activated sludge when
carbohydrate content (65). the VFA content of a low strength wastewater (1,000 to
Schmidt and Ahring (66) determined the ECP content 1,500 mg/L COD) changed from 15% to 85%. The authors
of six types of granules treating different wastewaters found that only fluffy flocs were formed when the feed
at different temperatures. It was found that granules contained 15% VFA and 85% glucose. The polysaccharide
treating complex substrates have higher total ECP and ECP was very high in the fluffy flocs as a result of
higher polysaccharide and protein contents, but lower the predominance of fermenters, and it decreased after
lipid contents than those fed acetogenic and methanogenic granules formed. The same changes were observed for
substrates. Methanogens and acetogens seemed to have protein ECP. The electrophoretic mobility of the biomass
a limited ability to produce ECP, particularly polysaccha- was less negative after granulation. This could relate
rides, possibly because the production and excretion of to the predominance of Methanosaeta, which has a less
ECP require too much energy. On the other hand, fer- negative surface in comparison with other anaerobic
menters were believed to be the major organisms, which microorganisms isolated from granular sludge (71).
 ANAEROBIC GRANULES AND GRANULATION PROCESSES 243

Beside fermenters, some other microorganisms are also sludge and crushed granular sludge). Some general
found to be important in producing ECP for granulation. guidelines for primary start-up of UASB reactors were
For example, M. arboriphilicus can synthesize all amino summarized by Lettinga (5). These guidelines reflect the
acids but cysteine. When hydrogen and nitrogen supply current understanding of the mechanisms involved in
is high and cysteine is limiting, M. arboriphilicus cells granulation. For instance, given the importance of nuclei,
secrete extra amino acids as peptide polymers. These it is recommended to use inert carrier particles or crushed
polymers are believed to be important for granulation (72). granules to promote start-up (40). Similarly, based on the
TEM observation of granules fed VFAs showed that cells fact that Methanosaeta spp. usually are present at high
in the granules were surrounded by ECP and attached levels in various granules and play an important role in
to each other by fibrous polymers (73). Chemical analysis granulation, it is generally recommended to maintain the
of the components of EPS of the granules demonstrated acetate level below 200 mg/L to promote the growth of
that they were the same as EPS of M. mazei and Methanosaeta. Furthermore, given that M. arboriphilicus
M. formicicum. The authors concluded that these two secretes ECP and can enhance granulation, Wentzel and
microorganisms, especially M. formicicum, contributed coworkers (72) suggested that a UASB should be operated
significantly to the production of the ECP of anaerobic in plug-flow or semi plug-flow pattern with high hydrogen
granules, and facilitated granulation. partial pressure (PH2 ) in the lower zone of the reactor.
The importance of the effect of polymers on granula- This should enhance the growth of M. arboriphilicus and
tion was also demonstrated by El-Mamouni and cowork- promote secretion of ECP, thus enhancing granulation. In
ers (74). UBF reactors were started up with syntrophic addition, because fermenters can produce large amounts
consortia as inoculum, or disintegrated biomass with nat- of ECP, wastewaters containing high-energy substrates
ural or synthetic polymers. The granulation rate of the (e.g., sugars) should be ideal for granulation. However,
reactor with natural polymer was the highest, whereas the because of the high growth rate of fermenters, it may be
one with syntrophic consortia was the lowest. Metabolic necessary to control the levels of fermenting bacteria (40).
activity tests on the three types of granules showed that For example, an excessive abundance of fermenters can
polymer enhanced granules had higher substrate activities result in fluffy granules as observed in a UASB reactor fed
on acetate and ethanol, but demonstrated lower substrate with 85% glucose and 15% VFAs (acetate and propionate)
activity on glucose compared to granules developed from at low COD concentrations (1,000 and 1,500 mg/L) (70).
syntrophic consortia. The mixing level and/or the upflow velocity can be adjusted
In summary, ECP contributes to granulation by to selectively remove some of the fermenters from fluffy
allowing microorganisms to stick together. The proper granules (19).
amount of ECP can make granules stronger. In addition, As an alternative to primary start-up, a reactor can be
it can make the granules more hydrophilic and prevent started up with (nonadapted) granular sludge (76). This
flotation (75). Fermenters can produce large amounts is called secondary start-up. With more and more full-
of ECP with high-carbohydrate content. Their presence scale UASB reactors in operation, excess granular sludge
could enhance granulation. However, excessive ECP is becoming available. As a result, secondary start-up is
(as in the flocs) causes repulsion between cells and becoming more popular.
prohibits granulation (38). Moreover, Methanobrevibacter As discussed in the previous sections of this article,
spp., Methanobacter spp., and Methanosarcina spp. also the microbial composition and granular structure change
can be important in secreting ECP for granulation. when feed composition and operating condition change.
These changes are likely to cause problems when
Colonization involves the attachment of different nonadapted granular sludge is inoculated into a new
organisms to the nuclei as a mechanism to be retained reactor. Problems often associated with secondary start-
in the reactor. In addition to interactions between up have been summarized by Lettinga in several
different organisms, cell surface characteristics play an reviews (2,5,76):
important role at this stage. Granulation can be enhanced
by stimulating the production of ECP by selecting for
populations that produce significant amounts of ECP and • Dispersed fermenters present in wastewater can act
by the addition of chemicals such as surfactants, cations, as carrier material for methanogens, or can attach to
or synthetic polymers. granules. Both phenomena result in flotation.
• Flotation can also be caused by the formation of
hollow spaces inside granules and/or the formation
START-UP PROCEDURES FOR REACTORS THAT UTILIZE
of poorly gas-permeable scales of organic (e.g., lipids)
GRANULAR SLUDGE
or inorganic (e.g., CaCO3 ) scales, which retard or
completely obstruct the escape of biogas bubbles.
The ultimate goal when studying granular sludge and
granulation processes is to change reactor designs • Granules may disintegrate, which can result in
and operational strategies to improve the start-up and a serious drop in the amount of granular sludge
performance of anaerobic wastewater treatment systems. biomass. Disintegration can be caused by changes in
There are two types of start-up procedures for UASB loading rate, substrate types, or level of acidification
reactors, namely, primary start-up and secondary start- of the substrate.
up. A primary start-up indicates that the inoculum • Granules can become too big, which may lead to
consists of nongranular sludge (e.g., anaerobic digester serious substrate transport limitations.
244 ANAEROBIC GRANULES AND GRANULATION PROCESSES

• It may also be difficult for organisms with low 6. L. T. Angenent and R. R. Dague, Med. Fac. Landbouwk en
growth rates to develop to high levels in nonadapted Toeg. Biol. Wet. Univ. Gent 61, 2077–2084 (1996).
granular seed sludge. For example, it may take a long 7. G. C. Banik et al., Structure and Methanogenic Activity of
time to increase the levels of organisms required for Granules from an ASBR Treating Diluted Wastewater at
the degradation of specific compounds present in a Low Temperatures, Proceedings of The 8th International
Conference on Anaerobic Digestion, May 25–29, vol. 2,
new wastewater or of thermophilic organisms when
Sendai, Japan, 1997, pp. 548–555.
starting up a thermophilic reactor using mesophilic
8. W.-M. Wu et al., Appl. Environ. Microbiol. 57, 3438–3449
granules.
(1991).
9. W.-M. Wu et al., Appl. Microbiol. Biotechnol. 38, 282–290
Because of these problems, in particular problems
(1992).
associated with granular structure, crushed granules may
10. B. Schink, Microbiol. Mol. Biol. Rev. 61, 262–280 (1997).
be a better inoculum than intact granules (76). Crushed
11. B. K. Ahring, Antonie van Leeuwenhoek 67, 91–102 (1995).
granules can provide nuclei and the specific microbial
populations necessary for granulation. 12. S. Uemura and H. Harada, Appl. Microbiol. Biotechnol. 43,
358–364 (1995).
13. S. Uemura and H. Harada, Environ. Technol. 14, 897–900
CONCLUSION (1993).
14. J. T. Grotenhuis et al., Appl. Environ. Microbiol. 57,
Three decades have passed since the development of 1942–1949 (1991).
the UASB reactor and the use of granular sludge to 15. F. A. Visser et al., Appl. Environ. Microbiol. 57, 1728–1734
treat wastewater were first reported. During this time (1991).
period, the UASB reactor and its derivatives have demon- 16. D. Zheng, Master thesis, University of Illinois, Urbana-
strated their potential as high-rate anaerobic wastewater Champaign, Ill., 1995.
treatment systems, and our understanding of granules 17. H. J. M. Harmsen et al., Appl. Environ. Microbiol. 62,
and granulation processes has improved significantly. 2163–2168 (1996).
This article reviews achievements in understanding the 18. D. Zheng and L. Raskin, Microb. Ecol. 39, 246–262 (2000).
microbial composition and the architecture of anaerobic 19. D. Zheng, Ph.D. thesis, University of Illinios, Urbana-
granules, the mechanisms of granulation, and the role of Champaign, Ill., 1999.
different microorganisms in granulation processes. These 20. Y. Sekiguchi et al., Microbiology 144, 2655–2665 (1998).
achievements in turn have contributed to the improve- 21. F. A. MacLeod et al., Appl. Environ. Microbiol. 56, 1598–
ment of start-up and operating strategies for reactors, 1607 (1990).
which utilize granular sludge. 22. S. R. Guiot et al., Water Sci. Technol. 25, 1–10 (1992).
Some of the recent findings can be attributed to the 23. H. K. Chui and H. H. P. Fang, J. Environ. Eng. (ASCE) 120,
development and application of new technologies for 1322–1326 (1994).
investigating complex microbial communities. In partic- 24. P. A. Alphenaar et al., Micron 25, 129–133 (1994).
ular, nucleic acid based techniques (e.g., rRNA targeted 25. J. W. Morgan et al., J. Chem. Technol. Biotechnol. 50,
probe hybridizations, PCR, and cloning) and techniques 211–226 (1991).
based on immunology (e.g., antibody techniques) have 26. H. J. M. Harmsen et al., Appl. Environ. Microbiol. 62,
made it possible to study microbial community composi- 1656–1663 (1996).
tion and structure without the limitations and biases of 27. K. Syutsubo et al., Water Sci. Technol. 36, 391–398 (1997).
cultivation. However, our current understanding of gran- 28. Y. Sekiguchi et al., Appl. Environ. Microbiol. 65, 1280–
ules and granulation processes is far from complete. For 1288 (1999).
example, we still need to answer questions such as what 29. S. Rocheleau et al., Appl. Environ. Microbiol. 65, 2222–
are the metabolic activity levels within granules, how 2229 (1999).
are substrates transferred into granules, and how do the 29a. C. M. Santegoeds et al., Appl. Environ. Microbiol. 65,
characteristics of microorganisms influence the granula- 4618–4629 (1999).
tion process and the properties of granules. The ability to 30. H. H. P. Fang, Microbial Distribution and Syntrophic Asso-
answer these questions largely depends on the availability ciation in UASB Granules, Proceedings of The 8th Interna-
of suitable techniques, such as in situ metabolic activity tional Conference on Anaerobic Digestion, May 25–29, vol. 1,
assays, in situ measurement of concentration profiles, Sendai, Japan, 1997, pp. 83–90.
and in situ microbial identification and determination of 31. Z. I. Bhatti et al., World J. Microbiol. Biotechnol. 11,
microbial characteristics. 631–637 (1995).
32. H. H. P. Fang et al., Water Sci. Technol. 30, 87–96 (1994).
33. A. J. L. Macario et al., J. Gen. Microbiol. 137, 2179–2189
BIBLIOGRAPHY
(1991).
1. G. Lettinga et al., Biotechnol. Bioeng. 22, 699–734 (1980). 34. J. E. Schmidt and B. K. Ahring, Immobilization Studies
with Pure Cultures of Methanogenic Bacteria in UASB
2. G. Lettinga, Water Sci. Technol. 33, 85–89 (1996).
Reactors, Proceedings of The 8th International Conference
3. J. B. van Lier et al., Water Sci. Technol. 30, 251–261 (1994). on Anaerobic Digestion, May 25–29, vol. 2, Sendai, Japan,
4. S. R. Guiot and L. van den Berg, Biotechnol. Lett. 6, 161–164 1997, pp. 587–596.
(1984). 35. P. N. L. Lens et al., Appl. Environ. Microbiol. 59, 3803–
5. G. Lettinga, Antonie van Leeuwenhoek 67, 3–28 (1995). 3815 (1993).
 ARBUSCULAR MYCORRHIZAE 245

36. J. Thaveesri et al., Appl. Environ. Microbiol. 61, 3681– 68. J. Quarmby and C. B. Forster, Enzyme Microb. Technol. 17,
3686 (1995). 493–498 (1995).
37. J. Quarmby and C. F. Forster, J. Chem. Technol. Biol. 63, 69. W. Verstraete et al., World J. Microbiol. Biotechnol. 12,
60–68 (1995). 221–238 (1996).
38. J. E. Schmidt and B. K. Ahring, Appl. Biochem. Biotechnol. 70. A. G. Brito et al., Biotechnol. Lett. 19, 363–367 (1997).
49, 229–246 (1996). 71. J. Grotenhuis et al., Appl. Environ. Microbiol. 58, 1054–
39. C. F. Forster, J. Biotechnol. 17, 221–232 (1991). 1056 (1992).
40. L. W. Hulshoff Pol, Ph.D. thesis, Wagneningen Agricultural 72. M. C. Wentzel et al., Water Sci. Technol. 30, 31–42 (1994).
University, The Netherlands, (1989).
73. M. C. Veiga et al., Appl. Environ. Microbiol. 63, 403–407
41. T. Imai et al., Advanced Start-up of UASB Reactors by (1997).
Adding of Water Absorbing Polymer, Proceedings of The
74. R. El-Mamouni et al., Chemical versus Microbial Assistance
8th International Conference on Anaerobic Digestion, May
of the Anaerobic Granulation Process, Proceddings of The
25–29, vol. 1, Sendai, Japan, 1997, pp. 405–412.
8th International Conference on Anaerobic Digestion, May
42. R. A. Wirtz and R. R. Dague, Water Environ. Res. 68,
25–29, vol. 2, Sendai, Japan, 1997, pp. 597–600.
883–892 (1996).
75. V. O’Flaherty et al., Appl. Microbiol. Biotechnol. 47,
43. W. M. Wiegant and A. W. A. de Man, Biotechnol. Bioeng. 28,
102–107 (1997).
718–727 (1986).
76. G. Lettinga et al., Anaerobic treatment in sustainable
44. M. Yoda et al., Water Sci. Technol. 21, 109–120 (1989).
environmental production concepts, The 8th International
45. H. H. Beeftink and P. Staugaard, Appl. Environ. Microbiol.
Conference on Anaerobic Digestion, Sendai, Japan, 1997.
52, 1139–1146 (1986).
46. J. H. Thiele et al., Biotechnol. Bioeng. 35, 990–999 (1990).
47. B. Vanderhaegen et al., Water Sci. Technol. 25, 21–30
(1992). ANAMMOX. See ACTIVATED SLUDGE — MICROBIOLOGY OF
48. J. K. Jhung and E. Choi, Water Res. 29, 271–277 (1995). NITROGEN REMOVAL
49. W. M. Wiegant, in G. Lettinga et al., ed., Granular Anaero-
bic Sludge: Microbiology and Technology, Puduc Wagenin-
gen, Lunteren, The Netherlands, 1988.
50. W.-M. Wu et al., Appl. Environ. Microbiol. 62, 2037–2044
(1996).
ANHYDROBIOSIS. See DESICCATION BY EXPOSURE TO
SPACE VACUUM OR EXTREMELY DRY DESERTS: EFFECT ON
51. R. W. Robinson, Appl. Environ. Microbiol. 52, 17–27 (1986).
MICROORGANISMS
52. J. T. C. Grotenhuis et al., in G. Lettinga et al., ed., Granular
Anaerobic Sludge: Microbiology and Technology, Puduc
Wageningen, Lunteren, The Netherlands, 1988.
53. P. A. Alphenaar et al., Bioresour. Technol. 43, 249–258
(1993). ANTHRAX. See BIOTERRORISM
54. R. El-Mamouni et al., Water Sci. Technol. 32, 173–177
(1995).
55. R. El-Mamouni et al., J. Biotechnol. 39, 239–249 (1995).
56. R. El-Mamouni et al., Appl. Microbiol. Biotechnol. 47,
189–194 (1997). ANTIBIOTIC RESISTANCE. See HETEROTROPHIC
57. E. M. Mahoney et al., Water Sci. Technol. 19, 249–260 BACTERIA AS INDICATORS OF WATER QUALITY; NOSOCOMIAL
(1987). INFECTIONS; FECAL CONTAMINATION, SOURCES OF
58. J. A. S. Goodwin et al., Enzyme Microb. Technol. 12,
877–884 (1990).
59. J. E. Schmidt and B. K. Ahring, Enzyme Microb. Technol.
15, 304–310 (1993).
60. C. F. Shen et al., Water Res. 27, 25–33 (1993). ANYDROBIOSIS. See FREEZE DRYING: PRESERVATIONS OF
MICROORGANISMS BY FREEZE-DRYING
61. J. Thaveesri et al., Appl. Microbiol. Biotechnol. 43, 1122–
1127 (1995).
62. D. Daffonchio et al., Appl. Environ. Microbiol. 61, 3676–
3680 (1995).
63. J. Thaveesri et al., Antonie van Leeuwenhoek 68, 329–337 AQUIFERS, ATLANTIC COASTAL PLAIN.
(1995).
See MICROBIOLOGY OF ATLANTIC COASTAL PLAIN AQUIFERS AND
64. J. W. Costerton et al., Annu. Rev. Microbiol. 35, 299–324 OTHER UNCONSOLIDATED SUBSURFACE SEDIMENTS
(1981).
65. X. S. Jia et al., Water Sci. Technol. 34, 309–316 (1996).
66. J. E. Schmidt and B. K. Ahring, Appl. Microbiol. Biotechnol.
42, 457–462 (1994).
67. C. F. Forster and J. Quarmby, Antonie van Leeuwenhoek 67, ARBUSCULAR MYCORRHIZAE. See MYCORRHIZAE:
103–110 (1995). ARBUSCULAR MYCORRHIZAE
246 ARCHAEA: DETECTION METHODS

ARCHAEA, DETECTION OF. See ARCHAEA IN MARINE walls are complex structures but do not contain muramic
ENVIRONMENTS acid common to bacteria (9). Cell membranes of the
Archaea are made of lipids containing ether linkages and
phytanyl chains (10). Transfer RNA lacks ribothymidine
in one very specific loop in the secondary structure of the
molecule (11). The subunit structure of RNA polymerase
ARCHAEA: DETECTION METHODS is unique (12), and several unique coenzymes have been
found among the methanogenic Archaea (11). Many of
PETER D. NICHOLS these characteristic molecular or biochemical traits can
CSIRO Marine Research and be exploited to detect Archaea, and methods to do so are
Antarctic CRC described in this chapter.
Hobart, Australia
CAROL A. MANCUSO NICHOLS CULTURING METHODS
Antarctic CRC and School of
Agricultural Science, University
Classical microbiological techniques have been applied
of Tasmania
extensively to the culturing of Archaea. Isolation media for
Hobart, Australia
Archaea are selective in themselves, and growth is a direct
method of detection as few other bacteria share the growth
During the last twenty years, our understanding of the optima common among the Archaea. However, there is
phylogeny of living things has undergone significant some overlap, and selective culturing methods have been
change. The once commonly held notion that there were used to detect Archaea in mixed communities. Many
two basic cell types, the complex eukaryotes and the Archaea are enriched in media containing antibiotics such
more simple, ancestral prokaryotes, was tested when as kanamycin, penicillin, or vancomycin, which inhibit the
molecular biochemical techniques became available to cell walls of bacteria.
study an integral part of the protein synthesis machinery,
Ribosomal RNA (rRNA) (1). Because rRNA is ubiquitous Bile Salts and Antibiotics
in nature and the product of highly conserved genes,
comparing the sequence of nucleotides in rRNA of different Halophilic Archaea grow at salt concentrations between
organisms provided an excellent means for assessing 1 M and 4.5 M (7). Halophilic bacteria can also exist at the
the relatedness of two organisms. The relative length lower end of this range, although some, such as members
of time since the two organisms diverged from a common of the family Halomonadaceae, do live in saturated
ancestor could also be estimated once the molecular clock is brines. Low concentrations of bile salts such as sodium
adequately calibrated against evolutionary events. Early deoxycholate or sodium taurocholate have been shown
studies showed that prokaryotes and eukaryotes had to lyse Halobacterium and related Archaea, whereas both
evolved from a common ancestor by separate pathways (2). halophilic bacteria and halococci remain intact. Antibiotics
Closer examination of prokaryotes revealed a group as such as anisomycin behave in a similar fashion (13). In
distinctly different, biochemically and structurally, from this way, halophilic Archaea can be detected indirectly
other bacteria as it was from eukaryotes (2,3). These and population sizes can be estimated among mixed
microorganisms formed the third phylogenetic Kingdom communities (14,15).
and were named Archaeabacteria (1) to reflect their primi-
tive evolutionary status; they are at least as old as the Most Probable Number (MPN)
other major bacterial groups. The term Archaeabacteria Methanogens produce methane by metabolizing carbon
was changed subsequently to the archaeal domain (4). substrates and hydrogen within reduced environments
The domain is divided into two Kingdoms (5). The first, (Eh below −330 mV). Although most methanogens are
Euryarchaeaota, contains members that are physiologi- mesophiles, marine species include psychrophiles, thermo-
cally diverse and includes the methanogens that produce philes, and hyperthermophiles and they are ubiquitous
methane from hydrogen and carbon dioxide or other simple despite their special growth requirements (16).
carbon compounds in a strictly anaerobic environment (6), The concentration of viable cells from these various
extreme halophiles that grow aerobically in highly saline environments can be roughly calculated using MPN
environments and maintain an extremely high internal techniques. This involves the mathematical estimation of
salt concentration (7), and thermophiles found in close the viable count from the fraction of multiple cultures that
proximity to terrestrial and shallow water–hot springs fail to show growth in a series of dilution tubes containing
and deep-sea hydrothermal vents. The second kingdom, a suitable medium (17). The procedure involves preparing
Crenarchaeaota, includes members of the extremely several replicate dilutions in a growth medium selective
thermophilic sulfur-metabolizing phenotype. Originally to the particular methanogen under study and noting the
known as thermoacidophiles, these bacteria have also been fraction of tubes showing bacterial growth. Tubes with no
well studied (8). growth are assumed to not have received even a single cell
In addition to their ability to inhabit specific ‘‘extreme’’ capable of growth. Accuracy in detecting and estimating
environments and the presence of identifying 16S rRNA methanogenic biomass using the MPN technique increases
sequences, there are several other characteristics held in with the use of increasing numbers of culture tubes. This,
common by the Archaea studied thus far. Archaeal cell combined with the use of anaerobic media, can result in
 ARCHAEA: DETECTION METHODS 247

this approach being a cumbersome technique when applied uncultured Archaea may be among the most abundant
to methanogenic Archaea. microorganisms on earth. The microbial community of
Despite these drawbacks, the MPN method has been Deep Lake in Vestfold Hill, Antarctica, was found to be
applied to many anaerobic environments, such as rice made up almost entirely of halophilic Archaea, including
fields (18), where selective media enabled the enumeration several novel lineages (28). In the past, conventional
of the populations from four major methanogenic trophic culturing methods had resulted in the isolation of one
groups. In another study, MPN counts showed that viable halophilic Archaeon from Deep Lake (29).
methanogens coexisted with homoacetogens at depths
of 45 m to 446 m below sea level in a deep-granite Denaturing Gradient Gel Electrophoresis (DGGE)
aquifer (19). In a third study, MPN of methanogens from
Another technique that avoids the need for cloning has
a rumen were used to elucidate the relationship between
been developed. It involves the fractionation of PCR-
Archaea and ciliate protozoans in the gut (20).
amplified 16S rDNA fragments on polyacrylimide gels
containing a denaturing reagent consisting of formamide
MOLECULAR TECHNIQUES and urea. The level of denaturation of individual sequences
is used to resolve mixtures of DNA fragments. Gel patterns
Despite the search for new strains, all Archaea isolated generated can be used to compare different environmental
using classical microbiological culturing techniques fall samples and individual bands can be directly analyzed
into one of the three basic phenotypes. Traditional by sequencing. This method has been applied to study
laboratory enrichment and culturing methods identify microbial communities in deep-sea sediments (30) in
microorganisms based on metabolism, morphology, and which archaeal 16S rRNA represents 2.5–8.0% of the
physiology (16). The current opinion is that this method total prokaryotic DNA.
of detection uncovers less than 0.1% of microbes in the
environment (21–23). Molecular techniques have provided Fluorescent In Situ Hybridization (FISH)
a means by which to determine the presence of Archaea Chemical staining has been used in the past to detect
and to characterize them while circumventing the need for cells in culture and in environmental samples. These
culturing (16,24). However, the physiology and ecological methods tend to be nonspecific (31,32). Fluorescent anti-
role of many of the ‘‘uncultivatable’’ Archaea still requires bodies used as stains can be highly specific (33–35), but
elucidation. the strain specificity is dependent on the use of well-
studied microbes (24). A staining method that involves
Molecular Cloning hybridization of oligodeoxynucleotide probes (24) has been
The advent of nucleic acid sequencing techniques have described. The probes binding to complementary regions
opened up the potential to study archaeal members of of 16S rRNA sequences (17–34 nucleotides in length) are
microbial communities from environments in which they chosen because they are specific to a particular phyloge-
were never before known to exist (25,26). Short segments netic group (e.g., Archaea). This molecular probe–staining
of deoxynucleotides have been used as probes to bind method is highly specific, but requires no previous knowl-
very specifically to targeted sequences of 16S rRNA. This edge of the organisms under study, and has successfully
allows the total chromosomal DNA of a specific group to been applied to archaeal cells (see FLUORESCENT PROBES FOR
IN SITU ANALYSES OF MICROBIAL COMMUNITIES). The probes are
be detected in a complex microbial population. Sections
of archaeal 16S rRNA genes bound by Archaea-specific tagged with a fluorescent label for microscopic examina-
primers are amplified by PCR (see RIBOTYPING METHODS tion. Cells are harvested from a culture or environmental
FOR ASSESSMENT OF IN SITU MICROBIAL COMMUNITY STRUCTURE,
samples, resuspended in buffer, and fixed with formalde-
this Encyclopedia). Individual fragments are then cloned hyde. Cells are then smeared on a slide coated with gelatin,
dried and fixed again (36). Smeared slides are treated with
and sequenced, and the sequences are then compared
buffered fluoro-labeled probes, incubated in hybridization
to known Archaeal rDNA sequences. In this way, the
mixture, and then viewed.
phylogeny of Archaeal members of a population can be
determined quickly and without the need for classical
Optical Tweezers
culturing techniques (16).
These molecular techniques were applied to marine An extremely sensitive microscopic technique has been
plankton (25). Sequences that were speculated to be from used to complement the accuracy of these molecular
a previously undescribed Archaeal group were found. detection methods. A highly focused infrared laser is
These microorganisms may have diverged from ancestors coupled to a computer-controlled inverse microscope
of other previously characterized microorganisms very and this system, called ‘‘optical tweezers’’ is used to
early in evolution. More recently, these techniques have micromanipulate single Archaeal cells from a mixed
been used to detect Archaea in cold Antarctic marine microbial population for culturing. Using this method, it
surface waters, where they were estimated to comprise can be demonstrated that cultures obtained subsequently
34% of the prokaryotic biomass (26). Archaea were found from this single cell correspond to the one detected in
to be widespread in the deep sea when this method was situ (16,37). In the past, such detection and description
applied to samples taken at great depth (1,000–3,000 m) of Archaea would have been possible only by in situ
from both the Atlantic and Pacific Oceans (27). These phylogenetic analyses. This new method was used to
researchers concluded that previously unknown and isolate anaerobic Archaea, which, at 83 ° C, produce
248 ARCHAEA: DETECTION METHODS

grapelike aggregates of coccoid cells that were previously then digested with trypsin (40,46,47). Purified cell walls
overgrown during enrichment by filamentous cells. can be further degraded into constituent amino acid and
amino sugar residues by acid hydrolysis. Identification
CELL ENVELOPE — CHEMICAL TECHNIQUES and quantification of amino acids and sugars are carried
out with an amino acid analyzer. Talosaminuronic
The cell envelope includes the cell wall and cell membrane acids may be qualitatively determined after short-term
constituents. The latter includes lipid in the membrane hydrolysis (4 M HCl, 30 minutes, 100 ° C) (40,45). After
bilayer and quinones. With reference to Archaea, these isolation by thin-layer chromatography, the acid and
specific constituents and methods to detect them are lactone forms of the acid are determined by an amino
described in the following section. acid analyzer (41). Neutral sugars are converted to their
alditol derivatives, which are then identified by gas
Cell Wall chromatography (GC) (48).

Methods to detect Archaea by measurement of cell wall S-Layers (Surface Layers)

components are neither direct nor as conclusive as either
molecular (previous section) or signature lipid techniques Protein S-Layer. Methanogens of the order Methanococ-
(following section). Nevertheless, a range of protocols exist cales have a cell envelope consisting of the cell membrane
for measuring cell wall constituents and methods to detect and a layer of nonglycosylated protein subunits arranged
them are described in the following paragraphs. hexagonally. This outer layer is referred to as an S-layer,
In bacterial cell walls, a polymer of peptidoglycan and differences in the molecular weights of S-layer pro-
(or murein) provides strength and rigidity to the teins are useful taxonomic markers (49). Methods for the
cell. Although it differs slightly from one bacterial isolation of S-layers and S-layer proteins of methanococci
species to the next, three main components form have been described (44). After mechanical disruption,
the basic structure: (1) N-acetylmuramic acid (NAM), cells are treated with DNAase and then a surfactant. The
(2) N-acetylglucosamine (NAG), and (3) a tetrapeptide S-layer is extracted by a modified Bligh-Dyer (50) extrac-
containing D-amino acids (38). tion. Dialysis is followed by gel electrophoresis to obtain
Archaea possess cell walls that are markedly different the S-layer protein. Protein is determined by measuring
from those of bacteria (39–41). Changes in chemical absorbance at 750 nm (51).
composition of the constituents are reflected in structural
differences. Cell walls of the Archaea, when they are Glycoprotein S-Layers. The surface layer of Halobac-
present, do not contain peptidoglycan as muramic acid, terium halobium is made up of glycoprotein subunits that
and D-amino acids are absent. Instead, glycoproteins or maintain the shape of the cell (52). This S-layer dissolves
polysaccharides and proteins give rigidity to the cell wall. in salt concentrations below 12% in contrast to S-layers of
The composition of the cell walls of the various members other cells, which tend to aggregate in water.
of the Archaeal group reflects the heterogeneity of the In Methanothermus spp., S-layer glycoproteins and
group. No common cell wall component, such as muramic glycoprotein sugars are characteristic. Extraction with
acid found in bacteria, is present in the Archaea (42). In trichloroacetic acid followed by reversed-phase chromatog-
Archaea such as Thermoplasma, the cell wall is completely raphy yields the purified glycoprotein (53). Molecular
absent (43). The gram reaction is highly variable among weights obtained through SDS–polyacrylamide gel elec-
the Archaea, and is dependent on the combination of cell trophoresis together with other features can be useful
wall components involved in this structure (44). taxonomically (44,54).
The glycoprotein of the S-layer in Sulfolobus acido-
Pseudomurein caldarius contains predominantly the polar amino acids
serine and threonine. To isolate and quantify these com-
The cell wall components of Archaea have not been ponents the acidic culture is neutralized, and after cen-
as thoroughly studied as those of bacteria, but there trifugation the buffered pellet is digested with DNAase.
is enough evidence to support their use as taxonomic Cell envelopes are disintegrated with heat (60 ° C) at pH 9,
markers (44). Archaea of the genus Methanopyrus and and the resultant glycoprotein is purified by molecular-
those of the order Methanobacteriales have cell walls sieve chromatography (55). After acid hydrolysis, amino
composed of pseudomurein (45). This polymer consists of acid residues are identified and quantified using an amino
peptide subunits (Lys, Glu, Ala or Thr) that cross-link acid analyzer (44).
glycan strands composed of an N-acetylated amino sugar
(glucosamine or galactosamine) that is common to bacteria
Methanochondroitin
and an N-acetyl-L-talosaminuronic acid that is unique to
Archaea (42,44). Methanochondroitin is a cell wall component unique to
Although there are many different variations in Methanosarcina, with a few exceptions (M. acetovorans
the peptide subunit of murein in bacteria, only one and M. frisia) that have only a single S-layer. A polymer
structure for pseudomurein has been found so far in of uronic acid and N-acetylgalactosamine residues present
Archaea. However, substitution in the peptide subunit in a 2 to 1 molar ratio gives cell aggregates their cuboid
can occur. For taxonomic purposes, elucidation of the appearance (56). Methanochondroitin is isolated according
chemical composition of pseudomurein is sufficient (45). to methods used for the isolation of pseudomurein.
Initially, the cells are mechanically disintegrated and Cells are mechanically disintegrated and then digested
 ARCHAEA: DETECTION METHODS 249

with trypsin (40,46,47). Uronic acids are subjected to signature lipids are shown in Figure 1. The isoprenoid
methanolysis and then derivatization before identification side-chains are usually of C20 and C40 chain length. When
by capillary GC (48). two C20 isoprenoid side chains are linked to a single
glycerol moiety, the structure is termed a diether (DE)
Heteropolysaccharide lipid (diphytanylglycerol, archaeol). When two C40 side
chains are linked to two glycerol units, the structures are
Methods have been described for the isolation and iden-
termed tetraether (TE) lipids (bidiphytanyldiglycerol, cal-
tification of heteropolysaccharide and constituent compo-
darchaeol). The diether and tetraether ‘‘lipid cores’’ refer
nents (41,57). To date they have not been demonstrated
to the hydrophobic or lipid portion of the membrane lipids.
to be taxonomically useful and their use for detecting
C15 and C25 side-chains may be present in the diether
Archaea may therefore be limited. Of interest is the obser-
lipids from selected Archaea, as may be a hydroxy group at
vation that Haloccocus morrhuae has a rigid cell wall of
the C3 position of the isoprenoid side chains. A variation to
a complex heteropolysaccharide that prevents cell lysis at
low-salt concentration. the DE core lipid occurs with the macrocyclic DE present
in deep-sea methanogens (60); a single C40 side chain is
Microscopy linked at both ends to one glycerol molecule. Variations
to the tetraether lipid structure include the presence of
Electron micrographs of thin sections of Archaeal cells between zero and eight cyclopentane rings.
reveal a variety of profiles, which give rise to two Other variations to the tetraether structure include the
basic gram reactions. Gram-positive Archaea are most presence of a cyclic C6 H11 O5 unit at the second glycerol,
similar to gram-positive non-Archaea cells. The cells calditoglyceroarchaeol (61,62). In Archaea, the diether and
have a pseudomurein layer surrounding the cytoplasmic tetraether lipids are generally present in phospholipid and
membrane and in some cases a heteropolysaccharide or glycolipid structures.
a single layer (S-layer) of protein or glycoprotein as well. Within the Archaea, the compositions of the lipid
Most Archaea are gram-negative and contain an S-layer cores, phospholipids, and glycolipids can be used to distin-
surrounding the cytoplasmic membrane. Cells containing guish between the three subgroups, extreme halophiles,
mitochondroitin and an underlying S-layer may gram methanogens, and extreme thermophiles or thermoaci-
stain positive or negative. dophiles, and to some extent between members of the
Thin sections of Archaea are prepared after cells different genera within each subgroup (63). The analysis
are fixed in gluteraldehyde, washed, dehydrated, and of pure cultures provides taxonomic information as the
embedded. After cutting, thin sections are contrasted with distribution of specific ether lipids may be restricted to
lead citrate, washed, dehydrated, and then contrasted certain subgroups of Archaea (Table 1). Application of the
again. Cells may also be subjected to platinum shadowing signature lipid assays to field samples then allows the
or negative staining before electron microscopy (58). detection of Archaea by microbial ecologists, organic geo-
chemists, and other scientists interested in understanding
Membrane Lipids the role of Archaea in the environment.
General Considerations. Along with the cell wall, the In their use of signature lipids, microbial ecologists
cell membrane is a major component of the cell envelope. are often more interested than organic geochemists in
A number of terms have been used to describe molecules the viable component of the community. As all living
used for the measurement of lipid components in cultured cells are surrounded by a membrane containing polar
Archaea and environmental samples. These include lipids, the component fatty acid and ether lipid side
molecular marker, biomarker, lipid biomarker, signature
lipid, signature lipid biomarker, and chemical fossil. In
this chapter, the term signature lipid is used as it is Table 1. Listing of Main Archaeal Groups and a Summary
more often the term preferred by microbial ecologists. The of Their Ether Lipid Composition (adapted from refer-
ence 63). See Table 2 for explanation of abbreviations.
main criteria for an ideal signature lipid are: (1) source
specificity and (2) conservative behavior. Ether Lipids Present
Source specificity refers to the link between a signature Group/Genus DE, Diether; TE, Tetraether
lipid and its source. Under ideal conditions, this link
should be direct and unique (59). In showing conservative Halophiles DE (diphytanylglycerol,
behavior, a signature lipid should be stable over Halobacterium, Haloferax, archaeol)
Haloarcula, Halococci
timescales relevant to the processes under study. Most
signature lipids will be subject to microbially-mediated Haloalkaliphiles DE
transformations and/or breakdown in the environment. Methanogens DE and TE
Methanococcus, (bidiphytanyldiglycerol,
To assess the potential use of a signature lipid as a
Methanosarcina, caldarchaeol). 3-OH group
quantitative tool, we need to understand all aspects of
Methanobacterium, may be present in the
its behavior in the environment. Methanothermus isopranyl side chains.
The recognition that Archaea contain unique lipid pro-
Thermophiles DE and TE. TE may contain
files comprising ether-linked isoprenoid alkyl side chains
Desulfurococcus, cyclopentane rings in the
relative to bacteria and eukaryotes has encouraged the Thermoplasma, Sulfolobus, isopranyl side chains.
assay of ether lipid distribution in pure isolates and Thermoproteus
field samples. Specific structures of selected archaeal
250 ARCHAEA: DETECTION METHODS

H2C O

H C O

H2COH

H2C O
C20−C25
H C O

H2COH

H2C O

H C O
OH
H2COH

CH2OR

O CH
CH2 O

O CH2
CH O

CH2OH

CH2OR
CH2 O
O CH

CH O
O CH2

CH2OH
Figure 1. Structures of ether lipid cores. Upper three diether (DE) core lipid structures:
2,3-O-diphytanylglycerol, archaeol; 2-O-sesterterpanyl-3-O-phytanylglycerol; 3-hydroxy-2-O-phyta-
nyl-3-O-phytanylglycerol. Lower two tetraether (TE) core lipid structures: bidiphytanyldiglycerol,
caldarchaeol; an example representative of a cyclopentanyl-containing caldarchaeol.

chains are considered to be several of the more important layer is obtained by addition of chloroform and water to
signature lipid classes (64). Unlike many other lipids, the Bligh-Dyer extraction (final solvent ratio, chloroform-
phospholipids are rapidly degraded within hours of cell methanol-water, 1/1/0.9, v/v/v). A representative scheme
death. Therefore, the examination of phospholipid-derived for the preparation of ether lipids is shown in Figure 2.
fatty acids of eubacteria and eukaryotes and phospholipid A combined lipid and DNA extraction method for
ether lipids of Archaea allows the viable community to environmental samples has been recently reported (67).
be investigated (65,66). The concept and use of signature At the phase separation stage of the Bligh and Dyer
lipids is discussed more fully in the entry LIPID BIOMARKERS procedure, lipids partitioned into the organic phase and
AND IMMUNOLOGICAL METHODS IN ENVIRONMENTAL MICROBIOLOGY.
DNA partitioned into the aqueous phase. The DNA was not
degraded during the lipid extraction procedure, although
Lipid Extraction and Fractionation DNA recovery was 40 to 50% compared to samples treated
Total lipids are extracted quantitatively from most by conventional DNA extraction techniques alone. An
Archaea using a modified Bligh-Dyer one-phase (chloro- advantage of the Bligh and Dyer procedure was that
form-methanol-water, 1/2/0.8, v/v/v) extraction proce- it also resulted in DNA extracts from field samples
dure (50,63). Halophiles with rigid cell walls (e.g., Halococ- containing lower amounts of humic material, compared
cus spp.) are subjected to sonication before extraction, and to conventional DNA extraction. The method shows utility
acidic conditions may be needed with certain methanogens for the co-recovery of both signature lipids and DNA from
(see following section). The lipid-containing chloroform a single sample; this is particularly useful when only
 ARCHAEA: DETECTION METHODS 251

Archaea and/ or environmental sample Degradative Procedures

Following extraction and column chromatographic frac-
Lipid extraction (Bligh and Dyer)
tionation of total lipid material, chemical breakdown of the
complex lipids is undertaken to produce the diether and
Total lipids tetraether lipid cores. A standard strong acid methanolysis
Column or thin layer chromatography treatment of the complex lipids with 0.6–2 N methanolic
HCl at 80–100 ° C for 1–3 hours has been often used to
produce the diether and tetraether side-chains.
Neutral lipids Polar lipids Identification of the hydrocarbon type and chain length
of diether and tetraether lipid cores is performed by
Acid methanolysis cleavage of the ether linkages with BCll3 , BBr3 , or HI.
removal of other lipids if needed
The ether lipid is made to react with the reagent for
Ether lipid cores 4 to 12 hours, typically at 100 ° C (69). The alkyl halides
or derived acetates produced by treatment with glacial
Chemical degradation acetic acid in an excess of silver nitrate or hydrocarbons
and/or direct derivitization produced by treatment with LiAlH4 , are then analyzed
by GC. The analyses of ether lipids by these cleavage
Analysis by GC/GC-MS, HPLC, SFC or other technique techniques are essential for their structural elucidation.
Figure 2. Flow diagram for the preparation and separation of However, such analyses involve many steps and generally
total lipids from bacteria, Archaea, and environmental samples. result in low recoveries and therefore have been deemed
as not appropriate for routine or quantitative analysis of
environmental samples (70).

Table 2. Listing of Selected Abbreviations Used with

Archaeal Components and Detection Methods Thin Layer Chromatography (TLC)

Abbreviation Description Preparative TLC has been used to fractionate both polar
lipids (e.g., individual phospholipids) and ether lipid cores
(diether and tetraether). For separation of the polar
Chemical Structures lipids, the chromatography is generally by silica gel, and
mixtures of chloroform/methanol/acetic acid/water (e.g.,
DE Diether
MDE Macrocyclic diether 85/30/15/5, v/v) (71) are used as the solvent system. In two-
OH Hydroxy-containing structure dimensional developments, ammonia replaces the acetic
PLEL Phospholipid ether lipid acid in the first development (71). Following preparation
TE Tetraether of the ether lipid cores by acid methanolysis, TLC has also
been widely used to separate the ether lipid cores. Again,
Instrumental and Other Methods Used to Separate
silica gel is the standard plate used with solvent systems
Archaeal components
composed of light petrol or hexane/ethyl ether/acetic acid
DGGE Denaturing gradient gel electrophoresis mixtures (e.g., 50/50/1, v/v; 70/30/1, v/v) (71).
FISH Fluorescent in situ hybridization A simple TLC procedure has been used to resolve
FTIR Fourier transform infrared spectroscopy mixtures of tetraether lipid cores containing up to eight
GC Gas chromatography cyclopentane rings (72). Silica gel F254 was used with
GC–MS Gas chromatography — mass spectrometry hexane/ethyl acetate (7/3, v/v) as the solvent system.
HPLC High performance liquid chromatography For most TLC systems, components are qualitatively
MPN Most probable number observed by using iodine vapor or dichlorofluoroscein
PCR Polymerase chain reaction
spray (all lipids), or more selective sprays (e.g., ninhydrin,
rRNA Ribosomal ribonucleic acid
SFC Supercritical fluid chromatography
amino groups; acid molybdate, phospholipids; α-naphthol,
TLC-FID Thin layer chromatography glycolipids).
A TLC–flame ionization detector (TLC–FID) analyzer
has been used to quantitatively analyze lipid mixtures
from a range of sample matrices (73). The analyzer uses
silica-gel coated glass rods with conventional development
small samples are available. This procedure was used of the rods in solvent tanks for component separation.
with samples containing bacterial lipid (i.e., phospholipid Detection of the separated lipid classes is by the sensitive
fatty acids, PLFA), and is yet to be applied to archaeal- FID; sensitivity is within the submicrogram to the 10-mcg
containing samples. range. The TLC-FID analyzer has not, to our knowledge,
Total lipids are separated into lipid classes (neutral been used for the separation of ether lipid cores. This
lipid, glycolipid, and phospholipid) by silicic acid column technique may be useful for both qualitative screening
chromatography (68), acetone precipitation, or other and quantitative analysis of both archaeal isolates and
procedures. TLC can also be used to separate individual environmental samples, including determination of the
components (see following section). concentrations of diether and tetraether.
252 ARCHAEA: DETECTION METHODS

Issues with Extraction and Other Processing Procedures methanogens (85). The diether core lipids were initially
obtained by acid methanolysis. The p-nitrobenzoyl ester
A few studies have reported the occurrence of unsatu-
of the diether core lipid was separated using a silica
rated diethers in Archaea. The halophilic Halobacterium
column and gradient elution (hexane to hexane-diethyl
lacusprofundi isolated from an Antarctic lake was first
ether, 80/20, v/v) with detection at 254 nm. The tetraether
reported to contain unsaturated diether lipid (74). Based
core lipids were not separated using this system.
on this finding, Methanococcoides burtonii was subse-
A further HPLC separation of the diether core lipids
quently noted to also contain unsaturated diethers (75).
was obtained using 9-anthroyl derivatives (82,86). The
Subsequently, it was also reported that novel 3-OH
derivatives were formed using 9-anthroyl nitrile with
hydroxydiether lipids (hydroxyarchaeol) were present in
purification needed. Separation was then obtained using
a number of methanogens (76,77), and that the standard
a C8 -bonded silica column employing a spectrofluoro-
strong acid methanolysis treatment (0.6–2 N methanolic
metric detector. The solvent was an isocratic acetoni-
HCl) of these lipids produced various products including a
trile/isopropanol mixture (80/20 v/v). The technique was
methoxy derivative, 3-mono-O-phytanyl-sn-glycerol (78),
also capable of separating the 3-OH diether core lipids.
and an unsaturated diether lipid (79). This finding was
The HPLC systems noted above only separate the
confirmed in further studies examining the ether lipid
diether core lipids. Several other procedures have been
composition of M. burtonii in which it was noted that the
developed that enable simultaneous separation of both
unsaturated ether lipids were most likely produced during
the diether and tetraether core lipids. The first HPLC
sample preparation (80,81). Based on these findings, we
system developed to separate both ether lipids used an
believe that the identification of the unsaturated diether
amino silica column, with an isocratic solvent, hexane/
in H. lacusprofundi may also be doubtful.
n-propanol (87). Unlike the HPLC systems that separate
Production of the artifacts from the 3-OH hydroxyether
the diether core lipids alone with detection of derivatized
lipids can be avoided by use of milder acid methanolysis
components, this system employed refractive index (RI)
conditions. The core ether lipids may be obtained
detection of underivatized diether and tetraether core
without degradation, and in good yield, by mild acid
lipids. An advantage of this system was an absence
methanolysis in 0.18% HCl (0.05 N) in methanol (71,79)
of impurities derived from the derivatization procedure.
or in 5% methanolic HCL/chloroform (1 : 27, v/v) at 50 ° C
However, the RI detector is prone to baseline drift
for 24 hours (82).
restricting the analysis to use of an isocratic solvent.
Low recovery of ether lipids using the Bligh and
The sensitivity of RI detection is also lower than
Dyer procedure has been noted for Methanobacterium
for other common HPLC detectors. This system was
thermoautotrophicum and other methanogenic species
shown to also separate the macrocyclic diether derived
containing highly polar and acidic lipids (82,83). These
from Methanococcus jannaschii, a thermophilic deep-
studies demonstrated that recovery of the ether lipids was
sea methanogen. Separation of the diether, tetraether,
increased by first extracting with the standard Bligh and
and macrocyclic diether obtained using this system is
Dyer procedure and then rapidly with a mildly acidic
illustrated in Figure 3. Component identification was
Bligh-Dyer one-phase solvent, chloroform/methanol/5%
further confirmed by FT–IR (DRIFT mode) analysis
aqueous trichloroacetic acid (1/2/0.8, v/v/v). This method
of HPLC fractions. The various ether lipid cores were
was also applied to environmental samples containing
differentiated by Fourier self-deconvolution of infrared
methanogens. The total lipid in the chloroform layer is
spectra. This HPLC technique has been applied in
washed using methanol and water to remove any traces
environmental studies to measure methanogenic biomass
of acid.
in an Antarctic lake and in digestors (87–89).
Compared with the considerable literature available
A second HPLC system capable of separating the
on PLFA abundance in bacteria, only limited data are
diether and tetraether core lipids has been reported (90).
available for phospholipid ether lipid (PLEL) concentra-
The system comprised an Inertsil ODS-2 column and
tions in Archaea. Conversion factors are routinely used to
UV detection at 254 nm of dinitrobenzoyl (DNB) deriva-
translate PLFA concentration results into either number
tives of the core ether lipids following gradient elution
of cells/g or number of cells/L; the average Bacterium is
(acetonitrile/2-propanol, 40/60–20/80, v/v). The technique
generally assumed to contain 100 µmoles of PLFA/g cells
(dry weight) (64,84). However, the as-yet-limited quantita- claimed to have greater sensitivity than the first HPLC
system developed (87). However, based on the chro-
tive data available for Archaea has shown large variations
between species. Therefore, care may now be needed in matograms presented, the diether was not adequately
the use of PLEL concentration data and conversion factors resolved from the internal standard (1,2-di-O-hexadecyl-
as is routinely performed with PLFA data. Notwithstand- rac-glycerol), and impurities introduced by the derivatiza-
ing this qualification, PLEL analysis will still prove to tion procedure tailed into both the internal standard and
be a reliable tool for determining and comparing viable diether core lipid peaks. Furthermore, the preparation of
archaeal populations in environmental samples. DNB derivatives involved three additional steps, and veri-
fication by TLC of derivatization was also required. Several
of the advantages claimed for this HLPC procedure have
High Performance Liquid Chromatography (HPLC)
to be weighed against possible disadvantages.
Several research groups have used HPLC procedures to Considerable advances have occurred in recent
separate the diether and tetraether core lipids. HPLC was years with evaporative light-scattering detectors (ELSD),
first used to separate the diether core lipid derived from including application with a range of lipid classes (91). To
 ARCHAEA: DETECTION METHODS 253

GE std alkyl chlorides, include use of a nonpolar column (e.g.,

SE30, OV17, HP1) with the GC oven initially run
isothermally, followed by temperature programming to
between 300 and 340 ° C (92–94). Detection is usually by
flame ionization detection (FID) or GC–mass spectrometry
(GC–MS).
A number of researchers have analyzed diether core
lipids, derived from cultured Archaea and environmental
samples, by GC. In the environmental samples, measure-
ment of the concentration of diether core lipids has been
used to estimate the biomass of methanogens (69,95) and
other Archaea (96). The protocol usually involves solvent
extraction of the total lipid, chemical breakdown of the
MDE complex lipids to produce the diether lipid core, derivati-
zation, purification by TLC (this step may be excluded),
and analysis by capillary GC. Measurement of diether
core lipids is related to cell numbers, methane pro-
duction and turbidity (69), and to production of organic
matter (95).
Conditions for the GC analysis of the diether core
lipids are similar to those noted earlier for the C40 hydro-
carbon. In particular, a high final oven temperature is
required (e.g., 300–340 ° C) to ensure that the diether
DE core lipids elute from the column. A number of inter-
nal standards have been employed in diether core lipid
TE analysis. These are generally phytanyl acetate prepared
from phytol (69) and 1,2-di-O-hexadecyl-rac-glycerol (as
the O-trimethylsilyl ether), following treatment with N,O-
bis(trimethylsilyl)trifluoroacetamide(BSTFA) (96). An is-
sue with the analysis of the 3-OH ether core lipids
0 30 is their incomplete derivatization when using sily-
Retention time (minutes) lating reagents such as BSTFA. Complete derivati-
Figure 3. HPLC chromatogram showing separation of the ether zation of hydroxyarchaeol to the di-TMSi derivative
lipid cores derived from the glycolipids of Methanococcus occurs on using BSTFA in pyridine/dichloromethane (1/2
jannaschii. Solvent, hexane/isopropanol, 99/1 v/v. Abbreviations: v/v) (97).
Std, internal standard; DE, diether; MDE, macrocyclic ether; TE, For cases in which GC has been used to determine
tetraether.
diether core lipid composition and content, the diethers
have been usually purified first. This removes the fatty
our knowledge, the use of HPLC with ELSD has not been acids that, in environmental samples, are derived from
applied to ether lipid analysis. Such an application has bacteria, eukaryotes, and other sources. A capillary GC
potential for routine analysis of ether lipids and is worthy procedure capable of simultaneous estimation of microbial
of investigation. Similarly, it would be particularly useful phospholipid-derived fatty acids and diethers was recently
if HPLC procedures capable of resolving both diether and reported (98) and applied to the microbial consortia of
tetra ether core lipids could be routinely interfaced to mass wetland sediments. See Figure 4 (99). The fatty acids (as
spectrometers to further facilitate the identification of the methyl esters) and diethers were not separated before
tetra ether core lipids in particular. With recent develop- analysis. In particular, a slow oven temperature-heating
ments in HPLC-MS, it is foreseen that such analyses will rate (from 110–220 ° C at 1 ° C/minute) was employed to
be increasingly used by researchers. gain separation of the novel signature lipids, C16 and C18
monounsaturated fatty acids derived from methanotrophs
Gas Chromatography (GC) (e.g., 16 : 1ω6c, 16 : 1ω8c, 16 : 1ω5c & t, 18 : 1ω8c), in
GC is the most widely used instrument employed by the same chromatogram as the more conventional C16
lipid researchers for the analysis of fatty acids and and C18 monounsaturated fatty acids and the diether
other lipid classes, including hydrocarbons and sterols. lipid core. The slower oven-heating rate was used
Many studies on ether lipids have used GC analysis to achieve separation of novel fatty acids that are
to measure the C20 and C40 hydrocarbon or other generally not resolved when using more standard oven-
products derived from the ether lipid alkyl chains heating programs (e.g., from 150 ° C to 250 ° C at 2 or
obtained using chemical degradation procedures. Typical 3 ° C/minutes).
operating conditions for analysis of the C20 and C40 Archaeal diether and tetraether lipid cores were
hydrocarbon and other products, such as alcohols or analyzed simultaneously by high-temperature capillary
254 ARCHAEA: DETECTION METHODS

C16 fatty acids

C19 fatty acids methyl ester std

C18 fatty acids

GE std

DE

50 60 70 80 90 100 110 120 130 140 150

Retention time (minutes)
Figure 4. Partial gas chromatogram (HP1) showing phospholipid fatty acids (as methyl ester) and
diether lipid (as OTMSi ether) from Billabong sediments. Abbreviations: GE std, glyceryl ether
internal standard; DE, diether. Adapted from P. Virtue, P. D. Nichols and P. I. Boon, J.Microbial.
Methods 25, 177–185 (1996).

GC (80). The high-temperature GC achieved sep-

aration of diether, tetraethers, and cyclopentane-
containing tetraethers. A short BPX5-fused silica capillary
column (3 m × 0.25 mm) and on-column injector were used
for the analyses. Initial research on high-temperature
GC separation of diether and tetraether core lipids used
novel Aluminum-clad capillary columns. Separation was TE
obtained, but the columns were brittle and their use was
not pursued. With other columns tested, loss of tetraether
was observed; this was most probably due to the break-
down of this component by active sites on the column. The
use of the BPX5 phase on a short capillary column over-
came these problems and the tetraether recovery remained DE
stable for continual running of the column over a period of
several months. The GC oven was temperature-programed
from 90 ° C to 190 ° C at 30° /minutes and then to 380 ° C at 0 4 8 12 16 20 24 28 32 36 40
10 ° C/minutes. The final temperature was maintained for Retention time (minutes)
15 minutes. Using FID, the instrumental limit of detec- Figure 5. Capillary gas chromatogram (BPX5) showing phospho-
tion for individual ether lipids was approximately 1–2 ng. lipid-derived ether lipid distribution of Methanobacterium ther-
Representative GC traces illustrating the separation of moautotrophicum strain Hverigerdi. Abbreviations: DE, diether;
ether lipids obtained using the high-temperature capillary TE, tetraether. Components eluting between four and six min-
column are shown in Figures 5 and 6. utes are fatty acid methyl esters. Adapted from P. D. Nichols,
P. M. Shaw, C. A. Mancuso and P. D. Franzmann, J. Microbiol.
It was recommended for high-temperature GC analysis
Methods 18, 1–9 (1993).
of samples containing both di- and tetraether core lipids
that appropriate standard mixtures containing known
proportions of the two ether lipid cores be run routinely.
Alternately, internal standards that elute close to the di- high-temperature GC procedure was at least comparable
and tetraether core lipids can be used to assess recoveries. to and generally better than other techniques in use,
GC is widely used in many research laboratories. and that these features provide new possibilities for
Therefore, the use of high-temperature capillary GC for measuring Archaea in industrial projects (80). It would
measurement of Archaeal diether and tetraether lipids be particularly useful if the high-temperature GC
provides considerable potential for routine analysis of procedure could be interfaced to a mass spectrometer
these microbial lipids in taxonomic, microbial, ecological, to further facilitate the identification of tetra ether core
and organic geochemical studies. It was reported that the lipids.
 ARCHAEA: DETECTION METHODS 255

Recent developments with GC analysis include multi-

dimensional GC and cryo-focusing. Both procedures can
be used to enhance component resolution. Should these
and other techniques become more widely used, they
could be readily applied to obtain increased resolution
cyp TE of the various ether lipid components when analyzed
TE by GC.

Supercritical Fluid Chromatography

Several researchers have used supercritical fluid chro-
matography (SFC) to analyze archaeal lipids. SFC is
a separation technique similar to GC and HPLC. GC
and HPLC employ a gas and liquid phase respectively
as the mobile phase, but a compressed gas is used in
SFC; carbon dioxide is the gas employed in most SFC
0 4 8 12 16 20 24 28 32 36 40 separations. It has been stated that the versatility of
Rentention time (minutes) SFC and its range of applications is directly related to
Figure 6. Capillary gas chromatogram (BPX5) showing ether the unique physicochemical properties of the supercriti-
lipid distribution of Thermoplasma acidophilum. Abbreviations: cal fluid.
TE, tetraether; cypTE, cyclopentane-containing tetraethers. Com- The first report of SFC applied to ether lipids
ponents eluting between 4 and 10 minutes include phytadienes demonstrated separation of a range of tetraethers from
and monophytanyl glyceryl diols. Adapted from P. D. Nichols, Thermoplasma and also tetraether-derived isopranyl
P. M. Shaw, C. A. Mancuso and P. D. Franzmann, J. Microbiol. hydrocarbons (101). Complete separation of the tetraether
Methods 18, 1–9 (1993).
core lipids that differed only by the number of cyclopentane
rings present in the isopranyl side chains was achieved.
With the phased used, the tetraether core lipids eluted
In a specific application, highly sensitive gas chro- in order of increasing cyclopentane content. The method
matography combined with negative chemical ionization was suitable for qualitative- and quantitative-screening
mass spectrometry (with selected ion-monitoring) enabled for ether lipids in living organisms and sedimentary
the detection, for the first time, of C15 and C25 homologs organic matter.
of the normal C20 isopranyl side chain in several strains A strategy was developed for use of SFC to separately
of methanogens (100). These methods would be applica- analyze the polar lipid, glycolipid, and lipid-extracted
ble to field samples for determining archaeal community residue fractions, and the approach was applied to
structure. seven strains of Archaea (70). The methodology was
Mass spectral data is often not reported for studies that suitable for integration with a protocol for the analysis
examine the distribution of core ether lipids in Archaea of bacterial lipids. Currently, the SFC methodology does
and environmental samples. The mass spectra of the not have sufficient resolution for the analysis of the more
diether lipid core (as the trimethyl silyl ether) derived complex fatty acid mixtures. The limited literature results
from the methanogen, Methanoccoides burtonii, is shown available to date for SFC assay of individual strains of
in Figure 7. Further details on formation of the major Archaea show that separation of individual ether lipid
mass fragments observed have been reported (96). structures is comparable to other chromatography (GC

130
100 ×10
Ion Abundance

50 620
412

Figure 7. Mass spectrum of the diether

426 lipid (as the trimethyl silyl ether) derived
369
278 from the methanogen Methanoccoides
309 724 burtonii. Instrument: VG Autospec
445 Ultima mass spectrometer. Mass spec-
504
0
trum kindly provided by Dr. Noel Davies
200 300 400 500 600 700 800 and Ms. Jenny Skerratt, University of
m/z Tasmania, Hobart, Tasmania, Australia.
256 ARCHAEA: DETECTION METHODS

and HPLC) procedures. However, the use of SFC has genera Acidianus, Desulfurolobus, Metallosphaera, Sty-
generally not been widespread and, to our knowledge, the giolobus and Sulfolobus are, with few exceptions, based
technique is not used in many laboratories performing on the benzothiophene nucleus, whereas members of the
microbial lipid assays. genera Archaeoglobus, Thermoplasma, and Thermopro-
teus contain only unsaturated or fully saturated isoprenoid
Other Signature Lipid Procedures naphthoquinones. The application of quinone methodology
Carbon isotopic fractionation has been used to deter- to environmental samples for measuring Archaea remains
mine carbon sources, including the contribution from to be explored.
methanogens, in environmental samples. It is gener-
ally considered that methanogens produce biomarkers, OTHER PROCEDURES
including ether lipids and pentamethyleicosane (PME)
that are 13 C-depleted compared to their carbon source. Methane Production
The advent of GC–isotope ratio mass spectrometry
(GC–IRMS) as a research tool now permits more rou- Several studies that have used ether lipids or other
tine examination of the isotopic composition and frac- techniques to estimate methanogenic biomass have also
tionation for cultured microorganisms and environmental measured methane. In marine sediments (86), a direct
samples. In a recent study, the finding of highly 13 C- correlation was found between methane concentration
depleted ether lipid and polyisoprenoid moieties, compared measured by a headspace procedure (104) and diether
to the 13 C-content of biomass of cultured methanogens core lipid measured by HPLC.
suggests that there is significant isotopic fractionation In water samples, methane is measured by purge
inherent in the lipid biosynthetic pathways of some and trap procedures. Such techniques have been applied
Archaea (81). to ocean waters (105). The increased recognition of the
Other instrumental procedures have been used in presence of methanogens in oxygenated surface oceanic
the structural elucidation of ether lipids. These include waters has occurred over the past decade. This observation
NMR (both 13 C and 1 H), probe MS, and fast-atom bom- was even termed the ocean methane paradox (106).
bardment MS (FAB MS). These procedures have been Methane measurements in northern Pacific Ocean waters
invaluable for the determination of precise structural showed that methane was supersaturated with respect
details (102). For example, in a report on the extremely to its equilibration with atmospheric methane (107). To
halophilic Archaea Halobacterium cutirubrum, FAB (tan- gain further insight into this interesting phenomenon,
dem) mass spectrometry was used to analyze intact application of the signature lipid approach in combination
polar ether lipids (103). Simple positive and negative ion with methane measurements and molecular probes would
mass spectra of the three major polar lipids extracted be extremely fruitful.
from H. cutirubrum were obtained that contained ions The direct measurement of methane production
with masses corresponding to cationized or deproto- rates can be used to detect archaeal activity. This
nated lipid molecules. Structural information was then is achieved by measurement of 14 CH4 formed from
obtained on individual components of the polar lipid NaH14 CO3 , [14 C] — formate and [14 C-2] — acetate in time-
mixtures by discrete selection of ion masses when the course experiments using the gas chromatography–gas
instrument was operated in tandem mass spectromet- proportional counting (GC–GPC) procedure (108). In an
ric mode. application of this method in a unique Antarctic lake (109),
These procedures are generally not routinely used to the sensitivity of the method was demonstrated and a
detect Archaea in environmental studies. However, their maximal methane production of 2.5 µmol kg−1 day−1 was
use is noted here, although they are not covered further estimated.
in this section. Further information on many of these A signature lipid and radiotracer analysis approach
methods is available (102). has been used to determine anaerobic conversion of
biomass to methane (110). Methane was measured by
Quinones GC — GPC, and both PLFA and PLEL (the latter by SFC)
were measured to determine the bacterial and archaeal
The structure and distribution of the various types of iso-
biomass, respectively, as well as the community structure.
prenoid quinones in Archaea has been reviewed (102).
In addition, radiolabel incorporation into PLFA and PLEL
Isoprenoid quinones are recovered from the neutral
fractions was also determined.
lipid fraction of total lipid extracts obtained from aer-
obic Archaea (Fig. 2) and their structures are based on
CoEnzymes
naphthoquinone or benzothiophenquinone chromophores.
Separation of components has been performed by Methanogens — Coenzyme F420 . Coenzyme F420 is the
TLC and HPLC procedures, with structural elucida- N-(NL-lactyl-γ -L-glutamyl)-L-glutamic acid phosphodi-
tion by NMR, MS, UV, and visible spectroscopy. Stud- ester of 7,8-didemethyl-8-hydroxy-5-deazariboflavin 5 -
ies of archaeal quinones have mainly focused on ther- phosphate (111). The enzyme functions as an electron
mophilic and halophilic strains; detailed studies of the carrier in methanogens for anaerobic respiration and cell
quinone composition of methanogens are yet to be per- carbon synthesis (112). On exposure to UV light, fluores-
formed (102). cence emitted by this molecule allows for tentative ID
Thermophilic Archaea are divided into two groups with as either coenzyme F420 or its analog has been found in
respect to their quinone composition. Members of the all methanogens analyzed to date (113). After purification
 ARCHAEA: DETECTION METHODS 257

by HPLC (114), quantification of the coenzyme has been comments on the draft manuscript. We sincerely thank CSIRO
used to measure methanogenic biomass in mixed sys- and Antarctic CRC colleagues, including Prof. Tom McMeekin,
tems (115). However, recent studies have shown that the Mark Rayner, and Jenny Skerratt, for their assistance and
levels of this cofactor and others varied significantly among support.
methanogens as well as within individual methanogens,
according to growth substrate (113).
BIBLIOGRAPHY
Fourier Transform–Infrared (FT–IR) Spectroscopy 1. C. R. Woese and G. E. Fox, Proc. Natl. Acad. Sci. U.S.A. 74,
FT–IR spectroscopy has been used by microbial ecolo- 5088–5090 (1977).
gists to examine microbial biomass and nutritional sta- 2. C. R. Woese, Sci. Am. 244, 98–102 (1981).
tus, including exopolymer production (116,117). Use of 3. G. E. Fox et al., Science 209, 457–463 (1980).
the FT–IR technique offers the potential for rapid and 4. C. R. Woese, O. Kandler, and M. L. Wheelis, Nature 351,
nondestructive analysis of bacteria and bacteria-polymer 528–529 (1991).
mixtures. The protein-derived amide I band at 1,680 cm−1 5. C. R. Woese, O. Kandler, and M. L. Wheelis, Proc. Nat.
defines biomass, and bands at 1,740 cm−1 — derived Acad. Sci. U.S.A. 87, 4576–4579 (1990).
from the storage product polyhydroxybutyrate — and 6. J. G. Zeikus, Bacteriol. Rev. 41, 514–541 (1977).
1,150 cm−1 are indicative of exopolymer-containing car- 7. D. J. Kushner, in R. S. Wolfe and R. C. Woese, eds., The
bohydrate. Bacteria, Academic Press, New York, 1985, pp. 171–214.
The ratio of the various IR bands provides a 8. K. O. Stetter and W. Zillig in R. S. Wolfe and R. C. Woese,
comparative tool for use in environmental studies. eds., The Bacteria, Academic Press, New York, 1985,
Differences in IR spectra for a range of bacteria have pp. 85–170.
also been reported, suggesting that the FT–IR technique 9. O. Kandler and H. Konnig, in R. S. Wolfe and R. C. Woese,
may be capable of providing information on community eds., The Bacteria, Academic Press, New York, 1985,
structure. This possible application of the technique is pp. 413–457.
certainly less developed than its use in examining biofilms 10. T. A. Langworthy, T. G. Tornabene, and G. Holzer, Zbl.
of the key constituents just noted. Bakt. Hyg., I. Abt. Orig. C3, 228–244 (1982).
To our knowledge, FT–IR has not been used to directly 11. N. Best, J. Bacteriol. 133, 240 (1978).
examine members of the Archaea. Given the promising 12. W. Zillig, R. Schnabel, J. Tu, and K. O. Stetter, Naturwiss.
findings to date for bacterial communities and noting the 69, 197–204 (1982).
distinctive biochemical properties of Archaea with respect 13. A. Oren, FEMS Microbiol. Ecol. 73, 187–192 (1990a).
to their lipid and cell envelope composition, use of this
14. A. Oren, J. Microbiol. Meth. 10, 183–187 (1989).
instrumental approach may offer potential for examining
15. A. Oren, FEMS Microbiol. Ecol. 73, 41–48 (1990b).
this unique kingdom.
16. R. A. Aravalli, Q. She, and R. A. Garrett, Trends Ecol. Evol.
13, 190–194 (1998).
CONCLUSION 17. L. Koch, in P. Gerhardt, R. E. Murray, W. A. Wood, and
N. R. Krieg, eds., Methods for General and Molecular Bac-
The routine measurement of Archaea in taxonomic teriology, American Society for Microbiology, Washington,
and environmental studies is now possible, based on DC, 1994, pp. 257–260.
developments in the past two decades. In environmental 18. C. Joulian, B. Ollivier, B. K. C. Patel, and P. A. Roger,
samples, in particular, the use of modern molecular FEMS Microbiol. Ecol. 25, 135–145 (1998).
and signature lipid techniques enables the detection 19. S. Kotelnikova and K. Pedersen, FEMS Microbiol. Ecol. 26,
and quantification of these important microorganisms. 121–134 (1998).
The choice of equipment available to the researcher 20. M. Tokura, K. Ushida, K. Miyazaki, and Y. Kojima, FEMS
will influence the selection of a suitable protocol with Microbiol. Ecol. 22, 137–143 (1997).
the signature lipid approach. Further research on 21. H. W. Jannasch and G. E. Jones, Limnol Oceanogr. 4,
molecular techniques is necessary before such methods 128–139 (1959).
can become routine and quantitative. It is expected that 22. R. L. Ferguson, E. N. Buckley, and A. V. Palumbo, Appl.
further developments will undoubtedly occur with these Envir. Microbiol. 57, 49–55 (1984).
procedures and will increase their resolving power and 23. R. I. Amann, W. Ludwig, and K. H. Schleifer, Microbiol.
widen their application. For example, it should be possible Rev. 59, 143–169 (1995).
to detect methanogens during climate change and in other 24. E. F. DeLong, G. S. Wickham, and N. R. Pace, Science 243,
research programs in which modeling studies of the role of 1360–1363 (1989).
this key group will increasingly need validation via field 25. J. A. Fuhrman, K. McCallum, and A. A. Davis, Nature 356,
measurements. 148–149 (1992).
26. E. F. DeLong, K. Y. Wu, B. B. Prezelin, and R. V. M. Jovine,
Acknowledgments
Nature 371, 695–697 (1994).
We are indebted to Prof. D. C. White for his long-standing
encouragement and support in our pursuit of the development 27. J. A. Furhman and A. A. Davis, Mar. Ecol.: Prog. Ser. 150,
and application of methods for the measurement of Archaea in 275–285 (1997).
environmental samples. Drs. Peter Franzmann, John Bowman, 28. J. P. Bowman, S. A. McCammon, S. M. Rea, and T. A.
Peter Rothlisberg, and an anonymous referee provided helpful McMeekin, FEMS Microbiol. Let. 183, 81–88 (2000).
258 ARCHAEA: DETECTION METHODS

29. P. D. Franzmann et al., Sys. Appl. Microbiol. 11, 20–27 57. J. Steber and K. H. Schleifer, Arch. Microbiol. 105, 173–179
(1988). (1975).
30. A. Vetriani et al., Appl. Environ. Microbiol. 65, 4375–4384 58. J. H. Konig and K. Stetter, Zbl. Bakt. Hyg., I. Abt. Orig. C3
(1999). 478–490 (1982).
31. L.-A. Meyer-Reil, Appl. Environ. Microbiol. 36, 506–512 59. R. P. Eganhouse, in R. P. Eganhouse, ed., Molecular Mark-
(1978). ers in Environmental Geochemistry, American Chemical
32. R. J. Bobbie and D. C. White, Appl. Envir. Microbiol. 39, Society, Washington, DC, 1997, pp. 1–20.
1212–1222 (1980). 60. P. B. Comita, R. B. Gagosian, H. Pang, and C. E. Costello, J.
33. A. J. L. Macario et al., Appl. Envir. Microbiol. 55, 1996– Biol. Chem. 259, 15234–15241 (1984).
2001 (1989). 61. M. Nishihara and Y. Koga, Biochim. Biophys. Acta 1,082,
34. E. Conway-deMacario, A. J. L. Macario, T. Mok, and T. J. 211–217 (1991).
Beveridge, J. Bacteriol. 175, 3115–3120 (1993). 62. A. Sugai et al., Lipids 30, 339–344 (1995).
35. S. R. James, H. R. Burton, T. A. McMeekin, and C. A. Man- 63. M. Kates, in R. P. Eganhouse, ed., Molecular Markers in
cuso, Ant. Sci. 6, 325–332 (1994). Environmental Geochemistry, American Chemical Society,
36. S. J. Giovanonni, E. F. DeLong, G. J. Olsen, and N. R. Pace, Washington, DC, 1997, pp. 35–48.
J. Bacteriol. 170, 720 (1988). 64. D. C. White et al., in R. P. Eganhouse, ed., Molecular Mark-
37. R. Huber et al., Nature 376, 57–58 (1995). ers in Environmental Geochemistry, American Chemical
38. M. J. Pelczar, E. C. S. Chan, and N. R. Krieg, in M. J. Society, Washington, DC, 1997, pp. 22–34.
Pelczar, E. C. S. Chan, and N. R. Krieg, eds., Microbiology 65. D. C. White et al., Oecologia 40, 51–62 (1979a).
Concepts and Applications, McGraw-Hill, Inc., Sydney, 66. D. L. Balkwill et al., Microb. Ecol. 19, 73–84 (1988).
Australia, 1994, p. 120. 67. S. R. Kehrmeyer et al., J. Microbiol. Methods 25, 153–163
39. O. Kandler and H. Hippe, Arch. Microbiol. 113, 57–60 (1996).
(1977). 68. J. B. Guckert, C. P. Antworth, P. D. Nichols, and D. C.
40. O. Kandler and H. Konig, Arch. Microbiol. 118, 141–152 White, FEMS Microbiol. Ecol. 31, 147–158 (1985).
(1978). 69. G. C. Smith and G. D. Floodgate, Continental Shelf Research
41. H. Konig and O. Kandler, Arch. Microbiol. 121, 271–275 12, 1187–1196 (1996).
(1979). 70. D. B. Hedrick, J. B. Guckert, and D. C. White, J. Lipid Res.
42. O. Kandler and H. Konig, in R. S. Wolfe and R. C. Woese, 32, 659–666 (1991).
eds., The Bacteria, Academic Press, New York, 1985, 71. M. Nishimura and Y. Koga, J. Biochem 101, 997–1005
pp. 413–457. (1987).
43. G. Darland, T. D. Brock, W. Samsonoff, and S. F. Conti,
72. A. Tricone et al., J. Gen. Microbiol. 134, 3159–3163 (1988).
Science 170, 1416–1418 (1970).
73. J. K. Volkman and P. D. Nichols, J. Planar Chromatogr.-
44. H. Konig, in M. Goodfellow and A. G. O’Donnell, eds.,
Mod. TLC 4, 19–26 (1991).
Chemical Methods in Prokaryotic Systematics, John Wiley
& Sons, New York, 1994, pp. 85–119. 74. P. D. Franzmann et al., Syst. Appl. Microbiol. 11, 20–27
(1988).
45. H. Konig, R. Kralik, and O. Kandler, Zbl. Bakt. Hyg., I. Abt.
Orig. C3, 179–191 (1982). 75. P. D. Nichols and P. D. Franzmann, FEMS Microbiol. Lett.
98, 205–208 (1992).
46. K. H. Schleifer and O. Kandler, Arch. Microbiol. 57,
335–364 (1972). 76. G. Ferrante, I. Ekiel, G. B. Patel, and G. D. Sprott, Biochim.
Biophys. Acta 963, 173–182 (1988).
47. K. H. Schleifer in G. Gottschalk, ed., Methods in Microbiol-
ogy, Academic Press, New York, 1985, pp. 123–156. 77. G. Ferrante et al., J. Lipid Res. 30, 1601–1609 (1989).
48. P. Albersheim, D. J. Nevins, P. D. Englich, and A. Karr, 78. G. D. Sprott, I. Ekiel, and C. Dicaire, J. Biol. Chem. 265,
Carbohydrate Res. 5, 340–345 (1967). 13735–13740 (1990).
49. H. Konig and K. Stetter, Syst. Appl. Microbiol. 7, 300–309 79. I. Ekiel and G. D. Sprott, Can. J. Microbiol. 38, 764–768
(1986). (1992).
50. E. G. Bligh and W. J. Dyer, Can. J. Biochem. Physiol. 35, 80. P. D. Nichols, P. M. Shaw, C. A. Mancuso, and P. D. Franz-
911–917 (1959). mann, J. Microbiol. Methods 18, 1–9 (1993).
51. G. Peterson, Anal. Biochem. 83, 346–356 (1972). 81. R. E. Summons, P. D. Franzmann, and P. D. Nichols, Org.
Geochem. 28, 465–475 (1998).
52. F. Wieland, J. Lechner, and M. Sumper, Zbl. Bakt. Hyg., I.
Abt. Orig. C3, 161–170 (1982). 82. S. Ohtsubo et al., FEMS Microbiol. Ecol. 12, 39–50 (1993).
53. J. Heukeshoven and R. Dernick, J. Chromatogr. 326, 83. M. Nishihara and Y. Koga, J. Biochem. 101, 997–1005
91–101 (1985). (1987).
54. E. Nußer et al., in U. B. Sleytr, P. Messner, D. Pum, and 84. D. C. White et al., in C. D. Litchfield and P. C. Seyfried, eds.,
M. Sara, eds., Proceedings of the Second International Methodology for Biomass Determinations and Microbial
Workshop on S-Layers in Prokaryotes, Vienna, Spinger Biomass in Sediments, ASTMSTP 63, American Society for
Verlag, Berlin, Germany, 1988, pp. 21–25. Testing and Materials, Washington, DC, 1979b, pp. 87–103.
55. H. Michel, D. C. Neugebauer, and D. Osterhelt, in W. 85. R. F. Martz, D. L. Sebacher, and D. C. White, J. Microbiol.
Baumeister and W. Vogell, eds., Electron Microscopy at Methods 1, 53–61 (1983).
Molecular Dimensions, Spinger Verlag, Berlin, Germany, 86. M. Sunamura, Y. Koga, and K. Ohwada, Mar. Biotech. 1,
1980, pp. 27–35. 562–568 (1999).
56. P. Kreisl and O. Kandler, Syst. Appl. Microbiol. 7, 293–299 87. C. A. Mancuso, P. D. Nichols, and D. C. White, J. Lipid Res.
(1986). 27, 49–56 (1986).
 ARCHAEA IN BIOTECHNOLOGY 259

88. C. A. Mancuso, P. D. Franzmann, H. R. Burton, and P. D. ARCHAEA IN BIOTECHNOLOGY

Nichols, Microb. Ecol. 19, 73–95 (1990).
89. P. D. Nichols, C. A. Mancuso, and D. C. White, Org. Geo- D. A. COWAN
chem. 11, 351–361 (1987). Department of Microbiology
90. K. Demizu et al., J. Ferment. Bioeng. 73, 135–139 (1992). Cape Town, South Africa
91. W. W. Christie, in W. W. Christie ed., Advances in Lipid S. G. BURTON
Methodology — One, The Oil Press, Glasgow, U.K., 1992, Department of Biochemistry
pp. 239–271. and Microbiology
92. T. A. Langworthy, P. F. Smith, and W. R. Mayberry, J. Grahamstown, South Africa
Bacteriol. 112, 1193–1200 (1972).
93. M. de Rosa, S. de Rosa, A. Gamborta, and J. D. Bu’Lock, The Archaea represent one of the three fundamental
Phytochemistry 16, 1961–1965 (1977). domains of life. Genotypically, they are as distinct from
94. M. de Rosa, S. de Rosa, A. Gamborta, and J. D. Bu’Lock, Bacteria because they are from higher organisms (the
Phytochemistry 19, 249–254 (1980). Eukarya). The domain Archaea is composed of at least
95. G. G. Pauly and E. S. Van Vleet, Geochim. Cosmochim. Acta three phenotypically distinct groups, each dominated by
50, 1117–2225 (1986). members inhabiting extreme environments (high tem-
96. P. Teixidor and J. O. Grimalt, J. Chromatogr. 607, 253–259 perature/low pH, anoxia, hypersalinity). The unique bio-
(1992). chemistry of some members of the Archaea (such as the
97. K.-U. Hinrichs et al., Organic Geochem. 31, 1685–1702 methanogens), coupled with the molecular and physio-
(2000). logical adaptations of others to physical and chemical
98. P. Virtue, P. D. Nichols, and P. I. Boon, J. Microbiol. Meth- extremes, has endowed this group of microorganisms with
ods 25, 177–185 (1996). significant biotechnological potential. The use of enzymes
99. P. I. Boon, P. Virtue, and P. D. Nichols, Mar. Freshwater in diagnostics and biocatalysis, fermentations in methane
Res. 47, 27–41 (1996). production and metal ion recovery, and cellular bioprod-
ucts as food additives and separation membranes are all
100. C. A. Mancuso et al., J. Lipid Res. 26, 1120–1125 (1985).
examples of the enormous capacity for biotechnological
101. S. J. DeLuca, K. J. Voorhees, T. A. Langworthy, and G.
exploitation of this diverse group of organisms.
Holzer, J. High Resol. Chromatogr. Commun. 9, 182–185
(1986).
102. M. de Rosa and A. Gambacorta, in M. Goodfellow and ARCHAEAL PHYLOGENETICS
A. G. O’Donnell, eds., Chemical Methods in Prokaryotic
Systematics, John Wiley & Sons, Chichester, U.K., 1994, The Archaea comprise one of the three fundamental
pp. 197–264. groups (domains or superkingdoms) of living organisms
103. H. L. Fredrickson et al., Biomed. Environ. Mass Spectrom. (Fig. 1). This fundamental tripartite topology is derived
18, 96–105 (1989). from small subunit rRNA sequence comparisons (1) and
104. M. E. Linstrom, Limnol. Oceanogr. 28, 1247–1251 (1983). is rooted by the inclusion of pairs of paralogous genes
105. D. Karl and B. D. Tilbrook, Nature 368, 732–734 (1994). (related genes that duplicated and diverged before the
106. R. P. Kiene, in J. E. Rogers and W. B. Whitman, eds., three domains separated) as out-groups (2,3). The rooting
Microbial Production and Consumption of Greenhouse of the phylogenetic tree places the Archaea as a very
Gases: Methane, Nitrogen Oxides, and Halomethanes, Am. deep branch on the Eukaryotic lineage, linking these two
Soc. Microbiol., Washington, DC, 1991, pp. 111–146. more closely in evolutionary terms than the Bacteria
107. B. D. Tilbrook and D. Karl, Mar. Chem. 368, 732–734 and the Archaea. Notwithstanding the importance of
(1994). genetic transfer between members of the three domains,
108. D. R. Nelson and J. G. Zeikus, Appl. Microbiol. 28, 258–261 the close evolutionary relationship between the Archaea
(1974). and the Eukaryotes has considerable significance in
109. P. D. Franzmann et al., Hydrobiologia 210, 191–201 (1991). terms of Biotechnology, for example, in the potential
use of Archaea as primitive models for eukaryotic
110. D. B. Hedrick and D. C. White, Biomass Bioenergy 5,
247–259 (1993).
cells.
Until relatively recently the Archaeal domain has
111. L. D. Eirich, G. D. Vogels, and R. S. Wolfe, Biochemistry 17,
been clearly separated, on the basis of the genotypic
4583–4593 (1978).
relatedness of cultured organisms, into two kingdoms,
112. G. D. Vogels, J. T. Keltjens, T. J. Hutten, and C. Vander-
the Crenarchaeotes and Euryarchaeotes (Fig. 1). The
drift, Zbl. Bakteriol. Hyg. I. Abt. Orgi. C3, 258–264 (1982).
crenarchaeote kingdom was thought to be phenotypi-
113. G. M. Gorris and C. Vanderdrift, Biofactors 4, 139–145 cally homogenous and composed largely of the so-called
(1994).
sulfur-dependent thermoacidophiles (Table 1). The eur-
114. P. van Beelen, W. J. Geerts, A. Pol, and G. D. Vogels, Anal. yarchaeote kingdom was more phenotypically diverse and
Biochem. 131, 285–290 (1983a). included the methanogens, the extreme halophiles, and
115. P. van Beelen, A.C. Dijkstra, and G. D. Vogels, Eur. J. Appl. various thermophiles, including the fermentative sulfate
Microbiol. Biotechnol. 18, 67–69 (1983b). reducers. For in-depth discussions of the phylogeny of
116. P. D. Nichols et al., J. Microbiol. Methods 4, 79–94 (1985). the Archaea, the reader is directed to some excellent
117. W. R. Dade et al., J. Geomicrobiol. 8, 1–8 (1990). reviews (4,5).
260 ARCHAEA IN BIOTECHNOLOGY

ces
Planctomy
Bacteria

m
mitoch

r iu
c te
Rho

emp
Agrobacterium

ba
ondrion
Es

ter
doc chia vib

vo

low t
ch

Me
Fla
ibac

Methanococcus
Sy Ch s

y c lu
eri ulfo
bu

Ther
tha

a
De pla
ne lo
o

sm
Th Marine
gl

Flex
ch
ro
s st

no
oc o

pla
ae x

moc
a

ba
Chl G oc h fer

mo
amy loeo cus Me rc lo

ct e
rio A Ha

occu
tha

er
dia ba

riu
Chloro cte no
bium r irillum
the Methanosp

m
Me

s
Leptonema tha rm
nop us
Clostridium yru lobus
s Sulfotium
Bacillus dic
rium Pyro
b te cter
a c Thermoproteus
Helio a Therm
rob ofilum
Ar th
us us a Roo Archaea
f l ex herm otog t
r o T m
lo er x
Ch Th u ife
Aq
Co

Ho
pri

mo
nu
s

Zea 0.1 changes per site

Cryptomonas
a
Achlyria
s ra
t a
Co rphy
Babe ium
sia

Po Gi
c
me

ard
ia
tellum
ra
Pa

Dictyos

Entamoeba

Phy

Tr
Naegleria

icho
sar

m
Tryp

um

on
Euglen

Enc

Eucarya
as
Va
a

eph
noso

in
a

im
alito

or
ma

ph
zoo

a
n

Figure 1. The universal phylogenetic tree. (Reproduced with permission from pace [18]).

The clear-cut twin kingdom structure of the domain numerous low temperature marine and terrestrial derived
Archaea has been substantially blurred, over the last phylotypes (17,18) in the hitherto exclusively thermophilic
decade of the twentieth century, by the application of crenarchaeota. The discovery of several sequences that
molecular phylogenetics to microbial population analysis. branch very deeply in the crenarchaeote lineage has
The direct extraction of microbial community DNA, led some to propose a third archaeal kingdom [the
coupled with PCR amplification of 16S rRNA genes using ‘‘Korarchaeota’’ (19,20)] and others to speculate on the
universal and Archaea-specific primers (14), has led to generation of treeing artifacts.
a multitude of phylotypes positioned between the two
kingdoms (15,16), generating a more homogeneous single ARCHAEAL DIVERSITY
domain. Molecular phylogenetic analysis of community
DNA has also added apparently new phenotypic and The intense interest in all aspects of the Archaea has led to
genotypic groups to the domain. One of the startling the isolation and characterization of hundreds of different
revelations of this method has been the appearance of genera and species. Tables 1 to 3 summarize the current
Table 1. The Thermoacidophilic Archaea

Upper Growth
Order Genus Species Habitata Temperature ( ° C)

‘‘Archaeoglobales’’ Archaeoglobus A. fulgidus D,S,O,B 92

A. profundus D,S 92
A. veneficus D 85
Ferroglobus F. placidus S 95
‘‘Igneococcales’’ Desulfurococcus D. amylolyticus D,T 97
D. mobilis T 95
D. mucosus T 97
D. saccharovorans T 97
Hyperthermus H. butylicus S 108
Pyrodictium P. abyssi D 110
P. brockii D,S 110
P. occultum D,S 110
Staphylothermus S. marinus D 98
Sulfophobococcus S. zilligii T 95
Thermodiscus T. maritimus S 98
Thermosphaera T. aggregans T 90
Sulfolobales Acidianus A. ambivalens T 95
A. brierleyi T 75
A. infernus T 95
Metallasphaera M. sedula T 80
M. prunae T 80
Stygioglobus S. azoricus T 89
Sulfolobus S. acidocaldarius T 85
S. metallicus T 75
S. shibatae T 86
S. solfataricus T 87
Pyrolobus P. fumarii
Thermococcales Pyrococcus P. furiosus D,S 103
P. horikoshii D 102
P. kodakaraensis
P. woesii D,S 103
Thermococcus T. alcaliphilus S 90
T. acidaminovorans S 93
T. aggregans D
T. barossii D >88
T. celer S 93
T. chitonophagus D 93
T. fumicolans D >90
T. hydrothermalis D >85
T. gorgonarius D >88
T. guaymasensis D
T. littoralis S 98
T. pacificus D >88
T. peptonophilus D 85
T. profundus D 90
T. stetteri D,S 98
Thermoproteales Aeropryum A. pernix T 100
Pyrobaculum P. aerophilum T 104
P. islandicum T,S 103
P. organotrophicum T 103
Thermofilum T. librum T 95
T. pendens T 95
Thermoproteus T. neutrophilus T 97
T. tenax T 97
T. uzoniensis T 97
a
D, deep marine hydrothermal vent; S, shallow marine hydrothermal vent; T, terrestrial;
O, subsurface oil reservoir, B, subsurface biosphere
Source: Data from References 6–13; http://www3.ncbi.nlm.nih.gov/htbin-post/Taxonomy/

261
262 ARCHAEA IN BIOTECHNOLOGY

Table 2. Halophilic Archaea

Optimum Salt
Order Genus Species Habitata Concentration (% w/w)

Halobacteriales Halobacterium H. salarinarium SF 25

H. cutirubrum SF 37
H. halobium SF 37
Halorubrum H. saccharovorum SL 20
H. sodomense SL 15
H. lacusprofundii SL 20
H. trapanicum SL 20
Haloferax H. volcanii SL 15
H. denitrificans SL 15
H. gibbonsii SL 15
H. mediterranei SL 15
Haloarcula H. vallismortis SL 20
H. marismortui SL 20
H. hispanica SL 20
H. japonica SL 20
Halobaculum H. gomorrense SL 20
Halococcus H. morrhuae SL, SP 20
H. saccharolyticus SL 20
H. salifodinae SM 20
Natronobacterium N. pharaonis SD 20
N. gregoryi SD 20
N. magadii SD 20
N. vacuolata SD 20
Natronococcus N. occultus SD 20

Source: Data from Reference 13; http://www3.ncbi.nlm.nih.gov/htbin-post/Taxonomy/

status of archaeal species diversity, divided as the three comprehensive reviews of the evolution, ecology, physiol-
primary phenotypic groupings, the Thermoacidophiles, the ogy, and biochemistry of the Archaea, the reader is directed
Halophiles, and the Methanogens. The reader should be to more specialized articles (6–13).
aware that this list represents only cultured archaea, that
it will be out of date before publication as a result of
UNCULTURED ARCHAEA
taxonomic revisions and new isolations, and that the true
archaeal diversity includes the hundreds or thousands of
The number of known and cultured Archaea is now far
new species and higher orders so far seen only as 16S
exceeded by the number of ‘‘new’’ species and higher
rRNA sequences or not yet observed. Given that some
taxa known only through their 16S rRNA sequences
of the uncultured phylotypes represent major phenotypic
(phylotypes). Phylotypic analyses of microbial community
groups (such as the low-temperature crenarchaeotes), the
DNA from established ‘‘archaeal’’ habitats such as thermal
‘‘mining’’ of the genomic resources of the Archaea has
pools (19) and hypersaline lakes (23) inevitably show
barely started.
numerous novel phylotypes, in which each represents a
The Archaea show a wide metabolic diversity, with a
new species, genus, or higher order of thermoacidophile
number of unique pathways. For example, the methano-
or halophile. Even more significantly, phylotypic analyses
genesis pathway of all methanogens (Fig. 2), reviewed in
of many other terrestrial, freshwater, marine, and other
detail by Daniels (21), contains completely novel sets of
habitats have shown that entire groups of Archaea await
enzymes and cofactors. Many of the cofactors are struc-
isolation. Table 4 shows a selection of the ‘‘uncultured’’
tural analogs of coenzymes in either bacteria or Eukarya,
phylotypes that have been revealed in recent years. Some
but some are unique to the methanogens. Among the ther-
of these phylotypes appear to be virtually ubiquitous and
moacidophiles and halophiles, a variety of chemolithoau-
in many, such as the low-temperature crenarchaeotes
totrophic, fermentative, and chemo-organotrophic modes
(Table 4), not a single example organism has been isolated.
of metabolism have been identified, for example, see Ref-
It is perhaps ironic that one such organism, a psychrophilic
erence 6. Most are not unique to the Archaea, although
symbiont of a marine sponge (24), is presently a target for
certain central catabolic pathways such as the modi-
genome sequencing (Table 5) despite never having been
fied Entner-Doudoroff pathway (in halophiles) and the
cultured.
nonphosphorylated Entner-Doudoroff pathway (in ther-
moacidophiles) seem to be common in the Archaea but very
rare outside this domain (22). Other key enzymes, such ACCESSING UNCULTURED GENOMES
as the thermostable hydrogenases of the chemolithoau-
totrophic thermoacidophiles, have unique properties of There is obviously a great need for new technologies to
considerable potential application (shown later). For more isolate and culture novel microorganisms. However, in
 ARCHAEA IN BIOTECHNOLOGY 263

Table 3. The Methanogens

Order Genus Species

Methanobacteriales Methanobacterium 14 (3 thermophilic)

Methanobrevibacter 7 (0 thermophilic)
Methanosphaera 1 (0 thermophilic)
Methanothermobacter 3 (3 thermophilic)
Methanothermus 2 (2 hyperthermophilic)
Methanococcales Methanococcus 11 (6 thermophilic)
Methanomicrobiales Methanocorpusculum 4 (0 thermophilic)
Methanocalculus 2 (1 halotolerant)
Methanoculleus 5 (1 thermophilic)
Methanofollis 2 (0 thermophilic)
Methanogenium 3 (0 thermophilic)
Methanomicrobium 1 (0 thermophilic)
Methanoplanus 3 (0 thermophilic)
Methanospirillum 1 (0 thermophilic)
Methanopyrales Methanopyrus 1 (1 hyperthermophile)
Methanosarcinales Methanomicrococcus 1 (0 thermophilic)
Methanococcoides 3 (0 thermophilic)
Methanohalobium 1 (1 halotolerant)
Methanohalophilus 4 (4 halotolerant)
Methanolobus 5 (2 thermophilic)
Methanomethylovorans 2 (0 thermophilic)
Methanophilus 1 (0 thermophilic)
Methanosaeta 2 (1 thermophilic)
Methanosarcina 8 (1 thermophilic)

Source: Data from http://www3.ncbi.nlm.nih.gov/htbin-post/Taxonomy/.

Note: Only validly named species are included: Many named genera have numerous
unnamed species and uncultured phylotypes.

recent years, methods have been developed that partially the rapid acquisition, via PCR cloning, of known genes
circumvent the current limitations of microbial isolation. and for the identification of novel or hitherto unidentified
On the basis of direct extraction of total ‘‘community’’ DNA genes. The latter requires powerful computational tools,
and the generation of multigenomic libraries (43,44), this but must always be confirmed by traditional biochemical
technology (Fig. 3) accesses a much higher proportion of methodologies.
the prokaryotic genomes present in any environmental Given the relatively small number of known Archaeal
sample than can be obtained from classical microbial species compared with known bacteria, the number
isolation. A second advantage is that any number of gene of Archaeal genomes being sequenced (Table 5) is
products may be targeted, limited only by the ingenuity disproportionately high. This preoccupation with the
and design of the assay method. Archaea, particularly the hyperthermophiles, reflects the
Current limitations in these methods, including uncer- biochemical and physiological novelty of the domain and
tainties in extraction efficiency, biases introduced during the widespread belief that hyperthermophile genes and
cloning and transformation, and low expression efficien- gene products will yield substantial rewards for basic
cies due to the presence or absence of upstream and down- science (such as new molecular mechanisms of protein
stream control sequences, will clearly be reduced by fur- stabilization) and for biotechnology.
ther developments in recombinant technology. Although The novelty of the Archaea is evident in the proportion
still relatively new, multigenomic cloning will undoubtedly of their genomes constituted by unassigned open reading
be adapted and optimized for eukaryote genomes (e.g., via frames (ORFs) (Table 5). These hypothetical protein
cDNA technologies), for expression screening of libraries, sequences (identified computationally by the presence of
and for the targeting of components of the ‘‘metabolome’’ putative start and stop signals) are categorized either
(the complement of metabolites generated by an organism) as conserved hypothetical proteins (observed in another
via specialized vectors accepting large DNA inserts and genome, but function unknown) or hypothetical (not
capable of expressing the components of partial or entire seen elsewhere, function unknown) proteins. Although
synthetic pathways. the proportion of the latter decreases with each new
genome sequence, functional annotation of the former is
ARCHAEAL GENOME SEQUENCES slow because it is primarily dependent on more detailed
biochemical or computational characterization.
The complete sequencing of an organism’s genome The mining of information from new genomes is
provides a comprehensive blueprint of the metabolic and clearly hampered by the limitations in assigning structure
functional characteristics of the organism. A genome and/or function to hypothetical protein sequences. New
sequence is also a valuable biotechnological resource for computational tools, such as more sophisticated and
 Co2

Formy-MFR dehydrogenase

CHO-MFR
(formyl-methanofuran)

FormyI-MFR:H4MPT formyltransferase

CHO-H4MPT
(formyl-tetrahydromethanopteran)

Methenyl-H4MPT cyclohydrolase

CH2 ≡ H4MPT
(methenyl-tetrahydromethanopteran)

Methylene-H4MPT dehydrogenase

CH2 = H4MPT
HCHO (methenyl-tetrahydromethanopteran)

Methylene-H4MPT reductase

CH3-H4MPT
(methyl-tetrahydromethanopteran)

Methyl-H4MPT:H-S-CoM methyltransferase

CH3OH CH3-S-CoM
(methyl-coenzyme M)

Methyl-CoM reductase
Heterodisulphide reductase

CH4
Figure 2. The methanogenesis pathway.

Table 4. Diverse Habitats Harboring Uncultured Archaeal Phylotypes

Habitat Putative Organism Type/Phylogeny Reference

Marine surface and Low-temperature (Group I) 25

aphotic picoplankton crenarchaeotes; low-temperature
(Group II) euryarchaeotes/methanogens
Marine abyssal water Unknown (group III) euryarchaeotes 26
Marine sediments Low-temperature (Group I) 27
crenarchaeotes; low-temperature
(Group II) euryarchaeotes/methanogens
Fish midgut Low-temperature (Group II) 28
euryarchaeotes/methanogens
Rice paddies Low-temperature (Group I) 29
crenarchaeotes; low-temperature
(Group II)
euryarchaeotes/methanogens; new
phyla
Terrestrial soils Low-temperature (Group I) 30
crenarchaeotes
Freshwater lakes Low-temperature (Group I) 31
crenarchaeotes
Deep subsurface Low-temperature (Group I) 32
biosphere crenarchaeotes

264
 Table 5. Archaeal Genome Sequences (from http://linkage.rockefeller.edu/wli/seq/)

Genome Size Full Sequence % Conserved % Hypothetical

Organism Metabolism Group Responsible for Sequence (Mb) Completed/Available? Hypothetical ORFs ORFs Reference

Aeropyrum pernix Aerobic heterotroph Biotechnology Center, National 1.67 Yes/Yes — — 33

Institute of Technology and
Evaluation, Japan
Archaeoglobus Anaerobic heterotroph TIGR, U.S.A. 2.18 Yes/Yes 27 26 34
fulgidus
Cenarchaeum Unknown TIGR, U.S.A. ? No/No — — 24
symbiosum
Halobacterium Halophilic heterotroph Max-Planck-Institute for 4 No/No – — —
salinarum Biochemistry, Germany
Halobacterium sp. Halophilic heterotroph University of Massachusetts/ 2.57 Yes/Yes — 36 35
NRC-1 University of Washington, U.S.A.
Methanobacterium Anaerobic autotroph Genome Therapeutics & Ohio State 1.75 Yes/Yes 28 27 36
thermoau- (methanogen) University, U.S.A.
totrophicum
Methanococcus Anaerobic methanogen TIGR, U.S.A. 1.66 Yes/Yes 6 56 37
jannaschii

265
Methanococcus Anaerobic methanogen University of Washington, U.S.A. ? No/No — — —
maripaludis
Methanogenium Anaerobic methanogen UNSW / AGRF, U.S.A. ? No/No — — —
frigidum
Methanosarcina Anaerobic methanogen Goettingen Genomics Laboratory, 2.8 No/No — — —
mazei Germany
Pyrobaculum Anaerobic Caltech / UCLA, U.S.A. 2.2 Yes/?? — — —
aerophilum chemolithoautotroph
Pyrococcus abyssi Anaerobic heterotroph GENOSCOPE, CNRS, France 1.8 Yes/Yes — — 38
Pyrococcus furiosus Anaerobic heterotroph Center of Marine Biotechnology / 1.91 Yes/No — — 39
Univ. Utah, U.S.A.
Pyrococcus Anaerobic heterotroph Biotechnology Center, National 1.74 Yes/Yes — — 39
horikoshii OT3 Institute of Technology and
Evaluation, Japan
Sulfolobus Aerobic heterotroph/ European and Canadian 3.05 Yes/Yes — — 40
solfataricus anaerobic autotroph consortium
Thermoplasma Microaerophilic Max-Planck Institute for 1.56 Yes/Yes 29 16 41
acidophilum heterotroph Biochemistry, Germany
Thermoplasma Microaerophilic Osaka University and AIST, Japan 1.58 Yes/Yes — — 42
volcanium GSS1 heterotroph
266 ARCHAEA IN BIOTECHNOLOGY

Microbial community various batch fermentations (Table 7; 48). These figures

are only broadly representative, because the degree of
fermentation development varies between organisms, and
Extraction technology the mode of fermentation (e.g., batch vs. fed-batch vs.
Multigenomic DNA Multigenomic mRNA continuous) has a marked effect on space-time yields (49).
Recent developments in fermentor technology, including
the use of gas-lift and membrane-dialysis fermentors, have
Cloning and vector cDNA technology resulted in substantially increased biomass yields (48).
technology
Multiplex DNS libraries For production of primary gene products (enzymes,
proteins), the obvious answer to problems of low native
biomass yields is the cloning and expression of the relevant
gene in a suitable heterologous host. Yields of recombinant
gene product are then dependent on expression levels and
Expression screening Hybridisation screening PCR screening biomass yields of the host organism.
Despite predictions that hyperthermophilic proteins
might not fold correctly in mesophilic hosts, there is ample
evidence that this is not the case. Many thermophilic
Bioactive gene products Sequence homologus and hyperthermophilic protein genes have now been
Figure 3. Schematic representation of the components of ‘‘mul- expressed in a variety of mesophilic hosts, including
tiplex cloning.’’ E. coli, Pichia pastoris, Saccharomyces cerevisiae and
insect cells, and, in most cases, the recombinant
reliable methods for searching for structural or functional enzymes are functionally identical to the native protein.
motifs, are required. Rapid methods for identifying the There is some evidence, largely anecdotal that some
functional properties of expressed ORFs are also essential. recombinant hyperthermophilic proteins are significantly
The development of automated microarray technologies for less thermostable than the purified native proteins, but no
monitoring binding or catalytic functions (45) would be a structural or mechanistic basis has yet been established.
significant advance. The production of recombinant extreme halophile
proteins in heterologous hosts is severely limited.
Many proteins from the haloarchaea have an obligate
FERMENTATION AND PRODUCTION requirement for 1 to 2 M intracellular salt concentrations
for correct folding and stability and are normally expressed
The successful and economically feasible exploitation of as insoluble, inactive inclusion bodies in nonhalophilic
Archaeal products, such as enzymes, structural proteins, hosts. Halophilic host-vector systems for expression of salt-
lipids, pigments, compatible solutes, and so on, will, in dependent proteins are currently under development (50).
many cases, be dependent on the ability to generate
high biomass yields in fermentation. Alternatively, for
PRODUCTS OF ARCHAEA
primary gene products, the ability to express Archaeal
genes in heterologous hosts at high levels is equally
Enzymes
important. Table 6 outlines some of the current limitations
associated with Archaeal fermentations and production of The Archaea present a unique source of novel enzymes
recombinant Archaeal products (47). because growth in extreme environments requires unique
One major problem associated with generation of high enzyme properties with respect to both activity and stabil-
biomass yields of two of the three groups of Archaea, the ity. The metabolic rates, and hence biochemical reaction
hyperthermophiles and the methanogens, is that many rates, in extremophilic cells are not significantly different
of these organisms are autotrophs. Chemoautotrophic from those in mesophilic cells, yet the extreme envi-
metabolism is generally energy inefficient, and typically ronmental conditions impose unusual constraints on the
results in low cellular doubling times and culture growth protein structure-function relationships of extremophilic
rates and poor biomass yields. In addition, many of these enzymes. Various studies, using techniques such as
organisms grow optimally at temperatures above 85 ° C. genome sequencing and protein crystallography, and com-
Apart from the practical and physical disadvantages parisons with mesophilic enzymes, have demonstrated
of high temperature growth (Table 6), it is probable that archaeal enzymes have certain structural properties
that high temperatures impose a significantly greater that allow them to maintain their catalytic activity under
metabolic burden on cellular constituents than low conditions of extreme temperature, pH, salinity, or chemi-
temperatures. The higher requirement for molecular cal environment. These properties make archaeal enzymes
repair and regeneration in hyperthermophiles is probably particularly attractive as candidates for industrial biocon-
a significant factor in low biomass space-time yields. version processes, in which stringent reaction conditions
Thermal degradation of key nutrients, such as vitamins, frequently prevail. Enzymes possessing enhanced stability
may also be a limiting factor. and selectivity, unusual or novel reactions and/or sub-
In consequence, biomass yields for Archaeal fermenta- strate specificities, are well-established goals in the field
tion are generally very much lower than for many other of biotransformations. Furthermore, the technologies of
microorganisms, as indicated by consensus figures from molecular biology now allow us to manipulate genes and
 ARCHAEA IN BIOTECHNOLOGY 267

Table 6. Limitations in the Production of Archaeal Biomass and Gene Products

Hyperthermophiles Methanogens Extreme Halophiles

Archaeal fermentations High temperatures increase Methanogenic metabolism is Very high chloride concentrations
wear and tear on steel vessels, inefficient are detrimental to steel vessels
bearings, lines, and detectors
Operating costs are high Methane end-product may pose Native enzymes are unstable in
safety issues low-salt buffers
Large-scale culturing of Many low-temperature Some halophilic archaea are
chemoautotrophs is difficult methanogens have very low genetically unstable
doubling times
Expression of Archaeal No high-level extremophilic host-vector expression systems are
genes in heterologous yet available.
hosts
Hyperthermophilic genes No known major problems for Most mesophilic hosts are
expressed in mesophilic hosts mesophilic methanogenic genes incompatible with correct
are not always as stable as folding of extremely halophilic
the wild-type gene products proteins, which typically
require molar intracellular salt
concentrations

Table 7. ‘‘Laboratory’’ Biomass Yield Ranges for Various Organisms

Maximum Batch
Fermentation
Growth Temperature Biomass Yields
Organism ( ° C) Metabolism (g/L w.w.)

Streptomyces 25 Heterotrophic, aerobic 20–100

Bacillus 30 Heterotrophic, aerobic 20–50
Escherichia coli 37 Heterotrophic, aerobic 20–100
Natronobacterium 37 Heterotrophic, aerobic 5–15
Bacillus stearother- 60 Heterotrophic, aerobic 5–10
mophilus
Thermus 70 Heterotrophic, aerobic 2–5
Methanothermus 85 Autotrophic, anaerobic 1–5
Sulfolobus 85 Heterotrophic, aerobic 2–10
Methanococcus 95 Autotrophic, anaerobic 0.1–1
Pyrococcus 98 Heterotrophic, anaerobic 2–10
Pyrodictium 98 Autotrophic, anaerobic 0.1–0.5

gene products to the extent that enzymes with desirable the extent that the process proceeds less efficiently at
extremophilic properties may be fine-tuned for specific lower temperatures.
biotechnological applications. Thermophilic enzymes have evolved to control cellular
metabolic activity optimally at the elevated temperatures
Enzyme Properties of their environment, generally by means of mechanisms
causing the proteins to have less conformational flexi-
Thermostability. The remarkable activity and stability bility than homologous mesophilic enzymes, and more
of thermophilic enzymes at high temperature suggests resistance to the unfolding processes that initiate denatu-
numerous advantages with regard to their biotechnolog- ration (53). Recent research has indicated that a number
ical application. Many industrial processes are ideally of relatively minor modifications to the higher order struc-
run at elevated temperatures, in which, for instance, tures of the proteins are responsible for these differences
increased reactant solubilities or equilibrium ratios may (Table 8), and protein sequences do not appear to vary sig-
be favored. Higher enzyme stability and resistance to inac- nificantly between functionally homologous thermophilic
tivation reduces the need for replacement or regeneration and mesophilic enzymes (54,55). In particular, the pres-
of the biocatalyst. Structural thermostability also confers ence of extended networks of ionic interactions on the
resistance to inactivation by chemical denaturants (deter- surfaces of the proteins, increased hydrogen bonding,
gents, chaotropic agents), oxidizing agents, proteases, and reduced sections of random loop structure, and structural
organic solvents, and hence to the harsh conditions that mechanisms resulting in compact protein packing, such as
may prevail in many industrial process situations (51,52). hydrophobic interactions and subunit association, have all
Conversely, of course, the high temperature-optima of been observed in thermophilic enzyme structures. In addi-
thermophilic enzymes constrains reaction conditions to tion, low proportions of thermolabile amino acids enhance
268 ARCHAEA IN BIOTECHNOLOGY

Table 8. Structural Modifications Leading to Increased Thermostability in Archaeal Proteins

Structural
Property Modification Example Reference

Primary structure Reduced number of thermolabile Pyrococcus furiosus citrate synthase 11

amino acids, e.g., Met, Cys, Asn, Gln
N- and C-terminal amino acid residues Methanothermus fervidus histone 56
immobilized into protein structure
Polar and ionic Extended ion-pair networks acting Glutamate dehydrogenases from a 57,58
interactions over extensive portions of protein range of thermophiles including P.
furiosus
High concentration of hydrogen bonds Glutamate dehydrogenase from P. 59
furiosus
Binding of calcium ions Extracellular protease Archaelysin 51
from Desulfurcoccus strain Tok12 S1
Protein folding structure Increased hydrophobic interactions Staphylothermus marinus surface 58
giving close packing and excluding protein
solvent
Increased inter-subunit interactions — —
and oligomer formation
Extended secondary structure — —
Few and /or short random loops Citrate synthase from P. furiosus 11
Solvent-filled hydrophilic cavities S. solfataricus —

stability in the primary structure and the absence of many in the successful expression of extreme halophile protein
glycine residues reduces peptide chain flexibility. Extrin- genes in common hosts, in purification procedures, and in
sic factors such as binding of calcium ions also contribute the development of biotechnological applications.
to thermostabilization in some proteins (51).
Wherever thermophilic enzyme genes are expressed in Organic Solvent-Tolerant Enzymes. The advantages in
mesophilic hosts, the thermostability of the recombinant biocatalysis of using organic or aqueous-organic media
enzyme confers the advantage of facile purification, in for enzyme catalyzed reactions (e.g., increased substrate
which contaminating mesophilic proteins can be removed solubility and/or stability, reduced water-mediated side
by heat shock. reactions, reduced microbial contamination) have provided
the incentive for seeking and characterizing solvent-
Halotolerance. The defining feature of the enzymes of tolerant enzymes. The stability of halotolerant enzymes
halophilic microorganisms is resistance to inactivation in originates at least partly from the ability to withstand very
high ionic strength (low water activity) solutions with salt low water activity; such adaptations can also confer solvent
concentrations of 2 to 5 M. Conversely, the presence of tolerance. The solvent tolerance of several enzymes from
a high ionic strength environment is mandatory for the halophilic Archaea has been reported (Table 9). Similarly,
correct folding and biological activity of many extreme the mechanisms leading to enzyme thermostability,
halophile proteins. One strategy for the survival of particularly reduced conformational flexibility, also confer
halophiles is the intracellular accumulation of inorganic resistance to solvent-mediated denaturation (53,60).
ions, such as K+ and Cl− , which means that intracellular
proteins must be halotolerant to the same extent as Cold-Tolerant Enzymes. Numerous potential applica-
extracellular enzymes. Significant adaptations of the tions of cold-tolerant enzymes, particularly in the food,
protein structures are required to give the proteins the detergent, and textile industries, have stimulated interest
unusual resistance to the unfolding and aggregation in psychrophiles (64,65). Advantages in the use of cold-
processes that typify mesophilic proteins exposed to tolerant enzymes include the facility to inactivate them
high ionic strength conditions. The halophilic proteins by moderate increases in temperature, increased reaction
are commonly characterized by a high proportion of yields with unstable substrates, and energy cost sav-
acidic amino acids (glutamate and aspartate), which ings. Among the very few archaeal cold-tolerant enzymes
contribute a high concentration of negative charge that have been reported is a psychrophilic DNA poly-
on the protein surface. This facilitates the binding merase from C. symbiosum, isolated by direct genomic
of hydrated counterions from solution (halophilicity), DNA cloning (60). The structural features that apparently
maintaining the protein surface hydration and decreasing facilitate activity at low temperature included extended
the surface hydrophobicity that would otherwise lead to charged surface loops, reduced numbers of hydrogen bonds
aggregation (54). and disulfide bridges, and increased ion-pair content, all
A high salt environment is an absolute requirement of which would lead to increased protein flexibility (66).
for many extreme halophiles, and their enzymes are This increased flexibility has potential value in biocatal-
frequently observed to be unstable in the absence of high ysis applications in which low water activity is used, but
ionic strength media. This can be a serious disadvantage broad substrate selectivity must be maintained.
 ARCHAEA IN BIOTECHNOLOGY 269

Table 9. Examples of Solvent-Tolerant Enzymes from the Archaea

Archaeal Group Enzyme Potential Application Reference

Halophiles Halobacterium cutirubrum Removal of H2 O2 from reaction mixtures 61

catalase
Haloferax mediterranei protease Peptide synthesis 60
Halobacterium halobium Peptide synthesis with advantage of broad 59
extracellular protease substrate specificity and very high activity
Thermophiles Sulfolobus solfataricus malate Biocatalytic redox reactions 60
dehydrogenase
Sulfolobus solfataricus alcohol Biocatalytic redox reactions 62
dehydrogenase
Thermococcus stetteri proteinase Stereospecific peptide or amide synthesis 63

Pressure-Tolerant Enzymes. High pressure is generally Enzymes that are stable and active at the high
regarded as a denaturing condition for proteins. However, temperatures required to gelatinize starch form one of
in the case of thermostable archaeal proteins, high the largest-scale industrial applications of biocatalysis.
pressure has been shown to increase the stability and The glycosyl hydrolases of importance in the starch
even reaction rate, as demonstrated for a protease from industry include amylases exhibiting both α-1, 4- and β-1,
M. jannaschii (11). Pressure-stabilization is thought to be 4 specificity, β-glucosidases that hydrolyze nonreducing
due to increased packing of hydrophobic core regions. The terminal 1,4-linkages, and pullulanases (debranching
DNA polymerase of Pyrococcus strain ES4 was used as an enzymes) required to release glucose monomers from 1,6-
example of a barophilic enzyme with high thermostability linked branches in the polymers (Fig. 4). Biotechnological
in a study that indicated that pressure-tolerance is related applications of bacterial thermostable glycosyl hydrolases
to restricted volume changes during protein unfolding (67). are well established, and as the availability of bulk, low-
cost archaeal enzymes improves, their potential should
be realized (Table 11). For example, the halotolerant
Applications — Low Specificity Enzymes amylases produced by the halophilic Archaea H. halobium
and Natronococcus could provide useful alternatives to
Glycosyl Hydrolases. Thermostable glycosyl hydrolases
the commercial enzymes currently in use (72). In addition,
derived from Archaea act on intracellular and extra-
more specialized reactions may be developed, such as
cellular polysaccharides, facilitating the utilization of transglycosylations catalyzed by P. furiosus β-glycosidase
external energy sources and maintaining a dynamic equi- for the production of modified chiral sugars used in
librium in the formation and break down of carbohydrate pharmaceuticals (53).
osmolytes (68). Heterotrophic hyperthermophiles that can
grow on starch or β-linked carbohydrates are common and Proteases. Thermostable proteases have particular
have been well documented (Table 10; 69,70). The func- application in protein hydrolysis because the elevated
tions of the glycosyl hydrolases are not always clear, as temperatures accelerate the unfolding of a protein poly-
in the case of the autotrophs such as M. jannaschii (71). mer and increase accessibility of the peptide to pro-
Hyperthermophiles produce numerous glycosyl hydro- tease activity. Conversely, proteases that are resistant
lases with varying activities and exhibit both α- and to denaturation by organic solvents find application
β-hydrolyzing specificities. Typically, extracellular hydro- in peptide synthesis carried out in low water activ-
lases are more stable than the intracellular enzymes (53). ity media. Thermostable proteases have been isolated

Table 10. Thermostable Glycosylases Produced by Archaea (66,67)

Enzymes Properties

α-specific Amylase Pyrococcus furiosus Topt 100 ° C Recombinant enzyme cloned into
mesophilic host, expressed extracellularly
Glucosidase Topt 115 ° C
Amylopullulanase Topt 100 ° C Extracellular recombinant enzyme
Glucosidase Sulfolobus Topt 105 ° C
solfataricus
Amylase Topt 115 ° C
Pullulanase Pyrococcus woesei Topt 120 ° C Recombinant enzyme
β-specific Glucosidase Pyrococcus furiosus Topt 105 ° C Recombinant, expressed in E. coli
Glucosidase Sulfolobus Topt 105 ° C
solfataricus
Mannosidase Pyrococcus furiosus Topt 105 ° C
Lamminarase First endoglucanase found in the Archaea
270 ARCHAEA IN BIOTECHNOLOGY

Starch granules Halotolerant proteases, produced by the halophilic

Hyperthermostable species, are utilized for proteolysis reactions in the
amylases would extend high-salt (up to 25% w/v salt) fermentations used to
Gelatinization the 105˚C ′jet cooking′ convert certain foods (commonly fish or vegetables) into
phase, reduce reaction
strongly flavored sauces such as nuoc mam and soy sauce.
time and limit the need
for cooling Members of the Archaeal Halococcus and Halobacterium
Released amylose
genera have been isolated from such fermentations,
having originated, putatively, from the salt added to
Amylase Liquefaction the fermentation (72). Halotolerant exoproteases are also
More thermostable
produced by H. salinarum and other Halobacteriaceae
Amyloglucosidase
and pullulanase strains.
Dextrins would increase
reaction rate, reduce Applications — High Specificity Enzymes
Amyloglucosidase, proces time
pullulanasey Saccharification DNA Polymerases. DNA polymerases catalyze intracel-
lular reactions involved with the replication and repair
Thermostable glucose of DNA. The biotechnological applications of DNA poly-
Glucose (xylose) isomerase merases, particularly in thermal cycling reactions for
would increase reaction amplification of DNA (the polymerase chain reaction),
rate and reduce
Glucose have led to very significant developments in molecu-
Isomerization process time.
isomerase lar biology. Although the first used and perhaps best
known DNA polymerase, Taq (from Thermus aquaticus),
High-fructose syrup is not a product of an archaeal microorganism, ther-
mostable DNA polymerases from archaeal thermophiles
Figure 4. Enzymes used in carbohydrate conversions. represent an important potential source of variety. The
heat-stable ‘‘Vent’’ and ‘‘Deep Vent’’ polymerases from
Thermococcus litoralis (73) and Pfu polymerase from P.
and/or cloned from a number of thermophilic Archaea furiosus have been characterized, cloned, and success-
(Table 12). fully expressed, and are widely used in molecular biology.

Table 11. Potential Biotechnological Applications of Glycosyl Hydrolases

Enzymes Sources Potential Application

Xylanases Thermococcus zilligii Xylans associated with lignin in wood pulping

effluent treatment (74)
Hemicellulases Thermatoga neopolitana Decreasing viscosity of mannan solutions, e.g.,
guar gum (70)
Cellulases Cellulose degradation to release glucose
Laundry detergent additives (68)
Stonewashing cloth
Amylase Natronococcus sp. Conversion of starch to maltotriose (75)
Amylase + pullulanases + Pyrococcus furiosus One-step starch-to-glucose process (70)
glucosidase
Pullulanases Pyrococcus furiosus, Glucose polymer 1,6-linkage debranching in
P. woesei starch hydrolysis
Amylase Halobacterium halobium Complete starch conversion to glucose
β-glycosidase Pyrococcus furiosus Transglycosylation of chiral sugars (70)

Table 12. Thermostable Proteases from Archaea

Archaeal Source/Protease Properties Ref

Aeropyrum pernix ‘Pernilase’ Intracellular;Topt 90 ° C; t1/2 12 minutes at 110 ° C 74

Aeropyrum pernix ‘‘Aeropyrolysin’’ Extracellular;Topt 100 ° C; t1/2 1.2 hours at 125 ° C 75
Sulfolobus acidocaldarius ‘‘Thermopsin’’ — 76
Sulfolobus solfataricus neutral protease Topt 87 ° C 77
D. mucosus ‘‘Archaelysin’’ Topt 100 ° C; purified protease active up to 125 ° C 78
Pyrococcus furiosus protease ‘‘Pyrolysin’’ Topt 115 ° C; protease in cell-free extracts still active 79
after 24 hours boiling with 1% SDS
Pyrococcus furiosus protease Pfp1 Topt 95 ° C; t1/2 19 minutes, 95 ° C 80
Pyrobaculum aerophilum ‘‘Aerolysin’’ — 81
Methanococcus jannaschii proteosome 700 kDa complex;Topt 100 ° C 82
 ARCHAEA IN BIOTECHNOLOGY 271

These enzymes have functional advantages over Taq poly- is potential value in the search for additional, more
merase, with higher thermostability and proofreading (3 - thermostable, ligase enzymes from the Archaea.
5 exonuclease) activity. More recently, other DNA poly-
merases from T. acidophilum, M. thermoautotrophicum, Specialist Enzymes for Biocatalysis. With the rapid
S. acidocaldarius, and S. solfataricus, have been expansion of biotransformation technology, the need for
reported (53). Although none of the aforementioned show biocatalysts with very specific characteristics now drives
these properties, any thermostable polymerase with par- the search for novelty in enzyme function. The exquisite
ticularly high fidelity DNA amplification or extended (e.g., stereospecificity of many enzymes is not readily matched
megabase) read-through will almost certainly be commer- in conventional synthetic chemistry, and enzymes pos-
cializable. sessing both high selectivity and stability under extreme
conditions potentially have a high-value market. For
Restriction Endonucleases. Restriction nucleases with example, the alcohol dehydrogenases of S. solfataricus
unusual site specificity are of particular value in and H. butylicus, which have thermal stability and sol-
molecular biology, for use in the cloning and/or the vent tolerance, have good stereoselectivity in the catalysis
restriction analyses of DNA. Such a nuclease, produced of secondary alcohol oxidations or carbonyl compound
by Halococcus acetoinfaciens, was reported by Obayashi reductions (53). Similarly, esterases with solvent toler-
and coworkers (76). Although many Archaea possess ance have application in stereochemical resolution and in
Type 2 endonuclease activities, few have unique (and transesterifications and ester hydrolyses. Thermostable
therefore particularly valuable) recognition- and/or cut- and halotolerant archaeal enzymes may well provide new
site specificities. In addition, there are currently no sources of these activities. Biocatalytic redox reactions
significant applications in which high thermostability in may also impose unusual constraints, such as specific
restriction endonucleases is perceived to be an advantage. redox potential or cofactor recycling. Enzymes with high
In most instances, the simple and effective removal of stability and activity and able to function in organic
activity after a restriction digest is paramount. solvent media, would be especially valuable in cofac-
tor regeneration. For example, pyruvate dehydrogenase
Inteins. Inteins (protein introns) are internal portions from P. furiosus (Topt >95 ° C), glutamate dehydrogenase
of protein sequences that are post-translationally excised from P. furiosus (Topt 95 ° C), glutaraldehyde-3-phosphate
although the flanking: regions are spliced together, dehydrogenase from P. woesei, and formaldehyde oxi-
making an additional protein product. Inteins have doreductase from T. litoralis (Topt 95 ° C) (50) may all be
been found in a number of homologous genes in yeast, candidates for developing such applications.
mycobacteria, and hyperthermophilic Archaea such as
P. furiosus and M. jannaschii (86). Inteins are probably Lipids
multifunctional, autocatalyzing their own splicing and, in The ‘‘core’’ lipids of the Archaea are largely based on
some cases, exhibiting endonuclease activity. saturated isoprenoid chains linked to a glycerol back-
Yeast intein genes have been used in the development bone by ether bonds. Common structures include the
of an innovative expression-purification system. The monomeric diphytanylglycerol ethers and the dimeric dibi-
IMPACTTM system, developed by New England Biolabs, phytanyldiglycerol tetraethers and dibiphytanyl glycerol
Inc. U.S.A., used a chimeric E. coli expression vector nonitol tetraethers (a generic structure is shown in Fig. 5).
incorporating fused genes encoding the yeast intein The former, termed archaeols, are found in all Archaea,
and a Bacillus chitin-binding protein. The heterologous whereas the latter, termed caldarchaeols and nonitolcal-
protein gene is inserted into a multiple cloning site at darchaeols, are found only in the thermophilic Archaea.
the N-terminal autocleavage site of the intein. After The caldarchaeols and nonitolcaldarchaeols exhibit fur-
capture of the expressed fusion protein (using a chitin ther modification by containing up to four cyclopentane
affinity column), the heterologous protein can be released rings in each of the C40 biphytanyl chains. The addition
autocatalytically after addition of a reducing agent. of cyclic structures in the transmembrane portion of the
This novel expression-purification system has many lipid appears to be a thermoadaptive response, resulting
possible applications. The potential use of a thermophilic in enhanced membrane packing and reduced membrane
archaeal intein gene rather than the mesophilic yeast fluidity (89).
gene may provide an alternative approach to the rapid Although few, if any, studies have been carried out
purification of heterologous thermophilic proteins on the chemical stability of purified ether lipids, the sta-
bility of artificial ether-lipid membranes (liposomes) has
DNA Ligases. The DNA ligase reaction system (ligase attracted considerable attention. Membranes composed of
chain reaction) allows defects in DNA to be detected
through mismatching during hybridization of DNA
samples with specifically selected primers. The reaction R
depends on thermostable ligase activity catalyzing the OH
O
annealing of previously heat-denatured oligonucleotides. O
Such a reaction has potential application in detection of O
O
genetic defects (87) or specific pathogens (88). Although OH
a thermostable DNA ligase from the bacterial Thermus Figure 5. Biphytanyl tetraether core lipid structure (where
thermophilus has been cloned and made available, there R = H, calditol, etc).
272 ARCHAEA IN BIOTECHNOLOGY

C40 membrane spanning (boliform amphiphilic) tetraether some species possessing multiple layers with higher
lipids maintain a constant thickness of 25 to 30Å (89), order structures, and/or layers incorporating unusual
somewhat thinner than typical C18 phosphodiester bilayer glycoprotein polymers.
membranes. Nevertheless, these archaeal membranes are The S-layer proteins of Archaea have a high propor-
much more physically stable than those formed from tion of hydrophobic amino acids and are often highly
phosphodiesters. For example, large (600 nm) vesicles gen- glycosylated, sulfated, and/or acidic. The glycosyl compo-
erated from T. acidophilum ether lipids were found to be nents involved frequently have complex composition, with
more resistant to physical disruption by high tempera- unusual methylation or acetylation of amino groups on the
ture and surface-active agents such as phenol, alcohols, polysaccharide backbone. For example, in H. halobium,
and detergents than dipalmitoyl phosphatidylcholine vesi- 50 sulfate groups are present per molecule, on acety-
cles (90). In a more extensive study of liposome stability, lated glucuronic acid and acetylated N-aminogalactosyl
Sprott and coworkers (91) compared liposomes prepared residues (100). In some methanogens, the S-layer pep-
from the ether lipid extracts of a number of archaea with tidoglycan subunits also contain the unusual polymer
egg phosphatidylcholine and dipalmitoyl phosphatidyl- pseudomurein, which contains the amino acids L-lysine,
choline liposomes. In virtually all instances, the archaeal L-glutamic acid, L-alanine, and sometimes ornithine,
ether lipid liposomes showed higher levels of stability to in addition to certain unusual N-acetylated amino sug-
temperature, pH, serum proteins, and long-term oxida- ars. The cell walls of Methanosarcina spp. contain
tive effects than the ester lipid vesicles. Not surprisingly, methanochondroitin, another specialized glycopeptide, in
the former were also highly resistant to the addition of which the carbohydrate residues are N-acetylated glu-
phospholipases. The use of fluorescent probe techniques cosamine and galactosamine (101). Such glycosylation and
to measure the thermal stability of S. acidocaldarius lipid chemical diversity patterns are far less common in the
liposomes over a temperature range of 25 to 85 ° C showed bacteria.
little variation in proton permeability across the tempera- The functions of S-layers are believed to include main-
ture range (92). tenance of shape, protection, cell adhesion, molecular
The remarkable physical and chemical stability of surface recognition, and molecular binding to the cell sur-
archaeal ether lipid liposomes, attributed to the presence face (102). The well-defined pore-size, 30 to 40 kDa (103),
of the ether linkage, to the intimate packing of the is likely to confer a molecular exclusion function, pro-
phytanyl chains, and to the reduced degree of molecular tecting the cell against attack by lytic enzymes, immuno-
translational freedom in a boliform structure, has provided gens, and biocides. Pseudomurein, for instance, is resis-
a significant incentive to the practical utilization of tant to cell-wall antibiotics, lysozyme, and proteases and
such liposomes. A number of applications, including drug methanochondroitin is resistant to cellulolytic activity.
delivery systems (91) and bioelectronics components (93), Whatever the structure or function of the S-layer, con-
have been proposed. tinuous recrystallization occurs, allowing the dynamic
The recent demonstration that Archaeal lipid fractions processes of protein turnover and cell expansion and divi-
could be formed into stable liposomes (termed archaeo- sion (98). The highly active expression systems required
somes) has led to a resurgence of interest in these for continuous in vivo S-layer production are of consider-
structures as vaccine and drug delivery systems (94). The able interest in the development of vectors for high-level
uptake of archaeosomes by phagocytic cells can be up to expression of heterologous proteins.
50-fold greater than normal liposomes (95), and toxicity
trials have not shown any adverse reactions. Mice treated Biotechnological Applications of S-Layers. The capability
with archaeosome preparations containing antigenic pep- of S-layer proteins to self-assemble and recrystallize
tides have been shown to give immune responses of the correctly and consistently is also observed in vitro, even
same order as when using conventional adjuvants (96). after disruption due to chemical treatment, which suggests
In addition, tissue distribution data following adminis- a number of potential biotechnological applications for S-
tration of archaeosomes by different routes suggests that layers. The cloning and overexpression of bacterial S-layer
proteins has facilitated their production for reassembly on
some degree of organ targeting may be possible (97).
foreign (non cell-surface) surfaces, in amounts sufficient
S-Layers for industrial applications.
Porous monolayers can be obtained by allowing S-layer
The cell walls of most Archaea are characterized by the proteins to self-assemble on a variety of different type
presence of a regular arrangement of paracrystalline, surfaces, including inorganic (e.g., silicon) and metallic
homogenous protein, or glycoprotein subunits, the Surface- (e.g., gold). The orientation of the protein subunits is such
layer (S-layer), as the outermost surface barrier of the that there is asymmetry in the monolayer, with respect
cell. The protein subunits are usually of high molecular to hydrophobicity of the component molecules, so that the
weight (40–200 kDa) and are arranged in lattices with layers assemble with a more hydrophobic face presented to
characteristic patterns of symmetry and complexity, the more hydrophobic portion of any biphasic environment
conferring varying degrees of porosity (98). In some such as an air-water or water-solvent interface or on
archaeal species, such as H. salinarum, H. halobium, solid or liquid supporting surfaces (104). Thus, selection of
and S. acidocaldaricus (99), the cell wall is composed the medium permits some control over the nature of the
only of an S-layer, a simple protein layer outside monolayer that forms and hence its function.
the plasma membrane. However, the composition and The exploitation of archaeal S-layers has been less
complexity of the S-layer varies considerably, with commonly reported, possibly because of difficulties in the
 ARCHAEA IN BIOTECHNOLOGY 273

large-scale fermentation or expression of archaeal pro- water-compatible molecules. Although significantly more
teins. Nevertheless, the progress of production technology information is known regarding the bacterial solutes,
is sufficiently rapid to justify the assumption that archaeal hyperthermophilic members of the archaeal group are
S-layer proteins will soon be accessible in quantities that known to accumulate di-myo-inositol-1, 1 -3, 3 -phosphate,
allow the industrial exploitation of these more complex cyclic 2, 3-bisphosphogycerate, or trehalose (109). In ther-
monolayers. The discussion in the following text therefore mophilic Archaea, these compatible solutes are considered
reports largely on results achieved with bacterial S-layer to have a role in protein thermostabilization, and in
proteins, but may be extrapolated to the application of halophiles, in osmoregulation and stabilization against
archaeal S-layers. high ionic strength denaturation. In certain halobacteria,
production of compatible solutes such as biodegradable
Ultrafiltration Membranes from S-Layer Proteins. Self- polyalkanoates and ectoines has been developed for com-
assembled monolayers obtained from S-layers have reg- mercial applications, but no such applications have yet
ular and well-defined dimensions (5–15 nm thick, with been developed for archaeal sources. Nevertheless, an
pores of 2–8 nm), making them particularly suitable understanding of the role of these solutes in protein
for use as ultrafiltration membranes for high-selectivity stabilization is likely to provide mechanisms whereby
molecular exclusion, with molecular cutoff of 30 to biotechnological and pharmaceutical products can be sta-
40 kDa (103). The consistency of their chemical composi- bilized.
tion and charge distribution, as well as physical structure,
give S-layer membranes more uniform characteristics
ARCHAEAL PROCESSES
than conventional synthetic ultrafiltration membranes.
The proteins are allowed to assemble on synthetic micro-
Methane Production
filtration membranes, and may be cross-linked chemically
to provide physical resilience. Further chemical modifi- The generation of methane through the mesophilic
cation of the protein monolayers can be used to produce (and sometimes thermophilic) anaerobic metabolism of
customized membranes with specific functions such as methanogenic Archaea is a naturally-occurring and
antifouling properties (105). widespread process and is particularly evident in swamps,
rice paddies, landfill sites, and in ruminant and mam-
Immobilization of Biomolecules on S-Layers. The func- malian digestive processes. Many of the consequences of
tional groups of S-layer proteins and their associated this process may be considered to be detrimental, partic-
carbohydrate moieties provide a variety of potential reac- ularly where unwanted methane production generates an
tion sites for immobilization of other biomolecules, such explosion hazard (as in landfill sites) or through the sup-
as enzymes, antibodies, or lipids (106). Such immobilized posedly deleterious effects on the ozone layer. However,
systems can be formulated on a micro- and nanoscale, under ‘‘controlled’’ conditions, Archaeal methanogenesis
as in the case of immobilization of enzymes for use as (biogas) has been exploited as an inexpensive and effec-
microscale biosensors in the detection of sucrose (98,107). tive source of energy for many centuries. The first biogas
Coimmobilizing multiple biomolecules allows application production plant was constructed in India in 1859 and
of the technology to biochemical reactions of considerable some 10 million small-scale digestion plants are now in
complexity, for example, in immunoassays (106). More use worldwide for local heating and lighting. Large-scale
sophisticated surfaces, formed by combining S-layers with energy production employing methanogenesis is now com-
bimolecular lipid membranes and incorporating various monplace in Denmark.
biochemical systems such as proton pumps, provide much Biogas production is based on the digestion of
potential for new applications (107). inexpensive renewable organic material, including animal
manure, organic industrial waste, and municipal solid
Nanotechnology. In more advanced applications, waste. On a larger scale, the biogas production process is
recrystallized monolayers can be laid down in predesigned frequently integrated into the treatment of industrial,
patterns on a silicon surface and etched by use of commercial, and municipal wastewaters and effluent
ultraviolet radiation, resulting in practical nanoscale streams. For additional information, the reader is directed
semiconductor components. Coatings of S-layer proteins to the following reviews (110,111).
can also be used to mask areas of silicon surfaces
before etching. Alternatively, nanometer scale metallic H2 Production
particles (3–5 nm) and lattices can be formed with
An innovative application of hyperthermophilic archaeal
preselected symmetry, by using S-layer assemblies as
enzymes to molecular hydrogen production has recently
templates (104).
been proposed. Because of its high energy content, hydro-
gen is a particularly favored fuel, although biological pro-
Compatible Solutes
duction methods have received relatively little attention.
An important characteristic of halophilic and hyperther- The discovery (112) of an NADPH-dependent hydrogenase
mophilic microorganisms is the accumulation of a range in the hyperthermophilic Archaeon P. furiosus has pro-
of low molecular weight compounds in cells, serving as vided an in vitro mechanism for H2 production. By coupling
osmoprotectants of cellular components, including pro- NADPH production to an enzymic generation system, such
teins (68,108). The chemical nature of these ‘‘compatible as a thermophilic glucose dehydrogenase, a process that
solutes’’ varies between species, but all are hydrophilic, uses an inexpensive feedstock (glucose) and generates
274 ARCHAEA IN BIOTECHNOLOGY

two high-value products (glucuronic acid and H2 ) can be 25

developed. Preliminary results that show high conversion Copper
efficiencies are promising (113).
Iron
Biomining: Biological Extraction of Metals 20

Metal ion concentration (g/L)

The ability of iron- and sulfur-oxidizing acidophilic
microorganisms to enhance the extraction of metals such
15
as copper, gold, and uranium (known as biomining or
bioleaching) has been employed commercially for some
years. Piles or dumps of low-grade mineral ores are typ-
ically acidified by addition of mineral acid, stimulating 10
the natural populations of FeII - and S-oxidizing organisms
(primarily the mesophilic bacteria Thiobacillus ferroox-
idans, Thiobacillus caldus, and Leptospirillum ferrooxi-
5
dans; 114). Where natural heating occurs, populations
of the moderately thermophilic bacterium Sulfobacillus
and archaeon Thermoplasma may become significant.
At higher temperatures (e.g., >60 ° C), the extremely 0
thermophilic Archaea, Sulfolobus, Acidianus, and Met- 0 50 100 150 200 250 250
allosphaera, dominate. The process of metal recovery is Time (hours)
dependent on oxidative solubilization of copper and iron Figure 6. Solubilization of chalcopyrite ore at 78 ° C using
sulfides (chalcopyrite and iron pyrite), releasing soluble thermophilic archaea (redrawn with permission).
metals (e.g., copper) and insoluble metal particles (e.g.,
gold and uranium).
CONCLUSION
For recovery of high-value metals, it is economically
feasible to leach mineral sulfide concentrates rather
The reader will note that one of the most used terms in the
than low-grade ores and to employ bioreactor technology
(typically simple but large stirred tank reactors) rather entry with regard to the value of Archaea in biotechnology
than ore piles. A number of pilot plants and large- is the term potential, reflecting the reality that while
scale demonstration plants are currently operating in much information is available and more is being generated
Brazil, South Africa, Ghana, and Australia (115,116), with very rapidly, the number of real commercial applications
throughputs of a few tens to more than a thousand tonnes is limited. This situation is strangely at odds with the
of ore per day (Table 13). level of interest in both academic and industrial sectors
The interest in thermophilic bioleaching is derived with respect to extremophiles in general and Archaea
from advantages in process economics, mineral solubility, in particular. One factor contributing to this is limited
and reaction rates from operating at elevated temper- success in developing efficient expression systems for the
atures (117). Norris and coworkers (118) have demon- purposes of enzyme production in recombinant hosts.
strated that efficiency of extraction increases proportion- Additionally, the operating conditions required for fer-
ally with temperature up to around 80 ° C. Furthermore, mentation processes for generating Archaeal bioproducts
when operating an exothermic batch reaction in a hot are substantially different from those of conventional
climate, the use of thermophilic organisms precludes the industrial processes and require novel approaches in terms
requirement for costly cooling technology. Possible disad- of reactor development, process engineering, and economy.
vantages lie in the sensitivity of archaeal cells to physical Both of these problem areas are currently being addressed
damage in stirred reactors (116) and in the sensitivity by modern technology developments. The section of the
of some archaeal strains to high concentrations of solu- entry describing Archaeal products, whether potential or
ble metal ions (118). However, the recent identification real at this stage, provides justification for the attention
of Sulfolobus strains capable of tolerating concentrations the field is receiving.
of copper ions up to 40g/L (Fig. 6) suggests that these
organisms may be well suited to large-scale metal ion BIBLIOGRAPHY
recovery.
1. C. R. Woese and G. E. Fox, J. Mol. Evol. 10, 1–6 (1977).
Table 13. Operating Mineral Processing Bioreac- 2. R. M. Schwarz and M. O. Dayhoff, Science 199, 395–403
tors (100,116) (1978).
3. N. Iwabe et al., Proc. N.Y. Acad. Sci. U.S.A. 86, 9355–9359
Throughput Reactor (1989).
Location (tonnes ore/day) Capacity (m3 )
4. C. R. Woese, in M. Kates et al., eds., The Biochemistry of
Archaea (Archaebacteria), Elsevier Science Publ., Amster-
Fairview, South Africa 40 900
dam, The Netherlands, 1993, pp. vii–xxix.
São Bento, Brazil 110 1,160
Wiluna, W. Australia 152 4,240 5. O. Kandler, in J. Weigel and M. W. W. Adams, eds., Ther-
Obuasi, Ghana 1,152 21,360 mophiles: The Keys to Molecular Evolution and the Origins
of Life, Taylor & Franscis, London, U. K., 1998, pp. 19–31.
 ARCHAEA IN BIOTECHNOLOGY 275

6. K. O. Stetter, Extremophiles: Microbial Life in Extreme 43. A. Henne et al., Appl. Environ. Microbiol. 66, 3113–3116
Environments, Wiley-Liss, New York, 1998, pp. 1–24. (2000).
7. R. F. Doolittle, Res. Microbiol. 151, 85–89 (2000). 44. D. A. Cowan, Trends Biotechnol. 18, 14–16 (2000).
8. J. R. Brown and W. F. Doolittle, Microbiol. Mol. Biol. Rev. 45. G. Zweiger, Trends Biotechnol. 17, 429–426 (1999).
61, 456–486 (1997). 46. T. Kawashima et al., Proc. Natl. Acad. Sci. U.S.A. 97,
9. M. Kates et al., eds., The Biochemistry of Archaea (Archae- 14257–14262 (2000).
bacteria), Elsevier Science, Amsterdam, The Netherlands, 47. O. Holst et al., Comp. Biochem. Physiol. A 118, 415–422
1993, p. 582. (1997).
10. R. M. Daniel and D. A. Cowan, Cell. Mol. Life Sci. 57, 48. M. Krahe et al., FEMS Microbiol. Rev. 18, 271–285 (1996).
250–264 (2000).
49. D. Riesenberg and R. Guthke, Appl. Microbiol. Biotechnol.
11. M. J. Danson and D. W. Hough, Trends Microbiol. 6, 51, 422–430 (1999).
307–314 (1998).
50. S. D. Nuttall et al., Biochem. J. 346, 251–254 (2000).
12. G. Schafer et al., Microbiol. Mol. Biol. Rev. 63, 570–595
51. D. A. Cowan, Comp. Biochem. Physiol. 118A, 429–438
(1999).
(1997).
13. W. D. Grant et al., Extremophiles: Microbial Life in Extreme
52. M. W. W. Adams, Annu. Rev. Microbiol. 47, 627–658 (1993).
Environments, Wiley-Liss, New York, 1998, pp. 93–132.
53. D. A. Cowan, in R. A. Herbert and R. Sharp, eds., Mol. Biol.
14. R. I. Amann et al., Microbiol. Rev. 59, 143–169 (1995).
Biotechnol. Thermophiles, Chapman & Hall, New York,
15. A.-L. Reysenbach et al., Appl. Environ. Microbiol. 60, 1992, pp. 1–43.
2113–2119 (1994).
54. D. W. Hough and M. J. Danson, Curr. Opin. Chem. Biol. 3,
16. K. Takai and Y. Sako, FEMS Microbiol. Ecol. 28, 177–188 39–46 (1999).
(1999).
55. R. Jaenicke and G. Bohm, Curr. Opin. Struct. Biol. 8,
17. K. L. Hershberger et al., Nature 384, 420 (1996). 738–748 (1991).
18. N. R. Pace, Science 276, 734–740 (1997). 56. S. Cavagnero et al., Biochemistry 34, 9865–9873 (1995).
19. S. M. Barns et al., Proc. Natl. Acad. Sci. U.S.A. 93, 57. K. S. P. Yip et al., Eur. J. Biochem. 255, 336–346 (1998).
9188–9193 (1996).
58. F. T. Robb and D. L. Maeder, Curr. Opin. Biotechnol. 9,
20. A. L. Reysenbach et al., Extremophiles 4, 61–67 (2000). 288–291 (1998).
21. L. Daniels, in M. Kates et al., eds., The Biochemistry of 59. J. Kim and J. S. Dordick, Biotechnol. Bioeng. 55, 471–479
Archaea (Archaebacteria), Elsevier Science Publ., Amster- (1995).
dam, The Netherlands, 1993, pp. 41–112.
60. G. A. Selleck and J. B. Chaudhuri, Enzyme Microb. Technol.
22. M. J. Danson, in M. Kates et al., eds., The Biochemistry of 25, 471–482 (1999).
Archaea (Archaebacteria), Elsevier Science Publ., Amster-
61. J. K. Lanyi and J. Stephenson, J. Bacteriol. 98, 611–616
dam, The Netherlands, 1993, pp. 1–24.
(1969).
23. S. Grant et al., Extremophiles 3, 139–145 (1999).
62. C. A. Raia et al., Biocatalysis 11, 143–150 (1994).
24. C. M. Preston et al., Proc. Natl. Acad. Sci. U.S.A. 93,
63. M. Klingeberg et al., Appl. Environ. Microbiol. 61,
6241–6246 (1996).
3098–3104 (1995).
25. R. Massana et al., Appl. Environ. Microbiol. 66, 1777–1787
64. C. Gerday et al., Trends Biotechnol. 18, 103–107 (2000).
(2000).
65. D. Nichols et al., Curr. Opin. Biotechnol. 10, 240–246 (1999).
26. B. L. Maidak et al., Nucleic Acids Res. 27, 171–173 (1999).
66. N. J. Russell, Extremophiles 4, 83–90 (2000).
27. C. Kato et al., Extremophiles 1, 117–123 (1998).
67. M. Summit et al., Extremophiles 2, 339–345 (2000).
28. J. O. McInerney et al., Appl. Environ. Microbiol. 61,
1646–1648 (1995). 68. K. Horikoshi, in K. Horikoshi and W. D. Grant, eds.,
Extremophiles: Microbial life in extreme environments,
29. H. Kim et al., Microbiology 146, 2309–2315 (2000).
Wiley-Liss, New York, 1998, pp. 155–181.
30. S. B. Bintrim et al., Proc. Natl. Acad. Sci. U.S.A. 94,
69. M. W. Bauer et al., Curr. Opin. Biotechnol. 9, 141–145
277–282 (1997).
(1998).
31. C. Schleper et al., Appl. Environ. Microbiol. 63, 321–323
70. D. Lévêque et al., Enzyme Microb. Technol. 26, 3–14 (2000).
(1997).
71. S. Janecek, Folia Microbiol. 43, 123–128 (1998).
32. D. P. Chandler et al., Microb. Ecol. 36, 37–50 (1998).
72. E. A. Galinski and B. J. Tindall, in R. A. Herbert and
33. Y. Kawarabayasi et al., DNA Res. 6, 83–101 (1999).
R. Sharp, eds., Molecular Biology and Biotechnology
34. H. P. Klenk et al., Nature 390, 364–370 (1997). of Extremophiles, Chapman & Hall, New York, 1992,
35. W. V. Ng et al., Proc. Natl. Acad. Sci. U.S.A. 97, pp. 76–114.
12176–12181 (2000). 73. U.S. patent 8:05,352,778, 1994, D. G. Comb et al.
36. D. R. Smith et al., J. Bacteriol. 179, 7135–7155 (1997). 74. A. M. Uhl and R. M. Daniel, Extremophiles 3, 263–267
37. C. J. Bult et al., Science 273, 1058–1073 (1996). (1999).
38. A. Chinen et al., Gene 259, 109–121 (2000). 75. T. Kobayashi et al., in K. Horikoshi and W. D. Grant, eds.,
39. D. L. Maeder et al., Genetics 152, 1299–1305 (1999). Extremophiles: Microbial Life in Extreme Environments,
40. Q. X. She et al., DNA Seq. 11, 183–192 (2000). Wiley-Liss, New York, 1998, pp. 155–181.
41. A. Ruepp et al., Nature 407, 508–513 (2000). 76. A. Obayashi et al., U.S. Patent 4:724 209 1989.
42. T. Kawashima et al., Proc. J. Acad. Ser. B. 75, 213–218 77. P. C. Croocker et al., Extremophiles 3, 3–9 (1999).
(1999). 78. Y. Sako et al., FEBS Lett. 415, 329–334 (1997).
276 ARCHAEA IN MARINE ENVIRONMENTS

79. X. L. Lin and J. Tan, Methods Enzymol. 248, 156–168 115. P. R. Norris et al., Extremophiles 4, 71–76 (2000).
(1995). 116. P. R. Norris and W. J. Ingledew, in R. A. Herbert and
80. P. Fusi et al., Experentia 47, 1057–1060 (1991). R. J. Sharp, eds., Molecular Biology and Biotechnology of
81. D. A. Cowan et al., J. Bacteriol. 172, 3654–3660 (1987). Extremophiles, Blackie, New York, 1993, pp. 115–142.
82. M. W. W. Adams, Annu. Rev. Microbiol. 47, 627–658 (1993). 117. D. W. Dew et al., in R. Amils and A. Ballester, eds.,
83. S. B. Halio et al., J. Bacteriol. 178, 2605–2612 (1996). Biohydrometallurgy and the Environment: Toward the
Mining of the 21st Century, Elsevier, Amsterdam, The
84. P. Volkl et al., Protein Sci. 3, 1329–1340 (1994).
Netherlands, 1999, pp. 229–246.
85. H. L. Wilson et al., J. Bacteriol. 182, 1680–1692 (2000).
118. P. R. Norris and J. P. Owen, in M. Ladisch and A. Bose, eds.,
86. X. Q. Liu, Ann. Rev. Genet. 34, 61–76 (2000). Harnessing Biotechnology for the 21st Century, American
87. I. Kalin et al., Mutat. Res. 283, 119–123 (1992). Chemical Society, Washington, D.C., 1993, pp. 445–448.
88. T. M. O’Connor et al., Thorax 55, 955–957 (2000).
89. A. Gliozi et al., Biochim. Biophys. Acta 735, 234–242 (1983).
90. K. Ring et al., in K. H. Schmidt, ed., Liposomes as Drug ARCHAEA IN MARINE ENVIRONMENTS
Carriers, Thieme, Stuttgart, Germany, 1986, pp. 101–123.
91. C. G. Choquet et al., Appl. Microbiol. Biotechnol. 42, ALISON E. MURRAY
375–384 (1974). Desert Research Institute
92. H. Komatsu and P. L.-G. Chong, Biochemistry 37, 107–115 Reno, Nevada
(1998).
93. M. De Rosa et al., Biosens. Bioelectron. 9, 669–675 (1994). Archaea were originally thought to exist solely in extreme
94. G. B. Patel and G. D. Sprott, Crit. Rev. Biotechnol. 19, anoxic, hot, or hypersaline environments. However,
317–357 (1999). technological advances in the past 20 years have led
95. L. Krishnan et al., J. Immunol. 166, 1885–1893 (2001). to a significant revision of this view. Archaea are now
96. L. Krishnan et al., Infect. Immun. 68, 54–63 (2000). recognized to be abundant in the largest habitable
97. A. Omri et al., J. Drug Target. 7, 383–392 (2000). regions on earth, the ocean, and in underlying sediments,
98. M. Sára et al., in T. J. Beveridge and S. F. Koval, eds.,
at times rivaling the abundance of the bacteria (1–3).
Advances in Bacterial Paracrystalline Surface Lay- Marine Archaea have been detected as free-living
ers, Plenum Publishing Corporation, New York, 1993, plankton associated with suspended particles in oxic
pp. 195–204. and anoxic zones of marine sediments, in the guts
99. E. Nußer and H. König, Can. J. Microbiol. 33, 256–261 of fish and holothurians, and as symbionts of marine
(1986). sponges. Phylogenetic diversity of the marine Archaea
100. T. J. Beveridge, Curr. Opin. Struct. Biol. 4, 204–212 (1994). is somewhat limited in comparison to the diversity of
101. U. B. Sleytr and T. J. Beveridge, Trends Microbiol. 7, marine organisms in the bacterial domain, although it
253–260 (1999). is evident that we do not currently know the extent
102. M. F. Mescher and J. L. Strominger, Proc. Natl. Acad. Sci. of archaeal diversity, as new groups are continually
U.S.A. 73, 2687–2691 (1976). being discovered (4–6). Although most of these newly
103. W. Baumeister et al., Can. J. Microbiol. 61, 1502–1506 described Archaea remain to be cultivated, promising
(1989). results from environmental genomic, stable isotopic, and
104. D. Pum and U. B. Sleytr, Trends Biotechnol. 17, 8–12 microautoradiographic analyses suggest that at least some
(1999). phenotypic characterization of these important organisms
105. M. Sára and U. B. Sleytr, J. Membr. Sci. 33, 27–49 (1987). is possible. Evidence suggests that the Archaea are
106. D. Pum et al., in U. B. Sleytr et al., eds., Immobilized probably important to present day biogeochemical cycles
Macromolecules: Application Potential, Springer-Verlag, and to past conditions on earth.
London, U. K., 1992, pp. 141–160.
107. D. Pum et al., Nanotechnology 2, 96–202 (1991).
NEWLY RECOGNIZED ARCHAEAL DIVERSITY
108. M. F. Roberts, Frontiers Biosci. 5, D796-D815A (2000).
109. L. O. Martins et al., Appl. Environ. Microbiol. 63, 896–902
Recent developments in the study of molecular evolu-
(1997).
tion have led to a unified view of life on earth in which
110. G. Lettinga and A. Van Haandel, in C. A. Covelo, ed.,
all life falls into three evolutionarily distinct forms, Bac-
Renewable Energy: Sources for Fuels and Electricity., Island
Press, New York, 1992, pp. 817–839.
teria, Archaea, and Eucarya (Fig. 1; 7,8) as opposed to
five kingdoms (Monera, Protista, Fungi, Plantae, and
111. C. Rivard, ‘‘Anaerobic Digestion of Municipal Solid Waste,’’
Second Biomass Conference of the Americas, National
Animalia) or two forms (Eucaryotic and Prokaryotic) as
Renewable Energy Laboratory, Portland, Ore., 1995, suggested by other theories based on morphology. Bio-
pp. 801–808. chemical, genetic, phenotypic, and now genomic data
112. F. O. Bryant and M. W. W. Adams, J. Biol. Chem. 264, collectively suggest that there are two kinds of prokary-
5070–5079 (1989). otes, the bacteria and the Archaea, and that the Archaea
113. J. Woodward et al., Nat. Biotechnol. 14, 872–874 (1996). are as distinct from the bacteria as they are from the
114. D. E. Rawlings et al., in R. Amils and A. Ballester, eds., Eukaryotes. Archaea were formally recognized to be dis-
Biohydrometallurgy and the Environment: Toward the tinct from bacteria in the late 1970s following the work
Mining of the 21st Century, Elsevier, Amsterdam, The of Carl Woese (7). There are a number of distinctions
Netherlands, 1999, pp. 777–786. on which these divisions have been based, supporting
 ARCHAEA IN MARINE ENVIRONMENTS 277

yces
Bacteria

Mitochon

m
Planctom

r iu
cte
Rh o

ba
m

temp
Esc Des ast us
docy chia vibr

vo
drion

te r
Agrobacteriu

Fla
her ulf

Me
ibac
Ch

clus

e low
Sy

tha
lor co

Methanococcus

s
ne

bu
Flex

a
Therm
op cc

Ar lasm
nob
ch

lo
Marin

og
l
o

Ch
lam Gloe o ax

ae
act

p
fer

mo
ydi ob M

ch
lo

ococ
io Ha

eriu
Chloro a ac et

er
bium ter ha

Th
no

m
cus
th llum
Leptonema Me
tha er
m Methanospiri
no us
Clostridium py Marine Gp. 1 low tem
rus p
Bacillus GpG .
p. 1 lo
w
p 2 t em
rium pSL 2SL 12 low tem p
bacte ter
Helio ac Gp. 2 p
rob 19
3 lo
Arth Pyr Sulfolo w tem
PS o p
pO s us ga Th dictiu bus
xu Roo m

Th pSL
er
6

fle herm oto

uif S 6

t mo

er 50
r o m Archaea
lo T er pr

m
Aq OP

Ch ote
EM1 ex

of
Th
pJ
pJP 2
us

ilu
p

P7

m
8
7
Co

Ho
prin

m
o
us

Zea
Cryptomonas 0.1 changes per site
a
Achly ia
s tar ra
Co phy
Babe ium

r
sia

Po
ec

G
ram

ia
rd
stelium

ia
Pa

Entamoeba
Dictyo

Phy

Tr
Naegleria

ich
saru

om

Eucarya
Trypan

on
Euglena

Enc

as
Va
eph

irim
oso

alito

orp
ma

ha
zoon

Figure 1. Universal phylogenetic tree based on small subunit rRNA sequences for the three
domains of life. (Reprinted (abstracted/excerpted) with permission from N. R. Pace, A molecular
view of microbial diversity and the biosphere, Science, copyright, 1997, American Association for
the Advancement of Science.)

their unique properties. For example, the Archaea har- lineages, the Crenarchaeota (traditionally thought of as
bor the replication machinery of eukaryotes, and the the sulfur-dependent hyperthermophiles), and the Eur-
basic metabolic attributes of the bacteria. Initially only yarchaeota (thought to comprise the methanogens, some
thought to occur in extreme (hot, saline, or anaerobic) thermophiles, and halophiles, which require salt at con-
environments representing ancient forms of life, represen- centrations greater than 1M).
tatives of the Archaea have been found globally in both From the perspective that scientists have been studying
terrestrial and marine environments, in oxygenated and marine bacteria for more than 200 years and the deep
anoxic habitats. The phylogeny of the archaeal division sea for 120 years, scientists may have unknowingly been
strongly supports the existence of two deep branching studying marine Archaea since the establishment of
278 ARCHAEA IN MARINE ENVIRONMENTS

marine microbiology. This is a result of the fact that from these studies is that the cultivated Archaea
under the microscope most prokaryotes basically look numerically represent a minority of the naturally
the same and Archaea cannot be readily distinguished occurring Archaea.
by direct cell counts. The history about the existence of
Archaea (as they are now formally recognized) in marine
DETECTION OF MARINE ARCHAEA
environments dates back to at least 1951 with the isolation
of a marine methanogen, Methanococcus vannielii from
Microorganisms are inherently difficult to recognize and
San Francisco Bay sediments (9). Marine methanogens,
study because of their small size and limited mor-
thermophiles, and hyperthermophiles have been isolated
phological characteristics. In fact, it is thought that
subsequently from extreme environments such as deep- microbiologists are able to cultivate less than 1% of
sea hydrothermal vents, firmly establishing that Archaea the microbial diversity on this planet (17). At times,
occupy marine ecosystems (10,11). Since then, a number sophisticated equipment and media are required for
of significant revelations about the extent of marine microbial cultivation. Methanogenic members of the Eur-
archaeal diversity, distribution, and abundance have yarchaeota were the first marine Archaea to be isolated.
irreversibly transformed the perception of the importance They are widely distributed in marine environments
of these organisms in marine environments, with the most (mostly sediments), and are thought to be the most
revolutionary discoveries following introduction of modern commonly cultivated archaeal groups in marine envi-
molecular biology to the study of marine environments. ronments (18). They are typically associated with low-
Before 1992, all Archaea were thought to exist solely in temperature anoxic sediments in which they carry out
extreme environments, in which they harbored unique the final step in carbon oxidation to methane. Isolation of
metabolisms and performed specialized biogeochemical marine methanogenic archaea was followed by character-
transformations for survival. On the whole, Archaea were ization of hyperthermophilic methanogens; these discov-
not considered to be important in marine environments eries precipitated a number of cultivation-based efforts at
because the habitats in which they flourished were hydrothermal vents (see HYDROTHERMAL VENTS: BIODIVERSITY
relatively scarce and geographically isolated. However, IN DEEP-SEA HYDROTHERMAL VENTS and HYDROTHERMAL VENTS:
after molecular tools enabled the discovery that there PROKARYOTES IN DEEP-SEA HYDROTHERMAL VENTS, this Ency-
are mesophilic and psychrophilic Archaea affiliated with clopedia).
both the Crenarchaeota and Euryarchaeota in seawater Cultivation approaches by necessity limit the diversity
and marine sediments (12–15), this view has changed sampled because the organisms are taken out of
dramatically. their environment. Thus, a major trend in the study
In recent years, molecular surveys have repeatedly of microorganisms living in the environment in the
indicated that cultivated microorganisms represent only past 15 years has turned to approaches that directly
a minority of the naturally occurring microbial diversity sample the diversity of organisms in natural ecosystems.
(16 and references therein); the planktonic Archaea are Significant advances in recognizing archaeal diversity can
exemplary of this observation. Today Archaea are often be attributed to recognition that small subunit ribosomal
found in molecular surveys of marine environments. rRNA (SSU rRNA) sequences can be used to study
Cultivation of the mesophilic and psychrophilic Archaea microbial diversity in natural environments (19). Thus far,
remains to be accomplished, although many experts it is thought that Archaea in the Crenarchaeota division
have extended efforts to do so. Although the closest harbor single ribosomal operons, whereas Euryarchaeota
relatives of the uncultivated marine crenarchaeotal can have as many as four. There is no single genomic
psychrophiles are the cultivated hyperthermophiles, there organization of these genes, because a variety of patterns
are a number of lines of evidence suggesting that (tandemly linked and randomly distributed) have been
these organisms belong to a phylogenetically distinct identified (20).
lineage. The marine crenarchaeotes appear to have a A number of SSU rRNA-based approaches have been
common ancestor with the hyperthermophiles and may adapted for studying marine Archaea. (1) SSU rRNA
have adapted to the colder habitats in which they have amplification and clone library generation and DNA
been detected. The approach for sampling organisms sequence analysis was performed in initial studies of
directly from the environment is tremendously valuable marine Archaea (i.e., 12,13). SSU rRNA sequences pro-
for detecting organisms and establishing relationships in vide robust data for biological diversity and phyloge-
an evolutionary framework, yet the molecular signatures netic studies. (2) Denaturing gradient gel electrophoresis
do not provide much in the way of physiological potential, (DGGE) analysis has been useful for assaying diversity
role, or capability. in a comparative framework. This approach was recently
This chapter addresses the topic of Archaea in marine used to describe archaeal diversity in deep marine sed-
environments from a global biological diversity and iments (4). (3) Terminal fragment length polymorphism
ecological perspective that has risen largely as a result (T-RFLP) analysis of SSU rRNA gene fragments is not as
of technological advances made with modern biology. sensitive as DGGE, although it is useful for discriminat-
Because the salinity of seawater does not approach ing between marine crenarchaeotal and euryarchaeotal
the levels required by the extreme halophiles, they sequences (21). A recent study demonstrated the utility of
are rarely recovered from marine samples (although this approach for detecting marine Archaea only after a 10-
they are common in ancient hypersaline seas), and cycle amplification (22) in which the different phylogenetic
thus will not be covered in this article. A key finding clusters could be resolved easily, and more closely related
 ARCHAEA IN MARINE ENVIRONMENTS 279

archaea could be resolved following multiple digestions.

Amplification-based approaches are typically not very use-
ful for quantifying specific prokaryotic groups, although
C40 H-H
recent work in this area shows promising results for esti-
mating the abundance of groups of organisms in natural
populations using highly sensitive nucleic acid detection
techniques such as the 5
nuclease assay (Taqman assay,
23,24). C40 1-R
For quantification, quantitative membrane hybridiza-
tions and fluorescently labeled in situ hybridization (FISH,
25,26) that was modified to increase sensitivity (1,2,27),
have proven to be quite effective for estimating natu- C40 2-R
ral abundances of marine Archaea. The first quantitative
marine Archaea data was determined using oligonu-
cleotide probes designed to be complementary to archaeal
SSU rRNA sequences (12). Application of oligonucleotide
hybridizations has enabled the detection of Archaea in
C40 3-Ri
numerous samples because it is amenable to multi-
Figure 2. Caldarchaeol-derived acyclic and cyclic biphytane
ple sample analysis and is relatively quick to perform.
structures derived from Antarctic picoplankton lipid extracts.
Membrane-based oligonucleotide hybridizations to marine Symbols: R, pentacyclic ring; i, isomer.
community rRNA has proven to be quite accurate in com-
parison to direct counts of whole cells that are fluorescently
labeled (1), and to gene copy numbers estimated by quanti- MARINE ARCHAEAL HABITATS, ECOLOGY, AND
tative PCR methodologies (23), although owing to intrica- DIVERSITY
cies with probe design, secondary structure conformation,
and sequence biases, overestimation, and underestimation Archaea have been isolated from or detected in a grow-
are known to occur. ing number of marine environments. The extremophiles
Another recent technological advance has enabled phy- (methanogens, thermophiles, or hyperthermophiles) are
logenetic identification of microbial cells (FISH) to be not considered to be numerically dominant in earth’s ocean
combined with metabolic activity (microautoradiography, systems as they are typically associated with hydrother-
28,29). The cells then are triple stained with 16S rRNA mal vents, anoxic marine sediments, or associations with
probe, DAPI (binds DNA), and the radiolabeled substrate guts of marine organisms. They are covered to some degree
(incorporated into proteins, DNA, or membranes), which later in the text and in other articles in this Encyclopedia.
Refer to Table 1 for a synthesis of the various studies that
can be visualized under phase contrast following emul-
have been completed in recent years reporting diversity
sification and developer treatment. This approach has
of marine Archaea using small subunit ribosomal RNA
enabled researchers for the first time to show incorpo-
sequences.
ration of a substrate into planktonic marine archaeal
cells (30). Marine Plankton
One of the unique and unifying characteristics of all
Archaea is that archaeal membranes are derived from Biological Diversity and Environmental Distribution.
ether linked isoprenoid-based alkyl chains, as opposed to Planktonic Archaea have been detected in both coastal
acyl-ester membrane lipids of all Bacteria and Eukaryotes. and open ocean ecosystems worldwide in which only
The major membrane lipid present in methanogens four divergent groups (Marine GI-IV) have thus far been
and halophiles is diphytanylglycerol (archaeol), whereas identified. All four groups remain uncultivated to date
hyperthermophiles and a few methanogens contain and are distantly related to known thermophiles and
methanogens. Cumulative evidence from studies of plank-
caldarchaeol, a dibiphytanyl-diglycerol tetraether (31).
tonic archaeal diversity indicate that the most commonly
Detection of archaeal lipids in marine waters and
recovered sequences are phylogenetically closely related in
sediments (32–34) has confirmed that (1) the Archaea are
a cluster known as the marine group I (Marine GI, 12), in
present in abundance in planktonic marine environments,
fact, these organisms might be one of the most abundant
(2) their lipids are similar to membranes of other archaea, microbial species on the planet (3).
and (3) because of the recalcitrant nature of the ether- The marine GI Archaea were originally detected
linked compounds, marine planktonic Archaea probably by phylogenetic analysis of environmentally-derived
contribute significantly to lipids in marine sediments, and sequences in 100 mm and 500 mm waters off the
hence, are present in the geologic record. Four major coast of California (13). At the same time marine GI
lipid compounds found exclusively in Archaea possessing and GII sequences and archaeal rRNA (using small
a C40 biphytane carbon skeleton containing zero to three subunit rRNA-targeted oligonucleotide hybridization)
cyclopentane rings have been detected from seawater were detected in coastal waters of the Santa Barbara
and sediment samples collected from diverse locations Channel and at Cape Cod (12). Interestingly, at that
(Fig. 2). time, no archaeal sequences were recovered from surface
280 ARCHAEA IN MARINE ENVIRONMENTS

waters of the central Pacific or Atlantic Ocean gyres. the anoxic zone of the Cariaco Basin (42). Most reports of
The GI sequences were distantly related to known marine Archaea suggest that they are free-living, although
sulfate-reducing hyperthermophiles grouping with the studies from the Pacific coast, Mediterranean coast of
Crenarchaeota, whereas the GII sequences were most Spain, and North Sea have found archaeal 16S rRNA
closely related to the Methanomicrobiales, a diverse group sequences to be associated with particles (12,21,39,51,54).
of methanogens (12). Since then numerous studies have An important confirmation that the planktonic archaea
most commonly detected close relatives of marine GI were indeed significant residents of the plankton resulted
Archaea in midwater to deep ocean depths (Table 1). from the gas chromatographic (GC) and GC-mass
Investigations of the extent of diversity of the marine spectroscopy (GC-MS) lipid analyses. Archaeal lipids
planktonic Crenarchaeota, in particular the marine GI containing a biphytane carbon (C40) skeleton, with zero to
cluster, have indicated that the diversity of this group three cyclopentane rings have been detected in seawater
is fairly restricted, perhaps only being represented by from Arthur Harbor, Antarctica, the Santa Monica Basin,
a few closely related species (52). Comparisons of 16S Cariaco Trench, and the Black sea, (Fig. 4), and the
rDNA (Fig. 3a) and ribosomal spacer (internal transcribed Cariaco Trench (34). Acyclic and cyclic biphytanes were
spacer, ITS) region suggest that there is a population level the most dominant lipids found in the Arabian Sea
structure (ecotypes) in this group that defines two surface- at water depths between 450 and 1,500 m (34). The
associated clades, an Antarctic clade, and a deep-ocean marine picoplankton and Cenarchaeum symbiosum were
clade, which harbors the greatest diversity. Diversity of found to contain caldarchaeol and derived acyclic and
the marine planktonic Archaea supports the view that cyclic dibiphytanes, in which the acyclic head to head
prokaryotes can have global distribution. It remains to be biphytane was the most abundant isoprenoid identified
shown how different the genomes and physiologies of such (Fig. 4). The value of these lipids as biogeochemical tracers
closely related ecotypes will be. is just beginning to be realized, indicating that these
Although it was initially thought that the planktonic organisms contribute significantly to carbon deposition in
archaeal diversity was restricted to the marine GI and GII, the sea (33).
this view has also changed. A cluster of euryarchaeaotal
Marine Archaeal Abundance. Several studies have
sequences classified as the marine GIII was reported from
reported direct counts of Archaea in marine coastal
sequences recovered from deep-sea samples collected in
and open ocean environments from surface waters to
the Pacific and Atlantic Oceans (36). Representatives from
the deep sea where values ranged from 1 × 103 to
this apparently deep-sea-adapted cluster have since been
nearly 2 × 105 Archaea × mL−1 seawater, representing 1
detected in the Aegean Sea (22), and at 3,000 m in the
to 60% of the DAPI-stained cells (Table 2). There is
Antarctic Polar Frontal region (Fig. 3b). Similarly, another
substantial variability in archaeal abundance (as well
marine archaeal lineage, referred to as marine GIV, was
as community diversity) with depth (Table 2). In the
detected in deep waters (3,000 m) in the Antarctic polar
coastal and continental shelf studies, it appears that the
front region. These deeply branching Euryarchaeota are
GI Archaea dominate below 60 m, whereas the GII cells
affiliated with the haloarchaea clade (Fig. 3b). Surveys
are mostly restricted to the upper 40 mm with seasonal
of various regions across the Drake passage, in the peaks in surface waters (1,53). In the open ocean, the
Mediterranean, and North Atlantic confirmed the presence GI Archaea peak between 150 and 500 mm and persist
of these organisms using primers specific to the marine in high relative abundance throughout the mesopelagic
GIV cluster in samples collected at depth (occasionally at and bathypelagic zones to the ocean bottom (5,000 mm
100 m, and more commonly below 500 m), and never in the Fig. 5). In a study at Station Aloha in the north pacific, the
upper water column [<100 m (6)]. Recent detection of the abundance of marine GI Crenarchaeota below 1,000 mm
marine GIV cluster is probably due to the divergence equaled the numbers of bacteria and approached up
of this group, which would not have been discovered to 40% of the DAPI-stained cells (Fig. 5). The marine
with several commonly used primer sets. Also, detailed GII cells at Station Aloha represented at most a few
phylogenetic investigations of the deep sea are scarce, and percentage of the DAPI-stained cells, although sporadic
it appears that although the marine GIV organisms are peaks in abundance were detected in the upper water
commonly present in deep-sea samples, they might not be column. Abundance of marine GII declined throughout
very abundant (6). the mesopelagic zone then increased again in deeper
Several studies have indicated that planktonic archaeal samples. The abundance of marine GI and GII planktonic
populations are vertically distributed in the water column Archaea have also been estimated using a Q-PCR Taqman
and that this distribution has phylogenetic coherence. assay, in which the marine GI and GII sequences
Studies in the Santa Barbara Channel (36,53) and could be detected at the level of 2.5 × 103 to 2.5 × 107
Monterey Bay (1,23) have established the ecological gene copies per reaction (sensitive enough to detect
distribution of the two groups of the marine Archaea. archaeal in one milliliter of seawater, in many locations,
The marine GI affiliated sequences appear to be more 23).
frequently isolated from deeper in the water column,
whereas marine GII Archaea have been typically detected Seasonal Variation. Most molecular-based studies have
intermittently in the surface waters (Table 1). Planktonic described archaeal abundance by single point surveys or
archaeal ecotypes have been shown to be associated with described distributions of the Archaea down the water
specific environments, such as a biogeographical region column, leaving far less known about the seasonal dynam-
studied in the Columbia River Estuary (CRE, 39) or in ics of the Archaea. Four temporal studies exist, which
Table 1. Summary of Marine Archaeal Groups Detected by Environmental rRNA Cloning and Sequencing Studies

Water Sediment
Habitat/Location Depth (m) Depth (cm) Temp. (° C) Archaeal Division/Group Reference

Seawater
North East Atlantic 500 Marine GI 35
North Atlantic off Scotland 0–500 7–13 Marine GI 35
Atlantic Ocean 1,000 7 Marine GI, GII 36
Atlantic Ocean coastal waters 15–20 Marine GI and GII 12
Atlantic Ocean 1,000 3–5 Marine GI 36
Pacific Ocean 500 and 3,000 6.8 and 1.6 Marine GI, GII, GIII 36
Santa Barbara Channel coastal waters 10–15 Marine GI and GII 12
Santa Barbara Channel 10–20 10–15 Marine GI 27
Santa Barbara Channel 0–200 Marine GI and GII 37
Pacific Ocean 100 7–9 Marine GI 38
Pacific Ocean 500 5–7 Marine GI 36,38
Pacific Ocean 3,000 3–5 Marine GI 36
Western Pacific Ocean 100 14 Marine GII 38
Western Pacific Ocean 500 5.5 Marine GI 38
Columbia River Estuary coastal waters Marine GI, GII, + others 39
Antarctic Peninsula 0 −1.8 Marine GI, GII 40
Antarctic 10–200 Marine GI 41
Antarctic, Polar Front 3,000 Marine GII, GIII, GIV 6
Southern Ocean coastal waters −1.8 Marine GI and GII 40
Mediterranean 5–25 Marine GI 41
Mediterranean, Biscane Bay 5–50 Marine GI 41
Mediterranean (Atlantic, 150–450 Marine GI 41
Pacific)
Eastern Mediterranean 10 13 Marine GI, GII, GIII 22
(Agean Sea)
Cariaco Basin 500–1,310 Marine GI, ANME 42

Sediment
Mariana Trench 11,000 Marine GI 43
NW Atlantic 1,500–4,500 0–27 Marine GI, marine benthic 4
groups (MBG) A–E
Buzzard’s Bay, Cape Cod 15 Marine benthic group B, D 15
Coastal Lagoon, SW France 1–2 Marine GI, distantly related GI 44
type, and euryarchaeotal
types
Eel River Basin, Pacific 521 13–15 and 22 Methanosarcinales, ANME-1 5

Hydrothermal Vent
Tachibana Bay, Japan 22 128 Korarchaeota, Crenarchaeotal 45
hyperthermophiles
Tachibana Bay, Japan 1–10 25–75 Marine GI 45
Suiyo Sea Mount, Myojin 972–1,398 100–300 Korarchaeota, Crenarchaeotal 18
Knoll, Iheya Basin hyperthermophiles
Suiyo Sea Mount, Myojin Marine GI 18
Knoll, Iheya Basin
Peles Vent, Loihi Seamount, Marine GI, Marine GII 46
Hawaii
North Gorda Ridge 2,000–2,800 Hyperthermophiles 47
Milos, Greece 0.2–11.7 21–100 Hyperthermophiles 48
Mid-Atlantic Ridge 112 at vent, Hyperthermophiles, 49
20–70 in Euryarchaeotal Marine
chamber sediment GII

Associated with other organisms

Holothurian Marine GI 50
Axinella mexicana, sponge Marine GI 27
Flounder, grey mullet Marine GI, GII, 21,51
Methanococcoides

281
282 ARCHAEA IN MARINE ENVIRONMENTS

(a)

Figure 3. Phylogenetic relationships among the marine archaeal clusters based on SSU rRNA
sequences. (a) Phylogeny of the marine GI sequences that group into a surface-associated cluster
(Crena-S1 and Crena-S2), an Antarctic cluster (Crena-A), and a deep-sea cluster (Crena-D).
Relationships for aligned sequences (600 bp) were determined by neighbor-joining analysis with
distances estimated using the Jukes-Cantor parameters, bootstrap values above 50% are indicated.
Reprinted (abstracted/excerpted) with permission from Garcia-Martinez and Rodriguez-Valera
Microdiversity of Uncultured Marine Prokaryotes: the SAR11 Cluster and the Marine Archaea
of Group I, Molecular Ecol., Copyright, (2000) Blackwell Publishers. (b) Phylogentic relationships
among the marine GII, GIII, and GIV (GIV shown in the circle). Maximum likelihood analysis
based on 1,047 aligned positions was used to determine the relationships of the sequences.
Representatives with full-length sequences are in bold. The scale bars correspond to 5 substitutions
per 100 positions. Reprinted (abstracted/excerpted) with permission from Lopez-Garcia et al.,
A novel haloarchaeal-related lineage is widely distributed in deep oceanic regions, Environ.
Microbiol., copyright, (2001), Blackwell Publishers.

paint partially divergent stories about the persistence austral winter to early spring, drops significantly with
of these organisms in marine waters. Antarctic archaeal the onset of spring, and stays low, (1 to 2%) until fall
vertical distribution differs from the patterns described (Fig. 6). A similar observation was made during the late
earlier, because the planktonic Crenarchaeota dominate winter to spring transition (October to November) at a sta-
the surface waters of the coastal Antarctic Peninsula. tion in coastal waters of the Gerlache Strait, Antarctica
The pattern of seasonal abundance is striking in these where relative proportions of the Archaea diminishing
waters, where the surface archaeal signal (approaching from 5 to 1% in surface waters, and 20 to 10% in deep
25% of the total rRNA signal), which is high in the (500 m) waters (56). The GI Archaea were dominant in
 ARCHAEA IN MARINE ENVIRONMENTS 283

(b)

Figure 3. (Continued)

C40 H-H (I)

C40 1R (II)
1.2
C40 2R (III)
C40 3Ri (IV)

1.0

0.8
Relative amount

0.6

Figure 4. Relative abundances of acyclic and cyclic biphy-

tanes in marine samples from the Black Sea sedi-
0.4 ment (BB sediment), Antarctic picoplankton lipid extract
(Antarctic), C. symbiosum, the marine GI sponge symbiont
(C. symbiosum), and the Santa Monica Basin suspended
particulate lipid samples (SMB3 and SMB4). Structures
0.2
for these compounds are shown in Figure 2. Reprinted
(abstracted/excerpted) with permission from DeLong et al.
Dibiphytanyl ether lipids in nonthermophilic Crenar-
0.0 chaeotes, Appl. Environ. Microbiol., copyright, 1998, Amer-
BS sediment Antarctic C.symbiosum SMB 3 SMB 4 ican society for microbiology.

the Gerlache Strait waters (69% of the archaeal signal), (and Eukaryal hybridization signal) and the planktonic
although a surface-associated GII rRNA hybridization archaeal hybridization signal (56,57,60). Hypotheses to
signal was also detected. In the Antarctic studies, neg- explain the difference in Antarctic archaeal vertical distri-
ative correlations were apparent between chlorophyll butions include increased biological competition in the
284 ARCHAEA IN MARINE ENVIRONMENTS

Table 2. Summary of Studies in Which Archaeal Abundances Were Characterized Using Various Molecular Techniques
in Marine Environments. The GI and GII Percentages are Related to the Total Archaeal Signal. Abbreviation for
the Methods used are as Follows: In situ (rRNA Targeted In situ Hybridization), RFLP Band (Restriction Fragment
Length Polymorphism Band Fragments), Q-hyb Membrane-Based Quantitative rRNA-Targeted Hybridization, Q-PCR
(Quantitative Polymerase Chain Reaction, Taqman Assay), RCA (Relative Clone Abundance Derived From Clone Library)

Location In situ - % of Total or Cells/mL Molecular Technique Reference

Plankton
Mediterranean 0 m 17% In situ 2
Mediterranean, 10 m 75 bands RFLP Band 22
Mediterranean 200 m 61% In situ 2
Mediterranean, 200 m GI and GII: 43% In situ 30
Woods Hole, 0 m Arch: 0.1% of total signal Q-Hyb 12
Atlantic central ocean, 0 m Arch 0% Q-Hyb 12
North Atlantic, coast of Scotland, near-shore Arch: 8%, Near-shore Arch 30% Q-PCR, RCA 35
North Atlantic, coast of Scotland, off-shore off-shore Arch: 28% RCA 35
North Sea, 1 m 3% In situ 55
Sargasso Sea, near Bermuda 10 m (28.C) 0 In situ 38
Monterey Bay (vertical profiles to 3,400 m) GI: 5 × 103 − 1 × 105 (1–20% max at In situ 1
60 m)
Monterey Bay (vertical profiles to 3,400 m) GII 9 × 103 − 1.8 × 105 (5–8% max at In situ 1
0–20 m)
Monterey Bay, 200 m GI and GII: 14.3% In situ 30
Santa Barbara Channel 0 m Arch: 2.3% of total signal Q-Hyb 12
Santa Barbara Channel, 0 m Arch: 8.01; GI: 0.63%; GII: 49% Q-Hyb 53
Santa Barbara Channel, 20 m Arch: 11.47%; GI: 1.95%; GII: 48% Q-Hyb 53
Santa Barbara Channel, 75 m Arch: 31.02%; GI: 22.27%; GII:12% Q-Hyb 53
Santa Barbara Channel, 150 m Arch 34.20; GI: 23.59%; GII: 8% Q-Hyb 53
Santa Barbara Channel, 300 m Arch 39.17%; GI: 30.49%; GII: 8% Q-Hyb 53
Playa del Rey, Coastal Pacific Ocean, 0 m 2% In situ 54
San Pedro Channel, California 100 m <5% In situ 2
San Pedro Channel, California 400 m 30% In situ 2
San Pedro Channel, California 600 m 40% In situ 2
Western Pacific, 100 m Arch: 23% RCA 38
Western Pacific, 500 m Arch: 60% RCA 38
North Western Pacific, Alaska Arch: 13.5% Q-Hyb 40
North Pacific, off Hawaii (0–5,000 m) Arch: 3 × 103 –5 × 104 , In situ 3
Central Pacific, 0 m Arch 0% Q-Hyb 12
Palmer Station, Antarctic, 0 m Arch: 18.5–26.2%, GI: 64%; GII: 36% Q-Hyb, RCA 40
Gerlache Strait, Antarctic Peninsula GI: 69% (average) Q-Hyb 56
Gerlache Strait, Antarctic Peninsula, 0 m Arch: 0–11% Q-Hyb 56
Gerlache Strait, Antarctic Peninsula, 75 m Arch: 5–21% Q-Hyb 56
Gerlache Strait, Antarctic Peninsula, 250 m Arch 7–30% Q-Hyb 56
Drake Passage Arch: 20–55%, GI: 93.4%; GII 2.9% Q-Hyb 57

Marine Sediments
Estuarine sediments, Japan, 15 cm Arch: 13.1%, Arch: 9.6% Q-PCR, RCA 24
Svalbard, Arctic sediments (0–15 cm) Arch: 0.4–6.4%; up to 1.9 × 108 , Arch: In situ, Q-Hyb 58
0.6–1.7%
Svalbard, Arctic sediments (0–31 cm) Arch: 1–4% Q-Hyb 59
Baltic Sea Arch: 1–12% Q-Hyb 59
NW Atlantic (0–2 cm) Arch: 2.5% Q-Hyb 4
NW Atlantic (8 cm) Arch: 8% Q-Hyb 4
NW Atlantic (0–13 cm) GI: 68% RCA 4
Eel River Basin, Pacific Ocean (22 & 13–15 cm) ANME: 82% at 22 cm & 55% at 13 cm RCA 5

Hydrothermal Vent
Effluent vent water, Iheya Basin, Japan Arch: 30%, Arch: 30% Q-PCR, RCA 24
Shallow vent sediments, Milos Greece Arch: 1–22% increasing with depth Q-Hyb 48

Antarctic summer, reaction to increased light irradi- In comparison with the dramatic seasonal variation
ance with the onset of summer, or other responses to seen in Antarctic archaeal rRNA abundance, studies in
the onset of summer, because Antarctic surface win- temperate waters off the coast of Santa Barbara and at
ter water is analogous to temperate and tropical deep Station Aloha in the North Pacific reported little correla-
water. tion between archaeal rRNA abundance and season (3,53).
 ARCHAEA IN MARINE ENVIRONMENTS 285

Figure 5. Temporal distribution of bacterial and crenarchaeotal cells indicated by contour plots
at the Hawaii Ocean Time-series station, ALOHA, in the North Pacific subtropical gyre. White
dots indicate sample depths. Contours represent percentages of the bacterial or crenarchaeotal
cells hybridizing with fluorescently labeled polynucleotide probes, in comparison to the numbers of
cells stained with the DAPI nucleic acid stain. Reprinted (abstracted/excerpted) with permission
from Karner et al. ‘‘Archaeal dominance in the mesopelagic zone of the Pacific Ocean,’’ Nature,
copyright, 2001, Macmillan publishers Ltd.

The differences in the latitude between the polar and rarely dominant in conditions of high biological activity or
more subtropical-temperate environments may partially prokaryotic abundance (53,60).
explain this discrepancy. In polar environments, the influ-
ence of strong periodicity in season has immeasurable
Marine Sediments
impact on biological organisms and processes. In the Santa
Barbara Channel (SBC) and the North Pacific, mesoscale Originally, methanogens were thought to exist only in
processes might play a more important role in influenc- anaerobic marine sediments. However, following the isola-
ing planktonic archaeal distribution and abundance. The tion of Methanococcus jannaschi at a marine hydrothermal
marine GI Archaea appear to consistently inhabit the vent and other thermophilic methanogens in subsequent
waters 75 mm or below, although during upwelling peri- studies, it is apparent that the marine methanogens
ods in the SBC when deep, nutrient-rich water is brought are widely distributed in anoxic marine habitats, where
to the surface, GI archaea were detected in surface waters. they have since been isolated from psychrophilic, ther-
The marine GII were detected on occasion, typically in mophilic, and hyperthermophilic marine environments. It
SBC surface waters (<20 m). Peaks in the marine GII is now thought that methanogens are the most cosmopoli-
rRNA signal succeeded phytoplankton blooms four out of tan of the presently cultured Archaea (61). Methanogens
nine times, suggesting a potential biological connection have been isolated from seawater particles (62,63), sug-
to their distribution. One trend seen with studies in the gesting the existence of anoxic microzones within the
Antarctic and in the SBC is that the GI Archaea are particles.
286 ARCHAEA IN MARINE ENVIRONMENTS

30

Archaeal HS1390 × 100 and group I HS1390 × 100

1996-97-Station N

25

20

Figure 6. Temporal variation in the plank- 15

tonic archaeal rRNA hybridization signal
(black boxes) and marine Group I hybridiza-
tion signal (open boxes) in the near-shore 10
waters of Arthur Harbor, Antarctica. Reprinted
(abstracted/excerpted) with permission from
Murray et al. ‘‘Seasonal and spatial variabil- 5
ity of bacterial and archaeal assemblages in the
coastal waters near Anvers Island, Antarctica,’’
Appl. Environ. Microbiol., copyright, 1998, 0
American Society for Microbiology. 1-Aug-96 1-Oct-96 1-Dec-96 1-Feb-97 1-Apr-97 1-Jun-97 1-Aug-97 1-Oct-97

Archaea have been detected in marine sediments of marine benthic Archaea was determined by FISH (58).
around the world in a number of molecular surveys in the Although generally a small fraction of the community, the
past six years. Sediment-associated Archaea consist of a Archaea were detectable mostly in the surface sediments
ubiquitous diverse consortium in which several previously (up to 6.4% of DAPI-stained cells), and decreased rapidly
unknown lineages have been repeatedly found. Similar in abundance, although they represented about 1% of the
to early findings in the plankton, molecular detection DAPI-stained cells even at 15 cm; however, difficulties
suggests that the cultivated groups of methanogens with detection of cells decreased substantially below
probably do not represent the abundance or diversity of 10 cm, probably because of low activities and rRNA
the benthic archaeal microbiota, although several studies amounts in those cells.
have detected some methanogen-related sequences. One The same archaeal lipids detected in the plankton have
of the first such studies conducted in salt marsh sediments also been found in abundance in marine sediments in
detected a diverse assortment of Euryarchaeal sequences the Cariaco Trench, Indian Ocean, Black Sea, Arabian
including some that clustered with different methanogenic Sea, and other geologic formations (32–34,65). Arabian
groups (Methanococcoides and Methanolobus) that either sea sediments had significant concentrations of acyclic
reduce carbon dioxide, or a suite of noncompetitive and cyclic biphytanes at depths between 0 to 141 cm (34).
substrates (14). Sequences affiliated with a halophilic In comparison to lipids detected in the plankton, marine
clade, which is surprising because halophiles that require sediments off California and Antarctica contained higher
at least five times the amount of salt that is found in the lipid concentrations, probably indicating their stability
ocean, were detected along with sequences clustering with and recalcitrant nature rather than the abundance of
the marine GII sequences found in the plankton. organisms in sediment environments, although this has
Kato and coworkers (43) sampled perhaps the deepest yet to be quantitatively determined (33). It may also
marine sediments on the planet in the Marianas Trench be possible to estimate the abundance of Archaea in
(11,000 m), and reported finding sequences that grouped marine sediments from lipid concentrations, as was
with the Marine GI cluster. Because the marine GIs suggested by a study conducted in Tokyo Bay in which
are thought to be planktonic in origin, the investigators high concentrations of archaeol, and caldarchaeol were
presume that the Archaea arrived from the middle detected in bay sediments (66). The Tokyo Bay study that
ocean with settling planktonic particles. Marine GI focused on quantifying archaeol, (to relate concentrations
affiliated sequences were also the most common phylotype to methanogens) also noted that dibiphytanyl glycerol
found in cold-seep sediments sampled in the Japan and cyclophytanyl glycerol were detected in appreciable
Trench [6,400 m (64)]. Diversity of Archaea in deep-sea concentrations in relation to the archaeol.
sediments sampled in the NW Atlantic appears to have
higher complexity, including four Crenarchaeota and Methane Hydrates. Methane hydrates consist of a frozen
two Euryarchaeota-affiliated sequences that are distantly combination of methane and water that can be found
related to each other (with 70 to 80% sequence divergence, in continental margin subduction zones, the deep sea,
4). Both RFLP analysis of clone libraries and DGGE and in the Polar Regions. Studies have found that active
analysis indicated that community diversity increased areas of microbial activity are associated with methane
with sediment depth. hydrates, although the rates of carbon remineralization
Using 16S rRNA-targeted oligonucleotide probes, are thought to be slow. These habitats have been known to
reports of archaeal abundances in Arctic sediments of harbor dense assemblages of chemoautotrophic sulfide-
Svalbard showed a trend of increasing archaeal rRNA oxidizing communities composed of Beggiatoa mats,
with depth (Table 2, 65). In the same region, abundance tubeworms, and clam fields (Calyptogena). The sulfide
 ARCHAEA IN MARINE ENVIRONMENTS 287

source supporting the chemoautotrophic community has exclusively marine origins (75). These groups all are salt
often been puzzling because sulfate-dependent carbon requiring, and do not appear to have a pressure require-
mineralization was not occurring in these sediments at ment. The predominant metabolism thus far known
appreciable rates. A relatively new, and very interesting involves the use of elemental sulfur as an electron acceptor,
story has unfolded in recent years regarding the diversity and organic carbon (amino acids/peptides, carbohydrates)
and metabolic capabilities of the microbial consortia as electron donor, and thus is heterotrophic. However,
associated with this habitat. Biogeochemical evidence autotrophic Archaea such as the methanogens (growing at
has suggested that methane is being consumed in temperatures from 2 to 100 ° C), and a couple of crenar-
anaerobic habitats before contacting the aerobic water chaeotal genera including Pyrolobus and Ferroglobus have
column or atmosphere. However, no organisms have also been isolated (75).
been isolated that perform this process (67). To get a Molecular surveys have also turned up new microbial
cultivation-free assessment of the microbial assemblage, diversity in hydrothermal environments (Table 1). One
Hinrichs and coworkers (5) characterized the small study at a shallow hydrothermal vent in Tachibana
subunit rRNA phylogenetic diversity and stable isotopic Bay, Japan (45) recovered organisms related to commonly
signatures of lipids found in methane hydrate deposits off cultivated hyperthermophiles in vent water samples,
California in the Eel River Basin. This study indicated as well as members of the Korarchaeota, a newly
the existence of a microbial community containing recognized third division of the Archaea, described by
sequences from known anaerobes including methanogens Barns and coworkers (76). Korarchaeotal sequences were
(Methanosarcinales), sulfate-reducing bacteria, gram- also recovered from a study at a deeper vent site (18).
positive bacteria, and a novel lineage distantly related to Marine GI archaeal sequences have been detected in the
the Methanosarcinales and Methanomicrobiales (5). The sediments associated with a hydrothermal vent (45), in
stable isotopic signatures of methanogenic archaeal lipid microbial mats associated with a venting sea mount (46),
biomarkers (archaeol and hydroxyarchaeol) were strongly in black smoker vent water, and a nearby chimney (18).
depleted in 13 C, indicating that they were derived from The explanation for this is either that the members of
methane, which has less 13 C than any other biologically this group actually can reside, and perhaps thrive, in
produced compound. Currently, effort is being extended to mesophilic-thermophilic sediments, or that because the
cultivate these organisms because they may be a key link marine group I have been thought to dominate the archaea
in the global methane cycle and past climates. found in the plankton, the sediment-derived sequences are
Rapid progress has led to a much better view of the from planktonic origin. At many hydrothermal vents, the
microbial processes occurring at methane hydrates and in bacterial diversity exceeds that of the archaeal diversity,
other anoxic marine sediments in general. Results from although within the Archaea, diverse sequences have been
several studies have detected Methanosarcinales-related reported (18,45,48). Between 3 to 26.5% of universally
sequences in methane-enriched environments (5,68–71) primed clone libraries were represented by archaeal
suggesting that these organisms, which were thought only sequences in a survey of several deep hydrothermal vent
to perform methanogenesis (using acetate as a carbon environments (Table 2), and archaeal rRNA abundances
source), may be operating methanogenesis in reverse. from 0.7 to 22.2% were detected from shallow vent
A key piece of evidence supporting the existence of a sites in sediments sampled in Milos, Greece (48). In situ
methane-oxidizing sulfate-reducing microbial consortia hybridization analyses confirmed cloning and sequencing
was demonstrated using FISH in several independent results that the black smoker chimney and hot sediments
studies, in which cells hybridizing to a Methanosarcinales- had high cell numbers, and relatively high levels of
targeted archaeal probe (ANME-probe) were aggregated archaeal diversity (Table 2).
in a cluster (1–11 mm in diameter) that was surrounded Although it is logistically challenging to do experi-
by cells hybridizing to probes for sulfate-reducing bac- mental research at the hydrothermal vents, Reysenbach
teria (70–73). These studies and others have also sug- and coworkers (49) succeeded in conducting short-term
gested that there are probably more archaeal species growth experiments in an in situ growth chamber at a
involved in the system because not all archaeal cells vent site on the Mid-Atlantic Ridge. This study reported
hybridize to the ANME-probe (72), the diversity of lipid substantial archaeal diversity in which representatives
structures and/or rRNA sequence diversity found indi- from the Thermococcales dominated the community (71%
cate the presence of other archaeal groups (70,74), and of the clones), followed by a significant representation
recently, a study combining FISH with the ANME-probe from the Archaeglobales group (22%). Thermococcales are
and secondary ion mass spectrometry presented striking known to be widespread at deep-sea hydrothermal vents.
evidence that the non-ANME archaeal cells had signifi- A couple sequences affiliated with the Thermoplasmales
cantly higher d13 CCratios than the Methanosarcinales-
(thermoacidophiles) were detected, which were not previ-
hybridizing clusters (71).
ously known to exist in hydrothermal vent ecosystems, and
a sequence that was in the marine GII lineage that appears
Hydrothermal Vents
to group with other deep-sea sediment Archaea (4).
Archaeal diversity has been studied in both deep One of the questions stemming from hydrothermal
(>4,000 m) and shallow (<100 m) hydrothermal vents vent research has been the origin and source of the
using a combination of isolation, enrichment, and hyperthermophilic microorganisms. In one case, Summit
cultivation-independent approaches. There are 18 marine and Baross (47) sampled a recently venting plume and
archaeal hyperthermophilic genera, 16 of which have were able to isolate thermophiles from 8 of 9 samples
288 ARCHAEA IN MARINE ENVIRONMENTS

at one plume, and 9 of 13 samples in a second plume, Probably the most promising evidence that the plank-
in contrast to background samples from the same area tonic GI Archaea are active came from a study in which
where they were unable to cultivate any thermophiles. FISH was used in tandem with microautoradiography.
Their conclusion was that there is a subsurface reservoir The results of this study suggest that 60% of the plank-
of thermophilic organisms, and suggested that these tonic Archaea were active, and that they incorporated
subsurface microbes likely thrive in the vicinity of mid- dissolved amino acids (30). This finding led the researchers
ocean ridges because of circulation and chemical energy to conclude that the archaeal cells were heterotrophic, and
sources. perhaps as active as the heterotrophic bacteria in the
same sample (by comparisons of the staining intensities).
ECOPHYSIOLOGICAL ATTRIBUTES OF MARINE ARCHAEA However, more work in this arena will be necessary to
determine whether the amino acids are being used in
Planktonic Archaeal Metabolism assimilatory or dissimilatory pathways.
It is possible that the marine GII, GIII, and GIV
Determining the metabolic role of the planktonic Archaea have different metabolic properties than the GI, given
has been the subject of many investigations, although the differences in their phylogenetic groupings as well as
few if any have made headway in accurately measuring their close relationship to the methanogens and halophiles.
growth rates, or even determining definitively their basic Fuhrman and Davis (36) postulated that the marine GII
metabolism. It is likely that the planktonic Archaea do and GIII could be methanogens — because oceanographers
influence oceanic biogeochemistry and elemental cycling have reported on midwater methane signals (77,78), but
because of their shear abundance (20% of the global little proof exists to support this hypothesis and some
picoplankton population). Regardless of their metabolism, reports have found the marine GIIs to bloom in surface
it is agreed that they contribute significantly to planetary waters, although these occurrences have been shown to
carbon biomass. It was suggested by Karner and occur following periods of upwelling (53). The potential
coworkers (3) that the GI Archaea are better adapted to that GIV phylotypes are halophilic is possible on the basis
the oceanic environment than most pelagic prokaryotes, of their close relationship to cultivated halophiles (6).
such that they have probably adopted a common adaptive
strategy that has enabled their expansive radiation.
Methanogenesis in Marine Environments
It is also the default assumption that, because of
the environmental conditions in which the marine GI Methanogenesis has been thought to occur in marine sedi-
Archaea prevail, they have aerobic heterotrophic lifestyles, ments either where there are low concentrations of sulfate,
although the evidence for this is scant (27,36). Scientific so that they are not in competition for substrates with
clues are mounting in a couple of directions, although sulfate reducers, or when the methanogens can use non-
until these organisms have been cultivated, or enough of competitive substrates such as methanol, methylamines,
their genome sequence is determined, this may remain or methionine. Both physiological and molecular data
largely guesswork. Stable isotopic analyses of membrane support both conditions. Methanogen-related sequences
lipids suggest that the archaeal lipids have 13 C 12 C ratios have been isolated from surface sediments. The topic of
ranging between 19.7 to 23 parts per mil (32,34). Hoefs methanogenesis will be covered elsewhere in this Encyclo-
and coworkers (32) proposed two alternative hypotheses pedia and will not be covered in more detail here.
to explain the results; (1) that the Archaea use dissolved
inorganic carbon as their carbon source, and do not Sulfur-Dependent Anaerobic Methane Oxidation
discriminate against 13 C as other primary producers do,
or (2) the Archaea carbon and energy source are acquired Anaerobic oxidation of methane (AOM) not only occurs
from low-molecular weight organics such as acetate and at methane hydrate sites, it is also suspected to occur in
methylated amines derived from algal carbohydrates and areas with substantial methane production and sulfate
proteins. The archaeal lipid signatures in the Arabian availability, including marine sediments, anoxic waters,
Sea were typically enriched in 13 C compared to algal deep continental margin sediments, and soda lakes (79).
sterols (34). Theoretical and field-based lines of evidence indicate that
One of the facets of Archaea is that they are AOM occurs in the presence of sulfate reduction by a
resistant to many antibiotics that inhibit bacterial consortium of organisms. Thermodynamics suggests that
activity. On the basis of this premise, experiments AOM would only be favorable if there was a removal
conducted on Antarctic planktonic microbiota set out to process for hydrogen, which could occur if sulfate reducers
measure the activity remaining following erythromycin were consuming the hydrogen. Lipid analyses from a well-
treatment of concentrated plankton in which the Archaea known methane-enriched mud volcano in the Eastern
represented between 1 to 9% of the assemblage (<1.6 µm Mediterranean, methane hydrates in the Eel River basin
fraction, (56), Massana and coworkers 1998). Although and Santa Barbara Channel, and seeps in the Black
these investigators failed to detect any activity signal Sea have indicated that archaeol and hydroxyarchaeol
higher than background in those experiments, they are present along with a number of other archaeal
suggested that the marine Archaea could be sensitive and sulfate-reducing bacteria derived lipids (68,70,74,80).
to erythromycin, perhaps do not incorporate leucine at These studies showed that not only were the archaeal
appreciable rates, were damaged in the concentration lipids strongly depleted in 13 C, but glycerol ethers and
procedure, or simply were not as active as their biomass fatty acid products from Bacteria were abundant, and
indicated. showed similarly depleted 13 C values in comparison to
 ARCHAEA IN MARINE ENVIRONMENTS 289

Eel river basin PC 26

Methanotrophic archaea
Bacterial syntrophic partner
Higher plants / algae

Eel river basin HPC 4

Figure 7. Carbon isotopic concentrations of sedimentary
lipids assigned to methanotrophic archaeal, bacterial syn-
troph, and plant/algal sources from three different methane
seeps. Note the similarities in 13 C depletion for the
Santa barbara basin PC 24 methanotroph and bacterial syntrophy in comparion to the
relatively 13 C enriched algal lipids, indicating that the
carbon source for the methanotroph and syntrophy was
methane, and not algae or terrestrially-derived carbon.
Reprinted (abstracted/excerpted) with permission from Hin-
−140 −120 −100 −80 −60 −40 −20 richs et al. ‘‘Molecular and isotopic analysis of anaerobic
methane-oxidizing communities in marine sediments’’, Org.
δ13C Geochem., copyright, 1999, Elsevier science.

control sites (Fig. 7). In addition, algae and plant lipids and 82 ° C in different cores, with an upper limit deter-
from the nearby sites was significantly enriched in 13 C mined to be 102 ° C, where Archaea such as Archaeglobus
in comparison to either the methanotrophs or bacterial profundus have been previously isolated (81). Dissimila-
syntrophs, indicating different carbon sources (CO2 versus tory reduction of iron is a relatively new discovery (82),
CH3 , Fig. 7). Control sites with similar carbon input and was recently shown to occur in the tandem with
as the hydrate sites consistently had less recalcitrant humic substance reduction (83). In this case, the extracel-
carbon in the sediments in comparison with methane lular quinones in the humics are reduced and the electrons
seep sites that had significant levels of residual carbon in are abiotically transferred to neighboring iron oxides, thus
the sediments, suggesting that sulfate removal processes, regenerating the humics.
rather than remineralization, are occurring at methane
seeps (68). A recent review proposed a mechanism for the
sulfur-dependent methane oxidation process, in which the ARCHAEAL ASSOCIATIONS WITH MARINE ORGANISMS
following reaction provided the highest energy yield of the
different reactions examined (79), and produces twice the Endosymbiosis has been a common feature of methano-
amount of energy predicted from a reverse methanogenesis genic euryarchaeotes (Methanomicrobiales and Methano-
reaction. bacteriales) including associations with ciliates, insects,
and a large number of vertebrates including humans. The
discovery that psychrophilic Crenarchaeota were found in
Methane Oxidation: 2CH4 + 2H2 O → CH3 COOH + 4H2 associations with marine organisms was initially quite
(1) surprising. Marine GI archaeal sequences were first found
Sulfate reduction: 4H2 + SO4 −2 + H+ → HS− + 4H2 O (2) to be associated with the midgut of a deep-sea deposit
feeding holothurian (50). Perhaps not as surprising in
Sulfate reduction: CH3 COOH + SO4 −2 → 2HCO3 − light of recent molecular surveys of marine sediment
+ HS− + H+ (3) habitats in which the Archaea appear to be common
constituents (although the GII-affiliated sequences and
Net reaction: 2CH4 + 2SO4 −2 → 2HCO3 − + 2HS− + 2H2 O other euryarchaeotal benthic Archaea appear to be
(4) more common), this was an important finding because
this was the first nonmethanogenic archaeal group to
Whether this reaction or something similar is actually be characterized in association with another organism.
occurring as a result of an archaeal-sulfate-reducing Since then a true symbiosis has been described for the
bacterium syntrophy remains to be seen. marine sponge A. mexicana, which hosts the marine GI
symbiont C. symbiosum in its tissues (27). The rationale
for determining that the association was symbiotic was
Archaeal Metabolism at Hydrothermal Vents
based on the high archaeal cell numbers, active cell
Hyperthermophilic microbes contribute significantly to division, and persistence of the association. Cenarchaeum
geochemical cycling in hot environments in present day symbiosum appears as a rod, with a specific nuclear
environments and in early earth conditions. There are region, which can be discriminated when using DNA and
three major metabolic processes currently thought to rRNA stains. Archaea have been detected in other marine
dominate hydrothermal vent habitats, reduction of sulfur sponges off the California coast, in the Antarctica, and in
compounds, methanogenesis, and iron reduction. Sulfate the Great Barrier Reef (84,85), although the associations
reduction rates measured in hydrothermal sediment sam- have not been described in detail. The exact nature of
ples collected in the Guaymas Basin had optimum of 70 ° C nutrient exchange, secondary metabolite production, or
290 ARCHAEA IN MARINE ENVIRONMENTS

other benefits from the archaeal-sponge interaction have taking place (87). The genome sizes of all archaeal
not been elucidated as of yet. autotrophs are reduced (approximately 1.6 Mbp), in com-
Marine GII archaeal sequences have also been found in parison with bacterial heterotrophs (approximately 2.5 to
the guts of marine fishes, flounder, and grey mullet, as well 6 Mbp). The first archaeal genome to be sequenced was
as flounder fecal pellets (21). Marine GI clones (86) were M. jannaschi, a methanogenic hyperthermophile, isolated
also recovered from the flounder digestive tract, although from a hydrothermal vent at a depth of 2,600 mm. The
marine GII sequences were more frequently represented novelty of finally having a whole genome sequence vali-
(22 of 29 clones). In a later study, Methanococcoides- dated suspicions that archaea are composed of bacterial
related archaeal sequences were detected in the flounder and Eukaryal genes. Although 56% of the M. jannaschi
digestive tract using a nested-PCR approach (51). The genome coded for hypothetical proteins, not matching
sequences from the fish digestive tracts were quite any sequences in sequence databases, the genes associ-
similar to those detected on suspended particles in the ated with central metabolism and cell division matched
surrounding water column, suggesting that a potential sequences of bacterial origin, whereas genes coding for
habitat for the anoxic methanogens lies in marine fish or information processing, such as DNA replication, tran-
other animal digestive tracts. Fecal pellets appear to be the scription, and translation were more eukaryotic like (88).
marine particulate source of the methanogens detected, One of the key features of an environmental genomics
although the alternative that the fish acquire the Archaea approach is that it provides access to the genomes of
via seawater ingestion cannot be discounted (51). uncultivated organisms. Several large archaeal DNA
fragments have now been sequenced from environmentally
ARCHAEAL GENOMICS AND BIOTECHNOLOGICAL derived samples (92–95). The first environmental contig
APPLICATIONS was assembled from a 40 Kbp sequence prepared from
a library assembled with DNA collected from a depth
A promising future for discovery in archaeal biology lies in of 200 m off the coast of Oregon. The results of this
genomics. The hopes entail developing an understanding of study confirmed that the genome fragment containing an
the basic biology of extremophiles in which comparisons to archaeal ribosomal operon contained several genes that
mesophilic and psychrophilic genomes will help determine had the highest similarity to archaeal sequences in the
how these organisms cope with the extreme conditions database, to the exclusion of the Bacteria or Eucarya (92).
they inhabit. Similarly, for the uncultivated Archaea, Evidence for the thermal sensitivity of enzymes from
environmental genomics may be one of the best ways the psychrophilic marine GI archaeon, C. symbiosum,
to advance knowledge concerning the basic metabolic resulted from a study in which the DNA polymerase
properties of these widespread yet poorly understood gene discovered by sequence analysis of a 42.4-kbp fos-
groups of organisms. mid was expressed in E. coli (93). The results indicated
Approximately ten archaeal genome sequences are that the DNA polymerase from C. symbiosum was inacti-
present in public databases, in which six are from vated at 40 ° C, which lent support to the nonthermophilic
organisms isolated from marine environments (Table 3). phenotype of this symbiont. Further screening of the
Genome organization is not well conserved among C. symbiosum-sponge genomic library indicated the pres-
those archaeal genomes sequenced. However, compara- ence of two highly related, but nonidentical rRNA sequence
tive genomics studies of these genomes are presently variants with 28 kbp of overlapping sequence (94). Results

Table 3. Archaeal Genomes Sequences from Completed Genomes and Environmentally Derived
Large DNA Fragments. Some Were Not Published (n.p.) at the Time of Printing. The Pyrococcus
Furiosus Genome is being Sequenced by the Center of Marine Biotechnology/University of Utah,
and the Pyrococcus abyssi Genome is being Sequenced by Genoscope

Species/Clone Name Environmental Origin of Organism/Sample Reference

Complete Genomes
M. jannaschii Submarine hydrothermal vent; East Pacific Rise, 88
Pacific Ocean
Archaeoglobus fulgidus Marine hydrothermal sediment; Vulcano island, Italy 89
Pyrococcus horikoshii Marine hydrothermal vent; Okinawa Trough vents, 90
Pacific Ocean
Aeropyrum pernix, K1 Marine solfataric vent, Japan 91
Pyrococcus furiosus Marine sand surrounding sulfurous volcanoes n.p.
Pyrococcus abyssi Deep-sea hydrothermal vent, South-east Pacific n.p.

Environmental Genome Fragments

Fosmid 4B7 Marine plankton, crenarchaeotal marine group I 92
Cenarchaeum symbiosum (Fosmids Marine sponge A. mexicana, symbiont 92,93
60A5, 101G10)
Fosmid 37F11 Marine plankton, euryarchaeotal marine group II 94
 ARCHAEA IN MARINE ENVIRONMENTS 291

indicate that there is microheterogeneity in the GI C. symbiosum DNA polymerase, G + C content of the GI
archaeal symbiont population within the sponge host and crenarchaeotal rRNA operon, detection of terrestrial GI
suggest that heterogeneity will be a complicating factor in Archaea, and so on indicated that the newly described,
genomic studies of natural populations. nonextremophilic Archaea belonging to the Crenarchaeota
A 60-kbp marine GII archaeal insert was detected from and Euryarchaeota branches of the Archaea are likely
a BAC library prepared from surface water picoplankton adapted to their cold, aerobic environments.
of the Monterey Bay (95). Sequence analysis of this The cold-adapted Archaea have probably been long-
large insert resulted in identification of 38 genes, with term residents of marine planktonic and sediment envi-
the majority of characterized genes having archaeal ronments, and have potentially played key roles in these
homologs, although eight genes appeared to have bacterial present day and potentially ancient ecosystems. Archaeal
homologs, suggesting the potential for lateral gene lipids have been detected in sediments from the Eocene
transfer. Structural analysis of protein domains of Era [50 million years ago, (32)], suggesting their long-
unknown genes suggests the potential for a membrane- term existence in marine environments. Furthermore,
associated proteolytic system to be encoded by a couple of new studies investigating deep cores samples in the North
the open reading frames, thus demonstrating the potential Atlantic provide compelling support for dominance of non-
for gene discovery in uncultivated organisms and the thermophilic Archaea in sediment horizons correlating
potential for predicting phenotypes from gene sequences. with oceanic anoxia (approximately 112 mya). Stable iso-
The marine Archaea are a valuable resource for topic analysis of the archaeal lipids indicated that they had
applications in the biotechnology field, providing enzymes chemoautotrophic lifestyles (102). In addition, the recog-
active from −2 to +130 ° C, biopolymers, and the nition that anaerobic methane oxidation is a significant
ability as whole cells to perform chemical manipulations archaeal-mediated biogeochemical process also suggests
that currently require application of a number of that the same process could have occurred on early earth
toxic chemicals. Hyperthermophilic archaeal enzymes when the atmosphere may have consisted primarily of
are being developed and used for their properties of methane in which the anaerobic methane oxidizers could
high-temperature activity and tolerance for activity in have played an important role in earth’s climate (79).
organic solvents (96). Recent studies indicate that marine Theories regarding the origins of life and early earth
hyperthermophilic diversity is still untapped, and the environmental conditions that led to the formation
potential for these organisms as a biotechnological of life have been an active area of research in the
resource, for example, in producing enzymes for research past decade. One popular theory postulates that the
applications (polymerases, restriction enzymes, etc.), for earliest forms of microbial life existed in hot anaerobic
the food industry, and in ‘‘green chemistry’’ is just environments. It is likely that the origin and evolution of
being realized (97,98). Through the study of homologs many metabolic reactions and pathways occurred in the
in genomic fragments of the psychrophilic Archaea, thermophiles and hyperthermophiles. For example, work
important clues to enzymatic temperature adaptation by Lovley and coworkers (83) demonstrating the ability
are likely to unfold (16,94). One of the areas of future of hyperthermophiles to reduce quinones in extracellular
research, in particular regards to extremophilic genomics humics may suggest a significant impact on early life
and developing an understanding of molecular adaptations on earth because extracellular quinones and iron oxides
to high temperatures will be predictive modeling and may have both been present in early earth conditions.
structural biology. Using signature 3-D structural themes The origins of the cold dwelling Crenarchaeota remain
in microbial proteins, modelers will be primed to apply a mystery as to the history of life on earth, since the
this approach to the understanding of enzyme structure discovery of these organisms opened up the potential for
for hyperthermophiles and other extremeophiles. psychrophilic origins from what was once argued to be
the hot origin of life for the Archaea. However, a recent
study conducted in Yellowstone appears to have recovered
ORIGINS AND EVOLUTION OF THE MARINE ARCHAEA
sequences that have a common ancestor (pSL12) with the
cold Crenarchaeota (76).
The ‘‘source’’ of the Archaea in marine planktonic envi-
ronments has been debated since their discovery nine
years ago. Theories that they arise from hydrother- CONCLUSION
mal vent plumes, from anaerobic gut environments in
marine fish or invertebrates (copepod guts) have mostly Marine Archaea have been isolated or detected from
been put to rest because their widespread distribution, environments with temperatures ranging the spectrum
abundance, and occurrence in aerobic cold habitats is found on this planet, from more than 100 ° C to less
commonplace (16). However, hyperthermophilic archaea than 0 ° C representing hyperthermophilic (>80 to 110 ° C),
can survive in low-temperature oxic seawater (99), and thermophilic (45 to 80 ° C), mesophilic (15 to 45 ° C),
have been found in hydrothermal vent plumes, which and psychrophilic (<15 ° C) lifestyles. Archaea have been
could potentially lead to the dispersal of these organ- found to exist as free-living organisms, attached to
isms far from the vent habitats (100,101). Similarly, particles, and associated with other marine organisms
methanogens could be resuspended from anoxic sedi- in a variety of habitats. Planktonic Archaea can be
ments, and have been isolated from oxic seawater (62,63). found throughout the water column from surface regions
Evidence including the C. symbiosum symbiosis and in coastal areas, to the deep abyss, and appear to be
other sponge-archaeal associations, thermolability of the globally distributed among the world’s oceans, including
292 ARCHAEA IN MARINE ENVIRONMENTS

the Southern Ocean, Atlantic, Pacific, North Sea, and 19. N. R. Pace, D. A. Stahl, D. L. Lane, and G. J. Olsen, Adv.
the Mediterranean. The planktonic Crenarchaeota may Microb. Ecol. 9, 1–55 (1986).
be one of the most abundant bacteria on the planet (3). 20. R. A. Garrett et al., Trends Biochem. Sci. 16, 22–26 (1991).
Archaea associated with marine sediments comprise a 21. G. Muyzer, E. C. D. Wall, and A. G. Uitterlinden, Appl.
diverse consortium in which the diversity is largely defined Environ. Microbiol. 59, 695–700 (1993).
by the geochemistry of the sediments. In contrast to 22. A. E. Murray, J. T. Hollibaugh, and C. Orrego, Appl. Envi-
the plankton, several members of sediment-associated ron. Microbiol. 62, 2676–2680 (1996).
archaeal communities have been cultivated, although 23. C. L. Moyer, J. M. Tiedje, F. C. Dobbs, and D. M. Karl, Appl.
molecular methods have revealed new diversity in nearly Environ. Microbiol. 62, 2501–2507 (1996).
all phylogenetic archaeal lineages. Diversity recognized 24. W.-T. Liu, T. L. Marsh, H. Cheng, and L. J. Forney, Appl.
thus far includes methanogens, relatives of halophiles, Environ. Microbiol. 63, 4516–4522 (1997).
anaerobic methane oxidizers, hyperthermophiles, and 25. M. J. E. C. van der Maarel, R. R. E. Artz, R. Haanstra, and
psychrophilic Crenarchaeota and Euryarchaeota. L. J. Forney, Appl. Environ. Microbiol. 64, 2894–2898
(1998).
The future holds promise for revealing detailed infor-
mation about the presently uncultivated GI and GII 26. M. M. Moeseneder, C. Winter, J. M. Arrieta, and G. J. Her-
ndl, J. Microbiol. Methods 44, 159–172 (2001).
planktonic lineages through applications of genomic tech-
nologies, which could eventually uncover mysteries of 27. M. T. Suzuki, L. T. Taylor, and E. F. DeLong, Appl. Environ.
Microbiol. 66, 4605–4614 (2000).
their physiological lifestyles, and the adaptations that
these organisms have undergone for survival in cold 28. K. Takai and K. Horikoshi, Appl. Environ. Microbiol. 66,
5066–5072 (2000).
environments. These organisms may hold the clues to
29. S. J. Giovannoni, E. F. DeLong, G. J. Olsen, and N. R. Pace,
understanding differences between thermophily and psy-
J. Bacteriol. 170, 720–726 (1988).
chrophily because the nearest cultivated relatives of the
30. E. F. DeLong, G. S. Wickham, and N. R. Pace, Science 243,
GI Crenarchaeotea are hyperthermophiles. Along similar
1360–1363 (1989).
lines of discovery, the interdisciplinary approaches being
31. R. Amann, W. Ludwig, and K. H. Schleifer, Microbiol. Rev.
applied to the anaerobic methane oxidizer consortia (69)
59, 143–169 (1995).
through combining stable isotopic signature analysis with
32. C. M. Preston, K. Y. Wu, T. F. Molinski, and E. F. DeLong,
molecular identifications promises to be a highly valuable
Proc. Natl. Acad. Sci. U.S.A. 93, 6241–6246 (1996).
approach that will likely be applied to increasingly more
33. H. Lee et al., Appl. Environ. Microbiol. 65, 1289–
microbial-biogeochemical investigations.
1297 (1999).
34. C. C. Ouverney and J. A. Fuhrman, Appl. Environ. Micro-
BIBLIOGRAPHY biol. 65, 1746–1752 (1999).
35. C. C. Ouverney and J. A. Fuhrman, Appl. Environ. Micro-
1. E. F. DeLong, L. T. Taylor, T. L. Marsh, and C. M. Preston, biol. 66, 4829–4833 (2000).
Appl. Environ. Microbiol. 65, 5554–5563 (1999). 36. M. Kates, in M. Kates, D. J. Kushner, and A. T. Matheson,
2. J. A. Fuhrman and C. C. Ouverney, Aquat. Ecol. 32, 3–15 eds., The Biochemistry of Archaea (Archaebacteria), Elsevier
(1998). Science Publishing, New York, 1992, pp. 261–296.
3. M. B. Karner, E. F. DeLong, and D. M. Karl, Nature 409, 37. M. J. L. Hoefs et al., Appl. Environ. Microbiol. 63, 3090–
507–510 (2001). 3095 (1997).
4. C. Vertiani et al., Appl. Environ. Microbiol. 65, 4375–4384 38. E. F. DeLong et al., Appl. Environ. Microbiol. 64, 1133–
(1999). 1138 (1998).
5. K.-U. Hinrichs et al., Nature 398, 802–805 (1999). 39. S. Schouten, M. J. L. Hoefs, and J. S. S. Damste, Org.
6. P. Lopez-Garcia, D. Moreira, A. Lopez-Lopez, and F. Rod- Geochem. 31, 509–521 (2000).
riguez-Valera, Environ. Microbiol. 3, 72–78 (2001). 40. J. O. McInerney, M. Mullarkey, M. E. Wernecke, and R. Po-
7. C. R. Woese and G. E. Fox, Proc. Natl. Acad. Sci. 74, 5088 well, Proc. R.l Soc. London. Ser. B 264, 1663–1669 (1997).
(1977). 41. E. F. DeLong, K. Y. Wu, B. B. Prezelin, and R. V. M. Jovine,
8. C. R. Woese, Microbiol. Rev. 51, 221–271 (1987). Nature 371, 695–697 (1994).
42. A. E. Murray et al., Aquat. Microb. Ecol. 18, 263–273
9. T. C. Stadtman and H. A. Barker, J. Bacteriol. 62, 269–280
(1999a).
(1951).
43. L. L. King, T. K. Pease, and S. G. Wakeham, Org. Geochem.
10. J. A. Romesser et al., Arch. Microbiol. 121, 147–153 (1979).
28, 677–688 (1998).
11. W. J. Jones et al., Arch. Microbiol. 136, 254–261 (1983).
44. R. Massana, A. E. Murray, C. M. Preston, and E. F. DeLong,
12. E. F. DeLong, Proc. Natl. Acad. Sci. 89, 5685–5689 (1992). Appl. Environ. Microbiol. 63, 50–56 (1997).
13. J. A. Fuhrman, K. McCallum, and A. A. Davis, Nature 45. A. E. Murray et al., Aquat. Microb. Ecol. 20, 129–145
(London) 356, 148–149 (1992). (1999b).
14. M. A. Munson, D. B. Nedwell, and T. M. Embley, Appl. 46. S. G. Acinas, R. Rodrigues-Valera, and C. Pedros-Alio,
Environ. Microbiol. 63, 4729–4733 (1997). FEMS Microbiol. Ecol. 24, 27–40 (1997).
15. C. Vetriani, A.-L. Reysenbach, and J. Dore, FEMS Micro- 47. J. A. Fuhrman, K. McCallum, and A. A. Davis, Appl. Envi-
biol. Lett. 161, 83–88 (1998). ron. Microbiol. 59, 1294–1302 (1993).
16. E. F. Delong, Science 280, 542–543 (1998). 48. R. Massana et al., Limnol. Oceanogr. 43, 607–617 (1998).
17. K. Takai and K. Horikoshi, Genetics 152, 1285–1297 (1999). 49. R. Massana, E. F. DeLong, and C. Pedros-Alio, Appl. Envi-
18. R. Amann, Syst. Appl. Microbiol. 23, 1–8 (2000). ron. Microbiol. 66, 1777–1787 (2000).
 ARCHAEA IN SOIL HABITATS 293

50. J. Garcia-Martinez and F. Rodriguez-Valera, Molecular 83. R. D. Pancost et al., Appl. Environ. Microbiol. 66, 1126–
Ecol. 9, 935–948 (2000). 1132 (2000).
51. J. A. Fuhrman and A. A. Davis, Mar. Ecol. Prog. Ser. 150, 84. L. Elsgaard et al., Geochim. et Cosmochim. Acta. 58,
275–285 (1997). 3335–3343 (1994).
52. B. C. Crump and J. A. Baross, FEMS Microbiol. Ecol. 31, 85. M. Vargas, K. Kashefi, E. L. Blunt-Harris, and D. R. Lovley,
231–239 (2000). Nature 395, 65–67 (1998).
53. M. J. E. C. van der Maarel, W. Sprenger, R. Haanstra, and 86. N. R. Pace, Science 276, 734–740 (1997).
L. J. Forney, FEMS Microbiol. Lett. 173, 189–194 (1999). 87. J. L. Stein et al., J. Bacteriol. 178, 591–599 (1996).
54. V. M. Madrid, G. T. Taylor, M. I. Scranton, and A. Y. Chis- 88. D. R. Lovley et al., Chem. Geol. 169, 289–298 (2000).
toserdov, Appl. Environ. Microbiol. 67, 1663–1674 (2001). 89. J. O. McInerney et al., Appl. Environ. Microbiol. 61,
55. F. O. Glockner, B. M. Fuchs, and R. Amann, Appl. Environ. 1646–1648 (1995).
Microbiol. 65, 3721–3726 (1999). 90. C. M. Preston, Ph.D. Thesis, University of California, Santa
56. A. E. Murray et al., Appl. Environ. Microbiol. 64, 2585– Barbara, Calif., 1998.
2595 (1998). 91. H. S. Webster et al., Environ. Microbiol. 3, 19–31 (2001).
57. W. B. Whitman, F. Pfeifer, P. Blum, and A. Klein, Genetics 92. C. J. Bult et al., Science 273, 1058–1073 (1996).
152, 1245–1248 (1999). 93. H. P. Klenk et al., Nature 390, 364–370 (1997).
58. F. J. Cynar and A. A. Yayanos, Curr. Microbiol. 23, 89–96 94. Y. Kawarabayasi et al., DNA Res. 5, 55–76 (1998).
(1991).
95. Y. Kawarabayasi et al., DNA Res. 6, 83–101, 145–152
59. J. M. Sieburth, Mar. Ecol.: Prog. Ser. 95, 67–80 (1993). (1999).
60. K. Sahm and U.-G. Berninger, Mar. Ecol.: Prog. Ser. 165, 96. C. Schleper, R. V. Swanson, E. J. Mathur, and E. F. De-
71–80 (1998). Long, J. Bacteriol. 179, 7803–7811 (1997).
61. A. Cifuentes et al., Appl. Environ. Microbiol. 66, 1715– 97. C. Schleper et al., J. Bacteriol. 180, 5003–5009 (1998).
1719 (2000). 98. O. Beja et al., Environ. Microbiol. 2, 516–529 (2000).
62. C. Kato, L. Li, J. Tamaoka, and K. Horikoshi, Extremophiles 99. O. Lecompte et al., Genome Res. 11, 981–993 (2001).
1, 117–123 (1997).
100. D. E. Robertson et al., Soc. Ind. Microbiol. News 46, 3–8
63. L. Li, C. Kato, and K. Horikoshi, Mar. Biotechnol. 1, (1996).
391–400 (1999).
101. H. Huber and K. O. Stetter, J. Biotechnol. 64, 39–52 (1998).
64. K. Ravenschlag, K. Sahm, and R. Amann, Appl. Environ.
102. M. T. Bustard, J. G. Burgess, V. Meeyoo, and P. C. Wright,
Microbiol. 67, 387–395 (2001).
J. Chem. Tech. Biotechnol. 75, 1095–1109 (2000).
65. M. Sunamura, Y. Koga, and K. Ohwada, Mar. Biotechnol. 1,
103. H. W. Jannasch, C. O. Wirsen, S. J. Molyneaux, and T. A.
562–568 (1999). Langworthy, Appl. Environ. Microbiol. 54, 1203–1209
66. B. Schink, Microbiol. Mol. Biol. Rev. 61, 262–280 (1997). (1988).
67. K.-U. Hinrichs et al., Org. Geochem. 31, 1685–1701 (2000). 104. R. Huber et al., Nature 345, 179–181 (1990).
68. K. A. Bidle, M. Kastner, and D. H. Bartlett, FEMS Micro- 105. J. F. Holden, M. Summit, and J. A. Baross, FEMS Micro-
biol. Lett. 177, 101–108 (1999). biol. Ecol. 25, 33–41 (1998).
69. V. J. Orphan et al., Appl. Environ. Microbiol. 67, 1922– 106. Kuypers et al., Science 293, 92–94 (2001).
1934 (2001). 107. V. J. Orphan et al., Science 293, 484–487 (2001).
70. A. Boetius et al., Nature 407, 623–626 (2000).
71. J. R. Marchesi et al., FEMS Microbiol. Ecol. 34, 221–228
(2001).
72. V. Thiel et al., Mar. Chem. 73, 97–112 (2001).
ARCHAEA IN SOIL HABITATS
73. M. W. W. Adams, J. Appl. Microbiol. Symp. Suppl. 85,
HOLLY M. SIMON
108S–117S(1999).
ROBERT M. GOODMAN
74. K. Takai and Y. Sako, FEMS Microbiol. Ecol. 28, 177–188
University of
(1999).
Wisconsin–Madison
75. S. M. Barns, C. F. Delwiche, J. D. Palmer, and N. R. Pace, Madison, Wisconsin
Proc. Natl. Acad. Sci. 93, 9188–9193 (1996).
76. C. L. Moyer, J. M. Tiedje, F. C. Dobbs, and D. M. Karl, Deep- Archaea comprise a heterogeneous group of microorgan-
Sea Res. II 45, 303–317 (1998). isms, many of which contribute substantially to terrestrial
77. S. M. Sievert, W. Ziebis, J. Kuever, and K. Sahm, Microbiol. biosphere processes through their unique metabolisms.
146, 1287–1293 (2000). Their ‘‘extremophile’’ image in popular literature as well
78. A.-L. Reysenbach, K. Longnecker, and J. Kirshtein, Appl. as in scientific circles, derives from the fact that archaea,
Environ. Microbiol. 66, 3798–3806 (2000). traditionally, have been isolated from ‘‘extreme’’ habitats,
79. M. Summit and J. A. Baross, Deep-Sea Res. II 45, that is, high sulfide, hypersaline, anoxic, or hyperthermal.
2751–2766 (1998). Today, molecular evidence shows that soils and aquatic
80. M. A. DeAngelis and C. Lee, Limnol. Oceanogr. 39, 1298– environments of a ‘‘nonextreme’’ nature are also habitats
1308 (1994). of novel archaea with as yet unknown metabolic proper-
81. D. M. Karl and B. D. Tilbrook, Nature 368, 732–734 (1994). ties.
82. D. L. Valentine and W. S. Reeburgh, Environ. Microbiol. 2, Here, we present an overview of soil as a habitat
477–484 (2000). and a review of what is known about archaea in
294 ARCHAEA IN SOIL HABITATS

soils. We describe what is known or deduced about the

ecological properties of archaea found in both extreme
and nonextreme soil habitats. Research on ecological
roles of archaea in soils is scarce except with regard to
methanogenesis, which is a globally important process
carried out exclusively by members of the Archaea.
We give an overview of the ecology of methanogenic
archaea; however, this topic is more thoroughly discussed
by R. Conrad (Flooded Soils) in another article in this
Encyclopedia on the microbiology of flooded soil habitats.

SOIL AND THE BIOSPHERE

Soil is a major reservoir of microbial biomass, comprising

from 10 to 50% of the Earth’s terrestrial, prokaryotic
carbon, with an estimated 26 Pg (1 Pg = 1015 gram) of
organic C in 2.6 × 1029 cells (1). The global impact of
prokaryotic metabolism in soil habitats is significant. For
example, symbiotic N2 fixation provides approximately
80% of the biologically fixed reduced nitrogen on land (2).
Microorganisms in both water-saturated, anoxic, and
drier, well-aerated soils are also important influences
on atmospheric chemistry; they make a net contribution
of about 60% of the CH4 present (3) and contribute
significantly to the cycling of carbon monoxide (CO),
carbonyl sulfide (OCS), nitrous oxide (N2 O), and nitric
oxide (NO) [(4) and refs. within]. A complete account of the
contribution of archaea to the biomass and metabolism of
soil prokaryotes awaits results and future studies on the
prevalence, metabolic repertoire, and microhabitats of soil
archaea.

Complexity of Soil
Although familiar and seemingly simple, soil is so complex
as a microbial habitat that many of its secrets remain
well guarded as the twentyfirst century dawns. Soil is a
heterogeneous mixture of living organisms, decomposing
organic materials, and inorganic substances (Fig. 1).
Differences in particle composition, size, shape, density,
and charge, combined with climatic variation, form the
basis for diversity in colonization of soil by different
life-forms. Plants, animals, and fungi, the latter two Figure 1. Electron micrograph of soil (scale bar 5 µm). Bac-
mainly microscopic, inhabit soil. Billions of prokaryotes teria (B) are usually associated with organic matter that still
contains carbohydrate (e.g., cell walls, W). Other highly lignified
also live in each gram of soil. Physical, chemical, and
and convoluted organic matter (O) does not support bacteria, but
biological heterogeneity in soil creates a network of
there are numerous microorganisms (0.3 µm diameter) scattered
microenvironments; these are strongly influenced by throughout the clay (*). P is a pore, approximately 1 µm in diam-
temperature and can vary widely in ionic strength and eter. Reproduced from R. C. Foster, Quaestiones Entomologicae
composition, pH, and water and redox potential. The 21, 609–633 (1985).
diverse prokaryotes that adapt to and exploit these
microniches match this diversity in habitat with equally
diverse metabolic capacities. and eukaryotic metabolisms influence the temperature
regulation of the Earth through CO2 turnover, whereas
Soil Microorganisms and Biosphere Functioning CH4 cycling (3) and the recycling of crucial minerals such
as N, S, and P are largely dependent on prokaryotic
Potential productivity in both natural and cultivated
activities (4).
systems is directly related to soil organic matter
concentration and turnover, which in turn supplies most
Diversity of Soil Microorganisms
macronutrients required for plant growth. Through their
roles in the decomposition of soil organic matter, soil Soil is an immense reservoir of microbial diversity. The
microorganisms are largely responsible for terrestrial rich diversity of soil microorganisms can be glimpsed
ecosystem development and functioning. Both prokaryotic through use of the agar plating technique; a broad
 ARCHAEA IN SOIL HABITATS 295

array of different colony morphologies can be recovered and order (24). Examples of characteristics exclusive to
upon plating a soil sample. Knowledge of the microbial archaea include the sequence and structure of particular
composition of natural environments was, for a long time, regions of ribosomal RNA (rRNA) (25) and the composition
based mainly on those organisms that were amenable to of membrane lipids and lipid linkages (26).
growth in vitro using such standard approaches. However,
it was also known that a large discrepancy existed Taxonomy and Phylogeny of Archaea
between culturable and total microscopic cell counts in
There are two phylogenetic divisions of the domain
natural habitats, and soil habitats in particular (5,6).
Archaea represented by cultured isolates, Crenarchaeota,
This observation suggested many unknown and not
and Euryarchaeota (27), and a third proposed division,
easily cultured microorganisms were present. Torsvik
Korarchaeota (28), for which only rRNA sequences have
and coworkers elegantly demonstrated this concept using
been recovered (28–30). All cultured representatives of the
DNA-DNA reassociation kinetics, showing that thousands
Crenarchaeota are extremely thermophilic sulfur metabo-
of prokaryotic genomes are present in a gram of
lizers. Examples include members of the genera Sulfolobus
soil, compared with a relatively less complex genome
and Thermoproteus. The Euryarchaeota comprise cul-
mixture derived from the culturable subset (7,8). A
tured members with a more diverse metabolic repertoire,
staggering diversity of microorganisms is also inferred
including methanogens, for example, Methanothermus and
from the results of culture-independent methods that
Methanosarcina, extreme halophiles (‘‘salt-loving’’) such as
exploit direct recovery of genomic DNA and analysis to
Halobacterium and Natronococcus, and the thermophilic,
derive phylogenetic types (phylotypes) from sequences of
wall-less acidophile Thermoplasma, in addition to sulfur-
ribosomal RNA; such analysis reveals that the majority of
metabolizing hyperthermophilic archaea, for example,
phylotypes recovered from soil are novel (9–14).
Archaeoglobus and Thermococcus.
The Archaea serve as an example to illustrate the
discrepancy between culture-based and nonculture-based
Archaea in Soil
knowledge derived from studies of microorganisms in the
environment. Most archaea have not been cultured from In this article, we describe the distribution of archaea
habitats considered to be nonextreme, such as mesophilic in what are considered conventional soils, both extreme
soil or oxygenated marine surface waters. On the basis of and nonextreme (see also METALS: MICROBIAL PROCESSES
culture-independent studies, it now appears that archaea AFFECTING METALS). Archaea typically have been isolated
are not only present in such habitats, but are reasonably from extreme (high sulfide or saline, elevated tempera-
(and in some cases very) abundant members of the tures) and anoxic environments, resulting in the generally
microbial assemblages there (Table 1). accepted idea that they represent a primitive group exiled
from the (now) more common aerobic and nonextreme envi-
AN OVERVIEW OF ARCHAEA IN SOIL ronments of contemporary earth. Until recently, this view
extended to soil habitats as well. Archaea were previously
Archaea: The Third Domain and almost exclusively cultured from soils of an extreme
nature, such as acidic hot soils heated by solfataric gases (a
The discovery of Archaea as a third domain of life solfatara is defined as a volcanic fissure that gives off sul-
revolutionized evolutionary thinking by establishing that furous vapors and steam) and evaporative saline soils. An
prokaryotes are not monophyletic (23). Although archaea exception to this observation is cultured isolates recovered
are prokaryotes and resemble bacteria in morphology from anoxic soils. Archaea isolated from these environ-
(single cells lacking internal compartmentalization), in ments encompass diverse groups and physiologies. More
evolutionary lineage they are closer to the Eukarya than recently, however, the known diversity among Archaea
to bacteria. Archaea, in fact, share some characteristics has been broadened considerably more by the inclusion of
with eukaryotes, for example, the homology of archaeal new members represented solely by small subunit (SSU)
and eukaryotic RNA polymerases (24), but in other ways rRNA gene sequences recovered from temperate habitats,
are similar to bacteria, for example, gene composition including mesophilic and low-temperature soils that have
not previously yielded isolates of Archaea.
Table 1. Estimates of Abundance of Nonthermophilic
Distribution of Archaea in Soil
Crenarchaeota
Habitat Abundance Ref. Archaea have been isolated by culturing from extreme
and anaerobic soils located around the globe (Table 2).
Oxygenated coastal 2% total rRNA 15 Examples include solfataric fields in Italy, the Azores,
waters United States, Iceland, and New Zealand, and rice
Antarctic surface waters 34% prokaryotic biomass 16 paddies in Japan and Italy. Even more novel SSU
Santa Barbara channel 20% total rRNA 17 rRNA sequences of uncultured Archaea have now been
Antarctic waters 1–24% total rRNA 18 recovered from soils all over the globe. Flooded soils
Pacific mesopelagic zone 30–50% prokaryotic biomass 19
and wetlands yield sequences representing members of
Axinella mexicana 60% prokaryotic cells 20
both Euryarchaeota and Crenarchaeota (11,31), whereas
(marine sponge)
Michigan soil 1.4% total rRNA 21 only sequences representing the Crenarchaeota have
Tomato roots 3.3–16% prokaryotic cells 22 been recovered thus far from dry soils (9,10,21,32,33).
The phylogeny of archaeal sequences from the various
296 ARCHAEA IN SOIL HABITATS

Table 2. Examples of Archaea Isolated from Soils

Division Order Family Isolate Type Soil Habitat Reference

Euryarchaeota Methanobacteriales Methanobacteriacea culture, SSU Italian rice paddy soil, 11,34–38
rRNA bog, pasture, and
sequence. marsh
Euryarchaeota Methanobacteriales Methanothermaceae culture, solfataric mud, 36,115
Iceland
Euryarchaeota Methanococcales Methanococcaceae culture, SSU black shore mud, San 37,39
rRNA Francisco Bay
sequence
Euryarchaeota Methanomicrobiales Methanomicrobiaceae SSU rRNA rice paddy soil, Italy 11,34,36
sequence,
Euryarchaeota Methanomicrobiales Unknown SSU rRNA rice paddy soil, Italy 11,34,36
sequence,
Euryarchaeota ‘‘Methanosarcinales’’ Methanosarcinaceae culture, SSU marshes, wetlands, 11,34,36–38,40
rRNA Italian rice paddy,
sequence, and garden soil
Euryarchaeota ‘‘Methanosarcinales’’ ‘‘Methanosaetaceae’’ SSU rRNA rice paddy soil, Italy 34,36
sequence,
Euryarchaeota ‘‘Methanosarcinales’’ Unknown SSU rRNA rice paddy soil, Italy 11,34,36
sequence,
Euryarchaeota Unknown Unknown SSU rRNA rice paddy soil, Italy 11,34,36
sequence,
Euryarchaeota ‘‘Thermoplasmatales’’ Unclear culture solfataras, coal refuse 41,42
piles
Euryarchaeota Halobacteriales Halobacteriaceae culture hypersaline soils in 43–45
Spain & former
USSR
Crenarchaeota Sulfolobales Sulfolobaceae culture acidic hot soils 46–48
Crenarchaeota Thermoproteales Thermoprotaceae culture Uzon caldera, Russia 41,49
Crenarchaeota Thermoproteales Desulfurococcaceae culture solfatars, Indonesia 41
Crenarchaeota Unknown Unknown SSU rRNA agricultural, boreal 9,11,33,34,36,50
sequence forest, rice paddy
soil

different soil habitats ranges from having close affiliation paddies, and landfills are major contributors to the
to members of known groups of Archaea, such as global CH4 cycle, contributing about 60% of atmospheric
Methanobacterium and Methanosaeta (11), to being highly methane (3).
divergent from any cultured archaea (9). Members of four of the five methanogenic orders
(Methanosarcinales, Methanococcales, Methanobacteri-
ales, and Methanomicrobiales) have been isolated by
GROUPS OF SOIL ARCHAEA culturing from soils. SSU rRNA gene sequences from mem-
bers of the families Methanosarcinaceae, Methanococ-
Methanogens caceae, and Methanobacteriaceae, as well as SSU rRNA
Methanogens are a phylogenetically congruent group gene sequences representing members of previously
of microorganisms that play critical roles in recycling unknown groupings, have been recovered from soils
elements through the biosphere. They provide the only (Table 2).
significant biogenic source of the hydrocarbon methane
(CH4 ), the second most abundant carbon-containing gas in
Morphology and Metabolism of Methanogenic Archaea
our atmosphere and a major greenhouse gas (3). Although
very diverse, both morphologically and physiologically, Methanogenic archaea exhibit diverse morphologies:
methanogens have a limited substrate range and are rods, cocci, spirals, pseudosarcinae, and multicellular
all strictly anaerobic, obligate methane producers (51). aggregates. They are either motile or nonmotile and they
Methanogenesis is important in anoxic environments stain either gram-negative or gram-positive. Carbon and
with low redox potentials, in which the reduction of energy sources include H2 + CO2 , formate, acetate, methyl
alternative oxidants such as nitrate, Mn(IV), Fe(III), and compounds (methanol, methylamines, methsulfides, and
sulfate has occurred (52). Waterlogged soils are among possibly methylselenides), methanol + H2 , or alcohols +
the most common habitats from which methanogens CO2 . A number of isolates are obligate or facultative
are isolated, along with anaerobic sediments, sludge autotrophs. Ammonia is the typical nitrogen source
digesters, insect guts, and mammalian rumina and (although some isolates can use amino acids or fix nitrogen)
intestines. Soil-dwelling methanogens from wetlands, rice and sulfur or sulfide is a source of sulfur (51).
 ARCHAEA IN SOIL HABITATS 297

Soil Habitats of Methanogenic Archaea contribute to erosion of the ozone layer. Over the last
decade, the atmospheric content of methane has been
Soil macroenvironments can be divided into the zones
measured at about 1.7 ppm, and is estimated to increase
above and below the water table (51). The zone below
by 0.5 to 1% a year (see also TRACE GASES SOIL). Terrestrial
the water table is saturated with water, whereas the
environments, particularly forests, are the major sources
vadose zone above is not. O2 diffuses rapidly through
of atmospheric methane. Although they are low-flux
soil and thus anoxic vadose zones are extremely rare,
systems relative to waterlogged soils, their contribution
although anoxic microenvironments can occur. Because
of methane to the biosphere is significant because of their
O2 diffuses more slowly through water than air, the
vast coverage area (56). Although methane production in
water-saturated zone is often anoxic, and representative
terrestrial ecosystems has not been well studied, two major
environments are waterlogged soils such as rice paddies,
terrestrial habitats of methanogens are termite guts and
stagnant marshes, and bayous. Groundwater, which
the rotting heartwood of trees. Better studied are high-
contains low amounts of O2 even when air-saturated
flux systems, including wetlands, rice paddy soils, and
(less than 0.3 mM O2 at 20 ° C), can easily become
freshwater sediments, which account for between 15 and
anoxic and suitable for the growth of methanogens,
45% of the total biogenic methane (54).
particularly in polluted areas. Although methanogens are
extremely sensitive to O2 in culture, they are protected
Thermoplasmales
in oxic zones in nature by the activities of O2 -utilizing
organisms (51). Variable-shaped, wall-less members of the genus Thermo-
Methanogens produce methane in three major ways, plasma have been isolated from solfataric fields in many
with acetate and H2 /CO2 being the most important geographic locations (Table 2). Originally found in smol-
methanogenic substrates in soils (52). Some produce dering coal refuse piles, solfataras appear to be the pri-
methane by the reduction of CO2 , using H2 , formate, or mary soil habitat of Thermoplasma species. Isolates have
certain alcohols as an electron donor and CO2 as the been cultured from these environments in wide geographic
electron acceptor. This method is common among rumen locations, including Italy, the Azores, United States, Ice-
methanogens. Methane production through the oxidation land, and Indonesia; the two characterized species are
of C-1 compounds is another, less common, process. Some Thermoplasma acidophilum and T. volcanium (42). In
of the substrate molecules are typically oxidized to carbon culture, these microorganisms grow heterotrophically as
dioxide and the remaining methyl groups are reduced facultative anaerobes by sulfur respiration. They grow
to methane. Methanogenesis from C-1 compounds occurs aerobically with yeast or other similar, complex extracts.
in environments that are enriched for these substrates, Growth temperatures range from about 45 ° C to 63 ° C
such as marine sediments. A third manner of methane for T. acidophilum (42) and from about 33 ° C to 67 ° C for
production occurs by an aceticlastic reaction, in which the T. volcanium (41,42). Neither natural growth factors nor
methyl carbon of acetate is reduced to methane and the ecological roles are currently known for Thermoplasma
carboxyl carbon is oxidized to carbon dioxide. The ability spp. found in conventional soils.
to catabolize acetate among the known methanogens is
limited to species of Methanosarcina and Methanosaeta Halobacteriales
(‘‘Methanothrix’’). Methane production from acetate is
General Characteristics. Cultured isolates of halophilic
found in habitats such as freshwater sediments where
archaea are currently placed in a single family, Halobac-
other anaerobes are limited for acetate catabolism by the
teriaceae, in the order Halobacteriales (57). Halophiles
availability of alternate electron acceptors.
comprise a variety of morphological types, including rods,
In nature, methanogens typically play a role in part
pleomorphic forms, angular cells, and cocci. Isolates from
of a degradative microbial consortium, performing the
saline soils, like other halophilic archaea, are red col-
final anaerobic step in the decomposition of organic
ored because of the presence of cellular pigments such
matter in soils (51). The major methanogenic precursors
as β-carotene and bacterioruberin (58). Some halophilic
in the complete decomposition of organic matter to
isolates are chemoorganotrophs and use only amino acids
CO2 and methane are acetate, formate, and H2 + CO2 .
or organic acids as carbon sources, but many have the
Initially, bacterial fermentation converts organic matter
ability to grow on carbohydrates. Originally reported as
into volatile organic acids. The longer-chain volatile
obligate or facultative aerobes, there is a growing body
organic acids (with three or more carbon atoms), such
of evidence suggesting that most species are facultative
as propionate and butyrate, must be further metabolized
anaerobes (58). Certain species also synthesize ATP using
to suitable methanogenic substrates. Interspecies electron
bacteriorhodopsin as a light-driven proton pump.
transfer, an association involving methanogens with
other, specialized groups of microorganisms, often helps
accomplish this. Habitats and Ecophysiology. Halophilic archaea are
found in hypersaline environments in which salt concen-
trations exceed 150 to 200 g/liter. Most exhibit optimum
Anthropogenic Increases in Methane Production
growth at NaCl concentrations of between 2 and 4 M,
A steady increase has occurred in the amount of whereas extremely halophilic species can grow even at the
atmospheric methane in conjunction with the rise in saturation point for NaCl in water (5.5 M) (57). Halophiles
human populations (53–55). Both agricultural practices fall into a number of ecophysiological groups on the basis
and landfilled waste are major sources of methane and of their response to salinity, magnesium concentration,
298 ARCHAEA IN SOIL HABITATS

and pH (58,59). There is a reasonably clear division, for Sulfolobales

example, between alkiliphilic and neutrophilic members
Taxonomy and Habitats. The order Sulfolobales com-
of Halobacteriaceae (no acidophiles have been reported).
prises a number of genera, including Sulfolobus; Metal-
Halophilic archaea are most commonly found in aquatic
losphaera; Acidianus; Desulfurolobus; ‘‘Stygiolobus’’ and,
environments, but representatives have also been isolated
potentially, Sulfurococcus. Commonly isolated from conti-
from saline soils (Table 2). Isolates have been recovered
nental solfataric water and mudholes of broad geographic
from saline soils of arid areas such as the Mediterranean
distribution, members of the order Sulfolobales are all
coast of Spain (43) and in soils of various regions of the for-
extremely thermoacidophilic aerobic organisms. The nat-
mer Soviet Union (45). Examples include, among others,
urally geothermally heated terrestrial environments in
members of the genera Halobacterium, Haloarcula, and
which they have been isolated tend to be acidic (as low
Halococcus. The ecology of halophilic soil archaea has not
as pH 0.5) with low salinity (0.1 to 0.5% salt). Sulfolobus
been studied, but by analogy to hypersaline aquatic envi-
and Acidianus spp. occur in diverse locations such as
ronments, the concentration of divalent cations, especially
the United States (46,65–67), Italy (68), the Azores (46),
magnesium and calcium, may be important (57,58).
New Zealand (69), Japan (70), the West Indies (66), and
Outward-directed sodium pumps in the cytoplasmic
Iceland (46,66,71). The typical irregular, lobed cells of
membrane of halophilic archaea are essential for regu-
members of the order can be present in quite high numbers
lating pH and for maintaining suitable intracellular ionic
(approximately 108 cells/ml) in these habitats, and are fre-
conditions. In addition, members of the Halobacteriaceae
quently recognized by the presence of an oily iridescent
maintain osmotic balance in hypersaline habitats by the
layer floating on the surface of muddy pools. Sulfolobus
accumulation of inorganic ions to reduce internal concen-
strains have been isolated as well from hot acid soils (72);
trations of Na+ (59). K+ appears to be the main compatible
such soils, in fact, being the dominant feature of solfataric
ion used. Adaptations to highly ionic environments are also
fields (66).
observed in the particular composition of halophilic mem-
In contrast to the cosmopolitan distribution of Sul-
branes and proteins (58). (See also SALINITY EFFECTS ON THE
folobus and Acidianus, Metallosphaera, and Stygiolobus
PHYSIOLOGY OF SOIL MICROORGANISMS).
isolates have been recovered on a much more limited
basis. Metallosphaera isolates have been found only within
Thermoproteales
Neapolitan solfataras (73), and Stygiolobus exclusively
General Characteristics and Habitats. Members of Ther- from the Azores (74,75). It is presently unclear whether
moproteales are strictly anaerobic gram-negative rods, their limited detection is due to geographic isolation or
discs, or spherical cells of varying sizes. These hyperther- selective cultivation.
mophilic microorganisms, typically, have optimum growth
temperatures between 85 ° C and 105 ° C. Members of Ther- Physiology of Sulfolobales. In culture, Sulfolobales pos-
moproteales that can be found in soils include those from sess diverse metabolic capabilities, spanning the biological
the genera Thermoproteus, Pyrobaculum, Desulfurococ- spectrum from autotrophy (76,77) to heterotrophy (66–79)
cus, and Thermofilum, all of which can be isolated from on a wide variety of complex organic substrates, sug-
mud holes and soils of continental solfataras in Italy, the ars, and amino acids, to possibly even mixotrophy (77).
Azores, Iceland, United States, and New Zealand (Table 2). All species are able to oxidize molecular sulfur to sul-
Many members of these and other genera are adapted for furic acid (73,76,80), with the exception of Stygiolobus.
life in high saline, thermal submarine environments. MoO4 2− (81) and Fe3+ (82) can both serve as alternative
electron acceptors; S2− (66), S4 O6 2− , and Fe2+ (65,82) can
Physiology and Ecology of Thermoproteales. Unfortuna- all replace sulfur as an energy source. The ability to oxi-
tely, there are no studies so far that describe the dize sulfur and ferrous iron results in the formation of
physiological ecology of members of Thermoproteales. In soluble sulfates and subsequent mobilization of metals
culture, all members are capable of sulfur respiration from sulfidic ores by some Sulfolobus species (65). This
using various organic substrates and of producing CO2 and other characteristics make these organisms impor-
and H2 S (60). Sulfur can be replaced by various sulfur tant candidates for microbial leaching (83). See Brierley
components under heterotrophic growth conditions. Some and Brierley (81) for a review of copper leaching studies
Thermoproteales are strict organotrophs, for example, performed with Sulfolobus.
Pyrobaculum organotrophum and members of the genus Acidianus and Desulfurolobus can both oxidize and
Thermofilum. Others, such as Thermoproteus tenax, reduce molecular sulfur (74,84). Growth and sensitivity
Thermoproteus neutrophilus, Pyrobaculum islandicum, profiles suggest that strict and facultative aerobic
Pyrodictium occultum, and Pyrodictium brockii, are able species of Sulfolobales are microaerophiles. Acidianus,
to grow chemolithoautotrophically with CO2 as the Desulfurolobus, and Stygiolobus species are capable of
sole carbon source (61), reducing sulfur with molecular growth anaerobically by S0/H2 -autotrophy, with H2 S being
hydrogen and producing H2 S. In T. neutrophilus, an produced (61).
unusual reductive citric acid cycle for carbon assimilation
was found with characteristics similar to those of pathways Ecology of Sulfolobus. Although isolated from hot
in eukaryotic microorganisms (62,64,114). Studies are acid soils, Sulfolobus could not typically be observed
needed to determine the relevance of the physiological directly in the soil or attached to soil particles (66).
characteristics of Thermoproteales members in culture to The organism was therefore quantified in such soils by
their ecology in situ. microscopy, using most-probable-number dilutions in an
 ARCHAEA IN SOIL HABITATS 299

autotrophic medium containing elemental sulfur at pH leading to the discovery of microbial complexity in the
3.0 (72). Sulfolobus numbers measured across thermal environment involved DNA-DNA reassociation kinetics
gradients at two sites in Yellowstone National Park after strand dissociation of the DNA in a sample (7,8).
reached approximately 102 to 103 per gram of soil at Control experiments demonstrated that reassociation
temperatures ranging from about 50 ° C to 70 ° C (72). The kinetics were affected by the diversity of genomic DNA
uptake of 14 CO2 was measured in these soils, and a good in a sample. Subsequently, approaches taking advantage
relationship was found between cell counts and rate of of phylogenetic conservation among the rRNA of various
CO2 fixation for higher numbers of sulfur oxidizers. Two groups of microorganisms have involved the use of
peaks were observed for 14 CO2 incorporation, one around polymerase chain reaction (PCR). Typically, DNA is
70 ° C with a preliminary assignment to Sulfolobus, and extracted from the environment of interest and SSU
another larger peak at around 30 ° C, which was cautiously rRNA gene sequences are amplified from the sample
attributed to Thiobacillus. Thiobacillus, the only other by PCR and analyzed by various methods. The most
organism detected by Fliermans and Brock (72) in these informative approach for using rRNA information to study
soils, demonstrated a growth peak at around 30 ° C and the microbial members of an ecosystem involves cloning
overlapping growth with Sulfolobus at around 55 ° C. There and sequencing (87) and is also the most labor-intensive.
was not an exact correlation between CO2 fixation and Less time-consuming, but also less informative, are
numbers of sulfur oxidizing microorganisms because no methods such as amplified-ribosomal-DNA-restriction
CO2 fixation peak was observed corresponding to peaks of analysis (ARDRA, 112), denaturing-gradient gel elec-
microbial growth at around 55 ° C. trophoresis (DGGE, 90), temperature-gradient gel elec-
trophoresis (TGGE, 111), single-strand conformational
Distribution of Sulfolobales and Thermoproteales in Sol- polymorphism (SSCP, 63), and terminal-restriction frag-
fataras. The distribution of Sulfolobales and Thermopro- ment length polymorphism (T-RFLP, 113) analyses. These
teales differs within the solfataric habitat. Two niches methods rely on size and/or nucleotide composition of the
have been identified in solfataric regions: (1) an upper, native or PCR-generated DNA fragments for discrimina-
oxic zone that is approximately 30-cm thick, highly acidic tion, but do not reveal precise phylogenetic information
(pH < 2) because of both biotic and abiotic oxidation, and without sequencing of specific bands or the use of quite
rust-colored because of the presence of ferric iron (66,85), specific primer sets.
and (2) a deeper, highly anaerobic zone reduced by volcanic Some methods, such as in situ hybridization (5,88),
gases that is higher in pH (pH > or = 4) and blackish-gray reverse-transcription PCR (89,90), TGGE or DGGE, and
in color because of the presence of heavy-metal sul- T-RFLP, are used to identify metabolically active or
fides (85). Sulfolobus are present in both niches, whereas numerically dominant populations. However, all but the
members of Thermoproteales are found exclusively in the first method rely on PCR, which is known to demon-
reduced zone. Unfortunately, there is no distribution data strate strong bias under certain circumstances (91–94).
available for other archaeal members in solfataric habi- Attempts to avoid PCR bias have been made, for example,
tats. with a recent approach that takes advantage of the abil-
ity of the bacterial artificial chromosome (BAC) vector
NONCULTURE-BASED APPROACHES TO MICROBIAL to accept large fragments of DNA (95,96) and act as
SAMPLING an expression vector for environmental DNA in a sur-
rogate host (97). Although molecular methods also have
The view that archaea are strictly extremophiles is limitations, the evidence from their concerted use is uncon-
now changing because of a ‘‘molecular revolution’’ (23) testable. Our knowledge of natural microbial populations
in microbial ecology that has led to the discovery of was previously severely limited through use of culture-
nucleic acid sequences and in situ evidence for previously based techniques only.
unknown members of the Archaea (as well as for
bacteria and Eukarya). The development and application
of a variety of molecular techniques not reliant on UNCULTURED SOIL ARCHAEA
culturing to detect and discriminate microorganisms
has led to a restructuring of our view of microbial Research by Bintrim and coworkers serves as one example
ecosystems. The most extensively used phylogenetic tools of a study examining the diversity of Archaea in a
are those based on the exceptional conservation of SSU mesophilic soil environment (9). Phylogenetic analysis
rRNA gene sequences among different organisms. Such of sequences recovered from a Midwestern (fallow)
approaches are powerful because they enable the recovery agricultural soil placed all 35 of the archaeal clones
of sequences representing specific microbial phylotypes recovered (designated SCA) in a clade that diverged
from natural environments, without cultivation of the deeply within the division Crenarchaeota. This and other
corresponding microorganisms (5). Use of this approach such discoveries (10,21,32,33,98) were surprising because
has demonstrated the existence and abundance of novel archaea had never before been suspected to be present in
archaea across a broad array of nonextreme terrestrial mesophilic dry soils.
and aquatic habitats (9–11,15,16,19,21,22,31–33,86). Crenarchaeota SSU rRNA sequences have now been
Because all techniques have biases, multiple recovered worldwide from terrestrial and aquatic low-
approaches are often utilized when analyzing the microbial to moderate-temperature environments. The sequences
diversity of a particular habitat. Initial approaches belong to at least four distinct subgroups and appear
300 ARCHAEA IN SOIL HABITATS

Euryarchaeota

Thermophillic Crenarchaeota
pSL69
pJP89
pJP44 SCA1180
pSL1 KHS-Cu19
pSL123 SCA1150
pLcmB389 SCA1173
pLcmB235 pLemA8
pLcmB108 pM17
pLcmB21 pP17

Freshwater cluster
pLcmB22 SCA1166
pGrfB98 SCA1151
pLcmB137 pGrfA22
pLcmB126 pGrfA17
pLcmB322 KBS-Nat4
pGrfB22 pGrfA33
pLAWIl PAD19
pLAW2 KBS-Nat12
pLAW9 pGrfA44
pGrfC62 SCA1170
pLAW12 pGrfA4
pSL12

Terrestrial cluster
KBS-Nat1
pSL77 KBS-Nat11
FFSB6 KBS-Cu17
FFSB1 SCA1158

FFSB cluster
FFSB2 FIE16
FFSB10 pGrfA19
FFSB11 SCA1154
FFSB5 PGrfA5
FFSB4 PGrfA42
FFSB3 PGrfA50
FFSB7 KBS-Cu14
LMA238 SCA1145
LMA137 KBS-Cu113
SB95-53 KBS-Nat2
NH49-14 KBS-Nat20
SB96-57 pGrfA24
SB95-61 KBS-Cu15
NH49-8 SCA1175
PVA-3 KBS-Cu11
NH49-9 pGrfA23
NH25-1 pGrfA53
NH49-4 pGrfA9
PVA-4 pGrfA62
Fosmid4B7
Marfine cluster

pGrfA15
C.symblosion
SB95-16
SB95-59
SB95-54
SB95-60
Figure 2. Phylogenetic tree show- LMA229
LMA226
ing the relationships of nonther- SBAR12
mophilic Crenarchaeota 16S rDNA SBAR5
sequences. The symbols indicate ANT12
WHARQ
the specificities of Crenarchaeota JM2
probe Cren499R (diamond), probes JM6
Cren667 and GI-554 (
), and probe JM3
•
Cren745 ( ). Cenarchaeum sym- JM4
JM1
biosum C. symbiosum. Reproduced JM8
with permission from Buckley et al. JM7
(1998).

to have a common ancestry (Fig. 2; 21). The major- from 2% (15) up to 34% (16) (Table 1). Thus, in both
ity of archaeal sequences recovered from soil clus- terrestrial and oceanic nonextreme habitats, archaeal
ter together in one clade — with the exception of a members constitute a significant proportion of micro-
group of sequences recovered from a single study of bial assemblages. Similar trends of novelty, diversity,
boreal soil from Finland (32) — which cluster in a sep- and abundance are also observed for SSU rRNA gene
arate group. Quantitative estimates of the total RNA sequences representing uncultured members of the Eur-
of crenarchaeota in agricultural and undisturbed soils yarchaeota (15,16,18,99).
range from 0.37 to 1.4% relative to total community
RNA (21) (Table 1). The closest affiliation of uncultured
The Rhizosphere (See also RHIZOSPHERE MICROBIOLOGY)
soil archaea is to a clade of marine, planktonic crenar-
chaeotes, designated as ‘‘Group 1 archaea’’ (15). Studies of Soil archaea thrive in another habitat previously thought
the biomass contribution of planktonic archaea ranged to be the domain of bacteria and eukaryotes, plant
 ARCHAEA IN SOIL HABITATS 301

roots and the rhizosphere. The rhizosphere extends (designated ARR/ABS, 11). Direct sequence comparisons
from the surface of the root (the rhizoplane) into the supported divergence of these two rhizosphere groups,
surrounding soil a few millimeters. It is a rich ecological yielding sequence similarity of 76 to 84% between the
niche abundantly exploited by soil microbes, in terms TRC and ARR/ABS sequences, compared with a range
of both population numbers and activity. Roots and of 91 to 99% sequence similarity between the TRC
rhizosphere habitats are nutrient-rich relative to those and uncultured clones from West Madison (designated
of bulk soil. This is in part due to the fact that SCA, 9) and Michigan (designated KBS, 21) soils. These
roots deposit from 5 to 30% of the total photosynthate studies, like those on mesophilic soil archaea, indicate
of the plant into the surrounding soil in a process that rhizosphere Crenarchaeota comprise a phylogenet-
known as rhizodeposition (100), and in part due to ically coherent, but quite diverse, group of microorgan-
sloughed, injured, or senescing plant cells that leak isms.
nutrients.
Although a far more nutrient-rich habitat than bulk Abundance of Crenarchaeota on Tomato Roots. In the
soil, the rhizosphere consists of networks of microenviron- first study to document colonization of plant surfaces
ments similar to those found in soils. Extreme niches exist by members of the Archaea, we showed that members
in the rhizosphere as certainly as they do in soil, because of the nonthermophilic Crenarchaeota colonize the roots
the rhizosphere environment is subjected to dramatic cli- of tomato plants grown in soil (Fig. 3; 22). Our results
matic events resulting in fluctuations of water and osmotic demonstrated the recruitment of the crenarchaeotes from
potential, salt concentrations, pH, and soil particle struc- soil and their growth, and apparent persistence, on
ture (101–103). The question remains to be answered, roots. The abundance of crenarchaeotes was documented
however, as to whether archaea occupy extreme niches on tomato roots by fluorescence in situ hybridization
in soil and rhizosphere, or if, instead, the archaea there (FISH) assays, using probes designed to be specific
compete successfully against bacteria within nonextreme for the SSU rRNA gene sequences of Crenarchaeota
niches. Specialized metabolic activities such as methane recovered from soil and tomato rhizospheres. We found
production or nitrogen fixation may allow archaea, which Crenarchaeota broadly distributed on both young and
typically grow more slowly than most bacteria, to overcome senescent roots. Their abundance was determined to be
competition by making essential trophic contributions about 3.3% relative to the number of total probe-positive
within the ecosystem. cells (bacteria and crenarchaeotes) on young roots and
about 16% on senescent roots of plants seven to eight
Archaea in the Rhizosphere weeks in age. Although they were present on all roots
examined from plants spanning one to eight weeks of
Root-Associated Methanogens. The relative contribution
age, their abundance increased 10-fold on senescent roots
of root-associated methane production to the atmosphere
compared with younger roots of about the same age,
can be very important for flooded soil environments.
with a concurrent increase in the frequencies of cells in
For example, root-associated methane production in
pairs (presumed to be dividing cells) and in colonies. By
rice paddy soils varied between 4 and 52% in one
contrast, bacterial abundance increased only fewfold on
study (104). In another study, removal of aboveground
senescent roots. Greater abundance of crenarchaeotes was
vegetation from natural wetlands resulted in more than
also observed in the older regions of nonsenescent roots,
an order of magnitude reduction in methane emissions,
whereas bacterial numbers decreased in these regions.
without a concurrent decrease of methane stored in
These results suggest that rhizosphere Crenarchaeota
soil (105).
may have roles in mediating later stages of root biology,
a function previously unsuspected for members of the
Root Colonization by Archaea of Unknown Pheno-
Archaea.
types. Additional evidence for archaeal associations with
the rhizosphere comes from studies in both flooded (11)
and dry (22) soils. SSU rRNA gene sequences were recov- CONCLUSION
ered from these environments using total rhizosphere
and/or rhizoplane DNA and primers hybridizing to the One of the great mysteries of the world lies beneath our
SSU rRNA genes of archaea in PCR reactions. In keeping feet in the guise of the mundane, a substance known
with the results from previous studies in soil, rhizosphere simply as soil or dirt. Through time and considerable
samples from rice plants grown in flooded soils harbored effort, humankind continues to push back the frontiers of
SSU rRNA sequences from both Euryarchaeota and Cre- science in the fields of chemistry, physics, and biology. We
narchaeota (11), whereas rhizosphere samples from plants have yet to fully understand, however, the dynamics of
grown in dry soils, thus far, have yielded evidence for Cre- these forces and how they interact to form that enduring
narchaeota only (22). rDNA clones recovered from tomato interface between the organic and inorganic components
rhizoplane (Lycopersicon esculentum) samples (designated of the terrestrial biosphere. Although scientific knowledge
TRC clones, with 91 to 99% sequence identity over 1,280 of the chemical and physical properties of soil is
nucleotide positions) clustered with uncultured soil Cre- substantial, our understanding of microbial composition
narchaeota from dry, mesophilic environments. Phyloge- and contribution to soil ecosystem dynamics is incomplete.
netic analysis also indicated that TRC sequences were The recent discoveries regarding archaea in soil and
distinct from sequences representing uncultured Crenar- other conventionally nonextreme habitats dramatically
chaeota clones from rice rhizospheres and rice paddy soil illustrate this point.
302 ARCHAEA IN SOIL HABITATS

a1 b1 c1 d1 e1

a2 b2 c2 d2 e2

a3 b3 c3 d3 e3

Figure 3. Fluorescence in situ hybridization of the tomato rhizosphere for Crenarchaeota.

One-week-old (A) tap and (B) lateral rootlets, (C and D) senescent rootlets, and (E) rhizosphere soil.
Oligonucleotide probes used were designed to be (1) specific for nonthermophilic Crenarchaeota
and (2) Bacteria. Images in row (3) show results of DAPI staining of the same fields shown in
rows (1) and (2). Scale bar represents 5 µm. Reproduced with permission from Simon et al. (2000).
See color insert.

Extreme or Not Extreme — That Is the Question state is known to exist for, at least, some culturable
microorganisms under certain conditions (106). This leads
The fact that isolation of archaea was initially successful
in primarily extreme or uncommon environments implied to the prediction that we may have mislabeled mesophilic
the failure of archaea to compete with bacteria in more soil habitats as ‘‘nonextreme,’’ and suggests that archaea
common habitats. Following this line of thought, archaea thrive in soil precisely because it contains niches of an
are successful because of their specialization. For example, extreme nature. The recent discovery of archaea in other
adaptation to extreme environments, in which growth of nonextreme habitats, such as low-to-medium temperature
bacteria and eukaryotes is limited, results in reduced marine environments, can also be rationalized using this
competition. Thus, archaea may survive by avoiding direct hypothesis. It can be argued that marine planktonic
competition with other microorganisms. On the other archaea are in relatively extreme environments as
hand, the discovery of archaea in mesophilic habitats well, because the majority of the marine ecosystem is
poses the question — do archaea thrive only in highly oligotrophic (contains 1–10 mg of carbon per liter), with
specialized niches, as dogma holds, or is our knowledge of microniches created, for example, by pockets of suspended
the extent of their ecological range limited by our inability
detrital matter (107). The answers to the questions of
to culture them? Soil is an excellent case in point. It is
what makes a habitat extreme (from a microbial point of
well known that microscopic counts of microbial cells in
soil far exceed, by as much as two orders of magnitude, view) lie, in part, within the physiological mysteries of the
the number of cells that can be cultured on typical as-yet uncultured archaea found thriving, unexpectedly,
agar medium (5,6). A hypothesis to explain this result in these so-called nonextreme environments. When we
states that a portion of cells in soil environments are unlock the metabolic secrets of these archaea, we
in a starvation state and that makes them recalcitrant will gain a better understanding of the niches they
to growth in vitro. Such a ‘‘viable but nonculturable’’ inhabit.
 ARCHAEA IN SOIL HABITATS 303

The Future BIBLIOGRAPHY

Although molecular techniques have added substantially 1. W. B. Whitman, D. C. Coleman, and W. J. Wiebe, Proc.
to our knowledge of the composition of microbial Natl. Acad. Sci. U.S.A. 95, 6578–6583 (1998).
assemblages in soils and expanded our knowledge 2. R. M. Atlas and R. Bartha. Microbial Ecology, Fundamen-
concerning the ecological range of archaea, there is more tals and Applications. The Benjamin/Cummings Publ. Co.
to learn in this regard as the list of habitats harboring Inc., New York, 1993.
these organisms lengthens. There are also a number 3. R. J. Cicerone and R. S. Ormeland. Global Biogeochem.
of challenges involved in describing microbial dynamics Cycles 2, 299–327 (1988).
in soil habitats that remain unsolved. For example, 4. R. Conrad, in J. G. Jones, ed., Advances in Microbial Ecol-
many questions remain relating to sampling; spatial ogy, vol. 14, Plenum Press, New York, 1995, pp. 207–250.
heterogeneity and scale are both extremely complex 5. R. I. Amann, W. Ludwig, and K.-H. Schleifer, Microbiol.
Rev. 59, 143–169 (1995).
problems that demand integrative and interdisciplinary
6. J. T. Staley and A. Kanopka, Annu. Rev. Microbiol. 39,
approaches for their solution. Approaches monitoring
321–346 (1985).
chemical and physical parameters, in combination with
7. V. Torsvik, K. Salte, R. Sorheim, and J. Goksoyr, Appl.
microbial analyses at the micro- as well as macroscale Environ. Microbiol. 56, 776–781 (1990).
level, are needed to address these issues and others. 8. V. Torsvik, J. Goksoyr, and F. L. Daae, Appl. Environ.
Advances in the areas of genomics and high throughput Microbiol. 56, 782–787 (1990).
monitoring of changes in microbial populations in response 9. S. B. Bintrim, et al., Proc. Natl. Acad. Sci. U.S.A. 94,
to ecosystem fluctuations will further illuminate microbial 277–282 (1997).
composition, metabolic potential, and dynamics in soil 10. J. Borneman and E. W. Triplett, Appl. Environ. Microbiol.
environments. Examples of such advances include obtain- 63, 2647–2653 (1997).
ing genetic and functional information from soil microor- 11. R. Großkopf, S. Stubner, and W. Liesack, Appl. Environ.
ganisms without culturing using metagenomics (97) and Microbiol. 64, 4983–4989 (1998).
other direct access methods (50,108–110) and the use of 12. C. R. Kuske, S. M. Barns, and J. D. Busch, Appl. Environ.
oligonucleotide arrays to monitor detection, abundance, Microbiol. 63, 3614–3621 (1997).
and gene expression of populations and assemblages of 13. W. Liesack and E. Stackebrandt, J. Bacteriol. 174,
5072–5078 (1992).
microorganisms.
Progress in defining ecological roles of archaea 14. E. Stackebrandt, W. Liesack, and B. M. Goebel, FASEB J.
7, 232–236 (1993).
and other microorganisms will be realized best by
15. E. F. DeLong, Proc. Natl. Acad. Sci. U.S.A. 89, 5585–5689
using classical culturing and molecular approaches in
(1992).
concert. For example, sequences recovered from previously
16. E. F. DeLong, K. Y. Wu, B. B. Prezelin, and R. V. M. Jovine,
unknown and numerically abundant archaea in soil Nature 371, 695–697 (1994).
environments provide information for the design of specific 17. R. Massana, et al., Limnol. Oceanogr. 43, 607–617 (1998).
oligonucleotide primers and probes hybridizing to the 18. A. E. Murray, et al., Appl. Environ. Microbiol. 64,
SSU rRNA gene sequences of those microorganisms. The 2585–2595 (1998).
oligonucleotides can then be used to monitor the growth of 19. M. B. Karner, E. F. DeLong, and D. M. Karl, Nature 409,
previously uncultured organisms in enrichment cultures. 507–510 (2001).
Growth of these archaea in pure culture will, in turn, 20. C. M. Preston, K. Y. Wu, T. F. Molinski, and E. F. DeLong,
potentially lead to an understanding of their metabolic Proc. Natl. Acad. Sci. U.S.A. 93, 6241–6246 (1996).
capabilities and potential impact on ecosystem dynamics, 21. D. H. Buckley, J. R. Graber, and T. M. Schmidt, Appl. Envi-
predictions about which can be tested using phylogenetic ron. Microbiol. 64, 4333–4339 (1998).
and substrate-level probes in situ. 22. H. M. Simon, J. A. Dodsworth, and R. M. Goodman, Envi-
In conclusion, many questions remain regarding ron. Microbiol. 2, 495–505 (2000).
the ecological distribution and roles of archaea. We 23. G. J. Olsen, C. R. Woese, and R. Overbeek, J. Bacteriol. 176,
1–6 (1994).
predict that archaea have far greater global impact
on biosphere composition and function than previously 24. W. Zillig, et al., in M. Kates, D. J. Kushner, and A. T. Math-
eson, eds., The Biochemistry of Archaea (Archaebacteria),
thought. Archaea will be found to perform fundamental
Elsevier, New York, 1993, pp. 367–92.
roles in conventionally nonextreme environments, as they
25. R. Amils, P. Cammarano, and P. Londei, in M. Kates,
do in typically extreme habitats. Their early evolution and D. J. Kushner and A. T. Matheson, eds., The Biochemistry
history placed them in these environments as fundamental of Archaea (Archaebacteria), Elsevier, New York, 1993,
nodes in the microbial network; their unique physiological pp. 393–409.
properties keep them there. 26. T. A. Langworthy, in C. R. Woese and R. S. Wolfe, eds., The
Bacteria, A Treatise on Structure and Function, Vol. 8,
Acknowledgments Academic Press Inc., Orlando, Fla., 1985, pp. 459–497.
Work in our laboratory has been supported by funding from 27. S. Winker and C. R. Woese, Syst. Appl. Microbiol. 14,
the McKnight Foundation, the National Science Foundation, the 305–310 (1991).
National Institutes of Health, and the David and Lucille Packard 28. S. M. Barns, C. F. Delwiche, J. D. Palmer, and N. R. Pace,
Foundation. Proc. Natl. Acad. Sci. U.S.A. 93, 9188–9193 (1996).
304 ARCHAEA IN SOIL HABITATS

29. C. F. Brunk and N. Eis, Appl. Environ. Microbiol. 64, 61. F. Fischer, W. Zillig, K. O. Stetter, and G. Schreiber, Nature
5064–5066 (1998). 301, 511–513 (1983).
30. A. -L. Reysenbach, M. Ehringer, and K. Hershberger, 62. S. Schäfer, T. Paalme, R. Vilu, and G. Fuchs, Eur. J.
Extremophiles 4, 61–67 (2000). Biochem. 186, 695–700 (1989).
31. Y. Kudo, et al., Biosci. Biotechnol. Biochem. 61, 917–920 63. F. Schweiger and C. C. Tebbe, Appl. Environ. Microbiol.
(1997). 64(4), 870–876 (1998).
32. G. Jurgens, K. Lindström, and A. Saano, Appl. Environ. 64. G. Strauss, W. Eisenreich, A. Bacher, and G. Fuchs, Eur. J.
Microbiol. 63, 803–805 (1997). Biochem. 205, 853–866 (1992).
33. T. Ueda, Y. Suga, and T. Matsuguchi, Eur. J. Soil Sci. 46, 65. C. L. Brierley and J. A. Brierley, Can. J. Microbiol. 19,
415–421 (1995). 183–188 (1973).
34. K. -J. Chin, T. Lukow, and R. Conrad, Appl. Environ. 66. T. D. Brock, Thermophilic Microorganisms and Life at High
Microbiol. 65, 2341–2349 (1999). Temperatures, Springer-Verlag, New York, 1978.
35. H. König, Can. J. Microbiol. 30, 1477–1481 (1984). 67. T. D. Brock, K. M. Brock, R. T. Belly, and R. L. Weiss, Arch.
36. T. Lueders and M. Friedrich, Appl. Environ. Microbiol. 66, Mikrobiol. 84, 54–68 (1972).
2732–2742 (2000). 68. M. deRosa, A. Gambacorta, G. Millonig, and J. D. Bu’Lock,
Experientia 30, 866–868 (1974).
37. I. R. McDonald, et al., Microb. Ecol. 38, 225–233 (1999).
69. D. B. Bohlool, Arch. Microbiol. 97, 181–194 (1975).
38. B. S. Rajagopal, N. Belay, and L. Daniels, FEMS Microbiol.
Ecol. 53, 153–158 (1988). 70. S. Yeats, P. McWilliam, and W. Zillig, EMBO J. 1,
1035–1038 (1982).
39. T. C. Stadtman and H. A. Barker, J. Bacteriol. 62, 269–280
(1951). 71. K. O. Stetter, and W. Zillig, in The Bacteria, A Treatise on
Structure and Function. C. R. Woese and, R. S. Wolfe, eds.,
40. T. N. Zhilina and G. A. Zavarzin, Int. J. Syst. Bacteriol. 37,
Archaebacteria, Vol. 3, Academic Press Inc., Orlando, Fla.,
281–283 (1987).
1985, pp. 85–170.
41. G. Huber, et al., Syst. Appl. Microbiol. 14, 397–404 (1991).
72. C. B. Fliermans and T. D. Brock, J. Bacteriol. 111, 343–350
42. A. Segerer, T. A. Langworthy, and K. O. Stetter, Syst. Appl. (1972).
Microbiol. 10, 161–171 (1988).
73. R. Huber, M. Kurr, H. W. Jannasch, and K. O. Stetter,
43. E. Quesada, A. Ventosa, F. Rodriguez-Valera, and A. Nature 342, 833–834 (1989).
Ramos-Cormenzana, J. Appl. Bacteriol. 53, 155–161 (1982).
74. A. Segerer, K. O. Stetter, and F. Klink, Nature 313, 787–789
44. T. Takashina, K. Otozati, T. Hamamoto, and K. Horikoshi, (1985).
Biodiversity Conserv. 3, 632–642 (1994).
75. A. Segerer, A. Trincone, M. Gahrtz, and K. O. Stetter, Int.
45. I. S. Zvyagintseva and A. L. Tarasov, Microbiology 56, J. Syst. Bacteriol. 41, 495–501 (1991).
664–668 (1988). 76. D. W. Shivvers and T. D. Brock, J. Bacteriol. 114, 706–710
46. A. Segerer, A. Neuner, J. K. Kristjansson, and K. O. Stetter, (1973).
Int. J. Syst. Bacteriol. 36, 559–564 (1986). 77. A. P. Wood, D. P. Kelly, and P. R. Norris, Arch. Microbiol.
47. K. O. Stetter, et al., Syst. Appl. Microbiol. 7, 393–397 (1986). 146, 382–389 (1987).
48. W. Zillig, et al., Syst. Appl. Microbiol. 8, 197–203 (1986). 78. M. deRosa, A. Gambacorta, and J. D. Bu’Lock, J. Gen.
49. E. A. Bonch-Osmolovskaya, et al., Arch. Microbiol. 154, Microbiol. 86, 156–164 (1975).
556–559 (1990). 79. D. W. Grogan, J. Bacteriol. 171, 6710–6719 (1989).
50. A. Henne, R. Daniel, R. A. Schmitz, and G. Gottschalk, 80. A. Segerer and K. O. Stetter, in A. Balows, et al., eds., The
Appl. Environ. Microbiol. 65, 3901–3907 (1999). Prokaryotes, Springer-Verlag, New York, 1999, pp. 684–701.
51. S. H. Zinder, in J. G. Ferry, ed., Methanogenesis, Chapman 81. J. A. Brierley and C. L. Brierley, in T. D. Brock, ed., Ther-
& Hall, New York, 1993, pp. 129–206. mophiles, General, Molecular and Applied Microbiology,
52. R. Conrad, FEMS Microbiol. Ecol. 28, 193–202 (1999). John Wiley & Sons, New York, 1986, pp. 279–305.
53. M. A. K. Khalih and R. A. Rasmussen, Chemosphere 29, 82. T. D. Brock and J. Gustafson, Appl. Environ. Microbiol. 32,
833–842 (1994). 567–571 (1976).
54. J. Lelieveld, P. J. Crutzen, and F. J. Dentener, Tellus Ser. B 83. C. L. Brierley, Sci. Am. 247, 42–51 (1982).
Chem. And Phys. Meteor. 50, 128–150 (1998). 84. W. Zillig, et al., Nature 313, 789–791 (1985).
55. D. I. Stern and R. K. Kaufmann, Chemosphere 33, 159–176 85. K. O. Stetter, in T. D. Brock, ed., Thermophiles, General,
(1996). Molecular and Applied Microbiology, John Wiley & Sons,
56. J. C. Sheppard, et al., JGR J. Geophys. Res. 87, 1305–1312 New York, 1986, pp. 39–74.
(1982). 86. C. Schleper, W. Holben, and H. P. Klenk, Appl. Environ.
57. D. J. Kushner, in C. R. Woese and R. S. Wolfe, eds., The Microbiol. 63, 321–323 (1997).
Bacteria, A Treatise on Structure and Function, Vol. 8, 87. P. Hugenholtz and N. R. Pace, Trends Biotechnol. 14,
Academic Press Inc., Orlando, Fla., 1985, pp. 171–214. 190–197 (1996).
58. B. Javor, in T. D. Brock, ed., Hypersaline Environments, 88. R. I. Amann and M. Kühl, Curr. Opin. Microbiol. 1, 352–358
Springer-Verlag, New York, 1989, pp. 101–124. (1998).
59. A. Oren in A. Oren, ed., Microbiology and Biogeochemistry of 89. G. F. Duarte, et al., J. Microbiol. Methods 32, 21–29 (1998).
Hypersaline Environments, CRC Press, Boca Raton, Mass., 90. A. Felske, A. Wolterink, R. Van Lis, and A. D. L. Akker-
1999, pp. 1–10. mans, Appl. Environ. Microbiol. 64, 871–879 (1998).
60. G. Huber and K. O. Stetter, in A. Balows, et al., eds., The 91. F. Mathieu-Daudé, J. Welsh, T. Vogt, and M. McClelland,
Prokaryotes, Springer-Verlag, New York, 1999, pp. 677–683. NAR 24, 2080–2086 (1996).
 ASSESSING MICROBIAL PROTEOMES IN THE ENVIRONMENT 305

92. E. Rubin and A. A. Levy, NAR 24, 3538–3545 (1996). catalysis, maintenance of the integrity of membrane recep-
93. A. G. C. L. Speksnijder, et al., Appl. Environ. Microbiol. 67, tors and general cellular architecture, and the mitigation
469–472 (2001). of stressful environmental conditions. The profile of pro-
94. M. T. Suzuki and S. J. Giovannoni, Appl. Environ. Micro- teins synthesized by microorganisms under indigenous
biol. 62, 625–630 (1996). conditions offers a window through which investiga-
95. U.-J. Kim, et al., Genomics 34, 213–218 (1996). tors may glimpse complex environmental modulation of
96. H. Shizuya, et al., Proc. Natl. Acad. Sci. U.S.A. 89, multilevel interactions between genetic endowment and
8794–8797 (1992). natural selection. The term proteome was coined by
97. M. R. Rondon, et al., Appl. Environ. Microbiol. 66, Wilkins and Williams and their coworkers to describe a
2541–2547 (2000). postgenomic field of research aimed at understanding the
98. K. L. Hershberger, et al., Nature 384, 420 (1996). disposition of genetic potential in the strategic survival
99. R. Massana, A. E. Murray, C. M. Preston, and E. F. DeLong, of organisms in variable environments (1–8). Microbial
Appl. Environ. Microbiol. 63, 50–56 (1997). proteomics is an integrative approach toward the analy-
100. J. M. Whipps and J. M. Lynch,. Ann. Proc. Phytochem. Soc. sis of protein molecules in microbial communities, with
26, 59–71 (1985). particular emphasis on understanding how genetic poten-
101. A. D. Rovira, Annu. Rev. Microbiol. 19, 241–266 (1965). tial is transformed into the observed array of diverse
but compatible structures and functions in both natural
102. A. D. Rovira, Bot. Rev. 35, 35–57 (1969).
and engineered environments (9–24). Proteomics research
103. A. D. Rovira, in D. L. Keister and P. B. Cregan, eds., The
encompasses both the genetic and environmental under-
Rhizosphere and Plant Growth, Kluwer Academic Publish-
ers, Boston, Mass., 1991, pp. 3–13. pinnings of microbial community integrity. Application
of proteomics in environmental microbiology research is
104. T. Minoda, M. Kimura, and E. Wade, J. Geophys. Res. 101,
21091–21097 (1996).
still evolving, but it is so far dominated by axenic culture
studies. However, investigations of microbial community
105. G. J. Whiting and J. P. Chanton, Global Biogeochem. Cycles
6, 225–231 (1992).
proteins are emerging in the literature (17,25–34). Early
reports focused on methods of quantitative recovery of
106. D. McDougald, S. A. Rice, D. Weichart, and S. Kjelleberg,
FEMS Microb. Ecol. 25, 1–9 (1998).
protein molecules from uncultivated but densely popu-
lated microbial communities (22,24), the radiolabeling of
107. S. Giovannoni and M. Rappé, in D. Kirchman, ed., Microbial
target proteins in heterogenous populations (21), and the
Ecology of the Oceans, John Wiley & Sons, New York, 2000,
pp. 47–84. direct assay for enzymatic proteins in situ (21). These
research programs fulfill the need for methods to comple-
108. O. Bèjá, et al., Science, 289, 1902–1906 (2000).
ment the remarkable progress that has been made toward
109. S. F. Brady and J. Clardy, J. Am. Chem. Soc., 122,
the elucidation of phylogenetic diversity and microbial
12903–12904 (2000).
community structures, made possible by lipid-based and
110. J. Stein, et al., J. Bacteriol. 178, 591–599 (1996).
nucleic acid–based techniques for analyzing heterogenous
111. G. Muyser, Curr. Opin. Microbiol. 2, 317–322 (1999). microbial systems such as soils and groundwater (35–39).
112. M. Heyndrickx, et al., J. Microbiol. Methods 26, 247–259 The methodological complementation is essential because
(1996). knowledge of microbial community structure does not nec-
113. T. L. Marsh, Curr. Opin. Microbiol. 2, 323–327 (1999). essarily lead to useful information on microbial community
114. S. Schäfer, C. Barkowski, and G. Fuchs, Arch. Microbiol. functions such as metabolic capacity, control of popula-
146, 301–308 (1986). tion dynamics for native and invasive species, sensitivity
115. G. Lauerer, et al., Syst. Appl. Microbiol. 8, 100–105 (1986). to variable environmental conditions, and intrinsic niche
diversity (40).

ARCHAEA IN SOILS. See SOIL BACTERIA

APPLICATIONS OF PROTEOME-BASED ASSESSMENTS IN
ENVIRONMENTAL MICROBIOLOGY

ARCHAEA, THERMOPHILIC. See THERMOPHILES, The practical rationale for proteome analysis in aquatic
DIVERSITY OF microbiology is the desire to understand and harness the
roles performed by protein molecules in a given sample of
water for the improvement of environmental and public
health. However, it is generally recognized that only
about 10% of the microbial diversity in natural samples
ASSESSING MICROBIAL PROTEOMES IN THE is extractable through traditional techniques involving
ENVIRONMENT laboratory cultivation. Additionally, methods developed
for direct nucleic acid assessments such as DNA and rRNA
OLADELE A. OGUNSEITAN fingerprints do not illuminate the response of the diverse
University of California microbial world to real-life environmental variations
Irvine, California (Fig. 1). Similarly, assessment of bacterial mRNA is
generally too temporary for its use in constructing a
Proteins are generally recognized as the molecular wheels correlative map between environmental conditions and
of cellular machinery. Their functions include enzymatic microbial response (41).
306 ASSESSING MICROBIAL PROTEOMES IN THE ENVIRONMENT

Microbial cell in a hypothetical proteome sampling space

GENOME

TRANSCRIPT
Housekeeping
Inducible functions functions
PROTEOME
Stress proteins
Enzyme activity
Maintenance of cellular
architecture
Transport
proteins

Ectoenzmes
Figure 1. Conceptual diagram showing the
transformation of genomic information into Mitigation of environmental Increasing competitive advantage (fitness)
an array of proteins molecules that may be threats
assessed within the framework of responses Excretion of catabolic enzymes
to influential environmental parameters. In Antibiotic resistance Recognition of genetic exchange opportunities
theory, all species in a given environmental Bacteriophage infection Chemotaxis
sample contribute to the collective proteome Toxic chemical degradation
pool but responses by particular species Ultraviolet light resistance Extracellular proteins in environmental samples
present in large numbers may dominate Survival of temperature fluctuation are released actively by cell
the outcome of assessments limited by the Survival of moisture level fluctuation secretion or passively by cell lysis. Also
stringency of molecular resolution. includes viral protein coats.

Microbial community proteome analysis has been designed to serve as biosensors of environmental
useful for understanding the ecotoxicological effects of change.
chemical pollutants, the activity of catabolic enzymes in 4. How does microbial community structure change in
environmental bioremediation, the tracking of indicator response to changes in biotic and abiotic modifica-
species and pathogens in potable water distribution tions to their habitat? For example, protein-profile
systems, and the tracking of genetically engineering assessments have been used to monitor the status
microorganisms and their protein products following of the microbial community necessary for successful
deliberate release into natural ecosystems. wastewater treatment in engineered processes and
Questions pursued by environmental microbiologists to diagnose microbial cellular states (42,43).
engaged in proteome assessments include the following: 5. What is the evidence for evolutionary divergence
and population disequilibrium in highly conserved
1. Are specific pathogens present in a given environ- niches that are inhabited by distinct phylogenetic
mental sample recognizable by defined protein pro- groups of microorganisms? Investigators frequently
files? For example, several waterborne pathogenic find that different versions of an enzyme (allozymes
viruses are detected in polluted water samples or isozymes) have evolved over time despite the
through antigen–antibody reactions that are based conservation of relevant catalytic function. For
on polypeptides in the structural backbone of the example, at least two versions of δ-aminolevulinate
capsid protein. Similarly, pathogenic bacteria can dehydratase, the key enzyme in porphyrin synthesis,
be detected through screening for proteinaceous tox- are known to respond differently to toxic metal
ins (12,17). exposures (25,26). At least eight versions of β-
2. When, and in what quantities, are certain enzymes glucosidase have been identified in marine microbial
capable of detoxifying chemical pollutants produced communities (33). The protocol of multilocus enzyme
in contaminated environments? For example, direct electrophoresis discussed later facilitates a rapid
assessment of mercuric reductase is possible in pro- assessment of these physiological ecology questions
tein extracts from mercury-contaminated freshwater through the analysis of protein molecules in their
systems (18). natural states.
3. How do specific microorganisms respond to speci-
fic environmental stimuli through inducible gene EXTRACTION OF PROTEINS FROM AQUATIC
expression? For example, the production of specific ENVIRONMENTS
‘‘head-shock’’ or stress proteins by bacteria culti-
vated under relatively uncomfortable environmental Several protocols have been developed to facilitate the
conditions has been a major application of proteome recovery of proteins directly from environmental samples.
assessment in environmental microbiology (34). In The paramount consideration in selecting a method
some cases, microbial stress proteins have been for recovery of proteins from environmental samples
 ASSESSING MICROBIAL PROTEOMES IN THE ENVIRONMENT 307

is whether the study is a general survey of protein sediments, freshwater, seawater, and wastewater (20). In
molecules or whether it focuses on a few targeted proteins general, sonication is the preferred method for cell-lysis
assayed by function. In the former case, there may be when the preservation of protein structure and function
no need to preserve the structure or catalytic function is essential. Chemical cell-lysis techniques including the
of recovered proteins and relatively harsh but thorough use of detergents and alkali or boiling are used when
extraction techniques may be used. When emphasis is the resolution of proteins requires denaturation (44–46).
placed on preserving protein activity, conservative cell- Because large numbers of polypeptides are expected in
lysis techniques are preferred. For the recovery of virus- complex microbial communities, sample fractionation is
associated or other cell-free proteins from aquatic samples, usually required before the attempt to separate pro-
a combination of ultracentrifugation, adsorption, and/or teins on the basis of molecular mass, ionic charge, or
precipitation is used (17) (Fig. 2). both. Sample fractionation can be achieved by differ-
Protein extraction methods have been developed for ential centrifugation, molecular sieves, and precipita-
a variety of environmental sample materials including tion (7).

Water sampling
Centrifuge 0.01−100 L of water samples at 25,000 g for 25 minutes at 4°C to
separate cell-associated proteins from extracellular proteins

Sample fractionation Stress dose-response experiments

Ultracentrifugation or

Protein extraction
(1) Cell-free fraction; (2) Denatured proteins; (3) Preserved proteins

Resuspend pellet in 1−5 ml cold buffer Concentrate protein molecules from the
containing 20 mM Tris-Cl, 1 mM supernatant by one of several alternative methods
Phenylmethyl sulfonyl fluoride, pH 7.4 (1) Liquid chromatography using for example,
preparative reverse phase coating, C4,
pH-stable polystyrene matrix columns
Sonicate to lyse cells at 4°C, Up to 10 (2) Polyethylene glycol (mol. wt. 4,000−6,000)
minutes at 20 kHz with a microtip adsorption
(3) Ultrafiltration using stirred prefabricated cells
(Amicon, Danvers, MA)
Centrifuge at 25,000−30,000 g for 30
minutes at 4°C to remove cell debris

Determine protein concentration spectrophotometrically [44]

Protein resolution
(1) Isoelectric focusing; (2) Gel or capillary electrophoresis; (3) Chromatography

Protein display
(1) Silver staining; (2) Enzyme staining; (3) Autoradiography

Protein identification
Mass spectrometry, Antigen−antibody reactions, Protein sequencing, or Enzyme Figure 2. Flow diagram of proce-
dures used in the construction of
microbial protein profiles in environ-
Posttranslation modification analysis mental assessments. The initial vol-
Glycosylation, Phosphorylation, Acylation, Alkylation, Carboxymethylation, Sulfation ume of water sample is determined
based on preliminary determinations
of cell density and potential contam-
Bioinformatics inating materials that may interfere
Protein database consultation for sequence comparisons with protein extraction and resolu-
tion.
308 ASSESSING MICROBIAL PROTEOMES IN THE ENVIRONMENT

RESOLUTION AND DISPLAY OF PROTEINS EXTRACTED one hour compared with the typical four to ten hours
FROM ENVIRONMENTAL SAMPLES required for slab gel electrophoresis. In general, capil-
lary electrophoresis resolves polypeptides according to
Gel Electrophoresis ionic charge, but the system is adaptable to the use
Polyacrylamide gel electrophoresis is the most widely used of various detectors, including conductivity, ultraviolet
method for protein resolution. The resolution power of and visible light absorbance, radioactivity, and fluores-
one-dimensional electrophoresis is limited to less than cence. It is also possible to couple capillary electrophore-
50 polypeptides, but it remains the most widely used sis with mass spectrometry for measurement of the
method when highly specific methods are available to molecular mass of resolved polypeptides (49). Capillary
select from for the polypeptide(s) of interest. In theory, electrophoresis has been used to study various aspects
two-dimensional polyacrylamide gel electrophoresis (2D- of proteome assessment in environmental microbiology,
PAGE) is capable of simultaneous separation of thousands including the construction of protein profiles for phy-
of proteins and is uniquely suitable for resolving logenetic differentiation (51), changes in bacterial outer
complex and highly diverse preparations (43,47). In membrane proteins in response to toxic chemicals (51),
practice, however, several factors can reduce the resolving and enzyme polymorphism in marine bacterial communi-
capacity of 2D-PAGE, particularly with respect to sample ties (33).
preparation and solubilization. The configuration of the
interface between the first dimension (isoelectric focusing
(IEF) to resolve polypeptides according to ionic charge) and Radioisotopes
the second dimension (polypeptide resolution according Radioisotopes have been used extensively in proteome
to molecular size) is also important. It is now possible research involving single microbial species and their use
to analyze single spots from a 2D-PAGE gel partly is increasing in in situ investigations. The radioisotope
because the availability of immobilized pH gradients have tracer approach takes advantage of the fact that many
eliminated the problem of gradient instability during microorganisms are able to incorporate externally supplied
IEF. In addition, the reproducibility of electrophoretic amino acids into polypeptide chains. For these organisms,
conditions and the results have improved with the 35
S-labeled amino acids, typically methionine, is added
commercial availability of ready-made gels in a variety to the growth medium. For organisms that engage in
of prescriptions.
de novo protein synthesis, radioisotope is supplied as
A few precautions enable a relatively trouble-free
H2 35 SO4 . When working with natural environmental
PAGE analysis. First, it is important that proteins
water samples, it is better to use a combination of
remain soluble throughout the electrophoresis process.
radiolabeled amino acids and sulfuric acid so as to cover
Therefore, chaotropes and surfactants such as thiourea
the entire range of microbial protein synthesis preferences.
and sulfobetaine are often used to enhance protein
The radioactive isotope tracer technique has been used to
solubility (8,46). Secondly, nucleic acids must be removed
from samples before electrophoresis to reduce the viscosity detect the synthesis of specific proteins in the immediate
of samples. A combination of carrier ampholites and aftermath of exposure of freshwater microorganisms to
ultracentrifugation may be used to remove nucleic acids if toxic chemicals (22).
the extraction is conducted at high pH. Alternatively, For most applications, cells pooled from aquatic sam-
endonuclease digestion is recommended when sample ples are usually incubated for up to an hour with
preparation is performed at high pH (40 mM Tris). radioactive precursors of protein synthesis. During the
Thirdly, it is important to eliminate disulfide bonds incubation period, cells may be exposed to a variety of
that act to reduce protein solubility. Therefore, chemical- stressful environmental conditions (e.g., toxic pollutant
reducing agents such as β-mercaptoethanol, dithiothreitol, exposure, temperature shift, high salt concentrations, or
or tributyl phosphine are used before isoelectric focusing. virus infection). Proteins are then extracted from the
Finally, when working with environmental samples, cells and resolved electrophoretically. The radiolabeled
humic and fulvic substances are usually coextracted polypeptides are subsequently detected by exposing gels
and must be removed. Several commercially available to an X-ray film for autoradiography, or if capillary
chromatographic matrix columns (for example, from electrophoresis is used, detection is achieved by direct
Biorad, Hercules, Calif.) may be used to reduce the measurement of radioactivity. Control experiments with
concentration of these compounds in samples in which cells incubated under relatively normal environmental
they present a problem. conditions allow the detection of inducible proteins synthe-
sized in response to specific environmental stress stimuli.
Capillary Electrophoresis Quantitative assessment of radioactivity incorporated into
The use of capillary electrophoresis equipment to resolve each polypeptide band is achieved with radioactive image
complex preparations of biomolecules is increasing (48). analyzers. Such quantitative data provide valuable cross-
Capillary electrophoresis of protein molecules is per- scale information on the maintenance of homeostasis in
formed in very thin (internal diameter of 25–100 µm) microbial communities under variable environmental con-
fused silica tubes ranging in length from 25 to 100 cm. ditions. Intensely labeled polypeptides of interest can be
The dimensions of the capillary allow the use of high isolated, sequenced, and compared with existing protein
electric fields, leading to very fast separation of pro- sequence databases for possible identification of func-
teins with a protocol completion time of less than tion (5).
 ASSESSING MICROBIAL PROTEOMES IN THE ENVIRONMENT 309

Immunofluorescence and Enzyme-Linked Immunosorbent but not much work has been done on the application of
Assay MLEE to whole microbial communities. In general, conser-
vative extraction of total proteins present in cells pooled
Antigen–antibody reactions targeted against ‘‘signature
from 1 to 100 L of water samples is achieved by sonication
proteins’’ are widely used for immunological detection
and centrifugation. Protein extracts are then resolved on
of microorganisms in environmental samples (52). The
gels made with starch or very thin slabs of polyacrylamide
increasing availability of genomic sequences is contribut-
under nondenaturing conditions. After electrophoresis, the
ing to rapid revision and developments in the use of anti-
location of enzymes that depend on the redox reactions can
sera for tracking and identifying microorganisms through
be ascertained by incubating the gel in solutions such that
immunofluorescence and enzyme-linked immunosorbent
the enzymatic reaction is coupled to the electrochemical
assays (ELISA) (52). Furthermore, because monoclonal
production of colorful dyes (45). A quantitative method
antibodies recognize very specific protein targets, their
has also been developed for routine screening of enzyme
use have virtually eliminated traditional limitations of
activity in natural microbial communities by immobilizing
immunologically based techniques involving polyclonal
proteins on nitrocellulose membranes before the develop-
antibodies. Monoclonal antibody-based protocols are also
ment of staining reactions (21). The membrane method is
amenable to quantitative assessment. Immunofluores-
rapid, but precludes the detection of multiple bands that
cence and immunomagnetic extraction of proteins from
aquatic sources are also undergoing rapid development, indicate molecular diversity MLEE.
but are currently limited by the fact that microbial anti- Another approach toward assessing microbial enzyme
gens exhibit different ranges of immunological specificity. profiles in aquatic environments involves the use of
In addition, the type-specific antigens may be shared by capillary electrophoresis coupled with the detection of
related microorganisms (52). fluorescent products of enzyme activity. For example,
investigation of β-glucosidase isozymes in the micro-
Enzyme Electrophoresis bial population present in 100 L of seawater revealed
unsuspected isozyme polymorphism (9,33). Pooled pro-
Microbial enzymes are invariably of central importance tein extracts from aquatic microorganisms can be
to the major research issues in environmental microbio- loaded onto capillary tubes filled with buffer contain-
logy. Hence, the influences of environmental conditions on ing fluorogenic substrates, for example, resorufin β-D-
enzyme activity in nature have implications for the under- glucoside for glucosidase activity. During electrophoresis,
standing and controlling of desirable processes ranging isozymes separate according to electrophoretic mobility.
from bioremediation of polluted local environments to the The electrical current through the capillary is then inter-
impact of global environmental change on ecosystem func- rupted for approximately 10 minutes to allow interaction
tions sustained by the microbial community. It is generally between enzyme and substrate. Fluorescent products
well understood that enzyme activities depend on assay are then detected at the anodic end of the capil-
conditions. Therefore, the forms and functions of specific lary.
enzymes in their natural habitat can be expected to vary
somewhat from assays conducted under laboratory condi- Mass Spectrometry
tions. Therefore, it is important to develop methods useful
for understanding enzyme diversity in microbial popu- The coupling of gel or capillary electrophoresis to mass
lations present in the environment. The fact that many spectrometry for the identification of polypeptides is
enzymes depend on redox reactions for effective catal- a rapidly evolving protocol (5,6). Standardized protocols
ysis has been exploited to produce colorimetric assays using this approach have generated reproducible maps of
amenable to rapid assessment of enzyme kinetics and 2D gels that allow comparisons across multiple experi-
enzyme polymorphism in environmental microorganisms. ments. Consequently, the identification of polypeptides
Such assays are linked to sensitive indicator dyes such through mass spectrometry can be achieved by matching
as formazan that assume color simultaneously with the profiles across phylogenetic boundaries. Both matrix-
progression of enzyme-catalyzed reactions (19,30,45). The assisted laser desorption-ionization mass spectrometry
technique can be accomplished on the same gel or mem- (MALDI-MS) and electrospray ionization mass spectrome-
brane used for protein resolution, thereby producing a try (ESI-MS) represent technological advancements in this
documented zymogram. category. Mass spectrometric analyses of peptides require
Because small variations in protein properties can enzymatic digestion of proteins. Therefore, not every
influence the course of microbial adaptation and evolu- routine gel-staining technique can be used for primary
tionary innovation, techniques such as multilocus enzyme resolution of protein molecules. For example, traditional
electrophoresis (MLEE) have been developed to capture silver-staining methods are generally not compatible with
the occurrence and distribution of allozymes (multiple enzymatic digestion unless they are modified to avoid the
enzymes with similar functions encoded by different alleles use of cross-linkers and strong oxidizing and reducing
at the same genetic locus) and isozymes (multiple enzymes agents. A compilation of protein-staining methods that
with identical functions encoded by different genetic loci) are compatible with mass spectrometry is available from
in microbial populations (31). Hence, MLEE yields direct Protana (Table 1). Mass spectrometric fingerprinting of
information on niche diversity and selective pressure at protein molecules allows rapid collection of information
the molecular level in a given microbial community. MLEE on a large number of proteins in a given sample, but
has also been used extensively to investigate the determi- ultimately, several pieces of data derived by using dif-
nants of population dynamics in single microbial species, ferent methods must be integrated to identify a given
310 ASSESSING MICROBIAL PROTEOMES IN THE ENVIRONMENT

Table 1. Selected World Wide Web Resources for Environmental Micro-

bial Proteomics Research
Category Web Address

Protein Sequence Databases

SWISS-PROT http://www.expasy.ch/sprot/sprot-top.html
NCBI http://www.ncbi.nlm.nih.gov/

Protein Patterns and Profiles

PROSITE http://www.expasy.ch/sprot/prosite.html
Prowl http://prowl.rockefeller.edu/
Protein prospector http://prospector.ucsf.edu/mshome.html

Metabolism Databases
ENZYME http://www.expasy.ch/sprot/enzyme.html
EcoCyc http://www.ai.sri.com.ecocyc/ecocyc.html

Two-dimensional Analysis
ECO2DBASE http://www.pcsf.brcf.med.mich.edu/eco2dbase
SWISS-2DPAGE http://www.expasy.ch/ch2d/ch2d-top.html
WORLD-2DPAGE http://www.expasy.ch/ch2d/2d-index.html

Three-dimensional Analysis
Swiss-SDImage http://www.expasy.ch/sw3d/sw3d-top.html
PDB http://www.pd.bnl.gov

Some Proteomics Companies

Ciphergen http://www.cyphergen.com
Kendrick laboratories http://www.msn.fullfeed.com/∼-kendrick/
Large scale biology corporation http://lsbc.com
Protana http://www.protana.com
Proteome Inc. http://quest7.proteome.com/
Proteome-systems http://www.Proteomesystems.com
Proteometrics http://www.proteometrics.com
Scanalytics http://scanalytics.com

protein unequivocally. A major follow-up aspect of polypep- screened with biomolecules from unknown samples for
tide mass fingerprinting is the search for characteristic the determination of protein function. Laser-based mass
matches in established protein databases (Table 1). detection is then employed to measure the molecular
mass of proteins. The stringency of molecular recognition
The ProteinchipTM Array in the array can be adjusted by washing and reading
procedures to optimize signal-to-noise ratios. Potential
The convergence of microchip electronics and high-fidelity
applications of the protein microarray in environmental
biosensor technology continues to revolutionize environ-
mental microbiological research. Microarray technology microbiology include the screening of proteins extracted
involves the immobilization of large numbers of biological from a natural microbial community for interaction with
molecules on tiny chip platforms for subsequent screening a given toxic metal. For such a task, metal ions are
with molecular probes. This approach was initially deve- immobilized in the microarray wells and protein extracts
loped to support large-scale genome sequencing efforts will be added, followed by a stringent wash and read
to facilitate the screening of large numbers of distinct process. For example, molecular diversity in enzymes such
DNA or mRNA molecules (for example, the GeneChipTM as δ-aminolevulinate dehydratase and mercuric reductase
system developed by Affymetrix). Similarly, sophisticated that interact with lead and mercury, respectively, can be
technology is now available for proteome assessments. rapidly assessed (18,25,26). Other biomolecules that can
The surface enhanced laser desorption-ionization (SELDI) be immobilized in microarray wells include membranes,
ProteinChipTM manufactured by Ciphergen (Palo Alto, CA) receptor molecules, and antibodies.
is designed to facilitate high throughput protein resolution
Neural Networks and Cluster Analysis for Protein Profile
and identification (49).
Analysis
The ProteinChipTM array is typically made of metallic
or plastic trays containing up to a thousand wells of On the basis of the molecular size of microbial genomes,
1 mm in diameter. The wells are chemically treated (e.g., a prediction can be made about the expected number of
with copper (II)) to capture trace quantities of unknown independent proteins present in a cell at a given time.
proteins (e.g., copper-binding proteins) in a given sample. Although the one gene–one protein conjecture has proven
Wells may also be coated with a known protein to be to be inexact, it can be surmised that between 500 and
 ASSESSING MICROBIAL PROTEOMES IN THE ENVIRONMENT 311

5,000 protein molecules are extractable, in theory, from maps, and three-dimensional structure databases. Table 1
each microbial component of environmental samples. The shows selected web site addresses that are useful
expected total number of proteins increases dramatically for consultation by both beginners and experienced
in environmental samples with highly diverse population proteomics researchers.
of species. The development of highly sophisticated pattern
recognition techniques is required for optimizing the
CONCLUSION
information generated by protein profiles in natural
microbial communities. Early attempts to identify micro-
organisms on the basis of protein profiles depended Accurate assessment of genetic expression is fundamen-
on statistical cluster analysis for recognizing matches tal to all applications of biotechnology that involve the
between different protein profiles (42). More recent activities of viable cells. Furthermore, in biotechnological
approaches to the interpretation of profiles from complex processes with an environmental dimension, the need
heterogenous microbial communities rely on strategies for consistent analysis of gene expression is paramount
based on neural network computing (53,54). Artificial because of the extensive range of environmental para-
neural networks, also known as parallel distributed meters that are known to influence the transformation of
processing, have an aptitude for comparative assessment genetic potential to metabolic and ecological advantage.
and retrieval of experiential information concerning The goals of assessing microbial proteomes in the environ-
large databases. Protein-profile management techniques ment include (1) monitoring the expression and activity
should be commensurate with the molecular resolution of specific enzymes or structural proteins in targeted pop-
capacity (Fig. 3). For example, profiles constructed using ulations that inhabit natural ecosystems, (2) identifying
one-dimensional polyacrylamide gel electrophoresis can specific indicator organisms present in complex heteroge-
be managed with visual or linear image analysis nous microbial communities, (3) monitoring changes in
programs, but two-dimensional electrophoresis requires the population structure of microbial communities as they
more sophisticated pattern recognition techniques. A respond to various environmental stimuli, and (4) tracking
profiling database constructed with three-dimensional the tempo and mode of molecular evolution in microbial
information, for example, on protein hydrophobicity, populations. With imminent technological advancements
ionic charge, and molecular mass, on large numbers in protein extraction and analysis and in the robustness
of organisms would be facilitated with neural network of protein characteristics databases, these already ambi-
computing for pattern recognition. tious goals will undoubtedly be expanded dramatically to
address the increasingly multifaceted research questions
INTERNET RESOURCES FOR ENVIRONMENTAL in environmental microbiology.
MICROBIAL PROTEOME ASSESSMENT
Acknowledgments
This work was supported in part by a research fellowship from
A comparative analysis of protein characteristics is central
the Josiah Macy, Jr., Foundation.
to proteomics research. Therefore, access to databases
available on the World Wide Web for protein conformation,
sequence, and function are becoming indispensible. In BIBLIOGRAPHY
general, web sites for proteome researches specialize in one
of six activities, namely, sequence depository and retrieval, 1. M. A. Traini et al., Electrophoresis 19, 1941–1949 (1998).
comparative analysis of patterns and profiles, metabolic 2. M. Wilm et al., Nature 379, 466–469 (1996).
functions, posttranslational modification, two-dimensional 3. M. R. Wilkins et al., Biotechnology 14, 61–65 (1996).

Protein-profile assessment tools

Neural networks
3D Combination

Protein Chip-MS
Computerized cluster Figure 3. Protein-profile management tool
analysis
requirement is proportional to the power of
molecular resolution. MS = mass spectrome-
try. 1D, 2D, and 3D refer to one-, two-,
and three-dimensional resolution, respec-
2D tively. The pattern recognition capacity of
Visual image analysis
protein-profile management tools increases
1D linearly on the vertical axis. Techniques higher
on the vertical scale can also accommodate
0 30 100 300 30,000 100,000
tasks performed by lower, less sophisticated
Number of polypeptides techniques.
312 ASSIMILABLE ORGANIC CARBON (AOC) IN DRINKING WATER

4. M. R. Wilkins et al., Electrophoresis 17, 830–838 (1996). 35. O. C. Dermer, V. S. Curtis, and F. R. Leach, Biochemical
5. M. R. Wilkins and K. L. Williams, J. Theor. Biol. 186, 7–15 Indicators of Subsurface Pollution, Ann Arbor Science, Ann
(1997). Arbor, Mich., 1980, pp. 1–203.
6. M. R. Wilkins et al., J. Mol. Biol. 289, 645–657 (1999). 36. R. H. Findlay and F. C. Dobbs, in P. F. Kemp et al., eds.,
Handbook of Methods in Aquatic Microb. Ecology, Lewis,
7. M. R. Wilkins et al., eds., Proteome Research: New Frontiers
Boca Raton, Fla., 1993, pp. 271–284.
in Functional Genomics, Springer Verlag, New York, 1997,
pp. 1–243. 37. P. A. Noble, R. Citek, and O. A. Ogunseitan, Electrophoresis
19, 528–535 (1998).
8. J. Yan et al., J. Chromatogr. A736, 291–3021 (1996).
38. O. A. Ogunseitan, Environ. Test. Anal. 5, 36–40 (1996).
9. M. Agis et al., Microb. Ecol. 36, 66–74 (1998).
39. O. A. Ogunseitan, Sci. Prog. 78, 183–204 (1995).
10. A. Blom, W. Harder, and A. Matin, Appl. Environ. Microbiol.
58, 331–334 (1992). 40. J. E. Brower, J. H. Zar, and C. N. von Ende, Field and
Laboratory Methods for General Ecology, WCB McGraw-Hill,
11. R. G. Burns, in W. G. Characklis and P. A. Wilderer, eds.,
Boston, Mass., 1998, pp. 171–198.
Structure and Function of Biofilms, John Wiley & Sons, New
41. V. Hatzimanikatis, L. H. Choe, and K. H. Lee, Biotechnol.
York, 1989, pp. 333–349.
Prog. 15, 312–318 (1999).
12. P. Cash, Anal. Chim. Acta 372, 121–145 (1998).
42. R. M. Niemi et al., Appl. Environ. Microbiol. 59, 2190–2196
13. R. J. Chrost, in J. Overbeck and R. J. Chrost, eds., Aquatic (1993).
Microb. Ecology, Springer Verlag, New York, 1990, pp. 47–78.
43. R. A. VanBogelen et al., Electrophoresis 20, 2149–2159
14. H.-G. Hoppe, in P. F. Kemp et al., Handbook of Methods in (1999).
Aquatic Microbial Ecology, Lewis, Boca Raton, Fla., 1993,
44. M. M. Bradford, Anal. Biochem. 72, 248–254 (1976).
pp. 423–431.
45. M. P. Deutscher, ed., Guide to Protein Purification, Academic
15. A. J. Jaan, B. Dahlof, and S. Kjelleberg, Appl. Environ.
Press, San Diego, Calif., 1990, pp. 68–421.
Microbiol. 52, 1419–1421 (1986).
46. G. P. Manchenko, Handbook of Detection of Enzymes on
16. O. A. Ogunseitan, J. LeBlanc, and E. Dalmasso, in P.
Electrophoretic Gels, CRC Press, Boca Raton, Mich., 1994,
Rochelle, ed., Environmental Molecular Microbiology: Pro- pp. 3–27.
tocols and Applications, Horizon Scientific Press, Norfolk,
47. B. Herbert, Electrophoresis 20, 660–663 (1999).
England, 2000, In Press.
48. K. Ou et al., Am. Biotechnol. Lab. 17, 16–20 (1999).
17. O. A. Ogunseitan, in K. Scow et al., Integrated Assessment
of Ecosystem Health, CRC Press, Boca Raton, Fla., 2000, 49. A. L. Lagu, Electrophoresis 20, 3145–3155 (1999).
pp. 207–222. 50. I. Kustos et al., Electrophoresis 19, 2324–2330 (1998).
18. O. A. Ogunseitan, Appl. Environ. Microbiol. 64, 695–702 51. T. W. Hutchens and T.-T. Yip, Rapid Commun. Mass Spec-
(1998). trom. 7, 576–580 (1993).
19. O. A. Ogunseitan, in K. Cooksey, ed., Molecular Approaches 52. I. Kusos et al., Electrophoresis 19, 2317–2323 (1998).
to the Study of the Ocean, Chapman & Hall, London, U.K., 53. S. F. Wright, in M. A. Levin, R. J. Seidler, and M. Rogul,
1998, pp. 147–159. eds., Microb. Ecology: Principles, Methods, and Applications,
20. O. A. Ogunseitan, in A. D. L. Ackkermans, J. D. van Elsas, McGraw-Hill, New York, 1992, pp. 45–63.
and F. J. De Bruijn, eds., Molecular Microbial Ecology 54. P. A. Noble, K. D. Beadle, and M. Fletcher, Appl. Environ.
Manual, Kluwer Academic Publishers, The Netherlands, Microbiol. 63, 1762–1770 (1997).
1998, pp. 4.1.6:1–12. 55. D. W. Rice and D. Eisenberg, J. Mol. Biol. 267, 1026–1038
21. O. A. Ogunseitan, J. Microbiol. Methods 28, 55–63 (1997). (1997).
22. O. Ogunseitan, Microb. Ecol. 31, 291–304 (1996).
23. O. Ogunseitan, J. Microbiol. Methods 17, 273–281 (1993).
24. C. H. Scilling, J. S. Edwards, and B. O. Palsson, Biotechnol. ASSIMILABLE ORGANIC CARBON (AOC) IN
Prog. 15, 288–295 (1999). DRINKING WATER. See BIODEGRADABLE DISSOLVED
25. O. A. Ogunseitan, S. Yang, and E. Scheinbach, Biol. Bull. ORGANIC CARBON IN DRINKING WATER
197, 283–284 (1999).
26. O. A. Ogunseitan, S. Yang, and J. Ericson, Soil Biol. Biochem.
32, 1899–1906 (2000).
27. E. Tanoue, M. Ishii, and T. Midorikawa, Limnol. Oceanogr.
41, 1334–1343 (1996). ASSIMILABLE ORGANIC CARBON (AOC) IN
28. S. Suzuki, K. Kogure, and E. Tanoue, Mar. Ecol. Prog. Ser. TREATED WATER: DETERMINATION AND
158, 1–9 (1997). SIGNIFICANCE
29. R. G. Quivey, W. L. Kuhnert, and R. C. Faustoferri, Methods
Cell Sci. 20, 165–179 (1998). DICK VAN DER KOOIJ
30. R. K. Selander et al., Appl. Environ. Microbiol. 51, 873–884 Kiwa Water Research
(1986). Nieuwegein, The Netherlands
31. A. Shevchenko et al., Proc. Natl. Acad. Sci. U.S.A. 93,
14440–14445 (1996). CONTROLLING MICROBIOLOGICAL WATER QUALITY
32. L. Tonella et al., Electrophoresis 19, 1960–1971 (1998).
33. M. Unanue, Microb. Ecol. 7, 36–48 (1999). The quality of drinking water at the consumer’s tap
34. K. Watson, Adv. Microb. Physiol. 31, 183–216 (1990). depends on a chain of processes, namely, the effects
 ASSIMILABLE ORGANIC CARBON (AOC) IN TREATED WATER: DETERMINATION AND SIGNIFICANCE 313

of water treatment on the raw water, the effects of biodegradable compounds by microorganisms, formation
distribution on the composition of treated water, and the of biofilms on the exposed surface detachment of (micro-
effects of the household plumbing system, respectively. )organisms and dead biomass from the pipe wall and
Water quality criteria related to health are defined in accumulation of sediments, respectively. The presence
national and international regulations and the water of biodegradable compounds is a major driving force in
industry has the obligation to provide a safe drinking these regrowth processes but other environmental condi-
water. Microbiological safety is of major importance tions (e.g., temperature, system hydraulics, and retention
because ingestion of pathogens can have a direct impact time) and physicochemical processes (adsorption, oxida-
on the health of the consumer. For this reason, multiple tion/reduction, and coagulation/sedimentation) also have
barrier approaches are used in water treatment in significant effects on the extent to which microorganisms
combination with frequent monitoring of indicator bacteria multiply. When a disinfectant is added to water before
in treated water and pathogens (protozoa and viruses) in distribution, the situation is even more complex because
raw water, respectively. A multiple barrier approach is the concentration of disinfectant generally changes in time
also used to ensure the safety of drinking water in the and differs among various locations in the system (22).
distribution system (1). Barriers against contamination of Bacteria present in the biofilm are released into the
drinking water in the distribution system include (1) a water and may be detected in routine monitoring of the
constant high pressure in the mains, (2) cross-connection microbiological quality of drinking water. Most bacteria
prevention measures, (3) safety procedures for activities predominating in regrowth processes do not seem to
in the distribution system, (4) reliable materials and constitute a direct hygienic or aesthetic problem (23,24),
construction techniques. Also, appropriate water quality but their ability to produce biomass under the prevailing
monitoring is important to check the effectiveness of these conditions is a main cause of the enables Parts of the
preventive measures. Maintaining a disinfectant residual biofilm sloughing from the surface may settle in pipes
in drinking water during distribution is considered as at low flow conditions and contribute to the formation of
an additional safety factor (2), but in a few European sediments. Biofilms and sediments of decaying biomass are
countries it is allowed to distribute drinking water without niches for undesirable bacteria including coliforms (25),
disinfectant residual (3–5). The main disadvantages of aeromonads (12), Legionella spp. (13,15), Mycobacterium
maintaining a disinfectant residual are the effect on spp. (11,14), but also fungi and yeasts may be present (26).
taste and odor of the water, and/or the formation of Certain invertebrates use biofilms and sediments as a food
disinfection by-products, including trihalomethanes with source (19,20). (See also this Encyclopedia Invertebrates
toxic properties (6–9). and protozoan in drinking-water distribution systems, Van
A main function of the disinfectant residual is the pre- Lieverloo et al.).
vention of multiplication of microorganisms in drinking The rate and extent of multiplication of microorganisms
water distribution (‘‘regrowth’’). Controlling regrowth is in water depend on the concentration and the composition
important to the water industry for the following main of the available food sources. These food sources provide
reasons (1) multiplication of coliform bacteria results in energy and the elements needed for the biomass synthesis.
noncompliance with regulations (10), (2) the multiplica- Heterotrophic bacteria utilize organic compounds as
tion of opportunistic pathogenic bacteria, including species a source of energy and carbon. In addition, certain
of Legionella, Mycobacterium, Pseudomonas, Aeromonas, inorganic compounds are required, in particular nitrogen
and Flavobacterium poses a potential health threat to the and phosphorus, and a number of other elements at
consumer (11–15), (3) heterotrophic plate count values much lower amounts, for example, sulfur. The need for
may exceed those defined in national legislation and may carbon, nitrogen, and phosphorus can be derived from the
cause problems in the food industry, (4) microbial activity gross composition of microbial biomass, C5 H7 NO2 P1/30 ,
and biomass may affect turbidity, color, taste, and odor of and the proportion (50%) of organic carbon used for
the water (16), (5) microbiological processes may acceler- dissimilation (27). Hence, organic compounds serving as
ate corrosion (Microbiologically Induced Corrosion, MIC) a source of energy and carbon are needed in much
of pipe materials (17,18), (6) invertebrates multiplying on larger amounts than the inorganic nutrients nitrogen and
bacterial biomass may cause consumer complaints (19,20). phosphorus (C : N : P = 100 : 10 : 1). Nitrate, a common
In the absence of a disinfectant residual regrowth compound in drinking water, is a good nitrogen source
is controlled by distributing biologically stable water, for many aquatic bacteria, although certain types
as obtained by a far-reaching removal of biodegradable (e.g., Aeromonas spp.) require ammonia-nitrogen for
compounds from water during treatment (21). Assessment growth (28). Phosphate is also a common compound in the
of the degree of biological (in)stability requires a method aquatic habitat, and it is only needed in very low quantities
for determining the concentration of growth-promoting
relative to the carbon source. Hence, the availability of
compounds in water or the microbial growth potential.
suitable organic compounds will be the growth-limiting
This entry describes the assimilable organic carbon (AOC)
parameter for heterotrophic bacteria in most water types.
method and its significance for defining biological stability.
Some studies suggest that phosphate may be growth
limiting in certain types of groundwater, which contain
REGROWTH PROCESSES AND GROWTH KINETICS a relatively high concentration of organic compounds (29).
The amount of biomass produced on a certain amount
Regrowth is a complex phenomenon affected by many of substrate depends on the yield (Y) values of the bacteria
variables. Key processes in regrowth are the uptake of for the present compound(s). Generally, 1 mg of substrate
314 ASSIMILABLE ORGANIC CARBON (AOC) IN TREATED WATER: DETERMINATION AND SIGNIFICANCE

carbon yields 1 mg of biomass (dry weight) during aerobic than the typical aquatic bacteria for identical substrates.
growth (27). Expressing the yield value as number of Still, the Ks values of 13 to 72 µg of glucose-C/L (31,32), and
bacteria (colony forming units, CFU) is more appropriate 50 to 150 µg of yeast extract/L (38) as reported for E. coli
for describing regrowth processes. Typical yield values are also relatively low. Despite these kinetics, E. coli and
for compounds serving as favorable growth substrates for P. aeruginosa do not belong to the indigenous bacterial
heterotrophic bacteria (low molecular organic compounds) community of drinking water. This may be due to the
of 4 × 106 to 2 × 107 CFU have been reported (see later). relatively slow growth of these bacteria at temperatures
Hence, utilization of 1 µg of C/L theoretically corresponds below 15 ° C (34) and the very low concentrations of easily
with a maximum colony count value of 104 CFU/mL. The available compounds in treated water.
substrate concentration also affects the rate of growth. As a result of maintenance energy requirements, a cer-
This relationship is described by the Monod equation (30): tain minimum substrate concentration (Smin ) is required
for survival without growth (35). Typical maintenance
V = Vmax × S/(Ks + S) (1), rates for bacteria due to endogenous substrate utiliza-
tion range from 0.005 to 0.02 h−1 (36), but a much higher
where: V = the growth rate (doublings/hour) in the expo- value (0.3 h−1 at 37 ° C) has been reported for E. coli (31).
nential growth phase at substrate concentration S; Vmax = The values for the substrate affinity of a number of bacteria
maximum growth rate (h−1 ); Ks = the substrate satura- presented in Table 1, are below the value of maintenance
tion constant, is the concentration of S at which V = rate. Consequently, these bacteria are not able to grow at
1/2Vmax . An example of such a relationship is presented in a concentration of 1 µg of C/L of the specified compounds.
Figure 1, which clearly demonstrates that the increase of In practice the situation is more complicated because
the growth rate is less than proportional to the increase of in most situations, several compounds will be present.
the substrate concentration. However, at very low values Most of the bacteria described in Table 1 are able to uti-
for S, that is, S Ks then: lize a number of substrates simultaneously when present
at an individual compound concentration of 1 µg C/L. A
V = Vmax /Ks × S (2) mixture of amino acids stimulated the multiplication of
A. hydrophila at an individual compound concentration as
low as 0.05 µg C/L (28).
and the growth rate is linearly related with the
The growth kinetics as presented in Table 1 demon-
substrate concentration S and also depending on the
strate that low concentrations of easily available organic
values for Vmax en Ks , which are constants for specific
compounds can cause a rapid and significant growth of
combinations of organism and substrate. The quotient
bacteria present in treated water. For this reason, the
Vmax /Ks (h−1 . . . µg−1 . . . L, the substrate affinity constant
method for assessing the regrowth potential of treated
is numerically identical to the growth rate calculated for
water as developed in the Netherlands was mainly directed
a substrate concentration of 1 µg C/L.
at determining the concentration of such compounds.
Bacteria representing the indigenous bacterial flora
of drinking water have Ks values for easily degradable Assessment of the Microbial Growth Potential of Treated
low molecular weight compounds as low as a few Water
micrograms of C/L. Relatively low Ks values have been also
been observed for the high molecular weight compounds Already at the end of the nineteenth century, studies
amylopectin and amylose (Table 1). Escherichia coli, were conducted on bacterial multiplication in treated
coliforms, and also P. aeruginosa have higher Ks values water (44). It was observed that the number of culturable
bacteria was much higher in stored samples of a high-
quality deep-well water with a low concentration of organic
0.5 compounds (up to 5 × 105 CFU/mL) than in river water
Vmax samples (4,300 CFU/mL). The assumption was made that
0.4
the well water obtained from the chalk contained a
relatively high concentration of biodegradable compounds
Growth rate V (h−1)

because of the absence of bacteria in this water directly

0.3 after abstraction. On the other hand, it was known at that
time that the relatively strong growth in treated water
0.2 V = VMAX × S/(Ks + S) stored in flasks with a cotton was related to compounds
present in the air. Also, the difference in growth between
0.1 well water and river water was most probably the result of
Ks the multiplication of a few types of bacteria contributing
0 to the HPC values in well water, whereas many of the
0 1000 2000 3000 4000 5000 bacteria present in river water were not able to grow on a
Lactate concentration (µg C/I) solid medium. These observations show that determining
Figure 1. The growth rate of an Escherichia coli strain (isolated
the growth potential using colony counts of the indigenous
from river water) at various concentrations of lactate in bacterial community has a number of limitations. In
autoclaved slow sand filtrate (+PO4− 3− P and NH4 + -N) incubated the Netherlands, Beijerinck (45) suggested in 1891 that
at 25 ° C. The constants of the Monod relationship (Ks and Vmax ) the growth promoting properties of drinking water be
are given in Table 1. studied by inoculating selected pure cultures in sterilized
 ASSIMILABLE ORGANIC CARBON (AOC) IN TREATED WATER: DETERMINATION AND SIGNIFICANCE 315

Table 1. Growth Kinetics of Bacteria Isolated from Drinking Water∗

Organism Compound Temp. (° C) Ks (µg C/L) Vmax (h−1 ) Vmax /Ks (h−1 ). µg−1 .L Reference

Aeromonas hydrophila Acetate 15 11 0.15 0.013 28

Aeromonas hydrophila Glucose 15 16 0.28 0.018 28
Aeromonas hydrophila Amylose 15 93 0.26 0.0028 37
Aeromonas hydrophila Oleate 15 2.1 0.23 0.109 28
Citrobacter freundii Glucose 15 95 0.17 0.0018 38
Enterobacter sp. Glucose 15 60 0.21 0.004 38
Escherichia. coli Lactate 25 142 0.37 0.0026 This paper
Flavobacterium sp. Glucose 15 3.3 0.21 0.063 39
Flavobacterium sp. Glucose 15 109 0.15 0.001 40
Flavobacterium sp. Maltose 15 23.7 0.37 0.016 40
Flavobacterium sp. Maltopentaose 15 5.7 0.44 0.077 40
Flavobacterium sp. Amylose 15 26 0.50 0.020 40
Flavobacterium sp. Amylopectin 15 11 0.48 0.044 40
Klebsiella pneumoniae Maltose 15 51 0.49 0.0096 41
Klebsiella pneumoniae Maltopentaose 15 92 0.41 0.0045 41
Pseudomonas fluorescens Acetate 15 4. 0.18 0.045 42
Pseudomonas fluorescens Glucose 15 57 0.22 0.004 42
Pseudomonas aeruginosa Acetate 15 28 0.09 0.003 34
Spirillum sp. Oxalate 15 0.24 0.016 43

Note: ∗ growth measurements conducted in slow sand filtrate

samples of the water that was to be tested. He also the ability to utilize certain carboxylic acids, as produced
pointed out the necessity to prevent contamination by by ozonation (49). Regrowth phenomena gained also atten-
volatile compounds. However, no results were presented tion in other European countries and in the United States,
and it seems that this method did not attract much and a series of methods for assessing the growth potential
attention. In 1928 a completely different approach was of drinking water have been developed (Table 2).
reported by Heymann (46) who determined the effect of Batch tests are most commonly used because such
a series of passages of water through sand columns on test can be conducted under controlled conditions in the
the concentration of organic compounds, measured as laboratory. These tests give information about the growth
permanganate value. However, this method also did not potential but the test conditions do not reflect the processes
gain wide application. occurring in the distribution system. Another limitation
Interest in the regrowth phenomenon strongly of the AOC and BDOC tests is that the growth potential
increased in the Netherlands in the 1970s as a result of of treated water may also be affected by the presence
the introduction of ozonation and granular activated car- of inorganic compounds (e.g., ammonia and sulfides) and
bon (GAC) filtration in water treatment for the removal methane, which are not included in these tests. A variety
of persistent organic pollutants. In 1978 a method for of techniques and devices is available to simulate the
determining the heterotrophic growth potential of treated biofilm formation as occurring in distribution systems, for
water was described, which was based on determining the example, Rototorque system (49), Robbins device, coupon
maximum level of growth of a selected pure culture in test (50), and a biofilm monitor (51,52). A combination of
samples of water collected and contained in thoroughly growth tests, determining the (effects of) concentrations of
cleaned glass-stoppered Erlenmeyer flasks (47,48). Subse- rapidly and more slowly available compounds, as well as
quently, this method, which was designated as the AOC chemical analysis (e.g., for ammonia and methane), may
method, was improved by including a bacterial strain with be needed to assess the biostability of treated water. In the
Netherlands biostability assessment of water is conducted
by using the AOC test and the BFR test (5).
Table 2. Characteristics of Methods for Assessing the
Microbial Growth Potential of Treated Water
Method∗ Mode Organisms Parameter Reference AOC METHOD

AOC Batch Pure culture CFU 43,47,48,53 Principle of the Test

AOC Batch Indigenous bacteria ATP 54
AOC Batch Pure cultures ATP 55 As explained earlier, easily biodegradable organic com-
BGP Batch Indigenous bacteria Turbidity 56 pounds may cause significant growth at a level of a few
BDOC Batch Indigenous bacteria DOC 57,58,59 micrograms per liter. Therefore, a test was developed
BDOC Column Indigenous bacteria DOC 60,62 aiming at determining low concentration of these com-
BFR Column Indigenous bacteria ATP 5,51,52,62 pounds, which were designated as assimilable organic
∗
AOC, easily Assimilable Organic Carbon; BGP, Bacterial Growth Poten-
carbon (AOC), derived from earlier descriptions where
tial; BDOC, Biodegradable Dissolved Organic Carbon; BFR, Biofilm For- the terminology ‘‘assimilable compounds’’ was used (46).
mation Rate. Assessment of the concentration of easily assimilable
316 ASSIMILABLE ORGANIC CARBON (AOC) IN TREATED WATER: DETERMINATION AND SIGNIFICANCE

organic carbon (AOC) is based on growth measurements colony counts (Nmax , CFU/mL) obtained in the water types
of a mixed culture of two select bacterial strains in a tested.
sample of pasteurized water collected and contained in a Spirillum species strain NOX had been isolated from
thoroughly cleaned glass-stoppered Erlenmeyer flask. On slow sand filtrate enriched with 25 µg C/L of oxalate,
the basis of the maximum colony counts of these organ- glyoxylate and formate, respectively (43). The strain is
isms, the original concentration of substrates is calculated specialized in the utilization of carboxylic acids, including
using the yield values of the bacteria for acetate. Hence, formate, oxalate, glycollate and glyoxylate, which are
the AOC concentration is expressed as µg of acetate-carbon not used by strain P17. Growth tests conducted with
equivalents/L. mixtures of substrates with an individual compound
concentration of 1 µg of C/L clearly revealed that only
carboxylic acids were utilized at this low concentration
Test Strains (Fig. 2b). As a result of its preference for carboxylic acids,
The strains used in the AOC test are P. fluorescens the Nmax value of strain NOX gives information about
strain P17, which is capable of utilizing a wide the concentration of carboxylic acids in the water tested.
Yield factors have been determined for acetate and for
range of low molecular weight compounds at very low
oxalate, respectively. The yield (CFU/µg of carbon) of
concentrations (42) and a Spirillum sp. strain NOX, which
strain NOX for acetate is about four times higher than
only utilizes carboxylic acids (43). The general properties
for oxalate-carbon. This difference can be explained by
of these organisms are described in Table 3, together
the low energy content of oxalate (27). The organism is
with another organism specialized in the utilization of
used in a mixed culture strain P17 for determining the
carbohydrates (63).
concentration of easily assimilable organic carbon (AOC)
P. fluorescens strain P17 originates from tap water in (drinking) water.
and belongs to one of the most commonly occurring Flavobacterium species strain S12 was obtained from
biotypes of the fluorescent pseudomonads in drinking slow sand filtrate incubated at 15 ° C after enrichment with
water (64). Growth measurements conducted in drinking 100 µg of starch-C/L (63). The organism is specialized in
water supplemented with mixtures of compounds revealed utilizing carbohydrates, including amylose, amylopectin,
that the organism can utilize most or all amino acids, a maltose, and maltodextrins, which are not used by strain
number of carboxylic acids, some carbohydrates and a P17 nor by strain NOX. Growth tests conducted with
number of aromatic acids, at a concentration of 1 µg of C/L mixtures of substrates with an individual compound
when present as a mixture (Fig. 2a; 42). In addition, the concentration of 1 µg of C/L clearly revealed that only
organism requires a simple nitrogen source and multiplies certain carbohydrates were utilized at this concentration.
rapidly on agar media. For these reasons, strain P17 was Nmax values of strain S12 therefore give information
selected for determining the concentration of easily AOC about the concentration of maltose- and maltodextrin-
in (drinking) water (48). The yield factor of strain P17 like compounds in the water tested. Yield factors of strain
for acetate-carbon (4.1 × 106 CFU/µg carbon) is used for S12 for starch, maltose, and maltodextrins range from
calculating the AOC concentration from the maximum 2.0 × 107 to 2.3 × 107 CFU/µg C (40). Concentrations of

Table 3. Properties of P. fluorescens Strain P17, Spirillum sp. Strain NOX, and
Flavobacterium sp. Strain S12, Respectively (42,43,63)

P. fluorescens Spirillum sp.∗ Flavobacterium sp.

Characteristic strain P17 strain NOX strain S12

Origin Tap water Slow sand Slow sand

filtrate filtrate
Shape Rod Curved rod Rod
Motility + + −
Gram stain − − −
Oxidase + + +
N-sources NO3 /NH4 NO3 /NH4 NO3 /NH4
Max. growth temperature (° C) 30 30 30
Arginine dihydrolase + + −
Denitrification + (N2 ) − −
O/F test with glucose +/− −/− −/−
Hydrolysis of:
− Gelatin + − −
− Tween-80 + − −
− Starch − − +
− Chitin − − −
Preferred substrates Versatile Carboxylic Carbohydrates
organism acids
∗
a recent 16SRNA sequencing analysis indicated that the organism is related to the genus Ultramicrobium
(unpublished observation).
 ASSIMILABLE ORGANIC CARBON (AOC) IN TREATED WATER: DETERMINATION AND SIGNIFICANCE 317

(a) Other strains can be used for growth measurements to

2.5×105 0.20
determine the utilization of present or added substrates
Nmax h−1 or to determine the growth potential of the water for
a specific organism. Typical examples of such strains
2.0×105
include Aeromonas species (28), representatives of the
0.15
coliform group, including E. coli (65,66; Fig. 1) and also
P. aeruginosa (43).
NMAX (CFU/ml)

1.5×105

V(h−1)
0.10
Test Conditions
Sampling and transportation of the samples and sample
1.0×105
treatment are essential to obtain an accurate AOC value.
For this reason, representative samples of the water to
0.05 be investigated are collected in duplicate in thoroughly
0.5×105 cleaned (including heating at 550 ° C for four hours)
Erlenmeyer flasks. In these flasks the water samples
are transported, pasteurized, and incubated, respectively.
0 0 Hence, contact with surfaces and air is restricted as much
Blank AA CA CHA AR TM
as possible. Pasteurization aiming at inactivating the
Substrate mixture
indigenous bacterial community as present in the water
samples is conducted by placing the flasks in a water bath
(b) 3.0×105 0.30 at 90 ° C. The flasks are removed from the water bath when
Nmax V water temperature has reached 60 ° C and subsequently
are placed in an incubator at 60 ° C for 30 minutes. After
2.5×105 0.25
cooling with cold tap water, the test strains are added to

2.0×105 0.20
NMAX (CFU/ml)

(a) 106
, Blank
V(h−1)

1.5×105 0.15 , 5 µg C/I

Colony count (CFU/ml)

, 25 µg C/I
105
, 50 µg C/I
1.0×105 0.10 , 100 µg C/I
104

0.5×104 0.05
103
0 0
Blank AA CA CHA AR TM
102
Substrate mixture 0 10 25 50
Time (days)
Figure 2. (a) Maximum colony counts and growth rates of
P. fluorescens strain P17 in the presence of mixtures of compounds
at an individual compound concentration of 1 µg C/L. Blank:
(b) 4.0×106
slow sand filtrate without added substrate; AA, 19 amino acids;
CA, 14 carboxylic acids; CHA, 6 carbohydrates/alcohols; AR,
7 aromatic acids; TM, total mixture (46 µg C/l). Error bars indicate
duplicate measurements. Data adapted from Fig. 3. (b) Maximum 3.0×106
Nmax (CFU/ml)

colony counts and growth rates of Spirillum sp. strain NOX

in the presence of mixtures of compounds at an individual
2.0×106
compound concentration of 1 µg C/L. Error bars indicate duplicate
measurements.
Strain NOX
1.0×106

carbohydrates available for strain S12 in drinking water Strain P17

usually are very low (<1 µg of C/L) and therefore, the 0
0 50 100 150 200 250
organism is not included in the test for determining the Acetate (µg C/I)
concentration of easily assimilable organic carbon (AOC)
in (drinking) water. In certain water types however, a Figure 3. (a) Growth curves of P. fluorescens strain P17 in
significant concentration of carbohydrates may be present, pasteurized slow sand filtrate (DOC = 2.3 mg/L) supplemented
for example, when algal growth occurred in the raw water with different acetate-C concentrations (at 15 ° C); (b) calibration
and/or when a starch-derived coagulant aid is used in curves of strain P17 and strain NOX for acetate. Yield values are
water treatment (40). presented in Table 4.
318 ASSIMILABLE ORGANIC CARBON (AOC) IN TREATED WATER: DETERMINATION AND SIGNIFICANCE

the flasks. Cultures of the test strains grown in tap water conducted with strain NOX 43. The relationship between
at an initial concentration of 1 mg of acetate-C/L, in which Nmax values and the acetate concentrations is used for
the maximum colony count had been reached are used as calculating the yield values (Fig. 3b) At acetate-carbon
inoculum. The inoculum gives is about 50 to 500 CFU/mL concentrations above 100 µg C/L, the Nmax value of strain
in the pasteurized samples. The flasks are incubated at P17 remains below the level expected on the basis of
15 ± 1 ° C in the dark, without shaking. the values observed at lower concentrations (data not
Membrane filters have been used to remove the shown). This is probably because the cells are larger
indigenous bacteria from the samples. However, organic when grown at higher substrate concentrations. This
compounds can release from: (1) the filters, (2) the phenomenon has not been observed with strain NOX.
equipment used for membrane filtration, and (3) the flasks Yield values of a number of test strains are given in
in which the filtered water is collected. Furthermore, Table 4.
membrane filtration does not always result in sterile AOC concentrations are determined using a mixed
samples because of the presence of very small bacterial inoculum of strain P17 and strain NOX. The AOC
cells (ultramicrobes). Hence, application of membrane concentration is determined from the Nmax values of each
filters may give unreliable results. Also, autoclaving is of the strains, using their Y value for acetate (Fig. 4). Next
not appropriate because it causes a strong increase (three- to the AOC concentration, information also is obtained
to fourfold) in the AOC concentration, depending on the about the AOC composition because strain NOX can only
water type (unpublished results). grow on carboxylic acids (40).
In certain samples, detoxification of the water is Figure 5 shows the contributions of strain NOX and
required, for example, in the presence of a disinfectant strain P17 to the AOC concentration when grown simulta-
residual, or in the presence of copper. Growth measure- neously in pasteurized slow sand filtrate supplemented
ments with thiosulphate, used to neutralize a disinfectant with yeast extract. Growth of strain NOX was not
residual showed an AOC increase of about 5 µg C/L at a enhanced by the addition of yeast extract. The Y value
concentration of 10 mg/L. EDTA and NTA do not affect the of strain P17 for yeast extract is 0.35 µg AOC/µ g YE.
AOC concentration as determined with strains P17 and
NOX (unpublished results). The use of NTA is preferred
because this compound is more easily biodegradable in the
environment than EDTA. 106 5.0
NOX
Growth Measurements and Calibration Curves
4.0 P17
The growth curve is determined by periodic colony counts 105
Colony count (CFU/ml)

for which purpose water volumes of 0.05 mL are spread in

triplicate over the surface of agar medium plates, followed
AOC (µg C/I)
3.0
by incubation at 25 ° C for 2 (P17) to 3 (NOX) days. A broth
4
agar medium (Lab Lemco Agar, Oxoid) is used routinely, 10
but the R2 A medium 67 has the same recovery for strain 2.0
P17 and strain NOX (data not shown). Colonies of strain
P17 are larger than those of strain NOX and produce a
fluorescent pigment on R2 A medium. 103
1.0
The yield (Y) values of strain P17 and strain NOX for
acetate are needed to calculate the AOC concentration
from the maximum colony counts. Typical growth curves 102 0.0
of strain P17 in slow sand filtrate supplemented with 0 5 10 15 20 25 SSF
different acetate concentrations are presented in Figure 3. Time (days)
Nmax values were reached after five to seven days at acetate Figure 4. Growth curves of test strains P17 and NOX in slow
concentrations greater than or equal to ≥10 µg C/L, and sand filtrate (SSF) and the AOC value calculated from the Nmax
after about 10 to 14 days at an initial acetate-carbon values of the strains. In these samples the concentration of AOC
concentration of 5 µg C/L. Similar experiments have been compounds utilized by strain P17 was less than 1 µg C/L.

Table 4. Yield Values of a Number of Bacterial Strains Used in Growth Experi-

ments (28,40,42,43,48)

Y value (CFU/µg C)
Organism Acetate Oxalate Glucose Starch Lactate

P. fluorescens strain P17 4.1 × 106 No growth Nd∗ No growth Nd∗

Spirillum sp. strain NOX 1.18 × 107 2.9 × 106 No growth No growth Nd∗
Flavobacterium strain S12 No growth No growth 1.8 × 107 2.2 × 107 No growth
A. hydrophila strain M800 6.8 × 106 No growth 8.0 × 106 Nd∗ Nd
Escherichia coli strain 8872∗∗ Nd∗ Nd 4.5 × 106 Nd∗ 6 × 106

Note: ∗ nd, not determined; ∗∗ tested at 25° C, see Fig. 1.

 ASSIMILABLE ORGANIC CARBON (AOC) IN TREATED WATER: DETERMINATION AND SIGNIFICANCE 319

30 300

25
250
AOC(P17/NOX)
20
AOC (µg C/|)

200

AOC (µg C/I)

15
AOC(P17) 150
10

5 AOC(NOX) 100

0 50
0 10 20 30 40 50
Yeast extract (µg/I)
Figure 5. AOC concentrations at various concentrations of yeast 0
1(P17) 2(NOX) 1+2 P17/NOX
extract added to slow sand filtrate. AOC-P17 is the AOC
concentration derived from the Nmax values of strain P17 grown in 35
the presence of strain NOX; AOC-NOX is the AOC concentration
calculated from the Nmax values of strain NOX. AOC-P17/NOX is 30
the total AOC concentration.

25
Effect of Separate Growth of Test Strains
AOC (µg C/I)

20
Both strain P17 and strain NOX are able to utilize
carboxylic acids. In a mixed culture, there is competition 15
for these substrates. When grown separately, compounds
available to the two strains are used by both strains.
10
Figure 6 shows the AOC concentrations determined in
two different water types by the individual strains and by
5
growth as a mixture, respectively. In both water types,
the AOC concentration mainly consists of compounds
available to strain NOX (carboxylic acids). From these 0
1(P17) 2(NOX) 1+2 P17/NOX
results, it can be calculated that 45 to 66 % of the
Figure 6. Effect of separate growth of the test strains in two
compounds potentially available to strain P17 were water types. (a) water after ozonation; (b) ozonated water after
utilized by strain NOX when grown as a mixed culture. GAC filtration. 1(P17), strain P17 alone; 2(NOX) strain NOX
In addition, a small fraction of the compounds available alone; 1 + 2, AOC calculated from 1 and 2; P17/NOX, strains
to strain NOX is utilized by strain P17 when grown as a P17 and NOX grown as mixed culture. Error bars give standard
mixed culture. deviations.

Reproducibility and Repeatability of AOC Determinations

tool (55,60). A major adaptation is the use of small-capped
A large number of AOC determinations have been containers, in combination with a single measurement
conducted since the development of the method. In all in a container (in triplicate) after a defined incubation
cases, the test was done in duplicate sample flasks. Hence, period. Table 5 compares the conditions of the AOC test
information is available about the reproducibility of the as described in standard methods (6g) with the AOC test
test. Figure 7 shows the Relative Standard Deviations that was described earlier. A further adaptation to the
(RSD, %) of a total of 430 AOC tests, conducted within a procedure described in standard methods is the use of
one-year period, most of which with an AOC value below ATP for determining the growth of the test strains. With
25 µg C/L. The STD values were below 10% for 60% of the ATP analysis, the AOC test can be conducted within
all samples tested. In 10% of the samples STD values two to four days (55).
were greater than 35% (Fig. 7b). The repeatability of the
test is shown by the constant yield values of the test
strains for acetate and the constant AOC values obtained AOC CONCENTRATIONS IN WATER
for one particular slow sand filtrate over a period of
more than 15 years (with values <5 µg C/L) (results not AOC Concentrations in Raw and in Treated Water
shown).
The AOC test had been developed for assessing the
biostability of treated water. However, it is also used
Comparison of Described AOC Procedures
in raw water and in water in various treatment stages.
The AOC test as described earlier, has been adapted by a A problem observed with the use of the AOC test in
number of researchers to improve its use as a monitoring raw water is the presence of spore-forming bacteria
320 ASSIMILABLE ORGANIC CARBON (AOC) IN TREATED WATER: DETERMINATION AND SIGNIFICANCE

(a) 25 one to two hours) can be applied to eliminate this

problem.
Percentage of samples

20 Typical AOC concentrations in river (Meuse) water

ranged from about 50 µg C/L to 400 µg C/L, depending on
15 the season, with the highest concentrations in spring and
in late summer (Fig. 8a). About 70 to 90% of these AOC
10 concentrations was available to strain P17, indicating
that the substrates available to strain P17 (e.g., amino
5 acids, peptides, carbohydrates, etc.) were present in much
higher concentrations than the carboxylic acids. The
0 concentration of dissolved organic carbon (DOC) was 3.2
0 25 50 75 100 125 to 4.2 mg/L and the AOC concentrations were 1.5 to 10%
AOC-concentration (µg C/I) of the DOC concentration.
In a survey conducted in 1986 to 1987 (53), AOC con-
centrations observed in drinking water in the Netherlands
(b) 100 ranged from 1 to about 60 µg acetate-carbon equivalents
Cum. percentage of samples

90 per liter (Fig. 8b). The highest values were observed in

75
surface-water supplied with ozonation a treatment ste. In
all types of treated water the fraction available to strain
60 NOX was the largest proportion of the AOC concentration,
50 indicating that carboxylic acids were the predominating
, AOC < 10 µg C/I compounds. The AOC concentration utilized by strain
25
, all samples (430) P17 was less than 1 µg C/L in most samples of treated
water. The data presented in Figure 8b and many obser-
vations conducted since then clearly demonstrated that
0 the AOC concentrations in treated groundwater water in
0 25 50 75 100 125
the Netherlands usually are below 10 µg of C/L. Also, in
Relative standard deviation (%)
surface water supplies with slow sand filtration as the
final treatment step such low AOC concentrations were
Figure 7. (a) AOC concentrations as measured in 430 samples
and (b) the relative standard deviation (RSD) of the AOC
found.
concentrations (relative to the average value of duplicate AOC concentrations in treated water are only a
measurements). small fraction (<1.7%) of the DOC concentration (Fig. 9).
The lowest AOC/DOC ratios have been observed in
groundwater supplies, in which AOC concentrations
or other types of bacteria surviving the pasteurization usually were below 10 µg of C/L. Minimum values for
procedure. These organisms hamper the AOC test either the AOC/DOC quotient were all close to 0.1%.
by growing on the plates and/or by growing in the AOC values in treated water observed in the Nether-
water samples and thus consuming an unknown part lands are low in comparison to values reported in
of the substrates potentially available to the test strains. other countries. Several surveys conducted in the United
Pasteurization at a higher temperature (e.g., 70 ° C for States revealed that AOC concentrations in treated water

Table 5. Comparison of AOC Procedures

Test Characteristic AOC (Original Method) AOC in Standard Methods (69)

Containers Borosilicate glass Borosilicate glass vial (45 mL)

Erlenmeyer (1 L); glass with TFE-lined silicone septa
stoppered
Sample volume 600 mL 40 mL
Surface/volume (cm−1 ) 0.63 (equivalent to pipe 1.5 (equivalent to pipe diameter
diameter of 6 cm) of 2.7 cm)
Test strains P17 and NOX P17 and NOX, separate growth
Biomass parameter Colony count Colony count or ATP (55)
Sample treatment 30 minutes at 60o C 30 minutes at 70o C
Addition of chemicals none Thiosulphate (50 mg/L)
Incubation period ≥14 days 9 days
Calculation of AOC Nmax values in duplicate Average of colony counts in
flasks triplicate vials on days 7,8,
and 9
Result (units) µg of acetate-c µg of acetate-C equivalents/L
equivalents/L
 ASSIMILABLE ORGANIC CARBON (AOC) IN TREATED WATER: DETERMINATION AND SIGNIFICANCE 321

(a) 400 in a number of water types. Hence, AOC values in treated

350 NOX water in the United States are much higher than those
in treated water in the Netherlands. These differences in
300 P17
AOC levels may be due to differences in water treatment
AOC (µg C/I)

250 (less biological treatment and addition of chlorine). Also,

200 the AOC test as used in the United States is not identical
to the one described in this report (Table 5), but a few
150
comparisons between the test methods did not show large
100 differences between the results (72). Still, the separate
50 growth of the test strain will have an effect in certain
0 water types (cf. Fig. 6).
0 10 20 30 40 50
Week nr Effects of Water Treatment

60
From the differences between the AOC concentration
(b)
Strain NOX
in raw river water and the AOC concentration in
AOC (µg acetate-C eq/I)

50 treated water, it can be concluded that surface water

Strain P17 treatment as applied in the Netherlands can cause an
40 AOC reduction of more than 90%. A variety of processes
have different effects on the AOC concentration Fig. 10.
30
The AOC concentration is reduced by biological processes
20 in filter beds when these processes are applied properly
and also during soil passage physicochemical processes
10 such as coagulation/sedimentation and adsorption also
can reduce the AOC concentration of water. On the
0
1 5 10 15 20 other hand, oxidation processes clearly cause an increase
Treatment plant of the AOC concentration. The effect of ozonation is
related to the degradation of large molecules of natural
Figure 8. (a) AOC concentration in river water in a one year organic compounds, such as fulvic and humic acids, into
period. (b) AOC concentrations in treated water of 20 different low molecular compounds, as has been demonstrated in
treatment plants. Hatched bars represent treated surface water; studies, using advanced chemical analytical techniques.
open bars represent treated groundwater. Groundwater was AOC concentrations increase with increasing ozone
treated without chemical disinfection. Nrs. 15 and 17 are slow dosage (74,75,76).
sand filtrates. Adapted from D. Van der Kooij, J. Am. Water Works
The effect of biological filtration processes on the
Assoc. 84(2), 57–65 (1992).
AOC concentration is of major importance in achieving
a biologically stable drinking water. In the Netherlands
such processes are included in all water treatment plants,
60 either as rapid sand filtration, GAC filtration, slow
sand filtration, or soil passage of surface water (river
50 bank filtration, artificial dune recharge). Fig. 11 shows
AOC-P17/NOX (µg C/I)

40
150
30
125
AOC-P17/NOX (µg C/I)

20
100
10
75
0
0 1 2 3 4 5 6 7 8
50
DOC (mg/I)
Figure 9. AOC concentrations in treated water as a function
25
of the DOC concentration. Open circles: groundwater supplies;
closed circles: surface water supplies. Line shows Lowest
0
AOC/DOC ratios. Data (adapted) from D. Van der Kooij, J. Am. RW CS O3 DMF GACF PCl2
Water Works Assoc. 84(2), 57–65 (1992).
Figure 10. The effects of water treatment on the AOC concen-
tration. RW, raw water river Meuse water after storage in open
reservoirs; CS, coagulation/sedimentation; O3 , ozonation; DMF,
ranged from about 20 to 200 or more than 300, respec- dual media filtration; GACF, granular activated carbon filtration;
tively (70,71). Median AOC values observed in these stud- pCl2, postchlorination. The water is distributed after postchlori-
ies were about 100 µg C/L. In one of these surveys, the nation. Black bar represents the AOC fraction utilized by strain
fraction of AOC available to strain P17 exceeded 50 µg C/L P17.
322 ASSIMILABLE ORGANIC CARBON (AOC) IN TREATED WATER: DETERMINATION AND SIGNIFICANCE

that the AOC reduction decreases with decreasing AOC

concentration in the influent. 60 , System A
From the data shown in this figure, it can be , System B
50

AOC (µg ac-C eq/I)

calculated that the AOC reduction in this type of , System c
process is negligible at AOC concentrations below 6 µg 40
of C/L. Similar observations have been reported by other
investigators (77). Figure 11 also shows that the presence 30
of a chlorine residual in the influent of rapid sand filters
decreased the AOC removal, most probably by hampering 20
the biological processes.
10
Membrane filtration processes have a huge potential
in water treatment because of the variety of available 0
membrane types, and the capacity of removing undesirable 0 10 20 30 40 50
particles (including bacteria) and dissolved compounds Distance from treatment plant (km)
without side-effects on water quality. Ultrafiltration cause Figure 12. The effect of distribution on the AOC concentration
an AOC removal up to 84% (from 25 to 3.8 µg C/L) in a in three different distribution systems. In all cases, ozonation
surface water treatment, probably by removing particulate was used as a treatment step. From: D. Van der Kooij, J. Appl.
compounds of biological origin (78). Reverse osmosis is Microbiol. Symp. Suppl. 85: 39S–44S (1999).
capable of reducing the AOC concentration to a level below
1 µg C/L. In this water, inorganic nutrients were added to
40
conduct the AOC test (unpublished result 5).
Decrease of AOC (µg C/I) δAOC = 0.6 AOCP − 4.1 (r=0.94)
Effects of Distribution 30

Observations on water sampled from a number of distribu-

20
tion systems have revealed that AOC concentrations may
decline rapidly in the distribution system, depending on
the initial AOC concentration (Fig. 12). Most likely, biofilm 10
processes play an important role in AOC uptake (53,79).
Figures 12 and 13 demonstrate that AOC reduction is very 0
limited at values close to and below 10 µg C/L.
−10
Effect of Storage of Water Samples on AOC and HPC Values 0 10 20 30 40 50 60
AOC (µg C/I) at treatment plant
Water samples collected in thoroughly cleaned Erlenmeyer
flasks were incubated without pasteurization to obtain Figure 13. Maximum AOC decrease observed in distribution
systems as a function of the AOC concentration of treated water.
information about the effect of the multiplication of
Open circles: surface water supplies; closed circles: groundwater
the indigenous population on the AOC concentration. supplies. Adapted from D. Van der Kooij, J. Am. Water Works
Figure 14 shows that the AOC concentration of river Assoc. 84(2), 57–65 (1992).
water (Fig. 9) collected from a storage reservoir initially
followed an exponential decrease. After three days at an
AOC concentration of about 10 µg C/L the rate of decrease declined sharply and followed a linear function with time
in the test period up to four weeks. In ozonated water,
the AOC concentration rapidly declined to a value of
150 15 µg of C/L. Thereafter the decline became linear with
dAOC=0.84 AOC − 5.17 (r2=0.99) time (Fig. 14(b)). In the stored river water, the fraction
125 of AOC utilized by strain P17 constituted 72% of the
initial AOC concentration. In ozonated water the fraction
dAOC (µg C/I)

100 available to strain P17 initially was less than 10% of the
total AOC concentration. Obviously, biomass compounds
75 (amino acids, peptides, etc.) predominated in stored river
water, whereas carboxylic acids were the major AOC
50
, RSF/DMF fraction in ozonated water.
, + Chlorine The following equations can be used to describe the
25
, GACF observed AOC decrease:
Phase 1 (up to five days); exponential decrease with
0
0 25 50 75 100 125 150 175 residual:
AOC(P17) (µg C/I) AOCT = AOC0 x e−kT + AOCR , (3)
Figure 11. AOC reduction (dAOC) as achieved with biologi-
cal filtration. RSF, rapid sand filtration; DMF, dual media Phase 2 (after about 5 days); linear decrease:
(anthracite/sand filtration); GACF, granular activated carbon fil-
tration. Adapted from Van der Kooij, 1984 (73). AOCT = AOC5 − Ux(T − 5) (4)
 ASSIMILABLE ORGANIC CARBON (AOC) IN TREATED WATER: DETERMINATION AND SIGNIFICANCE 323

(a) 35 This increase can be determined with several techniques,

including the heterotrophic plate count (HPC), the total
30
direct (microscopic) count (TDC), or the concentration
25 of adenosinetriphosphate (ATP). These techniques all
AOC (µg C/I)

have their specific advantages and limitations. The main

20
limitations of these parameters include: (1) the HPC value
15 of a mixed population of aquatic bacteria is always a
(small) fraction of the total active population, and requires
10
a long incubation period; (2) the TDC values include active
5 and inactive cells, which may also differ greatly in size;
(3) the ATP test is fast, but the ATP concentration declines
0
0 5 10 15 20 25 30 rapidly once the maximum level has been attained. For
a number of samples, maximum HPC values have been
(b) 125 determined in relation to the initial AOC concentration.
The results presented in Figure 15 clearly show that there
is a good correlation between the AOC concentration
100
and the maximum HPC values, with an average yield of
7.37 × 106 CFU/µg C. Obviously, the compounds serving as
AOC (µg C/I)

75 sources of carbon and energy for the indigenous population

are those compounds used by the AOC test strains P17
50 and NOX. However, differences between the bacterial
communities in the various water types result in different
25 yield values and differences in the fraction contributing to
the HPC value influence the Nmax values.
0
0 5 10 15 20 25 30
BIOLOGICAL STABILITY
Days

Comparison of AOC and BDOC

Figure 14. AOC decrease in water samples incubated at 15 ° C
in thoroughly cleaned Erlenmeyer flasks. (a) river water from an The AOC and BDOC methods are used to assess the
open storage reservoir; (b) water after ozonation. Symbols: , biological (in)stability of treated water. Table 2 gives an
•
AOC P17/NOX; , Fraction of AOC available to strain P17. Error indication of the difference between these two approaches.
bars give standard deviation of the AOC value (duplicate flasks). AOC values are a small fraction of the DOC concentration
of river water and even lower (0.1 to 1%) in treated water.
BDOC values may constitute a significant proportion (19 to
In these equations AOCT is the AOC concentration at time 54%) of the DOC concentration (3.5 to 13.3 mg/L) of river
T, AOC0 , the AOC concentration at the start (zero time), water (80). Surveys revealed that BDOC concentrations
AOC5 the AOC concentration at day five, and AOCR , is of drinking water from 79 utilities in the United States
the apparent residual AOC concentration, respectively. K ranged from 0.01 to 2.4 mg/L and were 0.4 to 52.8 % of the
is the decay constant (d−1 ) in the exponential phase, and DOC concentration (70). In another study, BDOC values
U is the uptake rate (µg . . . l−1 . . . d−1 ) in the linear phase. for treated water of 31 plants averaged by site ranged
The constants of the equations for the two water types are from 0.03 to 1 mg/L and constituted 5 to 21% of the DOC
given in Table 6. concentration (71). In both studies significant correlations
Simultaneously with the AOC decrease, the numbers of were found between concentrations of AOC and BDOC,
bacteria increase in the water samples during incubation. respectively. However, the correlation coefficients were

Table 6. Kinetics of the AOC Decrease in Samples (600 mL) of Two Water Types
Stored at 15 ° C in Thoroughly Cleaned Erlenmeyer Flasks

River Water from Open

Parameter∗ Storage Reservoir Water After Ozonation

DOC (mg/L) 3.4 2.9

AOC0 (at time zero) 29.1 ± 6.1 105 ± 11
(µg C/L)
AOCR (µg C/L) 9.6 ± 0.7 13.2 ± 0.8
K (day−1 ) 0.98 ± 0.11 0.90 ± 0.04 (r2 = 0.99)
(r2 = 0.99)
Uptake coefficient 0.2 ± 0.01 0.27 ± 0.02 (r2 = 0.99)
(µg C/L.day) (r2 = 0.99)

Note: ∗ DOC, dissolved organic carbon; AOC0 = AOC concentration at time zero; AOCR = residual
AOC concentration; K = exponential decay rate (d−1 ).
324 ASSIMILABLE ORGANIC CARBON (AOC) IN TREATED WATER: DETERMINATION AND SIGNIFICANCE

1. Samples collected from drinking water distribution

106
systems revealed that AOC concentrations did not
decrease at values close to or below 10 µg of C/L
(cf. Figs. 12 and 13). Such values are observed in
Nmaxaut (CFU/ml)

slow sand filtrate and in groundwater supplies in

the Netherlands, where rapid sand filtration is the
105 final treatment step;
2. AOC reduction in biological filtration (e.g., rapid
sand filtration) is very limited at values below 10 µg
r2 = 0.899
C/l. (Fig. 11);
3. AOC concentrations in samples of stored water
decline rapidly to a level of about 10 µg C/L (Fig. 14);
104
1 10 102 4. A statistically highly significant relationship was
AOC-P17/NOX (µg C/I) observed between the AOC concentration and the
Figure 15. The relationship between the AOC concentration and heterotrophic plate counts as determined on a
the maximum colony count (Nmax ) of the autochthonous bacterial nutrient poor medium (diluted-broth agar). At
community in water samples stored at 15 ° C in thoroughly cleaned AOC values below 10 µg of C/L, the increase of
Erlenmeyer flasks. HPC values were determined using diluted these colony counts during distribution remained
broth agar medium (spread plate method), incubated at 25 ° C for limited (53).
10 days. 5. In a number of water types, the AOC/DOC ratio was
as low as 1 µg of C/mg C. Values below this level
have not been observed, suggesting that this level is
relatively low, as a result of the wide range of AOC to indicative for organic compounds with a high degree
BDOC ratios. On average, the AOC concentration was of biostability. With DOC values below 10 ppm, this
about 20 to 30% of the BDOC concentrations in treated also gives AOC values below 10 µg of C/L.
water.
The main reasons for the difference between AOC and For limiting regrowth of coliforms in chlorinated
BDOC concentrations include: supplies AOC concentration of 50 to 100 µgC/l have been
derived from extended studies, which indicated that many
factors affected regrowth of these organisms (83).
• the use of an adapted microbial community in the
BDOC test; Biofilm Formation Rate
• the use of a large amount of biomass (bacteria
Practical experience in the Netherlands show that the
attached to sand) in the BDOC test.
90-percentile values of the heterotrophic plate counts
(22° C, 3 days incubation) remain below 100 CFU/mL in
Also, the basis of the methods differ. The AOC
water types with AOC levels below 10 µg of C/L. In
concentration gives information about the production of groundwater supply this level of AOC is achieved with
biomass, which is the parameter of concern. The difference traditional treatment processes, namely, aeration followed
between the Y values for acetate and oxalate (cf Table 4) by biological (rapid sand) filtration. In most supplies in
implies that the relationship between the amount of the Netherlands the level of 10 µg of C/L is attained,
utilized organic carbon (BDOC) and biomass formation but defining biological stability was found to be more
may differ between water types. In fact, using the Y complicated. Even at AOC values below 10 µg C/L and at
values for oxalate in the AOC test gives much higher low heterotrophic plate counts, aeromonads were found
AOC values, some of which being close to the BDOC to increase in numbers (expressed as CFU/100 mL) (5).
value (70). Observations indicated that these organisms multiplied
From various studies a BDOC concentration of 0.1 in biofilms and in sediment 28. Therefore, a biofilm
to 0.2 mg/L has been derived as a reference value for monitor had been developed for determining the Biofilm
biological stability (81,82). Formation Rate (BFR) values of treated water (25,55).
The data about growth kinetics (Table 1), suggest that Biofilm concentrations are determined by assessing the
the concentrations of easily biodegradable compounds concentration of adenosinetriphosphate (ATP) with the
in treated water should be less than 1 µg of C/L luciferin-luciferase test. In the Netherlands, the BFR
to prevent multiplication. However, complete growth values typically range from less than 1 pg ATP cm−2 d−1
inhibition cannot be achieved and is also not required. to a value greater that 100 pg ATP cm−2 d−1 (5). Values
The goal is to limit microbial growth to such an extent below 1 pg ATP cm−2 d−1 have been observed in slow
that water quality deterioration does not occur. Defining sand filtrate and in drinking water prepared from aerobic
an acceptable level of growth is complicated, and may be groundwater. These water types, with AOC values clearly
different for different systems and seasons. below 10 µg Cl−1 represent drinking water with the highest
In the Netherlands an AOC value concentration of 10 µg degree of biostability. Higher BFR values have been found
of C/L has been derived as the reference value below which in drinking water prepared from anaerobic groundwater,
treated water has a very limited regrowth potential. This but also these water types had AOC values below 10 µg
value is based on the following observations: C/L.
 ASSIMILABLE ORGANIC CARBON (AOC) IN TREATED WATER: DETERMINATION AND SIGNIFICANCE 325

Dosage of acetate at a concentration of 10 µg of C/L in treated water demonstrate that easily available
to a biofilm monitor gave a BFR value of about 360 pg compounds have been removed.
ATP/cm2 .d. This and other observations showed that a • AOC concentrations in treated water in the Nether-
concentration of 1 µg of acetate-C/L causes a BFR value of lands are below 10 µg C/l in the groundwater supplies
35 pg ATP/cm−2 . . .d−1 (84). and in most supplies as the result of the application
A significant relationship has been observed between of multiple biological processes in water treatment.
the level of regrowth of aeromonads in groundwater The much higher AOC values (median: 100 µg C/l)
supplies and the BFR value of the water leaving the as reported for treated water in the USA may be
treatment facility. At a BFR value of 10 pg ATP cm−2 d−1 , ascribed to the absence of biological filtration and
the risk of exceeding the guideline value for Aeromonas
the effect of the disinfectant on the AOC concentra-
(90-percentile value of 200 CFU 100 mL−1 ) is 20% (5).
tion;
Based on these observations the definition of biostabil-
ity has been extended and a two dimensional approach • A reference AOC value of 10 µg C/l has been
is used. Hence, AOC concentrations below 10 µg of C/L in derived from effects of biological filtration on the
combination with BFR values below 10 pg ATP/cm2 day AOC concentration, the AOC decline in distribution
represent treated water with a high degree of biostability. systems and the relationship between AOC and HPC
Only in a few water types (slow sand filtrate and aerobic values, respectively in unchlorinated supplies. AOC
groundwater) lower values can be achieved with biological values below 100 or 50 µg C/l have been suggested for
processes. limiting coliform regrowth in chlorinated supplies in
the USA.
Materials in contact with treated water • Even at AOC values below 10 µg C/l, certain micro-
organisms may multiply in the distribution system.
Materials in contact with drinking water can also affect
Assessment of the Biofilm Formation Rate in
bio(in)stability by releasing biodegradable compounds
combination with AOC values gives more complete
into treated water. For this reason methods have been
information about the biological stability of treated
developed to test the growth promoting properties of
materials. In the United Kingdom the MDOD (mean wate;
dissolved oxygen difference) test is used (85), and in • Materials of pipes and reservoirs may release
Germany a test based on measuring slime production is biodegradable compounds into treated water. Assess-
used (86,87). In the Netherlands the Biofilm Formation ment of the growth-promoting properties contributes
Potential (BFP) test is applied, which determines the to the selection of appropriate materials.
biofilm concentration (pg ATP/cm2 ) on the material as
a function of time in a batch test in slow sand filtrate (88).
Typical BFP values are below 10 pg ATP/cm2 for glass
Acknowledgments
and 20 to 50 pg ATP/cm2 for PVC, whereas values The AOC test has been developed within the framework of
between 500 to 3,000 have been observed for polyethylene the Joint Research Program of the Water Supply Companies
materials. Such BFP values can directly be compared in the Netherlands. Many persons, including Ton Visser, Peter
with biofilm concentration values as observed on the J. Oranje, Wim A. M. Hijnen, and Harm R. Veenendaal, have
surface of pipe segments collected from the distribution been involved in the investigations aiming at developing and
system. In addition, BFR values as observed in the improving the AOC test.
biofilm monitor also give information about the biofilm
concentration, which can be attained on the pipe wall
in contact with the water tested. In this way a simple BIBLIOGRAPHY
framework is obtained, which enables the assessment
and evaluation of the biostability of both water and 1. R. R. Trussel, J. Am. Water Works Assoc. 91, 46–54 (1999).
materials with a consistent approach, the Unified Biofilm 2. M. W. LeChevallier, J. Am. Water Works Assoc. 91(1), 86–90
Approach. (1999).
3. B. Hambsch, J. Am. Water Works Assoc. 91(1), 81–86 (1999).
CONCLUSION 4. O. Hydes, J. Am. Water Works Assoc. 91(1), 70–74 (1999).
5. D. Van der Kooij, J. H. M. van Lieverloo, J. A. Schellart, and
• The AOC method described above has proven to P. Hiemstra, J. Am. Water Works Assoc. 91(1), 55–64 (1999).
give highly reproducible results. Contamination of 6. R. J. Bull and F. C. Kopfler, Health Effects of Disinfectants
the water with biodegradable compounds during and Disinfection By-Products, AWWA Research Foundation,
sampling, transportation, handling and analysis is Denver (Co.), 1991.
a main concern, with glassware and air as potential 7. R. H. Burttschell, A. A. Rosen, F. M. Middleton, and M. B.
sources. The described procedures aim at preventing Ettinger, J. Am. Water Works Assoc. 51, 205–214 (1959).
such contamination. 8. P. E. Bryan, L. N. Kuzminski, F. M. Sawyer, and T. H. Feng,
• The two test strains are not capable of utlizing all J. Am. Water Works Assoc. 65, 363–368 (1973).
biodegradable compounds present in a water type. 9. J. Rook, J. Soc., Water Treatment Exam. 23, 234–243 (1974).
The compounds, which are utilized by the test strains 10. M. W. LeChevallier, J. Am. Water Works Assoc. 82(11), 74–86
are ubiquitous in water and low AOC concentrations (1990).
 ASTROVIRUSES 327

The Royal Society of Chemistry, Cambridge, U.K., 1997, ASTROVIRUSES

pp. 195–207.
67. D. J. Reasoner and E. E. Geldreich, Appl. Environ. Microbiol. SUZANNE M. MATSUI
49(1), 1–7 (1985). Stanford University School of
68. L. A. Kaplan, T. L. Bott, and D. J. Reasoner, Appl. Environ. Medicine
Microbiol. 59(5), 1532–1539 (1993). Stanford, California
69. Anonymous, in M. A. H. Franson, ed., Standard Methods for Veterans Affairs Palo Alto
the Examination of Water and Waste Water, American Public Health Care System
Health Association, Washington 1995, pp. 9-39–9-43. Palo Alto, California
70. L. A. Kaplan, D. J. Reasoner, and E. W. Rice, J. Am. Water JOHN E. HERRMANN
Works Assoc. 86(2)121–132 (1994). University of Massachusetts
71. C. J. Volk and M. W. LeChevallier, J. Am. Water Works Assoc. Medical School
92(5), 64–76 (2000). Worcester, Massachusetts
72. W. M. Lechevallier, Factors limiting microbial regrowth
in distribution systems, American Water Works Service
Astroviruses are 28-nm diameter, icosahedral viruses
Company, Inc. Voorhees, NY 1996.
that have a characteristic five- or six-pointed star-like
73. D. Van der Kooij, The Growth of Bacteria on Organic
surface structure when viewed by electron microscopy
Compounds in Drinking Water. Dissertation, Kiwa N.V.,
(EM) (Greek. astron, star; Fig. 1). Madeley and Cosgrove
Rijswijk, The Netherlands, 1984.
coined the term astrovirus in 1975 to describe 28-
74. D. Van der Kooij, The Effect of Treatment on Assimilable
nm diameter round viral particles they observed by
Organic Carbon in Drinking Water, in P. M. Huck and P. Toft,
eds., Treatment of Drinking Water for Organic Contaminants,
direct EM in the stools of infants with diarrhea (1,2).
Pergamon Press, New York, 1987, pp. 317–328. Along with the Picornaviridae and the Caliciviridae, the
75. D. Van der Kooij, Assimilable Organic Carbon (AOC) in
Astroviridae comprise a third family of nonenveloped
Drinking Water, in G. A. McFeters, eds., Drinking Water viruses whose genome is composed of plus-sense, single-
Microbiology; Progress and Recent Developments, Springer- stranded RNA. In addition to humans, astroviruses have
Verlag, New York, 1990, pp. 57–87. been isolated from numerous animal species, including
76. D. van der Kooij, W. A. M. Hijnen and J. C. Kruithof, Ozone lambs (3), calves (4), deer (5), piglets (6,7), kittens (8),
Sci. Eng. 11, 297–311 (1989). mice (9), dogs (10,11), and turkey poults (12,13).
77. S. Zhang and P. M. Huck, Wat. Res. 30(5), 1195–1207 (1996).
77. S. Zhang and P. M. Huck, Water Res. 30(5), 1195–1207
(1996).
78. M. M. Nederlof, J. C. Kruithof, J. S. Taylor, D. van der Kooij
and J. C. Schippers, Proceedings Conference on Membranes
in drinking water and industrial water production, Paris,
France, 3-October Desalination Publications, L’ Aquila, Italy,
2000, 1, pp. 453–465.
79. D. Van der Kooij, J. Appl. Microbiol. Symp. Ser. 85, 39S–44S
(1999).
80. J. C. Joret, Y. Levy and C. Volk, Water Sci. Technol. 24(2),
95–101 (1991).
81. P. Servais, P. Laurent and G. Randon, Impact of biodegrad-
able dissolved organic carbon (BDOC) on bacterial dynamics
in distribution systems, Proceedings American Water Works
Association Water Quality Technology Conference, American
Water Works Association, Denver Co. 1993, pp. 963–980.
82. P. Servais, P. Laurent, G. Billen, and D. Gatel, Revue des
Sciences de l’Eau. 8, 427–462 (1995).
83. M. W. LeChevallier, N. J. Welch and D. B. Smith, Appl.
Environ. Microbiol. 62(7), 2201–2211 (1996).
84. D. Van der Kooij, H. S. Vrouwenvelder, and H. R. Veenen-
daal, Water Sci. Technol. 32, 61–65 (1995).
85. J. S. Colbourne and D. A. Brown, J. Appl. Bacteriol. 47(2),
223–231 (1979).
86. DVGW, Vermehrung von Mikroorganismen auf Materialien
fur den Trinkwasserbereich Prüfung und Bewertung. Tech-
nische Regeln, Arbeitsblatt W270, 1990.
87. D. Schoenen and H. F. Schöler, Trinkwasser und Werkstoffe,
Gustav Fischer Verlag, Stuttgart, New York, 1983.
88. D. van der Kooij and H. R. Veenendaal, Water Sci. Technol: Figure 1. Electron micrograph of human astrovirus in a fecal
Water Supply, 1(3), 39–45 (2001). specimen. Courtesy of W. D. Cubitt.
328 ASTROVIRUSES

CLASSIFICATION at least three smaller proteins. In cells infected with

human astrovirus serotype 2, grown in trypsin-containing
Structure media, Sanchez-Fauquier and coworkers (27) used a
neutralizing monoclonal antibody (PL-2) to determine that
The average diameter given for astroviruses, 28 nm, is
an approximately 86-kDa protein (P86) is the precursor
based on the measurement of approximately 1,000 neg-
to VP26, a major component of viral particles. VP29, a
atively stained (phosphotungstic acid) astroviruses in
stools (14). The diameter of astrovirus particles may vary, minor component of virions and the product of alternative
depending on the source of virus and the method of prepa- processing of P86, and intermediate proteins ranging from
ration for EM. For example, bovine astrovirus serotype 2 P74 to P35 were also identified in infected cells by PL-2.
(US2) propagated in primary NBK cells had an average Although these results suggest processing of P86 to P26
diameter of 34 nm (range 30 to 37 nm) after fixation in and/or P29 occurs within infected cells, the presence of
glutaraldehyde/1% osmium tetroxide and staining with trypsin in the media impairs optimal interpretation of
uranyl acetate (15). Purified preparations of cell culture- the data.
adapted human astrovirus serotype 1 particles evaluated Using partially purified preparations of human astro-
by electron cryomicroscopy and image analysis demon- virus reference strains (serotypes 1 to 7) grown in cell
strated a rippled solid capsid shell (330 Å in diameter) culture, astrovirus types 1 to 4 consisted of three pro-
with 30 dimeric spikes extending 50 Å from the sur- teins, P1 (∼33 kDa), P2 (∼33 kDa), and P3 (25 to 28 kDa,
face (16). Most of the data regarding the buoyant density depending on serotype), that were immunoprecipitated
of astrovirus particles have been collected using cesium with homologous rabbit antisera (28). For reference types
chloride density gradients. However, separation of fecal 5 to 7, three infection-specific proteins were observed in
astrovirus in this medium and further concentration by radiolabeled samples, but only P2 and P3 reacted with
pelleting disrupts virus morphology (17). Highly purified type-specific rabbit serum (28). In a relatively trypsin-free
viral particles are needed for evaluation by electron cry- system, Bass and Qiu (29) found evidence for intracellular
omicroscopy. The inclusion of divalent cations (Ca++ and processing of the full-length ORF 2 product to a 79-kDa
Mg++ ) has found to stabilize viral particles undergoing moiety that was incorporated into viral particles. Infec-
isopycnic centrifugation in cesium chloride (16). A gra- tivity was significantly enhanced by treatment of these
dient of potassium tartrate progressively diluted with particles with trypsin, which also resulted in cleavage of
glycerol yields a positive-density, negative-viscosity gradi- the 79-kDa protein to 33-, 29-, and 26-kDa subunits.
ent in which virus morphology is maintained and from
which viral particles may be pelleted without delete- Genetics
rious effects (17). Peak density for astroviruses in this
medium was 1.32 g/mL (17). Buoyant densities of human The genome of astroviruses is a plus-sense, single-
astroviruses in cesium chloride have ranged from 1.35 to stranded RNA of approximately 6,800 nucleotides (nt)
1.40 g/mL (18–23). in length, excluding the poly(A) tail at the 3
-end. RNA
With cell culture-adapted astroviruses, four structural extracted from virions is infectious (30). Infectious RNA
proteins (36.5, 34, 33, and 32 kDa) were identified in also can be transcribed from a full-length cDNA clone
astrovirus serotype 4 that was propagated in cell culture of human astrovirus serotype 1 [AVIC (30)]. BHK cells,
in the presence of trypsin (20). It was postulated that previously thought not infectable with astrovirus (31),
the faint 36.5-kDa band might represent a precursor have been shown to support astrovirus replication and
protein that undergoes subsequent processing, in a maturation when transfected with RNA extracted from
manner similar to the VP0 protein of enterovirus (24). virions or transcribed from AVIC. Astrovirus produced in
Two additional proteins with 24 and 5.2 kDa were found these BHK cells have been shown to infect monolayers of
by Kurtz (25) in cells infected with astrovirus serotype 1. Caco-2 cells in trypsin-containing media.
In CaCo-2 cells infected with astrovirus serotype 1, three Complete nucleotide sequences of four strains of human
astrovirus-specific polypeptides (33.5, 31.5, and 24 kDa) astrovirus are currently available: (1) the serotype 1
were identified by reactivity with astrovirus serotype 1- Oxford reference strain propagated in LLCMK2 cells (32),
specific antiserum (23). In some preparations, a 27-kDa (2) a serotype 1 Newcastle strain recently isolated and
polypeptide was also noted, and it was hypothesized that maintained in CaCo-2 cells (33–35), (3) the serotype 2
this may be derived from one of the larger proteins by Oxford reference strain propagated in LLCMK2 cells (36),
proteolytic digestion. In studies of human serotype 5 (4) a serotype 3 strain (GenBank accession AF141381),
isolated from volunteer fecal specimens, a single 30-kDa and (5) a serotype 8 strain (37).
structural protein was immunoprecipitated (22). During infection of susceptible cells, it has been
In astrovirus-infected LLCMK2 cells in the absence noted that both the full genomic (6.8 kb) and a subge-
of trypsin, a single approximately 90-kDa protein was nomic (2.4 kb) RNA are produced (21,26). The viral
immunoprecipitated from infected cell lysates with genome comprises three open reading frames (ORFs). The
hyperimmune antiserum to purified human astrovirus two ORFs at the 5
-end of the genome, designated ORF 1a
serotype 2 (26). Pretreatment of the lysates with trypsin and ORF 1b by the International Committee on the Taxon-
resulted in immunoprecipitation of three proteins, a omy of Viruses [ICTV (38)], encode nonstructural proteins.
predominant 29-kDa protein as well as a 31- and a 20-kDa The third ORF, designated ORF 2, encodes a structural
protein. These results suggested that the approximately protein and is found at the 3
-end of the genome (32,34,38).
90-kDa protein may be proteolytically cleaved to yield This ORF is common to both the genomic and subgenomic
 ASTROVIRUSES 329

RNA. Sequences of the capsid-encoding ORF 2 are avail- amino acid residue. This substitution is seen in several
able in GenBank for human serotypes 1 through 8 and for viruses, sobemoviruses, luteoviruses, and arteriviruses,
feline, porcine, and turkey astroviruses. but the putative astrovirus protease bears stronger overall
A schematic representation of the astrovirus serotype resemblance to the calicivirus protease.
1 Oxford reference strain’s genome is shown in Figure 2. A bipartite nuclear localization signal is encoded
A 5
-untranslated region of 85 nt precedes ORF 1a, which downstream of the viral protease motif in ORF-1a (35,36).
is 2,763 nt in length. The length of this ORF can vary Recent studies of human astrovirus serotype 1 (Newcastle)
depending on the method by which the virus was initially have indicated that this signal is functional and directs
isolated. For example, ORF 1a of the Newcastle serotype ORF-1a-encoded proteins to the nucleus (41). Of note,
1 strain, isolated in Caco-2 is 45 nt longer than the Oxford a prior study of cell culture-adapted bovine astrovirus
serotype 1 strain, isolated in primary HEK cells (39). detected viral proteins in the nucleus of infected cells by
A 73-nt overlap exists between ORF 1a and ORF 1b immunofluorescence (15). Further studies are required to
and contains the sequences critical for (−1) ribosomal determine the role and importance of nuclear localization
frameshifting and translation of ORF 1b. There is an 8- in the astrovirus life cycle.
nt overlap between ORF 1b and ORF 2. ORF 2 varies Analysis of ORF 1a sequences of reference serotypes
in length among human astroviruses from 2,316 nt for 1 to 7 indicates two distinct genogroups that correlate
a serotype 4 strain to 2,391 nt in the Oxford serotype 2 with serotype (42). Types 1 to 5 (42) and 8 (43) are found
strain (36), but for serotype 1 (Oxford) is 2,364 nt. An 80- in one genogroup (designated A), whereas types 6 and
nt 3
-untranslated region is found between ORF 2 and the 7 fall into a separate genogroup (B). These phylogenetic
poly(A) tail at the extreme 3
-end of the genome. The length associations for ORF 1a sequence are different from that
of the astrovirus 3
-untranslated region and its predicted observed for ORFs 1b or 2 in which nearly equidistant
secondary structure are similar to those of certain clustering of serotypes is found with unrooted phylogenetic
picornaviruses (33), but there is minimal sequence identity trees. Whether these differences are attributable to
between these two viral groups. In a separate study, recombination remains to be studied.
the terminal 19 nt of ORF 2 and adjacent 3
-untranslated ORF 1b contains a motif consistent with an RNA-
region were found to be highly conserved among all eight dependent RNA polymerase. This motif is highly conserved
human astrovirus serotypes (40). Of note, similarities among astrovirus serotypes 1 to 8 (32,37,42). The astro-
in sequence and folding of the 3
-untranslated region virus polymerase motifs are most similar to those of plant
of human, sheep, pig, and turkey astroviruses, avian viruses, bymovirus, and potyvirus, and can be aligned
infectious bronchitis virus (a coronavirus) and equine with other polymerases of Koonin’s Supergroup I (44) that
rhinovirus serotype 2 (a picornavirus) were observed (36) includes picornaviruses, caliciviruses, and certain plant
and it has been suggested that these features may have viruses (36), although similarities between astrovirus and
arisen from recombination events. plant viruses in this region do not necessarily imply a
The polypeptide sequences of ORFs 1a and 1b were the common origin (36).
first indication that nonstructural proteins are encoded by Regions that may encode an RNA helicase, methyl-
these ORFs. The ORF 1a-encoded protein contains a viral transferase, or papain-like protease have not been identi-
serine protease motif that has features consistent with fied in the astrovirus sequences that are currently avail-
chymotrypsin-like proteases of other plus-stranded RNA able. It is unusual for a plus-strand RNA virus with a
viruses, including RHDV and feline calicivirus (36). An genome length of greater than 6,000 nt to lack a helicase
important difference found when comparing the astrovirus domain (36,45). Astrovirus may encode a VPg, as sug-
protease motif with that of the caliciviruses is the gested by the absence of a methyltransferase-encoding
substitution of a serine for a cysteine at the third catalytic region and the similarity of its RNA-dependent RNA

5′ 3′ AA
Genome
1 6797
Subgenomic RNA
AA
4314 6797
Figure 2. Genome organization of hu-
man astrovirus serotype 2 (36). Genomic
ORF 1b and subgenomic RNA are depicted in the
upper half of the figure. Below this, the
2773 Pol 4329 three ORFs with predicted locations for
transmembrane helices (MB), protease
ORF 1a RFS ORF 2
(Pro), nuclear localization signal (NLS),
ribosomal frameshift structure (RFS),
83 MB Pro NLS 2842 4325 6712 and RNA-dependent RNA polymerase
(Pol) are shown schematically. Nonstruc-
tural proteins are likely encoded in ORFs
1a and 1b. At least one structural protein
0 1,000 2,000 3,000 4,000 5,000 6,000
that is most likely a precursor to smaller
Nucleotides capsid proteins is encoded by ORF 2.
330 ASTROVIRUSES

polymerase with primarily VPg-containing viruses. A con- type species is the human astrovirus. It has recently
vincing VPg domain has not been identified, but it may be been proposed by the Astrovirus Study Group of the
located between the putative protease motif and upstream International Committee on Taxonomy of Viruses that
transmembrane alpha-helices, where a serine at aa posi- the Family Astroviridae be subdivided into two genera,
tion 420 that could link the VPg to genomic RNA is found. Mammoastrovirus and Aviastrovirus, on the basis of
ORF 1a also encodes an immunoreactive epitope (aa 757 phylogenetic clustering of capsid sequence, host of origin,
to 899) identified by antisera produced to purified viral and target organs. Mammoastrovirus would include
particles, but its significance is not known. astroviruses that infect mammals and primarily cause
The greatest sequence variability is found in ORF 2, gastroenteritis. Aviastrovirus would include astroviruses
which encodes a viral structural protein. A high degree of that infect avian species and may cause intestinal as well
conservation of the N-terminal 415 (or so) amino acids is as extraintestinal illness.
observed among the human serotypes (46). In this region, Astroviruses are resistant to many common chemical
the amino acid sequence for the feline and porcine strains agents, including chloroform, a variety of detergents
share certain conserved stretches of amino acid residues (nonionic, anionic, and Zwitterionic) lipid solvents, and
with the human strains and can be aligned with minimal acid conditions (pH 3) (20). Human astrovirus retains
gaps, whereas the avian strains cannot be aligned easily activity after 5, but not 10, minutes at 60 ° C. Viral
with the human and other mammalian strains. Beyond particles appear to be stable when stored at ultralow
this region, there is considerable variability in sequence temperatures (−70° to 85 ° C) for 6 to 10 years (50), but may
among the serotypes, requiring gaps to be introduced for be disrupted by repeated freezing and thawing. Methanol
optimal alignment. The C-terminal 8 amino acids are has been shown to be more effective in reducing astrovirus
highly conserved, particularly among the human strains. infectivity than equivalent concentrations of isopropanol
Information regarding processing of the nonstructural or ethanol (51).
proteins is very limited. A preliminary report by Gibson When astroviruses were first described during the
and coworkers (47) noted a ‘‘lack of processing’’ in vitro 1980s, five serotypes of human astrovirus were identified
of ORFs 1a and 1a/1b translation products containing the by immunofluorescence, neutralization, and immune
3C-like serine protease motif. By contrast, Willcocks and EM (52–54). Each strain was isolated from natural
coworkers (41) studied nonstructural protein processing in infections in the United Kingdom and adapted to growth
astrovirus-infected Caco-2 cells by Western blot analysis in cell culture (31). Serotypes 6 and 7 were identified in the
with antisera specific for the C-terminal 298 aa of the United Kingdom in 1989, 1991, and 1992 (55). An eighth
ORF 1a product. They detected proteins of 74, 34, 20, 6.5, serotype is also now described (56).
and 5.5 kDa, but no full-length ORF 1a product (p101). By serotyping isolates, astrovirus serotype 1 has gener-
In a separate study (48), processing of the ORF 1a ally been the most common type found (20,53,55–63), but
product was studied in a cell-free expression. ORF-1a the predominant serotype may vary depending on loca-
products identified by astrovirus-specific and N-terminal tion (64). In community-acquired astroviruses studied in
tag-specific antibodies included the full-length 101-kDa Oxford, United Kingdom, between 1975 and 1987 (20),
protein (p101), an N-terminal cleavage product (p64), 72% of the astroviruses detected were serotype 1.
and a C-terminal cleavage product (p38). Mutation and Serotypes 2 to 5 each accounted for 6 to 8% of the astrovirus
deletion analyses indicated the important role of the strains. A later survey by Lee and Kurtz (55) examined the
catalytic triad of the viral 3C-like serine protease (Ser-551, astrovirus serotypes found in the Oxford region between
Asp-489, and His-461) in processing of p101. 1976 and 1992. Serotype 1 was again found most fre-
In summary, the astrovirus genome is organized with quently and accounted for an average of 65% of the cases
the primarily nonstructural proteins encoded by the 5
1991 was the only year in which the number of serotype
two-thirds (ORFs 1a and 1b) and the structural protein 2 cases exceeded the number of serotype 1 cases. A third
encoded by the 3
one-third (ORF 2). This arrangement study from the United Kingdom also found a predominance
resembles the genome organization of caliciviruses (49). of astrovirus serotype 1 (63). By contrast, a longitudinal
However, several features distinguish these two viral study of diarrhea in a cohort of Mexican children found
families, including differences in the size, number, and that astrovirus serotype 2 was most common (35%) and
processing of structural proteins, the lack of a helicase serotypes 1 and 5 were least common (4% each) (64).
domain in astroviruses, the use of ribosomal frameshifting Animals hyperimmunized with each of the human
to translate the RNA-dependent RNA polymerase, and astrovirus serotypes 1 to 7 raise antisera that appear
distinctive morphologic features described in the preceding to react type-specifically in immunofluorescence tests
text. Astroviruses differ from picornaviruses in genome and IEM (20,55,65). Also, rabbit polyclonal antiserum
organization, lack of a helicase domain, and use of
neutralizes astrovirus in a serotype-specific manner by
translational strategies, such as ribosomal frameshifting
plaque assay (65). All eight serotypes of human astrovirus
and subgenomic RNA.
are recognized by a monoclonal antibody produced
by Herrmann and coworkers (55,66). This monoclonal
Biology and Antigenic Properties
antibody, which does not neutralize virus, is directed
Astroviruses appear to incorporate specific features of a at a viral structural protein (57), most likely an epitope
number of different viruses without strictly resembling within the highly conserved N-terminal half of the ORF-2
any one group and therefore have been accorded product (Geigenmüller and Matsui, unpublished data). An
a separate viral family, the Astroviridae (38). The EIA in which the monoclonal antibody serves as a capture
 ASTROVIRUSES 331

antibody and polyclonal antiserum is used as a detector world (63,78,121). In samples of gamma globulin pools
antibody has been developed (66) and employed in large- collected in the United States, antibodies to all five
scale seroepidemiologic studies described in the following of the originally identified serotypes of astrovirus are
text. A typing EIA that uses serotype-specific reference detectable (111). Antibodies to astrovirus tend to be
antisera for antigen capture and the group-reactive acquired in early childhood. A survey of 87 children less
monoclonal antibody as detector has been developed than 10 years old in the Oxford region of the United
to assess the antigenic types of astroviruses found in Kingdom revealed that antibody prevalence rises rapidly
clinical samples (66). Molecular methods for genotyping from 7% in 6- to 12-month-old infants to 70% by school age.
clinical isolates by reverse transcriptase polymerase chain Astrovirus antibodies could be detected in 75% of the 10-
reaction (RT-PCR) may also be used, using primers year-old children studied. Among young adults (nursing
from the more conserved N- or C-terminal regions of students), 77% had antibodies to astroviruses.
ORF-2 (11,59). A high concordance between results of
genotyping and serotyping has been observed. Clinical Symptoms
Astroviruses are a cause of gastroenteritis worldwide.
EPIDEMIOLOGY AND CLINICAL DISEASE Young children, especially those under six months of
age, are the ones primarily affected. However, all ages
The medical importance of astrovirus in humans was may be affected. Older children and adults have been
established initially in studies conducted in Thailand involved in outbreaks (109,110), and astrovirus infec-
and Guatemala where astroviruses were found to be the tions have been recognized in elderly, institutional-
second most common cause (after rotavirus) of viral diar- ized patients (63,79,89,105) and in immunocompromised
rhea in young children (67,68). This was made possible patients (20,73,75,81,122). Human astrovirus infection
by the development of a monoclonal antibody EIA that induces a mild, watery diarrhea that typically lasts for two
can detect all known human astrovirus serotypes (64,66). to three days, associated with vomiting, fever, anorexia,
This and other improvements in detection techniques, abdominal pain, and various constitutional symptoms that
along with increased awareness of astrovirus infection last for up to four days (20,111,123–126). Protracted diar-
have helped to define the epidemiology of this illness. rhea and viral shedding have been observed in some
Astrovirus infections have been found worldwide, pri- studies (20,107,108). In children, it may not be possi-
marily, but not exclusively, in young children with ble to distinguish diarrhea owing to astrovirus from that
diarrhea (19,67–104). Sporadic outbreaks of gastroen- owing to rotavirus on clinical grounds alone (20,67,68)
teritis because of astrovirus have been reported among and studies of hospitalized children have found astro-
elderly patients (79,86,97,105) and military recruits (106). virus to be the second most common viral pathogen,
Several studies associate astroviruses with diarrhea in after rotavirus (61,116). In general, astrovirus diarrhea
immunocompromised adults (73,81,107,108). Large, food- is milder and does not lead to significant dehydration
borne outbreaks, affecting thousands of individuals in or hospitalization (16,64,67,83,85,91,95,126). However,
Japan, have occurred among otherwise normal school-age in a significant percentage of rural Egyptian children
children and adults as well (109,110). with astrovirus gastroenteritis severe dehydration was
Most astrovirus infections are detected in the seen (127). In addition, severe infections have been noted
winter months in temperate regions and in in young adults infected with astrovirus serotype 4 (74).
the rainy season in more tropical climates, a Deaths associated with astrovirus infections are extremely
pattern that resembles rotavirus infections (20,68,111). rare but have been reported (100,105).
Both community-acquired (1,67,82,89,95,112,113) and The role of astrovirus in persistent diarrhea remains
nosocomial (69,70,76,77,82,83,87,90,95,103,112,114,115) to be established (103). Prolonged lactose intolerance
infections have been described. In Australia, astrovirus wand, less commonly sensitivity to cow’s milk, have
is the second most common cause of gastroenteritis in been problems for some patients (76,83,95). Intravenous
children, after rotavirus (61). Infection with serotype 1 immunoglobulin may be a useful adjunct in severely
predominates, particularly in the winter months when immunodeficient patients who fail to respond to conser-
gastroenteritis is common (61,116,117). In a population of vative measures (107,128), but larger prospective studies
Mayan infants, astrovirus was the most common enteric are needed to establish the efficacy of this approach.
pathogen identified in stool samples collected during a In animals, as in humans, astrovirus has usually been
prospective study of oral poliovirus immunogenicity (118). identified in association with diarrhea, but in ducklings
Astroviruses have been shown to be an important has also been associated with a rapidly fatal hepati-
cause of outbreaks of diarrhea in childcare centers among tis (129). Avian astroviruses have also been associated
children age three years or less (85,91). In eight outbreaks with nephritis in chicks (130) and immunosuppression
of astrovirus diarrhea in six childcare centers, 20% of the compounding enteritis in turkey poults (131).
children with diarrhea shed astrovirus (92). Although each
outbreak was associated with a single astrovirus serotype,
Viral Pathogenesis
two distinct serotypes were identified in sequential
outbreaks during one winter diarrhea season. Pathogenesis of astrovirus infections in humans has been
Serological studies indicate that astrovirus infections limited to a report of two children with diarrhea that
are common in the United States (111,119), Japan (19), correlated fecal shedding of astrovirus with identification
the United Kingdom (60,120), and other areas of the of astrovirus particles in intestinal epithelial cells. This
332 ASTROVIRUSES

suggested that virus replication occurs in intestinal pattern suggesting nucleolar involvement. This was fol-
tissue in humans (99). Astrovirus particles in the biopsy lowed by dense immunofluorescent granules appearing in
specimen were localized to the epithelial cells in the perinuclear region and diffuse staining of the cyto-
the lower regions of the villus in a patient with plasm. This pattern of infection has also been seen in cell
sucrase isomaltase deficiency and in the ‘‘exposed cultures infected with human astroviruses (Herrmann,
surface epithelium’’ of a second patient who had a unpublished data).
severe enteropathy because of sensitivity to cow’s milk
formula. The significance of astrovirus infection in Modes of Transmission
these patients was not definitive because they had
Human adult volunteer studies indicate that astroviruses
underlying gastrointestinal problems and fecal shedding of
can be transmitted through the fecal-oral route, (22,132)
Escherichia coli 086 or rotavirus. Human volunteer studies although few of the volunteers developed diarrhea with
have not examined the histological effects of astrovirus- fecal shedding of virus. It appeared that person-to-person
related diarrhea (22,132). spread was responsible for an outbreak of gastroenteritis
In animals, gnotobiotic lambs given ovine astrovirus in Marin County, California, in 1978 (97), that was
showed that mild transient diarrhea was caused by subsequently shown to be owing to astrovirus type
infection of mature enterocytes in the apical two-thirds of 5 (138,139). Astrovirus diarrhea does not usually develop
villi and subepithelial macrophages, followed by transient in adults, but adult caregivers of infected children,
villus atrophy and crypt hypertrophy (3,133). Infected including parents, teachers and medical personnel, may
intestinal cells were detected by immunofluorescence become ill (20,109,125,140). An outbreak among military
between 14 and 70 hours after inoculation. Most infected troops has also been described (106). This suggests that
cells were found 14 to 38 hours postinoculation (133) and these adults may have been exposed to a larger dose
during this time, aggregates of virus particles along the of astrovirus or through a variable route (e.g., fomites,
microvilli or in lysosomes and autophagic vacuoles were contaminated food, or water). Astrovirus has also been
observed by EM (133). detected in water from an area where an outbreak
Bovine astrovirus infection, which does not result of astrovirus gastroenteritis occurred (141), and 93% of
in symptomatic illness, causes infection of M cells and surfers in the United Kingdom had antibodies to human
absorptive enterocytes overlying the dome villi of Peyer’s astrovirus serotype 4, compared to 22% of age-matched
patches in the small bowel (134,135). Infected cells were controls (142). A study of sewage and environmental
sloughed, replaced by cuboidal cells, and formed an samples from a water treatment facility in the United
exudate with inflammatory mononuclear and eosinophilic Kingdom failed to detect astrovirus RNA by RT-PCR (143).
cells above the dome villi. The lamina propria was A case of astrovirus gastroenteritis was diagnosed at the
infiltrated with neutrophils and contained cells with end of the sampling period, but occurred outside the
degenerate nuclei. Lymphoid cell depletion was noted in surveyed area.
the central region of germinal centers beneath infected Large outbreaks of astrovirus gastroenteritis occurred
dome villi. in Japan that were thought to be food-borne (109,110).
These two studies with mammalian astrovirus infec- Food handlers should be instructed that shedding of
tions suggest that astroviruses enter cells through the astrovirus in the feces may begin a day before symptoms
apical surface. In polarized CaCo-2 cells, however, human and continue for several days after diarrhea resolves.
astroviruses isolated from wild-type infections appear to In immunocompromised individuals, viral shedding may
enter through the basolateral surface (23). Whether or not last for weeks after resolution of symptoms (107,108). In
this result reflects the arrangement of astrovirus receptors addition, foods, such as shellfish that have been implicated
in differentiated intestinal epithelial cells in vivo remains in outbreaks of astrovirus gastroenteritis should be
to be determined (136). carefully selected and prepared.
Entry of astroviruses into cells has also been studied
using Graham 293 cells, a transformed line of primary RESERVOIRS
HEK cells (137) that has been frequently used for
isolation of enteric adenoviruses. The effects on astrovirus Astrovirus infections appear to be species-specific, which
infection of lysosomotropic agents (ammonium chloride, suggests that animals are not a reservoir for human infec-
methylamine, and dansylcadaverine) and the ionophore tions. Where antigenic relationships among astroviruses
monensin were investigated. All inhibited viral infection, of different animal species have been examined, there
suggesting that a functional endocytic pathway is has been no evidence for cross-reactivity between species
necessary for delivery of infectious astrovirus into the (5,6,54,135,144). In general, infection of animals and cell
cytoplasm. culture is also species-specific. BHK-21 cells support prop-
Bovine astrovirus infection in cell culture has also agation of human astrovirus type 2 (145), and astrovirus
been studied. Bovine astrovirus was propagated in pri- isolated from red deer has been shown to infect BEK cells,
mary NBK cells and the expression of viral antigens was but serial propagation was not achievable (5). It is likely
followed by fluorescent antibody probes (15). Immunofluo- that not all the animal astroviruses have been exten-
rescence was first detected at 7 hours postinfection in the sively tested as serial passage has not been established
cytoplasm of infected cells. Shortly thereafter, immunoflu- for astroviruses from lambs (3), red deer (5), kittens (8),
orescent granules were identified in the nucleus in a dogs (10,146), turkeys (12), and ducks (129). To date, only
 ASTROVIRUSES 333

bovine (15) and porcine (7) astroviruses have been propa- Propagation of serotypes 1 to 4 was accomplished in a
gated in cell culture. hepatoma cell line, PLC/PRF/5, a line that was also used
to isolate human astrovirus directly from fecal samples
(four isolates) (151). Several other cell lines have been
DETECTION, OCCURRENCE, AND PERSISTENCE tested for astrovirus isolation and cultivation (145). When
laboratory-adapted human astrovirus serotypes 1 through
Methods of Detection 7 were tested for growth in 15 human, 7 simian, and 10
Electron Microscopy. Before immunoassay or PCR other nonprimate mammalian cell lines, propagation of
methods for astrovirus detection were developed, EM or all seven serotypes was successful in the human cell lines
immune electron microscopy (IEM) techniques were the Caco-2, T84, HT-29, and in the African green monkey
major assays used, and although insensitive, are still kidney cell line MA-104. Both primary and secondary
used by some for detection of astroviruses where other African green monkey kidney cells were more effective
assays are not available. EM may also be important than Rhesus monkey kidney cells for cultivation of
for standardizing new detection methods as they are astrovirus. Except for human foreskin cells, all of the
developed. Patients with diarrhea owing to astrovirus other human and simian cell lines supported growth of
frequently shed large numbers of viral particles, 1010 at least one astrovirus serotype. The only nonprimate cell
per milliliter, or 108 viable particles per milliliter (20). line that permitted sustained passage of astroviruses was
IEM techniques are useful to detect lower levels of BHK-21 (C13) cell line for astrovirus serotype 2. BHK cells
astroviruses (19,147,148). The sensitivity of EM has been transfected with RNA transcripts derived from a genomic
estimated to be 106 to 107 virus particles per gram of stool. cDNA of human astrovirus serotype 1 have been shown
Diagnosis by EM requires experience, as only 10% to support viral replication and morphogenesis (30). Of 17
of the particles in a given specimen may display human stool specimens that had previously been shown to
the distinctive astrovirus surface star (149). Oliver be astrovirus positive by ELISA, Caco-2 cells (13 isolates),
and Phillips conducted a retrospective analysis of T84 cells (12 isolates), and PLC/PRF/5 cells (12 isolates)
fecal viruses that were originally identified as small, were the cell lines most effective for isolation of human
round, featureless viruses by EM alone (96). When astroviruses from these stool specimens (145).
complementary immunologic techniques were applied to
careful reexamination of the specimens by direct EM, 14 Enzyme Immunoassay. A group reactive monoclonal
of the 53 samples (26%) first classified as small, round, antibody (8E7) was developed by Herrmann and coworkers
featureless viruses were shown to be astrovirus. Two and was used in an enzyme immunoassay (EIA) to
caliciviruses and a Norwalk-like virus were also found to capture viral antigen in stools (65,66). In this assay,
be misclassified. Another example of misclassification by polyclonal antiserum is used as the detector antibody
direct EM is the Marin County virus, originally classified in an indirect EIA. A modification of this EIA uses
in 1978 as a ‘‘Norwalk-like’’ virus (97). Subsequent a biotinylated polyclonal detector antibody (93). Both
analyses, however, proved that the Marin County virus EIAs are comparable to IEM in sensitivity (91%) and
was an astrovirus (serotype 5) (138,139). specificity (98%). EIAs have been useful in rapidly
detecting astrovirus antigen in studies where a large
Isolation in Cell Culture. Serial passage of astroviruses number of samples must be assayed (2,67,68,85,112). EIA-
in cell culture (primary HEK cells) was first described and RT-PCR-based methods to type human astroviruses
in 1981 by Lee and Kurtz (31). This was made possible found in clinical samples have been described more
by incorporating trypsin (10 µg/mL) in serum-free growth recently (59). An astrovirus antigen EIA that uses a group
media (31). After six passages in HEK cells, primary reactive monoclonal antibody [8G4] for both capture and
baboon kidney (PBK) cells and a continuous line of detection (biotinylated) has also been developed (112). The
LLCMK2 cells could also be infected, but not by sensitivity of EIA is estimated to be 105 to 106 viral
direct inoculation of PBK or LLCMK2 cells with fecal particles per gram of stool (66,93).
derived astroviruses. All five of the serotypes of human A commercial enzyme immunoassay based on the group
astrovirus first described (53) were adapted to growth in reactive monoclonal antibody 8E7 (59) is available for
LLCMK2 cells. the detection of astrovirus antigen in clinical specimens
Direct isolation of an astrovirus from stools in a (IDEIATM Astrovirus EIA, DAKO Diagnostics). In one
continuous cell line was first described by Willcocks and survey from Australia, this qualitative assay was found to
coworkers (23) for astrovirus serotype 1. The cell line have a sensitivity of 100% and specificity of 98.6% (117).
used, CaCo-2 cells, a continuous line of human colon
RNA Probes and Reverse Transcriptase-Polymerase Chain
carcinoma, has been subsequently found to allow isolation
Reaction (rt-PCR)
of all astrovirus serotypes, including the more recently
identified types 6, 7, and 8 (43,55,145). Use of this cell Hybridization. The use of nucleic acid probes for astro-
line may allow more rapid isolation and subsequent virus identification was first described in 1991 (93,152).
propagation of wild-type astroviruses (150). Growth in The sensitivity of RNA probes is approximately equal to
CaCo-2 cells may be also a way to distinguish astrovirus that obtained by EIA (93,152), thus for routine diagnosis
infection from infection with human caliciviruses, viruses there is no advantage of probes compared with monoclonal
that have not been successfully isolated or propagated in antibody immunoassays for astroviruses in gastroenteri-
cell culture. tis. The sensitivity is similar, but the probe techniques
334 ASTROVIRUSES

are considerably more complex and require more time to environmental source was described in 1996, in a sample
obtain results. of sewage polluted water (141). The virus was recovered
by passage of 500 L of water through positively charged
Polymerase Chain Reaction. Amplified techniques, such filters (Zeta plus MKII filters) followed by elution with
as the polymerase chain reaction (PCR), offer a more sensi- 0.05 M glycine buffer plus 3% beef extract, and organic
tive approach than direct hybridization, and PCR applica- flocculation. The final volume of the sample was 50 mL.
tions have been described for astroviruses (81,153,154). Viral nucleic acid could not be directly detected in the
The sensitivity in detecting astrovirus infections is concentrated samples, but was detected after passage
improved, and is estimated to be 10 to 100 particles per of the samples in CaCo-2 cells. It was estimated that
gram of stool (62). In one study of an outbreak of astrovirus for the wild-type astrovirus strain detected, there was
gastroenteritis at a day care center (155), the prevalence a minimum of 20 astroviruses per liter of in the
of astroviruses determined by EIA varied from 50% in unconcentrated sample. A second report was published in
affected rooms to 20% in rooms with older children. Analy- 2000 (160). Surface water samples were processed using
sis by RT-PCR indicated nearly all children were infected filtration through cartridge filters (type 1 MDS Zetapor
and viral shedding was detectable for weeks despite resolu- cartridge filters) followed by elution with beef extract
tion of symptoms. In addition, RT-PCR detected astrovirus and concentration by organic flocculation according to the
in 32% of the samples tested, compared to a detection rate method described for enteric viruses by the Environmental
of 10% by EIA. Protection Agency (161). Portions of concentrated samples
The complete sequences of four strains of astrovirus were inoculated onto monolayers of CaCo-2 cells and tested
(two serotype 1, one serotype 2 and one serotype 3) for astroviruses by RT-PCR. Of 29 water samples analyzed,
are currently available. Oligonucleotide primers selected 15 were positive for astroviruses.
from sequences at the 3
-end of the genome can
amplify astrovirus-specific products from all five reference Recovery from Sludge. Because astroviruses have a
serotypes (156). Primers derived from the highly conserved different capsid structure than enteroviruses, Chapron
RNA-dependent RNA polymerase motif may also be and coworkers (162) compared the prevalence of human
suitable candidates for amplification of different human astroviruses in sludges as compared to enteroviruses,
astrovirus serotypes by RT-PCR (32). RT-PCR has been using an EPA method designed for other enteric viruses.
used also to confirm astrovirus-positive samples detected Using 500 mL of sludge samples collected from sewage
by EIA (73,81,91,118). Good correlation has been observed plants in the United States from January to June 1999,
between these two methods for detection (78). viruses were concentrated from sludges according to the
EPA CFR part 503 regulations (EPA 1992). This procedure
Recovery and Concentration. As was discussed earlier, uses beef extract elution followed by organic flocculation,
the mode of transmission of astroviruses is considered and results in a final volume of 20 mL. For detection of
to be similar to that of other enteric viruses, which viruses, CaCo-2 cells were inoculated and virus detected
includes potential transmission in environmental samples. by RT-PCR/nested PCR. Of 16 sample sites analyzed,
The environmental samples most commonly assayed are 15 were positive for astroviruses and 16 were positive
those in which other enteric viruses have been isolated for enteroviruses. This study not only demonstrated that
or detected, including surface waters, drinking water, astroviruses can be recovered from sewage sludge, but
shellfish, and sewage samples. The stability of astroviruses also showed that astroviruses can be detected in sludge
in water also suggests the possibility of transmission samples by standard EPA procedures for the recovery of
by the water route. In dechlorinated water, astrovirus enteric viruses.
survival was comparable to rotavirus serotype 3 and
enteric adenovirus 40/41, both of which can be transmitted Recovery from Shellfish. Methods have been described
through water. Astrovirus and adenovirus were more for recovering astroviruses from samples of mussels (163)
resistant than poliovirus to inactivation by the addition of or mussels and oysters (164) seeded with laboratory
free chlorine to the water, but less resistant than human strains of viruses, but to date only one report has been
rotavirus or hepatitis A virus (157,158). published for naturally occurring astroviruses in shellfish.
Although studies on detection of astroviruses in In a three-year study of enteric viruses in shellfish, Le
environmental samples have been very limited in Guyader and coworkers (159) detected astroviruses in
comparison to those for other enteric viruses, the 17% of oyster samples and 50% of mussel samples. The
techniques for recovery of astroviruses are similar mussels were collected from beds frequently contaminated
and follow the same general procedures; extraction, by fecal coliforms. Astroviruses were recovered more
concentration, purification (or removal of inhibitors), and frequently from mussels than either hepatitis A virus
detection. For water samples, this entails adsorption of (13%) or Norwalk-like viruses (35%), both of which have
virus to filters followed by elution and concentration. been implicated in shellfish-borne outbreaks of disease.
For the one study on naturally occurring astroviruses Recovery of astroviruses was accomplished by extracting
in foods (shellfish) (159), virus was extracted from a nucleic acid for PCR from dissected tissues (stomach
food homogenate with organic solvents, followed by virus and digestive diverticula) of shellfish, a method that
precipitation and nucleic acid extraction for RT-PCR. has been described for recovery of Norwalk virus and
Hepatitis A (165). To concentrate virus, the tissues were
Recovery from Water Samples. The first description homogenized in buffered saline, extracted with chloroform-
of recovery of astroviruses naturally occurring in an butanol, and precipitated with polyethylene glycol. Viral
 ASTROVIRUSES 335

nucleic acids were extracted by the phenol-chloroform- 22. K. Midthun et al., J. Clin. Microbiol. 31, 955–962 (1993).
water technique and ultimately precipitated with ethanol 23. M. M. Willcocks et al., Arch. Virol. 113, 73–81 (1990).
for use in RT-PCR. As is the case for other enteric 24. B. L. Semler, R. J. Kuhn, and E. Wimmer, RNA-directed
viruses, the extracts need to be tested for the presence Virus Replication, in E. Domingo, J. J. Holland, and
of PCR inhibitors. A commercial kit that uses a guanidine P. Ahlquist, eds., RNA Genetics, vol. 1, CRC Press, Boca
thiocyanate silica procedure to extract viral RNA directly, Raton, Fla., 1988, pp. 23–48.
without mechanical homogenization of shellfish, is now 25. J. B. Kurtz, Astroviruses, in M. J. G. Farthing, ed., Viruses
available (Promega, Madison, Wis.). The virus of interest and the Gut, Proceedings of the Ninth British Society of
(astrovirus, Norwalk virus, hepatitis A virus) then can Gastroenterology, Smith Kline and French International
be amplified from the extracted RNA using virus-specific Workshop, Windsor, Berkshire, U.K., 1988, pp. 84–87.
primers in RT-PCR assays (164). Methods for isolation or 26. S. S. Monroe et al., J. Virol. 65, 641–648 (1991).
detection of astroviruses in foods other than shellfish have 27. A. Sanchez-Fauquier et al., Virology 201, 312–320 (1994).
not been reported. 28. G. Belliot, H. Laveran, and S. S. Monroe, Virus Res. 49,
49–57 (1997).
CONCLUSION 29. D. M. Bass and S. Qiu, J. Virol. 74, 1810–1814 (2000).
30. U. Geigenmüller, N. H. Ginzton, and S. M. Matsui, J. Virol.
The medical importance of astroviruses as major causes 71, 1713–1717 (1997).
of gastroenteritis has now been well established through 31. T. W. Lee and J. B. Kurtz, J. Gen. Virol. 57, 421–424 (1981).
clinical and epidemiological studies. The detection of astro- 32. T. L. Lewis et al., J. Virol. 68, 77–83 (1994).
viruses in water and shellfish samples, combined with 33. M. M. Willcocks and M. J. Carter, Arch. Virol. 124, 279–289
epidemiological evidence for food-borne outbreaks of astro- (1992).
virus gastroenteritis (109,110) suggests that astroviruses 34. M. M. Willcocks and M. J. Carter, FEMS Microbiol. Lett.
have been spread through environmental sources. With 114, 1–8 (1993).
the techniques for detection of astroviruses that are now 35. M. M. Willcocks et al., J. Gen. Virol. 75, 1785–1788 (1994).
available, it will be possible to assess the impact of these
36. B. Jiang et al., Proc. Natl. Acad. Sci. U.S.A. 90,
viruses in food and waterborne gastroenteritis. 10539–10543 (1993).
37. M. Méndez-Toss et al., J. Gen. Virol. 81, 2891–2897 (2000).
BIBLIOGRAPHY 38. S. S. Monroe et al., Astroviridae, in M. H. V. van Regenmor-
tel, C. M. Fauquet, and D. H. L. Bishop, eds., Virus Taxon-
1. C. R. Madeley et al., J. Hyg. [Cambridge] 787, 261–273 omy, Academic Press, San Diego, Calif., 2000, pp. 741–745.
(1977).
39. M. M. Willcocks et al., J. Virol. 68, 6057–6058 (1994).
2. D. Madeley, APMIS. 101, 497–504 (1993).
40. C. Monceyron, B. Grinde, and T. Ø. Jonassen, Arch. Virol
3. D. R. Snodgrass and E. W. Gray, Arch. Virol. 55, 287–291
142, 699–706 (1997).
(1977).
41. M. M. Willcocks, A. S. Boxall, and M. J. Carter, J. Gen.
4. G. N. Woode and J. C. Bridger, J. Med. Microbiol. 11,
Virol. 80, 2607–2611 (1999).
441–452 (1978).
42. G. Belliot, H. Laveran, and S. S. Monroe, Arch. Virol. 142,
5. S. Tzipori, J. D. Menzies, and E. W. Gray, Vet. Rec. 108, 286
1323–1334 (1997).
(1981).
43. G. Belliot et al., Abstracts of the American Society for
6. J. C. Bridger, Vet. Rec. 107, 532–533 (1980).
Virology Meeting, Amherst Mass., 1999, p. 176.
7. M. Shimizu et al., J. Clin. Microbiol. 28, 201–206 (1990).
44 E. V. Koonin, J. Gen. Virol. 72, 2197–2206 (1991).
8. Y. Hoshino et al., Arch. Virol. 70, 373–376 (1981).
45. G. Kadaré and A. -L. Haenni, J. Virol. 71, 2583–2590 (1997).
9. E. Kjeldsberg and A. Hem, Arch. Virol. 84, 135–140 (1985).
46. M. M. Willcocks et al., Epidemiol. Infect. 114, 385–391
10. J. A. Marshall et al., Aust. Vet. J. 61, 33–38 (1984).
(1995).
11. S. M. Matsui et al., Microbiol. Immunol. 42, 539–547 (1998).
47. C. A. Gibson et al., Adv. Exp. Med. Biol. 440, 387–391
12. M. S. McNulty, W. L. Curran, and J. B. McFerran, Vet. Rec. (1998).
106, 561 (1980).
48. D. Kiang and S. M. Matsui, (2001), submitted.
13. D. L. Reynolds and Y. M. Saif, Avian Dis. 30, 728–735
(1986). 49. M. J. Studdert, Arch. Virol. 58, 157–191 (1978).
14. C. R. Madeley, J. Infect. Dis. 139, 519–523 (1979). 50. F. P. Williams Jr., J. Med. Virol. 29, 192–195 (1989).
15. D. Aroonprasert et al., Vet. Microbiol. 19, 113–125 (1989). 51. J. B. Kurtz, T. W. Lee, and A. J. Parsons, J. Hosp. Infect. 1,
321–325 (1980).
16. M. Yeager et al., 2001, submitted.
52. R. W. Hudson, J. E. Herrmann, and N. R. Blacklow, Arch.
17. C. R. Ashley and E. O. Caul, J. Clin. Microbiol. 16, 377–381
Virol. 108, 33–38 (1989).
(1982).
18. E. O. Caul and H. Appleton, J. Med. Virol. 9, 257–265 53. J. B. Kurtz and T. W. Lee, Lancet 2, 1405 (1984).
(1982). 54. T. W. Lee and J. B. Kurtz, J. Hyg. [Cambridge] 89, 539–540
19. T. Konno et al., J. Med. Virol. 9, 11–17 (1982). (1982).
20. J. B. Kurtz and T. W. Lee, in G. Bock and J. Whelan, eds., 55. T. W. Lee and J. B. Kurtz, Epidemiol. Infect. 112, 187–193
Novel Diarrhoea Viruses. Ciba Foundation Symposium 128, (1994).
John Wiley & Sons, Chichester, 1987, pp. 92–107. 56. M. B. Taylor et al., J. Med. Virol. 64, 256–261 (2001).
21. S. M. Matsui et al., J. Virol. 67, 1712–1715 (1993). 57. T. O. Jonassen et al., J. Virol. Methods 52, 327–334 (1995).
336 ASTROVIRUSES

58. H. Mustafa, E. A. Palombo, and R. F. Bishop, J. Clin. 99. A. D. Phillips, S. J. Rice, and J. A. Walker-Smith, Gut 23,
Microbiol. 38, 1058–1062 (2000). A923–A924 (1982).
59. J. S. Noel et al., J. Clin. Microbiol. 33, 797–801 (1995). 100. P. B. Singh, M. A. Sreenivasan, and K. M. Pavri, Epidemiol.
60. J. Noel and D. Cubitt, Epidemiol. Infect. 113, 153–159 Infect. 102, 345–353 (1989).
(1994). 101. I. M. Spence, S. Afr. Med. J. 64, 181–182 (1983).
61. E. A. Palombo and R. F. Bishop, J. Clin. Microbiol. 34, 102. A. D. Steele et al., Ann. Trop. Pediatr. 18, 315–319 (1998).
1750–1753 (1996). 103. L. E. Unicomb et al., Pediatr. Infect. Dis. J. 17, 611–614
62. K. Saito et al., Microbiol. Immunol. (Japan) 39, 825–828 (1998).
(1995). 104. A. Y. Xu et al., Chin. Med. J. (Engl. Ed.) 94, 659–662 (1981).
63. S. A. Wilson and W. D. Cubitt, J. Virol. Methods 19, 105. S. M. Matsui et al., Gastroenterology 106, A730 (1994).
151–160 (1988). 106. G. Belliot, H. Laveran, and S. S. Monroe, J. Med. Virol. 51,
64. M. L. Guerrero et al., Pediatr. Infect. Dis. J. 17, 723–727 101–106 (1997).
(1998). 107. M. Björkholm et al., Int. J. Hematol. 62, 117–120 (1995).
65. J. E. Herrmann et al., J. Infect. Dis. 158, 182–185 (1988). 108. P. Coppo et al., Ann. Hematol. 79, 43–45 (2000).
66. J. E. Herrmann et al., J. Infect. Dis. 161, 226–229 (1990). 109. I. Oishi et al., J. Infect. Dis. 170, 439–443 (1994).
67. J. E. Herrmann et al., N. Engl. J. Med. 324, 1757–1760 110. E. T. Utagawa et al., J. Clin. Microbiol. 32, 1841–1845
(1991). (1994).
68. J. R. Cruz et al., J. Clin. Microbiol. 30, 1140–1144 (1992). 111. C. W. LeBaron et al., Morbid. Mortal. Wkly Rep. 39(RR-5),
69. H. Appleton and P. G. Higgins, Lancet 1, 1297 (1975). 1–24 (1990).
70. C. R. Ashley, E. O. Caul, and W. K. Paver, J. Clin. Pathol. 112. S. Shastri et al., J. Clin. Microbiol. 36, 2571–2574 (1998).
31, 939–943 (1978). 113. M. B. Taylor, F. E. Marx, and W. O. Grabow, Epidemiol.
71. R. M. Avery et al., J. Diarrh. Dis. Res. 10, 105–108 (1992). Infect. 119, 227–230 (1997).
114. R. Bennet et al., Acta Paediatr. 84, 667–671 (1995).
72. P. R. Bates et al., J. Med. Virol. 39, 224–228 (1993).
115. M. Riepenhoff-Talty et al., J. Clin. Microbiol. 17, 352–356
73. G. J. Cox et al., Gastroenterology 107, 1398–1407 (1994).
(1982).
74. W. D. Cubitt, Bailliere’s Clin. Gastroenterol. 4643–4656
116. A. Gaggero et al., J. Clin. Microbiol. 36, 3691–3693 (1998).
(1990).
117. C. J. McIver et al., J. Virol. Methods 84, 99–105 (2000).
75. W. D. Cubitt et al., J. Med. Virol. 57, 313–321 (1999).
118. Y. Maldonado et al., J. Infect. Dis. 178, 334–339 (1998).
76. H. Esahli et al., Pediatr. Infect. Dis. J. 10, 511–515 (1991).
119. D. K. Mitchell et al., Pediatr. Infect. Dis. J. 18, 249–254
77. E. L. Ford-Jones et al., Am. J. Epidemiol. 131, 711–718
(1999).
(1990).
120. J. Kurtz and T. Lee, Med. Microbiol. Immunol. 166, 227–230
78. R. I. Glass et al., Arch. Virol. [Suppl.] 12, 287–300 (1996). (1978).
79. J. J. Gray et al., J. Med. Virol. 23, 377–381 (1987). 121. M. P. G. Koopmans et al., Clin. Diagn. Lab. Immunol. 5,
80. G. Grohmann, in S. Tzipori et al., eds., Infectious Diarrhoea 33–37 (1998).
in the Young, Elsevier, Amsterdam, 1985, pp. 25–32. 122. D. J. Wood et al., J. Med. Virol. 24, 435–444 (1988).
81. G. S. Grohmann et al., N. Engl. J. Med. 329, 14–20 (1993). 123. N. R. Blacklow and H. B. Greenberg, N. Engl. J. Med. 325,
82. K. L. Kotloff et al., Pediatr. Infect. Dis. J. 11, 587–589 252–264 (1991).
(1992). 124. H. B. Greenberg and S. M. Matsui, Infect. Agents Dis. 1,
83. J. B. Kurtz, T. W. Lee, and D. Pickering, J. Clin. Pathol. 30, 71–91 (1992).
948–952 (1977). 125. J. Kurtz and W. D. Cubitt, Astroviruses and Caliciviruses, in
84. J. F. Lew et al., Pediatr. Infect. Dis. J. 9, 709–714 (1990). M. J. G. Farthing and G. T. Keusch, eds., Enteric Infection:
85. J. F. Lew et al., J. Infect. Dis. 164, 673–678 (1991). Mechanisms, Manifestations, and Management, Raven
Press, New York, 1989, pp. 205–215.
86. D. C. Lewis et al., J. Hosp. Infect. 14, 9–14 (1989).
126. X.-L. Pang and T. Vesikari, Acta Paediatr 88, 532–536
87. C. R. Madeley and B. P. Cosgrove, Lancet 2, 124 (1975).
(1999).
88. J. A. Marshall et al., J. Trop. Med. Hyg. 85, 213–215 (1982). 127. A. B. Naficy et al., J. Infect. Dis. 182, 685–690 (2000).
89. D. M. McLean, K. S. K. Wong, and S. K. A. Bergman, Can. 128. K. -Y. Yuen et al., Diagn. Microbiol. Infect. Dis. 30, 75–81
Med. Assoc. J. 1, 1035–1036 (1977). (1998).
90. P. J. Middleton, M. T. Szymanski, and M. Petric, Am. J. Dis. 129. R. E. Gough et al., Vet. Rec. 114, 279 (1984).
Child. 131, 733–737 (1977).
130. T. Imada et al., J. Virol. 74, 8487–8493 (2000).
91. D. K. Mitchell et al., J. Pediatr. 123, 725–732 (1993).
131. M. Koci et al., J. Virol. 74, 6173–6177 (2000).
92. D. K. Mitchell et al., J. Infect. Dis. 180, 514–517 (1999).
132. J. B. Kurtz et al., J. Med. Virol. 3, 221–230 (1979).
93. C. L. Moe et al., J. Clin. Microbiol. 29, 2390–2395 (1991). 133. D. R. Snodgrass et al., Arch. Virol. 60, 217–226 (1979).
94. S. S. Monroe et al., J. Med. Virol. 33, 193–198 (1991). 134. G. A. Hall, in G. Bock and J. Whelan, eds., Novel Diarrhoea
95. H. Nazer, S. Rice, and J. A. Walker-Smith, J. Pediatr. Viruses. Ciba Foundation Symposium 128, John Wiley &
Gastroenterol. Nutr. 1, 555–558 (1982). Sons, Chichester, 1987, pp. 192–217.
96. A. R. Oliver and A. D. Phillips, J. Med. Virol. 24, 211–218 135. G. N. Woode et al., J. Clin. Microbiol. 19, 623–630 (1984).
(1988). 136. M. M. Willcocks, M. J. Carter, and C. R. Madeley, Rev. Med.
97. L. S. Oshiro et al., J. Infect. Dis. 143, 791–795 (1981). Virol. 2, 97–106 (1992).
98. R. Pavone, et al., Ann. Trop. Paediatr. 10, 15–20 (1990). 137. G. Donelli et al., J. Med. Virol. 38, 271–277 (1992).
 ATTACHMENT OF MICROORGANISMS 337

138. J. E. Herrmann et al., Arch. Virol. 110, 213–220 (1990). 154. M. M. Willcocks, J. G. Silcock, and M. J. Carter, FEMS
139. J. E. Herrmann, R. W. Hudson, and N. R. Blacklow, Lancet Microbiol. Lett. 112, 7–12 (1993).
2, 743 (1987). 155. D. K. Mitchell et al., J. Infect. Dis. 172, 1437–1444 (1995).
140. H. Appleton, in G. Bock and J. Whelan, eds., Ciba Founda- 156. S. S. Monroe et al., J. Virol. 67, 3611–3614 (1993).
tion Symposium 12, John Wiley & Sons, Chichester, 1987, 157. F. X. Abad et al., Appl. Environ. Microbiol. 60, 2377–2383
pp. 108–125. (1994).
141. R. M. Pintó et al., Appl. Environ. Microbiol. 62, 1811–1813 158. F. X. Abad et al., Appl. Environ. Microbiol. 63, 3119–3122
(1996). (1997).
142. S. Myint, R. Manley, and D. Cubitt, Lancet 343, 1640–1641 159. F. Le Guyader et al., Appl. Environ. Microbiol. 66,
(1994). 3241–3248 (2000).
143. S. I. Egglestone et al., J. Appl. Microbiol. 86, 709–714 160. C. D. Chapron et al., Appl. Environ. Microbiol. 66, 2520–
(1999). 2525 (2000).
144. D. A. Harbour et al., Vet. Rec. 120, 555–557 (1987). 161. Environmental Protection Agency, Virus Monitoring Pro-
145. J. P. Brinker, N. R. Blacklow, and J. E. Herrmann, Arch. tocol for the Information Collection Requirements Rule.
Virol. 145, 1847–1856 (2000). EPA/814-B95-002, Government Printing Office, Cincinnati,
146. F. P. Williams Jr., Arch. Virol. 66, 216–226 (1980). Ohio, 1995.
147. L. Berthiaume et al., J. Gen. Virol. 55, 223–227 (1981). 162. C. D. Chapron, N. A. Ballester, and A. B. Margolin, J. Appl.
148. E. Kjeldsberg, Acta Pathol. Microbiol. Immunol. Scand [B] Microbiol. 89, 11–15 (2000).
85B, 351–354 (1977). 163. O. Traore et al., Appl. Environ. Microbiol. 64, 3118–3122
(1998).
149. C. R. Madeley and B. P. Cosgrove, Lancet 2, 451–452 (1975).
164. O. Legeay et al., J. Virol. Methods 90, 1–14 (2000).
150. R. M. Pintó, J. M. Diez, and A. Bosch, J. Med. Virol. 44,
310–315 (1994). 165. R. L. Atmar et al., Appl. Environ. Microbiol. 62, 254–258
151. M. B. Taylor, W. O. Grabow, and W. D. Cubitt, J. Virol. (1996).
Methods 67, 13–18 (1997).
152. M. M. Willcocks et al., Epidemiol. Infect. 107, 405–410
(1991). ATTACHMENT OF MICROORGANISMS.
153. T. O. Jonassen, E. Kjeldsberg, and B. Grinde, J. Virol. See ADHESION (PRIMARY) OF MICROORGANISMS ONTO SURFACES
Methods 44, 83–88 (1993).
 B
BACTERIA: STREAMS. See STREAM MICROBIOLOGY BASELINE STUDIES

A number of multiple site baseline studies have been

conducted in volunteer homes in the United States and in
the United Kingdom. In some of these studies, sites and
BACTERIAL CELL STRUCTURE. See SOIL BACTERIA surfaces throughout the home have been examined (1,2),
whereas in others, kitchens and bathrooms (3) or kitchens
only (4,5) have been examined. There does not appear to
be any significant difference in the data based on the
regional or national distributions of the homes in these
BACTERIAL CONTAMINANTS IN RESIDENTIAL surveys. This is perhaps not surprising as, in general, the
ENVIRONMENTS lifestyles are relatively similar in the developed world,
and most of the homes in these surveys probably fall
ELIZABETH SCOTT into a broadly similar socioeconomic distribution. Climatic
Newton, Massachusetts conditions do not appear to have made a significant impact
on the microbiology of surfaces inside the home.

Over the last thirty years, scientists have drawn

Sampling Techniques
attention to a number of actual and potential infection
concerns related to the presence of biocontaminants in A number of different sampling techniques have been
the home. The list of infections associated with the employed. Hard, flat surfaces such as counter tops have
home includes food-borne infections, shigella dysentery, been sampled with a combination of Rodac plates (1,2),
viral gastroenteritis, common cold virus infections, skin contact slides (2), and moist cotton swabs (1–5). A ‘‘sweep
infections, respiratory allergies, and zoonoses associated plate’’ method has been reported for dry fabrics (1) and for
with domestic pets. In addition, there are specific infection premoistened swabs. Kitchen rags and sponges have also
control issues related to the care of infants and young been sampled by wetting them with a sterile solution and
children, the elderly, and other immunocompromised then collecting a small quantity of this fluid by squeezing it
persons in the home. This has prompted an interest in out of the rag or sponge (3,4). Settle plates have been used
the study of microbial contamination at environmental for air sampling (2) and liquid samples from toilets and
sites and surfaces in residential settings. Most studies waste traps have been collected by pipette and transferred
have investigated bacterial contaminants, and there are to contact slides in their containers.
The different sampling techniques that have been
very few reported studies on fungi or viruses in residential
employed make it difficult to draw conclusions from
environments.
specific comparisons between the studies. However, as
This article focuses on bacterial contamination at
the following sections will show, it is possible to see
environmental surfaces in domestic settings and the
common trends in the results. While sampling techniques
different approaches that have been employed to elicit
have changed little, laboratory techniques have become
data on this topic.
much more sophisticated over the period in which
these studies took place. In addition to searching for
evidence of contamination with Salmonella species, more
BACTERIAL CONTAMINANTS recent studies have specifically looked for pathogens such
as Yersinia, Listeria, and Campylobacter. The role of
Studies of domestic bacterial contaminants have largely these organisms in food-borne illness and their culture
been conducted under one of the following approaches: requirements was not widely known before the last decade.
(1) baseline studies of either multiple or single sites
that collect information on a broad spectrum of bacterial Bacterial Species
species, (2) searches focused on collecting data on specific
The range of organisms examined in baseline surveys of
pathogens at environmental sites in the absence of ill-
all sites and surfaces in the home is given in Table 1.
ness in the home, (3) pathogen searches at environmental
sites related to the presence of a known human illness,
(4) cross-contamination studies designed to elicit informa- BACTERIAL CONTAMINATION IN DOMESTIC KITCHENS
tion and understanding on the potential for the transfer
of potential pathogens from and via environmental sites, In total, results for 28 different kitchen sites have been
or (5) disinfectant studies that attempt to evaluate house- reported, as shown in Table 2. A factor that seems to
hold disinfectant products under ‘‘in-use’’ conditions in the have a major impact on the type of species and levels of
home. contamination, at any given site, is whether the site is wet
338
 BACTERIAL CONTAMINANTS IN RESIDENTIAL ENVIRONMENTS 339

Table 1. Bacterial Species Iso- Table 2. Sites and Surfaces

lated from Sites and Surfaces Sampled in Domestic Kitchens
in the Home
Wet Sites
Enterobacteria
E. coli Sink surface and drain rim area
K. pneumoniae Sink drain
Klebsiella spp. Dishcloth/sponge
Proteus mirabilis Sink drainer
Salmonella spp. Sink brush and sink bowl
Citrobacter freundii Cleaning utensils
Citrobacter spp. Soap dish
Ent. Cloacae
Ent. Agglomerans Dry Sites
Enterobacter spp.
Serratia spp. Food Contact Surfaces
Pseudomonads Worktop and cutting board
P. aeruginosa Refrigerator surfaces
P. maltophilia Cooker top/kitchen table
P. cepacia Cutlery and crockery
P. putrifaciens Food shelf and vegetable rack
P. fluorescens
Pseudomonas spp. (nontypable) Hand Contact Surfaces
Aeromonas hydrophila Sink taps/faucet handles
Others Tea towel/hand towel
Yersinia enterocolitica Handles
Campylobacter spp.
Listeria monocytogenes General Surfaces
S. aureus Kitchen floors
Bacillus cereus Trash barrels
Bacillus spp. Window sill
Streptococci curtains
Micrococci radiator
Ceramic tiles

or dry. Dry sites can be further divided into food contact

surfaces, hand contact surfaces, and general surfaces.
coliform counts. Even so, for food contact surfaces, total
For all kitchen sites, species of Bacillus and coagulase-
negative micrococci are almost universally present. and fecal coliform counts range up to 106 per sample
Results for wet sites indicate some common trends area on counter tops and cutting boards, refrigerators,
although differing sampling, culturing, and counting and kitchen tables. The presence of E. coli is recorded
techniques are probably responsible for some of the on worktops and cutting boards, refrigerator surfaces,
variations in the frequency of occurrence and the cookers, clean cutlery/crockery, and on food shelves. Fecal
levels of contamination for any particular species. Other coliforms have also been recovered from 24% of cutting
parameters such as the occurrence of food preparation or boards, 20% of kitchen tables, and 10% of refrigerator
cleaning and disinfection before sampling may also impact surfaces in one study (5). Staphylococcus aureus was
the data. recovered from 7.4% of cutting boards in one study (2)
Highest recovery for Escherichia coli and other and Pseudomonas spp. have been recovered from 37% of
coliforms is associated with sinks, sink drains, and to refrigerator surfaces.
a lesser extent, draining boards. High total and fecal For hand contact surfaces, sink taps are often found to
coliform counts of 105 –107 per sample area also have been be contaminated with Enterobacteria. Fecal coliforms have
commonly recorded at sink surfaces (5) and around the rim been isolated from about 32% of sink taps with average
of the sink drain (3). Counts as high as 1011 per sample counts of 103 to 104 and S. aureus has been isolated from
area for E. coli have been recorded from sink surfaces (1). up to 48% of kitchen towels (1). Although some studies
Overall, the recovery of E. coli from sink surfaces ranged have found that hand contact surfaces such as handles
from 4% to 81% of samples. Recovery of E. coli from drains showed only species of Staphylococci and Bacillus, average
has been somewhat more consistent, ranging from 39% total and fecal coliforms of 106 and 102 , respectively, per
to 65% of samples, probably reflecting the influence of sample area have been recorded from the refrigerator
moisture on bacterial survival and recovery. handle (3).
High total and fecal coliform counts of 107 to 109 General sites such as kitchen floors and kitchen trash
also are associated with sponges, with more than 65% barrels generally show low levels of contamination with a
of samples yielding these organisms on occasion (5). The range of different organisms and high levels of Bacillus and
presence of E. coli and Pseudomonas aeruginosa also has Micrococci spp. Pseudomonas species have been recovered
been recorded in baseline studies of these items (1,2,4). from 37% of uncarpeted kitchen floors (2) and total and
Data for dry sites show a different picture with fecal coliforms counts of 104 and 102 per sample area,
lower recovery of Enterobacteria and generally, lower respectively, have been recorded (3).
340 BACTERIAL CONTAMINANTS IN RESIDENTIAL ENVIRONMENTS

Bacterial Contamination in Domestic Bathrooms and Toilets and bacillus. In general, the incidence of E. coli and other
Enterobacteria and also S. aureus is low at these sites and
Data are available for 25 bathroom and toilet sites,
contamination levels are also low.
as shown in Table 3. As with sites in the kitchen, the
influence of moisture on the occurrence of Enterobacteria
Sponges, Rags, and Other Fabrics
and on high levels of contamination is also observed
at bathroom and toilet sites. Highest counts have been A number of bacterial surveys of household sponges,
observed at diaper buckets, face cloths, cleaning cloths, rags, and other fabrics have been published over the last
bath surfaces, bath and basin drain rim areas, and drain 30 years. It has long been established that dishcloths, tea
pipes. Interestingly, toilet water samples generally yield towels (6) and hand towels (7) can support high levels of
low counts, indicating the effectiveness of toilet flushing bacterial contamination.
as a means of reducing counts here. Escherichia coli and Bacterial counts, carried out by both contact plate and
other Enterobacteria were most often isolated from the serial dilution methods on reusable kitchen rags, have
potty, the diaper bucket, the toilet brush, toilet water and shown that such items are contaminated on average with
toilet bowl, bath surface, bath and basin drain rim areas, more than a billion bacteria per rag after only one day’s
and face cloth. One study reported finding S. aureus on normal domestic use (8). After further use, counts can
44% of towels and 20% of bathroom floors (1). reach more than 10 billion per rag. Different bacterial
species have been isolated from sponges and dishrags (9)
Bacterial Contamination in General Living Areas including species of Salmonella.
Investigations have shown that soiled clothes are heav-
There is little published in the literature on bacterial
ily contaminated with potentially pathogenic microbes
contamination at general living area. A total of 11 sites
(10,11). Various studies have looked at the potential for
have been reported from only two studies (1,2), comprising
biocontaminants to survive the domestic laundry wash
the following: living rooms (carpet, curtains, upholstery,
processes. One study (12) described the spread of staphy-
table, window sill, stair rail, and telephone mouthpiece)
lococcal skin infections among families sharing laundry
and bedrooms (bed sheets, blankets or duvets, carpet, and
facilities in a closed student community. The machine
dressing table).
wash temperatures of up to 149 ° F/65 ° C were found to be
The predominant organisms at sites in living rooms
inadequate for disinfection. A review (13) of the changes
and bedrooms are species of pseudomonads, micrococci,
that have been made to household laundry practices in
recent times indicates the impact that these changes have
Table 3. Sites and Surfaces had on textile cleaning and hygiene assurance. Measures
Sampled in Domestic Bath- to reduce the environmental impact of household wash-
rooms ing machines have stressed the ability to reduce bacterial
contamination to safe levels.
Wet Sites

Toilet bowl PATHOGEN SEARCHES

Toilet water
Toilet brush A number of surveys have incorporated isolation meth-
Diaper bucket
ods targeted at specific primary and opportunistic
Basin surface
Bath surface
pathogens. Further, some of the organisms isolated in
Bath and basin drains baseline surveys (1–5) may be considered as pathogens,
Bath and basin drain rim area although they were not in that category at the time
Facecloth of sampling. Pathogens isolated in the baseline surveys
Toothbrush are S. aureus, Salmonella spp, P. aeruginosa, Bacillus
Bathroom cleaning cloth cereus, Listeria monocytogenes, Yersinia enterocolitica,
and Campylobacter.
Dry Sites Species of Listeria have been detected in 47.4%
of 213 houses investigated (14) with dishrags and the
Hand Contact Surfaces
area around the bath drain being most frequently
Toilet seat contaminated. Listeria monocytogenes was present in
Toilet handle
21.1% of homes. Highest numbers (104 per object)
Toilet door handle
Potty
were found in dishrags and washing-up brushes. Lower
Basin and bath/faucet handles numbers (up to 103 per object) were obtained from kitchen
Soap dish sinks, refrigerator vegetable compartment samples, and
Towels toothbrushes.

General Surfaces
CROSS-CONTAMINATION STUDIES
Floors
Countertop
Drapes For an infection to result from contact with contaminated
Window sill environmental surfaces, a number of criteria must be
in place. These include the presence of an infectious
 BACTERIAL CONTAMINANTS IN RESIDENTIAL ENVIRONMENTS 341

Pathogenic organism
Bacteria Fungi
Virus Parasite

Higher risk host Source / Reservoir

Very young Pregnant Infected person Surfaces
Shock or trauma Elderly Animals or pets Inanimate objects
Immunocompromised Chronic disease Food Equipment

Means or mode of transmission

Direct contact Indirect contact

Figure 1. Factors necessary for the transmission of infectious microorganisms via contaminated surfaces.

agent, an environmental source or reservoir for pathogenic ILLNESS STUDIES

contaminants, a means or mode of transmission, and
the presence of a host. Although anyone may potentially Studies in homes with active Salmonella infections
acquire an infection from a contaminated surface, more indicate that environmental contamination may play a role
often than not, the host is likely to be characterized as in the development of Salmonella infections, especially
at higher risk for infection. When linked together, these in young children (19,20). Infecting Salmonella serotypes
criteria increase the probability that disease will occur via have been isolated not only from human occupants,
contaminated surfaces, as shown in Figure 1 (15). but also from pets, kitchen sinks and other kitchen
A number of experiments designed to assess the poten- surfaces, floors (19,20), and from dishrags (personal
tial for cross-contamination in the domestic environment communication). In addition, serotypes of Salmonella
similar to those that have caused infection in the home
have been performed in homes. It should be noted that
have been detected in biofilms under the toilet rim several
many other experiments on cross-contamination have been
weeks after the infectious outbreak occurred (21). An
performed in laboratories, under conditions that mimic the
outbreak of Staphyloccus aureus skin infections among
home environment and that these are not reported here.
families sharing laundry facilities demonstrated the
Such studies have been reviewed by Scott and cowork-
potential cross-contamination risk posed via household
ers (15). fabrics (22).
The microbiological hazards of household toilets
droplet production was assessed in the mid 1970s (16).
Detection of seeded bacteria and viruses on surfaces in DISINFECTANT BENEFIT STUDIES
bathrooms after flushing indicated that such organisms
had become aerosolized and may present an aerosol With the growing interest in the positioning of disinfectant
infection hazard in older toilet designs. Subsequent studies products in the home, a number of studies have attempted
of domestic toilets and bathrooms have not detected large to evaluate the potential benefit of disinfectants and
disinfection processes in reducing microbial contamination
numbers of toilet contaminants on surrounding surfaces,
at environmental surfaces in the home. This is an area of
suggesting that under normal domestic circumstances,
consumer-related microbiology that is likely to continue
flushing is an effective means of removal of bacterial
to develop in order to fulfill requirements for product
contaminants.
development, testing, and support.
The potential for cross-contamination in kitchens
In-use tests performed in homes show that the
resulting from the preparation of frozen chickens has been targeted use of household disinfectant products on a
examined (17). Raw chickens were seeded with E. coli range of surfaces can produce substantial reductions in
K12 and prepared in 60 kitchens. Cross-contamination bacterial contamination (3,5,18,23) but that the maximum
occurred in a high proportion of the kitchens examined and protection afforded by disinfection is relatively brief;
contaminants were found on surfaces such as chopping three to six hours after disinfection, contamination
boards, dishrags, and sinks, even after rinsing and levels are only marginally less than those observed
washing-up actions occurred. The routine preparation of at pretreatment (23), indicating that the timing of
chicken that is naturally contaminated with Salmonella disinfectant application is a critical factor. In addition,
and Campylobacter spp. has been found to cause wide results for the use of detergent and hot water cleaning
dissemination of these organisms to hand and food contact at fixed surfaces such as counters show that reductions
surfaces throughout the home kitchen (18). in contamination levels cannot be reliably achieved (18)
342 BACTERIAL CONTAMINANTS IN RESIDENTIAL ENVIRONMENTS

and that there may even be an apparent increase in cross-contamination to other crucial surfaces such
contamination due to surfactant or mechanical break-up as high-risk foods or the hands.
and redistribution of cell aggregates (23).
In addition to simple disinfection benefit studies, the Effective hygiene practices should include steps to
data on bacterial contamination and cross-contamination limit the potential transfer of pathogenic microbes from
in the home is now being used to develop hygiene practice these areas to other more sensitive surfaces such as
models for surface cleaning and disinfection, using similar food and hands. This is especially important when
concepts to those employed in the food industry, using there are people present who are at higher risk for
the hazard analysis and critical control point concept infection (i.e., the young, the elderly, and those who
(HACCP) (15). are immunocompromised through existing illness and/or
treatment) or when there is potential for increased levels of
environmental contamination, such as might occur during
FUNGAL CONTAMINATION incidents of enteric infection in the home. Encouraging
the concept that home hygiene is a series of interrelated
There have been a small number of studies investigating practices and procedures all based on the same underlying
the fungal species that are present in the indoor air of microbiological principles allow the opportunity for a
houses (24–27). Species of Penicillium, Cladosporium, and rational approach based on risk assessment (15). The
Aspergillus are predominant and there is considerable hygiene processes used in the home include the use of
variation concentrations between houses. Studies have soap and detergents, the use of heat (especially for cooking
indicated a link between the presence of both airborne and washing), the use of drying, the use of mechanical
fungi and airborne bacteria in the home with the incidence action as part of the washing process, and the use of
of respiratory allergies such as asthma (28,29). In addition chemical disinfectants. An effective hygiene strategy for
to generally damp conditions, factors such as textile the home targets opportunities to prevent or reduce the
floor coverings and indoor storage of organic household risk of cross-contamination using effective processes and
waste have been linked to the presence of increased an understanding of best practices.
levels of fungal markers in homes (30). Air-conditioning
equipment, humidifier reservoirs, dehumidifier drip pans,
and showerheads can all harbor molds and mildews. (For CONCLUSION
more details, see FUNGI AND INDOOR AIR.)
The study of biocontaminants in the home is a relatively
new field but one that is likely to continue to expand,
VIRAL CONTAMINATION as scientists try to better understand the impact that
biocontamination in our immediate living environment
There are a number of published studies that indicate has on our overall health. In an era that has already
that various viral species survive for significant periods been described as ‘‘postantibiotic,’’ there is a growing
on dry surfaces and that contaminated surfaces can play indication that more emphasis will be placed on effective
a role in virus transmission in settings such as day-care hygiene practice in our daily life as a means of reducing
centers, as reviewed by Scott and coworkers (15). However, the potential for infection. But in order to develop
at this time, there are no published studies indicating the effective environmental hygiene practices, we need a
prevalence of viral species at environmental sites and better understanding of the nature of the microbial
surfaces in the home. contamination in the environment. At this time, there
is a body of data concerning bacterial contamination on
surfaces, however, the data are not yet comprehensive
HYGIENE PRACTICES IN THE HOME
and there are many gaps in the knowledge base. For
example, there is still little information available on the
It has been proposed that there are three general
indoor air quality of domestic dwellings or on the role that
categories of sites and surfaces in the home where the
the many commensal bacterial species may play in cell-
risks of bacterial contamination and cross-contamination
mediated disease and allergic responses. Further, there is
are highest (2). These are known as
even less information about the presence and impact of
fungi and algae in domestic settings. There is almost no
Reservoirs (toilets, diaper buckets, sinks, drains, etc.) published data on virus sampling in homes. Viruses are a
where bacterial contamination is high and there is major cause of domestic infections, and data on the role
often the potential for bacterial multiplication. of inanimate surfaces in the transfer of virus infections
Reservoir-disseminators (wet cleaning utensils such would be invaluable in further characterizing the risk
as rags, sponges, and mops) where bacterial posed by contaminated surfaces in the home. To expand
contamination is high and there is the potential the knowledge base in the emerging field of domestic
for direct transfer of this contamination to other microbiology, further research is needed to complete the
surfaces whenever these items are used. picture of biocontamination in the home, to develop and
Hand and food contact surfaces (kitchen counters, refine risk analysis for environmental surfaces in the
cutting boards, faucets, handles, laundry, floors, home, and to link the potential benefits of hygiene practice
etc.) where there are generally lower levels of in the home, either directly with reduced infections or with
contamination but there is a constant potential for other meaningful indicators.
 BACTERIAL PHYTOSTIMULATORS IN THE RHIZOSPHERE 343

BIBLIOGRAPHY BACTERIAL PHYTOSTIMULATORS IN THE

RHIZOSPHERE
1. J. E. Finch, J. Prince, and M. Hawksworth, J. Appl. Bacteriol.
45, 357–364 (1978). SAUL BURDMAN
2. E. Scott, S. F. Bloomfield, and C. H. Barlow, J. Hyg. Camb. DANIEL KADOURI
92, 279–293 (1982). EDOUARD JURKEVITCH
3. P. Rusin, P. Orosz-Coughlin, and C. Gerba, J. Appl. Micro- YAACOV OKON
biol. (1988). The Hebrew University of
4. A. P. Speirs, A. Anderton, and J. G. Anderson, Int. J. Envi- Jerusalem
ron. Health Res. 5, 109–122 (1995). Rehovot, Israel
5. K. L. Josephson, J. R. Rubino, and I. L. Pepper, J. Appl.
Bacteriol. 83(6), 737–750 (1997). From an agricultural point of view, ‘‘soil’’ refers to the
6. J. G. Davis, J. R. Blake, and C. M. Woodall, Med. Off. 120, broad environment from which plants obtain mechanical
29–32 (1968). and nutritional support (1). That part of the soil, which is
7. J. G. Davis, Med. Officer 111, 89–95 (1970). adjacent to the plant root and therefore under its influence
8. K. Redway et al., A Study of the Comparative Performance is generally denominated as the rhizosphere. As a result
and Hygiene of Kitchen Cloths and Paper Towels, 1996. of their activity, living roots secrete a wide variety of
9. C. E. Enriquez et al., Dairy, Food, Environ. Sanitation 17(1), compounds that alter the physical and chemical properties
20–24 (1997a). of their surrounding environment.
10. G. M. Ridenour, A Bacteriological Study of Automatic Clothes Defining the rhizosphere is still a matter of debate,
Washing, National Sanitation Foundation, Ann Arbor, Mich., and even more of choice. Throughout the history of
1952. rhizospheric microbiology research, investigators have
11. E. McNeil and E. A. Choper, Soap Chem. Special. 60, 51–100 been arbitrarily determining what they consider to be the
(1964). rhizosphere in their studies. In broad practical terms, the
12. R. B. Kundsin, Clin. Med. 3, 27–29 (1966). rhizosphere has been considered to be the soil attached
13. P. M. J. Terpstra, Int. Biodeter. Biodegrad. 41, 169–175 to the roots when extracting them from their growth
(1998). medium, whereas a more restricted view regards the
14. R. R. Beumer et al., Epidemiol. Infect. 437–442 (1996). rhizosphere as the relatively small amount of soil particles
15. E. Scott, D. J. Gaber, and T. M. Cusack, Chemical disinfec- that remain attached to the roots after they are washed
tion of microbial contaminants on surfaces, in S. S.Block, ed., from the residual ‘‘bulk’’ soil. This aspect becomes even
Disinfection, Sterilization, and Preservation, 5th ed., Williams more complicated when considering that the qualitative
& Wilkins, 2001. and quantitative distribution of the secreted compounds
16. C. P. Gerba, C. Wallis, and J. L. Melnick, Appl. Microbiol. from the roots is largely influenced by plant species
30(2), 229–237 (1975). and development, environmental conditions, and location
17. J. C. De Wit, G. Broekhizen, and E. H. Kampelmacher, J. along the roots (2). Hopefully, the increasing exploitation
Hyg. Camb. 82, 27–32 (1979). of modern molecular tools by microbial ecologists will be
18. T. A. Cogan, S. F. Bloomfield and T. J. Humphrey, Lett. Appl. helpful in delimiting the rhizosphere ‘‘borders.’’
Microbiol. 29, 354–358 (1999). In the rhizosphere, bacteria can be enriched by a factor
19. M. Van Schothorst, J. Huisman, and M. Van Os, Nederlands of 100 as compared with their number in the surrounding
Tijdschrift voor Geneeskunde 122, 1121–1125 (1978). soil (3). The main factors responsible for this ‘‘rhizosphere
20. G. E. Schutze et al., Pediatrics 94, 1018–1020 (1999). effect’’ are the exudation of photosynthetically derived
21. J. Barker, and S. F.Bloomfield, J. Appl. Microbiol. (in press). carbon sources (sugars, amino acids, organic acids,
22. R. B. Kundsin, Clin. Med. 3, 27–29 (1966). phenolic compounds) and the sloughing off of root cap
23. E. Scott, S. F. Bloomfield, and C. G. Barlow, J. Hyg. Camb. cells and older sections of the cortex (3,4). In turn,
92, 193–203 (1984). these secretions affect bacterial activity and, therefore,
24. C. A. Hunter et al., Int. Biodeter. 24, 81–101 (1988). the microbial composition of the rhizosphere. Thus, the
25. D. P. Strachan et al., Thorax 45, 382–387 (1990). bacterial population of the rhizosphere generally differs
from that present in soil areas not influenced by the
26. C.-S. Li, Y.-M. Kuo, and L.-Y. Hsu, Environ. Int. 20, 179–189
(1994). roots, not only quantitatively but also in its qualitative
composition and distribution (3), although in some studies
27. D.-W. Li and B. Kendrick, Mycologia 87(2), 190–195 (1995).
no differences were observed (5). Table 1 summarizes the
28. M. H. Garrett et al., Clin. Exp. Allergy 28, 459–467 (1988).
effects caused by roots on the rhizosphere.
29. B. Flannigan et al., Soc. Appl. Bact. Symp. Ser. 20, 61s–74s
Among the bacteria living in the rhizosphere, some are
(1991).
able to affect plant growth, either positively or negatively.
30. I. M. Wouters et al., Appl. Environ. Microbiol. 66(2), 627–631
Rhizospheric bacteria that favorably affect plant growth
(2000).
and yield of many commercially important crops are
designated plant growth promoting rhizobacteria (PGPR).
PGPR are able to exert positive effects on plants by
BACTERIAL PATHOGENS IN WASTE various mechanisms. These include those that can directly
STABILIZATION PONDS. See WASTEWATER cause plant growth promotion (PGP) by improving seed
STABILIZATION PONDS germination, root development, mineral nutrition, and
344 BACTERIAL PHYTOSTIMULATORS IN THE RHIZOSPHERE

Table 1. Factors Derived from the Activity of Living Roots inoculants of PGPR that positively affect plant growth
That Influence the Composition of Rhizosphere Microbial by direct mechanisms (phytostimulators). Rhizospheric
Communities bacteria capable of improving plant growth by biological
Root Activity Factors control of soilborne diseases, and those able to invade
the plant and develop nitrogen-fixing symbioses such as
Secretions/Exudates Release of low- or high-molecular Rhizobium and Frankia, will not be discussed. Rhizobia
weight compounds used as will be mentioned only in relation to their interactions
nutrients with PGPR in the growth promotion of legumes.
Effects on soil solution pH and
conductivity
Release of extracellular enzymes MECHANISMS OF ACTION OF PHYTOSTIMULATING
Lysates Release of compounds from BACTERIA
aging/dead epidermal and
cortical cells Biological Nitrogen Fixation (BNF)
Synthesis of mucilage Changes in hygroscopicity There are many free-living nitrogen-fixing rhizobacteria
(mucigel) contributing fixed nitrogen from the atmosphere to plants,
Increased root area available for including agronomically important crops (7,8). Bacterial
bacterial colonization
diazotrophs in the rhizosphere belong to a wide variety
Protection of rhizosphere bacteria
from bacterial predators
of genera (Table 3), the most investigated of which have
been Azotobacter and Azospirillum, although many others
Respiration Effects on air composition
have been reported to be associated with tropical grasses
and rice (7).
Nitrogen fixation by rhizospheric bacteria is sometimes
Table 2. Modes of Action of Plant Growth Promoting referred to as associative symbiosis, and has been
Rhizobacteria (PGPR) extensively estimated by a wide variety of methodologies,
Effects Mechanisms
including long-term nitrogen balances in soils, estimation
of nitrogen contents by the Kjeldahl method, acetylene
Direct effects — Production of plant growth regulators reduction assay, and the 15 N isotope dilution method (9).
Phytostimulation (PGRs): auxins, gibberelins, However, to date, quantification and interpretation of
cytokinins the results have been difficult and controversial. It is
Biological nitrogen fixation estimated that the contribution of nitrogen fixation by
(biofertilizers) free-living rhizobacteria in the rhizosphere of crops such
Indirect effects — Production of antibiotics and as wheat, sorghum, and maize is in most cases on the
Biological control bacteriocins order of 1 to 10 kg N ha−1 yr−1 (10). It may be higher in
Production of cyanide acid (HCN) rice because of the activity of nitrogen-fixing blue-green
Production of siderophores algae, and in some tropical grasses (7). Although positive,
Competition for colonization sites and
these amounts are, in fact, of minor importance when
nutrients
Systemic acquired resistance (SAR) of
compared with the amounts of nitrogen fertilizers applied
the plant against pathogens in modern agriculture, which are on the order of 150 to
Production of chitinases and 250 Kg N ha−1 yr−1 .
glucanases and hydrolysis of fungal Many agriculturally important grasses are associated
cell walls with endophytic diazotrophic bacteria (Table 3). They have
mainly been isolated from plants in which significant BNF
has been demonstrated, particularly in some Brazilian
sugarcane and rice cultivars (6,11) and in Kallar grass in
water utilization (phytostimulation). Others affect plant
water logged salty soil in Pakistan (12).
growth via indirect effects that involve suppression of
bacterial, fungal, and nematode pathogens (biological
Production of Plant Growth Regulators (PGRs)
control) (6). The main modes of action of PGPR are
summarized in Table 2. Evaluation of the contribution of BNF to plant growth is
Growing interest in sustainable agriculture has encour- complicated by the fact that most diazotrophic rhizospheric
aged PGPR research. Many groups are currently investi- bacteria and endophytes (Table 3), and many other soil
gating possible modes of action of these bacteria at the microorganisms, produce a wide variety of PGRs such as
molecular level using DNA recombinant techniques. How- auxins, cytokinins, gibberelins, ethylene, and abscisic acid
ever, because of the complexity of the rhizospheric environ- that may alter plant growth (8,13).
ment and the infinite possibilities of plant-soil-microbial A common observation following inoculation with
interactions, in most cases, the suggested mechanisms bacteria from the genus Azospirillum is an increase
have not yet been fully elucidated. Nevertheless, substan- in the number of root hairs and lateral roots. This
tial advances have been made in recent years, as reviewed leads to an improved uptake of water and nutrients,
later. resulting in significant increases in crop yields (14).
In this chapter, we will focus mainly on the ecology, Although it has been proposed that changes in root
modes of action, and utilization of agrobiotechnological morphology by Azospirillum are caused mainly by the
 BACTERIAL PHYTOSTIMULATORS IN THE RHIZOSPHERE 345

Table 3. Phytostimulating Bacteria and Their Modes of Action (6–8)

Organisms Mechanisms

Rhizosphere diazotrophs
Azospirillum brasilense Mainly through production of plant growth regulators (PGRs), mostly auxin.
Azospirillum lipoferum Proliferation of root hairs and branching. Improved mineral and water uptake.
Increased nodulation and biological nitrogen fixation (BNF) (by Rhizobium) in
legumes. BNF: 5 to 10 kg N ha−1 yr−1 in Gramineae. Increased crop yields in
inoculated fields at intermediate fertilizer levels: 10 to 30%
Azotobacter chroococcum BNF: 5 to 10 kg N ha−1 yr−1 . Yield increases of 10 to 20% (rhizosphere habitat
controversial)
Azotobacter paspali BNF: increased N-content of about 10% in Paspalum notatum cv batatais.
Production of PGRs. Proliferation of root hairs and branching
Pseudomonas stutzeri BNF in rice
Klebsiella pneumoniae and Klebsiella oxytoca BNF in maize and rice. Root hair proliferation in Poa pratensis
Enterobacter BNF in rice
Burkholderia BNF in maize

Endophytic diazotrophs
Acetobacter diazotrophicus BNF in sugarcane (for some cultivars, 50 kg N ha−1 yr−1 ), coffee, and sorghum
Herbaspirillum seropaedica BNF in sugarcane
Azoarcus BNF in Kallar grass and rice
Rhizobium leguminosarum bv trifolii Plant growth promotion in rice

Nondiazotrophic
Bacillus Promotion of nodulation and BNF (by Rhizobium) in chickpeas and common bean
Serratia liquefaciens and S. proteamaculans Promotion of nodulation and BNF (by Bradyrhizobium) in soybean
Pseudomonas fluorescens Promotion of nodulation and BNF (by Bradyrhizobium and Rhizobium) in
soybean and peas
Pseudomonas putida Promotion of nodulation and BNF (by Rhizobium) in common bean. Improved
phosphate uptake
Pseudomonas putida GR 12.2 Root elongation. ACC deaminase. Production of PGRs

bacterial production of indole-3-acetic acid (IAA) (14,15), this enzyme participates in root growth promotion by this
production of gibberellins and cytokinins may also be bacterium (17).
involved (14). Most beneficial bacteria synthesize IAA
via the indole-3-pyruvate pathway in which tryptophan Beneficial Effects of PGPR on the Rhizobium-Legume
is first transaminated to indole-3-pyruvic acid, then Symbiosis
decarboxylated to indole-3-acetaldehyde, which is oxidized The effects of PGPR on root hair proliferation and root
to IAA. Genes encoding indole-pyruvate carboxylase (ipdC) branching, mineral, and water uptake and growth can also
have been isolated from Azospirillum (16). An ipdC Tn5- be observed in legumes (8). Positive effects were observed
mutant of Azospirillum brasilense that shows 10% residual for several Azospirillum-inoculated legumes in the field
IAA production in comparison with the wild type has been and greenhouse, and under gnotobiotic conditions. The
obtained and shown to possess reduced ability to promote increases in dry-matter production and nitrogen content
root hair proliferation (15). were attributed to early nodulation, increased nodulation,
Involvement of ethylene in plant growth promotion and higher nitrogen fixation rates, and to a general
has also been investigated. It was demonstrated that improvement in root development (18).
a small number of soil bacteria contain the enzyme 1- Rhizobium infection occurs via the formation of
aminocyclopropane-1-carboxylate (ACC) deaminase. This infection threads in the root hairs. Stimulation of
enzyme hydrolyzes ACC, the immediate precursor of ethy- nodulation following inoculation with Azospirillum may
lene in plants, to yield ammonium and α-ketobutyrate. be due to the differentiation of a greater number of
Bacteria possessing ACC deaminase activity may there- epidermal cells into infectable root hairs (19). In addition,
fore grow using ammonium as the sole nitrogen source. A. brasilense was shown to cause an increase in the
The beneficial rhizobacterium Pseudomonas putida strain secretion of nod gene inducing flavonoids by common
GR 12–2 stimulates root elongation of different plants. bean and alfalfa roots, resulting in the appearance of
This bacterium was found to contain ACC deaminase. a greater number of upper nodules following coinoculation
Mutants lacking ACC deaminase activity were not able to with Rhizobium and Azospirillum in comparison with
promote root elongation of canola seedlings, implying that inoculation with Rhizobium alone (20,21).
346 BACTERIAL PHYTOSTIMULATORS IN THE RHIZOSPHERE

Several works have been published on the interaction Metabolic Versatility

of other rhizobacteria (e.g., P. putida, Pseudomonas
In aerobic environments, most bacteria are highly
fluorescens, Bacillus polymyxa, Enterobacter and Serratia)
versatile and can use a large variety of organic
with the Rhizobium-legume symbiosis (Table 3). Despite
compounds as carbon and energy sources. One of the
data supporting positive effects on nodulation, nitrogen
best-characterized examples of metabolic versatility are
fixation, and yield attributes, little is known about the
bacteria belonging to the genus Azospirillum. These are
mechanisms involved in these interactions, but it is likely
free-living nitrogen-fixing rhizobacteria found in close
that in some cases they are similar to those described for
association with plant roots, where they exert beneficial
Azospirillum (8).
effects on plant growth and yield of many agronomically
important crops (14). A variety of carbon sources may be
Phosphorus Solubilizers
used by azospirilla, (Table 4) for example, salts of organic
Phosphate deficiency in crops can be diminished by the acids such as malate, succinate, lactate, and pyruvate, and
use of bacteria that act directly as phosphate solubilizers various sugars or amino acids. Under aerobic conditions,
in the rhizosphere, and indirectly by stimulating root oxygen is the preferred respiratory electron acceptor.
excretion of organic acids and/or mycorrhizal associations. However, in the absence of oxygen, most strains can use
Enhanced excretion of organic acids results in increased nitrate, nitrite, and nitrous oxide as electron acceptors of
phosphate solubilization and hence in increased phosphate the respiratory chain (24,25).
uptake (22). In addition to the monomeric or dimeric carbohydrates,
Insoluble inorganic phosphorus is largely unavailable to A. brasilense and A. lipoferum can degrade polymers
plants, but many microorganisms can bring the phosphate such as xylan, starch, cellulose, and carboxymethyl
into solution. Species of Pseudomonas, Mycobacterium, cellulose, which are available in the rhizosphere of host
Micrococcus, Bacillus, and Flavobacterium are active in plants (26,27). Furthermore, A. irakense has been found
that conversion. These bacteria grow in media with to grow with pectin as sole carbon source, and some
Ca3 (PO4 )2 and apatite as sole phosphorus sources. Not Azospirillum spp. are able to grow on methanol or other
only do the microorganisms assimilate the element, C1 compounds (25).
they also render a large portion of it soluble. The Relatively high growth rates and the ability to
many phosphate-dissolving microorganisms found near effectively use a wide range of organic compounds seem to
roots may appreciably enhance phosphate assimilation by be important factors for rhizosphere competence. However,
higher plants (22). it was found that rhizosphere and soil bacteria do not
Inoculation with A. brasilense was shown to signifi- significantly differ in their rates and extent of growth
cantly enhance H2 PO4 uptake by maize in hydroponic in the presence of exudates from soybean seeds and
systems, on the order of 30 to 50% over noninoculated plants (28). In a large study searching for determinants of
controls. In field-grown sorghum and wheat, the increase rhizosphere competence, Jjemba and Alexander (29) could
was in the range of 10 to 30%. In that case, the increases not find a correlation between the growth rate of some
in phosphorus uptake could have resulted from increased of the tested bacteria and their ability to colonize the
root respiration (10). rhizosphere.

Accumulation of Storage Material

BACTERIAL FEATURES FOR RHIZOSPHERE COMPETENCE
Microorganisms can accumulate a variety of storage
Inoculation with PGPR is becoming a common practice compounds such as glycogen, polyphosphate, and sulfur
in agriculture. It is therefore important to understand compounds. It is thought that the accumulation of these
what microbial traits are involved in rhizosphere com- storage materials enhances competition and survival of
petence, which is the relative root-colonizing ability of a bacteria under stress.
rhizobacterium (23). The role of some intuitively defined A wide variety of microorganisms are known to produce
properties has been not fully demonstrated, with contra- intracellular energy and carbon storage compounds, gener-
dictory results found in the literature. In this section, the ally described in the past as being poly-β-hydroxybutyrate
focus will be on those bacterial features that have received (PHB). In recent years, it has been found that in most
the greatest attention. cases, the polymers are polyhydroxyalkanoates (PHAs)

Table 4. Utilization of Different Carbon Sources by Azospirillum spp. (25)

Carbon Source A. lipoferum A. brasilense A. amazonense A. halopraeferens A. irakense

D-Glucose + − + − +
Glycerol + + − + −
D-Mannitol + − − + −
Pectin − − − − +
D-Sorbitol + − − − −
Sucrose − − + − +

Note: + indicates utilization of the carbon source.

 BACTERIAL PHYTOSTIMULATORS IN THE RHIZOSPHERE 347

comprising copolymers containing different alkyl groups of siderophores, with organisms producing one or more
at the β-position (30). different kinds. Siderophores are present in the soil and
It has been proposed for diverse ecological systems in the rhizosphere, and their utilization is important for
that the accumulation, degradation, and utilization of the competitive growth of most microorganisms.
PHAs by several bacteria may favor their establishment, Fluorescent pseudomonads produce siderophores
proliferation, survival, and competition, especially in termed pseudobactins. They seem to play an important
competitive environments in which carbon and energy role in rhizosphere competition, enhancing their capability
sources are limiting factors, such as those encountered in as biocontrol agents and as growth promoters (39).
soil and in the rhizosphere (31). Scher (40) gave indirect evidence of the role of
In situ studies of the role played by PHAs on bacteria in siderophores in competing for iron in Fusarium-
the rhizosphere are very scarce, because of the lack of an suppressive soils. In that work, biocontrol of Fusarium
accurate methodology for measuring their levels. Although wilt by P. putida was lost when available iron was
mutants unable to synthesize PHAs have been isolated added, whereas when a Fe3+ -chelate with a high bind-
and genetically engineered (32–34), most PHA-negative ing constant was added, biological control was enhanced.
mutants were examined for their effects on symbiosis and It was also shown that a siderophore-negative mutant
cellular metabolism, whereas their role in rhizosphere of P. putida loses its suppressive property against car-
competence was not addressed. nation wilt (41). In potato stem cuttings treated with a
Electron microscopy studies of the rhizosphere have siderophore-negative mutant with and without the par-
shown many bacteria containing PHAs. The C:N ratio ent strain, the mutant strain multiplied on the cuttings
of the rhizosphere is estimated at 20. This ratio favors to higher levels when the parent strain was present (42).
PHA accumulation in Azospirillum and Azotobacter, and This indicates that the siderophore produced by the par-
in other rhizospheric bacteria in culture (31). ent strain made iron available to the mutant strain and
The possible functions of PHAs in A. brasilense played an important role in its development.
strain Cd were investigated, with PHB being the only Azospirillum lipoferum was shown to produce a
PHA accumulated by this bacteria (31,35). Degradation catechol-type siderophore under iron-starved conditions.
and synthesis of the polymer occurred in a biphasic In addition to their established role in iron transport,
pattern under starvation conditions and were affected the siderophores exhibited antimicrobial activity against
by the PHB content of the cells. Under starvation, various bacterial and fungal isolates, suggesting that the
the survival and respiration rate of bacteria containing high-affinity iron uptake system of A. lipoferum may be
about 40% PHB (of total dry weight) were higher of use in the competition among soil microbes for access
than in bacteria containing 5% PHB. Polymer-rich cells to available iron, thereby enhancing its survival in the
fixed atmospheric nitrogen in the absence of exogenous rhizosphere (43).
carbon and combined nitrogen. Under stress and adverse
conditions, such as UV-irradiation, desiccation, and Production of Antibiotics
osmotic pressure, PHB-poor cells died more rapidly than
One of the mechanisms used by PGPR, which helps
PHB-rich cells. It was concluded that PHB might provide
them ensure an ecological edge over other root-colonizing
A. brasilense with the ability to survive under conditions
microorganisms, is the excretion of antibiotic substances,
of starvation and stress, by serving as a sole carbon
which exclude or reduce pressure from competitors on
and energy source (35). Furthermore, in many cases,
colonized niches. A large array of different compounds
PHB accumulation is accompanied by flocculation (36),
showing antibiotic activity against plant pathogens have
which renders cells highly resistant to desiccation.
been isolated from PGPR (44–46). Until now, the bulk
These properties could conceivably be important for the
of the research on antibiotics was focused on their
proliferation of Azospirillum, providing a competitive edge
role in biocontrol, with their significance in rhizosphere
over other microorganisms in the rhizosphere (31,35).
competition not being specifically examined.
In Rhizobium bacteroids, PHAs are mobilizable,
Kloepper and Schroth (47) demonstrated that five
energy-yielding reserves that provide endogenous carbon
PGPR fluorescent pseudomonads strains that exhibited
substrates for support of nitrogen fixation when exogenous
antibiosis against the soft rot pathogen Erwinia caro-
carbon is not available. They may also serve as a carbon
tovora, and several other bacteria isolated from the rhi-
and energy supply outside the bacteria (37).
zosphere of radish plants, increased plant growth, but
mutants that lacked antibiotic production had no effect.
Production of Siderophores
The parent strain, but not the mutants lacking antibiotic
Despite the fact that iron is only a minor nutrient for production, reduced the number of gram-positive bacteria
the growth of most soil microorganisms, it is essential. and fungi in the rhizosphere.
Although it is abundant in terrestrial habitats, its It was shown that the higher the number of P.
availability is limited as a result of the low solubility of the fluorescens on the root, the lower the number of take-
ferric ion (1). Therefore, many microbes have developed all lesions found. When an antibiotic-negative mutant
pathways to obtain it. Many microorganisms overcome of P. fluorescens was tested, there was no suppression
this iron unavailability by producing siderophores, specific of the disease (48). Mazzola and coworkers (49) found that
low molecular weight Fe3+ -chelates, and a complementary when a phenazine-negative mutant strain of P. fluorescens
uptake system, the biosynthesis of which is regulated by was added together with its parent strain to pasteurized
the level of available iron (38). There exists a wide variety soil to which the take-all fungus had been added,
348 BACTERIAL PHYTOSTIMULATORS IN THE RHIZOSPHERE

populations of both the parent and mutant strain were involved in the first step, and that the second step is
maintained. However, when they were added to untreated mediated mainly by extracellular polysaccharides (54,55).
soil, the phenazine-mutant population declined, whereas In the last decade, outer membrane proteins (OMPs)
the parent strain proliferated, thereby illustrating the have been proposed to play a role as adhesins in invasion
importance of antibiotic production for the establishment and adhesion processes in various gram-negative bacteria.
of an organism in the rhizosphere. The major OMPs of P. fluorescens, Rahnella aquatilis,
However, when mutants of P. fluorescens defective in and A. brasilense were suggested to be involved in
in vitro antibiotic production and the parental strain the adsorption of these rhizospheric bacteria to plant
were coinoculated at a 1:1 ratio, no differences in root roots (56–58).
establishment were found between the strains, suggesting Bacteria able to adhere to the root surface are closer
that in this case, antibiotic production contributed little to to the source of root exudates than are bacteria at
rhizosphere competence (50). some distance from the root. Moreover, adhering cells
are presumably more likely to be transported with the
Resistance to Protozoa extending root. For adhering cells, the root may also serve
as a basis of physical support for bacterial proliferation.
Protozoa are unicellular eukaryotic, generally motile However, studies carried out on root adhesion have been
microorganisms. They usually obtain food by ingesting quite contradictory in supporting the importance of this
other organisms or organic particles and are found in trait for rhizosphere competence.
a variety of habitats, including the rhizosphere. These Chao and coworkers (59) showed that 82% of the
microorganisms are thought to play a major role in bacteria isolated from the rhizosphere exhibit a positive
controlling the population of bacteria residing in the agglutination reaction with pea roots, in contrast to only
soil (51). 30% of bacteria from bulk soil. However, it was suggested
The extent of soil colonization by Rhizobium phaseoli that root attachment by bacteria is, in most cases, a very
was shown to be inversely related to the presence of widely occurring property of low specificity, being greatly
large numbers of bacteria and protozoa. Colonization influenced by the medium ionic composition (60). Studies
of R. phaseoli was improved upon suppression of carried out with single-gene mutants of P. putida, differing
protozoa with protozoa inhibitors (52). Casida (53) added in root attachment ability, showed that some agglutinin-
representatives of several categories of bacteria to soil to negative strains still possess good root colonization
determine which of them might elicit responses from soil ability (61). Recently, Jjemba and Alexander (29) showed
protozoa. The protozoa did not respond to Cupriavidus that the ability of introduced bacteria to colonize the
necator, a potent bacterial predator, or to Micrococcus rhizosphere of soybeans does not correlate with the
luteus, one of its prey species. Cupriavidus necator also capacity of low or high densities of these bacteria to adhere
had no effect on the protozoa, showing that in this case to the roots. The aforementioned examples suggest that the
bacterial and protozoan predators do not interact. The soil ability to attach to the roots is not a crucial determinant
protozoa did not respond to Arthrobacter globiformis or for rhizosphere colonization, especially when considering
to Bacillus thuringiensis. Apparently, resistance of these that only a very small percentage of the cells are firmly
microbes to protozoa enhanced their survival. The addition attached to the roots (62).
of Bacillus mycoides and Escherichia coli caused specific
responses by soil protozoa (53). Motility and Chemotaxis
In summary, addition of bacteria to soil did not
cause a general increase in the numbers of protozoa. In Most of the known bacterial species are flagellated by
other studies, the extent of colonization of a nonsterile means of polar and/or lateral flagella. Flagella provide the
rhizosphere by tested bacteria was not correlated to their motility needed for the bacteria to reach the most favorable
densities in the rhizosphere containing large numbers of niche and to compete with other microorganisms for these
protozoa (29). In view of these findings, it seems that niches. It has been observed that in many cases when
protozoa may reduce the abundance of some species competition is absent, bacteria lose their motility and/or
of rhizospheric bacteria; however, its significance in flagellation ability (63).
controlling the population of bacteria in the rhizosphere is In P. fluorescens, many studies have suggested that
not yet well understood. motility does not affect root colonization following either
seed or soil inoculation (64–66). In contrast, De Weger and
coworkers (67) found that bacterial motility is required for
Root Adsorption and Colonization
root colonization of potato by P. fluorescens. A comparative
Bacterial adherence to plant roots is thought to be an study between a wild type and a nonmotile Tn5 mutant of
important trait in determining rhizosphere competence, P. fluorescens suggested that motility plays a role in the
and this property has been the subject of a large number movement of cells along roots more than in the movement
of studies. from bulk soil toward roots (50).
Attachment to plant roots by some rhizobacteria, Many studies have been carried out on the importance
including rhizobia, seems to occur in two distinguishable of motility for nodulation of legume roots by rhizobia.
steps: a rapid, weak, and reversible adsorption and a Although some of them indicated that motility is
slow, firm, and irreversible anchoring to the plant root not essential for root attachment or nodulation, some
surface (54). In the case of A. brasilense, it was suggested studies based on competition assays among wild-type
that an adhesin closely related to the polar flagellum is and nonflagellated and/or nonmotile mutant strains
 BACTERIAL PHYTOSTIMULATORS IN THE RHIZOSPHERE 349

suggested that motility confers a selective advantage when bacterial flocs can be produced on a large scale and be
competing with nonmotile strains (62,63). easily separated from the growth medium. This is of great
In A. brasilense, horizontal and vertical movements in interest for the production of bacterial inoculants.
different soil types depend mainly on the presence of plant On the basis of experimental evidence, Neyra and
roots and on the availability of water films in soil (68). On coworkers (80) proposed the generation of inoculants
the basis of observation of motile and nonmotile strains containing intergeneric coaggregates of A. brasilense and
of A. brasilense, Bashan and Holguin (69) concluded that Rhizobium leguminosarum biovar phaseoli to improve the
active motility is needed for interroot travel (movement growth of common bean plants.
among neighboring plants) by this bacterium. Despite much evidence indicating that the aforemen-
Besides locomotive properties, adhesive properties tioned traits improve bacterial survival under laboratory
have been attributed to bacterial flagella. As previously conditions, their real impact on bacterial survival in the
indicated, the first step in the root attachment process by rhizosphere has yet to be demonstrated. In addition, these
A. brasilense is mediated by an adhesin closely related to resistant forms are less metabolically active than the veg-
the polar flagellum. Azospirillum brasilense cells lacking etative forms, a fact that should be taken into account
the polar flagellum were unable to adsorb, and purified when preparing bacterial inoculants.
polar flagella adsorbed specifically to wheat roots (55).
Strongly related to flagellation is chemotaxis, which INTERACTIONS OF PGPR WITH BIOTIC AND ABIOTIC
allows bacteria to detect and move toward gradients FACTORS
of different compounds. In the rhizosphere, it may be
advantageous to respond positively to chemoattractants Because soil, the milieu in which roots develop, is
released by plant roots that may increase metabolic rates. inherently heterogeneous, and because of their impact
Some of the most studied PGPR with respect to on their immediate environment (see the introductory
chemotaxis are those belonging to the genus Azospirillum section), beneficial root-associated microorganisms meet a
that exhibit positive chemotaxis to a wide variety of complex habitat. A major component of this habitat is the
compounds such as organic acids, sugars, amino acids, indigenous microbial populations that vary in structure,
and aromatic compounds (70–72). These bacteria also spatial distribution, physiology, and impact on the plant’s
possess a strong aerotactic response toward reduced health. Not one methodological approach for studying
oxygen pressure in the root zone when nitrogen fixation is rhizospheric bacterial communities and their relationship
required (73,74). is necessarily better than another, but each, within its own
Using the β-glucuronidase (GUS) reporter system, bias, offers a different vantage point from which a more
various A. brasilense mutants were investigated for complete and accurate picture of ‘‘life in the rhizosphere’’
their capacity to initiate wheat root colonization at can be reconstructed.
the root hair zones. Only nonflagellated mutants and The physicochemical parameters that bacterial cells
a generally nonchemotactic mutant exhibited strongly encounter in the rhizosphere are important factors in
reduced root colonization ability as compared with the determining their fate. Such parameters, among others,
wild type, thus demonstrating the requirement of bacterial include hydrogen and redox potentials, water tension
motility in the establishment of the Azospirillum-plant and flow, atmospheric composition, viscosity, solution
root association (75). composition, nutrient fluxes, and charged surfaces.
Reliable devices enabling precise acquisition of this type
Conversion into Forms with Increased Resistance Under
of data at the relevant micrometer scale have been
Stress
successfully applied in microbial ecology (81) and are also
Under adverse environmental conditions, some bacteria being applied to study the rhizopheric environment.
are able to convert themselves into forms that are
more resistant than the vegetative form. For instance, Microsensors
Bacillus species, which are known to promote both
A thorough description of microsensor devices can be
phytostimulation and biocontrol effects in a wide range
found in another article of this book. Briefly, these are
of plants (62), are able to form endospores that are highly
mostly microelectrodes responding to specific solutes or
resistant to desiccation and heat.
gases that, owing to their small tip diameter, are able
Species of Azotobacter (e.g., A. chroococcum) are known
to perform measurements at the micrometer level. A
to form heat- and desiccation-resistant cysts endowed with
large variety of microsensors have been developed, such
a long life span. This property has been used to prepare
as ones for measuring pH, temperature, carbon dioxide,
nitrogen-fixing liquid bioinoculants useful for a variety
sulfide, nitrate, nitrite, oxygen, nitrous oxide, organic
of crops (76). Cyst formation was also reported for A.
carbon, and methane (81–83). They have been used to
brasilense and A. lipoferum (77).
study sediments, sludge, and plant rhizosphere, the latter
Cell aggregation is a widespread phenomenon in
mainly under flooded conditions (84–86).
the microbial world. Rhizospheric bacteria such as
Rhizobium, Pseudomonas, Azotobacter, and Azospirillum
Reporter Genes
are able to aggregate and flocculate under both soil and
laboratory conditions (58,78,79). There is evidence that Soil remediation has been an important incentive for
this property positively affects bacterial survival under the development of biosensors, namely, reporter genes,
adverse environmental conditions (77,78). In addition, cloned under the regulation of a promoter sensitive to
350 BACTERIAL PHYTOSTIMULATORS IN THE RHIZOSPHERE

specific deleterious organic or inorganic compounds. The plants alike. Mahaffee and Kloepper (94) investigated
type of genes employed usually yields a colorimetric (lacZ, whether the introduction of a bioluminescence gene
gus, xyl, and lux) or a bioluminescent (gfp) reaction, or in a rhizosphere isolate of P. fluorescens would cause
ice nucleation. This approach is also gaining acceptance detectable changes in microbial community composition.
in rhizosphere studies, with indicator organisms for No such shift could be found by fatty acid methyl ester
iron, oxygen, sucrose, and tryptophan already being analysis of 7,200 colonies isolated over a two-year field
developed (87–89). study.
In contrast, a population shift was observed in the
INOCULATION WITH PGPR IN A DENSE MICROBIAL rhizosphere of Lotus corniculatus plants engineered to
BACKGROUND produce opines; increased levels of bacterial strains able
to specifically grow at the expense of a particular opine
Bacterial inoculants, phytostimulators, or biocontrol were detected in the rhizosphere of the transgenic plants
agents, insomuch as they colonize the rhizosphere and producing that opine as compared with the wild-type
influence plant growth and health, certainly affect plants (95).
indigenous populations. Such effects may stem from
direct competition for resources and/or from products MOLECULAR METHODOLOGY FOR TRACKING
of metabolic activity of the introduced microbe, or from INTRODUCED PHYTOSTIMULATORS
indirect factors such as changes occurring in the root
system and in exudate release following inoculation. In Although most studies have been conducted using
turn, these indigenous communities may also impede or classical, culture-based techniques, culture-independent
promote the establishment of the introduced microbe. modern molecular techniques are being implemented to
Although few such studies have been performed, this track introduced bacteria, to monitor their physiological
knowledge is important for increasing the performance of status in situ, to understand their relationship with the
inocula and for evaluating the impact of the introduction surrounding bacterial communities and to analyze these
of exogenous bacteria on root communities. It may also communities. More information about these techniques
lead to the isolation of new, efficient PGPR. can be found in another article of this book and in Burdman
and coworkers (8).
Introduced Bacteria and Indigenous Rhizosphere
Populations: Support, Interference, or Indifference?
Reporter Genes
The influence of natural populations on introduced inocula
and vice versa is strongly dependent on the systems under Inocula of engineered PGPR can be tracked using reporter
study, that is, using PCR-based, culture-independent genes such as β-galactosidase (96) or GUS (97,98), the
ribosomal interspacer analysis (RISA); Robleto and latter being more adequate for direct detection on plant
coworkers (90) were able to show that inoculation of material because X-gal (5-bromo-4-chloro-3-indolyl-β-D-
common bean (Phaseolus vulgaris L.) rhizosphere with galactopyranoside), the substrate of β-galactosidase, can
a trifolitoxin (TFX) antibiotic-producing Rhizobium etli be cleaved by root enzymatic activities. Newer approaches,
leads to a detectable shift in bacterial populations, with combining both markers yielding stronger signals and
bands originating from members of the α-Proteobacteria more sensitive detection systems, such as coupled charge
subgroup disappearing. devices (CCD), result in greatly enhanced sensitivity.
Plants inoculated with a nonproducing TFX derivative For example, a 1,000-fold increase in sensitivity was
yielded a pattern identical to that of noninoculated plants. achieved using the bioluminescence luxAB gene system
Conversely, soil bacteria can reduce PGPR root coloniza- in rhizosphere-inoculated Pseudomonas spp. as compared
tion; Ikeda and coworkers (91) showed that individual with that achieved using the β-galactosidase gene (99).
strains of soil fluorescent pseudomonads exhibit such dele- The green fluorescence protein (GFP), product of
terious effects on a fluorescent pseudomonad PGPR. Com- the Aequoria victoria jellyfish gfp gene (100), has been
binations of individually nonsuppressive gram-negative demonstrated as perhaps the most versatile and efficient
bacteria could significantly suppress subsequent colo- reporter system; it is highly stable and does not require
nization of tomato (Lycopersicon esculentum) roots with external or internal reducing power, only illumination
this particular PGPR, suggesting that less competitive to fluoresce. A number of modified proteins have been
strains may hinder colonization of a competitive strain. engineered and exhibit absorption/emission spectral shifts
An inverse situation was found with Frankia, as Knowl- resulting in a change of color, fluorescence intensity, or
ton and coworkers (92) reported a significant increase in half-life (101,102). These genes have been successfully
nodulation of Alnus rubra under nonsterile conditions as used in a number of studies aimed at tracking introduced
compared with gnotobiotic growth with Frankia alone. bacteria on or in roots in various plant species (103,104).
Moreover, inoculation of pepper (Capsicum annum L.) Although quantitative uses are possible, bioluminescent-
with a mycorrhizal fungus resulted in a strong reduc- tagged PGPR have mainly been employed for spatial,
tion of 14 C root exudation, altering the composition of in situ studies of cell distribution along the root,
the rhizosphere, as detected using a bioluminescent pseu- in conjunction with epifluorescence or confocal laser
domonad (93). microscopy. Moreover, the physiological activity of a
Such studies are also important for assessing the particular introduced bacterium can be monitored at the
impact of genetically modified organisms, microbes, and single cell level using unstable variants of the GFP protein
 BACTERIAL PHYTOSTIMULATORS IN THE RHIZOSPHERE 351

in cells containing this particular gene under the control COMMERCIAL PRODUCTS: PRESENT AND FUTURE
of a ribosomal promoter (105).
The use of PGPR inoculants has become rather com-
Antibodies mon practice and is on the rise in many regions of
the world. Inoculation with PGPR also aims at reduc-
Antibodies raised against a phytostimulating bacterium ing the application of potentially polluting chemical
can be particularly useful for quantifying colonization fertilizers and pesticides (22,117). Various inoculants,
and for obtaining data on the spatial distribution of the referred to as biofertilizers, phytostimulators, and biopes-
inoculum, without relying on reporter genes, which may ticides, are commercially available under different brand
put a physiological burden on the cell. names (10,118,119).
The high specificity of monoclonal antibodies (mAbs) As an example, first-generation (wild-type) Azospiril-
raised against specific strains of A. brasilense was lum inoculants for maize (Azogreen ) are commercially
demonstrated (106). These mAbs were then tested in
available in Europe (Lipha, France). Results from field
situ in experiments in which the bacterial strains
experiments have consistently shown better utilization
were coinoculated on wheat roots grown in soil where
of nitrogen fertilizer by maize plants inoculated with
they could be specifically quantified in different root
A. brasilense (P. Wadoux, Lipha, Personal Communica-
compartments by chemiluminescent ELISA (106). In situ
tion).
detection and differentiation of coinoculated A. brasilense
Another company, Soygro (Pty) Ltd. (Potchefstroom,
strains Wa3 and Sp7 was achieved on a background
South Africa), is producing liquid and peat-based Azospir-
of root-associated bacteria concomitantly stained with
illum inoculants for 150,000 and 12,000 commercial
4’,6-diamidino-2-phenylindole (DAPI) and rRNA-directed
hectares of maize and wheat, respectively, and 500 exper-
fluorescent oligonucleotides (107).
imental hectares of sorghum. Field experiments with
Polyclonal antibodies (pAbs) have been used to identify
maize carried out over the last six years have shown
rRNA homology group I pseudomonad colonies grown from
yield increases of 10 to 30% above noninoculated plants.
rhizosphere material (108) and for in situ tracking of each
In the last few years, Azospirillum inoculants have also
component of a mixed inoculum of P. fluorescens in barley
been applied to legumes in South Africa. Average yield
roots grown under sterile conditions (109). Schlöter and
increases of 15 to 30% have been observed, following
Hartmann (110) used a monospecific pAb to quantify and
coinoculation with Azospirillum and Rhizobium, as com-
track R. leguminosarum biovar trifolii R39 inoculated
pared with inoculation with Rhizobium alone (T. E. M.
on roots of various legumes and nonlegumes using
Odendaal, Potchefstroom, Personal Communication).
chemiluminescent ELISA and fluorescein isothiocyanate
Large surfaces of maize and wheat in Mexico and
(FITC)-coupled antibodies, respectively.
Argentina are currently being inoculated with Azospiril-
lum, leading to significant increases in yield (J. Caballero-
rRNA Probes Mellado, CIFN, Cuernavaca, Mexico, Personal Commu-
Characteristic features of the rRNA molecules (16S and nication). Surprisingly, a company specializing in the
23S) form the basis for the development of oligonucleotide improvement of golf courses in the United States (EcoSoil
probes designed to match various levels of specificity, from Systems, Inc.) is now using Azospirillum to obtain better
the universal to the species-specific level (111), rendering quality of grass turf.
this tool instrumental in microbial ecology research. Important factors that can influence the efficiency
Fluorescence-labeled rRNA-directed probes can be used of microbial inoculants are summarized in Table 5.
in conjunction with scanning confocal laser microscopy to Successful field inoculation experiments appear to be
detect bacteria in the root system (85,112). They are useful those in which researchers paid special attention to
for studying the spatial distribution of targeted groups those factors, especially to the optimum number of cells
when applied in conjunction with scanning confocal laser in the inoculants and to the appropriate inoculation
or epifluorescence microscopy, or for direct quantification methodology, whereby an optimum number of cells
when cells are extracted from the rhizosphere. However, remained viable and available for root colonization (120).
because in most cases strain-specific probes cannot be The main obstacle impeding more intensive and
designed, tracking a specific strain by this means in a widespread use of PGPR inoculants has been the so-called
dense bacterial background may prove difficult. When inconsistency of results in field experiments, which could
used along with other detection methods, such as specific be explained by the complexity and the variability of the
antibodies directed toward the introduced bacterium, the soil-plant-microflora relationship at any experimental site.
relationship of the latter with surrounding rRNA-labeled However, such variability is also the lot of Rhizobium
bacterial populations can be observed, as Aßmus and inoculants, which have been commercially utilized for
coworkers (107) reported for A. brasilense (inoculated on about a century, and of chemical fertilizers and pesticides,
soil-grown wheat). which are not always efficient in the field.
Furthermore, and because the signal obtained is Biotechnological research objectives aiming at improv-
directly proportional to the rRNA cellular content (113), ing bacterial performance should also aim at diminishing
it is possible to assess the proportion of cells that inconsistencies. To this aim, the following approach may
remain physiologically active. This proportion appears be taken: first, a large number of strains need to be
to be greater in the rhizosphere (114,115) than in bulk screened, using selected crop plants grown under labo-
soil (116). ratory or greenhouse conditions. In the screening assays,
352 BACTERIAL PHYTOSTIMULATORS IN THE RHIZOSPHERE

Table 5. Factors Influencing the Efficiency of Microbial Inoculants

Factor Description

Inoculant quality Utilization of sterilized (gamma) carriers

Optimum bacterial concentration
Optimum physiological state of the cells
Quality control
Soil cultivation Cultivation practices leading to reduced natural microbial population
Preparation of a weed-free seedbed
Light irrigation just after planting/inoculation
Tillage implements, prevention of soil compaction (aerobic conditions)
Fertilization Nitrogen, phosphate: optimization according to the inoculant
Fungicides/Insecticides Prevention of negative effects on microorganism survival (e.g., physical/temporal
separation between application of chemicals and inoculation)
Environmental factors Soil characteristics at planting: pH, temperature
Flooded soil conditions, cool and hail storms that negatively influence bacterial activity

the different strains should be evaluated according to their Alternatives may be vermiculite, alginates, and liquid
capabilities to improve germination, seedling vigor, root formulations (8). With regard to application technology,
elongation, root branching, nitrogen fixation, and nodula- more extensive investigations need to be performed in
tion in the case of legumes. However, this procedure does order to compare the performances of seed and soil
not ensure that similar performance will be obtained in inoculation.
soil. Therefore, selected strains need to be further tested in Growth and development of bacteria under laboratory
pots and finally in the field. It is necessary to demonstrate, conditions is now rather well understood. The challenge
using Koch’s postulates, that the selected microorganism of the next decade will be to extend this knowledge
is responsible for the observed effect on the plant. One to microbial growth, survival, and exertion of beneficial
can then proceed toward the development of a commercial effects on the plants in situ, to understand the significance
product with the best strains. of factors limiting growth (e.g., nutrient limitation,
Another approach should aim at understanding the toxic products, protozoa) in the rhizosphere, and the
mechanisms involved in the production of the beneficial impact of abiotic factors (e.g., temperature, drought,
plant growth effects seen with certain bacterial strains. oxygen pressure) on bacterial survival and activity.
Such knowledge can provide a rational basis for the Moreover, the discovering of specific gene expression
improvement of inoculant production and storage pro- patterns in the rhizospheric environment and the
cedures, thus resulting in qualitatively superior products. understanding of the influence of environmental factors on
However, this approach is so expensive that it is not the essential functions for survival, and on traits involved
feasible for most agroindustrial products. in colonization and in plant growth promotion, will provide
Application of genetic engineering techniques for the answers necessary for the development of more reliable
development of better products with ‘‘superior’’ PGPR PGPR inoculants.
strains appears to be the most promising path. Extensive
research efforts are being carried out to isolate, clone, BIBLIOGRAPHY
and characterize genes involved in bacterial features
and activities related to survival, colonization, and 1. M. Alexander, Introduction to Soil Microbiology, 2nd ed.,
plant growth effects in the rhizosphere. Efforts are also John Wiley & Sons, New York, 1977.
being made to fuse reporter genes to those rhizosphere- 2. P. E. Maloney, A. H. C. Van Bruggen, and S. Hu, Microb.
related genes with the aim of following their activities Ecol. 34, 109–117 (1997).
and expression in situ. However, because of public 3. E. A. Curl and B. Truelove, The Rhizosphere, Springer-
disagreement and political developments with regard to Verlag, New York, 1986.
the release of biologically engineered material to the 4. J. M. Lynch and J. M. Whipps, Plant Soil 129, 1–10 (1990).
environment, at the present time, this approach cannot 5. B. M. Duineveld, A. S. Rosado, J. D. Van Elsas, and
be taken into account for applied purposes. J. A. Van Veen, Appl. Environ. Microbiol. 64, 4950–4957
Development of PGPR products includes paying (1998).
attention to survival and viability of the bacteria, not 6. J. W. Kloepper, in Y. Okon, ed., Azospirillum/Plant Associ-
only in soil, but also during storage. The aforementioned ations, CRC Press, Boca Raton, Fla., 1994, pp. 137–166.
properties (such as accumulation of storage material, 7. V. M. Reis, J. I. Baldani, V. L. D. Baldani, and J.
encystation, aggregation, etc.) may be extremely relevant Döbereiner, Crit. Rev. Plant Sci. 19, 227–247 (2000).
to the success of an inoculant. Carrier optimization also 8. S. Burdman, E. Jurkevitch, and Y. Okon, in N. S. Subba
has to be taken into account, as it is expected that a Rao and Y. R. Dommergues, eds., Microbial Interactions in
suitable carrier will provide good support and encourage Agriculture and Forestry, vol. 2, Science Publishers, Enfield,
bacterial survival during storage. For the Rhizobium U.K., 2000, pp. 227–248.
inoculants, the most ‘‘popular’’ carrier has been peat. 9. R. M. Boddey, Crit. Rev. Plant Sci. 14, 263–279 (1995).
 BACTERIAL PHYTOSTIMULATORS IN THE RHIZOSPHERE 353

10. Y. Okon and C. A. Labandera-Gonzalez, Soil Biol. Biochem. 40. F. M. Scher, in T. Swinburne, ed., Iron, Siderophores and
26, 1591–1601 (1994). Plant Diseases, Plenum Press, New York, 1986, pp. 109–117.
11. E. K. James and F. L. Olivares, Crit. Rev. Plant Sci. 17, 41. P. Lemanceau et al., Appl. Environ. Microbiol. 58,
77–119 (1998). 2978–2982 (1992).
12. B. Reinhold-Hurek and T. Hurek, Crit. Rev. Plant Sci. 17, 42. P. A. H. M. Bakker, J. M. Raaijmakers, and B. Schippers, in
29–54 (1998). L. L. Barton and B. C. Hemmings, eds., Iron Chelation in
13. M. Arshad and W. T. Frankenberger, Jr., Adv. Agron. 62, Plant and Soil Microorganisms, Academic Press, London,
45–151 (1998). U.K., 1993, pp. 269–281.
14. Y. Okon and J. Vanderleyden, ASM News 63, 366–370 43. S. Shah, V. Karkhanis, and A. Desai, Curr. Microbiol. 25,
(1997). 347–351 (1992).
15. S. Dobbelaere et al., Plant Soil 212, 155–164 (1999). 44. L. S. Thomashow and D. M. Weller, J. Bacteriol. 170,
16. M. Lambrecht, Y. Okon, A. Vande Broek, and J. 3499–3508 (1988).
Vanderleyden Trends Microbiol. 8, 298–300 (2000). 45. W. J. Howie and T. V. Suslow, Mol. Plant-Microbe Interact.
17. B. R. Glick, C. B. Jacobson, M. M. K. Schwarze, and 4, 393–399 (1991).
J. J. Pasternak, Can. J. Microbiol. 40, 911–915 (1994). 46. C. Keel et al., Mol. Plant-Microbe Interact. 5, 4–13 (1992).
18. S. Burdman et al., in C. Elmerich, A. Kondorosi, and 47. J. W. Kloepper and M. N. Schroth, Phytopathology 71,
W. E. Newton, eds., Biological Nitrogen Fixation for the 1020–1024 (1981).
21st Century, Kluwer Academic Publishers, Dordrecht, The 48. C. T. Bull, D. M. Weller, and L. S. Thomashow, Phytopathol-
Netherlands, 1998, pp. 609–612. ogy 81, 954–959 (1991).
19. E. Yahalom, Y. Okon, A. Dovrat, and H. Czosnek, Isr. J. Bot. 49. M. Mazzola, R. J. Cook, and L. S. Thomashow, Appl. Envi-
40, 155–164 (1991). ron. Microbiol. 58, 2616–2624 (1992).
20. S. Burdman et al., Appl. Environ. Microbiol. 62, 3030–3033 50. K. Toyota and K. Ikeda, Biol. Fertil. Soils 25, 416–420
(1996). (1997).
21. H. Volpin et al., Mol. Plant-Microbe Interact. 9, 388–394 51. M. Clarholm, Microb. Ecol. 7, 343–350 (1981).
(1996).
52. C. Ramirez and M. Alexander, Appl. Environ. Microbiol. 40,
22. Y. Okon, G. V. Bloemberg, and J. J. Ben Lugtenberg, in 492–499 (1981).
A. Altman, ed., Biotechnology in Agriculture, Marcel Dekker,
53. L. E. Casida, Appl. Environ. Microbiol. 55, 1857–1859
New York, 1998, pp. 327–349.
(1989).
23. D. M. Weller and L. S. Thomashow, in F. O’Gara,
54. P. De Troch and J. Vanderleyden, Microb. Ecol. 32, 149–169
D. N. Dowling, and B. Boesten, eds., Molecular Ecology of
(1996).
Rhizosphere Microorganisms, VCH, Weinheim, Germany,
1994, pp. 1–18. 55. L. Croes et al., J. Gen. Microbiol. 139, 2261–2269 (1993).
24. N. R. Krieg and J. Döbereiner, in N. Krieg and J. G. Holt, 56. R. De Mot and J. Vanderleyden, FEMS Microbiol. Lett. 81,
eds., Bergey’s Manual of Systematic Bacteriology, vol. 1, 323–328 (1991).
Williams & Wilkins, Baltimore, Md., 1984, pp. 94–104. 57. W. Achouak, R. De Mot, and T. Heulin, FEMS Microbiol.
25. A. Hartmann and W. Zimmer, in Y. Okon, ed., Azospiril- Ecol. 16, 19–24 (1995).
lum/Plant Associations, CRC Press, Boca Raton, Fla., 1994, 58. S. Burdman, Y. Okon, and E. Jurkevitch, Crit. Rev. Micro-
pp. 15–39. biol. 26, 91–110 (2000).
26. P. P. Wong, N. E. Stenberg, and L. Edger, Can. J. Microbiol. 59. W. L. Chao, E. B. Nelson, G. E. Harman, and H. C. Hock,
26, 291–296 (1980). Phytopathology 76, 60–65 (1986).
27. D. H. Halsall, G. L. Turner, and A. H. Gibson, Appl. Envi- 60. D. W. James, T. V. Suslow, and K. E. Steinback, Appl.
ron. Microbiol. 49, 423–428 (1985). Environ. Microbiol. 50, 392–397 (1985).
28. E. Hozore and M. Alexander, Soil Biol. Biochem. 23, 61. A. J. Anderson, P. Habibzadegah-Tari, and C. S. Tepper,
717–723 (1991). Appl. Environ. Microbiol. 54, 375–380 (1988).
29. P. K. Jjemba and M. Alexander, Soil Biol. Biochem. 31, 62. G. D. Bowen and A. D. Rovira, Adv. Agron. 66, 1–102
623–632 (1999). (1999).
30. L. L. Madison and G. W. Huisman, Microbiol. Mol. Biol. Rev. 63. S. Moens and J. Vanderleyden, Crit. Rev. Microbiol. 22,
63, 21–53 (1999). 67–100 (1996).
31. Y. Okon and R. Itzigsohn, FEMS Microbiol. Rev. 103, 64. W. J. Howie, R. J. Cook, and D. M. Weller, Phytopathology
131–140 (1992). 77, 286–292 (1987).
32. E. Husteda, A. Steinbuechel, and H. G. Schlegel, Appl. 65. J. H. Bowers and J. L. Parke, Soil Biol. Biochem. 25,
Microbiol. Biotechnol. 36, 87–93 (1993). 1693–1701 (1993).
33. S. Povolo et al., Can. J. Microbiol. 40, 823–829 (1994). 66. J. Boelens, M. V. Woestyne, and W. Verstraete, Soil Biol.
34. M. A. Cevallos et al., J. Bacteriol. 178, 1646–1654 (1996). Biochem. 26, 269–277 (1994).
35. S. Tal and Y. Okon, Can. J. Microbiol. 31, 608–613 (1985). 67. L. A. De Weger et al., J. Bacteriol. 169, 2769–2773 (1987).
36. B. H. Bleakley, M. H. Gaskins, D. H. Hubbel, and S. G. Zam, 68. Y. Bashan and H. Levanony, J. Gen. Microbiol. 133,
Appl. Environ. Microbiol. 54, 2986–2995 (1988). 3473–3480 (1987).
37. F. J. Bergersen and G. L. Turner, Proc. R. Soc. London, Ser. 69. Y. Bashan and G. Holguin, Microb. Ecol. 29, 269–281 (1995).
B 240, 39–59 (1990). 70. R. Barak, I. Nur, and Y. Okon, J. Appl. Bacteriol. 162,
38. J. B. Neilands, Annu. Rev. Biochem. 50, 715–731 (1981). 399–403 (1983).
39. J. W. Kloepper, M. N. Schroth, and T. D. Miller, Phy- 71. D. Heinrich and D. Hess, Can. J. Microbiol. 31, 26–31
topathology 70, 1078–1082 (1980). (1985).
354 BACTERIONEUSTON

72. I. B. Zhulin and J. P. Armitage, J. Bacteriol. 175, 952–958 103. D. J. Gage, T. Bobo, and S. R. Long, J. Bacteriol. 178,
(1993). 7159–7166 (1996).
73. Y. Okon, L. Cakmakci, I. Nur, and I. Chet, Microb. Ecol. 6, 104. G. V. Bloemberg, G. A. O’Toole, B. J. J. Lugtenberg, and
277–280 (1980). R. Koleter, Appl. Environ. Microbiol. 63, 4543–4551 (1997).
74. O. Reiner and Y. Okon, Can. J. Microbiol. 32, 829–834 105. C. Ramos, L. Molbak, and S. Molin, Appl. Environ. Micro-
(1986). biol. 66, 801–809 (2000).
75. A. Vande Broek, M. Lambrecht, and J. Vanderleyden, 106. M. Schlöter and A. Hartmann, Symbiosis 25, 159–179
Microbiology 144, 2599–2606 (1998). (1998).
76. S. Inamdar, R. U. Kanitkar, and M. G. Watve, Curr. Sci. 78, 107. B. Aßmus et al., Microb. Ecol. 33, 32–40 (1997).
719–722 (2000). 108. L. Kragelund, K. Leopold, and O. Nybroe, Appl. Environ.
77. L. Sadasivan and C. A. Neyra, J. Bacteriol. 163, 716–723 Microbiol. 62, 480–485 (1996).
(1985). 109. M. Hansen, L. Kragelund-Lene, O. Nybroe, and J. Sørensen,
78. J. B. Calleja, Microbial Aggregation, CRC Press, Boca FEMS Microbiol. Ecol. 23, 353–360 (1997).
Raton, Fla., 1984. 110. M. Schlöter and A. Hartmann, Appl. Environ. Microbiol. 63,
79. Z. Abbass and Y. Okon, Soil Biol. Biochem. 25, 1061–1073 2038–2046 (1998).
(1993). 111. R. Amann, W. Ludwig, and K. H. Schleifer, Microb. Rev. 59,
80. C. A. Neyra, A. Atkinson, and O. Olubayi, in I. Fendrik, 143–169 (1995).
M. Del Gallo, J. Vanderleyden, and M. De Zamaroczy, eds., 112. S. J. Macnoughton, T. Booth, T. M. Embley, and A. G.
Azospirillum VI and Related Microorganisms, Springer- Donnell, J. Microbiol. Methods 26, 279–285 (1996).
Verlag, Berlin, Germany, 1995, pp. 429–439. 113. R. Ruimy, V. Breittmayer, V. Boivin, and R. Christen,
81. N. P. Revsbech and B. B. Jørgensen, Adv. Microb. Ecol. 9, FEMS Microbiol. Ecol. 15, 207–214 (1994).
293–352 (1986). 114. B. Aßmus et al., Appl. Environ. Microbiol. 61, 1013–1019
82. H. T. Gollany, T. E. Schumacher, R. R. Rue, and S. Y. Liu, (1995).
Microchem. J. 48, 42–49 (1993). 115. E. Schallmach, D. Minz, and E. Jurkevitch, Microb. Ecol.
83. F. Neudoerfer and L. A. Meyer-Reil, Mar. Ecol. Prog. Ser. 40, 309–316 (2000).
147, 295–300 (1997). 116. H. Christensen, M. Hansen, and J. Sørensen, Appl. Environ.
84. L. Phillipot et al., FEMS Microb. Ecol. 21, 175–185 (1996). Microbiol. 65, 1753–1761 (1999).
85. G. Bettina and P. Frenzel, Soil Biol. Biochem. 30, 117. B. R. Glick, Can. J. Microbiol. 41, 109–117 (1995).
1903–1916 (1998). 118. P. M. Brannen and P. A. Backman, in M. H. Ryder,
86. N. P. Revsbech, O. Petersen, W. Reichardt, and A. Briones, P. M. Sterphens, and G. D. Bowen, eds., Improving Plant
Biol. Fertil. Soils 29, 379–385 (1999). Productivity with Rhizosphere Bacteria, CSIRO, Adelaide,
87. J. E. Loper and M. D. Henkels, Appl. Environ. Microbiol. Australia, 1994, pp. 83–85.
63, 99–105 (1997). 119. O. Mohammadi, in M. H. Ryder, P. M. Sterphens, and
88. O. Højberg et al., Appl. Environ. Microbiol. 65, 4085–4093 G. D. Bowen, eds., Improving Plant Productivity with
(1999). Rhizosphere Bacteria, CSIRO, Adelaide, Australia, 1994,
89. C. H. Jaeger, III et al., Appl. Environ. Microbiol. 65, pp. 282–284.
2685–2690 (1999). 120. J. Fages, Symbiosis 13, 15–26 (1992).
90. E. A. Robleto, J. Borneman, and E. W. Triplett, Appl. Envi-
ron. Microbiol. 64, 5020–5022 (1998).
91. K. Ikeda, K. Toyota, and M. Kimura, Can. J. Microbiol. 44, BACTERIONEUSTON. See NEUSTON MICROBIOLOGY:
630–636 (1998).
LIFE AT THE AIR –WATER INTERFACE
92. S. Knowlton, A. Berry, and J. G. Torrey, Can. J. Microbiol.
26, 971–977 (1980).
93. P. Marschner, D. E. Crowley, and R. M. Higashi, Plant Soil
189, 11–20 (1997).
94. W. F. Mahaffee and J. W. Kloepper, Can. J. Microbiol. 43, BACTERIOPHAGE AS INDICATORS
344–353 (1996).
95. P. Oger, A. Petit, and Y. Dessaux, Nat. Biotech. 15, 369–372 JUAN JOFRE
(1997). University of Barcelona
96. M. Höfte, M. Mergeay, and W. Verstraete, Appl. Environ. Barcelona, Spain
Microbiol. 56, 1046–1052 (1990).
97. S. B. Sharma and E. R. Signer, Gen. Dev. 4, 344–356 (1990). The use of indicator organisms to assess the microbio-
98. T. B. Hurek, B. Reinhold-Hurek, M. Van Montagu, and logical quality of water is well established and has been
E. Kellenberger, J. Bacteriol. 176, 1913–1923 (1994). practiced for almost a century. Early work focused on the
99. L. A. De Weger et al., Appl. Environ. Microbiol. 57, detection of bacterial indicators of fecal pollution, but now
3641–3644 (1991). it has been extended to bacterial indicators of sewage,
100. M. Chalfie et al., Science 263, 802–805 (1994). indicators of nutrient pollution, and so on. In the last
101. B. P. Cormack, R. H. Valvidia, and S. Falkow, Gene 173, 50 years, bacteriophages have been investigated as indi-
33–38 (1996). cators for different purposes. The most widely studied is
102. J. B. Andersen et al., Appl. Environ. Microbiol. 64, the use of phages infecting enteric bacteria as models
2240–2246 (1998). in water quality control. Other examples are the use of
 BACTERIOPHAGE AS INDICATORS 355

bacteriophages infecting Serratia marcescens or Staphy- Table 1. General Features of the Three Groups of Bacte-
lococcus aureus as tracers of groundwater movement, or riophages Proposed as Model Microorganisms for Water
bacteriophages infecting Vibrio parahemolyticus as indica- Quality Assessment
tors of eutrophication in marine waters. The scope of this Somatic F-specific Phages of
article is limited to a critical description of the relationship Feature Coliphages RNA-Phages B. fragilis
between the presence of bacteriophages of enteric bacteria
in water and their possible application as model microor- Homogeneity of the group + +++ +++
ganisms in water quality control. Different authors (1,2) Availability of +++ +++ +++
have listed the requisites for a model microorganism. How- standardized methods
ever, some potential extra advantages of bacteriophages Occurrence and levels in ++ + +
human feces
as indicators of water quality deserve comment. First,
Occurrence and levels in +++ + +
the limited importance, if any, of phenomena such as
animal feces
‘‘stress,’’ ‘‘injury,’’ or ‘‘reactivation’’ that frequently lead to Occurrence and levels in +++ +++ +++
misinterpretation of environmental data on bacterial indi- urban sewage
cators. Secondly, the samples for phage analysis can be Occurrence and levels in +++ ++ +
conserved for longer periods of time than samples for bac- abattoir sewage
terial analysis before the performance of assays. Finally, Probability of replication ++ + −
the somatic coliphages allow results to be obtained in four in environment
hours. Resistance to inactivation ++ + +++
in environment
Resistance to disinfection
MODEL, INDEX, AND INDICATOR CONCEPTS Chlorination + ++ +++
Ozonization ND +++ ND
The term indicator organism is not clearly defined, which UV irradiation + +++ ++
sometimes leads to confusion. Therefore, the concept of High-energy radiation ++ + ++
model microorganisms as used by other authors (1,2) Thermal treatment +++ + +++
will be followed. According to this usage, the index and Note: +++, high; ++, intermediate; +, low; −, very low; ND, no data
the indicator functions can be attributed to the model available.
organisms.
An index organism is one that is related to the
occurrence of the selected surrogate microorganism or quality assessment: somatic coliphages, F-specific RNA
microorganisms. The relationship may be direct, such as bacteriophages, and bacteriophages infecting Bacteroides
the index of human viruses or indirect, such as an index of fragilis. Table 1 summarizes some of the features of the
fecal pollution, sewage pollution, or types of fecal pollution three groups of bacteriophages that is more extensively
(i.e., human or animal). The criteria for an index organism described in the following text.
are very similar to those commonly used for bacterial fecal
indicators. Somatic Coliphages
The model organisms with an indicator function have a
much broader definition. An indicator is basically a model Somatic coliphages are bacteriophages, which consist of
that has behavioral characteristics similar to those of the a capsid containing single-stranded or double-stranded
surrogate microorganism or microorganisms, and has the DNA as the genome. The capsids may be of simple cubic
same or greater resistance to environmental stresses, but symmetry or complex structures with heads, tails, and
that does not necessarily originate from the same source tail fibers. They are classified into the following families:
as the pathogen (e.g., feces). This criterion, and others Myoviridae (dsDNA, long contractile tails, and capsids
mentioned earlier, define an indicator as a marker for the up to 100 nm), Siphoviridae (dsDNA, long noncontractile
efficiency of treatments or for the variation in numbers tails, and capsids 50 nm), Podoviridae (dsDNA, short
of the surrogate microorganisms in specific environments. noncontractile tails, and capsids 50 nm), and Microviridae
In general, the criteria for indicator organisms are less (ssDNA, no tail, and capsid 30 nm). All types are found
restrictive than the criteria for index organisms (1). in sewage, although the most abundant are Myoviridae
Ideally, a perfect model organism will fulfill both and Siphoviridae (3,4). Somatic coliphages attach to the
the index and the indicator functions. However, the bacterial cell wall and may lyse the host cell in 20
more is known about fecal pathogens and potential to 30 minutes under optimum conditions. They produce
model microorganisms, the more unlikely it seems that plaques of widely different size and morphology. The
a single fecal microorganism fulfilling all criteria under all methodology to detect them is very simple and results
circumstances, will be found. may be obtained in four hours.
Natural host strains of somatic coliphages include
BACTERIOPHAGES AS MODEL ORGANISMS IN WATER Escherichia coli and other closely related bacterial species.
QUALITY ASSESSMENT Some of these may occur in pristine waters, hence somatic
coliphages may also multiply in these environments
Three main groups of bacteriophages infecting enteric even in E. coli (5,6). Indeed, one of the drawbacks of
bacteria have been considered so far as potential somatic coliphages is their replication potential outside
model microorganisms for various aspects of water the gut. However, the contribution of this potential
356 BACTERIOPHAGE AS INDICATORS

replication outside the gut to their occurrence in natural They produce clear plaques, which do not differ very much
environments has never been quantified. in size or morphology. Bacteroides fragilis strains differ
The term somatic coliphages covers many types of widely in the numbers of phages that they recover from
phage with a wide range of characteristics, including municipal sewage, but most strains recover bacteriophages
differential resistance to inactivating factors. Moreover, that are very similar and belong to the family Siphoviridae.
different strains of E. coli and different assay media count, Bacteriophages infecting B. fragilis have not been reported
different numbers, and types of somatic coliphages (4,7). to replicate under environmental conditions. The method
Consequently, the information available on both the for detecting these bacteriophages is slightly more complex
presence and the behavior of somatic coliphages in water than methods for detecting the other groups because
environments has to be interpreted very cautiously, of the anaerobic nature of the host bacteria. However,
because the data reported in the literature had been a relatively simple standardized method is now being
obtained with different host strains of E. coli, media, and discussed (12). Results can be obtained in 18 hours.
assay conditions. To avoid confusion, the term somatic Bacteriophage B40-8 cited in the text belongs to this
coliphages, unless otherwise indicated, will be restricted group.
to phages infecting E. coli C as established by standard
methods (8) or ISO (9). Bacteriophages frequently used to
study somatic coliphage behavior are T-even and T-odd, BACTERIOPHAGES AS INDEXES
X147, and PRD-4.
Bacteriophages as Indexes of Fecal Pollution
F-Specific RNA Bacteriophages Many studies have been performed on the distribution of
F-specific RNA bacteriophages consist of a simple capsid bacteriophages specific to enteric bacteria in human and
of cubic symmetry of 21 to 30 nm in diameter and contain animal feces. Somatic coliphages using E. coli C as the host
single-stranded RNA as the genome. They belong to the have been isolated in variable percentages (up to 60%) of
family Leviviridae. They infect bacteria through the sex human stool samples, reaching numbers up to 109 plaque
pili, which are coded by the F plasmid first detected in forming units (PFU)/g (13,14). The percentages of isolation
E. coli K12. The F plasmid is transferable to a wide reported in mammals and birds range from 100% in pigs
range of gram-negative bacteria. The pili coded by the to 38% in rabbits (14,15) with counts up to 107 PFU/g.
F plasmid do not form below 25 ° C (10). Therefore, the F-specific bacteriophages rarely have been isolated
probability of these phages replicating in the environment from human feces, irrespective of the host strain
is small (10). The infection process is inhibited by the used. The maximum positive isolation described is
presence of RNase in the assay medium, which can be used approximately 30% (14–16). Such phages have been
to distinguish between the F-specific RNA bacteriophages isolated inconsistently from domestic and feral animal
and the rod-shaped F-specific DNA bacteriophages of the feces, but incidences are in general higher than in
family Inoviridae, which also infect the host cell, through humans (15,17), with densities up to 104 PFU/g, although
the sex pili. usually lower.
Strains (Salmonella typhimurium WG49 and E. coli Bacteroides fragilis RYC 2056 has been reported to
HS) tailored to detect F-specific bacteriophages also recover phages from 28% of human stool samples,
detect small percentages of somatic phages. All phages although it also allows the isolation of phages from
detected by the tailored strains are usually referred to as animals (18), with the maximum incidence, 30%, in pigs
F-specific bacteriophages. The number of F-specific RNA with maximum values of 3 × 102 PFU/g. Bacteriophages
bacteriophages is the difference between the number of infecting B. fragilis HSP40 have been isolated from 10 to
phages counted in the presence and in the absence of 13% of human stool samples (15,19) with maximum values
RNase in the assay medium. More than 90% of the phages above 108 (19). However, they have not been isolated from
detected in sewage by tailored strains are F-specific RNA animal feces (15,19).
bacteriophages. This percentage may be lower in receiving
waters. Bacteriophages as Indexes of Sewage Pollution
A standardized method for the detection and enumera-
Bacteriophages of the three groups mentioned earlier are
tion of F-specific RNA bacteriophages is now available (11).
consistently present in raw sewage and sewage effluents.
Results can be obtained in 12 hours.
Consequently, sewage pollution will lead to contamination
F-specific RNA bacteriophages mentioned in this entry
by the three groups.
are f2 and MS2.
Most reports indicate that somatic coliphages are the
most abundant in raw municipal sewage, with values
Bacteriophages Infecting Bacteroides fragilis
ranging from 106 to 107 PFU/100 ml, approximately less
The most abundant bacteriophages infecting Bacteroides than one order of magnitude lower than the numbers of
fragilis, one of the most common bacteria in the gut of fecal coliforms or E. coli (15,20,21). In addition, they are
warm-blooded animals, belong to the family Siphoviridae, the most abundant in abattoir wastewater, with values
with flexible tails (dsDNA, long noncontractile tails, and that are similar to those found in municipal wastewater,
capsids up to 60 nm). Phages infecting B. fragilis attach or values that keep the proportion to fecal coliforms as in
to the cell wall of the host bacteria and may lyse the municipal wastewater (13,22,23). Values from slurries of
host cell in 30 to 40 minutes under optimum conditions. different animals are also similar.
 BACTERIOPHAGE AS INDICATORS 357

Numbers in sewage effluents depends on the waste- Because of the difficulty of enumerating human viruses,
water treatment. However, they are present in the great studies establishing a correlation between human viruses
majority of sewage effluents. and bacteriophages are scarce. A certain degree of
F-specific RNA bacteriophages rank second in abun- correlation has been reported between: (1) enteroviruses
dance both in municipal raw sewage and in raw waste- and somatic coliphages in water at the different stages
water from abattoirs. The most frequent average values of a water treatment plant (29), (2) F-specific RNA
range from 5.105 to 5.106 PFU/100 ml, usually about bacteriophages and enteroviruses in fresh water (31),
one order of magnitude lower than values for somatic (3) F-specific RNA bacteriophages and enteroviruses
coliphages (13,17,18,20,22,24,30). The ratios between F- in shellfish (25), (4) F-specific RNA bacteriophages and
specific RNA bacteriophages and somatic coliphages are calicivirus in shellfish (32), and (5) enteroviruses and
of the same order of magnitude in both raw municipal rotaviruses and phages infecting B. fragilis HSP40 in
sewage and wastewater from most abattoirs. In some slur- marine sediments (33).
ries (e.g., cattle) and specific abattoir wastewater, their If bacteriophages are not isolated in all human stool
relative abundance may be lower. Numbers in sewage samples tested and if they may originate in sites other
effluent depend on the treatment, but these phages are than the human gut, there will not be an absolute and
always found in wastewater effluents. unequivocal correlation between the presence of viruses
Bacteriophages infecting B. fragilis RYC 2056 are and the presence of the different groups of bacteriophages.
found in municipal sewage (Europe, Africa, and America), However, taking into consideration that bacteriophages
frequently ranging from 104 to 105 PFU/100 ml, usually infecting enteric bacteria are always present in raw sewage
one order of magnitude less than F-specific RNA bacterio- that sewage constitutes the main input of fecal pollution
phages. Their ratio with respect to somatic coliphages and to the environment, and that some phages behave similar
F-RNA bacteriophages is remarkably constant in munici- to human viruses, the presence of certain phages in given
pal sewage (18,20). Host strain VPI 3625 has been shown numbers in waters may indicate the presence of human
to recover approximately 104 PFU/100 ml in the United viruses in such waters.
States (25), whereas strain HSP40 detected approximately
104 PFU/ml in some geographic areas, but much lower Bacteriophages as Indexes of Human and Animal Pollution
values in other areas (18,25).
Identifying the sources of fecal pollution is important
In abattoir effluents, phage numbers detected by
for water management. One reason, although not fully
host strain RYC 2056 range from 0 to 104 PFU/100 ml,
supported by epidemiological studies, is the perception
but their ratios to somatic coliphages and to F-RNA
that there is a higher health risk associated with human
bacteriophages are significantly lower when compared
fecal pollution than with fecal pollution of animal origin.
with urban sewage (18). Numbers in municipal sewage
But, irrespective of the health risk, tracking the origin
effluents depend on the treatment, but these phages are
of fecal water pollution will be a very useful tool for
regularly found in wastewater effluents.
water managers. A range of microorganisms and chemical
It can be concluded that numbers of the three
indicators have been examined as fecal source indicators,
groups of bacteriophages are fairly constant in raw
including phages.
sewage throughout the world, as are the numbers of
F-RNA bacteriophages are divided in four main
bacterial indicators. Furthermore, bacteriophages of the
subgroups that can be recognized by serotyping. Studies
three groups are consistently found in sewage effluents.
with a number of isolates indicate that serotypes II
Consequently, the three groups of phages may be
and III are mainly isolated from human feces, whereas,
considered as indexes of sewage pollution. On average,
serotypes I and IV are usually found in animal feces (34).
all the groups of phage are more abundant in raw sewage
More recently, it has been shown that the subgroups
than most pathogens.
can be grouped in four main genotypes, which, with few
Studies on the abundance of bacteriophages infecting
exceptions, show overall comparability with serotypes.
enteric bacteria in wastewater sludges (biosolids) are
Probes for each genotype allow plaque hybridization, and
relatively scarce. The accumulation of bacteriophages in
then it is easy to study the distribution of subgroups
both primary and secondary sludges is similar to that
in water samples. Subgroups II and III predominate in
of bacterial indicators, and the ratios of such indicators
water samples contaminated with human fecal pollution
to phages do not differ significantly from those in
and subgroups I and IV in animal feces and water samples
sewage (26,27). The three groups of bacteriophages seem
contaminated with animal fecal pollution (25,35,36).
to accumulate in a similar manner (26). However, sludges
Bacteriophages infecting B. fragilis HSP40 are detected
are subjected to additional treatment before release to
in samples with human fecal pollution and are practically
the environment. Phage concentrations in treated sludges
absent in samples with animal fecal pollution (15,19),
depend on the treatment (26).
but they are very scarce in wastewater in some geo-
graphic areas (18,25). Bacteriophages infecting B. fragilis
Bacteriophages as Indexes of Human Viruses
RYC 2056 are found at concentrations ranging from 104
Somatic coliphages (28), F-specific-bacteriophages (22), to 105 PFU/100 ml in municipal sewage of very different
and bacteriophages infecting B. fragilis (19) have been geographic origins, and at significantly lower levels in
proposed as potential indexes of the presence of human wastewater from abattoirs (18).
viruses in the environment on the basis of their More studies are needed on the ecology of these two
resemblance to human enteric viruses. groups of phages, but they seem to be one of the best ways
358 BACTERIOPHAGE AS INDICATORS

of distinguishing human from animal fecal pollution at resistance (e.g., the T-even and T-odd bacteriophages
present. among the somatic coliphages). When possible, this article
will refer to data based on changes of ratios among the
different model microorganisms or simulation experiments
BACTERIOPHAGES AS INDICATORS
with naturally occurring populations.
The indicator function requires that resistance to natural
inactivating factors and to treatment of the indicator Removal and Inactivation in Nature
should be similar to or slightly greater than that of the Sedimentation. Sedimentation of microorganisms dep-
surrogate microorganism or microorganisms. In addition, ends on size, whether they occur singly, in clumps, or are
pathogens are assumed not to replicate in the aquatic associated with suspended solids, among other factors. In
environment. Model organisms should not replicate in general, removal of microorganisms by sedimentation is
the environment, and consequently, if there are no new more efficient for larger particles. But these then survive
contributions, their numbers in any aquatic environment better in sediments, because they are bound to solids.
are assumed to decrease. However, they may be resuspended into the water column
The reduction of pathogens and model microorganisms if the sediments are disturbed. Adsorption and aggregation
in sewage and environmental waters, on their way from of viruses may also occur. The tendency to be adsorbed to
human or animal feces to humans through water and food particles has been described to be similar for viruses and
is a complex process in which many factors interact. There somatic coliphages (37). At least in waters with a high fecal
are three major groups of processes: pollution, the three groups of bacteriophages suggested
as model microorganisms settle similarly (38). However,
1. Physical removal by mechanisms such as sedimen-
their survival in sediments is expected to depend on the
tation, adsorption, and straining. These processes
type of microorganism.
do not inactivate the microorganisms. They merely
transfer the microorganisms to other environmen-
tal compartments (sewage sludges, sediments, soils, Adsorption to Soil. Adsorption of pathogens and indi-
etc.) where they are inactivated, but from which they cators to soil particles plays an important role in their
can return to the water. removal from water as it moves through the soil. Soil type
and composition, ionic environment, pH, moisture content,
2. Inactivation by direct or indirect effects of physical-
temperature, microorganism species, or strains all inter-
chemical factors on the integrity of the organisms.
act to affect the adsorptive capacity and die-off rate in soil
Die-off of phages and viruses in aquatic environ-
of pathogens and model microorganisms.
ments is influenced by many factors, such as sus-
Data on the presence of bacteriophages in groundwater
pended matter, sunlight, temperature, pH, and ionic
and on the comparison of numbers of bacteriophages,
environment. The disinfectants used by man are
viruses, and bacterial indicators are scarce, and most
included in this group.
often only presence/absence is given, and consequently,
3. Inactivations as a result of the biological activity results from different reports are very difficult to compare.
of other microorganisms. Antagonistic activity and Various studies performed across the globe indicate small
grazing by protozoa contribute to pathogen removal. and nonsignificant differences among the frequencies of
The main factor that seems to affect bacteriophages isolation of the three groups of bacteriophages. Thus, in
and viruses is the production of antiviral substances. a nationwide study performed in United States, F-specific
The relative importance of each of these processes is bacteriophages were more frequently isolated (18%) than
determined by a great number of variables, which are also somatic coliphages (7.6%) in concentrate equivalents of
interrelated. Two major factors are time and temperature. 4.5 L (39). F-specific coliphages were isolated 18.9%,
Each of the earlier mentioned processes is directly and B. fragilis phages 17.1%, and somatic coliphages 10.4%
positively affected by time, and the great majority is also of time in 100-ml samples of groundwater in Spain (40).
directly or indirectly affected by temperature. Removal Most model studies on the behavior of bacteriophages in
and inactivation are more rapid at higher temperatures soil were done with F-specific RNA bacteriophages, mostly
and increase with time. MS2 and f2, both belonging to subgroup I. A number
Separate studies have indicated that there are great of studies have consistently shown that F-specific RNA
differences between the removal and the inactivation bacteriophages adsorb soil particles poorly and survive
of seeded and naturally occurring bacteria, viruses, better in groundwater than enteric viruses. As a result,
and bacteriophages. Hence, realistic comparisons of the the F-RNA phages have been recommended as an indicator
behavior of the different groups of microorganisms of human virus transport in soil and of their presence in
can be made only on data with naturally occurring groundwater (41–43).
microorganisms. Unfortunately, extensive studies have
not yet been carried out, and those already published are Inactivation. Susceptibility or resistance to inactivation
difficult to compare because of differences in detection in the environment can be estimated either by studying
methodologies. Many studies were performed with a the ratios between the different microorganisms naturally
single-seeded phage, which in addition to the shortcoming occurring in sewage and water environments with remote
mentioned earlier may be neither the most abundant pollution or through the performance of experiments that
nor the most representative of a given group regarding are modeled on what happens in nature.
 BACTERIOPHAGE AS INDICATORS 359

Data available on phage inactivation in the environ- wetlands may be particularly useful in warm and temper-
ment frequently seem contradictory. This may be because ate climates to further treat secondary effluents. Again,
of the fact that various phages have been studied under information is very variable and partial, but in all these
different experimental conditions and using diverse detec- systems bacterial indicators are removed more efficiently
tion methodologies. Phages generally survive longer in than somatic coliphages (52,53). Seeded polioviruses and
aquatic environments than most bacterial indicators, with phage MS2 behave similarly (54). Data comparing the
the exception of clostridia. Differences between the sur- three groups of bacteriophages are very scarce, but all
vival of phages and bacterial indicators seem to be greater were inactivated more slowly than the bacterial indica-
in seawater than in fresh water, in warm water than in tors and the differences in their removal are smaller than
cold water, and in areas with abundant sunshine than differences between their removal and the removal of
in areas with scarce sunshine. Somatic coliphages and bacterial indicators (55).
bacteriophages infecting B. fragilis survive better than Filtration, coagulation/filtration, and postprecipitation
F-specific RNA bacteriophages in most water environ- processes are frequently used to remove suspended solids
ments, mainly in those with more inactivating power. from secondary effluents and precipitation (postprecip-
These assertions are supported by data on relative abun- itation) to remove N and P compound from secondary
dance of the different groups of microorganisms in different effluents. Filtration alone has little effect on the removal
aquatic environments (20,25,38,44–46) and by results of of bacteria and viruses. But coagulation/filtration and
model experiments (24,47,48). (post)precipitation have a significant effect on the reduc-
Data available from experiments with seeded cultured tion of microorganisms, improving the removal obtained
viruses and bacteriophages indicate that phage survive by activated sludge by 90–95%. The reduction of viruses,
more similarly to human viruses than bacterial indica- bacterial indicators, and bacteriophages appears to be
tors (24,47). similar (21,56).

Removal and Inactivation by Drinking Water Treatments

Removal and Inactivation by Wastewater Treatment
Most data on the presence of bacteriophages in drinking
Wastewater treatment processes, especially the biological
water refers to somatic coliphages. These have been
processes, are diverse in design and operating conditions. isolated in tap water samples in the absence or in
If we consider the different bacteriophage detection the presence of low levels of fecal coliform bacteria in
methodologies used in different studies, it is not surprising both developing and developed countries. Many of these
that data in the literature on bacteriophage removal by samples containing somatic coliphages contained residual-
wastewater treatment processes vary widely. Presumably, free and combined chlorine (1,57,58). After filtering
the bacteriophage removal rate is lower than that of nine samples of 20,000 liters of drinking water, no
bacterial indicators and closer to that of animal viruses. enteroviruses were isolated, somatic coliphages were
However, the proportion of phages removed depends on isolated in two samples and F-specific RNA bacteriophages
the treatment process and, in some cases, on the group of in one sample in Canada (59). In a three-year study
bacteriophages studied. performed in Israel on 1,136 tap water samples, the
The removal of bacteriophages by primary treatment is percentages of positive isolations were 0.7% for fecal
usually inefficient and erratic, and data reporting ranges coliforms, 5.6% for somatic coliphages, 7.1% for F-specific
from 10 to 90% without significant differences in the bacteriophages, and 5.0% for B. fragilis phages (40). In
removal rate of bacterial indicators and bacteriophages, Spain, in 410 samples including finished water from a
or between the different groups of bacteriophages (49,50). drinking water treatment plant and tap water samples the
However, treatment with lime shows a clear difference, percentages were 0.5% for fecal coliforms, 1.5% for somatic
because fecal coliforms, the most sensitive of the indicators coliphages, 2.2% for F-specific bacteriophages, and 5.2%
suffer a 99.85% reduction, and phages infecting B. fragilis, for phages infecting B. fragilis (7,40).
the most resistant of the phages, display a 91.8% Precise quantification of bacteriophage removal in
reduction (50). drinking water plants is difficult to achieve, either because
Most data on secondary treatments are related to the of the low levels of naturally occurring bacteriophages
activated sludge process. Again, available information is in source water or because of the treatment removes
very variable. Pathogen and indicator reductions depend all the bacteriophages. Bacteriophages are more difficult
on many factors that may have a differential effect to remove than coliform bacteria. Their comparative
on the various pathogens and model organisms. Well- removal depends on the proportion of physical removal
operated plants remove from 90 to 99% of bacteria and (coagulation, sedimentation, and filtration) and chemical
viruses (21,51). The different groups of bacteriophages are disinfection. Physical removal procedures frequently used
also removed in similar percentages (21,50,51). in drinking water treatment plants such as flocculation-
The typical concentrations of the different indicator sedimentation and filtration have little effect on removal
microorganisms in secondary effluents maintain the same of bacteriophages, less than 0.5 logs, in comparison with
relative proportion as in raw sewage. the removal of bacteria including Clostridium (29,58,59).
Effluents from activated sludge plants may be fur- Removal of bacteriophages by disinfection is greater than
ther in, that is, lagoons, oxidation ponds, and wetlands; physical removal. Bacterial indicators are inactivated
filtration, coagulation, and sedimentation; and disinfec- more efficiently than bacteriophages, with the exception
tion. Lagoons, oxidation ponds, and artificial or natural of clostridia, which may be more resistant than phage.
360 BACTERIOPHAGE AS INDICATORS

Levels of the three groups of phages seem to be reduced type of water, viral strain studied, and methodology
similarly by physical removal. In contrast, their removal used.
by disinfection depends on the procedure. The difference
in the fecal coliform removal and the most resistant group Chemical Disinfectants. The most widely used disinfec-
of bacteriophages in a given plant may be of several tants for water and wastewater are chlorine, chloramine,
orders of magnitude. Thus, prechlorination followed chlorine dioxide, bromide, bromine chloride, iodine, ozone,
by flocculation removed phages infecting B. fragilis and peracetic acid.
1.5 logarithmic units, when fecal coliforms removal was The available information on the effectiveness of these
4.9 logarithmic units (60). The differences in the extent disinfectants are with laboratory-grown strains. Bacterial
of removal of the different phages will be greater as the indicators, with the exception of spores of clostridia, are
treatment plant increases in complexity and will depend usually more sensitive to free chlorine than with viruses
on the type of disinfectant (7,59). and bacteriophages. But viruses and bacteriophages
Bacteriophages are expected to behave more similarly differ in their sensitivity. Thus, among viruses, human
to human viruses than bacteria, and many drinking water rotaviruses and hepatitis A virus (HAV) rank among the
treatments provide water without fecal indicators that still most resistant and polioviruses and SA11 among the most
contain infectious bacteriophages. It is suggested that the sensitive (62). Among phages, B40-8 infecting B. fragilis is
presence of bacteriophages of any of the three groups in more resistant than MS2 and f2, which are more resistant
drinking water indicates a potential risk of viral pollution, than polioviruses (63). In the presence of 2 ppm of free
and bacteriophage tests have been proposed as additional chlorine, B40-8 is as resistant as HAV (62). Data on
criteria for drinking water safety (1). somatic coliphages are very variable, but this variability
may depend on the morphological type. Thus, some of them
Inactivation by Sludge (Biosolid) Treatment are very sensitive and others significantly more resistant
than f2 and similar to SA11 (64).
Wastewater treatment generates huge amounts of
Comparative data on the chlorinated secondary efflu-
biosolids, where pathogens and indicators accumulate.
ents referring to naturally-occurring pathogens and indi-
According to their destination, sludges receive different
cator microorganisms are scarce, but in spite of the fact
treatments, the most frequent being storage, digestion,
that they were obtained in different treatment plants
and, to a lesser extent, disinfection by pasteurization,
and even on different continents, they show a similar
irradiation, or lime treatment.
pattern of inactivation. The difference of inactivation
The survival of viruses and bacteriophages in sludge
of coliform bacteria and fecal streptococci with respect
during storage depends on the temperature, as occurs for
to inactivation of phages is higher than in drinking
human viruses. At 4 ° C all survive quite well, whereas at
water. This fact probably indicates than phages are more
higher temperatures F-specific bacteriophages inactivate
resistant to monochloramines than bacteria. Regarding
more rapidly than somatic coliphages and bacteriophages
bacteriophages, those infecting B. fragilis are the most
infecting B. fragilis, the difference being significant at
resistant, followed by F-specific RNA and finally, somatic
37 ° C (26).
coliphages. Enteroviruses rank between F-specific RNA
Different kinds of digestion processes are used to
bacteriophages and bacteriophages infecting B. fragilis.
reduce the amount and improve the quality of sludges.
Sulfite-reducing clostridia behave similar to the most
Bacteriophage f2 has been reported to survive more
resistant bacteriophages (24,65,66).
successfully than enteroviruses and rotaviruses to both
Available data on the effect of ozone on bacteriophages
mesophilic and thermophilic digestion (61).
are scarce and contradictory. Since the aggregation state of
microorganisms is a key factor in determining inactivation
Inactivation by Disinfection
by ozone, data on naturally occurring viruses and phages
Disinfection is used both in wastewater and drinking are essential to assess the effect of ozone on these
water treatment. It is in these processes that major microbes. The effect of ozone in finished drinking water
differences in removal are given between pathogens, is difficult to measure, because usually ozone inactivates
including different strains or types of the same pathogen, all microorganisms to such an extent that they are never
between pathogens and indicators, and between indicators detected.
themselves. Resistance of phages will only be compared Very few data exist on removal of phages from
with that of bacterial indicators because our knowledge of secondary effluents by ozonation and, again, they are
comparative resistance with respect to other pathogens contradictory regarding the comparative resistance of
such as Giardia and Cryptosporidium is very limited bacterial indicators, viruses, and phages. Thus, some
and because the methods to determine their viability studies indicate that bacteria are more sensitive than
are difficult to perform. The difference in sensitivity to viruses and phages, but other studies indicate the
a given disinfection process deduced from experiments contrary. These contradictions may be because of the
with pure cultures, especially those seeded with the very fact that these studies were done with laboratory-grown
different types of somatic coliphages, may not reflect microorganisms.
what happens in the real world. Moreover, contradictory
results comparing a given virus and a given bacteriophage, Radiation. Different types of radiation are used for
for example, f2 and poliovirus, appear in the literature. disinfection. The most frequent are UV radiation, which is
These inconsistencies probably reflect differences in the used for low-turbidity water (drinking water, secondary
 BACTERIOPHAGE AS INDICATORS 361

effluents, shellfish depuration water, etc) and gamma accumulate. The extent of accumulation depends on many
and electron-beam irradiation, for water and sludges factors, such as water characteristics, physiological status
respectively. of the animal, and the identity of the microorganisms.
UV radiation is extensively used for disinfecting Bacteriophages accumulate in shellfish in a similar
drinking water and secondary effluents. Viruses and manner (25,72,73).
bacteriophages are less sensitive than bacterial indicators, Many studies have also shown differential rates of
although there are broad ranges of sensitivities between reduction of bacteria and viruses in depurating shell-
the strains of each group of microorganisms. Moreover, fish, with bacteria generally being reduced more rapidly.
binding to suspended solids protects both viruses and In depuration experiments with heavily contaminated
bacteria. Therefore, data obtained with seeded drinking oysters and mussels, F-specific bacteriophages are depu-
water cannot be extrapolated to secondary effluents. F- rated much less efficiently than E. coli and similarly to
specific RNA bacteriophages are more resistant to UV polioviruses (74,75).
irradiation than fecal coliforms, fecal streptococci, most Numbers of fecal bacteria, viruses, and bacteriophages
human viruses, bacteriophages infecting B. fragilis, and in shellfish collected in areas with different pollution levels
selected somatic coliphages (63,67,68). are in agreement with the earlier assertions. All the three
Few experiments have been performed with naturally- groups of bacteriophages are found in shellfish grown
occurring phages. F-specific phages have been reported to in waters with low levels of fecal pollution in amounts
be 2.3 times more resistant than E. coli in secondary efflu- that indicate that they accumulate and survive longer
ent (67). Other studies reported removals from filtrated than bacterial indicators, and more similar to human
secondary effluents of between 2 and 3 logs for fecal col- viruses (24,32,73). Data on somatic coliphages should
iforms, fecal streptococci and somatic coliphages, between be considered cautiously, because they may be able to
1 and 2 logs for F-specific RNA bacteriophages and inter- replicate in shellfish (6).
mediate values for B. fragilis bacteriophages (55); and
between 1 and 2 logs for fecal coliforms, fecal strepto- CONCLUSION
cocci and somatic coliphages from coagulated secondary
effluents (69). F-specific RNA bacteriophages thus appear Bacteriophages infecting different enteric bacteria have a
to be the most suitable indicator to evaluate UV radiation number of characteristics that make them potential useful
effects. model organisms in water quality control and assessment.
Gamma and electron beam radiation are slightly more Three groups of bacteriophages that infect enteric bacte-
effective in eliminating E. coli than in eliminating somatic ria are under consideration as potential model microor-
coliphages in sewage (70). The only data comparing the ganisms. These are somatic coliphages, F-specific RNA
different groups of phages are on phages seeded in bacteriophages, and bacteriophages infecting B. fragilis.
distilled and tap water, which indicates that MS2 is Bacteriophages belonging to the three groups of phages
more sensitive to ionizing radiation than phage B40- are irregularly found in feces, but they are very consis-
8 infecting B. fragilis, and this is more sensitive than tently found in sewage, in numbers that usually exceed
somatic coliphage X147. In this case, MS2 is as sensitive the numbers of pathogens by several orders of magnitude.
or more than E. coli (71). Standardized methods for detecting and enumerating the
three groups of bacteriophages are presently available.
Heat Treatment. Levels of fecal microorganisms in Methods for detecting and enumerating bacteriophages
sludges are lowered mainly through thermal treatment, are simple and can be carried out by any well-trained
either by mesophyllic and thermophilic digestion or microbiologist. They do not need sophisticated equipment
by even pasteurization. Bacterial indicators are more and can be performed in standard routine microbiology lab-
sensitive to heat treatment than viruses and phages, which oratories. They are relatively inexpensive and definitive
in turn differ in their resistance/sensitivity to heat. Thus, results can be obtained quickly.
bacteriophage f2 is more resistant than rotaviruses and Bacteriophages had been proposed as indicators of the
enteroviruses, but less so than parvoviruses (61). Data presence of fecal pollution, sewage, human viruses, and
comparing the resistance of naturally occurring phages human and/or animal fecal pollution. If bacteriophages
of all groups to pasteurization are not available. Data are not found in all stool samples and if bacteriophages
available for thermophilic and mesophyllic digestion do have a chance to originate in other sites than gut, there will
not allow to differentiate the thermal effect from chemical not be an absolute and unequivocal correlation between
effects (61). the presence of bacteriophages and the presence of fecal
pollution, sewage, human viruses, and human and/or fecal
Accumulation and Depuration of Fecal Pathogens by
pollution. However, if bacteriophages are always present
Shellfish
in sewage, and if sewage, raw or treated, constitutes the
The presence of fecal pathogens and indicators in shellfish, main input of fecal pollutants into the environment, the
providing there is no replication, depends on many factors, presence of given numbers of certain bacteriophages in
but, briefly, it is the result of accumulation and/or waters will indicate the likely presence of the surrogate
depuration. fecal microorganism or microorganisms.
There is no relationship between the levels of Moreover, bacteriophages may be used to indicate the
bacterial indicators and viruses in the water and behavior of pathogens other than bacteria during the
their concentrations in shellfish, where most of them processes of removal and inactivation in nature, water,
362 BACTERIOPHAGE AS INDICATORS

and sludge treatment processes. In general, the behav- 4. M. Muniesa, F. Lucena, and J. Jofre, J. Appl. Microbiol. 87,
ior of bacteriophages resembles more to the behavior 402–409 (1999).
of persistent pathogens (i.e., animal viruses and proto- 5. N. D. Seeley and J. B. Primrose, J. Gen. Virol. 46, 87–95
zoa) than the behavior of currently accepted bacterial (1980).
indicators, although different groups of bacteriophages 6. J. M. Vaughn and T. G. Metcalf, Water Res. 9, 613–616
show different degree of persistence to different remov- (1975).
ing processes and resistance to different inactivating 7. J. Jofre et al., Appl. Environ. Microbiol. 61, 3227–3231
factors. (1995).
Each one of the three groups of bacteriophages proposed 8. Anonymous, Standard Methods for the Examination of
as model organisms has advantages and drawbacks. Water and Wastewater, 18th ed., American Public Health
Somatic coliphages are the most abundant, and the method Association, Washington, DC, 1992.
for their detection and enumeration is the most simple 9. Anonymous, ISO 10705-2: Water Quality. Detection and Enu-
and fast, with results available in one working day. On meration of bacteriophages-part 2. Enumeration of somatic
the other hand, they are a heterogeneous group with coliphages, Geneva, Switzerland: International Organization
for Standardization, 2000.
important differences in resistance to inactivation among
its members, and it has been reported that they may 10. M.A. Woody and D. O. Cliver, J. Appl. Microbiol. 82, 431–440
(1997).
replicate outside the gut. This fact is without any doubt
the main shortcoming of somatic coliphages. However, 11. Anonymous, ISO 10705-1: Water Quality. Detection and Enu-
meration of bacteriophages-part1: Enumeration of F-specific
the contribution of this potential replication in the
RNA bacteriophages, Geneva, Switzerland: International
different water environments has never been quantified. Organization for Standardization, 1995.
Because of the ease of the method for their detection
12. Anonymous, ISO/DIS10705-4: Water Quality. Detection
and their abundance, this potential contribution should be and Enumeration of bacteriophages-part 2. Enumeration
estimated before rejecting them as model organisms for of bacteriophages infecting Bacteroides fragilis. Geneva,
water quality control. Switzerland: International Organization for Standardization,
F-specific RNA bacteriophages rank second in abun- 1999.
dance and they are a homogeneous group. The method 13. W. O. K. Grabow, C. S. Holtzhausen, and J. C. De Villiers,
for detecting them is simple and fast, although not as Research on Bacteriophages as Indicators of Water Qual-
much so for the somatic coliphages. They offer good per- ity, Project Report No. 321, Water Research Commission,
spectives as index organisms for viruses in groundwater Pretoria, South Africa, 1993.
and to monitor some water treatments, for example, UV 14. A. H. Havelaar, K. Furuse, and W. H. Hogeboom, J. Appl.
disinfection. On the contrary, their persistence in surface Bacteriol. 60, 255–262 (1986).
waters, mainly in moderate and warm climates, seems 15. W. O. K. Grabow et al., Water Sci. Technol. 31, 223–230
low, and their resistance to some inactivating treatments (1995).
is low or intermediate as compared with persistent virus 16. S. Osawa, K. Furuse, and I. Watanabe, Appl. Environ.
or protozoa. Microbiol. 41, 164–168 (1981).
Bacteriophages infecting B. fragilis rank third in 17. K. R. Calci et al., Appl. Environ. Microbiol. 64, 5027–5029
abundance. They are a homogeneous group. They are (1998).
more resistant than the other groups of phages to most 18. A. Puig et al., Appl. Environ. Microbiol. 65, 1772–1776
inactivating factors and treatments and they do not seem (1999).
to be able to replicate outside the gut. The method for their 19. C. Tartera and J. Jofre, Appl. Environ. Microbiol. 53,
detection requires anaerobiosis and results are available 1632–1637 (1987).
only after 18 hours. Their low numbers constitute their 20. J. Jofre et al., Determination of Bacteriophages in Bathing
main drawback. Waters, EUR Report EN 19506, Office for Official Publications
In view of the present knowledge, it seems unlikely of the European Communities, Luxembourg, Belgium, 2000.
that we can find a single group of bacteriophages 21. T. J. Nieuwstad et al., Water Res. 22, 1389–1397 (1988).
fulfilling all the requirements to be used as an universal 22. A. H. Havelaar and W. M. Hogeboom, J. Appl. Bacteriol. 56,
indicator, but this seems to be true for all the potential 439–447 (1984).
microbial indicators. Now that standardized methods are 23. C. Tartera, F. Lucena, and J. Jofre, Appl. Environ. Microbiol.
available, more research in the ecology of these groups of 55, 2696–2701 (1989).
bacteriophages should be done to guarantee the best choice 24. H. Chung and M. D. Sobsey, Water Sci. Technol. 27, 425–429
for the various potential applications of bacteriophages as (1993).
model microorganisms. 25. H. Chung et al., Water Sci. Technol. 38, 37–44 (1998).
26. J. Lasobras et al., J. Appl. Microbiol. 86, 723–729 (1999).
BIBLIOGRAPHY 27. F. P. Williams Jr. and C. J. Hurst, Water Res. 22, 847–851
(1988).
1. IAWPRC Study Group on Health Related Microbiology, Water 28. Y. Kott et al., Water Res. 8, 165–171 (1974).
Res. 25, 529–545 (1991). 29. R. Stetler, Appl. Environ. Microbiol. 48, 668–670 (1984).
2. R. Armon and Y. Kott, Crit. Rev. Environ. Sci. Technol. 26, 30. H. Havelaar and W. M. Hogeboom, Antonie van Leeuwenhoek
299–335 (1996). 49, 387–397 (1983).
3. H. W. Ackermann and T. Nguyen, Appl. Environ. Microbiol. 31. A. H. Havelaar, M. Van Olphen, and Y. Drost, Appl. Environ.
45, 1049–1059 (1983). Microbiol. 59, 2956–2962 (1993).
 BACTERIOPHAGE: BIOLOGY AND GENETICS 363

32. W. J. Doré, K. Henshilwood, and D. N. Lees, Water Sci. 66. F. Lucena and J. Jofre, in Microbiologia de las Aguas de
Technol. 38, 29–35 (1998). Abastecimiento (Tomo II), Asociación Española de Abastec-
33. J. Jofre et al., Water Sci. Technol. 21, 15–19 (1989). imiento de Agua y Saneamiento, Madrid, Spain, 1997,
pp. 37–47.
34. K. Furuse, in S. M. Goyal, C. P. Gerba, and G. Bitton, eds.,
Phage Ecology, Wiley & Sons, New York, 1987, pp. 87–124. 67. A. H. Havelaar et al., Ozone: Sci. Eng. 9, 353–368 (1987).
35. H. Fu-Chih et al., Appl. Environ. Microbiol. 61, 1960–3966 68. R. Sommer et al., Water Sci. Technol. 38, 145–150 (1980).
(1995). 69. H. Dizer et al., Water Res. 27, 397–403 (1993).
36. W. O. K. Grabow, C. S. Holzhausen, and C. J. Williers, Eval- 70. S. Farooq et al., Water Res. 27, 1177–1184 (1993).
uation of the Application of Bacteriophages as Indicators of 71. R. Sommer et al., Water Res. 35(3), 109–117 (2001).
Water Quality, Project Report No. 540/1/98, Water Research 72. W. Burkhard III, W. D. Watkins, and S. R. Rippey, Appl.
Commission, Pretoria, South Africa, 1997. Environ. Microbiol. 58, 826–831 (1992).
37. P. Payment et al., Can. J. Microbiol. 34, 907–910 (1988). 73. F. Lucena et al., Appl. Environ. Microbiol. 60, 2272–2277
38. R. M. Araujo et al., J. Appl. Microbiol. 82, 281–286 (1997). (1994).
39. M. Abbaszadegan, P. Stewart, and M. LeChevalier, in Water 74. W. J. Doré and D. N. Lees, Appl. Environ. Microbiol. 61,
Quality Conference Proceedings, American Water Works 2830–2834 (1995).
Association, Denver, Colorado, 1997. 75. U. F. Power and J. K. Collins, Appl. Environ. Microbiol. 55,
40. R. Armon et al., J. Appl. Microbiol. 83, 627–633 (1997). 1386–1390 (1989).
41. J. S. Meschke and M. D. Sobsey, Water Sci. Technol. 38,
187–189 (1998).
42. A. M. Nasser, Y. Tchorch, and B. Fattal, Water Sci. Technol. BACTERIOPHAGE: BIOLOGY AND GENETICS
27, 401–405 (1993).
43. M. V. Yates, C. P. Gerba, and L. M. Kelly, Appl. Environ. SCOT E. DOWD
Microbiol. 49, 778–781 (1985).
USDA-LIRU
44. R. G. Bell, Water Res. 10, 185–187 (1976). Lubbock, Texas
45. M. M. Moriñigo et al., Water Res. 26, 267–271 (1992).
46. B. O’Keefe and J. Green, Water Res. 23, 1027–1030 (1989). Bacteriophages (phage), by definition, are viruses that
47. R. Gironés, J. Jofre, and A. Bosch, J. Environ. Qual. 18, infect bacteria. In general, phages have a proteinaceous
34–39 (1989). outer shell with an inner core containing their genetic
48. L. W. Sinton et al., Appl. Environ. Microbiol. 65, 3605–3613 material. They usually infect a host bacterium by
(1999). attaching to the bacterial cell via specific host receptor
49. G. Ignazzitto et al., Water, Air, Soil Pollut. 13, 391–398 proteins and injecting their nucleic acids into the bacteria.
(1980). The phages then use the chemical energy and biochemical
50. J. Lasobras, Ph.D. Thesis, University of Barcelona, Barce- machinery of the host to produce new phage particles.
lona, Spain, 1997. The phage is ultimately released from the bacterial cell
51. V. C. Rao and J. L. Melnick, Aspects of Microbiology. Envi- by budding or extrusion from the plasma membrane or
ronmental Virology, ASM, Washington, DC, 1986. even complete lysis of the host cell. Phages are in most
52. T. Omura et al., Water Sci. Technol. 21, 119–124 (1989). instances, host-specific, with particular phage infecting
only certain bacterial species. Since their discovery in
53. P. Scheuerman, S. Farrah, and G. Bitton, Water Sci. Technol.
19, 539–546 (1987).
1915 by the British investigator Frederick W. Twort, and
the introduction of the term bacteriophage by Felix Hubert
54. R. M. Gersberg et al., Appl. Environ. Microbiol. 53, 731
(1987).
d’Herelle (1), bacteriophage have become important tools,
not only in the studies of bacterial genetics and cellular
55. A. Bourrouet, Ph.D. Thesis, Politechnical University of
mechanisms but also in the field of environmental
Catalunya, Barcelona, Spain, 2000.
microbiology. Similar to other disciplines, phage are used
56. A. M. Nasser, H. Dizer, and J. M. Lopez, Monogr. Virol. 15,
primarily as models or surrogates to assess the fate of
163–170 (1984).
pathogenic human enteric viruses, as pathogen indicators,
57. B. Dutka et al., Environ. Pollut. 63, 293–296 (1990).
as vectors for transport of genetic information, and even
58. R. S. Wentsel, P. E. O’Neil, and J. F. Kitchens, Appl. Environ. as biocontrol agents.
Microbiol. 45, 430–443 (1982).
Unlike molecular biology that makes use of the
59. P. Payment and E. Franco, Appl. Environ. Microbiol. 59, phage as tools for genetic engineering, environmental
2418–2424 (1993). microbiologists more often use phage such as MS2,
60. Z. P. Sun et al., Water, Air, Soil Pollut. 96, 175–183 (1997). PRD1, and PhiX174, which structurally and biochemically
61. S. K. Spillman et al., Appl. Environ. Microbiol. 53, 2077– resemble human pathogenic enteric viruses, such as
2081 (1987). Poliovirus. These spherical bacteriophage are used as
62. F. X. Abad et al., Appl. Environ. Microbiol. 60, 2377–2383 indicators or model viruses not only because they have
(1994). similar size and shape to important human pathogenic
63. C. Tartera, A. Bosch, and J. Jofre, FEMS Microbiol. Lett. 56, viruses but more importantly because they are much
313–316 (1988). easier and less expensive to assay. The coliphage MS2,
64. S. W. Dee and J. C. Fogleman, Appl. Environ. Microbiol. 58, in particular, has been used extensively as an indicator
3136–3141 (1992). and model for enteric viruses such as Poliovirus, which
65. A. H. Havelaar, Microbiol. Sci. 12, 362–364 (1987). is a member of the Picornaviridae family. Table 1 lists
364 BACTERIOPHAGE: BIOLOGY AND GENETICS

Table 1. Comparison of the Phage MS2 and the Entero-

virus Poliovirus Capsid The protein shell that encloses
MS2 Poliovirus and protects the nucleic acid
Feature Leviviridae Picornaviridae of the phage
Structural units The smallest functional
equivalent building units of
Particle the capsid
Weight (Mdalton) 3.9 8–9 Nucleocapsid The capsid together with its
Diameter (nm) 24 27b enclosed nucleic acid
Isoelectric point ∼3.9 ∼8.2a Virion The complete infective virus
Major capsid proteins, number 1 4 particle
Symmetry Icosohedral Icosohedral
Enveloped None None
Triangulation number T=3 T=1 These structures are highlighted in Figure 1.

RNA
TAXONOMY
Sense (positive sense acts as mRNA) Yes Yes
Molecular weight (Mdalton) 1.2 2.5
Monocistronic No Yes The International Committee on Nomenclature of Viruses
Overlapping genes Yes or No No was established in 1966 to develop classification schemes
Replication form infectious No Yes for viruses and phages (6). Since its establishment, there
has been dispute over the taxonomic systems proposed by
Multiplication the committee. One such system acts as an all-embracing
Translation sequential Yes No scheme for the classification of viruses and phages
Cleaved precursor No Yes into subphyla, classes, orders, suborders, and families
RNA enters preformed capsid No Yes (Table 2). In this type of scheme, descending hierarchical
divisions would be based on characteristics such as nucleic
Source: Adapted from H. Ackermann and M. S. DuBow, Viruses of
acid type, capsid symmetry, and presence or absence
Prokaryotes, vol. 1, General Properties of Bacteriophages, CRC Press, Boca
Raton, Fla., 1987, pp. 1–202; H. Ackermann and M. S. DuBow, Viruses of of an envelope. Although the debate over classification
Prokaryotes, vol. 2, Natural Groups of Bacteriophages, CRC Press, Boca schemes continues, David Baltimore (7) devised a scheme
Raton, Fla., 1987, pp. 1–242. for classification of viruses and phage, which is now
a
Floyd and Sharp (4) coming to be widely accepted. This scheme is based on the
b
Burke and coworkers (5)
relationship between the viral or phage genome and the
viral or phage mRNA. Table 3 shows the major families of
bacteriophage grouped according to the Baltimore scheme.
some of the similarities and differences between MS2 and The taxonomic nomenclature used to describe bacterio-
Picornavirus. phage is different from that used in the naming of animal
Many of the specific areas where phages are important and plant viruses. Animal and plant viruses are typically
in the field of environmental microbiology are covered in named in one of three ways: (1) after the disease; (2) after
other entries of this encyclopedia and therefore will only the discoverer of the virus; or (3) after the location where
be mentioned briefly in this entry. The main purpose of the virus was discovered. Viruses are also classified at the
this section is to describe the basics of bacteriophage struc- family (-viridae), genus (-virus), and species level.
tures, taxonomy, and genetics. This section concentrates Originally, bacteriophages were only classified into
on the spherical coliphage MS2 and the tailed lambda families. Exceptions now exist to this rule. Phages unlike
phage. The lambdoid phage may be important in environ- plant and animal viruses are named exclusively with
mental exchange of genetic material between bacteria, numbers and letters. For instance, PhiX174 was the
which could be important in many disciplines within 174th virus characterized in the 10th group of a large
environmental microbiology including bioremediation, in set of phage being classified, with phi (φ) being the
which the transport and exchange of genetic material can symbol that stands for phage. For instance, three genuses
help to increase the ability of naturally occurring bacteria Chlamydiamicrovirus, Microvirus, and Spiromicrovirus
to metabolize pollutants. The lambdoid phage may also are grouped into the family Microviridae. The Leviviridae
prove to be important in the environment, where genes are not only broken down into genuses but subgroups
encoding virulence factors such as Shiga toxin and antibi- as well. In the past, classification of the Leviviridae
otic resistance may be transferred from one bacterium to was based on serotyping, so classification schemes based
another. Such events could create a potentially pathogenic on six serogroups also exist. In the National Center
strain from an otherwise innocuous variety. for Biotechnology Information’s GenBank (October 1999),
the genera designations Allolevivirus and Levivirus are
listed under Leviviridae along with their serogroupings.
TERMINOLOGY Bacteriophage taxonomy continues to develop, mainly
assisted by the rapid development of genetic sequencing
Basic terminology for virus and phage include the technologies. For instance, the genome of a typical
following: phage can reasonably be determined in a matter of
 BACTERIOPHAGE: BIOLOGY AND GENETICS 365

(a) Icosahedral Table 2. Phage Families and Genera of Bacteriophage

nucleocapsid
Major
Bacteriophage Genus Examples of
Families Example Species (Host Organisms)

Myoviridae Myovirus • T4, T2, T6, Wphi, P2, 186,

Ac3, Acm1 •(Escherichia coli),
• PBSX, PBS2, beta22, 2C,
(Bacillus spp.),
• S-WHM1, S-PM2, N1
Naked icosahedral (Cyanobacteria),
• PSP3 (Salmonella spp.),
(b) Nucleocapsid • Twort (Staphylococcus spp.),
• RL1RES, RL2RES
(Rhizobium leguminosarum),
Sheath • HP1, S2, HP1 (Haemophilus
spikes Sheath fibers influenzae),
• A118, A500, A511 (Listeria
monocytogenes),
Siphoviridae Siphovirus • Lambda, 21, T5–933W
(E. coli)
• A118 (Listeria monocytogenes)
Tail spikes • B40–8 (Bacteroides fragilis)
Base • FC1 (Enterococcus spp.)
plates • phi-41 (Lactococcus lactis)
Complex virus • phi-C (Staphylococcus aureus)
• psiM2 (Methanobacterium
(c) thermoautotrophicum)
• SFI11 (Streptococcus
Icosahedral thermophilus
nucleocapsid
Podoviridae Podovirus • T7 H-19B, N4, (E. coli)
• phiAAU2 (Arthrobacter
Nucleic acid aureus)
• B103, phi29, H1, NF, Phi-15,
Nucleic acid spilling out of nucleocapsid Phi-29, (Bacillus spp.)
Figure 1. Schematic view of general bacteriophage and virus • Cf16 (Xanthamonas
structures. (a) Naked icosahedral structure, which is the basic campestris)
shape of many phage including Leviviridae and Microviridae. By • Cp-1, CP-7 (Streptococcus
definition, an icosahedron is composed of 20 facets, each shaped pneumoniae)
as an equilateral triangle, 12 vertices, 6 fivefold axes of symmetry • K11 (Klebsiella spp.)
passing through the vertices, 10 threefold axes extending through • Kvp18 (Kluyvera spp.)
each face and 15 twofold axes passing through the edges of an • LP7, P22, PS119 (Salmonella
icosahedron. Caspar and Klug (1962) later defined all possible spp.)
polyhedra by determining that because an icosahedron has 20 • Phi-13 (Staphylococcus
equilateral triangular facets, each of which is identical, there aureus)
must be 20 T structure units, where T is the triangulation • Phi H (Halobacterium
number given by the rule: T = Pf 2 , where P is any number of salinarium)
the series 1, 3, 7, 13, 19, 21, 31 and f is any integer. Typically, • Phi-vML3 (Lactococcus spp.)
for viruses the P value is 1 or 3. Morphological units can be • P1 Mycoplasma spp.)
clustered as 20 T trimers, 30 T dimers, or considered individually • phage V, Phage X (Shigella
as 60 T monomers. (b) Represents a complex structure such flexneri)
as Myoviridae possess. Indicated in this figure are the major Microviridae Microvirus • PhiX174, G4, S13, ST-1, U3,
features such as the tail spikes and base plates. (c) Represents a alpha 3, G14, PhiK (E. coli,
nucleocapsid with the nucleic acid spilling out. Salmonella spp., Shigella)
• PhiCPG1 (Chlamydia
psittaci)
days. Similarly, improvements in other molecular and
Corticoviridae Corticovirus • PM2 (Alteromonas spp.)
biochemical techniques are continuing to provide tools
• O6N-58P (Pseudomonas)
that allow for structural and topological characterization
Tectiviridae Tectivirus • PRD1 (E. coli, LT2,
of viruses and phages at the angstrom level of resolution.
Acinitobacter spp.,
Pseudomonas, and Vibrio),
STRUCTURE AND PATHOGENESIS • PR3, PR4, PR772, L17
(E. coli)
The architecture of bacteriophages and viruses in general • AP50 (Bacillus anthracis)
is an elegant molecular structuring of protein subunits, (continued overleaf )
366 BACTERIOPHAGE: BIOLOGY AND GENETICS

Table 2. (Continued) The description by Watson and Crick of the molecular

Major
structure of DNA (8) was followed by yet another macro-
Bacteriophage Genus Examples of molecular structural elucidation. Crick and Watson (9)
Families Example Species (Host Organisms) reasoned that the small genomes of viruses could encode
for only a few proteins from which they had to build
• PhiNS11 (Bacillus their protein shells. They predicted that to use identical
acidocaldarius) subunits to construct such a shell, it would, by necessity,
• Bam35 (Bacillus possess a cuboidal symmetry. These predictions were later
megatarium),
confirmed in principle and the basic symmetrical structure
Leviviridae Levivirus • MS2, f2, fr, JP501, GA, JP34, of viruses were shown more precisely to be icosahedral.
Ku1, TH1, BO1, fr1, PP7,
Although not the only shape employed, an icosahedral
TW19, (E. coli)
Allolevivirus • Q-beta, M11, ST, VK, NL95,
structure is utilized by a wide variety of bacteriophage
MX1 (E. coli) and animal viruses. The basic shapes of bacteriophage are
• PRR1 (Pseudomonas schematically depicted in Figure 2, and, Table 4 lists the
aeruginosa) structural properties for the various phage families. In
Cystoviridae Cystoviridae • Phi6 (Pseudomonas addition to the variety of structural types found among
phaseolicola) the various bacteriophage families, there is also a variety
Inoviridae Inovirus • Fd, I2–2, If1, Ike, M13 of nucleic acid types. Table 5 provides information on the
(E. coli), nucleic acids for the various bacteriophage families.
• Pf1, Pf3 (Pseudomonas Morphologically, phages can be classified into several
aeruginosa) groups:
• Cf1c, Cf1t, Cf (Xanthomonas
campestris)
• fs-2, v6, Vf33, Vf12 (Vibrio 1. Naked icosahedral such as MS2virus, and phiX174
cholerae) 2. Naked helical such as Xf2, Pf1, and v6
Plectrovirus • MV-L1 (Acholeplasma spp.) 3. Enveloped icosahedral such as phi6
• Phage 1 (Spiroplasma spp.)
4. Enveloped helical of which there are no known
Source: Adapted from H. Ackermann and M. S. DuBow, Viruses of bacteriophage
Prokaryotes, vol. 1, General Properties of Bacteriophages, CRC Press, Boca
Raton, Fla., 1987, pp. 1–202.
5. Complex such as T4 and lambda

PATHOGENESIS
which provides the basis for infectivity, specificity, and
self-assembly. The requirements of self-assembly and Bacteriophage can be considered pathogenic to bacteria
host infection specificity have generated a remarkably and thus they follow infection processes common to
wide array of phage viroid structural organization and viral pathogens, including attachment, penetration, and
geometric designs. Nevertheless, certain common features release. Table 6 provides information on the pathogenesis
and general principles can be applied to most phages. variables of phage families.

Table 3. Baltimore Classifications for Bacteriophage

Kind of Single- or Genome Sense Naked or

Genome Double-stranded (if single strand) Enveloped Families

RNA SS Positive Naked Leviviridae

Enveloped N/D
Positive but Enveloped N/D
reverse-transcribed
Negative, nonsegmented Enveloped N/D
Negative, segmented Enveloped N/D
DS N/A(segmented) Enveloped Cystoviridae
DNA SS Negative (most often) Naked Inoviridae Microviridae
DS DS N/A Naked Tectiviridae Siphoviridae
Podoviridae Myoviridae
Lipothrixviridae Microviridae
Fuselloviridae Corticoviridae
N/A Enveloped Plasmaviridae
N/A, but with reverse Enveloped N/D
transcriptase step in
replication

Source: Adapted from D. Baltimore, Trans. N. Y. Acad. Sci. 33(3), 327–332 (1971).
 BACTERIOPHAGE: BIOLOGY AND GENETICS 367

Corticoviridae

Myoviridae

Tectiviridae

Siphoviridae

Leviviridae

Figure 2. Schematic representations of the structure

and relative sizes of major bacteriophage families. Indi-
Podoviridiae Inoviridiae cated here are virus structures such as the complex
structure of Myoviridae, the tailed phages Siphoviri-
dae, simple icosohedral such as Microviridae, spiked
icosohedral such as Tectiviridae, and filamentous such
Microviridae Cystoviridae as Inoviridae.

Table 4. Comparative Structural Properties of Bacteriophage Families

Buoyant Mol.b
Symmetry of Lipid density Weight
Phage group Nucleocapsid (%) Enva Size (g/mL,CsCl) (MD)

Microviridae Cubic, − − 27 1.41 6.7

Icosahedron
Corticoviridae Cubic, 13 − 60 1.28 49
Icosahedron
Tectiviridae Cubic, 16 − 63 1.28 70
Icosahedron
Leviviridae Cubic, − − 24 1.43 4
Icosahedron
Cystoviridae Cubic 23 + 75 1.27 90
dodecahedron
Inovirus genus Helical − − 760–1915 × 6 1.30 12–23
Plectrovirus Helical − − 85–250 × 14 1.37 ?
genus
Plasmaviridae Pleomorphic 12 + 70–90 ? ?
Myoviridae Complex-tailed ? − Capsid 84 Tail 158 1.48 140
Siphoviridae Complex tailed ? − Capsid 62 Tail 195 1.50 84
Podoviridae Complex tailed ? − Capsid 58 Tail 195 1.48 88

Source: Adapted from H. Ackermann and M. S. DuBow, Viruses of Prokaryotes, vol. 1, General Properties of
Bacteriophages, CRC Press, Boca Raton, Fla., 1987, pp. 1–202;
H. Ackermann and M. S. DuBow, Viruses of Prokaryotes, vol. 2, Natural Groups of Bacteriophages, CRC Press, Boca
Raton, Fla., 1987, pp. 1–242.
a
Envelope.
b
Molecular.

BASIC GENETICS OF RNA AND DNA PHAGE also allowed for genetic manipulation of their hosts,
quickly becoming a necessary tool for molecular micro-
Because of the relatively simple genomes, rapid repli- biologists. Because of their utility and the need for basic
cation, simple assay procedures, and ease of manipula- understanding, bacteriophage genetics was quickly unrav-
tion, early bacteriophages such as lambda became the eled and now serves as a model for genetic control and as a
favorite tool of many molecular microbiologists. Phages mechanism for protein expression. In this section, we will
368 BACTERIOPHAGE: BIOLOGY AND GENETICS

Table 5. Properties of Bacteriophage Families Nucleic viral survival and UV exposure modeling. MS2, or more
Acids precisely the F+ coliphage of which MS2 is the model
Family or Total Mol. Weight G+C organism, have also been used extensively as indicator
Genus Nature (%) (Md)a (%) viruses. Indicators are nonpathogenic microorganisms
whose presence in an environmental sample indicates
Microviridae D1C 26 1.7 44 that human pathogens may also be present.
Corticoviridae D2C 13 5.8 43 As mentioned, MS2 is a small, icosahedral bacterio-
Tectiviridae D2L 14 9.7 51 phage containing a single-stranded linear RNA molecule.
Leviviridae R1L 31 1.2 51 The MS2 capsid is composed of a single protein except for a
Cystoviridae R2L 12 10.4 58
single copy of an accessory protein known as the A-protein
Inovirus D1C 6–21 1.9–3 40–60
Plectrovirus D1C 1.8
(maturation protein). MS2 is a male-specific phage that
F3 group D2L 10 usually infects F+ or Hfr strains of E. coli. F+ refers to
Plasmaviridae D2C 7.6 32 bacteria that possess the F-plasmid, which encodes for the
Myoviridae D2L 43 108 45 F-pilus, a conjugation structure that allows for transfer of
Siphoviridae D2L 49 35 50 genetic material between two bacteria. MS2 attaches to
Podoviridae D2L 43 32 48 the sides of the pilus using the A-protein and injects its
genetic material into the hollow conjugation tube, allowing
Source: Adapted from H. Ackermann and M. S. DuBow, Viruses of
Prokaryotes, vol. 1, General Properties of Bacteriophages, CRC Press, Boca
migration of the phage RNA up through the conjugation
Raton, Fla., 1987, pp. 1–202; tube into the host bacterium.
H. Ackermann and M. S. DuBow, Viruses of Prokaryotes, vol. 2, Natural Because the RNA of MS2 is coding sense (positive-
Groups of Bacteriophages, CRC Press, Boca Raton, Fla., 1987, pp. 1–242. stranded RNA), it serves as its own mRNA, and once it
Note: D, DNA: R, RNA; 1, single-stranded, 2, double-stranded, C,
gains entrance to the cytoplasm of the host bacterium, it
circular: L, linear.
a
Mega Daltons. can immediately begin production of a new phage. MS2
codes for four known proteins, the coat protein, the A-
protein (maturation protein), the replicase protein, and
describe the genetics of two bacteriophages: MS2, a single- the L protein (lysis protein). Figure 3 shows the genetic
stranded RNA coliphage, and lambda, a double-stranded map of MS2 indicating the relative positions of these genes.
DNA coliphage. Two of the genes are coded within one reading frame, while
the second two are coded in another. To further conserve
on the size of the genome, the L-protein’s reading frame
RNA Phage: MS2
overlaps with that of the coat and the replicase proteins.
As noted earlier, MS2 is physically similar to the Fiers and coworkers (10) were the first to determine
enteroviruses such as Poliovirus. Because of this and its the complete sequence of the 3,569 ribonucleotide bases
easy assay procedures, MS2 soon became the classic model of the MS2 genome. Even if the sequence information,
of choice of environmental microbiologists in disinfection mapping of open reading frames, and the RNA polymerase
and survival experiments such as aerosol or groundwater binding sites is given, the regulation of the genome and

Table 6. Pathogenic Characteristics of Bacteriophage

Virus
Host Adsorption Attachment Localization of Release
Phage Site Structure Infection Method

Tailed phage Cell wall, Tail Nucleoplasm Lysis

capsule, pili, periphery
flagella,
Surface
protein
Microviridae Cell wall Spikes Nucleoplasm Lysis
Corticoviridae Cell wall Spikes Plasma Lysis
membrane
Tectiviridae Pili, cell wall Spikes, ‘‘tail’’ Nucleoplasm Lysis
Leviviridae Pili Apical protein Cytoplasm Lysis
Cystoviridae Pili Envelope Nucleoplasm Lysis
Inovirus Pili Virus tip Nucleoplasm Extrusion
Plectrovirus Plasma ? Plasma Extrusion
membrane membrane
Plasmaviridae Plasma Envelope ? Budding
membrane

Source: Adapted from H. Ackermann and M. S. DuBow, Viruses of Prokaryotes, vol. 1, General
Properties of Bacteriophages, CRC Press, Boca Raton, Fla., 1987, pp. 1–202;
H. Ackermann and M. S. DuBow, Viruses of Prokaryotes, vol. 2, Natural Groups of Bacterio-
phages, CRC Press, Boca Raton, Fla., 1987, pp. 1–242.
 BACTERIOPHAGE: BIOLOGY AND GENETICS 369

1 3569

Coat protein Replicase protein

A-protein L-protein
Figure 3. Schematic view of the MS2 genome showing the four encoded proteins. Note that there
are two open reading frames that overlap. The replicase gene overlaps the lysis (L) protein and the
L-protein overlaps the coat protein. The assembly (A) gene does not overlap with any of the other
genes. The arrows indicate the reading direction of each gene and the scale at the top indicates
the relative length of the genome.

its four encoded proteins is still difficult to understand The replicase protein is required for the formation of
without an understanding of the secondary structure the negative-strand template RNA. Replication of phage
of the MS2 genome. It is the secondary structure that RNA into a negative strand is a necessary process
actually controls expression of the four encoded proteins. for the phage to generate its genome. The replicase
By using secondary structure to control expression levels and three host-derived proteins combine to form a
of individual proteins, the MS2 genome adds elegance holoennzyme, which converts positive-strand RNA into
and economy, forgoing the need to encode for additional negative-strand RNA (RNA-dependent RNA polymerase).
regulatory proteins that would increase the size of the These negative-stranded RNAs are then converted back
genome. Regulation is accomplished through a process into positive-stranded (original sense) RNA genomes,
known as translational coupling. Translational coupling which are packaged into capsids during maturation of
refers to a gene arrangement in which the translation of the phage. Similar to the A-protein, the expression of the
one protein coding sequence on a polycistronic mRNA replicase also is controlled by the RNA tertiary structure
(coding for more than one gene) is required for the and translation of the coat protein, although the replicase
translation of the second downstream or upstream coding is located downstream of the coat protein. Thus, as the
sequence. coat protein is translated and the ribosome dissolves the
To grasp the concept of how MS2 uses the secondary secondary structure around the coat protein termination
structure of its RNA, it is important to realize that site, it exposes the ribosomal binding site of the replication
ribosomes, which translate the MS2 RNA genome into enzyme, allowing for its translation.
proteins, have an inherent ability to relieve mRNA The L-protein is encoded in a different reading frame
secondary structure (disrupt hydrogen bonds). However, than the coat protein. In addition, the reading frames
this can only occur when the ribosome is bound of these two proteins overlap. Thus, after the ribosome
to mRNA and actively moves along during scanning completes translation of a coat protein and relieves the
or translation. Thus, ribosomes are unable to relieve secondary structure that obscures the L-protein initiation
secondary structure before binding and scanning. Thus, site, it is required to release the mRNA, move backward
nucleic acid secondary structure of mRNA can prevent (upstream) to the now exposed binding site of the L-
initiation of macromolecular synthesis but does not inhibit protein, and initiate translation in a new reading frame.
initialized synthesis (11). This regulatory system based It is reasonable to assume that if the ribosome does not
on secondary structure controls the production of the quickly find this new open reading frame, or it diffuses in
A-protein (attachment/maturation protein), the replicase the wrong direction after release from the stop site of the
protein, and the L-protein (lysis protein) of MS2. The coat protein that the secondary structure (hairpin loop)
fourth protein is the coat protein, which is for the most part that blocks translation, can quickly reform (13). Thus,
freely transcribed as it is needed in higher concentrations RNA secondary structure regulates translation of the L,
than the other proteins. A, and replicase proteins and represents the primary form
In the case of the A-protein, its coding region is preceded of repression, that ensures the proper balance protein
by an untranslated leader of 130 nucleotides, which folds expression.
into a cloverleaf, (three stem-loop structures) covering the This mechanism of genetic regulation may be more
A-protein’s translation initiation region. Thus, as the coat important for the L-protein than for the A or replicase
protein is translated and the ribosome moves into the proteins, as a premature accumulation of the L-protein
coding region of the A-protein, the cloverleaf structure would cause lysis of the host cell before sufficient coat
is dissolved exposing the A-protein ribosomal binding protein has been made and all phage effectively packaged.
site, allowing binding of a ribosome and translation of In essence, overexpression of the lysis protein lowers the
the protein. Groeneveld and coworkers (12) indicated that burst size (number of phage produced per bacterium).
translational starts of the A-protein gene only take place By nature, the binding affinity for the polymerase is
before reannealing of the cloverleaf structure and after relatively low at the L-protein ribosome binding site.
translation of the capsid protein. Decreasing this affinity further can also delay or even
370 BACTERIOPHAGE: BIOLOGY AND GENETICS

prevent accumulation of the L-protein, which in turn can stress-related signals and remains incorporated in the
inhibit subsequent lysis of the host cell (11). host genome. In either case, whether the phage goes into a
Competition for ribosomes is another control method lytic cycle or stays in the lysogeny cycle, the phage is still
used during translation of A, L, and replicase. The able to ensure its own survival and propagation, making
ribosome leaving the coat protein gene can only bind to it a very successful parasite of E. coli.
one of these additional protein genes. Thus, competition
for ribosomes represents yet another level of control (11). Structure. The lambdoid phages are considered ‘‘com-
Three levels of expression control in MS2 include the RNA plex,’’ meaning that in addition to their icosohedral capsid
secondary structure, consensus of ribosome binding sites, (which is like MS2’s capsid), they also possess an addi-
and competition for ribosomes during or after translation tional tail-like structure that is used for attachment of the
of the coat protein ensures that proper ratios of the other phage to the host cell during infection. The more complex
three proteins encoded by MS2 are achieved. structure of a lambdoid phage can be contrasted to the
It is interesting to note that MS2 can self-assemble in simple structure of the MS2 capsid (Fig. 2). The lambdoid
vitro with no special enzymatic process involved (14,15). virion has an isometric head with two major capsid pro-
During this maturation process, the coat proteins teins, E and D, two minor proteins B and C, and portal
accumulate around the positive-strand RNA, and the proteins W and FII. The capsid is attached to a long flex-
capsid shape begins to form automatically. Finally, the A- ible tail composed of several proteins, culminating in a
protein is added at one end and the virion is completed. The single fiber of J, which recognizes the outer membrane,
A-protein is required in the mature phage for attachment maltose porin protein LamB (2,3).
to the host pilus during infection and injection of the The genetics of bacteriophage lambda is well known,
genetic material. At this point during the phage life cycle, and most of this section will deal with the prototypical
sufficient L-protein has accumulated and the cell wall of lambda phage genetics. Lambda has a genome of about
the host is considerably weakened. As the strength of the 46 kb. In the capsid, the lambdoid chromosome is a
wall is compromised, water tension causes the eventual linear dsDNA molecule usually with 12 nucleotide ssDNA
bursting of the host cell and release of the now complete cohesive ends. Table 7 provides the major genes found
phage particles (11). in lambda phage along with the function of the encoded
proteins.
DNA Phage: Lambdoid Phage
Andre Lwoof discovered lambdoid phage at the Pasteur Early Infection. Relying on diffusion, as opposed to any
Institute, Paris, France. He discovered that certain directed locomotion or motility, the lambdoid bacterio-
isolates of Bacillus spp., when exposed to ultraviolet light, phage attaches to the LamB receptor protein, which is
ceased dividing and then burst open, releasing hundreds of located in the bacterial cell outer membrane and which
bacteriophage particles. Later in collaboration with Jacob was later discovered to be the maltose porin. LamB protein
and Monod, he showed that these bacterial strains carried is responsible for transporting maltose from the environ-
the bacteriophage in a dormant form, termed a prophage, ment into the cell. Following contact between the J-protein
and that the phage could be induced to switch from the on the tail of the lambdoid phage and the host cells LamB
lysogenic (nonproductive) to the lytic (productive) cycles. protein, a conformational change results in the injection
Lambda has since become one of the most studied and of the phage DNA through the core of the hollow tail
best understood models for genetic control. Even today, structure, through the LamB core, and ultimately into the
many of the intricacies of the genetic control of lambda cytoplasm of the host cell. On entry into the bacterial host
are still being elucidated. The genetics of lambda has cell, the linear genome circularizes as the complimentary
been extensively studied and provides a valuable model
for regulatory control of gene expression.
Table 7. Major Proteins Encoded by Lambdoid Phage
Similar to MS2, the lambdoid phage can infect E. coli,
replicate, synthesize new phage, lyse the host cell, and Genes Function
release progeny phage as part of a lytic growth cycle.
Unlike MS2, however, lambda can also infect the bacterial nul, A DNA packaging
cell and its DNA and can then become integrated in the W, B, C, D, E, FII Head formation
host cell’s genetic material, where it is propagated as a part Z, U, V, G, T, H, M, L, K, I, J Tail formation
int, xis Integration, excision
of the bacterial chromosome. Whether the lambdoid phage
exo, bet, gam Recombination
enters the lysogenic cycle or the lytic cycle is dependent on
cIII Stabilization of cII
many factors including the health of the host cell and the N Early antitermination
multiplicity of infection (MOI). In the case of the discovery cI Repression (primary)
of the phage, Lwoff had exposed the host bacteria to UV cro Repression (secondary)
light, which ultimately signaled the lysogenized phage cII Turn on cI, int
that its host was in danger. In response to the signal, the O, P Replication
lysogenized phage was induced into the lytic cycle. Many Q Late antiterminator
forms of stress send signals to the lysogenized phage S, R, Rz Lysis
causing it to leave its dormant stage and enter into a lytic Source: Adapted from A. Campbell, in Frederick C. Neidhardt et al., ed.,
cycle. In the event that the host cell is in a vegetative Escherichia Coli and Salmonella: Cellular and Molecular Biology, 2nd ed.,
state, the lambdoid phage does not receive the required ASM Press, Washington, D.C., 1996.
 BACTERIOPHAGE: BIOLOGY AND GENETICS 371

1 10,000 20,000
1 28,000 40,000 48,502

Head Tail Int Xis Clll N Cl Cro Cll O P Q S R Rz

proteins proteins
Attp
Cos Cos
PI TL PL PRM PR Tr2 PRE Tr2 PR’

Figure 4. Schematic view of the lambda genome showing the major genes associated with the
control of the lysogenic or lytic decision. To the left end of the genome, the bp scale indicated on
the top of the figure is condensed to provide clarity to the early genes. Arrows toward the bottom
indicate the direction of transcription, which originates from promotors (P). The genome is laid
out as it would be in the nucleocapsid before the sticky ends (cos) join after entry into the host.
Note the location of the attP site where integration with the host genome occurs.

3 overhanging cos sites (sticky ends) present on each end originating from PRm (leftward) because PRm is a weak
of the phage genome hybridize with each other and are promoter (16).
sealed by DNA ligase to form a circular genome. Thus, N, cro, and to a lesser extent cI are produced
During early infection, several components contribute in the cell soon after injection of the phage DNA. The N
to the decision of whether the phage will enter the lytic protein is an antiterminator, which associates with the
or the lysogenic cycle. These components include three terminators TL , TR1 , and TR2 and consequently allows
promoters (PL , PR , and PRm ), two operators (OL and OR ), transcription of longer mRNA. Additional phage proteins
and a group of regulatory proteins (cI, cII, cIII, cro, and are transcribed once sufficient N is produced, including
N). The location of these components on the phage genome cII from rightward transcription and cIII from leftward
can be seen in Figure 4. An introductory overview of these transcription. The N protein performs its antitermination
factors is as follows: function through a complex interaction with several host
proteins, RNA, and the host RNA polymerase. Thus, there
was an increase in the concentration of N, cro, cI, and now
• PL , PR , and PRm are promoters responsible for
the production of two new proteins, cII and cIII. These
transcription.
events lead the bacteriophage to the next stage in the
• OL and OR are regulatory protein binding sites. Lambda infection termed the decision-state.
• cI is the major regulatory protein and is transcribed
from two different promoters (PRm and PRE ). cI Making the Decision Between the Lytic and Lysogenic
represses transcription from PL and PR but allows State. At this point in early infection, lambda must make
transcription from PRm . the choice between a lytic and a lysogenic cycle. The
• cII and cIII encode activator proteins, which coop- decision is controlled by the regulators and operators of
eratively bind DNA regulatory sites and enhance lambda, but is ultimately decided by the environment and
transcription of the cI gene from PRE . consequently by the physiological state of the host. For
• cro is a repressor, which competes with cI to bind the example, if the host is experiencing any stress such as
DNA damage, starvation, and so forth, the lytic pathway
OR sites.
is favored. In other words, if the host has been damaged by
• N encodes an antiterminator protein, which acts by radiation from ultraviolet light or is experiencing a limited
replacing rho factor for host cell RNA polymerase. It supply of a growth factor such as tyrosine, the phage
modifies host RNA polymerase activity and prevents presumably decides to move onto a more suitable host.
termination at the termination sites (TL , TR1 , and Recent research has revealed that the regulatory molecule
TR2 ). responsible for the lytic-lysogenic signal may be Guanosine
tetraphosphate (ppGpp) (17), which is a nucleotide that is
After circularization of the phage genome within the synthesized by E. coli cells in response to amino acid or
host cell, the bacterial genetic machinery, considering carbon source starvation. Too low and too high levels of
the newly introduced circular DNA to be ‘‘self,’’ begins ppGpp resulted in less efficient lysogenization. If the host
transcription and translation from the phage genome. is growing, and more importantly, dividing normally and
As can be seen from Figure 4, this primarily results has not been exposed to any harmful chemicals or DNA-
in the production of three proteins. Transcription and damaging radiation, the phage considers the bacteria to
translation from PR (rightward) leads to production of cro, be a good host, and the lysogenic cycle is more likely
with transcription termination at the right transcription to be chosen. Lambda makes this decision using a very
terminator site (TL1 ). Transcription and translation from elegant mechanism of genetic balance among operators,
PL (leftward) leads to production of N and termination promoters, and regulatory proteins, centered around the
at the right transcription terminator site (TR ). There is intracellular levels of the cI protein. If the phage can
also a lower level of transcription and translation of cI produce sufficient cI before too much cro is built up, it will
372 BACTERIOPHAGE: BIOLOGY AND GENETICS

N OL1 OL2 OL3 Cl OR3 OR2 OR1 Cro

Figure 5. Schematic view of the focal area of the operator sites to which cro and cI bind. Arrows
indicate the direction of transcription. OL and OR both have three binding sites for cI, and each
binds one cro or cI dimer as described in the text. As indicated in the text, cI first binds to OL1 and
OR1 and then to the other operator sites in a cooperative manner. The carboxy-terminal domains
of the repressor (cI) dimers mediate this cooperativity. Cro has the opposite affinity binding first
to OR3 , which represses transcription of PRM .

enter lysogeny. The phage genome will be integrated into Lysogenic Decision. Lysogeny in lambda is character-
the bacterial genome as a prophage. If too much cro is ized by events that lead to integration or incorporation of
produced before cI reaches the critical level, the cell enters the lambda genome into the host chromosome. Integra-
the lytic cycle, resulting in expression of phage structural tion is accomplished through site-specific recombination
genes, replication of the phage genome, production of between the lambda genome and the host chromosome. A
mature phage, and lysis of the host cell. lambda-encoded recombinase enzyme known as integrase
Cro and cI compete for binding sites on OR . This is the (Int) mediates the recombination event. Int, together with
operator that ultimately decides the fate of the bacteria several host and phage accessory proteins, inserts the
and the bacteriophage. Two properties of the OR region phage genome into the host chromosome using specific
enable it to provide sensitive regulation of the phage life sites in both genomes known as attachment sites (att) (24).
cycle. First the OR region has three binding sites for cI The attachment sites on both genomes, attB for the bac-
or cro, which from left to right are termed OR3 , OR2 , teria and attP for the Phage, share 15 bp of identical
and OR1 (Fig. 5). Secondly, cro and cI, while they bind sequence. It is important to realize that even during opti-
to the same locations on the OR , do so with different mum environmental conditions, a substantial fraction of
affinities for different binding sites. The cI repressor binds lambda enter the lytic cycle. Some conditions, such as high
with affinity OR1 > OR2 , >OR3 ; binding of cro is with an multiplicity of infection or the presence of divalent cations
affinity OR3 , >OR1 , >OR2 . Occupancy of sites OR1 and OR2 or cyclic AMP can increase the proportion of cells that go
by cI facilitates the constitutive production of cI from toward lysogeny.
PRM as required for lysogenic growth, while precluding Thus, in what may be considered a rare occasion when
the concentration of cII builds up, transcription of cI
transcription of cro from PR . Similarly, cro bound to OR3
is enhanced and intracellular levels of the cI repressor
inhibits transcription of cI from PRM .
protein rise in relation to cro lambda enters the lysogenic
Transcription of cI originating from PRm by itself
cycle. The cI protein binds to OR and OL , preventing
cannot compete with the transcription of cro during early
transcription of all late infection (structural, reproductive,
infection. If the levels of cro reach a threshold level before
lysis) phage genes from the early promoters PL and
cI is able to turn off its synthesis by binding the OR
PR . At this point, the level of cI protein is maintained
sites, then the phage becomes committed to the lytic cycle.
automatically by a negative feedback mechanism at
There is, however, a very strong cascade of genetic events
PRM . Protein cII, during the time when it successfully
and kinetic reactions, which tip the balance back toward upregulated expression of cI, also turns on the promoters
cI if cellular conditions are adequate (18). For instance, that control the prophage int gene (25).
proteins cII and cIII assist cI in at least two ways: they Integration is very efficient and requires no external
stabilize and protect cI from degradation by host proteases energy sources. As the lambda DNA and the host
such as RecA and they interact with two additional chromosome come together, they form an integration
promoters on the lambda genome (PI and PRE ) (19,20). PI is complex, which mediates synapsis and strand exchange.
the promoter in charge of the proteins needed for lysogeny The integration complex consists of the attB, attP, the
(int and xis) and PRE is important in the decision-state bivalent DNA binding protein Int, and host proteins.
of lambda. PRE is located just downstream to the left of The host proteins include DNA gyrase, which introduce
the cII gene and controls leftward transcription of a cro negative supercoils into DNA (26), integration host factor
antisense RNA and the cI mRNA (positive-sense). Thus, (IHF) that bends attP DNA (27,28), and Fis, which assists
additional cI is transcribed from PRE , which becomes a in integration (29). Int cleaves each strand at attB and
stronger promoter when associated with cII/cIII (21,22), attP, creating strand exchanges by joining broken ends
which stabilizes cI and is thought to be the sensory from each (30). Following the initial exchange, there is a
molecule for multiplicity of infection (MOI). MOI is the ‘‘resolving’’ exchange between the other two DNA strands,
number of phage that infect a cell at one time. It has been resulting in an integrated phage genome.
observed that a high multiplicity of infection enhances the At this point, the genome of the prophage is almost
expression of the cIII and cII genes, progressively delays completely shut down. The cI protein has accumulated
lysis time, and increases the rate of lysogenization (23). and is bound to the operator sites on the phage genome.
 BACTERIOPHAGE: BIOLOGY AND GENETICS 373

There is a balancing act between cI production from the chromosome. A delicate equilibrium exists between stable
PRM promoter and the normal degradation of cI by the lysogeny and induction. Induction is the stimulation of
host. As long as cI is the only protein being expressed transcription resulting in excision of the phage genome
from the phage genome, the result is a lysogenized E. coli from the host chromosome and, ultimately, movement
bearing an integrated lambda prophage (31). into the lytic cycle. The health of the host and its
responses to environmental stimuli determine whether
Lytic Decision. In order to enter the lytic growth cycle, the phage will remain within the host genome or enter the
cro must bind to the three operator-sites in OL and OR , induction process. The primary mechanism that controls
repressing all transcription from PL and PR . This binding induction is the availability of cI. Physiological stress such
ultimately prevents expression of N, which consequently as UV damage results in the activation of a host cell
prevents expression of cII and cIII. Thus, cro turns off SOS response. The expression of the genes in the SOS
all early gene expression. During early infection, the regulatory network is controlled by a complex circuitry
transcript from PL , which produces cro and cII, also involving the RecA and LexA proteins (28,34–36). As
produces three other proteins, which accumulate in the indicated aRecA’s normal substrate is LexA but it
host. O, P, and Q proteins are essential for the development also nonspecifically degrades cI (lambda repressor),
of the lytic cycle. Q is an antiterminator, which allows RNA stimulating the dormant prophage to enter the lytic cycle.
polymerase to override the termination signal downstream The nonspecific cleavage of cI by aRecA renders it
of PR so that it can transcribe the lysis genes (S, R, and inactive and it falls off the operator sequences OL and
Rz) and the structural and accessory genes (Nu1, A, W, OR . This allows E. coli RNA polymerase to transcribe
etc.). O and P are utilized for replication of the phage the gene for cro and Xis and Int, which were under
genome. Thus, at the point where the lytic decision for the repression by cI. The Xis and Int proteins interact to
phage genomic DNA replication is well under way owing form an excision complex, which more or less reverses the
to the background levels of the O and P proteins, the copy integration process and effects induction of the prophage
number of late genes available for activation is increased. from the host chromosome. From this point, the lambda
Replication of the lambda genome early in the lytic genome circularizes and follows a lytic infection cycle as
cycle is bidirectional (Theta-form replication) and initiates described previously. It should be noted that induction is
from an origin within the O gene. Protein O recognizes irreversible. One probable explanation is that in a lysogen
the four 19-bp palindromes of the origin and protein P there is a MOI of only one (the prophage itself). After
binds to the host-derived primase DnaB, recruiting it induction, the lytic cycle, as described previously, leads
into the replication complex. Ultimately during viral DNA to morphogenesis, lysis of the host cell, and release of
replication, multigenomic tails of double-stranded DNA progeny phage particles.
are spun out from rolling circles. These long pieces of DNA
containing repeated lambda genomes become substrate CONCLUSION
for DNA packaging (16). During rolling, circle as the DNA
is spooled out, it is cut into correctly sized pieces by a Bacteriophages have been and will continue to be used by
phage-encoded enzyme called Ter. Ter cuts at the cos sites environmental microbiologists as indicators and model
releasing mature phage lambda DNA ready to be packaged organisms. Simple and inexpensive assay procedures
into the phage head. create a remarkably easy system that can provide a
Morphogenesis is the process of phage particle assembly wealth of information not only about the environmental
that takes place in the host cell nucleoplasm periphery fate of the phage themselves but also as a model system
and includes formation of the phage capsid, condensation for the study of eukaryotic viruses. From their use
of DNA, packaging of the DNA into the capsid, and as airborne indicators of fecal pollution, as models for
attachment of the tail. Morphogenesis results in mature studying the subsurface transport of viruses in aquifers, as
phages that are capable of infecting a new host. During indicators of fecal pollution in surface and drinking water,
morphogenesis of the phage particles, there is a concurrent as vectors for the transmission of genetic information
buildup of lysis proteins, which ultimately results in lysis between species of bacteria and as simple models to help
of the host cell and release of the infective lambda particles unravel the complexities of genetic control, bacteriophage
into the extracellular environment. Lambda has at least have been instilled into the core of most aspects of
two lysis genes, S and R (32). Protein R is an enzyme microbiology.
that cleaves between N-acetylglucosamine residue bonds
similar to the action of lysozyme. Protein S is a holin-
BIBLIOGRAPHY
like protein, which disrupts inner membranes of the
host, allowing R access to the N-acetylglucosamine in 1. F. d’Herelle, Acad. Sci. Ser. D. 165, 373 (1917).
the murein layer. Ultimately, the infection cycle results 2. H. Ackermann and M. S. DuBow, Viruses of Prokaryotes,
in the production and release of 50 to 100 infective phage Vol. 1, General Properties of Bacteriophages, CRC Press, Boca
particles per lysed cell (33). Raton, Fla., 1987, pp. 1–202.
3. H. Ackermann and M. S. DuBow, Viruses of Prokaryotes,
Induction: The Switch from Lysogenic to Lytic. Lysogeny, Vol. 2, Natural Groups of Bacteriophages, CRC Press, Boca
once established, is relatively stable. The cI repressor Raton, Fla., 1987, pp. 1–242.
continues to stimulate its own synthesis and the inserted 4. R. Floyd and D. G. Sharp, Appl. Environ. Microbiol. 38,
prophage is passively replicated as part of the bacterial 395–401 (1979).
374 BACTERIOPHAGE DETECTION METHODOLOGIES

5. K. L. Burke et al., Nature 332, 81–82 (1988). NENA NWACHUKU

6. P. Wildy et al., Prog. Med. Virol 9, 476–482 (1967). United States Environmental
7. D. Baltimore, Trans. N. Y. Acad. Sci. 33(3), 327–332 Protection Agency
(1971). Washington, D.C.
8. J. D. Watson and F. H. C. Crick, Cold Spring Harbor Symp.
Quant. Biol. 18, 123 (1953). Bacteriophages are viruses that infect bacterial cells. A
9. F. Crick, and J. Watson, Nature 177, 473 (1956). variety of different bacteriophages or phages (as they
10. W. Fiers et al., Nature 260, 500–507 (1976). are commonly referred to) that infect different bacterial
11. E. A. Birge, Bacterial and Bacteriophage Genetics, 3rd ed., cells have been reported in the literature. Phages that
Springer-Verlag, New York, 1994. infect the coliform group of bacteria are termed coliphages.
12. H. Groeneveld et al., RNA 1(1), 79–88 (1995). Coliphages that infect coliform bacteria that contain the
13. B. F. Schmidt et al., J. Mol. Biol. 195(3), 505–516 (1987).
F-pilus or sex pilus are termed male-specific phages or F-
specific phages. According to the International Standards
14. P. P. Hung et al., Science 166(913), 1638–1640 (1969).
Organization (ISO), the F-specific RNA bacteriophages
15. C. M. Ling et al., Biochemistry 8(11), 4464–4649 (1969).
are defined as, ‘‘bacterial viruses which are capable of
16. A. Campbell, in F. C. Neidhardt et al., eds., Escherichia Coli infecting a specific host strain with F-or sex-pili to produce
and Salmonella: Cellular and Molecular Biology, 2nd ed.,
visible plaques (clearance zones) in a confluent culture
ASM Press, Washington, D.C., 1996.
lawn grown under appropriate culture conditions. The
17. M. Slominska et al., Virology 262(2), 431–441 (1999).
infectious process is inhibited, however, by concentration
18. D. S. Burz and G. K. Ackers, Biochemistry 33, 8406–8416 (40–400 µg/mL) of RNase in the plating medium’’ (1).
(1994).
The ISO defines somatic coliphages as ‘‘bacterial viruses
19. M. O. Jones et al., Proc. Natl. Acad. Sci. U.S.A. 76, 150–154 which are capable of infecting selected Escherichia coli
(1979).
host strains (and related strains) by attachment to the
20. U. Schmeissner et al., Proc. Natl. Acad. Sci. U.S.A. 77, bacterial cell wall as the first step of the infectious process.
3191–3195 (1980).
They produce visible plaques in a confluent lawn of host
21. G. N. Gussin et al., J. Bacteriol. 174, 5156–5160 (1992). bacteria grown under appropriate culture conditions’’ (2).
22. A. Kihara et al., Proc. Natl. Acad. Sci. U.S.A. 94, 5544–5549 Bacteriophages that infect Bacteroides fragilis are defined
(1997). by the ISO as ‘‘bacterial viruses which are capable of
23. L. Tsui and K. Mark, Mol. Gen. Genet. 143, 269–278 (1976). infecting selected Bacteroides fragilis host strains by
24. A. M. Segall and H. A. Nash, EMBO J. 12, 4567–4576 (1993). attachment to the bacterial cell wall as the first step
25. M. Ptashne, A Genetic Switch: Phage 1 and Higher Organ- of the infectious process’’ (3). Various studies have shown
isms, 2nd ed., Cell Press, Blackwell Scientific Publications, the usefulness of bacteriophages as indicators of fecal
Cambridge, Mass., 1992. pollution caused by viruses. Coliphages may be good
26. M. Gellert et al., Proc. Natl. Acad. Sci. U.S.A 73, 3872–3876 indicators of enteric viral pollution because they are found
(1976). in relatively high numbers in human and nonhuman feces,
27. C. A. Robertson and H. A. Nash, J. Biol. Chem. 263, are persistent in the environment, do not multiply in the
3554–3557 (1988). environment, and the cost of coliphages assay is only a
28. E. C. Friedberg et al., DNA Repair and Mutagenesis, Ameri- fraction of an enterovirus tissue culture assay. Recently,
can Society for Microbiology, Washington, D.C., 1995. coliphages also have been suggested as conservative
29. S. Kim and A. Landy, Science 256, 198–203 (1992). indicators of enteric viruses because of their similarities in
30. H. A. Nash and C. A. Robertson, EMBO J. 8, 3523–3533 size, transport, survival, persistence, and high densities in
(1989). sewage and septic samples. Bacteriophages that infect the
31. H. A. Nash, in F. C. Neidhardt et al., eds., Escherichia Coli Bacteroides fragilis also have been suggested as potential
and Salmonella: Cellular and Molecular Biology, 2nd ed., indicators of fecal pollution (4). Bacteroides fragilis is
ASM Press, Washington, D.C., 1996. an obligate anaerobe found in high concentrations in
32. R. Young, Microbiol. Rev. 56, 430–481 (1992). human feces, and thus the presence or absence of these
33. M. T. Bonovitch and R. Young, J. Bacteriol. 173, 2897–2905 phages may be indicative of fecal pollution in a sample.
(1991). One salient feature of using Bacteriodes phages is that
34. J. W. Little and D. W. Mount, Cell 29, 11–22 (1982). because these bacteria are reportedly present only in
35. G. C. Walker, Microbiol. Rev. 48, 60–93 (1984). human feces, the presence of these phages in a sample
36. E. M. Witkin, Biochimie 73, 133–141 (1991) may be a strong indication of human fecal pollution. The
use of bacteriophages such as coliphages as indicators
of fecal pollution is based on the assumption that their
presence in water samples denotes the presence of bacteria
BACTERIOPHAGE DETECTION capable of supporting the replication of the phage (5).
METHODOLOGIES Bacteriophage as an indicator has an advantage. The
detection method is simple and inexpensive and the
results usually are available within 8 to 10 hours. This
SURESH D. PILLAI chapter provides an overview of some of the methods that
Texas A&M University have been used to detect and enumerate bacteriophages,
College Station, Texas especially the coliphages and the Bacteriodes phages in
 BACTERIOPHAGE DETECTION METHODOLOGIES 375

environmental samples. The methods that have been 4. The sample concentrate is neutralized as soon as
described pertain to only those assays in which the primary possible with 0.1 M HCl.
purpose is to determine the microbiological quality of an 5. This sample concentrate then is used for the bacte-
environmental sample. The reader is advised to consult riophage analysis using appropriate host bacteria.
other entries in this encyclopedia for a detailed discussion
on the classification of bacteriophages, their occurrence, In the original method, the host bacterium Salmonella
their persistence, and the rationale for using phages as typhimurium (strain WG 49) was employed for the
indicators of fecal pollution. detection of F-specific coliphages. This particular host
strain is resistant to nalidixic acid and kanamycin
RECOVERY OF BACTERIOPHAGES FROM GROUNDWATER and contains an E. coli plasmid (F 42lac : Tn5) that is
AND DRINKING WATER responsible for pili production and thus susceptible to
infection by F-specific coliphages (6). The host strain
One of the primary issues confronting the detection of is grown in a medium containing 1% tryptone, 0.1%
bacteriophages, especially the male-specific coliphages in glucose, 0.8% NaCl, 0.03% CaCl2 ·H2 0, 0.015% magnesium
environmental samples is the potentially large volume sulfate, 100 mg/L nalidixic acid, and 20 mg/L kanamycin
of samples that needs to be analyzed. The rationale for sulfate. The concentrations for the top and bottom agar
employing large sample volumes is based on the premise were 0.7% and 0.5%, respectively. Although the original
that enteric viruses are normally present in very low protocol suggested the use of S. typhimurium WG49 as
numbers. There are various procedures for the recovery, the host bacterium, the authors indicated that there may
detection, and enumeration of bacteriophages (2,6–8)). be interferences by somatic Salmonellaphages. There is
These protocols vary in terms of the recommended concern that the plasmid is not stable within the host
volumes to be analyzed, sample processing methods, bacterium and thus false negatives could arise if the WG49
and the recommended host bacteria. The terminologies strain is used. This has led to the use of alternate host
used in these procedures are based on the distinguishing bacteria such as E. coli 15597 as a tool to screen for male-
characteristics of the procedure, of the original publication, specific phages. Theoretically, it is possible to use any
or of the agency that published it. For example, the other suitable host bacteria and exploit the advantageous
‘‘AWWARF method’’ is based on a membrane filtration features of this protocol. In laboratory studies, the
(MF) approach developed by Sobsey and coworkers (9), observed decrease in phage recovery with increase in
whereas the ‘‘Environmental Protection Agency (EPA) sample volumes was attributed to the accumulation of
methods’’ rely on the concentration of large-volume ‘‘deposits’’ on the filter surface. Sobsey and coworkers (9)
(500 gal) samples and smaller (100 and 1,000 mL) recommended that sample volumes should be small
samples. The ‘‘ISO method’’ differs from the other methods enough to prevent deposit accumulations (or the use of 90-
primarily in the use of B. fragilis as the host bacteria. mm diameter filters). Sobsey and coworkers (9) suggested
that the filters could be placed ‘‘face down’’ onto agar
MF Technique media containing lawns of the specific host bacteria. To
aid in the detection of the enumeration of the plaques,
The MF method of Sobsey and coworkers (9) was
the use of tetrazolium dye (0.03% final concentration) was
based on a field-tested protocol that compared the
suggested.
recovery of male-specific coliphages in several source
The MF method has been used to screen groundwater
waters with concentrations of fecal-indicator bacteria.
serving public water supply wells along the Texas-Mexico
Erb and coworkers (10) reported on the detection of low
border region. In this study, 17 wells were sampled over a
concentrations of male-specific coliphages in swimming
span of three months, however, none of the samples were
pool waters using the MF method. However, they had to
positive for male-specific coliphages. The protocol was as
adjust the sample pH to 3.5 before filtration to achieve
follows:
efficient recoveries. They have reported that humic acids
in the water samples could interfere with adsorption of
1. Groundwater field samples (1,000 mL) were col-
phages to the filter. The MF-based method is unique in
lected in sterile containers filled under optimum
that it relies on the use of widely available inexpensive
conditions. The samples were adjusted initially to
membrane filters and filtration apparatus along with
0.05 M MgCl2 using sterile MgCl2 and thoroughly
manageable sample volumes (1,000 mL).
mixed.
Procedure 2. The samples were then filtered through a sterile
0.45-µm pore size filter at the rate of 100 mL per
1. A water sample (1,000 mL) initially is supplemented minute.
with MgCl2 to obtain a final concentration of 0.05 M 3. A sterile polypropylene tube was placed beneath
MgCl2 . the membrane and 5 mL of a beef extract elution
2. The sample is filtered through a 47-mm (0.45 µm solution (3% beef extract V; 0.3% Tween 80; pH
pore size), cellulose acetate membrane at a flow rate 9.5) was added to the membrane. After five-minute
not exceeding 100 mL per minute. incubation, this solution was filtered through the
3. The F-specific coliphages adsorbed to the membranes membrane (by vacuum application). To quickly bring
are eluted using 3 mL of a high pH buffer (3% beef the pH of the buffer to neutrality, 770 µl of 0.1 M HCl
extract V; 0.3% Tween 80; pH 9.5). was added to the tube containing the eluate.
376 BACTERIOPHAGE DETECTION METHODOLOGIES

4. This concentrated solution was then used for the of the eluate is mixed with 4 mL of glycerol and
bacteriophage analysis. Aliquots (0.5 mL) of the the sample is stored at −80 ° C until the sample is
solution were used in conventional double agar flocculated and further concentrated (details of these
overlay techniques for phage enumeration using host steps are presented elsewhere in the encyclopedia).
bacteria specific for F-specific phages (E. coli ATCC 3. The bacteriophage assay is then performed on the
15597) and somatic phages (E. coli ATCC 13706). concentrated sample using the appropriate host
Parallel studies, using spiked groundwater samples, bacteria and the conventional double agar overlay
showed very high phage recoveries. method.
5. In this protocol, a filter-elution step (to elute the 4. There is concern that the exposure of phage to
viruses off the filter) was employed, rather than high pH during the sample processing could lead
‘‘laying the filter facedown’’ on an agar plate. to significant loss of phage recovery. This has forced
This avoided problems associated with the plaque researchers to explore alternate sample processing
visibility and the need to use tetrazolium dyes for approaches.
plaque contrast purposes.

The U.S. EPA ICR Protocol ISO Method for Detection and Enumeration of
Bacteriophages
The U.S. EPA published a standardized procedure for
the enumeration of somatic and male-specific coliphages The method for the detection of F-specific bacteriophages
in the implementation of the Information Collection Rule already is an official standard for the ISO, whereas the
(ICR) (7). In this procedure, water samples are passed method for the detection and enumeration of somatic
across and in between a positively charged 1 MDS filter phages and B. fragilis phages currently is in the final
(CUNO Inc., Meridien, CT). However, for source water, stages of approval (Juan Joffre, Personal Communication).
the total amounts of sample that can be filtered through The detection and enumeration method for F-specific RNA
the 1 MDS filter will depend on the water quality in terms bacteriophages can be applied theoretically to all kinds
of suspended solids. The recommended sample volumes of water, sediments, and sludges where necessary after
are 200 L for source water and 1,500 L for finished dilution. The ISO method recommends a preconcentration
water. However, samples up to 1,500 L were used for if necessary for samples that may harbor low numbers of
groundwater-monitoring purposes (11). bacteriophage.

ICR Method ISO Method for the Enumeration of F-Specific RNA

Sampling Bacteriophages. The sampling, transport, and storage of
water samples are to be handled in accordance with
The routine precautions that one employs normally when the ISO, ISO 8199 (ISO 5667-1, ISO 5667-2 and ISO
sampling for enteric viruses are followed such as flow rate 5667-3 (1,11–13). The method recommends the use of
(not exceeding 3 gallons per minute) (7). Also, if the water S. typhimurium WG 49, E. coli K-12 Hfr, or E. coli HS
to be sampled contains a disinfectant such as chlorine, a (pFamp) R (14). The host cultures are grown in tryptone
2% sodium thiosulfate as a neutralizing agent is injected yeast extract glucose broth (TYGB) at 37 ° C. The quality
continually into the sampling stream at a rate of 10 mL/gal control check of the S. typhimurium strain (WG49) is
(2.6 mL/L or 30 mL per minute at a flow rate of 3 gal per performed by screening for the presence of lactose-negative
minute). The filter cartridge along with the filter is then
colonies (which would indicate plasmid segregation) and
aseptically packed in sterile foil and transported to the
inhibition zones around antibiotic disks impregnated with
laboratory on ice, and processed as soon as possible.
kanamycin and nalidixic acid. Less than 8% of the cells
should be lactose-negative with no inhibition zone around
Sample Processing
the nalidixic acid disk and an inhibition zone around
1. Filters are eluted initially using one liter of sterile kanamycin disks of less than 20-mm diameter. The ISO
beef extract glycine solution (1.5% beef extract method clearly details the protocol for samples containing
V, 0.05 M glycine in water, pH 9.4). The filter is high bacterial background, samples with low phage counts,
soaked in the elution solution for 15 minutes. The and a presence/absence test. The salient feature of the
solution then is forced from the housing into a protocol is that the method suggests confirmatory steps
sterile 2-L beaker using nitrogen gas or positive when sampling new sources, when there is an unexplained
air pressure. (It has been found that this particular overabundance of F-specific bacteriophages, or when there
step in the protocol is too cumbersome, prone to is indication that somatic phages (‘‘large, circular, clear
significant errors, and possibly could lead to cross- plaques with smooth edges’’) are being isolated. For
contamination). Currently in our laboratories, the confirmation purposes, the protocol recommends plating
elution solution is added to the filter holder and aliquots of the sample (in parallel) on a series of plates
rocked back and forth for 15 minutes using a rocking containing RNase at a final concentration of 40 µg/mL. The
platform. appropriate volume of a stock solution is added to soft agar
2. The eluate is then poured into a sterile 2-L beaker, just before adding the top layer. The protocol suggests that
the pH adjusted to 7.0–7.4 using 1 M HCl, and at times the RNase concentration may need to be as high
the solution stirred for 15 minutes. Forty milliliters as 100 µg/mL for selectivity.
 BACTERIOPHAGE DETECTION METHODOLOGIES 377

Protocol 6. Semisolid TYGA in bottles is melted and placed in

a 45 ° C water bath. Calcium-glucose solution then
1. Semisolid (soft agar) TYGA media (50 mL) in
is aseptically added (500 µL/50 mL)
bottles are melted and placed in a 45 ° C water
bath. (Because nalidixic acid is heat-stable, this 7. Aliquots (2.5 mL) of the TYGA media are placed
antibiotic could be added before autoclaving to into culture tubes with caps, placed in a 45 ° C
prevent background flora interferences). To these water bath.
bottles, 0.5 mL of calcium-glucose solution (CaCl2 · 8. To each tube, one milliliter of the host culture is
2H2 O 3 g; glucose 10 g; water 100 mL) and 2.5-mL added, mixed carefully to avoid bubble formation,
aliquots are distributed into sterile capped tubes. and the contents are poured atop a TYGA plate in a
9-cm petri dish. Care is taken to make sure that the
2. To each tube, one milliliter of the undiluted or
soft area is spread out evenly over the bottom agar
diluted sample is added, thoroughly mixed, and the
plate and is allowed to solidify on a level surface.
contents of the tube poured over a 9-cm bottom agar
TYGA plate. The plates are incubated for 20 hours 9. The plates are dried in a laminar flow cabinet or in
at 37 ° C. The protocol recommends that each sample a 37 ° C incubator for 30 minutes, while the plates
be tested at least in duplicate. are inverted and the lids are off.
10. A drop of the chloroform-treated culture is placed on
Samples with High Background Flora. For such samples, the inoculated plate using a fine capillary pipette.
nalidixic acid is added to semisolid soft TYGB to a final 11. The spot is allowed to dry and the plates are
concentration of 100 µg/mL. The conventional double agar incubated inverted (with the lids) at 37 ° C for
layer method then is adopted. 18 hours.
12. The plates then are examined for clearance zones,
Samples with Low Phage Counts. For samples in which which are indicative of F-specific phages.
the phage numbers are suspected to be low, the
The ISO (standard procedure) does mention that the
methodology calls for modifications as mentioned in the
protocol can be amended for use with larger samples (by
following text. Ten milliliters of semi solid TYGA is
using double-strength TYGB) and in a most probable
combined with one milliliter of host culture and 5 mL
number (MPN) format. Controls (both positive and
of sample dilutions (in duplicate) are poured over 50 mL
negative controls, especially spiked positive controls) are
of TYGA in a 14-cm petri dish. It is mentioned that this
strongly recommended. The results are expressed as
procedure is capable of detecting up to one phage particle
plaque-forming particles per milliliter
per 50 or 100 mL if 10–20 plates are inoculated in parallel.
N − NRNase
Confirmatory Test. Plates containing RNase (40 µg/mL) (pfp/mL) = ×F
n
are prepared by mixing the appropriate volume of RNase
solution to 2.5-mL semisolid TYGA in a tube and then where pfp/mL is the confirmed concentration of F-specific
pouring plates. Confirmatory tests are recommended RNA bacteriophages in milliliters, N is the total number
when examining new sampling points, and when large of plaques counted on W49 plates, NRNase is the total
circular clear plaques with smooth edges (probably somatic number of plaques counted on WG49 plates with Rnase, n
Salmonella phages) are routinely isolated. is the number of replicates, and F is the dilution factor.

Presence/Absence Test ISO METHOD FOR THE ENUMERATION OF SOMATIC

COLIPHAGES
Method
1. The host culture is prepared as described previ- The ISO method recommends the use of the E. coli
ously. Twenty-five milliliters (25 mL) of TYGB are strain c (ATCC 13706) for samples with low bacterial
added to a conical flask and prewarmed to room count such as drinking water and unpolluted natural
temperature. waters. For samples containing large numbers of bacteria
(e.g., wastewater and polluted source waters), the method
2. The host culture (0.25 mL) is added along with
recommends the use of the nalidixic acid–resistant E. coli
the calcium-glucose solution (250 µL) to the TYGB-
strain CN (ATCC 70078) also known as WG5. (15).
containing flask and incubated at 37 ° C for three
hours with shaking (100 rpm). Preparation of Host Culture. The host culture is grown
3. The water sample (1 mL) (prewarmed to room in modified Scholtens’ medium. The medium contains (per
temperature) is added to the flask and incubated 1,000 mL) peptone 2 g, yeast extract 3 g, meat extract 12 g,
for 18 hours at 37 ° C. NaCl 3 g, Na2 CO3 solution (150 g/L) 5 mL, and MgCl2
4. One milliliter (1 mL) of the culture is transferred solution (100 g MgCl2 ·6H2 O in 50 mL water) 0.3 mL. The
to a centrifuge tube, and 0.4 mL of chloroform pH of the medium after sterilization should be around 7.2.
is added, thoroughly mixed, and centrifuged at
3,000 X g for five minutes. Method
5. The host culture is prepared as described previ- 1. Molten semisolid modified Schölters’ agar (MSA)
ously. medium is placed in a 45 ° C water bath and 300 µL
378 BACTERIOPHAGE DETECTION METHODOLOGIES

of sterile calcium chloride solution (14.6 g/100 mL) The procedure can be used in an MPN approach to
is added to it. quantitatively estimate phage numbers. However, double-
2. Aliquots (2.5 mL) are placed into culture tubes with strength medium needs to be employed.
caps and placed in a 45 ° C water bath. The number of pfp is computed as follows:
3. To each of the culture tubes, one milliliter of the
N
water sample or its dilutions (prewarmed to room pfp/mL =
temperature) is added. nvF
4. The host culture (1 mL) is added to each of the where N is the number of plaques counted, n is the number
culture tubes containing the media and the sample, of replicates counted, V is the volume employed, and F is
and mixed well, taking care to prevent bubble the dilution factor
formation. The contents of the tube are then poured
over a bottom agar plate of MSA. ISO METHOD FOR THE ENUMERATION OF
5. The plates are allowed to incubate for approximately BACTERIOPHAGES INFECTING B. FRAGILIS
18 hours at 37 ° C.
The rationale for enumerating phages infecting B. fragilis
Samples with High Background Flora. For such samples, is that the bacterium (B. fragilis) is an obligate anaerobe
nalidixic acid at a final concentration of 250 µg/mL is added found only within the gastrointestinal tracts of warm-
to the top agar (semisolid MSA). blooded animals. Thus, detecting such phages in source
water or finished water is indicative that the sample has
Samples with Low Phage Counts. For such samples, the fecal contamination.
protocol suggests the use of 10 mL of semisolid media, The primary difference between this bacterial host
60 µL of calcium chloride solution, one milliliter of host and the other hosts is obligatory anaerobic incubation
culture, and 5 mL of sample in duplicate. The entire requirements of the host bacteria. Although the bacterium
contents are poured over a 15-cm petri dish containing is an obligate anaerobe, it does not require anaerobic
50 mL of MSA. handling conditions. Only the incubation has to be carried
out under anaerobic conditions. Anaerobic cabinets, jars,
Presence/Absence Test or bags can be employed for this purpose. However, when
Method using liquid media, it is critical that the containers are
completely filled (to avoid oxygen-rich headspaces) and
1. Modified Schölters broth (25 mL) is added to closed with a screw cap.
a sterile conical flask and prewarmed to room The specific host bacterial strain for the purposes of
temperature. this methodology is B. fragilis RYC2056 (ATCC 700786)
2. Calcium chloride solution (1 M) (150 µL) is added (16). The bacterium can be grown in (Bacteroides phage
aseptically to the flask. recovery medium broth) BPRMB. This medium contains
3. Then, 0.25 mL of the host culture is added and (per liter) meat peptone 10 g, casein peptone 10 g, yeast
followed by one milliliter of the sample (prewarmed extract 2 g, NaCl 5 g, monohydrated L-cysteine 0.5 g,
to room temperature) and incubated for 18 hours. glucose 1.8 g, MgSO4 ·7H2 0 0.12 g, and CaCl2 solution
(0.5g/mL) one milliliter.
4. An aliquot (1 mL) of the above culture is transferred
to a centrifuge tube and 0.4 mL of chloroform is
Preparation of Host Bacterial Culture. A vial of the stock
added, mixed well, and centrifuged at 3,000 x g for
culture is allowed to equilibrate to room temperature.
five minutes.
Using a cotton swab, the surface of the BPRM agar plate
5. Bottles containing molten 50 mL of semisolid MSA is streaked and incubated at 37 ° C for at least 40 hours
are placed in a 45 ° C water bath. under anaerobic conditions. Using a cotton swab, portions
6. Prewarmed calcium chloride solution (300 µL) is of the colonies are transferred from the plate to a 10-mL
added aseptically to the agar in the water bath and tube containing 10 mL of BPRMB. This is incubated at
2.5-mL aliquots are transferred to culture tubes 37 ° C for at least 24 hours. The optimum cell density
with caps. The culture tubes are placed in a 45 ° C for plaque formation and visualization is between 1 to
water bath. 4 × 108 CFU/mL. Appropriately designed side-arm flasks
7. To each of these tubes, one milliliter of the host are used to obtain optical density readings.
culture is added, mixed well, and the contents
poured atop bottom agar plates containing MSA. Method
8. The plates are incubated for 30 minutes in a 1. The host culture is prepared as described and
laminar flow hood or incubator at 37 ° C and air- allowed to equilibrate to room temperature.
dried with the lids off. 2. Bottles containing molten 50 mL of semisolid
9. One drop of the chloroform-treated sample culture BPRMA (top agar) is maintained at 45 ° C in a
is ‘‘spotted’’ on the above plates. water bath.
10. The plates are incubated at 37 ° C for 18 hours. 3. Ten milliliters (10 mL) of hemin solution (hemin
11. The presence of phages is indicated by the 0.1 g, 1 N NaOH 0.5 mL, 99.5 mL H2 O), one
appearance of clear zones near the spotted area. milliliter of (0.05 g·mL) CaCl2 solution, 25 mL of
 BACTERIOPHAGE DETECTION METHODOLOGIES 379

1 M Na2 CO3 , one milliliter of 100 µg/mL kanamycin 10. The contents of each tube are poured on to a 9-cm
sulfate, and 4 m of nalidixic acid (100 µg/mL) are petri dish and allowed to solidify on an even surface.
added to the medium already mentioned. 11. One drop of chloroform-treated culture is placed on
4. Aliquots (2.5 mL) are removed and placed into each plate and the plates incubated at 37 ° C for at
culture tubes with caps and placed in a 45 ° C least 18 hours under anaerobic conditions.
water bath. 12. The plates then are examined for characteristic
5. To each of these tubes, 1 milliliter of water sample clear zones or plaques.
is added and mixed well, avoiding the formation of
air bubbles. QUALITATIVE DETECTION OF COLIPHAGES IN SOURCE
6. The entire contents are poured on a bottom agar WATER — U.S. EPA METHOD 1601
plate containing BPRMA in a 9-cm petri dish.
7. The plates are allowed to solidify, and incubated Method 1601, a two-step enrichment procedure, is a
upside down under anaerobic conditions at 37 ° C for performance-based method for qualitatively detecting
18 hours. the presence or absence of male-specific and somatic
coliphages in groundwater and source water (17). This
8. After incubation, the number of plaques is counted
protocol has been tested in a multilaboratory round
to estimate the phage concentration per unit volume
robin testing protocol in the United States. Method 1601
of the sample.
specifies the host strains for isolating male-specific and
somatic coliphages.
Samples with High Background Flora. For such samples,
The attractive features of this protocol are that it can
it is recommended that kanamycin monosulfate at a final
be used either with 100-mL or 1,000-mL sample volumes.
concentration of 300 µg/mL be added to the soft agar.
The ability to use a large sample volume is advantageous
because it increases the probability of detecting phages
Presence/Absence Test. This test can be used to that may be present in the low concentrations. Somatic
qualitatively determine the presence or absence of coliphages are screened using E. coli CN-13 as the host
B. fragilis–specific phages in different samples volumes. strain, whereas male-specific coliphages are screened
The protocol given in the following text is for 100-mL using E. coli Famp as the host strain. E. coli CN-13 is a
sample volume. nalidixic acid–resistant mutant of E. coli (ATCC 700609),
whereas E. coli Famp is an ampicillin-streptomycin-
Method resistant mutant of E. coli. E. coli CN-13 was originated
1. Double-strength BPRMB (100 mL) is added to a at the University of Quebec in Canada, whereas E. coli
250-mL screw cap glass bottle and prewarmed to Famp was originated by Victor Cabelli at the University
room temperature. of Rhode Island.
2. The sample (100 mL) (also prewarmed to room
Method
temperature) is added to the bottle containing the
medium. 1. Logarithmic phase host bacterial cultures are pre-
3. The host culture in the exponential growth phase pared in tryptic soy broth (TSB) amended with the
containing approximately 108 cells per milliliter is appropriate antibiotics. E. coli CN-13 is grown in
added to the bottle. nalidixic acid–containing medium (10 mg/100 mL),
4. Additionally, medium is added to fill the bottle com- whereas E. coli Famp is grown in streptomycin-
pletely and the bottle is screwed tight. (To indicate ampicillin-containing medium (1.5 mg each antibi-
anaerobiosis, resazurin can be added. Resazurin at otic/100 mL).
a stock concentration of 0.025 g/100 mL is recom- 2. The cultures are incubated at 37 ° C on a shaker
mended. Anaerobiosis is indicated by a change of at 100–150 rpm for approximately four hours until
color from blue to straw color). the cultures are in log phase. The host cultures
5. The bottles are incubated at 37 ° C for 18 hours. when ready can be placed at 4 ° C for four hours
until they are ready to be used in the protocol.
6. To 1 milliliter of this enrichment culture, 0.4 mL
3. Spot Plate Preparation. Separate spot plates
of chloroform is added, mixed, and centrifuged at
are prepared for the somatic and male-specific
3,000 X g for five minutes.
coliphages. Two flasks of tryptic soy agar (TSA)
7. Semisolid BPRMA (50 mL) is melted using a boiling (100 mL) are autoclaved and cooled to 45 ° C in a
water bath and incubated at 45 ° C. Hemin, Na2 CO3 , water bath. To one of the flasks, 2 mL of the host
and antibiotics are added as per the concentrations culture and one milliliter of 10 g/mL nalidixic acid
mentioned earlier. solution are added. To the other flask, 2 mL of
8. Aliquots (2.5 mL) are placed in culture tubes with the host culture and one milliliter of 0.15-mg/mL
caps and the tubes are incubated at 45 ° C. ampicillin-streptomycin stock solution are added.
9. Host culture (1 milliliter containing approximately Twenty-milliliter aliquots are poured into 100-mm
108 cells per milliliter) is added and mixed well petri plates. The plates are appropriately labeled
with the media. Care is taken to make sure that no as being either CN-13 or Famp plates depending
air bubbles form. on the type of antibiotic that was added. The plates
380 BACTERIOPHAGE DETECTION METHODOLOGIES

are allowed to solidify and could be used that day basic principle of the method is as follows: host bacteria,
or stored at 4 ° C for up to a week. double-strength agar medium, and MgCl2 are added to
4. 100-mL Samples. Water samples (100 mL) are 100-mL water sample and the entire contents poured into
placed in sterile 125-mL bottles. To each of the 5 to 10 petri dishes. After overnight incubation, the total
bottles that contain the sample, 0.5 mL of the numbers of plaques are counted and tallied and the result
appropriate host bacterial culture, one milliliter reported as PFU(plaque-forming unit)/100 mL.
of the appropriate antibiotic solution, 1.25 mL of a
4 M MgCl2 ·6H2 O solution, and 5 mL of 10 X TSB Method
medium are added. The bottles are tightly capped, 1. The host bacteria and their respective antibiotic
mixed by inversion, and incubated for 24 hours at solutions are similar to the enrichment culture
37 ° C with no further mixing. described under Method 1601.
5. 1,000-mL Samples. Water samples (1,000 mL) are 2. The host bacterial cultures are prepared as
placed in sterile 2-L bottles. To each of the bottles described earlier. The cultures can be stored at
that contain the sample, 5 mL of the appropriate 4 ° C for no more than 48 hours.
host bacterial culture, 5 mL of the appropriate
3. Concentrated (2X) TSA (100 mL) is prepared and
antibiotic solution, 12.5 mL of a 4 M MgCl2 ·6H2 O,
allowed to cool to 45 ° C in a water bath. Sep-
and 50 mL of a 10 X TSB are added. The bottles are
arate flasks are prepared for the nalidixic acid
tightly capped, mixed by inversion, and incubated
amendment and for the ampicillin-streptomycin
for 24 hours at 37 ° C with no further mixing.
amendment.
6. Following incubation, 10 µL of the enriched sample
4. Two 100-mL samples are placed in separate 500-mL
is ‘‘spotted’’ on the appropriate spot plate that con-
flasks.
tains the appropriate host bacteria and antibiotic.
For example, 10 µL of the sample bottle that con- 5. To each of these flasks, 0.5 mL of 4 M MgCl2 ·6H2 O
tains the CN-13 host bacteria is spotted on the spot is added.
plate that has CN-13 and the nalidixic acid. 6. The flasks are placed into a 37 ° C water bath for
7. As many as 20 spots can be made on one 100-mm five minutes to prewarm the sample.
spot plate. The spots are allowed to absorb 7. To each of the flasks, 10 mL of the appropriate host
into the medium by incubating for approximately bacterium is added.
30 minutes. Care must be taken to make sure that 8. This sample/host bacterium mixture is then added
the spot inoculum is not allowed to ‘‘run’’ across the to the 100 mL of the 2 X TSA that was cooled to
plate. 4 ° C.
8. The plates are incubated at 37 ° C for 16 to 24 hours. 9. The contents of each of the flasks are poured into
9. Positive results are indicated when lysis zones (a 100-mm petri dishes at 20 mL per plate. Care must
circular zone of clearing) appear at the various be taken to make sure that the plates are labeled
spots. A positive result may also appear as one or properly to designate whether they contain CN-13
more small plaques or clearings at the spots. If or Famp host bacteria.
bacterial growth at the lysis zones interferes with 10. Positive and negative controls should be included
an accurate determination of whether the spot is with each assay.
‘‘positive’’ or ‘‘negative,’’ aliquots of the enrichment 11. The PFU/100 mL of either somatic or male-specific
sample can be centrifuged or filtered quickly and coliphages is calculated based on the total of the
the supernatant can be spotted on a fresh spot plaques obtained for all plates for that particular
plate. host bacterium.
10. Careful controls should be included in all exper-
iments because the potential for phage cross-
contamination is very high. The use of replicate OTHER METHODS
spots on separate plates also is recommended.
11. Because this is a qualitative procedure the results In addition to the protocols already mentioned, other
can be expressed only as either positive or negative methods have also been reported for the enumeration
for each 100- or 1,000-mL sample. of coliphages. Armon and Kott (19) have reported on a
simple, rapid, and sensitive presence/absence detection
QUANTITATIVE DETECTION OF COLIPHAGES IN SOURCE test for bacteriophages in 500 mL of drinking water.
WATER — U.S. EPA METHOD 1602 In this method, a probe filled with solidified soft agar
and bacterial host cells was immersed in a glass jar
Method 1602 is a quantitative method for the detection containing the water sample and bacterial host cells
of coliphages in water samples (18). However, unlike the (Fig. 1). The entire device was incubated in water bath
previously described Method 1601, this method allows at 36 ° C. Following this incubation (which ranged from
the quantitation of the coliphage numbers that are 90 minutes for E. coli CN13 to 8 hours for B. fragilis),
detected and also employs E. coli CN-13 and E. coli the probes were removed and incubated further. Using
Famp for somatic and male-specific coliphage detection, this method, the authors report their ability to detect
respectively. Method 1602 is only for 100-mL samples. The male-specific coliphages infecting E. coli Famp within six
 BACTERIOPHAGE DETECTION METHODOLOGIES 381

6. The culture is swirled and maintained at room

temperature without shaking for 10–15 minutes
and then returned to the shaker at 37 ° C for an
5 cm
22.5 cm
additional 105 minutes at 200 rpm.
7. The culture is placed in a sterile 50-mL centrifuge
tube and centrifuged at 5,000 rpm (Beckman model
TJ-6) at 6 ° C for 15 minutes.
Water level
8. The supernatant is vacuum filtered through a
0.2-µm pore size low protein-binding filter into a
sterile test tube.
Probe 20 cm Plaque
9. In a clean tube, 9 mL of Z-buffer (20) was combined
12.5 cm with one milliliter of the supernatant and 100 µL
of chlorophenol red β-D-galactopyranoside and
incubated at 37 ° C for 30 minutes.
4 cm
10. A positive coliphage test is observed by the
Water jar
immediate development of red color in an initially
Figure 1. Presence/absence–based bacteriophage detection yellow solution that intensifies over a 15-minute
apparatus. period and reaches a maximum within 30 minutes.
A negative test is indicated by an unchanged yellow
color.
hours and phages infecting B. fragilis within eight hours.
They successfully field-tested this approach on 45 different
types of aquatic samples. RECOVERY OF BACTERIOPHAGES FROM SOILS AND
BIOSOLIDS

COLORIMETRIC DETECTION OF BACTERIOPHAGES Mignotte and coworkers (22) tested eight virus extraction
protocols for recovering phages and mammalian viruses
Ijzerman and coworkers (20) reported on liquid colori- from urban sludge. The three extraction techniques
metric presence/absence coliphage detection method. This that resulted in the highest virus recoveries were as
methodology is based on the induction of β-galactosidase follows: beef extract (10%) solution at pH 9 combined
by E. coli. The release of this compound in the medium
with sonication, 0.3 M NaCl/7% beef extract solution at
because of cell lysis as a result of coliphage infections
pH 7.5, and freon and 0.1 M borate buffer/3% beef extract
permits the hydrolysis of a yellow chromogenic substrate
solution (pH 9) combined with sonication. It is obvious
that develops into a distinct red coliphage-positive sam-
that virus recoveries tend to be higher with organic
ple. A coliphage-negative sample will remain yellow. The
buffers at elevated pH levels. Lasobras and coworkers (23)
results can be obtained within 4.5 hours. The only appar-
also have reported on high recoveries of somatic, male-
ent disadvantage of this method is the requirement that
specific, and Bacteroides-specific phages from different
the water sample be concentrated to no more than 5 mL.
types of sludges encountered in water treatment plants.
Although the liquid colorimetric system was developed
using E. coli strain C (ATCC 13706), the method has the The protocol reported by Mignotte and coworkers (21) for
potential to be adapted to any of the other bacterial host optimum recovering phages from urban sludge is provided
strains. According to the authors, in combination with in the following text.
adequate sample concentration methods, this protocol can
detect 2 PFU/1,000 mL.
Method
Colorimetric Method Ten grams (10 g) of sludge is centrifuged initially at
1,500 g for 15 minutes.
1. E. coli (ATCC 13706) is inoculated into 25 mL of
The pellet is resuspended in 360 mL of 0.1 M borate
Luria broth supplemented with 5 mM of CaCl2 and
solution containing 3% beef extract (pH 9.0).
1.25 mM of MgSO4 ·7H2 O.
The mixture then is stirred at 500 rpm (New Brunswick
2. The culture is incubated at 37 ° C for one hour Scientific) for 15 minutes followed by sonication for one
at 200 rpm on a rotary shaker (Lab-Line Orbit minute at a 100 W, 0.9s setting.
Environ-Shaker, Melrose Park, Illinois). The sample is centrifuged at 10,000 x g for 45 minutes
3. The culture is then aseptically inoculated with at 4 ° C and the supernatant is neutralized to pH 7.2
25 µL of isopropyl-β-D-thiogalactoside (IPTG). Before bacteriophage analysis, the extract is decon-
4. The culture is incubated for an additional taminated by adding one-third volume of chloroform and
30 minutes at 37 ° C. vigorously mixing for 30 minutes at 4 ° C.
5. Concentrated water sample (1.25 mL) (obtained The sample is centrifuged at 1,500 x g for 10 minutes at
through any virus concentration protocols) is added 4 ° C. The chloroform phase is removed and the remaining
to the culture. phase is used for the bacteriophage assay.
382 BACTERIOPHAGE DETECTION METHODOLOGIES

RECOVERY OF COLIPHAGES FROM SHELLFISH Pillai and coworkers (29) employed the impinger, AGI-
30 samplers for 20 minutes at 12 L per minute using
Studies have shown that during depuration, FRNA phages 0.1% peptone. The 20-mL sample was concentrated down
are removed from the digestive tracts of a contaminated to 7 mL using Centriprep-50 concentrators (Millipore,
shellfish considerably more slowly than E. coli is removed. Bedford, Massachusetts). Aliquots of the concentrate
Because this relatively slow elimination has been shown were used in a conventional double agar layer format.
to be similar to the elimination of enteric viruses (24), It is extremely important to decide whether to employ
the presence of FRNA phages thus could be used as impaction or impingement approaches because the method
an indicator of enteric viruses. Dore and coworkers (25) of sample collection will dictate the extent of sample
recently reported on the isolation and use of male-specific processing and sample analysis.
RNA bacteriophages as candidate human enteric virus
indicators for bivalve molluskan shellfish. They described MOLECULAR DETECTION AND CHARACTERIZATION OF
a procedure for the detection of the FRNA phages using BACTERIOPHAGES
S. typhimurium WG49 as the host strain.
The complete nucleotide sequences of a number of bacte-
Method riophages are currently available. The availability of this
information has led to the development of oligonucletodie
1. Oysters are washed and scrubbed thoroughly under
probe and primer sequences for the detection and char-
running potable water.
acterization of bacteriophages. Molecular detection tech-
2. The oysters are opened aseptically with a flame- niques are not being used primarily for the rapid detection
sterilized shucking knife and the meat and of phages in environmental samples. These approaches are
intravalvular fluid are removed, diluted in 0.1% (w/v) being employed to find relationships between the presence
peptone water, and homogenized. or absence of mammalian enteric viruses and phages.
3. The diluted homogenates are centrifuged at Phylogenetically, F-specific RNA coliphages fall into
1,000 X g for five minutes at room temperature. four subgroups (30). Subgroups I and II are related and
4. The supernatant is decanted into a sterile bottle form the major group A. Subgroups III and IV are very
and a 1 : 10 dilution with peptone water (pH 7.2) similar and together make up major group B. Male-specific
is made. RNA coliphages are composed of serogroups I through
5. Ten milliliters of the undiluted supernatant and IV. It has been shown (30,31) that strains isolated from
4 mL of the 1 : 10 dilution are assayed for FRNA human feces usually are group II and III, whereas groups
phages by using the standard double agar overlay I and IV usually are found in animal feces. Although
method with 1 milliliter portions and 90-mm petri serotyping of F + RNA coliphages can distinguish human
dishes. One milliliter of the sample is mixed with fecal contamination from animal fecal contamination,
2.5 mL of the molten soft agar and 1 milliliter of the antisera for F + RNA coliphages are not readily available
host bacterium and poured over a prepared bottom and some isolates are difficult to stereotype (30,32). Given
agar plate. these reasons, the serotyping of phages is not being
discussed in detail in this chapter.
6. The presence of RNA phages was confirmed using
Genotyping of male-specific coliphages with oligonu-
plates amended with RNase in parallel.
cleotide probes is feasible (32,33). The oligoprobes were
7. The reported sensitivity was 30 PFU of FRNA/100 g end-labeled with digoxigenin and used in DNA-RNA
of shellfish. hybridizations and hybrids were observed by colorimetric,
immunoenzymatic detection.
RECOVERY OF BACTERIOPHAGES FROM AEROSOL The nucleotide sequence of the oligoprobes used for
SAMPLES genotyping is shown in Table 1.
The protocol for phage transfers and hybridization,
There have been a few studies reporting on the which is provided in the following text, is based on the
concentration of coliphages in bioaerosol samples. Most of report by Hsu and coworkers (32).
these studies have centered on municipal waste handling
operations such as biosolid land application sites and Method
effluent spray irrigation sites. Although the survival of
virus particles in bioaerosols is less than that of bacterial 1. Two microliter volumes of the phage isolates and
cells, detectable concentrations of both mammalian prototype strains that are representative of each
viruses and bacteriophages have been reported in the group are spotted on a lawn of host bacteria and
literature. There is no standard method for collecting incubated at 37 ° C overnight.
and concentrating viruses in bioaerosols. A number of 2. Plates are incubated at 4 ° C for 30 minutes.
different samplers have been employed (26–28). Carducci 3. The bacterial lawn is covered with a nylon mem-
and coworkers employed an impacting sampler (surface air brane for two minutes to adsorb phages from the
system SAS, PBI, Milan, Italy) to collect the air samples. plaques. Nylon membranes from Biodyne (Pall Bio-
A Rodac plate made of a phage agar base was used as support, East Hills, New York.), Boehringer Man-
the impacting surface. A top layer of phage agar (7 mL) heim Corp, Indianapolis, Indiana, or Gene Screen
combined with 1 milliliter of host bacterial culture was (NEN Research Products, Boston, Massachusetts)
poured over the Rodac plate to enumerate the phages. show efficient transfer capabilities.
 BACTERIOPHAGE DETECTION METHODOLOGIES 383

Table 1. Oligonucleotide Sequence Information for Group-Specific Probes (modified from

F. -C. Hsu et al., Appl. Environ. Microbiol. 61, 3960–3966 (1995))

Source
Probe Sequence Locus Phage

I CTA AGG TAT GGA CCA TCG Maturation protein MS2 (I)
AGA AAG GA
II CCA TGT TAT CCC CCA AGT Maturation protein GA(II)
TGC TGG CTA T
III ATA CTC AGT GAA (A/G)TA 5 nontranslated region Qβ
CTG CTG TGT
IV GGC ATA GAT TCT CCT CTG 5 nontranslated region SP (IV)
TAG TGC G
A AGC CCG ATC TAT TTT ATT Replicase MS2 and GA
GTT CTT CGG AAC
B TAA TTT TGC CAT GAT CGA Nontranslated region and coat Qβ and SP
ATT GAC CCA AAC protein

4. Blotting buffer (7.5 X SSC buffer with 4.6 M GENE AMPLIFICATION–BASED METHODS
formaldehyde) and heating at 65 ° C for 15 minutes
are used for the denaturation step. With the advent of gene amplification methods, the
5. The membranes are cross-linked using ultraviolet nucleotide sequence information that was available for
(UV) light treatment for five minutes and vacuum- the generation of oligonucleotide probes has been used
baked at 80 ° C for 15 minutes before hybridization. to develop PCR primers. Primers have been developed
6. Oligonucleotides are 5 end labeled using dioxy- against phages including MS2 and B. fragilis phages.
genin by terminal transferase using the Genius Puig and coworkers (34) recently published a report
labeling kit (Boeringer Manheim, Indianopo- describing the DNA amplification procedure for detecting
lis, Indiana). Unincorporated nucleotides were B. fragilisphages in water. In this work, primers were
removed using ethanol precipitation and probes designed from a 1.5-kb region that was specific to
are stored at −20 ° C until use. B. fragilis phages. The nested PCR assay was capable
7. The membranes are prehybridized for two hours at of detecting as few as 10–100 virus particles in
45 ° C in prehybridization solution [6 X SSC buffer seawater and sewage and 1–10 virus particles in river
(0.9 M sodium chloride; 0.09 M sodium citrate), 5 X water.
Denhardt’s solution (100 x Denhardt’s solution:for
500 mL: 10 g Ficoll 400; 10 g polyvinylpyrrolidone
MW 360000; 10 g BSA fraction V), 16 mM tris- CONCLUSION
HCl (pH 8.0), 0.1% Sodium dodecyl sulfate
solution (SDS), and 75 µg sheared salmon sperm A number of culture-based and molecular methods
DNA]. are available for the detection and characterization of
coliphages. Many of these methods have been used
8. Hybridization is performed in the same solution
successfully in multiple laboratories whereas some of
using 5 pmol of DIG-UTP-labeled probe overnight
them also have been rigorously evaluated as part of round
at 45 ° C.
robin multiple laboratory testing. Given the importance
9. The membranes are washed twice in 6 X SSC-0.01% of male-specific coliphages to serve as indicators of fecal
SDS at 40 ° C for 15 minutes. pollution, it is, however, necessary that these methods
10. The detection and color development is based on be tested, employed, and optimized using a variety of
the labeling kit manufacturer’s suggested protocol. environmental samples under different environmental
conditions. Molecular characterizations of male-specific
Using this approach, more than 80% of the phage RNA coliphages have demonstrated that it is possible
isolates could be genotyped directly from plaques, to identify the sources of the male-specific coliphages
providing evidence that genotyping is a much easier
once they are isolated. Thus, the ability to detect and
method of differentiating phages than sereotyping (which
characterize male-specific coliphages can assist in the
would have required extensive phage purifications). Hsu
detection of fecal contamination and in the identification
and coworkers (32) report that 17 isolates from piglet and
of possible sources.
porcine feces were classified as group-II phages (primarily
human), suggesting that group classification will not
always distinguish between human and porcine fecal BIBLIOGRAPHY
contamination. Because the dietary and living conditions
of pigs have historically involved exposure to human fecal 1. ISO, Water quality — Detection and enumeration of bacte-
wastes, this may account for the detection of group-II riophages, Part–I. Enumeration of F-specific RNA bacterio-
phages in pig feces (32). phages, ISO-10705, 1996.
384 BACTERIOPHAGE OF ENTERIC BACTERIA: OCCURRENCE AND PERSISTANCE IN THE ENVIRONMENT

2. ISO, Water quality — Detection and enumeration of bac- 34. M. Puig et al., Appl. Environ. Microbiol. 65, 1772–1776
teriophages, Part–2. Enumeration of somatic coliphages, (1999).
ISO-10705-2, 1999. 35. A. Carducci et al., Lett. Appl. Microbiol. 21, 207–209 (1995).
3. ISO, Water quality-Detection and enumeration of bacterio-
phages, Part–I. Enumeration of bacteriophages infecting
Bacteroides fragilis, ISO-10705-4, 1999.
4. C. Tartera and J. Joffre, Appl. Environ. Microbiol. 53, BACTERIOPHAGE OF ENTERIC BACTERIA:
1632–1637 (1987). OCCURRENCE AND PERSISTANCE IN THE
5. C. P. Gerba, in R. M. Maier, I. L. Pepper, and C. P. Gerba, ENVIRONMENT
eds., Environmental Microbiology, Academic Press, San
Diego, Calif., 1999, 491–504. MARYLYNN V. YATES
6. AWWARF. Male-specific coliphages as indicators of viral University of California
contamination of drinking water, 1995. Riverside, California
7. USEPA. ICR (Information Collection Rule) Microbial Labora-
tory Manual. Office of Research and Development, Washing- Bacteriophages, viruses that infect bacteria, are dis-
ton, D.C., 1996. tributed ubiquitously in nature. Their potential use as
8. EPA, 1990. indicators of fecal pollution and as models for the behavior
9. M. Sobsey, K. J. Schwab, and T. R. Handzel, J. Am. Water of enteric pathogens has been studied for many years. This
Works Assoc. 52–59 (1990). is due to the relative ease, cost, and speed of analysis for
10. P. Erb, A. M. Nasser, and B. Fattal, Water Sci. Technol. these organisms compared to that of pathogenic microor-
31211–31214 (1995). ganisms. These uses of bacteriophages will be the focus of
11. M. Abbaszadegan et al., Appl. Environ. Microbiol. this article.
65444–65449.
12. ISO 1999a. SURROGATES FOR HUMAN ENTEROVIRUSES
13. ISO 1999b.
14. J. Debartolomeis and V. Cabelli, Appl. Environ. Microbiol. Human enteric pathogens include viruses, bacteria,
57, 1301–1305 (1991). and parasites that infect the gastrointestinal tract of
15. W. O. K. Grabow and P. Cobrough, Appl. Environ. Microbiol. warm-blooded animals. Exposure to these pathogens is
52, 430–433 (1986). through the oral route, and they are eliminated from
16. A. Puig et al., Appl. Environ. Microbiol. 65, 1772–1776 the body in fecal material in the so-called ‘‘fecal-oral’’
(1999). route of transmission. The major source of pathogenic
17. USEPA. Method 1601. Male-specific (F+) and somatic microorganisms in domestic wastewater is the fecal
coliphage in water by two-step enrichment procedure. Draft- material of infected individuals; however, urine may also
EPA 821-R-00-009, 2000a. be a source of certain pathogenic viruses (1). There are
18. USEPA. Method 1602, Male-specific (F+) and somatic hundreds of different types of microorganisms that may be
coliphage in water by single agar layer (SAL) procedure. present in domestic wastewater. The numbers and types
Draft-EPA 821-R-00-009, 2000b. of pathogens found in wastewater will vary both spatially
19. R. Armon and Y. Kott, J. Appl. Bacteriol. 74, 490–496 and temporally, depending on the disease incidence in
(1993). the population producing the wastewater, season, water
20. M. A. Ijzerman et al., J. Virol. Methods, 45, 229–234 use, economic status of the population, and quality of the
(1993). potable water (2).
21. J. H. Miller, in J. H. Miller, ed., Experiments in Molecular Between 1971 and 1998, there were more than
Genetics, Cold Spring Harbor, New York, 1977, pp. 352–355. 650 outbreaks of waterborne disease reported in the
22. B. Mignotte et al., J. Virol. Methods 78, 71–80 (1999). United States (3). These outbreaks resulted in more than
23. C. J. Lasobras et al., J. Appl. Microbiol. 86, 723–729. 550,000 people becoming ill, and more than 125 deaths.
24. G. P. Richards, J. Food Prot. 51, 218–251 (1988). The majority (90%) of these outbreaks were caused by
25. W. J. Dore et al., Appl. Environ. Microbiol. 66, 1280–1285 microorganisms (see Fig. 1), most of them of enteric
(2000). origin.
26. Carducci, 1999.
27. K. P. Brenner et al., Appl. Environ. Microbiol. 54, 409–415 Indicators of Microbial Contamination
(1988). Because human pathogenic microorganisms have such
28. S. D. Pillai et al., Appl. Environ. Microbiol. 62, 296–299 an impact on public health, water microbiologists and
(1996). regulators have searched for suitable compounds or
29. Pillai, 1999. microorganisms that can indicate microbial and fecal
30. K. Furuse, in S. M. Goyal, C. P. Gerba, and G. Bitton, contamination of ground and surface water. Such a
eds., Phage Ecology, Wiley-Interscience, New York, 1987, compound or organism is called an indicator.
pp. 87–123. Male-specific coliphages, such as MS-2, are often
31. S. Osawa et al., Appl. Environ. Microbiol. 41, 164–168 (1981). naturally present in sources of human wastes (e.g., septic
32. F.-C. Hsu et al., Appl. Environ. Microbiol. 61, 3960–3966 tank effluent) and therefore their presence or absence
(1995). in groundwater may be indicative of water quality. The
33. J. Beekwilder et al., J. Appl. Bacteriol. 80, 179–186 (1996). suitability of a particular virus as an indicator is evaluated
 BACTERIOPHAGE OF ENTERIC BACTERIA: OCCURRENCE AND PERSISTANCE IN THE ENVIRONMENT 385

sense that it can replace the exact behavior of a human

enteric virus in the environment. However, because of the
Gastroenteritis Chemicals extremely complex sorption and transport mechanisms of
46% 10% viruses, recent studies have raised doubts regarding the
applicability of any single indicator’s ability to mimic the
behavior of a specific virus. Jin and coworkers (15) have
shown that MS-2 is relatively easily inactivated by the
Viruses air-water interface in unsaturated systems and that it
8% is sensitive to the presence of metal oxides (16). Penrod
and coworkers (17) compared the deposition kinetics of
bacteriophages MS-2 and λ and found that even subtle
differences in viral surface structures could significantly
influence the rate at which viruses were removed from
Parasites Bacteria the water phase by infiltration. Considering the complex
22% 14% sorption and retention mechanisms of viruses, it is
Figure 1. Causative agents on waterborne outbreaks (U.S.) unlikely that any single bacteriophage will adequately
1971–1998. mimic the transport behavior of viral pathogens in
the environment, particularly given that the surface
on the basis of relative insensitivity to inactivation (4) and structures of different waterborne human viruses are quite
dissimilar.
its ecological and morphological similarities to human
viral pathogens (5).
There are numerous review articles on indicators ECOLOGY OF BACTERIOPHAGE
available on this subject (6,7). Coliphages, particularly
RNA phage, have been proposed as suitable indicators Coliphages, the viruses that infect coliform bacteria,
for human enteric viruses (6). Among the coliphages, MS2 are the most widely studied bacteriophages in terms
appears to be the most suitable indicator (8,9). of their use as surrogates. Coliphages are found in
the feces of man and other warm-blooded animals, in
domestic wastewater, and in wastewater-contaminated
Models for Pathogen Transport
waters, thus suggesting the utility of these organisms
When considering indicators to use as models for microbial as indicators of the presence of fecal material that may
survival and transport in the environment, a worst-case contain pathogenic microorganisms and thereby pose a
indicator for transport and fate should: threat to human health.
The distribution of coliphages in several hundred
samples of feces from human and other warm-blooded
1. be unable to reproduce in the environment
animals is shown in Table 1. Reported concentrations
2. show similar or less sorption and retention than of coliphages in various sources are shown in Table 2.
pathogenic microorganisms in porous media under Clearly, the concentrations and presence of coliphages in
identical conditions feces is extremely variable. In addition, these viruses are
3. be at least as resistant to inactivation under natural found more commonly in the feces of lower animals than
conditions as pathogenic microorganisms in those of man.
4. be nonpathogenic to humans (only if used as tracers However, in contrast to human feces, coliphages
in the field) are consistently found in domestic raw wastewater.
Raw sewage has been found to contain between 105
Several markers have been used as indicators for and 106 plaque-forming units (pfu)/L (Table 2). Septic
domestic wastewater contamination (10,11). Examples tanks, which serve a smaller population than would be
of such markers are coprostanol or surfactant related
markers such as linear alkylbenzene. These markers
Table 1. Distribution of Coliphage in Human and Ani-
have not been tested yet for their sorption and transport mal Feces
properties in porous media and no comparative studies
with viruses are available at present. Source No. of Samples No. Positive % Positive
Based on the mechanisms of virus sorption and
Birds (zoo) 25 23 92.0%
retention and the experimental evidence of virus transport Fowl 30 9 30.0%
available, MS2 appears to be a reasonable candidate for Mammals (zoo) 97 72 74.2%
a transport model at this time. MS2 was found to be Horses 30 3 10.0%
a useful indicator for viruses because it showed similar Cows 20 6 30.0%
or less sorption and retention than other viruses under Pigs 11 10 90.9%
identical conditions (12–14). However, this conclusion is Humans 597 140 23.5%
based on the specific experimental conditions used in these
Source: Modified from (K. Furuse, Distribution of Coliphages in
studies, and there is no guarantee that MS2 behaves the Environment: General considerations, Chap. 3, in S. M. Goyal,
as a worst-case indicator under a variety of soil and C. P. Gerba, and G. Bitton, eds., Phage Ecology, John Wiley & Sons,
water conditions. MS2 is not an ideal indicator in the New York, 1987.)
386 BACTERIOPHAGE OF ENTERIC BACTERIA: OCCURRENCE AND PERSISTANCE IN THE ENVIRONMENT

Table 2. Concentrations of Coliphages in Feces liter (21). Little work has been done on the occurrence of
and Wastewater phage that infect indigenous freshwater bacteria. Phage
Source Typical Concentration % Positive that infect Aeromonas hydrophila have been recovered
from river water by several investigators (21). Concentra-
Animal feces 0–106 /g 12–75 tions of these phage were on the order of a few thousand
Human feces 0–107 /g 10–92 per liter of water. Pseudomonas aeruginosa phage have
Gray water 0–>1010 L 12 also been isolated from river water in the United Kingdom
Septic tanks 103 –106 /L 100 at a concentration of 3.5 × 103 per liter (23).
Raw sewage 105 –107 /L 100 In addition to surface water, the occurrence of phages
Primary 0.2–106 /L 100
in groundwater systems has been studied. There has been
effluent
a surge of interest in this area due to recent activity
Secondary 101 –104 /L 100
effluent at the U.S. Environmental Protection Agency (EPA). In
(activated 2000, the EPA proposed the Ground Water Rule, the
sludge) goal of which is to protect potable groundwater supplies
Oxidation pond 103 –107 /L 100 from fecal contamination (23a). The difficulty associated
effluent with monitoring more than 150,000 water systems for
enteric viruses spurred a great deal of interest in the
Source: Modified from (C. P. Gerba, Phage as Indicators of Fecal
identification of a suitable indicator for the presence
Pollution, Chap. 8, in S. M. Goyal, C. P. Gerba, and G. Bitton,
eds., Phage Ecology, John Wiley & Sons, New York, 1987.) of these microorganisms in groundwater. Due to the
similarity to enteric viruses in terms of their size, shape,
persistence in the environment, and transport behavior,
served by a municipal system, have also been found to bacteriophages are being considered as indicators for the
consistently contain coliphages. Concentrations in septic purposes of this regulation.
tank effluent have been reported to range between 103 One of the most extensive surveys of the occurrence
and 106 pfu/L (Table 2). A recent report of coliphage of enteric microorganisms in groundwater involved
analyses of septic tank effluent from a single septic the analysis of water samples from more than 400
tank revealed a mean concentration of 7 × 105 pfu/L (20). groundwater wells in the United States. This survey
The ubiquitous presence of the coliphages in untreated revealed the presence of coliphages in 20% of the
domestic wastewater is the rationale behind the use of samples (24). In this study, three different hosts, E. coli C,
these organisms as indicators of fecal contamination of E. coli C3000, and S. typhi murium WG49, were used
water. to test for the presence of coliphages in concentrated
Bacteriophages have been found in a variety of natural groundwater samples. Of the 444 samples analyzed, 4.1%,
waters including lakes, streams, rivers, wetlands, and 10.8%, and 9.5% were found to be positive using these
groundwater (19). In the majority of the studies that have three as the respective hosts. Only one sample was found
been conducted, strains of E. coli have been used as the to be positive for coliphages using all three hosts, while 92
host organism. Thus most of the available data on phage of the samples (20.7%) were found to be positive using at
occurrence in freshwater environments reflect coliphage least one host.
occurrence. Table 3 lists concentrations of coliphages Coliphages have also been isolated from groundwater
found at selected sites. in an area influenced by recharge of the subsurface with
Phage that infect other enteric bacteria have also been reclaimed water (22). Coliphages were detected on at least
isolated from freshwater. From rivers in the United King- one occasion in all twenty-six of the wells sampled. The
dom, bacteriophage that infect Klebsiella pneumoniae and shallowest of the wells had a perforated interval of 25.9
Salmonella typhi murium have been isolated at concentra- to 72.9 m; the deepest was perforated at 345.2 to 434.5 m.
tions ranging from approximately 1,000 to 1,000,000 per While some of the wells were located in close proximity

Table 3. Concentrations of Coliphages in Freshwater

Source Host pfu/L

pond, Florida E. coli B 1.9 × 104

lake, Florida E. coli B 1.5 × 101 –1.1 × 103
lake, Florida E. coli C 1.5 × 102 –1.3 × 104
river, U.K. E. coli W3110 3.5 × 103 –6.5 × 105
river, U.K. E. coli HfrH 1.8 × 104
river, South Africa E. coli K12Hfr 1.7 × 105
river, New York E. coli 5 × 101
river, Washington, DC E. coli C 1 × 104
river, Czechoslovakia E. coli B-39 2 × 104
river, California E. coli HS12 (pFamp)R 1.7 × 101 –1.5 × 102

Source: S. Farrah, Ecology of Phage in Freshwater Environments, Chap. 4, in

S. M. Goyal, C. P. Gerba, and G. Bitton, eds., Phage Ecology, John Wiley & Sons,
New York, 1987.
 BACTERIOPHAGE OF ENTERIC BACTERIA: OCCURRENCE AND PERSISTANCE IN THE ENVIRONMENT 387

(20 to 40 m) to the recharge basins, others were located

Inactivation rate (log10/d)

1.4
more than 100 m away; in one case the well was located
1.2
more than 1,700 m from the recharge basins. Estimated
1
travel times from the basins to these wells ranged from
0.8
24 to 5,852 days. Based on the data presented, it is clear
0.6
that the presence of coliphages in some of the wells is not
0.4
due to the presence of the recharge basins. The presence of
0.2
other potential sources of coliphage, such as seepage from
0
the domestic sewerage system, may be responsible for the
0 5 10 15 20 25 30 35
presence of these viruses in the groundwater.
Temperature (ºC)
Figure 2. Survival of MS2 coliphage in groundwater as a
PERSISTENCE OF BACTERIOPHAGES IN THE function of temperature.
ENVIRONMENT

There are a number of factors that affect the length of time due to denaturation of the viral capsid (25) as proteins
that bacteriophages persist in the environment. The most are denatured at high temperatures. However, it has
important factors and the effects they have on persistence been shown that the RNA released during thermal
are summarized in Table 4. Some of these factors will be denaturation remains infective as it is more resistant
discussed in more detail in the following text. For a more to heat inactivation than the protein capsid (26).
complete description of the effects of these factors on the
persistence of viruses in general, please see VIRUS SURVIVAL
Microbial Activity
IN SOILS.
There are conflicting reports regarding the role of microor-
Temperature ganisms in virus inactivation; many of these have been
reviewed (27,28). The survival of MS2 viruses in ground-
Temperature is probably the most important factor water that had been filtered to remove the indigenous
influencing virus inactivation in the environment (25). bacteria compared to their survival in nonfiltered water
In general, it has been found that the inactivation rate is shown in Figure 3 (data shown in Table 5). There is no
is significantly correlated with temperature, with faster clear trend — in some cases the phage survived longer in
inactivation rates occurring at the higher temperatures. the filtered water, in other cases survival was prolonged
Figure 2 presents a compilation of data (shown in in the presence of the indigenous bacteria.
Table 5) on the inactivation rate of MS2 coliphages as
a function of temperature. This is the general trend,
pH
although there are clearly situations in which viruses
are inactivated more rapidly at lower temperatures. The effect of pH on virus survival in soil has not
Those exceptions can likely be explained by differences in been extensively studied. It has been suggested that
other environmental characteristics, such as pH, microbial pH indirectly influences virus survival by controlling
activity, etc. adsorption onto soil particles, which, in turn, affects virus
The exact mechanism whereby temperature inactivates survival (29). The direct effects of pH on virus persistence
viruses in soils has not been determined, but several have been studied by a few investigators. Hurst and
theories have been proposed. The inactivation may be coworkers (30) studied the survival of poliovirus 1 and

Table 4. Factors Affecting the Persistence of Viruses in the Environment

Factor Effect on Persistence

Temperature Viruses persist longer at lower temperatures

Microbial activity Some viruses are inactivated more readily in the presence of certain
microorganisms, and sorption to the surface of bacteria can be protective
Moisture content Most viruses survive longer in moist soils and even longer under saturated
conditions; unsaturated soil may inactivate viruses at the air-water interface
pH Most enteric viruses are stable over a pH range of 3 to 9; however, survival may
be prolonged at near-neutral pH values
Salt species and concentration Certain cations may prolong survival depending on the type of virus
Virus association with soil Viral association with soil generally increases survival, although attachment to
certain mineral surfaces may cause inactivation
Soil properties Effects on survival are probably related to the degree of virus sorption
Virus type Different virus types vary in their susceptibility to inactivation by physical,
chemical, and biological factors
Organic matter Organic matter may prolong survival by competitively binding to air-water
interfaces where inactivation can occur; organic matter may also retard viral
infectivity
388 BACTERIOPHAGE OF ENTERIC BACTERIA: OCCURRENCE AND PERSISTANCE IN THE ENVIRONMENT

Filtered Non-filtered Table 5. Survival of Coliphage in Ground-

water
1.2
Inactivation Rate
Inactivation rate

1
Virus pH (log10/d) Temp. (° C)
(log10/d)

0.8
0.6 f2 7.8 0.390 9
0.4 f2 7.6 1.416 22
0.2 MS 2 7 0.180 4
MS 2 7.9 0.014 4
0
MS 2 8 0.020 4
0 5 10 15 20 25 30
MS 2 8.1 0.032–0.064 4
Temperature (ºC) MS 2 8.2 0.025 4
Figure 3. Survival of MS2 coliphage in groundwater in the MS 2 8.3 0.012 4
presence and absence of indigenous bacteria. MS 2 6 0.034 12
MS 2 7.3 0.037 12
MS 2 7.9 0.030 12
MS 2 8 0.093 12
two bacteriophages, MS2 and T2, in nine soils with pH MS 2 8.1 0.113–0.162 12
values ranging from 4.5 to 8.2. They found that virus MS 2 8.2 0.040 12
inactivation was significantly correlated (p < 0.05) with MS 2 8.3 0.028–0.095 12
soil saturation pH, with longer survival at the lower pH MS 2 7.7 0.114 13
values. MS 2 8 0.077 13
The exact mechanism whereby pH causes virus MS 2 8.1 0.075 17
MS 2 8 0.082 18
inactivation has not been fully elucidated, but results
MS 2 0.069–0.149 20
obtained by Salo and Cliver (31) suggested that virus MS 2 7.9 0.187 23
inactivation involves alterations in the virus capsid. These MS 2 8 0.244 23
investigators found that the RNA of the inactivated virus MS 2 8.1 0.278–1.196 23
particles became sensitive to ribonuclease at all pH levels MS 2 8.2 0.325–0.738 23
tested (pH 3 to 9), and at pH 5 and 7 the RNA was MS 2 8.3 0.244–0.262 23
hydrolyzed in the absence of ribonuclease. MS 2 0.103–0.16 23
Loveland and coworkers (32) studied the effects of MS 2 0.109–0.211 24.5
MS 2 7.3 0.129–0.207 25
pH on the reversibility of virus attachment to mineral
MS 2 7.5 0.186–0.34 25
surfaces. Their experiments on the attachment of PRD1 MS 2 7.6 0.313 25
to quartz and ferric oxyhydroxide-coated quartz indicated MS 2 7.7 0.249 25
that attachment is controlled by the isoelectric point of MS 2 7.8 0.39–0.54 25
the mineral surface. Attachment of PRD1 was observed MS 2 7.9 0.198–0.649 25
at pH values 2.5 to 3.5 units above the isoelectric point MS 2 6.8 0.293 26
of the mineral surface. Below this pH, the attachment MS 2 6.9 0.269 26
of PRD1 was found to be complete and irreversible, MS 2 7.2 0.211–0.254 26
MS 2 7.3 0.321–0.663 26
whereas above this pH, the adsorption was minimal and
MS 2 7.4 0.278 26
reversible.
MS 2 7.5 0.426–1.104 26
The mechanism(s) whereby pH affects virus adsorption MS 2 7.6 0.132–0.836 26
to soil particles has been explained in terms of the electro- MS 2 7.7 0.235–0.932 26
chemical nature of the virus and soil surfaces (29,33,34). MS 2 7.8 0.286–0.648 26
The outer surface of the enteric viruses is made of pro- MS 2 7.9 0.237–0.321 26
tein; therefore, the surface charge is influenced by the MS 2 0.113–0.325 26
ionization of the carboxyl and amino groups in the cap- MS 2 7.8 0.107–0.652 27
sid. The isoelectric point (pI) of many enteric viruses is MS 2 7.9 1.870 27
MS 2 8 0.144 27
below 7 (34); thus, at neutral pH, most viruses are nega-
MS 2 0.153–0.581 27
tively charged. Most soils are also negatively charged at MS 2 0.161–0.340 28.25
neutral pH, and virus adsorption is not favored because MS 2 7.8 0.132–0.637 30
of the mutual repulsion. If the pH of the environment MS 2 7.9 0.584–0.634 30
is decreased, the surface charge of the virus will become MS 2 0.214–0.706 30.5
positive (or less negative) due to increased ionization of
Source: M. V. Yates, Report to EPA, 1992 (unpub-
the amino groups and decreased ionization of the carboxyl lished).
groups. While soils also become less negatively charged
at lower pH values, soil pI values are generally lower
than those of viruses, thus they may still have a net complicating factors that can interfere with the mecha-
negative charge at acidic pH levels. This results in an nism discussed earlier. One is that a given virus may
electrostatic attraction between the virus particle and the have more than one pI: poliovirus 1 (Brunhilde strain)
soil, which leads to adsorption. In reality, the effect of pH has pIs at pH 4.5 and 7.0 (34). The factors responsible for
on virus adsorption is not so clear-cut. There are many passage from one form to another are not known at this
 BACTERIOPHAGE OF ENTERIC BACTERIA: OCCURRENCE AND PERSISTANCE IN THE ENVIRONMENT 389

time. Other soil factors such as cations and humic and Aggregation of Virus Particles
fulvic acids may also influence the net surface charge of
The formation of aggregates influences virus survival in
viruses.
natural waters. It has been suggested that this is because
virus particles within the aggregates are highly resistant
Salt Species and Concentration to destruction by environmental factors (25). It has been
The presence of certain chemicals may render a virus shown that aggregation renders virus particles more
more or less susceptible to inactivation. Burnet and resistant to inactivation by chemical disinfectants such
McKie (35) found that bacteriophage inactivation at 60 ° C as bromine (40), free chlorine, and monochloramine (41).
was partially prevented in the presence of 0.002 to Although there are no definitive reports on the effects of
aggregation on virus survival in soil, the results of studies
0.01 M CaCl2 , or BaCl2 . However, when the concentration
in aqueous media would suggest that viruses in aggregates
was increased to 0.15 M or greater, thermal inactivation
would survive longer in soils than would monodispersed
was increased.
viruses (33).
Thurman and Gerba (36) studied the effects of modify-
ing soil with several metals on the survival of MS2 and
poliovirus. The addition of aluminum metal, magnesium Soil Properties
oxide, and magnesium peroxide had a significant negative The influence of soil properties on virus survival is prob-
effect on the inactivation rate of MS2 when compared with ably related to the degree of adsorption. Hurst and
unmodified control soils. The addition of unrefined sub- coworkers (31) suggested that the correlation between
stances such as zeolite, bauxite, limonite, and glauconite pH and virus survival was probably mediated through
(which contain multivalent cations and oxides of iron and its influence on virus adsorption. A positive correlation
aluminum) did not have a significant negative effect on between virus survival and soil exchangeable aluminum
the inactivation rate of MS2. Rather, they seemed to have and a negative correlation with resin-extractable phospho-
a protective effect, as indicated by the lower inactivation rus were also attributed to influencing virus adsorption
rate as compared with the unamended control soil. Further onto the soil.
experiments by these investigators indicated that contact
between the aluminum and virus was necessary for inacti- Virus Type
vation of the viruses. They postulated that a combination
As is obvious from the previous discussions, different
of electrodynamic van der Waals interactions and electro-
viruses vary in their susceptibility to inactivation in
static double layer interactions promoted virus adsorption
the subsurface environment. In a comparative study of
to the surface of the aluminum, where inactivation of the
the survival of poliovirus, echovirus, and coliphage MS2
virus could then take place.
in several different groundwater samples, no significant
The role of soil cation exchange capacity (CEC) in
difference (p < 0.01) was found overall (6). There were,
virus adsorption has also been investigated. Burge and
however, differences in the inactivation rates of the viruses
Enkiri (37) found that the CEC of four of five soils was
in individual water samples.
correlated significantly (p < 0.05) with virus adsorption.
The survival of poliovirus 1 and f2 coliphage in dry
In contrast, Goyal and Gerba (38) did not find a significant
sand was compared by Lefler and Kott (42). Poliovirus
correlation between soil CEC and adsorption of 15 different
survived for at least 77 days at room temperature. The f2
viruses. Additionally, no correlation was found between
phage survived considerably longer, possibly for as long
virus adsorption and total phosphorus or between total
as one year under dry conditions. Based on the previous
and exchangeable iron, calcium, and magnesium.
discussions, it is possible that the survival times of both
viruses would be greatly increased if water were added to
Virus Adsorption to Soil the soil.
The adsorption of viruses to soils and other surfaces may
prolong or reduce survival, depending on the properties Organic Matter
of the sorbent. Murray and Laband (38a) found that The influence of organic matter on virus survival has not
poliovirus adsorption onto some inorganic substances, been firmly defined. In some studies, it has been found
such as CuO, results in decreased infectivity of the that proteinaceous materials present in wastewater may
virus. These investigators suggested that van der Waal’s have a protective effect on viruses; however, in others no
interactions between the virus and the particle surface effect has been observed (25).
induced spontaneous disassembly of the virus. The effects of natural humic material and wastewater
More commonly, however, adsorption to soils has sludge-derived organic matter on the transport of MS2
been found to prolong virus survival. The mechanisms bacteriophages in unsaturated soil was studied by
whereby adsorption to a solid surface prolongs or reduces Powelson and coworkers (43). The transport of MS2
virus survival have not been elucidated. However, Gerba was found to be higher in a loamy fine sand column
and Schaiberger (39) have suggested several possibilities, that had been treated with the organic material than
including interference with the action of virucides, in a parallel column that had not been treated. In
increased stability of the viral protein capsid, prevention a series of experiments, Bales and coworkers (13,44)
of aggregate formation, and adsorption of enzymes and studied the effects of hydrophobic organic material on
other inactivating substances. the attachment of bacteriophages MS2 and PRD1 and
390 BACTERIOPHAGE OF ENTERIC BACTERIA: OCCURRENCE AND PERSISTANCE IN THE ENVIRONMENT

poliovirus 1 to silica beads. These investigators found that BIBLIOGRAPHY

even very small amounts of hydrophobic organic material
(≥0.001%) can retard the transport of viruses in porous 1. C. J. Hurst, CRC Crit. Rev. Environ. Control 18, 317–343
media. (1989).
Several investigators have found that organic material 2. J. B. Rose and R. P. Carnahan, Pathogen Removal by Full
Scale Wastewater Treatment, Report to Department of
can act not only as a competitor for virus adsorption
Environmental Regulation, St. Petersburg, Fla., 1992.
sites, but also as an eluting agent, that is, it can cause
3. Centers for Disease Control and Prevention, MMWR 49,
sorbed viruses to desorb from the soil (34). Pieper and
SS4–(2000).
coworkers (45) studied the effects of wastewater-derived
4. M. V. Yates, C. P. Gerba and L. M. Kelly, Appl. Environ.
organic matter on the transport of bacteriophage PRD1 Microbiol. 49, 778–781 (1985).
in a sand and gravel aquifer on Cape Cod, MA. They
5. A. H. Havelaar, M. van Olphen and Y. C. Drost, Appl.
found that the removal of the virus was greater in Environ. Microbiol. 59, 2956–2962 (1993).
an uncontaminated area (83%) than in an area that 6. J. A. Snowdon and D. O. Cliver, Crit. Rev. Environ. Control
had been contaminated by secondary wastewater effluent 19, 231–249 (1989).
(42%). They concluded that the difference in removal 7. IAWPRC, Water Res. 25, 529–545 (1991).
was due to the higher content of organic matter in the 8. M. V. Yates, C. P. Gerba and L. M. Kelley, Appl. Environ.
contaminated zone (2.0 to 4.4 mg/L vs. 0.4 to 1.0 mg/L Microbiol. 49, 778 (1985).
dissolved organic carbon), which blocked the adsorption 9. V. S. Springthorpe, C. L. Loh, W. J. Robertson and S. A.
of the PRD1. The attachment of the viruses to the Sattar, Water Sci. Technol. 27, 413–420 (1993).
soil surfaces was reversible, as demonstrated by the 10. R. W. Walker, C. K. Wun and W. Litsky, Crit. Rev. Environ.
recovery of viruses upon addition of linear alkylbenzene Control 12, 91–112 (1982).
sulfonates (LAS) to the system. In the contaminated 11. C. M. G. Vivian, Rev. Sci. Total Environ. 53, 5–40 (1986).
zone, 87% of the attached PRD1 bacteriophages were 12. C. P. Gerba, S. M. Goyal, I. Cech and G. F. Bogdan, Environ.
remobilized. Jin and coworkers (14) also found that Sci. Technol. 15, 940–944 (1981).
addition of a solution of 1.5% beef extract, which is high 13. R. C. Bales et al., Water Resour. Res. 29, 957–963 (1993).
in organic matter, could remobilize X174 bacteriophages 14. Y. Jin, M. V. Yates, S. S. Thompson and W. A. Jury, Envi-
from Ottawa sand. ron. Sci. Technol. 31, 548–555 (1997).
Widespread use has been made of this property of 15. Y. Jin, Y. Chu and Y. Li, J. Contam. Hydrol. 43, 111–128
organic matter in the area of removing viruses from filters (2000).
to detect them in environmental samples. Mucks and other 16. Y. Chu, Y. Jin and M. V. Yates, J. Environ. Qual. 29(4),
soils with high organic matter content are poor adsorbers 1103–1110 (2000).
of viruses (46) and may not be suitable for wastewater 17. S. L. Penrod, T. M. Oslon, and S. B. grant, Langmuir 12,
application sites (33). 5576–5587 (1996).
In contrast to the other studies, Powelson and 18. K. Furuse, Distribution of Coliphages in the Environment:
coworkers (46) found that the type of effluent (primary, General considerations, Chap. 3, in S. M, Goyal, C. P Gerba
secondary, and tertiary) had no significant effect on the and G. Bitton, eds., Phage Ecology, John Wiley & Sons, New
York, 1987.
transport of three viruses (poliovirus 1, MS2, and PRD1)
19. C. P. Gerba, Phage as Indicators of Fecal Pollution, Chap. 8,
through columns of an alluvial coarse sand.
in S. M, Goyal, C. P Gerba and G. Bitton, eds., Phage
Ecology, John Wiley & Sons, New York, 1987.
20. D. C. DeBorde, W. W. Woessner, B. Lauerman and P. Ball,
CONCLUSION Water Res. 32, 3781–3875 (1998).
21. S. Farrah, Ecology of Phage in Freshwater Environments,
Bacteriophages have been isolated from wastewater, sur- Chap. 4, in S. M. Goyal, C. P, Gerba and G. Bitton, eds.,
face water, and groundwater. There is a great deal of Phage Ecology, John Wiley & Sons, New York, 1987.
interest in the potential for the use of coliphages as 23. W. A. Yanko et al., Water Res. 33, 53–64 (1999).
indicators of the presence of human enteric pathogens, 23a. U.S. Environmental ProtectionAgency, National Primary
as well as of the potential for fecal contamination Drinking Water Regulations: Ground Water Rule; Proposed
Rules, EPA-815-2-00-002, 2000.
of groundwater. While bacteriophages show a great
24. S. B. Primrose and M. Day, J. Appl. Bacteriol. 42, 417–421
deal of promise as indicators, all scientific and reg-
(1977).
ulatory communities have not yet accepted their use
24a. M. Abbaszadegan et al., Occurrence of Viruses in Ground
as such.
Water in the United States, American Water Works Service
The interaction of factors affecting the survival of Company, 1999.
coliphage in the environment is a complex issue that has 25. G. Bitton, Introduction to Environmental Virology, John
not yet been completely explained. The inactivation rates Wiley & Sons, New York, 1980.
of two coliphages and the temperatures and pH values 26. E. P. Larkin and A. C. Fassolitis, Appl. Environ. Microbiol.
at which those inactivation rates occurred are shown in 38, 650 (1967).
Table 5. Close examination of the data reveals that there 27. S. A. Sattar, Virus Survival in Receiving Waters, in
are trends, as discussed earlier, but quantitation of these M Goddard and M. Butler, eds., Viruses and Wastewater
trends is not yet possible. Treatment, Pergamon Press, New York, 1981.
 BACTERIOPHAGE OF ENTERIC BACTERIA: OCCURRENCE AND PERSISTANCE IN THE ENVIRONMENT 391

28. R. B. Kapuscinski and R. Mitchell, Water Res. 14, 363 Bull H. B., Symp. Quant. Biol. 6, 140–149 (1938).
(1980). Campbell-Renton M. L., J. Pathol. Bacteriol. 45, 237–251
29. C. P. Gerba G. Bitton, Microbial Pollutants: Their Survival (1937).
and Transport Pattern to Groundwater, in G. Bitton and Dimmock N. J., Virology 31, 338 (1967).
C. P Gerba, eds., Groundwater Pollution Microbiology, John
Drewry W. A. and Eliassen R., J. Water Pollut. Control Fed. 40,
Wiley & Sons, New York, 1984.
R257–R271 (1968).
30. C. J. Hurst, C. P. Gerba and I. Cech, Appl. Environ. Micro-
Duboise S. M., Moore B. E., Sorber C. A. and Sagik B. P., Crit.
biol. 40, 1067–1079 (1980).
Rev. Microbiol. 7, 245–258 (1979).
31. R. J. Salo and D. O. Cliver, Appl. Environ. Microbiol. 36, 68
Dulbecco R. and Ginsberg H. S., Virology, Harper and Row,
(1978).
Philadelphia, Pa., 1980.
32. J. P. Loveland, J. N. Ryan, G. L. Amy and R. W. Harvey,
Funderburg S. W., Moore B. E., Sagik B. P., and Sorber C. A.,
Colloids Surf. 107, 205–221 (1996).
Water Res. 15, 703–711 (1981).
33. M. D. Sobsey, Transport and Fate of Viruses in Soils,
Gerba C. P., Dev. Ind. Microbiol. 24, 247 (1983).
Microbial Health Considerations of Soil Disposal of
Domestic Wastewaters, U. S.Environmental Protection Gerba C. P. and Goyal S. M. Pathogen Removal From Wastewater
Agency, Cincinnati, Ohio, 1983. During Groundwater Recharge, in T Asano, ed., Artificial
Recharge of Groundwater, Butterworth Publishers, Boston,
34. C. P. Gerba, Adv. Appl. Microbiol. 30, 133 (1984).
Mass., 1985.
35. F. M. Burnet and M. McKie, Aust. J. Exp. Biol. Med. Sci. 7,
Gerba C. P., Goyal S. M., Cech I. and Bogdan G. F., Environ. Sci.
183 (1930).
Technol. 15, 940 (1981).
36. R. B. Thurman and C. P. Gerba, J. Environ. Sci. Health A22,
Gerba C. P. et al., Water Sci. Technol. 24, 95–102 (1991).
369–388 (1987).
37. W. D. Burge and N. K. Enkiri, J. Environ. Qual. 7, 73 Grant S. B., List E. J. and Lidstrom M. E., Water Resour. Res. 29,
(1978). 2067–2085 (1993).
38. S. M. Goyal and C. P. Gerba, Appl. Environ. Microbiol. 38, Jansons J., Edmonds L. W., Speight B. and Bucens M. R., Water
241 (1979). Res. 23, 293–299 (1989).
(a) J. P. Murray and S. J. Laband, Appl. Environ. Microbiol. Jansons J., Edmonds L. W., Speight B. and Bucens M. R., Water
37, 480 (1979). Res. 23, 301–306 (1989).
39. C. P. Gerba and G. E. Schaiberger, J. Water Pollut. Control Kinoshita T., Bales R. C., Maguire K. M., and Gerba C. P., J.
Fed. 47, 93 (1975). Contam. Hydrol. 14, 55–70 (1993).
40. D. C. Young and D. G. Sharp, Appl. Environ. Microbiol. 33, Koerner E. L. and Haws D. A. Long-term Effects of Land Appli-
168 (1977). cation of Domestic Wastewater: Vineland, New Jersey, Rapid
Infiltration Site, EPA-600/2-79-072, U.S. Environmental Pro-
41. M. D. Sobsey, T. Fujii and R. M. Hall, J. Am. Water Works
tection Agency.
Assoc. 83, 64–67 (1991).
McKay L. D. et al., Environ. Sci. Technol. 27, 1075–1079 (1993).
42. E. Lefler Y. Kott, Virus Retention and Survival in Sand, in
J. F Malina and B. P Sagik, eds., Virus Survival in Water Poletika N. N., Jury W. A. and Yates M. V., Water Resour. Res.
and Wastewater Systems, Center for Research in Water 31, 801–810 (1995).
Resources, Austin, Tex, 1974. Powelson D. K. and Gerba C. P., Water Res. 28, 2175–
43. D. K. Powelson, J. R. Simpson and C. P. Gerba, Appl. Envi- 2181 (1994).
ron. Microbiol. 57, 2192–2196 (1991). Powelson D. K., Simpson J. R. and Gerba C. P., J. Environ. Qual.
44. R. C. Bales et al., Environ. Sci. Technol. 25, 2088–2095 19, 396–401 (1990).
(1991). Schaub S. A. and Sorber C. A., Appl. Environ. Microbiol. 33, 609
45. A. P. Pieper et al., Environ. Sci. Technol. 31, 1163–1170 (1977).
(1997). Skilton H. and Wheeler D., J. Appl. Bacteriol. 65, 387–395 (1988).
46. R. L. Bixby and D. J. O’Brien, Appl. Environ. Microbiol. 38, Skilton H. and Wheeler D. J. Appl. Bacteriol. 66, 549–557 (1989).
840 (1979). Sobsey M. D., Dean C. H., Knuckles M. E., and Wagner R. A.,
47. D. K. Powelson, C. P. Gerba and M. T. Yahya, Water Res. Appl. Environ. Microbiol. 40, 92 (1980).
27, 583–590 (1993). Sobsey M. D. et al., Water Sci. Technol. 18, 97–106 (1986).
48. M. V. Yates, Report to EPA, 1992. (unpublished). U.S. Environmental Protection Agency, Fed. Reg. 57, 33, 960
(1992).
ADDITIONAL READING Vaughn J. M. and Landry E. F., Viruses in Soils and Groundwa-
ter, in G Berg, ed., Viral Pollution of the Environment, CRC
Adams M. H., J. Gen. Physiol. 31, 419–432 (1948). Press, Boca Raton, Fla., 1983.
Bales R. C., Gerba C. P., Grondin G. H., and Jensen S. L., Appl. Verwey E. J. W. and Overbeek J. T. G., Theory of the Stability of
Environ. Microbiol. 55, 2061–2067 (1989). Lycophobic Colloids, Elsevier, Amsterdam, The Netherlands,
Bitton G., Water Res. 9, 473–484 (1975). 1948.
Bitton G. and Harvey R. W., Transport of Pathogens Through Wallis C. and Melnick J. L., Tex. Rep. Biol. Med. 19, 683 (1961).
Soils and Aquifers, in R Mitchell, ed., Environmental Microbi- Wellings F. M., Lewis A. C., Mountain C. W. and Pierce L. V.,
ology, Wiley-Liss, New York, 1992. Appl. Microbiol. 29, 751 (1975).
Bitton G., Masterson N. and Gifford G. E., J. Environ. Qual. 5, Yahya M. T., Galsomies L., Gerba C. P. and Bales R. C., Water
370 (1976). Sci. Technol. 27, 409–412 (1993).
Bixby R. L. and O’Brien D. J., Appl. Environ. Microbiol. 38, 840 Yanko W., Progress Report, Soil Removal of Male-Specific
(1979). Coliphage During Groundwater Recharge With Reclaimed
392 BIOAEROSOL SAMPLING AND ANALYSIS

Water in Montebello Forebay Recharge Areas, County quarters may result in human exposure and adverse
Sanitation Districts of Los Angeles County, Whittier, Calif., health effects, ranging from lost productivity to severe
1994. illness. Detection and measurement of biocontaminants in
Yates M. V., Stetzenbach L. D., Gerba C. P., and Kelley L. M., J. indoor environments are needed to assess contamination
Environ. Sci. Eng. A25, 81–100 (1990). levels and estimate the resultant exposure of occupants. A
Yeager J. G. and O’Brien R. T., Appl. Environ. Microbiol. 38, 694 wide variety of sampling and analysis methods have been
(1979a).
used and new methods are being developed (3). However,
several problems remain to be solved. For instance, no
single sampling method is suitable for the collection and
BACTERIOPLANKTON. See PLANKTONIC analysis of all types of bioaerosols and no standardized
MICROORGANISMS: BACTERIOPLANKTON sampling protocols are currently available. Therefore
data between studies are often difficult to compare
because of differences in sampler design, collection time,
airflow rate, and analysis method. In addition, there
is currently no regulation limiting biocontamination or
BENTHIC ALGAE. See MEROPLANKTON bioaerosol concentrations for home, office, and school
environments. One reason for this is that human exposure
limits have not been established for bioaerosols because
of the lack of exposure, dose, and response data (1).
This complicates the use of sampling results for risk
BENTHIC-ASSOCIATED PRIMARY
assessment.
PRODUCTIVITY. See PRIMARY PRODUCTIVITY IN THE
Conventional biocontaminant monitoring relies on
MARINE ENVIRONMENT
the collection of airborne and surface microorganisms
and analysis of samples by either culture on artificial
growth media or microscopic assay. Although culture
and microscopy methods are generally effective analytical
BIOAEROSOL SAMPLING AND ANALYSIS tools, they are hampered by critical limitations. Culture
analysis methods underestimate concentrations because
MARK P. BUTTNER only culturable cells are enumerated and identified,
University of Nevada, Las whereas nonculturable organisms go undetected (4).
Vegas Microscopic assays are laborious and imprecise, and the
Las Vegas, Nevada level of identification is generally limited to the genus
level at best. The inaccuracy of conventional methods and
Aerobiology is the study of airborne microorganisms and lengthy analytical time required to characterize bioaerosol
their effects on human health and the environment. concentrations emphasizes the need for developing new
Historically, aerobiology has been concerned with the air- analytical techniques that can provide rapid and reliable
borne transmission of disease. However, in recent years, data for bioaerosol exposure monitoring.
research in this field has expanded due to an increased It is important for the investigator to consider carefully
awareness of the variety of health effects potentially aris- the objectives of sampling before any samples are
ing from exposure to airborne microorganisms. Today the taken. After determining what information is desired,
concentration and composition of airborne microorgan- an appropriate sampling and analysis method can be
isms is of interest in diverse areas, such as agricultural
selected and incorporated into the monitoring design.
and industrial settings, medicine, home and office envi-
The purpose of this chapter is to present various
ronments, and military research. Particular emphasis has
bioaerosol sampling and analysis methods to facilitate the
been placed recently on the areas of indoor environmental
selection of instrumentation and techniques. The collection
quality and exposure assessment.
methods used in bioaerosol sampling are presented with
The term bioaerosol is used to describe living
a description of sampling equipment currently available,
airborne particles or those that originate from living
including a discussion on sampling efficiency. Equipment
organisms (1). Microorganisms, including bacteria, fungi,
viruses, parasites and algae, as well as microbial calibration and sampling considerations such as collection
fragments, toxins and metabolites, may be present in times and the number of samples are also addressed.
bioaerosols. Bioaerosols can range in size from less than Analysis methods are presented in the second section
0.01 µm to more than 100 µm in diameter. Bioaerosols of the chapter, beginning with traditional culture and
in the respirable size range (≤10 µm) are of particular total count methods and concluding with some recently
concern to human health. developed biochemical and molecular techniques.
Human exposure to bioaerosols can result in a variety
of adverse health effects. Numerous microbial agents
representing every class of microorganisms have been BIOAEROSOL SAMPLING
associated with adverse health effects resulting from
bioaerosol exposure (2). Contamination and subsequent Bioaerosol sampling may be used to obtain useful data
dispersal of biocontaminants in the workplace and living for a variety of applications, such as (1) measuring
 BIOAEROSOL SAMPLING AND ANALYSIS 393

the presence and concentration of airborne biocontam- Impaction. The majority of bioaerosol samplers utilize
inants; (2) confirming dissemination from a biocontam- impaction as the collection principle. The impaction
inant source; (3) evaluation of bioaerosol exposure and method uses the inertia of the particles to separate them
health risks; (4) documentation of clearance to reoccupy from the air stream. The particles are deposited onto a
a space, following remediation efforts; and (5) compliance collection surface that usually consists of an agar medium
monitoring in industrial environments. Regardless of the for culture-based analysis or an adhesive-coated surface
application, the objective of bioaerosol sampling is the effi- for microscopic analysis. The impaction process depends
cient removal and collection of biological particles from on the inertial properties of the particle, such as size,
the air in a manner that does not affect the integrity density and velocity, and on the physical parameters of
of the sample. This is dependent on the characteristics the impactor such as inlet nozzle dimensions and air flow
of the microorganisms and on the physical features of pathway (7,8). The impactor sampler draws air through an
the sampling instrument (5,6). The selection of a sam- inlet nozzle toward a collection surface, and particles with
pler depends on a number of factors, such as sampler sufficient inertia impact while smaller particles remain
performance, expected bioaerosol concentration, and the in the air stream (Fig. 1a). Impactor samplers may be
analysis method. These factors are discussed later in the designed with multiple circular inlet nozzles; the sampler
chapter. is then referred to as a multiple-hole or sieve sampler. If
there are several stages with successively smaller nozzles,
the sampler is referred to as a cascade impactor. For
Sampling Methods and Equipment samplers in which air is drawn through a single nozzle, the
shape of the nozzle is usually rectangular and the impactor
There is a wide variety of commercially available is referred to as a slit sampler. Some impactors utilize
bioaerosol samplers (3,5,7). Bioaerosol samplers may be centrifugal impaction, which also uses inertial forces to
grouped according to their method of collection. The three separate the particle from the air stream, but in a radial
main collection methods used in quantitative bioaerosol geometry (Fig. 1b).
sampling are impaction, impingement, and filtration.
Electrostatic precipitation has also been used for the Multiple-Hole Impactor Samplers. The Andersen impac-
collection of airborne microorganisms. Nonquantitative tor sampler (Fig. 2) has been widely used for culturable
sampling by gravitational settling of particles will also be bioaerosol measurements (9). The sampler draws air
discussed. Several of the commercially available bioaerosol at a flow rate of 28.3 lpm and is operated using
samplers are listed in Table 1. an electric vacuum pump. The Andersen six-stage

(a) (b)

Air
steamline F Inertial Air
steamline
Particle trajectory Particle trajectory

F
Inertial

(d) F
(c)
Inertial
and
F others
Inertial

Particle
trajectory

Particle
trajectory
Air
steamlines

Figure 1. Mechanisms of particle removal used in bioaerosol sampling. (a) solid plate impaction;
(b) centrifugal impaction; (c) liquid impingement; (d) filtration; Finertial = inertial force (Adapted
from Nevalainen et al., 1993, with kind permission from Elsevier Science Ltd., The Boulevard,
Langford Lane, Kidlington OX51GB, United Kingdom).
Table 1. General Characteristics of Several Commercially Available Bioaerosol Samplers (see text for Manufacturer
Information; Adapted from Buttner et al., 1997, with kind permission from ASM Press, Washington, DC)

Sample
Air Flow Rate Analysis
Sampler Collection Medium (liters/minute) Method(s) Comments

Impaction

Multiple-Hole Impactors

Andersen viable impactors 1-, 2-, or Agar; 100-mm Petri dishes 28.3 C Particle size discrimination (2
6-stage and 6-stage models); vacuum
pump required; positive-hole
correction
Burkard portable air sampler Agar; l00-mm Petri dishes 10–20 C Battery operated; positive-hole
correction
MicroBio MB1 and MB2 55 mm contact plates 100 C Battery operated; positive-hole
correction
SAS Super 90, compact and high flow Agar; 55-mm contact plates 90 (compact, C Battery operated;
85-mm contact plates super 90) programmable for multiple
180 (high flow) samples; positive-hole
correction

Slit Impactors

Air-O-Cell sampling cassette Adhesive-coated surface 12–30 M Determination of total fungal

spores and pollen
Allergenco MK-3 Adhesive-coated glass slide 15 M Determination of total fungal
spores and pollen;
programmable for multiple
samples; electric or
battery-operated
Burkard spore traps, 24-hour, 7-day, Adhesive-coated surface, 10 M Determination of total fungal
and personal samplers tape, or glass slide spores and pollen; time
discrimination (24-hour and
7-day models); battery
operated (personal model)
Mattson/Garvin 220 and P-320 Agar; 150-mm Petri dishes 28.3 C Time discrimination up to
1 hour
New Brunswick slit-to-agar, STA-101, Agar; 150-mm Petri dishes 50 C Time discrimination up to
-203, and -303 1 hour; vacuum pump
required for some models
Rotorod Models 20 and 40 Adhesive-coated 48 M Determination of total fungal
polystyrene rods spores and pollen;
programmable for multiple
samples; electric or
battery-operated

Centrifugal Impactors

Biotest RCS, RCS plus and RCS high Agar; plastic strips 40 (RCS) C Battery operated; high volume
flow 50(RCS Plus) sampling (High Flow); strips
100(High Flow) available unfilled or with agar
medium
SASS 2000 Liquid 225 C,M,B,P,I High volume sampling; High
and low bioaerosol
concentrations; electric or
battery-operated

Impingement

All-glass impingers (AGI-30 and Liquid 12.5 C,M,B,P,I High and low bioaerosol
AGI-4) concentrations; vacuum pump
required

(continued overleaf )

394
 BIOAEROSOL SAMPLING AND ANALYSIS 395

Table 1. (Continued)

Sample
Air Flow Rate Analysis
Sampler Collection Medium (liters/minute) Methods Comments

BioSampler Liquid 12.5 C,M,B,P,I High and low bioaerosol

concentrations; short- and
long-term sampling; viscous
and nonviscous fluids;
vacuum pump required
Burkard multistage liquid impinger Liquid 20 C,M,B,P,I High and low bioaerosol
concentrations; particle size
discrimination; vacuum pump
required

Filtration
25, 37, or 47 mm filter cassettes Filter membrane 1–50 C,M,B,P,I Loss of culturable vegetative
cells; portable, useful for
personal monitoring; vacuum
pump required
Sartorius MD8 air sampler Gelatin membrane filter 42–133 C Gelatin filter reduces
desiccation stress; high
volume sampling; virus
collection

Electrostatic Precipitation

SCAEP PM-1B Liquid 115–371 C,M,B,P,I High volume sampling; High

and low bioaerosol
concentrations;

Note: C = Culture; M = Microscopy; B = Biochemical Assay; P = PCR; I = Immunoassay

impactor samplers designed for sampling culturable

airborne microorganisms. These include the Surface
Air System (SAS, International PBI, Milan, Italy,
distributed by Bioscience International, Rockville, Md.);
the Burkard portable air sampler (Burkard Manufacturing
Co., Ltd., Rickmansworth, Hertfordshire, England); and
the MicroBio MB1 and MB2 (Spiral Biotech, Inc.,
Bethesda, Md.) (Table 1).

Slit Impactors. Slit impactors are available for the

measurement of either culturable or total airborne
microorganisms. Many slit impactors have a moving
collection surface that allows enumeration of bioaerosol
concentration over time. Slit impactors that deposit the
bioaerosol onto an agar surface for the estimation of
culturable cells include the Mattson-Garvin air sampler
(Barramundi Corp., Homosassa, Fla.) and the slit-to-agar
sampler (New Brunswick Scientific Co., Inc., Edison, N.J.).
Figure 2. Andersen single-stage sampler (courtesy of Paula
Slit impactors that collect particles onto an adhesive-
Jacoby-Garrett, UNLV). coated surface are generally used for the microscopic
enumeration of fungal spores or pollen grains. These
include the Burkard spore traps and the personal air
sampler (Fig. 3) (Burkard Manufacturing Co., Ltd.; Spiral
impactor sampler (Andersen Instruments, Inc., Smyrna, Biotech, Inc.), the Air-O-Cell sampling cassette, (Zefon
Ga.) consists of six stages with decreasing nozzle International, Inc., St. Petersburg, Fla.), the Allergenco
diameter that collect progressively smaller particles air sampler (Allergenco/Blewstone Press, San Antonio,
onto agar plates. The number of colony-forming units TX) and the Rotorod (Sampling Technologies, Inc., St.
(CFU) that form on the agar collection plates provides Louis Park, MN) (Table 1).
bioaerosol size distribution information (9). Single-stage
and two-stage models of the sampler are also available. Centrifugal Impactors. The Reuter centrifugal sampler
There are several portable, battery-powered single-stage (RCS) and RCS plus (Biotest Diagnostics Corp., Denville,
396 BIOAEROSOL SAMPLING AND ANALYSIS

Figure 3. Burkard personal air sampler (courtesy of Paula

Jacoby-Garrett, UNLV). Figure 4. AGI-30 sampler (left) and SKC BioSampler (right)
(courtesy of Paula Jacoby-Garrett, UNLV).

N.J.) are portable battery-powered samplers that cen-

trifugally impact the microorganisms onto agar strips passage, making this sampler useful for studying the
for culture analysis. The agar strips containing selected respiratory infection potential of bioaerosols (10,11,12).
media formulations are available from the manufacturer. For other applications the inlet tube is washed with a
The Smart Air Sampler System (SASS) 2000 (Research known volume of collection fluid to recover nonrespirable
International, Woodinville, WA) is a wetted-wall cyclone airborne particles. The AGI-30 has an impaction distance
sampler that collects airborne particles in a thin liquid of 30 mm from the jet to the bottom of the sampler.
layer through centrifugal action. The AGI-4 model features a shorter distance of 4 mm
to improve particle collection efficiency over the AGI-
Liquid Impingement. Liquid impingement samplers 30. However, added sampling stress may result from
draw air through an inlet and particles are collected in impaction against the glass bottom of the sampler,
a liquid (Fig. 1c). The particle is removed from the air by resulting in a loss of cell viability (3,7). The Burkard
inertial force and impaction into a swirling or bubbling multi-stage liquid impinger (Burkard Manufacturing Co.,
liquid. Aggregates of cells may be broken apart by the Ltd.) is a stainless steel sampler that collects particles
action of the collection fluid. The collection of bioaerosol in three size fractions: >10 µm, 4 to 10 µm, and <4 µm.
particles in a liquid medium has several advantages, The BioSampler (Fig. 4) (SKC, Inc., Eighty Four, Penn.)
including the retrieval of bioaerosol particles over a wide combines impingement into a liquid with centrifugal
range of airborne particle concentrations, the ability to motion (13). The BioSampler collects particles by drawing
divide the sample for multiple analyses, and the option of air through three nozzles that are directed at an angle
applying a variety of analysis methods as discussed in the toward the inner sampler wall and the liquid swirls
following section. A liquid sample can be concentrated by upward on the inner wall of the sampler to remove
filtration or diluted by liquid addition, depending on the collected particles. Although other impingers are designed
concentration of collected microorganisms. In addition, for use with water-based collection fluids, the BioSampler
several culture media can be inoculated with aliquots can be used with viscous collection fluids (e.g., heavy white
of the collection medium for the culture of groups of mineral oil) to minimize sampling stress and evaporation
microorganisms with different nutrient requirements. loss of the collection buffer.
The AGI-30 all-glass impinger sampler (Fig. 4) (Ace
Glass, Inc., Vineland, N.J.) is a widely used sampler that Filtration. Filtration sampling is used to separate
has a curved inlet tube designed to simulate the nasal particles from the air stream by passing air through a
 BIOAEROSOL SAMPLING AND ANALYSIS 397

porous medium (Fig. 1d). Collection of particles depends Sampling Efficiency

on the aerodynamic diameter of the particle, the filter
No single sampler can be used for the collection and
pore size, and the airflow rate (14). Inertial forces and
analysis of all types of bioaerosols. The selection of the
other mechanisms such as interception, diffusion, and
appropriate sampler depends on a variety of factors
electrostatic attraction result in the collection of particles
including the sampling environment, the analysis methods
on the surface of the filter, including particles smaller
used, and the monitoring objectives. Consideration of the
than the pore size of the filter (12). Membrane filters
theoretical aspects of air sampling efficiency and the
housed in disposable plastic cassettes are typically used for
experimental results of performance studies may assist
filtration sampling (Table 1). The air sampling cassettes
in the selection of the appropriate sampling method for
may contain 25-, 37-, or 47-mm diameter filter membranes a particular monitoring situation. The performance of
and are available from a variety of manufacturers bioaerosol samplers can be divided into physical and
(e.g., Gelman Sciences, Ann Arbor, Mich.; Millipore biological components. The primary physical parameters
Corp., Bedford, Mass.; Nuclepore, Corning Costar Corp., that affect sampling efficiency are the inlet sampling
Cambridge, Mass.; Poretics Corp., Livermore, Calif.; SKC efficiency and the particle collection efficiency. These
Inc., Eighty Four, Penn.). Filter material, including criteria are presented briefly in the following section
polycarbonate, cellulose mixed ester or polyvinyl chloride and have been discussed in detail in the literature (7).
with a variety of pore sizes, may be used depending on Biological sampling efficiency refers to the effects of sample
the nature of the bioaerosol and the method of sample collection on the culturabilty or integrity of the cells or
analysis (7,12,14). cellular components for sample analysis.
The advantages of using filtration sampling for
bioaerosol monitoring are its simplicity, low cost, and Inlet Efficiency. The efficiency of the sampling inlet
versatility. Filtration sampling is adaptable to a variety affects the ability of the sampler to extract airborne
of assays but loss of viability of vegetative cells may particles from the environment. Ideally the inlet should
occur, presumably due to desiccation stress during entrain particles from the air without bias due to their
sampling (15,16,17). The MD8 air sampler (Sartorius AG, size, shape, or density. However, when sampling from a
Göttingen, Germany) collects airborne microorganisms on moving airstream, the air velocity and the orientation
a gelatin filter to reduce desiccation stress. The gelatin of the sampler may effect the inlet efficiency. The
membrane is incubated on an agar medium of choice for inlet characteristics of several bioaerosol samplers have
culture analysis. been calculated for different types of bioaerosol particles
sampled under various conditions (24). The results showed
Electrostatic Precipitation. This method of collection that certain sampling conditions could result in either
uses electrical forces to separate particles from the air overestimation or underestimation of the true bioaerosol
stream (18). The Space Charged Atomizing Electrostatic concentration. Therefore it is important that while
Precipitation (SCAEP) PM-1B sampler (Team Technolo- sampling in a moving airstream the samples be collected
gies, Newton Upper Falls, MA) operates by drawing air isokinetically, with the sampler inlet aligned with the
through a charged liquid spray that collects particles by airstream flow and with the velocity through the inlet
electrical forces and concentrates them into a liquid. equal to the airstream velocity (25).

Collection Efficiency. Collection efficiency, also termed

Gravity Sampling. Gravity sampling, also referred particle removal efficiency, is the ability of the sampler to
to as settling plate or depositional sampling, is a remove particles from the airstream and transfer them to
semi-quantitative collection method in which airborne the collection medium. Particle losses may occur between
microorganisms are collected by gravitational settling onto the sampler inlet and the collection device. The distance
an exposed agar plate. Although this method is simple between the inlet and the collection device is therefore
and inexpensive, collection of airborne microorganisms usually minimized to improve collection efficiency (7). For
using this method is affected by the size and shape impactor samplers a useful parameter for evaluating
of the particles and by the motion of the surrounding collection efficiency is the ‘‘cut-off diameter’’ or d50 . The
air (8). As a result, large particles are more likely d50 refers literally to the particle diameter at which 50%
to be deposited onto the collection surface (19). This of the particles are collected. However, because of the
can lead to misrepresentation of the prevalence of sharp cut-off characteristics of impactor samplers, the
airborne microorganisms and the exclusion of smaller d50 is generally considered the particle diameter above
particles from collection (20). Furthermore, the number which all particles are collected (26,27). The d50 values for
of colonies enumerated cannot be related to the bioaerosol several samplers have been calculated using the physical
concentration of the environment because the volume of dimensions of the impaction nozzle(s) and the airflow rate
air from which the particles originate is unknown. Gravity (Table 2). For membrane filters, the collection efficiency
sampling has been compared with various volumetric is approximately 100% for particles larger than the pore
culture sampling methods. The results show that the size (28).
airborne concentrations derived from gravity sampling
are not qualitatively or quantitatively accurate and do Biological Efficiency. Ideally a bioaerosol sampler
not compare favorably with those obtained by volumetric should collect all airborne microorganisms without alter-
sampling methods (20–23). ing the culturability or the biological integrity required
398 BIOAEROSOL SAMPLING AND ANALYSIS

Table 2. Calculated and Reported Cut Diameters (d50 ) for the stress of sampling may injure the collected microorgan-
Several Bioaerosol Samplers (Adapted from Buttner et al., isms, depending on their physiological characteristics (30).
1997, with kind permission from ASM Press, Washington, As a result, microorganisms may be in a noncultur-
DC) able state.
d50 (µm)a Many sampling methods rely on culturing the collected
microorganisms. Minimizing sampling stress is therefore
Sampler Calculated Reported critical. For agar or liquid collection media, it has been
observed that culturability of cells can be improved by
using certain culture media (31–33) or by the addition
Impaction
of certain compounds to the collection medium such as
Multiple-hole impactors osmoprotectants, which aid in the resuscitation of stressed
Andersen 6-stage viable impactor or damaged cells (34). Filtration sampling, although highly
efficient for the collection of airborne microorganisms, has
Stage 1 6.24,6.61b 7.0
the disadvantage of viability losses of vegetative cells,
Stage 2 4.21 4.7
Stage 3 2.86 3.3 presumably as a result of desiccation (15–17). Therefore
Stage 4 1.84 2.1 filtration sampling in combination with culture analysis
Stage 5 0.94 1.1 is generally used when bioaerosol concentrations are very
Stage 6 0.58 0.65 high and sampling times are short. Filtration sampling is
Andersen 2-stage viable impactor also used to monitor desiccation-resistant forms, such as
fungal spores or bacterial endospores, or in combination
Stage 0 6.28 8.0
Stage 1 0.83 0.95
with a total count method of analysis (35). The length of
Andersen single-stage viable impactor 0.58 0.65 collection time, discussed in the following section, also
Burkard portable air sampler 4.18,2.56c plays a major role in the efficacy of air sampling for the
MicroBio MB1 and MB2 1.8c retrieval of culturable microorganisms.
Surface Air System (SAS)
Super 90 1.94c 2.24c Sampler Performance. Numerous sampler comparison
Compact 1.97 2.0 studies have been published, evaluating overall sampler
High Flow 1.52,1.45b 2.0 performance and this work has been reviewed (3). The
Slit impactors Andersen 6-stage viable impactor and the AGI-30 liquid
Air-o-cell sampling cassette 2.3c impinger sampler have been suggested as reference
Allergenco MK-3 2.0c methods (5) due to their efficiency and wide spread
Burkard spore traps use, and many sampler efficiency studies have included
24-hour and 7-day these samplers in side-by-side comparisons. Some of the
Standard nozzle 3.70,5.2c factors that can negatively affect sampler performance are
High-efficiency nozzle 2.17 presented here and in the following sections.
Personal sampler 2.52 Agar impactor samplers collect microorganisms directly
Mattson-Garvin 220 and P-320 0.53 onto the culture medium. One of the disadvantages
Centrifugal impactors of this sampling method is that the nonculturable
Biotest RCS 7.5 3.8 component of the bioaerosol is not measured and can
Biotest RCS plus 6c 0.82c be a significant percentage of the total composition (35).
Another disadvantage of agar impaction is overloading
Impingement when bioaerosol concentrations are high, resulting in
All-glass impinger AGI-30 0.3 agar plates or strips that contain colonies that overlap
Biosampler <0.3d and are numerous to count (36). In addition, aggregates
Burkard multi-stage liquid impinger or clumps of microorganisms may impact at the same
Stage 1 10 place on the agar surface and be enumerated as a single
Stage 2 4 colony, resulting in underestimation of the bioaerosol
Stage 3 — concentration. Other problems that may occur with
a
impaction onto either agar or adhesive-coated collection
values obtained from Jensen et al., 1994, unless otherwise noted.
b
A. Nevalainen et al., 1992. surfaces and can lead to erroneous results are particle
c
K. Willeke and J. M. Macher, 1999. bounce, where particles rebound off the collection surface
d
K. Willeke et al., 1998. and reenter the airstream (8), and electrostatic forces that
attract biological particles to the surfaces of the collection
device (9).
Impingement of microorganisms by impaction onto a
for the detection and quantification of the microorgan- liquid-wetted surface may also result in culturability
isms or their metabolites. However, the infectivity and losses through sampling stress (9,37). Another difficulty
allergenicity of bioaerosols may be affected by numerous is that the microorganisms collected into a liquid may
chemical, physical, and biochemical factors (29). Airborne become reaerosolized from the bubbling collection fluid
microorganisms are subjected to a variety of environmen- during sampling. In addition, the collection efficiency of
tal stressors, such as UV radiation, chemical pollutants, liquid impingers may decrease significantly due to liquid
desiccation, and temperature extremes (4). In addition, evaporation over time (38).
 BIOAEROSOL SAMPLING AND ANALYSIS 399

Sampling Considerations sampling time for five bioaerosol samplers by using

the formula t = (δ)(A)/(Ca )(Q), where t is the sampling
In the development of a bioaerosol sampling design,
time, δ is the desired surface density, A is the area
the investigator should have an understanding of the
of the sampling surface, Ca is the average expected
objectives of sampling and the types of data that will
bioaerosol concentration, and Q is the sampler flow rate.
be obtained. The selection of the sampling and analysis
In this way the investigator determines the optimal
method(s) is a primary concern. Numerous other factors
number of colonies or cells on the collection surface and
should also be considered before sampling is initiated,
predicts the bioaerosol concentration. The drawback of this
including the length of the sample collection time, the
number of samples to be taken, and calibration of the method occurs when the actual bioaerosol concentration
sampling equipment. is substantially different from the estimate. For this
reason, more than one collection time is often employed
Collection Time. The goal of sample collection is to for sample collection to increase the likelihood of obtaining
obtain a representative sample of the bioaerosol of interest. useful data.
If sampling times are too short, too few microorganisms are Another factor that influences the selection of the sam-
collected to reliably estimate the bioaerosol concentration ple collection time is sampling stress, which can result
and composition. If sampling times are too long, the in the loss of culturability of airborne microorganisms
sampler may be overloaded with particles too numerous to or degradation of the sample. It has been observed that
enumerate and sampling stress may cause losses. In the increased sampling time has resulted in decreased via-
selection of sample collection times, parameters that must bility for aerosolized vegetative bacterial cells (10,39–41).
be considered are the expected bioaerosol concentration, Therefore a doubling of sampling time may not result in
the quantitation range of the sampler, and the effect of a doubling of CFU or other analytes, depending on the
sampling stress on the overall collection efficiency (3). stress tolerance of the bioaerosol components. In many
Bioaerosol concentrations can vary over several orders cases, taking several consecutive samples of short dura-
of magnitude, from <102 to >106 colony-forming units tion is preferable to fewer samples with a long collection
(CFU) per cubic meter of air, depending on the time. For further information on the selection of collection
environment. Concentrations of airborne microorganisms times, guidelines for the use of various bioaerosol samplers
can also fluctuate over time within the same environment. have been published (1,8).
Therefore a single air sample provides only a brief
glimpse of the environment in a particular time and Number of Samples. As previously stated, air sample
place. Although the ambient bioaerosol concentration collection periods are of relatively short duration and
is an unknown quantity, estimates of its concentration a single sample is of limited monitoring value due to
can be used to determine the sample collection time as the heterogeneity and variability of bioaerosols. Multiple
shown below. samples for each sampling method used should therefore
The quantitation range of a bioaerosol sampler can be be taken for bioaerosol assessment. Determining the
determined for a specific collection time using the sampler number of samples required for a particular situation
airflow rate (39). The lower detection limit (LDL) of a is dependent on several factors, including the objective of
sampler for a particular collection time may be calculated sampling, the statistical approach used to interpret the
assuming the detection of a single microorganism. Air data, the variability of the bioaerosol, the length of the
samples collected near the LDL of a sampler can contain sample collection period, and the available resources for
an insufficient number of particles to accurately represent sampling and analysis. The number of samples required
the true bioaerosol concentration. There is no definite should therefore be determined before sampling during
upper detection limit (UDL) for filtration or impingement the experimental design phase.
sampling methods because the liquid collection medium A high degree of variability has been shown between
or the liquid used to elute microorganisms from a filter paired samplers of the same type (21,42). Duplicate
membrane may be serially diluted prior to analysis. For samples are therefore often recommended, and the sample
impactor samplers the UDL is determined by the number measurement is reported as the average of the replicate
of sampling nozzles or the area of the collection surface. data points. Guidelines indicating the number of samples
Overcrowding of the impaction surface may result in required for monitoring in the occupational environment
errors in enumeration for microscopic and culture analysis have been published (1). For example, to estimate average
methods (see Sample Analysis in the following section). worker inhalation exposure, the ACGIH recommends
For multi-hole impaction samplers, the UDL is reached sampling in duplicate at least three times a day per
when a CFU develops under each of the sampling nozzles. site for at least three consecutive days for each sample
Other problems at or near the UDL include difficulties in type collected. Statistical methods can also be applied
resolving distinct particles or colonies (36) and inhibition to calculate the number of samples needed to estimate
of the growth of microorganisms as a result of competition confidence intervals around a mean or to compare mean
for space and nutrients. values from sampling locations.
One method for establishing the collection time for
a particular sampler is to determine the ideal surface Sampler Calibration. The concentration of a bioaerosol
density of microorganisms on the collection area and is calculated by dividing the number of microorganisms or
to estimate the order of magnitude of the bioaerosol. other analyte in the sample by the air sample volume (e.g.,
Nevalainen and coworkers (27) calculated the optimal CFU per cubic meter). Therefore the volumetric airflow
400 BIOAEROSOL SAMPLING AND ANALYSIS

rate of a sampler must be known and should be calibrated retrieval of culturable airborne fungi (31–33,44). Although
regularly. For impactors, the sampler’s d50 is a function of results and interpretations vary, a number of media
the airflow rate through the nozzle(s). A decrease in airflow have been recommended. Malt extract agar (MEA) is a
increases the d50 and the sample collection is shifted widely used fungal isolation medium. Of the formulations
toward larger particles (26). Conversely, an increase in available, unamended 2% MEA was found to promote
airflow decreases the d50 , resulting in increased small sporulation better than MEA amended with glucose
particle collection and potentially increased stress on and peptone (45,46). Rose bengal–containing agars (e.g.,
the microorganisms. Various calibration methods that rose bengal-streptomycin and dichloran-rose bengal-
measure the flow rate through a soap bubble meter or chloramphenicol) have been used to inhibit the spread of
an electronic calibration instrument are available (7). rapidly growing fungal genera (e.g., Rhizopus and Mucor),
allowing the enumeration and identification of other fungi
SAMPLE ANALYSIS in the sample. However, rose bengal is photoactivated
in direct sunlight, forming cytotoxic products that may
A wide variety of sample analysis methods can be inhibit fungal growth (47). Dichloran glycerol-18 is a
applied to air samples to provide information on the low water activity medium (aw = 0.955) developed for
concentration and composition of bioaerosol components. the isolation of xerophilic fungi (48) and it compares
The selection of an analysis method should be determined favorably with other media for culturing mesophilic
before monitoring is conducted because sampling and airborne fungi (33). Other commonly used fungal media
analysis methods are not necessarily compatible. Many include potato dextrose agar and Sabouraud’s agar (20,49).
impaction-sampling methods rely on the traditional It should be emphasized that many bacterial species are
analysis methods of culture or microscopy. Filtration capable of growth on general fungal media and the use
and impingement sampling methods are more flexible of antibacterial agents, such as chloramphenicol as an
with respect to sample analysis options. Limitations amendment to the medium, is recommended. Incubation
in the traditional analysis approaches have led to periods for fungi typically range from 3 to 10 days and
the development of alternative techniques, such as most airborne fungi are mesophilic and grow well at
biochemical, immunological, and molecular biological temperatures of 20 ° C to 25 ° C.
assays. New methods are continuously being developed For the culture of bacteria, several broad-spectrum
for the enhanced detection of bioaerosols (see ENHANCED media, including tryptic soy agar, nutrient agar, and
DETECTION OF AIRBORNE MICROBIAL CONTAMINANTS). casein soy peptone agar, are commonly used (12,43). These
media should be amended with antifungal agents such as
Culture cycloheximide to restrict the growth of fungi. Incubation
temperatures from 28 ° to 35 ° C for 1–7 days are usually
The majority of bioaerosol data reported has been
used for environmental and human source bacteria.
generated using culture analysis, and many of the
Important exceptions are the thermophilic actinomycetes,
currently available samplers are designed for collection
which are cultured at 55 ° C.
onto a nutrient agar surface. Airborne bacteria and
fungi collected in this way can be cultured directly,
whereas samples collected by impingement or filtration Enumeration. After the appropriate incubation period,
are processed and transferred to a culture medium. culturable microorganisms are determined by enumerat-
The major drawback of culture analysis is that only ing the colony-forming units (CFU). The concentration of
those cells that survive and reproduce to form visible culturable airborne microorganisms (CFU per cubic meter)
colonies are enumerated. Microorganisms that have been is calculated by dividing the number of CFU per sample
subjected to aerosolization and sampling stress may by the volume of air sampled. When microorganisms are
not respond to the artificial nutrient conditions in the collected using multiple-hole impactor samplers (Table 1),
laboratory. In addition, microorganisms exhibit a wide positive-hole corrections are generally applied to the data
range of nutritional and temperature requirements for before calculations to avoid underestimation of the cultur-
growth, and no formulation is capable of culturing every able bioaerosol concentration (9,21,50). These corrections
type of microorganism. For example, different media are are based on the probability of multiple impactions of
required for isolation of airborne bacteria and fungi. microorganisms through the same sampling hole. The
Therefore a common approach in bioaerosol monitoring magnitude and the standard deviation of the correction
is to choose general media that promote the growth increase with the number of CFU (7).
of the greatest diversity of species. For the isolation of When air samples are collected by impaction onto an
specific microorganisms or groups, conditions such as pH, agar plate and the bioaerosol concentration is unknown,
temperature, water activity, nutrients, antibiotics, light too few or too many colonies may be cultured, leading
and aeration, can be manipulated to select for the growth to enumeration errors. Colony counts that are too low
of the target microorganisms (43). Consequently, it may can be nonrepresentative of the population and exhibit
be necessary to perform replicate sampling using multiple high variability. High colony counts can lead to errors
types of nutrient media or to divide the samples prior to because of overlap of colonies and inhibitory effects of
inoculation. microorganisms on one another. When a limited amount
of overlap occurs, the true bioaerosol concentration can
Culture Media and Incubation. Several broad-spectrum be statistically calculated from the number of colonies on
media have been evaluated for their utility in the the agar (36,50). Enumeration errors for agar impaction
 BIOAEROSOL SAMPLING AND ANALYSIS 401

samplers can be reduced by selecting multiple sample microscopic analysis. Aliquots of collection buffer or the
collection times as previously discussed. For filtration or buffer used to elute cells from the filter membrane may
impingement sampling, the wash solution or collection be stained for epifluorescence microscopy by the acridine
buffer can be serially diluted to obtain counts within an orange direct count method (35,51,52). Fungal spores
acceptable range. may also be enumerated using the method. However,
some fungal spore types may resist staining or have
Identification. Identification of fungal isolates is based dark pigmentation that masks fluorescence (43). Cells
largely on microscopic determination of morphological and endospores may also be enumerated by bright-field
characteristics and requires training and expertise or phase-contrast microscopy using a hemocytometer or
in mycological techniques. Bacterial isolates may be some other cell counting chamber.
identified by several methods. Often culturable bacteria A disadvantage of microscopic analysis of air samples
are differentiated according to their Gram stain reaction. is the labor-intensive nature of microscopic enumeration.
Identification to the genus or species level requires the use Computerized image analysis systems that overcome this
of classical biochemical methods or various commercially limitation are available and can be used for enumeration.
available identification systems. An alternative to microscopic enumeration of total cells is
the use of flow cytometry, in which cell concentrations
Microscopy can be measured by light scattering and fluorescence
Microscopy is another commonly used method for the emitted from fluorochromes bound to cells. Impinger
enumeration of airborne microorganisms collected by samples labeled with DAPI yielded comparable results
impaction onto an adhesive-coated surface or by filtration. when enumerated by microscopy or flow cytometry (53).
Microscopy allows enumeration of total (culturable and Another difficulty of microscopic analysis is the level
nonculturable) microorganisms, unlike culture analysis of expertise required to identify fungal spores and pollen.
that only measures the culturable bioaerosol components. Misidentification and the inability to distinguish biological
This method can provide important exposure information from nonbiological particles are common sources of error
because some microorganisms that are nonviable or in this method (43).
nonculturable can still elicit an immune or toxic response
when inhaled. However, identification of microorganisms
beyond the genus level is usually not possible without Immunoassay
utilizing a taxon-specific technique such as immunospecific The limitations of traditional analysis methods and the
fluorescence staining (described in the following section). need for human exposure data for bioaerosol risk assess-
If information on viability or culturability is desired, ment have lead to the development of enhanced detection
microscopy can be used in conjunction with culture methods for air sample analysis. Immunoassay meth-
sampling and analysis. In addition, a variety of stains are ods have been recognized for their potential application
commercially available that differentiate respiring cells to bioaerosol analysis (19). Immunoassays rely on the
from nonrespiring cells.
binding of antibodies to a specific target antigen. Tar-
Microscopic enumeration is generally used for the
get antigens may be cell surface–associated proteins,
determination of total airborne fungal spores. Air samples
polysaccharides, or human allergens. The development
collected with slit impactors (e.g., Burkard spore traps,
of a specific antibody to the target antigen is required for
the Allergenco, and Air-o-Cell samplers) are prepared
these assays. The advantages of immunoassay methods
by mounting and staining the collection slide or tape,
for quantification of airborne allergens are their specificity
and fungal spores and pollen grains are enumerated
and sensitivity. The major limitation of immunoassays is
by light microscopy. The total number of spores or
that specific antigens for microorganisms are difficult to
pollen grains enumerated in a sample is divided by the
volume of air sample to estimate the total number of define and standardize (54). Variations of this method
spores or pollen grains per cubic meter of air. When that have been applied to bioaerosol analysis include
the sample is collected onto a moving surface, time fluorescence immunoassay, enzyme immunoassay, and
discrimination of fungal spore concentration is possible radioimmunoassay.
by enumerating spores present between specified points For fluorescence immunoassays, samples are stained
along the collection surface that represent discrete time with a fluorescence-labeled antibody that binds specifically
intervals. Several stains, including lactolphenol cotton to the antigens on the surfaces of the target organisms. The
blue, phenosaffranin, and basic fuchsin, are commonly samples are then analyzed by epifluorescence microscopy.
used to facilitate discrimination of spores or pollen grains Fluorescent labeling dyes include fluorescein, fluores-
from debris (43). Data may be reported as the total spore cein isothiocyanate, and rhodamine isothiocyanate (12).
concentration, the concentration of a specific target genus, Radioimmunoassays utilize radioactive-labeled antibod-
or the percent composition of fungal genera. However, ies and enzyme immunoassays utilize enzyme-labeled
differentiation of certain fungal genera is not possible by antibodies or antigens. The concentration of antigen is
spore morphology alone (e.g., Aspergillus and Penicillium measured by radioactivity or enzyme activity. These meth-
spp.). Therefore these genera are grouped when reporting ods have been applied to the measurement of airborne
percent composition of bioaerosol samples. allergens, such as dust mite allergen, animal dander, and
For the determination of total airborne bacteria, liquid β-1,3-glucan (55–58). Impaction, impingement, and filtra-
impingement or filtration sampling is generally used with tion samples may be analyzed using immunoassays (59).
402 BIOAEROSOL SAMPLING AND ANALYSIS

Biochemical Assay in a greenhouse when culture data were negative (76).

Using PCR, it is possible to obtain results within hours
Numerous cellular components or metabolites, such as
of sample collection, compared with days or weeks for
endotoxins and mycotoxins, may produce adverse health
culture methods. If there are multiple organisms of
effects when inhaled (60,61). Biochemical assays have
interest in a sample, multiplex PCR may be performed
been developed to measure these compounds directly in
in the same reaction mix to simultaneously amplify
an effort to relate environmental exposure to human
several DNA targets of interest. Liquid impingement
responses.
and filtration sampling methods have been used with
Endotoxins are lipid polysaccharides (LPS) present in
PCR analysis for the detection of target microorganisms
the cell walls of gram-negative bacterial cells. Elevated
(76–78).
levels of endotoxin due to the presence of airborne
A limitation of the PCR assay is the inability to dis-
gram-negative bacteria have been measured in bioaerosol
tinguish between culturable and nonculturable microor-
samples from a variety of agricultural, industrial, and
ganisms. In addition, the presence of PCR inhibitors in
office environments (62). The concentration of endotoxin in
environmental samples may reduce sensitivity or result
a sample may be measured using the Limulus amebocyte
in false negatives (80). A nucleic acid purification step or
lysate (LAL) test. The lysate of amebocytes from the
dilution of the sample may be required to remove environ-
horseshoe crab, Limulus polyphemus, gels in the presence
of LPS, and a variety of test systems have been developed mental interference to PCR amplification (81).
to quantify endotoxin using this method. Filtration
sampling is most often used to collect airborne endotoxin; Data Analysis
however, certain filter materials have been shown to
inhibit the LAL assay (63). In addition, interference with Data generated from air sampling studies can be used to
the LAL reaction by inhibitors present in environmental test hypotheses and answer scientific questions pertain-
samples must be accounted for in the measurement of ing to environmental biocontamination, biocontaminant
endotoxin content (64). dispersal, human exposure, and efficacy of remediation.
Mycotoxins associated with dusts from fungus- For data to be useful, quality assurance and statistical
contaminated grains are a known source of illness (65). considerations should be addressed before sample collec-
Exposure to Stachybotrys chartarum mycotoxins has been tion. A quality assurance plan should be developed to
associated with toxicoses in contaminated home and office minimize errors in sample collection and analysis (82,83).
environments (66–68). Filtration sampling using glass- Statistical considerations include determination of appro-
fiber filters or polycarbonate membranes has been used priate data reporting methods, statistical tests for data
to measure aerosolized mycotoxins of S. chartarum in analysis, and sample sizes. Statistical tests can be used to
the laboratory (69,70) and in the field (71). Other bio- compare means from various sampling locations and from
chemical methods that may be applied to the analysis of different sampling or analysis methods.
airborne fungal exposure include ergosterol (72,73) and There are several commonly used methods for data
β-1,3-glucan (56,74,75) assays. reporting (83). Often descriptive statistics, such as means,
standard errors, and confidence intervals, are used
Polymerase Chain Reaction for interpretation of bioaerosol concentration results.
Expressing data as the percent composition of genera
Polymerase chain reaction (PCR) is an example of a molec-
or groups is often helpful for population comparisons.
ular biological technique that has been applied to the
There are several acceptable approaches for reporting
detection of target microorganisms in air samples (76–78).
and analyzing bioaerosol data, and consideration of these
PCR is an enzymatic procedure used to rapidly amplify
factors in advance will increase the likelihood of obtaining
specific DNA sequences (79). Following amplification, the
meaningful data.
target DNA is present in sufficient quantity to allow
detection as low as the level of a single microorgan-
ism. The authenticity of the amplified DNA can be con-
firmed by sequencing, restriction analysis, or gene probe CONCLUSION
hybridization. The requirement for PCR detection of tar-
get microorganisms is that unique, non-cross-hybridizing There are many sampling and analysis methods available
sequences are available to serve as primers for amplifica- for bioaerosol monitoring. Increased awareness of the
tion. Quantitative PCR methods have been developed and variety of health effects potentially arising from exposure
are commercially available for the measurement of the to airborne microorganisms has led to the development
concentration of the DNA segments present in the original of numerous new analytical techniques that can provide
sample. rapid, reliable data for bioaerosol exposure monitoring.
The primary advantages of using PCR for the analysis Therefore it is important for the investigator to consider
of bioaerosols are sensitivity and speed. Because PCR carefully the objectives of sampling before any samples are
detection is not dependent on the culturable state of a taken. After determining what information is desired, an
microorganism, PCR can be used to detect microorganisms appropriate sampling and analysis method can be selected
that are nonculturable, difficult to culture, or grow slowly. and incorporated into the monitoring design. Reliable
For example, positive results with the PCR technique were bioaerosol data are needed to further the understanding of
demonstrated for the detection of bacteria aerosolized the effects of human exposure to airborne microorganisms
 BIOAEROSOL SAMPLING AND ANALYSIS 403

and aid in the development of standardized monitoring 24. S. A. Grinshpun, C. W. Chang, A. Nevalainen, and K. Wil-
protocols. leke, J. Aerosol Sci. 25, 1503–1522 (1994).
25. J. E. Brockmann, in K. Willeke and P. A. Baron, eds., Aerosol
Measurement: Principles, Techniques and Applications, Van
BIBLIOGRAPHY Nostrand Reinhold, New York, 1993, pp. 77–111.
26. V. A. Marple, K. L. Rubow, and B. A. Olson, in K. Willeke
1. J. M. Macher, ed., Bioaerosols: Assessment and Control, and P. A. Baron, eds., Aerosol Measurement: Principles,
American Conference of Governmental Industrial Hygienists, Techniques and Applications, Van Nostrand Reinhold, New
Cincinnati, Ohio, 1999. York, 1993, pp. 206–232.
2. L. D. Stetzenbach, in C. J. Hurst, ed., Manual of Environ- 27. A. Nevalainen, J. Pastuszka, F. Liebhaber, and K. Willeke,
mental Microbiology, ASM Press, Washington, D.C., 1997, Atmos. Environ. 26A, 531–540 (1992).
pp. 619–628. 28. M. Lippmann, in S. V. Hering, ed., Air Sampling Instruments
3. M. P. Buttner, K. Willeke, and S. A. Grinshpun, in C. J. for Evaluation of Atmospheric Contaminants, 7th ed.,
Hurst, ed., Manual of Environmental Microbiology, ASM American Conference of Governmental Industrial Hygienists,
Press, Washington, D.C., 1997, pp. 629–640. Cincinnati, Ohio, 1989, pp. 305–336.
4. C. S. Cox, Sci. Prog. Oxford 73, 469–500 (1989). 29. C. S. Cox, in C. S. Cox and C. M. Wathes, eds., Bioaerosols
5. M. A. Chatigny, J. M. Macher, H. A. Burge, and W. R. Solo- Handbook, CRC Press, Boca Raton, Fla., 1995, pp. 77–99.
mon, in S. V. Hering, ed., Air Sampling Instruments for 30. S. L. Stewart et al., Appl. Environ. Microbiol. 61, 1232–1239
Evaluation of Atmospheric Contaminants, 7th ed., American (1995).
Conference of Governmental Industrial Hygienists, Cincin- 31. H. P. Burge, W. R. Solomon, and J. R. Boise, J. Allergy Clin.
nati, Ohio, 1989, pp. 199–220. Immunol. 60, 199–203 (1977).
6. K. Willeke and P. A. Baron, eds., Aerosol Measurement: 32. K. L. Morring, W. G. Sorenson, and M. D. Attfield, Am. Ind.
Principles, Techniques and Applications, Van Nostrand Hyg. Assoc. J. 44, 662–664 (1983).
Reinhold, New York, 1993.
33. A. P. Verhoeff et al., Allergy 45, 275–284 (1990).
7. K. Willeke and J. M. Macher, in Bioaerosols: Assessment and
34. B. Marthi, in B. Lighthart and A. J. Mohr, eds., Atmospheric
Control, American Conference of Governmental Industrial
Microbial Aerosols: Theory and Applications, Chapman &
Hygienists, Cincinnati, Ohio, 1999, pp. 11-1–11-25.
Hall, New York, 1994, pp. 192–225.
8. A. Nevalainen et al., in K. Willeke and P. A. Baron, eds.,
35. U. Palmgren, G. Strom, G. Blomquist, and P. Malmberg, J.
Aerosol Measurement: Principles, Techniques and Applica-
Appl. Bacteriol. 61, 401–406 (1986).
tions, Van Nostrand Reinhold, New York, 1993, pp. 471–492.
36. C. W. Chang et al., Appl. Environ. Microbiol. 60, 3732–3738
9. A. A. Andersen, J. Bacteriol. 76, 471–484 (1958).
(1994).
10. C. S. Cox, The Aerobiological Pathway of Microorganisms,
37. M. E. Tyler and E. L. Shipe, Appl. Microbiol. 7, 337–349
John Wiley & Sons, Chichester, U.K., 1987.
(1959).
11. S. A. Grinshpun, C. W. Chang, A. Nevalainen, and K. Wil-
38. X. Lin, K. Willeke, V. Ulevicius, and S. A. Grinshpun, Amer.
leke, J. Aerosol Sci. 25, 1503–1522 (1994).
Ind. Hygiene Assoc. J. 58, 480–488 (1997).
12. P. A. Jensen, B. Lighthart, A. J. Mohr, and B. T. Shaffer, in
39. M. P. Buttner and L. D. Stetzenbach, Appl. Environ. Micro-
B. Lighthart and A. J. Mohr eds., Atmospheric Microbial
biol. 57, 1268–1270 (1991).
Aerosol: Theory and Applications, Chapman & Hall, New
York, 1994, pp. 226–284. 40. M. T. Hatch and H. Wolochow, in R. L. Dimmick and
A. B. Akers, eds., An Introduction to Experimental Aerobi-
13. K. Willeke, X. Lin, and S. A. Grinshpun, Aerosol Sci. Technol.
ology, John Wiley & Sons, New York, 1969, pp. 267–295.
29, 439–456 (1998).
41. M. W. Thompson et al., J. Aerosol Sci. 25, 1579–1593 (1994).
14. K. W. Lee and M. Ramamurthi, in K. Willeke and P. A.
Baron, eds., Aerosol Measurement: Principles, Techniques 42. L. L. Lembke, R. N. Kniseley, R. C. VanNostrand, and M. D.
and Applications, Van Nostrand Reinhold, New York, 1993, Hale, Appl. Environ. Microbiol. 42, 222–225 (1981).
pp. 179–205. 43. H. A. Burge, in H. A. Burge, ed., Bioaerosols, CRC Press, Boca
15. P. A. Jensen, W. F. Todd, G. N. Davis, and P. V. Scarpino, Raton, Fla., 1995, pp. 1–23.
Am. Ind. Hyg. Assoc. J. 53, 660–667 (1992). 44. T. Smid, E. Schokkin, J. S. M. Boleij, and D. Heederik, Am.
16. I. M. Lundholm, Appl. Environ. Microbiol. 44, 179–183 Ind. Hyg. Assoc. J. 50, 235–239 (1989).
(1982). 45. C. A. Hunter, C. Grant, B. Flannigan, and A. F. Bravery, Int.
17. P. S. Thorne, M. S. Kiekhaefer, P. Whitten, and K. J. Don- Biod. 24, 81–101 (1988).
ham, Appl. Environ. Microbiol. 58, 2543–2551 (1992). 46. D. P. Strachan, B. Flannigan, E. M. McCabe, and F. McGarry,
18. B. Crook, in C. S. Cox and C. M. Wathes, eds., Bioaerosols Thorax 45, 382–387 (1990).
Handbook, CRC Press, Boca Raton, Fla., 1995, pp. 269–283. 47. J. G. Banks, R. G. Board, J. Carger, and A. D. Dodge, J. Appl.
19. H. A. Burge and W. R. Solomon, Atmos. Environ. 21, 451–456 Bacteriol. 58, 392–400 (1985).
(1987). 48. A. D. Hocking and J. I. Pitt, Appl. Environ. Microbiol. 39,
20. W. R. Solomon, J. Allergy Clin. Immunol. 56, 235–242 (1975). 488–492 (1980).
21. M. P. Buttner and L. D. Stetzenbach, Appl. Environ. Micro- 49. S. R. Hirsch and J. A. Sosman, Ann. Allergy 36, 30–38 (1976).
biol. 59, 219–226 (1993). 50. J. M. Macher, Am. Ind. Hyg. Assoc. J. 50, 561–568 (1989).
22. W. J. Sayer, N. M. MacKnight, and H. W. Wilson, Am. J. 51. J. E. Hobbie, R. J. Daley, and S. Jasper, Appl. Environ.
Clin. Pathol. 58, 558–562 (1972). Microbiol. 33, 1225–1228 (1977).
23. W. J. Sayer, D. B. Shean, and J. Ghosseiri, J. Allergy 44, 52. M. Walters, D. Milton, L. Larsson, and T. Ford, Appl. Envi-
214–227 (1969). ron. Microbiol. 60, 996–1005 (1994).
404 BIOAEROSOLS IN AGRICULTURAL AND OUTDOOR SETTINGS

53. J. L. Lange, P. S. Thorne, and N. Lynch, Appl. Environ. BIOAEROSOLS IN AGRICULTURAL AND
Microbiol. 63, 1557–1563 (1997). OUTDOOR SETTINGS
54. T. M. Madelin and M. F. Madelin, in C. S. Cox and C. M.
Wathes, eds., Bioaerosols Handbook, CRC Press, Boca Raton, ESTELLE LEVETIN
Fla., 1995, pp. 361–386. The University of Tulsa
55. F. DeBlay, M. D. Chapman, and T. A. E. Platts-Mills, Am. Tulsa, Oklahoma
Rev. Respir. Dis. 143, 1334 (1991).
56. J. Douwes et al., Appl. Environ. Microbiol. 62, 3176–3182 Bioaerosols are biological agents carried in the air as large
(1996). molecules, volatile compounds, single particles, or clusters
57. C. M. Luczynska, Y. Li, M. D. Chapman, and T. A. E. Platts- of particles that are living or were released from a living
Mills, Am. Rev. Respir. Dis. 141, 361 (1990). organism. Included in this broad classification are viruses,
58. T. A. E. Platts-Mills et al., J. Allergy Clin. Immunol. 89, 1046 bacteria, fungal spores, algae, pollen grains, arthropod
(1992). particles, mammalian particles, complex metabolites, and
59. M. D. Chapman, in H. A. Burge, ed., Bioaerosols, CRC Press, volatile compounds. Particle sizes of bioaerosols range
Boca Raton, Fla., 1995, pp. 235–248. from approximately 0.05 to 100 µm and they occur both
indoors and outdoors with the greatest diversity typically
60. R. M. Castellan, S. A. Olenchock, K. B. Kinsley, and J. L.
Hankinson, N. Engl. J. Med. 317, 605–610 (1987).
occurring outdoors.
A tremendous variety of bioaerosols in the atmosphere
61. B. Crook, S. Higgins, and J. Lacey, in D. R. Houghton,
have been studied since the nineteenth century. Some of
R. N. Smith, and H. O. W. Eggins, eds., Biodeterioration 7,
Elsevier, London, U.K., 1987, pp. 791–797.
the early investigators include C.H. Blackley in England,
who first described chest tightness and ‘‘bronchial catarrh’’
62. R. R. Jacobs, Appl. Ind. Hyg. 4, 50–56 (1989).
brought on by inhaling fungal spores on moldy straw
63. D. K. Milton, R. J. Gere, H. A. Feldman, and I. A. Greaves, and D.D. Cunningham, who attempted to relate airborne
Am. Ind. Hyg. Assoc. J. 51, 331–337 (1990).
organisms to the incidence of disease in India. During
64. D. K. Milton, in H. A. Burge, ed., Bioaerosols, CRC Press, the early twentieth century, advances in aerobiology (the
Boca Raton, Fla., 1995, pp. 77–86. study of bioaerosols) were made by plant pathologists
65. B. Flannigan, Internat. Biodeterioration 23, 73–78 (1987). interested in the spread of agricultural pathogens,
66. W. S. Croft, B. B. Jarvis, and C. S. Yatawara, Atmos. Envi- specifically the movement of stem rust of wheat in the
ron. 20, 549–552 (1986). United States.
67. B. B. Jarvis et al., Appl. Environ. Microbiol. 64, 3620–3625 The development of the volumetric spore trap by Hirst
(1998). in 1952 was probably the most important catalyst in the
68. E. Johanning et al., Int. Arch. Occup. Environ. Health 68, scientific study of bioaerosols because it made it possible
207–218 (1996). to accurately estimate spore and pollen concentrations in
69. A. Pasanen et al., Atmos. Environ. 27A, 9–13 (1993). the air at any given time. Nearly 50 years later, spore
70. W. G. Sorenson et al., Appl. Environ. Microbiol. 53, 1370– traps based on Hirst’s design are among the most widely
1375 (1987). used instruments for studying bioaerosols. Analyses of
air samples have shown that fungal spores are the most
71. I. Yike, T. Allan, W. G. Sorenson, and D. G. Dearborn, Appl.
Environ. Microbiol. 65, 88–94 (1999).
abundant types of bioaerosol particles (1). This entry
will discuss fungal spores in the atmosphere and their
72. J. D. Miller et al., Internat. Biodeterioration 24, 103–120
importance in agricultural settings. Other bioaerosols are
(1988).
addressed in other sections of this volume.
73. J. D. Miller and J. C. Young, Am. Ind. Hyg. Assoc. J. 58,
39–43 (1997).
74. B. Fogelmark et al., Agent Actions 35, 50–56 (1992). SOURCES OF AIRBORNE FUNGI
75. H. Goto, K. Yuasa, and R. Rylander, Am. J. Ind. Med. 25,
81–83 (1994). Fungal spores are a normal component of the atmosphere,
and they occur in large numbers whenever the ground
76. A. J. Alvarez et al., Appl. Environ. Microbiol. 60, 374–376
(1994).
is free of ice and snow. Atmospheric concentrations may
exceed 200,000 spores/m3 of air (2), although the levels
77. M. H. Sawyer et al., J. Infect. Dis. 169, 91–94 (1994).
are typically an order of magnitude lower. Spores are
78. K. D. C. Stärk, J. Nicolet, and J. Frey, Appl. Environ. Micro- produced by fungal saprobes, pathogens, and symbionts;
biol. 64, 543–548 (1998).
and they are dispersed into the atmosphere by various
79. R. K. Saiki et al., Science 230, 1350–1354 (1985). methods through the action of both wind and rain. The
80. I. G. Wilson, Appl. Environ. Microbiol. 63, 3741–3751 (1997). enormous number of spores that are produced has been
81. A. J. Alvarez, M. P. Buttner, and L. D. Stetzenbach, Appl. well documented for many species (3–5).
Environ. Microbiol. 61, 3639–3644 (1995). Although fungi are everywhere in the environment,
82. L. D. Stetzenbach and A. Cross-Smiecinski, Quality Planning the major sources of the airborne spores are considered
for the Life Science Researcher: Meeting Quality Assurance to be from fungi growing on living plants and those
Requirements, CRC Press, Boca Raton, Fla., 1994. growing on leaf litter (1). Cladosporium and Alternaria,
83. J. M. Macher, Bioaerosols: Assessment and Control, American which frequently colonize both substrates, are typically
Conference of Governmental Industrial Hygienists, Cincin- the two most abundant genera of spores registered during
nati, 1999, pp. 13-1–13-16. atmospheric sampling. Lacey (6) found that natural and
 BIOAEROSOLS IN AGRICULTURAL AND OUTDOOR SETTINGS 405

cultivated grasslands were important spore reservoirs. He by nearby sources of spores (2). During crop harvesting or
showed that spores from Cladosporium, Alternaria, and mowing, remarkable levels of spores (up to 109 spores/m3 )
Epicoccum, which colonize senescent and dead grasses, are released into the atmosphere with Alternaria, Cla-
reached high concentrations in the atmosphere (up to dosporium, and Epicoccum, typically the most abundant
106 spores/m3 ) when a wheeled spore trap was drawn taxa (2,8,13).
over an area of grass. Yeasts and yeast-like fungi are Rain can passively disperse spores by various mech-
also abundant on leaf surfaces, and are the sources of anisms. Members of the dry air spora may be removed
yeasts in the outdoor air (7). Soil also contains many by raindrops, as they strike a leaf surface causing vibra-
fungi; soil moisture and wind speed appear to be the tion and puffing (8). This generally occurs with the initial
critical environmental factors that control their dispersal. raindrops, and it may result in increased atmospheric
Spores will only become airborne when soils are dry and concentrations of Cladosporium, Alternaria, and other
wind speeds high enough to entrain soil particles into the conidia. This may explain some anomalous results of
atmosphere (1). air sampling that show increases of these spores during
rain events (14,15). A similar tapping mechanism occurs
among puffballs. These fungi are unique among the basid-
DISPERSAL MECHANISMS iomycetes as their spores are passively dispersed. They are
released in a puff when a raindrop strikes the peridium of
The entrainment of spores into the atmosphere is the mature fruiting body.
dependent on the method of discharge by individual Spores of many fungi are dispersed by rain
types of fungi as well as meteorological factors such splash (9,16–18). Splash-borne spores are usually formed
as temperature, humidity, and wind speed. Spores are in mucilage, which inhibits their dispersal by wind and
discharged into the air either passively (relying on outside protects them from desiccation. During rain, the first
forces) or through active mechanisms (relying on drying drops dissolve the mucilage and leave a spore suspension
or the development of osmotic pressure). Although both available for splash dispersal by additional raindrops. For
wind and rain play an important role in some dispersal example, in Colletotrichum gloeosporioides (19), the num-
mechanisms, wind appears to be the major factor (11). The ber of conidia was found to vary from drop to drop. Splash
literature on this topic is extensive. Lacey (8) provided droplets from the first drop released some mucilage and
a general review, and McCartney and Fitt (9) recently only a few conidia; the largest number of conidia occurred
described the dispersal mechanisms for foliar fungal in the fifth drop. The size of the droplet may also affect the
pathogens. number of conidia as the smallest droplets are unable to
carry spores. Studies show that most spores are carried by
Passive Mechanisms droplets greater than 100 µm in diameter (8,12). In addi-
Wind dispersal is well studied and is the method resulting tion to precipitation, which is the most important medium
in the most abundant fungal spore types occurring in the for splash dispersal, overhead irrigation can also disperse
atmosphere. Many wind-dispersed spores are produced fungal spores using the same mechanisms.
on erect sporangiophores or conidiophores, which position Many splash-dispersed spores have thin colorless walls
the spores above the substrate and its boundary layer, and elongated shapes, such as conidia of Fusarium
thereby facilitating dispersal. Others are produced on leaf and Pseudocercosporella. Many frequently have adhesive
surfaces, which are well above the ground so that the properties that enable them to stick to the new host
spores are liberated directly into the turbulent layer of the surface. Although these spores lack the protective features
atmosphere (8). of thick, pigmented walls commonly found in members of
For many of these wind-dispersed spores, the atmo- the dry air spora, they are protected from desiccation by
spheric concentration depends upon the ease with which their mucilage during dry periods and by wet conditions
spores are detached from the reproductive structure or during dispersal and deposition (16). For plant pathogenic
hyphae at different wind speeds and levels of turbulence. fungi, splash dispersal is second to wind in importance
Minimum wind speeds needed to remove spores range for pathogen dispersal, but is only of significance for local
dispersal (16).
from 0.2 to 2.0 m/second (8,10). Wind gusts facilitate their
Raindrops are also efficient spore collectors, and
removal, as does the movement of the leaf itself. Aylor (11)
even a light rain can cleanse most spores from the
and Pedgley (12) have suggested that strong gusts are able
atmosphere (20). The surface properties of the spore are
to disperse spores from leaf surfaces even when the aver-
important for determining the outcome of this washout.
age wind speeds are too low to accomplish this. Warm, dry
Spores with a wettable surface are incorporated within
weather promotes wind dispersal for both pathogenic and
the raindrop, while those that are nonwettable remain on
saprobic fungi, and the entrained propagules are referred
the surface of the drop. Wettable spores will stay with the
to as the dry air spora. They include the most abundant
drop and settle when the raindrop comes to rest, while the
fungi in the atmosphere and include conidia of Cladospo-
nonwettable spores may be left on leaves or other surfaces
rium, Alternaria, Botrytis, Drechslera, Epicoccum, smut
as the raindrop rolls over the surface (4).
teliospores, and rust uredospores. Spores of the dry air
spora usually peak during the afternoon hours under con-
Active Mechanisms
ditions of low humidity and maximum wind speeds (5).
Although there are similarities in the dry air spora world- Spores of many fungi are dispersed by a variety of
wide, at any one time the air spora may be dominated active discharge mechanisms that propel spores into
406 BIOAEROSOLS IN AGRICULTURAL AND OUTDOOR SETTINGS

the turbulent layer of the atmosphere. The ballistics of rainless days as well (23). The average daily concentra-
some species are quite amazing, and these have been tion of Leptosphaeria spores in the Tulsa atmosphere
studied by many researchers throughout the twentieth during 1994 and 1995 was 69 spores/m3 . On rainless
century (3,4,8). days the average concentration was 44 spores/m3 but
In the majority of ascomycetes, the explosive discharge on rainy days it increased significantly to 109 spores/m3
of ascospores from the ascus is linked to available (t728 = −7.52, p < 0.01). These spores were present on 607
moisture. The absorption of water within the ascus days during the 2-year period, but rain only occurred on
leads to swelling and the development of high osmotic 271 days. The data suggest that dew is effective in stimu-
pressure (21). The resulting pressure causes the ascus to lating spore release in some Leptosphaeria species. Water
burst, explosively propelling the spores into the turbulent stored within the organism can also play a role in spore
layer of the atmosphere. Characteristic bursting methods release. Daldinia concentrica fruiting bodies were shown
(apical slit, hinged lid, apical cap, or minute pore) are to release ascospores during the night throughout the sum-
related to the taxonomic position of individual species (4). mer even in the absence of rain. In fact, when a mature
The osmotic force can be assessed by examining the fruiting body was detached from its substrate and brought
distance the spore is propelled, which ranges from less indoors, spore discharge continued for weeks from water
than 1 mm to 40 cm. Because of moisture requirement, stored in the stromal tissues (4). Other species not depen-
ascospores are usually abundant in the atmosphere during dent on direct wetting by rain include Epichloe typhina
and after rainfall or even during times of high humidity. and Bulgaria inquinans. The former, a grass pathogen,
Diurnal rhythms among the ascomycetes are complicated is apparently able to obtain sufficient moisture for spore
by this water requirement. Provided the water supply is release from the transpiration stream of its host, whereas
not limited, ascomycetes with a perithecium (flask-shaped the later is able to obtain sufficient moisture from the
fruiting body) tend to be either diurnal or nocturnal (4). mucilaginous layer of the apothecium to sustain spore
Many ascomycete fruiting bodies are xerotolerant but release for several days (4).
only release ascospores under wet conditions (4). During Discharge of basidiospores from most basidiomycetes
dry periods, the fruiting bodies dry out and become requires moisture but the mechanism is quite different
inactive. When wetted again, the fruiting bodies recover than that found in the ascomycetes. Money (26) has
quickly releasing spores into the atmosphere soon after recently reviewed this topic, which has been of intense
rainfall. This explains spectacular increases in spore levels interest to mycologists for over 100 years. Basidiospores
soon after rainfall. The amount of moisture necessary for form externally on basidia, with four basidiospores
this recovery is highly variable, and in some ascomycetes typically produced by each basidium. Each basidiospore
such as Mycosphaerella moisture from dew is sufficient for develops asymmetrically and is attached to the sterigma
spore liberation. In Venturia inaequalis, as little as 0.2 mm by a hilar appendage. Moisture from the atmosphere
of rain will lead to the release of abundant ascospores, but condenses around a crystal of mannitol and hexoses
dew is not effective. In other fungi more than 1.2 mm of excreted at the hilar appendage, and the resulting drop
rain are needed (4). of fluid, called Buller’s drop, quickly increases in size (26).
Ascomycetes also show differences in the speed of The moisture in Buller’s drop ultimately fuses with
response following wetting. Ophiobolus begins spore a film of water that surrounds the basidiospore and
release within 15 minutes (4). Release of Mycosphaerella produces a shift in the center of gravity (27). The result
musicola ascospores began within 10 minutes after of this transfer is that the spore is violently shot from
wetting infected leaf tissue, with 85% of the spores the basidium. This mechanism has been described as
being released from the perithecia within 2 hours (24). a surface tension catapult. Although the discharge only
Similar results were seen in Mycosphaerella citri and propels spores a fraction of a millimeter, the spores are
V. inaequalis (25). Ascospore release in M. citri began a liberated from the basidial layer and able to fall free
few minutes after rain and continued for about 1 hour. under gravity and be swept into the turbulent layer of the
However, a second rain event 3 hours later released few atmosphere.
spores, suggesting a depletion of mature spores after the The need for atmospheric moisture confines basid-
initial rain. In V. inaequalis, ascospore release began 5 iospore discharge to periods of high humidity. The presence
minutes after the start of rain and the maximum release of rain, however, may be a deterrent because liquid water
was approximately 1 hour later. In addition to spore would disperse the solutes from the hilar appendage. Var-
release directly related to wetting, other ascospores are ious studies have shown peak basidiospore concentrations
released only as the fruiting body begins to dry following during nighttime or early morning hours when humidi-
a rain event (4). ties are high (14,28–30). Haard and Kramer (28) studied
The moisture requirement can be met by several differ- spore discharge in a number of mushrooms and bracket
ent sources of water. For example, Richardson (22) found fungi and found three patterns of spore discharge. Many
that Didymella spores were present in the Edinburgh species exhibited a peak during the middle of the night,
atmosphere following rain as well as during periods of while others showed a double peak with spore release in
high humidity, but spore release only occurred at night. the early morning and also in the evening. A third group
Leptosphaeria is abundant in the atmosphere following composed of small mushrooms exhibited continuous spore
rain (4); however, a background level of Leptosphaeria release with the peak 24 to 48 hours after the onset of
was found in the Tulsa, Okla., USA, atmosphere on discharge.
 BIOAEROSOLS IN AGRICULTURAL AND OUTDOOR SETTINGS 407

The active discharge of both ascospores and basid- bioaerosols can be carried downwind for thousands of
iospores requires increasing atmospheric moisture; how- kilometers. One of the best-studied examples of long-
ever, the active dispersal of some oomycete sporangia distance transport is the movement of Puccinia graminis
occurs during drying conditions. Rapid decreases in rela- uredospores from southern United States and Mexico to
tive humidity during the morning hours are thought to be the wheat belt in the northern United States and Canada.
responsible for the hygroscopic twisting of the sporangio- This and other instances of long-distance dispersal of rust
phores in species of Phytophthora and Peronospora (4,8). It spores have been reviewed by Nagarajan and Singh (34)
is believed that during twisting, the sporangia from these and Pedgley (12).
fungi are flung into the air. Recently, it has been suggested During airborne transport, fungal spores are vulnerable
that electrostatic repulsion may also be involved, as the to various types of environmental damage. Exposure to
leaves and the sporangia become charged and repel each harmful UV radiation and extremes of temperature and
other during rapid drying conditions (31). humidity can decrease viability of pathogenic species.
Damage may be greater for species with thin-walled,
colorless spores that are easily plasmolyzed and lack
TRANSPORT IN THE ATMOSPHERE
the protection provided by melanin found in pigmented
spores (32). Despite the environmental hazards, a small
When spores are discharged from the parent hyphae or
percentage of spores are able to survive long-range
fruiting bodies, they pass through several layers of the
transport and cause infection (34). Although pathogenic
atmosphere. An extremely thin film of still air exists at
spores may lose viability during transport, it should be
the surface of the ground and on the surface of objects;
emphasized that spores probably retain their allergenic
air in this zone is held to the surface by molecular forces. properties even when no longer viable.
The laminar boundary layer occurs above this zone. Here,
airflow is typically parallel to the surface. The laminar
boundary layer is usually thin, 1 mm, but the thickness DEPOSITION
will vary with wind speed and topography (5). During
the night when air is very still the thickness of the Spores can be deposited by sedimentation, impaction,
laminar layer increases exponentially; however, during boundary-layer exchange, turbulent deposition, electro-
high wind speeds the thickness of the layer decreases to static deposition, and through rainfall (34). In still air,
less than 1 mm. Above the laminar layer is the turbulent spores would fall to the ground in response to gravity
layer in which air movement is constantly shifting and at a rate (based on Stokes’ law) that is proportional
unpredictable. The turbulence depends on wind speed to the square of the spore radius (5). However, in the
and direction, temperature, and the roughness of the natural environment air is seldom calm, and the effect
terrain. Bioaerosols are carried in the turbulent layer of sedimentation is insignificant at wind speeds at or
and the dispersion is related to the intensity of the above 2 m/second. Under normal atmospheric wind speeds,
turbulence (5,32). impaction and turbulent deposition are the most impor-
Plant pathogen spores within a crop must first pass tant mechanisms for deposition. Impaction occurs when
from the laminar layer on the leaf surface into the air currents approach a leaf or twig or other objects and
turbulent layer within the crop. Wind gusts and turbulence are deflected around it. Spores carried in the air diverge
enhance spore removal, but before spores reach the open from the deflected air stream by amounts depending on
air above the crop they must also pass through the the size of the spore and the size of the object. Impaction is
boundary layer surrounding the crop (11). Under normal most efficient when large spores impact on small objects at
turbulence as much as 90% of the spores are deposited high wind speeds (8). Very small spores, at or below 5 µm,
within 100 m with greater escape occurring under more would require wind speeds of 25 m/second for impaction
turbulent winds (8). The location of lesions within the to occur. Impaction may be one of the most important
canopy will also affect escape of the spores. Spores methods for spore deposition in crops. Turbulent deposi-
produced high in the canopy will have greater chance tion occurs when air blowing over a horizontal surface
of atmospheric entrainment as they are exposed to high deposits spores at a greater rate than would occur by grav-
winds and greater turbulence than spores lower in the ity. Turbulent deposition increases at higher wind speeds,
canopy. Consequently, disease will spread more rapidly although some bounce may occur as well (5,8).
through the field if infection occurs in the upper canopy
than if the pathogen is confined to the bottom or middle of TYPES OF FUNGAL SPORES IN THE AIR SPORA
the canopy (11).
Spores are transported vertically and horizontally by Many fungi are cosmopolitan; as a result, there are
wind. Spores can be carried upward by thermals and have similarities in the air spora worldwide. From agricultural
been recovered from various altitudes up to 5,000 m (5). to urban areas, and from arctic to tropical areas, it is
Pady and coworkers (33) reported the recovery of rusts, possible to find many of the same types of spores in the
smuts, Alternaria, Helminthosporium-type, and many atmosphere whenever the ground is free of ice or snow.
unidentified spores on slides exposed during flights over Although some spore types are common throughout the
Canada at 1,200 to 1,500 m. Gregory (5) reviewed the year, other taxa may peak in the spring, summer, or
numerous studies from the 1920s through the 1960s fall. In addition to seasonal variation, most fungi show a
that recovered a variety of fungal spores, especially diurnal rhythm with peaks occurring at specific times of
plant pathogens, from various altitudes. Horizontally, the day.
408 BIOAEROSOLS IN AGRICULTURAL AND OUTDOOR SETTINGS

Examination of air samples from most areas of the prominent components of the air spora, especially focusing
world will show that many of the spore types are ones on those implicated in plant diseases.
that potentially have an impact on humanity as aller-
gens, potential human pathogens, and devastating crop Dry Air Spora
pathogens. Approximately 20% of the human population
In dry weather the atmosphere in temperate areas is
suffer from respiratory allergies and many of these individ-
dominated by spores of fungal saprophytes and pathogens,
uals are hypersensitive to fungal spores. While exposure which are passively dispersed by the wind. Cladosporium,
to airborne fungal spores is typically associated with Alternaria, Epicoccum, Drechslera, Bipolaris, Pithomyces,
asthma and allergic rhinitis (hay fever), spores are also Curvularia, smut teliospores, and rust uredospores are
known to cause hypersensitivity pneumonitis (2). Aller- among taxa typically found in the dry air spora (Fig. 1).
gens have been reported from every major group of fungi,
but the most widely studied allergens are the asexual Cladosporium. Cladosporium is a large genus of
spores that are components of the dry air spora (35,36). anamorphic fungi with a wide variation in spore size
In 1952, Gregory and Hirst (37) suggested airborne basid- and shape (44,45). Several species of Cladosporium are
iospores as possible allergens, and in the past two decades, anamorphs of Mycosphaerella or Venturia spp. Spores are
clinical evidence has been collected on the occurrence of typically ellipsoidal to cylindrical and slightly pigmented
basidiospore-induced allergies (36,38). Far less is known (Fig. 2). Spores may be unicellular or have one or two
about the allergenicity of ascospores although anecdotal septa. The size range of spores in the genus is seen in
evidence has suggested these are also important triggers common species. Spores of C. sphaerospermum and C.
of allergic disease (39). cladosporoides range from 3–4.5 µm in diameter for the
Although the focus of this entry is outdoor bioaerosols, former to 3–7 × 2–4µm, for the latter; while those of
especially related to agricultural applications, concentra- C. macrocarpum average 15–25 × 7–10 µm.
tions of fungal spores in farm buildings are also of interest Conidia of Cladosporium species are generally the most
as they may impact respiratory diseases in farm workers abundant airborne spore type in temperate areas of the
and their families as well as storage disease in crop plants.
Wherever crops or hay are stored, airborne concentrations
of storage fungi are high, with levels up to 1010 spores/m3
recorded (2,40). Penicillium and Aspergillus are frequently
the dominant fungi found; however, the taxa depend on
the actual storage conditions. Pasenen and coworkers (41)
studied the airborne spore load in urban and rural areas
in Finland. They found that cow barns had levels up to
109 spores/m3 , with the highest levels associated with
areas where hay was handled. Airborne concentrations up
to 105 spores/m3 were found in farmhouses; these levels
were 10 to 1,000 times higher than levels measured in
urban dwellings. Pasenen and coworkers (41) suggested
that many of the fungal spores were introduced into the
farmhouses on the person or clothing of the residents.
Many fungal spores are allergenic, but only a small
group of fungi are considered human pathogens (42). Figure 1. Components of the dry air spora from Tulsa, Okla., air
The majority of mycoses (human fungal infections) are sample.
infections of the skin or mucous membranes, such as
athlete’s foot and thrush, with only a few fungi considered
systemic human pathogens. However, many common
airborne fungi can also cause serious human infections in
immunocompromised-compromised individuals. Species of
Aspergillus and Fusarium are two of the more common
types of fungi in this category (36). These infections
are of special concern because fungi do not respond to
conventional antibiotic treatment, and mortality rates
are high.
The majority of plant diseases are caused by fungi
with approximately 10,000 fungal species recognized as
plant pathogens. Although many important plant diseases
are caused by soil fungi, these seldom cause sudden,
widespread epidemics (43). By contrast, fungal pathogens
with an airborne dispersal phase can quickly spread from
the initial infection focus and cause widespread epidemics.
The remainder of this section will examine the most Figure 2. Cladosporium cladosporoides conidia from air sample.
 BIOAEROSOLS IN AGRICULTURAL AND OUTDOOR SETTINGS 409

world. Cladosporium species are common saprobes on leaf

surfaces, dead herbaceous and woody plants, soil, foods,
fabrics, and paint (44). Several species are plant pathogens
causing various diseases such as leaf spots, scabs, and
fruit rots. In addition, spores of Cladosporium are well
known to be important aeroallergens (35). Aerobiological
surveys including those conducted by the Aeroallergen
Network of the American Academy of Allergy, Asthma, and
Immunology (AAAAI) (46) routinely report the presence of
Cladosporium conidia in the atmosphere. This network
consists of a group of 84 certified air sampling and
reporting stations, primarily in the United States. In the
1999 network report, airborne fungal spore concentrations
were reported from 35 of the network sampling stations.
Cladosporium spores were the most abundant airborne
spore type reported at 23 stations and among the top six Figure 3. Alternaria sp. conidia from air sample.
at all 35 stations.
In many areas, airborne Cladosporium spores can be
detected year-round with the highest levels occurring from Alternaria conidia are passively released from infected
late spring though early fall. Summer peaks have been leaves or other substrates by moderate to strong gusty
found in England (47), Sweden (48), and Colorado (49). winds with minimum velocities of 2 to 3 m/second required.
However, Halwagy (50) showed year-round occurrence Dispersal typically occurs during midday when conditions
of Cladosporium in the atmosphere of Kuwait with are warm and dry with high wind speeds. Rotem (52)
highest levels found in the spring and secondary peaks reported that the greatest dispersal occurs during dry
occurring in the fall. Similar results were shown by weather, which immediately follows periods of rain or
Cosentino and coworkers (51) in Italy. These differences heavy dew. Dry windy periods can eventually deplete
may result from moisture and other climatic growth spore reserves on the leaves and inhibit sporulation,
factors. In the Tulsa, Oklahoma area, Cladosporium which requires moister conditions. Spores can also be
spores are also present in the atmosphere year-round dispersed into the atmosphere when washed or splashed
with the highest concentrations typically occurring from leaf surfaces by raindrops or irrigation. Timmer
in the late summer or fall (46). At times, hourly and coworkers (53) showed large numbers of conidia
concentrations of Cladosporium have exceeded 100,000 were released by simulated rainfall in an environmental
spores/m3 . Atmospheric concentrations of Cladosporium chamber, but they found no correlation between spore
spores typically peak around noon (14). concentration and rain during field studies. By contrast,
Although generally considered a member of the dry air high humidity inhibits the release of spores from
spora (8,47,49), various researchers have also described wet leaves and airborne spores are washed from the
dispersal by mist droplets in this genus. Experimental atmosphere by rain and irrigation (52).
studies with both mist-laden and dry air showed that mist Species of Alternaria are pathogenic to a number of
was no more efficient in dispersing spores than the dry crop plants, causing early blight, black spots, brown spots,
air for Cladosporium herbarum (4). In addition, studies seedling blight, black rots, or leaf spot diseases. They also
have shown transient increases in the concentration after occur as saprobes on a large variety of organic substrates
the start of rainfall (14,15). It has been suggested that including leaf surfaces, stored foods, soil, and textiles (44).
an electrostatic release or the vibration and puffing from Most pathogenic species have a cosmopolitan distribution
the first raindrops may be responsible for this increase (8). including A. solani on potato and tomato, A. brassicae
Others have suggested that washout is responsible for this on members of the Brassicaceae, A. porri on onions, and
increase in concentration (49). Wind gusts associated with A. alternata on a variety of host species (52). Alternaria
the frontal system may also play a role in this increase; pathogens are often secondary invaders attacking plants
however, different species of Cladosporium may respond under stress, especially those stressed by drought, insect
differently to precipitation. infestation, senescence, or even other fungi.
Crop losses caused by Alternaria pathogens tend to be
Alternaria. Alternaria spores are frequently the second less than losses owing to more serious pathogens such as
most abundant component of the dry air spora. Alternaria rusts or downy mildews; however, this genus is prominent
species are anamorphs of ascomycetes including members in the aerobiological literature because it is also recognized
of the genus Pleospora. The genus is characterized by as an important airborne allergen. Although there is no
distinctive large multicellular dictyospores, which are nationwide network of plant pathologists conducting air
beaked and produced in chains (Fig. 3). Alternaria spores sampling for the detection of fungal pathogens, data from
are commonly found in atmospheric concentrations of the Aeroallergen Network show that Alternaria conidia
several hundred to several thousand spores/m3 . In some were among the top 10 spore types (in terms of atmospheric
areas, these spores may be present throughout the concentrations) reported from all 35 stations in 1999. At 19
year; however, peak concentrations usually occur in late stations Alternaria spores were among the top five spore
summer or fall (46). types identified from the atmosphere (46). While some of
410 BIOAEROSOLS IN AGRICULTURAL AND OUTDOOR SETTINGS

the Alternaria conidia may be from saprobic species, these the famines that resulted from this plant disease, two
spores are a significant component of the air spora and million people died of starvation. Southern corn leaf
represent a ubiquitous threat to many crops. blight caused by Cochliobolus heterostrophus (Bipolaris
maydis) is another species in this group capable of causing
Helminthosporium/Drechslera-Type Spores. Several gen- devastating crop loss. The fungus causes small lesions on
era of anamorphic fungi form similar thick-walled, pig- the leaf that may be so abundant they almost cover the
mented, cylindrical conidia with rounded ends and a entire leaf surface (55). An epidemic of one race of this
distinctive attachment scar at the base. The conidia species quickly destroyed approximately 15% of the U.S.
are pseudoseptate but vary in shape with some being corn crop in 1970. The fungus attacked all corn hybrids
long (up to 160 µm) and tapering, while others are rela- with the Texas male sterility gene and the resulting crop
tively short (44) (Fig. 4). For many years, fungi with this loss was estimated at $1 billion.
type of spore were classified as species of Helminthospo- Tan-spot caused by Pyrenophora tritici-repentis (Drech-
rium; however, in the past 20 years the taxonomy in slera) is a serious disease of wheat throughout the Great
this group has undergone revision and several genera are Plains and is considered the most economically important
now recognized (54). The best known genera in this group wheat disease in North Dakota (56). Epidemics are ini-
are now Helminthosporium, Drechslera, Bipolaris, and tiated by both ascospores and conidia that are found on
Exserohilum. Teleomorphs are species of Cochliobolus (for overwintered crop debris. The conidia form the repeating
Bipolaris), Pyrenophora (for Drechslera), and Setosphaeria stage, which can rapidly spread the disease when environ-
(for Exserohilum). The genus Helminthosporium is gener- mental conditions are suitable. Francl (56) showed that
ally reserved for saprobic members of this group, and the greatest numbers of conidia were produced after rains
no teleomorph is recognized. Other genera are delimited as the wheat neared maturity. Conidia develop in the
by the shape of the conidium, morphology of the attach- dark, but peak dispersal occurs in the afternoon after leaf
ment scar, and the number and location of germ tubes surfaces have dried.
produced by the conidium. It should be noted that the Conidia from these fungi are common in the atmo-
separation into these genera is generally acknowledged sphere. They are reported among the top 10 spore types at
by mycologists and plant pathologists, but it is not uni- 12 stations from the Aeroallergen Network in 1999 (46).
versally accepted. Some books and journals still use the Moderate winds disperse the conidia that can travel
Helminthosporium binomials and others use Drechslera through microscale or mesoscale transport (9,56).
for all taxa (54). Curvularia is a closely related genus with a Cochliobo-
This group of fungi is known to cause leaf spots and lus teleomorph. Conidia have three or more transverse
blights and other diseases on wild and cultivated members septa and are typically curved owing to an enlarged central
of the Poaceae; they are especially important pathogens cell (Fig. 5). Like the other genera in this group, Curvu-
on cereal crops (54). Some fungi in this group are able laria is responsible for a number of leaf spot diseases and
to cause sudden, devastating diseases that have resulted is a common component of the dry air spora.
in major famines and great economic loss (55). Although
many other host plants are infected by these fungi, the Smuts. Smut fungi are plant pathogenic basidiomycetes
major losses have been to cereals. In addition to their role in the order Ustilaginales. There are approximately 1,200
as plant pathogens, members of this group are allergenic species of smut fungi classified in 50 genera; however, the
and are the leading cause of fungal sinusitis. In addition, majority of species are classified into two large genera,
they are capable of causing fungal infections in livestock Tilletia and Ustilago (57). About 4,000 species of host
and are also toxigenic (42,54).
plants are attacked by smut fungi including both crops
The great Bengal famine of 1943 was owing to the
and native plants. They are especially serious pathogens
destruction of rice plants by brown spot of rice, which
on cereal crops.
is caused by Cochliobolus miyabeanus (Bipolaris). During

Figure 4. Drechslera/Bipolaris type spore from air sample. Figure 5. Curvularia sp. spore from air sample.
 BIOAEROSOLS IN AGRICULTURAL AND OUTDOOR SETTINGS 411

generally ranged from 100 and 1,000 teliospores/m3 during

both seasons with, the mean daily concentrations at
291 spores/m3 and 356 spores/m3 for 1991 and 1992,
respectively. Peak concentrations were 1,874 spores/m3
on July 5, 1991 and 5,906 spores/m3 occurring on
May 12, 1992. Daily concentrations during the study
period experienced many fluctuations owing to variable
climatic conditions as well as the phenologies of both
the host species and pathogens (60). This last factor
was especially important because it resulted in different
species of smuts occurring at different times of the year.
In May and June, teliospores found in the atmosphere
included those belonging to species, which infect Bermuda
grass (Ustilago cynodontis), Johnson grass (Sphacelotheca
occidentalis), oat (U. kolleri), and wheat (U. tritici).
Figure 6. Ustilago maydis teliospores from infected corn. Smut spores identified during September and October
included U. maydis, which infects corn, and U. brumivora
and U. bullata pathogenic to several native Oklahoma
The major dispersal phase of smut fungi is the asexual grasses (60). Although this study was limited to the period
stage, which produces teliospores (55). Extraordinary of May to October for both years, other data from Tulsa
numbers of teliospores develop and are passively dispersed show that atmospheric smut spores are not limited to this
by wind (Fig. 6). Smut fungi have no repeating stage; host 6-month period (46).
plants produces only one generation of teliospores per In addition to the significance of smut fungi as plant
growing season (55). Spores of various smut species are pathogens, smut spores serve as important aeroallergens
produced during different seasons. Hamilton (47) found as they are abundant in the atmosphere for extended
that airborne spores of Tilletia showed highest levels periods of time (35,61). However, the full extent of
in late August and early September in England, while allergenicity to smut spores and the clinical significance
Ustilago spores peaked from mid-June and late July. She of various smut species is not known (60). Data from the
also showed that the atmospheric concentrations of these Aeroallergen Network showed that smut spores represent
genera depended on several meteorological variables. a significant component of the air spora and are, therefore,
Levels decreased during rain or high humidity and a significant source of aeroallergens as well as a significant
increased during periods of abundant sunshine, strong source of plant pathogen inoculum (46).
winds, and high barometric pressure, typical condition for
release of dry air spora (47). Smut spores typically show a Rusts. Members of the basidiomycete order Urediniales
mid-day peak but here the rhythm seems entirely owing to are commonly called rust fungi. About 6,000 species of rust
environmental conditions. Teliospores of an infected plant fungi have been identified; these fungi attack a wide range
tend to mature all at one time; however, their release of seed plants and can cause destructive plant diseases.
is spread over a long period and appears to be closely Many rust fungi have complex life cycles with several
correlated with wind velocity, which is often higher around spore stages. Probably the most important spore state is
noon (4). the uredial stage, which typically causes reddish (or rust-
Various other studies have also reported the presence colored) lesions on plants and is the source of the common
of smut spores in the atmosphere. Halwagy (50) showed name for these pathogens. Spores produced in these
that Ustilago spores were the second most common spore lesions are known as uredospores (also called urediospores
type identified in the Kuwait atmosphere. Misra (58) noted and urediniospores). Thousands of years before scientists
that smut spores peaked in India during dry conditions knew that fungi could cause plant disease, rust epidemics
in December and January, while Rubulis (59) reported were recognized and the reddish lesions noted on plants.
peaks during the late spring and fall in Sweden. In The epidemics were described in the writings of Aristotle
both of these studies, smuts were basically considered and Theophrastus who also realized that different plants
as one spore category with no attempt to differentiate the varied in the susceptibility to these diseases. The rust
species involved. Likewise, the data from the Aeroallergen fungi are responsible for some of the most important
Network from the AAAAI also report smut spores as a diseases of cereal crops, and it is estimated that 10% of
single category. In their 1999 report (46), smut spores the world grain crop is lost to rust fungi each year (55). In
were among the top 10 spore types, in terms of atmospheric addition to the cereals, rust fungi have caused devastation
concentration, reported from 32 out of 35 sampling stations to coffee, apple, pine trees, and other crops.
that provided counts for fungal spores. Because of the economic importance of these fungi, the
Crotzer and Levetin (60) examined the atmospheric long-distance transport of rust spores, especially wheat
levels of smut spores in Tulsa, Okla., during 1991 and rust, has been studied more extensively than that of
1992 and attempted to identify the most abundant smut any other fungal pathogen (34). Worldwide, rust fungi
spores. The authors found that smut spores were present are the most important and damaging diseases of wheat,
in the atmosphere every day from May through October with stem rust (P. graminis f.sp. tritici) and leaf rust (P.
in both 1991 and 1992. The average daily concentration recondita f.sp. tritici) the most prevalent types. In North
412 BIOAEROSOLS IN AGRICULTURAL AND OUTDOOR SETTINGS

America, leaf rust is currently the more important species, of kilometers (10,34). In southern Texas and Mexico where
but historically major epidemics of stem rust have been the climate is mild, the uredial stage can continue all
more damaging. In recent decades, the widespread use of winter and give rise to spring infections in northern
resistant wheat varieties has prevented major epidemics of states when the uredospores are carried by prevailing
stem rust. Nevertheless, approximately one million metric southerly winds. Uredospores can also be transferred
tons of wheat are lost to stem rust in North America each back to southern areas in late summer and fall. Initially
year, and in other wheat growing areas the losses are far demonstrated in 1923 by Stakman, the movement of
greater (55). rust uredospores along the ‘‘Puccinia Pathway’’ in North
Puccinia graminis f.sp. tritici, the fungus responsible America is one of the best known examples of long-distance
for stem rust of wheat has a complex life cycle that involves dispersal (64). In addition to P. graminis and P. recondita,
five spore stages on two separate host plants, wheat other fungal pathogens and the insect vectors of viral
and barberry (55,62). This fungus typically overwinters diseases are all linked to dispersal along this pathway (34).
as teliospores that give rise to basidiospores in the spring. In a review, Nagarajan and Singh (34) also described the
Basidiospores are capable of infecting young barberry long-distance transport of Puccinia uredospores in other
leaves and giving rise to spermatia and receptive hyphae parts of the world including parts of Europe, India, and
and later to aeciospores, also on barberry. Aeciospores Australia. They also provided evidence for the transoceanic
become airborne and can transfer the infection to wheat transport of uredospores between Australia and regions in
on which the spores germinate, and hyphae enter the southern Africa.
plant through stomata. Once established in wheat, the In addition to suitable wind patterns, successful
fungus produces uredia within 2 weeks; these appear on long-distance dispersal of rust spores is also contin-
the stem as long narrow pustules, which produce dark gent on source strength and viability. Nagarajan and
red powdery masses of uredospores. Large numbers of Singh (34) estimated that a field of wheat moder-
spores are produced during the uredial stage (Fig. 7). ately infected with stem rust would produce 4 × 1012
It is estimated that a mature uredium is capable of uredospores/day/hectare (34). They suggested that ure-
producing about 10,000 uredospores per day over a period dospores must reach a suitable host within 5 days after
of several weeks and an even mildly infected plant discharge to remain infectious. However, Eversmeyer and
with 50 uredia would produce 500,000 spores/day (63). Kramer (65,66) examined the survival of P. graminis and
Uredospores become readily airborne and infect new wheat P. recondita uredospores under a variety of temperature
plants, thereby forming the repeating stage of this fungus. conditions in the field and in environmental chambers.
Near the end of the growing season, the uredia turn In field experiments, they found that during freezing or
black as two-celled overwintering teliospores replace the subfreezing temperatures spore viability quickly declined;
uredospores (55,62). Wheat plants infected by P. graminis but during the warmer temperatures in spring, 10 to 20%
are severely weakened, but not destroyed, and the grain of the spores remained viable after 5 days. In addition, 1%
yield is significantly reduced (55). of the spores remained viable for 19 days. In their chamber
Efforts to eradicate barberry plants from wheat growing studies, spores exposed to constant temperatures between
areas have reduced the risk of infection from stem rust 10 and 30 ° C remained viable for up to 36 days (66). While
of wheat. Also, the alternate host for leaf rust, meadow environmental conditions during transport may be severe,
rue apparently does not occur in North America. However, these studies suggest that at least a small percent of the
for both pathogens the alternate host is less important enormous number of spores may survive long enough to
than uredospores as a source of inoculum. Uredospores reach a suitable host.
are easily carried by wind from one plant to another
giving rise to rust epidemics. In fact, uredospores can be Other Members of the Dry Air Spora. Other fungal spores
transported by prevailing winds for hundreds or thousands are also commonly found in the dry air spora. These include
conidia of Epicoccum, Pithomyces, Nigrospora, Botrytis,
Torula, Stemphylium, Oidium, Aspergillus, Penicillium,
and many others. Although many of these genera
are widespread saprobes, others are well known plant
pathogens. The genus Oidium represents the anamorphic
state of the powdery mildews, which are important
pathogens on many crops, especially the cereals (55).
Botrytis diseases are possibly the most common diseases
of vegetable crops, responsible for gray mold and rots of
fruits and vegetables in the field as well as in storage (55).
Epicoccum is a cosmopolitan genus of asexual fungi that
occurs as saprobes on a wide variety of substrates, but it
is also a secondary invader on many types of plants and is
frequently found on leaf spots (44).

Moist Air Spora

Figure 7. Puccinia graminis uredium with uredospores on In damp or wet weather, the atmosphere is typically
infected wheat. dominated by ascospores and basidiospores. As described
 BIOAEROSOLS IN AGRICULTURAL AND OUTDOOR SETTINGS 413

earlier, both spore types are actively released from the ascospores were found in orchards from mid-April to
fruiting body through mechanisms that require moisture. early-June (71). Once the primary infection is established,
Leptosphaeria ascospores are often abundant along with conidia of the anamorphic stage (Spilocaea pomi) begin
Venturia, Didymella, and spores from members of the developing leading to further disease spread. The more
family Diatrypaceae (39,67). Many ascospores found in primary inoculum present, the more rapidly the disease
the air spora are produced by important agricultural will build up and the more serious the epidemic will
pathogens. Basidiospores that are typical components of ultimately become. Several studies have recently focused
the atmosphere include those from saprobic species of on modeling the aerial dispersal of Venturia ascospores
mushrooms, bracket fungi, and puffballs (38) although to help evaluate the risk of infection and optimize
spores from a number of tree pathogens can also occur the application of fungicides to the periods when it is
in the atmosphere (30). In addition to ascospores and absolutely necessary (43,68,70,72).
basidiospores, asexual conidia of Fusarium, Cercospora,
and Colletotrichum are also abundant in moist weather. Fusarium. Fusarium species are well known soil fungi
This section will briefly consider two representatives of with several species responsible for a variety of plant
the moist air spora. For a fuller review of this topic, see diseases. Vascular wilts and other diseases caused by
Fitt and coworkers (16) and Huber and coworkers (17). Fusarium oxysporum have been recognized for over 100
years. Many subspecies of F. oxysporum attack a wide
Apple Scab. The ascomycete, V. inaequalis Scab is variety of crop plants. Although typically considered
responsible for apple scab, the most important disease a soilborne pathogen, there is considerable evidence
of apples with all commercial varieties susceptible to of airborne dispersal of conidia in many races and
infection (55). This pathogen occurs worldwide wherever subspecies. Katan and coworkers (73) described the
apple trees are cultivated; similar scab diseases also airborne collection of viable F. oxysporum f.sp. lycopersici
affect pear and hawthorn. The principal symptoms are conidia in greenhouses from diseased tomato plants.
the development of scab lesions on both leaves and fruit; They also obtained similar results with a variety of
however, the disease can also cause premature defoliation, other subspecies on other vegetables. Splash dispersal
premature fruit drop, and poor bud development for the of Fusarium macroconidia has been generally accepted as
following year (55). Control of apple scab requires repeated the major vehicle of disease spread (8,74) (Fig. 9).
applications of fungicides (8 to 20 times per season) to Fusarium head blight (FHB) of wheat and other small
protect the crop. Without fungicidal protection, 70 to 100% grains is caused by Fusarium graminearum. FHB, also
of the crop would be unsalable. known as scab, is increasing throughout the world and
In the spring, the primary inoculum consists of during the past decade there were several epidemics on
ascospores that develop in overwintered fruiting bodies. wheat in the United States and Canada (74). Fusarium
Immature fruiting bodies are found in dead leaves that graminearum regularly forms its sexual stage, Gibberella
overwinter on the ground in the orchard. They mature zeae, in nature, and both conidia and ascospores are
in the spring, and ascospores are actively discharged involved in disease spread. Ascospores are formed in
from these ascocarps during rain (16) (Fig. 8). Venturia perithecia and are forcibly shot into the air, while
ascospores are a major component of the air spora macroconidia are formed on sporodochia and are splash
in orchards during the spring and are carried by dispersed. Both stages occur on over-wintered crop debris
wind to young leaves in which they cause a primary and both require rainfall for dispersal. Spores infect the
infection (68–70). Ponti and Cavanni (70) showed that anthers on wheat inflorescences and the fungus spreads
the ascospores could typically be found in the air of in the developing grains. Symptoms are evident in two
Italian apple orchards during periods of rainfall from mid- to three weeks and include the premature senescence of
March to mid-June. In the northeastern United States, spikelets (74). The fungus survives on the wheat residue

Figure 8. Venturia sp. ascospores from air sample. Figure 9. Fusarium sp. macroconidia from culture.
414 BIOAEROSOLS IN AGRICULTURAL AND OUTDOOR SETTINGS

and gives rise to new infections the following spring. metalaxyl, the most widely used fungicide to control late
Fusarium graminearum also can exist as a saprobe blight (77,78).
on other residue especially corn stalks. The increased Long-distance dispersal of P. infestans has been well
cultivation of corn in wheat growing areas and reduced documented by two different modes of transport (75).
tillage agricultural systems are both possible reasons for Intercontinental migration has been associated with the
the recent epidemics. transport of infected plants or tubers by humans. This
occurred in the 1840s and again before the 1980 outbreak
Other Pathogens of the A2 mating type. Dispersal over tens of kilometers is
attributed to airborne sporangia, which are able to survive
The spore types described earlier are common components for hours at the reduced humidity encountered during
of the air spora and are often present in the atmosphere transport (75). The rapid progress of blight epidemics in
year-round. However, spores of other plant pathogens are the 1840s and in the 1990s illustrates the effectiveness of
not normally seen in the air spora unless the disease is airborne dispersal.
epidemic in the area. Two diseases of this type are late
blight of potato and tobacco blue mold. Blue Mold. Peronospora tabacina (syn P. hyoscyami
f.sp. tabacina) is an oomycete that is responsible for
Late Blight of Potato. Phytophthora infestans is the tobacco blue mold. This disease was first described in
pathogen responsible for late blight of potato (and also Australia during the nineteenth century, and the fungus
tomato). This oomycete occurs wherever these crops are was identified by Baily in 1890 (79). The first serious
grown, and all cultivars are susceptible. Late blight of epidemic in North America occurred in 1979. Prior to this
potato has been the most important disease of potato time the disease was confined to seedbeds. In 1979 and
since the 1840s, when it caused the destruction of the again in 1980, the infection rates were especially severe,
potato crop in Ireland and the resulting widespread with the epidemic advancing northward at rates of 10
famine. Phytophthora infestans can invade all parts of to 32 km/day in the eastern United States to southern
the plant resulting in the rapid death of infected parts. Canada and with crop losses estimated at $350 million.
Productivity of the potato plants is quickly reduced and Blue mold is able to attack both wild and cultivated
tuber destruction frequently occurs as well. Sporangia species of tobacco. Host plants and pathogen exist year-
are produced on aerial hyphae, which grow out of the round in tropical areas such as the Mediterranean and
stomata and are dispersed by wind, possibly carried for Caribbean basins. However, in temperate regions, tobacco
tens of kilometers (75). At low temperatures (10 to 15 ° C) is grown as an annual. Also, P. tabacina is not able to
and high humidity, sporangia germinate by producing overwinter in temperate zones. As a result, reappearance
numerous zoospores, while at higher temperatures each of the pathogen is owing to either contaminated seedlings
sporangium gives rise to a single germ tube that develops or the long-distance transport of inocula from tropical
into hyphae. As a result, during cool, wet periods, the regions.
production of zoospores leads to an amazingly rapid The asexual propagules of P. tabacina are the
spread of the disease. Without fungicidal protection, a sporangia. Infection can begin within four hours after
blighted field can be destroyed within a couple of days. a sporangium lands on the leaf surface. A symptom-free
Similar disease progression also occurs in infected tomato incubation period, typically five to seven days, ends with
fields (70). Forecasting models have been used for over 40 the appearance of yellow lesions and the development
years to help optimize the timing of fungicide application of new sporangia. Unlike many other oomycetes, no
to periods when weather conditions promote the disease zoospores are produced by Peronospora and the sporangia
spread (76). themselves (often referred to as sporangiospores or
Generally, Phytophthora infestans overwinters as conidia) are the only asexual stage. As described earlier,
mycelium in infected tubers. In countries where both sporangia are released from the sporangiophore by the
mating types have been identified, it can also overwinter twisting movements of the sporangiophore that occurs
as oospores in the soil. In the 1840s, when the disease during the drying of the leaf in the morning as humidity
first appeared in Europe and the United States, only one decreases and temperature increases (11,80). Because the
mating type (A1 ) of the fungus was present. The second sporangia are subject to desiccation during transport,
mating type (A2 ) was identified in 1950 in central Mexico cool, wet, overcast weather, favors the rapid advance of
and was confined to Mexico until 1980 when outbreaks the fungus, while clear, hot, dry weather halts disease
began occurring in Europe. It has been suggested that the spread (80).
A2 mating type was introduced in a shipment of potatoes Each spring in the eastern United States, weather
from Mexico to Europe in the late 1970s. In the past 20 conditions are favorable for the northward transport of
years, the A2 type has spread throughout the world (77,78). Peronospora sporangia with two possible pathways of
This has raised anxieties about the possibility of sexual disease spread (80). Along one pathway, the pathogen can
reproduction occurring in the field leading to new strains spread northward from Cuba into Florida and Georgia.
developing at a faster rate. The implications for potato In the other pathway, the source of inoculum is on
breeders searching for blight resistance are obvious. These wild Nicotiana repanda south central Texas or from
concerns have been strengthened by the recent outbreaks cultivated tobacco in Mexico. From here the disease
of late blight in the United States and Canada. The spreads to Kentucky and then east to North Carolina.
strains of P. infestans involved were nearly all resistant to These pathways show that long-distant transport of
 BIOAEROSOLS IN AGRICULTURAL AND OUTDOOR SETTINGS 415

sporangia can lead to blue mold outbreaks, and suggest BIBLIOGRAPHY

that forecasting systems can potentially provide time for
tobacco farmers to apply fungicides. This is the aim of 1. J. Lacey and J. Venette, in C. S. Cox and C. M. Wathes,
the North American Plant Disease Forecast Center from eds., Bioaerosols Handbook, Lewis, Boca Raton, Fla., 1995,
the Department of Plant Pathology at North Carolina pp. 407–471.
State University (81). Since 1995, this center has used 2. J. Lacey, in D. L. Hawksworth, ed., Frontiers in Mycology,
the HYSPLIT trajectory model to predict outbreaks of CAB International, Kew, UK, 1991, pp. 157–186.
blue mold in North America. Documented sources of blue 3. A. H. R. Buller, Researches on Fungi, vol. 2, Longmans, Green
mold are used in the model along with meteorological data & Co., London, 1922.
from NOAA’s ETA or AVN Models. This modeling system 4. C. T. Ingold, Fungal Spores: Their Liberation and Dispersal,
predicts the short-term weather and plots the trajectories Clarendon Press, Oxford, 1971.
of inoculum-laden parcels of air. The daily forecast 5. P. H. Gregory, The Microbiology of the Atmosphere, 2nd ed.,
Halstead Press, New York, 1973.
produced with HYSPLIT trajectories describes weather
conditions at the source and along the anticipated pathway 6. J. Lacey, Trans. Br. Mycol. Soc. 64, 265–281 (1975).
with emphasis given to those atmospheric conditions that 7. T. M. Madelin and M. F. Madelin, in C. S. Cox and
favor sporulation at the source, survival during transport, C. M. Wathes, eds., Bioaerosols Handbook, Lewis, Boca
Raton, Fla., 1995, pp. 361–368.
and deposition. An overall outlook describes the likelihood
of blue mold spread over the subsequent 48 hours. The 8. J. Lacey, Mycol. Res. 100, 641–660 (1996).
forecasters anticipate that the use of this model will 9. H. A. McCartney and B. D. L. Fitt, in D. G. Jones, ed., The
continue to provide valuable information in the efforts Epidemiology of Plant Disease, Kluwer Academic, Dordrecht,
The Netherlands, 1998, pp. 138–160.
to control blue mold epidemics (81). A recently developed
numerical model encompasses the release, transport, and 10. B. Hau and C. de Vallavielle-Pope, in D. G. Jones, ed., The
Epidemiology of Plant Disease, Kluwer Academic, Dordrecht,
deposition of blue mold sporangia (82). Since 1996, the
The Netherlands, 1998, pp. 323–347.
daily blue mold forecasts have been available on the
11. D. E. Aylor, Ann. Rev. Phytopathol. 28, 73–92 (1990).
Internet during the growing season (81) to provide on-
line forecasts for tobacco blue mold outbreaks throughout 12. D. E. Pedgley, in J. H. Andrews and S. S. Hirano, eds.,
Microbial Ecology of Leaves, Springer, New York, 1991,
continental North America. In 1998, the center was
pp. 43–59.
expanded to also include forecasts for downy mildew of
13. H. A. Burge, M. M. Muilenberg, and J. A. Chapman, in
cucurbits.
J. H. Andrews and S. S. Hirano, eds., Microbial Ecology of
Leaves, Springer, New York, 1991, pp. 222–236.
CONCLUSION 14. J. M. Hirst, Trans. Br. Mycol. Soc. 36, 375–393 (1953).
15. K. M. Venables et al., Clin. Exp. Allergy 27, 725–736 (1997).
A tremendous variety of fungal spores are present in the 16. B. D. L. Fitt, H. A. McCartney, and P. J. Walklate, Annu.
atmosphere. While many of the spores originate from Rev. Phytopathol. 27, 241–270 (1989).
saprobic species, spores of plant pathogens are often 17. L. Huber, L. V. Madden, and B. D. L. Fitt, in D. G. Jones,
a significant component of the air spora. These spores ed., The Epidemiology of Plant Disease, Kluwer Academic,
pose a constant threat to crops throughout the world. Dordrecht, The Netherlands, 1998, pp. 348–370.
Control of plant diseases often requires extensive use 18. A. Pielaat, L. V. Madden, and G. Gort, Phytopathology 88,
of fungicides. To optimize the application of fungicides, 1131–1140 (1998).
timing of the applications should be tied to periods of peak 19. A. H. Rajasab and H. T. Chawda, Grana 33, 162–165 (1994).
spore dispersal. This requires a thorough understanding 20. J. W. Deacon, Modern Mycology, 3rd ed., Blackwell Science,
of the aerobiology of the pathogen. Aerobiological studies Oxford, UK, 1997.
should be part of any effort to understand the distribution 21. E. Moore-Landecker, Fundamentals of the Fungi, 3rd ed.,
and epidemiology of agricultural pathogens that rely on Prentice-Hall, Englewood Cliffs, N.J., 1990.
air currents for dispersal. Unfortunately, no regional 22. M. J. Richardson, Mycol. Res. 100, 213–216 (1996).
or national sampling network exists for agricultural 23. E. Levetin, C. Rogers, C. Beasley, and M. Sterling, 13th
pathogens. The Aeroallergen Network of the AAAAI Conference on Biometeorology and Aerobiology, American
maintains a national network of air sampling stations, Meteorological Society, Boston, 1998, pp. 335–336.
and a similar network is needed for the plant pathology 24. D. S. Meredith, Annu. Rev. Phytopathol. 11, 313–342 (1973).
community. Studies during the last half century have 25. T. R. Gottwald, T. M. Trocine, and L. W. Timmer, Phy-
provided an understanding of the dispersal mechanisms topathology 87, 1078–1084 (1997).
used by fungi; however, research is still needed on 26. N. P. Money, Mycologia 90, 547–558 (1998).
many aspects of fungal biology to fully understand the 27. J. Webster, R. A. Davey, and J. C. R. Turner, Mycol. Res. 93,
effects of meteorological conditions on dispersal, transport, 297–302 (1989).
and survival in the atmosphere. This knowledge will 28. R. T. Haard and C. L. Kramer, Mycologia 62, 1145–1169
help in the development of better computer models for (1970).
forecasting disease epidemics. These models are in place 29. F. I. McCracken, Can. J. Bot. 50, 2111–2115 (1972).
for forecasting of blue mold and downy mildew of cucurbits 30. R. Craig and E. Levetin, Aerobiologia 16, 75–81 (2000).
from the North American Plant Disease Forecasting 31. C. T. Ingold and H. J. Hudson, The Biology of Fungi, 6th ed.,
Center. More systems of this type are needed. Chapman & Hall, London, 1993.
416 BIOAEROSOLS IN INDUSTRIAL SETTINGS

32. C. S. Cox, The Aerobiological Pathway of Microorganisms, 66. M. G. Eversmeyer and C. L. Kramer, Phytopathology 85,
John Wiley & Sons, New York, 1987. 161–164 (1995).
33. S. M. Pady, B. Peturson, and G. J. Gree, Phytopathology 40, 67. S. M. Hasnain, Grana 32, 315–318 (1993).
632–641 (1950). 68. D. E. Aylor, Plant Dis. 82, 838–849 (1998).
34. S. Nagarajan and D. V. Singh, Annu. Rev. Phytopathol. 28, 69. B. D. L. Fitt and H. A. McCartney, in K. J. Leonard and
139–153 (1990). W. E. Fry, eds., Plant Disease Epidemiology, vol. I, Macmil-
35. H. A. Burge, Clin. Rev. Allergy 3, 319–329 (1985). lan, New York, 1986, pp. 311–345.
36. E. Levetin, in H. Burge, ed., Bioaerosols, Lewis, Boca Raton, 70. I. Ponti and P. Cavanni, Aerobiologia 8, 94–101 (1992).
Fla., 1995, pp. 87–120. 71. D. E. Aylor, Phytopathology 86, 769–776 (1996).
37. P. H. Gregory and J. M. Hirst, Nature 170, 414 (1952). 72. D. M. Gadoury, A. Stensvand, and R. C. Seem, Phytopathol-
38. E. Levetin, Grana 30, 123–128 (1991). ogy 88, 902–909 (1998).
39. U. Allitt, Trans. Br. Mycol. Soc. 87, 147–154 (1986). 73. T. Katan, E. Shlevin, and J. Katan, Phytopathology 87,
40. J. Lacey and B. Crook, Ann. Occup. Hyg. 32, 515–533 (1988). 712–719 (1997).
41. A.-L. Pasenen et al., J. Am. Ind. Hyg. Assoc. 50, 631–633 74. T. C. Paulitz et al., Phytopathology 89, 74–83 (1999).
(1989). 75. W. E. Fry et al., Ann. Rev. Phytopathol. 30, 107–129 (1992).
42. D. H. Larone, Medically Important Fungi: A Guide to 76. D. R. MacKenzie et al., in D. C. Erwin, S. Bartnicki-Garcia,
Identification, 3rd ed., ASM Press, Washington, D.C., 1995. and P. H. Tsao, eds., Phytophthora: its Biology, Taxonomy,
43. A. L. Jones, in D. G. Jones, ed., The Epidemiology of Plant Ecology, and Pathology, APS Press, St. Paul, Minn., 1983,
Disease, Kluwer Academic, Dordrecht, The Netherlands, pp. 303–313.
1998, pp. 389–403. 77. W. E. Fry, Plant Dis. 81, 1349–1357 (1997).
44. M. B. Ellis, Dematiaceous Hyphomycetes, CAB International, 78. W. E. Fry and E. S. Mizubuti, in D. G. Jones, ed., The
Oxon, UK, 1971. Epidemiology of Plant Disease, Kluwer Academic, Dordrecht,
45. M. H.-M. Ho, R. F. Castaneda, F. M. Dugan, and S. C. Jong, The Netherlands, 1998, pp. 371–388.
Mycotaxon 72, 115–158 (1999). 79. G. I. Johnson, in W. E. McKeen, ed., Blue Mold of Tobacco,
46. AAAAI (American Academy of Allergy, Asthma, and APS Press, St. Paul, Minn., 1989, pp. 1–18.
Immunology), 1999 Pollen and Spore Report, American 80. C. E. Main and J. M. Davis, in W. E. McKeen, ed., Blue Mold
Academy of Allergy, Asthma, and Immunology, Milwaukee, of Tobacco, APS Press, St. Paul, Minn., 1989, pp. 201–216.
Wisc., 2000.
81. C. E. Main, T. Keever, and J. M. Davis, North American
47. E. D. Hamilton, Acta Allergol. 13, 143–175 (1959). Plant Disease Forecasting Center, 2000 (http://www.ces.
48. M. Hjelmroos, Grana 32, 40–47 (1993). ncsu.edu/depts/pp/bluemold/).
49. R. K. Katial, Y. Zhang, R. H. Jones, and P. D. Dyer, Int. J. 82. C. Yao, S. P. Arya, J. Davis, and C. E. Main, Atmos. Environ.
Biometeorol. 41, 17–22 (1997). 31, 1709–1714 (1997).
50. M. Halwagy, Mycol. Res. 93, 208–213 (1989).
51. S. Cosentino, P. L. Pisano, M. E. Fadda, and F. Palmas, Ann.
Allergy, 65, 393–400 (1990).
52. J. Rotem, The Genus Alternaria, APS Press, St. Paul, Minn.,
BIOAEROSOLS IN INDUSTRIAL SETTINGS
1994.
53. L. W. Timmer et al., Phytopathology 88, 1218–1223 (1998).
KENNETH F. MARTINEZ
DOUGLAS TROUT
54. J. L. Alcorn, Annu. Rev. Phytopathol. 26, 37–56 (1988).
RICHARD KANWAL
55. G. N. Agrios, Plant Pathology, 4th ed., Academic Press, San
NIOSH
Diego, 1997.
Cincinnati, Ohio
56. L. J. Francl, Can. J Plant Pathol. 19, 247–255 (1997).
57. K. Vanky, Illustrated Genera of Smut Fungi, Gustav Fischer
Verlag, Stuttgart, Germany, 1987. Microorganisms are normal inhabitants of the environ-
58. R. P. Misra, Perspect. Mycol. Res. 1, 243–252 (1987).
ment; they exist in the soil on which we walk, in the air
that we breathe, in the water that we drink, and even in the
59. J. Rubulis, in S. Nilsson and B. Raj, eds., Nordic Aerobiology,
1984, pp. 85–93. food that we eat. The saprophytic varieties (those utilizing
the organic products of other organisms, living or dead, as
60. V. Crotzer and E. Levetin, Aerobiologia 12, 177–184 (1996).
a food source) generally predominate. Under appropriate
61. J. Santilli, W. J. Rockwell, and R. P. Collins, Ann. Allergy 55,
469–471 (1985).
conditions (optimum temperature, pH, and with sufficient
moisture and available nutrients), saprophytic popula-
62. L. J. Littlefield, Biology of the Plant Rusts: An Introduction,
Iowa State University Press, Ames, Iowa, 1981.
tions can be amplified. In the air, they can exist as discrete
particles (e.g., as individual fungal spores) or they can be
63. A. P. Roelfs, in A. P. Roelfs and W. R. Bushnell, eds., The
Cereal Rusts, Volume II — Diseases, Distribution, Epidemi- attached to other particles or substrates in a medium
ology, and Control, Academic Press, Orlando, Fla., 1985, (solid, liquid, or gas). Many microbial species have been
pp. 3–37. documented to tolerate various environmental extremes
64. D. E. Pedgley, in K. J. Leonard and W. E. Fry, eds., Plant of temperature and pressure, allowing them to acclimate
Disease Epidemiology, vol. I, Macmillan, New York, 1986, to the most extreme conditions. Bacteria have been iden-
pp. 346–365. tified in the lower depths of the ocean, growing amidst
65. M. G. Eversmeyer and C. L. Kramer, Phytopathology 84, intense volcanic activity. Mycobacteria have been found
332–335 (1994). in mummified tissues. It is clear that microorganisms
 BIOAEROSOLS IN INDUSTRIAL SETTINGS 417

(fungi, bacteria, and viruses) are ubiquitous occupants the focus of this article, biotechnology will be presented
throughout many types of diverse environments. in terms of the production of goods using industrial
Many species of microorganisms are beneficial to the processes, specifically, fermentation technology. Aseptic
environment and to our existence. Microorganisms have fermentation techniques have been employed to produce
been used for centuries to produce alcoholic beverages and a variety of commodities on a large scale, including
to leaven bread. More recently, the metabolic systems of single-cell proteins, antibiotics, acetic acid, citric acid,
specific microbial populations have been used to produce lactic acid, amino acids, vitamins, and enzymes. During
pharmacologically important products (e.g., penicillin and industrial fermentation, microbial inocula (fungi or
citric acid) and a limited number of other chemical agents. bacteria) are used to convert a substrate to a desired
Unique uses of the microbial community include the product typically using a batch processing system in
application of bacterial species in the decontamination of three main stages: laboratory and inoculation, (or
hazardous substances (i.e., oil spills in the ocean) and as an microbial preparation and growth), fermentation (or
insecticide (i.e., Bacillus thuringiensis) to control cabbage product biosynthesis), and process recovery (or product
worm, cotton boll worm, and chicken louse. Our everyday extraction and purification).
encounters with the microbial community generally do During the laboratory and inoculation stage, cultures
not pose a significant health risk to the immunocompetent of microorganisms are developed, prepared, and grown
population (healthy adults). However, select occupational prior to the transfer to large-scale fermentation. All
environments present unique exposure concerns because pertinent microbiological operations within the laboratory
of the nature of the microorganisms encountered, the are conducted using sterile equipment with aseptic
microbial concentrations observed, and the susceptibility transfer techniques. To reduce the risk of contamination
of the exposed population. Within the health care industry, with foreign microbial strains, successive recultivations of
attention has been focused on human infections or cultures and numerous propagation steps are held to a
infectious agents, of which the obligate parasites and minimum.
facultative saprophytes (including the primary pathogens During the fermentation stage, multiple propagation
and opportunistic pathogens) are primary concerns (1). steps are again held to a minimum to reduce the risk of
Exposure to bioaerosols (defined here as microorgan- contaminating large quantities of culture media and to
isms and/or their products entrained in/on airborne parti- optimize the use of process equipment. The seed tank,
cles) has been a well recognized safety and health issue in containing a sterile nutrient medium, is inoculated with
agricultural settings, and more recently, in nonindustrial the selected microbial culture prepared in the laboratory.
indoor environments. However, bioaerosols are also impor- The seed tank is designed to promote the growth of the
tant health and safety issues in nonagricultural industrial microorganism population to a level necessary for proper
work settings, the topic of this article. Bioaerosols include inoculation of the fermentor tank. The batch mixture
microorganisms (both alive and dead) and their fragments in the seed tank is aerated and mechanically agitated
or products. Among the specific components of bioaerosols until the optimum level of culture growth is achieved.
that have been studied include microorganisms (such On completion of the cycle, the contents of the seed
as bacteria, fungi, and viruses), endotoxins, and [1 → 3]- tank are aseptically transferred to a larger seed tank
Beta-D-glucans. Endotoxins are lipopolysaccharide (LPS) or, as is generally the case, directly into the fermentor
compounds, which are part of the outer cell wall of all tank. The fermentor tank is where ‘‘fermentation’’ occurs
gram-negative bacteria. [1 → 3]-Beta-D-glucans are glu- and the product of interest is biologically synthesized.
cose polymers, the structural components of most fungal A submerged, batch fermentation process is typically
cell walls. employed using a deep-tank reactor vessel with a
Occupational exposure limits have not been established top-mounted agitator and bottom air sparger. Proper
for bioaerosols or for any of the specific components of temperature conditions are maintained with cooling coils
bioaerosols (2), although several reports have proposed inside (or a cooling jacket outside) the reactor vessel.
guidelines for occupational exposure limits to endotox- The fermentor tank, containing the presterilized nutrient
ins (3–5). One factor that limits our ability to establish medium from a batching tank and the inoculant microbial
guidelines for exposure to endotoxins (and for bioaerosols culture broth mixture from the seed tank, is aerated and
in general) has been laboratory variation in analysis; this mechanically agitated for continued microbial growth and
issue has recently been addressed for endotoxins in a biologic synthesis of the product.
collaborative study that found that both intra- and inter- The mechanical equipment components used in the
laboratory variation remain important problems (6). process recover stage are largely determined by the
We will now review several industries in which product and the microorganism used in the biosynthesis.
exposure to bioaerosols is an important safety and health
Products that are produced extracellularly are extracted
issue.
via solid/liquid separation techniques, which might
employ the application of simple unit operations such
BIOTECHNOLOGY as centrifugation and filtration. In some instances,
concentration and purification (polishing) may be required
Biotechnology has been broadly defined as any technique and can be achieved with such processes as ultrafiltration,
that uses living organisms (or parts thereof) to make vacuum evaporation, or the precipitation of proteins. The
or modify products, to improve plants and animals, or separation of intracellular products is more complicated
to develop microorganisms for specific uses (7). Within because the cells must first be disrupted. These processes
418 BIOAEROSOLS IN INDUSTRIAL SETTINGS

generate bioaerosols, which can be released to the 10,000

surrounding area. Plant 1

Bacterial concentration (CFU/m3)

The etiologic agents of disease may be a contaminant Plant 2
that has found a suitable environment to proliferate Plant 3
1,000
(as in agricultural environments) or may be an active
component of the manufacturing process. Topping and
coworkers (8) documented sensitization among workers
in a biotechnology plant producing citric acid using 100
Aspergillus niger. The results of this study indicated
that the risk was from exposures to not only fungal
spores but also proteinaceous products found in the 10
culture fluid. A follow-up longitudinal survey study of
the same workforce indicated no new skin sensitizations
believed to be the result of the implementation of control
1
strategies, that is, process enclosure, the application
Back Fermentor Seed Sample Separation
of local exhaust ventilation, the use of respiratory ground agitator agitator port process
protection, and worker/management education (9). In
Figure 1. Sample locations reflect worker exposures to process
the pharmaceutical industry, Lagier and coworkers
microorganisms.
identified a case of occupational asthma in a worker
exposed to penicillamine (10). The primary occupational
biological hazards in the biotechnology industry is
the potential for process microorganism and their results indicated that controls are most needed around
metabolic products to produce an immunologic response high-energy operations, including separation equipment
in susceptible individuals (11). Intermediate processing (i.e., filters and centrifuges), fermentor agitator shafts,
chemicals used during the manufacture of specific products and manual sampling ports. These operations may not
of biotechnology can also pose occupational exposure be amenable to complete sealing, enclosure, or isolation.
hazards (e.g., amyl acetate used to extract penicillin). Significantly lower bacterial concentrations were observed
It has also been suggested that although infections at a rotary vacuum drum filter compared with concentra-
from microbial production systems based on the use of tions at the filter press (1). These differences appeared to
recombinant DNA technologies are plausible, the risk is have resulted from the application of local exhaust ventila-
considered extremely low (12). tion, the inherently better containment characteristics of
The effective containment of the potential hazards in drum filters, and operator work practices (dislodging filter
the biotechnology industry is dependent on the equipment cake from the filter press plates at the end of each cycle).
designs employed in existing chemical process technology. Rotary vacuum drum filters have been reported to be the
The equipment designs must provide for containment of most widely used filters in the fermentation industry (15).
the microorganisms (viable and nonviable forms), biolog- In contrast, a centrifuge would be expected to produce
ically active products or intermediates, and processing large concentrations of microbial aerosols. However, as
chemicals such as extraction solvents. The level of con- reported in the enzyme manufacturing study, effective
tainment is determined by the anticipated risks associated process enclosure and the application of local exhaust
with these agents. Specific factors affecting containment ventilation at the biomass discharge point resulted in bac-
include the selection of appropriate fermentor and asso- terial concentrations significantly below those of the filter
ciated equipment, suitable exhaust gas treatment, design process (although above those of the rotary vacuum drum
considerations for vessel overpressurization and relief, filter).
suitable inoculation and sampling systems, and the col- Alternative methods of solids removal include pre-
lection and inactivation of condensate that may contain cipitation, coagulation, floculation and chromatography,
viable microorganisms (13). Consideration to these factors electrophoresis, and ultracentrifugation. These methods
are categorized as primary physical containment, which should be applied only after consideration of the agent risk
offers protection from the microbial agent to the process and the inherent containment abilities of the technology.
operator through isolation. Protection of the general envi- Rapid advancements in the biotechnology industry and
ronment is achieved through the application of secondary use of microorganisms that pose significantly increased
physical containment, which includes consideration of the risks have resulted in large-scale equipment that offers
building design, the appropriate use of ventilation sys- improved containment capabilities. In a study of biohaz-
tems, and waste disposal concerns. ards using a large-scale zonal centrifuge on moderate risk
Anticipating this reliance on chemical process technol- oncogenic viruses, it was determined that minimal risks
ogy, a 1988 study characterized the engineering controls to laboratory personnel existed during optimum opera-
used in conventional enzyme fermentation processes (14). tion (16). However, faulty seals did result in the detection
Sample locations selected to reflect worker exposures to of high concentrations of phage in the turbine air exhaust
process microorganisms included the laboratory (where and the seal coolant system.
culture transfers were conducted), inoculum and fermen- Exhaust gases from the fermentor tanks can be another
tor tanks (sample ports and agitator shafts), filtering major emission source for the production microorganisms.
operations, and background locations (Fig. 1). The study The aeration of fermentor broths produces a foam on
 BIOAEROSOLS IN INDUSTRIAL SETTINGS 419

the surface of the liquid that results in the continuous the machine or, as is found in larger machine shops, from a
production of bursting bubbles. It has been demonstrated central sump that distributes fluids to multiple machines.
that the droplets of dilute solutions formed by the bursting Occupational exposures to the fluids or contaminants
of bubbles can enrich the concentration of microorganisms within the fluids are influenced by the fluid composition,
by factors of 10 to 1,000 times (17,18). The control of the tool type and speed, the method of fluid application,
vented bioaerosols within the fermentor tanks produced the use of engineering controls (such as enclosure and
by this and other mechanical aerosolization processes is local exhaust ventilation), and the fluid maintenance
achieved primarily by the application of sterilizing (i.e., program.
high efficiency) filtration systems preceded by ‘‘roughing’’ There are four major types of MWFs — straight oils,
systems (e.g., cyclones, scrubbers, and/or condensers) (19). water-soluble oils, semisynthetic, and synthetic. Straight
Data from the enzyme manufacturing study revealed that oils (neat oils) are solvent-refined petroleum oils not
scrubbing systems alone may not be effective in controlling designed to be mixed with water. The other three types
vented bioaerosols from the fermentors. The seals around are water-based. Soluble oil MWFs are emulsions or are
the agitator shafts may be another emission source for the composed in part of water-soluble oils (30% to 85% by
process microorganisms. Double mechanical steam seals concentration); the concentrate is diluted with water prior
appeared to provide inherently better containment than to use. Semisynthetic MWFs are composed of 5 to 30%
packed seals. petroleum oils. Synthetic MWFs contain no petroleum
The work practices of the operators can also be oils. MWF aerosol refers to the mist generated during
a determining factor in the potential for occupational grinding and machining operations, and may contain a
exposure. During the collection of fermentor tank mixture of substances. Water-based MWFs are suitable
sample volumes in enzyme manufacturing, operators growth environments for many types of microbes owing
were observed purging the sample port with a ‘‘blast’’ to the abundance of water and organic substrates. As
of pressurized steam prior to the collection of a broth a result, many MWFs and MWF aerosols are routinely
sample. The steam served to clean and decontaminate the contaminated by bacteria and/or fungi.
interior surfaces of the pipes to ensure a pure sample for Historically, microbial contamination of MWF has been
subsequent analysis. However, the contact time between a problem, primarily because it can affect fluid quality
the steam and residual microbial populations in the and performance. Fluid degradation from microorganisms
sample line were not adequate to kill the microorganisms can cause changes in fluid viscosity, and the acid
and therefore resulted in a dissemination of viable aerosols products of fermentation may lower the pH of MWF,
into the surrounding environment. Work practices are causing corrosion and leaks in the MWF system (20).
most reliable when used in combination with effective The predominant microbial species routinely recovered
engineering measures such as isolation or automation. from MWFs are frequently the same as those recovered
For example, microbial exposures during the separation of from natural water systems (20). Although most bacterial
solids can be reduced by limiting operator interaction with species found in MWFs are gram-negative, the microbial
those processes or, if this is not possible, the observance of populations within such MWF samples are continually
proper and safe work practices. changing (20,21). The species identified in MWFs are
Large-scale spills of fermentor tanks can be effectively generally characterized as nonpathogens or opportunistic
controlled by concrete dikes constructed around the pathogens, of which the bacterial genus Pseudomonas is
periphery of the tank. Spilled material is directed to the most frequently isolated (22,23).
a sump and subsequently pumped to a holding tank Bacterial concentrations in MWFs often range from
for inactivation of the microorganisms. In-line sterilizers 105 to 108 colony forming units per milliliter (CFU/mL),
may also prove effective prior to the disposal of the but they can be as high as 109 CFU/mL (24–28).
biological material. Bulk samples of the inactivated It has been suggested that well-maintained MWF
material should be microbiologically examined to validate systems should have bacterial concentrations of less
the efficacy of the sterilization techniques. Spill responders than 106 CFU/mL (29). Only recently have gram-positive
should be equipped with personal protective equipment bacteria and Mycobacterium species been identified as
(PPE), including impervious clothing, gloves, autoclavable predominant species in the MWF of some plants (30).
boots, and appropriate respirators (e.g., a self-contained Data regarding airborne concentrations of mycobacteria
breathing apparatus). from contaminated fluids are limited, but one study
has reported concentrations ranging from 1,300 to
MACHINING 9,200 CFU/m3 (31). Occupational disease syndromes have
been suggested in workers exposed to metalworking fluids,
Machine tools use metalworking fluids (MWFs) during some of which have been reported to be contaminated
grinding, forming, treating, and machining operations. with mycobacteria (32,33). Similarly, recent evidence in
These fluids extend the life of the bits, blades, or other the scientific literature suggests an association between
cutting components of the machine tool by cooling and the aerosolization of mycobacterial laden hot tub waters
lubrication of the machined surface. The application of and the development of respiratory disease in the exposed
the fluid also assists in the removal of the metal chips, population (34). However, although the acid-fast organism
swarf, fines, and other residues and may serve to protect or Mycobacterium chelonae has been found to be present
treat the material surfaces being machined. The fluids are in MWF associated with outbreaks of hypersensitivity
disseminated from a dedicated sump located proximal to pneumonitis (HP) (30), the significance of finding any
420 BIOAEROSOLS IN INDUSTRIAL SETTINGS

particular fungal or bacterial species in MWF is not clear thoracic particulate mass∗ as a time-weighted average
at this time. (TWA) for up to 10 hours per day during a 40-hour
Due to the limited nature of microbial sampling, results week (22). The NIOSH recommended exposure limit was
of bulk sampling of MWFs at any given time may only established primarily to eliminate or reduce respiratory
represent a portion of the microecology in the MWF. health effects. Neither the Occupational Safety and Health
Most water-based fluids have low concentrations of fungi, Administration (OSHA) nor the American Conference
except when a bloom (which is often caused by a dramatic of Governmental Industrial Hygienists (ACGIH) have
decrease in bacterial contamination) occurs (28,35). MWFs exposure limits for MWF aerosols, although both have
that have high concentrations of gram-negative bacteria an eight-hour TWA limit of 5 mg/m3 for mineral oil mist.
frequently have high levels of endotoxin, which have been Currently, there are neither specific microorganisms
quantified in machining operations where water-based in MWFs linked to specific, exposure-related health
MWFs are used. effects nor specific criteria concerning the level of total
Exposure to MWF aerosols is known to be associated microbial contamination that may be related to potential
with increased prevalence of respiratory symptoms, health effects. Because the ecology of MWF systems
decreases in pulmonary function, and the occurrence of fluctuates, documenting the microbial exposures at the
occupational asthma and HP (22). The significance of time of symptom (or disease) onset is difficult. This
reported respiratory and irritant symptoms among MWF- points to the need for ongoing evaluations of the MWF
exposed workers in relationship to loss of pulmonary environment, which may be correlated with ongoing
function or illnesses such as asthma or HP is unclear surveillance of health effects among exposed workers.
at this time. HP, which is an immunologically mediated Proper management of the MWF and ventilation systems
in machining areas, as components of an overall MWF
inflammatory disease of the lung that occurs after
health and safety program, plays an important part in
repeated inhalation and sensitization to a wide variety
reducing exposures to MWF aerosols and minimizing
of microbial agents (bacteria, fungi, amoebae), animal
potential illness.
proteins, and low-molecular weight chemical antigens, has
been associated with exposure to MWFs in several recent
reports (30,36,37). Aerosolized endotoxins are suspect WASTE TREATMENT AND HANDLING
causative agents of occupationally related respiratory
effects (e.g., chronic bronchitis, abnormal cross-shift Because many types of waste inherently contain poten-
declines in pulmonary function, and asthma) among tially infectious organisms, persons working in both
workers exposed to MWF aerosol (22). wastewater and solid waste systems have potential expo-
Limited information has been published regarding the sure to bioaerosols. Important components of bioaerosols
concentrations of disseminated bioaerosols from MWF present in waste systems include a variety of microorgan-
operations. However, questions regarding the develop- isms (viruses, bacteria, fungi, and protozoan) and com-
ment of occupationally related respiratory disease and ponents of microorganisms (endotoxin, [1 → 3]-Beta-D-
exposure to MWF have prompted continued research glucan) (40). Airborne exposures that have most commonly
focused on environmental characterization of the micro- been monitored include total and respirable dust (some-
biological component. In one study evaluating worker times including analysis of particle size distribution), bac-
exposures to MWFs (including synthetics, mineral oil, teria, and endotoxin. Recent data have shown that aerosols
and rapeseed oil), airborne concentrations of gram- generated by various types of wastewater treatment pro-
negative bacteria and endotoxins from contaminated fluids cesses include varying concentrations of microorganisms
ranged up to 41,000 CFU/m3 and 600 nanograms (ng) (including bacteria, fungi, and viruses) (41,42). Generally,
per cubic meter, respectively (38). Additionally, MWF the solids-handling aspects of waste systems (the typical
workers responded with higher IgG antibody levels to household and industrial sewage treatment process) are
Stenotrophomonas maltophila, Pantoea agglomerans, and thought to involve the most likelihood for bioaerosol expo-
Comamonas acidovorans, the most common bacterial con- sure (40). However, the relationship between bioaerosols
taminants in the fluids, than did controls. In a separate produced in sludge production, processing, and handling
study characterizing MWF exposure indices and acute res- steps and potentially related health effects is not well
piratory effects, a thoracic fraction sampler [that permit- described. In one study, a variety of clinical and labora-
tory effects among sludge workers were associated with
ted the collection of personal breathing zone (PBZ) samples
aerosolized sewage sludge or a dust of dried sludge, both
for bioaerosols] was used to enumerate airborne bacteria
probably containing substantial concentrations of endo-
and endotoxin (39). PBZ bacterial concentrations of per-
toxin (43).
sonnel working around MWF processes were reported to
Recently, the process of land application of biosolids
extend up to 26,600 bacteria/m3 . Sampling for endotoxins
(sewage sludge treated to significantly reduce pathogen
resulted in PBZ concentrations ranging to 234 endotoxin
levels) has received considerable attention. In the United
units per cubic meter of air (EU/m3 ).
States, land application of biosolids is a process regulated
To prevent or greatly reduce the risk of adverse health
effects due to MWF exposure, the National Institute for
Occupational Safety and Health (NIOSH) recommends ∗Thoracic particulate mass is the portion of MWF aerosol that
that airborne exposures to MWF aerosols be limited penetrates beyond the larynx and may be deposited in the lung
to 0.4 milligrams per cubic meter of air (mg/m3 ) for airways and/or gas exchange region.
 BIOAEROSOLS IN INDUSTRIAL SETTINGS 421

by the Environmental Protection Agency (EPA) in to bioaerosols, which may contain a variety of potentially
which biosolids are applied to farmland, surface mines, infectious or toxic substances.
and other lands, following specific requirements (44).
However, potential occupational exposures during biosolid WOOD PROCESSING
application are not directly considered by the EPA. One
case study by NIOSH found that reports of gastrointestinal Various wood processing operations can place workers at
illness among five workers were temporally related risk for health effects due to bioaerosols. The degree of risk
to work in a land application process (45). In that depends on the particular operation (logging, sawmilling,
survey, air concentrations of bacteria ranged from 412 furniture manufacturing, carpentry), the types of woods
CFU per cubic meter (CFU/m3 ) to 2,356 CFU/m3 , with utilized, the conditions in which the wood is stored
a small percentage of the organisms being enteric before use, and the particular treatments applied to
organisms. Endotoxin levels ranged from 23 to 39 the wood (for example, fungicides) (51). A number of
endotoxin units per m3 (EU/m3 ). Another survey of studies assessing the relationship between exposure of
bioaerosol concentrations at a land application site found wood processing workers to bacteria, fungi, endotoxins,
concentrations of bacteria in the air of 105 CFU/m3 ; those and (1 → 3)-Beta-D-glucan, and respiratory and mucus
investigators noted that their data may not be relevant membrane symptoms were conducted by Alwis and
when considering occupational exposures in areas or coworkers (52–54). The workers in the studies worked
processes in which workers are performing job tasks in logging sites, sawmills, wood chipping mills, and
involving sludge agitation (46). Although direct contact joineries. Maintenance workers were used as controls.
with potentially infectious material is likely to be the The lowest fungal concentrations were observed in the
greatest potential hazard for biosolid application workers, sawmills (3,000 CFU/m3 ) and joineries (4,000 CFU/m3 ).
there is potential for respiratory exposure to bioaerosols. These process operations were equipped with the most
NIOSH is currently drafting recommendations, including efficient and well-maintained exhaust ventilation systems
eliminating potential hazards where possible and using and used vacuuming to remove dust from machinery and
appropriate engineering controls and personal protective the floor. Workers in the woodchipping mill were exposed
equipment, to help minimize occupational exposures of to the highest concentrations of fungi (74,000 CFU/m3 ).
workers exposed to biosolids (47). Aspergillus fumigatus and Penicillium spp. were the
Handling of solid waste also involves exposure to predominant genera at the logging sites where debarking
bioaerosols. A recent exposure study of workers at a refuse- was conducted; Penicillium spp. predominated in sawmills
derived fuel (RDF) plant found mean air concentrations of and the woodchipping mill. Joinery, sawmill, and chip
total bioaerosol (measured by epifluorescence microscopy) mill workers showed significant correlations between
to range from 106 to 108 organisms per m3 , with endotoxin mean personal exposure and work-related symptoms after
concentrations ranging from 5 to 346 EU/m3 (48). Com- controlling for smoking and age. The investigators pointed
posting is another area of solid waste treatment that has out that, although the exposures were considered low,
been evaluated. In a case study evaluating exposure and they were still associated with symptoms. A dose-response
respiratory illness of a worker handling composted wood relationship was shown between (1 → 3)-Beta-D-glucan
exposure and the occurrence of chronic bronchitis, throat
chips and leaves, high levels of fungi (106 to 108 CFU/m3 ),
irritation, and regular ear inflammation and infections.
total bacteria (105 to 108 CFU/m3 ), and endotoxin (636
A significant correlation was also found between mean
to 16,300 EU/m3 ) were documented (49). Another type of
personal exposures to endotoxin and chronic bronchitis.
solid waste handling process has recently been found to
Pulmonary function testing showed that respiratory
present potential hazards to waste workers. As a com-
symptoms were correlated with cross-shift decreases in
ponent of an evaluation of exposures at a medical waste
lung function.
processing facility, where at least one employee was epi-
Similar occupational exposures to microbiological
demiologically determined to have acquired tuberculosis agents have been described in related industries, that
due to occupational exposure, investigators conducted is, cellulose and paper manufacturing, which use timber
bioaerosol sampling. Air concentrations of total bacteria as a raw material resource. Pulp logs absorb moisture
in the area of the process with the highest concentrations during water-driven transport or during storage prior
ranged from 183 to 256 CFU/m3 . Speciation of bacteria to processing. During storage, the ample supply of
was limited to three organisms (Pseudomonas aerugi- water and organic nutrient material provide an optimum
nosa, Escherichia coli, and Staphylococcus aureus), which environment for the development of microbiological
were considered atypical environmental organisms and communities, especially, thermotolerant species. The
utilized as indicators of aerosolization of waste materials. inner wood of logs are less likely to contain levels of
Although air sampling for Mycobacterium tuberculosis was fungi that would pose a significant biohazard; however,
not conducted, the investigators concluded that employees bacteria are more likely to develop in this ‘‘core’’ wood (55).
in that facility could be exposed to this pathogen as a Fungi appear to develop more prominently during the
result of occupational bioaerosol exposure; recommenda- processing of waste wood and during removal of bark. The
tions were made to minimize the potential for exposure debarking of the pulplogs and subsequent storage of the
and infection (50). resulting biomass can result in the release of bioaerosols,
In summary, waste handling and treatment of a variety including fungi, bacteria, and endotoxins. A NIOSH
of types of waste involves potential exposure of workers study of a paper manufacturing process revealed elevated
422 BIOAEROSOLS IN INDUSTRIAL SETTINGS

concentrations of thermoactinomycete species during the Avoidance of exposure to the offending agent(s) is the
debarking and subsequent transport and storage of primary means of prevention of these disorders and can
removed biomass material (56). Concentrations of bacteria usually be accomplished through product substitution or
were two orders of magnitude greater in the transfer engineering controls. PPE may be suitable for occasional
tower and the biomass storage area than those observed brief exposures. If symptoms persist despite engineering
in outdoor control samples with thermoactinomycetes controls and PPE, an affected worker will need to be
comprising greater than 90% of the genera identified. removed completely from the exposure situation (61). This
The lower concentrations of bacteria in the debarking has been shown to be particularly important for HP
process were attributed to the application of water, which as repeated episodes of HP can lead to chronic lung
reduce the dissemination potential of bacterial spores. disease and recurrent low level exposures can lead to the
Another reported study of a cellulose factory described insidious development of chronic interstitial lung disease
elevated exposure concentrations of bacteria and fungi to with fibrosis.
barking workers of 46,000 and 5,900 CFU/m3 (geometric
means), respectively (57). Caterpillar drivers operating TEXTILE INDUSTRY
in wood chip piles were exposed to concentrations of
bacteria averaging 1,500 CFU/m3 and fungi averaging Recognition of adverse respiratory effects related to work
45,000 CFU/m3 . Additionally, IgG antibody levels were in the textile industry dates back many years (62).
significantly higher among the wood chip workers as Byssinosis is an occupational lung disease caused by
opposed to the barking workers. It was suggested that inhalation of cotton, hemp, or flax dust, which in the
dry wood chip material can penetrate to the lower parts of early stage is characterized by distinctive chest tightness
the lung more readily than wet aerosols generated in the associated with decreased lung function. The relationship
barking process. of respiratory effects and workplace exposures to cotton
Exposure to bioaerosols have also been documented in dust vary depending on the stage of processing workers
the paper recycling industry. In a Danish study of six are performing, with higher risk areas being those
recycling plants, two plants were reported with very high areas where the cotton is first processed (4,63). OSHA
counts of airborne microorganisms noting that several has regulations, directed primarily at the prevention
values exceeded the Danish threshold limit value of of byssinosis that limit occupational exposure to cotton
10,000 CFU/m3 (58). The highest microbial concentrations dust (64,65), and decreases in exposure to cotton dust in
were observed in the effluent water treatment area with the textile industry has been associated with a decrease
a mean concentration of 261,000 CFU/m3 ; lesser mean in the incidence of byssinosis (62). There has been ongoing
concentrations of 34,000 and 8,700 CFU/m3 were observed research related to bioaerosol exposure in the textile
around paper machines and near winding and cutting industry. Important issues being addressed by some of that
machines, respectively. Samples for endotoxins at one research include the issue concerning byssinosis occurring
plant near the paper machine revealed concentrations, among persons with exposures below the limits set by
which exceeded the Danish threshold limit value of OSHA and the issues of other respiratory disorders such as
100 ng/m3 (mean value of 294 ng/m3 ). High endotoxin chronic bronchitis and long-term decrements in pulmonary
concentrations may have been the result of intensive reuse function, which have been associated with exposure to
of process water. cotton dust. Many investigators have evaluated the role
Health effects associated with exposure to wood dust of endotoxin as the etiologic agent within cotton dust,
include occupational asthma, HP, organic dust toxic syn- and acute pulmonary responses (decreased pulmonary
drome (ODTS), chronic bronchitis, and upper respiratory function) have been related to the endotoxin concentration
and mucus membrane irritation (59). Although exposures within the cotton dust (4,66). More recently, results of
to the dusts of many different woods are thought to be an 11-year longitudinal study indicate that chronic lung
associated with occupational asthma, a specific sensitiz- function decrements are more closely related to cotton
ing agent directly related to the occurrence of asthma dust exposure than exposure to associated endotoxins (67).
has been identified for only one type of wood (plicatic This raises the question of whether exposure to bioaerosol
acid in the case of western red cedar) (59). A variety of components (e.g., endotoxin) and vegetable components of
fungi have be found to be associated with HP among dusts such as cotton dust may lead to different health
workers in this industry, including Cryptostroma corticale effects.
(maple bark), Aureobasidium pullulans (redwood), and Studies designed to characterize occupational expo-
Alternaria, Aspergillus, and Thermoactinomyces (moldy sures to microbiological agents in cotton mills have been
wood chips) (59). Penicillium, Graphium, Pullularia, and reported. In a study of two cotton mill plants in China,
Trichoderma species have also been found to be causative samples for airborne dust and endotoxins were collected
antigens (60). ODTS is characterized by the occurrence of using vertical elutriators and 5 micrometer (µm) polyvinyl
a flulike illness (fever, chills, myalgia, malaise, nonpro- chloride filters (68). Dust concentrations were similar for
ductive cough) within a few hours of inhalation of wood the two plants ranging from 0.15 to 2.34 mg/m3 at the first
dusts contaminated with fungi, bacteria, and/or endotox- plant and 0.19 to 2.50 mg/m3 at the second plant. How-
ins. Although the clinical picture differs in some respects ever, endotoxin concentrations between the two plants
from HP (59,61), it has been postulated that HP and ODTS were different, with the first plant exhibiting higher lev-
may represent parts of a spectrum of responses to exposure els, ranging from 20 to 920 ng/m3 , and the second plant,
to bioaerosols (49). ranging from 2 to 550 ng/m3 . In a separate study of two
 BIOAEROSOLS IN INDUSTRIAL SETTINGS 423

cotton mills in England (one mill was characterized as the potential for bioaerosol exposures. The discussions
new while the other had been long established), air sam- provided here are not meant to be an exhaustive
ples were collected for culturable fungi, actinomycetes, review; for example, in-depth studies at other types
and bacteria and for spores counts of fungi, actino- of industrial sites have provided information useful in
mycetes, and bacteria (69). Bulk samples of the cotton characterizing exposures and health effects related to
were also collected and microbiologically analyzed. Cotton bioaerosol exposure (76,77). Further research is needed
dust samples revealed spore counts averaging 2,400,000 to help clarify the health implications of exposure to
spores per gram (spores/gm) of material, with approx- bioaerosols in many industrial settings.
imately two-thirds comprising ‘‘actinomycete + bacteria’’
classified spores. Culture techniques revealed predomi- BIBLIOGRAPHY
nance by bacteria with lesser numbers of actinomycetes,
including Faenia rectivirgula, Streptomyces albus, S. 1. H. A. Burge, Occup. Med.: State Art Rev. 4(4), 713–721 (1989).
griseus, and Thermoactinomyces spp., and fungi, including 2. J. Macher et al., eds., Bioaerosols: Assessment and Control,
species of Aspergillus, Penicillium, and Cladosporium. As American Conference of Governmental Industrial Hygienists,
with the bulk sample analyses, ‘‘actinomyces + bacteria’’ Cincinnati, Ohio, 1999.
classified spores were the most numerous spores iden- 3. R. Rylander, Scand. J. Work Environ. Health 20(special
tified with a mean concentration of 220,000 spores/m3 issue), 116–122 (1994).
for ‘‘actinomyces + bacteria’’ and 50,000 spores/m3 for 4. R. M. Castellan, S. A. Olenchock, K. B. Kinsley, and J. L.
fungi in the new mill, and 2,770,000 spores/m3 for Hankinson, N. Engl. J. Med. 317(10), 605–610 (1987).
‘‘actinomyces + bacteria’’ and 50,000 spores/m3 for fungi 5. S. A. Olenchock, Appl. Occup. Environ. Hyg. 9(1), 62–64
in the long established mill. This was consistent with the (1994).
culturable air sample analyses, with bacteria being the 6. D. T. W. Chun et al., Appl. Occup. Environ. Hyg. 15(1),
most numerous on most culture plates. 152–157 (2000).
7. Office of Technology Assessment, Commercial Biotechnology:
An International Analysis, Report No. OTA-BA-218, Office of
FOOD PROCESSING Technology Assessment, Washington, D.C., 1984.
8. M. D. Topping et al., Br. J. Ind. Med. 42, 312–318 (1985).
The relationship of bioaerosol exposure as a cause of health
9. A. Seaton and D. Wales, Occup. Environ. Med. 51, 54–56
effects has been evaluated in a variety of food processing (1994).
plants, including trout processing facilities (70) and beet
10. F. Lagier, A. Cartier, J. Dolovich, and J. Malo, Thorax 44,
sugar refineries (71). The latter survey documented expo- 157–158 (1989).
sure to fungal bioaerosol in a workplace where employees
11. J. H. Duffus and C. M. Brown, Ann. Occup. Hyg. 29(1), 1–12
had been diagnosed with occupational asthma, and pro- (1985).
vided recommendations for detailed engineering controls
12. L. Andrup, B. H. Nielsen, and S. Kolvraa, Scand. J. Work
designed to minimize bioaerosol exposure. Respiratory Environ. Health 16, 85–95 (1990).
effects related to exposure to flours and grains have been
13. J. Van Houten, in W. C. Hyer, ed., Bioprocessing Safety:
researched for many years; it is clear that work-related Worker and Community Safety and Health Considerations,
respiratory symptoms and illnesses have been documented American Society for Testing and Materials, ASTM STP 1051,
among workers in both agricultural and nonagricultural Philadelphia, Pa., 1990, p. 91.
industries in which exposure to flours and grains occurs. A 14. K. F. Martinez, J. W. Sheehy, J. H. Jones, and L. B. Cusick,
broad range of etiologic agents are thought to be involved Appl. Ind. Hyg. 3(6), 177–181 (1988).
with these health effects, including many types of microor- 15. P. A. Belter, in H. J. Peppler and D. Perlman, eds., Microbial
ganisms (such as bacteria, fungi, wheat weevils, and Technology: Fermentation Technology, Academic Press, New
mites) and additives (such as enzymes and other pro- York, 1979, p. 403.
teins) (72). A series of evaluations have been performed 16. C. J. Baldwin, J. F. Lemp, and M. S. BarBeito, Appl. Micro-
concerning the potato processing industry (73,74) and a biol. 29(4), 484–490 (1975).
recent report evaluating workers in that industry demon- 17. D. C. Blanchard and L. D. Syzdek, Appl. Environ. Microbiol.
strated that decrements in lung function were associated 43, 1001–1005 (1982).
with endotoxin exposure above 53 EU/m3 (75). In addi- 18. S. Wangwongwatanta, P. V. Scarpino, K. Willeke, and
tion to the components of bioaerosols mentioned earlier, P. A. Baron, Aerosol Sci. Technol. 13, 297–307 (1990).
another area of research relevant to the food processing 19. P. Sayer, J. Burckle, G. Macek, and G. LaVeck, in
industry includes evaluation of microbial volatile organic B. Lighthart and A. J. Mohr, eds., Atmospheric Microbial
compounds as etiologic agents for respiratory and other Aerosols: Theory and Application, Chapman & Hall, New
symptoms. York, 1994, p. 331.
20. E. C. Hill, in K. W. A. Chater and E. C. Hill, eds., Monitoring
and Maintenance of Aqueous Metal-working Fluids, John
CONCLUSION Wiley & Sons, London, U.K., 1984, 97–113.
21. J. P. Byers, ed., Metalworking Fluids, Marcel Dekker, New
The preceding discussions demonstrate that exposure to York, 1994, p. 323.
bioaerosols can occur in a wide variety of nonagricultural 22. NIOSH, Criteria for a Recommended Standard: Occupa-
industrial settings and that occupational safety and health tional Exposure to Metalworking Fluids, DHHS (NIOSH)
professionals in these settings must continually consider Publication No. 98-102, Department of Health and Human
424 BIOAEROSOLS IN INDUSTRIAL SETTINGS

Services, Public Health Service, Centers for Disease Control 48. S. Mahar, S. J. Reynolds, and P. S. Thorne, Am. Ind. Hyg.
and Prevention, National Institute for Occupational Safety Assoc. J. 60, 679–683 (1999).
and Health, Cincinnati, Ohio, 1998. 49. S. Weber et al., Am. J. Ind. Med. 24, 365–374 1993.
23. M. K. Lonon, M. Abanto, and R. H. Findlay, Am. Ind. Hyg. 50. NIOSH, Hazard Evaluation and Technical Assistance Report:
Assoc. J. 60, 480–485 (1999). Stericycle, DHHS (NIOSH) Publication No. 98-0027-2709,
24. E. O. Bennet, Lubrication Eng. 28(6), 237–247 (1972). U.S. Department of Health and Human Services, Public
25. I. Mattsby-Baltzer et al., Appl. Environ. Microbiol. 55(10), Health Service, Centers for Disease Control and Prevention,
2681–2689 (1989). National Institute for Occupational Safety and Health,
Cincinnati, Ohio, 1998.
26. S. Foxall-VanAken et al., Appl. Environ. Microbiol. 51(6),
1165–1169 (1986). 51. D. Enarson, in R. Rylander and R. Jacobs, eds., Organic
Dusts: Exposure, Effects, and Prevention, Lewis Publishers,
27. I. Salmeen, J. A. Brown, Jr., S. Foxall-VanAken, and
Ann Arbor, Mich., 1994..
R. H. Olsen, Tribology International 20, 218–221 (1987).
52. K. U. Alwis, J. Mandryk, and A. D. Hocking, Appl. Occup.
28. M. D. Wort, G. I. Lloyd, and J. Schofield, Tribology Interna-
Environ. Hyg. 14(9), 598–608 (1999).
tional February, 35–37 (1976).
53. J. Mandryk, Am. J. Ind. Med. 35, 481–490 (1999).
29. L. A. Rossmoore and H. W. Rossmoore, in J. Byers, ed.,
54. J. Mandryk, Ann. Occup. Hyg. 44(4), 281–289 (2000).
Metalworking Fluids, Marcel Dekker, New York, 1994,
pp. 247–271. 55. J. Dutkwiewicz, Biodeter. Res. 2, 533–547 (1989).
30. K. Kreiss and J. Cox-Ganser, Am. J. Ind. Med. 32, 423–432 56. NIOSH, Hazard Evaluation and Technical Assistance Report:
(1997). Champion International, DHHS (NIOSH) Publication No. 90-
0375-2334, U.S. Department of Health and Human Services,
31. J. S. Moore et al., Am. Ind. Hyg. Assoc. J. 61, 205–213 (2000).
Public Health Service, Centers for Disease Control and
32. A. Freeman et al., Am. J. Ind. Med. 34, 387–392 (1998). Prevention, National Institute for Occupational Safety and
33. B. G. Shelton, W. D. Flanders, and G. K. Morris, Emerg. Health, Bucksport, Me., 1993.
Infect. Dis. Mar–April 5(2), serial online — 4 screens (1999). 57. M. Kotimaa, P. Koskela, and L. Saloranta, Proceedings of
34. J. Embil et al., Chest 111(3), 813–816 (1997). the VIIth International Pneumoconiosis Conference, ‘‘Microbe
35. H. W. Rossmore, Comprehensive Biotechnology, in M. Moo- exposure and the occurrence of antibodies against the expos-
Young, C. L. Cooney, and A. Humphrey, eds., The Principles ing microbes among wood workers in cellulose industry,’’
of Biotechnology: Engineering Considerations, Pergamon DHHS (NIOSH) Publication No. 98-108 Part 1, U.S. Depart-
Press, New York, 1985, pp. 249–268. ment of Health and Human Services, Public Health Service,
Centers for Disease Control and Prevention, National Insti-
36. CDC, MMWR 45(28), 606–610 (1996).
tute for Occupational Safety and Health, Cincinnati, Ohio,
37. M. C. Zacharisen et al., J. Occup. Environ. Med. 40(7), 1990.
640–647 (1998).
58. B. A. Rix and E. Lynge Am. J. Ind. Med. 30(4), 142–147
38. S. Laitinen et al., Int. Arch. Occup. Environ. Health 72, (1996).
443–450 (1999).
59. D. Enaarson and M. Chan-Yeung, Am. J. Ind. Med. 17, 33–38
39. L. Abrams et al., Appl. Occup. Environ. Hyg. 15(6), 492–502 (1990).
(2000).
60. D. M. Halpin, B. J. Graneek, M. Turner-Warwick, and
40. C. S. Clark, in R. Rylander and R. Jacob, eds., Organic Dusts: A. J. Newman Taylor, Occup. Environ. Med. 51, 160–164
Exposure, Effects and Prevention, Lewis Publishers, Ann (1994).
Arbor, Mich., 1994. 61. J. Balmes and C. Scannell, in J. Ladou, ed., Occupational and
41. G. Brandi, M. Sisti, and G. Amagliani, J. Appl. Microbiol. 88, Environmental Medicine, Appleton and Lange, New York,
845–852 (2000). 1997.
42. K. F. Fannin, S. C. Vana, and W. Jakubowski, Appl. Environ. 62. K. H. Kilburn, in W. M. Rom, ed., Environmental and Occu-
Micro. 49(5), 1191–1196 (1985). pational Medicine, Lippincott-Raven, New York, 1998,
43. I. Matsby and R. Rylander, J. Occup. Med. 20, 690–692 pp. 449–455.
(1978). 63. S. N. Raza et al., Occup. Environ. Med. 56, 514–519 (1999).
44. US EPA, Environmental Regulations and Technology: Control 64. OSHA, Occup. Exposure Cotton Dust 43, 27351–27417 (1978).
of Pathogens and Vector Attraction in Sewage Sludge, US 65. OSHA, Occup. Exposure Cotton Dust 50, 51120–51179 1985.
EPA Publication No. EPA/625/R–92/013, Washington, D.C.,
66. R. M. Castellan et al., Ann. Intern. Med. 101, 157–163 (1984).
1999.
67. D. C. Christiani et al., Am. J. Ind. Med. 35, 321–331 (1999).
45. NIOSH, Hazard Evaluation and Technical Assistance Report:
68. S. M. Kennedy et al., J. Am. Thorac. Soc. 135(1), 194–200
Bio-Solids Land Application Process, DHHS (NIOSH) Pub-
(1987).
lication No. 98-0118-2748, U.S. Department of Health and
Human Services, Public Health Service, Centers for Disease 69. J. Lacey and M. E. Lacey, Ann. Occup. Hyg. 31(1), 1–19
Control and Prevention, National Institute for Occupational (1987).
Safety and Health, Cincinnati, Ohio, 1999. 70. D. Sherson, I. Hansen, and T. Sigsgaard, Allergy 44, 336–341
46. S. D. Pillai, K. W. Widmer, S. E. Dowd, and S. C. Ricke, App. (1989).
Environ. Microbiol. 62(1), 296–299, (1996). 71. P. A. Jensen et al., Am. Ind. Hyg. Assoc. J. 54(12), 742–748
47. NIOSH, Hazard ID Workers Exposed to Class B Biosolids, (1993).
U.S. Department of Health and Human Services, Public 72. S. Brooks, in W. M. Rom, ed., Environmental and Occu-
Health Service, Centers for Disease Control and Prevention, pational Medicine, Lippincott-Raven, New York, 1998,
National Institute for Occupational Safety and Health, pp. 481–524.
Cincinnati, Ohio, 2000 draft. 73. J. P. Zock et al., Clin. Exp. Allergy 26, 542–548 (1996).
 BIOAEROSOLS: TRANSPORT AND FATE 425

74. A. Hollander, D. Heederik, and H. Kauffman, Occup. Envi- Table 1. Airborne Human Pathogens from Nonhuman
ron. Med. 51, 73–78 (1994). Sources
75. J. P. Zock, A. Hollander, D. Heederik, and J. Douwes, Am. J. Agent Disease Source
Ind. Med. 33, 384–391 (1998).
76. D. K. Milton et al., Am. J. Ind. Med. 28, 469–488 (1995). Bacillus anthracis Woolsorter’s disease, Contaminated hides,
77. D. K. Milton et al., Am. J. Ind. Med. 29, 3–13 (1996). pneumonic and bone meals
cutaneous anthrax
Histoplasma Histoplasmosis Soil enriched with
capsulatum bird droppings
BIOAEROSOLS: MODELING. See MODELING THE Coxiella burnetii Q fever Contaminated meats,
TRANSPORT OF BIOAEROSOLS animal products
Chlamydia psittici Ornithosis Dried droppings from
infected fowl

Source: Table Reproduced from Burrell (1991) with permission of the

publisher (1).
BIOAEROSOLS: TRANSPORT AND FATE

FRANCES L. STITES Table 2. Common Airborne Opportunistic Infections from

Dugway Proving Ground Nonhuman Sources
Dugway, Utah
Agent Disease Source
The study of bioaerosols (microorganisms aerosolized
Cryptococcus Cryptococciosis Pigeon dropping
by natural or artificial means) and how they behave neoformans contamination
in the environment has long interested the scientific Acanthamoeba spp. Various amoebic Natural water
community. In the late 1800s, it was discovered that infections sources
diseases could spread from person to person through Legionella spp. Legionaires’ disease Natural water
the air. The process of how diseases were actually sources
transmitted was not understood however, and the Atypical Tuberculosis like Water vegetation
effort to elucidate that process was the birth of the mycobacteria illnesses
scientific study of aerobiology. Aerobiology is defined Aspergillus spp. Allergy, Widespread in
as the study of the transport and fate of aerosolized bronchopulmonary decaying
aspergillosis, vegetation
microorganisms. Since then the study of bioaerosols
mycetomas,
has continued and much has been learned in the last hypersensitivity
century. However, as with many subjects of scientific pneumonitis
interest, there is still much to learn and understand
with regard to the transport and fate of bioaerosols. Source: Table Reproduced from Burrell (1991) with permission of the
publisher (1).
Within the last decade, interest in bioaerosol studies has
grown considerably because of the increase in perceived
threats of biological warfare and biological terrorism.
Therefore, the study of bioaerosols has two purposes, aerosolized microorganism and all of the external
the prevention and prediction of the spread of airborne factors that influence that survivability. These two
contagious diseases and the detection of aggressive parameters of biological aerosol study are interrelated.
biological warfare/terrorism threats. Both transport and fate are influenced by many of the
A large number of human pathogens are aerosolized same parameters including meteorology, forces involved
from a wide variety of sources. Airborne human pathogens in aerosol generation and dispersion, and time of day.
and common airborne microbes that cause opportunistic In this article, the transport and fate of bioaerosols
infections are listed in Tables 1 and 2, respectively. are investigated with an in-depth examination of their
The physics of biological aerosol stability and transport contributing factors, as they are currently understood.
have long been established. Modern efforts are geared It is important to note that all of the studies cited in
toward clarifying some of the more obscure effects the this chapter used culture-based analysis, and this assay
environment has on the bioaerosol as a whole. As the data method favors the results discussed.
from current efforts are collected and added to the body of
knowledge it helps to increase our understanding of a very
complex subject that has perplexed scientists for many TRANSPORT
decades.
Transport and fate are two fundamental parameters of According to Bartlett (2), there are three distinct stages
bioaerosols that must be understood by aerobiologists. involved with the transport of biological aerosols: disper-
Transport involves the origin of the bioaerosol and sal, diffusion, and deposition. Dispersal is the aerosoliza-
meteorological influences, as well as physical obstacles tion of the microbe, diffusion is the movement of the
that may affect the path the bioaerosol travels. The microbe over a distance, and deposition is the settling
second parameter of interest is bioaerosol fate. This of the microbe onto a surface. All of these stages are
relates to the biological survival or viability of the influenced by meteorological conditions. In order to fully
426 BIOAEROSOLS: TRANSPORT AND FATE

appreciate their functions in aerosol transport, each of the several patterns were observed (10). The average counts of
three stages will be investigated. aerosolized bacteria in the four locations were: agricultural
areas, 99 bacteria/m3 ; coastal areas, 63 bacteria/m3 ; city
Dispersal parks, 763 bacteria/m3 ; and city streets, 850 bacteria/m3 .
In all areas, the highest counts were observed in summer
Dissemination of microorganisms occurs when adhesion
and autumn with the lowest counts observed in winter.
(i.e., electrostatic) forces attracting a particle to a surface
Rain and low relative humidity (RH) corresponded to
are overcome by wind, mechanical, or other forces. This
decreased bacterial counts.
action is termed dispersal. Any activity or event that
Humans also generate natural bioaerosols as they go
provides the energy necessary to elevate a microorganism
into the atmosphere is capable of producing a bioaerosol. about their everyday activities. General movement (e.g.,
Naturally occurring bioaerosols are generated by a bed making, closing curtains, dressing, walking, talking,
variety of means. One example of natural aerosolization coughing, or laughing), can generate significant amounts
is injection into the atmosphere by a bubble traveling of bioaerosols. A cough can generate 3,000 droplets, talking
through a liquid and bursting on the surface (i.e., waves for five minutes may also generate 3,000 droplets, and
breaking on the shore, wastewater treatment plants). sneezing can generate as many as 40,000 droplets (11,12).
Aerosol particles generated by bursting bubbles have a Table 3 lists a variety of sources for bioaerosols of health
higher concentration of microorganisms than a sample of significance and the activities creating the aerosols.
the same volume taken from the liquid suspension that the Artificial bioaerosols can be generated using equipment
bubble traveled through (3–8). Aerosol droplets containing specially designed to serve that purpose. Some examples
Serratia marcescens generated by the bubbling action and uses of aerosol generators include, atomizers used for
contain concentrations of bacteria greater than the source the generation of mondisperse aerosol standards for the
solution by a factor of 1,000 (4). Other studies indicated calibration of particle size analyzers (13) and atomizers
that viruses could be concentrated up to 50 times with used in the agricultural industry for spraying trees and
respect to the source solution concentration (6,7). During plants for pest control (14). Other aerosol generators
the bubbling action the microbes collected in the resulting include the Collison nebulizer commonly used to generate
aerosol particle are concentrated in the surface microlayer aerosols used in respiratory studies (15), the vibrating
and are drawn into the particle during formation (5). orifice aerosol generator (16) and the spinning top aerosol
Additionally, it was found that aerosols generated generator (17), both used to generate monodisperse
from five-day-old Escherichia coli cultures, showed an aerosols for calibration.
aerosol particle concentration 15 times greater than Artificial bioaerosol generation is sometimes an unfor-
those generated from one-and-a-half-day-old cultures (5). tunate by-product of industrial and agricultural activity.
However, the researchers did not observe this same effect In one community, in which contaminated wastewater
with all of the bacteria used in their study. They felt that was used for spray irrigation, the number of enteric com-
increased lipid content in the cell envelope could account municable diseases was two to four times higher than
for the organisms being present in the surface microlayer. in communities that did not use wastewater for irriga-
They theorized that the lipid content of the cell envelope tion (18). It was determined that the bacterial levels in
in E. coli might increase as the organism ages, which the air were directly related to the bacterial levels in the
would account for the increase in particle concentration wastewater being sprayed. Wastewater treatment plants
over time. are also a significant source of bioaerosols. Aerators used
Bioaerosols can be dispersed by natural events to oxygenate the wastewater generate bubbles that burst
including rain or snow impacting on the surface of creating aerosols (19) as described earlier.
pooled water (4) and wind blowing across a surface where Artificial bioaerosols are intentionally generated in the
microorganisms are present. For this last mechanism agricultural industry for use in pest and disease control.
the force required to remove a particle from a surface For example, Erwinia sp. are commonly sprayed onto fruit
increases as the particle size decreases (9). Bartlett (2) trees in an aerosol form to help control fire blight (20–23).
explains that almost all naturally occurring organisms Generation of bioaerosols by either natural or artificial
are released within three to four meters from the ground means exerts some degree of stress on the microbes.
with a majority being released less than 1 meter from the Strange and Cox (24) state that the injury to an aerosolized
surface. It is this region that is most strongly affected microbe results mainly from shear and impaction forces
by the daily cycle of heating and cooling. During the with the extent of the injury depending on the method of
day this region can become very unstable so organisms aerosol generation, the microbial species being generated,
will be rapidly mixed vertically through a substantial and the physiological state of the microbe. However, in
depth of the atmosphere. After dark, strong inversions a more recent study using Pseudomonas syringae and
develop and the movement of particles in the lower Erwinia herbicola and an aerosol generation method
atmospheric layers is restricted. Material that finds its dissimilar to that discussed by Strange and Cox, it
way above the inversion becomes entrained in a low- was found that the mechanism of aerosol generation
level jet of air and can be transported to substantial itself did not seem to adversely affect the organisms
distances (2). being aerosolized (25). This information could lead one to
In a three-year study of the natural fluctuation conclude that the damage to an aerosolized microbe occurs
of aerosolized microbes in four localities (agricultural after aerosolization; however, further study is needed in
districts, coastal areas, city parks, and city streets) this area.
 BIOAEROSOLS: TRANSPORT AND FATE 427

Table 3. Health Significance of Various Sources of Indoor Aerosols

Examples of Organisms of Potential Risk to

Sources Activities Creating Aerosols Health Significance Human Health

Human:
Desquamated skin Motion (clothed and Staphylococci Low
unclothed), showering, bed
making
Respiratory tract Talking, coughing, sneezing, Staphylococci, streptococci, Moderate to high
blowing nose respiratory viruses,
Mycobacterium tuberculosis,
Yersinia pestis, Haemophilus
pertussis
Gastrointestinal tract Toilet Escherichia coli, enteroviruses Low
Ventilation:
Penetration from exterior Air movement from cooling Legionella species, B. anthracis, Low to high
towers and other exterior and various others
aerosols
Interior systems Operation Pseudomonas species, Low
staphylococci, fungi
Humidifiers Operation Pseudomonas species, Moderate
Acinetobacter species,
Serratia species,
Actinomycetes species
Industrial:
Wool, goat hair, cotton Textile and furniture Bacillus anthracis, C. burnetii, High
manufacture gram-negative organism
Slaughtered animals and Meat packing and rendering Coxiella burnetii, Brucella High
birds plants species, Chlamydia psittaci
Vegetables Freezing, canning, drying Streptococcus species Low
Dairies Milk processing Coxiella burnetii, streptococci Low to high

Source: Table Reproduced from Spenklove and Fannin (1983) with permission of the publisher (12).

Diffusion North American continent in summer. They feel this may

also explain the ubiquity and persistence of the disease.
After dispersion into the surrounding atmosphere, the
Cold temperatures at high atmospheres, where the virus
aerosol diffuses by natural meteorological means. Diffu-
is theorized to travel, may increase the survival of the
sion is the spreading of aerosol particles caused by thermal
virus. An important factor in long-range transmission is
energy possessed by all particles, expressed as Brownian
the ability of the microbe to survive in the aerosolized
motion, or by the action of meteorological conditions (e.g., state (26); this issue will be further explored in the fate
wind mixing). Bioaerosols can be transported very long dis- section of this chapter.
tances if the conditions are right. To enable long-distance The particle size of the generated aerosol is very
transport of naturally occurring aerosols, it is necessary important in the diffusion process. Not only does it
that the aerosol source be abundant, the microbes must be determine how far a particle will travel but it influences
resistant or protected from the atmospheric environment, the survival of the bioaerosol. As the aerosol particle
the aerosol particle size not too large (e.g., <10 µm), and increases in size, the survival rate of the microbes
the meteorological conditions be favorable (e.g., high wind contained within the particle also increases (24). However,
speed and suitable stratification) (26). if a particle is too large it will not travel a great distance.
Terrestrial-based bacteria have been found 160 km out A 100-µm particle has a gravitational settling velocity
to sea and marine-based bacteria found 54 km inland (27). of 10 to 50 cm per second, whereas a particle less than
There are also reports of various plant pathogens (28) 1 µm has a gravitational settling velocity of 0.003 cm per
and the causative agents of Foot and Mouth disease second (33). Small particulates of 1 to 2 µm can remain
(FMD) (29–31) and New Castle disease (31) traveling suspended in room air for more than 90 minutes (12).
great distances. Bacterial spores aerosolized during a sand However, a large particle can remain airborne and travel
storm near the Black Sea were found to have traveled a significant distance if the atmospheric turbulence has a
1,800 km to Sweden and Finland (32). Stem rust disease velocity component with updrafts exceeding one meter per
has been documented to travel from the Mississippi Valley second (33). These updrafts will counteract gravitational
to central and northern Canada (33). settling of larger particles.
One group of researchers (34) theorize that influenza Cool humid weather, and overcast skies seem to favor
viruses are transmitted over large distances from the far the transport of bioaerosols. This effect may be due to the
east because of seasonal atmospheric circulation patterns protection afforded microbes from the damaging conditions
that may help explain the absence of influenza on the of desiccation, heat, and UV radiation (31).
428 BIOAEROSOLS: TRANSPORT AND FATE

Deposition Relative Humidity

Deposition (the settling or impaction of an aerosol particle RH is defined as the ratio of the amount of water vapor
onto a surface) and terminal velocity (the rate at which a actually present in the air to the greatest amount possible
particle settles influenced by gravitational force) are both at the same temperature. In a study performed by Walter
governed in part by the mass and size of the particle (9). and coworkers (25), it was found that the RH level was
Particles 1–5 µm generally follow the streamlines of more detrimental to bacterial survival than temperature
the surrounding air, whereas larger particles have the or the physical stress of aerosolization.
momentum to deviate from the streamlines and impact Changes in water content represent the most funda-
surrounding surfaces resulting in deposition of the mental potential stress experienced by airborne microbes.
particle (9). When particles enter the region near a surface, In general, aerosolized bacteria have an increased sur-
van der Waals forces (weak attractions among electrically vival/viability at low and high RH levels than at mid-range
neutral molecules or parts of molecules) and electrostatic humidities (24). The actual effect of desiccation on an
forces (electrical charge forces of attraction and repulsion) aerosolized microbe is difficult to determine because of
can influence deposition (9). concurrent inactivation of the microbe by oxygen and
Deposition of bioaerosols is also influenced by meteoro- other factors (33). In order to negate the effect of oxygen
logical conditions. For example, they will be restricted from on aerosolized microbes, researchers conducted studies in
deposition if vertical mixing and winds are strong enough enclosed chambers in which oxygen was replaced by an
to keep the particles aloft. Particles are washed from inert gas such as nitrogen. They were then able to observe
the atmosphere by precipitation, both rain and snow (10). the effect of different levels of RH on bioaerosols.
It has been shown that the rate at which particles are In the early to mid 1960s it was a common practice
removed by rain increases as the particle size increases (2). to evaluate aerosol survival at only three RH values
It was shown with FMD that when winds are blowing corresponding to high (80%), medium (50%), and low
from infected areas, outbreaks tend to occur where rain (30%) RH. During a study using E. coli, it was observed
has fallen (31). It is theorized that this may be a result of that critical narrow RH bands exist where loss of
the FMD virus-bearing particles being washed out of the viability of E. coli was much greater than in adjacent
atmosphere. regions (24,33,38). Figure 1 displays the data from that
Deposition in the respiratory tract depends on the size study (33). The aerosols were generated in a nitrogen
of the particle (27,35). 0.5–15 µm is the optimum size atmosphere and RH was increased in 2% increments. This
for deposition in the lungs (12,24,31). Respirable particles phenomenon of narrow bands of instability was also seen
are considered to be in the size range of 1 to 5 µm with freeze-dried bacteria and viruses (24).
(the size of most bacterial and fungal spores). These Many theories exist regarding the effect desiccation,
particles will penetrate as far as the distal alveoli and caused by low RH, has on aerosolized microbes. In 1959,
respiratory bronchioles (1). Therefore, it is easy to see that Webb (39) proposed that cell death, after aerosolization,
the infectivity of a given microorganism is in some ways was due to water molecule movement in and out of a cell in
dependent on the aerosol particle size. Particle size can
also influence the dose of an organism required to cause
infection. As the particle size increases, so does the dose 100
required to cause disease (36). This may be due to the
fact that the size of the particle determines where in the
respiratory tract the particle is deposited (27,36).

FATE 10
% Viability

Before 1965, there was little understanding of the

survival of microbes in aerosols (33). Since then important
research has been completed to explore the effects various
factors have on bioaerosols. According to Theunissen, and
coworkers (37) the fate of aerosolized microorganisms 1
depends on five factors: (1) relative humidity (RH),
(2) temperature, (3) oxygen concentration, (4) presence of
UV radiation, and (5) constituents of the aerosol itself. In
addition to these five factors, the type of microorganism
being aerosolized and how much time the microbe spends
in the aerosol are important. Interactions among these
0.1
factors complicate interpretation making it difficult to 100 80 60 40
ascertain the effect each individual factor has on the % RH
aerosol stability of the microbe. However, it is generally Figure 1. Aerosol survival of E. coli B sprayed suspension in
agreed that the factor that most influences the survival or distilled water and stored in nitrogen atmospheres. Reprinted
viability of the bioaerosol is the RH level. from Cox (1989) with the permission of the publisher (33).
 BIOAEROSOLS: TRANSPORT AND FATE 429

an equilibrium system resulting in the collapse of natural ∆ log

cellular protein structures. He also reported at that time
that sensitivity of a bacterium to RH differed markedly
from organism to organism. 3
In the 1960s, more specific knowledge regarding the
effect of RH was attained. Webb and coworkers (40,41)
suggested that the death of airborne bacteria was the
DNA−
direct result of the loss of water, bound to nucleoproteins.
Anderson and Cox (27,42) published results supporting RNA−
that theory. They went on to propose that the structures of 2
macromolecules such as proteins, structural elements, and
nucleic acids are affected by water loss. They theorized that
the physical structure of proteins and deoxyribionucleic
acid (DNA) and ribonucleic acid (RNA) depend on water
content. However, Cox (27) stated that desiccation does not
markedly denature nucleic acid but instead inactivates
surface proteins. At low RH levels water loss might 1
increase the concentration of toxic compounds inside DNA+
the cell or crystallize them, potentially damaging the RNA+
metabolic processes of the airborne microbe (27,37,42).
The reactions of bacteria and viruses to the relative
humidity of the environment are different. This is most
likely due to the difference in structural makeup of the 0
bacterial cell wall and whether or not the viral envelope 0 10 20 30 40 50 60 70 80 90
contains lipids. Envelopes on viruses, which are composed %RH
of a phospholipid bilayer, with inserted protein molecules, Figure 2. Inactivation of aerosolized viruses after one hour of
are analogous to the outer phospholipid membrane of storage in aerosol at a given RH. DNA− are lipid free-enveloped
gram-negative bacteria and mycoplasmas (33). Generally, DNA viruses, DNA+ are lipid containing-enveloped DNA viruses,
these phospholipid membranes have an inherent thermo- RNA− are lipid free-enveloped RNA viruses, and RNA+ are lipid
dynamic instability that can lead to structural changes containing-enveloped RNA viruses.  log: the difference between
log of titre calculated and log of titre one hour after atomization.
with little agitation (33). For bacteria, sensitivity to stress
Reprinted from De Jong, et al. (1973) with the permission of the
appears to be related to gram stain reaction. Gram-positive publisher (47).
organisms are more stable in aerosols than gram-negative
organisms. Gram-positive organisms are impermeable to
ions and small molecules unlike gram-negative organ- the outer phospholipid bilayer leading to a cross-
isms, which may account for the differences in sensitivities linking reaction between protein moieties, thus enhancing
between the two groups (41). Interestingly, in one study it inactivation. This occurs mainly at mid to high RH
was observed that culturable gram-positive cocci outnum- levels (27). Conversely, the reactions of the surface
bered gram-negative organisms of any kind in samples proteins of lipid free enveloped viruses occur most rapidly
taken from urban air (43). at low RH levels. In bacteria, Cox (27,33) proposed that the
Gram-negative bacteria and viruses with envelopes lipoproteins found in gram-negative organisms denature
containing lipids demonstrate phase changes in the outer most readily at mid to high RH levels and other proteins
phospholipid bilayer membrane when exposed to various denature most readily at low RH levels.
RH levels (33). In viruses, those with envelopes containing Another inactivation mechanism is Maillard reac-
lipids survive better at a low RH levels and those viruses tions (37). Maillard reactions are reactions between lipids
without lipids in their envelopes survive better at a high and proteins or proteins and proteins and result in
RH(44–46) (Fig. 2). the removal of water. These reactions occur at low to
Israeli and coworkers (48) studied freeze-dried microor- medium (0% to 50%) RH levels and may possibly increase
ganisms and concluded that for many bacteria, the phos- with increasing temperature.
pholipid bilayer biomembrane undergoes conformational Webb (39) reported that the death of a bacterial
changes from the liquid-crystalline phase to gel phase as population in an aerosol proceeds in two stages. The
a result of water loss. The physical change of cell protein first stage is a rapid initial kill within the first second
conformation leads to a loss of viability of the airborne after aerosolization and the second stage is a subsequent
microbe. Theunissen and coworkers (37) also state that slower death. The research of many others (37,45,49–51)
the transition of lipids from a liquid-crystalline phase to supports Webb’s initial finding, in both aerosolized
a gel phase is a possible inactivation mechanism. They bacteria and viruses. Akers (45) expanded on Webb’s
believe this phase change would occur when the airborne theory by proposing that the quick initial death may be due
microbe is exposed to an environment of low temperature to the osmotic shock resulting from the humidity change.
and low RH. From the earlier discussion, it is apparent that RH
Cox (33) found that aerosolized gram-negative and directly plays an integral role in the survival and viability
lipid-containing viruses demonstrate phase changes in of airborne microbes. RH also affects the response of the
430 BIOAEROSOLS: TRANSPORT AND FATE

aerosol to other factors such as temperature and UV combination of RH and SO2 was greater than the sum
radiation (42). In a study with Mycoplasma pneumoniae, of the separate effects (45).
a high RH (95%) was observed to increase recovery Carbon monoxide (CO) has been observed to provide
by approximately ten percent after exposure to UV protection to aerosolized S. marcescens at 88% RH and
radiation (52). higher but at other RH levels, mainly below 75%, the death
rate was increased (57). The protection by CO at high RH
Oxygen and Pollutants levels may be due to the inhibition of a small energy-
requiring death mechanism or the rerouting of conserved
In 1914, it was observed that oxygen could cause a
energy to repair mechanisms, although the article did
loss of viability in freeze-dried bacteria (24). Enhanced
not describe or name the small energy-requiring death
bacterial survival was observed when oxygen was replaced
mechanism (57).
with nitrogen (42) or helium (24), resulting in inert
In a study with Flavobacterium, a dependence of nitrous
atmospheres.
dioxide (NO2 ) toxicity on the RH level was observed (58).
Oxygen inactivation is only seen in some species of
At lower RH levels, more NO2 was needed to produce
vegetative bacteria and algae (27). Oxygen susceptibility
equivalent losses that were observed at higher RH levels.
usually increases with the degree of desiccation of the
At 85% RH it is possible for atmospheric water to convert
aerosolized bacteria, increasing oxygen concentration, and
NO2 to nitric or nitrous acids, both of which are known to
the length of exposure time (27,48). The toxic effect
denature proteins (45).
of oxygen was observed to occur below 70% RH (53).
Atmospheric oxygen is not toxic to viruses in the aerosol
Open-Air Factor (OAF)
state (46). Figure 3 illustrates the results from a study
conducted in which Serratia marcescens was aerosolized The term open-air factor (OAF) is derived from the fact
into different combinations of nitrogen and oxygen at that significant increases in biological inactivation were
30% RH and 25 ° C (54). As the amount of oxygen in observed when aerosols were exposed to HEPA-filtered
the atmosphere decreased, the viability of S. marcescens outdoor air compared with clean, inert laboratory-supplied
increased. In this same study little or no toxic action of air (9).
oxygen was detected at RH levels above 65%. For many decades the causative agents of the OAF
Lighthart and coworkers (55) observed that the sulfur and the action of these agents have been speculated upon.
dioxide (SO2 ) found at urban air concentrations affects The death action of the OAF is still not fully understood.
the survival/viabilility of S. marcescens. Exposure of Currently, the OAF is thought to be the result of olefins
S. marcescens to urban levels of SO2 resulted in a marked reacting with the ozone (24,27,33). Olefins are unsaturated
decrease in viable bacteria. They also found that cells open-chain hydrocarbons that come from a variety of
in recently generated aerosols (i.e., aerosols aged 0 to sources. The question of why the OAF is observed on
1 hours) generally were more sensitive to SO2 , than cells bioaerosols outside and not inside is addressed by Cox (33).
in older aerosols (aerosols aged 1 to 5 hours). SO2 at However, the components of the OAF may require time to
3.6 ppm, in combination with simulated solar radiation, condense or buildup on the surface of the particle.
significantly reduced the viability of the Venezuelan Druett (59) observed, through experimentation, that
Equine Encephalitis (VEE) virus in aerosol (56). It was germicidal action arose from the reaction of the ozone
observed in one study that inactivation due to the with the olefin fraction of unburned fuel. Testing the OAF
effect on aerosolized microbes is very difficult because
the influence is rapidly lost once outside air is enclosed
% O2 in a test chamber (60). This loss is presumably due to
100 0 the adsorption of the ozone + olefin gaseous complex to
the walls of the holding container (60). Ozone + olefin
products have been reported (27,60) to have short half-
1.0 lives that may explain loss of reactivity as outside air
50
Percent viability

travels through pipes and particulate filters.

2.0 The concentration of OAF, at a given location depends
upon wind direction, factors influencing vertical mixing
of air masses, time of day, etc. (27). OAF toxicity is at
5.0
its greatest in air masses that have crossed urban and
industrial areas (27,59). In an OAF study of air passing
10.0 over industrial areas, urban areas, and the sea it was
20.9 found that the OAF was absent in air passing over the sea
0 and the highest microbial death rates were observed when
0 10 20 30 40 50 60 air passed over cities and industrial areas (61). Because of
Time (min) the many variables involved, the toxicity of OAF can vary
Figure 3. Aerosol survival (log scale) of S. marcescens 8UK from one day or minute to the next (62). The effect of the
sprayed different nitrogen + oxygen mixtures at 30% relative OAF is increased by high RH levels (33).
humidity and 25.0 ° C. Points are experimental; lines are The toxic effect of OAF on some bacteria and viruses
calculated. Reprinted from Cox, et al. (1974) with the permission results in the inactivation of coat constituents and nucleic
of the publisher (54). acids (27). The effectiveness of the OAF is due less to high
 BIOAEROSOLS: TRANSPORT AND FATE 431

toxicity than to the readiness of the components of the generated from the wet state (42). No explanation for this
OAF to condense on aerosol particles (27). effect has been determined. However, it has been shown
In some microbes, nucleic acids and coat proteins are that dry disseminated aerosols undergo a spontaneous
seriously affected by the OAF (33). However, microbes reactivation when held in the dark after UV radiation.
that tend to be resistant to UV radiation also appear to This may have something to do with the increased survival
be resistant to the OAF (33). Cox (33) believes this may be of bioaerosols generated from the dry state.
related to the lipid content, whereas, organisms without In a study in which airborne bacteria were collected
surface lipids have a greater resistance to OAF and UV continuously (64), microbes collected at night were more
radiation than those with surface lipids, namely, gram- sensitive to solar radiation than those collected on cloudy
negative bacteria and viruses with envelopes containing days. Bacteria collected during sunny days had the lowest
lipids. The causative virus of FMD, a lipid free enveloped sensitivity to solar radiation. As the solar radiation
virus, has been given as an example of a microbe that can intensity increased, the time to reach a lethal dose
travel great distances (at least 100 miles) while retaining decreased and cells died more rapidly. The organisms that
its viability and infectivity. This is possibly because of its were the most resistant to solar radiation were found in the
OAF resistance (33). upper atmosphere where the least amount of atmospheric
protection exists (i.e., ozone). However, these UV-resistant
Temperature organisms are rare.
The sole effect of temperature on aerosolized microbes Bacteria with pigmentation survive UV radiation
is difficult to ascertain because RH levels are related exposure better than those without pigmentation (43,65).
to environmental temperature. However, several studies When taking culture samples of indigenous background
with airborne viruses, in which the RH level was held microbes, pigmented bacteria are more prevalent than
constant, show the rate of aerosol decay increasing with nonpigmented bacteria (43).
increased temperature (10,24,42,45). In general, as the Berendt and coworkers (56,66), in the course of
temperature decreases the survival/viability of aerosolized their research, saw a reaction between sulfur dioxide
microbes increase. and UV radiation. They found that under certain
circumstances, sulfur dioxide (0.4 ppm concentration)
Repair reduced the virucidal properties of UV radiation. They
theorized that this effect was due to the radiation and the
According to Cox (33), the ability of microbes to survive the sulfur dioxide acting in an antagonistic manner, thereby
airborne state depends on their ability to repair damage. reducing a possible enhanced combined effect. Some of the
This ability depends on the chemical environment as wavelengths of solar radiation may be absorbed by sulfur
well as the genetic makeup of the microbe (27). Repair dioxide or sulfur dioxide may react with UV radiation,
of damage caused by solar radiation may be activated by leading to the production of sulfuric acid that has little
exposure to wavelengths of light longer than the damaging influence on viral inactivation (56,66).
wavelength (photo reactivation repair) or spontaneously
in the dark (dark repair) (63). This may account for the
rise in the airborne bacterial concentration seen in the Protection
late afternoon or after sunset. The survival of aerosolized microbes can be related
to the method of propagation (11), the spray fluid
UV Radiation (24,40,42,45,46,67,68), culture age (24,44,69–71), growth
Electromagnetic radiation can reduce the survival of media (24), and suspending media (40,45,71). For
microbial aerosols (42). Hypothetically, airborne inacti- instance, at high RH levels, it was observed that
vation of bacteria after UV light irradiation is due to rup- aerosolized E. coli survival was better if the organism
tured hydrogen bonds among water molecules or between was grown on agar rather than in liquid media (42). The
water molecules and cellular nucleoproteins or thymine survival response of an aerosolized virus to RH levels
dimers (42,45). UV radiation is also known to increase the is dependent greatly on the composition of the spray
mutation rate of bacteria in aerosols (42), possibly leading fluid (45,46,71).
to inactivation. The stage of growth of an organism also affects the
Many factors determine whether or not an aerosolized aerosol stability. Escherichia coli and S. marcescens both
microbe is sensitive to UV radiation. UV radiation effects exhibit increased decay rates when the organisms are
are dependent on bacterial or viral species, physiological harvested in the transition from lag to log phase, during
state, light intensity and duration, and wavelength (64). growth (42). Organisms harvested during the stationary
For example, many viruses (especially enveloped viruses) phase of growth survive the best (38,69,72). Aerosols
are sensitive to visible light of short wavelength (31), generated from E. coli harvested during the stationary
but a FMD virus is less sensitive (more resistant) to phase displayed a survival rate of 40%–50% in contrast
sunlight (30). to cells generated during the exponential or declining
Larger particles survive solar radiation better than growth stages in which survival rates of 6% and 1%
smaller particles (42). This may be because of the were respectively observed (70). Aged organisms have
protective action provided to organisms located in the a greater resistance to osmotic shock than log phase
interior of the particle. Bacterial aerosols generated from organisms (42,44). This may be because of the ‘‘mature’’
the dry state are less sensitive to radiation than those cells having low metabolic activity (70). Organisms that
432 BIOAEROSOLS: TRANSPORT AND FATE

grow slowly are more resilient than those that grow Table 4. Bioaerosol Stressors and Most Probable Target
quickly (24,70). Molecules
Adding a variety of compounds to either the growth Stress Most Probable Target Molecules
media or to the slurry before aerosolization, can affect
the survival of an aerosolized organisms. At low RH RH and temperature Outer membrane phospholipids,
levels, the addition of salts to the slurry has been proteins
shown to protect enteric viruses (73). Polyhydroxy and Oxygen Phospholipids, proteins
amino compounds (41,45) as well as proteins (45), and Ozone Phospholipids, proteins
Open-Air Factor Phospholipids, proteins, nucleic
sugars (24,27,33,38,42) can enhance aerosol survival of
(O3 + olefins) acids
bacteria and viruses. These compounds replace water γ -rays, X rays, UV radiation Phospholipids, proteins, nucleic
in protein structure during periods of desiccation. Raf- acids
finose and trehalose stabilize desiccated phospholipid
membranes and inhibit membrane fusion (27,33). Cox and Source: Table Reproduced from Cox (1989) with the permission of the
publisher (33).
coworkers (24,42) lists amino acids, antibiotics, aromatic
compounds, dyes, metal-chelating agents, polyhydric alco-
hols, salts, spent growth media, and sugars as protective
Although many of the processes involved with
additives. They state that relatively simple compounds
bioaerosol inactivation are unknown, some generalizations
such as inositol and sugars (di- and trisaccharides) offer
regarding these microbes can be made (37):
the best microbial protection over the widest range of
conditions. Israeli and coworkers (74) observed that tre-
1. Gram-negative bacteria survive best in conditions of
halose, added to drying media, protected biomembranes
low temperature and high RH levels.
and proteins from inactivation during the freeze-drying
2. Gram-positive bacteria survive better and longer
process. In their study with E. coli it was found that
than gram-negative bacteria in aerosols.
trehalose protected bacteria from known environmental
hazards such as visible light, air, and RH levels. They 3. The sensitivity of gram-negative bacteria to oxygen
stated that trehalose imparts protection by stabilizing the depends on species.
liquid-crystalline phase of the outer bacterial membrane, 4. Viruses with lipids in the envelope are more stable
probably replacing the water molecules in the phospho- than viruses without lipids in the envelope. Viruses
lipids. This prevents these molecules from denaturing, with lipid-containing envelopes survive better at RH
thus inactivating the bacteria. In a similar study con- levels less than 50%, whereas viruses without lipid-
ducted by Larson (75), using Francisella tularensis, it was containing envelopes survive better at RH levels
observed that the effect of RH levels on biological decay greater than 50%.
was essentially eliminated by the additives raffinose and 5. Normal survival characteristics can be altered by
dipyridyl. Inositol has been shown to provide a broad the addition of certain compounds.
range of protection from RH effects, specifically in aerosols
of E. coli and S. marcescens (76). Inositol in the absence It is obvious that the interactions affecting the
of water can form strong reversible hydrogen bonds with transport and fate of bioaerosol particles have many
cellular proteins, thereby stabilizing the protein struc- variables resulting in very complex relationships. Each
ture (41). Catalase added to collection media caused a of these variables affect the survival and viability of
significant increase (>63%) in the colony forming abilities the aerosolized microbe making it difficult to determine
of airborne bacteria (77). Catalase acts by removing or exactly what effect each variable actually has on the
degrading hydrogen peroxide that is lethal to cells. Cel- bioaerosol particle. Even the mechanisms and affects of
lular catalase activity is reduced by stress that results in factors that have been studied extensively are, in many
the accumulation of hydrogen peroxide. instances, not fully understood. Further research and
Webb found that some bacteria that possess antibiotic study is necessary to add to the body of knowledge that
resistance have a protein alteration that affords them currently exists. Advances in technology, which helps us
increased aerosol stability (41). He further discovered to design equipment and execute tests, make attaining
that challenging an aerosol-sensitive species to certain this information possible.
antibiotics resulted in adaptations in which the organism
became considerably more stable during air storage. Acknowledgments
Antibiotics that produced this effect were those that Special thanks to my mentor, Dr. Alan Jeff Mohr. Without his
interfered with protein synthesis. support and guidance this entry would not have been possible.

BIBLIOGRAPHY
CONCLUSION
1. R. Burrell, Environ. Health Perspect. 95, 29–34 (1991).
Table 4 is a summary of bioaerosol stressors and most 2. J. T. Bartlett, in J. F. Ph. Hers and K. C. Winkler, eds.,
probable target molecules. It is generally agreed upon Airborne Transmission and Airborne Infection, John Wiley
that the major stressor of any bioaerosol is desiccation but & Sons, New York, 1973, pp. 385–391.
it is difficult to isolate the effect of this stressor because of 3. S. R. Burger and J. W. Bennett, Appl. Environ. Microbiol. 50,
the complicating factors that are invariably present. 487–490 (1985).
 BIOAEROSOLS: TRANSPORT AND FATE 433

4. D. C. Blanchard and L. Syzdek, Science 170, 626–628 (1970). 35. M. K. Owen, D. S. Ensor, L. E. Sparks, Atmos. Environ. 26A,
5. T. W. Hejkal, P. A. LaRock, J. W. Winchester, Appl. Environ. 2149–2162 (1992).
Microbiol. 39, 335–338 (1980). 36. W. C. Day and R. F. Berendt, Infect. Immun. 5, 77–82 (1972).
6. E. R. Baylor, V. Peters, and M. B. Baylor, Science 197, 37. H. J. Theunissen et al., Appl. Environ. Microbiol. 59,
763,764 (1977). 2589–2593 (1993).
7. E. R. Baylor et al., Science 198, 575–580 (1977). 38. C. S. Cox, J. Gen. Microbiol. 43, 383–399 (1966).
8. M. Aubert and J. Aubert, in J. F. Ph. Hers and K. C. Winkler, 39. S. J. Webb, Can. J. Microbiol. 5, 649–669 (1959).
eds., Airborne Transmission and Airborne Infection, John 40. S. J. Webb, R. Bather, R. W. Hodges, Can. J. Microbiol. 9,
Wiley & Sons, New York, 1973, pp. 66–68. 87–92 (1963).
9. A. J. Mohr, in C. J. Hurst et al., eds., Manual of Environ- 41. S. J. Webb, Can. J. Microbiol. 6, 89–103 (1960).
mental Microbiology, ASM Press, Washington, D.C., 1997, 42. J. D. Anderson and C. S. Cox, in P. H. Gregory and J. L. Mon-
pp. 641–650. teith, eds., Aiborne Microbes, Cambridge University Press,
10. A. Bovallius et al., Appl. Environ. Microbiol. 35, 847–852 London, U.K., 1967, pp. 203–215.
(1978). 43. R. L. Mancinelli and W. A. Shulls, Appl. Environ. Microbiol.
11. E. C. Cole and C. E. Cook, Am. J. Infect. Control 26, 453–64 35, 1095–1101 (1978).
(1998). 44. K. C. Winkler, in J. F. Ph. Hers and K. C. Winkler, eds.,
12. J. C. Spendlove and K. F. Fannin, Public Health Rep. 98, Airborne Transmission and Airborne Infection, John Wiley
229–244(1983). & Sons, New York, 1973, pp. 1–11.
13. R. N. Berglund and B. Y. H. Liu, Environ. Sci. Technol. 7, 45. T. G. Akers, in J. F. Ph. Hers and K. C. Winkler, eds., Air-
147–153 (1973). borne Transmission and Airborne Infection, John Wiley &
14. M. W. van Vliet and J. J. C. Picot, Atomos. Spray Technol. 3, Sons, New York, 1973, pp. 73–81.
123–134 (1987). 46. J. E. Benbough, J. Gen. Virol. 10, 209–220 (1971).
15. H. W. Young et al., J. Clin. Microbiol. 5, 131–136 (1977). 47. J. C. De Jong, T. Trouwborst, K. C. Winkler, in J. F. Ph. Hers
16. K. H. Leong, J.Aerosol Sci. 17, 855–858 (1986). and K. C. Winkler, eds., Airborne Transmission and Airborne
Infection, John Wiley & Sons, New York, 1973, pp. 124–130.
17. J. P. Mitchell, J. Aerosol Sci. 15, 35–45 (1984).
48. E. Israeli, J. Gitelman, and B. Lighthart, in B. Lighthart and
18. B. Teltsch and E. Katzenelson, Appl. Environ. Microbiol. 35,
A. J. Mohr, eds., Atmospheric Microbial Aerosols, Chapman
290–296 (1978).
& Hall, New York, 1994, pp. 166–191.
19. J. C. Spendlove et al., APCA J. 31, (1981).
49. C. S. Cox, Appl. Microbiol. 21, 482–486 (1971).
20. A. Kerr et al., in Z. Klement, K. Rudolph, D. C. Sands, eds.,
50. A. I. Donaldson, J. Gen. Virol. 15, 25–33 (1972).
Methods in Phytobacteriology, Akademia Kiado, Budapest,
51. A. J. Mohr, Aerosol Stability of Reovirus, Utah State Univer-
Hungary, 1990, pp. 369–380.
sity, Logan, Utah, 1984.
21. J. L. Vanneste, J. Yu, and S. V. Beer, J. Bacteriol. 174,
52. D. N. Wright and G. D. Bailey, Can. J. Microbiol. 15,
2785–2796 (1992).
1449–1452 (1969).
22. S. A. I. Wright et al., Appl. Environ. Microbiol. 67, 284–292
53. C. S. Cox, M. C. Bondurant, and M. T. Hatch, J. Hyg. (Lon-
(2001).
don) 69, 661–672 (1971).
23. L. P. Kearns and H. K. Mahanty, Appl. Environ. Microbiol.
54. C. S. Cox, S. J. Gagen, and J. Baxter, Can. J. Microbiol. 20,
64, 1837–1844 (1998).
1529–1234 (1974).
24. R. E. Strange and C. S. Cox, in T. R. G. Gray and J. R. Post-
55. B. Lighthart, V. E. Hiatt, and A. T. Rossano Jr., J. Air Pollut.
gate, eds., The Survival of Vegetative Microbes, Cambridge
Control Assoc. 21, 639–642 (1971).
University Press, Cambridge, Mass., 1976, pp. 111–154.
56. R. F. Berendt, E. L. Dorsey, and H. J. Hearn, Proc. Soc. Exp.
25. M. V. Walter et al., Appl. Environ. Microbiol. 56, 3468–3472
Biol. Med. 139, 1–5 (1972).
(1990).
57. B. Lighthart, Appl. Microbiol. 25, 86–91 (1973).
26. A. Bovallius, R. Roffey, E. Henningson, Ann. N. Y. Acad. Sci.
353, 186–200 (1980). 58. M. A. Chatigny et al., in J. F. Ph. Hers and K. C. Winkler,
eds., Airborne Transmission and Airborne Infection, John
27. C. S. Cox, The Aerobiolgical Pathway of Microorganisms,
Wiley & Sons, New York, 1973, pp. 94–97.
John Wiley & Sons, London, U.K., 1987.
59. H. A. Druett, in J. F. Ph. Hers and K. C. Winkler, eds.,
28. J. C. Zadoks, in J. F. Ph. Hers and K. C. Winkler, eds., Airborne Transmission and Airborne Infection, John Wiley
Airborne Transmission and Airborne Infection, John Wiley & Sons, New York, 1973, pp. 141–149.
& Sons, New York, 1973, pp. 392–395.
60. J. E. Benbough and A. M. Hood, J. Hyg. (London) 69,
29. R. F. Sellers et al., in J. F. Ph. Hers and K. C. Winkler, eds., 619–626 (1971).
Airborne Transmission and Airborne Infection, John Wiley &
61. G. De Mik and I. De Groot, in J. F. Ph. Hers and K. C. Wink-
Sons, New York, 1973, pp. 405–412.
ler, eds., Airborne Transmission and Airborne Infection, John
30. A. I. Donaldson and N. P. Ferris, J. Hyg. (London) 74, Wiley & Sons, New York, 1973, pp. 155–158.
409–416 (1975).
62. G. J. Harper, in J. F. Ph. Hers and K. C. Winkler, eds.,
31. N. S. Hyslop, Trop. Anim. Health Prod. 4, 28–40 (1972). Airborne Infection and Airborne Transmission, John Wiley
32. A. Bovallius et al., Appl. Environ. Microbiol. 35, 1,231,1,232 & Sons, New York, 1973, pp. 151–154.
(1978). 63. B. Lighthart and B. T. Shaffer, Atmos. Environ. 28, 1,267–
33. C. S. Cox, Sci. Prog. 73, 469–499 (1989). 1,274 (1994).
34. G. W. Hammond, R. L. Raddatz, D. E. Gelskey, Rev. Infect. 64. Y. Tong and B. Lighthart, Atmos. Environ. 31, 897–900
Dis. 11, 494–497 (1989). (1997).
434 BIOAUGMENTATION

65. B. T. Shaffer and B. Lighthart, Microb. Ecol. 34, 167–177 their concentrations) and the hydrogeology of the site,
(1997). as well as nontechnical driving forces such as regulatory
66. R. F. Berendt, E. L. Dorsey, and H. J. Hearn, Proc. Soc. Exp. demands, property owner interests, and ultimately the
Biol. Med. 138, 1005–1008 (1971). cost of cleanup in terms of both money and time.
67. G. E. Hess, Appl. Microbiol. 13, 781–787 (1965). The remediation of sites contaminated with chlorinated
68. F. L. Schaffer, M. E. Soergel, and D. C. Straube, Arch. Virol. volatile organic compounds (CVOCs) has been a focus
51, 263–273 (1976). area for the development of bioaugmentation technologies.
69. A. D. Brown, Aust. J. Biol. Sci. 7, 444–451 (1954). CVOCs are a family of chemicals that have been used
70. F. A. Dark and D. S. Callow, in J. F. Ph.Hers and K. C. Wink- extensively as industrial solvents and cleaning agents,
ler, eds., Airborne Transmission and Airborne Infection, John and include perchloroethylene (PCE), trichloroethylene
Wiley & Sons, New York, 1973, pp. 97–99. (TCE), dichloroethylene (DCE), vinyl chloride (VC),
71. G. J. Harper, Arch. Ges. Virus-Forsch 13, 64–71 (1963). trichloroethane (TCA) dichloroethane (DCA), carbon
72. P. Hambleton et al., J. Hyg. (London) 90, 451–460 (1983). tetrachloride (CT), and chloroform (CF). The widespread
73. D. J. Adams et al., Appl. Environ. Microbiol. 44, 903–908 use of CVOCs, improper disposal practices, and their
(1982). chemical properties and stability have led to them
74. E. Israeli, B. T. Shaffer, and B. Lighthart, Cryobiology 30, becoming common groundwater contaminants throughout
519–523 (1993). the United States (1).
75. E. W. Larson, in J. F. Ph. Hers and K. C. Winkler, eds.,
Airborne Transmission and Airborne Infection, John Wiley REMEDIATION HIERARCHY
& Sons, New York, 1973, pp. 81–86.
76. L. J. Goldberg and I. Ford, in J. F. Ph. Hers and K. C. Wink- Within the group of in situ bioremediation technologies
ler, eds., Airborne Transmission and Airborne Infection, John
there has developed a hierarchy of alternative remedies
Wiley & Sons, New York, 1973, pp. 86–89.
based on ease of cleanup and cost. The first level
77. B. Marthi et al., Appl. Environ. Microbiol. 57, 2775–2776
of the hierarchy is intrinsic biodegradation whereby
(1991).
indigenous microflora destroy the contaminant of concern
before it creates a significant risk to downgradient
receptors. This has become a natural first choice of
BIOAEROSOLS, VIRUS. See VIRAL AEROSOLS remediation alternatives where applicable because it
requires no intervention; just monitoring of contaminant
concentrations and modeling of the groundwater flow and
natural degradation rates.
The second choice in the hierarchy, biostimulation
BIOAEROSOLS: WASTEWATER TREATMENT involves the stimulation of indigenous microbial popu-
PLANTS. See WASTEWATER AND BIOSOLIDS AS SOURCES OF lations to allow them to destroy the target chemicals. In
AIRBORNE MICROORGANISMS this case, the observation is made that a natural popula-
tion exists within the contaminated zone; however, specific
nutrients, growth substrates, inducers, and/or oxygen are
insufficient for microbial activity. Thus, through the intro-
duction of the correct co-substrate, the native degradative
BIOAUGMENTATION population can be stimulated to grow and destroy the
target contaminant.
MARY F. DEFLAUN In cases in which natural attenuation or biostimulation
ROBERT J. STEFFAN do not work because of insufficient or unacclimated
Envirogen, Inc. indigenous microflora, bioaugmentation of the subsurface
Lawrenceville, New Jersey is an option. Selected strains of bacteria with the desired
catalytic capabilities can be injected directly into the
Bioaugmentation is the addition of exogenous organisms contaminated zones along with any required nutrients
to an environment to complement or replace an indigenous to effect the biodegradation of the target chemicals.
population. In the context of remediating polluted Two distinct approaches have developed in the area of
environments, bioaugmentation typically involves the bioaugmentation for remediation of CVOCs. In the first
addition of bacteria that have been selected for their ability approach (Type 1), degradative organisms are added to
to transform a target contaminant. In the last several complement the native microbial population. The added
years, bioaugmentation for pollution remediation has microorganisms can be selected for long-term survival
become increasingly popular. Consequently, the relative and the ability to occupy a selective niche within the
conceptual simplicity of the technology has spawned contaminated environment, and stimulants or selective
an industry involved in the sale of mixed or pure co-substrates, can be added to aid survival. Thus, the goal
cultures of bacteria with the promised ability to degrade of the Type 1 approach is to achieve prolonged survival
nearly any contaminant in virtually any environment. In and growth of the added organisms and the concomitant
actuality, the rational selection of bioaugmentation as a prolonged degradation of the target pollutants.
remedial alternative for a given site requires a detailed In the second approach (Type 2), large numbers
evaluation of the pollution profile (i.e., chemicals and of degradative bacteria are added to a contaminated
 BIOAUGMENTATION 435

environment as catalysts that will degrade a significant Furthermore, because the oxidation of CVOCs does not
amount of the target contaminant before becoming allow regeneration of sufficient energy to continue the
inactive or perishing. Attempts can be made to increase oxidation process, supplemental energy substrates are
the expression of the degradative enzymes or to maximize needed to prolong degradation. Finally, the production
catalytic efficiency or stability, but the long-term survival, of oxygenase enzymes known to transform CVOCs is
growth, and establishment of the biocatalyst are not typically tightly regulated at the genetic level. That is,
the primary goals of the treatment approach. Additional degradative genes must be induced, and the inducers
microbial catalysts can be added as needed to further are typically the natural substrate for the enzyme. For
the degradation process. Although this review focuses example, CVOC-degrading toluene monooxygenase genes
on bioaugmentation for the remediation of chlorinated are induced by toluene. Thus, in addition to supplying
solvents, in particular TCE, examples of field-scale a substrate for energy production, an inducing substrate
applications of bioaugmentation for other compounds are must also be available.
discussed. To circumvent these limiting characteristics of natural
CVOC-degrading microorganisms, the organisms may be
ORGANISMS FOR BIOAUGMENTATION modified to improve their utility for bioaugmentation.
These modifications can often be accomplished by chemical
Several classes of microorganisms can degrade TCE under mutagenesis and/or natural selection, or they can be made
the proper conditions (2). TCE can be biodegraded by aer- by using recombinant DNA techniques.
obic bacteria, which oxidize aromatic hydrocarbons (3–6),
methane (methanotrophs; 7–10), ammonia (11), and
propane (12). All of these bacteria initiate the degrada- CONSTITUTIVE DEGRADATIVE ACTIVITY
tion of growth substrates and CVOCs by incorporating
atmospheric oxygen through the action of enzymes known A significant breakthrough in the development of organ-
as oxygenases, and in each case the degradation appears isms for remediating CVOC-contaminated aquifers has
to be co-metabolic; that is, the degradative organism does been the selection of organisms that constitutively produce
not gain sufficient metabolic energy from the oxidation of degradative enzymes. In these organisms degradation of
the CVOC to support growth of the organisms, but rather, the target pollutant is uncoupled from growth of the organ-
the CVOCs are fortuitously oxidized by enzymes that have isms, thereby allowing the organisms to grow on alterna-
evolved to degrade other substrates. Furthermore, the tive substrates while degrading the pollutant. Problems
oxygenase enzyme systems typically are only induced by of competitive substrate inhibition or the requirement for
growth in the presence of the specific hydrocarbon (e.g., using toxic or explosive co-substrates can potentially be
toluene-4-monooxygenase is induced by toluene). eliminated.
Anaerobic microorganisms can reductively dechlorinate In the first example of creating organisms that degrade
and ultimately mineralize many of the CVOCs. This can CVOCs constitutively, Shields and Reagin (1992) (21) used
occur co-metabolically or by dehalorespiration, a process transposon mutagenesis to generate insertional toluene-
that involves the chlorinated compound acting as the orthomonooxygenase (TOM) mutants of Burkholderia
electron acceptor (13). Dehalorespiration in particular can cepacia G4, and then selected spontaneous revertants of
effect dehalogenation at relatively high rates (14). Some the mutants that expressed TOM constitutively. TOM
of the lower chlorinated CVOCs can also be used as is an enzyme that catalyzes the degradation of certain
sole sources of carbon and their degradation does not CVOCs. One resulting variant of this selection process,
require a co-substrate. There is only one strain that has B. cepacia PR1, degraded TCE in the absence of inducing
been identified to date, that can degrade PCE all the substrates such as phenol and toluene, and was extremely
way to ethene (15). In general, the stepwise pathway stable over more than 100 generations of growth. In
of degradation is mediated by a number of different further studies by the same group (22) a similar variant,
bacteria. A number of these consortia have been enriched B. cepacia PR1301 , was developed by traditional chemical
from environments in which reductive dechlorination is mutagenesis. The latter strain is more applicable for
occurring (16–19). bioaugmentation because it is not considered genetically
engineered, and it has been evaluated in a series of
IMPROVING ORGANISMS FOR BIOAUGMENTATION laboratory and field demonstrations of bioaugmentation
(see below).
The successful application of bioaugmentation for reme- In similar studies, a constitutive soluble methane
diation of contaminated aquifers is limited by several monooxygenase (sMMO) variant of Methylosinus tri-
physiological characteristics of CVOC-degrading organ- chosporium OB3b was developed by chemical mutagenesis
isms. For example, many bacteria in nature produce a and revertant selection (23). The resultant strain pro-
glycocalyx layer and/or a variety of surface structures or duced sMMO, a broad-substrate range CVOC-degrading
polymers that aid in their adhesion to surfaces (20). As a enzyme, even in the presence of relatively high copper
result, bacteria tend to be very adhesive and they resist concentrations, and during growth on methanol. Consti-
migration through tight matrices such as those existing tutive variants of CVOC-degrading phenol hydroxylase-
in aquifer soils. Also, the initial oxidation of CVOCs is producing organisms (24), toluene oxidizers (25) and
an energy and reductant requiring process, and thus, co- isopropylbenzyne-degrading organisms (26) have been
factors such as NADH are required for degradation (2). developed by using various recombinant DNA techniques.
436 BIOAUGMENTATION

Like B. cepacia PR1301 , any of these mutants could poten- The foams flow effectively around heterogeneities in the
tially be added to contaminated aquifers and fed nonin- aquifers, and remove more DNAPL than either water or
ducing, noncompetitive, growth substrates to prolong their liquid surfactant. By incorporating strain ENV435 into
degradative activity in situ. the foam, the strain was distributed evenly throughout a
In related work, McClay and coworkers (1995) (27) model aquifer, and it resulted in significantly greater TCE
showed that TCE could act as an inducer of TCE removal than treatment with foam only. Incorporating
degradation activity in the toluene-oxidizing bacteria growth substrates, important nutrients, inducing sub-
P. mendocina KR1 and strain ENVPC5. Likewise, if P. strates, or oxygen-generating compounds into the foam,
putida F1, P. pickettii PKO1, or B. cepacia G4 were grown even greater degradation might be expected.
in the presence of TCE, the toluene oxidation genes Liquid surfactants and low ionic strength solutions also
were induced (28), and a wild-type toluene dioxygenase- have been tested for enhancing transport of bacteria for
producing P. putida degraded TCE in the absence bioaugmentation. Early work using glass beads as a model
of inducing substrates (29). Either of these strains porous media indicated that a significant increase (0.16
could potentially be used for bioaugmentation where to 60 m) in bacterial transport would result from some
noninducing substrates are added to support degradative of these chemical treatments (39). However, subsequent
activity and TCE acts as the inducing substrate. Again, work using the same bacteria and chemical additions, but
competitive inhibition between the growth substrate and with soils as the porous media, resulted in only moderate
the target CVOCs should be limited, but the added increases in transport (40). These moderate increases in
substrate could increase competition from nondegrading transport would be sufficient to create small bioactive
organisms. zones (∼1 m), but would not be sufficient for dispersion of
the bacteria throughout a plume.
Other efforts to enhance bacterial transport have shown
ENHANCED TRANSPORT
that under conditions of controlled starvation bacterial
cells become significantly smaller than active vegetative
Adhesion-deficient variants of CVOC-degrading bacteria
cells. These ‘‘ultramicro’’ bacteria produce less of a biofilm
should allow improved distribution of the biocatalysts
than vegetative cells and are able to penetrate deeper into
throughout contaminated aquifers, and minimize prob-
consolidated materials than full-size cells (41,42). Once in
lems of wellhead plugging during injection of organisms.
place, the cells can be revived by adding the appropriate
Previous studies have demonstrated the feasibility of
nutrients and/or carbon source. A similar approach could
generating adhesion-deficient variants of a number of dif-
be applied with spore-forming cells (43).
ferent bacterial strains by transposon mutagenesis, chem-
Others have evaluated the use of encapsulation to
ical mutagenesis, or by selection of naturally occurring
improve survival and transport of bacteria added to the
adhesion-deficient variants from a population (30–35).
environment. In a study by Crawford and coworkers,
In one study, greater than 91% of a highly adhesive
the encapsulation of a PCP-degrading Flavobacterium
strain of Pseudomonas fluorescens (Pf0–1) was consis-
improved its survival in groundwater (44). In this study,
tently retained by a 3 cm tall, 12 g sand column (36). With
encapsulation procedures were modified from previously
this degree of attenuation, in situ transport would be lim-
used techniques in order to create much smaller capsules
ited to approximately 0.2 m in an unconsolidated sand
that would both improve oxygen diffusion to the cells and
aquifer. However, an adhesion-deficient variant (Pf0–5)
improve the transport of the encapsulated bacteria in the
with 40% retention in the same sand column would be
subsurface.
capable of traveling approximately 1 m in a sandy aquifer.
By using this simple adhesion assay, transposon-
generated adhesion-deficient mutants and natural vari- ENERGY ENRICHMENT
ants of P. cepacia G4, M. trichosporium OB3b, P. cepacia
PR1301c , P. mendocina KR1, and P. cepacia ENV BF1 All known aerobic CVOC-degrading microbes require a
have been selected (37). Some of the variants adhere to cosubstrate for sustained enzyme activity and metabolic
the sand columns at less than 10%, while the wild-type survival. However, adding a cosubstrate to an aquifer
strains adhere at greater than 90%. Model aquifer studies can stimulate the growth of other non-TCE-degrading
have demonstrated that several of these strains migrate organisms and create problems with biofouling and oxygen
through a 25 cm sand column at approximately the same depletion. Furthermore, some of the desired cosubstrates
rate as a conservative tracer. Thus, these metabolically for CVOC degradation may be explosive (e.g., methane,
proficient, adhesion-deficient strains are more suitable for propane, or butane) or toxic (e.g., phenol, toluene, or
in situ applications than the wild-type organisms because isopropylbenzene) and their use may be regulated or
they are less likely to plug the injection well during appli- require the use of expensive equipment or excessive
cation and they should travel farther into the subsurface, monitoring.
thereby increasing the effective zone of remediation. An alternative approach for enhancing and maintaining
In a related approach, Rothmel and coworkers biological activity in situ is to utilize biocatalysts
(1998) (38) used surfactant foams to distribute strain that are enriched in energy reserves. The production
ENV435 throughout a model aquifer. Because of their of energy storage polymers, most commonly poly-beta-
viscosity and scouring properties, surfactant microbubble hydroxybutyric acid (PHB), by bacteria is a long-studied
foams (a.k.a. colloidal gas aphrons) are useful for remov- phenomenon (45). PHB is produced naturally by many
ing dense nonaqueous liquids (DNAPLs) from aquifers. bacteria, typically under conditions of a nutrient limitation
 BIOAUGMENTATION 437

and the presence of excess carbon, and it may account for often for destroying targets that are only degraded co-
up to 80% of the bacterial cell dry weight. Utilization of metabolically. In the latter case, an additional substrate
PHB as a reducing power substrate by methanotrophic must be added to support the growth and degradative
bacteria degrading TCE has been demonstrated (46,47). activity of the added organism.
Likewise, cells containing PHB may survive longer in A laboratory study performed by Krumme and col-
groundwater than cells without PHB (48). Thus, selecting leagues (53) compared the TCE degrading activity and
a degradative organism that produces storage polymers survival of wild-type B. cepacia G4 (54), with that of a
may be advantageous over selecting an organism that does transposon-generated mutant, strain PR1 that consti-
not produce such polymers. In work performed by Steffan tutively expresses the TCE-oxidizing enzyme toluene-o-
and coworkers (1999) (49), B. cepacia ENV435 was grown monooxygenase (TOM). Both strains survived for extended
to optical densities (O.D.550 ) of greater than 65 (100 g/L periods (>10 weeks) in sterile aquifer material, but the
wet weight) and so they contained as much as 60% dry population of each strain decreased by about three orders
weight of PHB. of magnitude during the same period. When PR1 was
An additional advantage of using energy-enriched added to microcosms at an initial cell concentration of
organisms for in situ remediation, versus feeding indige- 5 × 108 CFU/g of aquifer material, it degraded TCE from
nous or introduced organisms, may be that it forces more 60 µM to less than 0.1 µM within 24 hours. When PR1
efficient utilization of added oxygen and inorganic nutri- was added at lower cell densities, it continued to degrade
ents. The degradative organisms carry an internal food TCE over the entire 10-week incubation, but the observed
reserve with them into the aquifer. The food reserve is not degradation rates were low relative to those observed with
available to less efficient indigenous organisms and thus the high cell density inoculum.
should not stimulate increased oxygen demand or biofoul- In a related laboratory study, Munakata-Marr and
ing associated with feeding indigenous microbes. The only coworkers (1996) (22) coupled bioaugmentation with
additional oxygen that may be needed is that required biostimulation in aquifer microcosms. Aquifer microcosms
to support oxidation of target contaminants and cellular were augmented with either B. cepacia G4 or a chemically
respiration of the added strain. induced mutant of the strain, B. cepacia PR1301 , which
constitutively produces TOM. The augmented microcosms
MIXED CULTURES AND CONSORTIA were then supplied with either phenol or sodium lactate,
and TCE degradation in the microcosms was compared
Most commercial-scale bioremediation applications of to degradation in microcosms that did not receive the
bioaugmentation involve the addition of mixtures or degradative organisms.
consortia of bacteria. This is particularly true in the area of When strain G4 and phenol were added in small
hydrocarbon treatment where a number of vendors market columns to sterile aquifer material containing 250 µg/L
products composed of uncharacterized mixtures of bacteria TCE, there was no breakthrough of TCE during the first
that have been cultured and prepared for application in 37 days of incubation and several exchanges of TCE-
the field. Few mixed cultures or consortia of organisms contaminated water. Without the addition of G4, TCE
have been developed for remediation of CVOCs. This may degradation was not observed until approximately 19 days
be due, in part, to the fact that bacteria are not known to after biostimulation with phenol. The TCE concentration
use CVOCs as a growth substrate. Therefore, it is difficult in microcosms augmented with strain PR123 and lactate,
to enrich degradative organisms by using target CVOCs however, were not significantly different than in sterile
as a selective substrate. It may be feasible, however, to controls. These results suggested that bioaugmentation
create stable mixtures of CVOC-degrading organisms by could be effective, but that in this system, the constitutive
enrichment on substrates known to co-metabolize CVOCs. expression of TOM during lactate feeding was not an
Such stable mixtures have been maintained for extended effective mechanism for maintaining degradative activity.
periods on methane, propane, toluene, and phenol (50–52) In related experiments, microcosms were augmented
in the laboratory. Likewise, the stimulation of indigenous continuously with lower concentrations of degradative
microbial communities by adding selective substrates to organisms (22). Bioaugmentation with phenol-grown G4
a contaminated environment undoubtedly selects for the alone generated TCE removals similar to those observed
enrichment of stable consortia that can sustain prolonged in microcosms fed only phenol. However, if lactate was
co-metabolic degradation of CVOCs. The use of such added with the phenol-grown G4, TCE degradation was
mixtures in bioaugmentation may provide more stable 70% greater than in microcosms receiving only G4 or
populations of organisms to help prolong degradation only phenol. Similarly, if strain PR1301 was added to
in situ. microcosms with phenol, TCE degradation was about
twice that observed in phenol-only microcosms, and it
proceeded without a lag period. In each case, the addition
EXAMPLES OF TYPE 1 BIOAUGMENTATION
of organisms resulted in a significant oxygen demand that
could potentially limit the effectiveness of the technology.
Type 1 bioaugmentation involves the addition of degrada-
tive organisms to a contaminated environment with the
Field Application — TCE
goal of establishing an active degradative microbial popu-
lation. Although this technique has been used extensively In an early in situ bioaugmentation field study, B. cepacia
for destroying contaminants that can act as growth sub- G4, was injected directly into a TCE-contaminated
strates for the added organisms, it has been applied less sandy aquifer with tryptophane added as an inducing
438 BIOAUGMENTATION

substrate (55). Data from the study strongly suggested did not prevent it from being rapidly outcompeted by
that, although the organisms degraded TCE, movement of native microbes.
the microorganisms was severely retarded by the aquifer The oxidation of CVOCs might further hinder the
material, thereby limiting the effective treatment area. ability of an added strain to survive in a natural
Tracer test results estimated the hydraulic flow of the environment. The oxidation of CVOCs is known to result
aquifer to be approximately 48 feet per day. Nonretarded in toxicity to the degradative organisms, presumably due
organisms, therefore, should have appeared in the nearest to the production of highly reactive epoxide molecules
down gradient recovery well (10 ft.) about 8 to 10 hours that can react with and harm cellular components causing
after injection, but none of the injected microorganisms cellular inhibition (50,57,58). When four strains of toluene-
were observed in the monitoring well until six days oxidizing bacteria were placed in a reactor with a low feed
after injection. No injected organisms were detected at rate of toluene, each strain was maintained at relatively
a monitoring well located 75 feet downgradient from the constant levels (59). When TCE was added to the reactor,
injection well. Thus, it appears as though the injected however, P. putida mt-2, which does not degrade TCE
cells acted essentially as an in situ biofilter. Clearly more rapidly, became the dominant organism in the reactor.
aquifer material could be treated if the organisms or the Kinetic analysis of the three strains that perished revealed
aquifer conditions were altered to reduce adsorption and that their affinity for toluene was greatly reduced after
retardation of bacterial movement. exposure to TCE. Remarkably, this same selection did not
Despite the poor distribution of the biocatalysts,
appear to occur when toluene was used to stimulate TCE
TCE degradation by the injected strain G4 reduced
degradation during a prolonged field study, as indicated by
groundwater TCE concentrations from 2,500 µg/L to
continuous TCE degradation for more than one year (60).
466 µg/L for a period of about eight hours. TCE
In a related study, the addition of toluene or phenol to a
concentrations increased, however, to 3,280 µg/L the
TCE-contaminated aquifer resulted in the selection of a
following day, presumably due to a nutrient feed pump
native toluene oxidizing bacterial population in which 55%
failure. After another eight hours, TCE concentrations
of the strains contained toluene-ortho-monooxygenase,
again declined and remained below 300 µg/L with a mean
TCE concentration of 135 ± 72 µg/L over the first five which is known to oxidize TCE. A more detailed analysis
days following G4 injection. The mean TCE concentration of the native strains revealed that a significant number
during the following 10 days was 78 ± 64 µg/L. of the strains (20%) were poor TCE degraders, and none
Although the authors did not estimate the total amount degraded TCE as well as strain G4 (61). Nonetheless,
of TCE degraded, their data suggests that degradation was CVOC degradation in the aquifer was relatively constant
significant. The groundwater flow rate was 48 ft/day and through several months of biostimulation (62).
the receiving zone (area in which bacteria were injected) An area that has not been well investigated is
was 20 ft wide and 14 ft thick with a porosity of 25%. The the selection of organisms with an inherent selective
authors thus concluded that approximately 25,000 gallons advantage for long-term survival, or the development
of groundwater went through the test area per day. The of organisms, which have a selective advantage or that
average decrease in TCE concentration during the first can be provided with an environment that gives them a
five days (2,500 µg/L to 135 µg/L) represents an apparent selective advantage. For example, carbon tetrachloride
TCE removal of about 1,100 g TCE. However, because (CT) degradation by Pseudomonas sp. strain KC was
the authors estimated that the nutrient addition system found to be greater under denitrifying conditions than
diluted the groundwater by 30%, the apparent amount aerobic conditions (63,64). Furthermore, degradation was
of TCE degraded was approximately 800 g. Applying the inhibited by iron, but by adjusting the pH of growth media
same calculation for data obtained during the subsequent to 8.3, conditions under which iron is precipitated, this
10 days during which TCE declined to an average of inhibition could be overcome (65). Thus, strain KC had
78 µg/L, the apparent amount of TCE degraded during a competitive advantage over other aquifer organisms if
the 10-day period was 1,600 g. Thus, the total amount grown under nitrate-reducing conditions at high pH. By
of TCE degraded during the 15 days could have been as adjusting the pH and redox potential of a contaminated
much as 2,400 g. aquifer, one could create a selective niche for added
strain KC. These characteristics allowed Criddle and
Survival of Introduced Bacteria coworkers (66) to demonstrate the use of bioaugmentation
The ability of an exogenous organism to survive and for the remediation of a CT-contaminated aquifer in School
compete for resources against indigenous organisms has Craft, Michigan.
long been an area of interest in both macro- and In the School Craft demonstration, NaOH-amended
microbial ecology, and in most cases added organisms groundwater was used to maintain slightly alkaline con-
do not fare well. When three well-characterized toluene- ditions (pH>7.6), KC was injected, and acetate was added
degrading bacteria, P. putida PaW1, B. pickettii PKO1, in weekly pulses to maintain degradative activity. During
and B. cepacia G4 were added to a fluid bed bioreactor a period of poor chemical delivery incomplete degradation
with toluene as a feed source, strain PaW1 became of CT occurred and chloroform concentrations increased.
the predominant organism in the reactor (56). When However, data from downgradient monitoring wells indi-
groundwater strains were allowed to enter the reactor, cated that as long as the appropriate amendments were
however, even strain PaW1 was rapidly replaced. Thus, maintained, bioaugmentation was effective for CT reme-
even precolonization of the reactor with the added strain diation (66).
 BIOAUGMENTATION 439

In a related approach, Lajoie and coworkers (1992) (67) Anaerobic Bioaugmentation

isolated a bacterial strain, termed ‘‘field application
For some sites and contaminants anaerobic degradation is
vector,’’ that was resistant to a surfactant that it could
more applicable than aerobic. This may be because one or
also use as a carbon source. By cloning degradative
more of the contaminants only degrades anaerobically
genes into the resistant organism, the researchers could
or because the site is already anaerobic and adding
add surfactants to an environment to create a selective enough oxygen to reverse this condition is technically
niche, and then add the resistant strain containing the infeasible. PCE only degrades anaerobically to TCE and
degradative genes (68). Such strains may be of use in by successive dechlorinations these chlorinated solvents
bioaugmentation of aquifers during or after surfactant can be degraded to the innocuous product ethene. If
or foam flushing designed to remove free product PCE is present, or if the aquifer is already anaerobic,
contaminants (69,70). anaerobic biostimulation is a viable remedial option. Often
the addition of an electron donor (substrate) will be
Field Application — MTBE sufficient to stimulate the complete degradation of the
higher chlorinated solvents to ethene, especially if there
Type 1 bioaugmentation is most applicable to contami- is some evidence of dechlorination already occurring. In a
nants that can serve as a sole source of carbon and energy previously aerobic aquifer at Dover Air Force Base, DE, the
for the degradative strain so that the addition of a carbon addition of a substrate (lactate) and nutrients (nitrogen
source is not necessary for growth and maintenance of and phosphorus) were successful in turning the aquifer at
the added strain in situ. Methyl tert-butyl ether (MTBE), the test site anaerobic and also stimulating the anaerobic
a component of reformulated gasoline, is a contaminant dechlorination of TCE to cis-1,2 dichloroethene (cDCE).
that can serve as a sole substrate for bacteria (71–73), Further dechlorination, however, was not observed until
but these bacterial strains are not ubiquitous. Biostim- the site was inoculated with an ethene-forming microbial
ulation approaches for MTBE degradation have utilized consortium (78). A lag period of 90 days occurred between
propane to stimulate indigenous propane-oxidizing bacte- injection of the bacteria and the first appearance of VC, the
ria to degrade MTBE co-metabolically, but this approach is degradation product of cDCE, and an additional 60 days
not applicable to every site (74). Bioaugmentation, there- before complete conversion to ethene. This lag period can
fore, has been utilized to remediate this highly soluble be attributed to the small amount of bacterial inoculum
contaminant. (<35 g dry weight) and the slow growth rate of this culture.
Three demonstrations of this technology have been
conducted at Port Hueneme Naval Facility in Oxnard, EXAMPLES OF TYPE 2 BIOAUGMENTATION
California. In the first demonstration, injection of a
microbial consortium (MC-100) along with oxygenation Type 2 bioaugmentation involves the addition of degrada-
of the aquifer, was compared to oxygenation only and tive organisms as biocatalysts that can degrade a signif-
no treatment, in three adjacent test plots (75). Although icant amount of target compound and then perish. The
oxygenation alone appeared to stimulate indigenous assumptions are made that the added organisms will not
MTBE-degrading strains after a considerable lag time, the compete well against native organisms, and/or that the
plot with added bacteria had a faster and greater decrease added organisms can be grown more efficiently ex situ
in MTBE concentrations than in the oxygen-only plot. (e.g., in a fermentor) than in situ.
Two other demonstrations are currently being conducted
with pure cultures of MTBE-degrading bacteria (76,77). In Situ Biofilter
Strain PM-1 was injected into a test plot in November
In one field-scale demonstration of Type 2 bioaugmen-
1999 next to an oxygen-only plot (77). Difficulties in
tation (79), approximately 5.4 kg (dry weight) of the
oxygen delivery and degradation of MTBE by indigenous CVOC-degrading methanotroph Methylosinus trichospo-
populations stimulated by the oxygen injection have made rium OB3b were injected into a TCE-contaminated aquifer
data interpretation difficult, and the contribution by PM1 near Chico Municipal Airport in Chico, California. The
to degradation of the MTBE in this demonstration has cells were used to establish an in situ biofilter (a.k.a.
yet to be determined. Strain ENV425 is a propane- biocurtain or biobarrier) of attached resting state cells
oxidizing bacteria that degrades MTBE co-metabolically. (i.e., the cells were not fed after injection) through
Bioaugmentation with this strain coupled with oxygen which contaminated groundwater was directed. Analy-
and propane injection is also being tested at Port sis of groundwater extracted from the aquifer indicated
Hueneme (76). that approximately 50% of the bacteria injected into the
Hydrogenophaga flava ENV735 is an isolate that aquifer were extracted by the pumping action, whereas
degrades MTBE as a sole food source (73). This strain was the other 50% presumably became attached to the aquifer
injected into a fractured bedrock aquifer contaminated solids surrounding the extraction well.
with MTBE that previous tests had shown did not a TCE degradation by the in situ biofilter was very lim-
decrease in concentration during previous tests when ited. Initially, the TCE concentration in the extracted
only oxygen was added (76). The addition of ENV735 groundwater was reduced from about 425 µg/L to less
with oxygen was shown to effect immediate decreases than 10 µg/L, but it remained at these low concentra-
in the concentration of MTBE throughout the injection tions for only about two days. The TCE concentration
zone (unpublished results). then increased until, after 21 days, it was near the initial
440 BIOAUGMENTATION

(untreated) concentration. Integration of the TCE concen- fractures. Approximately 550 L of a high cell density
tration over time revealed that only about 20 g of TCE culture (∼1010 CFU/ml) of strain ENV435 was injected
were degraded during the 40-day study. This represented into the aquifer formation during the fracturing process,
approximately 40% of the TCE extracted through the in and results of plate count analysis demonstrated that the
situ biofilter, and only about 4% of the calculated degra- organisms were dispersed throughout the aquifer in a
dation capacity of the injected organisms. radius of about 25 feet from the fracture/injection well.
One of the greatest apparent limitations of the in Cell numbers in groundwater collected from monitoring
situ biofilter approach is that contact between the added wells were as high as 108 CFU/ml almost immediately
organisms and the contaminant is limited by the same after injection.
physiochemical properties that limit any pump-and-treat TCE concentrations in the formation rapidly decreased
technology, namely, the TCE present in the aquifer from between 20 to 30 mg/L to less than 5 mg/L within a
must undergo repeated sorption and desorption as it few days after injection. A decrease in TCE degradation
moves toward the biofilter and its associated extraction rate with time correlated with decrease in the viable
system (80). The rate-limiting step, as in any pump- ENV435 population during the same two-week period.
and-treat system, therefore, is transport of the TCE to It was estimated that during that study approximately
the treatment system. The biofilter in effect becomes 825 g of TCE were degraded by approximately 46,000 g
the treatment process for the pumping system, and in (wet weight) of ENV435 cells. This corresponds to an
order to be a viable technology it must be as reliable and apparent transformation capacity (Tc) of about 0.018 mg
inexpensive as competing treatment technologies. Given TCE/mg cells, which is greater than that estimated for a
the long treatment times associated with pump-and-treat toluene-degrading enrichment culture (0.0073) but lower
systems, it is unlikely that in situ biological systems than that reported for a phenol-degrading enrichment
can become as reliable or cost-effective as a standard culture (0.031) (51,52).
ex situ stripping tower and/or activated carbon adsorption
system. In fact, in the demonstration reported by Duba Field Application — Porous Aquifer
and coworkers, regulatory treatment standards were met
In a related study, strain ENV435 was injected during four
for only two days.
separate tests into a semi-confined CVOC-contaminated
In situ biofilters are an attractive remedial option,
aquifer located in Central New Jersey (49). The ground-
however, because they rely on the natural tendency of
water within the treatment zone was contained by recir-
bacteria to stick to solid surfaces, and because the bacteria
culating groundwater from a downgradient recovery well
are immobilized, addition of cosubstrates, nutrients, or
into a series of upgradient injection wells. A hydraulically-
oxygen and contact between the bacteria and the additives
isolated control plot was operated identically to the test
is easily implemented. One strategy that has been
plot, but without the addition of microorganisms. Results
developed to overcome the pump-and-treat limitations of
of the study indicate that in situ bioaugmentation can be
this technology is the use of an electric current to deliver
an effective treatment alternative for CVOC-contaminated
the contaminant to treatment zones created by hydraulic
aquifers.
fracture injection. In two field tests of this technology
In the first field experiment, ENV435 was injected
for TCE remediation (81,82) the treatment zones were
into six injection wells located approximately 12 m
composed of carbon and iron filings. However, bench
upgradient of the recovery well. Groundwater samples
scale studies have demonstrated that inoculation of these
were recovered from monitoring wells located 2, 5,
treatment zones with degradative bacteria is an effective
and 8 m downgradient from the injection wells. Strain
technology for the biodegradation of the electrokinetically-
ENV435 was enumerated by selective plating, and CVOC
delivered contaminants (83).
concentrations were determined by gas chromatography.
In another experiment, strain ENV435 was injected into
Field Application — Bedrock Aquifer
each of nine monitoring wells (2 depths each) throughout
In another field-scale demonstration of bioaugmen- the test plot.
tation, strain ENV435 was injected directly into a During the first experiment, the adhesion-deficient
bedrock aquifer (84). Strain ENV435, is a variant of strain ENV435 was distributed throughout the test plot
the toluene/TCE-degrading strain B. cepacia G4 (54). The aquifer in a pattern similar to the distribution of a bromide
strain had been mutated to select a new strain, B. cepacia tracer. Sodium bromide was distributed throughout the
PR1301 that constitutively expressed the toluene monooxy- test plot within about 13 days after injection, whereas
genase genes that were responsible for TCE degrada- strain ENV435 was distributed throughout the plot within
tion (22). An adhesion-deficient variant of strain PR1301 20 days. The calculated retardation factor for ENV435 in
was selected by successive passage of the culture over the fine sand layer of the aquifer was only 1.25. These
a sand column (85). Finally, before introduction into a results suggested that adhesion-deficient bacteria can be
contaminated aquifer, the strain was grown to high cell effectively dispersed throughout an aquifer. However,
density under conditions that promoted the production even better distribution is likely when the adhesion-
of high levels of poly-β-alkonoates that can serve as a deficient strain is injected into multiple injection points
high-energy storage material to prolong degradative activ- throughout the contaminated aquifer (e.g., experiment 2).
ity (46). In this case, the radius of influence of the strains can
To facilitate injecting the strain into the aquifer, be calculated to determine the spacing required between
pneumatic fracturing was used to expand bedrock injection points (49).
 BIOAUGMENTATION 441

When CVOC concentrations were monitored in the needed additions (cosubstrates, nutrients, oxygen, etc.)
ENV435-amended test plot, a rapid decrease in CVOC with the contaminants is key to the effectiveness of the
concentrations followed the injection of ENV435. An anal- remedy.
ysis of the ratio of degradable CVOCs in the groundwater
(TCE, DCE, and VC) to nondegradable CVOCs (PCE and BIBLIOGRAPHY
TCA) confirmed that the observed decrease in CVOC con-
centrations was caused by biodegradation and not other 1. J. J. Westrick, J. W. Mello, and R. F. Thomas, J. Am. Water
factors such as dilution. During the first experiment, this Works Assoc. 76, 52–59 (1984).
ratio was lowest at the first monitoring well and increased 2. B. D. Ensley, Annu. Rev. Microbiol. 56, 283–299 (1991).
at the downgradient wells suggesting that the degrada- 3. B. Dabrock, J. Riedel, J. Bertram, and G. Gottschalk, Arch.
tive activity of the injected cells decreased with time and Microbiol. 158, 9–13 (1992).
distance during the 15-day travel time through the aquifer. 4. M. J. K. Nelson, S. O. Montgomery, and P. H. Pritchard,
When the strain was injected into monitoring wells Appl. Environ. Microbiol. 54, 604–606 (1988).
spaced across the entire test plot (experiment 2), CVOC 5. L. P. Wackett and D. T. Gibson, Appl. Environ. Microbiol. 54,
concentrations were reduced throughout the entire test 1703–1708 (1988).
plot within only a couple of days after injection. Further- 6. R. B. Winter, K. M. Yen, and B. D. Ensley, Biotechnology 7,
more, the ratio of degradable CVOCs to nondegradable 282–285 (1989).
CVOCs decreased at every monitoring point. Thus, if eco- 7. M. M. Fogel, A. R. Taddeo, and S. Fogel, Appl. Environ.
nomically and logistically feasible, it appears that the best Microbiol. 51, 505–511 (1986).
treatment strategy for applying Type 2 bioaugmentation 8. B. G. Fox, J. G. Borneman, L. P. Wackett, and J. D. Lip-
is to use multiple biocatalyst injection points spaced to scomb, Biochemistry 29, 6419–6427 (1990).
allow rapid dispersion of the organisms throughout the 9. C. D. Little et al., Appl. Environ. Microbiol. 51, 951–956
contaminated zone of an aquifer. (1988).
The bacterial transport and CVOC degradation data 10. R. Oldenhuis, J. Y. Oedzes, J. J. van der Waarde, and
gathered during the field demonstration allowed the D. B. Janssen, Appl. Environ. Microbiol. 57, 7–14 (1991).
development of treatment and cost models. The treatment 11. D. T. Arciero, T. Vanelli, M. Logan, and A. B. Hooper,
model incorporated ENV435 transport characteristics Biochem. Biophys. Res. Commun. 159, 640–643 (1989).
to determine the required spacing of injection points, 12. L. P. Wackett, G. A. Brusseau, S. R. Householder, and R. S.
and knowing the spacing requirements, allowed the Hanson, Appl. Environ. Microbiol. 55, 2960–2964 (1989).
calculation of cost of system installation. For a site with 13. M. D. Lee et al., Annu. Rev. Microbiol. 52, 423–452 (1998).
the dimensions of 152 m × 152 m × 6 m, and a CVOC
14. J. Gerritse et al., Appl. Environ. Microbiol. 65, 5212–5221
concentration of 1,000 µg/L, estimated cleanup costs would (1999).
be approximately $7/m3 ($5/yd3 ) and it would require
15. X. Maymo-Gatell, Y. Chien, J. M. Gossett, and S. H. Zinder,
only about one year to remediate the site. These costs Science 276, 1568–1571 (1997).
are comparable to or less than competitive technologies
16. D. T. Adamson, and G. F. Parkin, Environ. Sci. Tech. 34,
including air sparging/soil vapor extraction with off-gas 1059–1965 (2000).
treatment.
17. M. R. Harkness et al., Environ. Sci. Tech. 33, 1100–1109
(1999).
CONCLUSION 18. S. J. Flynn et al., Environ. Sci. Tech. 34, 1056–1061 (2000).
19. Y. Yang and P. L. McCarty, Environ. Sci. Tech. 32,
Bioaugmentation, like most remedial technologies, is site- 3591–3597 (1998).
and contaminant-specific. In other words, whether or not 20. J. W. Costerton, R. T. Irvin, and K. J. Cheng, Annu. Rev.
bioaugmentation will be effective and how it will be imple- Microbiol. 35, 299–324 (1981).
mented depends of the hydrogeology and geochemistry 21. M. S. Shields and M. J. Reagin, Appl. Environ. Microbiol. 58,
of the site as well as the mixture and concentration 3977–3983 (1992).
of the contaminants. Extremely high concentrations and 22. J. Munakata-Marr et al., Environ. Sci. Tech. 30, 2045–2052
free product contamination can be toxic or inhibitory to (1996).
bacteria and not amenable to bioaugmentation without 23. P. A. Phelps, S. K. Agarwal, G. E. Speitel, and G. Georgiou,
pretreatment to lower concentrations. For contaminants Appl. Environ. Microbiol. 58, 3701–3708 (1992).
that are strongly sorbed to aquifer solids, injecting bacteria 24. Y. Kim, P. Ayoubi, and A. R. Harker, Appl. Environ. Micro-
that will disperse or grow throughout the contaminated biol. 62, 3227–3233 (1996).
zone would be more effective than an in situ biofilter 25. M. Tugusheva and R. J. Steffan, Annual Meeting of the
because of the length of time, that it would take for desorp- American Society for Microbiology (1993).
tion of the contaminant from the aquifer solids to occur. If 26. U. Pflugmacher, B. Averhoff, and G. Gottschalk, Appl. Envi-
bacteria are dispersed throughout the contaminated zone, ron. Microbiol. 62, 3967–3977 (1996).
the ongoing degradation will increase the concentration 27. K. McClay, S. H. Streger, and R. J. Steffan, Appl. Environ.
gradient and desorption will occur more rapidly. On the Microbiol. 6, 3479–3481 (1995).
other hand, if there is an ongoing source, or if the con- 28. J. G. Leahy, A. M. Byrne, and R. H. Olsen, Appl. Environ.
taminants are highly soluble, an in situ biofilter may be Microbiol. 62, 825–833 (1996).
an effective application of this technology. As with all in 29. D. Heald and R. O. Jenkins, Appl. Environ. Microbiol. 60,
situ remedial technologies, contact of the bacteria and any 4634–4637 (1994).
442 BIOAUGMENTATION: COLD-ADAPTED MICROBES

30. A. J. Anderson, H. Habibzadegah-Tari, and C. S. Tepper, 61. M. S. Fries, L. J. Forney, and J. M. Tiedje, Appl. Environ.
Appl. Environ. Microbiol. 54, 375–380, (1988). Microbiol. 63, 1523–1530 (1997).
31. F. Caccavo, P. C. Schamberger, K. Keiding, and P. H. 62. G. D. Hopkins, and P. L. McCarty, Environ. Sci. Tech. 29,
Nielsen, Appl. Environ. Microbiol. 63, 3837–3843 (1997). 1628–1637 (1995).
32. M. M. Cowan and M. Fletcher, J. Microbiol. Methods 7, 63. C. S. Criddle, J. T. DeWitt, D. Grbić-Galić, and P. L.
241–249, (1987). McCarty, Appl. Environ. Microbiol. 56, 3240–3246 (1990).
33. J. L. R. Crews, S. Colby, and A. G. Matthysse, J. Bacteriol. 64. T. A. Lewis and R. L. Crawford, Appl. Environ. Microbiol. 59,
172, 6182–6188 (1990). 1635–1641 (1993).
34. M. E. Fuller et al., Water Resources Res. No. 936, 2417–2431 65. G. M. Tatara, M. J. Dybas, and C. S. Criddle, Appl. Environ.
(2000). Microbiol. 59, 2126–2131 (1993).
35. J. Gong and C. W. Forsberg, Appl. Environ. Microbiol. 55, 66. M. J. Dybas et al., Appl. Environ. Sci. Tech. 32, 3598–3611
3039–3044 (1989). (1998).
36. M. F. DeFlaun et al., Appl. Environ. Microbiol. 56, 112–119 67. C. A. Lajoie, S. Y. Chen, K. C. Oh, and P. F. Strom, Appl.
(1990). Environ. Microbiol. 58, 655–663 (1992).
37. M. F. DeFlaun, C. W. Condee, and B. D. Ensley, In situ reme- 68. C. A. Lajoie, A. C. Layton, and G. S. Sayler, Appl. Environ.
diation, scientific basis for current and future technologies Microbiol. 60, 2826–2833 (1994).
(Part 1), in G. W. Gee and N. R. Wing, eds., Thirty-Third 69. I. Okuda, J. F. McBride, S. N. Gleyzer, and C. T. Miller,
Hanford Symposium on Health and the Environment, Battelle Environ. Sci. Tech. 30, 1852–1860 (1996).
Press, Washington, D.C., 1994, pp. 249–258. 70. K. D. Pennell, G. A. Pope, and L. M. Abriola, Environ. Sci.
38. R. K. Rothmel, R. W. Peters, E. St. Martin, and M. F. Tech. 30, 1328–1335 (1996).
DeFlaun, Environ. Sci. Tech. 32, 1667–1675 (1998). 71. R. A. Deeb et al., Environ. Sci. Tech. 35, 312–317 (2001).
39. M. J. Gross and B. E. Logan, Appl. Environ. Microbiol. 61, 72. J. P. Salanitro, L. A. Diaz, M. P. Williams, and H. L. Wis-
750–1,756 (1995). niewski, Appl. Environ. Microbiol. 60, 2593–2596 (1994).
40. Q. Li and B. E. Logan, Water Res. 33, 1090–1100 (1999). 73. R. J. Steffan et al., Bioremediation and Phytoremediation of
41. F. Caccavo, N. B. Ramsing, and J. W. Costerton, Appl. Envi- Chlorinated and Recalcitrant Compounds, Battle Press, Ohio,
ron. Microbiol. 62, 4678–4682 (1996). 2000.
42. F. A. Macleod, H. M. Lappin-Scott, and J. W. Costerton, 74. P. B. Hatzinger et al., Soil Sediment Water 81–84 (2001).
Appl. Environ. Microbiol. 54, 1365–1372 (1988). 75. J. P. Salanitro et al., Environ. Sci. Tech. 34, 4152–4162
43. L. Pang, M. Close, and M. Noonan, Groundwater 36, 112–122 (2000).
(1998). 76. R. J. Steffan et al., In Situ and On-Site Bioremediation, Sixth
44. K. E. Stormo and R. L. Crawford, Appl. Environ. Microbiol. International Symposium (2001).
58, 727–730 (1992). 77. K. M. Scow et al., In situ and On-Site Bioremediation, Sixth
45. M. Lemoigne, Bull. Soc. Chim. Biol. 8, 770–782 (1926). International Symposium (2001).
46. S. M. Henry and D. Grbić-Galić, Appl. Environ. Microbiol. 57, 78. D. E. Ellis and E. J. Lutz, Environ. Sci. Tech. 34, 2254–2260
236–244 (1991). (2000).
79. A. B. Duba et al., Environ. Sci. Tech. 30, 1982–1989
47. H. Henrysson and P. L. McCarty, Appl. Environ. Microbiol.
(1996).
59, 1602–1606 (1993).
80. W. P. Ball and P. V. Roberts, Environ. Sci. Tech. 25,
48. N. I. Lopez et al., FEMS Microbiol. Ecol. 16, 95–101 (1995).
1223–1236 (1991).
49. R. J. Steffan et al., Environ. Sci. Tech. 33, 2771–2781 (1999).
81. S. V. Ho et al., Environ. Sci. Tech. 33, 1086–1091 (1999a).
50. L. Alvarez-Cohen and P. L. McCarty, Appl. Environ. Micro-
82. S. V. Ho et al., Environ. Sci. Tech. 33, 1092–1099 (1999b).
biol. 57, 1031–1037 (1991).
83. S. V. Ho et al., Environ. Sci. Tech. 29, 2528–2534 (1995).
51. H. L. Chang and L. Alvarez-Cohen, Biotechnol. Bioeng. 45,
440–449 (1995a). 84. M. Walsh et al., in H. I. Inyang and C. J. Bruell, eds.,
Remediation in Rock Masses ASCE Press, Reston, Va., 2000,
52. H. L. Chang and L. Alvarez-Cohen, Environ. Sci. Tech. 29,
pp. 152–168.
2357–2367 (1995b).
85. M. F. DeFlaun et al., Appl. Environ. Microbiol. 65, 759–765
53. M. L. Krumme, K. N. Timmis, and D. F. Dwyer, Appl. Envi-
(1998).
ron. Microbiol. 59, 2746–2749 (1993).
54. M. J. K. Nelson, S. O. Montgomery, E. J. O’Neill, and P. H.
Pritchard, Appl. Environ. Microbiol. 52, 383–384 (1986). See also BIOREMEDIATION: AQUATIC ECOSYSTEMS.
55. M. J. Nelson, J. V. Kinsella, and T. Montoya, Environ. Prog.
9, 190–196 (1990).
56. A. Massol-Deya et al., Appl. Environ. Microbiol. 63, 270–276 BIOAUGMENTATION: COLD-ADAPTED
(1997).
MICROBES. See USE OF COLD-ADAPTED MICROORGANISMS
57. L. P. Wackett and S. R. Householder, Appl. Environ. Micro- IN BIOTECHNOLOGY
biol. 55, 2723–2725 (1989).
58. C. M. Yeager, P. J. Bottomley, and D. J. Arp, Appl. Environ.
Microbiol. 67, 2107–2115 (2001).
59. A. E. Mars et al., Appl. Environ. Microbiol. 64, 208–215
(1998). BIOAVAILABILITY OF ORGANIC SUBSTRATES.
60. P. L. McCarty et al., Environ. Sci. Tech. 32, 88–100 (1998). See ACTIVITY AND CARBON TRANSFORMATIONS IN BIOFILMS
 BIOCHIP-BASED DEVICES AND METHODS IN MICROBIAL COMMUNITY RIBOTYPING 443

BIOCHIP-BASED DEVICES AND METHODS IN often involves approaches based on ‘‘molecular biology’’ or
MICROBIAL COMMUNITY RIBOTYPING ‘‘analytical chemistry.’’ It is now recognized that many of
the classical schemes for differentiation of bacteria provide
SHENGCE TAO little insight into their genetic relationships and to some
JING CHENG extent are scientifically incorrect. The new information
Beijing National Biochip has resulted in renaming of certain bacterial species
Research and Engineering and, in some instances, in the complete reorganization of
Center relationships within and among many bacterial families.
Beijing, China The aim of typing is to identify stable groups in a
Tsinghua University
Beijing, China
population of the same species using markers that are
reproducible, accurate, and conveniently detected. The
XUEMEI MA complex characteristics (phenotypic patterns) are intu-
Beijing National Biochip itively recognized as good indicators of relationship. The
Research and Engineering characters of bacteria that are traditionally used for typing
Center
include biotypes, antigenic structure, phage susceptibility,
Beijing, China
and so forth. Genotypic information or sequence informa-
tion is superior to phenotypic information in two ways for
Studies in environmental microbiology are often limited by
relating and classifying microorganisms: (1) the results
the inability to unambiguously identify and directly quan-
can be more readily, reliably, and precisely interpreted,
tify the enormous diversity of natural microbial popula-
and (2) the results are innately more informative of evo-
tions. Our perspective on microbial diversity has improved
lutionary relationships than the phenotypic information.
enormously over the past few decades. In a large part, this
has been due to molecular phylogenetic studies of related The elements of a sequence, nucleotides, or amino acids
organisms. Phylogenetic tree based on gene sequences are are restricted in number and are well defined. The subjec-
maps that are used to articulate the elusive concept of tivity that goes into the judgments ‘‘same,’’ ‘‘similar,’’ and
biodiversity. The phylogenetic classification of microbes so forth at the phenotypic level is replaced by simple, more
based on rRNA gene (rDNA) sequences offers a well- objective judgments and mathematical definitions.
defined framework, which can be used to develop molecular In the past several years, sequence analysis of the
tools for microbial identification. The rDNA molecules typ- entire genome of one representative (i.e., a strain) out
ically consist of highly conserved sequences interspersed of a few bacterial species has been achieved. For species
with regions of more variable sequences. Together with identification, it would be ideal to compare the sequences
the development of the PCR technique and a number of of the entire bacterial chromosomal DNA. However, this
high-resolution molecular typing systems, this knowledge is currently not a practical approach, because enormous
has provided an approach for microbial community typ- amount of efforts has to be put in to sequence millions
ing. Many devices and methods have been developed and of nucleotides for each strain. An alternative way is to
applied for microbial community typing, which include assess the genomic similarity by the content of guanine
biosensors (1) capillary electrophoresis (2), the biochip (G) + cytosine (C), usually expressed as a percentage (%
technology (3) and others (4). The traditional typing meth- G+C). This has lately been replaced by two different
ods used in microbiology consist of (1) Culture followed approaches, namely, hybridization or sequence analysis
by morphological and physiological analysis, (2) Molecular of the genetic coding segment for 16S rRNA.
biology methods [such as polymerase chain reaction (PCR),
restricted fragment length polymorphism (RFLP), ampli- Ribosomal RNA
fication fragment length polymorphism (AFLP), and PCR-
The story of the ribosomal RNA (rRNA) begins in 1960s,
RFLP of polymorphic loci, etc.]. Compared with the newly
but not much progress was made until the early 1970s,
developed biochip methods, the traditional methods are
laborious and time-consuming. The advantages of the when Woese realized that rRNA is universal to life and
biochip-based systems are obvious: sensitive, specific, fast, that much historical information could be stored in it.
portable, and parallel sample analysis capability. In this Comparative analyses of small-subunit (SSU) rRNA (16S
article, we will first present a brief overview of microbial or 18S rRNA) and other gene sequences show that life
community ribotyping in environmental microbiology and falls into three primary domains — Bacteria, Eukarya, and
then focus on the application of biochip-based devices and Archaea (5). Since then, the three-domain theory based on
methods in microbiology, especially in microbial commu- rRNA has been widely accepted.
nity ribotyping in environmental microbiology. The ribosome is a molecular machine in which all the
proteins are synthesized. In the past few years, sequence
analysis of 16S rRNA has become the ‘‘gold standard’’
BACKGROUND OF MICROBIAL COMMUNITY
in bacterial taxonomy. This is due to its molecular
RIBOTYPING
ubiquitousness and the functional constancy. The 16S
rRNA changes very slowly in sequence and can be easily
Microbial Community Typing
traced through experimentation. Moreover, as the central
Modern taxonomic approaches often employ technically component of the highly complex translation apparatus,
more complex methodologies and are concerned with rRNA is among the most refractory of molecules to the
profiling the structural composition of bacteria. This vagaries of horizontal gene flow and so was considered
444 BIOCHIP-BASED DEVICES AND METHODS IN MICROBIAL COMMUNITY RIBOTYPING

likely to avoid the phylogenetic hodgepodge of reticulate nucleic acid hybridization analysis in a high-throughput
evolution (6). and low-cost way.
The nucleotide sequence of the 16S rRNA gene is about Active Biochips. The active biochip (7) refers to
1,500-letters long, and consists of many domains. The 16S devices on which elements with different functions have
rRNA changes very slowly in sequence, which makes it an been fabricated to generate forces (i.e., electric force,
accurate chronometer. The sequence of 16S rRNA provides magnetic force, acoustic force, etc.) for active manipulation
a measure for genomic similarity above the level of the of cells and molecules.
species and allows comparison of relatedness across the Most biological molecules are charged and can be moved
entire bacterial kingdom. Closely related bacterial species and concentrated by electric fields. Cells or microparticles
often have identical rRNA sequences. The technique with molecules attached can be manipulated by inducing
thus provides complementary information to DNA-DNA the dipole on the particles, which is independent
hybridization. Exploiting this basic research in the clinical of the native charges carried by these materials.
laboratory has allowed the development of probes for Research reports on active biochips are currently limited.
improved identification of bacterial pathogens. Companies such as Nanogen, Caliper technologies, Aclara,
A large collection of SSU rRNA-targeted DNA probes and AVIVA Biosciences are the leaders in this category.
has been developed for studies in environmental microbi- Among these, AVIVA is the only company making chips
ology. They have been designed to identify phylogenetic capable of producing multiple forces either sequentially
groups of different evolutionary depths, corresponding in or simultaneously depending on the actual needs of the
general to the taxonomic ranks of species, genus, family, experiment. The advantage an active biochip has over
and higher. In particular, the use of group-specific DNA a passive chip is the speed, accuracy, and sensitivity
probes complementary to the SSU rRNA has provided a in reaction and detection. Using active biochip-based
comprehensive framework for studies of microbial pop- analytical systems, the entire sample-to-result process
ulation structure in complex systems. However, to fulfil could be completed in minutes.
such studies, it is highly desirable to conduct independent
hybridization of multiple environmental samples to multi- Chip-Based Sample Preparation
ple DNA probes. Biochip as a new platform can meet just
Traditionally, the sample preparation step involves
this requirement.
numerous operations (e.g., pipetting, centrifugations,
extraction, precipitation, etc.) on a large quantity of
BIOCHIP-BASED DEVICES AND METHODS samples. They are often labor-intensive and time-
consuming. Recently, chip-based small disposable devices
Biochip is a fingernail-sized solid device used for have been developed for handling biological samples. Many
the analysis of biomolecules (e.g., DNA, RNA protein reports (8,9) have been dedicated to cell manipulation
etc.) in parallel. A typical analytical system usually and nucleic acid extraction, that is, separation and
consists of three classical steps, for example, sample transportation of cells and isolation of DNA and RNA
preparation, chemical reaction, and detection. The total molecules.
integration of these three steps has been the dream An effective microscopic sample preparation method
for many years for both academic researchers and should be sensitive to certain cell properties and selective
entrepreneurs. The marriage between molecular biology for cell types to allow for discrimination and separation of
and the semiconductor industry for the first time brings target cells, requiring little or no sample pretreatment.
this hope to the scientific community. The use of biochips So far, only a few microscopic-scale cell-manipulation
will substantially increase the research speed in molecular concepts have been demonstrated including diffusion-
biology and they will find their way in environmental based particle separation and detection; microtitration of
microbiology, especially in bacteria community typing, cell samples that exploits the cell size and mechanical
which is now mainly based on the analysis of sequence properties (10); and electronic separation of cells that
diversity of 16S rRNA. Biochip-based systems can be made makes use of the electrical properties of cells (11). Figure 1
small and compact and thus suitable for field and point- shows the separation of Escherichia coli. from blood using
of-care use for which hand-held analyzers are required. a microfabricated dielectrophoretic chip.
The San Diego–based company AVIVA Biosciences is
Classification of Biochips known for its unique multiple-force chips (Fig. 2). These

Passive Biochip. The passive biochip specifically refers

to devices on which no functional elements have been
fabricated to actively generate forces for molecule
manipulation. The reaction takes place simply on the
basis of the specific binding of complementary DNA or
RNA and the specific recognition of antigen and antibody
via random diffusion process. The results obtained
from the passive biochip are displayed in an arrayed
format. When compared with the traditional single-
probe hybridization method, DNA microarray technology Figure 1. The separation of E coli. from blood using a microfab-
demonstrated its advantage of being able to perform ricated dielectrophoretic chip. See color insert.
 BIOCHIP-BASED DEVICES AND METHODS IN MICROBIAL COMMUNITY RIBOTYPING 445

or proteinase K digestion, chemical labeling reaction, and

so on.
The nucleic acids extraction studies can be divided
into two types: high-concentration inputs and low-
concentration inputs (at or below 105 copies/mL). These
two regimes have relevance to different clinical situations.
Samples involving genomic DNA, such as whole blood,
typically possess large quantities of DNA. On the other
hand, samples for diagnosing infectious diseases often
contain minute amount of pathogenic nucleic acids of
interest. Serum, plasma, and urine are examples for
the latter case. For studies on high concentration, the
quantities of DNA are large enough to conduct direct
fluorescence assay. For low concentrations, however,
amplification reactions such as PCR must be used to first
amplify a target sequence.
PCR chips shown in Figure 3 have been used to perform
nucleic acid amplification (13). It has been assumed since
the first illustration of microchip-based PCR in the early
1990s that features such as low reagent consumption,
low-volume sample requirement, and rapid cycling would
Figure 2. A microfabricated multiple-force active biochip capa-
be a consequence of this technology. PCR reactions
ble of isolating various bioanalytes from a complex biological
previously taking up to three hours were shown to be
sample. With permission from AVIVA Biosciences. See color
insert. feasible with time as little as 20 minutes (14). It will
require considerable effort to overcome many difficulties in
fluidic control, efficient thermocycling, surface chemistry,
chips can produce a variety of physical forces and are specimen introduction, and quantitation of amplicons.
useful for actively manipulating a variety of bioanalytes However, it is quite clear now that microchip-based
such as cells, proteins, DNA, and RNA. The combined use amplification will play a major role in molecular biology
of AVIVA biochips with microbeads provides the capability and find wide application in environmental microbiology.
for separating and manipulating diversified bioanalytes
from crude biological samples. Extraction, purification, Chip-Based Detection
and concentration of nucleic acids from complex biological
The detection of nucleic acids is usually achieved by
samples are the initial steps in any nucleic acid–based
using two approaches. One approach is a ‘‘separation-
assay. This is necessary because the nucleic acids from
based detection’’ such as electrophoresis, chromatography,
live organisms are protected inside the living cell. Once
spectrometry; the other one is a hybridization-based
the cell is broken open, nucleic acids will be released
approach.
into solution and become accessible for analysis. However,
other biological chemicals may also be present in the
Chip-Based Electrophoresis Systems. Capillary elec-
solution. Many of these biochemicals, such as proteins
trophoresis (CE) is an attractive technique. In the last
and metal complexes, tend to bind with the nucleic acids
decade, chip CE has been made possible in separating
and therefore interfere with the followed DNA or RNA
and analyzing DNA samples within a few minutes or even
amplification reaction. Hence, to successfully carry out
seconds. It may eventually prove to be a valuable tool for
nucleic acid amplification, it is essential to eliminate those
microbiologists.
contaminating materials.
Oxidized silicon surface is useful for capturing nucleic Different materials (e.g., silicon, glass, and plastics)
acids. Christel described such a system for rapid nucleic have been used as substrata for fabricating chip CE
acids analysis (12). The extraction, purification, and con- devices. A variety of fabrication processes have been
centration of DNA from test samples have been accom- developed to accommodate the complicated requirements.
plished using fluidic microchips with high surface area to Biochip-based CE systems have demonstrated their uses
volume ratios. Short (500 bp) and medium size (48,000 bp)
DNAs have been captured, washed, and eluted using the
silicon dioxide surfaces of these chips. DNA quantities
approaching 40 ng/cm2 of binding area were captured
from input solutions in the 100 to 1,000 ng/ml concen-
tration range. The extraction efficiency was about 50%
with diluted samples of interest for pathogen detection.

Chip-Based Reactions
Chip-based reactions are usually represented by various Figure 3. Microfabricated silicon-glass PCR chips. See color
enzymatic reactions such as PCR amplification reaction, insert.
446 BIOCHIP-BASED DEVICES AND METHODS IN MICROBIAL COMMUNITY RIBOTYPING

in diverse applications such as the separation of amino complex mixtures. In one experiment, they successfully
acids, DNA or RNA fragment sizing, DNA sequence identified different strains of E. coli found in bacterial
analyses and CE-based laboratory-on-a-chip systems. cultures. This array is able to simultaneously identify
A fused silica CE chip has been fabricated and used a potentially large number of compounds and, combined
for fast-DNA profiling (15) with a replaceable denaturing with the precise exclusivity of each switch, adds up to a
polyacrylamide matrix. Baseline resolution was achieved recipe for a powerful and wide-ranging laboratory on a
with four amplicons representing loci of the CSF1PO, dime-sized slice of silicon.
TPOX, THO1, and vWA genes in less than two minutes.
This represents a 10- to 100-fold increase in the speed of Micro-Total Analysis System
separation compared with the conventional capillary or
slab gel electrophoresis systems. Ribosomal RNA samples The detection of microorganisms is important in disease
were reportedly separated in an injection-molded plastic diagnosis, in guaranteeing the safety of food and water
microchannel with a cross section of 100 × 40 µm and with supply, and in ensuring public safety from the threat of
an effective length of one centimeter. biological warfare agents. In many of these applications,
the time taken to provide the result is a critical issue. A
portable system that is easy-to-use, and yet would provide
Hybridization-Based Biochips. Hybridization analysis
a rapid result, would be of great benefit. A fully inte-
offers many advantages over other methods of DNA
grated biochip-based micro-Total Analysis System (µTAS)
identification, such as comparative size or relative migra-
or ‘‘lab-on-a-chip’’ possessing the aforementioned merits
tion determination, by providing insight into the specific
usually consists of (1) chip-based sample treatment, such
nucleotide sequence being investigated. Specifically, with
as cell separation, cell fractionation, and nucleic acid
hybridization, spatial position of sequence-specific probes
extraction; (2) chip-based biochemical reaction, such as
the corresponding signal intensity generated by these
PCR, immunoassay, and fluorescent labeling; (3) chip-
probes provide the basis for analyzing complex genetic
based detection, such as electrophoresis, chromatogra-
mixtures for sequence-specific characteristics.
phy, spectrometry, and ‘‘hybridization-based’’ approaches.
Microchip-based nucleic acid arrays now permit the
However, to complete such a system will require the
rapid analysis of genetic information by hybridization.
use of other microelectromechanic system (MEMS)-based
Many of these devices take advantage of sophisticated
devices including microheaters, valves, pumps, and detec-
silicon manufacturing processes developed by the semi-
tors. Such a system, once constructed, would allow for the
conductor industry over the last 40 years. With these
generation of very sophisticated biochemical and biomed-
devices, parallel hybridization was made possible with
ical information with simplified steps. Figure 4 is the
immobilized capturing probes. Stringency and rate of
schematic of such a system.
hybridization are generally controlled by temperature and
The first Lab-on-a-Chip system was developed and
salt concentration of the solutions used for hybridization
reported by Cheng and coworkers (17). Using the same
and wash. The passive microarray approaches have the fol-
type of photolithographic and deposition techniques
lowing limitations: (1) All nucleic acids are exposed to the
employed in the microelectronics industry, Cheng and
same conditions simultaneously — capture probes should
colleagues have created a 1 cm2 silicon chip that
be made with similar melting temperatures to achieve
functions as a lab. The manufactured bioelectronic chip
similar hybridization stringency; (2) The rate of hybridiza-
consists of an addressable array of 25 platinum 80-
tion is proportional to the initial concentration of the
µm diameter microelectrodes covered by an agarose
interrogated solution — high concentrations are therefore
permeation layer in a 4.84-µl flow cell. The alternating
required to achieve rapid hybridization; (3) Hybridization
current field that is set up in the chip separates cells by
conditions are difficult to control — single base discrim-
dielectrophoresis. Different cell types locate to different
ination is generally restricted to oligomers of 20 bases
regions of the microelectrode array. Weakly held cells
or less.
located in interelectrode regions of low field strength
Many of the limitations with passive hybridization
techniques can be overcome by using active biochips, which
can generate desired forces, thus facilitating (1) the rapid
Sample input
transport and selective addressing of DNA molecules to the Chemical reaction
test on the chip surface; (2) speedy hybridization process; Sample preparation
and (3) rapid discrimination of single base mismatches in
target DNA sequences.
A research group at Yale University has produced an
RNA-based microarray that promises to put a powerful
diagnostic lab on a dime-sized chip (16). To create the
prototype, they placed the RNA switches (a molecular
switch is a molecule that can be turned on or off by another Detection
molecule or compound) on a gold-coated silicon surface
and arranged them in clusters. Each switch was designed
to bind only to a specific molecule — its ‘‘target‘‘ — and
Reagents
then release a signal that identifies the target molecule.
They tested the array of RNA switches on a variety of Figure 4. Illustration of a lab-on-a-chip system.
 BIOCHIP-BASED DEVICES AND METHODS IN MICROBIAL COMMUNITY RIBOTYPING 447

can be selectively removed by washing. The chip isolates

E. coli cells from a whole blood sample in about 4 minutes
through the use of dielectrophoresis, followed by electronic
lysis and proteolytic digestion. The lysate contained
a spectrum of nucleic acids including RNA, plasmid
DNA, and genomic DNA and was further examined by
electronically enhanced hybridization assay on a second
bioelectronic chip. Cultured cervical carcinoma cells were
similarly isolated from normal human blood cells by this
group using the same device in a modified manner.

THE APPLICATIONS OF BIOCHIPS IN MICROBIOLOGY

Study the Relationships Between the Microbes and the Hosts

For more than a century, microbiologists studied
pathogens in pure culture, whereas cell biologists stud-
ied mammalian cells in tissue culture. A few years ago,
researchers realized that these laboratory methods of
growing pathogens and their hosts were quite artificial
and had very little in common with real life, wherein
pathogens and hosts coexist, interact, and compete in
conditions that are often far from optimum. To better
mimic what happens in real life, the study of the inter-
action between microbes and host cells was proposed,
taking advantage of technological progress that allowed
cocultures of microbes and cells to be handled together.
For this reason, a new discipline (cellular microbiology)
was born (18) and a powerful new approach to cellular
microbiology was described (19). Instead of studying one
parameter at a time, microarray-based methods allow the
study, within a single experiment, of all the host genes and
those of the bacteria whose expression is modified during
host–pathogen interaction: the global picture of the dialog Figure 5. Schematic description of how microarrays can be used
between the pathogen and the host in one experiment. to study host cell–pathogen interactions (20). See color insert.
High-density genome-wide microarrays of eukaryotic
cells are commercially available. Homemade chips can
also be printed when the genes are available. In the a Cy3-labeled reproducible control, which is then mixed
past few years, sequencing of the entire genome of one with Cy5-labeled cDNA prepared from the experimental
representative (i.e., a strain) of a few bacterial species has condition, including uninfected cells. This allows an
been achieved. These include Helicobacter pylori, Neisseria invariant comparator, so that any experimental profile
meningitidis, Mycobacterium tuberculosis, and E. coli. can be compared with any other. After hybridization, laser
Chips containing all genes of the sequenced bacterial scanning, image detection, and analysis of each chip, the
genomes are either commercially available or homemade, ratio of the targets in Figure 5b and c provides a precise
and many more will be available in the near future. indication of the relative expression of the genes of the
A typical microarray experiment is shown in mammalian cells grown in the conditions in Figure 5b
Figure 5 (20). RNA is prepared from bacteria culture in and c. More specifically, for the labeling reaction [RNA
standard laboratory conditions (a), from host cells grown in Figure 5b (Cy3) or RNA in Figure 5c (Cy5)], the red
under optimum conditions in tissue culture (b), and from spot indicates gene activation after bacterial addition to
cells that have been infected with bacteria (c). RNA is con- the cells, and a green dot refers to the downregulation of
verted into fluorescent cDNA by incorporation of different the gene, whereas a yellow spot indicates no change in
fluorochromes during the reverse-transcription reaction gene expression. The final results of the analysis are the
(typically by using Cy3- and Cy5-dCTP). The labeled expression ratios of Figure 5b or c obtained from the mean
cDNAs are then hybridized to microarrays containing value of two labeling conditions.
bacterial probes (d) or eukaryotic probes (e). Biochip may be used in the study of infected tissues and
In the experiment shown in Figure 5e, the RNA whole organisms. The possibility of studying pure cultures
prepared in (b) and labeled with Cy3 (green) is mixed of pathogens while they interact with pure cultures of
with the RNA prepared in (c) and labeled with Cy5 (red). mammalian cells is exciting and represents a major step
The two fluorescent cDNAs are then mixed together and forward from traditional studies, in which pathogens and
hybridized to the chip. Sometimes, RNA prepared from host cells were studied separately. However, even the
a pool of control conditions or cells is used to prepare in vitro infection of host cells is far from a real-life
448 BIOCHIP-BASED DEVICES AND METHODS IN MICROBIAL COMMUNITY RIBOTYPING

scenario in which pathogens infect animals and their sequencing projects because of the problems resulting from
tissues. There is no doubt that the technology is available inefficient discrimination between perfect and mismatched
for studying host cell and pathogen-gene expression in duplexes, secondary structure within the single-stranded
whole organisms. For instance, bacteria recovered from DNA (or RNA), and the presence of tandem repeats in the
infected tissues (blood, cerebrospinal fluid, etc.) should be DNA to be sequenced. Yet, it has proven to be an excellent
suitable for sample preparation. Similarly, macrophages method for sequence analysis of mutations, gene polymor-
and lymphocytes from infected organisms and tissue from phisms, and other genetic changes and has been further
patients with chronic infections can be recovered and used extended to the studies in microbial ecology.
for sample preparation.
So far, published studies describe only the host response Nitrifying Bacteria Typing. Nitrifying bacteria have
to infection. Many studies regarding the bacterial response proved particularly difficult to study using cultivation
to host cell contact are in progress. techniques because of their long generation times
As an example (19), DNA microarray has been used and poor counting efficiencies. Thus, a rapid, culture-
for studying the transcriptional responses of respiratory independent enumeration technique for nitrifiers could
epithelial cells to Bacillus pertussis. Bacillus pertussis is greatly facilitate research in their ecology.
the causative agent of whooping cough, which has many Guschin and coworkers (21) described the utility of
well-studied virulence factors and a characteristic clinical parallel hybridization of environmental nucleic acids to
presentation. Despite this information, it is not clear how oligonucleotides immobilized in a matrix of polyacrylamide
B. pertussis interacts with host cells in disease generation. gel pads on a glass slide (oligonucleotide microchip).
In this example, the authors examined the interaction of Oligonucleotides complementary to small-subunit rRNA
B. pertussis with a human bronchial epithelial cell line sequences of selected microbial groups, encompassing key
(BEAS-2B) and measured host transcriptional profiles by genera of nitrifying bacteria, were shown to selectively
using high-density DNA microarrays. retain labeled target nucleic acid derived from either
The results showed that B. pertussis induces mucin DNA or RNA forms of the target sequences. The
gene transcription by BEAS-2B cells and then counters utility of varying the probe concentration to normalize
this defense by using mucin as a binding substrate. hybridization signals and the use of multicolor detection
A set of genes is described for which the catalytic for simultaneous quantitation of multiple probe-target
activity of pertussis toxin is both necessary and sufficient populations were demonstrated. The technique should
to regulate transcription. Host genomic transcriptional provide a powerful format for the systematic exploration
profiling, in combination with functional assays to of natural microbial diversity.
evaluate subsequent biological events, provides insight
into the complex interaction of host and pathogen. Mycobacterium Species Identification and Rifampin Resis-
tance Testing. One of the applications of high-density DNA
Biochip-Based Microbial Community Ribotyping probe arrays for bacteriology focuses on fastidious bac-
Current analytical techniques used for the determination terium diagnostics. Using oligonucleotide as the probe,
of microorganisms are rarely automated, usually require the potential of the microarray strategy for parallel test-
incubation in a suitable medium to increase the number of ing of different targets has been demonstrated in this
organisms, and generally employ visual and/or chemical study, as the same hybridization conditions could be used
methods to identify the strain or subspecies of interest. The for the three genes tiled on the mycobacterial array (16S
inherent delay in such methods frequently necessitates rRNA, rpoB, katG).
medical intervention before definitive identification of the Tuberculosis (TB), caused by members of the
nature of an infection. In industrial, public health, or M. tuberculosis complex, is one of the most common human
clinical environments, such delays may have unfortunate infectious diseases, causing three million deaths a year
consequences. Therefore, there is a concrete need for worldwide. Although the disease is associated with impov-
any possible convenient systems that are capable of erished economic conditions, TB is on the rise in many
rapid detection of microorganisms. Ribotyping is mainly industrialized nations. The spread of TB is due to immi-
based on the sequence information of the bacterial 16S gration, the emergence of drug-resistant strains, and
rRNA and, once combined with oligonucleotide microchip the AIDS epidemic. Reduced and compromised immune
techniques, should have widespread applications in function as found in newborns, infants, and immuno-
environmental and medical microbiology. suppressed individuals allows opportunistic infections
Oligonucleotide microchips are the widely used format caused by mycobacteria other than M. tuberculosis, such
for determinative and environmental studies in microbiol- as M. avium-intracellulare, M. chelonae, M. fortuitum,
ogy. Oligonucleotide microchips were originally developed M. Kansasii, M. xenopi, M. marinum, M. scrofulaceum,
for rapid sequence analysis of genomic DNA, that is, and M. zxulgai.
sequencing by hybridization with oligonucleotides in a The increasing number of mycobacterial infections
matrix. For example, a single-stranded DNA fragment has made it clinically important to quickly identify
about 200-bases long could be hybridized in parallel mycobacteria at the species level. The diagnosis of a
with all the 65,536 possible DNA 8-mers. In principle, pathogenic versus a nonpathogenic species not only has
the DNA sequence of the fragment can be reconstructed epidemiological implications but also is relevant to the
from the pattern of hybridization to the chip. However, demands of patient management. Individuals with highly
this method has not yet been adapted for large-scale contagious infections may be isolated to prevent the spread
 BIOCHIP-BASED DEVICES AND METHODS IN MICROBIAL COMMUNITY RIBOTYPING 449

of the disease. Antibiotic treatments may vary according fractionated through differential binding of double- and
to the species encountered. single-stranded forms of nucleic acids to the silica. The
The emergence of drug-resistant M. tuberculosis has procedure involves sequential washing of the column
created additional concern in the event of TB diagnosis. with different solutions. No vacuum filtration steps,
Recently, a number of genetic changes leading to rifampin, phenol extraction, or centrifugation is required. After
isoniazid, pyrazinamide, and ethambutol resistance have hybridization, the overall fluorescence pattern is captured
been characterized and allowed for the development of as a digital image or is recorded by a polaroid camera.
probe-based assays for antibiotic resistance. This three-component system was used to discriminate
Traditionally, species identification within the genus E. coli, Bacillus subtilis, Bacillus thuringiensis, and
Mycobacterium and subsequent antibiotic susceptibility human HL60 cells. The procedure is rapid: beginning
testing rely on time-consuming, culture-based methods. with whole cells, it takes approximately 25 minutes to
Despite the recent development of DNA probes, which obtain labeled DNA and RNA samples and an additional
greatly reduce assay time, there is a need for a single- 25 minutes to hybridize and acquire the microarray image
platform assay that is capable of answering the multitude using a stationary image analysis system or the portable
of diagnostic questions associated with this genus. imager. This is the first passive chip-based total analysis
Troesch and coworkers (22) described the use of a DNA system for environmental microbiology; the portable
probe array on the basis of two sequence databases: one system may be further used in microbial identification
for the species identification of mycobacteria (82 unique in medical, agricultural, or environmental applications.
16S rRNA sequences corresponding to 54 phenotypical Although the system is not perfect, it has demonstrated the
species) and the other for detecting rifampin resistance promise of applying biochip technology in environmental
(rpoB alleles). They designed an array that contained microbiology.
all the 16S rRNA polymorphisms over a 200-bp region
present in a mycobaterial database. The array also
contained 51 M. Tuberculosis rifampin resistance-causing PERSPECTIVE
rpoB mutations in a 200-bp region and 2.2 kb of
the M. Tuberculosis wild-type katG gene–this fragment It is increasingly evident that only a small fraction of the
spanning codons in which mutations were previously microbial world has been characterized, although there
shown to confer resistance to isoniazid. An experimental is little agreement about the full extent of undescribed
protocol was optimized to perform hybridization on the diversity. The development of rRNA-targeted probes that
probe array, scanning, and data analysis in one hour. The selectively target characterized phylogenetic groups now
hybridization-based probe array assay used here showed provides a format to systematically address this question.
a specificity matching the sequence polymorphisms of the These probes are fully compatible with the microchip
16S rRNA marker. In this study, a total of 26 out of 27 format.
species were correctly identified and so were all of the Biochip is one of the most promising biotechnologies
rpoB mutants, with the only discrepancy being an artifact for the twenty-first century, which has the advantage of
because of an error in the reference sequence. This parallel combining a rapid, high-throughput and portable platform
testing format opens new perspectives in terms of patient for nucleic acid hybridization. In addition to their use
management for bacterial diseases by allowing a number in identifying described phylogenetic groups, the great
of genetic tests to be simultaneously run. capacity of the microchip array could be used to evaluate
sequence motifs of the novel microbial populations. Thus,
Portable System for Microbial Sample Preparation and we anticipate that the microchip will have application not
Oligonucleotide Microoarray Analysis. The limitation with only to microbial community ribotyping but also in ongoing
microarray technology is the lack of portability and studies of global diversity.
inexpensive devices for the acquisition of hybridization
patterns. The group headed by Mirzabekov has developed Acknowledgments
a three-component system and a rapid inexpensive We thank the National Natural Science Foundation of China,
with contract number 39825108, 39889001, and the National
procedure for analysis of different microorganisms based
Key Fundamental Science Development Program, with contract
on microchip biotechnology (23).
number G19990116, for their financial support.
The system consists of three main components: (1) a
universal syringe-operated silica minicolumn for succes-
sive DNA and RNA isolation, fractionation, fragmentation, BIBLIOGRAPHY
fluorescent labeling, and removal of excess free label
and short oligonucleotides; (2) microarrays of immobi- 1. A. E. G. Cass, Biosensors: A Practical Approach, IRL press,
lized oligonucleotide probes for 16S rRNA identification; London, U.K., 1990.
and (3) a portable battery-powered device for imaging the 2. K. R. Mitchelson and J. Cheng, Capillary Electrophoresis of
hybridization of fluorescently labeled RNA fragments with Nucleic Acids, Humana Press, Totowa, N.J., 2000.
the arrays. 3. J. Cheng and L. J. Kricka, Biochip Technology, Harwood
The minicolumn combines a guanidine thiocyanate Academic Publishers, Newark, N.J., 2001.
method of nucleic acid isolation with a newly developed 4. D. L. Wise, Bioinstrumentation and Biosensor, Marcel
hydroxyl radical-based technique for DNA and RNA Dekker, New York, 1991.
labeling and fragmentation. DNA and RNA can also be 5. C. R. Woese, Microbiol. Rev. 51, 221–271 (1987).
450 BIOCONTROL, MICROBIAL AGENTS IN SOIL

6. C. R. Woese, Proc. Natl. Acad. Sci. U.S.A. 95, 11043–11046 ecologically safer biological methods of plant disease
(1998). control (2–6).
7. J. Cheng et al., Anal. Chem. 70, 2321–2326 (1998). Biological control can be defined as ‘‘any conditions
8. M. A. Northrup et al., Anal. Chem. 70, 918–922 (1998). under which survival or activity of a pathogen is reduced
9. J. Yang et al., Anal. Chem. 71, 911–918 (1999). through the agency of any other natural or modified
10. P. Wilding et al., Anal. Biochem. 257, 95–100 (1998). living organisms and genes or gene products (delivered by
11. X. B. Wang et al., J. Phys. D: Appl. Phys. 26, 1278–1285 organisms) as well, with the result that there is reduction
(1993). in incidence of the disease caused by the pathogen’’ (2).
12. L. A. Christel, J. Biomech. Eng. 121, 22–27 (1999). Chemicals produced and delivered by living organisms
represents biological control, whereas chemicals extracted
13. J. Cheng et al., Nucleic Acids Res. 24, 380–385 (1996).
from living organisms represents chemical control. A
14. P. Belgrader et al., Anal. Chem. 73, 286–289 (2001).
beneficial organism used to protect plants is referred
15. M. Ogura et al., Clin. Chem. 44, 2249–2255 (1998).
to as a ‘‘biological control agent’’ (BCA), ‘‘biopesticide,’’
16. S. Seetharaman et al., Nat. Biotechnol. 19, 336–341 (2001). ‘‘microbial pesticide’’ or, simply as ‘‘antagonist,’’ because
17. J. Cheng et al., Nat. Biotechnol. 16, 541–546 (1998). it ‘‘antagonizes,’’ or interferes with, the target organisms
18. P. Cossart et al., Science 271, 315–316 (1996). that are damaging the plant.
19. C. E. Belcher et al., Proc. Natl. Acad. Sci. U.S.A. 97, Antagonists are generally naturally occurring, mostly
13847–13852 (2000). soil microorganisms, which have some trait or character-
20. R. Rappuoli, Proc. Natl. Acad. Sci. U.S.A. 97, 13467–13469 istic that enables them to interfere with pathogen or pest
(2000). growth, survival, infection, or plant attack. Usually, they
21. D. Y. Guschin et al., Appl. Environ. Microbiol. 63, 2397–2402 have little effect on other soil organisms, leaving the nat-
(1997). ural biology of the ecosystem more balanced and intact
22. A. Troesch et al., J. Clin. Microbiol. 37, 49–55 (1999). than would a broad-spectrum chemical pesticide. Several
23. S. G. Bavykin et al., Appl. Environ. Microbiol. 67, 922–928 plant diseases caused by soilborne plant pathogens have
(2001). been shown to be inhibited by soils. Soil suppressive-
ness may be constitutive, an inherent property of the soil
regardless of its cropping history, or adaptive, whereby
soil suppression is only achieved after a specific cultural
BIOCONTROL, MICROBIAL AGENTS IN SOIL practice, such as monocropping, has been adopted. Soil
suppressiveness of Fusarium wilt of melons (7) and of
ILAN CHET Gaeumannomyces graminis var. tritici (Ggt)-take-all of
LEONID CHERNIN wheat (2) are examples of constitutive and adaptive soil
The Hebrew University of suppression, respectively. Soil suppression is considered
Jerusalem a rather complex phenomenon in which the antagonists
Rehovot, Israel
establish themselves in the soil and interfere with the
growth or survival of the pathogenic microorganism. Sev-
Plant pathogens affect crops throughout the world and eral microorganisms can be involved in soil suppression of
are a major obstacle to effective food supplies worldwide. a single disease. Studies of why some soils can naturally
Losses due to plant diseases may be as high as $30 to $50 suppress plant diseases while others cannot have provided
billion in cultivated and stored crops and 20% of total food the basis of most research on BCAs.
production annually. The most efficient way of preventing Biological approaches to the control of soilborne plant
crop diseases caused by these pathogens is to use chemicals pathogens were discovered more than 70 years ago
to disinfect the soil against pathogens and pests. However, in the classical works of G. Sanford and R. Weindling,
pesticides, common in agricultural practice, create a however, it was only in the 1960s that biological control
health hazard for many nontarget organisms. Their use research took on a more practical approach (8). Since that
becomes more limited every year because of concern of time, many agrochemical and biotechnological companies
legislators for human and animal health, as well as for throughout the world have increased their interest and
quality of life and the environment. In addition, excessive investment in the biological control of plant diseases
application of pesticides has led to an increased proportion and pests. The current market for BCAs is estimated
of pathogens, which are resistant to the chemicals. Thus, at only $500 million, which is about 1% of the world’s
each year, diseases cause millions of dollars worth of crop total output for crop protection. The largest share of this
damage, despite extensive use of pesticides. Moreover, market involves biopesticides marketed for insect control
many nonspecific chemical pesticides currently widely (mainly products based on Bacillus thuringiensis strains
used against soilborne diseases (e.g., methyl bromide) which produce a protein toxin with strong insecticidal
are soon to be taken off the market as a result of activity) and these bioinsecticides represent around 4.5%
registration cancellation (1). Therefore, the development of the world’s insecticide sales. Other agents used for
of safer, environmentally friendly control alternatives that biocontrol exist on a much smaller scale commercially.
promote sustainable agroecosystems is urgently required However, the biopesticides market is expected to grow
and has become a research priority in developed countries. over the next 10 years at a rate of 10 to 15% per annum,
Agricultural biotechnology offers a new approach to the versus 1 to 2% for chemical pesticides (1). Despite the
problem: the development of alternative, efficient, and fact that the total register of commercially available
 BIOCONTROL, MICROBIAL AGENTS IN SOIL 451

biocontrol products is still small, their number has T. harzianum, capable of lysing mycelia of S. rolfsii and
increased more than twofold in the last few years. R. solani by application to pathogen-infested soil, signifi-
For plant pathogens alone, the current list of microbial cantly reduced the diseases caused by these fungi in bean,
antagonists available for use in commercial disease cotton, tomato, and eggplant seedlings under greenhouse
biocontrol includes around 40 preparations (3,5). These conditions. A seed treatment was developed to reduce
are all based on the practical application of several species the amount of Trichoderma added to the soil to control
of nonpathogenic bacteria and fungi able to suppress plant soilborne plant pathogenic fungi. Trichoderma hamatum
pathogen growth (Table 1). The main topics discussed in conidia applied in the laboratory to seeds of pea and radish
this review are the following: the: biocontrol of soilborne provided protection to seeds and seedlings from Pythium
diseases and nematodes by fungal and bacterial biocontrol spp. and R. solani, respectively, almost as effectively as
agents, mechanisms of biological control, formulation, and fungicide seed treatment. A T. harzianum isolate applied
delivery of biocontrol agents, combined usage of biocontrol to either soil or rooting mixture efficiently controlled
agents, integrated pest management, and molecular damping-off induced by P. aphanidermatum in several
approaches for improvement of biocontrol agents. crops with disease reduction of up to 85%. Commercial
preparations based on strain T. harzianum T-22 (Table 1)
are sold to the greenhouse, row crop, and turf industries.
BIOCONTROL OF SOILBORNE DISEASES
Applied as a seed treatment, as a drench or incorporated
into the soil as granules, the strains significantly increase
Fungi as Biocontrol Agents
yield of field corn, wheat, potato, and other crops. Granules
Fungi are relatively easy to grow and formulate for large- of T-22 added to potting materials used in the production
scale application. The first fungal antagonist to have of vegetables and ornamentals reduced the incidence of
been introduced commercially was Peniophora (Phlebia) diseases caused by Fusarium and other pathogens. The
gigantea, used for the control of Heterobasidium (Fomes) strain was shown to be highly rhizosphere competent,
annosum root rot on pines (9). Since then several other resulting in long-term root colonization and substantial
fungi have been marketed as BCAs (Table 1). Among advantages to plant growth and productivity (13).
them, the necrotrophic mycoparasitic fungus Gliocladium Biological control of wilt pathogens that penetrate the
virens Miller, Giddens, and Foster (=Trichoderma virens vascular system of their hosts (e.g., Fusarium oxysporum
Miller, Giddens, Foster, and von Ark) is a common soil or Verticillium dahliae) and that may attack the plant
saprophyte and one of the most promising and studied at various growth stages is considered to be less feasible
fungal BCAs. Trichoderma, a genus of hyphomycetes and than control of plant pathogens, which induce damping-
an anamorphic Hypocreaceae (class Ascomycetes) closely off. In the latter case, the host is only susceptible to
related to Gliocladium, is also common in the environ- the pathogen for a limited time, at the early growth
ment, especially in soils. Species of this genus have been stage of the plant, after which it is not susceptible to
used in the production of cellulolytic and hemicellulolytic infection. Thus, for biocontrol of damping-off, protection
enzymes, biological control of plant disease, biodegrada- is required for only a limited time. This might be the
tion of chlorophenolic compounds, and soil bioremediation. reason for the slower progress in developing introduced
The first report on a biological control experiment using antagonists for biocontrol of wilt pathogens. Nevertheless,
Trichoderma spp. under natural field conditions came T. harzianum T-35 isolated from the rhizosphere of
from Wells and coworkers (10), who used Trichoderma cotton plants grown in fields infested with Fusarium
harzianum grown on an autoclaved mixture of ryegrass significantly reduced the disease incidence caused by
seeds and soil to control Sclerotium rolfsii Sacc. Since that F. oxysporum f. sp. vasinfectum, F. oxysporum f. sp.
time, more Gliocladium and Trichoderma isolates have melonis and F. roseum Culmorum in cotton, melon, and
been obtained from natural habitats and used in biocon- wheat, respectively, in greenhouse trials when applied
trol trials against several soilborne plant pathogenic fungi as a peat-bran preparation, a conidial suspension or a
under both greenhouse and field conditions (11,12). In seed coating. In trials over two successive growing seasons
particular, isolates of G. virens, Gliocladium roseum, Tri- against Fusarium crown rot of tomato in fields naturally
choderma viride, T. harzianum Rifai, and Trichoderma infested with F. oxysporum f. sp. radici lycopersici, this
hamatum have been reported to be antagonists of many isolate provided a significant increase in the total yield
phytopathogenic fungi, including Botrytis cinerea, Fusar- of tomatoes (11). This and several other isolates of
ium spp., Phytophthora cactorum, Pythium ultimum, Trichoderma are produced commercially for the biocontrol
Pythium aphanidermatum, Rhizoctonia solani, Sclerotinia of soilborne and aerial fungal diseases (Table 1).
sclerotiorum, S. minor, S. rolfsii, and several other fungal Other fungal BCA, Pythium nunn, a necrotrophic
pathogens that cause soilborne and foliage diseases in a fungus with a limited host range, was isolated from
wide variety of economically important crops under a wide soil suppressive to a plant parasitic Pythium sp. The
range of environmental conditions. The antagonists kill fungus efficiently suppresses growth of P. ultimum,
the host by direct hyphal contact, causing the affected P. aphanidermatum, R. solani, and Phytophthora para-
cells to collapse or disintegrate; vegetative hyphae of all sitica by penetration and/or lysis of the host cell wall
species have been found susceptible. Thus, treatment of at the site of interaction. It can be used alone or in
soil infested with propagules of R. solani or P. ultimum combination with T. harzianum. The fungus Talaromyces
with G. virens resulted in a significant reduction in the flavus (Klocker) A.C. Stolk and Samson is a mycopara-
number of viable R. solani sclerotia. Several isolates of site of several soilborne plant pathogenic fungi, including
Table 1. Examples of the Microbial Products Containing Living Organisms Currently Employed for Biocontrol of Crop
Diseases and Nematodes
Examples of
Biocontrol Target Examples of
Organism Trade Name Formulation Application Pathogens/Effects Target Crop

Bacteria
Agrobacterium Galltrol-A, Culture grown on Bacterial suspension Crown gall disease Fruit, nut, and
radiobacter strain Nogall, agar, culture applied to seeds, caused by A. ornamental
K84 Diegall, suspensions seedlings, cuttings, tumefaciens nursery stock
Norbac 84 C roots, stems, and as soil
drench
Bacillus subtilis HiStick N/T, Dry powder Water-based slurry seed Alternaria spp., Barley, beans,
strains GB03, Kodiak, usually applied treatment; water-based Aspergillus spp., corn, cotton,
FZB24 and others Kodiak HB, with chemical suspension for Fusarium spp., R. peanut, pea,
Kodiak AT, fungicides in-furrow treatments, solani, Streptomyces potatoes, rice,
Rhizo-Plus, hopper box treatment, scabies Verticillium soybean,
Rhizo-Plus seed treatment, soil ornamental
Konz, drench, dip, and plants
Serenade, addition to nutrient
Subtilex, solutions
System 3
Burkholderia cepacia Dany, Intercept Peat-based dried Applied to seeds with a Fusarium spp., Pythium Alfalfa, barley,
biomass from sticking agent in sp.; R. solani, beans, clover,
solid planter box (aqueous nematodes cotton, peas,
fermentation; suspension formulation sorghum,
aqueous is for use in drip vegetables,
suspension irrigation or as a wheat
seedling drench)
Pseudomonas Spot-Less Liquid Over-head irrigation Anthracnose, Dollar spot, Turf grass
aureofaciens strain Michrochium patch
Tx-1 (pink snow mold), P.
aphanidermatum
Pseudomonas Cedomon Seed treatment Seed dressing Fusarium sp., Leaf stripe Barley and oats;
chlororaphis and spot. potential for
wheat and other
cereals
Pseudomonas BlightBan A506 Wettable powder Bloom time spray of the Frost damage, Erwinia Almond, apple,
fluorescens A506 flower and fruit amylovora, and apricot,
russet-inducing blueberry,
bacteria cherry, peach,
pear, potato,
strawberry, and
tomato
Pseudomonas syringae Bio-save 100, Frozen cell Pellets added to water to Botrytis cinerea, Pome fruit, citrus
strains ESC10, Bio-save 1000 concentrated produce liquid Geotrichum candidum,
ESC11 pellets suspension, Mucor pyroformis,
postharvest application Penicillium spp.,
to fruit as drench, dip,
or spray
Streptomyces Mycostop Powder Drench, spray, or through Alternaria brassicola, Field, ornamental,
griseoviridis strain irrigation system Botrytis spp., Fusarium and vegetable
K61 spp., Phomopsis spp., crops
Phytophthora spp.,
Pythium spp., and

Fungi
Ampelomyces AQ10 Water-dispersible Spray Powdery mildew Apples, cucurbits,
quisqualis granules grapes,
ornamentals,
strawberries,
and tomatoes
(continued overleaf )

452
Table 1. (Continued)

Examples of
Biocontrol Target Examples of
Organism Trade Name Formulation Application Pathogens/Effects Target Crop

Coniothyrium Contans, Kony Water-dispersible Spray, granules Sclerotinia sclerotiorum, Canola, sunflower,
minitans granule incorporated into soil or S. minor peanut, soybean,
soilless mix vegetables, and
ornamental
flowers in
greenhouse
production
Fusarium oxysporum, Biofox C Spores, dust, or In drip to rock wool; Fusarium oxysporum, Asparagus, basil,
non-pathogenic Fusaclean alginate incorporate in potting Fusarium moniliforme carnation,
granules mix; soil incorporation; cyclamen,
seed treatment gerbera, and
tomato
Gliocladium Primastop Wettable powder Drench and incorporation Botrytis spp., Didymella Greenhouse crops
catenulatum spp., Pythium spp.,
Rhizoctonia spp.,
Gliocladium virens SoilGard Alginate granules Granules incorporated in Pythium spp. and R. Ornamentals, food
strain GL-21 soil or soilless growing solani crops
media prior to seeding
Paecilomyces lilacinus Paecil Dry spore Seedling or soil drench Various nematode spp. Banana, wheat,
concentrate and potatoes
Phlebia gigantea Rotstop, P.g. Spores in inert Spray, chain saw oil Heterobasidium annosum Trees
Suspension powder
Pythium oligandrum Polyver-sum Wettable powder Root and stem drench, Aphanomyces spp., Canola, Cereals,
spray Alternaria spp., fruits, forest
Botrytis spp., Fusarium nurseries,
spp., G. graminis , legumes,
Pythium spp., vegetables, and
Phytophthora spp., R. ornamental
solani, S. cepivorum, plants
Tilletia caries
Talaromyces flavus Protus WG Water-dispersible Soil or seed treatment, Rhizoctonia solani, V. Cucumber, tomato,
isolate V117b powder soil drench, root dip dahliae, V. alboatrum and rapeseed oil
application
Trichoderma Trichoderma- Peat-bran Incorporated into soil or Fusarium spp., Pythium Nursery and field
harzianum strain Y 2000 preparation potting medium spp., R. solani, S. rolfsii crops
Trichoderma PlantShield, Wettable powder, Granules added Fusarium spp., Pythium Bean, cabbage,
harzianum strain RootShield, granules, or dry in-furrow, by broadcast spp., R. solani, corn, cotton,
T-22 (KRL-AG2) RootShield powder application to turf, Sclerotinia cucumber,
Home and mixed with greenhouse homoeocarpa peanut, potato,
Garden, T-22 soil, or by mixing sorghum,
Planter Box, powder with seeds in soybean, sugar
Turf Shield, planter box, or in beet, tomato,
commercial seed turf, greenhouse
treatment slurry ornamentals
T. harzianum strain Trichodex Wettable powder Spray Botrytis cinerea, Cucumber, grape,
T39 Collectotrichum spp., nectarine,
Fulvia fulva, Monilia soybean,
laxa, Plasmopara strawberry,
viticola, sunflower,
Pseudoperonospora tomato
cubensis, Rhizopus
stolonifer, S.
sclerotiorum
T. viride Pers ex Gray EcoSOM, Trieco Powder Dry or wet seed, tuber, or Fusarium spp., Pythium Tobacco, cereals,
set dressing or soil spp., Rhizoctonia spp. citrus, coffee,
drench, cotton, grapes,
spread/broadcast over oilseeds,
field soybean, tea,
tomato,
vegetables, and
others
(continued overleaf )

453
454 BIOCONTROL, MICROBIAL AGENTS IN SOIL

Table 1. (Continued)

Examples of
Biocontrol Target Examples of
Organism Trade Name Formulation Application Pathogens/Effects Target Crop

Trichoderma Bio-Fungus, Granular, wettable Applied after fumigation; Armillaria, Flowers,

harzianum other Root Pro, powder, sticks, incorporated in soil; Botryosphaeria, strawberry,
strains, Supresivit and crumbles sprayed or injected Chondrosternum, trees, vegetables
Trichoderma spp. Fusarium, Nectria,
Phytophthora,
Mixture of T. Trichopel, Pythium
harzianum and T. Trichoject, spp.,Rhizoctonia,
viride Trichoseal, Sclerotinia,
Trichodo-wels Verticillium
Mixture of T. Binab T Wettable powder Spray, mixing with Pathogenic fungi that Flowers, fruit,
harzianum and and pellets potting substrate, cause wilt, take-all, ornamentals,
T. polysporum mixing with water and root rot, and internal turf, and
painting on tree decay of wood products vegetables
wounds, inserting and decay in tree
pellets in holes drilled wounds
in wood

Modified from D. Fravel’s list of commercial biocontrol products for use against soilborne crop diseases, updated on September 20, 2000 and available through
the Internet (http://www.barc.usda.gov/psi/bpdl/bpdlprod/bioprod.html).

S. sclerotiorum, R. solani, and Verticillium spp. The mech- Bacteria as Biocontrol Agents
anism by which T. flavus controls plant pathogenic fungi
Many strains of soilborne and rhizospheric bacteria have
is not always clear and in some cases, it may involve
been used for biocontrol of plant disease. The bacteria are
a combination of antibiosis and mycoparasitism. The
easy to grow and can be applied as a seed treatment or as a
necrotrophic mycoparasite Coniothyrium minitans Camp-
soil drench, and mechanisms by which they reduce disease
bell has been found to be a natural destructive mycopara-
have been studied. The bacteria appear to protect plants
site of S. sclerotiorum, which kills both hyphae and sclero-
against a wide range of pathogens and the potential for
tia of this phytopathogen. Infection by C. minitans actually
provided natural biological control of S. sclerotiorum and commercial utilization is promising. Bacterial BCAs can
Sclerotium cepivorum under greenhouse conditions and in reduce disease by different modes of action, including
the field. Coniothyrium minitans could be recovered from production of antibiotics, toxins and lytic enzymes, direct
the soil more than a year after its application, and during parasitism of pathogen hyphae or propagules, competitive
this time it spread throughout the greenhouse, indicating exclusion by occupation of infection sites and/or depletion
its potential for long-term control. The fungus Fusarium of nutrients, plant growth enhancement and induction of
chlamydosporum was shown efficient against groundnut host resistance mechanisms (3).
rust caused by Puccinia arachidis, whereas the soilborne Most bacterial BCAs are gram-negative. Among them,
fungal BCAs, Trametes versicolor, and Pleurotus eryn- Agrobacterium radiobacter strain 84 is probably the best
gii were effective against F. oxysporum f. sp. lycopersici example of the application of a bacterial antagonist for the
race 2 on tomato. The fungus Penicillium purpurogenum biocontrol of an economically important plant pathogen
protected peach and tomato against the plant pathogens Agrobacterium tumefaciens, a causal agent of crown gall
Monilinia laxa and F. oxysporum f. sp. lycopersici. Iso- disease in a large number of crops and considered one of
lates of T. flavus and Stenotrophomonas maltophilia were the most destructive agricultural pathogens. Introduction
shown to antagonize S. rolfsii and P. ultimum, respec- of this strain into agricultural practice was the first case
tively. The biotrophic mycoparasite Sporidesmium scle- of successful commercialization of a BCA (16).
rotivorum Uecker, Ayers, and Adams is a dermatiace- Rhizospheric bacteria belonging to the fluorescent
sus hyphomycete, which was isolated from field soil. Pseudomonas spp. are receiving increasing attention
The fungus has been found to be an obligate parasite for the protection of plants against soilborne fungal
on sclerotia of S. sclerotiorum, S. minor, S. trifoliorum, pathogens (3). These bacteria, defined as root-inhabiting,
Sclerotium cepivorum, and B. cinerea under natural con- plant growth-promoting rhizobacteria (PGPR) have been
ditions (14). The fungus Ampelomyces quisqualis Ces. has shown to improve plant health and increase yield.
been reported as a biotrophic mycoparasite and BCA of PGPR may exert their effect directly and/or indirectly
many fungi, which cause powdery mildew. An isolate of by the suppression of plant pathogens, thereby providing
A. quisqualis M-10 obtained from an Oidium sp. infecting biological control (17). Direct promotion of plant growth
Catha edulis in Israel was developed as a BCA (prepa- may be exerted by several mechanisms, for example,
ration AQ10–Table 1) of several powdery mildew fungi via the production of auxin and other plant hormones,
belonging to the genera Oidium, Erysiphe, Sphaerotheca, synthesis of siderophores, biological nitrogen fixation,
Podosphaera, Uncinula, and Leveillula (15). solubilization of minerals such as phosphorus, or via
 BIOCONTROL, MICROBIAL AGENTS IN SOIL 455

enzymatic activities such as 1-aminocyclopropane-1- nematodes are root-knot (Meloidogyne), which have
carboxylate (ACC) deaminase, an enzyme, which decreases multiple generation per year and are the leading cause
the endogenous concentration of ethylene. PGPR strains of crop loss due to plant parasitic nematodes, and cyst
producing this enzyme may have a competitive edge over (Heterodera and Globodera) nematodes that have a single
other microorganisms in the rhizosphere because they generation per year and cause damage mainly in the early
can use ACC as a source of nitrogen (18). Several other root-infection process.
mechanisms have been proposed to account for PGPR’s Natural enemies of nematodes include viruses, rick-
disease-suppressive ability, including antagonism, and ettsias, bacteria, and nematophagous fungi. Antagonists
induced resistance. However, before the deliberate use most likely to be receptive to management for the biocon-
of PGPR and other bacteria as biofertilizers and BCAs, trol of nematodes are trapping fungi, fungal egg pathogens
it is necessary to know some key parameters, such or parasites, endoparasitic fungi, fungal pathogen or
as root colonization capacity, location of infection site, parasites of females, endomycorrizial fungi, plant growth-
and degree of persistence of the inoculum. The most promoting bacteria, obligate bacteria parasites. The para-
suitable techniques for studying these parameters are sitoid endospore-forming bacterium Pasteuria penetransis
microscopical and immunological. able to survive in the soil under extreme conditions and
One of the most intensively studied biocontrol systems destroys nematodes by parasitic action. This bacterium is
using rhizobacteria was generated after the isolation of an obligate parasite of many nematode species thought
Pseudomonas fluorescens strains from a soil naturally
to be responsible for the decline in nematode populations
suppressive to Gtg-take-all of wheat. These fluorescent
in some stable monocultures (e.g., sugarcane) and peren-
rhizosphere-colonizing bacteria, when applied as seed
nial cropping systems (e.g., grapes). Pasteuria penetrans
treatments to a soil conducive to the diseases, provided
has been an effective biocontrol agent against root-knot
significant disease control. Other pseudomonad species
nematodes. An application of endospores to soil prevent
with biocontrol activity have been described. Among
juvenile females attacking roots. The limitation of the bac-
them strains of Burkholderia (Pseudomonas) cepacia,
terium practical sense is that it can be cultured only in its
Pseudomonas aureofaciens, P. chlororaphis, P. corrugata,
P. putida, and P. syringae (3,5). nematode host or in association with host cells.
Besides Pseudomonas spp., several other gram- Other genera of bacteria known to reduce nema-
negative bacteria exhibit highly efficient biocontrol activ- tode populations mainly by colonizing the rhizosphere
ity. For example, the bacterium Enterobacter cloacae was of the host plant include Agrobacterium, Alcaligenes,
shown to be a capable BCA of different rot and pre- Bacillus, Clostridium, Desulfovibrio, Pseudomonas, Ser-
emergence damping-off diseases incited by Pythium spp., ratia, Streptomyces, and Telluria. spp. Application of
as well as of fusarium wilt and of some other plant some of these bacteria has shown very promising
diseases caused by fungal pathogens (19,20). Biocontrol results (24). Fungi described as efficient for nematode bio-
potential was also demonstrated for strains of Serra- control include mainly nematode-trapping species (e.g.,
tia marcescens and Enterobacter agglomerans producing Arthrobotrys oligospora) possessing the ability to capture,
chitinolytic enzymes (21,22). kill, and utilize nematodes by using specially adapted
Gram-positive bacteria of Bacillus spp. are widely mycelial structures called traps, and egg-parasitic species
dispersed in nature, easy to reproduce, have a long (e.g., Verticillium chlamidosporium, Paecilomyces lilaci-
shelf life when sporulated, and are nonpathogenic. nus T. harzianum, and Fusarium spp.) whose action in
The ability to produce endospores provides bacilli with reducing the nematode populations is rather dramatic
a high tolerance to heat and desiccation. Therefore, compared with other parasitic or predacious fungi. How-
they have a significant ecological advantage over other ever, the endoparasitic fungi are most specific for their
bacterial antagonists (23). Several strains of Bacillus are nematode host. They produce conidia or resting spores
now produced commercially as BCAs (Table 1). Because that are surrounded by adhesive material that allows the
Bacillus subtilis is a spore-forming organism, it is spores to attach to the nematode cuticle.
extremely tolerant of environmental stresses, including The most important obligate parasites of female cyst
seed treatment pesticides, soil and seed pH, cultivar nematodes of the genus Heterodera are Nematophthora
effects, edaphic factors, and long-term storage. It is readily gynophila and Catenaria auxiliaris. Thus, N. gynophila
produced with current fermentation technology. Other takes up to seven days to kill female cyst nematodes.
BCAs, such as Pseudomonas spp., do not readily adapt to The resting spores can survive at least five years
large-scale production methods, and stability is a limiting in the absence of cyst nematode females. Combined
factor. activities of N. gynophila and facultative parasite of
females V. chlamidosporum are largely held responsible
BIOCONTROL OF NEMATODES for natural reduction in Heterodera avanae, a cereal cyst
nematode.
Nematodes are small, round worms, which are common Endomycorrhizial fungi limit the densities of plant
inhabitants of many natural environments. Some of them parasitic nematodes on a broad range of host plants. Myc-
are pathogenic to plants. Plant parasitic nematodes are orrhizial fungi Gigaspora and Glomus decrease damage
major agronomic pests, with the annual global loss caused by root-knot nematodes, especially where trans-
in agriculture of U.S. $100 billion worldwide caused plants are used in crop production. Endotrophic mycor-
by nematode damage to plants. The most important rhizial fungus Glomus mossea was shown efficient toward
456 BIOCONTROL, MICROBIAL AGENTS IN SOIL

Meloidogyne incognita and Rotylenchulus reniformis in phytopathogenic fungi (e.g., Fusarium spp.) and bacteria
cotton, cucumber, tomato and so on (25,26). (e.g., Pseudomonas syringae). As a result, these plant
pathogens end up in an iron-deficient environment
deleterious to their growth.
MECHANISMS OF BIOLOGICAL CONTROL
A competitive exclusion mechanism was shown respon-
sible for the biological control of P. ultimum -incited seed
Biocontrol strategies may be divided into two broad cate-
infections by E. cloacae in which E. cloacae prevents the
gories. One follows a fundamentally ecological approach.
germination of the seed-rotting oomycete P. ultimum spo-
This strategy has been termed classical biocontrol or bio-
rangia by the efficient metabolism of fatty acid components
control with one-time introduction (27). The other uses
of seed exudate and thus prevents seed infections. It was
microorganisms as biopesticides and resembles, in some
conclusively demonstrated (a) that unsaturated fatty acids
important respects, the approach similar to chemical pes-
from seeds and roots are required to elicit germination
ticide treatment (which aims at control for a limited period
of P. ultimum around planted seeds, and (b) that very
of time). This strategy has also been referred to as aug-
rapid metabolism of unsaturated fatty acids that leach
mentative biocontrol. In the ecological approach, selected
from seeds is the primary mechanism by which E. cloacea
organisms must be able to function in the same environ-
controls Pythium seed rot (28).
mental niche as that of the pathogen they are meant to
Antagonists capable of more efficiently utilizing essen-
control. Biological control of plant diseases can be achieved
tial resources (e.g., carbon, nitrogen, volatile organic mate-
directly, through introduction of specific microbial antag-
rials, plant residues, iron, and microelements) effectively
onists into the soil or plant material, or indirectly, by
compete with the pathogen for its ecological niche and col-
changing the conditions prevailing in the plant’s envi-
onization of the rhizosphere and/or phyllosphere, leaving
ronment, and thus the microbiological equilibrium of its
the pathogen less efficient to grow in the soil or colonize the
ecosystem. In some cases, a combination of indirect and
plant. An important attribute of a successful rhizosphere
direct approaches leads to the best biocontrol results.
BCA is the ability to remain at a high population density
on the root surface, providing protection of the whole root
Direct Biological Control for the duration of its life. The ability of a microorganism
The direct approach includes four general direct mech- applied as a seed treatment to proliferate and establish
anisms of biological control of plant diseases by soil along the developing root system has been termed ‘‘rhi-
antagonistic microorganisms: (1) ‘‘competition’’ with the zosphere competence’’ (29). For example, competition for
pathogen for limited resources, such as space or nutrients, carbon and nitrogen in the rhizosphere, as well as rhizo-
this is considered an important mechanism of biocon- sphere competence, may be involved in the biocontrol of
trol by bacterial and fungal antagonists; (2) ‘‘antibiosis,’’ F. oxysporum by T. harzianum (11).
which is the inhibition or destruction of an organism by
the secondary metabolites produced by another organ- Antibiosis. This mechanism includes small toxic mole-
ism; (3) ‘‘production of lytic enzymes’’ (e.g., chitinases, cules, volatiles, and lytic enzymes that kill the target
glucanases, cellulases, and proteases) able to degrade the organism. Antibiosis is restricted for the most part to
pathogen cell wall and/or to destroy its metabolic activi- those interactions that involve a chemically heterogeneous
ties; (4) ‘‘predation or parasitism,’’ the phenomenon that group of low molecular weight (less than 1 kDa) diffusible
occurs when one fungus exists in intimate association with organic compounds (known as antibiotics) produced by
another from which it derives some or all its nutrients a microorganism, which at low concentrations, inhibit
while conferring no benefits in return. the growth or metabolic activities of another microorgan-
ism (30,31). In recent years, genetic and molecular studies,
Competition. Many plant pathogens require exogenous specific and sensitive assays, and detection systems have
nutrients to successfully germinate, then penetrate and provided evidence of the contribution of antibiotics to
infect the host tissue. In the rhizosphere, competition pathogen suppression and disease reduction in vitro and in
for space and nutrients is of major importance. Therefore, situ, under greenhouse and field conditions. Under natural
competition for limiting nutritional factors, mainly carbon, conditions, antibiotics participate in soil suppressivenes
nitrogen, and iron, may result in the biological control and the antagonism of phytopathogens. However, even in
of plant pathogens. Indeed, competition is one of the cases where antifungal metabolite production by an agent
main mechanisms involved in controlling plant pathogens. definitely reduces disease, other mechanisms may also be
Most work on bacterial competition has concentrated on operating.
iron (Fe). Many plant-beneficial rhizospheric bacteria,
for example, those that belong to the pseudomonads Antibiotics Produced by Bacterial Biocontrol Agents. A
(P. fluorescens and P. putida, to name a few) produce large number of antibiotics produced by bacterial BCAs
fluorescent siderophores, that is, low molecular weight appear to be involved in the biocontrol of plant pathogens.
microbial iron-transport cofactors. These high-affinity Many individual strains can produce a set of antibiotics
ferric iron (Fe3+ ) chelators specifically enhance the and secondary metabolites important in biocontrol.
bacteria’s iron acquisition by binding to membrane-bound One of the best known is biocontrol of A. tumefaciens
siderophore receptors. Siderophore production by the strains carrying the nopaline-type Ti plasmids, which
fluorescent pseudomonads leads to severe iron depletion are causative agents of crown gall disease in many
in the rhizosphere and can limit the growth of some dicotyledonous plants (16). The soilborne nonpathogenic
 BIOCONTROL, MICROBIAL AGENTS IN SOIL 457

strain A. radiobacter K84 produces the bacteriocin agrocin fenpiclonil and fludioxonil, with increased stability under
84, a low–molecular weight adenosine derivative with a light due to substitution of the chlorine on the pyrrole
very specific host range against pathogenic A. tumefaciens. ring with a cyano group, were developed by Ciba-Geigy
A large number of pseudomonads active as biocon- (Novartis) company as environmentally safe agricultural
trol strains are able to produce a diverse array of potent fungicides.
antifungal metabolites. The most well-characterized are Plt, an antibiotic that consists of a resorcinol ring linked
simple metabolites such as phenazines (Phz), 2,4- to a bi-chlorinated pyrrol moiety, is produced by biocontrol
diacetylphloroglucinol (Phl), pyoluteorin (Plt), pyrrolni- strains Pf-5 and CHAO of P. fluorescens, and has been
trin (Prn), hydrogen cyanide (HCN), and some oth- shown to be most toxic to Pythium spp. However, the role of
ers (30,32–35). Plt in biological control varies depending on the patterns of
Pnz are nitrogen-containing heterocyclic molecules, two its gene expression and production by the bacteria, which
of the main ones being phenazine-1-carboxylic acid (PCA) differ in the rhizosphere and spermosphere of various
and 2-hydroxy-phenazine-1-carboxylic acid (2-OH-PCA). plant hosts.
The antibiotics have fungistatic effect on G. graminis HCN is a volatile compound shown to be enzymatically
var. tritici (take-all disease of wheat) growth. They generated from glycine by several P. fluorescens strains
act as electron donor or acceptors, which interfere and this trait is linked to the bacteria’s ability to
with normal membrane function in target organisms suppress the black root rot disease of tomato and tobacco,
resulting in increased level of toxic intracellular active caused by the fungus T. basicola. Production of HCN is
oxygen products. Phz production by strain 30–84 of stimulated by iron. HCN has a direct effect on plant roots
P. aureofaciens contributes to its ecological competence (stimulates root hair formation) and may induce plant
on plant roots in competition with indigenous microflora defense mechanisms.
in natural soils. Therefore, Phz production was shown to Antibiotics providing strain-producer biocontrol activ-
be required for both BCA survival and pathogen inhibition. ity have been described in several gram-positive bacteria,
Phl, a member of the phloroquinoline class of mainly of Bacillus spp. Thus, B. subtilis strains with
compounds, is a major factor in the biological control antifungal activity have been shown to produce the anti-
of Thielaviopsis basicola (black root rot), G. graminis fungal lipopeptides iturin A and surfacin. Antibiotics
var. tritici (take-all disease of wheat), and a range of active against various plant pathogens have also been
other plant pathogens. Its antifungal activity may be found in strains of Bacillus cereus, Bacillus pumilus, and
realized via disruption of the fungal membrane. Phl was Bacillus licheniformis. Many of these bacilli are generally
found in the rhizosphere of plants inoculated with Phl- soil-inhabiting bacteria or exist as epiphytes and endo-
producing bacteria and in the rhizosphere of wheat grown phytes in the spermosphere (23). For this reason, Bacillus
in suppressive soil. Production of Phl by bacteria may species among other strains in the genera Pseudomonas,
activate the host plant’s defense systems. Burkholderia, Enterobacter, and Trichoderma spp. known
Prn, or 3-chloro-4-(2 -nitro-3 -chlorophenyl)-pyrrole, as excellent root colonizers, both internally and exter-
synthesized from tryptophan, possesses wide range of nally, are good candidates for use as BCAs against soil
antimicrobial activity. This secondary metabolite may borne pathogens. Examples of bacterial antibiotics that
interfere with normal membrane function in the target may be involved in disease biocontrol are presented in
organism. Prn-producing strains of Burkholderia cepacia, Table 2.
P. fluorescens, E. agglomerans, and Serratia plymuthica
were shown able to control P. ultimum, R. solani, and Antibiotics Produced by Fungal Biocontrol Agents. Spe-
some other fungi. However, Prn is highly sensitive to sun- cies of Gliocladium, and Trichoderma, T. flavus, Chaeto-
light and therefore was found to be unsuitable for use mium globosum, and some others produce a range
as an agricultural fungicide. Derivatives of Prn, called of antibiotics, which are active against pathogens in

Table 2. Examples of Bacterial Antibiotics and Other Secondary Metabolites That May Be Involved in Crop
Disease Biocontrol
Metabolite Pathogen Disease

Phenazines Gaeumannomyces graminis var. tritici Take-all of wheat

Pyrrolnitrin Alternaria spp., B. cinerea, Pythium spp., Seedling diseases, gray mold, root rot and
R. solani, S. sclerotiorum, Fusarium damping-off, white mold, dry rot of
sambucinum, V. dahliae potato, wilt
Oomycin A Pythium spp. Damping-off
Pyoluteorin Pythium ultimum Damping-off
2,4-Diacetylphloroglucinol Rhizoctonia solani, P. ultimum, T. Root rot and damping-off, Black root rot of
basicola tobacco
Hydrogen cyanide Thielaviopsis basicola Black root rot of tobacco
Agrocin 84 Agrobacterium tumefaciens Crown gall of fruit trees
Pseudobactin 10 Fusarium oxysporum lini Flax wilt
Ammonia Pythium spp. Damping-off

Source: Modified from N. Dowling and F. O’Gara, Trends Biotech. 12, 133–141 (1994), J. M. Ligon et al., Pest Manage. Sci. 56, 688–695 (2000).
458 BIOCONTROL, MICROBIAL AGENTS IN SOIL

vitro and, consequently, antibiosis has been suggested enables the bacteria to use living hyphae rather than chitin
to be involved in biological control by fungal antago- as the actual growth substrate, because chitin is an impor-
nists (31,36). Several secondary metabolites produced by tant constituent of most fungal cell walls. For example,
Gliocladium (= Trichoderma) virens were shown to be a strain of S. marcescens, isolated from the rhizosphere
important in mycoparasite biocontrol activity. Thus, pro- of plants grown in soil infested with S. rolfsii Sacc., was
duction of the antifungal antibiotics gliotoxin (an epid- found to be an effective BCA under greenhouse conditions
iothidiketopiperasine) and gliovirin (a diketopiperazine) against this pathogen and R. solani Kuhn. A chitinase(s)
has been associated with efficacy of G. virens as a BCA of produced by the bacterium caused degradation of S. rolfsii
seedling diseases incited by Pythium spp. and R. solani, hyphae in vitro, which provides evidence that this enzyme
which cause seedling damping-off and root rot, respec- plays a role in biocontrol. Chitinolytic strains of A. caviae
tively, of many host plants. Gliotoxin appears to be the and E. agglomerans were able to control R. solani and
key factor in G. virens biocontrol activity against both F. oxysporum f. sp. vasinfectum in cotton and S. rolfsii in
these fungal pathogens, whereas gliovirin appears to beans under greenhouse conditions. In many cases, how-
kill the first fungus but not the last. Strains of Tricho- ever, the inhibition of fungal growth was not accompanied
derma produce more than 40 compounds with antibiotic by bacterial chitinase production, indicating that other cell
activity. Many of them, including volatile alkyl pyrones, wall–degrading enzymes (β-1,3-glucanase and protease)
peptaibols, diketopiperazines, isonitriles, polyketides, and and/or antibiotics may also be involved in the antagonistic
steroids are thought to be responsible for biocontrol activ- activities of chitinolytic bacteria against fungi (38).
ity of some Trichoderma isolates. Some antibiotics are
synergistic with cell wall–degrading enzymes in their Lytic Enzymes Produced by Fungal Antagonists. The
action against fungal phytopathogens. It is worth noting, ability to produce lytic enzymes has been shown to be
however, that some highly effective biocontrol strains of a crucial property of the mycoparasitic fungi G. virens
Trichoderma may not produce the antibiotics shown to and Trichoderma spp., and many bacterial antagonists of
be important for the biocontrol efficiency of other Tricho- plant pathogenic fungi. In general, chitinolytic enzymes
derma strains. To illustrate, gliotoxin has been shown to are not only involved in the destruction of the host cell
be important for activity of the G. virens strain, which is wall; they may also play a role during the initial stages
the active component of SoilGard (Table 1), but it is not
of mycoparasitism. However, the relevance and role of
produced by T. harzianum T-22, which is described as one
enzymes and toxic metabolite(s) of the mycoparasitic
of the most efficient BCA (13).
fungi’s antagonism toward plant pathogens can vary
among independent isolates and should be reinvestigated
Cell Wall Lysis. Most fungal plant pathogens have cell
for each individual case. Moreover, the ability of lytic
walls, which contain chitin as a structural backbone
enzymes to provide biocontrol depends on the type of
and laminarin (β-1,3-glucan) as a filling material. The
plant being protected and the fungal pathogen. Thus, the
other minor cell wall components are proteins and
lipids. Chitin, an unbranched insoluble homopolymer cell walls of basidiomycetes (e.g., R. solani and S. rolfsii)
consisting of β-1,4-linked N-acetylglucosamine (GlcNAc) and ascomycetes (e.g., Sclerotinia sclerotiorum) contain
units, is the second (after cellulose) most common mainly β-glucan and chitin, but no cellulose. In contrast,
biodegradable polysaccharide in nature, being the main oomycetes (e.g., species of Pythium and Phytophtora)
structural component of cell walls of most fungi and are composed of β-glucan (laminarin), cellulose and
arthropods (insects, nematodes, and other invertebrates) less than 1.5% chitin. Therefore, the cell wall lytic
including many agricultural pests. Many species of enzymes chitinase and β-1,3-glucanase have frequently
bacteria, streptomycetes and actinomycetes, fungi, and been associated with biocontrol activity against the
plants produce chitinolytic enzymes, which catalyze the first groups of fungal phytopathogens, whereas β-1,3-
hydrolysis of chitin (37). Chitinolytic enzymes (chitinases) glucanases and cellulases (β-1–4-glucanases and β-1–4-
cleave a bond between two consecutive GlcNAc units. glucosidases) can provide biological control of the fungi
These enzymes play an important physiological and belonging to the second group (38). Protease production
ecological role in ecosystems as recyclers of chitin, by is common in microorganisms, including fungi, among
generating carbon and nitrogen sources. In recent years, them Trichoderma strain T39 that serves as a BCA of the
soilborne microorganisms that produce chitinases have fungus B. cinerea (39) and is commercially produced as the
become considered as potential BCA against fungal biopreparation ‘‘Trichodex’’ (Table 1).
pathogens, insects, and nematodes, which cause diseases Chitinases and other lytic enzymes also play an
and damage in agricultural crops. important role in the biocontrol of plant pathogenic
nematodes. The inhibition of Globodera rostochiensis egg
Lytic Enzymes Produced by Bacterial Antagonists. The hatch by the chitinase-producing bacteria S. maltophilia
ability to produce lytic enzymes is a widely distributed was suggested as a biocontrol strategy for the defense
property of soil, marine, and rhizospheric bacteria. Many of potato crops against potato cyst nematodes (40).
of these are potential BCA of chitin-containing plant Another approach to controlling plant parasitic nematodes
pathogens. The list of chitinolytic bacterial antagonists involves amending the soil with chitin that stimulates
includes Aeromonas caviae, E. agglomerans, a Strep- chitinolytic activity. This treatment drastically increases
tomyces sp., Pseudomonas stutzeri and P. fluorescens, the populations of chitinolytic bacteria and fungi (41).
S. marcescens, and S. plymuthica, among others. Chiti- The nematode’s surface is composed of a multilayered
nase production may be part of a lytic system, which cuticle, which consists mainly of collagen. Collagenolytic
 BIOCONTROL, MICROBIAL AGENTS IN SOIL 459

activity was shown therefore to play a role in the mycoparasites, in contrast to necrotrophic ones, tend to
antagonism of some fungi and bacteria toward plant have a more restricted host range and produce specialized
pathogenic nematodes. Soil amendment by collagen structures to absorb nutritients from their host. Because
sharply decreased the root-galling index of tomatoes of their nature, only a few examples of biotrophic
inoculated with Meloidogyne javanica. This led to mycoparasites (S. sclerotivorum and A. quisqualis) as BCA
the enrichment of collagenolytic microorganisms, which exist (45).
drastically reduced the number of galls caused by
M. javanica. The effect was even more pronounced Indirect Biological Control
when the amendment was supplemented with the
Indirect approaches include: (1) use of organic soil amend-
collagenolytic fungus Cunninghamella elegans (42). The
ments which enhance the activity of indigenous microbial
proteases produced by the fungus P. lilacinus and some
antagonists against a specific pathogen and where chang-
others were shown to play a role in this fungal nematode-
ing conditions in the plant’s environment may affect
parasitic activity (43). Some of the fungal parasites of plant the parasite, its host, or the saprophytic microorgan-
parasitic nematodes act by triggering induced resistance in isms potentially involved in disease control (46); (2) plant
the plant, by inhibition of the nematodes’s host-recognition growth promotion by root-colonizing bacteria and fungi
process or by producing toxins able to immobilize the which are able to stimulate plant growth and to increase
nematode, following by penetration of hyphae through the tolerance to biotic and abiotic stresses through enhanced
cuticle (25,26). root and plant development (17); (3) induced resistance,
which involves the stimulation of plant self-defense mech-
Predation or Parasitism. This is a complex mechanism, anisms against a particular pathogen by prior inoculation
which occurs when the BCA feeds directly on or inside of the plant with a nonvirulent strain.
the pathogen. Parasitism is defined as a direct attack on Successful induced resistance has been documented for
fungal cells followed by feeding by the parasite. When both bacterial and fungal pathogens. Induced resistance
one fungus feeds on another fungus, it is generally called is a plant response to challenge by microorganisms or
mycoparasitism. Competition for substrates already taken abiotic agents, such that following the induced challenge,
up by other fungi is an ecological feature shared by de novo resistance to pathogens is shown in normally
many mycoparasites. Briefly, mycoparasitism is a complex susceptible plants. Induced resistance can be ‘‘localized’’
process, which involves ‘‘recognition’’ of the host, positive when it is detected only near the inducing factor, or
chemotropic growth, attachment, and de novo synthesis of ‘‘systemic,’’ when resistance occurs subsequently at sites
a set of cell wall–degrading enzymes that aid the parasite throughout the plant. Both localized and systemic types of
in penetrating the host and completing its destruction. induced resistance are nonspecific and can act against
Lectins, the sugar-binding proteins or glycoproteins of a wide range of pathogens. The difference between
nonimmune origin, which agglutinate cells and are these two types of induced resistance is that localized
involved in interactions between the cell’s surface resistance occurs in many plant species, whereas systemic
components and its extracellular environment, have been resistance is limited to some plants. During localized
shown to play a role in the recognition and contact resistance, the plant reacts to the environmental stimulus
between necrotrophic mycoparasites of Gliocladium and by activating a variety of defense mechanisms, which
Trichoderma spp. and soilborne pathogenic fungi. This manifest themselves in various biochemical and physical
contact, in turn, initiates a signal transduction cascade changes. The general plant defense response consists of
toward the second most important step of mycoparasitism, the induction of pathogenesis-related (PR) proteins and
the induction of lytic enzymes able to degrade fungal cell the deposition of structural polymers, such as callose and
walls (4). lignin.
Barnett and Binder (44) divided mycoparasitism into: Little is known about the distribution and/or the
(1) necrotrophic (destructive) parasitism, in which the biochemistry of the interaction. Special emphasis has
relationship results in death and destruction of one or been placed on the accumulation of hydrolases such as
more components of the host thallus and (2) biotrophic chitinases and glucanases with antimicrobial potential.
(balanced) parasitism, in which the development of the Another group of enzymes includes peroxidases, which
parasite is favored by a living rather than dead host play a key role in the plant resistance process, being
structure. Necrotrophic mycoparasites tend to be more involved in phenolic compounds synthesis and structural
aggressive, have a broad host range extending to wide barrier formation. In some systemically protected plants
taxonomic groups, and are relatively unspecialized in (e.g., tobacco and cucumber), increases in newly formed
their mode of parasitism. The antagonistic activity of PR proteins have also been recorded, and these may
necrotrophic mycoparasites is attributed to the production be chitinase-, glucanase-, or osmotin-like proteins. The
of antibiotics, toxins, or hydrolytic enzymes in proportions induction of systemic resistance responses by sometimes
that cause the death and destruction of their host. only a single local lesion indicates involvement of an
Necrotrophic mycoparasitic fungi, being more common, immunity signal(s), which are appeared at the induction
saprophytic in nature and less specialized in their mode of site and then transduced to the target parts where it
action, are easier to study. As a result, most mycoparasites conditions the plant for resistance. Several endogenous
(P. nunn, T. flavus, C. minitans, Gliocladium spp., and plant compounds, salicylic acid, and jasmonic acid among
Trichoderma spp.), used as BCA in greenhouse or field them are promising candidate signals for induced systemic
trials to date have been necrotrophs. The biotrophic resistance. For example, application of exogenous salicylic
460 BIOCONTROL, MICROBIAL AGENTS IN SOIL

acid elicits PR protein patterns similar to those induced

by pathogens, and the treated plants develop resistance to
viral, bacterial, and fungal pathogens (47,48).
All known BCAs utilize one or more of these general
indirect or direct mechanisms. At the product level, this
includes the production of antibiotics, siderophores, and
cell wall–lytic enzymes, and the production of substances
that promote plant growth. Additionally, successful
colonization of the root surface is considered a key property
of prospective antagonists (17). The best BCAs may use
two or three different mechanisms. Antagonists can also
be combined to provide multiple mechanisms of action Figure 1. Dual cultures of mycoparasitic fungus T. harzianum
against one or more pathogens. Plant genes also play a (lower part) with the plant pathogenic fungus R. solani (upper
role in biocontrol efficiency. Plant health depends, in part, part) growing simultaneously at 28 ° C in a petri plates with
on associations with disease-suppressive microflora, but potato-dextrose agar (PDA). Left and right plates are two and
little is known about the role of plant genes in establishing seven days after contact.
such associations.

Selection of Biocontrol Agents

Despite the increasing interest in studying the basic
and applied aspects of biocontrol of plant pathogens,
the commercial usage of biocontrol facilities is still very
limited. The development of reliable biocontrol systems
depends on an accurate assessment of the problems and
on the limiting factors that interfere with them. One of
the most prominent problems associated with biocontrol
appears to be the inconsistent performance of microbial
antagonists under field conditions. Several reasons can
be attributed to this phenomenon, including screening
techniques, ecological factors, mechanisms of control, and
Figure 2. Assay of antifungal activity on petri plates. Two agar
problems associated with formulation and large-scale disks from an actively growing culture of P. aphanidermatum
production (11). were placed on either side of the bacterial growth area, and the
Different screening methods for the selection of BCAs plates were incubated at 28 ° C for two days. A suspension of
have been described. Hence, using in vitro methods many bacterial strain E. agglomerans was seeded in a line through the
strains of microorganisms with antagonistic properties center of a PDA plate and incubated at 28 ° C for 24 hours (right
have been isolated. In planta, screening methods have plate). In a control plate (left plate) where sterile water was used
been developed to optimize selection of antagonists effec- instead of bacteria, the fungal mycelium covered the entire agar
tive against a range of soilborne pathogens. Screening surface.
in the field follows screens in the laboratory or green-
house. Ecologically adapted antagonists against seed and
soilborne pathogens can be isolated from the surface of T. harzianum) (Fig. 1) or antibiosis causing the inhibition
seeds and roots of the plants susceptible to the desired of the pathogen (e.g., P. aphanidermatum) by the antibi-
pathogen. A frequently used approach is to look for healthy otics or siderophores produced by the antagonists (e.g.,
plants in a field with diseased plants of the same species. some bacteria) (Fig. 2). However, the results obtained
Thus, effective isolates have been found in soils suppres- with such plate tests do not necessarily indicate that a
sive to take-all of wheat caused by fungus G. graminis strain with good antagonistic activity in vitro will serve
var. tritici (49), or to Fusarium wilt (7). However, knowl- as an efficient BCA under natural environmental con-
edge about the origin of an antagonist will always be ditions. A soil plate method revealed a good correlation
relevant in connection with risk evaluation, and wild-type between the competitive saprophytic ability of a range of
candidates isolated from natural locations can be expected mycoparasitic Trichoderma isolates and their ability to
to give less complications in the registration process (50). destroy host (target) fungi in soil. In this method, an agar
Screening methods can be focused mainly on informa- plate is first colonized by a host fungus, then particles
tion about mechanisms of antagonism obtained in vitro or of soil or plant material are placed on the surface. Only
try to mimic greenhouse or field conditions. For example, fungi that are able to derive nutrients from the precol-
two-component screening methods (e.g., dual cultures of a onizing fungus and to tolerate its waste products will
candidate antagonist and a pathogen on agar) are based be able to grow. Mycoparasites isolated by this method
on interaction studies. The potential antagonists are typi- include G. roseum, a Papulaspora sp., Pythium spp., and
cally ranked according to their ability to inhibit the growth Verticillium biguttatum (50).
of the pathogen, as expressed by an inhibition zone. Dual The introduced microbial antagonist will have to over-
cultures on agar are able to reveal the mycoparasitism come competition from the indigenous soil microflora.
of the pathogen (e.g., R. solani) by the antagonists (e.g., Therefore, mechanisms other than those that can be
 BIOCONTROL, MICROBIAL AGENTS IN SOIL 461

assessed in in vitro assays are important under field con- alginate-encapsulated cells or spores survive well in the
ditions. The ability of the antagonist to metabolize the soil. Encapsulation of microorganisms in an alginate
exudate molecules produced by a protected plant, to colo- matrix was considered to be an effective way of delivering
nize the plant rhizosphere and to protect infection sites in useful microorganisms slowly into the soil for prolonged
situ appears to be key factors in the biocontrol of soilborne periods.
plant pathogens. Soil pH may also be an important factor Various BCAs have been encapsulated in amylopectin
determining the success or failure of introduced antag- (pregelatinized starch) to improve their survival under
onists. Competition and other antagonistic interactions severe environmental conditions. Several other granular
(e.g., antibiosis and parasitism), between the introduced formulations, including ‘Pesta’ granules in which fungal
BCA and the pathogen may also take place in the rhizo- propagules are encapsulated in a wheat gluten matrix,
sphere. Other mechanisms, such as induced resistance and peat or alginate prills, vermiculite, and application onto
plant growth promotion, can also be taken into account by crop residue were evaluated (52). Some formulations
screening programs on the basis of approaches related to provide rapid release of bacteria early in the growing
the plants. season whereas other formulations provide slow-release of
bacteria, particularly for weeds that emerge later in the
FORMULATION AND DELIVERY OF BIOCONTROL growing season. Industrially available products include
AGENTS formulated preparations of biocontrol fungi, bacteria,
and streptomycetes (Table 1). For example, G. virens,
Maintaining adequate viability of microorganisms formulated in alginate prill form (preparation GlioGardTM )
throughout processing, storage, and application for bio- and incorporated into soilless potting media was shown
control is critically important for commercial reasons. effective for disease control of bedding vegetable plants and
Unlike chemical pesticides that begin to degrade after ornamental seedlings against P. ultimum and R. solani.
application, BCA must survive and begin to proliferate. Another example is alginate prill formulations of T. flavus
To compete with chemical pesticides, appropriate tech- with organic carriers for biocontrol of V. dahliae. In
nology must be developed to improve the BCA stability, this case, pyrophyllite clay (Pyrax) and milled corncobs
bioactivity, and shelf life. Therefore, formulating biocon- used to make alginate prill with ascospores of T. flavus
trol organisms effectively is the key to their successful consistently enhanced the biocontrol activity of the fungus.
use. The formulation must be compatible with the require-
ments of the microorganisms, and even enhance their COMBINED USAGE OF BIOCONTROL AGENTS
performance, and it should be compatible with regular
agricultural practice, to be both acceptable and useful to Most approaches for biocontrol of plant diseases use a
the farmer. To be distributed through the normal com- single BCA as antagonist to a single pathogen. Indeed,
mercial chain, the preparation should ideally have a shelf microorganisms as BCAs typically have a relatively nar-
life of one to two years. The right formulation can help row spectrum of activity compared with synthetic pes-
protect the microorganisms biologically and physically. To ticides and often exhibit inconsistent performance in
achieve this protection, the formulation should be able practical agriculture resulting in the limited commercial
to induce low metabolic activity in the microorganism use of biocontrol approaches for plant pathogens suppres-
involved. This is usually brought about by reducing mois- sion (49). Usually, it is very difficult to find a single strain
ture potential by drying or by incorporating the agent into that is able to control a wide spectrum of pathogens under
an oily microaerophilic environment. Surfactants, stick- various rhizospheric and soil conditions. However, this
ers, emulsifiers, and spreaders are added to formulations goal can be achieved by genetic modification of the BCA by
to improve their application to the target. adding disease suppression mechanisms (e.g., genes encod-
Commercial formulations of microbial agents have been ing additional antibiotics or cell wall lytic enzymes), which
used in the biocontrol of plant diseases and plant parasitic operate against more than one pathogen, or by developing
nematodes. Products aimed at soil pathogens can be strain mixtures with superior biocontrol activity. Sev-
formulated for bulk application in the plant growth media eral strategies for forming mixtures of BCAs have been
or for seed treatment, with or without a food base. The proposed, including mixtures of organisms with differen-
type of delivery system used to apply the BCA to its tial plant-colonization patterns; mixtures of antagonists
specific habitats important for its success. For example, a that control different pathogens; mixtures of antagonists
mixture of peat and wheat bran (1 : 1) was shown efficient with different mechanisms of disease-suppression; mix-
medium for T. harzianum growth and sporulation. This tures of taxonomically different organisms; mixtures of
formulation has been applied to seedlings before planting, antagonists with different optimum temperature, pH, or
to seeds in furrows, and in broadcast applications (51). The moisture conditions for plant colonization. Mixtures of
most extensively and successfully used bioencapsulation antagonists are considered to account for the protec-
material is alginate. A method of forming alginate beads tion afforded by disease-suppressive soils. In the absence
for encapsulation of biocontrol strains of Trichoderma of cross-inhibition of antagonists, combining antagonists
and Gliocladium was developed. The fungal conidia, results in improved biocontrol. Therefore, successful devel-
chlamydospores, or fermentation biomass are mixed in opment of strain mixtures requires compatibility of the
a 1% sodium alginate solution and added dropwise into coinoculated microorganisms (53).
a 2 to 5% calcium chloride solution. The formed pellets Combinations of BCAs for plant diseases include mix-
or granules are dried for use as a BCA. The resultant tures of fungi, mixtures of fungi and bacteria, and mixtures
462 BIOCONTROL, MICROBIAL AGENTS IN SOIL

of bacteria. Thus, combined usage of bacteria and fungi for pseudomonads was observed in the rhizosphere of plants
plant disease control have been shown in some exper- that were grown in solarized soils and was found to be due
iments to be more efficient than a single antagonist. to the improved capacity of these bacteria to compete for
For example, E. agglomerans together with some Tricho- the plant exudates (55).
derma spp. strains were shown to be the most promising
antagonists from suppressive soils against Phythophtora MOLECULAR APPROACHES FOR IMPROVEMENT OF
capsici in pepper. Enterobacter cloacae and Pseudomonas EXISTING, AND CREATION OF NEW BIOCONTROL
spp. strain–induced suppression of Rhizoctonia damping- AGENTS
off were consistently more effective in combinations with
T. hamatum relative to the fungal isolate alone. The inte- Improvement of Bacterial Biocontrol Agents
grated approach based on using binuclear Rhizoctonia spp.
and E. cloacae strains in combination, led to protection Some BCAs have been genetically modified to enhance
of cucumber against the disease in greenhouse. Several their biocontrol capabilities or other desirable characteris-
bacterial and fungal BCAs, including Trichoderma, Glio- tics. Strain improvement is one of the most exciting ways
cladium, Pseudomonas, Bacillus, Agrobacterium, Enter- of achieving highly efficient biocontrol of plant diseases.
obacter, and Serratia, may protect plants against diseases Molecular approaches and genetic engineering techniques
via synergistic action. Therefore, it has often been stated have recently been applied to gain a better and more basic
that mixtures of BCAs are required for successful long- understanding of the mechanism of action of BCAs, and
term control because the individual components that to develop superior and improved strains of BCAs with
colonize different crops are adapted to different environ- enhanced activity (18,32).
ments or have different functions (49). However, single For example, synthesis of agrocin 84 by the BCA
isolates of Trichoderma and other microbial antagonists A. radiobacter K84 is directed by a gene located in the
efficient for biocontrol of a number of plant pathogens also extrachromosomal DNA known as plasmid pAgK84. This
have been described (11,13). plasmid also carries the genes needed for resistance to
agrocin 84 and is able to transfer from one bacterium
to another by a mechanism known as conjugation.
INTEGRATED PEST MANAGEMENT Consequently, pAg84 may be transferred to pathogenic
A. tumefaciens, which would then be resistant to agrocin
Biological control is just one component of a strategy 84. To prevent this resistance, a transfer-deficient mutant
termed integrated pest management (IPM). IPM is of strain K84 was constructed. Agrobacterium radiobacter
defined as the reduction of pest populations by natural strain K1026 is identical to the parental strain, except
enemies, and it typically involves an active human that the agrocin-encoding plasmid cannot be transferred
role. IPM strategy involves the coordinated use of from cell to cell (16).
pest and environmental information with available pest In systems in which antibiotics play a primary role
management methods to prevent unacceptable levels of in the strains biocontrol activity, molecular approaches
pest damage by the most economical means, and with the can be used to enhance the bacteria’s biocontrol efficacy
least possible hazard to humans and the environment. by increasing levels of antibiotic synthesis, either by
In short, the goal of the IPM approach is to manage increasing the copy number of the biosynthetic genes
pests and the environment so as to balance costs, benefits, or by modifying the regulatory signals that control
human health, and environmental quality. Thus, IPM their expression. This is particularly interesting for
is biologically based, cost-effective, and its site-specific bacteria in which the biosynthetic genes are arranged
risk-reduction system makes use of multiple methods in in either operons or clusters, as was shown for the
the decision-making process (54). Combinations of BCAs genetics of Phz, Phl, and Prn production in Pseudomonas
with low, otherwise inefficient doses of chemical pesticides spp. strains (56). For example, increased production
are often considered an important IPM approach to of Plt and Phl and superior control of P. ultimum
achieving the most capable protection of crop disease. damping-off of cucumber was achieved by increasing the
For example, under field conditions, a combination of number of antibiotic biosynthesis genes in P. fluorescens
lowered doses of methyl bromide (MB) and T. harzianum strain CHAO. Constitutive synthesis of oomycin A in
exhibited a significant synergistic effect on damping-off P. fluorescens strain HV37a was achieved by insertion
of carrot seedlings caused by R. solani, and its effect on of a strong tac promoter from Escherichia coli upstream
growth, yield, and disease control was similar to that of of the afuE locus encoding the antibiotic. The resultant
the much higher recommended MB dosage. Integrated recombinant strain produced a significantly higher level
control of V. dahliae in potato by T. harzianum and of oomycin A than the parental strain and provided
the fungicide captan increased both marketable and greater control of P. ultimum infection (57). Another
total potato yield (36). Integration of induced systemic approach is to introduce biosynthetic genes into a
resistance into pest management systems can further lead strain deficient in antibiotic production, or into one
to the decreased usage of the conventional pesticides (48). that produces a different antibiotic, to increase the
Combinating BCAs with soil solarization (solar heating) spectrum of activity. Thus, cloned Phz biosynthetic
technique is an additional efficient way of using natural genes were transferred from P. fluorescens 2–79, which
conditions to protect crops against pathogens and to exhibits poor rhizosphere competence, into P. putida and
enhance plant growth, even in the absence of known P. fluorescens strains, which exhibit superior rhizosphere
major pathogens. The rapid establishment of fluorescent competence. The recombinant strains that synthesized
 BIOCONTROL, MICROBIAL AGENTS IN SOIL 463

Phz in vitro are potentially superior BCAs of pathogenic of T. harzianum has been shown to be triggered in
fungus G. graminis var. tritici because of their ability mycoparasitic interactions. Therefore, duplication of this
to colonize the wheat rhizosphere. When genes conferring highly conserved gene appears to be a potential means
Phl synthesis were mobilized from a P. aureofaciens strain of improving the biocontrol capability of Trichoderma
into a P. fluorescens strain that normally produces only species.
Phz and inhibits this fungus, the obtained recombinant Proteinases were also overexpressed in Trichoderma
strain was shown to be additionally suppressive against with the aim of improving the isolate’s biocontrol
P. ultimum and R. solani through expression of an activity. Under greenhouse conditions, T. harzianum
additional antibiotic (49). transformants with an increased copy number of the
The chitinase gene chiA encoding one of the basic proteinase gene prb1 that were incorporated into
chitinases from biocontrol strains of S. marcescens pathogen-infested soil reduced the disease caused by
and E. agglomerans was isolated and cloned into R. solani in cotton plants up to fivefold as compared with
E. coli (58,59). Escherichia coli strains transformed with the wild-type strain. In all these studies, the Trichoderma
the chiA gene expressed and excreted the corresponding strains were genetically improved to serve as better BCAs
protein into the growth medium. Transformed E. coli was of fungi, such as R. solani, that contain chitin and β-glucan
effective in inhibiting S. rolfsii on beans and R. solani on as major cell wall components (4). A similar approach
cotton, but to a lesser degree than the original chitinolytic was approved for obtaining Trichoderma strains with
strains. The genetically engineered E. coli, a nonsoil bac- high activity against plant pathogenic oomycetes, such
terium, served here as a model system to demonstrate the as Pythium spp., containing cellulose as the main cell
role of chitinase in controlling a chitin-containing plant wall component. The difference in this case, however,
pathogen. It is suggested that the introduction of such was that the Trichoderma strains were designed to
engineered genes into soil bacteria will increase control overproduce cellulytic enzymes instead of chitinolytic
efficiency by combining high expression of a gene encoding ones. Hypercellulolytic strains of T. longibrachiatum
a lytic enzyme with rhizosphere competence. obtained by genetic engineering methods were shown
The current knowledge of regulatory mechanisms to be significantly more effective in controlling Pythium
governing the expression of various antifungal substances damping-off disease on cucumber than the wild-type
may help in the construction of strains with enhanced strain (60).
biocontrol activity. Manipulation of regulatory systems The major advantage of such genetic manipulations
responsible for the production of lytic enzyme and is the ability to isolate genes from one strain and to
antibiotics has resulted in significant improvement of introduce them into other varieties of fungi or bacteria.
the bacteria’s biocontrol potential. The advantages of this This enhances the potency of BCAs and makes a single
approach were demonstrated by increasing the doses of strain effective, stable, and consistent against more than
genes encoding the GacA-GacS system of global regulation one plant pathogenic fungus, without the hazardous effects
or sigma factors of transcription in some biocontrol of chemical pesticides.
strains of P. fluorescens (35,56). The findings supported The current strategies to improve BCAs activity
the idea that regulatory signals from the environment through genetic engineering technology are summarized
may contribute to the success of a bacterium in competition in Table 3.
with plant pathogens and other rhizosphere inhabitants
under natural conditions (33).
CONCLUSION
Improvement of Biocontrol Potency of Trichoderma spp.
Because of their beneficial activity, BCAs can serve as suit-
Introduction of foreign genes that could potentially able alternatives to some commercial fungicides. However,
enhance the biocontrol capacity of Trichoderma (4). For more studies are required to develop screening techniques,
example, genes encoding cell wall lytic enzymes were used appropriate formulations, and delivery systems to enhance
as a tool to increase the chitinase activity of one of the most application and survival, and to elucidate the molecular
potent biocontrol species, T. harzianum. Two different basis of their mode of action against plant pathogens.
strategies were used. According to one approach, the This, in turn, will yield genetic information to produce
chitinase gene of S. marcescens was introduced into the superior biocontrol strains, and give a better understand-
genome of T. harzianum under the control of a constitutive ing of ecological factors that may interfere with biocontrol
eukaryotic promoter, providing a high level of expression activity. Because these developments are time-consuming
of the foreign bacterial gene into a new fungal host. and often, do not produce immediate results, more empha-
Following another strategy, a T. harzianum strain was sis should be directed to extending the use of integrated
transformed by a vector that contained the coding region control as part of a comprehensive IPM program, thus
of the 42-kDa endochitinase gene of this fungus under introducing biocontrol into mainstream agriculture. The
the control of a strong cellulase promoter from a relative biocompatibility of these agents with fungicides is expected
fungus Trichoderma reesei (=T. longibrachiarum). A 10- to enhance their efficacy, and there are a few examples
fold increase in chitinase activity was achieved. Using the available to demonstrate this.
same the endochitinase gene of T. harzianum with its own Most microorganisms selected as potential BCAs never
regulatory sequences, transformants of T. hamatum with become products because of problems related to their
an approximately fivefold increase in chitinase activity production and downstream processing. The antagonistic
were obtained. The 42-kDa endochitinase encoding gene activities in vitro or in vivo considered important for the
464 BIOCONTROL, MICROBIAL AGENTS IN SOIL

Table 3. Examples of Genetic Engineering Approaches to Improve Biocontrol Agents by Enhancement of

Antimicrobial Compound Production

Mechanism of Biocontrol
Strategy Approach Enhancement

Altered expression Insertion of biosynthesis genes upstream of a Constitutive and/or increased production of
stronger promoter; introduction of new/or antibiotics, siderophores, or lytic enzymes
additional regulatory genes, construction
of regulatory mutants
Gene dosage Increase copy number of gene(s) in wild-type Overproduction of antibiotics and other
strains antimicrobial compounds
Heterologous expression Introduction of new genes encoding Conversion of non-producing strains to
antimicrobial compounds producing strains, expansion of assortment
of antimicrobial compounds produced by a
single biocontrol agent
Deletion Specifically delete transfer (tra) genes from Eliminates potential transfer of agrocin 84
plasmid encoding agrocin 84 resistance gene to target pathogen

Source: Modified from N. Dowling and F. O’Gara, Trends Biotech. 12, 133–141 (1994).

biocontrol potential of these strains are, in fact, often for soilborne diseases, with a few reports of reduction
not sufficient to develop an efficient biocontrol product. of foliar fungal pathogens. It is worth noting, however,
To become a product, the microorganisms have to fit the that even the most efficient BCAs can be overcome
following requirements. They must be easily produced in by heavy disease pressure. BCAs cannot cure already
fermentors, that is, a high amount of cells produced per existing infections; hence they must be used strictly as a
unit volume in a short time with low-cost culture medium; preventive measure (13). There are several areas, in which
they must be easily converted to an appropriate form for additional research is required, and these include further
longer shelf life; they must maintain biocontrol properties improvement of the agents to achieve long-term effects
after large-scale production and drying. Because of their and to enhance their survival and efficiency for routine
high production costs, biotechnological products require a application in agricultural practice for the benefit of the
large market to be beneficial. Therefore, it is worth being farmers.
able to use the same product against a large number of
diseases.
BIBLIOGRAPHY
The greatest value of biocontrol agents versus chemical
pesticides is that chemicals usually give a high level of 1. J. J. Menn and F. R. Hall, in F. R. Hall and J. J. Menn, eds.,
protection but strongly rhizospheric competent organisms, Biopesticides: Use and Delivery, Humana Press, Totowa, N.J.,
in particular, can provide a somewhat lower level of benefit 1999, pp. 1–10.
for extended periods, that is, up to several months. The 2. R. J. Cook, Annu. Rev. Phytopathol. 31, 53–80 (1993).
point is that biologicals and chemicals both have strong 3. J. M. Whipps, Adv. Bot. Res. 26, 1–134 (1997).
points and neither can do what the other does best (13).
4. A. Herrera Estrella and I. Chet, in A. Altman, ed. Agri-
The introduction or application of BCAs for pest and cultural Biotechnology, Marcel Dekker, New York, 1998,
disease control will result in yields that are pesticide-free pp. 263–282.
and hence biosafe. Pesticide-free products have already 5. M. Wilson and P. A. Backman, in F. R. Ruberson, ed., Hand
been introduced into the marketplace, particularly in Book of Pest Management, Marcel Dekker, New York, 1998,
developed countries. The market share of environmentally pp. 309–335.
friendly fresh produce is constantly increasing, and 6. L. G. Copping and J. J. Menn, Pest Manage. Sci. 56, 651–676
pesticide-free fruits and vegetables are achieving being (2000).
in higher prices than conventional produce. At present, 7. C. Alabouvette and P. Lemanceau, in F. R. Hall and
biological control may lack, in some cases, the rapid J. J. Menn, eds., Biopesticides. Use and Delivery, Humana
protective effect exhibited by the best pesticides, and Press, Totowa, N.J., 1999, pp. 117–135.
it may also result in somewhat lower efficiency and 8. J. K. F. Baker and W. C. Snyder, eds., Ecology of Soilborne
narrower host specificity than chemical control. Currently Plant Pathogens. Prelude to Biological Control, University of
however, several BCAs with a rather wide range of California Press, Berkeley, CA, 1965, pp. 571
biocontrol activity are available. The most extensively 9. J. Rishbeth, Ann. Appl. Biol. 52, 63–77 (1963).
studied bacterial organisms, including Pseudomonas spp., 10. H. D. Wells, D. K. Bell, and C. A. Jaworski, Phytopathology
B. subtilis, and Enterobacter spp., have been reported to 62, 442–447 (1972).
reduce several soilborne and aerial diseases on various 11. A. Sivan and I. Chet, in R. Mitchell, ed., New Concepts in
crop species. Fungal BCAs, including the extensively Environmental Microbiology, Wiley-Liss, New York, 1992,
studied Gliocladium and Trichoderma species, in addition pp. 335–354.
to other weakly virulent or saprophytic fungi, have been 12. A. Tronsmo and L. G. Hjeljord, in G. J. Boland and
reported to reduce infection or reproduction of many L. D. Kuykendall, eds., Plant-Microbe Interactions and Bio-
pathogens. Most of the applications of these fungi are logical Control, Marcel Dekker, New York, 1998, pp. 111–126.
 BIOCORROSION: ROLE OF SULFATE REDUCING BACTERIA 465

13. G. E. Harman, Plant Dis. 84, 377–393 (2000). 42. S. Galper, E. Cohn, Y. Spiegel, and I. Chet, J. Nematol. 23,
14. P. B. Adams, Annu. Rev. Phytopathol. 28, 59–72 (1990). 269–274 (1991).
15. R. Hoffstein and A. Chapple, in F. R. Hall and J. J. Menn, 43. P. J. M. Bonants et al., Microbiology 141, 775–784 (1995).
eds., Biopesticides: Use and Delivery, Humana Press, Totowa, 44. L. Barnett and F. L. Binder, Annu. Rev. Phytopathol. 11,
N.J., 1999, pp. 77–102. 273–292 (1973).
16. M. H. Ryder and D. A. Jones, in D. Hornby, ed., Biological 45. J. Inbar and I. Chet, Crit. Rev. Biotechnol. 17, 1–20 (1997).
Control of Plant Pathogens, CAB International, Oxon, U.K., 46. A. J. Hoitink, M. J. Boehm, and Y. Hadar, in H. A. J. Hoitink
1990, pp. 45–63. and H. M. Keener, eds., Science and Engineering of Compost-
17. W. Kloepper and C. J. Beauchamp, Can. J. Microbiol. 38, ing: Design, Environmental, Microbiological and Utilization
1219–1232 (1992). Aspects, Renaissance Publications, Worthington, OH, 1993,
18. R. Glick and Y. Bashan, Biotechnol. Adv. 15, 353–378 (1997). pp. 601–621.
19. Y. Hadar, G. E. Harman, A. G. Taylor, and J. M. Norton, 47. L. C. van Loon, P. A. H. M. Bakker, and C. M. J. Pieterse,
Phytopathology 73, 1322–1325 (1983). Ann. Rev. Phytopathol. 36, 453–483 (1998).
20. K. P. Hebbar, A. G. Davey, J. Merrin, and P. J. Dart, Soil 48. S. Tuzun and E. Bent, in A. A. Agrawal, S. Tuzin, and
Biol. Biochem. 24, 989–997 (1992). E. Bent, eds., Induced Plant Defenses Against Pathogens and
Herbivores, APS Press, St. Paul, Minn., 1999, pp. 95–115.
21. A. Ordentlich, Y. Elad, and I. Chet, Phytopathology 78, 84–88
(1988). 49. E. Mathre, R. J. Cook, and N. W. Callan, Plant Dis. 83,
972–983 (1999).
22. L. S. Chernin, Z. F. Ismailov, S. Haran, and I. Chet, Appl.
Environ. Microbiol. 61, 1720–1726 (1995). 50. M. B. Knudsen et al., Eur. J. Plant Pathol. 103, 775–784
(1997).
23. M. Shoda, J. Biosci. Bioeng. 89, 515–521 (2000).
51. W. J. Connick Jr., J. A. Lewis, and P. C. Quimby Jr., in
24. Z. A. Siddiqui and I. Mahmood, Bioresour. Technol. 69,
R. R. Baker and P. E. Dunn, eds., New Directions in Biolog-
167–179 (1999).
ical Control. Alternatives for Suppressing Agricultural Pests
25. Y. Oka et al., Pest Manage. Sci. 56, 983–988 (2000). and Diseases, Alan R. Liss, New York, 1990, pp. 345–372.
26. B. R. Kelly, Annu. Rev. Phytopathol. 38, 423–441 (2000). 52. R. D. Lumsden, J. A. Lewis, and D. R. Fravel, in F. R. Hall
27. H. M. T. Hokkanen and J. M. Lynch, eds., Biological Control and J. W. Barry, eds., Biorational Pest Control Agents.
Benefits and Risks, Cambridge University Press, Cambridge, Formulation and Delivery, American Chemical Society,
U.K., 1995, p. 304. Washington, D.C., 1995, pp. 166–182.
28. K. van Dijk and E. B. Nelson, Appl. Environ. Microbiol. 66, 53. G. S. Raupach and J. W. Kloepper, Phytopathology 88,
5340–5347 (2000). 1158–1164 (1998).
29. J. S. Ahmad and R. Baker, Phytopathology 77, 182–189 54. M. J. Way and H. F. van Emden, Crop Protection 19, 81–103
(1987). (2000).
30. L. S. Thomashow and D. M. Weller, in G. Stacey and 55. J. Katan, in R. Hall, ed., Strategies for Management Soilborne
N. T. Keen, eds., Plant-Microbe Interactions, vol. 1, Chapman Plant Pathogens, APS Press, St. Paul, Minn., 1996,
& Hall, New York, 1995, pp. 187–235. pp. 250–278.
31. R. Howell, in G. E. Harman and C. K. Kubichek, eds., Tricho- 56. D. Haas, C. Blumer, and C. Keel, Curr. Opinion Biotechnol.
derma and Gliocladium, vol. 2, Taylor and Francis, London, 11, 290–297 (2000).
U.K., 1998, pp. 173–184. 57. N. Gutterson, Crit. Rev. Biotechnol. 10, 69–91 (1990).
32. N. Dowling and F. O’Gara, Trends Biotech. 12, 133–141 58. R. Shapira, A. Ordentlich, I. Chet, and A. B. Oppenheim,
(1994). Phytopathology 79, 1246–1249 (1989).
33. L. S. Pierson, in G. J. Boland and L. D. Kuykendall, eds., 59. L. S. Chernin et al., Appl. Environ. Microbiol. 63, 834–839
Plant-Microbe Interactions and Biological Control, Marcel (1997).
Dekker, New York, 1998, pp. 355–392. 60. Q. Migheli et al., Phytopathology 88, 673–677 (1998).
34. I. Bender, V. Rangaswamy, and J. Loper, Annu. Rev. Phy-
topathol. 37, 175–196 (1999).
35. J. M. Ligon et al., Pest Manage. Science 56, 688–695 (2000).
36. K. Sivasithamparam and E. L. Ghisalberti, in C. P. Kubicek
BIOCORROSION: ROLE OF SULFATE
and G. E. Harman, eds., Trichoderma and Gliocladium,
REDUCING BACTERIA
vol. 1, 1998, pp. 139–191.
37. M. S. Manocha and V. Govindsamy, in G. J. Boland and
IWONA B. BEECH
L. D. Kuykendall, eds., Plant-Microbe Interactions and Bio- University of Portsmouth
logical Control/Books in Soils, Plants, and the Environment, Portsmouth, United Kingdom
Marcel Dekker, New York, 1998, pp. 309–327.
38. L. Chernin and I. Chet, in R. P. Dick and R. G. Burns, eds., In natural environments and artificial systems, corrosion
Enzymes in the Environment, Marcel Dekker, New York, reactions under a biological influence usually involve
2001, pp. 171–225. surface-associated microbial growth, namely, a biofilm.
39. Y. Elad and A. Kapat, Eur. J. Plant Pathol. 105, 177–189 Terms such as microbially influenced corrosion (MIC) or
(1999). biocorrosion are frequently employed to describe metal
40. D. Cronin, Y. Moenne-Loccoz, C. Dunne, and F. O’Gara, Eur. deterioration in the presence of microorganisms (1).
J. Plant-Pathol. 103, 433–440 (1997). Corrosion under a biological influence frequently occurs
41. S. U. Sarathchandra et al., Biol. Fertil. Soils 22, 221–226 as a localized attack, which results from the activity
(1996). of physiologically diverse microbial species associated
466 BIOCORROSION: ROLE OF SULFATE REDUCING BACTERIA

with the metal substratum. Microorganisms implicated • An electrolyte (a solution capable of conducting
in biocorrosion have frequently been grouped by their electrical flow).
metabolic demand for different respiratory substrates. • A cathodic reactant
Sulfate-reducing bacteria, sulfur-oxidizing bacteria, iron- • An anodic reactant
oxidizing/reducing bacteria, manganese-oxidizing bacte-
• Electronic contact between the anode and cathode
ria, as well as bacteria secreting organic acids and
exopolymers or slime have all been named as culprits Schematic diagram of a basic corrosion cell for a ferrous
in MIC failures. These organisms coexist within biofilms, metal under anoxic conditions is represented in Figure 1.
often forming synergistic communities able to affect elec- The process of corrosion consists of two half-reactions:
trochemical processes through cooperative metabolism, an anodic reaction involving the ionization (oxidation) of
which individual species have difficulty initiating. the metal and a cathodic reaction involving the reduction
The first documented case describing biocorrosion dates of a chemical species in contact with the metal surface. At
to 1891. This report proposed that the increase of corrosion the anodic site the metal dissolution occurs and metallic
of lead cable could be due to bacterial metabolic products. ions are formed. A cathodic site is required as an electron
However, it was not until 1934 that the corrosion of ferrous sink for the return of current to the metal. This reaction is a
metal buried in anaerobic clay soil was solely attributed to reduction reaction or an electron-consuming reaction. The
the activity of obligate anaerobic microorganisms known rate of corrosion is determined by the rate of the cathodic
as sulfate-reducing bacteria (SRB) (2). Since then SRB reaction, and as the electric current must migrate through
became the most investigated group of microorganisms the solution to the metal at the cathode; the cathodic
implicated in corrosion damage of iron and ferrous alloys current must be equal to the anodic current.
in a broad range of aquatic and terrestrial habitats varying In aerated and oxygen-free solutions the anodic reaction
in nutrient content, temperature, pressure, and pH values. is metal dissolution:
Biocorrosion of metals and their alloys is of great
importance to many industries such as oil and gas,
Me → Me2+ + 2e− (1)
shipping, power, and water distribution sectors. In
particular, the activity of sulfate-reducers can have
For aerated solutions at neutral or alkaline pH, the main
considerable impact on oil and gas industry, leading
cathodic reaction is the reduction of oxygen
to severe economical losses due to system failures and
equipment deterioration and resulting in serious health
and safety hazards and environmental risks (3,4). It has O2 + 2H2 O + 4e− → 4OH− (2)
been proposed that approximately 20% of all corrosion
damage to materials fabricated with metal is microbially In acidic or deaerated solutions, protons or, when
influenced or enhanced (5). MIC is therefore the subject of present, hydrogen sulfide may act as electron acceptor,
intense research efforts aiming to develop environmentally as demonstrated in reactions (3) and (4):
acceptable prevention and protection measures and to
improve the effectiveness of already existing control 2H+ + 2e− → 2H → H2 (3)
strategies. −
2H2 S + 2e → 2HS + H2 −
(4)
Several theories have been presented to explain
mechanisms by which sulfate-reducers can influence
In an intermediate pH solution, that is, one of low pH but
corrosion of metallic materials. In this article all proposed
aerated, protons are more readily available than oxygen
models of SRB-mediated corrosion are described along
and, therefore, proton reduction will be the main electron-
with the current hypothesis based on the identified
consuming route.
reactions of primary importance leading to the coupled
Although the anodic and cathodic reactions can occur at
electrochemical and biotic electron transfer processes (6).
different locations on the metal, typically they are adjacent
to each other. The sites of the anodic and cathodic reactions
FUNDAMENTAL ASPECTS OF CORROSION

The term corrosion is defined as the physicochemical Electrolyte

interaction between a metal and its environment, which
High oxygen (cathode) High oxygen (cathode)
results in changes in the properties of metal (7). These
O2 O2
changes may lead to impairment of the function of the low oxygen
metal, the environment, or the technical system of which (anode)
Fe2+
these form a part. Corrosion is an electrochemical process
defined as a chemical reaction involving the transfer of
electrons through a series of oxidation and reduction
reactions. e− Fe e−
For electrochemical corrosion to occur, the following
components have to be present:
Metal

• An anode site Figure 1. Basic corrosion cell for iron-based alloy under aerated
• A cathode site conditions.
 BIOCORROSION: ROLE OF SULFATE REDUCING BACTERIA 467

are referred to as the anode and cathode, respectively. The In the presence of biofilms, the following factors are likely
irregular distribution of corrosion cells across the metal to affect metal deterioration (1):
surface causes localized attack in the form of pitting and/or
crevice corrosion. When the half-cells are closely spaced, a • Microbial activity and microbial metabolic products.
general corrosion is observed. • The passage of charged entities through the exopoly-
As already stated, corrosion process causes materials mer matrix.
fabricated of pure metals and/or their mixtures (alloys) to • The degree of conductivity of the EPS matrix.
undergo chemical oxidation from ground state to an ionized • The binding of metal ions by the EPS matrix.
species. The oxidation reaction, in most instances, slows
• Destabilization of corrosion inhibitors.
to a low rate after a period of time because the oxidation
products (corrosion products) adhere to the metal surface The main importance to the corrosion process is the
and form a layer, which serves as a diffusion barrier to control the biofilm exerts on the transport of ions, namely,
other reactants. These products form a protective barrier regulating the exchange of ions between the metal surface
to further oxidation of the underlying metal. Changes in and the aqueous environment by either promoting or
the environmental conditions can affect the stability of the retarding metal ion transfer. Although in most natural
protective layers and, hence, the overall susceptibility of and artificial environments, microorganisms and products
the metal to corrosion. of their metabolism, including EPS, are present in
greater quantities at the metal surface compared to the
bulk phase, water still constitutes 95% or more of the
BIOCORROSION biofilm matrix. The biofilm can, therefore, still act as
a modified electrolyte. The high growth rates, varied
Due to their presence on the surface and their ability metabolic products, many of which are corrosive, and the
to carry out specific biochemical reactions, microorgan- high surface-to-volume ratio of microorganisms can result
isms can alter the physical/chemical conditions at the in interactions, which would modify physicochemical
metal/substratum interface and modify electrochemical parameters at metal solution interfaces to such a degree
reactions, which are fundamental to all corrosion pro- that corrosion rates can be accelerated by factors of 103 to
cesses. Microbial colonization of the surface is facilitated 105 (15). In contrast, certain types of biofilms may produce
by the production of extracellular polymeric substances a barrier effect resulting in a reduction of the chemical
(EPS) that are composed of a mixture of macromolecules activity and hence a substantial decrease in the corrosion
such as proteins, polysaccharides, nucleic acids, and lipids. rate of the metal (1).
EPS facilitate bacterial attachment to the substratum and
form the biofilm matrix ((8) and references therein). The
SULFATE-REDUCING BACTERIA
presence of exopolymers on the metal surface can modify
both morphology and chemistry of corrosion products alter-
Sulfate-reducing bacteria are readily isolated from
ing their corrosion characteristic, often rendering these
biofilms formed on different type of surfaces in natural
products more aggressive.
habitats and industrial systems. These bacteria thrive
The selective binding of different metal ions by EPS
within anoxic niches, which exist within biofilms as a
leading to the formation of metal concentration cells
result of oxygen depletion due to the respiration of aerobic
within the biofilm matrix has been proposed as one of
or facultatively anaerobic microorganisms.
the mechanisms of biocorrosion. The presence of such
cells could promote electron transfer processes between Phylogeny
the biofilm and base metal (9,10). A number of studies
have demonstrated the involvement of exopolymers The family of taxonomically diverse SRB is currently dis-
secreted by different genera of aerobic bacteria in metal tributed within two domains: Archaea and Bacteria (16).
deterioration (11,12). In the former domain, sulfate-reducers are represented by
Microbial metabolic processes, including the consump- microorganisms of the Archaeglobus genus. In the King-
tion of oxygen and production of acids and sulfides, dom Bacteria, SRB are divided into three distinct groups,
promote the establishment of localized chemical gradients the gram-positive spore-forming SRB with Desulfotumac-
at the metal surface. The existence of such gradients leads ulum being a predominant genus, the gram-negative
to the formation of electrochemical cells, which stimulates mesophilic SRB (δ-subclass of the Proteobacteria) of which
anodic and/or cathodic reactions, the ultimate consequence Desulfovibrio is one of the better characterized genera and
being loss of metal from the discrete locations on the sur- the thermophilic SRB represented by two genera Ther-
face. A number of biocorrosion mechanisms have been modesulfobacterium and Thermodesulfovibrio. Although
identified that reflect the variety of physiological activities SRB are currently divided into four phylogenetic groups,
carried out by different types of microorganisms (13,14). new divisions could be added, as several new genera of psy-
The interaction of a biofilm harboring many species chrophilic SRB were recently isolated from permanently
of aerobic, facultatively anaerobic, and strictly anaerobic cold Arctic marine sediments (17).
microorganisms and the underlying substratum produces
Physiology
a unique physical and chemical environment. Thus, con-
ditions at the substratum surface can differ significantly Lactate, pyruvate, and malate are frequently used as
from those in the bulk phase or at a biofilm-free surface. carbon source and electron donors and are oxidized
468 BIOCORROSION: ROLE OF SULFATE REDUCING BACTERIA

to acetate and carbon dioxide (18,19). Some SRB can APS is rapidly converted to sulfite and AMP by the
utilize high molecular weight fatty acids or simple cytoplasmic enzyme APS reductase.
aromatic compounds such as benzene or phenol and
are known to participate directly or indirectly in the APS → AMP + SO3 2−
degradation of variety of substances, including saturated
hydrocarbons (20). During their growth, SRB produce A number of different sulfite reductases have been
a large amount of hydrogen sulfide that assure the identified in SRB (24). The most commonly recognized
maintenance of an anoxic atmosphere. types, particularly among the genus Desulfovibrio, are the
Although all SRB species known to date have been biosulfite reductases, desulfoviridin, and desulforubidin.
described as strictly anaerobic, some genera are able Sulfite is reduced via a number of intermediates, including
to survive for long periods in the presence of oxygen, metabisulfite (S2 O5 2− ), dithionite (S2 O4 2− ), trithionate
indicating the existence of enzymatic defense mecha- (S3 O2− 2−
6 ), and thiosulfite (S2 O3 ), to form the sulfide ion.
nisms against oxygen radicals (21,22). Indeed, the pres- The latter can be further converted to hydrogen sulfide
ence of enzymes catalase and superoxide dismutase, (H2 S) in the presence of external hydrogen (H+ ) ions.
which prevent accumulation of H2 O2 and eliminate Apart from sulfate, some Desulfovibrio species can grow
toxic respectively, constitutively expressed during anaer- with nitrate or fumarate as alternative electron acceptors
obic growth of Desulfovibrio gigas, was recently con- releasing ammonia and succinate, respectively (20). At low
firmed (23). dissolved oxygen concentrations certain SRB are able to
respire with Fe(III) or even oxygen with hydrogen acting
as electron donor (25,26). Although some gram-negative
Respiration SRB can reduce Fe(III) there are no known SRB that are
capable of performing this reaction using ferric ion as a
Sulfate-reducing bacteria are chemolithotrophic organ-
sole electron acceptor.
isms that are able to carry out dissimilatory reduction of
sulfur compounds such as sulfate, sulfite, thiosulfate, and
Hydrogenase Enzymes
even sulfur itself to sulfide by a series of energy-conserving
reactions (Fig. 2; 18). Hydrogen metabolism plays a central role in energy-
The initial step in the dissimilatory sulfate reduction generating mechanisms of sulfate-reducing bacteria. The
is the transport of exogenous sulfate across the bacterial capability of SRB to produce and consume H2 means that
membrane into the cells. This process may be inhibited they possess the enzyme hydrogenase, which catalyzes
by selenate, which is the structural analog of sulfate. the reversible oxidation of hydrogen according to the
Once inside the cell, sulfate dissimilation proceeds by reaction (5):
the action of adenosine triphosphate (ATP)-sulfurylase H2 ↔ 2H+ + 2e− (5)
which combines sulfate with ATP to produce the highly
activated molecule adenosine phosphosulfate (APS), as Hydrogenase was first discovered by Stephenson and
well as pyrophosphate (PP) that may be subsequently Stickland in 1931 (27). Since then hydrogenase activity
cleaved to yield inorganic phosphate. was found in a large number of anaerobic and aerobic
prokaryotes, as well as in eukaryotic cells.
Two mechanisms involving H2 -metabolism have been
ATP + SO4 2− → APS + PP offered for energy coupling in Desulfovibrio during

SO42− enters cell

outside cell SO42− + ATP APS + PP 2P
2e
SeO42−
inhibits SO32− + AMP

H+
S2O52−

2e

S2O42−

2H+ S3O62−
H2S S2− S2O32−
2e 2e
Figure 2. A proposed cyclic pathway for dissimilatory outside
sulfate reduction (18). cell
 BIOCORROSION: ROLE OF SULFATE REDUCING BACTERIA 469

growth on organic substances in the presence of sulfate: conditions (Figs. 3 and 4). The number of SRB detected
(1) generation of a proton gradient by the periplasmic in a system, as either suspended or attached populations,
hydrogenase itself linked to obligate H2 -cycling (28) does not necessarily correlate with the extent or rate
and (2) proton translocation typically associated with of corrosion and different types of SRB dissimilar in
a trace H2 -transformation model (29,30). Both models their physiology vary in the ability to influence metal
propose that bioenergetic mechanisms are dependent deterioration (38–40). However, no clear consensus has
on the existence of at least two functionally distinct yet been reached linking specific bacterial metabolic rates
hydrogenases: one located in the cytoplasm involved in to observed corrosion rates.
H2 -production and another in the periplasm involved in Several key theories such as:
H2 -consumption.
Three types of hydrogenases have been isolated from • The depolarizing effect of hydrogenase enzyme.
the sulfate-reducing bacteria of the genus Desulfovibrio. • Cathodic depolarization due to biogenically produced
According to the type of metal present, hydrogenases iron sulfide.
are generally divided into: the iron hydrogenase ([Fe]
• Anodic depolarization due to the formation of iron
hydrogenase), the nickel-(iron-sulfur)-containing hydroge-
sulfide corrosion products.
nase ([NiFe] hydrogenase) and the nickel-(iron-sulfur)-
selenium containing hydrogenase ([NiFeSe] hydrogenase). along with reports on the generation of corrosive
They differ in their subunit and metal compositions, phosphorous compounds and binding of metal ions by
physicochemical characteristics, amino acid sequences, SRB exopolymers have been proposed to explain the
immunological reactivities, gene structures, and their mechanisms by which these bacteria can influence the
catalytic properties (31). The [NiFe] hydrogenases have corrosion reaction (41–43). It is now accepted that one
been found in all species of Desulfovibrio investigated to predominant mechanism may not exist and that a number
date. This class of hydrogenases can be localized in the of processes are involved (44–46).
periplasm, the cytoplasm, and/or the membrane depend-
ing on the SRB species and are particularly resistant
Hydrogenase Enzyme as Depolarizing Agent — Early Studies
to inhibitors such as CO and NO− 2 (32). [NiFeSe] hydroge-
nases, which contain nickel and selenium in equimolecular In 1934, von Wolzogen Kühr and van der Vlugt proposed
amounts, are only found in some species of Desulfovibrio. what is now referred to as the classical mechanism of
The [Fe] hydrogenase has the highest specific activity anaerobic corrosion of ferrous metals in the presence of
in the evolution and consumption of hydrogen and this SRB, also known as the cathodic depolarization theory (2).
enzyme is the most sensitive to CO and NO− 2 . It is not The essential step in this theory involves the removal of
present in all species of Desulfovibrio (33). hydrogen (cathodic depolarization) by the activity of the
The three classes of hydrogenases are not uniformly hydrogenase enzyme. The electron removal as a result of
distributed in all Desulfovibrio and an individual species hydrogen utilization forces more iron to be dissolved at the
may contain one (e.g., [NiFe] hydrogenase in D. gigas), two anode. The main steps involved are listed in the following
(e.g., [NiFe] and [NiFeSe] hydrogenases in D. baculatus) text (reactions 6 to 12) and depicted in Figure 5.
or three hydrogenases (e.g., [Fe], [NiFe], [NiFeSe]
hydrogenases in D. vulgaris Hildenborough). Most of the anodic reaction:
hydrogenases isolated from the genus Desulfovibrio are
confined to the periplasmic space; however, membrane- 4Fe → 4Fe2+ + 8e− (6)
bound and cytoplasm-located enzymes have also been dissociation of water:
described (34).
All three types of hydrogenases catalyze the same 8H2 O → 8H+ + 8OH− (7)
basic reaction, the reversible activation of the hydrogen cathodic reaction:
molecule by a heterolytic mechanism. The proton-
deuterium exchange reaction, the para-ortho hydrogen 8e− + 8H+ → 8H (8)
conversion, the hydrogen production from reduced methyl
cathodic depolarization:
viologen and the hydrogen utilization with various
artificial and natural electron acceptors can determine SO4 2− + 8H → S2− + 4H2 O (9)
the activity of hydrogenases, which varies with pH and
corrosion products:
temperature (35,36).
Apart from their participation in energy production Fe2+ + S2− → FeS (10)
(i.e., the involvement in hydrogen cycling and the
generation of a proton gradient), the hydrogenases of SRB corrosion products:
are proposed to play an important role in biocorrosion (37). 3Fe2+ + 6OH− → 3Fe(OH)2 (11)
overall reaction:
MODELS OF SRB-INFLUENCED CORROSION
4Fe + SO4 2− + 4H2 O → 3Fe(OH)2 + FeS + 2OH−(12)
The role of SRB has been documented in pitting corrosion
of iron-based alloys in both aquatic and terrestrial Using cathodic polarization measurements, a range
environments under anoxic, as well as oxygenated of studies demonstrated that the SRB species with
470 BIOCORROSION: ROLE OF SULFATE REDUCING BACTERIA

2.000 µM
µH
4
3
2
1

(a)
1000.00 nH

Figure 3. Atomic force microscopy image of a

µH
three-day-old biofilm formed by the marine 2.5
sulfate-reducing bacterium Desulfovibrio indonensensis 2.0
on AISI 316 stainless steel revealing (a) a group of cells 1.5
forming a discrete colony on the steel surface and
1.0
the abundance of irregularly distributed iron sulfide
0.5
particles, and (b) detailed topographic features of a single
SRB cell and its close association with FeS particles. (b)

3Fe2+
3Fe(OH)2 6OH−

FeS
S2− Sulphate
reducing
Fe2+
bacteria
SO42−→ S2− + 4O
8H + 4O → 4H2O
8H+
8H2O → 8OH− + 8H+ → 8H + 6OH− + 2OH−

4Fe → 4Fe2+

Metal

Figure 4. Scanning electron micrograph demonstrating localized +8e− +8e−

Anode Cathode
corrosion on the surface of AISI 316 stainless steel following the
removal of a 14-day-old biofilm formed in a pure culture of the
Figure 5. Schematic representation of the cathodic depolariza-
marine sulfate-reducing bacterium Desulfovibrio indonensensis.
tion process of ferrous metal in the presence of sulfate-reducing
Corrosion is visible as pitting attack of varying intensity.
bacteria.
 BIOCORROSION: ROLE OF SULFATE REDUCING BACTERIA 471

detectable hydrogenase activity were able to depolarize frequently found where SRB have been detected. On con-
the mild steel cathode. Depolarization did not occur in tinued exposure to SRB, mackinawite alters to greigite
the cultures of SRB where hydrogenase activity was and smythite and finally to pyrrhotite. Greigite is associ-
not recorded (47–49). Further support of the model was ated with generalized corrosion of iron, while smythite is
provided by demonstrating the dissolution of iron from indicative of pitting attack. The presence of mackinawite
a mild steel electrode electrically connected to a similar and greigite among corrosion products of ferrous materials
electrode in contact with a culture of hydrogenase-positive is generally considered as proof that SRB participated in
Desulfovibrio spp. on an agar surface containing benzyl the corrosion reaction (56). Although pyrite (FeS2 ) is not
viologen (BV), acting as electron acceptor. The SRB a common iron corrosion product, SRB can produce pyrite
oxidized cathodic hydrogen and transferred the electrons from mackinawite in contact with elemental sulfur.
to the redox dye, BV instead of sulfate (50,51). Mineral signatures of both crystalline and amorphous
However, the latter experiment met with criticisms iron sulfides indicative of SRB-mediated corrosion are
since it was likely that hydrogenase-positive SRB commonly detected using X-ray diffraction (XRD) and
catalyzed not only the consumption of hydrogen but energy dispersive X-ray analysis (EDX). Little is known
also the liberation of hydrogen by reduced BV. The about subsequent crystallization of amorphous sulfides,
apparent hydrogenase-dependent cathodic depolarization although biomineralization around SRB colonies, or within
was claimed to be an experimental artifact (15). A number biofilm matrix, may be a key process.
of follow-up studies produced inconclusive or contradicting The quantitative importance of cathode depolarization
results leading to the revision of the proposed SRB by solid FeS was demonstrated by the addition of
corrosion model based solely on the hydrogenase activity. chemically prepared FeS to mild steel test coupons in the
presence and absence of Desulfovibrio spp. and replacing
sulfate with fumarate as terminal electron acceptor for
Iron Sulfide as Depolarizing Compound
bacterial reduction processes. The extent of corrosion
Low corrosion rates were measured in semicontinuous as assayed both by polarization and by weight loss
or continuous hydrogenase-positive SRB cultures grown measurements, was proportional to the FeS added and
with low ferrous iron concentration in lactate medium dependent on direct contact between the sulfide and the
containing sulfate due to the formation of a protective metal surface (57).
iron sulfide film. When this film ruptured, much higher Comparison of the corrosion of mild steel by chemically
corrosion rates were recorded. Moreover, in the media with prepared FeS and biogenically derived FeS indicated
a high ferrous ion concentration, high corrosion rates (on that, in media of high Fe2+ concentration, most of the
the order of 1 mm y−1 ) were observed for both hydrogenase- corrosion in pure SRB cultures was attributed to the
positive and -negative SRB strains. These results biogenically derived FeS (58–60). The corrosion caused
prompted investigation into the involvement of biologically by the biogenic FeS appeared identical to that caused by
generated sulfides in the SRB-mediated biocorrosion FeS from inorganic reactions (61).
phenomenon (52). Detailed studies demonstrated that the Thus, a new model emerged, which proposed that
initial sulfide film formed during corrosion was in the corrosion of iron-based alloys in the presence of SRB
form of a continuous, adherent, and protective layer of proceeded under reduced conditions through cathodic
mackinawite (Fe1−x S). However, this film showed a degree stimulation of electrochemicals cells established between
of physical disruption in a time-dependent modification areas of unreacted metal (anode) and deposits of various
to greigite (Fe3 S4 ). In the presence of higher soluble biogenically generated reduced ferrous sulfide corrosion
iron concentrations there was a similar loss of protective products (cathode).
mackinawite film with the conversion to smythite (Fe9 S11 )
and pyrrothite (FeS1+x ). In each case, this loss of the Iron Sulfide and Hydrogenase as Depolarizing Agents
uniform protective mackinawite film generated active It has been noted that the presence of SRB is required
electrochemical cells between areas of unreacted and to maintain the chemical integrity and electrochemical
unprotected steel and deposits of the various ferrous activity of iron sulfide species. In bacteria-free cultures
sulfides, with the resulting corrosion occurring by cathodic the depolarizing capacity of FeS diminished with time,
stimulation. possibly as a result of the bonding of atomic hydrogen
Many sulfides under near-surface natural environmen- within the FeS crystals lattice. However, the corroding
tal conditions may only be produced by enzymatically ability of the FeS was restored in active SRB cultures.
catalyzed microbiological pathways on specific substrates A theory was constructed proposing that the cathodic
such as metals. For example, formation and stability of reaction, namely, H2 evolution resulting from reaction of
the sulfide mineral, mackinawite, is dependent on a con- Fe2+ with bacterially produced S2− occurred on the FeS
tinuously acting hydrogen sulfide source. Although abiotic deposits (41). It further postulated that SRB present as
sources of sulfide may also lead to mackinawite forma- biofilms on the surface of FeS continually regenerated or
tion in the absence of enzymatically governed biological depolarized the FeS by the removal of atomic hydrogen as
reactions (53), the most common source of sulfide respon- a result of hydrogenase activity. This mechanism would
sible for the vast majority of industrially related corrosion explain the continued high rates of mild steel corrosion in
is that produced by SRB (54,55). Although mackinawite biologically active, SRB-harboring systems.
formation is not favored under standard thermodynamic Studies on physiology of SRB provided evidence that in
conditions common to many corrosion scenarios, it is oxygen-free environments bacteria were able to utilize
472 BIOCORROSION: ROLE OF SULFATE REDUCING BACTERIA

cathodic H2 with sulfate acting as the sole source of foil was exposed to oxygen on an intermittent basis
energy (62–64). However, this process, which required the was 50 times higher than that reported under anaerobic
activity of hydrogenase, was influenced by the availability conditions (74).
of organic electron donors. It was concluded that in the It is believed that under alternating aerated and anoxic
presence of SRB, anoxic aqueous environments rich in conditions, the corrosion mechanism of mild steel is related
anaerobically degradable organic matter (e.g., interior of to the large Fe-S pool, which provides both surface area for
sewer pipes) should be more corrosive than environments colonization and hydrogen to SRB, as well as potentially
that are mainly inorganic (e.g., in geothermal heating corrosive elemental sulfur and polysulfide (46). The latter
plants). These findings led to reevaluation of the role can be produced as a result of sulfide reacting with
played by hydrogenase in the corrosion process. oxygen. When anoxic sites with SRB become aerobic, new
minerals are formed from the original corrosion products.
Enzyme Hydrogenase as Depolarizing Agent — New
The oxidation process yields lepidocrocite or goethite
Approach
and rhombic or orthorhombic sulfur. Mixed reducing
Although the role of iron sulfide in the SRB-mediated and oxidizing conditions, such as those commonly found
corrosion process could not be dismissed, it has been in certain pipeline systems, often produce the partially
argued that the oxidation of cathodically produced H2 oxidized iron oxide magnetite along with lepidocrocite and
was the determining step. Experiments were conducted goethite (56).
to demonstrate that the separation of bacteria from the It has been suggested that at the aerobic/anoxic inter-
surface by a dialysis membrane prevented the utilization face, the corrosion reaction sequence due to accumulation
of cathodic H2 by the SRB (65). Furthermore, the current of FeS might proceed in two stages. Firstly, the iron sul-
observed in the presence of biofilms formed by the fide on the steel surface may form a protective film of
hydrogenase-positive SRB did not appear when biofilms mackinawite, thereby exerting anodic control of metal
were developed by hydrogenase-negative SRB cultures. It dissolution by limiting ferrous ion diffusion through the
has also been demonstrated that the hydrogenase activity sulfide film. Subsequent breakdown of this protective film
did not depend on the presence of viable SRB cells and that would create an electrochemical cell with sulfide-free steel
the cell-free extracts of hydrogenase enzyme increased the acting as the anode, iron sulfide acting as the cathode, and
corrosion of mild steel (66,67). Thus, it was determined sulfur serving as the electron acceptor to promote cathodic
that regardless of the presence of living cells, hydrogenase depolarization.
might still retain its depolarizing capabilities, which could
be expressed under appropriate reducing conditions.
Studies on mixed populations of SRB in biofilms Anodic Depolarization
isolated from corroded and undamaged oil pipelines Careful analysis of cathodic depolarization theory and
revealed that the biofilm with detectable hydrogenase in particular the role of sulfides led to an alternative
activity was associated with a significantly higher hypothesis. The latter proposed that SRB promoted
corrosion rate (7.70 mm year−1 ) than the noncorrosive corrosion by a mechanism of anodic depolarization,
biofilm (0.48 mm year−1 ). Both biofilms had comparable essentially dependent on sulfide production. This model
number of SRB cells per unit area; however, no measurable postulates that an anode is first created by local H+
hydrogenase activity could be found in the latter type of production at a focus of SRB metabolic activity, with
biofilm. It was therefore concluded that the susceptibility ensuing metal dissolution (75,76). A key element in
of a system to biocorrosion depended on the microbial generating the necessary kinetic conditions is a localized
makeup of the mixed population of SRB and the activity acidification at the anode resulting from the formation of
of the hydrogenase enzyme (68). iron sulfide corrosion products according to reaction (13):
Subsequent investigations revealed that the hydro-
genase activity of SRB is regulated by dissolved Fe2+
availability, which is subject to repression/derepression Fe2+ + HS− → FeS ↓ +H+ (13)
mechanism and depends on SRB species (69,70). It is
therefore likely that apart from other parameters affect- The reaction also has the secondary effect of removing
ing enzyme performance, bacterial uptake of cathodically HS− , hence reducing the effectiveness of the H2 S/HS−
generated H2 governed by the activity of hydrogenase, buffer system, which would resist local acidification. The
and hence the rate of cathodic depolarization, fluctuates availability of sulfide and free iron would determine the
depending on the ecology and physiology of SRB species nature of the corrosion products formed. If the local sulfide
within biofilms and the level of bioavailable Fe2+ ions. concentration was low, as would be the case with the
To date, hydrogenase contribution to the SRB-enhanced reaction aforementioned, in the presence of high soluble
corrosion process has not been assessed in a quantitative iron concentration, the product formed would most likely
manner. be mackinawite, which, as already described, is considered
to be nonprotective. Where sulfide was in excess, the
The Influence of Oxygen
product would be the more protective pyrite. Although
Numerous observations confirmed that the presence of pitting corrosion of steel can theoretically be explained
dissolved oxygen played an important part in the SRB- by the mechanism of anodic depolarization, the processes
influenced corrosion of ferrous metals (71–73). A corrosion under which locally acidic conditions are maintained needs
rate recorded when the vessel containing SRB and steel to be confirmed.
 BIOCORROSION: ROLE OF SULFATE REDUCING BACTERIA 473

Corrosive Phosphorus Compounds microbial consortia thriving as biofilms on surfaces

of metallic substrata can influence corrosion processes
It has been suggested that not only iron sulfides, but also
mainly due to the generation of sulfide ions. The
a highly active phosphorus compound produced by SRB,
production of the latter varies with SRB species and
could cause anaerobic corrosion of ferrous alloys (77). The
is linked to the type and activity of hydrogenase
corrosion reaction yielded iron phosphide as a corrosion
product. Cathodic depolarization experiments on an agar enzyme(s) present in all SRB. The enzymatically catalyzed
surface in the absence of an electron acceptor resulted reactions, which are influenced by bacterial ecology
in the blackening of the agar at areas in contact with and physiological and environmental factors, facilitate
SRB. On extraction and analysis by X-ray diffraction, the oxidation of molecular hydrogen evolved at the cathode
black precipitate was found to contain iron phosphide of the electrochemical corrosion cell, leading to cathodic
(Fe2 P). It was concluded that the phosphorus compound depolarization. Metal that dissolves at the electrochemical
was produced as a result of both the activities of anode can combine with the biogenically generated
SRB and the effect of H2 S on phosphate, phosphite, sulfide to form corrosion products such as metal sulfides.
and hypophosphite. Although subsequent studies have The latter will precipitate within the anoxic zones of
excluded inorganic phosphate as the source of phosphorus the biofilm, causing further stimulation of either the
metabolized by the SRB into the corrosive phosphorous cathodic reaction by serving as an electron sink and/or
product (78), it has been found that corrosive activity by influencing the anodic reaction, namely, further metal
of hydrogenase can be augmented by the availability dissolution, due to local acidification at the anode.
of inorganic phosphate (79). Organic phosphate in the Depending on their physicochemical characteristics, the
form of phytic acid, a common metabolite of plants, was accumulating iron sulfides can be either protective or
proposed as the form of phosphate metabolized by the corrosive. A number of factors, for example, the traces of
SRB into the corrosive phosphorus product in the field. oxygen, levels of unbound iron ions, the concentration of
To date, however, the role of phosphorous compounds in soluble sulfide and the presence of other microorganisms
SRB-influenced corrosion remains ambiguous. such as sulfur-oxidizing bacteria, methanogenic bacteria,
or iron-oxidizing/reducing bacteria determine both the
The Role of Extracellular Polymeric Substances metabolic activity of SRB and the type of metal sulfide
species formed.
The ability of SRB to produce EPS in batch cultures In the presence of SRB-harboring biofilms the highest
were first documented by Ochynski and Postgate (1963) corrosion rates were recorded when there are fluctuat-
who also reported the carbohydrate composition of ing aerobic-anaerobic (O2 /AnO2 ) conditions. Under such
exopolymer using paper chromatography (80). Subsequent conditions, sulfide ions and ferrous ions are prone to
investigations demonstrated that both freely suspended biotic/abiotic oxidation with oxygen acting as ultimate
and surface-associated SRB are able to synthesize electron acceptor, which would lead to the development
exopolymers varying in their chemical composition of corrosion products such as ferric oxides/hydroxides and
depending on SRB species and growth conditions (81–83). elemental sulfur. A descriptive model of SRB-mediated
Electron and atomic force microscopy allowed visualization corrosion is depicted in Figure 6 (46). The model demon-
of EPS in SRB biofilms (84). Binding of metal ions, in strates that corrosion in the presence of SRB occurs by a
particular Fe, Cr, Mo, and Ni by SRB exopolymers has
also been documented (85). Additional study confirmed
the direct involvement of a high molecular weight,
thermostable protein-carbohydrate complex produced by a Aerobic
marine SRB in pitting corrosion of mild steel (86). Further biofilm
investigation demonstrated SRB species specificity in FeO (OH)
binding of Fe2+ ions originating from the corroding surface
O2 SO2−
4
of mild steel by exopolymers released into the planktonic
phase of bacterial culture media (87). It is therefore SO2− 0
Fe3+
4 ,S
conceivable that in addition to already listed processes,
complex organic macromolecules excreted by SRB could HS− Fe2+
FeS FeS2
contribute to corrosion at the biofilm/metal interface by
mechanisms outlined in previous sections of this entry.
SO2−
4
SRB
H2 Anaerobic
CONCLUSION Fe2+
biofilm

A hypothesis coupling electrochemical and physiological e− Fe

electron transfer processes has been presented to rational-
ize present understanding of the SRB-influenced corrosion
Metal
mechanisms and to create a framework for future stud-
ies (6). The key features of this hypothesis are summarized Figure 6. A diagram of the corrosion of ferrous alloys in an
below. aerobic/anaerobic biofilm system due to the metabolic activity of
Phylogenetically and physiologically diverse anaerobic sulfate-reducing bacteria. In this model oxygen acts as terminal
sulfate-reducing bacteria associated with mixed species electron acceptor (46).
474 BIOCORROSION: ROLE OF SULFATE REDUCING BACTERIA

process of electron transfer from the base metal to oxygen 17. C. Knoblauch, K. Sahm, and B. B. Jorgensen, Int. J. Syst.
as ultimate electron acceptor, through a series of cou- Bacteriol. 49, 1631–1643 (1999).
pled redox reactions of electrochemical, biotic, and abiotic 18. J. R. Postgate, The Sulphate Reducing Bacteria, 2nd ed.,
character. Cambridge University Press, London, U.K., 1984.
It is unequivocally accepted that SRB-influenced dete- 19. L. L. Barton, ed., Sulfate-Reducing Bacteria, Plenum Press,
rioration of iron and ferrous alloys does not involve any New York, 1984.
new mechanisms of corrosion. The observed metal deteri- 20. F. Widdel, in A. J. B. Zehnder, eds., Biology of Anaerobic
oration in SRB-active environments is due to microbially Microorganisms, John Wiley & Sons, New York, 1988,
influenced changes in surface conditions of the metal- pp. 469–586.
lic substratum that promote the establishment and/or 21. J. A. Hardy and W. A. Hamilton, Curr. Microbiol. 6, 259–262
maintenance of cathodic and/or anodic reactions not nor- (1981).
mally favored under otherwise similar conditions in the 22. H. Cypionka, F. Widdel, and N. Pfenning, FEMS Microbiol.
absence of these bacteria. Notwithstanding considerable Ecol. 31, 39–45 (1985).
research efforts undertaken to elucidate the role that 23. W. G. Santos et al., J. Bacteriol. 182, 796–804 (2000).
sulfate-reducers play in corrosion, there is still a number 24. I. Moura et al., J. Am. Chem. Soc. 110, 1075–1082 (1988).
of ambiguities related to the contribution of biologically
25. E. E. Roden and D. R. Lovley, Appl. Environ. Microbiol. 59,
catalyzed reactions, including the quantitative contribu-
734–742 (1993).
tion of enzymic activities and significance of exopolymers,
26. W. Dilling and H. Cypionka, FEMS Microbiol. Lett. 71,
to the overall corrosion reaction. Undoubtedly, the better
123–128 (1990).
understanding of relationships, both direct and indirect,
between the electrochemical corrosion reactions and spa- 27. M. Stephenson and M. Stickland, Biochem. J. 25, 215–219
(1931).
tial and temporal processes associated with metabolic
activities of SRB within biofilms would aid in resolving 28. R. M. Fitz and H. Cypionka, Arch. Microbiol. 152, 369–376
such uncertainties. (1991).
29. V. Niviere et al., Arch. Microbiol. 155, 579–586 (1991).
BIBLIOGRAPHY 30. F. S. Lupton, R. Conrad, and J. G. Zeikus, J. Bacteriol. 159,
843–849 (1984).
1. W. G. Characklis and K. C. Marshall, Biofilms, John Wiley & 31. L.-F. Wu and M. A. Mandrand, FEMS Microbiol. Rev. 104,
Sons, New York, 1990. 243–27 (1993).
2. C. A. V. Von Wolzogen Kuhr and L. S. Van der Vlugt, Water 32. A. E. Przybyla, J. M. Robbins, N. Menon, and H. D. Peck Jr.,
(den Haag) 18, 147–165 (1934). FEMS Microbiol. Rev. 88, 109–316 (1992).
3. W. A. Hamilton, Annu. Rev. Microbiol. 39, 195–217 (1985). 33. M. W. W. Adams, Biochim. Biophys. Acta 1020, 115–145
4. J. M. Odom and R. Singleton, The Sulphate-Reducing Bacte- (1990).
ria: Contemporary Perspectives, Springer-Verlag, New York, 34. G. R. Bell, J. LeGall, and H. D. Peck Jr., J. Bacteriol. 120,
1992. 994–997 (1974).
5. H.-C. Flemming, in E. Heitz, W. Sand, and H.-C. Flemming, 35. G. Fauque et al., FEMS Microbiol. Rev. 54, 299–344 (1988).
eds., Microbial Deterioration of Materials, Springer, Heidel- 36. B. R. Glick, W. G. Martin, and S. M. Martin, Can. J. Micro-
berg, Germany, 1996, pp. 5–14. biol. 26, 1214–1223 (1980).
6. W. A. Hamilton, in L. Evans, eds., Biofilms: Recent Advances
37. I. Pankhania, Biofouling 1, 27–47 (1988).
in their Study and Control, Harwood Academic Publishers,
Singapore, 2000, pp. 419–434. 38. B. Little and P. Wagner, Materials Perform. 36, 40–44 (1997).
7. J. C. Scully, Fundamentals of Corrosion, 3rd ed., Pergamon 39. C. C. Gaylarde, Int. Biodet. Bodegrad. 30, 331–338 (1992).
Press, Oxford, U.K., 1990. 40. I. B. Beech et al., Int. Biodet. Biodegrad. 34, 289–303 (1994).
8. J. Wingender, T. R. Neu, and H.-C Flemming, in J. Wingen- 41. R. A. King and J. D. A. Miller, Nature 233, 491–492 (1971).
der, T. R. Neu, and H.-C. Flemming, eds., Microbial Extracel- 42. W. P. Iverson, Adv. Appl. Microbiol. 32, 1–13 (1987).
lular Polymeric Substances: Characterization, Structure, and
43. J. M. Odom and R. Singleton, The Sulphate-Reducing Bacte-
Function, Springer, Berlin, Germany, 1999, pp. 1–15.
ria: Contemporary Perspectives, Springer-Verlag, New York,
9. G. G. Geesey et al., Wat. Sci. Technol. 20, 161–165 (1988). 1992.
10. G. G. Geesey, and Lang, J. in T. J. Beveridge and R. J. Doyle, 44. W.-C Lee, Z. Lewandowski, P. H. Nielsen, and W. A. Hamil-
eds., Bacterial Interactions with Metallic Ions, John Wiley & ton, Biofouling 8, 165–194 (1995).
Sons, New York, 1989, pp. 223–247.
45. W. A. Hamilton, Geomicrobiol. J. 15, 19–28 (1998).
11. G. G. Geesey, M. W. Mittelman, T. Iwaoka, and P. R. Griff-
iths, Materials Perform. 25, 37–40 (1986). 46. P. H. Nielsen et al., Biofouling 7, 267–284 (1993).
12. T. E. Ford, J. S. Maki, and R. Mitchell, in NACE Corrosion 47. G. H. Booth and A. K. Tiller, Trans. Faraday Soc. 56,
’87, Paper No. 380, NACE, Houston, Tex., 1987. 1689–1696 (1960).
13. N. J. E. Dowling, M. W. Mittelman, and D. C. White, in 48. G. H. Booth and A. K. Tiller, Trans. Faraday Soc. 58,
J. G. Zeikus, ed., Mixed Cultures in Biotechnology, McGraw- 2510–2516 (1962).
Hill, New York, 1991, pp. 341–372. 49. G. H. Booth and F. Wormwell, Proceedings of the First
14. I. B. Beech, Biofuture 186, 36–41 (1999). International Congress on Metallic Corrosion, Institute of
15. J. A. Costello, S. Afr. J. Sci. 70, 202–204 (1974). Materials, London, U.K., 1963, pp. 341–344.
16. H. F. Castro, N. H. Williams, and A. Ogram, FEMS Micro- 50. W. P. Iverson, Science 151, 986–988 (1966).
biol. Ecol. 31, 1–9 (2000). 51. G. H. Booth and A. K. Tiller, Corros. Sci. 8, 583–600 (1968).
 BIODEGRADABILITY: METHODS FOR ASSESSING BIODEGRADABILITY UNDER LABORATORY AND FIELD CONDITION 475

52. G. H. Booth, P. M. Shinn, and D. S. Wakerley, Proceedings 84. I. B. Beech, C. W. S. Cheung, D. B. Johnson, and J. R. Smith,
of the First International Congress on Metallic Corrosion, Biofouling 10, 65–77 (1996).
Cannes, France, 1965, pp. 542–554. 85. I. B. Beech and C. W. S. Cheung, Int. Biodet. Biodegrad. 35,
53. B. D. Craig, Corrosion 47(5), 329 (1991). 59–72 (1995).
54. M. B. McNeil and B. J. Little, Corrosion 46,(7), 599–600 86. I. B. Beech, V. Zinkevich, R. Tapper, and R. Gubner, Geomi-
(1990). crobiol. J. 15, 119–132 (1998).
55. M. B. McNeil, J. M. Jones, and B. J. Little, NACE Corrosion 87. I. B. Beech, V. Zinkevich, R. Tapper, and R. J. Avci, J. Micro-
91, Paper No. 580, NACE, Houston, Tex., 1991. biol. Meth. 36, 3–10 (1999)..
56. T. R. Jack, M. J. Wilmott, and R. L. Sutherby, Mat. Perform.
34, 19–22 (1995).
57. G. H. Booth, L. Elford, and D. S. Wakerley, Br. Corros. J. 3, BIODEGRADABILITY: METHODS FOR
242–245 (1968). ASSESSING BIODEGRADABILITY UNDER
58. R. A. King, C. K. Dittmer, and J. D. A. Miller, Br. Corros. J. LABORATORY AND FIELD CONDITION
11, 105–107 (1976).
59. R. A. King, J. D. A. Miller, and J. S. Smith, Br. Corros. J. 8, EUGENE L. MADSEN
137–141 (1973). Cornell University
60. R. A. King and D. S. Wakerley, Br. Corros. J. 8, 41–45 (1973). Ithaca, New York
61. J. S. Smith and J. D. A. Miller, Br. Corros. J. 10, 136–143
(1975). Herein are described basic principles pertinent to
62. R. Cord-Ruwisch and F. Widdel, Appl. Microbiol. Biotechnol. the design, implementation, and interpretation of both
25, 169–174 (1986). laboratory and field determinations of biodegradability.
63. J. A. Hardy, Br. Corros. J. 18, 1901–1903 (1983). The information is applicable to both organic and inorganic
64. I. P. Pankhania, A. N. Mossavi, and W. A. Hamilton, J. Gen. contamination problems whose solutions can be addressed
Microbiol. 132, 3357–3365 (1986). microbiologically. However, the focus here is primarily on
65. S. Daumas, Y. Massiani, and J. Crousier, Corros. Sci. 28, assessing the biodegradability of organic compounds.
1041–1050 (1988).
66. C. P. Chatelus et al., Appl. Environ. Microbiol. 53, 1,708– DEFINITIONS OF BIODEGRADABILITY AND
1,710 (1987). BIODEGRADATION
67. J. Boivin, E. J. Laishley, R. D. Bryant, and J. W. Costerton,
NACE Corrosion 90, Paper No. 128, NACE, Houston, Tex., ‘‘Biodegradability’’ embodies qualities representing the
1990. susceptibility of substances to alteration by microbial
68. R. D. Bryant et al., Appl. Environ. Microbiol. 57, 2804–2809 processes (1). The substances may be organic or inorganic.
(1991). The alteration may be brought about by (1) intra-
69. R. D. Bryant, F. V. O. Kloeke, and E. J. Laishley, Appl. Envi- or extracellular enzymatic attack that is essential for
ron. Microbiol. 59, 491–495 (1993). the growth of the microorganism(s) (e.g., the attacked
70. C. W. S. Cheung and I. B. Beech, in A. K. Tiller and C. A. C. substances are used as a source of carbon, energy,
Sequeira, eds., Microbial Corrosion. Proceedings of the 3rd nitrogen, or other nutrients or as a final electron
International EFC Workshop, Portugal, 1994, Institute of acceptor), (2) enzymatic attack that is beneficial because it
Materials, European Federation of Corrosion Publication serves some protective purpose (e.g., mobilization of toxic
Nr. 15 London, U.K., 1995, pp. 169–180. mercury away from the vicinity of the cells), (3) enzymatic
71. W.-C Lee et al., Biofouling 7, 197–216 (1993). attack that is of no detectable benefit to the microorganism
72. W.-C Lee et al., Biofouling 7, 217–239 (1993). (e.g., cometabolic reactions in which a physiologically
73. J. McKenzie and W. A. Hamilton, Int. Biodet. Biodegrad. 29, useful primary substrate induces the production of
285–297 (1992). enzymes that fortuitously alter the molecular structure
74. J. A. Hardy and J. L. Brown, Corrosion 40, 650–654 (1984).
of another compound), and (4) nonenzymatic reactions
stemming from by-products of microbial physiology that
75. J.-L. Crolet, Oceanol. Acta 15, 87–94 (1992).
cause geochemical change (e.g., consumption of oxygen,
76. J.-L. Crolet, J. Materials Sci. 28, 2589–2606 (1993).
production of fermentation by-products, or an alteration
77. W. P. Iverson and G. J. Olson, Microbial Corrosion, Metals in pH).
Society, London, U.K., 1983, pp. 46–53. ‘‘Biodegradation’’ of organic compounds is the partial
78. W. P. Iverson, Mat. Perform. 37, 46–49 (1998). simplification or complete destruction of their molecular
79. R. Bryant and E. Laishley, Appl. Microbiol. Biotechnol. 38, structure by physiological reactions catalyzed by microor-
824–827 (1993). ganisms (2–7). Biodegradation is routinely measured by
80. F. W. Ochynski and J. R. Postgate, in C. H. Openheimer and applying chemical and physiological assays to labora-
C. C. Thomas, eds., Marine Microbiology, Springfield, Illinois, tory incubations of flasks containing pure cultures of
1963, pp. 426–441. microorganisms, mixed cultures, or environmental sam-
81. I. B. Beech and C. C. Gaylarde, Int. Biodet. 27, 95–107 (1991). ples (e.g., soil, water, sediment, or industrial sludges).
82. I. B. Beech, C. C. Gaylard, J. J. Smith, and G. G. Geesey, When attempting to measure biodegradation or judge
Appl. Microbiol. Biotechnol. 35, 65–71 (1991). the biodegradability of substances, the investigator must
83. V. I. Zinkevich et al., Int. Biodet. Biodegrad. 36, 163–172 define the environmental context so that potential reac-
(1996). tants and products can be identified. Microorganisms
476 BIODEGRADABILITY: METHODS FOR ASSESSING BIODEGRADABILITY UNDER LABORATORY AND FIELD CONDITION

can catalyze only reactions that are thermodynamically or do not make contaminant compounds disappear, as
possible. Furthermore, reaction mechanisms are largely judged by analytical chemical criteria. Phase 2 begins with
constrained by precedents set during the evolution of phys- the isolation of pure cultures of contaminant-degrading
iological and biochemical functions. Because of ongoing microorganisms. Once these have been obtained, refined
microbial evolution and biochemical research, our under- physiological, enzymatic, and molecular biological assays
standing of mechanisms by which microorganisms degrade may then be performed. As DNA sequences of genes
substrates continues to expand. that code for metabolic pathways become increasingly
available, molecular procedures will continue to gain
prominence in biodegradation protocols. One of the
LABORATORY METHODS
final goals of the procedures shown in phase 2 is
understanding the molecular basis for gene expression
Principles for Measuring Biodegradability in the Laboratory and regulation.
Biodegradation methodologies are designed to confirm,
Design and Implementation of Biodegradation Assays Using
demonstrate, and explore both the net chemical changes
Environmental Samples and Pure Cultures
and the associated intracellular details pertinent to
how microorganisms influence the fate of contaminants. The traditional black-box approach to biodegradation
The procedures span a broad range of disciplines and assays asks the question, ‘‘Are microorganisms within
sophisticated protocols. Figure 1 provides an overview this complex microbial community (e.g., derived from
of the variety of objectives, disciplines, and protocols soil, water, sediment, or industrial sludge) able to
that play key roles in biodegradation research. The two metabolize the compound of interest?’’ To answer this
phases that serve as main divisions in Figure 1 result question, one aseptically gathers samples from a given
from the degree to which scientific detail is pursued. field location, dispenses known weights or volumes of
Phase 1 procedures treat samples of soil, sediments, water, the samples to replicated vessels, handles the samples
or industrial effluents simply as ‘‘black boxes’’ that do in a variety of ways that include a treatment that

Phase 1: Laboratory enrichment and demonstration of net metabolic activity

1. Soil, sediment, water, or industrial effluent in field site

2. Aseptically remove, contain, transport to laboratory

3. Divide into replicate live and abiotic treatments

4. If appropriate, add radiolabeled or unlabeled organic compound of interest

5. Use tools of analytical chemistry or physiology to periodically measure consumption of parent

compound and coreactants or production of metabolites or physiological endproducts

6. Compare time-course data obtained from live and abiotic treatments

7. Interpret and consider more reductionistic procedures shown in Phase 2.

Phase 2: Isolation of pure cultures and examination of physiological, biochemical, and

molecular basis of pollutant metabolism
8. Isolate pure cultures capable of expressing metabolic activity determined in Phase 1

9. Characterize growth, cell yield, sequential induction, and other physiological characteristics of
the microorganisms during pollutant metabolism

10. Extract and identify metabolites, enzymes, and cofactors associated with pollutant metabolism

11. Cell-free examination of metabolites, enzymes, and cofactors

12. Determine portion of genomic or plasmid DNA that codes for pollutant metabolism by
screening a cloned DNA library, by transposon mutagenesis, or other procedures

13. Conduct hybridization, restriction mapping, and sequencing DNA analyses seeking open
reading frames, homology with similar genes, and other key insights

14. Elucidate details of gene expression and regulation via a variety of genetic and molecular
techniques that include transposon mutagenesis, construction of expression clones,
insertional inactivation, and inducer/reporter experiments.
Figure 1. Two phases of procedures for understanding biodegradation processes. Phase 1 begins
with environmental samples. Phase 2 proceeds through biochemical and molecular aspects of
pollutant metabolism by single microorganisms (from Ref. 1).
 BIODEGRADABILITY: METHODS FOR ASSESSING BIODEGRADABILITY UNDER LABORATORY AND FIELD CONDITION 477

has been either sterilized or poisoned, incubates the biodegradation protocols. Step 1, a background issue
test samples under laboratory conditions, and employs considering information use, is fundamental to all related
over time both chemical and physiological assays that experimental decisions. The degree to which experimental
monitor the fate of the test compound within experimental minutiae of a given testing protocol must be initially
vessels (Fig. 1; phase 1). The objective of this general considered is commensurate with the scrutiny that the
experimental design for biodegradation procedures using final data will undergo. Artifacts and biases in data are
environmental samples or pure cultures is remarkably virtually unavoidable in biodegradation assays (see later);
simple, yet there is a substantial series of obstacles that thus, it may be wise to simply accept methodological
must be overcome before one obtains clear data that limitations rather than worry about initial potential
truly test a given set of specific hypotheses. Every design technical design flaws that may later have no practical
parameter selected for inclusion in a biodegradation assay impact.
can influence the resultant data. Therefore, decisions Once the reason for conducting the biodegradation
made in implementing biodegradation assays should be assay has been put in perspective (Table 1, step 1), another
well reasoned. Table 1 summarizes many of the practical background issue, that of physiological conditions, should
and theoretical decisions that must be made in developing be confronted. Step 2 appears in Table 1 to acknowledge

Table 1. Steps and Decisions Essential for Implementing Biodegradation Assays (After Ref. 7)

Step Decisions

1. Background: Determine how the resultant Objectives range from information about crude ‘‘biodegradation
data will be interpreted and used potential’’ to tests of specific hypotheses about physiological or
biochemical factors governing biodegradation processes
2. Background: Select the physiological The pivotal physiological concern is defining the mechanism by which
conditions under which pollutant metabolism the compound(s) is metabolized. Of primary importance is
is to be measured discriminating among such possibilities as cometabolic reactions, its
use as an electron acceptor, and its use as a carbon and energy
source. Other concerns address conditions in experimental flasks
such as nutrient sufficiency, which final electron acceptor regimen
should dominate, what pollutant concentration ranges should be
examined, and if conditions should change (batch culture) or remain
constant (continuous culture) during the assay
3. Practice: Select and aseptically prepare or For assays using environmental or industrial samples, aseptic
sample the microorganisms whose sampling techniques involve use of tools (such as flame-sterilized
physiological activity is of interest scoops, spatulas, and knives) and sample placement within sterilized
glass or plastic containers. For assays using pure cultures of
microorganisms, the microorganisms must be aseptically grown
under conditions that carefully define the cell physiological status
(e.g., stage of growth, cell numbers, induced enzyme systems,
nutritional state) desired by the investigator
4. Practice: Select the physical apparatus and Glass (or plastic) vessels must be assembled. These contain the test
hence the physiological setting for compound(s), the microorganisms being studied, and any
biodegradation reactions to occur accompanying components of soils, sediments, sludges, and water in
various ratios. Experimental hardware may be fitted with a variety
of gas and water exchange assemblies for maintaining physiological
conditions and assaying reaction progress
5. Practice: Select a metabolic activity assay that The general assay categories are physiological assays (e.g.,
is sensitive, effective, convenient, inexpensive, respirometry or growth) and chemical assays (which include gas
and compatible with experimental objectives chromatography, gas chromatography-mass spectrometry,
high-performance liquid chromatography, and radiotracer
techniques)
6. Practice: Aseptically prepare stock solutions of The validity of the results from biodegradation assays using
14 C-labeled organic compounds. Check 14 C-labeled substrates is dependent on substrate radiopurity and

radiopurity aseptic preparation of stock solutions

7. Practice: Complete the experimental design a. Concentration of the test substrate(s)
parameters for the assay vessels and the b. Number of replicated flasks per treatment
assays themselves c. Whether flasks can be sampled repeatedly or if they require sacrifice
at each sampling period
d. Frequency of sampling
e. Method of preparing abiotic controls
f. Methods for separating radioactive parent and product compounds
from one another
478 BIODEGRADABILITY: METHODS FOR ASSESSING BIODEGRADABILITY UNDER LABORATORY AND FIELD CONDITION

the fact that biodegradation is only a small portion of information congeals as a coherent picture (it makes
the perhaps thousands of physiological reactions occur- sense).’’ There needs to be consistency, redundancy, and
ring simultaneously when both pure cultures and mixed convergence of all types of evidence from as many of the
microbial populations in environmental samples are incu- appropriate scientific disciplines as are available.
bated in the laboratory. These physiological processes feed Because the key players in bioremediation are microor-
one another, interact in complex ways, and can be gov- ganisms, it is essential that the process makes sense to
erned by many of the sometimes inadvertent physical and the microorganisms themselves, in the physiological and
chemical manipulations made while preparing, incubat- thermodynamic contexts where they reside. Contexts for
ing, and sampling assay vessels. Uncertainties become bioremediation processes range from a variety of field sites
particularly striking when one is attempting to trou- in which organic contaminants have spilled accidentally
bleshoot failed attempts to demonstrate biodegradation (e.g., marine coastlines, desert soils, freshwater streams,
activity. The interplay between fundamental knowledge of or anaerobic deep subsurface sediments) to various aero-
physiology and experimental design parameters demands bic or anaerobic engineered stirred and staged bioreactor
that a variety of issues be confronted: (1) The mechanism systems. Regardless of the particular context, each must
by which the compound is metabolized (e.g., as a car- be scrutinized as a habitat for microbial metabolism in
bon source, as a nitrogen source, or as a cometabolic which individual cells can develop into populations and
substrate whose transformation will occur only when complex ecological communities whose fundamental phys-
another compound is supplied), (2) inclusion versus exclu- iological needs for adenosine triphosphate generation,
sion of potential growth-limiting vitamins and minerals, carbon assimilation, terminal electron acceptors, other
(3) inclusion versus exclusion of air in the headspace of inorganic and organic nutrients, and dynamic intercellular
the reaction vessel, (4) the solid-to-liquid ratios used in interactions (competition, synergism, interspecies hydro-
test vessels containing soil, sediments, or sludges, (5) the gen transfer, commensalism, predation, parasitism, etc.)
multiple roles of compounds in physiological reactions demand constantly improving sets of hypotheses aimed
(for instance, nitrate can serve as both a nitrogen source at refining our understanding of bioremediation. Once
and a final electron acceptor), and (6) the fact that the the fundamental thermodynamic, nutritional, and ecolog-
compound whose biodegradation is being tested may be ical bases for the sought metabolic functions are initially
toxic at high concentrations or fall below some minimum conjectured, a series of hypotheses will naturally unfold
threshold value for uptake and cell growth at low concen- that provide a means for documenting the bioremediation
trations. Background considerations raised in steps 1 and process of interest on a site-specific and case-specific basis.
2 guide most of the practical steps needed for completing Table 2 contains four examples of contaminants in
the implementation of the biodegradation assays (Table 1, field sites whose physiological contexts dictate how
steps 3 to 7). Detailed considerations pertinent to steps 3 microorganisms can metabolize the offending organic
to 7 (Table 1) have been discussed previously (1,7,8). compounds. Knowledge from laboratory-based (using
environmental samples, mixed cultures, and/or pure
cultures) assays provides the biochemical basis for
FIELD METHODS mechanisms operating in the field. Answers to key
questions, such as ‘‘Are the contaminants suitable
Assessing Biodegradation in the Field carbon and electron sources?,’’ ‘‘Which physiological
There is a fundamental paradigm for verifying that electron acceptors (oxygen, NO3 − , Fe3+ , Mn4+ , SO4 2− ,
the biodegradation processes we hope are occurring are carbon dioxide) are required coreactants?,’’ ‘‘Are the
actually occurring in field sites. The paradigm begins contaminants, themselves, final electron acceptors?,’’ and
by modestly admitting that both microorganisms and ‘‘What competing reactions may slow or prevent the sought
their habitats are incomplete puzzles. Our task is to biodegradation?,’’ provide a framework that launches a
relentlessly find new ways to create the puzzle pieces broad array of possible assays that can argue for or against
describing microbiological processes and to assemble the successful establishment of a given biodegradation
them logically. The scientific disciplines that contribute process (1,7,8). The example contaminated sites of Table 2
information and techniques toward creating the puzzle range from aerobic soil to aerobic and anaerobic aquifers.
pieces include microbiology, geochemistry, hydrology, The assays range from field-based oxygen probes, to
biochemistry, soil science, physiology, molecular biology, counts of contaminant-degrading bacteria, to laboratory
biodegradation assays, and to molecular biological assays
analytical chemistry, computer modeling, and both
of DNA and ribonucleic acid. The following sections of this
environmental and chemical engineering. It must be
article elaborate on the reasoning and detailed analyses
recognized that each of these disciplines is actively
required for the generation and testing of hypotheses that
being advanced and therefore, contributes a dynamic
allow bioremediation technology to be verified. Related
spectrum of expertise to bioremediation, ranging from
aspects of bioremediation and biodegradation have been
theoretical and basic knowledge to applied and practical
reviewed in several recent books (2,21–29).
instrumentation.
Verifying field biodegradation is perhaps best achieved
A Three-Step Strategy for Verifying Bioremediation
in two mutually supportive ways (9): (1) succinctly using
common sense and (2) using elaborate reasoning and The reasons for establishing sound scientific criteria
analyses (see later). The succinct answer is ‘‘We know that for microbiological involvement in contaminant loss are
bioremediation is taking place when all of the available (1) biodegradation processes are often unique in their
Table 2. Examples of Contaminated Sites, Hypothesized Key Bioremediation Processes, and the Corresponding
Field and Laboratory Measurements that Allow Site-Specific and Case-Specific Verification of Microbiological
Destruction of Contaminants (After Ref. 9)

Example Hypothesized Key Supportive Field and

Sites Bioremediation Processes Laboratory Measurements

Aerobic soil Heterotrophic microorganisms are growing • Coincident depletion of petroleum components
contaminated with using petroleum components as carbon and and oxygen in the field
petroleum products energy sources (10,11,12). Metabolism in this • Corresponding production of carbon dioxide
context relies on oxygen, both in the attack of • High numbers of petroleum-degrading aerobic
aliphatic and aromatic compounds, and as a heterotrophs inside but not outside the
final electron acceptor in respiratory chains contaminated areas
• If petroleum has a distinctive 13 C/12 C ratio,
this should be reflected in the carbon dioxide
• Adding nitrogen or phosphorus fertilizer to
replicated plots may relieve nutrient
limitation, hence enhance loss from field plots
compared to unfertilized controls
• Genes involved in the catabolism of petroleum
components should be expressed in high
abundance inside but not outside the
contaminated zone
Anaerobic aquifer Dehalorespiring bacteria are using chlorinated • Dechlorinated daughter products,
contaminated with aliphatic compounds as final electron trichloroethene (TCE), dichloroethenes (DCE),
perchloroethylene acceptors (13,14). Dechlorination reactions are vinylchloride (VC) within contaminant plume
governed by: complex microbial and chemical • Products of complete detoxification, such as
interactions that generate physiological ethene, should be inside and not outside the
electron sources (especially hydrogen gas); the plume
presence of alternative electron acceptors • Adaptation of site microorganisms to
(NO3 − , Mn4+ , Fe3+ , SO4 2− , carbon dioxide) in dehalorespiration can be documented by
site sediments and waters; and ecological and finding dechlorination activity in site samples
physiological competition among the from inside but not outside the contaminant
microorganisms carrying out the metabolism plume
that links electron donors and • Immunological or polymerase chain
acceptors (15,16) reaction–based data demonstrating the
presence of dehalorespiring enzymes, genes,
and characteristic bacteria inside but not
outside the plume
Aerobic aquifer TCE destruction is achieved cometabolically by • Microcosms prepared with site samples
contaminated with aerobic microorganisms supplied with a consume TCE only when supplied with both
TCE primary carbon source. Oxygenase enzymes oxygen and the primary substrate
(involved in metabolizing primary substrates • In recirculating field site waters, TCE loss is
such as methane, propane, toluene, and enhanced only when both oxygen and the
phenol) fortuitously convert TCE to unstable primary substrate are supplied
compounds that spontaneously hydrolyze to • Assays for oxygenase enzymes, genes, and
nontoxic and/or readily biodegradable appropriate metabolites (e.g.,
components (17–19) trans-dichloro-ethylene oxide) reveal high
abundances inside but not outside the
contaminated zone
Anaerobic aquifer Aromatic fuel components, especially toluene, • Microcosms containing site sediments
contaminated with serve as growth substrates for anaerobic incubated under sulfate reducing conditions
jet fuel microorganisms that utilize sulfate as a final produce 14 CO2 from 14 C-labeled toluene and
electron acceptor benzene (20)
• Sulfate is depleted along the groundwater flow
path in the field site
• Dissolved inorganic carbon (e.g., carbon
dioxide) increases along the flow path in field
sites (20)
• Contaminant plume has ceased advancing
despite a constantly dissolving reservoir of jet
fuel (20)
• A solute transport model accounts for
dispersion, flow velocity, and adsorption, and
produces biodegradation rate estimates that
are consistent with microcosm estimates (20)

479
480 BIODEGRADABILITY: METHODS FOR ASSESSING BIODEGRADABILITY UNDER LABORATORY AND FIELD CONDITION

capacity to break intramolecular bonds of contaminant All site characterization data must be interpreted in
compounds; thus, contaminants can be destroyed and not terms of the physiological processes that produce and con-
simply transferred from one location to another, as is the sume geochemical constituents. Final electron acceptors
case in many other pollution control technologies; (2) when that dominate the physiological reactions of field sites (or
the mechanism of pollutant destruction is certain, key site discrete zones therein) provide useful criteria for cate-
management decisions about process enhancement can gorizing biogeochemical regimes as aerobic, denitrifying,
be made; and (3) for bioremediation to meet pollution- iron reducing, manganese reducing, fermentative, dehalo-
control needs of the society, the industry must adopt genating, sulfate reducing, or methanogenic (41–43).
some standards for uniformity and quality control so that These physiological regimes are often separated in space
credibility and reliability can be attained (25–27). and/or time in field sites and largely determine the
However, the question remains: What is an adequate mechanisms of contaminant biodegradation. Information
proof of bioremediation? The legal system of the United establishing the physiological regime(s) operating at par-
States provides a variety of categories of certainty in ticular field sites is provided by field measurements of
interpreting evidence. The categories depend on the type the contaminants themselves, of concentration gradients
of the case and the significance of the issues. Among the of coreactant final electron acceptors (e.g., oxygen, NO3 − ,
different burdens of proof are (1) proof beyond reasonable Fe3+ Mn4+ , SO4 2− , halogenated compounds), and of end
doubt, (2) proof in a clear and convincing manner, and products of microbial metabolism (e.g., carbon dioxide,
(3) proof beyond a preponderance of doubt. This article Fe2+ , Mn2+ , S2− , N2 O, NH4 + , organic acids, reductive
neither intends, nor is able, to dictate to regulatory or dehalogenation daughter products, and methane) along
legal agencies what level of proof should be deemed site transects. In this regard, Lovley and coworkers (44)
adequate for bioremediation technology practitioners. and Chapelle and coworkers (45) have devised a gas sam-
Nonetheless, approaches are discussed here that can be pling bulb protocol for anaerobic groundwaters in the field,
used to distinguish biotic and abiotic reactions affecting which, in combination with hydrogen gas determinations
contaminants at field sites in which bioremediation and Winkler titrations for oxygen (40), provides definitive
technology is being applied (for additional discussion see information on dominant anaerobic redox couples.
Refs. 1,6,7,11,12,25,26,30–37). The goal of establishing site-specific historical records of
The consensus of a National Research Council the behavior of contaminants, coreactants, and metabolic
(NRC) (25) committee in recommending criteria proving products is a simple one. Theoretically, compilation of
in situ bioremediation is as follows: such field data documents the effects of contaminant-
attenuating processes over time. In conjunction with other
assays (see later), field data can be interpreted in ways
1. Develop historical records documenting loss of
that may implicate biodegradation as a cause of pollutant
contaminants from field sites.
losses. However, as in understanding in situ physiological
2. Perform laboratory assays unequivocally showing regimes, obtaining robust, interpretable field-monitoring
that microorganisms in site-derived samples have data may be an elusive goal if contaminant characteristics
the potential to metabolize the contaminants under and site conditions are complex. The overall objective is to
expected site conditions. establish a site-monitoring regime using a network of con-
3. Demonstrate that the metabolic potential measured sistent sampling locations, which affords the acquisition
under criterion 2 is actually expressed in the field. of contaminant concentration and other measurements
To achieve this, microbiological mechanisms of con- that are comparable over time. If the distribution of con-
taminant attenuation must be distinguished from taminants at the site and factors influencing contaminant
abiotic ones. Evidence deemed suitable for these transport (e.g., climate, hydrology, commercial, or indus-
purposes will vary according to the contaminants trial activities) are erratic, then the pertinent database
and conditions found at each site. on contaminant behavior may be so noisy as to mask any
trends. However, in many sites the type of contaminant
Implementing Step One: Site Monitoring to Understand monitoring protocols required by concerned regulatory
Site Biogeochemistry and Establish Historical Trends of Con- agencies can be integrated over time and can sometimes
taminant Behavior. It must be recognized that virtually all produce a clear historical record of diminishing contam-
locations in the biosphere (from the poles to the equator, inant concentrations from year to year. When such data
contaminated or pristine sites, engineered bioreactors, exist, they assist in meeting the first criterion for proving
or the deep sea) are inhabited by microorganisms. Fur- in situ bioremediation.
thermore, whenever physiological resources are available,
microbial metabolic activity will occur. Thus, site charac- Implementing Step Two: Laboratory Assays Demonstrat-
terization is designed to assess the resources and guide the ing That Microorganisms in Site Samples Have the Potential
documentation of their exploitation by microorganisms. to Transform the Contaminants Under Expected Site Condi-
There is a critical need to relate results of geochemi- tions. The biodegradation assays discussed in this article
cal measurements performed on field samples directly are designed to ask the question: ‘‘Are microorganisms
back to in situ processes and conditions. For details within samples of soil, water, or sediment microbial com-
of completing in situ analyses, avoiding site-sampling munities able to metabolize the compound(s) of interest
artifacts, and understanding site biogeochemistry see Ref- under conditions that are relevant to the specific field site
erences 7,38– 40. of interest?’’ To answer this question, one gathers samples
Table 3. Strategies for Obtaining Evidence for Field Expression and Biodegradation Processes (After Ref. 9)

Type of Strategy Principles and Examples References

Internal conservative tracers Assess loss of certain compounds relative to the persistence of less-biodegradable, but 47–57
similarly transported, compounds. Examples include using ratios of
straight-to-branched chain alkanes (C17 /pristane, C18 /phytane) and ratios of other
compounds to hopane in crude oil; ratios of lower to higher chlorinated congeners in
PCB mixtures (trimethyl benzene can serve as a conservative tracer in
enzene-toluene-ethylbenzene-xylene (BTEX) plumes); ratios of nonchlorinated to
chlorinated aromatics in mixed solvents; and selective metabolism of one
stereoisomer of particular pesticides (e.g., σ -chlorocyclohexane)
Added conservative tracers In some field sites, contaminant mixtures may lack internal tracers but be amenable to 25,58,59
the addition of materials that provide a baseline measure of various transport
processes. Examples include helium to assess oxygen loss or carbon dioxide
production in groundwater, propane to assess toluene loss from a stream, and
bromide to assess groundwater flow
Added radioactive tracers In rare instances, regulatory authorities have allowed intentional field release of 60–63
radioactive (e.g., 14 C-labeled) pollutants in field sites. Subsequent recovery of 14 CO2 ,
14 C-metabolites, and 14 C-parent compounds provide, definitive proof of metabolic and

other field processes

Added stable isotopic tracers Pollutant compounds that are nonradioactive, but isotopically labeled with deuterium 64,65
or 13 C, have been released in field sites. Subsequent stable isotopic analyses of field
samples for labeled CO2 , metabolites, and/or the parent compound provide proof of
metabolic and other field processes
Stable isotopic fractionation CO2 has different 13 C/12 C ratios depending on the 13 C/12 C signature of the substrates 49,66–70
patterns respired and the 13 C-enriching process of methanogenesis. When site-specific
signatures of both inorganic and organic carbon reservoirs have been characterized,
the relative contribution of pollutant biodegradation to the pool of CO2 can be
discerned. The radioactive (14 C) component of CO2 is also revealing because
petroleum contaminants contain no 14 C
Detection of intermediary When sufficient biochemical knowledge of pollutant biodegradation has accrued, 71–77
metabolites particular metabolites can be targeted using a combination of careful sampling and
analytical chemistry. Detection of stable (dead-end) metabolites and transient
metabolites (indicative of ‘‘real-time’’ biodegradation) has been reported. The
metabolites include trans- dichloroethylene oxide, dihydrodiols of aromatic
compounds, DDE, and hydroxylated pesticides
Replicated field plots Some relatively homogeneous field sites are amenable to randomized block designs of 47,53,78–81
treatments that stimulate microbiological activity. Comparing the loss of pollutants
from plots with and without nutrients and/or inocula can demonstrate effectively the
role of microorganisms in field biodegradation
Microbial metabolic Naturally occurring microbial communities that grow in response to pollutant exposure 16,82
adaptation have predictable characteristics relative to adjacent unexposed communities.
Adaptation is reflected in laboratory or field measure of: qualitative pollutant
metabolian or rates of pollutant metabolism; numbers of specific pollutant degraders;
and enhanced concentrations of protozoan predators of bacteria inside but not outside
contaminant plumes
Molecular biological Based on molecular biological characterization of pure cultures capable of pollutant 83–90
indicators metabolism, a variety of assays consistent with established genetic sequences and
their expression can be devised. These include polymerase chain reaction (PCR)
amplification of structural genes, messenger ribonucleic acid (mRNA) extracted from
field sites, reverse-transcriptase PCR detection of mRNAs, nucleic acid sequencing,
immunodetection of enzymes and metabolites, and 16S ribosomal RNA analysis of
the composition of microbial communities
Gradients of coreactants Ongoing in situ metabolism of pollutants consumes physiological final electron 20,91,92
and/or products acceptors and generates metabolic endproducts that reflect site-specific pollutant
metabolism. Chemical gradients in field sites should be apparent using measures
that include oxygen, NO3 − , Mn4+ , Fe3+ , SO4 2− , CO2 , NO2 − , N2 O, Mn2+ , Fe2+ ,
methane, hydrogen, pH, and alkalinity
In situ rates of respiration A subset of the previous entry that has been effectively applied to engineered 93,94
bioremediation of subsurface sites involves cessation of an oxygen (or air) sparging
regime, followed by insertion of an oxygen probe that documents real-time oxygen
consumption. This respiratory activity should be high inside but not outside the
contaminated area. The conserved gas, helium, can be included in the sparging step
to account for diffusional O2 loss
(continued overleaf )

481
482 BIODEGRADABILITY: METHODS FOR ASSESSING BIODEGRADABILITY UNDER LABORATORY AND FIELD CONDITION

Table 3. (Continued)

Type of Strategy Principles and Examples References

Mass balances of Under well-defined hydrogeologic regimes, fluxes of water contaminants and 56,92,95–97
contaminants, coreactants, physiological electron donors or acceptors can be quantified in a cross-sectional
and products (total analysis of site sampling stations. The stoichiometry of all appropriate aerobic,
expressed assimilative anaerobic, isotopic fractionation, and inorganic equilibria reactions can serve to
capacity) predict and distinguish biotic from abiotic processes and to identify contributions
from a variety of microbiological groups
Computer modeling that This approach considers quantitative aspects of fluid flow, dilution, sorption, 92,98–101
incorporates transport and volatilization, mixing, microbial growth, and metabolic reaction stoichiometries to
reaction stoichiometries of achieve an integrated and predictive tool for understanding all processes influencing
electron donors and the fate of pollutant compounds. This approach resembles the previous entry, but is
acceptors implemented on a larger scale and uses more sophisticated computations

aseptically from a given field site or bioreactor, dispenses specific microbiological processes (stable isotope fraction-
known weights or volumes of the samples to replicated ves- ation, detection of intermediary metabolites, stimulation
sels, handles the samples in a variety of ways that include of microbial activity in replicated field plots, metabolic
a treatment that has been either sterilized or poisoned, adaptation, in situ respiration, and gradients of coreac-
incubates the test samples under laboratory conditions, tants and/or products) that are manifest as observable
and employs both chemical and physiological assays that geochemical changes in the field. The molecular biological
monitor the fate of the test compound within experimental strategy in Table 3 is an elegant, emerging approach that
vessels over time (Fig. 1, phase 1). These procedures have is constantly strengthened by genetic links that are forged
been described earlier. between information from pure cultures and real-world
mixed microbial communities. The linkages are limited by
Implementing Step Three: Evidence for Field Expression the relatively small database of genetic sequences perti-
of Biodegradation Potential. Three sources of uncertainty nent to pollutant metabolism and the uncertain metabolic
must be confronted and overcome when demonstrating diversity that may arise when genes of unrelated lin-
that microorganisms are the active agents of pollutant eage may have converged on the same metabolic function.
loss in bioremediation projects. The final two strategies in Table 3 (computer modeling
and mass balances of contaminants, reactants, and prod-
1. We must acknowledge that extrapolation from ucts) attempt to account quantitatively for both transport
laboratory-based metabolic activity assays to the and metabolic processes within entire field sites or along
field is usually unwise because of the propensity distinct transects therein.
of microorganisms in field samples to respond to
laboratory-imposed physiological conditions that are CONCLUSION
unlikely to match those in the field perfectly (6,46).
2. The spatial heterogeneity of field sites may impede or Understanding and proving biodegradation processes
completely prevent trends in the behavior of environ- under laboratory and field conditions is a science of ongo-
mental contaminants from being discerned (12,20). ing discovery. This discovery requires a close dialog among
3. The action of a multitude of both abiotic and many disciplines. It must be recognized that only under
biotic processes may contribute simultaneously to relatively rare circumstances is a proof of field bioreme-
pollutant attenuation (6,20). diation unequivocal when a single piece of evidence is
relied on. In the majority of cases, the complexities of con-
To contend with these challenges, several strategies taminant mixtures, their hydrogeochemical settings, and
have been developed for verifying the success of field biore- accounting for competing abiotic mechanisms of contami-
mediation efforts in truly activating pollutant-destroying nant loss make it a challenge to document biodegradation
microbial processes in field sites and bioreactors. These processes. Unlike controlled laboratory experimentation
(comprehensively codified in Table 3) are simple, logical wherein measurements can usually be interpreted easily,
expressions of the fundamental paradigm for verifying cause-and-effect relationships are often very difficult to
bioremediation introduced previously and in Table 1. The establish at field sites. Furthermore, certain bioremedia-
strategies that appear in Table 3 are firmly based on the tion data that may be convincing for some authorities may
physiological principles that distinguish between biotic not be convincing for others. Thus, in documenting biore-
and abiotic contaminant attenuation processes. Four of mediation, the several approaches described previously
the strategies involve tracers (internal conservative, added should be independently pursued: a consistent, logical case
conservative, added radioactive, and added stable iso- relying on convergent lines of independent evidence should
topic) that either account for or circumvent problems be built. The three-step strategy for verifying bioremedia-
arising from abiotic changes in field concentrations of tion described previously has been augmented recently by
contaminants and related metabolites. Six of the strate- that of a new NRC Committee (26). What might be consid-
gies in Table 3 rely on detailed prior knowledge of ered a fourth step and an overarching goal is assuring the
 BIODEGRADABILITY: METHODS FOR ASSESSING BIODEGRADABILITY UNDER LABORATORY AND FIELD CONDITION 483

public that bioremediation in specific sites will be reliable, 26. National Research Council, Natural Attenuation for Ground-
sustainable, and quantitatively complete. water Remediation, National Academy Press, Washington,
D.C., 2000.
27. B. E. Rittmann et al., In Situ Bioremediation, 2nd, ed.,
BIBLIOGRAPHY Noyes Publications, Park Ridge, Neb., 1994.
28. L. Y. Young and C. E. Cerniglia, eds., Microbial Transfor-
1. E. L. Madsen, in C. J. Hurst et al., eds., Manual of Environ- mation and Degradation of Toxic Organic Chemicals, Wiley-
mental Microbiology, ASM Press, Washington, D.C., 1997, Liss, Inc., New York, 1995.
pp. 709–720. 29. I. M. Head, Microbiology 144, 599–608 (1998).
2. M. Alexander, Biodegradation and Bioremediation, Aca- 30. A. Heitzer and G. S. Sayler, TIB Tech. 11, 334–343 (1993).
demic Press, New York, 1999. 31. J. M. Tiedje, In Situ Bioremediation: When Does It
3. L. P. Wackett and C. D. Hershberger, Biocatalysis and Work?, National Academy Press, Washington, D.C., 1993,
Biodegradation, ASM Press, Washington, D.C., 2001. pp. 110–120.
4. R. M. Atlas, Curr. Opin. Biotechnol. 3, 220–223 (1991). 32. J. A. McDonald and B. E. Rittmann, Environ. Sci. Technol.
5. D. T. Gibson, ed., Microbial Degradation of Organic Com- 27, 1974–1979 (1993).
pounds, Marcel Dekker, New York, 1984. 33. J. F. Quensen III and J. M. Tiedje, in D. Sheehan, ed.,
6. E. L. Madsen, Environ. Sci. Technol. 25, 1662–1673 (1991). Methods in Biotechnology, 2. Bioremediation Protocols,
7. E. L. Madsen, in R. Burlage et al., eds., Techniques in Humana Press Inc., Totowa, N.J., 1997, pp. 241–253.
Microbial Ecology, Oxford University Press, New York, 1998, 34. J. P. Salanitro, Ground Water Monit. Remediation 13,
pp. 354–407. 150–161 (1993).
8. D. D. Focht, in R. W. Weaver et al., eds., Methods of 35. M. J. R. Shannon and R. Unterman, Annu. Rev. Microbiol.
Soil Analysis. Part 2. Microbiological and Biochemical 47, 715–738 (1993).
Properties, Soil Science Society of America, Madison, Wis., 36. P. J. Swannell et al., Microbiol. Rev. 60, 342–365 (1996).
1994, pp. 407–426. 37. R. Atlas and R. Unterman, in A. L. Demain and J. E. Davies,
9. E. L. Madsen, in J. J. Valdes, ed., Bioremediation, Kluwer eds., Manual of Industrial Microbiology, 2nd ed., ASM Press,
Academic Publishers, Netherlands, 2000, pp. 101–122. Washington, D.C., 1999, pp. 666–681.
10. R. M. Atlas and C. E. Cerniglia, Bioscience 45, 332–338 38. E. R. Weiner, Applications of Environmental Chemistry: A
(1995). Practical Guide for Environmental Professionals, Lewis
11. H. S. Rifai et al., in R. E. Hinchee et al., eds., Intrinsic Publishers, Boca Raton, Fla., 2000.
Bioremediation, Battelle Press, Columbus, Ohio, 1995, 39. L. H. Keith, Environmental Sampling and Analysis: A
pp. 1–29. Practical Guide, Lewis Publishers, Boca Raton, Fla., 1991.
12. J. T. Wilson and M. D. Jawson, in H. D. Skipper and 40. American Public Health Association (APHA), in A. E. Green-
R. F. Turco, eds., Bioremediation Science and Applications, berg et al., eds., Standard Methods for the Examination of
Soil Science Society of America, Inc., Madison, Wis., 1995, Water and Wastewater, 18th, ed., APHA Washington, D.C.,
pp. 293–303. 1992.
13. X. Maymo-Gatell et al., Science 276, 1568–1571 (1997). 41. E. J. Bouwer, in R. D. Norris et al., eds., Handbook of
14. P. L. McCarty, Science 276, 1521–1522 (1997). Bioremediation, Lewis Publishers, Boca Raton, Fla., 1994,
pp. 149–175.
15. D. E. Fennell et al., Environ. Sci. Technol. 31, 918–926
(1997). 42. H. F. Hemond and E. J. Fechner, Chemical Fate and Trans-
port in the Environment, 2nd, ed., Academic Press, New
16. R. M. Yager et al., Environ. Sci. Technol. 31, 3138–3147
York, 2000.
(1997).
43. J. J. Stumm and W. Morgan, Aquatic Chemistry, 3rd, ed.,
17. G. D. Hopkins et al., Environ. Sci. Technol. 27, 2542–2547
John Wiley and Sons, New York, 1995.
(1993).
44. D. R. Lovley et al., Environ. Sci. Technol. 28, 1205–1210
18. S. M. Pfiffner et al., J. Ind. Microbiol. Biotechnol. 18,
(1994).
204–212 (1997).
45. F. H. Chapelle et al., Environ. Sci. Technol. 31, 2873–2877
19. L. Semprini and P. L. McCarty, Ground Water 30, 37–44
(1997).
(1992).
46. E. L. Madsen, in G. Stotzky and J.-M. Bollag, eds., Soil
20. F. H. Chapelle et al., Ground Water 34, 691–698 (1996).
Biochemistry, vol. 9, Marcel Dekker, Inc., New York, 1996,
21. H. D. Skipper and R. F. Turco, eds., Bioremediation Science pp. 287–370.
and Applications, Soil Science Society of America, Inc.,
47. J. R. Bragg et al., Nat. (Lond.) 368, 413–418 (1994).
Madison, Wis., 1995.
48. M. R. Harkness et al., Science 259, 503–507 (1993).
22. R. L. Crawford and D. L. Crawford, eds., Bioremediation:
Principles and Applications, Cambridge University Press, 49. A. W. Jackson et al., Environ. Sci. Technol. 30, 1139–1144
New York, 1996. (1996).
23. P. E. Flathman et al., Bioremediation: Field Experience, 50. D. H. Kampbell et al., J. Hazard. Mater. 49, 197–204 (1996).
Lewis Publishers, Boca Raton, Fla., 1994. 51. P. Ludwig et al., Mar. Chem. 38, 13–23 (1992).
24. R. D. Norris et al., Handbook of Bioremediation, Lewis 52. R. C. Prince et al., Environ. Sci. Technol. 28, 142–145
Publishers, Boca Raton, Fla., 1994. (1994).
25. National Research Council, In Situ Bioremediation: When 53. P. H. Pritchard and C. F. Costa, Environ. Sci. Technol. 25,
Does It Work? National Academy Press, Washington, D.C., 372–379 (1991).
1993. 54. V. A. Tett et al., FEMS Microbiol. Ecol. 14, 191–199 (1994).
484 BIODEGRADABLE DISSOLVED ORGANIC CARBON (BDOC)

55. T. H. Wiedemeier et al., in R. E. Hinchee et al., eds., Intrin- 86. A. Ogram et al., Appl. Environ. Microbiol. 61, 763–768
sic Bioremediation, Battelle Press, Columbus, Ohio, 1995, (1995).
pp. 31–51. 87. S. Selvaratnam et al., Appl. Environ. Microbiol. 61,
56. R. A. Williams et al., J. Ind. Microbiol. Biotechnol. 18, 3981–3985 (1995).
177–188 (1997). 88. M. S. Shields and S. C. Francesconi, in R. L. Crawford
57. C. Zipper et al., Appl. Environ. Microbiol. 62, 4318–4322 and D. L. Crawford, eds., Bioremediation: Principles and
(1996). Applications, Cambridge University Press, New York, 1996,
58. H. Kim et al., Environ. Sci. Technol. 29, 108–116 (1995). pp. 341–390.
59. W. D. Robertson, Environ. Toxicol. Chem. 13, 183–191 89. M. S. Wilson et al., Appl. Environ. Microbiol. 65, 80–87
(1994). (1999).
60. F. Führ et al., in H. Frehse, ed., Pesticide Chemistry — 90. J. Zhou et al., Appl. Environ. Microbiol. 63, 2384–2390
Advances in Industrial Research, Development, and Leg- (1997).
islation. Proceedings of the Seventh International Congress 91. B. C. Alleman and A. Leeson, eds., Natural Attenuation of
of Pesticide Chemistry (IUPAC), VCH Publishers, Weinheim, Chlorinated Solvents, Petroleum Hydrocarbons, and Other
Germany, 1990, pp. 37–48. Organic Compounds, Battelle Press, Columbus, Ohio, 1999.
61. F. Führ et al., eds., The Lysimeter Concept: Environmental 92. H. S. Rifai et al., in R. E. Hinchee et al., eds., Intrinsic
Behavior of Pesticides, American Chemical Society, Wash- Bioremediation, Battelle Press, Columbus, Ohio, 1995,
ington, D.C., 1998. pp. 53–58.
62. K. Lee et al., Microb. Ecol. 11, 337–351 (1985). 93. G. B. Davis et al., in R. E. Hinchee et al., eds., Monitoring
63. W. G. MacIntyre et al., Water Resour. Res. 29, 4045–4051 and Verification of Bioremediation 3(5), Battelle Press,
(1993). Columbus, Ohio, 1995, pp. 193–201.
64. J. Thierrin et al., Ground Water 33, 469–475 (1995). 94. D. X. Li, in R. E. Hinchee et al., eds., Monitoring and Ver-
65. J. Thierrin et al., Hydrol. Sci. J. 38, 309–322 (1993). ification of Bioremediation 3(5), Battelle Press, Columbus,
66. P. K. Aggarwal and R. E. Hinchee, Environ. Sci. Technol. Ohio, 1995, pp. 115–126.
25, 1178–1180 (1991). 95. P. Höhener et al., J. Microbiol. Methods 32, 179–192 (1998).
67. M. E. Conrad et al., Environ. Sci. Technol. 31, 1463–1469 96. L. Semprini et al., Water Resour. Res. 31, 1051–1062 (1995).
(1997). 97. D. A. Wolfe et al., Environ. Sci. Technol. 28, 561A–568A
68. E. L. Grossman, in D. J. Hurst et al., eds., Manual of (1994).
Environmental Microbiology, ASM Press, Washington, D.C., 98. B. A. Bekins et al., J. Contam. Hydrol. 14, 279–294 (1993).
1997, pp. 565–576. 99. C. Y. Chiang et al., Ground Water 27, 823–834 (1989).
69. D. Hunkeler et al., Environ. Sci. Technol. 33, 2733–2738 100. M. M. Lang et al., Ground Water 35, 565–573 (1997).
(1999).
101. L. Semprini and P. L. McCarty, Ground Water 29, 365–374
70. J. E. Landemeyer et al., Environ. Sci. Technol. 30, 1120–
(1991).
1128 (1996).
71. H. R. Beller and A. M. Spormann, J. Bacteriol. 179, 670–676
(1997).
72. H. R. Beller et al., Environ. Sci. Technol. 29, 2864–2870 BIODEGRADABLE DISSOLVED ORGANIC
(1995). CARBON (BDOC). See ASSIMILABLE ORGANIC CARBON
73. W. P. Flanagan and R. J. May, Environ. Sci. Technol. 27, (AOC) IN TREATED WATER: DETERMINATION AND SIGNIFICANCE
2207–2212 (1993).
74. R. N. Lerch et al., Environ. Sci. Technol. 29, 2759–2768
(1995).
75. L. Semprini et al., Ground Water 28, 715–727 (1990).
76. L. Semprini et al., Water Resour. Res. 31, 1051–1062 (1995). BIODEGRADABLE DISSOLVED ORGANIC
77. M. S. Wilson and E. L. Madsen, Environ. Sci. Technol. 30, CARBON IN DRINKING WATER
2099–2103 (1996).
78. J. A. Glaser and R. T. Lama, in H. D. Skipper and R. F. CHRISTIAN J. VOLK
Turco, eds., Bioremediation of Science and Applications, Indiana-American Water
Soil Science Society of America, Inc., Madison, Wis., 1995, Company, Inc.
pp. 117–133. Muncie, Indiana
79. R. T. Lamar et al., Soil Biol. Biochem. 26, 1603–1611 (1994).
80. R. L. Raymond et al., Appl. Environ. Microbiol. 31, 522–535 Natural organic matter (NOM) present in surface waters
(1976). and sometimes in groundwaters can be the root cause of
81. R. K. Steffan et al., Environ. Sci. Technol. 33, 2771–2781 various problems in drinking water. Organic material can
(1999). be responsible for undesirable water taste, odor, and color.
82. E. L. Madsen et al., Science 252, 820–833 (1991). Organic material lead to the formation of disinfection
83. F. J. Brockman, Mol. Ecol. 4, 567–578 (1995). by-products (after reaction with the disinfectant during
84. R. S. Burlage, in D. J. Hurst et al., eds., Manual of Environ- water treatment) and to bacterial proliferation within
mental Microbiology, ASM Press, Washington, D.C., 1997, distribution systems. The presence of low concentrations
pp. 115–123. of organic nutrients in water is sufficient to support
85. J. T. Fleming et al., Environ. Sci. Technol. 27, 1068–1074 biofilm growth within the pipe network, potentially
(1993). leading to a deterioration of bacterial water quality.
 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER 485

Therefore, the control of organic matter concentration inert gas to remove the inorganic carbon, before TOC anal-
is becoming recognized as an important part of the ysis of TOC. When these are purged, the remaining TOC
operation of drinking water plants and distribution is referred to as nonpurgeable TOC or as nonpurgeable
systems. The following section will list methods available DOC if the sample is filtered. Organic matter levels typi-
for characterizing organic matter and its biodegradable cally range from 0.1 mg/L to 10 to 20 mg/L. UV absorbance
fraction, present the impact of water treatment on organic at 254 nm, measured using a spectrophotometer, provides
materials, and describe the role of biodegradable organic some information on the aromaticity of the organic matter,
matter (BOM) in bacterial regrowth. This section focuses because aromatic rings absorb UV light (1). Fluorescence
on one type of measurement for bacterial substrate called indicates the level of polymerization of organic molecules
biodegradable dissolved organic carbon (BDOC). The next present in water (1).
section is devoted to a second type of measurement for
bacterial nutrients named assimilable organic carbon Extraction Methods Through Macroreticular Resins. Ex-
(AOC). Limited AOC data is presented to compare both traction techniques using selective resins fractionate
measurements. organic matter into two parts: humic (humic and ful-
vic acids) and nonhumic substances, which correspond
to hydrophobic and hydrophilic fractions, respectively (3).
MEASUREMENT OF ORGANIC MATTER AND ITS
Humic substances are complex macromolecules obtained
BIODEGRADABLE FRACTION after polymerization of organic molecules, following
biodegradation and chemical, biological, and photochemi-
Chemical Characterization of Organic Matter cal oxidation of animal and vegetable wastes. They include
Most organic matter are naturally present in water, but aromatic structures linked together with aliphatic chains.
one small portion is artificial and originates from pollution The extraction method using the XAD-8 resin (3) is the
and human activity. Chemical characterization of organic most common technique for isolating aquatic humic sub-
matter is complex. To date, it has not been possible to stances. After acidification of a water sample to pH 2,
identify all organic molecules in water. The difficulty of humic substances adsorb to the XAD-8 resin, whereas
this approach arises from the great complexity of the hydrophilic acids pass through the resin. Half of the
organic matrix, constituted by a mixture of hundreds of nonextractable chemicals includes hydrophilic acids and
simple molecules or heteropolymers, and the very low the other half represents well-identified chemicals such as
concentration of these compounds. Consequently, organic carbohydrates, amino acids and proteins, carboxylic acids,
matter characterization is only possible through various and trace compounds. After an additional step of elution
fractionations and measurement techniques that give with caustic soda, humic substances are divided into two
partial information on organic matter levels, chemical fractions: humic acids precipitating in acid (pH 1) and ful-
functions, or molecules, depending on the methodology vic acids that remain soluble in acidic medium. Generally,
used (1). fulvic acids have lower molecular weights and a higher
number of carboxyl functional groups than humic acids.
Fulvic acids represent the major part (80 to 85%) of the
Measurement of Surrogate Parameters. The measure-
humic substances (3). Humic substances and hydrophilic
ment of total organic carbon (TOC), dissolved organic
acids are responsible for 50 and 30% of the DOC, respec-
carbon (DOC), ultraviolet absorbance at 254 nm (UV254),
tively. The proportion of humic material in water can also
or fluorescence provides some insight into the organic lev- be assessed by calculating the specific UV absorbance
els and properties of water. TOC and DOC measurements (SUVA). The SUVA value corresponds to the ratio
are the most common techniques for monitoring organic between UV absorbance at 254 nm and DOC concentration
matter levels in waters. Expressed as TOC, organic mat- (SUVA = UVabs [cm−1 ] × 100/DOC [mg/L]). SUVA values
ter can be fractionated into DOC, colloidal compounds, higher than 4 to 5 generally describe a relatively hydropho-
and particulate organic carbon (POC). The dissolved and bic DOC, mainly containing aquatic humic substances
particulate fractions are distinguished on the basis of the of high molecular weight and aromaticity. On the con-
porosity of the filters used to separate these two frac- trary, SUVA data less than 3 indicate that the DOC
tions. The typical exclusion limit of the particulate form is hydrophilic, low in molecular weight, and low in
of organic matter that has been adopted ranges from 0.2 charge density and aromaticity (4). Elemental analysis
to 0.45 µm. TOC is generally composed of 90% DOC and determination that defines the percentage of chemical
approximately 10% POC (2). Part of the colloidal carbon is elements such as carbon, hydrogen, nitrogen, oxygen, or
categorized as dissolved, whereas the other part is quan- sulfur within molecules showed that humic substances are
tified as particulate matter. TOC analyzers are common mainly composed of carbon (average of 50%), oxygen (35
instruments used to assess organic matter levels. They are to 40%), hydrogen (5%), and nitrogen (1%) (5).
based on the following principle: organic compounds are
oxidized to carbon dioxide using a combination of oxidizing Identification of Functional Groups, Structural or Mole-
agents (persulfate) and UV light or high temperature. The cular Size (Weight) Characterization. Up to one-third of the
amount of carbon dioxide produced can then be measured organic molecules common in water can be identified using
using nondispersive infrared absorption or conductivity. various technical analyses. Certain organic molecules can
As most waters contain carbonates and bicarbonates, the be characterized by spectrophotometric measurement (13 C
concentrations of these inorganic compounds are either nuclear magnetic resonance) or by a pyrolysis gas chro-
measured or the samples are acidified and purged with an matography/mass spectrometry (GC/MS) technique. The
486 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER

pyrolysis GC/MS method reveals fingerprints indicative of • Inoculation and Incubation of Water Samples.
specific polymers such as polysaccharides, amino-sugars, Inoculum density varies from a few hundred to
proteins, and polyhydroxy-aromatic compounds (6). Amino millions of bacteria, depending on the method.
acids (total and free), carbohydrates (total and free), Incubation periods, which allow test bacteria to grow
aldehydes, and carboxylic acids can also be identified and to degrade organic molecules present in the water
and quantified by specific chemical analyses (7–9). The sample, vary from a few hours to one month.
apparent molecular weight of organic fractions can be • Measurement of the Test Parameter for Biodegra-
assessed either by gel permeation liquid chromatography dation. Bacterial growth or carbon utilization is
(Sephadex) (10) or after ultrafiltration fractionation (1). monitored during the incubation of inoculated water
samples. Increase in bacterial counts can be evalu-
Biodegradable Organic Matter Estimation (BOM) ated using bacteria numeration on agar, microscopic
numeration in epifluorescence, measurement of cell
Organic matter can be divided into two fractions. The
energy such as adenosine triphosphate (ATP), or tur-
first, called BOM, can be utilized by bacteria as a
bidity monitoring. Carbon consumption by bacteria
source of energy and carbon. The second is refractory
is assessed by monitoring DOC concentrations using
to biodegradation (e.g., nonbiodegradable) and has little
a TOC analyzer.
effect on bacterial growth (11). Several biological tests
have been developed to assess the level of BOM in • Calculation of the BOM Concentration. Depending
water (11,12). The specific experimental setting varies on the parameter used to evaluate the BOM, results
from one assay to another; however, they all include the can be expressed using an index of biodegradability
following steps (13). (micrograms acetate carbon equivalents per liter),
cell production (growth yield or maximum growth), or
carbon consumption (expressed in milligrams carbon
• Preparation of Glassware. Carbon contamination
per liter).
must be avoided when performing BOM measure-
ments because water samples generally contain low
There are two established categories of biological tests:
carbon levels. The glassware is rendered carbon free
(1) AOC and (2) BDOC (Fig. 1). AOC refers to the frac-
after heat treatment at 550 ° C for several hours or by
tion of TOC that can be utilized by bacteria for growth.
washing in strong oxidant solution (e.g., sulfochromic
The inoculum is a mixture of pure bacterial strains
acid and persulfate).
cultivated under laboratory conditions (Pseudomonas flu-
• Pretreatment of Water Samples. Depending on the orescens P17, Spirillum NOX) (14–19) and environmental
methodology and sample type, pretreatment of sam- bacteria (20). Bacterial growth is monitored in the water
ples is required to inactivate indigenous bacte- samples using colony counts or ATP measurements. The
ria present in water, remove suspended matter, maximum growth (Nmax ) observed during the incubation is
or neutralize any toxic chemical-like disinfectant converted into AOC using the growth yield of the bacteria
residuals. Inactivation of microorganisms can be per- from calibration curves performed with known concentra-
formed using pseudopasteurization. Water samples tions of standard organic compounds (acetate and oxalate)
are heated at 60 to 70 ° C for 30 minutes to inactivate (Fig. 1). When turbidity is measured for monitoring bacte-
autochthonous bacteria without drastically altering rial growth, two parameters are determined: the growth
the organic matrix. Sample bacteria can also be yield (µ) recorded during the exponential growth phase
retained after filtration using a 0.2-µm pore size and the growth factor (log Y/Yo ) corresponding to the
membrane. Filtration is also performed for raw water increase in bacterial population (21). The concept of AOC
samples containing suspended materials. Filtration has generated many different methods that are summa-
membranes used to filter water samples are made of rized in Table 1. In contrast, BDOC content represents
material such as fiberglass or polycarbonate that do the fraction of DOC that can be assimilated and minerali-
not release organic compounds. Treated waters often zed by heterotrophic microbes. The inoculum consists of
contain disinfectants that could affect the microor- environmental bacteria suspended in water or fixed on a
ganisms used in the bioassay. Disinfectant residuals surface. BDOC is the difference between the initial DOC of
are neutralized using sodium thiosulfate. the water sample and the minimum DOC observed during
• Preparation of a Bacterial Inoculum. There is no the incubation period (23–27) (Table 2, Fig. 1). In gen-
standard inoculum. Inoculum can be composed of eral, AOC molecules correspond to the ‘‘easily’’ assimilable
pure cultures of one or several bacterial strains used organic compounds and generally represent a relatively
at certain concentrations or mixed natural flora. small portion of the BDOC.
Pure culture inocula involve bacterial strains well Microorganisms need various nutrients to grow in
adapted to low nutrient concentration conditions in water. The ratio in which compounds of carbon, nitrogen,
water. Natural inocula are generally composed of and phosphorus are required is 100 : 10 : 1 (C/N/P). Both
many unknown species. In this case, it is difficult AOC and BDOC assays quantify BOM (carbon), which
to assess the number of active cells, the microbial is typically the limiting nutrient for bacterial regrowth
species involved, and their reactivity toward organic in drinking water distribution systems. However, some
molecules. Inoculum may also be in suspension or nutrients other than organic carbon (phosphorus) may
attached to a solid media, such as biofilms attached limit the regrowth of heterotrophic bacteria in the humus-
to sand. rich water of some boreal regions such as Finland or
 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER 487

AOC BDOC
Inoculation:- Pure strains, or Inoculation: Suspended or
- Natural flora fixed natural flora
Bacterial growth monitoring
Doc reduction monitoring
Doc (mg/l)
Bacteria
Nmax Initial
DOC BDOC

Minimum
DOC
BDOC = ini DOC - mini DOC
0 Time 0 Time
Monitored parameter: Monitored parameter:
- Plate counts, or - DOC concentration
- ATP Results: mg C / l Figure 1. AOC and BDOC concepts for
Results: µg acetate C eq. / l biodegradable organic matter determination.

Norway (28,29). In addition, a bioassay has been developed

to determine if carbon or phosphorus is the limiting
Inoculum
nutrient for regrowth (30). 2 ml of river water
filtered through
BDOC Methodologies a 2 µm membrane

BDOC Determination Using Suspended Bacteria. A fil-

tered water sample (0.2-µm pore size filter) is seeded with
suspended bacteria and incubated at 20 ± 2 ° C in the dark
for 28 days (Fig. 2). Bacterial inoculum consists of indige- 200 ml water sample
nous bacteria contained in a small volume of surface water. filtered through a 0.2 µm
membrane
Inoculum water is filtered through a 2-µm porosity poly-
carbonate membrane to remove particles and any protozoa
that could graze bacteria. The inoculation ratio is 1 mL of
surface water/100 mL of water sample (ratio of 1%, v/v).
BDOC can be determined by two procedures. In the first
one, bacterial biomass and mortality rate are monitored,
the integrated flux of mortality (F) during the incubation
period is then calculated and divided by the growth yield Incubation: 28 days at 20°C in the dark
(Y) to give an estimate of BDOC (BDOC = F/Y) (24). This T0 Measure initial DOC
approach, which is very sensitive but also long, is used
when BDOC levels are very low. The second procedure is T28 days Measure minimum DOC
simpler and is more commonly used. The decrease in DOC
concentration is monitored until a plateau is reached and
BDOC is the difference between the initial and final DOC
values (24).
BDOC = DOC initial - DOC minimum
BDOC Determination Using Sand-Attached Bacteria. This Figure 2. Experimental setup to measure BDOC levels with
method is also based on the decrease in DOC concentra- suspended bacteria.
tions, but the water sample is incubated at 20 ± 2 ° C in
the presence of sand colonized by bacteria (23,31) (Fig. 3).
Water samples can be aerated during incubation with analyzed on a daily basis with a TOC analyzer. The incuba-
organic carbon–free air. The inoculum is quartz sand tion is continued until a minimum DOC value is reached.
(with a granulometry of 1 to 2 mm) colonized by bacterial BDOC concentration is defined as the difference between
biomass sampled from a sand filter of a water-treatment the initial and minimum DOC values. The minimum DOC
plant without a prechlorination stage (no oxidant resid- value observed after biodegradation, which corresponds to
ual present) or colonized sand from a freshwater fish the nonbiodegradable DOC, is defined as the refractory
tank. A sand/water sample ratio of 1 : 3 (w/v) is used for dissolved organic carbon (RDOC).
sample inoculation (e.g., 100 g of sand per 300 mL test
water). Using such an abundant fixed flora allows a rapid BDOC Determination Using Bioreactors. The plug flow
response; the test is completed after an incubation period biofilm reactor was developed to measure BOM levels
of a few days. The DOC values in the water sample are within a few hours (26,27). A BDOC reactor consists of
488 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER

Table 1. AOC Methodologies (Adapted from Huck (11))

Sample Inoculum Incubation Parameter Result

Authors (Reference) Preparation (Concentration) Conditions Measured Expression

Vander Kooij and Pasteurization Pure strains 20 days 15 ° C CFU/mL AOC calibration in
coworkers (10,11) P. fluorescens P17 and known solutions of
Spirillum NOX sodium acetate (µg C
(500 CFU/mL) eq. acetate/L)
Kaplan and Pasteurization P. fluorescens P17 and 9 days 20 ° C CFU/mL AOC
coworkers (12) Spirillum NOX (500 to
1,000 CFU/mL)
LeChevallier and Pasteurization P. fluorescens P17 and 5 days 22 ° C ATP AOC
coworkers (14) Spirillum NOX (104
CFU/mL)
Bradford and Filtration P. fluorescens P17 12 hours 20 ° C Cell elongation AOC
coworkers (19) Microscopic
counts
Kemmy and Filtration Mixture of four strains: 6 days 20 ° C CFU/mL AOC calibration in
coworkers (15) P. fluorescens + solution of mixed
Curtobacterium + organic compounds
Corynebacterium + 1 (µg C/L)
species of coryneform
type
Stanfield and Filtration Bacteria from
Jago (16)
Raw water Until max. ATP AOC
Sand-filtered water growth 20 ° C Standard conversion
factor (µg C/L)
Werner and Filtration Water sample bacteria 60 hours 20 ° C Turbidity Growth yield (µ), growth
Hambsch (17) retained on the filter factor (log Y/Yo )
(5 × 104 cell/mL)
Reasoner and Filtration Enterobacter cloacae, 5 days 20 ° C CFU/mL log N5 /N0
Rice (18) Escherichia coli,
Klebsiella oxytoca

Air flow meter

Pipet

Water sample
(300 ml)
Distilled
water

Biological sand
(100 g)
Incubation flask Washing bottles
Figure 3. Experimental setup to measure BDOC Air pump
levels with bacteria attached to sand. Thermostated room 20 ± 2 °C

two glass chromatography columns (2.5 × 60 cm, Kontes, are sampled for at least two hours after the beginning
Vineland, New Jersey) containing porous glass particles of the test. BDOC concentrations are determined from
(Siran; 1- to 2-mm diameter, 60- to 300-µm pore size, changes in DOC concentrations from the reactor inflow
Schott America, Yonkers, New York) (Fig. 4). The reactor and outflow.
column media is heavily colonized with bacteria from
the environment. Bioreactors are continuously fed with
Factors Affecting BOM Results
water at a flow rate of 4 mL/minute. The total hydraulic
retention time of the BDOC reactor is two hours for both BOM levels can be estimated using various bioassays, and
columns. The BDOC test begins by loading the bioreactor each test has specific strengths and weaknesses (11,12).
reservoir with a water sample. Column inflow and outflow Most of the methods have not been standardized, and
 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER 489

less than 0.5 mg/L) (33). A large survey that evaluated

Waste BDOC concentrations in 31 surface water treatment
Outlet
plant effluents showed that the geometric mean of BDOC
concentrations was 0.32 mg/L for all the sites (34). On an
average, the BDOC represented 5 to 21% of the DOC.
When the BDOC levels were tabulated site by site, the
1 2
mean concentrations of plant effluents ranged between
0.03 and 1.03 mg/L, whereas the DOC levels ranged from
0.61 to 4.53 mg/L (34) (Fig. 5).
AOC concentrations typically vary from a few to several
Inlet
hundred micrograms per liter. Figure 5 shows AOC results
of a survey conducted at 95 sites. AOC concentrations
Peristaltic 3-port entering distribution systems were generally high. Fifty
Water pump valve percent of the sites had plant effluent AOC levels
reservoir above 100 µg/L. When data were averaged for each
site, the mean AOC levels ranged from 18 to 214 µg/L,
Figure 4. Experimental setup to measure BDOC levels with a depending on the site (34). These finished water results
biofilm reactor. are comparable to other AOC data collected from 53 source
waters, including surface and groundwaters in North
America (35). Raw water AOC concentrations varied from
many factors such as the inoculum (pure strain versus 18 to 322 µg C-acetate/L or from 2.4 to 44% of the
mixed inocula from the environment), the glassware used, DOC content. Sixty-two percent of the raw water AOC
the monitored parameter (bacterial growth versus DOC), values were above 100 µg/L. Bradford and coworkers (19)
the use of suspended or fixed bacteria, and the experience reported AOC variations from 94 to 275 µg/L (mean
of the laboratory performing the test can influence the of 168 µg/L) in a surface water reservoir and 75 to
final test results (12,32). 731 µg/L (mean of 317 µg/L) in river water (Santa Ana
River, California). The mean AOC concentration for three
Nutrient Levels in Water. Organic matter and nutrient groundwater samples was 54 µg/L. Much lower AOC
concentrations vary dramatically between water sources, concentrations (<10 µg/L) were recorded after bank or
over the seasons, and from one year to another. In slow sand filtration (14).
general, DOC levels are low (<1 mg/L) in groundwater.
Surface waters generally contain 2 to 10 mg/L of DOC (2); Parameters Affecting BDOC Results. BDOC concentra-
however, higher values (20 to 30 mg/L) have been recorded tions vary, depending on the method applied and the
in some waters from Southeastern United States with operating conditions (inoculum size, incubation time,
high-humic content. BDOC concentrations typically range sample aeration) (Table 2). For the BDOC determina-
from less than 0.1 mg/L to 1 to 2 mg/L. For example, tion using bacteria fixed on sand, both inoculum size
a study conducted on various water types showed and aeration have been shown to affect the result of
that river waters had DOC values of 3.2 to 4.0 mg/L BDOC measurement (31,36) (Table 3). A sand/water ratio
and BDOC concentrations above 1.2 mg/L (33). After of 100 g : 300 mL (w/v) allows a rapid decrease in DOC
treatment, BDOC levels ranged between 0.2 and 1.0 mg/L concentration and optimum biodegradation of the organic
(DOC between 0.9 and 2.9 mg/L). Waters with low organic matter. However, levels of BDOC tend to increase with
matter levels had BDOC levels lower than 0.1 mg/L the sand/water ratio, showing sorption of a low fraction
(mineral, spring, and ultrapure waters with DOC levels of DOC (0.1 to 0.2 mg of DOC/100 g of sand). Aeration

Table 2. BDOC Methodologies (Adapted from Huck (11))

Author Sample Incubation Parameter Result

(References) Preparation Inoculation Conditions Measured Expression

Servais and Filtration Suspended 28 days 20 ° C DOC BDOC = DOCi − DOCf

coworkers (24) bacteria from
river water
Joret and Levi (23), Volk – Bacteria fixed on 1 week 20 ° C DOC BDOC =
and coworkers (31) sand DOCi − DOCmini
Mogren and – Bacteria fixed on 5 days 20 ° C DOC BDOC = DOCi − DOC 5
coworkers (25) sand days
Frias and coworkers (26), Filtration Column with 2.5 hours 20 ° C DOC BDOC =
Kaplan and bacteria fixed on DOCinflow − DOCoutflow
coworkers (27) porous glass
particles

Note: DOCi : initial DOC concentration, DOCf : final DOC concentration, DOCmini : minimum DOC concentration.
490 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER

2 5

BDOC DOC
4
BDOC (mg/l) 1.5

DOC (mg/l)
3
1
2

0.5
1

0 0
1 4 7 10 13 16 19 22 25 28 31
250
P17 NOX Total AOC
200
AOC (µg/l)

150

100

50

0
1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91
Figure 5. DOC, BDOC, and AOC concentrations for different plant effluents (from Volk and
LeChevallier (34)). The x-axis represents the site number from 1 to 31 and from 1 to 95 for AOC,
respectively, after ranking.

Table 3. Factors Affecting BDOC Results

Method Parameter Effects on Measured BDOC Concentration

Suspended Inoculum size None

bacteria method
Aeration None
Incubation period BDOC level increased
Disinfectant BDOC level decreased
neutralizer
Sand method Inoculum size BDOC level increased
Incubation period None
Aeration Faster test, possible increase in
BDOC level
Bioreactor method Contact time Possible increase in BDOC level

BDOCsuspended bacteria < BDOCsand = BDOCbioreactor

mixes the water and provides a more uniform exposure to biodegradation may still be incomplete after one month of
the sand, promotes faster kinetics of DOC consumption, incubation. BDOC levels increased by 0 to 125% when the
increases dissolved oxygen concentrations, and sometimes incubation was extended to 85 to 120 days (31). Finally,
leads to a greater consumption of DOC (31). precautions have to be taken regarding the dechlorination
In the method using suspended bacteria, inoculum size agent used to neutralize disinfectants. MacLean and
(from 0.04 to 5%, v/v) and aeration (4 L/hour) have not been coworkers (37) reported that sodium thiosulfate levels
shown to have any effect on BDOC determination (31,36). higher than 20 mg Na2 S2 O3 /L interfered with the 30-day
On the other hand, incubation time affects the result of BDOC test using suspended bacteria. During incubation,
BDOC measurement (Table 3). For certain water types, the presence of Na2 S2 O3 promoted the growth of sulfur
 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER 491

oxidizing bacteria resulting in the production of sulfuric gram of sand versus 106 bacteria per milliliter of river
acid and a depressed pH. The low pH inhibited the activity water). An attached inoculum might be able to degrade a
of heterotrophic bacteria in utilizing organic carbon (37). larger range of organic compounds. Thus, molecules that
The BDOC column method is attractive for use in are refractory to biodegradation by a suspended inoculum
water utilities because by simply measuring the DOC could be biodegradable by a fixed inoculum. This observa-
at the inlet and outlet of the columns, BDOC can be tion could be because of a greater bacterial diversity in the
determined rapidly on a continuous basis. The problem sand biofilm than in a river sample, a higher adaptation
with the bioreactor method, however, is that a long of sand bacteria (fixed species are selected according to
period (between six and nine months) is required for their ability to degrade a large spectrum of organic com-
the columns to be colonized with the variety and pounds), and an advantage caused by the fixation because
density of bacteria necessary to rapidly utilize trace cometabolism and synergy mechanisms can occur on sand
levels of organic carbon. Bioreactors are also highly particle biofilms.
site-specific — changing feed water source may require Both the bioreactor and the sand methods yield similar
an acclimatization period of several months. Columns BDOC results. A study involving 52 samples showed that
that were colonized with water from the Mississippi the differences between the BDOC levels obtained from
River required up to seven months to equilibrate once the bioreactor and sand methods were less than 0.1 mg/L
shipped to new locations with different types of source for 50% of the samples (34). The difference was between
waters (34). Consequently, bioreactors must be colonized 0.1 and 0.2 mg/L for 17% of the samples. In comparison,
and used to perform BDOC measurements with the Joret (39) reported differences less than 0.1 mg/L and
same water matrix. NOM is composed of a mixture between 0.1 and 0.2 mg/L for 49 and 30% of the samples,
of hundreds of simple molecules and complex polymers respectively. Kaplan and coworkers (27) and Prevost and
such as humic substances that are highly specific to a coworkers (40) found that the bioreactor results were
particular watershed. When colonization occurs, organic higher than the BDOC levels determined using suspended
matter leads to the development of unique and specific bacteria. This observation is not surprising, as the
populations of microbial species. This high specificity of BDOC-sand method gives higher results than the BDOC-
bacteria to the site’s organic material was reported to suspended bacterial method (31).
be a limitation of the BDOC-column methodology (27,38).
Bioreactors do not perform well as a stand-alone analyzer Variation Between AOC and BDOC Data. The AOC test
for samples from various origins. During interlaboratory measures a biological response to assimilable carbon, but
experiments of bioreactors colonized with different water unlike BDOC, it is not a direct measure of the carbon level
sources, maximum biodegradation was observed at the itself. Correlations observed between AOC and BDOC
site in which the microflora was specific to tested water can be strong in some cases and weak in others (12).
samples. At the other sites, when bioreactors were Figure 6 shows a comparison of the AOC and BDOC
exposed for a few hours to the new water conditions, values based on a monthly analysis of 31 plant effluent
the reactor biofilm could not metabolize any of the samples (34). Although the AOC and BDOC values were
organic matter. In addition, the bioreactor columns require significantly correlated (p < .01), the variation between
constant attention and maintenance. After colonization, the test results was high (correlation coefficient of only
bioreactors can operate several years, but they require 0.45). When the test results were grouped, a more clear
a continuous flow of low-turbidity water that does not relationship was observed between the two tests (Fig. 6).
contain disinfectants (feed water can be filtered river Volk and coworkers (33) observed an excellent correlation
water, granular activated carbon (GAC)-filtered water, or (r = 0.996) between the AOC and BDOC levels in a series
dechlorinated tap water). Routine checks include pump of dilutions of a particular type of river water. However,
function, absence of blockage, absence of leaks, and the correlation between the tests was not as strong
flow through the reactor. If a feed water reservoir is (r = 0.77) when performed on 31 different water types.
used, the water within the reservoir must be periodically Kaplan and coworkers (35) reported a significant but weak
replenished. correlation (r = 0.59) between the AOC and BDOC data
in a survey involving 109 samples from 79 drinking water
Variation Between the Different BDOC Methods. The supplies. AOC is an index of regrowth potential, which is
estimation of BDOC concentration depends on the applied strongly related to specific groups of organic molecules.
method. BDOC concentrations obtained with a suspended Moreover, the origin and changes in the characteristics
and a fixed/sand inoculum are well correlated. However, of organic matter might affect the relationship between
the BDOC values are generally higher (twice as high) these two methods of BOM measurement. Because NOM
while using attached (i.e., fixed/sand) bacteria (31). Math- characteristics vary from one water type to another, and
ieu (13) also observed that the BDOC values measured seasonally for a given water type, it is not surprising that
using bacteria attached to sand were 1.2 to 1.5 times waters with similar BDOC concentrations might have low
higher than those obtained with suspended bacteria. Sev- or high levels of AOC, depending on the type of organic
eral hypotheses could explain discrepancies between the molecules present in the water constituting BDOC. BDOC
suspended and fixed/sand inoculum. The bacterial con- concentrations are generally much higher than the AOC
centration in freshly inoculated water samples is 100 to levels. This is probably due to a difference in the range
1,000 times higher when using colonized sand than when of metabolic capabilities contained in the AOC test, which
using river water as the inoculum (107 to 108 bacteria per uses just two bacterial stains (P17 and NOX), versus the
492 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER

500

AOC (µg/l)
50

y = 0.24 x + 2.20 (n = 359, r = 0.45)

5
0.01 0.1 1
BDOC (mg/1)

190

170
AOC concentration (µg/l)

150

130

110

90

70

50
0 − 0.2 0.2 − 0.5 0.5 − 0.8 >0.8
Figure 6. Relationship between AOC and BDOC
concentrations (from Volk and LeChevallier (34)). Range of BDOC concentration (mg/l)

BDOC-sand bioassay with a wide variety and a greater Da (43). Finally, the BDOC fraction determined using the
density of microorganisms capable of degrading a large suspended bacteria assay (24) can be further fractionated
spectrum of organic compounds. into three pools of substrates (S, H1, and H2). The H1 frac-
tion represents the macromolecular substrate that can be
Composition of Biodegradable Molecules. AOC measures rapidly hydrolyzed by bacteria exoenzymes, whereas H2
the ‘‘easily’’ assimilable carbon, whereas the BDOC frac- includes polymeric organic molecules slowly hydrolyzed.
tion includes easily and slowly biodegradable compounds. The H1 and H2 fractions are transformed into the S frac-
Because of the complexity of the organic matter, it has not tion, including monomeric substrates directly available
been possible to determine precisely the chemical char- for bacteria. This partitioning is achieved by the use of
acterization of the biodegradable fraction. Biodegradable a biomass flux technique (44) and by the use of the HBS
organic materials have generally been considered to be Model (H: high molecular weight polymers, B: bacteria;
carbohydrates and small molecular weight compounds, S: direct substrates) (45). The polymeric BOM (H1 and
whereas complex molecules such as humic substances in H2 fractions) is hydrolyzed by the bacterial exoenzymes
water were assumed to be essentially nonbiodegradable. that are in concentrations directly proportional to bac-
However, some studies have shown that this classification terial biomass, according to Michaelis-Menten kinetics.
should be used with caution (41–43). It has been found The kinetics includes the maximum rates of exoenzy-
that biodegradable compounds include both low and high matic hydrolysis of H1 or H2 (e1max , e2max ) and the
molecular weight molecules. Some low molecular weight half-saturation constants (KH1 and KH2 ). The exoenzymatic
molecules can be directly utilized for metabolism. Other hydrolysis rates for H1 and H2 are, respectively, equal to
compounds, such as humic substances, which constitute e1max H1/(H1 + KH1 ) and e2max H2/(H2 + KH2 ). The exoenzy-
most of the DOC, can be partially degraded by bacte- matic hydrolysis of macromolecules yields monomeric sub-
ria after enzymatic action. A study showed that 27% strates (S), such as amino acids, which are rapidly utilized
of humic substances in stream water were utilized by by bacteria, according to another Michaelis-Menten kinet-
biofilm bacteria (43). On an average, BDOC was com- ics. For substrate S, the direct substrate uptake is equal to
posed of 75% humic substances, 30% carbohydrates, 4% bmax S/(S + KS )B. The kinetics is characterized by the fol-
amino acids, and 39% of molecules greater than 100,000 lowing parameters: maximum rate of substrate uptake per
 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER 493

unit of bacterial biomass (bmax ) and the half saturation con- They can also be contaminated by synthetic chemicals,
stant of direct substrate uptake (KH ). The numerical values including organic solvents and herbicides/pesticides.
of the different kinetic parameters were determined for a Surface water supplies originate from creeks, streams,
large spectrum of aquatic environment and correspond to rivers, lakes, and reservoirs. Surface waters have the
general physiological characteristics. These processes are characteristic of varying quality after heavy rains or
used to describe BDOC removal during biological filtration over the seasons. Surface waters must be treated
(Model CHABROL) (46), or BDOC consumption in distri- to remove microorganisms and turbidity. In addition,
bution systems (Model SANCHO) (47). These models are they contain organic matter and numerous inorganic
described later in this section. and organic contaminants related to natural or human
While BDOC can include larger molecules, humic activities (urban, industrial, and agricultural pollution).
substances, and some humic bound compounds, it appears The kinds of treatment available to produce drinking
that for most water samples, the AOC is mainly composed water can involve aeration, adsorption, chemical oxidation
of small (<1,000 Da) and nonhumic molecules (34). and disinfection, settling, filtration, membrane filtration,
softening, or ion exchange. Most surface waters are
REMOVAL OF ORGANIC MATTER DURING DRINKING conventionally treated to remove high levels of microbes
WATER TREATMENT and particles. The basic conventional treatment train
involves the following four steps: pretreatment, settling,
Drinking Water Treatment and Water Quality filtration, and posttreatment before distribution (Fig. 7).
Preoxidation is used to control bacteria and algae
Most types of water require treatment to remove certain growth and oxidize iron, manganese, taste, odor and
impurities and pollutants, add chemical substances, color molecules, as well as micropollutants. Primary
or change water characteristics in order to protect oxidants include chlorine, chlorine dioxide, potassium
public health, improve water aesthetics, and comply permanganate, ozone, and hydrogen peroxide. Chlorine
with drinking water regulations. Overall, the quality has been the most extensively used, but some concerns
of water is improved after treatment. Raw water type about the health effects of chlorine disinfection by-products
(ground versus surface water) and the type and levels have increased the use of alternative oxidants. Chemically
of impurities present in the source water determine the enhanced settling allows removal of particulate and
treatment processes that need to be implemented. In colloidal matter that would not precipitate within a
general, groundwaters have a uniform quality and contain reasonable time by gravity alone. After physicochemical
no turbidity (cloudiness), few microorganisms, and little reaction, the nonsettlable solids are converted into
organic matter. However, groundwaters often have high heavier and settlable solids that can be removed by
hardness and might contain various contaminants such sedimentation. Target elements include clay and silt
as iron, manganese, hydrogen sulfide, and radionuclides. particles, microorganisms, inorganic solids, as well as color

KMnO4
CL2, CIO2 CL2 NaOH
O3 NH3 Fluoride
Powdered CIO2
activated carbon
Corrosion inhib.
Lime Coagulant
Coagulant aid Distribution

Pre
Posttreatment/
treatment Coagulation Flocculation
Filtration Storage

Water Sedimentation
supply

Screening

Removal of: Removal of:

Suspended solids Disinfection
Remaining
Colloids suspended matter Corrosion control
Removal of:
Fe, Mn Dissolved compounds Nutrients pH adjustment
Color, odor, taste Organic compounds Reduce tooth cavities
Pesticide, herbicide removal
Industrial solvents

Disinfection
pH adjustment

Figure 7. Conventional water treatment steps.

494 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER

and organic matter. Settling involves three sequential ozone or chlorine forms biodegradable compounds from
steps of coagulation, flocculation, and sedimentation relatively nondegradable DOC.
(Fig. 7). Coagulation consists of adding and rapid mixing
of chemical coagulants into raw water. Alum salts, iron Settling. Coagulation, flocculation, and sedimentation
salts, and synthetic polymers are the most common are critical steps in surface water clarification. Histori-
coagulants used in water treatment. Particles in water cally, most coagulation processes were primarily designed
are usually negatively charged and tend to repel one for particle and turbidity removal. However, some plants,
another. Positively charged coagulants neutralize particle especially those having to treat highly colored water, were
charges and promote coagulation. Destabilized particles designed to remove organic matter. NOM and BDOC levels
aggregate during flocculation. Slow mixing promotes can be reduced by coagulation through colloid destabi-
particle aggregation and leads to the formation of bigger lization, precipitation, coprecipitation, or adsorption onto
flocs. The flocs are allowed to precipitate out of the water floc (48,49). Organic matter concentration is generally
by gravity during sedimentation (no mixing). Filtration reduced after settling. The elimination of organic material
is the final step in water clarification. Suspended matter is impacted by many factors such as the characteristics
remaining after settling is removed when passed through of the NOM, coagulation conditions (coagulant type and
a bed of granular media filter at a certain velocity. The dose, coagulation pH), the nature and concentrations of
most common materials used in filter media are sand, inorganic compounds, and the design and operation of
anthracite, garnet, and GAC. Filter materials can be the sedimentation basin. Consequently, the removal of
used alone (mono-media filter) or in combination (dual organic matter by coagulation varies widely, and is gener-
or multimedia filters). Following settling, a major goal ally between 10 and 90%.
of filtration is to physically remove suspended solids,
Organic Matter Characteristics. In general, higher
remaining precipitates, and floc that did not settle, thus
organic matter removal is observed for samples with high
improving the finished water turbidity. Moreover, if media
DOC values. For example, a study showed that water
include GAC, dissolved organic compounds can be removed
samples with DOC concentrations greater than 4 mg/L
through adsorption and biodegradation. GAC is generally
exhibited DOC and BDOC removals of 55 and 58%, respec-
used to remove micropollutants through adsorption.
tively, versus 31 and 18% for waters with DOC levels below
However, after a few months, the adsorption capacity of
4 mg/L (50). Lind (51) reported NOM removals as high as
GAC is exhausted and it must be regenerated or replaced. 90% for high-TOC water samples (>10 mg/L) from the
An alternative is to use GAC without regeneration, to use southeastern United States. Coagulation mostly removes
the benefit of microbial flora that colonize GAC grains the hydrophobic organic fraction (humic substances) and
in the absence of disinfectants. After filtration, a final large molecules. Hydrophilic compounds are little affected
disinfection (usually chlorine alone or in combination with by coagulation (1,4,52,53). Moreover, the removal of NOM
ammonia to form chloramines) is fed to control bacteria is affected by other physical and chemical properties of
levels in the distribution system. Additional treatment NOM, including the solubility of organic compounds, the
steps can be added to the basic conventional treatment charge density of molecules, or the functional group com-
train to deal with specific pollutants or improve water position (1,4,54–58).
quality. Powdered activated carbon (PAC) can be fed
during pretreatment to remove taste and odor compounds Coagulation pH. The pH of coagulation is also a
as well as micropollutants. A second-stage filtration very influential parameter governing organic removal.
(postfilter adsorber) may be required when waters contain Typically, the best removals of humic substances have
high levels of organic compounds (nutrients, taste and been obtained between pH 4 and 5 with ferric and between
odor, or synthetic organic compounds). pH 5 and 6 with alum (59–61). A series of coagulation tests
performed on a bench scale (jar tests) showed that the pH
Organic Carbon Transformation During Water Treatment of coagulation for optimized organic matter removal with
ferric chloride and alum ranged from 4.4 to 6.7 and from
Treatment of organic carbon is important because these 5.6 to 7.1, respectively. Polyaluminum chloride (PACI),
materials can be responsible for water taste, odor, another type of coagulant, can be used at a higher pH (6.0
and color. Organic matter will affect oxidant demand, to 6.8). Lowering the pH of coagulation with sulfuric
disinfectant by-product levels, micropollutant removal, acid led to additional DOC and BDOC removal (versus
residual disinfectant stability, and bacterial regrowth in coagulation without pH adjustment) of 6 to 17% and
distribution systems. The reduction of organic matter 0 to 13%, respectively (50). In comparison, Chowdhury
concentrations and its biodegradable fraction during water and coworkers (62) reported a TOC removal of 9% under
treatment is one strategy to control bacterial regrowth in plant conditions, but TOC removals increased to 25%
the distribution system. A variety of treatment processes when carbon dioxide was added to lower the pH from
can be used to control BOM levels in drinking water. 8.0 to between 7.0 and 7.3. Lind (51) also found that TOC
The degree of organic carbon removal during drinking removal was improved at lower pH values with alum
water production depends on several parameters such as and ferric chloride, except in cold, low alkalinity, and low
the treatment train design, operational conditions, and TOC water. At lower pH, the dissolved organic matter
the seasonal fluctuations of the source water quality. In precipitates and becomes particulate organic matter.
general, coagulation, biological filtration, or membrane Organic particles are then removed by sweeping or
filtration reduces BOM levels, whereas oxidation with adsorption on floc.
 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER 495

Coagulant Type. In addition to pH, the amount of treatment. When there is no prechlorination stage before
organic material removed is also dependent on the type filtration, a large amount of biomass can accumulate on
of coagulant used. Studies reported in the literature filter media in the form of a biofilm. This biofilm can absorb
are contradictory, suggesting that the performance of and assimilate biodegradable materials contained in the
a particular coagulant is dependent on the specific water. Organic removal during filtration is impacted by
characteristics of the organic matrix and the test the organic matter’s quality and quantity, and will vary
conditions. Several studies have demonstrated that iron- by season, media type, contact time, and backwashing
based coagulants are superior to alum salts (50,52,63). strategy.
Inorganic coagulants (ferric or alum) also seem superior
to synthetic organic polymers for coagulation of organic Temperature. Water temperature can limit bacterial
matter (64). Researchers (57) compared the performances activity within biological filters. Moreover, season affects
of ferric chloride versus alum or PACI for removal of DOC the composition and concentration of the organic matrix
and BDOC in different surface waters. They showed that and its subsequent biodegradation kinetics. Huck and
iron salts typically resulted in greater removal of DOC coworkers (69) suggested that biological contactors should
and BDOC, as high as 28 and 21%, respectively. PACI be operated seasonally, depending on seasonal water
resulted in the lowest removal of DOC and BDOC. In other characteristics and utility criteria. Bacterial density and
experiments performed on high-TOC water (Hillsborough biomass activity have shown differences between winter
River, Tampa, Florida), the average TOC removal was 47% and summer periods (71). The removal of DOC, BDOC,
with alum and 65% with ferric sulfate (65). Alternatively, and trihalomethanes (THMs) was found to be significantly
a field study involving 46 different plants showed that greater during warm temperatures (71–73). Merlet and
TOC removal was best with PACI or alum with sulfuric coworkers (72) showed that DOC removal ranged from
acid (51). The average TOC removal was 39% with low 0 to 15% in winter and 25 to 30% in summer. In a pilot
basicity PACI, 32% with alum, and only 13% with study (73), two filters containing similar bacterial biomass
ferric salts.
were loaded with the same water at 8 and 20 ° C . Bacterial
activity led to a 60% removal of BDOC at 20 ° C and a 27%
BOM Removal. As seen for DOC, reduction in BDOC removal at 8 ° C . Consequently, it was necessary to double
levels varies from one water sample to another, and the contact time to have similar BDOC removal efficiency.
the characteristics of the biodegradable materials govern
their removal during coagulation. Limited studies have
been performed to assess biodegradable DOC removal Filtration Media. Various filtration media can be used
by coagulation. Removal of BDOC ranged from 50 to alone or in combination; however, the medium type
86% using alum at a full-scale treatment plant using is critical. Filtration using biologically active carbon is
alum and no preoxidation (66). Croué and coworkers (67) more efficient than biological sand filters for the removal
reported that removal of BDOC by ferric chloride and of biodegradable organic carbon (74). GAC also showed
alum correlated to removal of DOC, and ranged between better removal of DOC, BDOC, oxalate, aldehyde, and
38 and 88%. When coagulation was optimized or enhanced glyoxalic acid than anthracite (75,76). This difference
for removal of dissolved organic materials for various can be explained by higher densities of fixed bacteria
surface waters, the removal of DOC could be increased (four times as much) on carbon filters. Compared
from 25 to 43% (50). Similarly, removal of BDOC could to microporous activated carbon, macroporous carbon
be improved from 30 to 38% through the application provides the bacteria with a larger number of attachment
of enhanced coagulation. However, when coagulation is sites protected from the abrasive action of backwash. The
optimized, the additional NOM removal preferentially more abundant biomass allows for stable BDOC removal
impacts the refractory (nonbiodegradable) fraction of DOC as well as sustained nitrification, even in cold waters (72).
as opposed to the biodegradable DOC fraction (50). Some First-stage filters that are designed to remove particles can
biodegradable compounds such as proteins are removed also achieve significant BOM removal through biological
very well by precipitation. Hureiki and coworkers (68) activity. However, the accumulation of floc and particles
reported that coagulation removed 34 to 72% of total could act as a diffusion barrier for BOM molecules
amino acids. However, AOC is typically not affected by and affect biomass activity, especially under cold-water
coagulation, probably because AOC target small molecular conditions (76).
weight, nonhumic compounds that are not amenable
to coagulation. AOC removal during settling could be Contact Time. The water-filter media contact time
attributed to phenomena other than physicochemical influences the amount of organic compounds removed.
reactions. Huck and coworkers (69) found AOC removals The higher the contact time, the better the removal.
ranging from 0 to 85% by clarification; however, the DOC removal was doubled (from 7 to 15%) when empty
reduction in AOC levels may have been caused by bed contact time (EBCT) was increased from 5 to
biological activity because no oxidation step was applied 20 minutes (72). Some organic compounds also require
before coagulation, and there were long residence times longer contact times to be removed by biological filtration.
within the sedimentation basin (70). Biologically active filtration is very effective on aldehyde
removal. However, dialdehydes such as glyoxal require a
Filtration. The removal of the biodegradable fraction slower filtration rate for their removal than formaldehyde
of organic matter is one of the main goals of biological and acetaldehyde (77).
496 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER

Backwashing. Effective backwashing is essential for concentration at the inlet and outlet of the filter divided by
long-term successful filter service and for avoiding the the EBCT. The removal rate is proportional to the influent
formation of mudballs or filter cracks. In the case of concentration at the inlet of the filter. The slope of the
biological filtration, the role of backwash is of primary relationship between the removal rate and the BOM level
importance for removal of nonbiological and biological is defined as the average specific removal (ras ) expressed
particles without removing all the fixed bacteria. Different in units of inverse minutes. The relationship also allows
filter backwash methods are available, using water alone one to determine a minimum concentration below which
or in combination with air. While using an upflow wash, the biodegradable materials could not be reduced by
water is introduced into the bottom of the filter, and biological filtration. This concentration was estimated as
the filter medium gradually assumes a fluidized state as 20 to 25 µg/L and 3 to 4 µg/L following first-stage filters
the backwash water flow increases and the bed expands. and GAC contactors, respectively (86). The CHABROL
Particles accumulated are washed out of the filter media. Model is another tool used to predict BDOC removal
Backwash flow is continued with full fluidization until during rapid sand or GAC filtration (46,71,87) (Fig. 8). The
the wastewater is clear. The effect of backwashing on model includes the following processes occurring during
biofilter bacteria varies depending on the type of support biological filtration:
media and the backwash strategy. Backwashing with
chlorinated water may remove a significant portion of the • Interaction Between Bacteria and Organic Matter.
fixed biomass and decrease the performance of the filter for Bacteria can directly assimilate monomeric sub-
removal of organic carbon (78,79). In the case of anthracite, strates (S fraction) and utilize easily and slowly
backwashing with chlorinated water may remove a biodegradable organic compounds (fractions H1 and
significant portion (24%) of the fixed biomass (79). This H2, respectively), after exoenzymatic hydrolysis.
had a short-term effect on the removal efficiency of • Interaction Between Bacteria and the Filter Media.
certain organic molecules such as aldehydes. Similarly, Bacteria can be biologically fixed (population B1) or
Characklis (78) showed that chlorinated backwash of adsorbed to the solid support (population B2) and
pilot sand/anthracite filters decreased TOC removal and suspended in the liquid phase (B3).
microbiological activity in the filter, as evidenced by the • Mortality and Grazing of Bacteria. The model
number of heterotrophic bacteria fixed to the media. As for describes bacterial biomass fate and BOM removal
GAC filters, chlorinated and nonchlorinated backwashes during biofiltration. The relationship between experi-
lead to the same amount of fixed biomass and similar mental results collected in full scale or pilot treatment
TOC removals (80). Servais and coworkers (71) reported plants and data predicted by the deterministic model
that in the case of activated carbon, the loss of biomass are generally satisfactory. The CHABROL Model can
appeared to be negligible, and it was less than 5% in be used to design filters producing biologically sta-
warm water. A combination of water and air can be used ble water.
to improve backwash efficiency. Pilot studies showed that
backwashes with a combination of water-air scour and Ozonation. Ozone has been widely applied at different
with water only yielded similar postbackwash AOC and stages of the treatment trains in Europe and more
TOC removals (81). recently in the United States. The use of ozone in pre-
or posttreatment considerably improves the quality of
Modeling. Several models were developed to predict the drinking water. Ozone is a powerful disinfectant that
performance of biological filters. Biofiltration processes are effectively oxidizes many chemical pollutants, removes
based on biofilm process principles. Biofilm kinetics has color, tastes, and odors, and enhances coagulation (88).
been well studied in the wastewater field. Ritmann (82) However, ozonation of NOM leads to the production of
applied steady state biofilm modeling to interpret field a large number of biodegradable compounds such as
data from drinking water biofilters. The principle of carboxylic acids, keto-acids, and aldehydes. The oxidation
steady state biofilm is that the growth of new biofilm of organic matter by ozone leads to an increase in
caused by substrate utilization is balanced by the biofilm molecular polarity, encourages the formation of low
losses caused by decay, detachment, and predation. The molecular weight compounds at the expense of larger
application of steady state biofilm kinetics to drinking molecules, and reduces the aromaticity (reduction in
water filters requires additional considerations. First, UV absorbance at 254 nm). Ozone reacts with organic
under low nutrient conditions, the presence of soluble contaminants in water via two major chemical pathways.
microbial products might also affect the quality of the filter It reacts directly as molecular ozone (O3 ) through highly
effluent. Moreover, backwash effects must be included. selective and fast reactions. Ozone acts directly to form
Biofilm detachment is high during backwash, whereas it carbonyl groups by acting as a dipole agent on C=C double
is relatively low between backwashes when biofilm growth bonds, as an electrophilic agent on aromatic molecules
may accumulate. Finally, biofilms composed of multiple by ring hydroxylation, or as a nucleophilic agent on
bacterial species are exposed to a wide range of inorganic C=N double bonds. Ozone can also react indirectly as
and organic nutrients (82–85). Huck and coworkers (86) a free radical (hydroxyl radical OH• ) arising from ozone
predicted the removal of biodegradable materials such as decomposition to form carbonyl compounds. The radical
AOC and BDOC, disinfection by-product precursors, and action of ozone is less selective than the direct action. Many
chlorine demand by first-order models. The removal rate compounds are only partially oxidized by ozone or are
of biodegradable substances is defined as the difference in refractory to the action of ozone alone. Several processes
 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER 497

Dissolved organic carbon

Bacteria

B
Non-BDOC BDOC
Filter
influent

H1 Bacteria GAC
Adsorption
H2 production
B
S Uptake
Adsorbed Res- B1/B2
DOC piration

CO2
Figure 8. CHABROL Model and
functioning of GAC filters (adapted
Backwashing from Servais and coworkers (71)).
H1: Rapidly hydrolysable polymeric
B3 BDOC; H2: slowly hydrolysable poly-
Filter
effluent meric BDOC; S: direct substrate;
Non- B: total filter bacteria; B1: biologically
BDOC
BDOC fixed bacteria; B2: adsorbed bacteria;
B3: free bacteria.

have been combined with ozone to increase the oxidation milligram of ozone added (93,94). At higher ozone doses,
level of molecules and to oxidize a larger spectrum of BDOC production is low (<0.05 mg BDOC/mg O3 ). BDOC
organic compounds (88–91). The ozone-hydrogen peroxide levels typically increase from 20 to 90% after ozonation of
and ozone-UV systems provide nonselective degradation various river or sand-filtered waters. Applied ozone dose
by enhancing free radical oxidation. The ozone-hydrogen also shows a greater effectiveness than contact time. For
peroxide combination leads to the degradation of alcohols, equivalent applied ozone dose-contact time, a short con-
aliphatic acids, aldehydes, and some micropollutants such tact time and a high ozone dose lead to higher BDOC
as atrazine and chlorinated solvents that are refractory to formation than a long contact time associated with lower
the action of ozone alone. The oxidation of these compounds ozone dose (93). Figure 9 also compares the influence of
with free radical action may be accompanied by significant ozonation used alone or in combination with hydrogen
TOC reduction when hydrogen peroxide-ozone is applied peroxide or a catalyst (titanium dioxide) on a fulvic acid
at an optimal ratio of 0.35 to 0.45 g/g in waters of neutral solution. Maximum BDOC formation was obtained with
pH. Ozone has also been combined with a catalyst such ozone-hydrogen peroxide and ozone alone. BDOC forma-
as titanium dioxide in an attempt to mineralize a greater tion was 30% lower in the presence of a catalyst than with
fraction of organic matter to CO2 and H2 O (92). ozone because catalytic ozone induced oxidation of ozone
Several authors have noted the increase in biodegrad- by-products into CO2 (95).
ability of model substances or dissolved NOM follow- Ozonation is often combined with biological filtration to
ing ozonation (88,91,93,94). BDOC concentrations of the remove biodegradable molecules produced after oxidation.
ozonated water are a function of the raw water BDOC and When the filter influent water is ozonated, the biological
ozone dosage. After ozone application, BDOC is formed degradability and thus filter biological activity are
within the first minute of treatment. The best ozonation enhanced. Although ozone increases biodegradability by
conditions for producing BDOC are generally those that lowering the RDOC, ozone followed by biological filtration
simultaneously provide maximum UV reduction measured removes an additional DOC and AOC/BDOC fraction,
as changes in absorbance at 254 nm (maximum changes reducing the number of chlorine-reactive sites, which is
in aromaticity) and leave traces of ozone residuals (Fig. 9). proportional to the risk of THM formation (71,96–101).
Ozone treatment increases the biodegradability of organic By-products formed during ozonation, such as carboxylic
molecules in two stages: first, the rate of BDOC formation acids, aldehydes, and peroxides (some of which are
is high at low/intermediate oxidant levels. Then higher health concerns), can be removed through biological
doses have little effect on the biodegradability of organic filtration (88,102–104).
substances because all the molecules likely to be BDOC
precursors have been transformed (93) (Fig. 9). Optimum
BDOC formation is generally obtained with moderate Membrane Filtration. Different membrane technologies
ozone doses of 0.5 to 1.0 mg O3 per mg DOC. Within the are available for the production of potable water, includ-
approximate range of 0 to 1 : 1 ozone to DOC ratio, several ing reverse osmosis (removal of ions and organic mat-
studies reported a 0.15 to 0.20 mg BDOC increase for each ter), nanofiltration (NF) (removal of polyvalent ions
498 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER

8 70

60
Removal of UV
absorbance
6

Ozone consumption (mg/l)

50

UV 254 reduction (%)

40
4
30

20
2 O3 consumption

10

0 0
0 2 3.5 5 6.5
Ozone dose (mg/l)

1.2 1

1 BDOC formation
0.8
DOC removal (mg/l)

0.8

BDOC (mg/l)
0.6
0.6
0.4
0.4 DOC removal

0.2
0.2

0 0
0 2 3.5 5 6.5
Figure 9. Effects of different oxidation treat-
ments on UV absorbance, oxidant residual, Ozone dose (mg/l)
BDOC, and DOC levels for a fulvic acid solu- O3 O3 + H2O2 O3 + cata
tion (O3 + cata: ozone with a titanium dioxide
Treatment
catalyst).

and organic compounds larger than 400 Da), ultrafil- that under high hardness and low pH conditions of feed
tration (UF) (removal of colloids and molecules larger water, NF allowed the passage of the AOC fraction even
than 10,000 Da), and microfiltration (MF) (removal of though DOC and BDOC rejection was near complete. This
particles greater than 0.2 µm, including turbidity, para- also highlights the complementary nature of AOC and
sites, bacteria, and some viruses). MF and UF may not BDOC as parameters.
be suitable for organic matter removal unless pretreat-
ments such as ozone and PAC are applied (105–107).
The addition of PAC increased DOC removal efficiencies Chlorination. Few studies documented the effects of
of ultrafiltration membranes to 40 to 50% (106). Ozona- chlorination on organic matter. It has been reported
tion improved DOC removal by the PAC/UF process up that biodegradability tends to increase after chlorina-
to 80% (108). NF membranes produce high removal effi- tion (96,100,113). Hambsch and coworkers (113) reported
ciencies of DOC (from 50 to 99%), BDOC, precursors of higher biodegradability of humic substances after chlori-
chlorination by-products, pesticides/herbicides, color, and nation.
UV absorbance (109–111). However, main organic con-
stituents in the membrane permeate following NF were
identified as amino acids and sugars that are highly BOM Changes During Conventional Treatment. BOM
biodegradable. These data suggest that not all BOM levels can be altered (increased or decreased) depend-
molecules could be retained during membrane filtration ing on treatment practices. Figure 10 shows the fate of
treatment (111,112). Full-scale and lab data (112) showed organic compounds at different points within a typical
 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER 499

conventional treatment, before and after GAC imple- BACTERIAL REGROWTH IN DISTRIBUTION SYSTEMS
mentation. Removal of biodegradable substrates is typ-
ically not favored during conventional treatment with Biofilm Formation and Bacterial Regrowth
sand/anthracite. While organic matter concentrations can BOM that is not removed during water treatment can
be reduced, those of biodegradable substrates can be result in the growth of biofilm bacteria in the distribu-
increased when using sand or sand/anthracite filters. tion system (Fig. 11). The consequences of this bacterial
Overall, after prechlorination and anthracite/sand filtra- growth can include: degradation of bacterial water qual-
tion, the site recorded a DOC removal of 25% and an ity, the acceleration of corrosion rates, the development
increase of AOC concentration of 50% through the treat- of tastes and odors, customer complaints, and the devel-
ment process (Fig. 10). Removal of DOC occurred during opment of a food chain leading to the presence of inver-
coagulation and settling, but removal was primarily of tebrates (114–117). The degradation of bacterial water
the refractory part (nonbiodegradable) of the DOC. Aver- quality can be observed by growth of heterotrophic bacte-
age raw water AOC level increased from 122 to 174 µg/L ria such as Flavobacterium, Acinetobacter, Pseudomonas,
after preoxidation and settling processes, suggesting that Bacillus, Moraxella, and Staphylococcus (118–120) and
chlorine pretreatment resulted in a large production sometimes by the occurrence of coliforms (Klebsiella,
of AOC. DOC, BDOC, and AOC were not removed by Escherichia coli, Enterobacter, etc.) (121–124). Coliform
mixed media anthracite/sand filtration because a contin- regrowth can cause a violation of drinking water reg-
uous chlorine residue was maintained through the filters. ulations. Suspended bacteria present in the water col-
In contrast, GAC implementation modified organic fate umn come mostly from biofilms formed on the surface
during treatment and BOM was removed by biological of distribution pipes, suspended particles, or sediments
GAC filtration. Overall, the plant recorded a reduction accumulated in storage tanks (124,125). Microbial accu-
in AOC and BDOC levels after implementing GAC fil- mulation in biofilms is governed by the three processes
tration (Fig. 10). AOC was increased by preoxidation; of adhesion, growth, and detachment (125–129) (Fig. 11).
however, AOC levels were reduced to 58% by GAC fil- Bacteria adhering to the pipe surface have several ori-
tration. BDOC concentrations also decreased after GAC gins. They may come from the source water through the
filtration. Biological activity can take place during con- treatment plant (130), regrowth of biofilm cells existing
ventional treatment when chlorine levels are low on top already within the distribution system, and introduction
of the GAC filters. Solutions for conventional treatment during main repair or cross connection. Growth of bacte-
plants with high BOM levels could include the conversion ria fixed on pipes is a combination of cell multiplication
of existing sand and sand/anthracite filters to sand/GAC and mortality. Biofilm levels at steady state in pilot or
filters and possibly delay the preoxidation step until after full-scale distribution systems generally range between
filtration, to limit BOM levels entering the distribution 105 and 107 bacteria/cm2 . The detachment of biofilm
system. bacteria into the water column can be induced by fluid

200
DOC, BDOC, RDOC (mg/l)

3.5 Sand/anthracite filtration

3 150
AOC (µg /l)

2.5
2 100
1.5
1 50

0.5
0
0
Raw Settled Filtered Pl. Effluent
120
3
DOC, BDOC, RDOC (mg/l)

Sand/GAC filtration
100
2.5
80
AOC (µg /l)

2
60
1.5
40
1
20
0.5
0
0
Raw Settled Filtered Pl. Effluent Figure 10. Changes in DOC, BDOC, and AOC concentrations
during water treatment using sand/anthracite or GAC filtra-
RDOC BDOC AOC
tion. DOC = RDOC + BDOC where RDOC: refractory DOC.
500 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER

Higher bacterial level

Output Lower organic matter level
Higher particle level

Influencing factors Biofilm

Detachment
Erosion Sediments
Introduction of bacteria in the network
Releasing
Biodegradable organic matter level
Residual disinfectant
Temperature Mortality
Residence time and flow conditions Growth
Pipe material
Particle presence Distribution
Deposition pipe
Adhesion

Bacteria
Input Organic matter

Figure 11. Biofilm in distribution systems. Particles

shear stress at the biofilm surface or other water quality concentration of 0.16 mg/L or less in finished water in the
changes, such as increased disinfectant levels. It is possi- absence of chlorine residue. Volk and Joret (135) indicated
ble to correlate the bacterial density in the biofilm to the that BDOC levels should be less than 0.15 mg/L at 20 ° C
dynamics of suspended bacteria in the bulk water phase. (warm temperature) and less than 0.30 mg/L at 15 ° C for
Servais and coworkers (131) suggested a ratio of approx- achieving biological stability in the Paris suburb an dis-
imately 60 between biofilm and suspended bacteria for tribution systems. Coliform occurrences were related to
small diameter pipes in several full-scale distribution sys- the consumption of more than 0.10 to 0.15 mg/L of BDOC
tems. Consequently, a biofilm density of 107 bacteria/cm2 within the distribution system (135). Van der Kooij (134)
could correspond to 5 × 104 bacteria/mL. Mathieu and showed that heterotrophic bacterial levels in nonchlori-
coworkers (132) reported that a biofilm of 106 bacteria/cm2 nated systems did not increase when AOC levels were
resulted in suspended bacterial concentrations of 103 to lower than 10 µg/L. LeChevallier and coworkers (133) sug-
105 bacteria/mL in pilot system studies. gested that regrowth of coliform bacteria in chlorinated
systems may be limited by AOC levels less than 50 to
BOM and Bacterial Regrowth 100 µg/L. All the above objectives are difficult to consis-
The level of BOM is an important parameter affecting tently achieve and require very high levels of treatment.
bacterial regrowth because all heterotrophic bacteria use In addition to the amount of nutrients, the composition
organic carbon for the production of new cell material of biodegradable organic materials is also an important
or for utilization as energy (132–134). BOM is gradually factor for controlling microbial growth. Amino acids are
consumed as the water travels through the distribution only a small fraction of the NOM, but they allow a high
system. When both BDOC levels and temperatures regrowth potential (high biomass production per unit of
are high in the water treatment plant effluent, BDOC substrate) and are highly reactive with chlorine (104,136).
consumption within the distribution system can be very Amino acids, carbohydrates, and humic substances were
rapid (36,131). There is generally a trend observed in found to support heterotroph growth. However, coliform
biofilm detachment and organic loading. bacteria were able to persist in water containing amino
The control of bacterial regrowth can be achieved acids or carbohydrates, but not with humic substances
when the amount of BOM entering the distribution sys- alone (104). The presence of disinfectant modifies the affin-
tem is limited. The BOM level required for biostability ity of biofilm bacteria toward various organic substrates.
depends on the amount of disinfectant present. Block and In the absence of any disinfectant, biofilm cells preferen-
coworkers (130) recommended an absence of biodegrad- tially utilize amino acids, then carbohydrates, and finally
able organic material after water treatment to limit humic substances. However, in the presence of chlorine,
bacterial regrowth. Servais and coworkers (131) associa- biofilm affinity switches to amino acids, then humic sub-
ted biological stability (corresponding to no BDOC con- stances, and finally carbohydrates. Humic substances may
sumption within the distribution system) with a BDOC adsorb onto the biofilm and protect the microorganisms
 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER 501

from disinfection. In addition, some of the humic sub- be used because of the improved biostability (low BDOC)
stances can be oxidized and converted into more easily of the water. Another pilot study using annular reactors
degradable compounds after reaction with chlorine. More- to simulate distribution systems evaluated the effect
over, the presence of a disinfectant also affects substrate of reducing nutrient levels after biological filtration on
consumption and the regrowth of heterotrophic bacteria bacterial water quality in drinking water (142). A system
in the system. In the presence of chlorine, biofilms show fed initially by conventionally treated water received
greater carbon removal, higher specific biofilm growth biologically treated water. BOM levels were reduced
rate, but lower biofilm yield, and lower biofilm densi- approximately by half after biological treatment. On an
ties than unchlorinated biofilms (137,138). The authors average, biofilm densities were reduced by 1 log unit
suggest that there is a protective mechanism of bacteria by biological treatment. Interestingly, the effect of the
cells, which may require more substrate in the presence treatment change on bacteria levels was not immediate. It
of a disinfectant. Bacteria may produce larger amounts required approximately six months of biological treatment
of extracellular polymeric substances (EPS) to react with before there was an observable impact on bacterial water
and neutralize chlorine (explaining the higher substrate quality.
uptake and carbon requirement). Acting as a sacrificial Field experience shows that distribution systems rep-
barrier, EPS are critical in limiting chlorine penetration resent complex reactors and that a variety of parameters
into the biofilm (138). influence biofilm growth. Factors such as water temper-
Biofilm density and bacterial water quality are related ature, disinfectant type and residual, selection of the
to the amount of biodegradable material entering the sys- pipe material, corrosion control, and hydraulic conditions
tem. Servais and coworkers (131) observed a relationship may be more influential than the levels of organic mat-
between biofilm bacteria and the concentrations of BDOC ter for regulating the biological activity of the biofilm.
at the point of entry of several full-scale distribution sys- These (and other) parameters probably influence microbial
tems. Another study compared bacterial water quality in water quality data as much as the amounts of nutri-
two pilot distribution systems supplied with different lev- ents variable. Pipe characteristics strongly influence the
els of organic matter (139). The first system was fed with density of fixed biomass. Iron pipe surfaces have been
ozonated water, whereas the second was supplied with bio- shown to stimulate bacterial growth (117,122,143,144).
logically filtered water. Lowering the nutrient levels with Camper (143) found that more heterotrophic bacteria grew
biological filtration led to lower biofilm densities. Biofilm on mild steel surfaces than on polycarbonate. Although
counts were 106 to 107 CFU/cm2 for the ozonated system, growth rates were higher in the polycarbonate reactors,
compared to 105 to 106 CFU/cm2 for the system fed with mild steel surfaces contained 10 times more heterotrophs
biologically filtered water. and 2- to 10-fold more coliform bacteria. Biofilm levels
It is critical to produce low nutrient level waters. are also governed by the pipe conditions. Corrosion, pit-
Several studies evaluated the effects of reducing nutrient ting, and tuberculation are fundamental to the presence
levels on bacterial water quality after a treatment of biofilms, their metabolic activity, and the release of
change. In two studies (140,141), bacterial water quality bacteria into the water column (144). Pipe conditions also
changes were monitored after implementation of NF indirectly impact biofilm bacteria by affecting disinfec-
in a pilot or full-scale system that was initially tion efficiency, especially with free chlorine. There is a
supplied with surface water treated with ozonation and relationship between the corrosion of iron surface and
biological filtration. In the pilot study (140), BDOC levels the protection of biofilm bacteria from chlorine disinfec-
decreased from 0.25 mg/L in the ozonated/filtered water tion (117,144). Rompre and coworkers (139) reported that
to less than 0.1 mg/L after NF. The authors did not free chlorine produced a rapid decrease in biofilm den-
observe drastic changes in the biofilm or suspended sity (>1 log CFU/cm2 ) in polycarbonate annular reactors,
bacteria levels after the treatment change. At the whereas the chlorine did not affect the biofilm in gray-
beginning of the study, total and culturable counts were iron reactors. Moreover, organic matter tends to adsorb to
4.9 × 106 cells/cm2 and 4 × 105 CFU/cm2 for the biofilm corrosion products on iron pipe surfaces and the elevated
and 2.6 × 105 cells/mL and 103 CFU/mL in bulk water, concentration of organic molecules can stimulate bacterial
respectively. Bacterial concentrations slightly decreased regrowth (137,138). A large survey conducted to under-
after 6 weeks of supplying nanofiltered water (suspended stand coliform regrowth showed that coliform occurrence
bacteria of 1.4 × 105 cells/mL and 259 CFU/mL; biofilm rates were higher in systems with a large proportion of
of 2.3 × 106 cells/cm2 or 1.2 × 105 CFU/cm2 ). After one unlined cast iron pipes subject to corrosion, suggesting
year of supplying nanofiltered water, biofilm levels that corrosion control is an important factor for limiting
were 1.9 × 105 CFU/cm2 and suspended bacterial levels coliform regrowth (145). The use of phosphate-based cor-
were 50 CFU/mL. Similarly, bacterial water quality rosion inhibitors was associated with lower coliform levels
changes were not obvious when the same treatment in several systems. Coliform regrowth can also be related
conversion was performed in a full-scale distribution to several engineering and operational factors, including
system supplying a suburban community (141) in Paris, water filtration, the use of ozone without biological treat-
France. Biofilm densities were low before the treatment ment, a large proportion of storage tanks, the presence
change because of the high chlorine levels (0.85 mg/L) of uncovered finished water reservoirs, corrosion control,
used to combat microbial growth caused by the high BDOC and distribution system flushing (145). Removal of BOM
levels (0.75 mg/L). Following the application of NF, biofilm through effective treatment and filtration is critical for con-
levels remained low, but a low disinfectant residual could trolling bacterial water quality. Coliforms occurred more
502 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER

frequently in systems using unfiltered surface waters or include chlorine decay, trihalomethane propagation, or
using ozone without biological filtration. Finally, the con- fluoride tracer analysis. PICCOLO software is another
figuration and management of the distribution system example of a residence time model that integrates the
also influenced water quality. Increased levels of coliforms prediction of free chlorine decay within the Paris suburbs
were observed in systems having a large proportion of network (149).
storage tanks. Storage tanks are necessary to meet water Biofilm development on a surface is the result of
demand, and provide adequate pressure and fire pro- several chemical, physical, and microbial processes,
tection. However, the disinfectant can dissipate quickly including an initial colonization phase of the pipe
and microbial water quality can deteriorate when water surface (transport and adsorption of organic material
is stagnant in storage tanks. Systems with open fini- and bacteria from the bulk water to the surface and
shed water reservoirs are prone to animal contamination adhesion of bacteria), the growth of bacterial cells
and algal bloom, and more likely to experience coliform leading to a multilayer biofilm embedded in a polymer
regrowth (145). Annual flushing of the distribution is a matrix, and the detachment of bacteria. Although
maintenance practice performed to remove accumulated fundamental aspects of these biofilm processes have been
sediments, limit red water complaints, and reduce coliform investigated (150), models for predicting bacterial water
occurrences. quality are limited. The Biofilm Accumulation Model
(BAM) models the processes of nutrient (AOC) transport
Distribution System Modeling and consumption, biofilm development, and subsequent
Several models have been developed to predict water growth and decay of bacterial cells, detachment, and
behavior and degradation of bacterial water quality transport of microorganisms (148). The validation of
during distribution. Although different factors may be the BAM has been performed in a pipe loop pilot
listed and examined individually, in reality, all these system.
factors are interrelated, sometimes with competing The SANCHO Model was adapted from biofilm pro-
effects. For instance, increased water temperature favors cesses observed during biological filtration (151). It is
microbial growth, but higher temperatures also improve based on the relationship between bacteria, disinfectant,
disinfection efficiency (decreasing bacterial regrowth). and BDOC levels in distribution systems (Fig. 12). The
Warm temperatures lead to higher reaction rates that may components of the model include the attachment of bac-
result in more rapid dissipation of disinfectant residuals teria onto the surface pipe (reversible adsorption followed
and stimulate pipe corrosion. Higher disinfectant doses by biological attachment), the interaction between bacteria
can reduce bacterial growth but increase nutrient levels and organic matter (direct assimilation or exoenzymatic
and stimulate iron corrosion. Because of the complexity of hydrolysis of organic material, growth of fixed and bulk
the distribution networks and the variety of confounding bacteria, and bacterial mortality), and free chlorine behav-
effects, the existing models are overly simplified or apply ior (chemical demand, disinfection of bacteria). The model
only to a single site (146). Several models are available successfully evaluates the spatial fluctuations of BDOC,
for predicting hydraulic flows and retention times (147). chlorine, and suspended and fixed bacteria concentrations
These models can incorporate a reactive component such in a water mass circulating in a series of pipes with
as chlorine decay or bacterial growth rates. The EPANET decreasing diameters. The model can be used to determine
Model simulates hydraulic and water quality behavior treatment goals to prevent the degradation of microbial
in distribution systems (148). The water quality modules water quality in any network (47,151,152).

Input
Internal processes Output
(H1, H2, B3)

BDOC
H2 CO2
Cl2 Cl2
Free
bact.
H1 S B3 Mortality
E

Cl2
Figure 12. SANCHO Model (from Servais
and coworkers (151)) H1: Rapidly hydrolyz- B2
able polymeric BDOC; H2: slowly hydrolyz- Mortality B1
able polymeric BDOC; S: direct substrate;
B: total filter bacteria; B1: biologically fixed
bacteria; B2: adsorbed bacteria; B3: free bac- Fixed bacteria
teria.
 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER 503

Because coliform occurrences are not related to a temperatures at or above 15 ° C, BDOC consumption
single variable, it is necessary to simultaneously examine greater than 0.15 mg/L, a chlorine residual lower than
the interaction of multiple parameters to predict the 0.10 mg/L, and a logarithm of bacterial concentration
propensity for coliform regrowth events. Tree-based higher than 5.2. The decision tree presented in Figure 13
statistical models have been used to define predictor graphically depicts the combinations of the threshold
variables for coliform regrowth (146). A tree-based model criteria. The sum of the positive criteria (number of
is a descriptive summary of how a set of predictor threshold variables simultaneously exceeded) is indicated
variables relate to an outcome variable (e.g., presence for each combination. Four positive criteria recorded
or absence of coliforms). Coliforms can be predicted using for a distribution network lead to the most favorable
a large set of variables, including AOC level, turbidity, situation for coliform occurrence (high temperature, low
temperature, disinfectant type, residual, pH, phosphates, disinfectant residual, high nutrient consumption, and
TOC, and alkalinity. Although it may be possible to elevated bacterial concentration). Alternatively, a ranking
develop extremely sophisticated models using a large of zero combines all the conditions for coliform control
number of variables to predict coliform occurrences, such (low temperature, high chlorine residual, low nutrient
models are too complicated to be applied to most water consumption, and low bacteria levels). The probability
utilities. Volk and Joret (135) developed the ALCOL Model of experiencing coliform bacteria when various threshold
to evaluate the risk of occurrences of total coliform bacteria criteria are exceeded is shown in Table 4. The majority of
in full-scale distribution systems in the suburbs of Paris, positive coliform samples occurred when the four criteria
France, on the basis of temperature, BDOC consumption, were exceeded. The probability of coliform occurrence was
disinfectant residual, and suspended bacteria counts 64% when the four criteria were positive, but nil when
(microscopic counts by epifluorescence) in the distribution only one or none of the thresholds was exceeded. Such
network. The analysis of coliform occurrence data has a model for the prediction of coliform episodes may be a
shown no simple linear relationship between coliform- useful tool to control bacterial regrowth. When a system
positive samples and these four water quality parameters. is considered to be at high risk, various actions have to
However, it is possible to determine a threshold value be undertaken to control bacterial water quality. Punctual
above which coliforms occur more commonly. The actions such as increased flushing, boosting disinfection
model uses the following four critical thresholds: water residuals, and decreased storage tank residence time can

> 5.2 4 T
Low Bacteria h
3 r
< 5.2
>0.15 mg/l Disinfectant e
residual > 5.2 3 s
Bacteria h
High 2 o
< 5.2 l
>15 °C BDOC > 5.2 3 d
Low Bacteria
consumption 2
Disinfectant < 5.2 V
residual > 5.2 2 a
<0.15 mg/l Bacteria r
High 1
< 5.2 i
T °C a
> 5.2 3 b
Low Bacteria l
>0.15 mg/l 2 e
Disinfectant < 5.2
> 5.2 s
residual 2
Bacteria
High 1 E
<15 °C < 5.2 x
BDOC > 5.2 2
Low c
consumption Bacteria e
Disinfectant 1
< 5.2 e
residual > 5.2 d
<0.15 mg/l 1
Bacteria e
High 0 d Figure 13. Decision tree for predicting coliform
< 5.2 regrowth (ALCOL Model, 135).

Table 4. Risk of Coliform Occurrences in a Distribution System

Frequency (%)
Number of Total Number Number of of Coliform
Risk Positive Criteria of Events Coliform Episodes Observations

Exposed 4 14 9 64
Low risk 2 or 3 87 3 3
No risk 0 or 1 22 0 0
504 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER

be effective in certain areas of the distribution system. 4. J. K. Edzwald, Proceedings of the American Water Works
Treatment modifications such as increasing plant effluent Association, Water Quality and Technology Conference, San
disinfectant residuals or improving BOM elimination Francisco, Calif., November 6–10, 1994.
can also be implemented at the production facility. 5. M. Dore, Chimie des Oxydants et Traitement des Eaux,
However, the current model is site-specific and situations Lavoisier, Paris, France, 1989.
in which coliform regrowth does not occur while the model 6. A. Bruchet, C. Rousseau, and J. Mallevialle, J. Am. Water
predicted coliforms are possible. Further refinement of the Works Assoc. 82(9), 66–76 (1990).
models may make these tools more accurate and reliable. 7. P. Lindroth and K. Mopper, Anal. Chem. 51, 1667–1674
Individual models for problem distribution systems may (1979).
be warranted and could integrate hydraulic models, 8. C. M. Burney and J. McN. Sieburth, Mar. Chem. 5, 15–28
pipeline configurations, and operational characteristics of (1977).
the network. Additionally, the weight of each parameter 9. W. H. Glaze, M. Koga, and D. Cancillia, Environ. Sci.
can be defined to refine the risk calculation. Technol. 23(7), 838–840 (1989).
10. F. H. Frimmel and S. Hesse, Proceedings of Workshop on
NOM in Drinking Water, International Water Supplier
CONCLUSION Association, Poitiers, France, September 18–19, 1996.
11. P. M. Huck, J. Am. Water Works Assoc. 82, 78–86 (1990).
Because a high BOM level is one factor related to bacte- 12. J. C. Joret and M. Prevost, Biodegradable Organic Matter
rial regrowth, BOM removal should be emphasized during in Drinking Water, in preparation, 2001.
water treatment to improve water quality. Several bioas- 13. L. Mathieu, Ph.D. Thesis in ‘‘Chimie et Microbiologie de
says are available to determine the levels of biodegradable l’Eau,’’ University of Nancy I, Paris, France, 1992.
organic materials in water. Measured BOM concentra- 14. D. Van der Kooij, A. Visser, and W. A. M. Hijnen, J. Am.
tions are often related to the applied methodology and the Water Works Assoc. 74, 540–545 (1982).
laboratory conducting the test. Each method has its own 15. D. Van der Kooij and W. A. M. Hijnen, Appl. Environ.
advantages and disadvantages. There is a weak correla- Microbiol. 47, 551–559 (1984).
tion between AOC and BDOC concentrations. The BDOC 16. F. A. Kemmy, J. C. Fry, and R. A. Breach, Water Sci.
test attempts to identify the entire pool of BOM, includ- Technol. 21, 155–159 (1989).
ing both very labile constituents and slowly biodegradable 17. L. A. Kaplan, T. L. Bott, and D. J. Reasoner, Appl. Environ.
molecules (requiring the successive steps of cell adsorp- Microbiol. 59, 1532–1539 (1993).
tion, exoenzymatic hydrolysis, and cell consumption), 18. M. W. LeChevallier, N. E. Shaw, L. A. Kaplan, and T. L.
whereas the AOC test detects only easily assimilable Bott, Appl. Environ. Microbiol. 59, 1526–1531 (1993).
organic compounds. Because both tests emphasize dif- 19. S. M. Bradford, C. J. Palmer, and B. H. Olson, Water Res.
ferent fractions of BOM, both parameters should be 28, 427–435 (1994).
monitored when studying nutrient changes during water 20. G. Stanfield and P. H. Jago, The Development and Use of
treatment and distribution. Bacterial regrowth and col- a Method for Measuring the Concentration of Assimilable
iform occurrences increase with temperature, the absence Organic Carbon in Water, WRC Environment, PRU 1628-M,
Manheim, U.K., 1987.
of disinfectant, and elevated levels of BOM. Beyond these
three major factors, coliform regrowth is also dependent 21. P. Werner and B. Hambsch, Water Supply 4, 227–232
(1986).
on other physicochemical, engineering, and operational
parameters such as pipe material, hydraulic shear, and 22. D. J. Reasoner and E. W. Rice, Proceedings of the Workshop
on Measurement of AOC in the Field of Drinking Water
corrosion. Therefore, it is difficult to predict the effects of
Treatment, Karlsruhe, Germany, November, 1989.
a treatment change at a specific site, depending on the
23. J. C. Joret and Y. Levi, Tribune Cebedeau 39, 3–9 (1986).
weight of the different factors regulating bacterial growth.
As a first step, a water utility can better limit biofilm prob- 24. P. Servais, G. Billen, and M. C. Hascoet, Water Res. 21,
445–450 (1987).
lems when simultaneously addressing the following three
issues: nutrient levels, corrosion, and disinfection. High 25. E. M. Mogren, P. Scarpino, and R. S. Summers, Proceedings
of the Annual American Water Works Association Conference,
organic matter removal during water treatment and effec-
Cincinnati, Ohio., June 17–21, 1989.
tive corrosion control during distribution would improve
26. J. Frias, F. Ribas, and F. Lucena, Water Res. 26, 255–258
disinfection and limit bacterial regrowth within the distri-
(1992).
bution system.
27. L. A. Kaplan et al., An Immobilized Biofilm Reactor for the
Measurement of Biodegradable Organic Matter in Drinking
BIBLIOGRAPHY Water: Validation and Test Field, American Water Works
Association Research Foundation, Denver, Colo., 1996.
1. G. L. Amy, Proceedings of the Workshop on NOM in Drink- 28. M. J. Lehtola, I. T. Miettinen, T. Vartiainen, and P. Marti-
ing Water, American Water Works Association Research kainen, J. Appl. Environ. Microbiol. 65, 2032–2034 (1999).
Foundation, eds., Chamonix, France, 1993. 29. C. Charnock and O. Kjonno, Water Res. 34, 2629–2642
2. E. M. Thurman, in M. Nijhoff, ed., Development in Biochem- (2000).
istry: Organic Geochemistry of Natural Water, Dr. W. Junk 30. A. Sathasivan, S. Ohgaki, K. Yamamoto, and N. Kamiko,
Publishers, Dordreicht, Germany, 1985. Water Sci. Technol. 35(8), 37–44 (1997).
3. E. M. Thurman and R. L. Malcolm, Environ. Sci. Technol. 31. C. Volk, C. Renner, C. Robert, and J. C. Joret, Environ.
45, 463–466 (1981). Technol. 15, 545–556 (1994).
 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER 505

32. J. C. Block et al., Water Res. 26, 481–486 (1992). 58. G. W. Harrington and P. H. Singer, Proceedings of the
33. C. Volk, C. Renner, and J. C. Joret, Sci. Eau. 5, 189–205 NOM Workshop, International Water Supplier Association,
(1992). Poitiers, France, September 18–19, 1996.
34. C. J. Volk and M. LeChevallier, J. Am. Water Works Assoc. 59. P. A. Chadik and G. L. Amy, J. Am. Water Works Assoc.
75(10), 532–546 (1983).
92(5), 64–76 (2000).
60. M. Jekel, Water Supply 3, 61–66 (1985).
35. L. A. Kaplan, D. J. Reasoner, E. W. Rice, and T. L. Bott, Sci.
Eau. 5, 207–225 (1992). 61. S. W. Krasner and G. L. Amy, J. Am. Water Works Assoc.
87(10), 93–107 (1995).
36. C. Volk, Controle de la Matiere Organique Biodegradable au
62. Z. K. Chowdhury, S. P. Papadimas, and E. B. Olivieri, Pro-
Cours des Traitements d’ Oxydation et de la Distribution
ceedings of the American Water Works Association Water
des Eaux d’ Alimentation, Ph.D. Thesis in ‘‘Sciences et
Quality Technology Conference, New Orleans, La., Novem-
Techniques de l’ Environnement,’’ Ecole Nationale des Ponts
ber 12–16, 1995.
et Chaussees, Paris, France, 1994.
63. F. Julien, B. Gueroux, and M. Mazet, Water Res. 28,
37. R. G. MacLean et al., Water Res. 30, 1858–1864 (1996). 2567–2574 (1994).
38. L. A. Kaplan and J. D. Newbold, Water Res. 29, 2696–2706 64. G. Crozes, P. White, and M. Marshall, J. Am. Water Works
(1995). Assoc. 87(1), 78–89 (1995).
39. J. C. Joret, Validation d’un Reacteur en Continu de Deter- 65. J. M. Gianatasio, M. A. Hook, and L. J. Hjersted, Proceed-
mination Rapide du CODB des eaux Potables , Anjou ings of the American Water Works Association Water Quality
Recherche, Maisons-Laffitte, France, 1994. Technology Conference, New Orleans, La., November 12–16,
40. M. Prevost, G. Dubreuil, R. Desjardin, and R. G. MacLean, 1995.
Proceedings of the American Water Works Association Water 66. M. Prevost, Etude de l’ Impact de la Filtration Biologique
Quality Technology Conference, Denver, Colo., November sur la Qualite de l’ Eau Potable, Ph.D. Thesis in Civil
9–12, 1997. Engineering, Ecole Polytechnique, Montreal, Canada, 1991.
41. L. A. Kaplan and T. L. Bott, Freshwater Biol. 13, 363–377 67. J. P. Croué et al., Sci. Eau. 8, 463–479 (1995).
(1983). 68. L. Hureiki, C. Gauthier, and M. Prevost, Sci. Eau. 9,
42. R. M. W. Amon and R. Benner, Limnol. Oceanogr. 41, 41–51 297–318 (1996).
(1996). 69. P. M. Huck, P. M. Fedorack, and W. B. Anderson, J. Am.
43. C. J. Volk, C. B. Volk, and L. A. Kaplan, Limnol. Oceanogr. Water Works Assoc. 83, 69–90 (1991).
42, 39–44 (1997). 70. J. Easton and P. Jago, Proceedings of the American Water
Works Association Water Quality Technology Conference,
44. P. Servais, P. Laurent, and D. Gatel, Proceedings of the
Miami, Fla., November 7–11, 1993.
American Water Works Association Water Quality Technol-
ogy Conference, New Orleans, La., November 12–16, 1995. 71. P. Servais, G. Billen, C. Ventresque, and G. Bablon, J. Am.
Water Works Assoc. 83, 62–68 (1991).
45. P. Servais, Ph.D Thesis, Universite Libre de Bruxelles,
72. N. Merlet, M. Prevost, Y. Merlet, and J. Coallier, Sci. Eau.
Belgium 1986.
5, 143–164 (1992).
46. G. Billen, P. Servais, P Bouillot, and C. Ventresque, Aqua
73. P. Servais, G. Billen, P. Bouillot, and M. Benezet, Aqua 41,
41, 231–241 (1992).
163–168 (1992).
47. P. Laurent, P. Servais, M. Prevost, and D. Gatel, J. Am.
74. G. P. Bablon, C. Ventresque, and R. Benaim, J. Am. Water
Water Works Assoc. 89(7), 92–103 (1997). Works Assoc. 80, 47–53 (1988).
48. S. J. Randtke, Proceedings of the Workshop on NOM 75. D. A. Reckhow et al., Proceedings of the Annual American
in Drinking Water, American Water Works Association Water Works Association Conference, Vancouver, B.C.,
Research Foundation, eds., Chamonix, France, September Canada, June 18–22, 1992.
19–22, 1993. 76. P. Niquette et al., J. Am. Water Works Assoc. 88(1), 86–97
49. J. G. Jacangelo, J. DeMarco, D. M. Owen, and S. J. Randtke, (1998).
J. Am. Water Works Assoc. 87, 64–77 (1995). 77. H. S. Weinberg, W. H. Glaze, S. W. Krasner, and M. J. Scli-
50. C. J. Volk et al., Water Res. 34, 3247–3257 (2000). menty, J. Am. Water Works Assoc. 85, 72–85 (1993).
51. C. B. Lind, Proceedings of the American Water Works 78. W. G. Characklis, Bacterial Regrowth in Distribution Sys-
Association Annual Conference, Anaheim, Calif., June tems, American Water Works Association Research Founda-
17–21, 1995. tion, Denver, Colo., 1988.
52. S. J. Randtke, J. Am. Water Works Assoc. 80(5), 40–56 79. R. J. Miltner et al., Proceedings of the American Water Works
(1988). Association Water Quality Technology Conference, Toronto,
Canada, November 15–19, 1992.
53. M. E. Tryby, R. J. Miltner, and R. Scott Summers, Proceed-
ings of the American Water Works Association Water Quality 80. F. A. DiGiano et al., Microbial Activity on Filter Adsorbers,
Technology Conference, Miami, Fla., November 7–11, 1993. American Water Works Association Research Foundation,
Denver, Colo., 1992.
54. R. L. Sinsabaugh, R. C. Hochn, W. R. Knocke, and A. E. Lin-
81. R. Ahmad, A. Amirtharajah, A. Al-Shawwa, and P. M. Huck,
kins, J. Am. Water Works Assoc. 78(5), 74–82 (1986).
Proceedings of the American Water Works Association
55. E. Lefebvre and B. Legube, Water Res. 27, 433–447 (1993). Water Quality Technology Conference, San Francisco, Calif.,
56. W. J. Huang and H. H. Yeh, Proceedings of the American November 6–10, 1994.
Water Works Association Water Quality Technology Confer- 82. B. E. Rittmann, J. Am. Water Works Assoc. 82(12), 62–66
ence, Miami, Fla., November 7–11, 1993. (1990).
57. M. Edwards, M. M. Benjamin, and J. E. Tobiason, J. Am. 83. B. E. Rittmann and J. A. Manem, Biotechnol. Bioeng. 39,
Water Works Assoc. 86(1), 105–116 (1994). 914–922 (1992).
506 BIODEGRADABLE DISSOLVED ORGANIC CARBON IN DRINKING WATER

84. P. B. Sáez and B. E. Rittmann, Biotechnol. Bioeng. 39, 110. L. Tan and G. L. Amy, J. Am. Water Works Assoc. 83, 74–78
790–793 (1992). (1991).
85. E. Namkung and B. E. Rittmann, Biotechnol. Bioeng. 29, 111. K. M. Agbekodo, B. Legube, and P. Cote, J. Am. Water Works
335–342 (1987). Assoc. 88, 67–74 (1996).
86. P. H. Huck, S. Zhang, and M. L. Price, J. Am. Water Works 112. I. Escobar and A. Randall, J. Am. Water Works Assoc. 91,
Assoc. 86(6), 62–66 (1994). 76–87 (1999).
87. P. Laurent et al., Water Res. 33, 1387–1398 (1999). 113. B. Hambsch, P. Werner, and F. H. Frimmel, Acta Hydro-
88. B. Langlais, D. A. Reckhow, and D. R. Brink, Ozone in Water chem. Hydrobiol. 21, 167–173 (1993).
Treatment: Application and Engineering, Lewis Publishers, 114. M. J. Allen, R. H. Taylor, and E. E. Geldreich, J. Am. Water
Chelsea, Mich., 1991. Works Assoc. 72, 614–625 (1980).
89. H. Paillard, M. Dore, and M. M. Bourbigot, Proceedings 115. R. V. Levy, F. L. Hart, and R. D. Cheetham, J. Am. Water
of the 10th Ozone World Congress, International Ozone Works Assoc. 78, 105–110 (1986).
Association, Monaco, France, March 19–21, 1991. 116. J. Coallier, P. Lafrance, D. Duchesne, and J. Lavoie, Sci.
90. J. Hoigne and H. Bader, Water Res. 10, 377–386 (1976). Tech. Eau. 22, 63–72 (1989).
91. C. Volk et al., Ozone Sci. Eng. 15, 405–418 (1993). 117. M. W. LeChevallier, C. D. Lowry, R. G. Lee, and D. L. Gib-
bon, J. Am. Water Works Assoc. 85, 111–123 (1993).
92. H. Paillard, F. Carbonnier, and P. Roche, Proceedings of
118. M. M. Bourbigot, A. Dodin, and R. Leritier, Water Res. 18,
the American Chemical Society Conference Emerging
585–591 (1984).
Technologies for Hazardous Waste Management, Atlanta,
Ga., October 1–3, 1991. 119. M. W. LeChevallier, T. M. Babcock, and R. G. Lee, Appl.
Environ. Microbiol. 53, 2714–2724 (1987).
93. C. Volk et al., Ozone Sci. Eng. 15, 389–404 (1993).
120. P. Payment, F. Gamache, and G. Paquette, Can. J. Micro-
94. R. D. Paode et al., J. Am. Water Works Assoc. 89(6), 79–93
biol. 34, 1304–1309 (1988).
(1997).
121. M. A. Goshko, W. O. Pipes, and R. R. Christian, J. Am.
95. C. Volk, P. Roche, J. C. Joret, and H. Paillard, Water Res.
Water Works Assoc. 75, 371–374 (1983).
31, 650–656 (1997).
122. M. W. LeChevallier, J. Am. Water Works Assoc. 82, 74–86
96. D. Van der Kooij, Treatment of drinking water for organic (1990).
contaminants, in P. M. Huck and P. Toft, eds., Proceedings of
123. A. K. Camper, G. A. McFeters, W. G. Characklis, and W. L.
the 2nd National Conference on Drinking Water, Pergamon
Jones, Appl. Environ. Microbiol. 57, 2233–2239 (1991).
Press, Edmonton, Canada, N.Y., 1987.
124. E. E. Geildreich, Microbial Quality of Water Supply in
97. C. Ventresque, G. Bablon, and A. Jadas-Hecart, Ozone Sci.
Distribution Systems, CRC Press, Boca Raton, Fla.,
Eng. 12, 91–107 (1990).
1996.
98. M. Prevost, R. Desjardins, D. Duchesne, and C. Poirier, 125. E. Van der Wende, W. G. Characklis, and D. B. Smith,
Environ. Technol. 12, 569–580 (1991). Water Res. 23, 1313–1322 (1989).
99. M. Prevost et al., Proceedings of the International Ozone 126. J. D. Bryers and W. G. Characklis, Biotechnol. Bioeng. 24,
Association Conference, Lille, France, May 12, 1995. 2451–2476 (1982).
100. M. W. LeChevallier, W. C. Becker, P. Schorr, and R. G. Lee, 127. J. D. Bryers, Biotechnol. Prog. 3, 57–68 (1987).
J. Am. Water Works Assoc. 84, 136–146 (1992).
128. K. Haudidier et al., Water Sci. Technol. 20, 109–115
101. G. E. Speitel et al., J. Am. Water Works Assoc. 85, 86–95 (1988).
(1993).
129. J. C. Block et al., Biofouling 6, 333–343 (1993).
102. S. W. Krasner, M. J. Sclimenti, and C. J. Hwang, Experi- 130. J. C. Block, P. Servais, and P. Werner, Proceedings of the
ences with Implementing a Laboratory Program to Sample American Water Works Association Technology Conference,
and Analyse for DBP, Disinfection By-Products: Current on ‘‘Bacterial Regrowth — Bugs, Molecules and Surfaces,’’
Perspectives, American Water Works Association Research Big Sky, Mont., August 27–28, 1993.
Foundation, Denver, Colo., 1989.
131. P. Servais et al., Proceedings of the American Water Works
103. H. S. Weinberg, W. H. Glaze, S. W. Krasner, and M. J. Scli- Association Water Quality Technology Conference, Miami,
menty, J. Am. Water Works Assoc. 85, 72–85 (1993). Fla., November 7–10, 1993.
104. A. K. Camper et al., Proceedings of the American Water 132. L. Mathieu et al., Sci. Eau. 5, 91–112 (1992).
Works Association Water Quality Technology Conference, 133. M. W. LeChevallier, W. Shulz, and R. G. Lee, Appl. Environ.
San Diego, Calif., November 1–3, 1998. Microbiol. 57, 857–862 (1991).
105. J. M. Laine, M. M. Clark, and J. Mallevialle, J. Am. Water 134. D. Van der Kooij, J. Am. Water Works Assoc. 84, 57–65
Works Assoc. 82, 82–87 (1990). (1992).
106. J. Mallevialle, V. Mandra, I. Baudin, and C. Anselme, Pro- 135. C. Volk and J. C. Joret, Sci. Eau. 7, 131–152 (1994).
ceedings of the Filtech Conference, Karlsruhe, Germany,
136. C. Gauthier, M. Prevost, J. Mailly, and A. Rompre, Pro-
October, 1993.
ceedings of the International Water Supplier Association
107. Y. Richard, Proceedings of the International Water Supplier Workshop on NOM in Drinking Water, Poitiers, France,
Association Conference, Zurich, Switzerland, May 16–20, September 18–19, 1996.
1994. 137. P. Butterfield, B. D. Ellis, A. K. Camper, and W. J. Jones,
108. C. Moulin, M. M. Bourbigot, and M. Faivre, Key Engineering Proceedings of the American Water Works Association Water
Materials, Trans Tech. Publication, Switzerland, 1994, Quality Technology Conference, Denver, Colo., November
pp. 61–62. 9–12, 1997.
109. J. S. Taylor, D. M. Thompson, and J. K. Carswell, J. Am. 138. P. Butterfield, A. K. Camper, W. J. Jones, and B. D. Ellis,
Water Works Assoc. 79, 72–89 (1987). Proceedings of the American Water Works Association
 BIODEGRADATION: REDUCTIVE DEHALOGENATION AND METABOLISM OF CHLORINATED ORGANICS BY ANAEROBES 507

Water Quality Technology Conference, Tampa, Fla., Octo- BIODEGRADATION: LANDFILLS. See LANDFILLING
ber 31–November 3, 1999. OFMUNICIPAL SOLID WASTES: MICROBIOLOGICAL PROCESSES AND
139. A. Rompre et al., Proceedings of the American Water Works ENVIRONMENTAL IMPACTS
Association Water Quality Technology Conference, New
Orleans, La., November 12–16, 1995.
140. I. Sibille et al., Proceedings of the American Water Works
Association Water Quality Technology Conference, New
Orleans, La., November 12–16, 1995. BIODEGRADATION: OXYGENASE ENZYMES.
141. P. Laurent et al., J. Am. Water Works Assoc. 91(10), 63–72 See OXYGENASE ENZYMES: ROLE IN BIODEGRADATION
(1999).
142. C. Volk and M. LeChevallier, Appl. Environ. Microbiol. 65,
4957–4966 (1999).
143. A. K. Camper, Factors Influencing Biofilm Growth in
Drinking Water Distribution Systems, Ph.D. Thesis in Civil BIODEGRADATION: REDUCTIVE
Engineering, Montana State University, Bozeman, Mont., DEHALOGENATION AND METABOLISM OF
1995. CHLORINATED ORGANICS BY ANAEROBES
144. M. W. LeChevallier et al., Microbial Impact of Biological
Filtration, American Water Works Association Research STEPHEN H. ZINDER
Foundation, Denver, Colo., 1998. Cornell University
145. M. W. LeChevallier, N. Shaw, and D. B. Smith, Appl. Envi- Ithaca, New York
ron. Microbiol. 62, 2201–2211 (1996).
146. M. W. LeChevallier, N. Shaw, and D. B. Smith, Factors INTRODUCTION
Limiting Microbial Regrowth in Distribution Systems: Field
Experiments, American Water Works Association Research
Of the enormous number of different halogenated organic
Foundation, Denver, Colo., 1997.
compounds that have been chemically synthesized, several
147. AWWA and USEPA, Proceedings of the Conference On
have found widespread use, particularly as solvents and
Water Quality Modellings in Distribution Systems, American
biocides. This use has led to their being accidentally or
Water Works Association Research Foundation, Denver,
Colo., February 4–5, 1991.
intentionally released into the environment, mainly to
groundwater. Although there is good evidence for natural
148. A. K. Camper, N. Wright, W. L. Jones, and A. B. Cunnin-
production of chlorinated organic compounds (1), many of
gham, Proceedings of the American Water Works Association
Water Quality Technology Conference, San Francisco, Calif.,
these anthropogenic organic halogen compounds appear
November 6–10, 1994. to be resistant to chemical and biological degradation
and therefore persist in the environment. This chapter
149. O. Wable et al., T.S.M Eau. 6, 311–314 (1992).
discusses the roles of anaerobic microorganisms in the
150. W. G. Characklis and K. C. Marchal, Biofilms, John Wiley
biodegradation of these compounds and focuses mainly
& Sons, New York, 1990.
on chlorinated compounds, which are the most pervasive
151. P. Servais, P. Laurent, D. Gatel, and M. Prevost, Proceed-
organohalogen contaminants and have received the
ings of the American Water Works Association Water Quality
greatest attention. Additionally, reactions carried out by
Technology Conference, San Francisco, Calif., November
6–10, 1994.
microorganisms in pure culture are emphasized.
Chlorinated organic compounds of environmental sig-
152. P. Servais, P. Laurent, G. Billen, and D. Gatel, Sci. Eau. 8,
nificance (Fig. 1) include chlorinated methane, ethane,
427–464 (1995).
ethene, and benzene solvents; the biocide pentachlorophe-
nol; the herbicide 2.4-dichlorophenoxyacetic acid (2,4-D);
polychlorinated biphenyls (PCBs); the insecticide DDT
BIODEGRADATION: COMPOSTS. See COMPOST: [1,1,1-trichloro-2,2-bis (chlorophenyl) ethane]; and dioxin
BIODEGRADATION OF TOXIC ORGANIC COMPOUNDS contaminants such as 2,3,7,8-tetrachlorodibenzodioxin
(TCDD). The chlorinated solvents found widespread use
because they are nonflammable and were once consid-
ered relatively nontoxic; however, many are currently
considered to be carcinogens, teratogens, neurotoxins,
or estrogen mimics. Moreover, because many chlorinated
BIODEGRADATION: FUEL OXYGENATES. compounds are lipophilic, they tend to accumulate in fat
See MICROBIAL DEGRADATION OF FUEL OXYGENATES deposits in members of higher echelons of the food chain
including humans.
When chlorinated solvents are released into soil, they
form dense nonaqueous phase liquids (DNAPLs) or, in
combination with other compounds such as hydrocarbons
can form light nonaqueous phase liquids (LNAPLs). The
BIODEGRADATION, HALOGENATED DNAPLs, in particular, tend to accumulate as neat solvent
AROMATICS. See FATE AND MICROBIAL DEGRADATION OF at the bedrock or soil interface and then slowly leach into
HALOGENATED AROMATICS groundwater forming a plume. Because the maximum
508 BIODEGRADATION: REDUCTIVE DEHALOGENATION AND METABOLISM OF CHLORINATED ORGANICS BY ANAEROBES

Cl Cl Cl (a) Hydrogenolysis

Cl C CH3 R Cl + 2H R H + HCl
Cl C Cl Cl C H

Cl Cl Cl (b) Vicinal dichloro-elimination

Carbon Chloroform 1,1,1-Trichloro-
tetrachloride (CT) (CF) ethane (TCA) Cl C C Cl + 2H C C + 2HCl

H H
Cl Cl Cl H Figure 2. Reductive dechlorination of chlorinated organic com-
Cl C C Cl C C C C pounds. (a) Hydrogenolysis. (b) Vicinal dichloro-elimination. If
Cl Cl Cl Cl the carbon–carbon bond is double in part B, a triple bond is
H H
formed.
1,2-Dichloro- Tetrachloroethene Trichloroethene
ethane (DCA) (PCE) (TCE)
degrade poorly. Anaerobic processes, on the other hand,
Cl Cl Cl were originally considered slow, unreliable, and applicable
Cl Cl Cl Cl to only a few compounds. Significant improvements
in anaerobic techniques, however, have led to studies
showing that there is a much greater metabolic diversity of
Cl Cl Cl Cl anaerobes than was originally believed and that anaerobes
Cl Cl Cl were capable of utilizing highly chlorinated compounds,
Hexachlorocyclo- Hexachloro- 1,4 (para)-Dichloro-
often better than more lightly halogenated ones.
hexane (Lindane) benzene (HCB) benzene Aerobic metabolism of chlorinated organic com-
pounds generally proceeds via destabilization of the car-
Cl Cl bon–chlorine bond by addition of proximal hydroxyl or
epoxy groups by oxygenases, or sometimes by replacing
Cl Cl
the chlorine with a hydroxyl group by hydrolytic enzymes.
Under anaerobic conditions, electrons derived from the
Cl Cl Cl oxidation of organic matter are available for reductive
reactions, such as reductive dechlorination. The most
OH OCH2COOH
common type of reductive dechlorination is essentially
Pentachlorophenol 2,4-Dichlorophenoxy- a hydrogenolytic reaction replacing a Cl with an H, lib-
(PCP) acetic acid (2,4D) erating HCl (Fig. 2a). In certain compounds containing
Cl chlorines on adjacent carbons, there can also be a vic-
Cl(1−5) inal dihalo elimination (Fig. 2b) in which two chlorines
Cl(1−5) Cl C Cl are released as 2HCl and another carbon–carbon bond
Cl C Cl is formed.
H
Polychlorinated Dichlorodiphenyl- COMETABOLIC REDUCTIVE DECHLORINATION
biphenyls(PCBs) trichloroethane (DDT)

Cl O Cl Early studies showed that reduced transition metal com-

plexes including biological tetrapyrroles such as hemes
(Fe2+ ), and vitamin B12 and other corrinoids (Co1+ ) could
Cl O Cl carry out reductive dechlorination and other dehalo-
genations (2). Later, reduced cofactor F430 (Ni1+ ) from
2,3,7,8-tetrachloro-
dibenzodioxin (TCDD) methanogens (3) was also shown to carry out reductive
dehalogenations. Thus it was not surprising that various
Figure 1. Examples of synthetic organochlorine compounds of
environmental significance.
reductive dehalogenation reactions were demonstrated in
organisms rich in these cofactors, such as sulfate-reducers
(hemes), acetogens (corrinoids), and methanogens (corri-
noids and F430 ). However, in all these cases, the reac-
allowable concentrations of many of these solvents in tions represented a low fraction of the organisms’ overall
drinking water are one to five parts per billion, a metabolism and the organisms received no tangible ben-
large amount of groundwater can be contaminated by efit from them. These reactions are therefore considered
a relatively small amount of these compounds. cometabolic. Cometabolic reductive dechlorination may be
Research on biodegradation of chlorinated compounds quantitatively important in anaerobic habitats in which
initially focused on aerobic metabolism, and much these organisms are naturally abundant such as in anoxic
exemplary scientific data has been obtained about the sediments.
organisms, pathways, and genes involved in aerobic A few examples of this type of dechlorination are given
degradation processes. However, it became increasingly in the following section. Gantzer and Wackett (4) tested
apparent that many highly chlorinated compounds several compounds and found that hematin, vitamin B12 ,
 BIODEGRADATION: REDUCTIVE DEHALOGENATION AND METABOLISM OF CHLORINATED ORGANICS BY ANAEROBES 509

and coenzyme F430 could dechlorinate tetrachloroethene COOH

(PCE), hexachlorobenzene (HCB), and carbon tetrachlo-
ride (CT), whereas ferredoxins, which contain iron-sulfur
centers, and azurin, which contains copper, could not.
Indeed, vitamin B12 and F430 could reduce PCE completely Cl
to ethene. All the dechlorinations followed first-order H2
kinetics, and each subsequent dechlorination was slower
1
by more than an order of magnitude than that preceding it. HCl
In another example, the enzyme carbon monoxide dehy-
COOH
drogenase from Methanosarcina thermophila could also
reduce PCE to ethene (5), and a similar amount of reduc-
tive dechlorination was obtained from corrinoids in con-
centrations equal to that of the enzyme. Similarly, crude
extracts of the methanogen Methanobacterium thermoau-
totrophicum were shown to reduce 1,2-dichlorothane(DCA) 2
to ethene (vicinal dihalo-elimination) and to chloroethane
(hydrogenolysis) (6). A large fraction of this activity could 3H2 + CO2 3CH3COOH
be attributed to the enzyme methyl-coenzyme M methyl
reductase, which contains cofactor F430 . Interestingly, this 3 4
enzyme could also slowly reduce to ethene bromoethane
sulfonic acid, a molecule that is often used in ecological 3.5 CH4 + 3.5 CO2
studies as a specific inhibitor of methanogenesis because
Figure 3. Conversion of 3-chlorobenzoate to CH4 , and CO2 by
of its inhibition of the methyl reductase. a microbial consortium. Predominant microorganisms involved:
It has been shown that a variety of obligate and facul- (1) Reductive dechlorinator (Desulfomonile tiedjei (DCB-1));
tative anaerobes can carry out reductive dechlorination of (2) Benzoate oxidizer (Syntrophus buswelii); (3) H2 consuming
the pesticide Lindane (hexachlorocyclohexane, Fig. 1) (7). methanogens (Methanobacterium formicicum, and Methanospir-
A culture of Clostridium rectum was shown to conserve illum hungatei); (4) Acetotrophic methanogen, (Methanosaeta
more ATP from pyruvate oxidation when it reduced Lin- (Methanothrix sp.). After Dolfing et al. (10).
dane to 1,2,4 trichlorobenzene or pentachlorocyclohexane
to 1,4-dichlorobenzene (8). In this case, the Lindane appar-
ently was serving as an ‘‘electron dump,’’ allowing greater were used by hydrogenotrophic methanogens to produce
oxidation of the pyruvate, rather than as a respiratory 0.5 moles of CH4 from CO2 . Another three moles of CH4
electron acceptor. Enhanced growth by C. rectum in the and CO2 were produced from acetate by a Methanosaeta-
presence of Lindane could not be demonstrated, presum- like methanogen. This consortium must function as a
ably because of toxicity. unit because benzoate oxidation is only thermodynami-
cally favorable if H2 is removed by methanogens and the
REDUCTIVE DECHLORINATION AND dechlorinator and acetate is removed by the acetotrophic
DEHALORESPIRATION methanogens (11). Moreover, the dechlorinator is depen-
dent on benzoate oxidation to provide reducing power
A water shed paper in the field of reductive dechlori- (H2 ) needed for its dechlorination, whereas the benzoate
nation was that of Suflita and coworkers (9) in 1982. It oxidizer is dependent on the dechlorinator for its sub-
described enrichment cultures derived from lake sedi- strate.
ments and sewage sludge that carried out the reductive Desulfomonile tiedjei strain DCB-1 was eventually
dechlorination at relatively high rates of several chlo- isolated from the consortium (12). It grew only poorly,
rinated, brominated, and iodinated benzoates, but not even in rich medium, until its nutrient requirements were
fluorinated benzoates. This was the first description of determined to include nicotinic acid, 1,4-naphthoquinone,
reductive dechlorination of halogenated aromatic com- and thiamine. Once the organism grew well in a defined
pounds. Perhaps even more significant, however, was medium, it was clearly demonstrated to be able to
their statement that a mixed culture that completely use 3-chlorobenzoate as a respiratory electron acceptor
converted 3-chlorobenzoate (3-CB) to methane and car- for growth (13). Desulfomonile tiedjei also used sulfate,
bon dioxide with benzoate as an intermediate had been thiosulfate, and several metachlorinated benzoates as
maintained for two years. This was the first suggestion electron acceptors (Table 1).
that reductive dechlorination could promote growth of a Using reductive dechlorination as a respiratory anaer-
microorganism. obic electron-accepting reaction has been called dehalores-
Further studies led to a better characterization piration as well as halorespiration (14), but because the
of the microbiology of the 3-CB–utilizing consor- latter term could also apply to reduction of compounds
tium (Fig. 3) (10). A dechlorinating organism reductively such as perchlorate, the former one is used here. A con-
dechlorinated 3-CB to benzoate, which was then converted siderable amount of energy is available for conservation
to H2 , CO2 by a benzoate oxidizer resembling Syntrophus from reductive dechlorination. For example, the oxidation-
buswellii. One of the moles of H2 was used for the reduc- reduction potential of chloroethene reductions varies from
tive dechlorination of 3-CB, whereas the other two moles 0.37 to 0.58 v (2) and that for chlorinated aromatics varies
510 BIODEGRADATION: REDUCTIVE DEHALOGENATION AND METABOLISM OF CHLORINATED ORGANICS BY ANAEROBES

Table 1. Representative Organisms Capable of Respiratory Reductive Dehalogenation

Representative Electron Other Electron Phylogenetic

Organism Dehalogenations Donors Acceptors Position

Desulfomonile 3-chlorobenzoate to H2 , formate, pyruvate Sulfate, thiosulfate Delta Proteobacteria

tiedjei (12) benzoate
Isolate 2CP-1 (50) 2 chlorophenol to phenol Acetate Oxygen Delta Proteobacteria
Desulfitobacterium 2, 4, 6-trichlorophenol H2 , butyrate, pyruvate Sulfite, fumarate, Firmacutes
dehalogenans (27) to 2,4-dichlorophenol nitrate,
Dehalospirillum PCE to cis-DCE H2 , formate, pyruvate Fumarate, nitrate Epsilon Proteobacteria
multivorans (74)
Desulfitobacterium Pentachlorophenol to Pyruvate, Sulfite, thiosulfate, Firmacutes
frappieri (26) 3-chlorophenol, PCE nitrate
to TCE/cis-DCE
Dehalobacter PCE to cis-DCE H2 None Firmacutes
restrictus (25)
Enterobacter strain PCE to cis-DCE Formate, pyruvate, Oxygen, nitrate Gamma Proteobacteria
MS-1 (30) acetate
Desulfuramonas PCE to cis-DCE Acetate, pyruvate Polysulfide, fumarate, Delta Proteobacteria
chloro ethenica (29) Fe(III)
Dehalococcoides PCE to ETH H2 None Chloroflexi
ethenogenes (31) 1,2-DCA to ETH(VC)
Desulfovibrio strain 2,4,6-tribromophenol to H2 , formate, lactate Sulfate, thiosulfate, S° Delta Proteobacteria
TBP-1 (53) phenol
Trichlorobacter Trichloroacetate to H2 S (acetate) S° , fumarate Delta Proteobacteria
thiogenes (47) dichloroacetate
Dehalococcoides strain Tetra- and H2 None Chloroflexi
CBDB1 (56) trichlorobenzenes to
tri- and
dichlorobenzenes

from 0.27 to 0.48 v (15), close to the nitrate/nitrite cou- Cl Cl 2H HCl Cl H 2H HCl H H
ple (0.43 v). Reductive dechlorination is therefore a much C C C C C C
more favorable electron-accepting reaction than is sul-
Cl Cl Cl Cl Cl Cl
fate reduction (−0.22 v) or methanogenesis from H2 /CO2
(−0.24 v). PCE TCE DCEs
Desulfomonile tiedjei was officially described in 1991 2H
and had been in pure culture for several years before that,
yet it was not until the mid-1990s that several additional
dehalorespiring organisms were isolated (Table 1). Nearly HCl
all were most closely related to the gram-negative sulfur H H H H
compound reducers in the delta and epsilon subphyla of C C C C
the Proteobacteria (purple bacteria and relatives) or to the HCl 2H
H H H Cl
Desufotomaculum group in the Firmacutes (low %G + C
gram-positive bacteria). These groups contain organisms ETH VC
that are versatile at using electron acceptors. Figure 4. Reduction of tetrachloroethene (PCE) to ethene (ETH).
Intermediates include trichloroethene (TCE), dichloroethene
isomers (represented by the cis isomer), and vinyl chloride (VC).
CHLOROETHENE UTILIZATION

The solvents PCE and trichloroethene (TCE) were on an enrichment culture derived from a sewage digestor,
used extensively in the second half of the twentieth first demonstrated complete reductive dechlorination of
century, particularly for degreasing metal parts and dry chloroethenes to ethene (ETH) (Fig. 4), and reduction
cleaning clothes. They are among the most frequently as far as ethane has also been described (19). In the
encountered groundwater pollutants. Aerobically, TCE former system, it was shown that in the anaerobic
can be oxidized to an epoxide form by a variety of enrichment culture, when methanol was fed as an electron
oxygenase enzymes such as methane monooxygenase or donor, reductive dechlorination of PCE could occur at
toluene dioxygenase (16), whereas PCE appears resistant unprecedentedly high rates (20). It was subsequently
to oxidative attack. Several studies showed reductive demonstrated that H2 was the direct electron donor for
dechlorination by anaerobic microbial communities to dechlorination even when methanol was being supplied
the extent of vinyl chloride (VC) (17), a known human to the culture (21). The dechlorination of PCE, TCE,
carcinogen. Freedman and Gossett (18), in their studies and cis-DCE all followed zero-order kinetics in the
 BIODEGRADATION: REDUCTIVE DEHALOGENATION AND METABOLISM OF CHLORINATED ORGANICS BY ANAEROBES 511

concentration range assayed, and rates of utilization Thus far, D. ethenogenes is the only isolate capable
of these chloroethenes were similar (22), in marked of reductive dechlorination of chloroethenes past DCE.
contrast to reductive dechlorination carried out by reduced Although D. ethenogenes cannot grow reducing VC to
transition metal cofactors (4). A more defined culture using ethene, there is good evidence that other organisms in
H2 as the electron donor was shown to require vitamin enrichment cultures are capable of growth from reductive
B12 , a cofactor that presumably was supplied by methanol- dechlorination of VC (35,36).
utilizing methanogens and acetogens in the methanol-PCE
enrichment culture. CHLORINATED METHANE AND TRICHLOROACETATE
Several laboratories isolated organisms that could UTILIZATION
reduce PCE to TCE (23) or more often to cis-DCE. These
include Dehalobacter restrictus (24,25), a gram-positive Three of the chloromethanes, carbon tetrachloride
organism capable of utilizing only PCE or TCE as electron (CT or tetrachloromethane), chloroform (CF), and
acceptors and H2 as the electron donor (Table 1), hence dichloromethane (DCM or methylene chloride), are widely
its species epithet. Dehalospirillum multivorans is more utilized solvents and are pervasive groundwater contami-
metabolically versatile than D. restrictus and is a member nants. Chloromethane is produced biogenically by several
of the epsilon subphylum of the Proteobacteria. Several fungi and algae (1). Aerobically, CT and CF are barely
strains of Desulfitobacterium are capable of dechlorinat- utilized, whereas DCM can be utilized as the sole source
ing PCE and TCE (23,26–28). Desulfitobacterium is a of carbon and energy by several aerobic methylotrophic
gram-negative endospore former and is phylogenetically microorganisms (37). For anaerobic metabolism, one could
related to the sulfate-reducer Desulfosporosinus (Desulfo- imagine a reduction series for chloromethanes similar
tomaculum) orientis and to D. restrictus. Desulfuromonas to that for chlorinated ethenes in which CT is sequen-
chloroethenica (29) is one of the few dechlorinators capable tially dechlorinated to CF, DCM, CM, and finally CH4 .
of using acetate as an electron donor and can also reduce However, utilization of chloromethanes as respiratory
polysulfide compounds. Enterobacter strain MS-1 (30) is a electron acceptors has not been described, although it
facultative aerobe and reduces PCE and TCE when utiliz- is thermodynamically feasible. Rather, these CT and CF
ing nonfermentable substrates in the absence of another compounds appear to be cometabolized by a variety of
electron donor such as nitrate or oxygen. Interestingly, organisms, particularly those with high levels of corri-
an Enterobacter strain from the American type culture noids, that can chemically convert them to CO2 , CO, or
collection that presumably was naive to any exposure formate (38–43). Pseudomonas stutzeri strain KC secretes
to chloroethene solvents was capable of PCE reductive a factor that converts CT to CO2 along with small amounts
dechlorination. From these and other studies it can be of CS2 and COS when grown under nitrate-reducing and
concluded that organisms that reduce PCE as far as DCE iron-limited conditions. This factor has been identified as
are cosmopolitan, and at least some of them are relatively pyridine-2, 6-bis(thiocarboxylate) (44), a compound previ-
easy to culture. ously isolated from a Pseudomonas strain and considered
The enrichment culture has been studied from the Gos- to be a metal ion chelator.
sett laboratory, eventually describing in 1997 (31) the iso- Both DCM and CM are used as substrates by
lation of an organism, tentatively named Dehalococcoides anaerobic acetogenic microorganisms. In both cases,
ethenogenes strain 195, capable of reductively dechlorinat- the chloromethanes serve as substrates rather than as
ing PCE to etherene (ETH). Dehalococcoides ethenogenes respiratory electron acceptors. Dehalobacterium formi-
had a very complex nutrition, requiring acetate, vitamin coaceticum (45) converts DCM to acetate by the equation:
B12 , and extracts of mixed cultures containing it for its
growth. Of many substrates tested, it used only H2 as an 3CH2 Cl2 + 2CO2 → 2HCOO− + CH3 COO− + 9H+ + 6Cl−
electron donor and only chloroethenes or 1,2-dichlorothane (1)
(DCA), which it reduces primarily to ethene, as electron The DCM is at the same oxidation state as formaldehyde
acceptors. Conversion of VC to ethene is the rate-limiting and feeds into the acetogenic pathway in a similar manner.
step, and subsequent results (32) showed that it cannot It is to be noted that formate rather than H2 was produced
grow using VC or trans-DCE as electron acceptors. Dehalo- by this organism, which is unusual. Dehalobacterium
coccoides ethenogenes is a small disc-shaped organism that formicoaceticum grew poorly without a methanogen or
is resistant to peptidoglycan synthesis inhibitors such as sulfate-reducer to remove the formate it produces. The
vancomycin and ampicillin and apparently lacks a pepti- only tested substrate that D. formicoaceticum could use
doglycan cell wall. Initial phylogenetic analysis of its 16S was DCM. Dibromomethane was utilized in short-term
rDNA showed that it was in the bacteria but not closely assays but did not serve as a growth substrate, presumably
related to any isolated species. More recent analysis (33) because of toxicity. The organism’s 16S rDNA sequence
shows that its 16S rDNA resides in a deeply branching sub- showed that it is a member of the Firmacutes.
phylum of the Chloroflexi (green nonsulfur bacteria), most Strain MC1 is a gram-positive organism that forms
closely related to sequences that were derived from uncul- cocci in chains and ferments CM to acetate, CO2 , and HCl
tured organisms in natural habitats (33). Interestingly, by the equation:
one of the most closely related sequences to that from
D. ethenogenes was derived from an enrichment culture 4CH3 Cl + 2CO2 + 2H2 O → 3CH3 COO− + 7H+ + 4Cl−
carrying out reductive dechlorination of PCBs (34). (2)
512 BIODEGRADATION: REDUCTIVE DEHALOGENATION AND METABOLISM OF CHLORINATED ORGANICS BY ANAEROBES

Strain MC uses the CM as a source of a methyl group to Cl

ferment, much like it uses methanol (46). It was quite Cl Cl
versatile at substrate utilization, growing on sugars,
methanol, methoxylated aromatics, H2 /CO2 , lactate, or Cl
pyruvate. However, it did not utilize DCM. Cl Cl
An organism has been isolated that reductively
OH
dechlorinates the trichloromethyl group of Trichloroac-
etate (TCA) (47), which has been used as an herbicide. OH
Trichlorobacter thiogenes has an unusual metabolism for
its reductive dechlorination of TCA to dichloroacetate Cl
in that it uses sulfide as the source of electrons for Cl Cl Cl Cl
the reduction, thereby oxidizing it to elemental sulfur,
which it stores as intracellular granules. Sulfide was
regenerated from the sulfur by electrons derived from Cl
oxidation of acetate, a metabolic mode similar to its close OH
OH
relative Desulfuromonas in the delta subphylum of the
Proteobacteria.

Cl Cl
CHLOROPHENOL AND BROMOPHENOL UTILIZATION
Cl Cl Cl Cl
Chlorophenols have been used as biocides, and there is
(Lag)
particularly widespread use and environmental contam-
ination by pentachlorophenol (PCP). 2,4-Dichlorophenol
is formed after hydrolysis of the common herbicide OH OH
2,4-dichlorophenoxyacetic acid (2,4-D) (Fig. 1). Moreover, Figure 5. Reductive dechlorination of pentachlorophenol (PCP)
chlorophenols and related methoxylated compounds are to 3-chlorophenol by Desulfitobacterium frappieri (26). Addition of
formed during lignin degradation by fungi (48) and are, H, and removal of HCl as in Figure 3 are not shown for simplicity.
therefore, natural compounds found in soil. Aerobic organ-
isms that can degrade lightly chlorinated phenols such as
by marine sediments has been documented, and the
2,4-dichlorophenol are relatively abundant, and there are
sulfate-reducer Desulfovibrio strain TBP was isolated from
aerobes that can degrade PCP (49).
Reductive dechlorination of chlorophenols has been an anaerobic bromophenol-degrading enrichment culture
found in anaerobic sediments and enrichment cultures and derived from estuarine sediments (53) and was shown to
in at least two genera that are capable of conserving energy completely debrominate 2,4,6-tribromophenol to phenol as
from the reductive dechlorination of chlorophenols. One of a dehalorespiratory process.
these, called Isolate 2CP-1, dechlorinated 2-chlorophenol
to phenol (50). This organism was a facultative aerobe and REDUCTIVE DEHALOGENATION OF OTHER
is phylogenetically related to the fruiting myxobacteria, ORGANOHALOGEN COMPOUNDS BY MIXED
an unusual affiliation for a reductive dechlorinator, MICROBIAL POPULATIONS
although in the delta subphylum of the Proteobacteria.
Desulfomonile tiedjei could dechlorinate PCP to 2,4,6- There is good evidence of microbial reductive dechlo-
trichlorophenol, but the activity needed a chlorobenzoate rination of several other organochlorine compounds by
inducer and concentrations greater than 10 µM inhibited anaerobic microcosms or by mixed microbial cultures.
rather than stimulated growth on pyruvate, suggesting 1,1,1-trichloroethane (methylchloroform, TCA) (Fig. 1)
that the activity is fortuitous (51). was a commonly used solvent that has been shown to
Several Desulfitobacterium strains can use chlorophe- be transformed to 1,1-dichloroethane and chloroethane by
nols (23,26,27). Most strains only affect ortho dechlo- mixed anaerobic microbial populations (2). It also under-
rination of chlorophenols, whereas D. frappieri (26) is went a slow abiotic elimination reaction leading to the
capable of converting PCP nearly stoichiometrically to formation of 1,1-dichloroethene, which was then biologi-
3-chlorophenol (Fig. 5). It attacks the ortho chlorines first, cally transformed to VC (2).
and then after a lag period, during which the organ- 1,2-dichloroethane (DCA) (Fig. 1) is commonly used
ism is presumably inducing a different dehalogenase(s), it for making plastics and as a solvent and is read-
attacks the meta and para chlorines. Desulfitobacterium ily biodegradable under aerobic conditions. Dehalococ-
frappieri can dechlorinate many different chlorophenols coides ethenogenes could utilize DCA as an electron
except monochlorophenol and certain dichlorophenols. It acceptor (32), converting it to approximately 99% ethene
also dechlorinates several chlorocatechols and chloroani- (dihalo-elimination) and 1% VC. Mixed cultures containing
lines and both dechlorinates and demethoxylates several Dehalococcoides ethenogenes also carried out a rapid con-
methoxylated chlorophenols (52). version of the fumigant 1,2-dibromoethane to ethene (22),
Bromophenols are produced by certain marine inverte- but attempts to grow the pure culture using this com-
brates, apparently as a means of controlling microorgan- pound as a respiratory electron acceptor failed. Another
isms in their burrows (1). Degradation of bromophenols chlorinated ethane, the solvent 1,1,2,2-tetrachloroethane
 BIODEGRADATION: REDUCTIVE DEHALOGENATION AND METABOLISM OF CHLORINATED ORGANICS BY ANAEROBES 513

(PCA), was dechlorinated mainly to cis and trans isomers Ni and coworkers (62) purified the membrane-bound
of DCE and to VC by freshwater tidal sediments (54). 1,2- 3-chlorobenzoate reductive dehalogenase (RD) from Desul-
dichloropropane, an antinematode fumigant, was metabo- fomonile tiedjei. The dehalogenase that was found only in
lized by mixed cultures to propene, with monochlorinated cells grown in the presence of 3-CB was relatively oxygen
propenes detected only in early stages of enrichment (55), tolerant and used reduced methyl viologen as an electron
and apparently did not serve as intermediates in propene donor (as do all the RDs characterized so far), but not
formation. any known physiological electron donors. The protein had
Chlorinated benzenes have been shown to be reduc- two subunits with molecular weights near 64 and 37 kDa.
tively dechlorinated, and reactions leading from hex- The spectrum of the purified protein indicated the pres-
achlorobenzene (Fig. 1) to monochlorobenzene have been ence of a heme group. In a subsequent study, evidence for
demonstrated (14), although usually not all in the same an electron transport chain to the reductive dehalogenase
sample. Often a large amount of 1,3,5-trichlorobenzene involving a novel quinone and a low potential (−0.34 v)
accumulates in cultures containing hexachlorobenzene, cytochrome c was obtained (63).
suggesting that dechlorination occurs most readily when There have also been studies on the biochemistry of
two chlorines are adjacent to each other. Dehalococcoides reductive dechlorination in chloroethene dechlorinating
strain CBDB1, which showed 98.3% identity in its 16S organisms. The PCE RD from Dehalobacter restrictus was
rDNA sequence with D. ethenogenes, could dechlorinate isolated (64) and found to be membrane associated. It
1,2,3-trichlorobenzende, 1,2,4-trichlorobenzene, and some consisted of a single subunit with a molecular weight near
of the tetrachlorobenzene isomers (56). The organism did 60 kDa. Both visible and electron paramagnetic resonance
not dechlorinate PCE. Conversion of monochlorobenzene (EPR) spectral evidence showed the presence of iron-
to benzene is slow, if carried out at all, by mixed cultures. sulfur centers and a corrinoid, with oxidation-reduction
The insecticide DDT (Fig. 1) is no longer used potentials of 150 mv for Co(II) to Co(III) and −350 mv
in the United States but is still used elsewhere in for Co(I) to Co(II). Incubation of the Co(I) enzyme with
the world to protect crops and to eliminate malaria- PCE yielded the Co(II) form. On the basis of the lack of
bearing mosquitoes. Through chemical and biological reductive dehalogenation when whole cells were treated
processes, it has been shown that DDT’s trichloromethyl with reduced methyl viologen, which is considered to be
group under anaerobic conditions can be completely unable to permeate the cell membrane, it was concluded
dechlorinated (57). The fate of the rest of the molecule that the PCE dehalogenase was on the inside face of
under anaerobic conditions is not known. PCBs and dioxins the membrane and the hydrogenase was on the outside
(Fig. 1) are often resistant to aerobic biodegradation face (65). Evidence for the involvement of a quinone in the
because of their high level of chlorination. Evidence for electron transport chain was also obtained.
reductive dechlorination of these important environmental The biochemistry of reductive dehalogenation in
contaminants has been obtained using microcosms from Dehalospirillum multivorans has also been intensively
contaminated sites (14,58–60), as is discussed in greater studied. The PCE RD from this organism was purified as
depth in another chapter in this volume. a soluble cytoplasmic corrinoid-iron/sulfur-protein with a
molecular weight near 58 kDa (61). The gene encoding
this protein, pceA, was isolated and found to have
BIOCHEMISTRY AND MOLECULAR BIOLOGY OF little sequence similarity with previously described genes
REDUCTIVE DECHLORINATION in the databases, including those for other corrinoid
proteins. Cotranscribed with pceA was a second gene,
With isolates in hand that carry out specific reductive pceB, that encoded a small (8.3 kDa) hydrophobic protein
dechlorinations, the enzymes that are responsible for the hypothesized to serve as the membrane anchor for the
process and the genes encoding them can be studied. These RD. The pceA gene was expressed in Escherichia coli, but
studies are providing information about the molecular no enzyme activity could be detected, probably because
mechanisms of reductive dehalogenation and about the the expressed protein lacked correctly inserted prosthetic
evolution of the process and the organisms carrying groups. Thus, if one wanted to transfer a reductive
it out. It is conceivable that some day they may dehalogenase to another organism, it would need to be
provide material for genetically engineering organisms one that assembles the RD properly.
for reductive dechlorination. Two RDs were purified from a mixed culture containing
Because transition metal complexes are capable Dehalococcoides ethenogenes (66). The first enzyme was a
of catalyzing reductive dechlorination, they are good PCE RD that reduced PCE to TCE and had a molecular
candidates for prosthetic groups of reductive dehalogenase weight near 51 kDa, whereas the second enzyme had a
enzymes, which can provide these complexes with a milieu molecular weight near 61 kDa and reduced TCE to DCEs,
that can accelerate their catalysis. This prediction has held then to VC, and slowly reduced VC to ethene. Thus, PCE
true because the reductive dehalogenases characterized reduction to ethene is at least a two-step process in this
thus far contained either heme or corrinoid prosthetic organism. Inhibitor studies indicated that both RDs were
groups. Moreover, the dehalogenases carry out the corrinoid proteins.
terminal reactions of a respiratory electron transport An ortho chlorophenol RD was isolated from Desul-
chains to develop a proton motive force, and research is fitobacterium dehalogenans. This membrane-associated
being done on electron transport from the electron donors protein had a molecular weight near 48 kDa. As for the
to the reductive dehalogenases (61). haloalkene RDs, it contained iron/sulfur centers and a
514 BIODEGRADATION: REDUCTIVE DEHALOGENATION AND METABOLISM OF CHLORINATED ORGANICS BY ANAEROBES

corrinoid group. Its cloned gene, cprA, was approximately TCE biodegradation, it has been shown that because of the
30% identical at the amino acid level with pceA from more favorable energetics of reductive dechlorination ver-
Dehalospirillum multivorans, much closer than to any sus methanogenesis, reductive dechlorinators had a much
other proteins. Moreover, associated with it was a gene lower threshold for H2 than methanogens [Loffler, 1999
cprB that encoded a small protein homologous to pceB. #415]. Moreover, electron donors that lead to low H2 con-
Also of interest is that the cprA gene product had a signal centrations, such as butyrate or propionate, known to be
sequence for secretion as well as a twin-arginine motif poor substrates for methanogenic consortia (11), allowed
that is considered to be a signal for secretion of a cofactor- much higher ratios of dechlorination to methanogenesis
containing polypeptide into the periplasm. The pceA gene than other electron donors, such as lactate or ethanol,
product in D. multivorans also has these periplasmic leading to higher concentrations of H2 (70). It should be
localization signals, in conflict with biochemical evidence noted that in some cases, however, lactate or ethanol can
placing it in the cytoplasm. be converted to propionate or butyrate, so that they may
Thus, the RDs from Dehalospirillum multivorans and actually be good substrates, and it is always advisable
Desulfitobacterium dehalogenans form a distinct gene to test dechlorination in microcosms before doing pilot or
family that may have evolved for reductive dechlori- full-scale tests.
nation. A perusal of the BLAST microbial genomics Early in situ studies, such as the one in which benzoate
site (http://www.ncbi.nlm.nih.gov/Micro blast/unfinish- was added to a contaminated aquifer in Texas (71) and
edgenome.html) of the unfinished genome of Dehalococ- allowed complete conversion of PCE to ethene, indicated
coides ethenogenes indicates that it has several homologues that stimulation of in situ anaerobic biodegradation was
of these corrinoid-containing RDs. A transposon mutage- technically feasible. One important question is whether
nesis system has been developed for D. dehalogenans (67) the appropriate dehalogenating organisms are present
and should prove useful in defining genes involved in at the site. For example, it appears that organisms
reductive dechlorination. that reduce PCE and TCE to DCE are cosmopolitan,
whereas those carrying out further reduction, such as
Dehalococcoides ethenogenes, are not (72). Both microcosm
BIOREMEDIATION STUDIES
studies and pilot scale studies for a contaminated site at
Dover Air Force Base, Delaware, indicated that conversion
Bioremediation, especially in situ bioremediation, is con-
of TCE did not proceed past DCE despite the feeding of
sidered a promising technology for treating groundwater
large quantities of a lactate/yeast extract mixture. An
sites contaminated with chlorinated organic compounds.
enrichment culture developed from a site in Florida that
There are both oxidative and reductive strategies for
carried out complete reductive dechlorination of TCE to
the bioremediation of chlorinated organic compounds.
ethene was added to Dover microcosms and complete
Many lightly chlorinated compounds degrade naturally
dechlorination was effected (72). Several hundred liters
and it is mainly a matter of ensuring that the right
of this culture were then added to the site (after
physiochemical conditions are present to allow for their
pathogen tests and Environmental Protection Agency
biodegradation. Another aerobic approach takes advan-
clearance), after which complete conversion of TCE to
tage of the nonspecificity of certain oxygenases so that
ETH was detected (73). This is a successful example
TCE, for example, has been treated in pilot studies by
of bioaugmentation, the addition of microorganisms to
adding methane or toluene plus oxygen to groundwater
hasten biodegradation. It should be pointed out that
aquifers to stimulate the appropriate organisms (68).
often it is inappropriateness of the physiochemical milieu,
Approaches to anaerobic bioremediation of chlorinated
such as pH or lack of nutrients, rather than lack of
organics involve both natural attenuation and accelerated
an appropriate organism that limits biodegradation of
bioremediation (69). In the former case, other compounds,
a pollutant.
such as hydrocarbons or organic solvents or natural
It should be noted that reductive dehalogenation in
organic compounds, are present along with chlorinated
contaminated aquifers requires the interactions between
organic compounds. Degradation of these nonchlorinated
dehalogenators and other microorganisms in microbial
compounds leads to anoxic conditions under which they
communities, such as that depicted in Figure 3. The
can then be fermented to products, such as H2 , acetate,
understanding of the diversity of reductive dechlorinators
lactate, or butyrate, which are then utilized as electron
and the other members of the communities is still
donors by organisms capable of reductive dehalogenation.
rudimentary. It will be fruitful to apply both cultural
Lightly chlorinated products may migrate into aerobic
methods and molecular biological tools [Flynn, 2000 #457;
zones in which they may be degraded. By careful analysis
Loffler, 2000 #437] to characterize these communities so
of a contaminant plume, one can sometimes document
that it will be possible to understand and manipulate them
that the organohalogens are present at sufficiently low
better.
concentrations to meet regulatory guidelines by the time
they migrate off site or reach a receiving body.
If natural attenuation is not rapid enough, steps can CONCLUSION
be taken, such as addition of an electron donor, and
sometimes other nutrients, to accelerate the process. It is It is now clear that many chlorinated organic molecules are
desirable to have an electron donor that supports reductive susceptible to reductive dechlorination, including many
dechlorination but does not support a competing process, highly chlorinated molecules that are resistant to aerobic
such as methanogenesis, as well. In the case of PCE and attack. Particularly attractive for development are those
 BIODEGRADATION: REDUCTIVE DEHALOGENATION AND METABOLISM OF CHLORINATED ORGANICS BY ANAEROBES 515

processes in which the chlorinated organic can serve as 33. P. Hugenholz, B. R. Goebel, and N. R. Pace, J. Bacteriol. 180,
a respiratory electron acceptor because it then supports 4765–4774 (1998).
growth of the organisms carrying out the dechlorination. A 34. T. R. P. Holoman et al., Appl. Environ. Microbiol. 64,
greater understanding of both respiratory and cometabolic 3359–3367 (1998).
reductive processes and the organisms carrying them out 35. F. E. Löffler, K. M. Ritalahti, and J. M. Tiedje, Appl. Environ.
will increase that ability to optimize the destruction of Microbiol. 63, 4982–4985 (1997).
chlorinated organic molecules in the environment. 36. B. M. Rosner, P. L. McCarty, and A. M. Spormann, Appl.
Environ. Microbiol. 63, 4139–4144 (1997).
BIBLIOGRAPHY 37. R. Galli, G. Stucki, and T. Leisinger, Experientia 38, 1,378
(1982).
1. G. W. Gribble, Environ. Sci. Technol. 28, 310A–319A (1994). 38. M. D. Mikesell and S. A. Boyd, Appl. Envir. Microbiol. 56,
2. T. M. Vogel, C. S. Criddle, and P. L. McCarty, Environ. Sci. 1198–1201 (1990).
Technol. 21, 722–736 (1987). 39. L. C. Sepulveda-Torres et al., Arch. Microbiol. 171, 424–9
3. U. E. Krone et al., Biochemistry 28, 10061–10065 (1989). (1999).
4. C. J. Gantzer and L. P. Wackett, Environ. Sci. Technol. 25, 40. L. Semprini et al., in R. E. Hinchee and R. F. Olfenbuttel,
715–722 (1991). eds., On-Site Bioreclamation — Processes for Xenobiotic, and
5. P. E. Jablonski and J. G. Ferry, FEMS Microbiol. Lett. 96, Hydrocarbon Treatment, Butterworth-Heinemann, Boston,
55–60 (1992). 1991, pp. 41–58.
6. C. Holliger et al., J. Bacteriol. 174, 4,435–43 (1992). 41. J. G. Becker and D. L. Freedman, Environ. Sci. Technol. 28,
1942–1949 (1994).
7. G. Jagnow, K. Haider, and P. Ellwardt, Arch. Microbiol. 115,
285–292 (1977). 42. D. M. Bagley and J. M. Gossett, Appl. Environ. Microbiol. 61,
3195–3201 (1995).
8. N. Ohisa, K. Kurihara, and M. Nakajima, Arch. Microbiol.
131, 330–333 (1982). 43. C. Egli et al., FEMS Microbiol. Lett. 68, 207–212 (1990).
9. J. M. Suflita et al., Science 218, 1115–1117 (1982). 44. T. A. Lewis et al., Environ. Microbiol. 2, (2000).
10. J. Dolfing and J. M. Tiedje, FEMS Microbiol. Ecol. 38, 45. A. Mägli, M. Wendt, and T. Leisinger, Arch. Microbiol. 166,
293–298 (1986). 101–108 (1996).
11. B. Schink, Microbiol. Rev. 61, 262–280 (1997). 46. J. Traunecker, A. Preuß, and G. Diekert, Arch. Microbiol.
12. K. A. DeWeerd et al., Arch. Microbiol. 154, 23–30 (1990). 156, 416–421 (1991).
13. W. W. Mohn and J. M. Tiedje, Arch. Microbiol. 153, 267–271 47. H. De Wever et al., Appl. Environ. Microbiol. 66, 2297–2301
(1990). (2000).
14. W. W. Mohn and J. M. Tiedje, Microbiol. Rev. 56, 482–507 48. E. de Jong and J. A. Field, Annu. Rev. Microbiol. 51, 375–414
(1992). (1997).
15. J. Dolfing and B. K. Harrison, Envir. Sci. Technol. 26, 49. G. J. Stanlake and R. K. Finn, Appl. Environ. Microbiol. 44,
2213–2218 (1992). 1421–1427 (1982).
16. B. D. Ensley, Ann. Rev. Microbiol. 45, 283–299 (1991). 50. J. R. Cole et al., Appl. Environ. Microbiol. 60, 3536–3542
17. G. Barrio-Lage et al., Environ. Sci. Technol. 20, 96–99 (1986). (1994).
18. D. L. Freedman and J. M. Gossett, Appl. Environ. Microbiol. 51. W. W. Mohn and K. J. Kennedy, Appl. Environ. Microbiol. 58,
55, 2144–2151 (1989). 1367–1370 (1992).
19. W. P. de Bruin et al., Appl. Environ. Microbiol. 58, 1,996– 52. D. Dennie et al., Appl. Environ. Microbiol. 64, 4603–4606
2,000 (1992). (1998).
20. T. D. DiStefano, J. M. Gossett, and S. H. Zinder, Appl. Envi- 53. A. W. Boyle, C. D. Phelps, and L. Y. Young, Appl. Environ.
ron. Microbiol. 57, 2287–2292 (1991). Microbiol. 65, 1133–1140 (1999).
21. T. D. DiStefano, J. M. Gossett, and S. H. Zinder, Appl. Envi- 54. M. M. Lorah and L. D. Olsen, Environ. Sci. Technol. 33,
ron. Microbiol. 58, 3622–3629 (1992). 227–234 (1999).
22. V. Tandoi et al., Environ. Sci. Technol. 28, 973–979 (1994). 55. F. E. Löffler et al., Appl. Environ. Microbiol. 63, 2870–2875
23. J. Gerritse et al., Arch. Microbiol. 165, 132–140 (1996). (1997).
24. C. Holliger et al., Appl. Environ. Microbiol. 59, 2991–2997 56. L. Adrian et al., Nature 408, 580–583 (2000).
(1993). 57. J. M. Aislabie, N. K. Richards, and H. L. Boul, N. Z. J. Agri.
25. C. Holliger et al., Arch. Microbiol. 169, 313–321 (1998). Res. 40, 269–282 (1997).
26. B. Bouchard et al., Int. J. Syst. Bacteriol. 46, 1,010–5 (1996). 58. Wittich, Appl. Microbiol. Biotechnol. 49, 489–499 (1998).
27. I. B. Utkin, C. Woese, and J. Wiegel, Int. J. Syst. Bacteriol. 59. D. L. Bedard, H. Van Dort, and K. A. Deweerd, Appl. Envi-
44, 612–619 (1994). ron. Microbiol. 64, 1786–1795 (1998).
28. E. Miller, G. Wohlfarth, and G. Diekert, Arch. Microbiol. 168, 60. A. L. Barkovskii and P. Adriaens, Appl. Environ. Microbiol.
513–519 (1997). 62, 4556–4562 (1996).
29. L. R. Krumholz, Int. J. Syst. Bacteriol. 47, 1262–1263 (1997). 61. C. Holliger, G. Wohlfarth, and G. Diekert, FEMS Microbiol.
30. P. K. Sharma and P. L. McCarty, Appl. Environ. Microbiol. Rev. 22, 383–398 (1998).
62, 761–765 (1996). 62. S. Ni, J. K. Fredrickson, and L. Xun, J. Bacteriol. 177,
31. X. Maymó-Gatell et al., Science 276, 1568–1571 (1997). 5135–5139 (1995).
32. X. Maymó-Gatell, T. Anguish, and S. H. Zinder, Appl. Envi- 63. T. M. Louie and W. W. Mohn, J. Bacteriol. 181, 40–46 (1999).
ron. Microbiol. 65, 3108–3113 (1999). 64. W. Schumacher et al., FEBS Lett. 409, 421–425 (1997).
516 BIODETERIORATION OF MINERAL MATERIALS

65. W. Schumacher and C. Holliger, J. Bacteriol. 178, 2,328– BIOFILMS AND MINERAL MATERIALS
2,333 (1996).
66. J. K. Magnuson et al., Appl. Environ. Microbiol. 64, 1,270– Biofilms are not confined to solid-liquid interfaces. They
1,275 (1998). can also be found at solid-air interfaces (1). Almost all
67. H. Smidt et al., J. Bacteriol. 181, 6882–6888 (1999). surfaces in nature, including mineral materials, may be
68. P. L. McCarty et al., Envi. Sci. Technol. 32, 88–100 (1998). colonized by microorganisms; this includes buildings (2)
69. M. D. Lee, J. M. Odom, and R. J. Buchanan Jr., Ann. Rev. and monuments (3,4). Microorganisms are able to obtain
Microbiol. 52, 423–452 (1998). calcium, aluminum, silicon, iron, and potassium from the
70. D. E. Fennell, J. M. Gossett, and S. H. Zinder, Environ. Sci.
substratum by biosolubilization, a process that involves
Technol. 31, 918–926 (1997). the production of various organic and inorganic acids
by microorganisms, bacteria, algae, lichens (symbiotic
71. R. E. Beeman et al., in R. E. Hinchee, A. Leeson, L. Semprini,
and S. K. Ong, eds., Bioremediation of Chlorinated, and Poly-
associations between fungi and algae), and fungi. These
cyclic Aromatic Hydrocarbon Compounds, Lewis Publishers, acids attack and degrade calcareous minerals by the
Boca Raton, Florida, 1993, pp. 14–27. dissolution of calcium ions and lead to the eventual
72. M. R. Harkness et al., Environ. Sci. Technol. 33, 1100–1109
formation of degraded products such as gypsum, calcite,
(1999). glauconite, dolomite, ettringite, and quartz (5). In addition
to liberating nutrients, the production of acids lowers the
73. D. E. Ellis et al., Environ. Sci. Technol. 34, 2254–2260
(2000).
pH of the surrounding environment.
Microbial colonization of a surface can be associated
74. H. Scholz-Muramatsu et al., Arch. Microbiol. 163, 48–56
with the formation of mucilaginous matrices, which may
(1995).
also be involved in degradative processes. Capsules and
extracellular polymeric materials are produced by many
bacteria and algae, but little attention has been paid
BIODEGRADATION: WETLANDS. See WETLANDS: to the role of the fungi in the structure or nutrition
BIODEGRADATION OF ORGANIC POLLUTANTS of biofilms (6). Fungi have been reported as members
of natural biofilms (7) and their participation in these
structures on mineral materials deserves more attention.

BIODETERIORATION. See WEATHERING, MICROBIOL DETERIORATION PROCESSES

The processes involved in nonmicrobiological weathering

of rocks are better understood than those that occur in the
presence of microorganisms.
BIODETERIORATION OF MINERAL MATERIALS
NONBIOLOGICAL WEATHERING PROCESSES
CHRISTINE GAYLARDE
UFRGS Apart from degradation by living organisms, rock, stone,
Porto Alegre, Brazil and other constructional materials are weathered, that is,
GLYN MORTON reduced to a more dispersed state, by rain, wind, gravity,
UCLAN ice, temperature change, and reaction with atmospheric
Preston, United Kingdom gases. There are two general types of weathering:
mechanical, involving only physical forces, and chemical,
in which the chemical structure of the material is changed.
Mankind uses rocks, stones, and other natural and manu- In cold and/or dry climates, mechanical degradation is
factured products as construction materials. Both in their usually more important, whereas in the humid tropics and
natural state and when incorporated into a built structure, semitropics, chemical degradation is predominant.
they are subject to the deteriorative and degradative action The products of mechanical weathering include frag-
of the environment and microorganisms. These processes ments of rock and materials formed by expansion and
are normally referred to as weathering. It is often difficult contraction of the rock, freezing and thawing cycles,
to distinguish between nonbiological and biologically medi- and earth movements. Products of chemical weathering
ated weathering of mineral materials. The two processes include the soil in its many forms; these chemical trans-
can occur concurrently, each contributing to the overall formations follow mechanical weathering.
deleterious effects. Mineral materials are used extensively
in the construction industry and include a range of famil- Chemical Weathering
iar building materials such as stone, bricks, roofing tiles Chemical weathering involves the migration of ions
and shingles, asbestos, mineralized felts, fascias (which within the rock structure and within the surrounding
are often natural stone or ceramic in nature), cement- environment. Ions may be mobilized by solution in water,
based renderings, plasters and mortars, and adhesives which is enhanced by the presence of inorganic acids
and groutings. Concrete is used as reinforced structural such as carbonic acid, sulfurous acid, sulfuric acid, and
elements and as building blocks. nitrous acid, and HNO3 , derived from the gases carbon
 BIODETERIORATION OF MINERAL MATERIALS 517

dioxide, sulfuric dioxide, nitrogen monoxide, and nitrogen decreased, according to the nature of the deposits, leading
dioxide, respectively. In the cases of iron and manganese, to localized dissolution or salting (13). Often, the end result
the chemical elements in the structure can be oxidized by of these processes is a catastrophic failure of the surface
atmospheric oxygen or reduced by volcanic H2 S. Silicate layer, known as spalling. Until such spalling takes place,
rocks are complex salts of silicic acids. Cations such as the structure appears to be in sound condition, although
magnesium and potassium can be leached from the rock, the superficial rind or patina may hide a deep zone of rotten
leaving silica as the final product. Clays result from stone beneath the surface. After the initial catastrophic
the weathering of aluminosilicates, and other silicon- failure, the degraded mineral layer is subject to rapid
containing minerals include quartz (SiO2 ) and kaolinite and sustained erosion. Once this loose, friable surface is
(Al4 Si4 O10 (OH)8 ). Iron oxides and bauxite are other exposed, colonization by higher plants, with their extended
resistant products of the weathering of siliceous rocks. root systems, may lead to more rapid deterioration.
Limestones and dolomites are degraded by the dissolution
of calcium and magnesium carbonates in the structure,
leaving behind the less soluble impurities such as clay and SILICON AND SILICATES
quartz (sand).
Silicon is one of the most abundant elements of the
earth’s crust (about 28%) and is an important part of rock
BIOLOGICAL WEATHERING — BIODETERIORATION AND mineral structure, being present in most of the materials
BIODEGRADATION that comprise the crust; carbonates, phosphates, sulfates,
sulfides, chlorides, and coal are some of the more common
Krumbein and Jens (8) quote the Bible as an early examples of rocks in which silicon is not significant.
publication on the infectious nature of decay in houses Sandstone, slate, schists, and granitic rocks, together with
(Leviticus XIV, 33–57), which also indicated the difficulty ceramics, mainly in the form of brick or flooring and roofing
of treatment. Colonization and biodeterioration of building tiles, are siliceous materials widely used in buildings.
materials are determined by environmental conditions, the In view of the prevalence of silicon in the natural world,
most significant being physical parameters (temperature, it is not surprising that there exist microorganisms capable
moisture, light intensity) and the chemical nature of the of both degrading silicon minerals and absorbing silicon.
surface. It is generally accepted that microorganisms Silicon is a structural element in diatoms, the cell walls
do not directly use the mineral components of building of which consist of up to 95% hydrated, amorphous silica.
materials, but may solubilize them as a result of their
They take up the element in the form of orthosilicate
normal metabolic activities, allowing the resultant ions to
(H4 SiO4 ) (14). Less well studied has been the uptake of
be absorbed by the cells in some cases. It is axiomatic that
silicon by bacteria, but this has been shown to occur
the only microorganisms able to colonize clean building
from silica gel, quartz, or sodium silicate (15), whereas
surfaces are those that do not require organic materials
some fungi can accumulate silicon from organic silicon
for growth, namely, the autotrophs. These include the
complexes (16). All these processes are linked to the
photosynthetic algae and cyanobacteria and lithotrophic
biodegradation of silicon minerals.
bacteria. The literature usually cites photosynthetic
Bacteria and fungi are able to solubilize silicon and are
organisms as the primary colonizers (9,10,11), but there is
important agents in the weathering of rock silicates and
little direct experimentation to provide evidence for this.
aluminosilicates, both in nature and when used as building
Most of the information on the biological weathering
materials. The solubilization mechanisms include:
of minerals comes from the literature on soil formation
and describes processes at the soil/rock boundary.
1. Chelation. Gluconate, for example, produced by var-
Estimates of the rates of weathering of rocks on the
ious bacteria, can complex the cationic components
continental platforms give half-life values between 30 and
of silica, solubilizing kaolinite, and quartz at neu-
1,800 million years for complete weathering, depending
tral pH levels (17). Organic acids produced during
on the site and composition of the rock (12). This is
bacterial and fungal metabolism can cause the dis-
equivalent to 1 mm/y to 1 mm/1,000 y. Such processes
solution of quartz by chelation, which is more rapid
may be of importance in the deterioration of structures
built below the ground, but the literature in this in neutral than in acid solutions (18). Silicates, as
respect is mostly restricted to the behavior of metal anions, may form soluble, hexacoordinate siliconate
and concrete pipes. Above the ground, biodeterioration complexes with natural phenolics in soil (19).
is the result of activities of superficial (epilithic) and 2. Acid degradation. Many bacteria and fungi produce
deeper (endolithic) organisms. Colonizing organisms must inorganic and organic acids that solubilize silicon
be able to withstand dry conditions, the effects of high minerals. Walch and Ullman (20) showed that
temperatures and light at the surface. Changes in acidity organic acids are more effective than hydrochloric
resulting from anthropogenic and natural inputs are and nitric acid in solubilizing feldspars, but
also important. Microorganisms may increase the rate sulfuric acid produced by sulfur-oxidizing bacteria
of transformation by their metabolic activities, such as the is extremely active against the aluminosilicates
uptake of ions and the production of acids and chelating present in concrete (21).
agents. They may also alter the water permeability of 3. Alkaline degradation. Silicates and aluminosilicates
the minerals by the deposition of polymeric materials are readily mobilized under alkaline conditions
and surfactants. Water transport may be increased or because of the lability of the Al-O and Si-O bonds.
518 BIODETERIORATION OF MINERAL MATERIALS

A number of bacteria make their surroundings more These rocks usually consist of a mixture of all three
alkaline by the release of ammonia from proteins components with varying proportions of other minerals,
or urea and this has been shown, in the case of such as iron oxides. Their properties and classification
Sarcina ureae, to degrade nepheline, plagioclase, and depend on their chemical constituents and the form of
quartz (22,23). Because silicates are relatively stable the mineral components. Only limestone and sandstone
in dilute ammonia, it is reasonable to postulate that are commonly used as building materials. Shales are used
chelation is also important. in their metamorphic form, slate. Limestone is the most
4. Reaction with extracellular polysaccharide. The abundant of the nonclastic (nondetrital) rocks and is the
extracellular polymers released by fungi, and largest reservoir of carbon at the earth’s surface. It is the
especially bacteria, can react with siloxanes to most important type of crushed stone used in the building
form the more soluble organic siloxanes (16,24). The industry, constituting about 75% of all the commercial
reaction seems to be pH-dependent and is favored products sold and is the principal raw ingredient used for
by acid conditions (25). Extracellular microbial Portland cement, in addition to being the source of lime
polymers on the surface of a mineral material used to make plasters and mortars. Sandstone, on the
may also cause increased water residence time, other hand, is a detrital sedimentary rock formed by the
facilitating the chemical reactions that characterize cementation of sand grains that are commonly composed
low temperature silicate weathering (26). of quartz, but which may include many other minerals.
5. Depolymerization of polysilicates (siloxanes, meta- The overall stability of the rocks depends on the types of
silicates). Two bacteria (a Proteus and a Bacillus minerals and fragments of other rocks that are present
species) have been shown to be capable of this and this will obviously affect their biodegradability.
transformation (27), but in view of the requirement Rock and stone may be biodegraded by bacterial, fungal,
of microorganisms to absorb silicon in its monomeric and algal acids (3,10,30) and by mechanical penetration
form, it seems likely that this activity is not by fungi and phototrophs (31,32). Microorganisms may
restricted to these genera. be classified according to their mode of growth into
6. Where anaerobic conditions exist within the rock, epi- or endolithic, that is, growing upon or within the
both volatile and soluble alkoxysiloxanes may be stone, respectively. Aesthetic biodeterioration is produced
formed (28). by surface discolorations, which occur on microbial
colonization either on or within the stone.
7. Under dry conditions, many microorganisms pro-
duce osmotic protectants to resist desiccation. These
are polyols such as glycerol, mannitol, alpha- Role of Fungi in the Biodeterioration of Mineral Materials
glycerylglucose and sucrose, and other low molec- Fungi have been associated with the mechanical degrada-
ular weight organic molecules including proline and tion of stone (33,34) and many fungi isolated from stone
betaines (29). Polyols cause the expansion of mica- buildings have been shown to produce acids that could
ceous minerals by intercalation among the polysilox- degrade calcareous structures (4,35–38). For instance,
ane planes of the mica crystals. Such expansion will a dark-pigmented mitosporic fungus and the bacterium
induce stress in the rocks, leading to physical weath- Bacillus cereus, both isolated from limestone build-
ering and spalling, and may increase the access of ings in Uxmal in the Yucatan peninsula, Mexico, have
acids and chelating agents to ions deep in the crys- been shown to reduce an initial pH of 7.0 in a glu-
tals. Such polyols may also react directly with silanol cose/mineral salts medium to values between 3.0 and
groups present at the crystal edges to cause solution 4.0 pH units (39). A range of fungi have been shown
of silicon. to solubilize stone by organic acid production (40–43).
Dark-pigmented mitosporic fungi can also actively pene-
Chelation and transport of iron and manganese to the
trate limestone, causing ‘‘biopitting’’ (44) and have been
surface of rocks, where it is deposited as an oxide film,
shown to be particularly important in arid and semiarid
usually called a patina, create a barrier to the evaporation
environments of hot and cold deserts because of their abil-
of water. Where this barrier is absent, the excess water
ity to resist high temperatures, desiccation, and osmotic
evaporates and deposits salts, which can cause severe
stress (45). However, the published literature does not
local spalling. Once this process is initiated, the protective
suggest that fungi are present in high numbers on ancient
film cannot be reformed, and severe pitting may result.
limestone structures. Populations of 102 to 105 colony-
Spalling on the underside of clay roofing tiles is a common
forming units (cfu) g−1 in stone are common (38,46,47).
problem, which does not become apparent until the entire
It has been suggested that algae can contribute to fungal
area needs to be replaced; this is associated with the
stone decay by providing organic substrates for the produc-
growth of microorganisms and of mosses and lichens.
tion of acidic metabolites (3,48), and fungal activity rather
than numbers may be the more important parameter.
LIMESTONE, SHALE, AND SANDSTONE Alternatively, the low fungal numbers may be explained
by the detection techniques used. Slow growing fungi,
The three most abundant sedimentary rocks of the earth’s which are considered to be important in decay processes,
crust, accounting for 95% of all sediments, are shale, are outgrown in normal culture techniques by less rel-
sandstone, and limestone, formed from silicates, silica, evant, rapidly growing species, such as Aspergillus and
and by the biological deposition of calcium carbonate. Penicillium (49), and thus are often not detected.
 BIODETERIORATION OF MINERAL MATERIALS 519

There is no doubt that fungi are important in the organisms on building surfaces than any other class (10). It
aesthetic biodeterioration of building surfaces. Melanotic has also been suggested that they have the most important
fungi and actinomycetes, and cyanobacteria with dark- influence on the weathering of exposed natural stone and
pigmented cells or capsules, are frequently reported on rock (31,57), penetrating rock in dry environments at the
stone structures (47,50). Such growths cause much of rate of 5 µm/y (58). In dry, exposed situations, the majority
the discoloration seen on stone buildings (Figs. 1–8). of the organisms are found within the rock.
Warscheid and Krumbein (51) have discussed this phe- Many cyanobacteria are able to withstand high levels
nomenon with respect to natural stone surfaces. of UV illumination (59–63), enabling them to exist at
In the modern world, high levels of particulates in upper elevations. Nevertheless, most grow preferentially
the atmosphere, particularly those contained in fumes in shady situations, along with the algae. Chemical
from fossil fuels, adsorb to the outer surfaces of buildings and microbiological analyses have shown that superficial
and cause a gray-black discoloration (52). This can be growth is heavier on interior walls of ancient limestone
mistaken for fungal growth. Winkler (53) discussed the buildings (47,64). This is due to the more sheltered
apparent increase in the weathering rate of stone in conditions, reducing illumination below inhibitory levels
urban locations after industrialization toward the end and maintaining humidity, and due to the presence of
of the ninetienth century. It has been suggested that organic nutrients supplied by the droppings of birds,
the layers of organic pollutants may act as nutrient bats, etc. In buildings of the ancient Mayan site of
sources for the growth of heterotrophic microorganisms Uxmal, in Mexico, for example, the microbial biomass,
(bacteria and fungi), thus accelerating both the aesthetic expressed as total phospholipid fatty acids (PLFA) per
and physicochemical deterioration processes (30,54,55). gram of stone, showed that indoor surfaces had an average
The adsorbed pollutants are a diversity of organic value of 73.7 nmol g−1 , compared with 18.4 nmol.g−1
materials, including fatty acids and aliphatic and aromatic for outdoor sites (47). The PLFA profiles indicated the
hydrocarbons (54), and these obviously modify the nature presence of bacteria, actinomycetes, cyanobacteria, and
of the stone surface. fungi. Chlorophyll concentrations, showing the presence
Fungi will not grow, even on susceptible substrates, of phototrophic organisms, were similarly higher on
unless there is sufficient moisture. Fungi require water interior rather than outdoor surfaces (1.17 mg.g−1 , as
for the diffusion of nutrients into their cells and for the compared with 0.22 mg.g−1 ), suggesting higher algal and
production of enzymes. Water activity (Aw ) is numerically cyanobacterial populations on interior walls. This was
equal to RH/100 (e.g., 75% RH = Aw 0.75); on this basis confirmed by direct examination and culture techniques,
an RH of 70% is near the lowest limit for fungal growth. when it was found that the biomass was mostly that
For most environmental work the term water potential is of cyanobacteria (47). The major genus detected was
preferred, and it is measured in megapascals (1 mMPa is Xenococcus, with Synechocystis and Gloeocapsa the next
equivalent to 9.87 atm, or 10 bar). most abundant. The latter two cyanobacterial genera have
Fungi require water for the diffusion of nutrients into been stated to be capable of limestone boring activity
the cells and for the release of enzymes. However, water (presumably penetration of the substratum effected by the
can be present in the environment and still be unavailable dissolution of the limestone), along with Stigonema and
too because it is bound by external forces: Schizothrix (31).
Although there seems little doubt that cyanobacteria
1. the osmotic potential (solute binding forces φπ ) are frequently predominant on buildings in Latin America,
2. matrix potential (physical binding forces φm ) algae have been cited as the dominant microorganisms in
3. turgor (φp ) some other countries, with Trentepohlia frequently being
detected in Singapore (65) and Chlorella and Stichococcus
4. gravimetric potential (φg ).
in Russia (66). It is difficult to know whether these varying
results come from the use of different detection techniques
Their effects are additive, so they are denoted by the
in the various locations. However, the recent design of
general term water potential φ.
specific primers for the detection of cyanobacteria by
molecular methods will assure that such information in
φ = φπ + φm + φp + φg the future is reliable and repeatable.
The growth of cyanobacteria and algae on the surfaces
Role of Cyanobacteria and Algae in the Biodeterioration of
of buildings leads directly to aesthetic deterioration of
Mineral Materials
the surface. Photosynthetic biofilms may be gray, black,
Cyanobacteria (blue-green bacteria, previously called brown, red, or green, depending on the predominant
Cyanophyceae, or blue-green algae) resemble the oldest strain(s) of organisms and the environmental conditions
fossil organisms on this planet and are found in (Figs. 3 and 6). Humidity, temperature, and the degree
an enormous variety of habitats. As photosynthetic of exposure to light can all affect growth and pigment
microorganisms, along with the algae, they do not require production in algal and cyanobacterial cells. Under drier
organic materials for growth and are common inhabitants conditions, the biofilms are generally gray in color,
of all water bodies and soil. Cyanobacteria survive whereas more humid areas are more frequently green.
repeated cycles of drying and rehydration (56), making The filamentous alga, Trentepohlia, found to be the most
them particularly important on exposed surfaces and they common colonizer of buildings in Singapore (65), stores
are probably of greater ecological importance as pioneer orange/red carotenoids within its cells under appropriate
520 BIODETERIORATION OF MINERAL MATERIALS

Figure 1. Discolored stone statue in Karlovy Vary, Czech

Republic, showing growth of mainly mitosporic fungi and Figure 3. Fountain in Machu Picchu, Peru, showing intense
cyanobacteria. See color insert. growth of algae and cyanobacteria in the splash zone. See color
insert.

a thick fungal growth on top of a cyanobacterial/algal

layer will frequently be black (Fig. 8). However, the effects
of cyanobacterial growth on building surfaces are not
merely aesthetic. Danin and Caneva (57) suggest that
cyanobacteria contribute to the decay of calcareous stones
as follows:

1. Attachment of cyanobacterial cells in small fissures;

2. Growth within the fissure;
3. Water uptake and expansion of cell mass, exerting
pressure within the structure;
4. Precipitation of carbonates and oxalates around the
cells;
Figure 2. Mayan building in Chichen Itza, Mexico, showing
blackening of limestone by epilithic and endolithic fungi and 5. Opening of the fissure because of these internal
cyanobacteria. See color insert. pressures;
6. Entry of dust, pollen grains, etc.;
conditions, and this leads to the production of pink or 7. Partial death of cyanobacterial cells and establish-
orange biofilms. The presence of other microorganisms ment of heterotrophic bacteria, fungi, and small
also affects the appearance of the biofilm. For example, animals such as mites within the fissure;
 BIODETERIORATION OF MINERAL MATERIALS 521

Figure 4. Spalled soapstone on the external wall of a church in

Ouro Preto, Brazil. Cyanobacteria and actinomycetes were the
major microorganisms detected below the detached layer. See
color insert.

Figure 6. Intense algal growth behind a lamp attached to a

building in Sao Joao del Rei, Brazil. Note black growth of
mitosporic fungi in the rain runoff area below the lamp. See
color insert.

the cyanobacterium Synechococcus GL24. These spherical

cells possess on their external surface a so-called S-
layer, which, in spite of its hydrophobic nature, is able
to bind calcium ions at negatively charged sites (67).
The bound calcium complexes with carbonate ions when
the pH level is more than 8.3, and this high pH is
readily produced as follows: bicarbonate ions are taken
up by the cells and converted by the enzyme carbonic
anhydrase into carbon dioxide and OH− (69); the carbon
dioxide is incorporated into the cell mass and the OH−
ions are released and concentrated around the cells (70),
Figure 5. Salting on the surface of mortar in the tomb of Sevilha, raising the local pH levels. The cells of Synechococcus
Necropolis of Carmona, Spain. See color insert.
can become encrusted with calcite within eight hours in a
suitable environment and must continually shed patches
of mineralized S-layer to remain viable (71). The whole
8. Increasing internal pressure on the superficial layer
of the structure leading eventually to its detachment procedure is then repeated, allowing the further uptake
(spalling) (Fig. 4). of calcium. Mobilization of calcium ions by such metabolic
activity and ion transport, in addition to the trapping of
Cyanobacterial cells on limestone buildings are often released particles of calcite, either biotic or abiotic, in the
seen covered by calcareous deposits, suggesting migration gelatinous cyanobacterial sheath (72,73), is an important
of calcium from neighboring sites (47,67,68). The biological mechanism of stone degradation by cyanobacteria. The
immobilization of carbon as carbonates, which can be same processes occur with eukaryotic algae, and electron
deposited both intra- and extracellularly, occurs in micrographs showing algal and/or cyanobacterial cells
some bacteria, fungi, algae, and some metazoa, and in residing in micropits in limestone or mortar have been
cyanobacteria. Phototrophs deposit CaCO3 in the light published (47,74). Siliceous rocks are also more readily
and solubilize it at night because of changing bicarbonate degraded at high pH levels, as previously stated, because
concentrations. This process has been well studied in of the lability of the constituent cations under these
522 BIODETERIORATION OF MINERAL MATERIALS

Nevertheless, Van der Oost and coworkers (77) showed

that Cyanothece carries out mixed acid fermentation and
could, therefore, induce acid degradation of stone. Mao-
Che and coworkers (78) demonstrated the release of a
carbonate-digesting liquid from the apical cells of one
of the Pleurocapsales group; although the nature of the
substance was not determined, it was suggested that
it could have acidic properties. However, there is little
evidence that cyanobacteria produce acids in situ, and
Waterbury (79) suggests that acid decay of calcareous
materials in the presence of cyanobacteria is due to the
concurrent presence of heterotrophic bacteria.
Such associations of algal and cyanobacterial cells with
heterotrophic bacteria can enhance the weathering of rock
minerals through bacterial organic acid production (80).
It is also known that fungal microflora can metabolize
the organic matter secreted by phototrophic biofilms (81)
and contribute to stone decay through the release of acidic
metabolites (48). Urzi (46) has challenged the accepted
view that phototrophic colonization of stone is followed
by the growth of chemoorganotrophs, but Gaylarde
and Gaylarde (82) demonstrated that the treatment of
external walls with an algicidal substance (a copper-
containing compound), used at concentrations too low
for bacteriostatic and fungistatic activity, inhibited
fungal and bacterial growth. This indicates that the
inhibition of algal colonization reduces the growth of fungi
and heterotrophic bacteria, and supports the accepted
colonization sequence, although on chemically polluted
Figure 7. Growth of fungi (upper black layer) and lichen (lower buildings organic materials may act as nutrients for
pale green layer) on the quartzite course of a church in Tiradentes, nonphototrophs. Under these conditions, heterotrophic
Brazil. See color insert. bacteria could, indeed, be the primary colonizers of
buildings (83).

Role of Lichens in the Biodeterioration of Mineral Materials

Lichens, symbiotic associations of fungi and algae or
cyanobacteria, grow well on sun-exposed stone surfaces
and in relatively unpolluted atmospheres (Fig. 7) (84).
Ancient mortars are readily colonized by lichens because
of their composition and physical structure, through the
pores of which the hyphae may pass. Mortars, either
ancient or modern, are generally the initial point of lichen
colonization for other adjacent building materials, such as
marble (85).
Lichens induce the biodegradation of mortar and stone
by direct chemical attack of organic acids, principally
oxalic acid (5,85–87). Large and small pits are formed on
Figure 8. Intense black discoloration caused by growth of mainly
the surface of the material beneath the lichen thallus;
mitosporic fungi and cyanobacteria on the east-facing painted wall hyphae and hyphal bundles produce micropits 5 to 200 µm
of a church in Ouro Preto, Brazil. See color insert. in diameter, whereas the lichen fruiting bodies lead to
pits of 0.5 to 1 mm in size (85). The penetration of the
lichen structure into the rock or mortar produces a
conditions and because the hydrolysis of polysilicates is very strong association between the two and removal
favored at high pH levels. of the biological growths may damage the underlying
The suggestion that algae and cyanobacteria may also material even further. However, some materials are
dissolve stone by acid production (75) is more debatable. little penetrated by lichens, the degree of internalization
As discussed earlier, photosynthetic organisms actively being determined by the mineralogical composition of
concentrate carbon dioxide within the cells as bicarbonate, the substratum (85). A discussion of the effects of lichen
increasing the external pH level to as high as 10.5 within growth on various rock-forming minerals is presented in
48- hour laboratory growth in natural lake water (71,76). Jones and coworkers (88).
 BIODETERIORATION OF MINERAL MATERIALS 523

Lichen colonization does not necessarily lead to of molecular biological detection methods, which allow the
biodegradation of the structure and, in some cases, may identification of nonculturable organisms (115,116). Mem-
actually protect the surface against the scouring action bers of the family Archaea and many actinomycetes have
of wind and weather through the insoluble oxalate layer been detected on degraded wall paintings (117) and the
deposited beneath the thallus (89–91). Lichen growth may same techniques suggest that the major heterotrophic bac-
also be aesthetically desirable in some cases, covering ugly teria colonizing rocks and monuments are actinomycetes,
buildings and providing attractive coloration on otherwise mainly related to the genus Geodermatophilus (118,119).
drab constructions (92).

Role of Nonphotosynthetic Bacteria in the Biodeterioration CEMENT AND CONCRETE

of Mineral Materials
It is perhaps inevitable that civilizations in their search
Bacteria have been implicated in the biodegradation for an acceptable substitute for stone as a building
of mineral building materials because of the high material eventually devised a medium with many of
numbers of these prokaryotic organisms associated the characteristics and properties of natural stone. To
with deteriorated built structures (42,93–95). However, the prerequisite properties of durability and strength
Somova and coworkers (66) reported that saprophytic were added the extra advantages of a material that
bacteria were more common on undegraded, rather than was relatively inexpensive to manufacture and one that
degraded, masonry and consider that algae are more could be easily transported. Furthermore, in the hands of
important. imaginative architects and competent builders, a new era
The two groups of bacteria discussed in the literature of building was born in which the scale of building was no
are autotrophic and heterotrophic organisms. longer a major problem.
Cement can be described as a material that possesses
Autotrophic Bacteria. Bacteria that oxidize inorganic both adhesive and cohesive properties that enable it to
sulfur and nitrogen compounds are able to degrade stone bond various mineral fragments into one complex whole.
substrates by the production of inorganic acids (sulfate Since the introduction of Portland cement, concrete has
and nitrate). There is some debate about the overall become one of the most widely used synthetic materials
importance of these microorganisms in the decay of stone, within the construction industry (120). Portland cement
which is well discussed in the article by May and cowork- (Table 1) is the most commonly used cement mixture used
ers (30). However, the acceleration of decay by nitrifying in construction. However, other types of cements exist
bacteria has been demonstrated in the laboratory for lime- and are widely used. The chemical composition of cement
stone (96), ultrabasic rocks (97), and concrete (98). They varies according to the role required by that specific batch.
have also been implicated in the deterioration of plas- It is the unique chemical composition of cement that
ter (99) and asbestos cement roofs (100) in stables. Weight confers surface properties and tensile strengths peculiar to
losses have been detected in concrete blocks incubated with that specific formulation. In reality however, the silicates
sulfur-oxidizing bacteria (101,102). There is little doubt found in cement are not pure compounds; they contain
that these acid-producing bacteria can cause stone decay, minor oxides in solid solution that confer vital effects
but their role in the real world remains to be confirmed. on the atomic arrangements, crystal form, and hydraulic
properties of the silicates (120).
Chemoorganotrophic (Heterotrophic) Bacteria. The role Concrete is composed of cement powder, water, and
of heterotrophic bacteria in the formation of soil by rock aggregates of various sizes, such as sand or gravel (121).
breakdown is well accepted (103). The rock is degraded The main constituents of cement powder are lime, silica,
by bacterial acids and alkalis and by chelation. There is alumina, and iron oxide. The cement paste hydrolyses
little work, however, concerning the identity of organisms on curing to form hydrated calcium silicates (C-S-H
responsible for built stone decay, although many authors gels) and Portlandite (Ca(OH)2 ) (120,122–125), and the
have reported high numbers of heterotrophic bacteria on arrangement of these crystals allows channels to form
decaying stone (40,93,104,105). Weight loss experiments within the concrete structure. The formation of these
have shown the involvement of such bacteria in the channels within the concrete permits the capillary action of
degradation of sandstone (95,106,107), and there are a water, which may contain microorganisms. Thus, as water
number of publications on the staining of monuments by permeates deeper into the concrete, fissures are formed,
bacterial pigments (108–110). allowing the deposition of organic material and the further
Actinomycetes of various genera have been found
frequently on ancient wall paintings and monu-
Table 1. Main Compounds of Portland Cement
ments (66,111–113). They contribute to biodegradation
by the mechanical penetration of the mycelia into the Name of Compound Oxide-Composition Abbreviation
structure. The genus Geodermatophilus has recently been
considered to be of particular interest (66,113). The other Tricalcium silicate 3CaO·SiO2 C3 S
bacterial genus that is frequently isolated from rocks and Dicalcium silicate 2CaO·SiO2 C2 S
caves is Bacillus (114), although this group may not be Tricalcium aluminate 3CaO·Al2 O3 C3 A
Tetracalcium aluminoferrite 3CaO·Al2 O3 Fe2 O3 C4 AF
important in decay processes (113).
Our knowledge of the bacterial populations present on A. M. Neville, Properties of Concrete, 2nd ed., Pitman Publishing, London,
stone surfaces is currently being modified by the results U.K., 1977.
524 BIODETERIORATION OF MINERAL MATERIALS

ingress of microorganisms into the concrete (126). In acid to nitric acid (2). The first group includes species of
addition to the main compounds of cement, various minor Nitrosomonas and Nitrosovibrio, which oxidize ammonia
components are present, for example, MgO, TiO2 , Mn2 O3 , via hydroxylamine to nitrous acid. The second group
K2 O, and Na2 O. These components may serve as important that affects the oxidation of nitrous acid to nitric
sources of nutrients for colonizing microorganisms. acid includes members of the genera Nitrobacter and
In 1945, Parker (127) discovered that concrete was Nitrosovibrio (139). The action of nitric acid upon the
being broken down by the metabolic activities of calcareous binding material of concrete results in the
microorganisms. He isolated Thiobacillus from areas production of calcium nitrate. This is a soluble salt that
of badly corroded concrete. This and subsequent stud- is either lost from the concrete, resulting in the formation
ies (127–129) showed that Thiobacilli were the causative of corrosion pits, or remains, thus adding salt to the pore
agents of microbially induced corrosion of concrete struc- water.
tures. This view has been substantiated in recent years by
several others working in the same field (130–137). Sulfate-Reducing Bacteria
The sulfate-reducing bacteria are usually associated
Biogenic Sulfuric Acid Corrosion of Concrete
with metal corrosion rather than with the deterioration
The decay of concrete by microorganisms is dependent on of concrete, and therefore could be involved in the
the production of corrosive metabolites that can solubilize deterioration of the bars in steel-reinforced concrete.
minerals (138). However, sulfate-reducing bacteria are widespread in
Freshly prepared concrete has a pH level between 12 nature and active in locations made anaerobic by microbial
and 13 compared with the highly acidic pH (2.5 to 0.5) of digestion of organic materials. In sewers, the involvement
corroded concrete. The original pH must be reduced (by of sulfate-reducing bacteria in concrete decay is mediated
chemical carbonation) before microorganisms can begin by gas, produced as a result of the proteolytic action
to colonize and grow. The major organisms involved in of sulfate-reducing bacteria living in anaerobic slimes or
acid concrete decay are the thiobacilli; their incidence and biofilms formed below the sewage level. (2).
activity are described in Sand and Bock (2) and Milde and
coworkers (130). Biogenic Organic Acid Corrosion
An example is their involvement in the deterioration
Fungi. Fungi isolated from the surfaces of monu-
of concrete in sewage pipelines. Parker (127) reported
ments, facades of buildings, and degrading concrete
that the bacteria use hydrogen sulfide released from the
include many genera that are representative of air
sewage and oxidize it to sulfuric acid. Sand and Bock (2),
and soil flora (140–143). Biodeterioration of stone by
reporting on the findings of a research project undertaken
fungal organic acids has been described (141,144). Di-
in Hamburg to establish the mechanism of corrosion found
and trivalent cations are reported to be solubilized
in several pipelines constructed between 1967 and 1971,
by chelate-forming organic acids (145). Krumbein and
recorded that thiobacilli were present in high numbers
Schonborn-Krumbein (146) suggest mechanisms involving
at the corrosion sites. The incidence of Acidithiobacillus
‘‘siderophores’’ of fungi in the etching of microfractures,
thiooxidans was found to be high at sites that showed
whereas Krumbein and Petersen (145) report on the bio-
heavy corrosion. The bacterium was not detected at sites
transfer of cations involved in exfoliation induction in
where corrosion was not detected. Thus, it was considered
sandstone monuments. For concrete, however, this level
appropriate to adopt this microorganism as an indicator of
of research activity is lacking.
corrosion rather than pH values resulting from bacterial
sulfuric acid generation, which are partially buffered by
Lichens. Oxalic acid has been one of the most studied
different concrete types. Volatile sulfur compounds that
acids produced by lichens (147). It is produced by the
are dissolved in the sewage escape into the atmosphere
fungal partner of the lichen and is more prevalent
within the sewer. In this atmosphere, hydrogen sulfide is
in mature, calcium-loving species. (147,148). Lichens
auto-oxidized to molecular sulfur, which adheres to the
known to produce calcium oxalate are highly effective
concrete walls of the sewage pipes where it is oxidized
biodeteriogens in a range of rock types with oxalates
to sulfuric acid. This acid reacts with the calcitic binding
formed at the thallus/substratum interface. These have
material of the concrete thereby causing its destruction.
closely related chemical compositions to the rocks that
These workers also detected two other sulfur compounds
they colonize (148). Tiano (147) cites authors who discuss
on the sewer walls, namely, sulfur dioxide and thiosulfate.
the importance of the pedogenic activity of lichens.
Thiosulfate is a reaction product of sulfur dioxide with
The incidence of lichens on cement-based materials, for
molecular sulfur and forms a good substrate for the growth
example, weathered roofing tiles, is common in the United
of thiobacilli.
Kingdom and elsewhere, but there is little evidence of
biodeterioration problems.
Biogenic Nitric Acid Corrosion
Although thiobacilli are considered to be responsible for Algae and Cyanobacteria. The hygroscopic sheath cov-
the degradation of concrete above ground, two groups ering many algal cells growing on the surface of stones
of nitrifying bacteria are considered to play a major often seals moisture into the material, thus allowing for
role in the degradation processes below ground. They extended growth periods during dry spells (140). Repeated
are responsible for the oxidation of ammonia via nitrous freezing and thawing of the hydrated algae contributes
 BIODETERIORATION OF MINERAL MATERIALS 525

further to deterioration. Lithophilous algae and cyanobac- is a promising technique for probing building surfaces
teria can be considered as the pioneering inhabitants of and identifying changes in fungal biodiversity caused, for
stone surfaces (147). The colonizers of such surfaces are example, by the use of biocides. The future will see a huge
closely related to those present in nonflooded soils (32). increase in our knowledge of the types of organisms whose
Controversy exists as to whether these microorganisms activities are really important in the deterioration and
have a biodeteriogenic effect on the surfaces that they col- degradation of these structures.
onize (10,121). One point of view is that their deteriorative
effect is negligible (149,150). However, there are reports BIBLIOGRAPHY
of the positive role of these microorganisms on the weath-
ering of stone in nature (151,152) and on monuments and 1. L. H. G. Morton and S. Surman, Int. Biodet. Biodeg. 34(3/4),
sculptures (151,153,154). Such speculation applies equally 203–221 (1994).
to weathered concrete surfaces. 2. W. Sand and E. Bock, Geomicrobiol. J. 9, 129–138 (1991).
3. R. J. Koestler, A. E. Charola, M. Wypyski, and J. J. Lee, in
Chemoorganotrophic Bacteria. Chemoorganotrophs ex- G. Felix ed., Vth International. Congress in Deterioration
crete organic acids, including acetic and benzoic acid (155), & Conservation of Stone, vol. 2, Presses Polytechniques
which can affect the structural properties of concrete. The Romandes, Lausanne, Switzerland, 1985, pp. 617–626.
metabolic activity of these microorganisms will depend on 4. P. S. Griffin, N. Indictor, and R. J. Koestler, Int. Biodet. 28,
nutritional levels that prevail, in addition to a range of 187–207 (1991).
environmental conditions (147). However, there is little 5. C. Ascaso, J. Galvan, and C. Rodriguez-Pascual, Pedobiolo-
published work on their role in concrete degradation. gia 24, 219–229 (1982).
6. M. V. Jones, in J. Wimpenny, P. Handley, P. Gilbert, and
H. Lappin-Scott, eds., The Life and Death of Biofilms,
CONCLUSION Bioline, Cardiff, U.K., 1995, pp. 157–160.
7. M. E. Aktar and W. I. Kelso, Biol. Fertil. Soils 16, 305–312
The deterioration of stonework, brickwork, and other (1983).
mineral materials by microorganisms has an immediate 8. W. E. Krumbein and K. Jens, Oecologia 50, 25–38 (1981).
impact on the construction and conservation industries 9. W. E. Krumbein, Deutsche Kunst-und Denkmalpflege 31,
because of the financial implications of remedial treat- 54–71 (1973).
ments. The incidence of lichens, cyanobacteria, algae, 10. C. Grant, Int. Biodet. Bull. 18, 57–65 (1982).
and fungi on stonework and monuments is often con-
11. C. Saiz-Jimenez, in J. Szege, ed., Soil Biology and Conserva-
sidered merely to be aesthetically undesirable, rather tion of the Biosphere, Akademiai Kiado, Budapest, Hungary,
than severely damaging to these structures. However, 1984, pp. 757–767.
there is little doubt that physical and chemical attack 12. J. Vezier, in A. Lerman and M. Meybeck, eds., Physical and
can also occur through the activities of organisms growing Chemical Weathering in the Geochemical Cycle, Kluwer
on the surface (epilithic) or in deeper layers (endolithic) Academic Publishers, Norwell, Mass., 1988.
of rock and stone structures. The role of bacteria in the 13. W. H. Dukes, Architects’ J. 156, 433–438 (1972).
breakdown of natural rock and the genesis of soils has 14. J. C. Lewin, in R. A. Lewin ed., Physiology and Biochemistry
long been accepted and the same processes can partici- of Algae, Academic Press, New York, 1965, pp. 447–455.
pate in the biodegradation of stone-built structures. The 15. W. Heinen, Acta. Bot. Neerl. 17, 105–113 (1968).
actinomycetes have recently been highlighted as impor- 16. L. Holzapfel and W. Engel, Z. Naturforsch. 9b, 602–606
tant bacterial colonizers of rocks and buildings of cultural (1954).
heritage and future research will certainly focus on the 17. P. Vandevivere, S. A. Walch, W. J. Ullman, and D. L. Kirch-
degradative activities of these organisms. man, Microb. Ecol. 27, 241–251 (1994).
Biofilms, as regions where the metabolic activities of 18. P. C. Bennett, M. E. Melcer, D. I. Siegel, and J. P. Hassett,
microorganisms are concentrated, have been accepted as Geochim. Cosmochim. Acta. 52, 1521–1530 (1988).
of major relevance in the biodeterioration of materials, and 19. J. J. Flynn and F. P. Boer, J. Am. Chem. Soc. 91, 5756–5761
this is an area of increasing importance in the study and (1969).
control of biodeterioration of construction materials. Tech- 20. S. A. Walch and W. J. Ullman, Geochim. Cosmochim. Acta.
niques recently introduced into biofilm research, such as 57, 2725–2736 (1993).
atomic-force microscopy, environmental-scanning electron 21. W. Sand and E. Bock, Environ. Technol. Lett. 5, 517–528
microscopy, confocal-laser-scanning microscopy, green flu- (1984).
orescent protein, fluorescent in situ hybridization (FISH), 22. T. V. Aristovskaya and R. S. Kutuzova, Pochvovedenie 12,
denaturing gradient gel electrophoresis (DGGE), and 59–66 (1968).
other molecular biology methods, are already being used 23. R. S. Kutuzova, Mikrobiologiya 38, 596–602 (1969).
on artificial and natural mineral materials. Rolleke and 24. Z. A. Avakyan, T. A. Pivovarova, and G. I. Karavaiko, Mi-
coworkers (156) using 16S rDNA analysis, showed the krobiologiya 55, 369–374 (1986).
presence of the Archaea genus Halomonas on a mural in a 25. S. A. Walch and P. Vandevivere, Geomicrobiol. J. 12,
German church, confirming the presence of extremophiles 227–238 (1995).
(in this case, salt-tolerant bacteria) on buildings. Recently, 26. W. W. Barker, S. A. Walch, and J. F. Banfield, in J. F. Ban-
Saad and coworkers (157) have suggested the use of field and K. H. Nealson, eds., Gemicrobiology, Interactions
17S rDNA analysis by DGGE to identify fungal gen- Between Microbes and Minerals, Mineralogical Society of
era on painted surfaces. Early results suggest that this America, Washington, D.C., 1997, pp. 267–307.
526 BIODETERIORATION OF MINERAL MATERIALS

27. A. M. Lauwers and W. Heinen, Arch. Microbiol. 95, 67–78 54. C. Saiz-Jimenez, Aerobiologia 11, 161–175 (1995).
(1974). 55. P. Young, New Scientist 152, 36–38 (1996).
28. R. Huppmann, H. W. Lohoff, and H. F. Schroeder, Frese- 56. B. A. Whitton, in N. H. Mann and N. G. Carr, eds., Pho-
nius’ J. Anal. Chem. 354, 66–71 (1996). tosynthetic Prokaryotes, Plenum Press, New York, 1992,
29. A. D. Hocking, in D. H. Jennings, ed., Stress Tolerance of pp. 1–51.
Fungi, Marcel Dekker, New York, (1993). 57. A. Danin and G. Caneva, Int. Biodet. 26, 397–417 (1990).
30. E. May et al., Biodet. Abs. 7, 109–123 (1993). 58. A. Danin, Z. Geomorphol. N. F. 27, 413–21 (1983).
31. L. Hoffmann, Bot. Rev. 55, 77–105 (1989). 59. F. Garcia-Pichel, N. D. Sherry, and R. W. Castenholz, Pho-
32. J. J. Ortega-Calvo, H. Hernandez-Marine, and C. Saiz- tochem. Photobiol. 56, 17–23 (1992).
Jimenez, Int. Biodet. 28, 165–186 (1991). 60. F. Garcia-Pichel, C. E. Wingard, and R. W. Castenholz,
33. A. A. Lepidi and G. Schippa, in V. Romanowski ed., 1st Appl. Environ. Microbiol. 59, 170–176 (1993).
International Symposium on Deterioration of Building
61. T. Matsunaga et al., Appl. Microbiol. Biotechnol. 39,
Stones, Les Imprimeries Reunies de Chambery, Chambery,
250–253 (1993).
France, 1973, pp. 143–148.
62. N. M. Chazal and G. D. Smith, Microbiology 140, 3,183–
34. M. Bassi, N. Barbieri, and R. Bonechchi, Arte Lombarde
3,189 (1994).
68/69, 117–121 (1984).
63. A. Roy, P. Tripathy, and S. P. Adhikary, Arch. Hydrobiol.
35. M. E. K. Henderson and R. B. Duff, J. Soil Sci. 14, 236–246
Suppl. 120, 147–61 (1997).
(1963).
64. X. Ariño and C. Saiz-Jimenez, in J. Rieder ed., Proceedings
36. S. B. Weed, C. B. Davey, and M. G. Cook, Soil Sci. Soc. Am.
of the 8th International Congress on Deterioration and
Proc. 33, 702–706 (1969).
Conservation of Stone, vol. 1, Berlin, Germany, 1996a,
37. M. A. Resende et al., in C. C. Gaylarde, E. L. S. de Sá,
pp. 725–731.
and P. M. Gaylarde, eds., Biodegradation & Biodeterio-
ration in Latin America, Mircen/UNEP/UNESCO/ICRO- 65. Y. C. Wee and K. B. Lee, Int. Biodet. Bull. 16, 113–117
FEPAGRO/UFRGS, Porto Alegre, Brazil, 1996, pp. 65–67. (1980).
38. P. M. Gaylarde et al., Biofouling, 17, 41–45 (2001). 66. N. G. Somova, T. G. Dobrovol’skaya, G. M. Zenova, and
R. N. Ivanovskii, Microbiology 67, 569–574, translated from
39. H. A. Videla, P. S. Guiamet, and S. Gomez de Saravia, Int.
Mikrobiologiya, 67, 687–693 (1998).
Biodet. Biodeg. 46, 335–341 (2000).
67. S. Schultze-Lam and T. J. Beveridge, Appl. Environ. Micro-
40. D. M. Webley, M. E. K. Henderson, and I. F. Taylor, J. Soil
biol. 60, 447–453 (1994).
Sci. 14, 102–112 (1963).
68. C. Ascaso, J. Wierzchos, and R. Castello, Int. Biodet. Biodeg.
41. F. E. W. Eckhardt, in G. Felix, ed., Proceedings of the 5th
42, 29–38 (1998).
International Congress of Deterioration & Conservation of
Stone, vol. 2, Presses Polytechniques Romandes, Lausanne, 69. A. G. Miller and B. Colman, Plant Physiol. 65, 397–402
Switzerland, 1985a, pp. 643–652. (1980).
42. F. E. W. Eckhardt, in J. I. Drever, ed., The Chemistry of 70. A. G. Miller, G. S. Espie, and D. T. Canvin, Can. J. Bot. 68,
Weathering, Reidel Publishers, Dordrecht, The Netherlands, 1291–1302 (1990).
1985b, pp. 161–173. 71. S. Douglas and T. J. Beveridge, FEMS Microbiol. Ecol. 26,
43. M. K. Kanatchinova and B. K. Kasymbekov, Mikrobiol. 74–88 (1998).
Fitopatol. 20, 47–52 (1986). 72. A. Pentecost, Proc. R. Soc. London B 232, 125–136 (1987).
44. K. Sterflinger and W. E. Krumbein, Geomicrobiol. J. 14, 73. A. Pentecost, J. Gen. Microbiol. 134, 2665–2671 (1988).
219–225 (1997). 74. X. Ariño and C. Saiz-Jimenez, Aerobiologia 12, 9–18
45. K. Sterflinger, Antonie Van Leeuwenhoek 74, 271–281 (1996b).
(1998).
75. J. W. Gillatt, JOCCA — Surf. Coatings Int. 74, 6–12 (1991).
46. C. Urzi, in R. Guerrero and C. Pedos-Alio eds., Trends
76. A. Kaplan and L. Rheinhold, Ann. Rev. Plant Physiol. Plant
in Microbial Ecology, Spanish Society for Microbiology,
Mol. Biol. 50, 539–570 (1999).
Madrid, Spain, 1993, pp. 667–672.
77. J. Van der Oost et al., Arch. Microbiol. 152, 415–9 (1989).
47. O. Ortega-Morales et al., Curr. Microbiol. 40, 81–85 (2000).
78. L. Mao-Che et al., Biodegradation of shells of black pearl
48. G. Gómez-Alarcón, M. L. Muñoz, and M. Flores, Int. Biodet.
oyster, Pinctada margaritifera var, cumingri, by microbes
Biodeg. 34, 169–180 (1994).
and sponges of French Polynesia, Mar. Biol. 126(3), 509–515
49. U. Wollenzien, G. S. de Hoog, W. E. Krumbein, and C. Urzi, (1996).
Sci. Total Environ. 167, 287–294 (1995).
79. J. B. Waterbury, in J. T. Staley, M. P. Bryant, N. Pfennig,
50. S. Letznicka, A. Strzelczyk, and D. Wandrychowska, 6th
and J. G. Holt eds., Bergey’s Manual of Systematic Bacte-
International Congress on Deterioration and Conservation
riology, vol. 3, Williams & Wilkins, Baltimore, Md., 1989,
of Stone, Nicholas Copernicus University, Torun, Poland,
pp. 1765–1769.
1988, pp. 102–110.
80. E. I. Friedmann and R. Weed, Science 236, 703–709 (1987).
51. H. Heitz, W. Sand, H. C. Flemming, eds., Microbial Influ-
enced Corrosion of Materials, Springer, Berlin, Germany, 81. C. Nalewajko, K. Lee, and P. Fay, Microb. Ecol. 6, 199–207
1996, pp. 273–295. (1980).
52. K. L. Gauri, Sci. Am. 238, 104–110 (1978). 82. P. M. Gaylarde and C. C. Gaylarde, RILEM Workshop on
Microbial Impacts on Building Materials, Sao Paulo, Brazil,
53. E. M. Winkler, in R. Rossi-Manaresi ed., The conservation
2000, published on CD-ROM.
of Stone II, Proceedings of the 2nd International Symposium,
Centro per la Conservazione della Sculture all’Aperto, 83. E. Zanardini et al., Int. Biodet. Biodeg. 45, 35–42 (2000).
Bologna, Italy, 1976, pp. 27–36. 84. F. Leblanc and D. N. Rao, Bryologist 76, 1–17 (1973).
 BIODETERIORATION OF MINERAL MATERIALS 527

85. X. Ariño, A. Gomez-Bolea and C. Saiz-Jimenez, Int. Biodet. 113. C. Urzi, M. Garcia-Valles, M. Vendrell, and A. Pernice,
Biodeg. 40, 217–224 (1997). Geomicrobiology 16, 39–54 (1999).
86. D. Jones, M. J. Wilson, and J. M. Tait, Lichenologist 12, 114. I. Gonzalez et al., J. Microbiol. Methods 36, 123–127 (1999).
277–289 (1980). 115. D. M. Ward, R. Weller, and M. M. Bateson, Nature (London)
87. C. Ascaso and J. Wierzchos, Bot. Acta. 107, 251–256 (1994). 345, 63–65 (1990).
88. D. Jones, M. J. Wilson, and W. J. McHardy, in D. Houghton, 116. G. Muyzer, S. Hottentrager, A. Teske, and C. Wawer, in
R. Smith, and H. Eggins, eds., Biodeterioration 7, Elsevier A. D. L. Akkermans, J. D. van Elsas, and F. J. de Bruijn,
Science, New York, 1988, pp. 129–134. eds., Molecular Microbial Ecology Manual, Kluwer Aca-
89. A. O. Lloyd, in A. H. Walters and E. H. Hueck-van der demic Publishers, Dordrecht, The Netherlands, 1996,
Plas, Biodeterioration of Materials, vol. 2, Elsevier Applied pp. 3.4.4.1–3.4.4.2.2.
Science London, U.K., 1972, pp. 185–191. 117. S. Rolleke et al., Appl. Environ. Microbiol. 62, 2059–2065
90. X. Ariño, J. J. Ortega-Calvo, A. Gomez-Bolea, and C. Saiz- (1996).
Jimenez, Sci. Total Environ. 167, 353–363 (1995). 118. M. Eppard et al., Arch. Microbiol. 166, 12–22 (1996).
91. I. Grondona, E. Monte, V. Rives, and M. A. Vicente, Mycol. 119. D. Daffonchio et al., Int. Biodet. Biodeg. 44, 161 (1999).
Res. 101, 1489–1495 (1997). 120. A. M. Neville, Properties of Concrete, 2nd ed., Pitman
92. F. H. Brightman and M. R. D. Seaward, in M. R. D. Sea- Publishing, London, U.K., 1977.
ward, ed., Lichen Ecol., Academic Press, London, U.K., 1977, 121. E. H. Hueck van der Plas, Int. Biodet. Bull. 4(1), 11–28
pp. 253–293. (1968).
93. S. G. Paine, F. V. Linggood, F. Schimmer, and T. C. Thrupp, 122. G. Dreux, Connaissance du beton Societe de diffusion des
Phil. Trans. Roy. Soc. London B 222, 97–127 (1933). techniques du batement et des traveux publics, Lahure, Paris,
94. J. Pochon and C. Jaton, Chem. Ind. 9, 1587–1589 (1967). France, 1964.
95. S. Tayler and E. May, Int. Biodet. 28, 49–64 (1991). 123. J. Dubois, Chantiers Magaz. 80, 85–89 (1977).
96. J. Kauffmann and P. Toussaint, Corr. Anticorr. 2, 240–244 124. V. W. Berner, Waste Manage. 12, 201–219 (1992).
(1954). 125. A. Darimont, Microsc. Res. Tech. 25, 179–180 (1993).
97. E. V. Lebedeva, N. N. Lyalikova, and Y. Y. Bugelskii, Mikro- 126. F. M. Lea, The chemistry of Cement and Concrete, 3rd ed.,
biologiya 47, 898–904 (1978). Edward Arnold, London, U.K., 1970, pp. 591–593.
98. E. Bock et al., in D. Houghton, R. Smith, and H. Eggins, 127. C. D. Parker, Aust. J. Exp. Biol. Med. Sci. 23, 81–90 (1945).
eds., Biodeterioration 7, Elsevier Science, New York, 1988, 128. C. D. Parker, Nature 159, 439–440 (1947).
pp. 346–440. 129. C. D. Parker and J. Prisk, J. Gen. Microbiol. 8, 344–364
99. R. Wasserbauer and J. Novotny, Pozemni Stavby 34, 69–70, (1953).
(1986). 130. K. Milde et al., J. Gen. Microbiol. 129, 1327–1333 (1983).
100. R. Wasserbauer, Z. Zadak, and J. Novotny, Int. Biodet. 24, 131. G. I. Karavaiko, T. V. Zherebyat’eva, Doklady Akademmi
153–165 (1988). Nauk SSSR 306, 355–359 (1989).
101. E. Fjerdingstad, Water Res. 3, 21–30 (1969). 132. T. V. Zherebyat’eva, E. V. Lebedeva, and G. I. Karavaiko,
102. P. S. Gore and R. V. Unnithan, Indian. J. Marit. Sci. 6, Geomicrobiol. J. 9, 119–127 (1991).
57–65 (1977). 133. R. L. Islander et al., J. Environ. Eng. 117(6), 751–771
103. J. Berthelin, in W. Krumbein, ed., Microbial Geochemistry, (1991).
Blackwell Scientific, Oxford, U.K., 1983, pp. 223–262. 134. T. Mori, K. Nonaka, K. Tazaki, and Y. Hikosaka, Water Res.
104. F. E. W. Eckhardt, in W. Krumbein, ed., Environmental Bio- 26(1), 29–37 (1992).
geochemistry and Geomicrobiology, Vol. 2, The Terrestrial 135. M. Takeo et al., Chem. Express 7(8), 669–672 (1992).
Environment, Ann Arbor Science, Ann Arbor, Mich., 1978, 136. E. Bock and W. Sand, J. Appl. Bacteriol. 74, 503–514 (1993).
pp. 675–686.
137. R. N. Coleman and I. D. Gaudet, Water Res. 27(3), 412–418
105. F. J. Lewis, Ph.D. thesis (CNAA), Portsmouth Polytechnic, (1993).
Portsmouth, U.K., 1987.
138. F. J. Lewis, E. May, and A. F. Bravery, in L. H. G. Morton,
106. P. A. Wood and I. C. Macrae, Int. Biodet. Bull. 8, 25–27 ed., Biodeterioration of Constructional Materials, Biodeteri-
(1972). oration Society, Kew, Surrey, U.K., 1987, pp. 45–53.
107. F. J. Lewis, E. May, B. Daley, and A. F. Bravery, in 139. L. M. Prescott, J. P. Harley, and D. A. Klein, Microbiology,
L. H. G. Morton, ed., Biodeterioration of Constructional 4th International ed., WCB/McGraw Hill, New York, 1999.
Materials, Biodeterioration Society, Kew, Australia, 1988,
140. A. Perichet, in L. H. G. Morton, ed., Biodeterioration of
pp. 45–53.
Constructional Materials, Biodeterioration Society, Kew,
108. M. Bassi, A. Ferrari, M. Realini, and C. Sorlini, Int. Biodet. Surrey, U.K., 1987, pp. 103–112.
22, 201–205, (1986).
141. K. Petersen, J. Kurkoczin, A. B. Strzelczyk, and W. E.
109. R. Cook and G. Martin, in J. Black, ed., Recent Advances Krumbein, in D. R. Houghton, R. N. Smith, and H. O. W.
in the Conservation and Analysis of Artifacts: University Eggins, eds., Biodeterioration 7, Elsevier Science, London,
of London Institute of Archaeology Jubilee Conservation U.K., 1987, pp. 123–128.
Conference, July 1987, 1987, pp. 359–364.
142. C. Grant, C. A. Hunter, B. Flannigan, and A. F. Bravery,
110. R. J. Koehler and K. Krannich, Neue Museumskunde 4, Int. Biodet. 25, 259–284 (1989).
272–274 (1981).
143. K. McCormack et al., in C. C. Gaylarde, E. L. S. de Sá, and
111. Y. P. Petushkova, N. N. Lyalikova, and M. N. Poglasova, P. Gaylarde, eds., Labs 2, Biodegradation and Biodeteriora-
Mikrobiologiya 58, 1021–1030 (1989). tion in Latin America, MIRCEN, UNEP/UNESCO ICRO -
112. G. Weirich, Mater. Org. 24, 139–159 (1989). FEPAGRO/UFRGS, Buenos Aires, Brazil, 1996, pp. 68–69.
528 BIODIVERSITY IN SOILS: USE OF MOLECULAR METHODS FOR ITS CHARACTERIZATION

144. M. E. Williams and E. D. Rudolph, Mycologia 6, 648–680 soil and affecting the microclimate, create the unique
(1974). rhizoplane and rhizosphere habitats. The increased
145. W. E. Krumbein and K. Petersen, Mikroorganismen beschle- microniche complexity (5) at or near root surfaces selects
unigen den Zarfall mittelalterlicher Wandgemdid, Wand- for different microorganisms than those dominant in
malereischdden Arbeitshefte Denkmalpflege Neidersach- the bulk soil (6). Seasonal changes in weather and the
senm, 8, 115–121 (1990). concurrent vegetation cycles add a temporal heterogeneity
146. W. E. Krumbein and C. E. Schonburn-Krumbein, Bauten- to this complex system, so that no two-soil samples
schultz/Bausanierung. 1, 14–23 (1987). have the same microbial community. As a consequence,
147. P. Tiano, in K. L. Garg, N. Garg, and K. G. Mukerjii, eds., soil probably harbors most of the planet’s undiscovered
Recent Advances in Biodeterioration and Biodegradation, microbial biodiversity.
vol. 2, Naya Prokash, Calcutta, India, 1993, pp. 301–323.
In one study of genotypic diversity in a forest soil,
148. H. G. M. Edwards, D. W. Farwell, and M. R. D. Seaward, DNA renaturation kinetics were used to estimate that
Int. Biodet. 27, 1–9 (1991).
4 × 103 to 7 × 103 differently sized microbial genomes
149. W. H. Dukes, Architects’ J. 156, 429–432 (1972). were present in 1 g of soil (7). This represents as many
150. A. Paleni and S. B. Curri, in A. H. Walters and E. H. Hueck- as 13,000 species (8). Obviously, these species are not
Van der Plas eds., Biodeterioration of Materials, Applied present in equal numbers, but vary from a single cell to
Science, London, U.K., 1973, pp. 392–400. as many as 108 cells per gram of soil. In a similar study
151. A. Danin, Zeitschrift fur Geomorphologie. Neue Folge 27, employing DNA hybridization techniques, it was shown
413–421 (1983). that the diversity varied with soil type (9).
152. H. A. Viles, Earth Surfaces Processes Landforms 12, Two qualities of microbial communities can be
319–330 (1987). addressed in molecular microbial ecology studies:
153. G. Hyvert, Studies Conserv. 18, 131–155 (1973). (1) diversity, and (2) community structure. Although
154. G. Gaccione, M. L. V. Rinaldi, and C. Giacobini, in R. Rossi diversity has been defined in a number of ways (3,10,11),
Manaressi, ed., Conservation of Stone–1, Bologna, Italy it is often used in molecular microbiological studies, albeit
1976, pp. 245–256. incorrectly in a classical ecology framework, as a synonym
155. M. Wilimzig and E. Bock, in A. Bousher, M. Chandra, for species richness (12). This is the number of differ-
and R. Edyvean, eds., Biodeterioration and Biodegradation ent species or, more precisely, the number of different
9, Institute of Chemical Engineers, Rugby, U.K., 1995.
sequence types (phylotype richness) detected. Because of
pp. 195–198.
the special problems involved in quantitative analysis of
156. S. Rolleke et al., Appl. Environ. Microbiol. 62, 2059–2065
molecular biology data, we shall adopt the convention
(1996).
that diversity studies needs to only include information
157. D. S. de Sousa Saad et al., in 4 Labs, Fourth Latin American
on species presence. In contrast, analysis of community
Biodeterioration and Biodegradation Symposium, Buenos
structure includes quantitative information on the abun-
Aires, Argentina, 16–20th, April 2001. Published on CD
ROM. dance of different taxa or physiological groups, and is thus
more related to classical definitions of ecosystem diversity
such as Shannon’s index.
The assessment of the diversity and structural
BIODIVERSITY IN SOILS: USE OF MOLECULAR composition of soil microbial communities is extremely
METHODS FOR ITS CHARACTERIZATION dependent on the methodological tools used. Traditionally,
characterization of microbial communities was based
WERNER LIESACK on cultivation of the respective members. However,
PETER F. DUNFIELD this technique identifies only that tiny portion of the
Max-Planck-Institut für community which can grow on the media used, which
Terrestrische Mikrobiologie is often less than 1% of the viable species present (13–15).
Marburg, Germany Cultivation studies may fail to detect abundant or even
dominant members of soil microbial communities and
Soil is a matrix of organic and inorganic particles formed thus are strongly biased (16–20). This approach is also
by the combined action of biotic and abiotic processes. time consuming and labor intensive. As a consequence,
Geographic, geologic, hydrologic, climatic, vegetation, the past decade has witnessed the development of a
faunal, and anthropogenic factors together determine variety of molecular methods that bypass cultivation
a soil’s properties and, in turn, its resident microbial and instead base community analysis on DNA or RNA
community (1,2). The soil mineral composition, salinity, extracted directly from a given environment. This article
pH, nutrient availability, organic content, temperature, will focus on the current state-of-the-art molecular
and water content, among other factors, define which ecology techniques available for cultivation-independent
ecological niches are present. High spatial heterogeneity investigation of the phylogenetic and functional diversities
within a soil is a result of the vertical horizon structure of the prokaryotic domains Bacteria and Archaea in soils.
and of heterogeneity in such factors as plant cover, It is organized into the following sections:
microtopology, and perturbations by fauna. Therefore,
just as plant communities show local variations in • Extraction of total nucleic acids (DNA, RNA) from
species composition within one vegetation type (3), so soils
does the microbial community vary within one soil • Fractionation of total DNA based on mol% G + C
type (4). Plants, besides adding organic material to content
 BIODIVERSITY IN SOILS: USE OF MOLECULAR METHODS FOR ITS CHARACTERIZATION 529

• Phylogenetic diversity assessment by retrieval and of direct DNA extraction techniques is usually preferred
comparative sequence analysis of small subunit for studying soil microbial diversity. One advantage of
(SSU) ribosomal DNA (rDNA) the indirect approach is that uncultured bacterial cells
• Retrieval and comparative sequence analysis of might still be viable. This was observed to be especially
genes coding for key enzymes that perform defined true for a cell purification procedure involving lectin-
functions mediated bead extraction (31). The applicability of cell
• Molecular profiling of soil microbial communities extracts for parallel study of physiology and ecology has
using phylogenetic and functional markers also been demonstrated for uncultured methane-oxidizing
bacteria (36).
• Microbial activity at the functional level
• Metagenome analysis
Direct Methods. In comparative studies, direct in situ
cell lysis procedures have produced DNA yields that are
EXTRACTION OF TOTAL NUCLEIC ACIDS FROM SOIL higher by 60% (37), one order of magnitude (22) or 20 to
70 times (25), compared with yields obtained by indirect
The extraction and purification of high-quality nucleic cell extraction procedures. Consequently, DNA extraction
acids, either DNA or RNA, is a prerequisite for protocols reported in the last decade are primarily based on
almost all molecular ecology studies on indigenous direct cell lysis within the soil or sediment (24,25,38–48).
microbial communities (fluorescence in situ hybridization A critical factor in determining the DNA yield by
[FISH] is an exception). In principle, studies using direct extraction methods is the choice of lysis protocol.
extracted DNA assess the phylogenetic diversity (rDNA) Mechanical disruption of the cells should facilitate a
or genetic potential (functional gene markers) of soil quantitative extraction of environmental DNA (24,25) and
microbial communities. Studies on rRNA preferentially allow the detection of the overall genetic potential in
assess the active members of a community, because the sample, including dormant microorganisms. It may
many species vary their ribosome content in positive also recover DNA from moribund populations and even
correlation with their cellular activity (21). rRNA studies extracellular DNA from naturally lysed cells. On the
should therefore reveal adaptational shifts in community other hand, some DNA released during the direct lysis
structure and activity more readily than DNA studies do. procedure may bind to soil colloids and thereby reduce
In addition, rRNA-based approaches have the advantage the overall recovery of soil DNA (49–52). This DNA
of a natural amplification of target numbers, as one cell adsorption is directly dependent on soil composition.
can contain thousands of ribosomes. Although rRNA is A high content of clay minerals has an especially
the main component of total RNA extracted from natural strong negative influence on the recovery of total DNA.
environments, messenger RNA (mRNA) can also be Frostegård and coworkers (47) reported that this problem
extracted. Studies using mRNA are, however, technically could be overcome by pretreating the soil with RNA to
far more problematic (see section Analysis of Microbial saturate the adsorption sites. RNA pretreatment of soils
Activity at the Functional Level). having a high clay mineral content increased the DNA
recovery to levels obtained from soils with less clay content.
Extraction of Total DNA Of the lysis treatments available, those using the
DNA extraction should be carried out as soon as possible detergent sodium dodecyl sulfate (SDS) in combination
after soil samples have been collected from the field, as with bead mill homogenization give the highest DNA
changes in community structure can occur during storage. yields from soil (41,45,48). Moré and coworkers (41)
In one study, storage of soil samples for three weeks at 4 ° C reported that the sequential application of SDS treatment
significantly reduced the fraction of high molecular weight and bead mill homogenization was more efficient than
DNA retrieved (22). Two different procedures–indirect SDS treatment followed by three cycles of freezing
and direct–can be applied for the extraction of soil and thawing. The improved extraction efficiency was
DNA. Indirect methods separate intact cells from the soil attributed to a higher proportion of cells being lysed.
matrix prior to the lysis step (pioneered by Holben and Viable counts of sediment microorganisms revealed
coworkers (23)), whereas direct methods lyse cells within that 2 and 8%, respectively, survived the bead mill
the soil (pioneered by Ogram and coworkers (24)). homogenization and freeze-thaw procedures. Bacillus
endospores were not lysed by freeze-thawing treatment
Indirect Methods. Indirect methods use differential (94% survival rate), but were almost completely lysed
centrifugation steps to separate the cell fraction from by mechanical disruption (2% survival rate). Similarly,
the soil matrix. The cell fraction is then lysed, and total Kuske and coworkers (45) reported that only bead mill
DNA isolated and purified (25–29). The main drawback homogenization was effective for DNA extraction from
of this method is the bias inherent in removing cells from Bacillus spores. However, the study by Moré and
soil particles. Although several physical and chemical coworkers (41) also showed that small bacterial cells (0.3
dispersion techniques have been employed to efficiently to 1.2 µm) were more difficult to lyse than were larger
desorb cells from soil or sediment particles (27,30–32), bacteria (2 to 10 µm). After an SDS treatment and 10-
DNA obtained by these extraction procedures might not minute bead mill homogenization of one sample, the
be representative of the entire microbial community, that remaining unlysed 6% of the original number of bacterial
is, might exclude DNA from microbes that attach strongly cells, as determined by direct counting with acridine
to soil particles (24,33–35). As a consequence, the use orange, were all in an approximate size range of 0.3 to
530 BIODIVERSITY IN SOILS: USE OF MOLECULAR METHODS FOR ITS CHARACTERIZATION

1.2 µm. A major portion of soil bacteria (50 to 70% of of artificially introduced bacteria (58). DNA extracted from
the total bacterial numbers in nonrhizosphere soil) are crude cell fractions might also be more contaminated
less than 0.5 µm in diameter (53) and thus fall within with humic acids than DNA extracted using in situ cell
the size range of unlysed cells in the study of Moré and lysis (37). Procedures that have been applied, individually
coworkers (41). or in combination, to separate impurities from envi-
Miller and coworkers (48) evaluated the effectiveness ronmental DNA include: (1) adsorption of impurities on
of various DNA extraction procedures on samples of polyvinylpolypyrrolidone (PVPP) (23,44,59); (2) treatment
varying organic mater content, including two silt loam with lysozyme (37); (3) cesium chloride density gra-
soils and a silt loam wetland sediment. Again, bead mill dient centrifugation (22,24,25); (4) Sephadex G-75 (44),
homogenization in a lysis mixture containing chloroform, Sephadex G-200 (22,48,60,61), and Sepharose 4B (61,62)
SDS, NaCl, and phosphate-Tris buffer (pH 8) was found gel filtration; (5) precipitation of DNA with potas-
to give the highest DNA yield and the most complete sium acetate, cesium chloride, and spermine-HCl (39);
cell lysis. The recovery of high molecular weight DNA (6) agarose gel electrophoresis (43,48,58,63); (7) reversible
was greatest when using brief, low-speed bead mill binding of DNA on silica-based support material (43,48);
homogenization. Under optimal physical lysis conditions, (8) a magnetic capture-hybridization PCR technique (64);
the maximum molecular size of the DNA extracted and (9) affinity matrix purification (46). This last method
was 16 to 20 kb, which was similar to that obtained is a sequential injection fluid system combining affinity
in previous studies using bead mill homogenization purification with novel renewable-surface microcolumn
[e.g., 20 to 25 kb (37,44)]. More intensive use of bead technology. The affinity matrix used for purification
mill homogenization increases the risk of severely is based on the universal SSU rDNA oligonucleotide
shearing environmental DNA (24,33). Highly sheared, low 1392r (65), which is covalently attached to 60-µm
molecular weight DNA might still be useful for polymerase microbeads by a dT8 linker. Besides DNA oligonucleotide
chain reaction (PCR) amplification of SSU rRNA or probes, peptide nucleic acid clamps have been assessed for
functional marker genes, but there is an increased risk their use in affinity purification of DNA and RNA from
of chimeric gene fragments being formed during PCR (see soils and sediments (66).
section Analysis of SSU rRNA and Its encoding gene). An evaluation of four different DNA purification meth-
An alternative to bead mill homogenization is soil ods (silica-based DNA binding, agarose gel electrophoresis,
grinding (35,43,47,54). Frostegård and coworkers (47) ammonium acetate precipitation, and Sephadex G-200 gel
suggested that this method is equally efficient as, or even filtration) found that Sephadex G-200 gel filtration was
more efficient than, bead mill homogenization. Using soil the most effective at removal of PCR inhibitors from
grinding only, 16 to 59 µg of DNA per gram of soil was crude DNA extracts of a silt loam soil (48). However,
obtained from different dried soils, a yield comparable to CsCl/ethidium bromide density gradient ultracentrifuga-
that obtained with extraction protocols involving bead mill tion was better than Sephadex G-200 gel filtration at
homogenization (depending on the soil this ranges from removing humic acids from crude soil DNA extracts in a
1.5 to 53 µg DNA per gram of dry soil (22,39,41,44,48,55). separate study (22).
Soft lysis techniques without mechanical treatment Commercial kits for extraction of nucleic acids from
facilitate the retrieval of environmental DNA of high soils also commonly include purification steps. Purified
molecular weight (37,56). For example, a combined extracts from two commercial kits, the FastDNA Spin
lysozyme and SDS treatment extracted soil DNA of size 40 kit (Bio101, La Jolla, California) and the UltraClean
to 80 kb (37). DNA of this size is suitable for metagenome Soil DNA kit (MoBio Laboratories, Inc., Solana Beach,
analysis (see section Metagenome Analysis). However, California) gave similar PCR product yields (indicating
DNA obtained by soft lysis techniques might be biased similar coextraction of impurities) as soil DNA extracted
toward easily lysed microbial populations, especially with an SDS-bead mill homogenization method and
gram-negative microorganisms (43), and thus might not purified on Sepharose 4B spin columns (62). However,
be representative of the entire microbial community. there were differences in community fingerprints obtained
by amplified rDNA restriction analysis (ARDRA) and
Purification of Crude Extracts. The extent to which ribosomal intergenic spacer analysis (RISA) (see section
purification is necessary depends upon the characteristics Microbial Community Fingerprinting) of these PCR
of the soil under investigation, especially its humic acid products, indicating that different biases were inherent
content. Watson and Blackwell (57) reported that a brown- at some stage in the three different extraction/purification
tinted organic component of soil inhibits PCR and is tightly methods.
associated with the soil microorganism fraction. It was In summary, various protocols are available for cell
found to be distinctly different from humic and fulvic lysis and for purification of crude extracts of soil DNA.
acids in both its solubility properties and in its Fourier Selection of the most appropriate methods depends on
transformed infrared and nuclear magnetic resonance the aim of the study and on the characteristics of
spectra. This soluble inhibitor complexes with proteins, the soil being investigated, especially its humic acid,
and thus the authors postulated that it inhibits PCR organic matter, and clay mineral content. If complete
through interaction with Taq DNA polymerase. extraction of the gene pool for studies such as PCR-based
In general, molecular detection of gene sequences from assessment of microbial diversity is the main criterion,
indigenous microbial populations demands a more inten- harsh techniques, which extract DNA fragments of up
sive purification of extracted DNA than does the detection to 25 kb in size, should be applied. These include bead
 BIODIVERSITY IN SOILS: USE OF MOLECULAR METHODS FOR ITS CHARACTERIZATION 531

mill homogenization and soil grinding. In contrast, if taxonomic groups vary in G + C content by no more than 3
the focus of the study is metagenomic analysis, soft lysis to 5% (78,79). The G + C profiling percentage provides
techniques, which allow the extraction of large 40- to 100- a coarse level of resolution in cultivation-independent
kb DNA fragments, are more appropriate. Comprehensive community analysis, which is not achieved by the other
manuals are available, which have detailed descriptions available techniques (80). It allows a comprehensive and
of different protocols for the extraction and purification of quantitative analysis of the structure and dynamics of
DNA from bulk soil (37,47,67,68) and from other habitats, complex microbial communities in soils (80–82) and other
such as aquatic systems, the rhizoplane, rhizosphere, and environments (29,83,84). However, the method requires
phyllosphere (67). a relatively large amount of DNA (50 to 100 µg) and an
ultracentrifuge to separate various G + C fractions.
Extraction of Total RNA The separation is achieved through the binding of bis-
benzimidazole (Hoechst reagent number 33258) to adenine
Because RNA is more labile than DNA, more care
and thymidine. The bound bis-benzimidazole changes the
has to be taken in both extraction and subsequent
buoyant density of DNA, and this change is proportional
molecular analyses. All solutions and glassware have
to its A + T (hence, G + C) content (77). A gradient of
to be rendered ribonuclease (RNase)-free by diethyl
DNA fragments sorted by G + C content is established by
pyrocarbonate treatment (69), and only certified RNase-
equilibrium density-gradient (CsCl) centrifugation, then
and deoxyribonuclease (DNase)-free plasticware should be
DNA fractions with different G + C contents are collected
used. As a consequence, relatively few protocols have been
and quantified by spectrophotometry. After purification,
published for extraction of total RNA from soils (32,70–75)
individual DNA fractions can be further analyzed using
or sediments (76).
molecular ecology techniques. For example, comparative
Protocols have recently been developed for rapid
sequence analysis of SSU rRNA genes retrieved from
coextraction of rRNA and DNA (32,75,76). Cell lysis
DNA fractionated based on %G + C content allowed the
is achieved by bead mill homogenization, either of
characterization of dominant and rare members of a young
soil samples directly (75,76) or of crude cell extracts
Hawaiian soil bacterial community (81). Other molecular
obtained by using sodium pyrophosphate to disrupt cell-
techniques that have been applied to G + C fractionated
soil aggregates followed by several washing steps (32).
DNA include functional gene probing (80,85), SSU rDNA-
Removal of humic acids and other contaminants from
based fingerprinting (81), and amplified rDNA restriction
total nucleic acids is achieved using phenol-based
analysis (81,82).
extraction buffers in combination with various treatments
already described for extraction/purification of total
DNA, for example, inclusion of polyvinylpyrrolidone or ANALYSIS OF SSU rRNA AND ITS ENCODING GENE
hexadecyltrimethylammonium bromide in the extraction (rDNA)
buffer, phenol/chloroform extraction, and Sephadex G-
75 spin column filtration. Separation of rRNA from Methodological Aspects
DNA is achieved either by treatment with RNase-free Comparative analysis of SSU rRNA sequences has become
DNase (32,75) or by Sephadex G-75 spin column filtration the central tool for determining phylogenetic relationships
using different concentrations of the elution buffer (76). or, to put it otherwise, for reconstructing the history
A protocol described by Felske and coworkers (74) of life (86–91). It led to the proposal of the domains
recovers intact rRNA from soils by isolating ribosomes. Archaea, Bacteria, and Eucarya as the three main
After mechanical cell lysis, the glass beads, soil particles, divisions of the universal tree of life (92). Many properties
and cell debris are removed by centrifugation at low speed of SSU rRNA make it an ideal marker biomolecule
(15,000 to 30,000x g), and then the ribosome fraction is for the inference of phylogenetic relationships: (1) it is
pelleted by ultracentrifugation at 100,000x g for 2 hours universal to life, (2) it is highly conserved structurally
at 2 ° C. The inclusion of PVPP in the extraction buffer to and functionally, (3) it is, as the central component of
adsorb humic acids was essential for high efficiency of RNA the highly complex translation apparatus, one of the
recovery. The RNA yield obtained by the ribosome isolation most refractory biomolecules, and has probably never
technique was 0.2 µg of highly purified RNA per gram (dry (or very infrequently) been transferred horizontally from
weight) of soil. Yields from direct RNA extraction methods one species to another. SSU rRNA–based trees therefore
are clearly higher, ranging between 0.25 and 5.0 µg RNA depict evolutionary lineages of organisms and not of
per gram (dry weight) of soil, depending on the extraction the gene only, (4) it contains sufficient nucleotides to
procedure and the soil type (32,75). infer phylogenetic relationships, (5) different regions of
the gene vary at different rates (93,94), thereby allowing
FRACTIONATION OF TOTAL DNA BY MOL% G-C the inference of phylogenetic relationships from the
CONTENT domain to the species level. Highly invariant, slowly
evolving positions confer valuable information on widely
Soil DNA extracts can be used directly for molecular separated groups or organisms, whereas positions of
community studies. Alternatively, the DNA can first higher variability are useful for the elucidation of
be separated into fractions of different guanine plus more recent branchings (89,94), and (6) highly conserved
cytosine (G + C) content (77). The rationale behind this regions present targets for universal primers that allow
fractionation is that the G + C content of prokaryotic DNA PCR amplification of almost–full-length SSU rDNA from
varies from approximately 24 to 76%, but that particular most members of either the domain Bacteria (65,95–98)
532 BIODIVERSITY IN SOILS: USE OF MOLECULAR METHODS FOR ITS CHARACTERIZATION

or the domain Archaea (99,100). Primers for amplification used as a synonym to OTU (104)). One such technique is
of partial SSU rDNA from most members of Bacteria and ARDRA. In ARDRA, the SSU rDNA clones are digested
Archaea together have also been designed (65,95,97). by hexameric and/or tetrameric restriction endonucleases
The ease with which SSU rDNA sequences can be and the resulting digests are separated into size fragments
studied has led to a rapid increase in the number of by agarose gel electrophoresis. Restriction fragment length
sequences now available in public-domain databases. For polymorphisms (RFLPs) are used to assign SSU rDNAs
example, the SSU rDNA database of the Ribosomal to particular OTUs (80,82,104,128). Representative clones
Database Project (RDP) (101) currently contains more of individual OTUs can then be selected for sequence
than 16,000 aligned prokaryotic SSU rRNA sequences analysis (82,104). This approach is rapid, efficient, and
(∼75% are longer than 899 bp). Most of these have been thoroughly assesses the species richness in SSU rDNA
retrieved directly from environmental samples during clone libraries. However, a single OTU defined by ARDRA
cultivation-independent studies into microbial diversity. may consist of several diverse SSU rDNA sequences.
These cultivation-independent studies generally pro- Dunbar and coworkers (104) reported that partial SSU
ceed by retrieval of total community DNA, PCR amplifi- rDNA sequences (427 nucleotides analyzed), which had
cation of SSU rDNA, generation of an SSU rDNA clone been grouped into the same OTU based on identical RsaI-
library, and analysis of the clone library by comparative BstUI RFLP patterns, had an average sequence similarity
sequence analysis. PCR is carried out using the univer- of only 87%. Thus, the screening of clone libraries by
sal primer sets noted above, targeting Bacteria (102–106), ARDRA provides only a broad, preliminary estimation of
Archaea (100,107–109), or both domains (110,111). Nowa- microbial diversity. McCaig and coworkers (105) grouped
days the most-widely used system for generation of SSU SSU rDNA sequences randomly sampled from clone
rDNA clone libraries is the TA cloning kit (Invitrogen libraries into OTUs defined by a 97% level of sequence
Corp., Carlsbad, California). In this kit the linearized similarity. Sequences exhibiting similarities above 97%
cloning vector carries single 3 thymidine residues to may or may not correspond to microbial strains of the same
take advantage of a particular characteristic of the Taq genospecies, but sequence types below this similarity value
polymerase, the addition of a single adenosine overhang never correspond to strains of the same genospecies (129).
at each 3 terminus of the double-stranded PCR prod- The number of unique OTUs retrieved in relation to the
ucts (112). This system avoids the need for endonuclease total number of randomly sampled sequences (collector
restriction of the PCR product prior to the ligation reac- curve), provides information on how well the species
tion, and results in a high cloning efficiency by avoiding richness of clone libraries has been assessed (105).
the loss of information, which may occur because of It should be noted that the relative frequencies of
restriction. The current state-of-the-art tool for admin- defined OTUs in SSU rDNA clone libraries do not provide
istration and analysis of SSU rDNA sequence data is any quantitative information on community structure.
the ARB software package developed by O. Strunk and Although the relative abundance of a single OTU relative
W. Ludwig (Technische Universität München, Germany). to others is frequently compared across samples to show
It is available online at http://www.mikro.biologie.tu- community differences in the relative importance of this
muenchen.de/pub/ARB/. The program includes various OTU (see section Microbial fingerprinting techniques),
tools for phylogenetic analysis, including the phylogeny one OTU cannot be directly compared to another because
inference package (PHYLIP version 3.2 113) and max- many biases lead to the preferential recovery of some
imum likelihood treeing methods (fastDNAml (101)). sequences over others. Bias can occur at all methodological
Details on how to construct phylogenetic trees with SSU levels from environmental sampling to data analysis (130),
rDNA sequence data and the various treeing algorithms but are especially severe during extraction of total soil
available are reviewed elsewhere (94,114,115). DNA (47,62), during multitemplate PCR (131–136), and
Total RNA, instead of DNA, can be extracted from because of possible contaminant sequences (137).
soils, and the SSU rRNA reverse transcribed (116–119). Another problem in interpreting environmental SSU
Especially in soil environments, diversity assessment rDNA sequences is the formation of chimeric SSU rDNA
with rDNA is experimentally less problematic than (hybrids) during multitemplate PCR (136,138–141).
with rRNA. However, organisms detected by retrieval Chimeric SSU rDNA sequences result in artificial lineages
of SSU rDNA may be ecologically unimportant because during treeing analysis, therefore any conclusion that
DNA persists in moribund and dead cells (120) and can a new microbial lineage has been detected in an
be present as extracellular DNA adsorbed to mineral environment must always be made with care. It should
surfaces (24,121,122). In contrast, RNA is highly labile. be checked by separate analysis of the 5 and 3 sequence
Ribosome levels, and therefore rRNA levels, have been halves (106,108,142), by use of the chimera check program
correlated with cellular activity (21,123–126). Thus, it of RDP (101), and by a careful secondary structure
can be argued that retrieval and analysis of SSU rRNA analysis.
provides a more representative view of the metabolically
active members of microbial communities (116–119).
Phylogenetic Diversity
Assessment of the SSU rDNA diversity in a clone
library is usually carried out by partial or full-length The first cultivation-independent phylogenetic surveys
sequence analysis of randomly selected clones, or by of microbial diversity were published in 1990, regard-
techniques which rapidly assign clones to operational ing samples collected from hot springs of the Yellow-
taxonomic units (OTUs) (127) (the term phylotype is often stone National Park (143) and the Sargasso Sea (144).
 BIODIVERSITY IN SOILS: USE OF MOLECULAR METHODS FOR ITS CHARACTERIZATION 533

The first studies on soil were published shortly there- of bacterial descent: Acidobacterium capsulatum (160),
after, using samples from subtropical Australia (63,145). Holophaga foetida (161), and Geothrix fermentans (162).
Since then, surveys of microbial diversity have been However, members have been detected both by SSU
conducted on diverse soil environments including: trop- rDNA (82,102–105,111,149,150,157,158) and by SSU
ical Hawaiian forest and pasture soil (82), mature forest rRNA (117,119,147) retrieval from various soil types
soil and active pasture in Eastern Amazonia (111), arid and geographic regions (Fig. 1). Based on the three
pinyon pine-juniper woodland soils in Arizona (102,104), cultured representatives and on the many envi-
grassland soils in Scotland (105,146) and the Nether- ronmentally retrieved SSU rDNA sequences, up to
lands (119,147,148), agricultural soils in Wisconsin (110) eight monophyletic subdivision-level lineages within the
and the Netherlands (149), German moorland soils pol- Holophaga/Acidobacterium group have been identified.
luted with zinc (150) or polychlorinated biphenyls (117), The three cultured organisms provide clues about the
Italian rice paddy soil (106,118,142,151), and Siberian physiological versatility within the group: A. capsulatum
tundra soil (152). is a moderately acidophilic aerobic heterotroph, whereas
These cultivation-independent studies have dra- H. foetida and G. fermentans are strict anaerobes that
matically changed our view of microbial diversity ferment aromatic compounds and acetate, respectively.
(88,91,107,153,154). They have contributed to the Geothrix fermentans is also able to grow by dissimila-
tripling of identifiable bacterial divisions (also called tory iron reduction. These diverse phenotypes, and the
main lines of bacterial descent or phyla) during the widespread occurrence of SSU rDNA sequences belong-
last 12 years (154–156). Only 23 of the 36 to 40 ing to the Holophaga/Acidobacterium division in diverse
presently recognized bacterial divisions are currently environments, suggests that members of this group
represented by cultured species (154,155). The total are active in many soil ecosystems. Besides terrestrial
number of divisions may in fact be well over 40, judg- soil (103,154,158), species have been detected in a peat
ing by various division-level lineages represented by bog (163), acid mine drainage (164), marine samples (165),
only single or few known sequence types (142,154,157). activated sludge (103), bioreactors (166), a contaminated
Major bacterial groups commonly detected by aquifer (155), and a hot spring (153), as well as in sedi-
cultivation-independent phylogenetic surveys of soils ments, water column and lake snow aggregates of fresh-
are: Holophaga/Acidobacterium (103,154,158); Proteobac- water lakes (103,167), and on sediment particles sampled
teria (82,102,104,105,110,111,117,119,147,150,152,157); from the permanent ice cover of an Antarctic lake (168).
gram-positives with low G + C content (104), includ- Cultivation-independent molecular phylogenetic sur-
ing clostridial sublineages (106,110,111) and Bacil- veys have also helped to shape the present percep-
lus (102,106,110,111,119,147,152,157); Actinobacteria tion of diversity in Archaea. Cultivated members of
(102,105,116,117); the Cytophaga/Flexibacter/Bacter- the kingdom Euryarchaeota are physiologically diverse
oides group (102,104–106,110,111,150); Nitrospira and include halophiles, thermophiles, and methanogens,
(104,149); Planctomycetales (102,110,111,117,152); and whereas cultivated members of the kingdom Crenar-
Verrucomicrobia (104–106,111,157,159). chaeota are more homogeneous, consisting only of sulfur-
The abundance and activity of each bacterial division dependent extreme thermophiles, such as Pyrodiction
depends strongly on physicochemical soil properties. occultum and Thermoproteus tenax (91). Consequently, it
For instance, the assessment of bacterial community was assumed at one time that Crenarchaeota are restricted
structure along a vertical oxygen gradient in flooded paddy to extreme high-temperature habitats (92). However,
soil microcosms suggested that α- and β-Proteobacteria cultivation-independent retrieval of crenarchaeal SSU
colonized the oxic soil especially, whereas clostridial rDNA sequences has revealed this assumption to be
subgroups predominated in the anoxic soil zone (118). false. Crenarchaota are widely distributed, inhabiting not
Buckley and Schmidt (159) determined the relative only high-temperature habitats but also moderate- and
abundance of rRNA from Verrucomicrobia in soil microbial low-temperature environments, such as marine waters,
communities by quantitative dot blot hybridization using terrestrial soils, marine and freshwater sediments, and
a Verrucomicrobia-specific oligonucleotide probe. A total also living in association with metazoa (169,170). Pelagic
of 85 soil samples were taken to assess rRNA changes crenarchaeota may even be one of the ocean’s most
with soil depth and moisture content and in relation abundant cell types (171). Crenarchaeota sequences from
to plant community composition and soil management low-temperature habitats form several deeply branching
history over a period of 2 years. The authors concluded clades clearly distinct from the monophyletic group of
that Verrucomicrobia were significantly affected by cultured thermophilic Crenarchaeota. Representatives of
environmental characteristics that changed in relation these groups have still not been obtained in pure culture.
to time, soil history, and soil depth. A statistically Terrestrial soil environments in which SSU rDNA
significant amount of the variation in verrucomicrobial of nonthermophilic crenarchaeota have been detected
rRNA abundance could be correlated to changes in soil include boreal forest soil in northern Finland (109,172);
moisture. agricultural soils in Germany (173), Wisconsin (174,175),
A detailed description of the diversity and environmen- and Michigan (176); forest soil in Eastern Amazonia (111);
tal range of all major bacterial divisions is beyond the and Italian rice paddy soil (177). Crenarchaeal SSU rDNA
scope of this article, so the Holophaga/Acidobacterium retrieved from soils form two distinct clades named the
division will be used as an example. There are presently FFSB cluster and the Terrestrial cluster (176). Quantita-
only three cultured representatives of this main line tive dot blot hybridization with an SSU rRNA–targeted
 Grassland peat TM2 (Lower Saxony, Germany)
Acid mine drainage TRB82 (Iron Mountain, California, USA)
Acidobacterium capsulatum
Grassland peat TM29 (Lower Saxony, Germany)
Temperate forest soil Ep_T1.153 (United Kingdom)
Zinc-contaminated Maatheide soil 614 (Belgium)
Temperate forest soil Ep_T1.152 (United Kingdom)
Subtropical forest soil MC13 (Queensland, Australia)
Potato rhizosphere TLu-I-68 (North Rhine-Westphalia, Germany)
I Temperate forest soil Ep_T1.154 (United Kingdom)
Mature forest soil P28 (Amazonia, Brazil)
Freshwater sediment RB29 (South Carolina, USA)
Potato rhizosphere TLu-I-40 (North Rhine-Westph., Germany)
Mature forest soil M10 (Amazonia, Brazil)
Subtropical forest soil MC26 (Queensland, Australia)
Subtropical forest soil MC27 (Queensland, Australia)
Active pasture P15 (Amazonia, Brazil)
Pinyon-juniper forest soil (Arizona, USA)
II Zinc-contaminated Maatheide soil 613 (Belgium)
Hot spring sediment OPB3 (Yellowstone Park, USA)
Grassland peat TM1 (Lower Saxony, Germany)
Grassland peat DA052 (The Netherlands)
Mature forest soil M26 (Amazonia, Brazil)
Agricultural soil 11-14 (Bavaria, Germany)
Agricultural soil RB41 (Bavaria, Germany)
Agricultural soil 32-11 (Bavaria, Germany)
Potato rhizosphere TLu-I-35 (North Rh.-Westph., Germany)
Volcanic cinders S023 (Arizona, USA)
Volcanic cinders S024 (Arizona, USA)
Agricultural soil RB30 (Bavaria, Germany)
Active pasture P12 (Amazonia, Brazil)
Pinyon rhizosphere C105 (Arizona, USA)
Potato rhizosphere TLu-II-80 (N. Rh.-Westph., Germany)
Agricultural soil 11-24 (Bavaria, Germany)
III Pinyon rhizosphere S111 (Arizona, USA)
Active pasture P8 (Amazonia, Brazil)
Active pasture P21 (Amazonia, Brazil)
Agricultural soil kb2426 (Bavaria, Germany)
Pinyon rhizosphere C112 (Arizona, USA)
Mature forest soil M19 (Amazonia, Brazil)
Grassland peat DA023 (The Netherlands)
Agricultural soil RB40 (Bavaria, Germany)
Pinyon-juniper forest soil C028 (Arizona, USA)
Grassland soil DA008 (The Netherlands)
Agricultural soil RB27 (Bavaria, Germany)
IV Active pasture P16 (Amazonia, Brazil)
Agricultural soil mb1228 (Bavaria, Germany)
Pinyon rhizosphere S125 (Arizona, USA)
Agricultural soil 32-20 (Bavaria, Germany)
Potato rhizosphere TLu-I-22 (North Rhine-Westph., Germany)
V Holophaga foetida
Geothrix fermentans

0.10

Figure 1. Phylogenetic dendrogram of the Acidobacterium/Holophaga division. Cultivated

organisms are shown in bold. For each environmental sequence the habitat source, clone name,
and geographic region from which the sequence was obtained are given. Subdivision-level lineages
with bootstrap values of greater than 90% in 1,000 data resamplings are indicated by roman
numerals (I to V). The tree was rooted using the SSU rDNA of Escherichia coli as the outgroup
reference. The scale bar indicates 0.1 change per nucleotide.

534
 BIODIVERSITY IN SOILS: USE OF MOLECULAR METHODS FOR ITS CHARACTERIZATION 535

oligonucleotide probe specific for all known nonther- inferred from the phenotypes of the closest phylogenetic
mophilic crenarchaeotal sequences assigned as much as neighbors that are available for study in pure culture.
1.42±0.42% of the SSU rRNA extracted from agricul- Consequently, assays have been developed, which target
tural soils to this group (176). FISH analysis with a functional genes indicative of biogeochemically important
set of SSU rRNA–targeted oligonucleotide probes specific groups. Ideally these genes code for an enzyme that is both
for Archaea, Crenarchaeota, or nonthermophilic Crenar- universal and specific for the microbial group of interest,
chaeota showed that Crenarchaeota colonize terrestrial so that the presence of the phenotype is inherent from
plant roots at an unexpectedly high frequency (175). the recovery of the sequence. For example, the amoA
Archaeal SSU rDNA directly retrieved from Ital- gene is universal to autotrophic ammonium oxidizers, and
ian rice paddy soil included, besides sequences from possession of an amoA gene implies that the organism is
well-known methanogens, various novel lineages of Eur- an ammonium oxidizer (181,182). The sequence phylogeny
yarchaeota (108). One of these lineages, termed rice clus- of the functional gene also ideally corresponds closely to
ter I, formed a distinct clade within the phylogenetic the SSU rDNA–based phylogeny, that is the selective
radiation of the orders Methanosarcinales and Metha- pressure acting on the gene should be primarily neutral or
nomicrobiales and thus was considered a novel taxonomic negative and no interspecies gene transfer should occur.
group of methanogens at the family or order level. Mem- Some functional genes that have been used as
bers of rice cluster I also occur in close association with rice phylogenetic markers are listed in Table 1.
roots (108,178). A methanogenic phenotype for rice clus- Further promising genes have also been identified,
ter I species was further supported by parallel phylogenies for example the hydroxylamine oxidoreductase gene
of SSU rDNA and sequences encoding the methanogen- (hao) (226). Many, on the other hand, such as the
specific gene marker methyl coenzyme M reductase sub- plasmid genes encoding for 2,4-dichlorophenoxyacetic
unit A (179) (see section Analysis of Functional Genes). acid–degrading enzymes (227) and the nod genes of
Kim and coworkers (180) have also detected a novel clade rhizobia (228), have been shown through phylogenetic
of deeply branched Archaea in total DNA extracted from analysis to be very mobile across species, and the retrieval
various Japanese rice paddy soils. This novel group could of these sequences from nature would not allow a clear
be assigned neither to the Euryarchaeota nor the Crenar- deduction of the bacterial species from which they came.
chaeota. Several of the markers included in Table 1 also suffer to
Comparative sequence analysis of environmental SSU some extent from this drawback, but have nevertheless
rDNA has therefore provided, and will continue to provide, been applied for general phylogenetic deduction.
insight into microbial diversity. It does not, however, allow Examples of good functional phylogenetic markers are
conclusions about the metabolic capacities, phenotypes, or the pmoA and amoA genes. The phylogenies of these
ecological niches of microorganisms. This is especially true genes are excellent mirrors of SSU rDNA phylogeny,
for groups from which no representatives have yet been and allow identification to the species level (181–184).
cultured (20,142,169). The amoA gene, originally applied phylogenetically by
Rotthauwe and coworkers (181), codes for a subunit of
ANALYSIS OF FUNCTIONAL GENES ammonium monooxygenase, the enzyme that essentially
defines the group of autotrophic aerobic ammonia oxi-
Phenotypes of the organisms from which SSU rDNA dizers. Community analysis of amoA products amplified
sequences are retrieved cannot be known for certain, only from environmental DNA has been performed with a wide

Table 1. Functional Genes Useful as Phylogenetic Markers

Selected
Gene Enzyme Target Group References

pmoA Particulate methane Methane oxidizers 183–190

monooxygenase
mmoX Soluble methane Methane oxidizers 191–195
monooxygenase
amoA Ammonia monooxygenase Ammonia oxidizers 181,182,196–203
mxaF Methanol dehydrogenase Methylotrophs 186,204–207
nifH Dinitrogenase Nitrogen-fixers 208–214
reductase H
mcrA, mcrI Methyl coenzyme M Methanogens 179,215,216
reductase
rbcL Form I ribulose Autotrophs 217
bisphosphate (phytoplankton)
carboxylase/oxygenase
nirK, nirS Nitrite reductase Denitrifiers 218,219
nosZ Nitrous oxide reductase Denitrifiers 220–222
narG Nitrate reductase Denitrifiers 223
dsrA, dsrB Dissimilatory sulfite Sulfate reducers 224,225
reductase
536 BIODIVERSITY IN SOILS: USE OF MOLECULAR METHODS FOR ITS CHARACTERIZATION

combination of methods, including cloning followed by DGGE is the electrophoretic separation of sequence
sequencing (181,196–200), denaturing gradient gel elec- types in an increasing gradient of chemical denaturant
trophoresis (DGGE) separation and hybridization anal- (urea plus formamide). The manner in which double-
ysis (201), RFLP of cloned sequences (202), and terminal stranded DNA denatures in the gradient, and hence its
restriction fragment length polymorphism (T-RFLP) (203). electrophoretic mobility, is controlled by the arrangement
The pmoA assay targets subunit A of particulate methane of strong G−C bonds versus weak A−T bonds. TGGE is
monooxygenase, which is functionally and evolutionar- a variant in which the denaturation is controlled via a
ily related to the amoA gene (183). This assay has been temperature gradient rather than a chemical gradient.
applied to identify methane-oxidizing species in various The principles and technical aspects of these two methods
environments and enrichment cultures (185–188), and have been well reviewed (230–232). A similar technique
the discovery of novel pmoA sequences in forest soils is single-strand conformation polymorphism (SSCP), in
has suggested the existence of not-yet cultivated species which separation is achieved in a polyacrylamide gel
of methane-oxidizing bacteria (189,190). Unfortunately, based on the different secondary conformations of different
there are pitfalls in the use of functional gene markers. single-stranded DNA sequences (233).
The pmoA gene, for example, is neither completely univer- Details of the T-RFLP method are covered by recent
sal nor completely specific. Although particulate methane reviews (234,235). T-RFLP is a simplification of ARDRA,
monooxygenase was originally thought to be present in all a technique used primarily to identify single DNA
methane-oxidizing bacteria, a species has recently been phylotypes by their unique patterns of restriction digests
isolated, which seems to contain only the alternative, sol-
in an agarose gel. ARDRA can also give community-specific
uble form of the enzyme (229). Furthermore, the amoA
patterns for mixed PCR products (80,236), but these can
gene can also be amplified by pmoA primers. This is not
be extremely complex for environmental samples and
a problem if the sequences clearly fall within a known
are therefore of limited use in phylogenetic deduction
group such as the Nitrosospira. However, the interpreta-
or assessment of species richness. In T-RFLP, one
tion of novel or unusual sequences (189,190,200) can be
PCR primer is labeled with a fluorescent dye. After
problematic.
restriction digestion of the PCR product followed by
gel or capillary electrophoreses, the terminal fluorescent
MICROBIAL COMMUNITY FINGERPRINTING fragments can be detected using a laser (237). In this way
T-RFLP analysis detects a single-size fragment (T-RF)
Several molecular fingerprinting techniques are available for any given sequence type. Typical T-RFLP patterns
to characterize and compare microbial communities. These for SSU rDNA PCR products are shown in Figure 2. The
analyses begin with a PCR product (either SSU rDNA comparison of DNA extracted from BIOLOG plates versus
or a suitable functional gene) obtained from complex soil again demonstrates the greater complexity obtained
community DNA, which is then separated into a pattern with cultivation-independent analysis (238). Other, less-
of different phylotypes based on some properties of popular, size-based fingerprinting techniques include
the component DNA sequences. The most commonly length heterogeneity PCR (239) and RISA (111,240,241),
used methods are DGGE, temperature gradient gel in which the length variability of SSU rRNAs, or of the
electrophoresis (TGGE), and T-RFLP. intergenic region between the SSU and large subunit

Figure 2. Comparison of SSU rDNA-based

T-RFLP community fingerprint patterns obtained
from a potato-rhizosphere soil sample after
cultivation on a BiOLOG GN microtitre
plate (a) or obtained by a cultivation-independent
approach (b). The 404-bp T-RF corresponds
to α-Proteobacteria and the 489-bp T-RF
to β- and γ -Protoebacteria. The T-RFs
indicative of bacterial groups detectable
only by the cultivation-independent approach
correspond to various major bacterial groups
including the Holophaga/Acidobacterium cluster,
the Cytophaga/Flexibacter/Bacteroides division,
Planctomycetales, and δ-Proteobacteria.
 BIODIVERSITY IN SOILS: USE OF MOLECULAR METHODS FOR ITS CHARACTERIZATION 537

rRNA genes, are the respective criteria for the separation seem to give consistent and comparable community profile
of phylotypes. patterns.
The simplest analysis of the patterns obtained by com- In choosing a fingerprinting method, one must consider
munity fingerprinting methods is visual: that particular the aim of the study. Of the methods described, T-
species/phylotypes are detectable or not detectable in RFLP best lends itself to quantitative analyses, but is
a community (190,213,242). This is the primary goal of least useful for phylogenetic deduction. T-RFs of SSU
studies analyzing single samples, and is generally suffi- rDNA can be compared to the RDP to analyze possible
cient when comparing few samples (142,202). For example, phylogenetic affiliation (101). Unfortunately this is often
the disappearance or appearance of species with time not unequivocal, as many T-RFs are not species- or
has been followed with DGGE (186,243), and changes in even genus-specific (256). This problem can be partially
bacterial community structure along a vertical oxygen solved by comparing T-RFs instead with clone libraries
gradient were clearly demonstrated with T-RFLP (118). generated from the same sample (142,177), but this
Presence/absence data can also be used to calculate simple procedure may defeat the desired simplicity of T-RFLP
similarity coefficients between community profiles, which analysis. Alternatively, PCR assays specific to particular
can be used in cluster analysis (244). However, better microbial groups can reduce the total diversity of the
numerical quantification of the patterns facilitates more T-RFLP profiles. Assays have been developed for the
intensive analysis. Quantification is easiest and most reli- functional genes nifH (212), amoA (203), nirS (250), and
able with T-RFLP, where the intensity of the fluorescence mcrA (179). Group-specific SSU rDNA primers may also be
signal is used as a measure of the abundance of any prod- applied. Nevertheless, a major advantage of DGGE/TGGE
uct relative to the others, and a threshold signal/noise and SSCP is still that individual bands can be excised
level can be set to score the presence or absence of a from the gel and sequenced for direct identification.
peak (236,245). Quantification in DGGE or TGGE is gen- A problem with DGGE/TGGE is that they are often
erally accomplished by image analysis software, and may too sensitive for the determination of total community
or may not include a densitometric comparison of band fingerprints, since nearly identical sequences can give
strengths (246–248). separate bands. The situation is especially problematic
Relative abundances of T-RFs are often compared for SSCP, where a single sequence can have multiple
across communities, to conclude that particular T-RFs secondary conformations. The nonspecificity of T-RFs
comprise a greater or lesser proportion of the total product might therefore be considered an advantage when dealing
in one sample compared to another (118,177,219). Of with complex communities, since the overall redundancy
course, caution must be exercised in making conclusions of the fingerprint is decreased and broader OTUs are
because of well-known PCR biases, as well as the detected.
fact that data are relative and therefore all cross-
correlated. If sufficient replicate data are available, a ANALYSIS OF MICROBIAL ACTIVITY AT THE
multivariate statistical analysis can determine whether FUNCTIONAL LEVEL
the patterns observed significantly differ across samples
or treatments (236). This analysis also takes into account PCR-based retrieval of functional genes gives insight
the cross-correlation of the various T-RFs and allows into the species composition of microbial communities,
the determination of which T-RFs contribute most to the and their potential to carry out certain functions, but
overall variability of the patterns (236). it does not provide information on in situ activity. This
Phylotype richness–the total number of DGGE bands gap can be bridged by investigating gene expression
or T-RFLP peaks–is also often reported (142,235,242,249, at the transcription level. The hybridization of poly-
250). Combining this with relative abundance data then A mRNAs to oligo(dT) probes allows direct extraction
allows the calculation of classical ecological diversity and purification of eukaryotic mRNA from complex
indices such as Shannon’s index and Simpson’s index. matrices and has been performed for soils (257–259).
This frequently give useful results (251–253), although However, bacterial transcripts lack polyA tails and are not
because gel bands or T-RFs do not represent standard detectable with these methods. In addition to the general
phylogenetic units, the resolution may not be fine enough methodological problems involved in the extraction of
to distinguish among very similar samples (245). Cluster total RNA from soils (as mentioned earlier), the transient
analysis using matrices of pairwise correlation coefficients character of prokaryotic mRNA transcripts and their low
or any of a variety of similarity coefficients (244) is cellular abundance relative to rRNA (260) account for the
a popular and powerful method to determine whether paucity of papers reporting the detection of prokaryotic
certain classes of communities are distinguishable within mRNA transcripts from soils (72,73,260–264) and aquifer
a large group of samples (222,235,246–248,250–252,254). sediment (265).
Cluster analysis of various marine bacterioplankton Because of the present interest in biotechnology,
communities gave similar results with T-RFLP and most of these studies have focused on the detection
DGGE data, although the T-RFLP detected more total of mRNA transcripts for genes coding biodegradation
phylotypes (255). T-RFLP analysis of a variety of soil enzymes. They have included: (1) the detection of mRNA
samples also gave similar community clustering patterns transcripts for naphthalene dioxygenase and mercury
as those obtained from analysis of SSU rDNA gene clone resistance genes by Northern blot hybridization (73),
libraries (245). Different fingerprinting methods therefore (2) quantification of mRNA transcripts of naphthalene
538 BIODIVERSITY IN SOILS: USE OF MOLECULAR METHODS FOR ITS CHARACTERIZATION

dioxygenase (261,262) and soluble methane monooxy- DNA fractions. The major conclusion of this study was that
genase (265) using an RNase protection assay, and bacterial functional redundancy increases with regrowth
(3) reverse transcriptase-PCR amplification of mRNA of plant communities.
transcripts for β-glucuronidase (264). These methods of Stable isotope probing takes this analysis a step further
mRNA analysis require a priori information about the and directly links genomic DNA with the metabolism
gene sequences of interest to design specific DNA probes of defined microbial groups. With the BrdU method
(hybridization assays) or primers for mRNA detection and the soil can be manipulated to favor the growth of
measurement. In this way they are simply an exten- different bacterial groups (e.g., particular substrates can
sion of the previously described techniques for detection be added) and the actively growing groups identified,
and characterization of functional and catabolic DNA but with the stable isotope method the incorporation of
sequences (263). a substrate can be followed directly. This technique is
Fleming and coworkers (260) combined the differential based on the capability of 13 C-DNA, produced during
display (DD) technique (266,267) and the RNA arbitrarily growth on a 13 C-enriched carbon source, to be resolved
primed PCR method (268,269) to detect cryptic or from 12 C-DNA by density-gradient centrifugation. Using
unknown RNA sequences transcribed under in situ this technique, methanol-utilizing microorganisms were
conditions in soils. An arbitrary primer for the reverse investigated through the addition of 13 CH3 OH to a
transcriptase step and the same arbitrary primer in forest soil (206). Phylogenetic analysis of SSU rRNA
conjunction with a Shine-Dalgarno (SD) primer in the PCR genes recovered from the purified 13 C-labeled DNA
step were used to monitor differential bacterial mRNA fraction identified CH3 OH-utilizing bacteria from two
transcription between control and toluene-contaminated distinct lineages, the α-subclass of the Proteobacteria and
soil microcosms. The SD primer was used to amplify, the Acidobacterium division. The conclusion that active
by PCR, a wide range of prokaryotic mRNA sequences. methylotrophs belong to the α-subclass of Proteobacteria
The use of this arbitrary primer in combination with was also supported by a parallel analysis of mxaF
the SD primer for DD resulted in highly reproducible sequences (see Table 1).
RNA fingerprints. However, one major shortcoming of the The BrdU and stable isotope probing techniques each
method was a high percentage of false positive bands. provide cultivation-independent means of investigating
Of the putative clones screened, only one in 12 was the effect of environmental conditions, such as tempera-
confirmed to be unique and differentially expressed in ture, water potential, pH, and substrate availability on the
the toluene-induced soil microcosms. Another limitation identity of active microorganisms, and thereby allow anal-
of this technique is the short lifetime of mRNA. Half- ysis of functionally relevant ecological diversity. However,
lives of mRNA molecules can be as long as 50 minutes, both techniques have methodological shortcomings that
but typically range between 0.5 and 2 minutes. Thus, a necessitate care in their application and in the interpreta-
large fraction of the prokaryotic mRNA pool may not be tion of the data obtained. A major limitation of the BrdU
accessible for study, even after induction. Nevertheless, strategy is the range of organisms capable of BrdU uptake
the DD technique has great potential for studying gene and incorporation. In one study, only two of four different
expression in soil environments, and potentially allows bacterial strains tested incorporated BrdU (271). One of
the detection of novel and differentially expressed gene these belonged to the α-Proteobacteria and the other to the
sequences. γ -Proteobacteria, whereas failing to assimilate BrdU were
A promising alternative to mRNA-based techniques a flavobacterium and a gram-positive bacterium. Stable
for analysis of microbial communities is the use of isotope probing also has limitations that may render it
biomarkers, such as bromodeoxyuridine (BrdU) (270,271), unsuitable in certain situations. For instance, simulta-
or substrates labeled with stable isotopes (206), which neous growth of target microorganisms on an unlabeled
are incorporated in the genomic DNA of active microbial (12 C) substrate will dilute the proportion of 13 C that is
populations only (206,240,270). Labeled DNA can be incorporated into their DNA. This will reduce the propor-
isolated from the target group of microorganisms and then tion of DNA that becomes isotopically labeled, making it
characterized taxonomically by gene probing and sequence more difficult to identify the microorganisms that are the
analysis. primary users of a labeled substrate (206).
Culture-independent identification of soil microorgan- Other innovative cultivation-independent methods
isms that grow in response to specific stimuli can be have been developed for linking the functional activities
performed using BrdU (270). After incubation of a soil of microorganisms with their taxonomic identity. These
with BrdU, total DNA is extracted and the newly synthe- include the combination of FISH with microautoradiogra-
sized DNA isolated by immunocapture of the BrdU-labeled phy (272,273) and the coupling of radio- or stable-isotope
DNA. This approach was used to measure bacterial analyses with characterization of microbial phospho-
functional redundancy in soils along a reclamation- lipid ester-linked fatty acids (PLFAs) (189,274,275). For
revegetation gradient (240). Samples were taken on a example, bacteria oxidizing atmospheric methane in soil
transect from a mine site to a soil under a new-growth samples from Greenland, Denmark, the United States, and
forest, and from an adjacent pristine forest. The diversity Brazil were characterized by incubation of the soils under
14
of bacterial groups responding to the substrates L-serine, CH4 and subsequent analysis of the radioactively labeled
L-threonine, sodium citrate, and α-lactose hydrate were PLFAs (189,275), and in parallel by analysis of methane
determined by incubating soils with BrdU plus the various monooxygenase gene libraries (189). The 14 C-PLFA fin-
substrates, and then performing RISA of the BrdU-labeled gerprints of the soil methanotrophs were clearly different
 BIODIVERSITY IN SOILS: USE OF MOLECULAR METHODS FOR ITS CHARACTERIZATION 539

from the PFLA profiles of the type I γ -proteobacterial the Cytophaga/Flexibacter/Bacteroides group, and the
methanotrophic bacteria, but similar to PLFA profiles of Holophaga/Acidobacterium cluster. Members of these
the α-proteobacterial type II methanotrophs. These 14 C- phyla are typical inhabitants of agricultural soil. Complete
PLFA fingerprint results corresponded well to the results sequence analysis of one BAC clone insert identified
of monooxygenase sequence analyses. Phylogenetic treeing various open reading frames, including a gene cluster
analysis identified a novel cluster of pmoA sequences possi- similar to the phosphate transport cluster (pstCAB-phoU)
bly belonging to an unknown group of methane-oxidizing of E. coli (283). This finding demonstrated the potential
bacteria. These sequences were more closely related to of metagenomic DNA analysis to characterize complete,
those of type II α-Proteobacteria methanotrophs than to intact operons.
those of the type I γ -Proteobacteria. Screening the two BAC libraries identified clones
expressing DNase, antibacterial compounds, lipase, and
amylase of metagenomic DNA. This demonstrated that
METAGENOME ANALYSIS
BAC libraries contain heterologous DNA sequences that
can be expressed in E. coli at detectable levels. The anal-
Cultivation-independent surveys of microbial communi-
ysis of metagenomic BAC libraries has the potential not
ties based on single gene phylogenies provide information
only to provide insight into the genomic potential and eco-
on the diversity of microorganisms at the phylogenetic
logical role of soil microorganisms–cultivated or not–but
or functional level, but allow little inference about the
also to provide enzymes with biotechnological applications.
detailed physiology or biochemistry of these organisms.
This was demonstrated by comparative sequence analysis
Recent improvements in molecular analytical techniques
and expression in E. coli of biotin biosynthesis operons
and bioinformation sciences have provided methods suit-
recovered from various soil enrichment cultures by direct
able for the analysis of the metagenome–the total gene
cloning of total DNA (284).
pool of a complex microbial community (276). Metage-
nomic analysis begins with the cloning of large fragments
of DNA extracted directly from the microbial commu- FINAL COMMENTS
nity. Analysis of these cloned fragments can provide a
link between phylogenetic and functional information and This article has attempted to provide an overview
allow for the characterization of total gene operons isolated of the current cultivation-independent state-of-the-art
directly from an environment. Pioneered by DeLong and methodological toolbox used to describe phylogenetic and
coworkers, this new approach has been applied to charac- functional diversity, structural composition and dynamics,
terize genome fragments of uncultivated nonthermophilic and metabolic activities of soil microbial communities.
marine crenarchaeota (277), including Cenarchaeum sym- New methods developed during the last few years or
biosum (278,279), a psychrophilic archaeon that lives in currently under development might soon be added to
specific association with the marine sponge Axinella mex- those described here. One candidate is quantitative PCR,
icana (280). which can be used to determine the exact number of
More recently, bacterial artificial chromosome libraries functional or phylogenetic target genes in a sample (e.g.,
(BACs) from soil metagenomes have been constructed (56). the TaqMan technique (285)). However, development of a
BACs are modified plasmids that contain an origin bias-free application of quantitative PCR to environmental
of replication derived from the E. coli F factor. The samples is still not complete (286,287). Another technique
replication of the BAC vector is strictly controlled, keeping with great potential is the hybridization of fluorescently
the replicon at one or two copies per cell. BAC vectors can labeled total RNA to microarrays of immobilized rRNA-
stably maintain inserts as large as 600 kb (281). It has targeted oligonucleotide probes. This technique may soon
also been suggested that BAC vectors display a low level allow a rapid and detailed assessment of the diversity and
of chimerism as a result of two genes carried on the activity of microbial communities (288,289).
plasmid (parA and parB) (282). Various aspects of soil microbial diversity have been
A critical step of soil metagenome analysis is the covered by other reviews, which we recommend because
extraction of high molecular weight DNA suitable for the they nicely complement the present article. These reviews
construction of BAC libraries (see section Extraction of have focused on the exploration of bacterial diversity
Total Nucleic Acids from Soil). The total fragment mixture and functioning in relation to bioindication in agroecosys-
of DNA extracted from an agricultural soil was cloned into tems (290,291) and on the impact of chemical pollutants
BAC library SL1, which had clone inserts ranging in size on soil microbial diversity (290,292,293). Finally, an argu-
from 10 to 60 kb, but mostly between 20 and 30 kb (56). ment in favor of adopting a natural species concept for
DNA fragments greater than 40 kb were then selected prokaryotes (294), that is, one that takes the ecological
from the same extract by preparative gel electrophoresis niche into consideration, is also recommended for further
prior to construction of BAC library SL2. This SL2 library reading.
had clone inserts ranging between 10 and 80 kb in size.
The average insert size was 44.5 kb, and more than 60% BIBLIOGRAPHY
of the inserts were larger than 40 kb (56).
A phylogenetic survey of BAC library SL1 identified 1. M. Alexander, Introduction to Soil Microbiology, 2nd ed.,
seven genome fragments containing SSU rRNA gene John Wiley & Sons, New York, 1977.
sequences. These belonged to four different bacterial 2. T. Hattori and R. Hattori, Microbial Life in the Soil, Marcel
phyla: the Proteobacteria, low G + C gram-positives, Dekker, New York, 1973.
540 BIODIVERSITY IN SOILS: USE OF MOLECULAR METHODS FOR ITS CHARACTERIZATION

3. M. Begon, J. L. Harper, and C. R. Townsend, Ecology, 3rd 31. J. Porter, R. Pickup, and C. Edwards, Soil Biol. Biochem.
ed., Blackwell, Oxford, U.K., 1996. 29, 91–100 (1997).
4. P. J. Harris, in K. Ritz, J. Dighton, and K. E. Giller, eds., 32. G. F. Duarte et al., J. Microbiol. Methods 32, 21–29 (1998).
Beyond the Biomass, John Wiley & Sons, Chichester, U.K., 33. L. G. Leff, J. R. Dana, J. V. McArthur, and L. J. Shimkets,
1994, pp. 239–246. Appl. Environ. Microbiol. 61, 1141–1143 (1995).
5. R. C. Foster, Biol. Fertil. Soils 6, 189–203, (1988). 34. G. S. Sayler and A. C. Layton, Annu. Rev. Microbiol. 44,
6. H. Bolton, Jr., J. K. Frederickson, and L. F. Elliot, in 625–648 (1990).
F. B. Metting, eds., Soil Microbial Ecology, Marcel Dekker, 35. J. Kozdrój and J. D. van Elsas, Biol. Fertil. Soilss 31,
New York, 1993, pp. 27–63. 372–378 (2000).
7. V. Torsvik, J. Goksøyr, and F. L. Daae, Appl. Environ. 36. A. Prieme, J. I. B. Sitaula, A. K. Klemedtsson, and L. R.
Microbiol. 56, 782–787 (1990). Bakken, FEMS Microbiol. Ecol. 21, 59–68 (1996).
8. V. Torsvik et al., in K. Ritz, J. Dighton, and K. E. Giller, 37. M. Krsek and E. M. H. Wellington, J. Microbiol. Methods
eds., Beyond the Biomass, John Wiley & Sons, Chichester, 39, 1–16 (1999).
U.K., 1994, pp. 39–48. 38. P. A. Rochelle, J. C. Fry, R. J. Parkes, and A. J. Weightman,
9. B. S. Griffiths, K. Ritz, and L. A. Glover, Microb. Ecol. 31, FEMS Microbiol. Lett. 100, 59–65 (1992).
269–280 (1996). 39. K. Smalla et al., J. Appl. Bacteriol. 74, 78–85 (1993).
10. R. K. Peet, Annu. Rev. Ecol. Syst. 5, 285–307 (1974). 40. Y. L. Tsai and B. H. Olson, Appl. Environ. Microbiol. 57,
11. J. L. Harper and D. L. Hawksworth, in D. L. Hawksworth, 1070–1074 (1991).
eds., Biodiversity Measurement and Estimation, Chapman 41. M. I. Moré et al., Appl. Environ. Microbiol. 60, 1572–1580
& Hall, London, U.K., 1996, pp. 5–12. (1994).
12. T. M. Embley, R. P. Hirt, and D. M. Williams, in D. L. 42. T. Volossiouk, E. J. Robb, and R. N. Nazar, Appl. Environ.
Hawksworth, eds., Biodiversity Measurement and Estima- Microbiol. 61, 3972–3976 (1995).
tion, Chapman & Hall, London, U.K., 1996, pp. 21–33. 43. J. Zhou, M. Ann Bruns, and J. M. Tiedje, Appl. Environ.
13. R. L. Ferguson, E. N. Buckley, and A. V. Palumbo, Appl. Microbiol. 62, 316–322 (1996).
Environ. Microbiol. 47, 49–55 (1984). 44. D. W. Cullen and P. R. Hirsch, Soil Biol. Biochem. 30,
14. L. R. Bakken and R. A. Olsen, Microb. Ecol. 13, 103–114 983–993 (1998).
(1987). 45. C. R. Kuske et al., Appl. Environ. Microbiol. 64, 2463–2472
15. R. Amann, W. Ludwig, and K.-H. Schleifer, Microbiol. Rev. (1998).
59, 143–169 (1995). 46. D. P. Chandler, B. L. Schuck, F. J. Brockman, and C. J.
16. R. Sørheim, V. L. Torsvik, and J. Goksøyr, Microb. Ecol. 17, Bruckner-Lea, Talanta 49, 969–983 (1999).
181–192 (1989). 47. Å. Frostegård et al., Appl. Environ. Microbiol. 65, 5,409–
17. V. Boivin-Jahns et al., Appl. Environ. Microbiol. 61, 5,420 (1999).
3400–3406 (1995). 48. D. N. Miller, J. E. Bryant, E. L. Madsen, and W. C. Ghiorse,
18. D. E. Caldwell, G. M. Wolfaardts, D. R. Korber, and J. R. Appl. Environ. Microbiol. 65, 4715–4724 (1999).
Lawrence, in J. H. Hurst et al., eds., Manual of Environ- 49. M. G. Lorenz and W. Wackernagel, Microbiol. Rev. 58,
mental Microbiology, ASM Press, Washington, D.C., 1997, 563–602 (1994).
pp. 79–90. 50. E. Paget, L. J. Monrozier, and P. Simonet, FEMS Microbiol.
19. D. P. Chandler et al., FEMS Microbiol. Ecol. 23, 131–144 Lett. 97, 31–40 (1992).
(1997). 51. A. V. Ogram et al., Appl. Environ. Microbiol. 60, 393–396
20. W. Liesack et al., in J. D. van Elsas, J. T. Trevors, and (1994).
E. M. H. Wellington, eds., Modern Soil Microbiology, Marcel 52. C. Crecchio, and G. Stotzky, Soil Biol. Biochem. 30,
Dekker, New York, 1997, pp. 375–439. 1061–1067 (1998).
21. R. Wagner, Arch. Microbiol. 161, 100–106 (1994). 53. R. A. Olsen and L. R. Bakken, Microb. Ecol. 13, 59–74
22. C. C. Tien, C. C. Chao, and W. L. Chao, J. Appl. Microbiol. (1987).
86, 937–943 (1999). 54. L. Ranjard et al., Eur. J. Soil Biol. 34, 89–97 (1998).
23. W. E. Holben, J. K. Jansson, B. K. Chelm, and J. M. Tiedje, 55. J. D. van Elsas, G. F. Duarte, A. S. Rosado, and K. Smalla,
Appl. Environ. Microbiol. 54, 703–710 (1988). J. Microbiol. Methods 32, 133–154 (1998).
24. A. Ogram, G. S. Sayler, and T. Barkay, J. Microbiol. Meth- 56. M. R. Rondon et al., Appl. Environ. Microbiol. 66, 2,541–
ods 7, 57–66 (1987). 2,547 (2000).
25. R. J. Steffan, J. Goksøyr, A. K. Bej, and R. M. Atlas, Appl. 57. R. J. Watson and B. Blackwell, Can. J. Microbiol. 46,
Environ. Microbiol. 54, 2908–2915 (1988). 633–642 (2000).
26. S. D. Pillai et al., Appl. Environ. Microbiol. 57, 2283–2286 58. J. B. Herrick, E. L. Madsen, C. A. Batt, and W. C. Ghiorse,
(1991). Appl. Environ. Microbiol. 59, 687–694 (1993).
27. D. W. Hopkins, S. J. MacNaughton, and A. G. O’Donnell, 59. M. Berthelet, L. G. Whyte, and C. W. Greer, FEMS Micro-
Soil Biol. Biochem. 23, 217–225 (1991). biol. Lett. 138, 17–22 (1996).
28. A. Faegri, V. L. Torsvik, and J. Goksøyr, Soil Biol. Biochem. 60. Y. L. Tsai and B. H. Olson, Appl. Environ. Microbiol. 58,
9, 105–112 (1977). 2292–2295 (1992).
29. W. E. Holben, K. Noto, T. Sumino, and Y. Suwa, Appl. 61. C. R. Jackson et al., Appl. Environ. Microbiol. 63, 4,993–
Environ. Microbiol. 64, 2528–2532 (1998). 4,995 (1997).
30. C. S. Jacobsen and O. F. Rasmussen, Appl. Environ. Micro- 62. F. Martin-Laurent et al., Appl. Environ. Microbiol. 67,
biol. 58, 2458–2462 (1992). 2354–2359 (2001).
 BIODIVERSITY IN SOILS: USE OF MOLECULAR METHODS FOR ITS CHARACTERIZATION 541

63. W. Liesack and E. Stackebrandt, J. Bacteriol. 174, 5,072– 95. W. G. Weisburg, S. M. Barns, D. A. Pelletier, and D. J.
5,078 (1992). Lane, J. Bacteriol. 173, 697–703 (1991).
64. C. S. Jacobsen Appl. Environ. Microbiol. 61, 3347–3352 96. E. Stackebrandt and W. Liesack, in C. Kessler, eds.,
(1995). Non-Radioactive Labeling and Detection of Biomolecules,
65. D. J. Lane, in E. Stackebrandt and M. Goodfellow, eds., Springer-Verlag, New York, 1993, pp. 232–239.
Nucleic Acid Techniques in Bacterial Systematics, John 97. R. Devereux and S. G. Willis, in A. D. L. Akkermans,
Wiley & Sons, Chichester, U.K., 1991, pp. 115–175. J. D. van Elsas, and F. J. de Bruijn, eds., Molecular Micro-
bial Ecology Manual, Kluwer Academic Publishers, Dor-
66. D. P. Chandler et al., Appl. Environ. Microbiol. 66, 3,438–
drecht, Netherlands, 1995.
3,445 (2000).
98. J. R. Marchesi et al., Appl. Environ. Microbiol. 64, 795–799
67. A. D. L. Akkermans, J. D. van Elsas, and F. J. de Bruijn,
(1998).
Molecular Microbial Ecology Manual, Kluwer Academic
Publishers, Dordrecht, Netherlands, 1995. 99. S. M. Barns, R. E. Fundyga, M. W. Jeffries, and N. R. Pace,
Proc. Natl. Acad. Sci. U.S.A. 91, 1609–1613 (1994).
68. W. E. Holben, in C. J. Hurst et al., eds., Manual of Environ-
mental Microbiology, ASM Press, Washington, D.C., 1997, 100. R. Großkopf, P. H. Janssen, and W. Liesack, Appl. Environ.
Microbiol. 64, 960–969 (1998).
pp. 431–436.
101. B. L. Maidak et al., Nucleic Acids Res. 29, 173–174 (2001).
69. D. D. Blumberg, Methods Enzymol. 152, 20–24 (1987).
102. C. R. Kuske, S. M. Barns, and J. D. Busch, Appl. Environ.
70. D. Hahn, R. Kester, M. J. C. Starrenburg, and A. D. L.
Microbiol. 63, 3614–3621 (1997).
Akkermans, Arch. Microbiol. 154, 329–335 (1990).
103. W. Ludwig et al., FEMS Microbiol. Lett. 153, 181–190
71. M. A. Moran, V. L. Torsvik, T. Torsvik, and R. E. Holben,
(1997).
Appl. Environ. Microbiol. 59, 915–918 (1993).
104. J. Dunbar et al., Appl. Environ. Microbiol. 65, 1662–1669
72. S. Selenska and W. Klingmüller, Microb. Releases 1, 41–46 (1999).
(1992).
105. A. E. McCaig, L. A. Glover, and J. I. Prosser, Appl. Environ.
73. Y.-L. Tsai, M. J. Park, and B. H. Olson, Appl. Environ. Microbiol. 65, 1721–1730 (1999).
Microbiol. 57, 765–768 (1991).
106. U. Hengstmann, K.-J. Chin, P. H. Janssen, and W. Liesack,
74. A. Felske, B. Engelen, U. Nübel, and H. Backhaus, Appl. Appl. Environ. Microbiol. 65, 5050–5058 (1999).
Environ. Microbiol. 62, 4162–4167 (1996).
107. S. M. Barns, C. F. Delwiche, J. D. Palmer, and N. R. Pace,
75. R. I. Griffiths, A. S. Whiteley, A. G. O’Donnell, and M. J. Proc. Natl. Acad. Sci. U.S.A. 93, 9188–9193 (1996).
Bailey, Appl. Environ. Microbiol. 66, 5488–5491 (2000). 108. R. Großkopf, S. Stubner, and W. Liesack, Appl. Environ.
76. K. J. Purdy, T. M. Embley, S. Takii, and D. B. Nedwell, Microbiol. 64, 4983–4989 (1998).
Appl. Environ. Microbiol. 62, 3905–3907 (1996). 109. G. Jurgens and A. Saano, FEMS Microbiol. Ecol. 29,
77. W. E. Holben and D. Harris, Mol. Ecol. 4, 627–631 (1995). 205–213 (1999).
78. M. Goodfellow and A. G. O’Donnell, in M. Goodfellow and 110. J. Borneman et al., Appl. Environ. Microbiol. 62, 1,935–
A. G. O’Donnell, eds., Handbook of New Bacterial Systemat- 1,943 (1996).
ics, Academic Press, San Diego, Calif., 1993, pp. 3–56. 111. J. Borneman and E. W. Triplett, Appl. Environ. Microbiol.
79. P. Vandamme et al., Microbiol. Rev. 60, 407–438 (1996). 63, 2647–2653 (1997).
80. J. M. Tiedje et al., Appl. Soil Ecol. 13, 109–122 (1999). 112. J. M. Clark, Nucleic Acids Res. 16, 9677–9686 (1988).
81. K. Nüsslein and J. M. Tiedje, Appl. Environ. Microbiol. 64, 113. J. Felsenstein, Cladistics 5, 164–166 (1989).
1283–1289 (1998). 114. D. L. Swofford, G. J. Olsen, P. J. Waddell, and D. M. Hillis,
82. K. Nüsslein and J. M. Tiedje, Appl. Environ. Microbiol. 65, in D. M. Hillis, C. Moritz, and B. K. Mable, eds., Molecular
3622–3626 (1999). Systematics, Sinauer Associates, Sunderland, Mass., 1996,
pp. 407–514.
83. J. A. Apajalahti et al., Appl. Environ. Microbiol. 64,
4084–4088 (1998). 115. M. Steel and D. Penny, Mol. Biol. Evol. 17, 839–850 (2000).
84. J. W. Santo Domingo et al., Mol. Ecol. 7, 761–767 (1998). 116. A. Felske et al., Microbiology 143, 2983–2989 (1997).
85. W. E. Holben et al., in R. Guerrero and C. Pedros-Alio, 117. B. Nogales, E. R. B. Moore, W.-R. Abraham, and K. N. Tim-
eds., Trends in Microbial Ecology, Spanish Society for mis, Environ. Microbiol. 1, 199–212 (1999).
Microbiology, 1993, pp. 367–370. 118. H. Lüdemann, I. Arth, and W. Liesack, Appl. Environ.
Microbiol. 66, 754–762 (2000).
86. C. R. Woese, Microbiol. Rev. 51, 221–271 (1987).
119. A. Felske et al., Appl. Environ. Microbiol. 66, 3998–4003
87. G. J. Olsen and C. R. Woese, FASEB J. 7, 113–123 (1993).
(2000).
88. G. J. Olsen, C. R. Woese, and R. Overbeek, J. Bacteriol. 176,
120. K. L. Josephson, C. P. Gerba, and T. L. Pepper, Appl. Envi-
1–6 (1994).
ron. Microbiol. 59, 3513–3515 (1993).
89. W. Ludwig and K. H. Schleifer, FEMS Microbiol. Rev. 15,
121. M. G. Lorenz and W. Wackernagel, Appl. Environ. Micro-
155–173 (1994).
biol. 53, 2948–2952 (1987).
90. C. R. Woese, Proc. Natl. Acad. Sci. U.S.A. 97, 8392–8396
122. M. J. L. Coolen and J. Overmann, Appl. Environ. Microbiol.
(2000). 64, 4513–4521 (1998).
91. N. R. Pace, Science 276, 734–740 (1997). 123. J. G. Kramer and F. L. Singleton, Appl. Environ. Microbiol.
92. C. R. Woese, O. Kandler, and M. L. Wheelis, Proc. Natl. 59, 2430–2436 (1993).
Acad. Sci. U.S.A. 87, 4576–4579 (1990). 124. P. F. Kemp, S. Lee, and J. LaRoche, Appl. Environ. Micro-
93. G. J. Olsen and C. R. Woese, Can. J. Microbiol. 35, 119–123 biol. 59, 2594–2601 (1993).
(1989). 125. S. Lee and P. F. Kemp, Limnol. Oceanogr. 39, 869–879
94. W. Ludwig et al., Electrophoresis 19, 554–568 (1998). (1994).
542 BIODIVERSITY IN SOILS: USE OF MOLECULAR METHODS FOR ITS CHARACTERIZATION

126. H. Bremer and P. P. Dennis, in F. C. Neidhart et al., 157. A. Macrae, D. L. Rimmer, and A. G. O’Donnell, Antonie Van
eds., Escherichia coli and Salmonella: Cellular and Leeuwenhoek 78, 13–21 (2000).
Molecular Biology, ASM Press, Washington, D.C., 1996, 158. S. M. Barns, S. L. Takala, and C. R. Kuske, Appl. Environ.
pp. 1553–1569. Microbiol. 65, 1731–1737 (1999).
127. C. L. Moyer, F. C. Dobbs, and D. M. Karl, Appl. Environ. 159. D. H. Buckley and T. M. Schmidt, FEMS Microbiol. Ecol.
Microbiol. 60, 871–879 (1994). 35, 105–112 (2001).
128. L. Marilley, G. Vogt, M. Blanc, and M. Aragno, Plant Soil 160. N. Kishimoto, Y. Kosako, and T. Tano, Curr. Microbiol. 22,
198, 219–224 (1998). 1–8 (1991).
129. E. Stackebrandt and B. M. Goebel, Int. J. Syst. Bacteriol. 161. W. Liesack, F. Bak, J.-U. Kreft, and E. Stackebrandt, Arch.
44, 846–849 (1994). Microbiol. 162, 85–90 (1994).
130. F. v. Wintzingerode, U. B. Göbel, and E. Stackebrandt, 162. J. D. Coates, D. J. Ellis, C. V. Gaw, and D. R. Lovley, Int. J.
FEMS Microbiol. Rev. 21, 213–229 (1997). Syst. Bacteriol. 49, 1615–1622 (1999).
131. M. T. Suzuki and S. J. Giovannoni, Appl. Environ. Micro- 163. H. Rheims, F. A. Rainey, and E. Stackebrandt, J. Ind.
biol. 62, 625–630 (1996). Microbiol. 17, 159–169 (1996).
132. D. P. Chandler, J. K. Fredrickson, and F. J. Brockman, Mol. 164. K. J. Edwards et al., Geomicrobiol. J. 16, 155–179 (1999).
Ecol. 6, 475–482 (1997). 165. J. A. Fuhrman and A. A. Davis, Mar. Ecol.-Prog. Ser. 150,
133. M. C. Hansen, T. Tolker-Nielsen, M. Givskov, and S. Molin, 275–285 (1997).
FEMS Microbiol. Ecol. 26, 141–149 (1998). 166. T. M. LaPara, C. H. Nakatsu, L. Pantea, and J. E. Alleman,
134. M. F. Polz and C. M. Cavanaugh, Appl. Environ. Microbiol. Appl. Environ. Microbiol. 66, 3951–3959 (2000).
64, 3724–3730 (1998). 167. S. Spring, R. Schulze, J. Overmann, and K.-H. Schleifer,
135. A. G. C. L. Speksnijder et al., Appl. Environ. Microbiol. 67, FEMS Microbiol. Rev. 24, 573–590 (2000).
469–472 (2001). 168. D. A. Gordon, J. Priscu, and S. Giovannoni, Microb. Ecol. 39,
136. X. Qiu et al., Appl. Environ. Microbiol. 67, 880–887 (2001). 197–202 (2000).
137. M. A. Tanner, B. M. Goebel, M. A. Dojka, and N. R. Pace, 169. K. L. Hershberger et al., Nature 384, 420 (1996).
Appl. Environ. Microbiol. 64, 3110–3113 (1998). 170. E. F. DeLong, Curr. Opin. Genet. Dev. 8, 649–654 (1998).
138. W. Liesack, H. Weyland, and E. Stackebrandt, Microb. Ecol. 171. M. B. Karner, E. F. DeLong, and D. M. Karl, Nature 409,
21, 191–198 (1991). 507–510 (2001).
139. E. D. Kopczynski, M. M. Bateson, and D. M. Ward, Appl. 172. G. Jurgens, K. Lindström, and A. Saano, Appl. Environ.
Environ. Microbiol. 60, 746–748 (1994). Microbiol. 63, 803–805 (1997).
140. G. C. Wang and Y. Wang, Microbiology 142, 1107–1114 173. R.-A. Sandaa, Ø. Enger, and V. Torsvik, Appl. Environ.
(1996). Microbiol. 65, 3293–3297 (1999).
141. G. C. Wang and Y. Wang, Appl. Environ. Microbiol. 63, 174. S. B. Bintrim et al., Proc. Natl. Acad. Sci. USA 94, 277–282
4645–4650 (1997). (1997).
142. M. Derakshani, T. Lukow, and W. Liesack, Appl. Environ. 175. H. M. Simon, J. A. Dodsworth, and R. M. Goodman, Envi-
Microbiol. 67, 623–631 (2001). ron. Microbiol. 2, 495–505 (2000).
143. D. M. Ward, R. Weller, and M. M. Bateson, Nature 345, 176. D. H. Buckley, J. R. Graber, and T. M. Schmidt, Appl. Envi-
63–65 (1990). ron. Microbiol. 64, 4333–4339 (1998).
144. S. J. Giovannoni, T. B. Britschgi, C. L. Moyer, and K. G. 177. T. Lueders and M. Friedrich, Appl. Environ. Microbiol. 66,
Field, Nature 345, 60–63 (1990). 2732–2742 (2000).
145. E. Stackebrandt, W. Liesack, and B. M. Goebel, FASEB J. 178. S. Lehmann-Richter et al., Environ. Microbiol. 1, 159–166
7, 232–236 (1993). (1999).
146. A. E. McCaig, S. J. Grayston, J. I. Prosser, and L. A. Glover, 179. T. Lüders, K. J. Chin, R. Conrad, and M. Friedrich, Environ.
FEMS Microbiol. Ecol. 35, 37–48 (2001). Microbiol. 3, 194–204 (2001).
147. A. Felske, A. Woltering, R. Van Lis, and A. D. L. Akker- 180. H. Kim et al., Microbiology 146, 2309–2315 (2000).
mans, Appl. Environ. Microbiol. 64, 871–879 (1998).
181. J.-H. Rotthauwe, K.-P. Witzel, and W. Liesack, Appl. Envi-
148. A. Felske et al., FEMS Microbiol. Ecol. 30, 137–145 (1999). ron. Microbiol. 63, 4704–4712 (1997).
149. E. Smit et al., Appl. Environ. Microbiol. 67, 2284–2291 182. U. Purkhold et al., Appl. Environ. Microbiol. 66, 5,368–
(2001). 5,382 (2000).
150. H. Brim et al., Can. J. Microbiol. 45, 326–338 (1999). 183. A. J. Holmes, A. Costello, M. E. Lidstrom, and J. C. Murrell,
151. K.-J. Chin et al., Appl. Environ. Microbiol. 65, 5042–5049 FEMS Microbiol. Lett. 132, 203–208 (1995).
(1999). 184. I. R. McDonald and J. C. Murrell, FEMS Microbiol. Lett.
152. J. Zhou et al., Microbiology 143, 3913–3919 (1997). 156, 205–210 (1997).
153. P. Hugenholtz, C. Pitulle, K. L. Hershberger, and N. R. 185. A. M. Costello and M. E. Lidstrom, Appl. Environ. Micro-
Pace, J. Bacteriol. 180, 366–376 (1998). biol. 65, 5066–5074 (1999).
154. P. Hugenholtz, B. M. Göbel, and N. R. Pace, J. Bacteriol. 186. T. Henckel, M. Friedrich, and R. Conrad, Appl. Environ.
180, 4765–4774 (1998). Microbiol. 65, 1980–1990 (1999).
155. M. A. Dojka, J. K. Harris, and N. R. Pace, Appl. Environ. 187. P. F. Dunfield et al., Appl. Environ. Microbiol. 65, 1,009–
Microbiol. 66, 1617–1621 (2000). 1,014 (1999).
156. P. Hugenholtz et al., Appl. Environ. Microbiol. 67, 411–419 188. M. G. Wise, J. V. McArthur, and L. J. Shimkets, Appl.
(2001). Environ. Microbiol. 65, 4887–4897 (1999).
 BIODIVERSITY IN SOILS: USE OF MOLECULAR METHODS FOR ITS CHARACTERIZATION 543

189. A. J. Holmes et al., Appl. Environ. Microbiol. 65, 3,312– 220. D. J. Scala and L. J. Kerkhof, FEMS Microbiol. Lett. 162,
3,318 (1999). 61–68 (1998).
190. T. Henckel, U. Jäckel, S. Schnell, and R. Conrad, Appl. 221. D. J. Scala and L. J. Kerkhof, Appl. Environ. Microbiol. 65,
Environ. Microbiol. 66, 1801–1808 (2000). 1681–1687 (1999).
191. I. R. McDonald, E. M. Kenna, and J. C. Murrell, Appl. Env- 222. D. J. Scala and L. J. Kerkhof, Appl. Environ. Microbiol. 66,
iron. Microbiol. 61, 116–121 (1995). 1980–1986 (2000).
192. S. N. Dedysh, N. S. Panikov, and J. M. Tiedje, Appl. Envi- 223. L. G. Gregory, A. Karakas-Sen, D. J. Richardson, and S.
ron. Microbiol. 64, 922–929 (1998). Spiro, FEMS Microbiol. Lett. 183, 275–279 (2000).
193. C. B. Miguez et al., Appl. Environ. Microbiol. 65, 381–388 224. M. Wagner et al., J. Bacteriol. 180, 2975–2982 (1998).
(1999). 225. D. Minz et al., Appl. Environ. Microbiol. 65, 4666–4671
194. T. Shigematsu et al., Appl. Environ. Microbiol. 65, 5,198– (1999).
5,206 (1999). 226. M. A. Bruns, M. R. Fries, J. M. Tiedje, and E. A. Paul,
195. A. J. Auman, S. Stolyar, A. M. Costello, and M. E. Lidstrom, Microb. Ecol. 36, 293–302 (1998).
Appl. Environ. Microbiol. 66, 5259–5266 (2000). 227. T. Vallaeys et al., FEMS Microbiol. Ecol. 28, 373–382
196. S. Juretschko et al., Appl. Environ. Microbiol. 64, 3,042– (1999).
3,051 (1998). 228. G. Laguerre et al., Appl. Environ. Microbiol. 62, 2,029–
197. T. A. Mendum, R. E. Sockett, and P. R. Hirsch, Appl. Envi- 2,036 (1996).
ron. Microbiol. 65, 4155–4162 (1999). 229. S. N. Dedysh et al., Int. J. Syst. Evol. Microbiol. 50, 955–969
198. Y. Sakano and L. Kerkhof, Appl. Environ. Microbiol. 64, (2000).
4877–4882 (1998). 230. H. Heuer and K. Smalla, in J. D. van Elsas, J. T. Trevors,
and E. M. H. Wellington, eds., Modern Soil Microbiology,
199. J. R. Stephen et al., Appl. Environ. Microbiol. 65, 95–101
Marcel Dekker, New York, 1997, pp. 353–373.
(1999).
231. G. Muyzer and K. Smalla, Antonie Van Leeuwenhoek 73,
200. S. C. Nold, J. Zhou, A. H. Devol, and J. M. Tiedje, Appl.
127–141 (1998).
Environ. Microbiol. 66, 4532–4535 (2000).
232. G. Muyzer, Curr. Opin. Microbiol. 2, 317–322 (1999).
201. G. A. Kowalchuk et al., Appl. Environ. Microbiol. 65,
396–403 (1999). 233. F. Schwieger and C. C. Tebbe, Appl. Environ. Microbiol. 64,
4870–4876 (1998).
202. G. A. Kowalchuk et al., Environ. Microbiol. 2, 99–110
(2000). 234. T. L. Marsh, Curr. Opin. Microbiol. 2, 323–327 (1999).
203. H.-P. Horz, J.-H. Rotthauwe, T. Lukow, and W. Liesack, J. 235. A. M. Osborn, E. R. B. Moore, and K. N. Timmis, Environ.
Microbiol. Methods 39, 197–204 (2000). Microbiol. 2, 39–50 (2000).
236. A. J. Martinez-Murcia, S. G. Acinas, and F. Rodriguez-
204. I. R. McDonald and J. C. Murrell, Appl. Environ. Microbiol.
Valera, FEMS Microbiol. Ecol. 17, 247–256 (1995).
63, 3218–3224 (1997).
237. W. T. Liu, T. L. Marsh, H. Cheng, and L. J. Forney, Appl.
205. A. Fesefeldt and C. G. Gliesche, Syst. Appl. Microbiol. 20,
Environ. Microbiol. 63, 4516–4522 (1997).
387–396 (1997).
238. T. Lukow, P. F. Dunfield, and W. Liesack, FEMS Microbiol.
206. S. Radajewski, P. Ineson, N. R. Parekh, and J. C. Murrell,
Ecol. 32, 241–247 (2000).
Nature 403, 646–649 (2000).
239. M. T. Suzuki, M. S. Rappe, and S. J. Giovannoni, Appl.
207. G. Eller and P. Frenzel, Appl. Environ. Microbiol. 67,
Environ. Microbiol. 64, 4522–4529 (1998).
2395–2403 (2001).
240. B. Yin et al., Appl. Environ. Microbiol. 66, 4361–4365
208. H. Hennecke et al., Arch. Microbiol. 142, 342–348 (1985).
(2000).
209. T. Ueda, Y. Suga, N. Yahiro, and T. Matsuguchi, J. Bacte-
241. L. Ranjard et al., Microb. Ecol. 39, 263–272.
riol. 177, 1414–1417 (1995).
242. G. F. Duarte et al., Appl. Environ. Microbiol. 67, 1,052–
210. T. Kudo et al., Extremophiles 2, 155–161 (1998). 1,062 (2001).
211. A. S. Rosado, G. F. Duarte, L. Seldin, and J. D. van Elsas, 243. H. Schäfer, P. Servais, and G. Muyzer, Arch. Microbiol. 137,
Appl. Environ. Microbiol. 64, 2770–2779 (1998). 138–145 (2000).
212. M. Ohkuma, S. Noda, and T. Kudo, Appl. Environ. Micro- 244. P. Legendre and L. Legendre, Numerical Ecology, 2nd ed.,
biol. 65, 4926–4934 (1999). Elsevier, Amsterdam, New York, 1998 (in English).
213. C. R. Lovell, Y. M. Piceno, J. M. Quattro, and C. E. Bagwell, 245. J. Dunbar, L. O. Ticknor, and C. R. Kuske, Appl. Environ.
Appl. Environ. Microbiol. 66, 3814–3822 (2000). Microbiol. 66, 2943–2950 (2000).
214. F. Poly et al., Appl. Environ. Microbiol. 67, 2255–2262 246. V. C. Ferrari and J. T. Hollibaugh, Hydrobiologia 401,
(2001). 55–68 (1999).
215. E. Springer, M. S. Sachs, C. R. Woese, and D. R. Boone, Int. 247. T. Zhang and H. H. P. Fang, Biotech. Lett. 22, 399–405
J. Syst. Bacteriol. 45, 554–559 (1995). (2000).
216. M. Ohkuma, S. Noda, K. Horikoshi, and T. Kudo, FEMS 248. B. M. Duineveld et al., Appl. Environ. Microbiol. 67,
Microbiol. Lett. 134, 45–50 (1995). 172–178 (2001).
217. S. L. Pichard, L. Campbell, and J. H. Paul, Appl. Environ. 249. C. H. Nakatsu, V. Torsvik, and L. Øvreås, Soil Sci. Soc. Am.
Microbiol. 63, 3600–3606 (1997). J. 64, 1382–1388 (2000).
218. G. Braker, A. Fesefeldt, and K.-P. Witzel, Appl. Environ. 250. G. Braker et al., Appl. Environ. Microbiol. 67, 1893–1901
Microbiol. 64, 3769–3775 (1998). (2001).
219. G. Braker et al., Appl. Environ. Microbiol. 66, 2096–2104 251. C. A. Eichner, R. W. Erb, K. N. Timmis, and I. Wagner-
(2000). Döbler, Appl. Environ. Microbiol. 65, 102–109 (1999).
544 BIOEMULSIFIERS

252. F. Ampe and E. Miambi, Int. J. Food Microbiol. 60, 91–97 285. W. Ludwig and K.-H. Schleifer, System. Appl. Microbiol. 23,
(2000). 556–562 (2000).
253. D. B. Hedrick et al., J. Microb. Methods 41, 235–248 (2000). 286. S. Becker, P. Böger, R. Oehlmann, and A. Ernst, Appl.
254. A. Gelsomino, A. C. Keijzer-Wolters, G. Cacco, and J. D. van Environ. Microbiol. 66, 4945–4953 (2000).
Elsas, J. Microbiol. Methods 38, 1–15 (1999). 287. J. R. Stults et al., Appl. Environ. Microbiol. 67, 2,781–
255. M. M. Moeseneder et al., Appl. Environ. Microbiol. 65, 2,789 (2001).
3518–3525 (1999). 288. D. Y. Guschin et al., Appl. Environ. Microbiol. 63, 2,397–
256. J. Dunbar, L. O. Ticknor, and C. R. Kuske, Appl. Environ. 2,402 (1997).
Microbiol. 67, 190–197 (2001). 289. S. G. Bavykin et al., Appl. Environ. Microbiol. 67, 922–928
257 B. W. Bogan, B. Schoenike, R. T. Lamar, and D. Cullen, (2001).
Appl. Environ. Microbiol. 62, 2381–2386 (1996). 290. V. Torsvik, F. L. Daae, R.-A. Sandaa, and L. Øvreås, J.
258. R. T. Lamar et al., Appl. Environ. Microbiol. 61, 2,122– Biotechnol. 64, 53–62 (1998).
2,126 (1995). 291. A. C. Kennedy, Agric. Ecosyst. Environ. 74, 65–76 (1999).
259. C. C. Tebbe et al., Appl. Environ. Microbiol. 61, 4296–4303 292. J. T. Trevors, Water, Air, Soil Pollut. 101, 45–67 (1998).
(1995). 293. J. Kozdrój and J. D. van Elsas, J. Microbiol. Methods 43,
260. J. T. Fleming, W.-H. Yao, and G. S. Sayler, Appl. Environ. 197–212 (2001).
Microbiol. 64, 3698–3706 (1998). 294. D. M. Ward, Curr. Opin. Microbiol. 1271–1277 (1998).
261. J. T. Fleming, J. Sanseverino, and G. S. Sayler, Environ.
Sci. Technol. 27, 1068–1074 (1993).
262. J. Sanseverino et al., Biodegradation 4, 303–321 (1993).
BIOEMULSIFIERS. See BIOSURFACTANTS: TYPES,
263. J. T. Fleming and G. S. Sayler, in R. A. Minear, A. M. Ford,
SCREENING METHODS, AND APPLICATIONS
L. L. Needham, and N. J. Karch, eds., Applications of
Molecular Biology in Environmental Chemistry, Lewis
Publishers, Boca Raton, Fla., 1995, pp. 45–58.
264. T. A. Mendum, R. E. Sockett, and P. R. Hirsch, FEMS
Microbiol. Lett. 164, 369–373 (1998).
BIOFILM DETACHMENT
265. R. D. Stapleton et al., J. Microbiol. Methods 32, 165–178
(1998).
BRUCE E. RITTMANN
266. R. Fislage et al., Nucleic Acids Res. 25, 1830–1835 (1997).
CHRYSI S. LASPIDOU
267. P. Liang and A. B. Pardee, Science 257, 967–971 (1992).
Northwestern University
268. M. McClelland and J. W. Welsh, PCR Methods Appl. 4, Evanston, Illinois
S66–S81 (1994).
269. K. K. Wong and M. McClelland, Proc. Natl. Acad. Sci. USA
A biofilm consists of microorganisms attached to a
91, 639–643 (1994).
solid surface called the substratum (1–7). Usually, the
270. J. Borneman, Appl. Environ. Microbiol. 65, 3398–3400
microbial cells are embedded in a matrix of organic
(1999).
polymers produced by the cells. Biofilms are ubiquitous
271. E. Urbach, K.-L. Vergin, and S. J. Giovannoni, Appl. Envi-
in nature and are increasingly important in engineered
ron. Microbiol. 65, 1207–1213 (1999).
processes. Some biofilms are viewed as ‘‘good,’’ and we
272. N. Lee et al., Appl. Environ. Microbiol. 65, 1289–1297
try to promote their accumulation. Examples of good
(1999).
biofilms include those that are exploited in fixed-film
273. C. C. Ouverney and J. A. Fuhrman, Appl. Environ. Micro-
processes used to treat contaminated water, wastewater,
biol. 65, 1746–1752 (1999).
and air (4,5,8), those that attach to stream beds and
274. H. T. S. Boschker et al., Nature 392, 801–805 (1998).
aquatic vegetation, leading to self-purification of water
275. Roslev and N. Iversen, Appl. Environ. Microbiol. 65, bodies (9), and those responsible for engineered or intrinsic
4064–4070 (1999).
bioremediation of contaminated groundwater (5,8,10).
276. J. Handelsman et al., Chem. Biol. 5, R245–R249 (1998). Other biofilms are viewed as ‘‘bad,’’ and we try to remove
277. J. L. Stein et al., J. Bacteriol. 178, 591–599 (1996). or prevent them. Bad biofilms include those that foul ship
278. C. Schleper et al., J. Bacteriol. 179, 7803–7811 (1997). hulls and pipelines, thereby increasing friction loss and
279. C. Schleper et al., J. Bacteriol. 180, 5003–5009 (1998). corrosion, those that cause ‘‘souring’’ of oil wells, and those
280. C. M. Preston, K. Y. Wu, T. F. Molinski, and E. F. DeLong, that cause medical problems, such as dental caries and
Proc. Natl. Acad. Sci. USA 93, 6241–6246 (1996). infection related to implants (1,2,6,7).
281. R. Zimmer and A. M. V. Gibbins, Genomics 42, 217–226 Biofilms are highly diverse (1,3,4,11). Some are physi-
(1997). cally very thin, consisting only of small clusters of cells
282. H. Shizuya et al., Proc. Natl. Acad. Sci. U.S.A. 89, on a substratum. Other biofilms are quite thick, having a
8794–8797 (1992). depth of as much as a few millimeters. While the physical
283. B. L. Wanner, in F. C. Neidhardt, eds., Escherichia coli and structure of some biofilms is dense and homogenous,
Salmonella: Cellular and Molecular Biology, 2nd ed., vol. 1, other biofilms are physically heterogeneous and have
ASM Press, Washington, D.C., 1996, pp. 1357–1381. clusters or streamers of biomass intermingled with open
284. P. Entcheva et al., Appl. Environ. Microbiol. 67, 89–99 channels (11). While single-species biofilms are useful for
(2001). research studies, naturally occurring biofilms often show
 BIOFILM DETACHMENT 545

a high degree of ecological diversity that can exhibit (a)

fascinating layering or sub-clustering.
One of the reasons that biofilms are so diverse is that
many processes occur together as the biofilm forms (2,12):
for example, microbial growth and death, attachment, and
detachment. Furthermore, biofilms are characterized by
strong gradients (2,5,8). Thus, the processes occur in very
different ways at the outer surface of the biofilm, compared (b)
to near the substratum.
The subject of this article is biofilm detachment, which
is the physical movement of microbial cells from the biofilm
matrix. Detachment results in a loss of material from the
biofilm, and these materials generally move to the liquid
that is in contact with the biofilm. The article begins by
describing the ways in which detachment can occur. Then, Figure 1. Schematic of two detachment mechanisms: (a) erosion
it discusses why detachment is such an important process and (b) sloughing.
for understanding and controlling biofilms. Finally, the
article summarizes what is known about how forces acting
on the biofilm and physiology of the microorganisms in the per unit of substratum surface area is called biofilm
biofilm affect detachment rates. accumulation, and it can be represented mathematically
Whether the goal is to encourage or discourage biofilm by Xf Lf , in which Xf is the biomass density of the
accumulation, detachment is one of the key determinants biofilm (mg biomass/cm3 ) and Lf is the biofilm thickness
for how much biofilm accumulates, as well as its physical (cm) (8,14). The accumulation results from a balance
and microbiological characteristics. Biofilm detachment among gain by synthesis from substrate (food) utilization,
is an emerging research field. Although some broad loss by decay (or death), and loss by detachment. For a
concepts are coming into view, most of what needs to steady-state biofilm, that balance leads to (8,14)
be known about what affects detachment rates will be
discovered in the future. This article provides a structure YJ
Xf Lf = (1)
for comprehending how physical forces and physiology b + bdet
interact to control biofilm determinant.
in which Y = true yield of biomass synthesized per
substrate utilized (mg biomass/mg substrate), J = flux
DETACHMENT PATTERNS of substrate into the biofilm (mg substrate/cm2 − d),
b = specific decay rate (d−1 ), and bdet = specific detach-
Detachment can occur with three broad patterns. The first ment rate (d−1 ).
pattern is erosion, which is a continuous process by which Equation 1 makes it clear that the amount of biofilm
relatively small pieces of biofilm are removed from the present on the surface (Xf Lf ) increases if more food is
biofilm’s outer surface (2,8,12,13). Each erosion event has available (making J larger) or if the detachment rate (bdet )
a small impact on the biofilm, but, because the events occur is smaller. For situations in which biofilm accumulation
continuously, erosion can be a significant loss mechanism. is to be encouraged (say, to treat wastewater), bdet may
The second pattern, sloughing, is an abrupt, intermit- have to be decreased. On the other hand, when biofilm
tent loss of a large segment of the biofilm (2,8,12,13). The accumulation is to be decreased (say, to minimize fouling
sloughed segment extends deep into the biofilm, often all or infection), it may be preferred to increase bdet .
the way to the substratum. One sloughing event can cause In principle, biofilms are totally eliminated if the
a profound change to the physical and microbiological substrate concentration is so low that the positive growth
structure of the biofilm. Sloughing is associated with a rate from synthesis is always less than the sum of the
breakdown of the structural integrity of the biofilm, allow- loss rates from decay and detachment. This threshold
ing a ‘‘fault’’ to occur along lines of structural weakness. concentration is called Smin , and it can be computed
Figure 1 illustrates the difference between the erosion and from (8,14):
sloughing mechanisms of detachment.
The third pattern can be termed scouring, and it b + bdet
Smin = K (2)
involves removal of large segments of biofilm through Yqmax − (b + bdet )
the action of very strong physical forces, such as abrasion
or scraping. Predation (sometimes called grazing) also in which K = the Monod half-maximum-rate concentra-
removes biofilm mass, but we do not include it among the tion (mg substrate/cm3 ), and qmax = the maximum specific
detachment mechanisms reviewed in this article. rate of substrate utilization (mg substrate/mg biomass −
d). The major trend is that increasing bdet makes Smin
IMPORTANCE OF DETACHMENT AND ITS PATTERN larger, thereby making it more difficult to sustain a biofilm.

Biofilm Accumulation Average Specific Growth Rate and Activity

The most obvious impact of detachment is to control the Communities of slowly growing microorganisms accumu-
total amount of biofilm present. The total mass of biofilm late a significant amount of inactive or inert solids because
546 BIOFILM DETACHMENT

low substrate concentration and endogenous decay lead Suspended Biomass Concentration
to a residue of ‘‘dead’’ or ‘‘dormant’’ biomass (8). The
When biofilms dominate the biomass accumulation in a
inert solids also include extracellular polymeric substances
system, the concentration of suspended microorganisms is
(EPS) that hold the biofilm together and tend to accumu-
controlled by the rate of detachment from the biofilm. This
late most near the substratum. Therefore, slowly growing
is expressed mathematically as
communities are less ‘‘active’’ than are fast-growing com-
munities in which less inert biomass accumulates.
For steady-state biofilms, the average specific growth X = bdet Xf Lf A/Q (3)
rate (µave , d−1 ) is equal to the specific detachment rate (8),
or µave = bdet . Thus, the metabolic activity of a biofilm in which X = the steady state increase in suspended-
is controlled strongly by the detachment rate (8,15–21). biomass concentration caused by detachment (mg
Low detachment rates lead to biofilms having a low biomass/cm3 ), A = the surface area of the biofilm (cm2 ),
average specific growth rate and being enriched in inactive and Q = the volumetric flow rate (cm3 /d). Equation (3)
biomass. makes it clear that increasing bdet or A/Q makes biofilm
detachment a more dominant influence on the concentra-
tion of suspended microorganisms.
Ecological Selection
Detachment rates and patterns affect ecological selection Substratum Properties
in profound ways. In multispecies biofilms, the most slowly As detachment decreases and Xf Lf becomes larger, the
growing species tend to accumulate deep inside the biofilm physical and chemical characteristics of the substratum
when erosion is the detachment pattern (15–21). Being can be altered. Examples of changes to substrata are
deep inside the biofilm protects them from detachment
losses that occur from the outer surface. Protection
• increased diameters and decreased density of sub-
effectively lowers bdet for these slow growers, which
stratum particles, such as sand and activated
increases their Xf Lf (Eq. 1) and decreases their Smin
carbon (20,23–25);
(Eq. 2) (11).
• decreased flow-path sizes in pipes and channels;
Detachment by erosion leads to a layering of microbial
types. The fastest growing species predominate near the • increased friction losses in pipes, channels, and
outer surface, where they have the advantage of the porous media (26,27);
best access to their substrate. On the other hand, the • decreased or increased surface roughness (2,27,28);
slowest growers predominate well away from the outer • changes to surface charge or hydrophobicity; and
surface, where they have the advantage of protection • increased heat-transfer resistance (2,7).
from erosion. Examples of this sort of layering include
heterotrophs protecting nitrifiers in aerobic biofilms (15,9)
and sulfate-reducers protecting methanogens in anaerobic FACTORS AFFECTING DETACHMENT RATES
biofilms. The slowest growing ‘‘species’’ is the inert biomass
produced as part of endogenous decay or dormancy; it What controls biofilm detachment is far from well
predominates at the substratum (22). understood today. Detachment is an active research
In the short term, an increased rate of erosion removes area that generates conflicting points of view. Three
the fastest growing, most active biomass, which resides characteristics of the field lead to today’s tumultuous state.
near the outer surface. However, more erosion of active
bacteria from the outer surface allows the substrate to 1. Biofilms are incredibly diverse in terms of substrata,
penetrate deeper into the biofilm. This may activate hydrodynamics, microbial ecology, and substrate
dormant bacteria near the substratum, and it causes the loading. The controlling mechanisms surely differ
active bacteria near the substratum to grow faster, which widely from one system to another. Thus, one com-
gradually pushes the inert biomass to the outer surface, mon ‘‘detachment theory’’ should not be expected.
where it erodes (15). Therefore, the long-term effect of 2. Experimental methods to measure the biofilm
erosion is most profound on the slowest growing species. properties relevant to detachment are primitive.
Although it requires some time for the full effects to be 3. Theoretical models for linking observed detachment
realized, an increase to the specific erosion rate lowers rates to biofilm properties are crude. So far, these
the accumulation of slow growers (including the inert of connections rely mostly on statistical correlations,
biomass) most dramatically (15). rather than mechanistic principles.
When detachment is by sloughing or scouring, the lay-
ering effects associated with erosion may be reduced or Despite much uncertainty, detachment is not totally
eliminated. Detachment that extends down to the sub- mysterious. Certain factors that affect detachment rates
stratum removes all types of biomass. These detachment are emerging, and they are the subjects of this section.
patterns can be quite devastating to slow-growing species The factors can be grouped into two categories: physical
(such as nitrifiers), whose survival in the biofilm depends forces acting on the biofilm and the physiology of the
on protection from detachment. microorganisms in the biofilm.
 BIOFILM DETACHMENT 547

Physical Forces (a)

The first factor recognized to affect the detachment rate is
the tangential shear stress (29), τ (dyne/cm2 ). Shear stress
is proportional to the friction losses for water passing by
the substratum, and it acts at the biofilm’s outer surface. In
general, a greater shear stress increases the detachment
rate, although the effect is not necessarily linear. Bakke
and coworkers (30) and Peyton and Characklis (31) found
an approximately linear relationship between bdet and τ ,
(b)
although there was considerable variation. Rittmann (32)
found that bdet was proportional to τ 0.58 for thin biofilms
(Lf ≤ 30 µm), and Chang and Rittmann (28) found that
this relationship was quantitatively accurate for biofilms
on smooth surfaces. When the biofilm was thicker than
30 µm, the specific detachment rate declined, presumably
because the biomass deep in the biofilm was protected
from erosion (29). When the biofilm was grown on a highly
irregular surface, bdet was essentially zero until the biofilm
accumulated enough to fill the crevices that protected it
from shear stress (28). Once the biofilm fully covered the
surface, the detachment rate was similar to that of smooth
surface (28).
The several studies that examined shear stress show
that it is an important factor for determining bdet as long
Figure 2. Graphic representation of the two main mechanisms
as the biofilm is actually exposed to the shear stress and
through which axial stresses can occur on biofilm: (a) abrasion,
other factors do not overwhelm it. Protection of all (28) or
shown as compressive forces acting on biofilm caused by
part (29) of the biofilm from the shear stress can result the collision of two bioparticles, and (b) pressure fluctuations
in substantially reduced detachment rates. Likewise, the (represented by arrows) on a bioparticle caused by turbulent
effects of shear stress can be ‘‘swamped’’ by other factors. eddies.
For example, Chang and coworkers (32) saw that bdet was
inversely correlated to τ for fluidized-bed biofilm reactors.
This correlation was the result of other factors (discussed The studies in which abrasion and/or turbulence
in the following section) controlling bdet , Xf Lf , and τ . created strong axial stress show that axial stress can
The second factor that affects the detachment rate is the overwhelm shear stress if the axial stress is large enough.
axial force acting on a biofilm. This force could be realized For example, the specific detachment rate increased up to
as tensile or compressive stress, σ (dyne/cm2 ), which acts 25-fold over what was expected from shear stress alone in
axially on the biofilm. Such stresses can occur through the two-phase fluidized bed (8,32). Aeration increased bdet
two main mechanisms: abrasion, caused by collisions by approximately another factor of 2 (8,23). Although the
with other particles, and pressure fluctuations, caused effect of abrasion can be profound, we have a weak basis
by turbulent eddies (Fig. 2). for quantifying it or its impact on bdet .
Abrasion effects were studied by varying the carrier Detachment is a structural failure of the biofilm matrix,
concentration (Cp , g/l) in a fluidized-bed biofilm reac- which means that the stress at the point of failure
tor (32). Using Cp as a measure of the frequency of exceeds the yield stress, σy . When the actual stress in
particle-to-particle contacts, Chang and coworkers (32) the biofilm exceeds σy , the biofilm breaks. Several groups
showed that bdet was strongly correlated to the abrasion have measured yield stresses by different methods. Ohashi
intensity. and Harada (33) applied tensile stress and found σy in the
Turbulent effects can be gauged by the Reynolds range of 0 to 1,750 N/m2 .
number, Re = υ L/v, in which υ = the liquid velocity Two models have been proposed to explain the way
(cm2 /s), L = a characteristic size dimension, such as in which the biofilm matrix behaves as the shear stress
carrier diameter or pipe diameter (cm), and v = the approaches σy . One model (33) views the biofilm matrix as
liquid’s dynamic viscosity (cm2 /s). In the same fluidized- a Bingham fluid (34), a useful concept in polymer science.
bed studies (32), bdet also correlated strongly with Re. A Bingham fluid exhibits non-Newtonian behavior. For
Two other studies with fluidized beds confirmed a positive stresses less than the Bingham yield stress, τ0 , it does not
correlation between bdet and Re (24,25). deform. When the stress reaches τ0 , the Bingham fluid
Turbulence effects can be greatly increased by adding flows with a viscosity that depends on τ0 .
energy inputs, such as from mechanical mixing or aeration. The second model (35) views the biofilm matrix as a
When fluidized-bed biofilm reactors were aerated to create viscoelastic structure, which can be characterized by an
a three-phase operation (20,23), the detachment rate elastic (Young’s) modulus (E) and a shear modulus (G). E is
increased significantly and was controlled primarily by the ratio of the applied axial stress to the strain (ε) caused
the gas velocity, although Cp and Re were still significant. by extension of the material (Fig. 3). G is the ratio of the
548 BIOFILM DETACHMENT

may be more important in some cases (38,39). EPS

formation is important in biofilm detachment because
EPS forms the matrix that gives the biofilm structural
integrity and its viscoelastic properties. It seems that
EPS production depends on the substrate-consumption
rate (40,41). Microorganisms seem to produce less EPS
Stress s

when rapidly consuming substrate. As the substrate

concentration is higher near the outer surface of the
biofilm, the EPS content may be lower there. As EPS is
critical for cohesive strength, the outer part of the biofilm
may be structurally weaker (e.g., a lower σy ), and this may
help explain how erosion comes about.
Slope = s / e = E
Biofilm Density. The density of the biofilm seems to play
Strain e a crucial role in biofilm strength and detachment rate. In
Figure 3. The stess-strain curve of a material, which defines its
general, denser biofilms are stronger (e.g., a higher σy )
Young’s modulus E. and have smaller specific detachment rates (bdet ). Ohashi
and coworkers (42) found that the EPS content of the
biofilm was directly proportional to the density. Kwok and
applied shear stress to the shear strain caused by bending coworkers (24), Chang and coworkers (32), and Nicolella
the material. Strain is the ratio of the distortion distance and coworkers (25) found that thinner, denser biofilms had
(L) to the total length of the material (L), whereas shear smaller bdet values. Stoodley and coworkers (35) found that
strain is the gradient of deformation under shear. In an E increased for biofilms exposed to increased shear stress.
ideally elastic solid, strain is directly proportional to stress, Differences in biofilm density are often explained by
and E or G is constant. If the stress is released, the solid assuming the presence of different types of microorgan-
returns to its original shape. Stoodley and coworkers (35) isms. However, evidence is strong that hydrodynamic
performed in situ deformation studies that gave average conditions have a direct effect on biofilm density. Vieira
E and G values of 40 and 27 N/m2 , respectively. and coworkers (43) showed that the density of a pure-
Biofilms can also exhibit plastic behavior, which is a culture biofilm increased with increasing shear stress
deviation from ideal elastic behavior. Plastic behavior can on the biofilm. Chang and coworkers (32) and Kwok and
be viewed as the material ‘‘flowing’’ like a liquid that coworkers (24) found that biofilms in fluidized-bed biofilm
has a very large dynamic viscosity. Creep, an example of reactors became denser when the particle concentration
plastic behavior in biofilms (35), refers to the continued (Cp ), a surrogate for abrasion forces, was larger.
deformation of the material over time, although the The evidence on the effects of shear stress and abrasion
applied stress is constant. On release of the stress, suggests that biofilms exposed to detachment forces adapt
the plastic material that has undergone creep does not to become denser and stronger. How this adaptation
return to its original shape, but remains deformed. Creep occurs is not known. One hypothesis is that strong
probably occurs because intermolecular bonds between detachment forces remove low-density new growth from
polymers in the biofilm matrix ‘‘unzip’’ gradually, allowing the outer surface (24). The continued removal of this
the molecules to slip past each other in viscous flow. When less dense layer enables the substrate to diffuse into
the stress is removed, the polymer strands contract and the inner biofilm layers, where microcolonies grow within
reaggregate with neighboring strands in a new position. the strong EPS matrix. A second hypothesis is that the
Flemming and Wingender (36) have termed this sort of microorganisms produce more and/or stronger EPS in
breaking and reforming of bonds ‘‘fluctuating adhesion response to the detachment force. This could occur by
points’’ and attributed it to the large number and widely selecting for microorganisms that produce strong EPS (in
ranging type of bonds formed by EPS. The formation of a multispecies biofilms) or by a regulatory response. A third
new EPS may stabilize the new position further. hypothesis is that a denser biofilm is the result of a gradual
displacement of loosely bound water out of the matrix.
Physiology Another factor that helps explain the differences in den-
sity across a biofilm is the buildup of inert solids, such as
The physiology of the microorganisms in the biofilm inactive biomass. Long-term operation with a long solid’s
appears to affect the strength of the biofilm material and, retention time (i.e., a small bdet ) allows inactive biomass
therefore, detachment rate and pattern. Five physiological solids to accumulate, particularly near the substratum.
factors are reviewed: EPS, density, heterogeneity, specific Mature biofilms, therefore, have a relatively large frac-
growth rate, and quorum sensing. tion of inactive volatile solids (20,23,42). Although inactive
biomass cannot produce EPS, it may be well embedded in
EPS. The cells in a biofilm use EPS to adhere to existing EPS, possible allowing a closer packing arrange-
each other and the substratum (37). The acronym EPS is ment if loosely bound water can be squeezed out more
sometimes used to mean extracellular polysaccharides, but easily. In fact, the EPS is probably part of the inert
this usage does not accurately represent the composition biomass.
of the polymers. Although sugars are important in Other inert solids that can build up deep in a biofilm
the EPS, proteins and nucleic acids are also key and are inorganic precipitates, including metal carbonates,
 BIOFILM DETACHMENT 549

sulfates, sulfides, phosphates, and hydroxides. Often, key factor affecting the detachment rate (31,51,52). This
these are formed because microbial reactions alter the relationship appears reasonable based on the observation
chemistry within the biofilm and cause super-saturation that EPS formation is inversely related to the substrate-
for a solid phase (22,44–48). Precipitation of inorganic utilization rate (40,41). However, conclusions about the
solids could create a very dense and strong structure (44), role of the specific growth rate must be viewed cautiously
essentially ‘‘cementing’’ the biofilm in place. Alternately, because the average specific growth rate for a steady-state
inorganic deposits could make the biofilm brittle, perhaps biofilm equals the specific detachment rate: µave = bdet .
making the biofilm susceptible to sloughing (35). This equality opens up the possibility that observations of
One very interesting adaptation finding (35) is on the a relationship between µave and bdet are a tautology, not
importance of the shear stress applied during growth of cause and effect.
the biofilm (τg ). When exposed to other shear stresses, Some studies suggest that a high µ is not responsible
biofilms behaved as elastic solids when the applied τ was for high detachment rates. Sawyer and Hermanawicz (53)
less than τg . However, they behaved like plastic fluids for report an inverse relationship between detachment rate
τ > τg . In other words, τg was a good surrogate for the and substrate availability. Gjaltema and coworkers (54)
Bingham yield stress, τ0 . For, τ > τg , the elastic modulus found that the effect of Cp on the detachment rate in
(E) increased, indicating that the stiffness of the biofilm fluidized beds differed depending on substrate availability.
was increasing with strain once the biofilm became plastic. When substrate was available to give high biofilm growth
rates, increasing Cp resulted in a slower detachment rate, a
Heterogeneity. Heterogeneity within the biofilm may finding consistent with others (24,25,32). When substrate
affect the detachment mode. Ohashi and Harada (49) was not available for growth, a higher Cp caused a faster
studied the in situ behavior of biofilm development and detachment rate.
detachment using a video camera set up above an open-
channel reactor in which they maintained uniform flow.
Quorum Sensing. The final physiological factor is
They observed major sloughing events for mature biofilms.
quorum sensing, which involves chemical signals by which
After 34 days of biofilm accumulation, massive sloughing
microorganisms communicate with each other (55,56).
events removed about 25% of the observable biofilms. They
Certain bacteria continually produce signal molecules,
found that sloughing was associated with the formation of
cavities within the biofilm. particularly homoserine lactones. When the density of
Stoodley and coworkers (35,50) also used video imaging bacteria is high enough in an aggregate (such as a biofilm),
to evaluate the behavior of filamentous streamers that the signal molecule achieves a high enough concentration
oscillate rapidly in the flow. These streamers cause that it triggers a rapid community response. One known
increased head loss. The streamers exhibited hysteresis response is to control aggregation or disaggregation of
for high applied shear stress. Hysteresis was more evident biofilms of Pseudomonas aeruginosa (55). Although the
near the base of the streamer than the tip. Streamers range of applicability of quorum sensing is undefined, it
exposed to prolonged oscillations may experience fatigue, could dramatically increase or decrease detachment for
which involves thinning and finally detachment. Fatigue situations in which it acts.
failure requires some time; as a result, it may be possible
for the microorganisms to produce EPS that repairs small
CONCLUSION
‘‘cracks’’ or ‘‘thin regions.’’ However, it probably will be
unable to repair cracks or thin regions that result from
Detachment is one of the most important phenomena
high frequency oscillations.
affecting the accumulation, physical properties, and
As a biofilm grows on the substratum, it starts with
ecology of biofilms. The physical and physiological factors
sparsely distributed cell clusters. If the clusters grow
that control detachment are just beginning to come into
together, they create a relatively planar biofilm. The shear
view, and this article defines a framework for designing
and axial stresses acting on individual cell clusters will
and interpreting the research that will illuminate
vary until the biofilm ‘‘fills in.’’ If the clusters are far
detachment. Understanding detachment is at the core
enough apart, the velocity and stress profiles around the
of answering fundamental questions about biofilms: for
clusters can reestablish between clusters so that each
example, Why do some have an open structure, while
cluster experiences similar stresses. The roughness of the
others are dense? Why do biofilms sometimes experience
clusters may increase shear stress and turbulence. On the
other hand, when the clusters are close together, clusters massive sloughing events, while others only erode? How
in the ‘‘wake’’ behind an upstream cluster may experience do physical forces interact with biochemical reactions to
significantly different stresses than the upstream cluster. control where different microorganisms reside in a biofilm?
Once the biofilm fills in to become more or less planar, the
surface may be hydrodynamically smoother, causing the BIBLIOGRAPHY
shear stress and turbulence to decline. Therefore, shear
and axial stresses depend on the degree of surface coverage 1. P. A. Wilderer and W. G. Characklis, in W. G. Characklis and
by the biofilm and the distribution pattern of clusters in P. A. Wilderer, eds., Structure and Function of Biofilms, John
heterogeneous biofilms. Wiley & Sons, New York, 1989, pp. 5–18.
2. C. J. Gantzer et al., in W. G. Characklis and P. A. Wilderer,
Specific Growth Rate. The specific growth rate (µ) of eds., Structure and Function of Biofilms, John Wiley & Sons,
microorganisms in the biofilm has been implicated as a New York, 1989, pp. 73–90.
550 BIOFILMS, ACTIVITY IN

3. W. G. Characklis and K. C. Marshall, in W. G. Characklis 32. H. T. Chang, Biotechnol. Bioeng. 38, 499–506 (1991).
and K. G. Marshall, eds., Biofilms, John Wiley & Sons, 1990, 33. A. Ohashi and H. Harada, Water Sci. Technol. 29, 281–288
pp. 3–15. (1994).
4. H. Lappin-Scott, P. Gilbert, M. Wilson, and D. Allison, Com- 34. E. C. Bingham, Fluidity and Plasticity, McGraw-Hill, New
munity Structure and Cooperation in Biofilms, SGM Sympo- York, 1922.
sium 59, Cambridge University Press, London, U.K., 2000.
35. P. Stoodley, Z. Lewandowski, J. D. Boyle, and H. M. Lappin-
5. J. D. Bryers, ed., Biofilms II. Process Analysis and Applica-
Scott, Biotechnol. Bioeng. 65, 83–92 (1999).
tions, John Wiley & Sons, New York, 2000.
36. H.-C. Flemming and J. Wingender, Water Sci. Technol. (2001)
6. L. V. Evans, ed., Biofilms; Recent Advancements in Their
in press.
Study and Control, Harwood Academic Publishers, Amster-
dam, The Netherlands, 2000. 37. B. E. Christensen, and W. G. Characklis, in W. G. Characklis
and K. G. Marshall, eds., Biofilms, John Wiley & Sons, New
7. J. T. Walker, J. J. Jass, and S. Surman, eds., Biofouling
York, 1990, pp. 93–130.
of Industrial Waters and Processes, John Wiley & Sons,
Chichester, U.K., 2000. 38. V. Lazarova, D. Bellahcen, J. Manem, D. A. Stahl, and
8. B. E. Rittmann and P. L. McCarty, Environmental Biotech- B. E. Rittmann, Water Sci. Technol. 39(7), 4–11 (1999).
nology: Principles and Applications, McGraw-Hill, New York, 39. H.-C. Flemming, this volume (2001).
2000. 40. J. A. Robinson, M. G. Trulear, and W. G. Characklis, Biotech-
9 C. J. Gantzer, B. E. Rittmann, and E. E. Herricks, Water Res. nol. Bioeng. 26, 1409–1417 (1984).
25, 15–20 (1991). 41. M. G. Trulear and W. G. Characklis, J. Water Pollut. Control
10. B. E. Rittmann et al., In Situ Bioremediation, 2nd ed., Noyes Fed. 54, 1288–1301 (1982).
Publ., Park Ridge, N.J., 1994. 42. A. Ohashi, T. Koyama, K. Syutsubo, and H. Harada, Water
11. P. Stoodley, J. D. Boyle, D. DeBeer, and H. Lappin-Scott, Sci. Technol. 39(7), 261–268 (1999).
Biofouling 14(1), 75–90 (1999). 43. M. J. Vieira, L. P. Melo, and M. M. Pinheiro, Biofouling 71,
12. W. G. Characklis in W. G. Characklis and K. G. Marshall, 67–80 (1993).
eds., Biofilms, John Wiley & Sons, New York, 1990, 44. A. J. Cooke, R. K. Rowe, B. E. Rittmann, and I. R. Fleming,
pp. 195–231.
Water Sci. Technol. 39(7), 60–67 (1999).
13. B. E. Rittmann, in W. G. Characklis and P. A Wilderer, eds.,
45. R. K. Rowe et al., in Proc. Sardnia’97, Sixth International
Structure and Function of Biofilms, John Wiley & Sons, New
Landfill Symposium, S. Margherita di Pula, Cagliari, Italy,
York, 1989, pp. 49–58.
October 1997, pp. 383–390.
14. B. E. Rittmann, Biotechnol. Bioeng. 24, 1341–1370 (1982).
46. K.-C. Lee, Autohydrogenotrophic Denitrification of Drinking
15. B. E. Rittmann and J. A. Manem, Biotechnol. Bioeng. 39, Water Using a Hollow-Fiber Membrane Biofilm Reactor,
914–922 (1992).
Ph.D. Dissertation, Dept. Civil Engr., Northwestern Univ.,
16. J. C. Kissel, P. L. McCarty, and R. L. Street, J. Environ. Eng. Dec. 1999.
110, 393–411 (1984).
47. E. Arvin, Water Sci. Technol. 15(3,4), 43–52 (1985).
17. O. Wanner and W. Gujer, Biotechnol. Bioeng. 28, 314–328
48. H.-C. Flemming and Leis, this volume (2001).
(1986).
49. A. Ohashi and H. Harada, Water Sci. Technol. 30(11), 35–45
18. H. Furumai and B. E. Rittmann, Water Sci. Technol.
(1994).
29(10,11), 439–446 (1994).
19. Y. Watanabe, S. Okabe, K. Hirata, and S. Masuda, Water Sci. 50. P. Stoodley, Z. Lewandowski, J. D. Boyle, and H. M. Lappin-
Technol. 31(1), 170–177 (1995). Scott, Biotechnol. Bioeng. 57, 536–544 (1998).
20. B. E. Rittmann, F. Trinet, D. Amar, and H. T. Chang, Water 51. G. E. Speitel Jr. and F. A. DiGiano, J. Environ. Eng. 113,
Sci. Technol. 26(3,4), 585–594 (1992). 464–475 (1987).
21. O. Wanner and W. Gujer, Water Sci. Technol. 17, 27–44 52. L. Tijhuis, M. C. M. van Loosdrecht, and J. J. Heijnen,
(1985). Biotechnol. Bioeng. 45, 481–487 (1995).
22. T. C. Zhang and P. L. Bishop, Water Res. 28, 2267–2277 53. L. K. Sawyer and L. Hermanowicz, Water Sci. Technol.
(1994). 37(4,5), 211–214 (1998).
23. F. Trinet et al., Water Sci. Technol. 23, 1347–1354 (1991). 54. A. Gjaltena, J. L. Vinke, M. C. M. van Loosdrecht, and
24. W. K. Kwok et al., Biotechnol. Bioeng. 58, 400–407 (1998). J. J. Heijnen, Water Sci. Technol. 36(1), 224–228 (1997).
25. C. Nicolella, S. Chiarle, R. DiFelice, and M. Rovatti, Water 55. D. G. Davies, Science 280, 295–298 (1998).
Sci. Technol. 36(1), 221–228 (1997). 56. Jaeger and Wingender, this volume (2001).
26. B. E. Rittmann, Water Resources Res. 29, 2195–2202 (1993).
27. A. B. Cunningham, in W. G. Characklis and P. A. Wilderer,
eds., Structure and Function of Biofilms, John Wiley & Sons,
New York, 1989, pp. 19–32. BIOFILMS, ACTIVITY IN. See ACTIVITY AND CARBON
28. H. T. Chang and B. E. Rittmann, Environ. Sci. Technol. 21, TRANSFORMATIONS IN BIOFILMS
280–288 (1987).
29. B. E. Rittmann, Biotechnol. Bioeng. 24, 501–506 (1982).
30. R. Bakke, W. G. Characklis, M. H. Turakhia, and A. I. Yeh,
in W. G. Characklis and K. C. Marshall, eds., Biofilms, John
Wiley & Sons, New York, 1990, pp. 487–520.
31. B. M. Peyton and W. G. Characklis, Biotechnol. Bioeng. 41, BIOFILMS, ALGAL BIOFILMS. See DIATOMS IN
728–735 (1993). BIOFILMS
 BIOFILMS: BACTERIAL-FUNGAL BIOFILMS 551

BIOFILMS: BACTERIAL-FUNGAL BIOFILMS Bulk fluid

Planktonic phase - bacterial and yeast
SARA K. ROBERTS cells, fungal spores and hyphae
HILARY M. LAPPIN-SCOTT
University of Exeter Conditioning film
Exeter, United Kingdom surface

Naturally occurring biofilms comprise a variety of different Adhesion and

organisms including bacteria, fungi, algae, and protozoa. attachment -EPS
Costerton and coworkers (1) commented on the ubiquitous Cell detachment and production
nature of bacterial biofilms, and although they are less biofilm maturation
studied, so too are bacterial-fungal biofilms. There are
many definitions of a biofilm; however these definitions Growth and division
generally only refer to bacterial biofilms. Attempts have - microcolony
been made recently to account for the presence of other formation
organisms. For the purpose of this review, a biofilm is
defined as microbial cells immobilized at an interface,
surrounded by an exopolymeric matrix of microbial origin.
The simplest biofilm model to study is the monoculture Recruitment
or pure culture biofilm. This model alone offers numerous and growth
avenues for research, so the possibilities in studying mixed
organism biofilms are vast. This in part may explain why
bacterial-fungal biofilms are so little studied. Fungi play
an important role in the biodeterioration of materials:
They are implicated in several nosocomial infections and
are associated with other health issues, and they are
used in many industrial and biotechnological processes.
The importance of understanding mixed organism biofilm Plateau
measurement

formation, especially with the emergence of nosocomial

Biofilm

infections and problems to industry, is apparent. The aim

of this article is to discuss our current understanding Log
of bacterial-fungal biofilms and to examine methods for Lag
their study. It is also the aim to highlight areas of
bacterial-fungal biofilm dynamics that necessitate further
investigation. Time
Figure 1. A schematic illustration proposing the life cycle and
growth pattern, typically a sigmoidal growth curve, of a mixed
DEVELOPMENT OF BACTERIAL-FUNGAL BIOFILMS organism biofilm.

In order to discuss bacterial-fungal biofilms, it is important

to briefly mention the fundamental differences between factors in great detail but to describe the events that lead
bacteria and fungi. Fungi are eukaryotic organisms and to bacterial-fungal biofilm formation (Fig. 1).
can be differentiated from bacteria by the fact that fungal The conditioning film is a crucial component in the
cells are normally larger and contain a nucleus, vacuoles, attachment of cells to a surface. When a surface is
and mitochondria. Collectively, fungi are a large group of exposed to the environment, for example, in aqueous
functionally diverse species and can be classified into three systems, a layer of organic deposits, including proteins
groups of major practical importance: moulds, yeasts, and and humic acids, forms. This is termed the conditioning
mushrooms. film. Molecules and small particles (<0.01–0.1 µm) rapidly
In spite of the variation in the definitions available adhere to the surface and can induce changes in the
to describe biofilms, and the inherent heterogeneity of surface properties. These range from the acquisition of a
biofilms, reports of biofilms always describe the following: net negative charge to changes in the hydrophobicity of
A biofilm is always associated with a surface-interface the surface. The composition of the conditioning film will
at which cells accumulate and biofilm microorganisms vary with the type of surface and the environment. It is
produce an exopolymer matrix within which the cells are generally assumed, though with no conclusive evidence
enclosed. There are a number of factors that will affect that the conditioning film across the surface is uniform in
bacterial-fungal biofilm formation and development. To both composition and coverage. It has been proposed that
date it is assumed that the same factors that affect the strength of biofilm formation becomes dependent on
bacterial biofilm formation will also affect bacterial- the cohesiveness of the conditioning film to the surface,
fungal biofilm formation. These include biological factors, rather, than on the direct bacterial contact with the bare
including competition, chemical factors, including the substratum surface (2).
nature of the surface, and physical factors, including flow Most surfaces and cells have a net negative charge;
conditions. It is not the aim of this article to describe these therefore, when a cell comes into contact with a surface,
552 BIOFILMS: BACTERIAL-FUNGAL BIOFILMS

a repulsion barrier is established between the cell and caplike appendages, which unfurl to form viscous threads,
the surface. For attachment to take place this must sticky appendages surrounding the spore, and tufts of
be overcome. There are a number of attractive forces, fibrillar appendages.
including van der Waals forces and hydrogen bonding, Fungal spores, like bacteria, make repeated contact
which play a role in bacterial attachment but are less with surfaces. Fungi are thigmotropic, that is, they have
defined for fungal cells. The three step mechanism the ability to respond to localized physical contact, and
proposed by Busscher and Weerkamp (3) takes into as a result the fungus is able to release spore adhesins
account the different forces, which arise at different stages on contact with a surface, which then attaches it to the
of attachment. This hypothesis is based on the distance surface. The chemical structure of the adhesins range
between the cell and the surface. At distances greater from only carbohydrate or protein to a mixture of protein
than 50 nm only van der Waals forces operate; this stage and carbohydrate. Once attached to a surface, the fungal
is reversible. At distances between 10 and 20 nm both spore swells by adsorption of water, and germ-tubes
van der Waals forces and electrostatic repulsion forces develop and elongate to form young hyphae. Both the
are active. Adhesion at this stage is initially reversible but germ-tube and hyphae are enveloped in a mucilaginous
has the potential to become irreversible. In the third stage, sheath. This mucilaginous sheath has been likened to
where the separation distance is less than 15 nm van der the exopolymeric substances produced by bacteria, and
Waals forces, electrostatic forces, and specific interactions, is believed to play a similar role including anchorage,
including the production of adhesive exopolysaccharides protection from desiccation, adhesion sites for other
lead to irreversible bonding of the cell to the conditioning microorganisms and serving as a source of support and
film/surface. nutrition (5). Aquatic hyphomycetes have been shown
Fungi have been described as excellent colonizers of to be specifically adapted to their habitat by having a
surfaces (4). Fungal adhesion has a number of stages high germination rate and the production of tetraradiate,
that appear to be in common with bacterial biofilm branched, and filiform spores. The yeast Candida albicans
formation. Spores, like bacteria, have a negative surface produces pseudohyphae, which not only provide anchorage
potential. Consequently, the adhesion of fungal cells to a for the cells but also produce exoenzymes that aid in the
surface is subject to the same electrostatic forces operating penetration of the cell to tissue lining. It may be that
with bacteria. The initial contact of fungal spores to a all fungi have specific adaptations to their habitat but
surface may be described as passive, when spores are as yet many remain undiscovered. These habitat specific
entrapped or adhere to a surface by appendages, and appendages may limit the colonization and attachment of
active when the spore is stimulated to produce adhesive the fungi to a specific environment and it may be that the
mucilage (5). Spores may become entrapped on jagged habitat induces the expression of certain phenotypes. The
surfaces where they are held by physical forces. Passive success of biofilm formation, pure culture or otherwise,
attachment may be mediated by a number of spore is therefore dependent on the success of the organism to
appendages including mucilaginous sheaths, hamate, or adapt to its environment.

(a) (b)
B

F
I

B I

Figure 2. A mixed organism biofilm after 7 h growth on stainless steel using image analysis. The
biofilm consisted of seven organisms isolated from an industrial system: Stenotrophomonas
maltophilia, Pseudomonas alcaligenes, Flavobacterium indologenes, Alcaligenes denitrificans,
Rhodotorula glutinis, F. oxysporum, and F. solani. (a) B, biofilm; I, interstitial channels; M,
microcolony; S, stainless steel surface. Bar = 200 µm. (b) by varying the plane of focus different
section of the biofilm microcolony can be visualised. F, fungal biomass; arrows indicate mycelial
network. Bar = 50 µm.
 BIOFILMS: BACTERIAL-FUNGAL BIOFILMS 553

Investigations into the development a biofilm has

revealed that a typical three-stage sigmoidal growth curve S
occurs, much the same as a planktonic growth curve.
The model proposed by Bryers and Characklis (6) consists
of (1) initial biofilm formation (lag); (2) exponential
accumulation (log) and (3) steady state (plateau phase)
(Fig. 1). It is not known whether a hierarchical order of
colonization exists between bacteria and fungi. However,
the complexity of naturally occurring biofilms suggests
that this would occur. It is recognized that the order in
which cells attach to a surface may affect the attachment
of other cells; it has been demonstrated that this is a
species-specific event and would depend on the nature
of the surface (7). However, once the biofilm has reached
steady state conditions, the order of colonization may not
be of great importance compared with its overall function.
Figure 4. Filamentous fungal attachment (F. oxysporum and
BACTERIAL-FUNGAL BIOFILM ARCHITECTURE F. solani) to a PVC surface in flowing conditions. The fungal
hyphae extend out into the bulk fluid and are positioned in the
The range of biofilm structures that have been identified direction of flow (indicated by arrow). B, bulk fluid; F, fungal
vary from a thin layer of attached cells to more hyphae. Bar = 50 µm.
complex attached communities containing multiple species
interacting with each other. Biofilm structure and
architecture is strongly connected to the function required virtually impossible to obtain a consensus biofilm model.
for the survival and maintenance of the microorganisms In practice different models are available for different
within and has been shown to be linked strongly to growth conditions and a consensus of variables that
nutrient availability and flow conditions (8) (Fig. 3 and influence biofilm architecture.
Fig. 4).
As yet, there is no consensus model for the structure
of a typical biofilm. Initially a two-dimensional model for CHARACTERIZING BACTERIAL-FUNGAL BIOFILMS
biofilm formation was proposed (9). With the increased
knowledge of biofilm structure it has become apparent Mixed organism biofilms can be described as diverse
that biofilms are inherently heterogeneous, so three- functioning communities. A community describes an
dimensional models depicting biofilm structure have also association of interacting populations, usually defined
been put forward (10). Research has shown that there are by the nature of their interaction or the place where
many conditions, which contribute to biofilm architecture they live (11). The development of a mixed biofilm
and structure. The combination of species specificity leads to a variety of complex relationships involving
dictating biofilm architecture and the different variables, interspecies and intraspecies interactions. The number of
which also influence biofilm structure, means that it is interactions increases when there are more species within
the mixed population. The simplest interaction model
consists of a two-membered, or binary, population. In
this scenario there are only three possible responses that
a growing population can make to the presence of a second
population, that is, beneficial, detrimental, or neutral (12).
The interactions can occur in a reciprocal manner
and the outcome consists of a maximum of six types
of interaction: neutralism, mutualism, commensalism,
amensalism, predator-prey relationships and competition
(Table 1). However, another classification can be added
to include a direct interaction, if it involves physical
contact between individuals, or indirect if the nonliving
B environment is a necessary intermediary between the two
populations (13). The binary system is useful for defining
simple interactions; however, in natural communities,
interactions will be more complex, and more than one
type can occur with one species.
Although biofilms are considered in terms of microbial
Figure 3. Filamentous fungal attachment (Fusarium oxysporum ecology, the complex nature of mixed organism biofilms
and Fusarium solani) to a PVC surface in flowing conditions. The necessitates descriptions, which are applied to macroe-
fungal mycelia are positioned in the direction of flow (indicated cology. By measuring the diversity of a community, both
by arrow). B, bulk fluid; S, PVC surface. Bar = 50 µm. the variety of species and the relative abundance of each
554 BIOFILMS: BACTERIAL-FUNGAL BIOFILMS

Table 1. Interaction Terms Used to Describe the Effect may cycle quite regularly and population cycles will vary
That One Population Has on Another from species to species. The degree of variation in the
size of the population depends on both the magnitude of
Interaction Description
disturbance and on the inherent stability of the population.
Neutralism Occurs when neither population When a mixed organism biofilm reaches steady state or a
is affected by the presence of climax community it could be said that if all parameters
the other remain constant and there is no disturbance to the system
Mutualism Occurs when both members that it is a stable community.
derive some advantage by the
presence of the other
population THE OCCURRENCE OF BACTERIAL-FUNGAL BIOFILMS
Commensalism Occurs when one member of the
population benefits and the Most habitats contain a wide structural and functional
other member does not diversity of microorganisms, which is in part related to
Amensalism Occurs when the growth of one the heterogeneity of the habitat and the evolution of
of the populations is species, which exist there. Presently there are limited
restricted and the second reports describing bacterial-fungal biofilms, although this
population is unaffected is changing rapidly. Bacterial-fungal biofilms have been
Predator-prey Occurs when one member gains
isolated from a range of diverse locations including
directly at the expense of
another
photoprocessing tanks (16), automobile air conditioning
Competition Occurs when both populations systems (17) and even the black slime that appears on
are limited in terms of growth the base of shower curtains is teeming with bacteria and
or final population size by a fungi. There are many other habitats that are currently
common dependence on an being investigated with respect to microbial adhesion and
external factor required for biofilms; These include soil particles, plant surfaces, and
growth animal guts. Microbial adhesion and physiology are much
more difficult to investigate in these habitats because of
their diversity and range of conditions and knowledge of
species is taken into account. It has previously been dis- these areas should increase with advancing technology.
cussed that many microorganisms exhibit certain traits The reader should also be made aware that in many
that are particular to a habitat, and habitat specializa- examples describing bacterial biofilms the surrounding
tion increases in direct relation to diversity. That is, the environment will invariably consist of other organisms
more diverse a habitat, the more diverse are the organ- and these will include fungi. It should also be noted that in
isms. If a habitat is highly specialized, then diversity is the literature, bacteria and especially fungi are discussed
decreased but habitat specialization is increased. Diversity without reference to a biofilm mode of growth when it is
indices have been devised to express the diversity of sam- highly likely that in nature they will exist within a biofilm.
ple by using a single number. The most frequently used Biofilms play an integral role in the system in
is the simple totaling of species numbers to give species which they are present and are considered to either
richness (14). One of the indices that combine species rich- have a beneficial or detrimental effect on that system.
ness with relative abundance is the Shannon diversity The terms fouling and biofouling describe the presence
index (15). This method could be used to assess not only of undesired surface associated microbial growth and
the diversity of a sample but to also to monitor the effect are used frequently in the scientific literature. The
of disturbance to a system, for example, an antimicrobial uncontrolled and undesirable accumulation of biofilms
regime or predation. in medical and industrial systems has three primary
The reason that certain species grow together in a effects: physical damage, for example, corrosion and
particular environment will usually be because they tooth decay; decrease in the function of the system, for
have similar requirements for existence in terms of example, reduced efficiency of heat exchangers, and the
environmental factors such as nutrient availability and establishment of a reservoir for potential pathogens.
temperature. If species do grow in association with each
Medical Bacterial-Fungal Biofilms
other, many of them should have similar responses and
tolerances to environmental pressures. Over a period of Many recurrent infections are associated with bacterial
time, community composition will often change according biofilms and because of the nature of the medical
to the principles of succession. Succession involves the environment, these infections are mainly attributed to
immigration and extinction of species together with a single species. Any medical implant ranging from a
changes in their relative abundances. Primary succession catheter to a replacement heart valve has the potential
occurs on a bare surface. Organisms that are the first to be colonized by bacteria and fungi. Detachment of the
to colonize an area are termed primary colonizers and organisms from the biofilm often results in septicaemia,
several successive groups of species, termed secondary which may be responsive to drug therapy; however the
and tertiary colonizers, will then establish. The end result biofilm generally is not so responsive and therefore acts
of succession is known as a climax community. These as a continuous source of infection. As a result, implant-
stages are in common with the stages involved in microbial associated infections are difficult to treat and in most cases
colonization of a surface. Populations within a community the implant must be removed.
 BIOFILMS: BACTERIAL-FUNGAL BIOFILMS 555

The occurrence of monoculture biofilms in the medical of other organisms including filamentous gram-negative
environment is well documented and there are now Fusobacterium species (Fusobacterium nucleatum) and
reports emerging concerning the infectious nature of spirochetes such as Borrelia species. Dental biofilms
fungal biofilms. Many fungi that are now increasingly illustrate the complex nature of mixed organism biofilms
responsible for nosocomial infections were previously and highlight some of the aspects covered in the literature.
thought to be of low virulence. The increase in the number
of fungal-associated infections is related to the increase Environmental and Industrial Bacterial-Fungal Biofilms
in the use of antibiotics. There are many examples of
Biofilms that accumulate on suspended particles in rivers,
pathogenic fungi in the literature, for example, C. albicans lakes, and marine systems are considered beneficial
and C. neoformans. It is unlikely, however, that in a to their environments because they play an essential
medical environment the causative agent of infection purification role by removing suspended, settled, and
would be attributed to bacterial-fungal biofilms. Infections dissolved organic material. Freeman and coworkers (21)
are caused separately by bacteria and fungi, and presently described river biofilms as a trophic link between dissolved
it is considered extremely rare if found otherwise, nutrients in the water column and the higher trophic levels
although reports have demonstrated that C. albicans is of the ecosystem. These river biofilms possess the ability
the dominant and causative organism in a bacterial- to trap nutrients by ion exchange processes and thereby
fungal biofilm isolated from a voice prosthesis. However, increase the nutrient concentration within the biofilm.
with the increase in antibiotic resistance it may not be Natural biofilms are able to biodegrade organic compounds
long before bacterial-fungal biofilms are the combined and transform inorganic compounds as part of their
causative agents of infection. natural metabolic pathway. The role of bacterial-fungal
Biofilm formation by C. albicans has generated much biofilms in the degradation of environmental pollutants
interest because it is a major fungal pathogen of humans is not well understood. The ability of fungi to degrade
and can be fatal to immunocompromised patients, for environmental pollutants has been well documented but
example, those with acquired immunodeficiency syndrome these studies rarely include a biofilm model. It is assumed
(AIDS). Candida albicans is a dimorphic, diploid fungus. that in a natural environment such as the soil habitat,
The pathogenicity of this fungus is thought to be related bacteria, and fungi do form biofilms (22) and therefore
to the production of pseudohyphae. The pseudohyphae the degradation of environmental pollutants occurs in a
produce exoenzymes, which penetrate cell tissue and biofilm state where it is also assumed that there is a higher
medical implants. Analysis of C. albicans biofilms have level of enzyme activity. The degradative capabilities of
shown that it consists of a dense network of yeasts, hyphae, biofilms have been exploited in wastewater treatment
and pseudohyphae (18). Another adaptive mechanism systems including trickling filters and fixed or fluidized
exhibited by C. albicans is the existence of different strains bed reactors. Such systems are based on particles that
within the same location. The ability of the organism provide a large surface area for biofilm formation. The
to achieve this is thought to be governed by its genetic mixed biofilms that develop on these surfaces break down
makeup. There is no sexual stage associated with this a wide range of chemical substances found in wastewater.
fungus and it has what are termed as specific repetitive Bacterial-fungal biofilms have been recognized to play a
sequences, which contribute to the organism’s ability to role in the accelerated deterioration of buildings, including
increase strain variation. It is likely that the biofilm mode the well known and dreaded dry rot fungus. The degree
of growth enhances the pathogenicity and recalcitrance of colonization and subsequent deterioration is dependent
of C. albicans by offering protection from antimicrobial on the moisture content of the construction materials.
agents and by facilitating the optimum utilization of The formation of bacterial-fungal biofilms in buildings has
nutrients. many serious health implications including the harboring
Dental biofilms are also examples of mixed organism of allergenic fungal and actinomycete spores. Stachybotrys
biofilms. Colonization of the oral cavity is a continuous atra grows on damp wallpaper and other domestic
process. The relatively smooth surfaces of the teeth surfaces. It produces potent macrocyclic trichothecene
are exposed to frequent mechanical cleaning by the toxins and is associated with chronic health problems.
tongue, cheek, saliva, or toothbrush; however, these The building can also act as a reservoir for diseases
actions have proved ineffective at preventing tooth including Legionnaire’s disease. It is important therefore
decay/biofilm removal alone. The very structure of that the potential to form bacterial-fungal biofilms is
teeth, a nonshedding surface, favors the retention of assessed in the architectural design and that plans should
food and thus provides numerous sites for harboring include minimizing the ingress of water into the building
dental biofilms and it has been reported that there construct.
are between 200 to 500 resident bacterial species in In industrial systems, biofilms are particularly delete-
the mouth. Species of Candida including C. albicans rious because of their ability to corrode the surface with
have also been isolated from dental biofilms. It has which they are associated, a phenomenon known as micro-
been found that the initial attachment of bacteria to bial induced corrosion (MIC). The corrosion of metal ships,
the conditioning film is relatively species-specific and pipelines, and oil rigs is an extremely costly problem.
involves the cooperation and coaggregation of only a Marine biofilms that form on ship hulls increase drag and
few gram-positive streptococci including Streptococcus frictional forces, which results in an increase in fuel con-
sanguis, S. sobrinus, S. mutans and S. mitis (19,20). The sumption and a decrease in engine efficiency. In water
growth of this microbial film encourages the proliferation distribution pipelines the formation of biofilms reduces
556 BIOFILMS: BACTERIAL-FUNGAL BIOFILMS

the pipe diameter, which affects flow and in cooling sys- Once any type of biofilm is established in a system,
tems an estimated £500 million annually is lost due to an complete eradication is very difficult and in some instances
inefficient operating system. irremediable. The biggest problem associated with the con-
trol of biofilms is that biofilm cells are less susceptible to
CONTROL OF BACTERIAL-FUNGAL BIOFILMS antimicrobial agents than their planktonic counterparts.
The mechanisms for this increased resistance remains to
The control of biofilms has generated a huge commercial be elucidated. The production of exopolymeric substances
market in designing and developing new antimicrobial by biofilm cells are believed to play an important role in
regimes. A wide variety of biofilm control strategies have the protection of the cells from antimicrobial agents and
been tested with varying degrees of success. There are two was thought to be the basis of biofilm resistance (24). It is
main approaches to the control of biofilms: the prevention now believed by many research groups that the reduced
of initial colonization and subsequent biofouling, and penetration of antimicrobial agents through the extra-
the development of removal/control strategies against cellular polymers are insufficient to account for much of
an existing biofilm. Biocides and antibiotics have been the observed resistance (24,25). There is no doubt that
incorporated into different surfaces in an attempt the exopolymeric matrix can function as an ion exchange
to prevent microbial colonization. However, to date column and excludes large, highly charged molecules; how-
these types of control strategies tend only to prevent ever, most solutes will equilibrate across it and access the
biofilm formation for a limited period and subsequent resident population (24). The extent of penetration and
biofouling occurs as if there was no antimicrobial agent concentration of the antimicrobial agent that reaches the
in place. Present day control strategies of industrial population is dependent on the thickness of the biofilm.
biofilms are aimed at maintaining partial control and This is a transport limitation phenomenon common to
a minimum contamination level. In many cases, infection control strategies, which aim to remove an established
by antibiotic-resistant organisms, for example, methycilin biofilm.
resistant Staphylococcus aureus and vancomycin resistant In addition to the physical barriers that a biofilm
Enterococcus faecium, biofilm eradication is not achieved presents, individual cells possess structures, which pre-
and infection is fatal. Fatalities in these incidences are vent the biocide reaching its target(s). In the case of
higher in immunocompromised patients. The ultimate gram-negative bacteria the presence of the lipopolysac-
aim is obviously to design an antimicrobial regime that charide cell envelope wall offers a supplementary barrier.
will permanently prevent biofouling. This control strategy This structure has a significant moderating influence
would be one that was entirely specific for the target on the penetration of both hydrophilic and hydropho-
organisms and would have no other detrimental effect. bic molecules, establishing a molecular weight cutoff
This is currently a problem that needs to be addressed. for the passage of molecules through water filled pores
For example, infections caused by C. albicans can be (porins) and requiring optimum lipophilic properties for
treated with antifungal antibiotics including amphotericin the progress of hydrophobic biocides (23,26,27). The walls
B, however, the effect of the antibiotic on the patient often of fungal hyphae and spores also contain a variety of
has more serious side effects than the original infection. components, which help protect them from chemical treat-
Paints that have been used to protect ship hulls have been
ment. Melanin components are chemically very stable and
shown to have an irreversibly detrimental effect on marine
are frequently found in many surface moulds. The pro-
life, namely, in generations of dog whelks, which have lost
duction of melanin in the cell wall in response to an
the ability to reproduce.
antimicrobial regime has been shown to contribute to
Antimicrobial agents may exert what are termed static
the resistance of fungal hyphae to lytic enzymes (28),
or cidal effects although the mechanism of action of each
oxidizing agents such as hypochlorite (29), and radia-
may differ (23). The term static is used to describe the
tion (30). Regardless of cell wall structure, all cellular
prevention or inhibition of growth by an antimicrobial
agent measured under conditions in which growth would components offer opportunities for nonspecific binding of
normally occur. The effect is reversible as removal or the antimicrobial agent, thereby depleting the chemical
neutralization of the agent will allow the cell to reestablish challenge (23).
growth. Cidal effects are the result of damage to cellular Another problem that is of primary concern for the
structures and function and cidal effects are irreparable control of fungal biofilms is their location. Many types
and irreversible. In order for the antimicrobial compound of fungi, such as the moulds, are found to grow on the
to have its desired effect, the compound must be taken interior and exterior of most buildings and the success
up by the cell, move to the target sites and then react of colonization by these fungi is related to the moisture
with the target site. Different antimicrobial agents have content of the building construct. If a building is poorly
different modes of action and it would be impossible to insulated there will be an increase in the amount of
cover them all within this chapter. It is common practice condensation produced, which in turn will increase the
to use a combination of antimicrobial agents to achieve diversity of the biofilm. Introducing equipment such as
a synergistic effect, for example, to use a combination dehumidifiers into the home environment is not always a
of antibiotics, which will have a complimentary and practical solution. Fungi are able to penetrate deeply into
enhanced effect without disrupting the mode of action of the building construct and emerge on wallpaper and wood,
either of the antibiotic. This is referred to as combination and it is not always possible because of safety reasons
therapy. to provide a chemical solution for domestic purposes. The
 BIOFILMS: BACTERIAL-FUNGAL BIOFILMS 557

areas of contamination may also be located in places that difference between planktonic and biofilm cultures. It
are not suitable for disinfection on a daily basis. is now widely accepted that this is not the case. If the
There are many other proposed mechanisms, which are level of contamination in industrial systems was assessed
thought to play a crucial role in the protection of biofilm by monitoring the number of planktonic cells, a serious
cells from an antimicrobial regime. It is thought that underestimate of the level of contamination would ensue
because of the structure of the biofilm, nutrient gradients as it has been reported that typically for every one
will exist where cells located in the upper sections of planktonic bacterium there are between 1,000 and 10,000
the biofilm will be exposed to higher levels of nutrients cells attached to the surface (31).
compared to those cells located at the surface, which are The ultimate aim in designing any experimental
assumed to be exposed to less nutrient rich conditions. apparatus is to include the major components of the
This will affect a number of physiological actions including system under investigation so as to provide the most
enzyme activity. As a result, the growth rate of the biofilm accurate model possible. Table 2 highlights some of the
is believed to be slower than planktonic cells, which measurements that can be evaluated from experimental
have the high doubling-time capacity. Other defensive model biofilms. The results obtained would then be
mechanisms activated by bacteria and fungi in response more reflective of in vivo conditions. In practice, most
to stressful conditions include the viable but nonculturable experimental biofilm models only represent certain aspects
state (bacterial) and the production of spores. of a system under investigation and tend to fall into
It is widely recognized that bacteria exposed to stressful in to two main categories: replicative, which encompass
conditions experience what is termed as sub lethal injury. a wide range of complex environmental variables, and
This is when a cell maintains minimal metabolic activity investigative, which are usually simpler and enable the
but viability is still retained. The genes that are thought control of a variety of influencing factors (32,33). The
to regulate these responses are expressed on entry to techniques used to study biofilms can either be destructive,
stationary growth phase where a slow growth rate would which involves the disruption and destruction of the
allow the cell an adaptation period to many stressful biofilm, or they are nondestructive, which permits the
conditions including antimicrobial regimes and heat shock. monitoring of the biofilm in its natural state. There are
In order to monitor the success of an antimicrobial regime, currently a wide range of experimental models available
an accurate recovery protocol must be established to take to study biofilms and it is not our aim to discuss
into account cells that are stressed. Quantification of cells
them all, but rather to discuss those that would be
in a stressed state by conventional methods, including
significant in increasing our understanding of bacterial-
plate counts, may lead to the underestimation of the level
fungal biofilms.
of contamination.
The Modified Robbins Device (MRD, Fig. 5) was initially
developed to study biofouling in industrial pipelines (34).
METHODS TO STUDY BACTERIAL-FUNGAL BIOFILMS Since then it has been used to study biofilms from a number
of different systems. The MRD consists of an enclosed
It has been common practice in the past to directly lumen with 25 replaceable sampling studs. Different
compare laboratory planktonic data directly to biofilm surfaces can be attached to the studs and removed for
conditions based on the assumption that there was no analysis by a number of different techniques. The MRD can

Table 2. Examples of Some of the Measurements That Can Be Assessed from

Experimental Model Biofilms

Measurement Method

Microbial activity Viable cell counts Culture techniques

Metabolic activity ATP levels
Fluorescent staining
Molecular techniques
Biofilm constituents Exopolysaccharide Protein content
Total carbohydrate
Fluorescent staining
Biofilm growth Thickness Image analysis
Biomass Culture techniques
Quantification Fluorescent staining
Surface area
Biofilm formation Structure Microscopy
Development — Image analysis
— Scanning electron microscopy
— Transmission electron microscopy
— Confocal scanning laser microscopy
— Environmental scanning electron microscopy
— Fluorescent microscopy
558 BIOFILMS: BACTERIAL-FUNGAL BIOFILMS

Stud/sampling port. qualitative information regarding surface colonization

is also possible (35). The flat plate reactor (Fig. 6) was
developed at the Center for Biofilm Engineering, Bozeman,
Montana. In this apparatus, different surfaces can be
examined and compared (35). The surface can also be
removed for other types of analyses. The disadvantage
Experimental surface attached. The surfaces are 1cm in diameter of the flat plate reactor is that it is constrained
and are fixed to black rubber backing discs which fit directly into by the channel thickness. This model of flow cell is
the stud. The surfaces are easily removed from the stud for analysis.
limited by the nature of the working distances of
Figure 5. Diagrammatic representation of an MRD showing the microscope objectives, resulting in a compromise
direction of flow (block arrows) and position of studs.
between magnification and hydrodynamics (37). Use
of the square glass tube reactor can overcome this
problem and facilitates higher magnification of the
(a) biofilm, allowing the flow cell to be viewed from
above (35). The square glass tube reactor flow cell (Fig. 6)
was initially designed to replicate biofilm fouling in
Surface
industrial pipelines. Stoodley and coworkers (38) have
Perspex base characterized the flow hydrodynamics of this system
using the relationship between the friction factor and
the Reynolds number, and found that they fit established
equations describing laminar and turbulent flow through
a smooth pipe. This system also permits visualization of
biofilm formation under both laminar and turbulent flow
(b) concurrently (38).
3 mm Another technique that has also been instrumental in
increasing our understanding of biofilm formation and
Figure 6. Diagrammatic representation of (a) a flat plate reactor structure is confocal scanning laser microscopy (CLSM).
and (b) a square glass tube reactor. Block arrows indicate CLSM operates using a krypton/argon laser, which excites
direction of flow. The required reactor would be place on a
fluorophore dyes present within the sample. The resulting
microscope stage and images captured via a camera and analysed
fluorescence is detected by photomultiplier tubes and a
using the appropriate software.
digital image is obtained. Alteration of the focal depth
(z plane) and the subsequent acquisition of the x-y plane
be incorporated into a variety of experimental models such images (parallel to the surface) enables a series of optical
as a batch recirculating and flow through culture systems. sections to be collected that are then processed by a
It can be connected to a chemostat as a recirculating or flow computer using image analysis software to create a three-
through culture system, which would allow the control of dimensional image. CLSM is an effective tool to study a
growth conditions and thereby permit the assessment of wide range of biofilm features, which include physiological
growth rate as a function of biofilm resistance. profiles and structural heterogeneity that would be
The main drawback with the MRD system is that particularly pertinent to mixed organism biofilms. CLSM
it is a destructive technique and involves scraping of permits the visualization of biofilm formation in situ and
the biofilm for analysis of biofilm accumulation and therefore allows the study of the biofilm in its natural
metabolic counts. Other drawbacks with the MRD system hydrated state. A number of other techniques could also be
are the inability to visualize the biofilm in situ, and used in combination, including microelectrodes to monitor
possible nutrient gradients along the length of the pH, and oxygen and nutrient gradients, resulting in a
lumen from the inlet to the outlet. Hydrodynamics comprehensive profile of the biofilm. Biofilms that have
around the sampling stud may also be compromised (35). been examined using CLSM are shown to be heterogeneous
Nevertheless, the information obtained from the MRD and consist of microcolonies or cell clusters separated by
model is valuable and allows comparisons to be drawn interstitial voids and channels (8). These channels have
with other techniques. been shown to act as transport channels (39) and therefore
Although valuable information can be obtained from the use of CLSM would allow the study of transport
models that involve the disruption of the biofilm, phenomenon within the biofilm, which would be relevant
the continuous and nondestructive analysis of biofilms in assessing the delivery of an antimicrobial agent to the
is essential in understanding biofilm processes (36). biofilm.
In situ monitoring techniques have advanced with It has been previously discussed that microorganisms
recent developments in image analysis technology. The are not culturable by traditional culture methods. Recent
advantages of this technique include the monitoring advances in molecular biology have circumvented this
of the biofilm in a hydrated state, and real-time problem. These methods include gene probing, DNA
analysis. Using digital image capture, accumulation hybridization, polymerase chain reaction (PCR) and
rates can be calculated by comparing captured images reporter gene technology. In 1994, a novel marker system,
with those at the outset of the experiment, allowing the green fluorescent protein (GFP) became available (40).
quantification of the growth kinetics of the biofilm; The GFP is a 27-kDa polypeptide, which converts the blue
 BIOFILMS, CONDITIONING FILMS 559

chemiluminescence of the Ca2+ -sensitive photoprotein, 15. M. Kent and P. Coker, Vegetation Description and Analysis:
aequorin, into green light (40). This fluorescence was first A Practical Approach, John Wiley & Sons, Chichester, U.K.,
observed in the jellyfish Aequorea victoria. GFP has been 1992.
successfully used in eukaryotic systems and there are an 16. K. T. Elvers, K. Leeming, C. P. Moore, and H. M. Lappin-
Scott, J. Appl. Microbiol. 84, 607–618 (1998).
increasing number of reports of the success that GFP has
to prokaryotic systems. GFP is stable in the presence of 17. R. B. Simmons, L. J. Rose, S. A. Crow, and D. G. Ahearn,
Curr. Microbiol. 39, 121–145 (1999).
many denaturants, high temperatures (65 ° C) and a range
of pH (6–12). It is therefore an attractive marker system 18. G. Baillie and L. J. Douglas, Antimicrob. Agents Chemother.
42, 2146–2149 (1998).
to monitor individual cells within a biofilm. GFP is now
commercially available in two other colors, red and yellow, 19. J. Pratten, P. Barnett, and M. Wilson, Appl. Environ. Micro-
biol. 64, 3515–3519 (1998).
and has also been applied to fungal cultures.
20. D. J. Bradshaw, P. D. Marsh, G. K. Watson, and C. Allison,
Biofouling 11, 217–226 (1997).
CONCLUSION 21. C. Freeman et al., Hydrobiology 20 (1995).
22. R. G. Burns, Structure and Function of Biofilms, John Wiley
There still remains many aspects of bacterial-fungal & Sons, Chichester, U.K., 1989.
biofilms to investigate. This paper has provided areas for 23. S. P. Denyer, Int. Biodeter. Biodegr. 36, 227–245 (1995).
discussion and has given insights into the complex nature 24. D. G. Allison and P. Gilbert, J. Ind. Microbiol. 15, 311–317
of studying mixed organism biofilms. Characterizing (1995).
bacterial-fungal biofilms in terms of macroecology may aid 25. W. W. Nichols, Bacterial Biofilms in Medicine and Industry,
our understanding of such heterogeneous communities. Bioline, Cardiff, Wales, 1993.
Fungal colonization of a surface has many stages in 26. P. Gilbert and N. Wright, Preservatives in the Food, Pharma-
common with bacteria, including the production of ceutical and Environmental Industries, Blackwell, Oxford,
exopolymeric substances. Bacterial-fungal biofilms cause U.K., 1987.
considerable problems in many systems, particularly 27. A. D. Russel, J. Appl. Bacteriol. 71, 191–201 (1991).
within industry. They may act as a source of pathogens and 28. A. T. Bull, Arch. Biochem. Biophys. 137, 345–346 (1970).
because of their inherent resistance to many antimicrobial 29. E. S. Jacobsen and S. B. Tinnel, J. Bacteriol. 175, 7102–7104
regimes are a serious health hazard. The treatment (1993).
of bacterial-fungal biofilms is additionally problematic 30. Y. L. Wang and A. Casadevall, Appl. Environ. Microbiol. 60,
because the biofilm cells have an altered physiology and 3864–3866 (1994).
morphology, which promotes their success in a particular 31. S. A. Blekinsopp and J. W. Costerton, Trends Biotechnol. 9,
habitat. The advances in technology are aiding in our 138–143 (1991).
understanding of biofilms with the ultimate aim to control 32. J. Rayner and H. M. Lappin-Scott, Environmental Monitor-
biofilm formation in the first instance. ing of Bacteria, Humana Press, Totowa, N.J., 1999.
33. P. Gilbert, Life and Death of Biofilm, BioLine, Cardiff, Wales,
1995.
BIBLIOGRAPHY
34. M. J. Bale, J. C. Fry, and M. J. Day, Appl. Environ. Microbiol.
1. J. W. Costerton, G. G. Geesey, and K.-J. Cheng, Sci. Am. 238, 54, 972–978 (1988).
86–95 (1978). 35. L. Hall-Stoodley, J. Rayner, P. Stoodley, and H. M. Lappin-
2. R. Bos, H. C. van der Mei, and H. J. Busscher, FEMS Scott, Environmental Monitoring of Bacteria, Humana Press,
Microbiol. Rev. 23, 179–230 (1999). Totowa, N.J., 1999.
3. H. J. Busscher and A. H. Weerkamp, FEMS Microbiol. Rev. 36. J. C. McCoy, J. D. Bryers, J. Robbins, and J. W. Costerton,
46, 165–173 (1987). Can. J. Microbiol. 27, 910–917 (1981).
4. M. V. Jones, The Life and Death of Biofilm, BioLine, Cardiff, 37. W. Characklis, Biofilms, Wiley & Sons, New York, 1990.
Wales, 1995. 38. P. Stoodley, Z. Lewandowski, J. D. Boyle, and H. M. Lappin-
Scott, Biotechnol. Bioeng. 57, 536–544 (1998).
5. E. B. G. Jones, Mycol. Res. 98, 961–981 (1994).
39. D. deBeer, P. Stoodley, and Z. Lewandoski, Biotechnol. Bio-
6. J. Bryers and W. Characklis, Water Res. 15, 483–491 (1981).
eng. 44, 636–641 (1994).
7. S. McEldowney and M. Fletcher, Arch. Microbiol. 148, 57–62
40. M. Chalfie, Photochem. Photobiol. 62, 651–656 (1994).
(1987).
8. P. Stoodley, J. D. Boyle, D. deBeer, and H. M. Lappin-Scott,
Biofouling 14, 75–90 (1999).
9. P. A. Wilderer and W. G. Characklis, Structure and Function
of Biofilms, John Wiley & Sons, New York, 1989. BIOFILMS, CARBON TRANSFORMATIONS IN.
10. J. W. Costerton et al., J. Bacteriol. 176, 2137–2142. See ACTIVITY AND CARBON TRANSFORMATIONS IN BIOFILMS
11. R. E. Ricklefs, Ecology, 3rd Ed., W. H. Freeman and Com-
pany, New York, 1990.
12. J. H. Slater and A. T. Bull, Companion to Microbiology,
Longman Group Ltd., London, U.K., 1978.
13. A. G. Fredrickson, Annu. Rev. Microbiol. 31, 63–87 (1977).
14. A. E. Magurran, Ecological Diversity and its Measurement, BIOFILMS, CONDITIONING FILMS.
Croom Helm, London, U.K., 1988. See CONDITIONING FILMS IN AQUATIC ENVIRONMENTS
560 BIOFILMS IN NATURAL AND DRINKING WATER SYSTEMS

BIOFILMS: EXTRACELLULAR ENZYMES. and genetic exchange are facilitated by the close and
See EXTRACELLULAR ENZYMES IN BIOFILMS ordered spatial arrangement of the cells (3,4). Biofilm
bacteria are, to a certain extent, sheltered from harm-
ful substances such as disinfectants, antibiotics, anti-
bodies, toxins, and also from viruses. Bacterial growth
rates are often enhanced within biofilms because of the
BIOFILMS, EXTRACELLULAR POLYMERIC accumulation of nutrients at interfaces and the coop-
SUBSTANCE. See EXTRACELLULAR POLYMERIC SUBSTANCES erative mobilization of nutrients by different organisms
(EPS): STRUCTURAL, ECOLOGICAL AND TECHNICAL ASPECTS of the microbial community. Thus, microbiologists have
to study biofilm communities (1–4) to catch a glimpse
of the true nature of microbial communities in the
environment.

BIOFILMS, FORMATION RATE IN WATER

DISTRIBUTION SYSTEMS. See ASSIMILABLE ORGANIC BIOFILM COMPONENTS
CARBON (AOC) IN TREATED WATER: DETERMINATION AND
SIGNIFICANCE The structure and composition of biofilms can vary widely,
depending on the environmental conditions and the popu-
lation structure of the microbial community. The extracel-
lular polymeric substances (EPS) excreted by the attached
bacterial population, forming a slimy, highly hydrated
BIOFILMS IN NATURAL AND DRINKING matrix that acts as glue for all biofilm components, should
WATER SYSTEMS certainly be regarded as a key constituent of biofilms.
EPS generally consist of polysaccharides, proteins, nucleic
WERNER MANZ acids, and heterogeneous polymeric compounds (5). The
ULRICH SZEWZYK volume of the EPS matrix, in which the microorganisms
Technical University of Berlin are embedded, normally clearly exceeds the volume of the
Berlin, Germany cells. Because of the reduced diffusion of substances, the
EPS matrix acts as a protective barrier against harmful
JOHN R. LAWRENCE
chemical compounds and viruses, retains extracellular
National Water Research
Institute
enzymes, and leads to the establishment of gradients
Saskatoon, Canada across the biofilm (6).
Bacteria generally form the dominant population, but
eukaryotic organisms are also present in most biofilms,
The concept of bacterial behavior and growth has long
although they have been less extensively studied than
been influenced by studies of bacteria grown in liquid
the prokaryotic community. These protozoa have an
culture in the laboratory. In natural and engineered
important function as predators, influencing not only
aquatic systems, however, microorganisms tend to grow
biofilm thickness and structure but also the species
preferentially in organized communities at almost all
composition of the bacterial population (7). Apart from
available interfaces. These organized structures, which
protozoa, fungi can occur in biofilms, as can algae in
are best defined as ‘‘cells immobilized at a substratum
phototrophic biofilms. Although the EPS matrix was
and frequently embedded in an organic polymer matrix
thought to prevent viruses from infecting biofilm bacteria,
of microbial origin’’ (1), are commonly referred to as
bacteriophages were recently found to be able to infect cells
biofilms. Costerton and coworkers (2) later broadened
and multiply in single-species laboratory biofilms (8,9).
the definition to include microbial aggregates in the
Most biofilms are dominated by organic matter, namely
free-water phase (‘‘matrix-enclosed bacterial populations microbial cells and their excretions, and usually to a minor
adherent to each other and/or to surfaces or interfaces’’), extent by depositions of inorganic material like quartz
however, the widespread occurrence and importance of grains and iron particles. In some cases, however, biofilms
eukaryotic microorganisms in biofilms is neglected in this can also be largely dominated by inorganic compounds.
definition. Biofilms are ubiquitous in all aquatic habitats, In drinking water biofilms, for example, certain types of
ranging from thick cyanobacterial mats, biofilms covering bacteria can be responsible for massive deposits of iron
stones and pebbles in rivers and lakes, immobilized and manganese particles (7).
cells in wastewater treatment reactors, surface-associated
growth in drinking water distribution systems, to
biofilms of medical relevance, such as dental plaque, FORMATION AND STRUCTURE OF BIOFILMS
bacterial growth on gut endothelial cells or on artificial
implants.
Colonization of Surfaces and Biofilm Development
Unlike bacteria multiplying as single cells in the
planktonic phase, biofilm bacteria grow in a highly The physicochemical properties of a surface submersed
structured microenvironment, where diffusion of sub- in an aquatic system not only influence the adsorption
stances is affected by the properties of the biofilm and of organic molecules and ions to the surface, the so-called
where metabolic interactions, cellular communication, ‘‘antenna-effect,’’ but also affect bacterial adhesion (10,11).
 BIOFILMS IN NATURAL AND DRINKING WATER SYSTEMS 561

Bacteria in aquatic systems can reach a surface either in many natural ecosystems, and in laboratory model
by random Brownian displacement, by fluid dynamic biofilms, to the layered, densely packed biofilms, forming
forces transporting small particles to the viscous boundary cyanobacterial mats or dental plaque. The architecture of
layer at a surface, by random flagellar- or pili-based biofilms is controlled by many parameters. The kind and
motility, or by active chemotactic movement toward concentration of nutrients present in a habitat certainly
the nutrient-enriched layer at the surface. The initial has a tremendous impact on biofilm thickness and over-
contact is followed by reversible or irreversible adhesion all structure (22), as well as on the species composition
of the cell to the surface. To describe this process, the of the microbial community. Mass transfer of nutrients,
Derjaguin-Landau and Verwey-Overbeek (DLVO) theory, metabolites or inhibitors, diffusion of oxygen, salts, and
which was developed for colloid chemistry and describes ions create gradients across the biofilm and induce the
changes in Gibbs energy of interaction as a function of formation of microniches, influencing microbial growth
the distance between two objects, has been invoked. In rates, production of extrapolymeric substances, and inter-
contrast to colloid particles, bacteria might play a more actions between the microorganisms. Hydrodynamic forces
active role in adhesion because of cellular appendages can become important in shaping the biofilm architec-
such as fimbriae, flagella, pili, cell surface adhesins, and ture. When grown under turbulent flow conditions, where
because of their motility. At small distances of separation, high-shear forces act on the biofilm, structural changes
short-range forces such as hydrogen bonding and ion from loose cell aggregates toward streamlined filaments
forces become effective. The physiological status of a and ripples can be observed. It has also been shown
cell can have a significant influence on its attachment that structural rearrangements occur within biofilms in
behavior, although the effect may differ from species response to environmental conditions, that is, turbulent
to species. Starved cells often show reduced motility, flow (23). Moreover, the microorganisms themselves shape
changes in cell surface hydrophobicity, and a reduction in their environment. Eukaryotic predators, including amoe-
cell size. For starved cells, both increased and decreased bae and other protozoa, and bacterial parasites such as
ability for adhesion has been reported (12,13), and these Bdellovibrio spp. may control the biomass and species com-
differences might be due to species-specific behavior position of biofilms. Bacterial shape, motility, production
and to differing starvation conditions generated in the of extracellular polymers, and cell signaling between bac-
experiments. Bos and coworkers (14) have discussed the teria, all play a crucial part for the architecture of mature
initial interactions of microorganisms with surfaces and biofilms. Substratum properties such as hydrophobicity,
the various methods for studying adhesion, aggregation, roughness, and electrostatic forces, on the other hand,
and coadhesion. play a substantial role in the initial stages of biofilm for-
Exopolymers produced by the adherent cells accu- mation but appear to be less significant in established,
mulate and act as glue between cells and the surface steady state biofilms.
and lead to an irreversible attachment of the bacte-
ria (6,15). Colonization of the surface may then proceed
Detachment and Dispersal
in a number of ways, depending on surface character-
istics and on the colonization behavior of the attached (see BIOFILM DETACHMENT, this Encyclopedia)
species (16). For example, in streams, a range of coloniza- Hydrodynamic forces, which influence the biofilm
tion behaviors have been described, including formation structure, are mainly responsible for the detachment and
of tightly packed, spreading, or branching colonies (17). sloughing off of clumps of cells or small parts of the
In the marine environment, a Psychrobacter sp. ini- biofilm. These large particles will be transported by the
tially formed microcolonies, followed by the development flow and sediment or attach to a substratum when flow
of compact, multilayered biofilms at hydrophobic sur- conditions permit it, thus contributing to the colonization
faces but developed only long, multicellular chains at of new surfaces. Biofilms, however, are not only subjected
hydrophilic surfaces (18). Time-lapse microscopy was used to passive detachment processes. Several studies have
and Pseudomonas aeruginosa was found to initially form demonstrated that individual cells might actively leave
a monolayer of cells on the surface, followed by forma- biofilms. Cell division can cause the active detachment
tion of microcolonies through aggregation of cells from of cells, involving different dispersal strategies, with
the monolayer, which led to a biofilm structure of dis- daughter cells leaving the biofilm, whereas the parental
persed microcolonies in the monolayer of cells (19). In cell remains firmly attached (17,24) or detachment taking
highly oligotrophic systems, like drinking water or sea- place during cell division, followed by dispersal or
water, biofilms can be restricted to a monolayer of reattachment (25).
cells (20,21).
Genetics of Biofilm Formation
Structure of Mature Biofilms
Attachment of bacteria is thought to induce an alteration of
The structure of mature biofilms in natural or engineered bacterial gene expression and metabolic activity. A positive
systems is as diverse as the habitats themselves, ranging effect of surfaces on bacterial growth had already been
from thin monolayers of dispersed cells in extremely olig- observed in the 1930s (26). Cells grown on solid surfaces
otrophic habitats such as drinking water distribution or on agar show physiological properties and behaviors
systems, patchy arrangements of matrix-embedded cell that are quite different from cells growing in the liquid
clusters interspersed with water-filled channels observed phase; a striking example is the coordinated multicellular
562 BIOFILMS IN NATURAL AND DRINKING WATER SYSTEMS

behavior displayed in swarmer cell formation of bacteria inability to build biofilms: the mutant strain forms only
such as Serratia liquefaciens (27), or induction of lateral a dispersed monolayer of cells and does not develop the
flagella in Vibrio parahaemolyticus (28). Molecular tools dense monolayer punctuated by microcolonies typical of
applied in pure culture studies have helped to identify the wild-type strain. This phenotype had so far been
a number of genes required for biofilm formation and observed in type-IV pilus mutants, which were defective
genes with altered levels of expression between sessile for twitching motility. Apparently, the Crc protein acts
and planktonic growth. Genes whose expression is as part of a signal transduction pathway that regulates,
altered in biofilms can be identified by using gene among other traits, pilus expression and thereby biofilm
reporter technology (29), whereas genes with an essential development (34).
function in biofilm formation can be detected by random Flagellar-mediated motility has been demonstrated as
transposon mutagenesis, followed by screening for biofilm an important factor in biofilm development for a number
formation defective mutants and determination of the of gram-negative bacteria including E. coli, P. aeruginosa,
genes involved (30). P. fluorescens, and Vibrio cholerae. For each of these
Altered gene expression has been reported for organisms, mutations in genes involved in flagellar-
Escherichia coli by analysis with reporter genes. Expres- mediated motility hindered biofilm formation. Apart from
sion of 38% of the genes was found to be affected by growth motility based on flagella action, twitching motility has
within biofilms. Some genes such as a gene involved in flag- been found to be essential for P. aeruginosa biofilm
ellin synthesis, were downregulated in sessile bacteria. formation. Twitching motility is based on type-IV pili and
Other cellular components showed enhanced expression; seems to be widespread among gram-negative bacteria.
examples are the OmpC porin, the high-affinity transport Twitching plays a role in the establishment of biofilm
system of glycine betaine, the colanic acid exopolysaccha- structure, especially in the formation of microcolonies,
ride, tripeptidase T, and the nickel high-affinity transport whereas flagellar action apparently is necessary to
system (31). These changes in gene expression show that generate cell surface contacts. Motility-defective mutants
bacteria rapidly adapt to the environmental conditions might, however, regain their biofilm formation capacity by
encountered in a biofilm, where osmolarity and cell density mutations in other genes such as the ompR gene discussed
are increased, and oxygen less available. in the following section, or they can be restored by addition
Genes with an essential function for biofilm formation of specific nutrients (32).
have been discovered by isolation of mutant bacteria The ability to adhere to a surface or to another cell
and analysis of the affected genes. Biofilm formation is crucial for the formation of biofilms. For example, cell-
can be triggered by external environmental cues and cell (intra or interspecific or multigeneric) coaggregation
by bacterial cell-cell signaling, which is also important among biofilm bacteria has been shown to be significant
in the establishment of the complex three-dimensional in a number of systems including aquatic habitats (35).
structure of biofilms. For the initial stages of biofilm Mutant strains of Staphylococcus epidermidis have been
formation, motility mediated by flagella or pili is essential, isolated, which were affected in their primary attachment
whereas adhesins and EPS play a crucial role for to polystyrene, but remained able to form multilayered
interactions between cells and between cells and the cell clusters. These mutants lacked four cell surface
surface (32). proteins, which are postulated to function as adhesins.
The significance of cell signaling has been recently Other mutants were defective for intercellular adhesion
demonstrated for a Pseudomonas aeruginosa mutant, and thus unable to form cell clusters and biofilms as a
which had a mutation in the lasI gene. LasI directs the result of a failure to produce a specific polysaccharide (36).
synthesis of an acylated homoserine lactone (AHL), which A mutant strain of E. coli has been isolated, which
acts as a signaling molecule between cells and activates the showed an increased biofilm-forming ability compared to
transcriptional activator LasR. The mutant P. aeruginosa classical laboratory strains. The mutant E. coli displayed
could only form flat, undifferentiated biofilms, which a single point mutation in the regulatory protein OmpR,
showed an increased sensitivity toward a biocide. When resulting in an increased expression of the particular
a synthetic signal molecule was added, biofilm growth fimbriae type Curli. Apparently, Curli acts as a cell surface
appeared to be normal (33). In Serratia liquefaciens, adhesin and is under the control of the EnvZ-OmpR two-
swarming motility has been found to be controlled by two component regulatory system (37). Other examples are
separate regulons, the fhlDC operon, a flagellar master genes coding for the mannose-sensitive type I pilus and
regulator, and the swrI gene, a putative homoserine the pilus adhesin FimH, which are essential for E. coli
lactone synthase. It appears that the AHL cell-signaling biofilm formation or the mannose-sensitive hemagglutinin
system controls the production of an extracellular pilus in a V. cholerae strain (32). For P. aeruginosa, genes
biosurfactant, which enables swarm cells to travel on top of the alginate synthesis pathway were found to be
of surfaces and facilitates the movement of differentiated upregulated on attachment and biofilm formation, and
swarm cells, whereas swarm cell differentiation itself expression of a putative alternative sigma factor required
(i.e., cell elongation and hyperflagellation) is controlled for expression of the critical alginate biosynthetic gene
by the flhDC operon (27). As mentioned in the preceding algD has been reported (38,39). Alginate is produced
section, environmental cues such as nutrient availability, in large quantities by mucoid P. aeruginosa strains,
not only affect the metabolic status of cells but also which is a critical pathogenic determinant during chronic
biofilm formation. In P. aeruginosa, a mutation in the infections in cystic fibrosis patients. In a V. cholerae
catabolite repression control (Crc) protein causes an strain, production of an exopolysaccharide encoded by the
 BIOFILMS IN NATURAL AND DRINKING WATER SYSTEMS 563

vps locus confers biofilm-forming capacity and chlorine successfully applied to investigate principles of microbial
resistance. Mutations within this locus result in an community ecology in a hot spring microbial mat (45),
inability to form biofilms (40). which is presented in more detail in the following section
on examples of natural biofilms.
As an alternative to the DGGE technique, restriction
IDENTIFICATION OF BIOFILM BACTERIA
fragment length polymorphism (RFLP) analysis of rRNA
genes amplified from microbial communities can be
(see BIOFILMS: BACTERIAL-FUNGAL BIOFILMS, this Encyclope-
applied to obtain specific fingerprints. After amplification
dia)
and subcloning of the PCR products, RFLP patterns
for the rRNA clones are determined with restriction
rRNA-Based Phylogeny and In Situ Hybridization
endonucleases (46). Alternatively, RFLP patterns can
The most remarkable recent advance in microbiology was be obtained for total 16S rRNA from a microbial
undoubtedly the development of molecular tools for the community (47). A modification to the RFLP technique,
identification and in situ detection of bacteria based termed terminal-RFLP (T-RFLP), consists of labeling one
on their evolutionary history. Sequencing of ribosomal of the PCR primers, followed by restriction digestion and
RNA (rRNA) genes from diverse organisms, which fragment detection by capillary electrophoresis, where
revealed the tripartite organization of life according to only the end-labeled fragment of each PCR product is
the domains Bacteria, Archaea, and Eucarya, eventually detected. DGGE and T-RFLP have been compared in a
led to the design of short, labeled oligonucleotide recent study, revealing that although both techniques
probes complementary to specific target sites on 16S or show similar results, T-RFLP was more sensitive than
23S prokaryotic rRNA. Ribosomal RNAs contain highly DGGE (47). Because both techniques rely on PCR-
conserved and variable sequence regions, which together procedures, they are subject to the same problems as
with the natural amplification effect due to their high cloning procedures with regard to chimeric sequences and
copy number in the cell, makes them ideally suited quantification of species abundance.
for in situ hybridizations with oligonucleotide probes,
targeting phylogenetic groups ranging from domains to Detection of Known Bacterial Strains
species (41,42).
For a first analysis of the phylogenetic diversity of For the detection of known cultivated bacteria in microbial
a microbial biofilm community, in situ hybridizations biofilm communities, the application of labeled antibodies
with already existing probes are performed in a top-to- or species-directed PCR procedures might be useful.
bottom approach, where oligonucleotides with increasing Both methods, which depend on the prior cultivation
phylogenetic narrowness will be applied. If a complete of the species of interest, are mainly used for the
analysis of the microbial community is attempted, bulk detection of pathogenic bacterial species. Thus, Legionella
DNA is extracted from the habitat, rRNA amplified pneumophila could be visualized by immunogold- or
by PCR with universal primers, subcloned, subjected fluorescein-labeled monoclonal antibodies in drinking
to DNA-sequencing, and finally compared to existing water biofilms (48). Problems with the use of antibodies
sequences in published databases. However, the data thus in natural biofilms are posed by hindrance of antibody
obtained have to be interpreted with caution because penetration into the biofilm and possible cross-reactivity
of the possibility of amplification and cloning biases, with antigens presented by other species.
resulting in misjudgments of species abundances or PCR procedures are helpful if the bacteria of interest
the formation of chimeric artifacts (43). The sequenc- are suspected to be present in such low numbers that
ing data can subsequently serve for the design of detection by antibodies, in situ probing, or fingerprinting
new hybridization probes and their application in the is impossible. Examples of this include detection of
microbial community, thus obtaining quantitative infor- Helicobacter spp. or Mycobacterium spp. by rRNA
mation on the population composition. Hybridization amplification from microbial communities in drinking
probes also allow the monitoring of enrichment cultures water systems. Specific PCR can be coupled with dot-blot
and enable the isolation of specific species of inter- hybridizations, reducing the possibility of false-positive
est. detection of pathogens (49,50). For the differentiation of
a known pathogen from other closely related bacterial
Fingerprinting Techniques: DGGE and RFLP species, PCR amplification of specific genes, differentiating
these strains from others can be a very useful tool. In a
DNA strands of the same length but different sequences recent study, a multiplex PCR approach was developed for
can be separated by exploiting differences in their migra- the simultaneous identification of Listeria monocytogenes
tion behavior in denaturing gradient gel electrophoresis and five other species of this genus, based on species-
(DGGE). Genes coding for 16S rRNAs obtained by PCR specific amplification of the iap gene (51).
amplification procedures from microbial communities can
be subjected to this technique. The banding patterns or
Cultivation Strategies
‘‘fingerprints’’ thus obtained offer the possibility to rapidly
detect changes in population structures of microbial com- Cultivation of bacteria from natural microbial commu-
munities. Apart from the visualization of such profiles, nities usually yields less than 1% of the bacterial
specific DNA bands of interest can be excised and fur- population, even though the proportion of metaboli-
ther analyzed by DNA sequencing (44). DGGE has been cally active bacteria in situ far exceeds this percentage,
564 BIOFILMS IN NATURAL AND DRINKING WATER SYSTEMS

a phenomenon described as the ‘‘great plate-count in modified direct viable count assays (54). Kane and
anomaly’’ (52). Comparison between cultivation and in coworkers applied a probe designed for an environmentally
situ detection methods in different habitats revealed derived rRNA sequence for the enrichment and isolation of
drastic cultivation-induced population shifts; the compo- a sulfate-reducing bacterium (56). In cyanobacterial mats,
sition of the isolates obtained did not reflect the in situ an organism with little numerical relevance was found to
population composition (45,53,54). In one case, however, competitively exclude other species that were more domi-
bacteria able to form colonies on solid media were reported nant in situ, when in laboratory culture. Only by dilution
to account for a large fraction of the community (55). of the inocula to extinction before enrichment, could these
Although PCR-based studies hold the potential to describe specific cyanobacteria that were highly relevant in situ,
microbial diversity without cultivation-dependent biases, be recovered. Enrichments from highly diluted inocula
it has to be kept in mind that this technique is subjected also allowed the isolation of in situ abundant chemoorgan-
to its own limitations (43,45). Furthermore, cultivation is otrophic bacteria from this environment (45). Ward and
indispensable to correlate phylogenetic information with coworkers stress the point that the incongruity between
physiological properties of the respective organisms. Cul- culture and molecular methods should not lead to the
tivation of in situ relevant bacterial species can be feasible conclusion that many or most bacteria in natural envi-
if it is combined with and monitored by molecular iden- ronments are ‘‘uncultivable’’ (45). They should rather be
tification techniques. In young drinking water biofilms, regarded as ‘‘uncultivated’’ because the most likely expla-
isolation of the dominant species was successful, based nation for cultivation failures might be our inability to
on findings gained by in situ probing and activity mea- understand and reproduce the real environmental niches
surements, which had revealed that most of the bacteria of bacterial species.
were affiliated to the b-subclass of Proteobacteria and that Different methods of bacterial identification in biofilms
most cells could utilize the medium used for cultivation are summarized in Table 1.

Table 1. Methods of Bacterial Identification in Biofilms

Level of
Method Parameter Measured Measurement

Light microscopy Transmission of light Individual cell

Epifluorescence microscopy Emission of light
Confocal laser scanning microscopy Transmission, emission, and reflection
(CLSM) of light
Scanning electron microscopy (SEM) Transmission and reflection of
short-wavelength light and electron
beams
Environmental scanning electron
microscopy (ESEM)
Total cell count determination: Fluorescence emission after binding of Individual cell
nucleic acid–specific fluorochromes
DAPI (4’,6-diamidino-2-phenylindole)
AO acridine orange
Hoechst 33258
Hoechst 33342
SYTO stains
SYTOX stains
SYBR Green
PicoGreen
Propidium Iodide
FISH (fluorescence in situ Fluorescence emission after Individual cell
hybridization) hybridization of DNA-probes to
ribosomal RNA
GFP (green fluorescent protein) Fluorescence emission after expression Individual cell
of the GFP
Microcalorimetry Heat originated from microbial Community
metabolism
FTIR spectroscopy Absorption spectra (amide I and II Individual cell
bands of proteins) obtained by their
attenuated total reflection
RFLP (restriction fragment length Fragment lengths after restriction of Community and
polymorphism) specifically amplified genes individual cell
DGGE (denaturing gradient gel Migration of specifically amplified genes Community and
electrophoresis) of similar length along a denaturing individual cell
gradient
Immunostaining Fluorescence emission after binding of Individual cell
antibodies to their antigens
 BIOFILMS IN NATURAL AND DRINKING WATER SYSTEMS 565

ASSESSMENT OF METABOLIC ACTIVITIES suggesting that the number of metabolically active cells in
a microbial community might be underestimated by this
Direct Viable Counts method (63).
Among microbiologists, it is widely accepted that ‘‘a single, Microautoradiography
viable microorganism is represented by a cell capable of
dividing and forming at least one live daughter cell when The uptake of specific substrates by microorganisms under
it is placed in a favorable environment’’ (57). According in situ conditions can be studied by using radiolabeled sub-
to the definition given by Oliver (58), bacteria that strates in combination with microautoradiography (64).
show metabolic activity but are incapable of undergoing This method has become widely used in ecological stud-
the sustained cellular division required for growth on ies and allows conclusions about the metabolic activity
a favorable artificial medium are termed viable but of microbial populations. A correlation between active
nonculturable cells (VBNC, see VIABLE BUT NOT CULTURABLE substrate uptake and identity of the organisms, how-
(VBNC) MICROORGANISMS, this Encyclopedia). To determine ever, is not possible based on microautoradiography alone.
the metabolic potential for bacteria to undergo cell A combination with immunofluorescent-labeling of spe-
division, Kogure and coworkers (59) developed the direct cific organisms was developed (65) but suffers from the
viable count (DVC) method, where bacterial samples are drawbacks of antibody application in natural microbial
incubated with appropriate nutrients in the presence of an communities mentioned in the preceding section. Recently,
antibiotic, which prevents DNA synthesis. Because other microautoradiography was successfully combined with
cellular functions are not affected, bacteria will elongate rRNA hybridizations and applied for the characterization
and can subsequently be enumerated by microscopy. of seawater picoplankton (66) and activated sludge com-
The difference between heterotrophic plate counts and munities (67). For single-cell resolution in sludge samples,
elongated cells is commonly considered the portion of a flocs were cryosectioned and examined by confocal laser
given bacterial population in a VBNC state and according scanning microscopy. Uptake of specific substrates could
to this, DVC has become the most widespread method thereby be combined with the analysis of the phylogenetic
to determine the number of cells in a VBNC state. A identity of the organisms.
major recent modification to this technique involves the
Other Methods
application of a cocktail of five antibiotics, which greatly
reduces the probability of cell growth due to the presence In recent years, a number of techniques have been devel-
of bacteria with antibiotic resistance and allows longer oped for the assessment of the general metabolic potential
incubation times (60). Another modification complements of bacteria. Most of them can be grouped into meth-
enumeration of elongated cells with in situ hybridizations, ods detecting activities of enzymes, synthesis of macro-
enabling one to monitor the metabolic potential of specific molecules, or membrane potentials. They include detection
species or groups of organisms of interest (54). of esterase activity (68), alkaline phosphatase (69), micro-
Additional methods for determination of viable and colony formation (70), membrane potential (71,72), mem-
nonviable cells include a commercial live-dead stain brane integrity (73), or DNA- and RNA-degradation (74).
(Molecular Probes, Eugene, Ore.), and some measures Specific activities of bacteria can be visualized in
of respiratory activity (see the following section) may also situ by the detection of target mRNA sequences. The
be used to assess activity. Because of the complexity of mRNA of interest is amplified by in situ PCR and
determining the status of bacterial cells, a combination subsequently visualized by labeled probes complementary
of techniques may be required, particularly, in complex to the amplified DNA (75,76). Reporter genes might be
communities. very useful to determine specific metabolic activities of
genetically modified organisms in model biofilms. The
Respiratory Activity reporter gene has to be fused to a gene or promoter of
interest in a specific bacterial strain; the most commonly
Bacterial respiratory activity can be assessed on the basis used reporter genes are the green fluorescent protein
of the reduction of tetrazolium salts to formazan crystals (GFP) gene, the E. coli lacZ gene, the lux gene coding
by the activity of cellular dehydrogenases. The tetrazolium for bacterial bioluminescence, or the firefly luciferase
salt (CTC) has replaced salts used in earlier studies gene (77,78, see GREEN FLUORESCENT PROTEIN (GFP), this
because of the easy detection of its fluorescent-reduced Encyclopedia).
form (61). CTC can be applied to biofilm samples, with or At the level of activities of microbial communities,
without addition of nutrients, yielding either the number determination of the ATP content is one of the most
of actively respiring bacteria under in situ conditions or frequently used methods, allowing, for example, study
the number of cells with the potential for respiratory of the influence of water distribution pipe materials on
activity when nutrient limitation is lifted. Some authors, bacterial activity and growth (79). For some applications,
however, reported that they could find no difference a combination of in situ probing and microelectrode
between incubation with or without nutrients (62). CTC measurements can be very useful. In trickling filters, the
has recently been shown to have a toxic effect on several distribution of sulfate-reducing bacteria was correlated
bacterial metabolic processes. Samples subjected to a CTC- to gradients of oxygen and sulfide (80), and biofilms in
assay yielded lower heterotrophic plate counts, displayed a nitrifying bioreactor were examined for distribution
lower bacterial production, and lower respiratory activity patterns and activities of nitrifying bacteria and gradients
as measured by glucose incorporation and respiration, of oxygen, ammonium, nitrate, and nitrite (81).
566 BIOFILMS IN NATURAL AND DRINKING WATER SYSTEMS

Table 2. Methods for Assessing Bacterial Metabolic Activities in Biofilms

Level of
Method Metabolic Parameter Measured Measurement

RH-795 staining Integrity of cell membrane Individual cell

Rhodamine 123 staining
Live/Dead Bacteria LightTM (Propidium
Iodide/SYTO 9)
ATP measurement Overall adenylate content Community
Microelectrodes Redox potential, microbial activity Community
Microautoradiography Substrate utilization Community and
individual cell
FISH (fluorescence in situ hybridization) Cellular ribosome content Individual cell
FISH-MAR (fluorescence in situ Cellular ribosome content and substrate Individual cell
hybridization microautoradiography) utilization
DVC (direct viable count) Substrate utilization and elongation Individual cell
PAC (probe active count) Substrate utilization and cellular Individual cell
ribosome content
SFDA (sulfofluorescein diacetate) Hydrolytic enzymatic activity (esterases) Individual cell
CFDA (carboxyfluorescein diacetate)
TTC (triphenyl tetrazolium chloride) Cellular dehydrogenase activity Individual cell
(respiratory activity)
INT (2-p-(iodophenyl)-3-(p-nitrophenyl)-5-
tetrazolium chloride)
CTC (5-cyano-2,3-ditolyl tetrazolium
chloride)
gfp gene (Green fluorescent protein) Gene expression Individual cell
lacZ gene (β-Galactosidase)
lux gene (Bacterial luciferase)

Different methods for assessing bacterial metabolic the Flavobacterium cells. Cysteine serves as a nutrient for
activities in biofilms are summarized in Table 2. Legionella and as a mild reducing agent (83).
Using GFP as a reporter system, metabolic interaction
could also be demonstrated for degradation of benzyl
COMMUNITY INTERACTIONS
alcohol in mixed culture biofilms. Gene fusions of gfp
with either the upper-pathway promoter or the meta-
Metabolic Interactions
pathway promoter of the toluene degradation pathway
In nature, many degradative processes cannot be per- were generated and introduced into P. putida. GFP
formed by pure cultures of bacterial species, but require expression from the upper-pathway promoter could be
the metabolic interactions of different bacterial species. induced in P. putida pure-culture biofilms and in mixed-
Biofilms, where bacteria are in close proximity, are ide- culture biofilms by addition of benzyl alcohol. Expression
ally suited for interactions of their community members. from the meta-pathway promoter, however, was only
Bacterial interactions in biofilms are not limited to com- induced when an Acinetobacter sp. was added to the biofilm
petition for available substrates, but often involve a more community. Apparently, the Acinetobacter sp. degraded
cooperative behavior such as cometabolism or symbio- benzyl alcohol to benzoate, which leaked from the cells
sis. Well-known examples are the anaerobic degradation into the surrounding medium and could serve as substrate
of organic matter, where at the end of the food-chain, and inducer of GFP expression from the meta-pathway
H2 -producing acetogens and H2 -consuming methanogens promoter in P. putida (84). In a flow cell-grown model
interact as symbiotic partners; the nitrification process system the effect of a commensal interaction on the
that requires the combined action of ammonia- and nitrite- spatial structure of the biofilm community consisting of
oxidizing bacteria; or the degradation of many xenobiotic two organisms was investigated using fluorescent in situ
compounds. In a defined biofilm system in which a drink- hybridization, confocal laser scanning microscopy, and a
ing water isolate was fed with benzoate as sole carbon GFP reporter system (85). Depending on the nature of
source, growth of an added E. coli strain could be demon- substrate given in the model system, different colonization
strated within the biofilm, although the strain was unable patterns could be visualized. When the consortium was
to grow on benzoate in pure culture. Apparently, metabo- fed with chlorobiphenyl, which could only be utilized
lites excreted by the drinking water bacterium allowed by one model organism, microcolonies consisting of that
not only survival but also growth of E. coli (82). Another bacterium and the associated second organism occurred.
well-studied example for interspecies exchange of metabo- In contrast to this, the two species formed separate
lites is the coexistence of Legionella pneumophila and microcolonies when fed with citrate, which could be
Flavobacterium breve, caused by the release of cysteine by metabolized by both organisms.
 BIOFILMS IN NATURAL AND DRINKING WATER SYSTEMS 567

Cell Signaling in microbial communities, and is also thought to play a

considerable role in biofilms. Even though the probability
In recent years, it has become clear that to react to
of random encounters of cells is lower in biofilms than
changes in the environment, bacteria must possess a
in the planktonic phase, the high cell density and stable
form of intercellular communication. One of the important
spatial localization might facilitate intra and interspecies
environmental factors for bacteria is the cellular density
gene transfer. The frequency of gene transfer can be
of their own population. Among many gram-negative
affected by a number of factors, such as cell densities (90),
bacteria, a ‘‘quorum-sensing’’ system has evolved, based
growth phase (91), temperature, pH, concentrations of
on diffusible, membrane-permeant acylated homoserine
nutrients, ions, and oxygen (57,92,93), as well as contact
lactone (AHL) pheromones that act as cell density cues
time (93) and spatial arrangement (90). Gene transfer
or ‘‘autoinducers’’ and cause changes in gene expression.
methods are described in detail by Christensen and
AHLs are constitutively expressed at a low basal level
coworkers (78). The frequency of gene transfer is usually
but expression is upregulated when the cell density,
determined by plating on selective media, where only
and hence AHL concentration, reaches a certain level.
transconjugants can grow, raising the question whether
Quorum sensing has been discovered in Vibrio fischeri,
the number of culturable transconjugants reflects the
where luminescence is controlled by two proteins, LuxI
number of in situ occurring gene transfer events. In
and LuxR. LuxI is an AHL-synthetase and LuxR is a a recent study on interspecies gene transfer between
transcriptional activator, which promotes transcription E. coli and Ralstonia eutropha in biofilms, detection of
of luciferase genes when bound to the AHL-autoinducer. transconjugants was achieved in situ by a combination of
In P. aeruginosa, two analogous quorum-sensing systems in situ probing and GFP expression. Interestingly, the rate
las and rhl, have so far been discovered. LasI and of gene transfer was found to be thousandfold higher than
RhlI direct the synthesis of specific AHLs, which the rate determined by classical plating techniques (93).
induce the transcriptional activators LasR and RhlR, However, if interspecies plasmid transfer was studied on
respectively. Genes regulated by quorum sensing include, floating filters, no difference between conjugations was
among others, genes coding for an endotoxin, secretion found between in situ detection and traditional plating
proteins, a protease, and the stationary-phase sigma techniques (94). This discrepancy might be explained
factor. Thus, production of virulence factors appears by different experimental setups or they could possibly
to be under the regulation of quorum-sensing systems hint at physiological differences between planktonic- and
in P. aeruginosa. Similar quorum-sensing systems have biofilm-associated cells (93). It must also be kept in mind
been found in numerous gram-negative bacteria, whereas that in situ detection of gene transfer does not allow
in gram-positive bacteria, quorum sensing is based on any conclusions with regard to plasmid stability in the
peptide signal molecules, which are recognized by sensor recipient. Because in situ monitoring of plasmid transfer
kinases that interact with cytoplasmic response regulator is based on GFP expression from the transferred plasmid,
proteins (86). AHLs not only serve as quorum sensors but it would be expected that only bacteria that keep the
some AHLs also stimulate their own production and can plasmid maintain green fluorescence. The method can be
establish positive autoinducer feedback loops. further refined by using unstable variants of the GFP.
Apart from the obvious importance of quorum sensing
for gene expression in biofilms, where cell densities are Interactions Between Eukaryotes and Bacteria
very high, defects in quorum sensing seem to have a crucial
impact on the biofilm formation potential of bacteria. As Grazing of bacteria by various eukaryotic microorganisms
already mentioned earlier, a P. aeruginosa lasI mutant is widespread in microbial communities. Moreover,
formed only undifferentiated, flat biofilms, which were predation was found to have a significant impact on
more sensitive to biocide action than the corresponding bacterial morphologies and on the taxonomic composition
wild-type biofilm (33). of bacterial populations. In planktonic populations,
Recently, a marine red alga has been found to produce protozoan grazing could be shown to trigger a population
antagonists of AHLs. These halogenated furanones appar- shift from small, fast-growing bacteria affiliated to
ently inhibit AHL-mediated gene expression by displacing the b-subclass of Proteobacteria and the Flavobacteria-
AHL from its receptor protein, the transcriptional acti- Cytophaga phylum toward more grazing-resistant, larger
vator LuxR in E. coli. By producing furanones, the alga cells affiliated the a- and g-subclasses of Proteobacteria
controls bacterial colonization of its surfaces. Again, these and filamentous bacteria of other phylogenetic groups (95).
findings demonstrate the importance of cell signaling for Grazing by protists may also influence the nature of the
colonization and biofilm formation (87). Evidence has also exopolymer matrix of biofilms as shown by Möller and
been found for the role of AHLs in the formation of natural coworkers (96).
and urethral catheter biofilms (88,89). Comparisons between grazing of planktonic ver-
sus attached bacteria showed protozoan preference for
attached bacteria in some cases, in other studies, a protec-
Gene Transfer
tive effect of adhesion was reported (97). A recent study
(see GENE EXCHANGE IN BIOFILMS, this Encyclopedia) attempted to directly quantify protozoan feeding rates
Genes can be horizontally transferred between bacteria on biofilm bacteria by using unstained, viable prey. In
by conjugation, transduction, and transformation. Gene this model system, the grazing rate of a small ciliate
transfer by conjugation is thought to be one of the major on P. putida cells was determined to be higher for sus-
mechanisms for the establishment of new genetic traits pended than for attached bacteria (97). Analysis of the
568 BIOFILMS IN NATURAL AND DRINKING WATER SYSTEMS

Ungrazed Grazed Ungrazed Grazed

25 µm

Figure 1. Stereo pair of false color confocal laser scanning micrographs, showing the effects of
snail grazing on the distribution of bacteria (stained with Syto 9, green), algae (autofluorescence,
red), and extracellular polymeric substance (stained with fluorescent Triticum vulgaris lectin,
blue) in a river biofilm. See color insert.

bacterial community composition during a diatom bloom or salt marshes. They are characterized by their layered
suggested that protozoan grazing, together with increased structure, where production of organic matter through
viral abundance, not only leads to a phylotype-shift of photosynthetic microorganisms is located in the upper-
the bacterial community but also to a pronounced bac- most layer, and aerobic and anaerobic decomposition of
terial colonization of larger suspended particles (98). The the organic substances takes place in deeper layers of
study of Hahn and coworkers (99) showed that grazing by the mat. The top layer of a microbial mat typically con-
Ochromonas spp., resulted in the development of a floc- tains cyanobacteria and sometimes also diatoms and small
forming subpopulation in Pseudomonas sp. MWH1 that algae, whereas anoxigenic photosynthesis takes place in
was interpreted as a successful grazing defense. the deeper levels. Here, purple phototrophs and filamen-
Interactions between eukaryotes and bacteria, how- tous green phototrophs can be found, together with aerobic
ever, are not restricted to grazing. A well-studied example and anaerobic chemoorganotrophs as well as sulfate-
for parasitic growth of a prokaryote within a eukaryotic reducing and methanogenic bacteria. Many parameters
host is represented by members of the family Legionel- vary vertically across the mats, including light inten-
laceae. Legionella pneumophila, the causative agent of sity, oxygen, carbon dioxide, nitrogen, phosphate, sulfur,
Legionnaires’ disease, grows intracellularly in a number of and organic nutrients (101). One of the most extensively
amoeba species and in other protozoa-like ciliates. These studied systems are hot spring cyanobacterial mat com-
hosts not only provide nutrients for Legionella organ- munities, which have recently been covered in a review
isms but also protect them from adverse environmental by Ward and coworkers (45). At the outset of investiga-
conditions. Legionella spp. present in drinking water sys- tions on the microbial ecology of these mats, studies based
tems may be enclosed in cysts of amoebae and thereby on microscopical observation and cultivation identified
be protected from chlorination, ozone, and other adverse two main bacterial species, a unicellular cyanobacterium
conditions (7). Recently, Mycobacterium avium, an oppor- and a filamentous green nonsulfur bacterium. Other
tunistic human pathogen that can survive and multiply bacterial species cultivated from the mats were aerobic
within drinking water biofilms, has been shown to grow chemoorganotrophs, gram-positive fermentative bacteria,
on metabolites excreted by Acanthamoeba polyphaga. sulfate-reducing bacteria, and methanogens. Data gained
Mycobacterium avium was even able to multiply within the by analysis of pigment and lipid biomarker abundances
outer walls of the double-walled cysts of the amoeba (100). suggested that cyanobacteria and green nonsulfur bacte-
Interactions between plants and bacteria also occur, as ria predominated over chemoorganotrophic bacteria, and
described above for a marine alga, which produces fura- even fewer terminal members of the anaerobic food chain,
nones to inhibit colonization by bacteria. such as methanogens or sulfate-reducers, were present.
Interactions with micro and macroinvertebrates also Direct retrieval of 16S rRNA sequences, using different
influences development of microbial biofilms, through molecular methods, however, did not yield sequences that
disturbance, sloughing, and active grazing of bacterial could be ascribed to previously isolated organisms but only
and exopolymer components of biofilms. For example, the sequences that were 93% or less similar to their closest
impact of snail, grazing on the biofilm architecture, and cultivated or uncultivated relatives. This divergence is
abundance of bacteria, algae, and extracellular polymeric thought to be because of competitive exclusion of species
substance is shown in Figure 1. that are numerically dominant by bacteria, which per-
form well in laboratory culture but are not abundant in
situ. As mentioned earlier, a solution for recovery of dom-
EXAMPLES OF NATURAL BIOFILMS
inant strains may be dilution-to-extinction enrichment.
DGGE analysis proved to be a very useful tool for the
Cyanobacterial Mats
investigation of the spatial distribution of populations in
Cyanobacterial mats are typical microbial biofilm com- the mat. Vertical distribution could be demonstrated for
munities in extreme environments such as hot springs cyanobacterial populations and for green sulfur and green
 BIOFILMS IN NATURAL AND DRINKING WATER SYSTEMS 569

nonsulfur bacteria. Although one cyanobacterial popula- algae even occur in the hyporheic zone (108,109), which
tion (termed type B) could be detected throughout the top means the interstitial layer of the river bottom that
0 to 0.5 µm layers, the other population (termed type A) may be in contact with both the river water and the
was found to be restricted to a well-defined layer about groundwater flow underneath. The continuously changing
600 µm from the surface, suggesting a vertical strati- riverine ecosystem biofilms contribute significantly to the
fication and adaptation to different light intensities of distribution and recycling of nutrients and inorganic sub-
cyanobacterial populations. Mechanical disturbance of the stances (110), and heterotrophic activity is considered to be
mat, caused by removal of the top green cyanobacterial almost exclusively associated with sediments (111–113).
layer, allowed the investigation of recolonization behav- Marxsen (114) discussed the significance of sediment
ior. DGGE analysis revealed that a new cyanobacterial biofilms for the carbon balance of streams. Dissolved
population colonized the disturbed mat, rather than the organic matter (DOM) is converted by microorganisms
previously dominant species that remained in adjacent in both hyporheic and epilithic biofilms (115–117). Naegli
undisturbed sites, indicating that as in macroecology, bac- and Uehlinger (118) showed for a prealpine river that
terial species might exist, which are effective at rapid the epilithic and hyporheic biofilm community contributed
growth rates (so-called r-strategists) and colonization. 4 to 19% and 76 to 96% of the total respiratory activ-
DGGE surveys also revealed cyanobacterial population ity as determined by oxygen consumption, respectively.
changes along the thermal gradient that occurs in the Although most studies focused on the upper layers of the
effluent channel of the Octopus Spring location (45). The sediment microhabitat, even deeper sediment zones can be
detected sequences appeared to be from dominant mem- populated by metabolically active bacteria (106,108,113).
bers of the population because they could also be recovered From a macroecological point of view, riverine biofilms
from cyanobacterial isolates obtained from highly diluted play an important role as a trophic resource for macroin-
inocula. The detected sequences showed a high level of 16S vertebrates in lotic systems (119,120). Biofilm bacteria
rRNA sequence relatedness, suggesting that the species and particle-associated microorganisms themselves are
might have formed through adaptive radiation from a subjected to a top-to-bottom control by grazing pro-
common ancestor (45). tists (121,122), which have repeatedly been shown on
surfaces in rivers (123).
Riverine Biofilms The biofilm organisms are typically embedded in or
In the global hydrological cycle from atmosphere to land associated with a complex heterogenous matrix of extra-
and oceans, running freshwater systems (lotic systems) cellular polymeric substances (EPS) secreted by the organ-
are the major linkages between terrestrial and aquatic isms themselves (124–127). The EPS matrix enhances the
habitats including surface and groundwater. The common immobilization of dissolved organic carbon (DOC), and
feature of river ecosystems is the continuous movement even particulate organic carbon, and allows the exchange
of the bulk water phase. Microbiological investigations of of matter within syntrophic associations of microorgan-
lotic systems revealed the important role of the complex isms. Indeed, Freeman and Lock (128) demonstrated that
microbial river communities in global nutrient cycles. river biofilms were remarkably resilient to a loss of avail-
It has become apparent that microbial transformations able carbon in the overlying waters, indicating that river
(the so-called microbial loop, 102) have a serious impact biofilms store energy that is subsequently available during
on the flux of organic carbon in running waters (103). starvation intervals.
Consequently, attention is now focusing on the role of Bacteria that can attach and multiply under the high-
bacterial production and the microbial food web as well shear forces normally present in rivers are obviously
as their impact on the self-purification capacity of lotic adapted to live in this environment but only little is
systems. known about the physiological plasticity and the ecological
In a recent study, the planktonic bacteria in the impact of planktonic and attached bacteria within these
bulk water phase of the Japanese rivers Minoh and ecosystems. It is generally accepted, however, that the
Neya were investigated as one distinct compartment of activity of attached microbes dominates in small streams,
the lotic community, using rRNA-targeted fluorescent in whereas planktonic bacteria are of greater importance in
situ hybridization (104). However, biofilm bacteria can large rivers because of the different ratios of the surface
comprise the majority of bacterial numbers (105) and (of sediments or suspended particulate matter) to the
most bacterial production and transformation in running volume of the floating water (110). However, the microbial
waters occurs in biofilms or communities associated components of lotic ecosystems have rarely been studied
with suspended particulate matter (106,107), whereas and it is not known to date whether there are specific
relatively little production occurs in the free water bacteria in rivers that can be distinguished from those
column. The balance of planktonic, suspended floc, and organisms that are passively transported downstream.
biofilm communities in rivers is dependent on interactions The overall structure and bacterial composition of specific
between the nutrient status, flow regime, and order of the lotic biofilm communities was described by Lock (110).
specific river under investigation. Nutritional relationships among bacteria and algae in
Epilithic biofilms that grow on the surface of rocks river biofilms were studied by Haack and McFeters (129),
or stones cover most substratum in the littoral zone suggesting a coupling of photosynthetic and heterotrophic
(i.e., those parts of the water body that occur in shal- populations. Leff and coworkers investigated the seasonal
lower waters of unspecified depth) and the streambed. changes in planktonic bacterial assemblages of two
These complex mixed assemblages of bacteria, fungi, and Ohio streams by measuring the abundance of total
570 BIOFILMS IN NATURAL AND DRINKING WATER SYSTEMS

bacteria, colony forming units, and the population of (TRITC)-conjugated lectins, indicating the presence of dif-
Pseudomonas cepacia by hybridization with a species- ferent carbohydrates within the EPS-matrix.
specific probe (130). The same experimental setup was used in another
Because of the obvious difficulties involved with biofilm study investigating algal, bacterial, and exopolymeric
investigations in situ within lotic systems, there is a strong components of fully hydrated natural biofilm communities
need for suitable laboratory models that particularly by Lawrence and coworkers (135). Multispectral imaging
mimic the development of attached communities under in conjunction with nucleic acid stains, fluor-conjugated
turbulent flow regimes and constant shear forces. lectins, and autofluorescence was performed in this study
Lawrence and Caldwell (17) used continuous flow cells to evaluate biofilm community composition. Biofilms
to visualize the rapid colonization of the substratum by a were treated with the nucleic acid stain SYTO 9 to
diversity of lotic bacteria. Mohamed and coworkers (131) quantify the bacterial biomass and fluor-conjugated lectins
used flow cells to investigate the impact of nutrients that (i.e., Triticum vulgaris lectin) to identify and allow
may control the growth of biofilm bacteria obtained from quantification of exopolymeric substances. Digital image
the Fraser River, Canada. As an important result of the analysis of confocal laser scanning sections in each of
study, the phosphorus limitation of heterotrophic river the channels was used to determine parameters such as
biofilm bacteria could be shown for the first time. biofilm depth, bacterial cell area (biomass), exopolymeric
Rotating annular bioreactors (RAB) were applied area, and algal biomass at various depths and locations.
for different studies including structural features and The results of multichannel imaging, showing an example
metabolism (132,133). Neu and Lawrence (134) used RAB of a rotating annular bioreactor generated river biofilm
for cultivation and structural examination of stream stained with fluor-conjugated lectins, a nucleic acid stain,
biofilms by confocal laser scanning microscopy (CLSM). and autofluorescence imaging of algae are shown in
The biofilms were generated in the reactor by feeding Figure 3.
with natural river water from the South Saskatchewan In a further study, the phylogenetic composition, three-
River system as inoculum and sole source of nutrients dimensional structure, and dynamics of the bacterial
and carbon. The reactor contained removable polycarbon- communities in river biofilms, generated in the RAB
ate strips that allowed sampling of biofilm material at system were studied by a combination of fluorescent in
desired intervals. The architecture of the biofilms was in situ hybridization (FISH) and confocal laser scanning
contrast to previous views of a homogeneous biofilm struc- microscopy. Digital image analysis was performed to mea-
ture and more recent biofilm models with channels and sure specific cell numbers and specific cell area after in
inverted biomass distribution. There was a development situ probing (136). During the different stages of biofilm
of ridges, oriented to the flow direction, that were colo- development, individual cells within the α-, β-, and γ -
nized by individual bacterial cells and microcolonies, as Proteobacteria as well as the Cytophaga-Flavobacterium
shown in Figure 2. As suggested by the authors, the differ- group could be visualized, demonstrating that they formed
ences may be attributed to the presence and coadsorption major parts of the attached community. During initial
of humic and detrital material in natural lotic systems. biofilm development, β-Proteobacteria accounted for the
After an initial phase of biofilm development, both the greatest part of the bacterial population. This was fol-
highest biomass and the numbers of putatively living cells lowed by a rapid increase of α-Proteobacteria and bac-
were generally distributed within the outer regions of the teria affiliated to the Cytophaga-Flavobacterium group.
biofilm. The heterogeneity of the biofilms was also visu- In mature biofilms, α-Proteobacteria and Cytophaga-
alized by the application of a panel of fluorescein isothio- Flavobacteria continued to be the prevalent bacterial
cyanate (FITC) and tetramethylrhodamine isothiocyanate groups. β-Proteobacteria constituted the morphologically
most diverse group within the biofilm communities and
more narrow phylogenetic staining revealed the impor-
tance of distinct β1-Proteobacteria for the composition of
the microbial community. The importance of members of
the β-Proteobacteria is also reported for sessile microbial
communities that grew on glass plates exposed in a highly
polluted river in Germany (137). In addition, Brümmer
and coworkers showed the seasonal dynamics of Plancto-
mycetes and Cytophaga-Flavobacteria that both showed
maximum levels of relative abundance in July, and, in case
of Cytophaga-Flavobacteria, a second peak in winter (137).
Besides the attached biofilm communities located at
virtually all surfaces and interfaces in the rivers, bacteria,
algae, and freshwater fungi also constitute a major part of
50 µm the macroscopic mixed assemblages (suspended particu-
late matter) present in the bulk water phase. In accordance
with microbial aggregates described in oceans and estuar-
Figure 2. A red-green anaglyph stereo projection of a series of ies (107,138–140), in the pelagic zone of lakes (141–143)
confocal laser scanning micrographs of a river biofilm stained with and rivers (144–146), the suspended particulate aggre-
Syto 9, and TRITC-labeled T. vulgaris lectin. See color insert. gates found in rivers have been recently termed river
 BIOFILMS IN NATURAL AND DRINKING WATER SYSTEMS 571

snow aggregates possess a higher nutrient concentra-

Bacteria (green)
tion (149) and the associated bacteria show a higher rate
of biological activity (150–152). Within suspended partic-
ulate matter, the formation of chemical gradients and
anoxic microhabitats has been described (153). The river
snow microhabitats are characterized by the occurrence
of photosynthetic organisms (algae, cyanobacteria) and a
broad range of different heterotrophic organisms includ-
ing aquatic freshwater fungi (154,155). In the microbial
aggregates, there is a pronounced patchiness of nutrients
and consequently, these ‘‘hot spots’’ of high nutrient con-
centrations (156) and microbial life (157) show a high rate
Polymer (red)
of heterotrophic metabolism and accelerate the mineral-
ization process (158).
With regard to heterotrophic bacteria, the phycosphere
of algae in which photosynthetic products are released is a
particular hot spot of nutrients on a subcellular scale (159).
Because of the physiological plasticity of the mobile
biofilm organisms, the river snow aggregates can also
be regarded as miniaturized wastewater treatment
plants and fulfill an important function within the
self-purification processes in lotic systems. The high
sorption capacity of the extracellular polymeric substances
Algae (far red)
within the mobile river snow aggregates significantly
contribute to the horizontal and vertical shift of inorganic
and organic substances and link the pelagic and
benthic compartments. Neu (147) examined the fully
hydrated architecture of living river snow aggregates
from the river Elbe, Germany, by multichannel confocal
laser scanning microscopy. Simultaneous application of
different fluorescently labeled lectins as carbohydrate-
specific probes and general nucleic acid stains, visualized
the heterogeneity of the EPS matrix in between the
cellular constituents of the river snow community. CLSM
Merged
performed in the reflection mode and in the far red channel
gave additional information about the mineral content in
the aggregates and the presence and spatial distribution
of chlorophyll-containing organisms.
The phylogenetic diversity and physiological versatility
of the Elbe river snow aggregates were investigated at the
cellular level, using aerobic and anaerobic cultivation tech-
niques, 16S rDNA based phylogeny, and fluorescent in situ
hybridization (FISH) by Böckelmann and coworkers (148).
25 µm
FISH analysis of the river snow community revealed pop-
Figure 3. A series of confocal laser scanning micrographs, ulation dynamics to be governed by seasonal factors and
showing multichannel imaging of a single location in a river more than 70% of the total bacterial cell count could be
biofilm, visualizing Syto 9 stained bacteria recorded with the hybridized with a bacteria-specific probe, indicating the
green channel (top image), T. vulgaris labeled extracellular high physiological potential of the river snow commu-
polymeric substance recorded with the red channel (second
nity. During all seasons, β-Proteobacteria constituted the
micrograph), and the algae imaged by autofluorescence recorded
numerically most important bacterial group, forming up to
with the far-red channel of the microscope (third micrograph).
The bottom image shows the false color presentation of merging 54% of the total cell counts. In contrast to this, the relative
all three images, visualizing bacteria (green), algae (red), and abundance of other major bacterial lineages ranged from
extracellular polymeric substance (blue). See color insert. 2% for the order Planctomycetales to 36% for Cytophaga-
Flavobacteria. Cultivation of river snow bacteria under
aerobic and anaerobic conditions with a variety of differ-
ent media resulted in the isolation of 40 new bacterial
snow (147,148). Like other biofilms, these river snow strains. Phenotypic and phylogenetic analysis revealed
aggregates, which may be considered as mobile biofilms, them as mostly unknown organisms affiliated to different
consist of inorganic and organic compounds, extracellular bacterial phyla. Some Elbe river snow isolates were affili-
polymeric substances (EPS), and the attached microbial ated to distinct freshwater phylogenetic clusters similar to
populations. In comparison to the bulk water phase, river organisms and environmental clones isolated from other
572 BIOFILMS IN NATURAL AND DRINKING WATER SYSTEMS

freshwater systems (160,161). The isolation of river snow tubercles) represent new habitats, where gradients for
bacteria with the potential for degradation of xenobiotics oxygen, redox potential, temperature, nutrients and tox-
routinely monitored in the Elbe river water indicated the ins develop, thus offering growth conditions completely
great physiological versatility of the river snow associated different from the free-water phase. Some bacteria can
microorganisms. also promote corrosion by creating areas with different
Microbial community analysis by cloning and sequenc- gradients in oxygen, minerals, and metals and also cat-
ing of 16S rRNA genes was conducted by Crump and alyze reactions associated with corrosion processes such
coworkers to study the microbial ecology of free-living and as conversion of divalent iron to trivalent forms (171).
particle-attached bacteria in the Columbia River and its In these new microhabitats, bacteria including pathogens
estuary (162,163). Both studies showed the great diver- may survive and even grow better, although the conditions
sity of bacterial sequences and the presence of putatively in the flowing water are inhospitable (7,166,172).
ubiquitous clades of bacteria within the aggregates. Inter- Bacterial species isolated using conventional tech-
estingly, not only eubacterial sequences but also new niques from drinking water biofilms were typically
clusters of archaeal 16S rRNA sequences could be iden- identified as members of fast-growing genera such as
tified, reflecting the great genetic diversity within the Pseudomonas, Acinetobacter, Alcaligenes, Flavobacterium,
riverine microbial aggregates. Moraxella, Arthrobacter (164,166) mostly affiliated to the
g-subclass of Proteobacteria. In contrast to this, exam-
Drinking Water Biofilms inations with oligonucleotide probes in biofilms from
different drinking water systems revealed the predom-
Drinking water distribution systems belong to the
inance of members of the b-subclass of Proteobacte-
most oligotrophic aquatic habitats, and bacterial growth
ria (20,173–175).
is severely limited by low contents of assimilable
Recently, biofilm formation has been extensively
organic carbon (164) or by phosphorus availability (165).
studied in a house installation system in Berlin, Germany,
Nevertheless, basically all existing interfaces are
with regard to population dynamics, species composition,
partially covered by biofilms, from the waterworks
and metabolic potential of the microbial community
through the distribution system to house installa-
of young biofilms. Surfaces were found to be rapidly
tions. The existence of these biofilms was reported
colonized, and a monolayer of rod-shaped β-Proteobacteria
and documented several decades ago, based mostly
developed on the substrata. This young biofilm was
on microscopic or electron microscopic examinations
grazed by amoebae, leading to a more diverse biofilm
of pipe materials and encrustations from pipe sur-
structure and population composition. The metabolic
faces (164,166).
potential of the population, assessed by CTC-reduction
In drinking water, biofilms are usually very thin and
and a modified DVC-assay (176), decreased with biofilm
patchy, often consisting of a monolayer of cells. Thick
development. The in situ dominant species could be
biofilms can only develop if the colonized pipe material
successfully isolated from the early biofilm formation
releases biodegradable dissolved organic carbon (BDOC),
stage and described as members of the new genus
which is the substrate, allowing bacterial development
Aquabacterium within the b-subclass of Proteobacteria.
in the drinking water. Apart from the availability of
The occurrence of the genus Aquabacterium could
nutrients, the treatment of the raw water, use of
also be shown for other drinking water distribution
disinfectants, temperature, type of surface material, and
systems (167), regardless of the raw water source used
hydraulics of the system may affect the densities of biofilm
for the drinking water supply. The species composition in
bacteria (167–169). Bacteria growing in drinking water
established biofilms, however, is expected to be far more
biofilms deteriorate the quality of the bulk water, and
complex (20,54,167).
detachment of bacteria from the biofilms can additionally,
negatively affect the water quality. Consequently, pipe
and fitting materials used in drinking water systems have CONCLUSION
to be subjected to extensive testing. The unwanted growth
of bacteria within drinking water distribution systems can • Biofilms are complex microbial communities and
be limited by distributing biostable drinking water that highly heterogenous in their structure, composition,
contains no easily degradable substrates and by applying and function.
biostable materials in contact with drinking water (170). • Biofilms are formed by bacteria, algae, fungi, protozoa,
The most often used methods for the assessment of the as well as organic compounds and inorganic materials.
biostability of the materials are based on measurement
• Biofilms are colonizing virtually each known interface
of respiration activity in the presence and absence of test
of natural and engineered systems.
materials, on the detection of cell numbers by means of
ATP content, and on scraping of the biofilm from the • The individual biofilm-associated cells within the
substratum and measuring its volume (7). biofilm communities are linked by extracellular
Drinking water biofilms can also contain high propor- polymeric substances (EPS).
tions of inorganic compounds, mostly consisting of iron and • There is an complex exchange of genetic material
manganese. The deposits are caused, at least partly, by between the individual members of the biofilm
microorganisms (e.g., the genera Hyphomicrobium, Side- communities.
rocapsa, Gallionella, Leptothrix, Pedomicrobium). These • The individual cells in the biofilm communities are
inorganic depositions and encrustations (e.g., flocs and communicating to each other by chemical transmitters
 BIOFILMS IN NATURAL AND DRINKING WATER SYSTEMS 573

that regulate properties of the individual members 23. P. Stoodley, Z. Lewandowski, J. D. Boyle, and H. M. Lappin-
(e.g., individual cell shape, motility) and of the whole Scott, Environ. Microbiol. 1, 447–455 (1999).
bacterial population (e.g., colony shape, cell density). 24. S. Kjelleberg, B. A. Humphrey, and K. C. Marshall, Appl.
• In nature, biofilm communities often form complex Environ. Microbiol. 43, 1166–1172 (1982).
stratified microhabitats and foodwebs (e.g., photosyn- 25. K. Power and K. C. Marshall, Biofouling 1, 163–174 (1988).
thetic cyanobacterial communities and heterotrophic 26. C. E. ZoBell and D. Q. Anderson, Biol. Bull. 71, 324–342
sulfate-reducers). (1936).
• Biofilms in riverine systems play an important role in 27. M. Givskov et al., J. Bacteriol. 180, 742–745 (1998).
nutrient and energy metabolisms, can be regarded as 28. L. McCarter, M. Hilmen, and M. Silverman, Cell 54,
hot spots of microbial activity, and are responsible for 345–351 (1988).
the self-purification ability of rivers. 29. C. Prigent-Combaret and P. Lejeune, Methods Enzymol.
• In drinking water distribution systems, biofilms are 310, 56–79 (1999).
hygienically and technically important in terms of 30. O’Toole et al., Methods Enzymol. 310, 91–109 (1999).
affecting both the quality of the drinking water and 31. C. Prigent-Combaret et al., J. Bacteriol. 181, 5993–6002
the materials used for pipes, reservoirs, and fittings in (1999).
the distribution system. 32. L. A. Pratt and R. Kolter, Curr. Opin. Microbiol. 2, 598–603
(1999).
BIBLIOGRAPHY 33. D. G. Davies et al., Science 280, 295–298 (1998).
34. G. A. O’Toole et al., J. Bacteriol., 182, 425–431 (2000).
1. W. G. Characklis and K. C. Marshall, eds., Biofilms, John 35. C. M. Buswell et al., J. Appl. Microbiol. 83, 477–484 (1997).
Wiley & Sons, New York, 1990, pp. 3–15. 36. C. Heilmann et al., Infect. Immun. 64, 277–82 (1996).
2. J. W. Costerton et al., Annu. Rev. Microbiol. 49, 711–745 37. O. Vidal et al., J. Bacteriol. 180, 2442–2449 (1998).
(1995).
38. D. G. Davies, A. M. Chakrabarty, and G. G. Geesey, Appl.
3. M. E. Davey and G. A. O’Toole, Microbiol. Mol. Biol. Rev. 64,
Environ. Microbiol. 59, 1181–1186 (1993).
847–867 (2000).
39. D. W. Martin et al., J. Bacteriol. 176, 6688–6696 (1994).
4. J. Wimpenny, W. Manz, and U. Szewzyk, FEMS Microbiol.
Rev. 24, 661–671 (2000). 40. F. H. Yildiz and G. K. Schoolnik, PNAS 96, 4028–4033
(1999).
5. T. R. Neu, Microbiol. Rev. 60, 151–166 (1996).
41. C. R. Woese, Microbiol. Rev. 58, 1–9 (1994).
6. P. Gilbert, D. J. Evans, and M. R. W. Brown, J. Appl.
Bacteriol. Symp. Suppl. 74, 67S–78S (1993). 42. R. I. Amann, W. Ludwig, and K.-H. Schleifer, Microbiol.
Rev. 59, 143–169 (1995).
7. U. Szewzyk, R. Szewzyk, W. Manz, and K. H. Schleifer,
Annu. Rev. Microbiol. 54, 81–127 (2000). 43. F. von Wintzingerode, U. B. Göbel, and E. Stackebrandt,
FEMS Microbiol. Rev. 21, 213–229 (1997).
8. M. M. Doolittle, J. J. Cooney, and D. E. Caldwell, Can. J.
Microbiol. 41, 12–8 (1995). 44. G. Muyzer, E. C. de Waal, and A. G. Uitterlinden, Appl.
Environ. Microbiol. 59, 695–700 (1993).
9. K. A. Hughes, I. W. Sutherland, and M. V. Jones, Microbiol-
ogy 144, 3039–3047 (1998). 45. D. M. Ward et al., Microbiol. Mol. Biol. Rev. 62, 1353–1370
10. M. Fletcher and K. C. Marshall, in K. C. Marshall, ed., (1998).
Advances in Microbial Ecology, Plenum Press, New York, 46. C. L. Moyer, F. C. Dobbs, and D. M. Karl, Appl. Environ.
1982, pp. 199–236. Microbiol. 60, 871–879 (1994).
11. K. C. Marshall, Can. J. Microbiol. 34, 503–506 (1988). 47. M. M. Moeseneder et al., Appl. Environ. Microbiol. 65,
12. S. Kjelleberg and M. Hermansson, Appl. Environ. Microbiol. 3518–3525 (1999).
48, 497–503 (1984). 48. J. Rogers and C. W. Keevil, Appl. Environ. Microbiol. 58,
13. P. M. van Schie and M. Fletcher, Appl. Environ. Microbiol. 2326–2330 (1992).
65, 5082–5088 (1999). 49. K. Hulten, T. Nyström, and L. Engstrand, J. Appl. Micro-
14. R. Bos, H. C. Mei, and H. J. Busscher, FEMS Microbiol. biol. 85, 282–286 (1998).
Ecol. 23, 179–230 (1999). 50. T. Schwartz et al., J. Microbiol. Methods 34, 113–123
15. M. C. M. van Loodsrecht et al., Microb. Ecol. 17, 1–15 (1998a).
(1989). 51. A. Bubert et al., Appl. Environ. Microbiol. 65, 4688–4692
16. J. R. Lawrence, D. R. Korber, G. Wolfaardt, and D. E. Cald- (1999).
well, Adv. Microbiol. Ecol. 14, 1–75 (1995). 52. J. T. Staley and A. Konopka, Annu. Rev. Microbiol. 39,
17. J. R. Lawrence and D. E. Caldwell, Microbiol. Ecol. 14, 321–346 (1985).
15–27 (1987). 53. M. Wagner et al., Appl. Environ. Microbiol. 59, 1520–1525
18. H. Dalton et al., J. Bacteriol. 176, 6900–6906 (1994). (1993).
19. G. A. O’Toole and R. Kolter, Mol. Microbiol. 30, 295–304 54. S. Kalmbach, W. Manz, and U. Szewzyk, Appl. Environ.
(1998). Microbiol. 63, 4164–4170 (1997b).
20. S. Kalmbach, W. Manz, and U. Szewzyk, FEMS Microbiol. 55. J. Pinhassi, U. L. Zweifel, and A. Hagström, Appl. Environ.
Ecol. 22, 265–280 (1997a). Microbiol. 63, 3359–3366 (1997).
21. K. Pedersen, Appl. Environ. Microbiol. 44, 1196–1204 56. M. D. Kane, L. K. Poulsen, and D. A. Stahl, Appl. Environ.
(1982). Microbiol. 59, 682–686 (1993).
22. J. W. T. Wimpenny and R. Colasanti, FEMS Microbiol. Ecol. 57. D. B. Roszak and R. R. Colwell, Microbiol. Rev. 51, 365–379
22, 1–6 (1997). (1987).
574 BIOFILMS IN NATURAL AND DRINKING WATER SYSTEMS

58. J. D. Oliver, in S. Kjelleberg, ed., Starvation in Bacteria, 92. T. A. Khalil and M. A. Gealt, Can. J. Microbiol. 30, 733–737
Plenum Press, New York, 1993, pp. 239–272. (1987).
59. K. Kogure, U. Simidu, and N. Taga, Can. J. Microbiol. 25, 93. M. Hausner and S. Wuertz, Appl. Environ. Microbiol. 65,
415–420 (1979). 3710–3713 (1999).
60. F. Joux and P. LeBaron, Appl. Environ. Microbiol. 63, 94. C. Dahlberg, M. Bergström and M. Hermansson, Appl.
3643–3647 (1997). Environ. Microbiol. 64, 2670–2675 (1998).
61. G. G. Rodriguez et al., Appl. Environ. Microbiol. 58, 95. K. Jürgens et al., Appl. Environ. Microbiol. 65, 1241–1250
1801–1808 (1992). (1999).
62. B. H. Pyle, S. C. Broadway, and G. A. McFeters, Appl. Env- 96. S. Möller et al., Appl. Environ. Microbiol. 63, 2432–2438
iron. Microbiol. 61, 4304–4309 (1995). (1997).
63. S. Ullrich et al., Appl. Environ. Microbiol. 62, 4587–4593 97. H. Eisenmann et al., Appl. Environ. Microbiol. 64,
(1996). 1264–1269 (1998).
64. T. D. Brock and M. L. Brock, Nature 209, 734–736 (1966). 98. L. Riemann, G. F. Steward, and F. Azam, Appl. Environ.
65. C. B. Fliermans and E. L. Schmidt, Appl. Microbiol. 30, Microbiol. 66, 578–587 (2000).
676–684 (1975). 99. M. W. Hahn, E. R. B. Moore and M. G. Höfle, Microb. Ecol.
66. C. C. Ouverney and J. A. Fuhrman, Appl. Environ. Micro- 39, 175–185 (2000).
biol. 65, 1746–1752 (1999). 100. M. Steinert et al., Appl. Environ. Microbiol. 64, 2256–2261
67. N. Lee et al., Appl. Environ. Microbiol. 65, 1289–1297 (1998).
(1999). 101. R. Guerrero and J. Mas, in Y. Cohen and E. Rosenberg,
68. S. De Rosa, F. Sconza, and L. Volterra, Water Res. 32, eds., Microbial Mats: Physiological Ecology of Benthic
2621–2626 (1998). Microbial Communities, American Society for Microbiology,
69. K.-T. Huang, K. D. Xu, G. A. McFeters, and P. S. Stewart, Washington D.C., 1989, pp. 37–51.
Appl. Environ. Microbiol. 64, 1526–1531 (1998). 102. F. Azam et al., Mar. Ecol. Progr. Ser 10, 257–263 (1982).
70. O. Nybroe, FEMS Microbiol. Ecol. 17, 77–84 (1995). 103. J. D. Allan, Stream Ecology: Structure and Function of
71. F. P. Yu and G. A. Mc Feters, J. Microbiol. Methods 20, 1–10 Running Waters, Chapman & Hall, London, U.K., 1995.
(1994). 104. T. Kenzaka et al., Microbiology 144, 2085–2093 (1998).
72. S. C. Williams et al., J. Microbiol. Methods 32, 225–236 105. J. W. Costerton and R. R. Colwell, Native Aquatic Bacteria:
(1998). Enumeration, Activity and Ecology, ASTM Press, 1979.
73. D. R. Korber et al., Appl. Environ. Microbiol. 62, 3939–3947 106. R. T. Edwards et al., Benthol. Soc. 9, 216–228 (1990).
(1996). 107. H. Zimmermann, Aquat. Microb. Ecol. 13, 37–46 (1997).
74. D. Weichardt et al., Appl. Environ. Microbiol. 63, 108. M. Pusch, Hydrobiologia 323, 107–118 (1996).
2754–2758 (1997).
109. H. Fischer et al., Arch. Hydrobiol. 137, 281–300 (1996).
75. R. E. Hodson et al., Appl. Environ. Microbiol. 61, 4074–4082
110. M. A. Lock, in T. E. Ford, ed., Aquatic Microbiology, Black-
(1995).
well, Cambridge, 1993, pp. 113–138.
76. T. Tolker-Nielsen, K. Holmström and S. Molin, Appl. Envi-
111. G. W. Minshall, Ecol. Monogr. 53, 1–25 (1983).
ron. Microbiol. 63, 4196–4203 (1997).
112. R. J. Naiman, Ecology 68, 1139–1156 (1987).
77. D. R. Korber et al., Methods Enzymol. 310, 3–20 (1999).
113. M. Pusch and J. Schwoerbel, Arch. Hydrobiol. 130, 35–52
78. B. B. Christensen et al., Methods Enzymol. 310, 20–42
(1994).
(1999).
114. J. Marxsen, Arch. Hydrobiol. 111, 339–350 (1988).
79. D. van der Kooij et al., Water Res. 29, 1655–1662 (1995).
115. D. M. Fiebig and M. A. Lock, Freshwater Biol. 26, 45–55
80. N. B. Ramsing, M. Kotuhn, and B. B. Jorgensen, Appl.
Environ. Microbiol. 59, 3840–3849 (1993). (1991).
81. A. Schramm et al., Appl. Environ. Microbiol. 65, 3690–3696 116. D. M. Fiebig and J. Marxsen, Freshwater Biol. 28, 129–140
(1999). (1992).
82. U. Szewzyk et al., FEMS Microbiol. Ecol. 13, 169–176 117. D. M. Fiebig, Limnol. Oceanogr. 42, 763–768 (1997).
(1994). 118. M. W. Naegli and U. Uehlinger, J.N. Am. Benthol. Soc. 16,
83. R. M. Wadowsky and R. B. Yee, Appl. Environ. Microbiol. 794–804 (1997).
46, 1447–1449 (1983). 119. K. R. Carman and J. B. Guckert, J.N. Am. Benthol. Soc. 13,
84. S. Möller et al., Appl. Environ. Microbiol. 64, 721–732 80–88 (1994).
(1998). 120. R. T. Edwards and J. L. Meyer, Freshwater Biol. 24,
85. A. T. Nielsen et al., Environ. Microbiol. 2, 59–68 (2000). 453–462 (1990).
86. B. L. Bassler, Curr. Opin. Microbiol. 2, 582–587 (1999). 121. L. A. Carlough and J. L. Meyer, Limnol. Oceanogr. 36,
873–883 (1991).
87. M. Manefield et al., Microbiology 145, 283–291 (1999).
122. J. Iriberri et al., Microb. Ecol. 26, 111–124 (1993).
88. R. J. C. McLean, M. Whitely, D. J. Stickler, and W. C.
Fuqua, FEMS Microbiol. Lett. 154, 259–263 (1997). 123. M. A. Sleigh et al., Hydrobiologia 248, 53–64 (1992).
89. D. J. Stickler, N. S. Morris, R. J. C. McLean, and C. Fuqua, 124. G. G. Geesey, ASM News 48, 9–14 (1982).
Appl. Environ. Microbiol. 64, 3486–3490 (1998). 125. I. W. Sutherland, CRC Crit. Rev. Microbiol. 10, 173–201
90. B. Normander et al., Appl. Environ. Microbiol. 64, (1983).
1902–1909 (1998). 126. T. R. Neu and K. C. Marshall, J. Biomater. Appl. 5, 107–133
91. A. Muela et al., Appl. Environ. Microbiol. 60, 4273–4278 (1990).
(1994). 127. T. R. Neu and K. C. Marshall, Biofouling 3, 101–112 (1991).
 BIOFILMS IN NATURAL AND DRINKING WATER SYSTEMS 575

128. C. Freeman and M. A. Lock, Limnol. Oceanogr. 40, 273–278 152. M. W. Silver and A. L. Alldredge, J. Mar. Res. 39, 501–530
(1995). (1981).
129. T. K. Haack and G. A. McFeters, Microb. Ecol. 8, 115–126 153. H. Ploug et al., Aquat. Microb. Ecol. 13, 285–294 (1997).
(1982). 154. H. Zimmermann and H. Kausch, Arch. Hydrobiol., Spec.
130. L. G. Leff et al., Freshwater Biol. 39, 129–134 (1998). Issues Advanc. Limnol. 48, 85–92 (1996).
131. M. N. Mohamed et al., Microb. Ecol. 36, 121–130 (1998). 155. H. Zimmermann-Timm et al., Estuaries 21, 613–621 (1998).
132. A. Gjaltema et al., Biotechnol. Bioeng. 44, 194–204 (1994). 156. F. Azam et al., in R. Guerro and C. Pedro-Alio, eds., Trends
in Microbial Ecology, Spanish Society for Microbiology,
133. T. A. Larsen and P. Harremoes, Water Res. 28, 1443–1452
Barcelona, 1993, pp. 410–414.
(1994).
157. K. Lochte, Habilitationsschrift der Universität Bremen
134. T. R. Neu and J. R. Lawrence, FEMS Microbiol Ecol. 24,
1994.
11–25 (1997).
158. D. C. Smith et al., Nature 3,359, 139–142 (1992).
135. J. R. Lawrence et al., J. Microbiol. Methods 32, 253–261
(1998). 159. J. J. Cole, Ann. Rev. Ecol. Syst. 13, 291–314 (1982).
160. G. Zwart et al., Syst. Appl. Microbiol. 21, 546–556 (1998).
136. W. Manz et al., Microb. Ecol. 37, 225–237 (1999).
161. S. C. Nold and G. Zwart, Aquat. Ecol. 32, 17–35 (1998).
137. I. H. M. Brümmer, W. Fehr, and I. Wagner-Döbler, Appl.
Environ. Microbiol. 66, 3078–3082 (2000). 162. B. C. Crump et al., Appl. Environ. Microbiol. 65, 3192–3204
(1999).
138. L. J. Albright et al., Appl. Environ. Microbiol. 51, 614–621
(1986). 163. B. C. Crump and J. A. Baross, FEMS Microbiol. Ecol. 31,
231–239 (2000).
139. M. Simon et al., Mar. Ecol. Prog. Ser. 65, 205–211 (1990).
164. M. W. LeChevallier et al., Appl. Environ. Microbiol. 53,
140. A. L. Alldredge and M. W. Silver, Prog. Oceanogr. 20, 41–82 649–654 (1987).
(1988).
165. I. T. Miettinen et al., Nature 381 654–655 (1996).
141. H. W. Paerl, Limnol. Oceanogr. 19, 966 (1973).
166. B. H. Olson and L. A. Nagy, Adv. Appl. Microbiol. 30,
142. M. Simon, Limnol. Oceanogr. 32, 591–607 (1987). 73–132 (1984).
143. P. Weiss et al., Appl. Environ. Microbiol. 62, 1998–2005 167. S. Kalmbach et al., Water Res. 34, 575–581 (2000a).
(1996).
168. P. Niquette et al., Water Res. 34, 1952–1956 (2000).
144. I. G. Droppo and E. D. Ongley, Water Res. 28, 1799–1809 169. O. M. Zacheus et al., Water Res. 34, 63–70 (2000).
(1994).
170. D. van der Kooij et al., J. Water SRT-Aqua 48, 31–37 (1999).
145. B. Hoch et al., Aquat. Microb. Ecol. 9, 279–288 (1995).
171. B. Holden et al., Water Sci. Tech. 32, 213–220 (1995).
146. B. Berger et al., Aquat. Microb. Ecol 10, 37–44 (1996).
172. C. D. Norton and M. W. LeChevallier, Appl. Environ. Micro-
147. T. R. Neu, Aquat. Microb. Ecol. 21, 85–95 (2000). biol. 66, 268–276 (2000).
148. U. Böckelmann et al., FEMS Microbiol. Ecol. 33, 157–170 173. W. Manz et al., Appl. Environ. Microbiol. 59, 2293–2298
(2000). (1993).
149. A. L. Shanks and D. Trent, Limnol. Oceanogr. 24, 850–854 174. T. Schwartz et al., Water Res. 32, 2787–2797 (1998b).
(1979). 175. R. Macdonald and V. S. Brözel, Water Res. 34, 2439–2446
150. D. Kirchman and R. Mitchell, Appl. Environ. Microbiol. 43, (2000).
200–209 (1982). 176. S. Kalmbach et al., in H.-C Flemming, T.Griebe, and
151. A. L. Alldredge and C. C. Gotschalk, Deep-Sea Res. 36, U. Szewzyk, eds., Biofilms: Investigative Methods and Appli-
159–171 (1989). cations, Technomic, Lancaster, 2000b, pp. 107–121.
 B(contd.)
BIOFILMS IN THE FOOD INDUSTRY typhimurium, and Yersinia enterocolitica produce biofilms
readily, causing severe sanitation problems on surfaces in
GUN WIRTANEN the food industry (15–21). Experiments have been carried
TIINA MATTILA-SANDHOLM out with L. monocytogenes, Pseudomonas fluorescens, and
VTT Biotechnology Y. enterocolitica, which cause contamination of materials
Espoo, Finland used in gaskets (22). Pseudomonas spp. have also been
found to attach easily to surfaces of hydrophobic materials
This entry deals with biofilm formation, detection, and (e.g., polystyrene. rubber- and teflon-based materials) used
elimination on surfaces in the food processing environ- in gaskets. Some microbes are also able to decompose rub-
ment. Biofilms have been used in research on sanitation, ber and use it as a source of energy (23–24). Biofilm-bound
that is, cleaning and disinfection, in the food industry bacteria can accelerate corrosion and material deterio-
since late 1980s. Microbes in microcolonies have a ten- ration, for example, in sensors and detectors (25–28).
dency to form protective extracellular matrices, which are Microbial corrosion has been observed in processing equip-
called biofilms. The microcolony formation is the start- ment, appearing either in combination with or without
up in the biofilm formation. The biofilm formation occurs biofilm formation. Corrosion leads to huge losses in differ-
under suitable conditions on any — both inert and liv- ent industrial areas, for example, piping and cooling water
ing — surface. Besides a surface, the microbes need only systems (26,28).
water to start-up this formation. The mechanisms of In the food industry, equipment design plays the most
the microbial attachment and biofilm buildup can be important role in combating biofilm formation (29,30).
divided into the following steps: (1) preconditioning of The choice of surface materials is of great importance in
the surface by macromolecules, (2) transport of cells to designing and building equipment and processing lines for
the surface, (3) reversible and irreversible adsorption to industrial use (31). Dead ends, corners, cracks, crevices,
the surface, (4) cell replication, (5) transport of nutrients gaskets, valves, and joints are vulnerable points for
and metabolism, (6) production of extracellular polymers, biofilm accumulation (29,32–33). The process equipment
and finally, (7) detachment. The food hygiene legislation, is easy to clean if the surface materials are smooth
the hygienic design of machinery together with increas- and in good condition (Fig. 1). It should be cleaned
ing public awareness of product quality, make testing of frequently to avoid the accumulation of biofilm. The choice
process surfaces and equipment cleanliness important.
Microbiological, physical, chemical, and microscopy meth-
ods can be used in studying the biofilm buildup and its (a) (b)
elimination. Advanced molecular biology methods can also
be used in the detection and enumeration of specific organ-
isms on surfaces. Determination of disinfectant efficacy is,
however, still performed in suspensions, which do not
mimic the conditions on surfaces where the agents are
expected to inactivate microbes within biofilms as well.
The elimination of biofilm is a very difficult and demand-
ing task because many factors such as temperature, time,
mechanical, and chemical forces affect the detachment.
(c) (d)

FORMATION OF BIOFILMS IN THE FOOD INDUSTRY

Biofilm formation causes problems in areas such as

industrial water systems (1–3), medicine (4–6), and in
the food-processing industry (7–12). Biofilm formation has
both positive and negative implications in food related
processes (13). It has been used effectively in the food
industry in the production of vinegar (12) and beer (14).
0.7 cm
Biofilms can generally be produced by any microbes
Figure 1. Epifluorescent microscopy photographs of 2 and
under suitable conditions, although some microbes nat-
10 days old L. monocytogenes biofilms stained with acridine
urally have a higher tendency to produce biofilm than
orange. The biofilms had been grown on standard as well
others. The most common food-borne biofilm producers as electrochemically polished 2B stainless steel surfaces (AISI
belong to the genera Alcaligenes, Bacillus, Enterobac- 304): (a) 2-day-old and (b) 10-day-old L. monocytogenes biofilm on
ter, Flavobacterium, Pseudomonas, and Staphylococ- standard 2B, (c) 2-day-old and (d) 10-day-old L. monocytogenes
cus (10). It is somewhat alarming that bacteria such biofilm on electrochemically polished 2B surfaces. The scale
as Listeria monocytogenes, Escherichia coli, Salmonella marker is equivalent to 0.7 cm. See color insert.
577
578 BIOFILMS IN THE FOOD INDUSTRY

of materials and their surface treatments, for example, and salad products (20,42,44–49). Biofilms cause prob-
grinding and polishing, are important factors in inhibiting lems at various sites: air-handling systems (46,50–51),
the formation of biofilm and in promoting the cleanability blancher extractors (46), conveyors (12,46), cooling sys-
of surfaces (34). Treatments of surface materials to reject tems (46,50), floors (12,52), floor drains (12,46), food
biofilms can be performed actively to remove or passively contact surfaces of stainless steel (35,42,46), gas-
retard biofilm reoccurrence. The cleanliness of surfaces, kets (45,48–49), heat exchangers (47,53), manufacture
training of personnel, and good manufacturing and design line for paper-based packaging material (28,54), milk-
practices are the most important tools in combating biofilm transfer lines (45,47), poultry-processing equipment (55),
problems in the food industry (30). rubber-fingered pluckers (44), mixers, slicers, packag-
Besides causing problems in cleaning and hygiene (35), ing machines (46), pasteurizers (47), ultrafiltration and
biofilm can cause energy losses and blockages in condenser reverse osmosis membranes (38,47), vegetable lines (46).
tubes, cooling fill materials, water, and wastewater We can see from the list that problems from biofilm can
circuits and heat exchangers (36). Biofilms can also cause occur anywhere in the food process if the design and
health risks by releasing pathogens in drinking water maintenance is improper.
distribution systems. Biofilms may enter a food processing
system by causing reduced effectiveness in ion exchange Open Process Lines in Food Processing
and membrane processes (37–38). In the food processing
Although there is a long history of biofilm impacting the
water supply systems, biofilms cause problems in granular
food industry, actual biofilm research in this industrial
activated carbon columns, reverse osmosis membranes,
area began in the late 1980s (8,56). The occurrence of
ion exchange systems, degasifiers, water storage tanks,
slime-forming microbes is a major problem in the sanita-
and microporous membrane filters. The bacterial level
tion and disinfection of process equipment. Good surfaces,
of the purified water used and the defects found in
instruments, and equipment hygiene in the food indus-
microelectronic devices correlate directly. Conductance,
try promotes the quality of the products processed (55,57).
electromigration, and corrosion in the oxide layers have
Commonly found microbes on food contact surfaces are
been observed in devices manufactured with insufficiently
enterobacteria, lactic acid bacteria, micrococci, strepto-
pure water (2). Accumulation of mixed population biofilms
cocci pseudmonads, and bacilli. Problems with the accu-
containing sulfate-reducing bacteria causes corrosion in
mulation of particulates and cells will occur whenever
industrial water systems. A summary of biofilm effects in
cleaning is for any reason inappropriate (52). The most
various processes follows:
useful material in processing equipment is steel, which can
be treated with, for example, mechanical grinding, brush-
• Reduced product quality in food production and
ing, lapping, and electrolytic or mechanical polishing. In
processes producing goods for food handling, for
electrolytic polishing, a preground surface is treated in
example, package material
an electrolyte bath to obtain an even surface. The surface
• Reduced water quality (e.g., pathogenic organisms) structure of stainless steel is very important in avoid-
in cooling systems and in water distribution systems ing biofilm formation (Fig. 2). It has been reported that
• Contamination of immobilized systems, for example, although the grain boundaries of AISI (American Iron
loss of specific chemical transformations and Steel Institute) 316L acid-resistant stainless steel
• Material deterioration in metal condenser tubes, constitute 3 to 20% of the total surface area, over 90% of
pipelines, and cooling systems the adherent bacteria were found attached to the grain
• Energy losses in heating of the process equipment boundaries (58).
• Reduction in the effectiveness on remote sensors,
sight glasses, etc.
(a) (b) (c)
Disinfection after removal of biofilms using suitable
cleaning procedures is required in food plants where wet
surfaces provide favorable conditions for the growth of
microbes (22). In practice, a biofilm left on improperly
cleaned surfaces is a barrier between microbes and
the disinfectants, antibiotics, or biocides used against
(d) (e) (f)
them (39–42). The permeability barrier of the biofilm
considerably affects the efficiency of these agents. The
effect of many antibiotics is based on inhibition of active
growth. It has been said that most of the bacteria in
biofilms are no longer growing actively and their resistance
is therefore altered. The resistance of microbes in a
Figure 2. Epifluorescent microscopy photographs of 2-day-old
biofilm cannot be proved without controlling their growth
Bacillus subtilis biofilms stained with acridine orange. The
rate (43). biofilms had been grown on various stainless steel surfaces (AISI
The food industries in which research on biofilms 304): (a) standard 2B, (b) standard 4N, (c) standard BA, (d) glass
has been carried out are those producing canned prod- blasted 2B, (e) brushed 2B and (f) electrochemically polished 2B
ucts, meat and poultry products, pastries, biscuits, pizza, surfaces. The scale marker is equivalent to 0.5 cm. See color
fish cakes, cheese, milk products, beer, spices, vegetable insert.
 BIOFILMS IN THE FOOD INDUSTRY 579

Inadequate cleaning and sanitation of surfaces coated membranes (65). When crust has fouled the probe no
with biofilm cause contamination because the biofilm probe for measuring properties in biofilms can provide
protects the microbes against sanitizers and disinfec- an accurate description of the process (29). The pene-
tants (22,59). Many equipment faults can be avoided by tration of disinfectants and heat are hampered by the
using good design practice guidelines, which are summa- porous matrix on the surface, which diminishes the effect
rized in Table 1 (10). of sanitation and sterilization procedures. The bacterial
The lubricants used in conveyors can function as a slime of Bacillus sp. improved the heat resistance of
contamination source, especially in dairies and brew- the bacterium, extending the autoclaving time required
eries. Oil-based lubricants containing water are very for successful sterilization to several hours. Biofilm com-
sensitive to microbial growth. Acinetobacter spp., Alcali- ponents can also alter the resistance to steam and
genes spp., Pseudomonas spp., and sulfate-reducing formalin (10).
bacteria have been isolated from lubricants, and the
biofilm formed in the lubricant can indirectly pro- Cooling Systems
mote corrosion (60). Listeria monocytogenes, an oppor- Biofilm formation in process water systems has been
tunistic pathogen, has been isolated from lubricants known for a long time (3,66–68) and the adhesion of
in dairies (61). Problems with biofilms are also met in microbial cells to the surfaces in water distribution
engineering works in which lubricants based on veg- systems is also well known (1,69–71). The formation of
etable oils are used. Synthetic lubricants containing biofilm takes several days at least, or weeks, in low-
biocides have been developed, and the problems associ- nutrient water systems. Biofilms formed in slow-flowing
ated with contamination of lubricants has been decreas- systems are structurally different from those formed in
ing (62). systems with high flow rates (72–73). The number of free
planktonic cells in water does not necessarily correlate
Closed Equipment in Food and Drink Production
with the amount of biofilm on the pipe surfaces. The
Poorly designed sampling valves can destroy an entire numbers of microbial cells adsorbed to surfaces may
process or give rise to incorrect information because of be as much as 500 to 50,000 times the numbers of
biofilm effects at measuring points. Owing to their con- planktonic cells in water (74). Biofilm formation appears
struction, valves are vulnerable to microbial growth and in sinks, faucets, valves, and in different joint areas
thus constitute a hygiene risk (29,63). Hoses, tubes, fil- through the piping systems (3,75). It has been found that
ters, etc. containing polyvinylchloride increase the risk temperatures below 50 ° C promote biofilm formation (72).
of contamination (64). Biofilm is easily formed on mea- The average temperature in cooling water systems is
suring probes and the inner parts of equipment because 35 ° C, which is close to the optimum temperature for
these are usually difficult to clean (45,47). Microbial accu- most microbial growth. From a hygienic point of view,
mulation is known to be a problem on reverse osmosis biofilm can include organisms that cause infections
and diseases. Problems with Legionella pneumophila, a
Table 1. Factors in Good Design Practice (10)
pathogenic organism that is able to form biofilm, can
occur in hot water systems (76–78). Biofilm formation
Design Area with Relevant Factors can be decreased by more than 50% if the oxygen level
in the water system is minimized, and by more than
• Organization and personnel responsibility 80% when the amounts of both oxygen and nutrients
functions, tasks, education, and courses
are minimized (72). Anaerobic conditions favor biofilm
• Production and production control
formation by anaerobic organisms, which can increase
production: process parameters, installations, and
maintenance of hygiene corrosion (73,79).
production control: working plan, material flow, and
disturbance reports Ventilation and Air-Handling Systems
• Equipment, pipelines, and equipment layout Air quality in food production facilities is very important
equipment: materials, surface finishing, accessories for the final product quality. The microbial population
and joints, checking and cleanability, clearing and
in the air channels depend on the environment, filtra-
cleaning and insulation
pipelines: accessories, materials, surface finishing,
tion membranes, and the sitings of airholes (51,80,81).
clearing and cleaning, consoles and insulation Formation of biofilm in air-conditioning systems does
equipment layout: cleanability and cleaning of not occur without a water reservoir of some kind. Nor-
equipment, servicing of equipment, rationalization of mally, there is no water in the air-conditioning sys-
operations and systematic layout tems, but it can accumulate unintentionally through
• Buildings and structures, electrification, and condensation. The pathogen L. pneumophila has been
automation isolated from water systems that were connected to
buildings and structures: materials, painting, an air conditioner (82). When the air-conditioning sys-
sewerage, tipping, joints, cleanability of surfaces, tem is to be cleaned and disinfected, it is very impor-
layout of the connections of different production
tant that the disinfection medium penetrates the biofilm
units and air-conditioning
and does not simply flow through the system with the
electrification: shielding of equipment and lighting
automation: control of production and process air (10).
reliability A study carried out in various industrial ventila-
tion ducts has shown that biofilms can accumulate
580 BIOFILMS IN THE FOOD INDUSTRY

within the ducts. The occurrence of bacterial and fun- DETECTION AND ELIMINATION OF BIOFILMS
gal biofilms takes place within a few weeks. Biofilms
form very easily in humid surroundings where oil Detection of Biofilm
has been left on the surfaces. Biofilm growth was
Many reviews have been published on the formation
prevented in clean room ducts by using effective fil-
and detection of biofilms in the laboratory (34,42,93,94).
ters and appropriate maintenance procedures (50). It
Methods of studying biofilm formation include microbiolog-
is therefore very important to monitor clean room
ical, physical, chemical, and microscopical methods, and,
behavior, as the majority of particles in clean rooms
methods commonly used in molecular biology (Table 2;
originate from the personnel (81,83). The air filtra-
34,42,94–97). The biofilm consists of 85 to 96% water,
tion equipment must be chosen carefully to suit the
which means that only 2 to 5% of the total biofilm vol-
processing environment (83,84). The membranes in the
ume is detectable on dry surfaces (4). Monitoring practices
air-conditioning system and the walls in the air-
based on sampling of the liquid phase does not reflect the
conditioning channels are places where biofilms start to
location or extent of microbes growing on surfaces. The
grow (80,85).
counts of planktonic cell in the processing fluids should
not be used in detecting biofilms because they do not
Production of Paper and Packaging Material represent the sessile organisms (98).
Inadequate cleaning and sanitation of surfaces coated
Biofilm formation causes major problems in paper with biofilms represents a source of contamination within
machines and it affects the production of good quality the process (42). Practical methods for assessing microbes
paper-based packaging material for the food indus- and organic soil on processing surfaces are needed
try (28,54). The paper machines are connected to water- to establish the optimum cleaning frequency of the
circulation systems, which are prone to rapid biofilm equipment. These systems should, preferably, provide
growth (86). The microbes, mostly bacteria and fungi, information online, nondestructively and in real time,
enter the process with the nutrients and from the air (87). about microbial growth required within the process.
The use of recycled fibers and changes in the acidity of Online monitoring techniques transferred from various
the process to neutral or alkaline have increased the total industrial fields, for example, glass fiber optics, infrared
amount of bacteria in the paper-making processes (88). It techniques, ion-mobility techniques, bioluminescence,
should be remembered that each paper-making machine microelectrodes, and heat transfer, could be used for
has its own microbial flora, which depends on process- assessing the quality of raw material and products and
ing parameters such as pH and temperature, the season, for control of microbial and chemical contaminants on
and the additives used. The slime forming microbial processing surfaces. Advanced molecular biology methods
flora can be divided into primary slime formers, for can be used in the detection and enumeration of specific
example, bacteria of the Bacillus, Pseudomonas, and organisms on surfaces (34,99–101).
Enterobacter genera and fungi of the Aspergillus, Mucor, Both open and closed process equipment should
and Penicillium genera, and secondary slime-formers, be easily cleanable, and therefore, cleanability testing
methods are needed. Development of practical detection
for example, bacteria of the Alcaligenes, Flavobacterium,
methods for research on biofilm buildup is needed in
Klebsiella, Micrococcus, and Staphylococcus strains, and
order to use the knowledge of biofilm accumulation
fungi of the Paecilomyces, Trichoderma, and Trichospo-
in the maintenance of process equipment. The natural
rum strains (89). Practical experiments show that biofilm
luminescence of marine bacteria has been used to assess
in pulp- and paper-processing equipment detaches eas-
biofilms in marine systems and for on-line evaluation
ily from the surface and the loose slime causes holes
of biofilm formation and antifouling coatings (102,103).
in the paper (90). Bacterial spores withstand the heat
The transformation of luminescence-encoding lux-genes
treatment during the production and therefore the Bacil-
to food-spoiling bacteria or opportunistic pathogens
lus genera, which are spore formers, are the main
makes it possible to use bioluminescence in cleanability
source of contaminants in final paper and cardboard
studies (104). Further efforts are needed to optimize
products (91). The slime in paper-machine biofilms con-
the microbial cleanability assay, using, for example, the
tains 1011 to 1012 cfu/g dry mass. The microbial growth
Photobacterium leiognathi procedure (42), and to adapt
causes problems in corrosion, process flow, contamina-
the assay to both open and closed systems.
tion, and paper quality (92). It also causes odor problems,
which can destroy both the air quality in the pro-
Chemical Agents Used in the Elimination of Biofilms
cess facilities and the quality of the end products (88).
Bacterial biofilms of anaerobic sulfate-reducing bacte- Efficacy of disinfectants and antimicrobial agents are
ria, thermophilic bacteria, and bacteria producing acids usually determined in free-cell suspensions, which do
and reducing compounds (5) cause corrosion of surfaces not mimic the growth conditions on surfaces where the
in the paper machines. Microbial corrosion causes more agents are required to inactivate the microbes (105–108).
than 50% of all corrosion problems in pipelines in paper- The agent must reduce the microbial population by 5 log
processing machines. Efforts to combat microbial growth units in suspensions in order to be considered effective.
and biofilm formation in paper machines are important The goal for reduction of surface-attached bacteria with
because the costs of maintenance of equipments are disinfectants is 3 log units (22). The standard suspension
high (28,92). tests have proved sufficiently reliable because the
 BIOFILMS IN THE FOOD INDUSTRY 581

Table 2. Summary of Main Methods Used in Process Hygiene (34)

Detection Method The Application Assesses

Conventional cultivation based on swabbing and Living bacteria on various types of growth media
contact agar techniques Limitation: sensitivity, time consuming
Impedance measurement Activity of viable and injured bacteria directly and indirectly
after reactivation in liquid growth media
Limitation: sensitivity
ATP measurement Total hygiene; large scatter in analysing biofilms because of
different metabolic status of cells
Limitation: sensitivity
Measurement of protein residues amount of protein-based soil
Limitation: sensitivity
Microscopical techniques using various staining Limitation: not for routine use on-site
procedures:
• epifluorescence microscopy Area coverage; microbial distribution; cell type and cell viability;
identification of microbes
• electron microscopy Morphology of microbes; cell distribution; identification of
microbes
• scanning confocal laser microscopy Biofilm structures; ecology of biofilms; identification of microbes
• interference reflection microscopy Adhesive interactions between surfaces and microbes
• differential interference contrast microscopy Individual microbial cells (not for quantitative analysis)
Flow cytometry Microbial cell densities
Limitation: detachment of cells
Polymerase chain reaction (PCR)-based methods Unculturable microorganisms; identification of bacteria in
complex ecosystems
Limitation: information on presence, but not on numbers and
Viability of target organisms

variations of results are within acceptable limits when the maintenance of the equipment, because the debris
replication is adequate. There can, however, be problems left on surfaces can act as nutrients for the buildup of
with the repeatability and reproducibility of suspension new biofilm (36), or may corrode the equipment material.
tests if the organic load in the suspension varies. It is To minimize biofouling, it is best to clean the equipment
obvious that the surface tests are even more difficult at frequent and regular intervals, using suitable efficient
to perform than suspension tests because of the carrier cleaning procedures for the process before the biofilm has
material used and the viability of dried cells on the the opportunity to develop (113,114).
surfaces (109,110). Efficient cleaning of surfaces depends on surface
Various disinfectants have been developed to destroy tension, viscosity, chemical reactivity, size of biofilm
microbes (111). Microbes have, nevertheless, been found particles, solubility, and living microorganisms (115,116).
in disinfectant solutions, which means that microbial con- The cleaning agents should be surface-active, soluble,
taminants can be spread on the surface with the solution. easy-to-rinse, noncorrosive, and easy-to-use. They should
Findings as early as 1967 reported that chlorhexidine
also emulsify fats and oils, suspend insoluble particulates,
mixtures were contaminated with Pseudomonas spp.,
and dissolve mineral deposits. Soaps and synthetic
which also have been found in concentrated iodine solu-
detergents clean surfaces in a similar manner, the
tions (112). Serratia marcescens was found to be viable
surfactant molecule having hydrophilic and hydrophobic
even after 27 months in a disinfectant containing 2%
ends. The hydrophobic end dissolves oily soils (116).
chlorhexidine. A concentration of 0.1% chlorhexidine is
The alkaline and acidic detergents are mixtures of
sufficient to kill the cells of S. marcescens if they are
freely suspended in liquid (74). Microbial contamination several classes of compounds: alkalis, acids, tensides,
has also been found in solutions of aldehydes and qua- and surface-active agents (116). Currently used cleaning
ternary ammonium compounds (82). Other microbes that agents contain chelators, for example, EDTA, surfactants,
have been isolated from disinfectants include Alcaligenes detergent builders, antideposition agents, enzymes, and
faecalis, Enterobacter cloacae, E. coli, Flavobacterium disinfectants (1,116,117).
meningosepticum, and Pantoea agglomerans. The chelators, however, are not biocides and should
therefore be combined with other antibacterial com-
Cleaning Agents. Cleaning agents are applied to remove pounds (82). Chelators bind magnesium and calcium ions,
soil, microbes, and biofilms from surfaces. The effects of the thereby destabilizing the outer membranes of microbial
cleaning agents on biofilms have been thoroughly investi- cells and extracellular matrices. Chelators have been used
gated (22,57,59,106). Removal of biofilm is important for in the breakage of the biofilm layers (118,119). Tensides
582 BIOFILMS IN THE FOOD INDUSTRY

are large molecules with polar and nonpolar groups form- induce biofilm buildup, for example, of Pseudomonas
ing hydrophobic and hydrophilic parts, depending on the sp. (70). Pseudomonas spp. are relatively resistant to chlo-
group. The tensides are divided into four types: anionic, rine treatments and can even multiply when chlorine has
cationic, nonionic and amphoteric tensides. Tensides are been used. Free chlorine (1.0 mg/L) showed no effect on
normally used in cleaning agents to achieve suitable clean- coliforms growing in biofilms, and even increased concen-
ing and foaming effects. The amphoteric tensides contain trations (2.0 mg/L) did not kill E. coli grown in biofilms.
both anionic and cationic parts and can therefore function The material on which microbes form biofilm has a major
both in acidic and alkaline environments or as double ions influence on the effect of the disinfectants used. Cap-
in the boundary region (120). sulated Klebsiella pneumoniae grown on glass surfaces
has been shown to have a 150-fold resistance to chlorine
Disinfectants. Disinfection is required in food plant compound to those grown in suspensions. When a low-
operations in which wet surfaces provide favorable nutrient liquid is used, the resistance factor is increased
conditions for the growth of microorganisms (57,121). The to about 600. On metal and carbon surfaces, the corre-
aim of disinfection is to reduce the surface population of sponding resistance ratios were about 2,400 and 3,000
viable microorganisms after cleaning, by destruction or times the value in suspension, respectively (70). The effect
removal, and to prevent microbial growth on surfaces of surfactant sanitizers against L. monocytogenes grown
during the interproduction period (122). Disinfectants in biofilms was decreased (107,128); however, sodium
do not penetrate the polysaccharide and glycoprotein hypochlorite and quaternary ammonium compounds have
matrix left on the surfaces after an ineffective cleaning proved effective against 24-hour old biofilms (129).
procedure, and thus do not destroy all the living Iodophors are used extensively in the food industry.
cells in biofilms (7,11). Disinfectants are most effective Iodine oxidizes essential parts of the microbial cells. Like
in the absence of organic material, for example, fat-, chlorine-containing products, iodophors are active against
sugar- and protein-based materials (59). Microbes are less gram-positive and gram-negative bacteria, yeasts, and
likely to survive disinfection after an effective cleaning. molds (116,122,125). Bacterial spores, however, are highly
The efficiency of disinfectants is generally controlled resistant to iodophors. Iodophors cannot be used in food
by interfering organic substances, pH, temperature, plants in which starch-containing products are produced
concentration, and contact time (22,123). The desired because iodine forms a purple complex with starch (116).
characteristics of disinfectants are the same as for cleaning Quaternary ammonium compounds are used as sanitizers
agents. They must be effective, safe, and easy-to-use, and in dairies and in the food industry because they have
easily rinsed off surfaces, leaving no toxic residues or good wetting properties and are nonspecifically described
residues that affect the sensory values of the product (116). as cationic surface-active agents in which the cationic
The use of disinfectants in food plants depends on part is hydrophobic. The greatest effect of quaternary
the material used and the adhering microbes (116). It ammonium compounds is observed against gram-positive
is therefore recommended that disinfectants used in bacteria, whereas gram-negative organisms, many of them
the sanitation of food-processing equipment should be significant in the contamination of food, may not be
tested on surfaces in circulating systems. Disinfectant affected (116).
testing in a circulating system provides many replicates, Hydrogen peroxide has been found to be effective in
which makes the evaluation easier and more rapid removing biofilms from equipment used in hospitals. The
to perform. The variety in soil types, cleaners, and effect of hydrogen peroxide is based on the production of
cleaning conditions makes it impossible to give an free radicals, which affect the polysaccharides and glyco-
exact, overall statement on required temperatures and proteins in the biofilm (130). The biocidal effect of peracetic
concentrations (124). Disinfectants approved for use in the acid on microbes in biofilms was shown to vary (131–133).
food industry contain chlorine-based compounds, iodine Aldehydes did not disrupt the biofilm, but rather seemed
compounds, alcohols, quaternary ammonium compounds to improve its stability. The biofilm must be disrupted in
or surfactants (1,116,125,126). The concentration of the some way before chemical agents such as peracetic acid
disinfectant can be altered to provide shock treatments. and aldehydes can be used effectively (131). The effect of
It has also been suggested that the disinfectants should ozone treatment has been found to vary depending on
be changed continuously. In the maintenance of process the processing circumstances and the microbes tested; for
hygiene, disinfectants should be chosen according to the example, ozonation proved very effective in the treatment
process (127). of cooling water systems (134,135). Physical disinfection
Chlorine or chlorine-based compounds that are can also be performed using ultraviolet irradiation, pulsed
approved for use in food plants, for example, gaseous laser beams, or steam disinfection (116).
chlorine, sodium hypochlorite, calcium hypochlorite,
chloramine-T, and chlorine dioxide. The antibacterial Elimination of Biofilms in Industrial Systems
active moiety is formed when chlorine gas or a hypochlorite
compound is added to water and produces hypochlorous General Aspects of Cleaning. Physical, chemical, and
acid. Stabilized hypochlorites are used when disinfection microbiological cleanliness are essential in food plants.
of long duration is required (116). The range of microor- Factors governing the selection of detergents and disin-
ganisms killed or inhibited by chlorine-based compounds fectants in the food industry are that the agent should
is probably broader than that killed by any other approved be efficient, safe, not damage or corrode equipment,
sanitizer. Stressing factors such as chlorination can also be easy-to-rinse, and not affect the sensory values of
 BIOFILMS IN THE FOOD INDUSTRY 583

the product (59,113). Physical cleanliness means no vis- Elimination of Biofilm in Open Systems. Gross soil should
ible waste, foreign matter, or slime on the equipment be removed by dry methods, for example, brushing,
surfaces. Chemically clean surfaces are surfaces from scraping, or vacuuming, and visible soil rinsed off with low-
which undesirable chemical residues have been removed, pressure water. The cleaning effect is increased by using
whereas microbiologically clean surfaces imply freedom water of sufficient volume and temperature. However, a
from microbes (124). Attached bacteria or bacteria in pure-water washing system is not practical because of its
biofilms can be a problem in food processing and grow and ineffectiveness and cost limitations. Pressure cleaning is
multiply after insufficient cleaning and contaminate the used to remove the remainder of the soil and most of
product (35). Once a biofilm is firmly established, clean- the microorganisms present on open surfaces (122). The
ing, and disinfection becomes much more difficult (22,136). elimination of biofilm from open systems, for example,
Caution should be exercised in selecting appropriate dis- surfaces in food-processing equipment, has not been widely
infectants for contaminated surfaces. reported. Biofilms were used in cleaning and disinfection
The key to the effective cleaning of a food plant is the studies for the food industry (8,56). After a production
understanding of the type and nature of the soil (e.g., run, the equipment should be dismantled and the cleaned
sugar, fat, protein, and mineral salts) and of the microbial utensils should be stored on racks and tables, but not
growth on the surfaces to be removed. The accessibility on the floor. The cleaning of open-process surfaces is
and type of equipment and accessories to be cleaned carried out using either foam or gel cleaning. Foams are
and the availability of suitable cleaning agents are also most effective in situations in which contact with the soil
important (116). Intelligent integration and coordination for an extended period of time is necessary. The foam-
between cleaning programs and manufacturing operations units are constructed to form foam of varying wetness and
are essential to achieve both successful cleaning and durability depending on the cleaning to be performed. The
application of gels extends the contact time with a soiled
business profit (116). An efficient cleaning procedure
surface and can be used with low-pressure systems (116).
consists of a sequence of rinses and detergent and
The cleaning is mostly carried out in combination with
disinfectant applications in various combinations of
a final disinfection because there are likely to be viable
temperature and concentration (113,131). This controls
microorganisms on the surfaces that could harm continued
the development of biofilms on equipment surfaces without
production (57,122,131).
corroding the surfaces (113,114). An independent quality-
control system to monitor the cleaning results for a food
plant can be integrated in the program based on Hazard Elimination of Biofilm in Closed Systems. In the cleaning
Analysis of Critical Control Points (HACCP). The goal to regime applicable to closed systems, prerinsing with cold
water is carried out to remove loose soil. The cleaning-
achieve a clean food plant must be desired by the plant
in-place (CIP) treatment is normally performed using
management, which has to invest the necessary time and
hot cleaning solutions (29), but cold solutions can also
money to accomplish it. The personnel must be properly
be used in the processing of fat-free products. The
trained and responsible enough to maintain a good level of
warm alkaline cleaning solution, normally of 1% sodium
plant hygiene. The tools and materials must be properly
hydroxide (NaOH), is heated to 75 to 80 ° C and the
designed for the plant equipment and the methods used
cleaning time is 15 to 20 minutes. The equipment is rinsed
must suit the process (122,124).
with cold water before the acid treatment is performed
Factors affecting the cleaning process are based on
at approximately 60 ° C for five minutes. The effect of
mechanical and chemical impact, holding time, and
chlorine-based agents can be divided into three phases:
temperature (22,59,116). The basic task of detergents loosening of the biofilm from the surface, breakage of the
is to reduce the interfacial tensions of soils so that biofilm, and the disinfecting effect of the active chlorine
the soil becomes miscible in water (116). The effect of (137). The cleaning solutions should not be reused in
the surfactants is increased by the mechanical effect processes aiming at total sterility because the reused
of turbulent flow or water pressure, or of abrasives, cleaning solution can contaminate the equipment (29).
for example, salt crystals. A prolonged exposure of the Studies on CIP cleaning showed that the ‘‘cleaning
surfaces to the detergent makes the removal efficient. effect ‘‘decreased markedly when the washing time and
Detergents to be used in the cleaning of open systems the flow rate were reduced, indicating that the cleaning
are formulated to be effective at room temperature as a whole was affected by mechanical effects, time,
or at slightly elevated temperatures in the range 35 temperature, and chemical parameters (57,59). Turbulent
to 50° C (116). In closed systems, the detergents are flow is therefore a decisive factor in combination with
formulated to be used at temperatures in the range of chemical compounds in the elimination of biofilms from
55 to 80° C. Fats are easily removed at temperatures closed systems. To ensure the removal of soil and to avoid
slightly above their melting point. Sugars and other biofilm formation and soil sedimentation, the minimum
carbohydrates are water-soluble at elevated temperatures, flow velocity in the CIP treatment must be at least
but temperatures causing caramelization should be 1.5 m/s, but 2.0 m/s is recommended. The flow should
avoided. Proteins are denatured at elevated temperatures be turbulent with a Reynold’s number from at least
and may adhere strongly to surfaces at high-temperatures. 10,000 to preferably 30,000, to ensure good radial mixing
Cleaning and disinfection procedures can be optimized and heat, mass, and momentum transfer (138). Chelating
with pilot-scale equipment for both closed and open agents in the cleaning solution enhance the removal
processes. of biofilms from processing surfaces. The CIP-systems
584 BIOFILMS IN THE FOOD INDUSTRY

used today are based on a combination of alkaline-acid effect in open systems can be enhanced using either
treatments and time-temperature treatments. Problems increased chemical effect through double foaming or
caused by equipment constructions, valves, and surface added mechanical forces through scrubbing. In closed
materials cannot be eliminated with CIP because the processes, the removal of biofilms from surfaces can be
CIP treatment was not designed to eliminate biofilms. performed using efficient flowing condition in combination
In normal operation, it is impossible to clean pipelines with effective cleaning agents. Strong agents are used
completely by mechanical treatment because there are to combat microbial deposits on equipment surfaces in
always bends, corners, pockets, and cracks where biofilm problem areas. Satisfactory elimination of biofilms using
remains (70). The depth of dead zones in the system should only disinfectant treatment cannot be achieved. The
be less than two pipe diameters if the dead zone cannot effects of cleaning procedures used in the food industry
be avoided, to ensure adequate cleaning throughout the can be evaluated using life cycle assessment (LCA).
system using CIP procedures. The maintenance programs All environmental aspects, including the process and
should guarantee that excessive biofilm accumulation energy consumption in producing the cleaning chemicals,
does not occur. Tanks should be cleaned by applying the transportation, properties of chemicals before and after
cleaning solution through properly installed, removable cleaning, amount of water used, effects of organic and
spray balls or nozzles. The design should ensure that inorganic loads on the wastewater and on the recipient
the part of the tank directly above the spray ball is also (e.g., nutrient load, pH, temperature).
cleaned (29). Drainage, minimization of internal probes,
crevices and stagnant areas, arrangement of valves, BIBLIOGRAPHY
couplings, and instrument ports and instrumentation
should be planned carefully so that the equipment is easily 1. H.-C. Flemming, in H.-C. Flemming and G. G. Geesey, eds.,
cleanable (29). Biofouling and Biocorrosion in Industrial Water Systems,
Legislation on food hygiene and the hygienic design Springer-Verlag, Berlin, Germany, 1991, pp. 47–80.
of food machinery, together with the public awareness of 2. M. W. Mittelman, in H.-C. Flemming and G. G. Geesey, eds.,
product quality and manufacturers’ desires to improve Biofouling and Biocorrosion in Industrial Water Systems,
product safety, makes reliable cleanability testing an Springer-Verlag, Berlin, Germany, 1991, pp. 133–154.
important issue. In this type of testing, it must be 3. J. C. Block, in L. F. Melo, T. R. Bott, M. Fletcher, and
possible to assess the relative cleanability of various B. Capdeville, eds., Biofilms — Science and Technology,
equipment components to facilitate the design, testing, Kluwer Academic Publishers, Dordrecht, The Netherlands,
and maintenance of hygienic food processing equipment. 1992, pp. 469–485.
The assessment must use standardized test procedures 4. J. W. Costerton, R. T. Irvin, and K.-J. Cheng, Annu. Rev.
with a sound scientific basis (139). The aim of the Microbiol. 35, 299–324, (1981).
European Hygienic Equipment Design Group (EHEDG), 5. J. W. Costerton et al., Annu. Rev. Microbiol. 41, 435–464
which is an independent consortium of representatives (1987).
from research institutes, the food industry, equipment 6. T. R. Neu, H. C. van der Mei, and H. J. Busscher, in
manufacturers, and government organizations is to L. F. Melo, T. R. Bott, M. Fletcher, and B. Capdeville, eds.,
develop hygienic equipment on a scientifically and Biofilms — Science and Technology, Kluwer Academic Pub-
lishers, Dordrecht, The Netherlands, 1992, pp. 21–34.
technologically sound basis.
7. R. D. Pontefract, Can. Inst. Food Sci. Technol. J. 24,
113–117 (1991).
CONCLUSION 8. J. T. Holah, R. P. Betts, and R. H. Thorpe, J. Appl. Bacteriol.
65, 215–221 (1988).
Cleaning in the food industry should be on the basis of 9. J. T. Holah and L. R. Kearney, in L. F. Melo, T. R. Bott,
systematic planning. Biofilm formation in these systems M. Fletcher, and B. Capdeville, eds., Biofilms — Science and
is a symptom of disturbance in the process because biofilms Technology, Kluwer Academic Publishers, Dordrecht, The
can be produced under suitable growth conditions by any Netherlands, 1992, pp. 35–41.
microbe. Biofilms are less likely to accumulate in well- 10. T. Mattila-Sandholm and G. Wirtanen, Food Rev. Int. 8,
designed systems, which are effectively cleaned. Detection 573–603 (1992).
of buildup on equipment surfaces at an early stage 11. B. Carpentier and O. Cerf, J. Appl. Bacteriol. 75, 499–511
enables effective countermeasures and thus results in an (1993).
improvement of the process hygiene. Methods used for 12. E. A. Zottola and K. C. Sasahara, Int. J. Food Microbiol. 23,
monitoring the microbial load are based on conventional 125–148 (1994).
cultivation techniques using various types of growth media 13. A. C. L. Wong and O. Cerf, Bull. Int. Dairy Fed. 302, 40–50,
and for the total hygiene ATP (adenosine triphosphate). (1995).
Careful thoughts must be given to the cleaning procedures, 14. J. Kronlöf, Immobilized Yeast in Continuous Fermentation
including the program, cleaning agents, disinfectants, and of Beer, The Technical Research Centre of Finland, VTT
cleaning equipment. Results indicate that low-pressure Publications 167, Espoo, Finland, 1994.
cleaning is not sufficiently effective to remove biofilms, 15. D. M. Helke, E. B. Somers, and A. C. L. Wong, J. Food Prot.
unless, the cleaning agent is effective. The efficiency of 56, 479–484 (1993).
cleaning agents is assessed by their ability to remove 16. T.-J. Ren and J. H. Frank, J. Food Prot. 56, 573–576 (1993).
biofilm from process surfaces together with their ability 17. A. B. Ronner and A. C. L. Wong, J. Food Prot. 56, 750–758
to kill the bacteria present in the biofilm. The cleaning (1993).
 BIOFILMS IN THE FOOD INDUSTRY 585

18. K. Y. Kim and J. F. Frank, J. Food Prot. 57, 720–726 (1994). 45. J. W. Austin and G. Bergeron, J. Dairy Res. 62, 509–519
19. I. C. Blackman and J. F. Frank, J. Food Prot. 59, 827–831 (1995).
(1996). 46. H. Gibson, J. H. Taylor, K. E. Hall, and J. T. Holah, Biofilms
20. S. H. Flint, J. D. Brooks, and P. J. Bremer, J. Appl. Micro- and Their Detection in the Food Industry, CCFRA, Chipping
biol. 83, 508–517 (1997). Campden, Gloucestershire, U.K., 1995.
21. C. Ntsama-Essomba et al., Vet. Res. 28, 353–363 (1997). 47. S. H. Flint, P. J. Bremer, and J. D. Brooks, Biofouling 11,
22. T. M. Mosteller and J. R. Bishop, J. Food Prot. 56, 34–41 81–97 (1997).
(1993). 48. I. Klemetti, M. Puustinen, and M. Koivisto, in G. Wirtanen,
23. A. H. L. Chamberlain and S. Johal, in D. R. Houghton, S. Salo, and A. Mikkola, eds., 30th R3-Nordic Contamination
R. N. Smith, and H. O. W. Eggins, eds., Biodeterioration 7, Control Symposium, Libella Painopalvelu Oy, VTT Sympo-
Elsevier Publishers Ltd., Essex, U.K., 1988, pp. 57–61. sium 193, Espoo, Finland, 1999, pp. 139–143.
24. B. J. Zyska, in D. R. Houghton, R. N. Smith, and 49. E. Storgårds, H. Simola, A.-M. Sjöberg, and G. Wirtanen,
H. O. W. Eggins, eds., Biodeterioration 7, Elsevier Trans. IChemE C 77, C2:146–155 (1999).
Publishers Ltd., Essex, U.K., 1988, pp. 535–552. 50. K. Tarvainen, G. Wirtanen, and T. Mattila-Sandholm, Heal-
25. W. G. Characklis, in H.-C. Flemming and G. G. Geesey, eds., thy Buildings, Technical University of Budapest, Budapest,
Biofouling and Biocorrosion in Industrial Water Systems, Hungary, 1994, pp. 217–222.
Springer-Verlag, Berlin, Germany, 1991, pp. 7–27. 51. J. T. Holah et al., Airborne Microorganism Levels in Food
26. A. H. L. Chamberlain, in L. F. Melo, T. R. Bott, M. Fletcher, Processing Environments, CCFRA, Chipping Campden,
and B. Capdeville, eds., Biofilms — Science and Technology, 1995.
Kluwer Academic Publishers, Dordrecht, The Netherlands, 52. E. Mettler and B. Carpentier, J. Food Prot. 61, 57–65 (1998).
1992, pp. 207–217. 53. L. F. Melo and M. M. Pinheiro, in L. F. Melo, T. R. Bott,
27. W. Lee and W. G. Characklis, Corrosion 49.186–199 (1993). M. Fletcher, and B. Capdeville, eds., Biofilms — Science and
28. L. Carpén, in G. Wirtanen, S. Salo, and A. Mikkola, eds., Technology, Kluwer Academic Publishers, Dordrecht, The
30th R3-Nordic Contamination Control Symposium, Libella Netherlands, 1992, pp. 499–509.
Painopalvelu Oy, VTT Symposium 193, Espoo, Finland, 54. L. Raaska, in G. Wirtanen, S. Salo, and A. Mikkola, eds.,
1999, pp. 153–161. 30th R3-Nordic Contamination Control Symposium, Libella
29. Y. Chisti and M. Moo-Young, J. Ind. Microbiol. 13, 201–207 Painopalvelu Oy, VTT Symposium 193, Espoo, Finland,
(1994). 1999, pp. 145–152.
30. J. Holah and A. Timperley, in G. Wirtanen, S. Salo, and 55. D. Lindsay, I. Geornaras, A. v. Holy, and A. v. Holy, Micro-
A. Mikkola, eds., 30th R3-Nordic Contamination Control bios 86, 105–116 (1996).
Symposium, Libella Painopalvelu Oy, VTT Symposium 193, 56. J. T. Holah, R. P. Betts, and R. H. Thorpe, Int. Biodeteriorat.
Espoo, Finland, 1999, pp. 11–39. 25, 147–154 (1989).
31. M. Pirbazari, T. C. Voice, and W. J. Weber, Jr., Waste Haz- 57. R. E. Brackett, J. Food Prot. 55, 808–814 (1992).
ard. Mater. 7, 239–250 (1990). 58. J. D. Bryers and A. Weightman, SIM News 45, 103–111
32. EHEDG, Trends Food Sci. Technol. 4, 375–379 (1993). (1995).
33. EHEDG, Trends Food Sci. Technol. 4, 225–229, (1993). 59. M. H. Czechowski and M. Banner, Tech. Q. Mast. Brew.
34. G. Wirtanen, Biofilm Formation and Its Elimination from Assoc. Am. 293, 86–88 (1990).
Food Processing Equipment, The Technical Research Centre 60. W. A. Hamilton, in H.-C. Flemming and G. G. Geesey, eds.,
of Finland, VTT Publication 251, Espoo, Finland, 1995. Biofouling and Biocorrosion in Industrial Water Systems,
35. S. K. Hood and E. A. Zottola, Food Control 6, 9–18 (1995). Springer-Verlag, Berlin, Germany, 1991, pp. 187–193.
36. W. G. Characklis, Biotechnol. Bioeng. 23, 1923–1960 (1981). 61. K. Rossmoore, in D. R. Houghton, R. N. Smith, and
H. O. W. Eggins, eds., Biodeterioration 7, Elsevier
37. H.-C. Flemming, G. Schaule, and R. Mcdonogh, in
Publishers Ltd., Essex, U.K. 1988, pp. 242–247.
L. F. Melo, T. R. Bott, M. Fletcher, and B. Capdeville, eds.,
Biofilms — Science and Technology, Kluwer Academic 62. E. Henriksson and A. Haikara, Mallas ja olut, 132–139
Publishers, Dordrecht, The Netherlands, 1992, pp. 487–497. (1990).
38. G. Mucchetti, Int. Dairy Fed. Special Issue 9,504, 118–124 63. EHEDG, Trends Food Sci. Technol. 5, 169–171 (1994).
(1995). 64. D. Price and D. G. Ahearn, J. Clin. Microbiol. 26, 1650–1654
39. S. Kinniment and J. W. T. Wimpenny, Int. Biodeteriorat. 26, (1988).
181–194 (1990). 65. H. F. Ridgway and J. Safarik, in H.-C. Flemming and
40. W. W. Nichols, in W. G. Characklis and P. A. Wilderer, eds., G. G. Geesey, eds., Biofouling and Biocorrosion in Industrial
Structure and Function of Biofilms, John Wiley & Sons, New Water Systems, Springer-Verlag, Berlin, Germany, 1991,
York, 1989, pp. 321–331. pp. 81–111.
41. W. W. Nichols, Rev. Med. Microbiol. 2, 177–181 (1991). 66. G. G. Geesey et al., Can. J. Microbiol. 23, 1733–1736 (1977).
42. G. Wirtanen et al., Biofilms and biofouling in J. Walker, 67. R. Lewin, Science 224, 375–377 (1984).
S. Surman, and J. Jass, eds., Biofouling in Industry and 68. G. G. Geesey, in H.-C. Flemming and G. G. Geesey, eds.,
Process Engineering, vol. 1, John Wiley & Sons, Chichester, Biofouling and Biocorrosion in Industrial Water Systems,
U.K., 2000, pp. 175–203. Springer-Verlag, Berlin, Germany, 1991, pp. 155–164.
43. P. Gilbert, P. J. Collier, and M. R. W. Brown, Antimicrob. 69. M. W. LeChevallier, T. M. Babcock, and R. G. Lee, Appl.
Agents Chemother. 34, 1865–1868 (1990). Environ. Microbiol. 53, 2714–2724 (1987).
44. S. Notermans, J. A. M. Dormans, and G. C. Mead, Biofoul- 70. M. W. LeChevallier, C. D. Cawthon, and R. G. Lee, Appl.
ing 5, 21–36 (1991). Environ. Microbiol. 54, 649–654 (1988).
586 BIOFILMS IN THE FOOD INDUSTRY

71. Y. Nohata and H. Taguchi, Mater. Perform. 34(3), 43–46 97. E. Storgårds, Process Hygiene Control in Beer Production
(1995). and Dispensing, Otamedia Oy, VTT Publications 410, Espoo,
72. P. C. Miller and T. R. Bott, J. Chem. Technol. Biotechnol. Finland, 2000.
32, 538–546 (1982). 98. T. E. Cloete, F. Smith, and P. L. Steyn, Int. Biodeteriorat 25,
73. J. E. Duddridge et al., Engineering Foundation Conference 115–122 (1989).
on Fouling of Heat Exchanger Surfaces, United Engineering 99. R.-F. Wang, W.-W. Cao, and C. E. Cerniglia, J. Appl.
Trustees, New York, 1983, pp. 717–726. Microbiol. 83, 727–736 (1997).
74. J. W. Costerton and E. S. Lashen, The NACE Annual Con- 100. T. Tolker-Nielsen, K. Holmstrom, L. Boe, and S. Molin, Mol.
ference and Corrosion Show, NACE Publication Department, Microbiol. 27, 1099–1105 (1998).
Houston, Tex., 1983, Paper no. 246. 101. T. Tolker-Nielsen, K. Holmstrom, and S. Molin, Appl. Envi-
75. J. T. Holah and R. H. Thorpe, J. Appl. Bacteriol. 69, ron. Microbiol. 63, 4196–4203 (1997).
599–608 (1990). 102. M. W. Mittelman et al., J. Microbiol. Methods 18, 51–60
76. G. Bezanson, S. Burbridge, D. Haldane, and T. Marrie, Can. (1993).
J. Microbiol. 38, 328–330 (1992). 103. S. Duncan, L. A. Glover, K. Killham, and J. I. Prosser, Appl.
77. J. Rogers et al., Appl. Environ. Microbiol. 60, 1842–1851 Environ. Microbiol. 60, 1308–1316 (1994).
(1994). 104. A. J. Walker, J. T. Holah, S. P. Denyer and T. G. S. A. B.
78. J. Rogers et al., Appl. Environ. Microbiol. 60, 1585–1592 Stewart, Colloids Surf. A: Physicochem, Eng. Aspects 77,
(1994). 225–229 (1993).
79. W. G. Characklis, in W. G. Characklis and K. C. Marshall, 105. H. Anwar et al., Antimicrob. Agents Chemother. 33,
eds., Biofilms, John Wiley & Sons, New York, 1990, 1824–1826 (1989).
pp. 195–231. 106. H. Anwar, M. K. Dasgupta, and J. W. Costerton, Antimi-
80. P. Hugenholz, M. A. Cunningham, J. K. Hendrikz, and crob. Agents Chemother. 34, 2043–2046 (1990).
J. A. Fuerst, Lett. Appl. Microbiol. 21, 41–46 (1995). 107. J. F. Frank and R. A. Koffi, J. Food Prot. 53, 550–554 (1990).
81. B. Ljungqvist and B. Reinmüller, in T. Ohlsson, ed., 108. J. Gillatt, Int. Biodeteriorat. 27, 383–394 (1991).
Advances in Aseptic Processing and Packaging Technologies, 109. P. Maris and R. Fresnel, Sci. Aliments 12, 721–728 (1992).
Kompendiet, Göteborg, Sweden, 1995. 110. S. F. Bloomfield, J. Appl. Bacteriol. 76, 86–94 (1994).
82. M. Heinzel, in K. R. Payne, ed., Industrial Biocides, John 111. M. R. W. Brown and P. Gilbert, J. Appl. Bacteriol. 74,
Wiley & Sons, Chichester, U.K., 1988, pp. 52–67. 87S–97S–(1993).
83. P. van Beek, 12th International Symposium on Contamina- 112. R. L. Anderson et al., Am. J. Public Health 80, 17–21 (1990).
tion Control, The International Confederation of Contami-
113. D. G. Dunsmore, A. Twomey, W. G. Whittlestone, and
nation Control Societies, Tokyo, Japan, 1994, pp. 369–373.
H. W. Morgan, J. Food Prot. 44, 220–240 (1981).
84. M. Lehtimäki, 26th Nordic R3-Symposium, R3 Nordic,
114. G. de Goederen, N. J. Pritchard, and A. P. M. Hasting, in
Helsinki, Finland, 1995, pp. 170–175.
H. G. Kessler and D. B. Lund, eds., Fouling and Cleaning
85. P. Hugenholz and J. A. Fuerst, Appl. Environ. Microbiol. 58, in Food Processing, Druckerei Walch, Augsburg, Germany,
3914–3920 (1992). 1989, pp. 115–130.
86. D. M. Holt, in D. R. Houghton, R. N. Smith, and H. O. W. 115. S. J. Lewis and A. Gilmour, J. Appl. Bacteriol. 62, 327–333
Eggins, eds., Biodeterioration 7, Elsevier Publishers Ltd., (1987).
Essex, U.K., 1988, pp. 493–506. 116. J. A. Troller, Sanitation in Food Processing, Academic Press
87. M. C. Hughes, Appita J. 46, 194–197 (1993). Inc., San Diego, Calif., 1993, pp. 30–70, 263–286.
88. P. H. Sorrelle and W. E. Belgard, Pulp & Paper 66(5), 57–64 117. C. Johansen, P. Falholt, and L. Gram, Appl. Environ.
(1992). Microbiol. 63, 3724–3728 (1997).
89. M. C. Hughes-van Kregten, Appita J. 41, 470–474 (1988). 118. I. N. Izzat, E. O. Bennett, J. E. Gannon, and I. U.
90. O. M. Väisänen, E.-L. Nurmiaho-Lassila, S. A. Marmo, and Onyekwelu, Tribol. Int. 14, 171–173 (1981).
M. S. Salkinoja-Salonen, Appl. Environ. Microbiol. 60, 119. M. H. Turakhia and W. G. Characklis, Biotechnol. Bioeng.
641–653 (1994). 33, 406–414 (1989).
91. P. Väätäinen, AEL-Insko seminar Prossesivesijärjestelmät: 120. B. M. Milwidsky and D. M. Gabriel, Detergent Analysis,
prosessiveden käsittely ja kierrätys, AEL-Insko, Vantaa, George Godwin, London, U.K., 1982, pp. 85–104.
Finland, 1994 (in Finnish). 121. T. R. Bott, Food Cont. 2, 44–49, (1991).
92. J. Mentu, in G. Wirtanen, L. Raaska, M. Salkinoja-Salonen, 122. J. T. Holah, in L. F. Melo, T. R. Bott, M. Fletcher, and
and T. Mattila-Sandholm, eds., Future Prospects of Biofoul- B. Capdeville, eds., Biofilms — Science and Technology,
ing and Biocides, The Technical Research Centre of Finland, Kluwer Academic Publishers, Dordrecht, The Netherlands,
VTT Symposium 165, Espoo, Finland, 1996, p. 23. 1992, pp. 645–659.
93. M. Fletcher, Methods Microbiol. 22, 251–283 (1990). 123. P. Wood, M. Jones, M. Bhakoo, and P. Gilbert, Appl. Envi-
94. T. L. Ladd and J. W. Costerton, Methods Microbiol. 22, ron. Microbiol. 62, 2598–2602 (1996).
285–307 (1990). 124. W. A. Gould, Current Good Manufacturing Practices: Food
95. D. E. Caldwell, D. R. Korber, and J. R. Lawrence, J. Appl. Plant Sanitation, 2nd ed., CTI Publications, Baltimore, Md.,
Bacteriol. 74, 52S–66S–(1993). 1994, pp. 189–215.
96. S. Salo, E. Storgårds, and G. Wirtanen, in G. Wirtanen, 125. J. T. Holah, C. Higgs, S. Robinson, D. Worthington, and
S. Salo, and A. Mikkola, eds., 30th R3-Nordic Contami- H. Spenceley, Lett. Appl. Microbiol. 11, 255–259 (1990).
nation Control Symposium, Libella Painopalvelu Oy, VTT 126. C. D. Norton and M. W. LeChevallier, Am. Water Works
Symposium 193, Espoo, Finland, 1999, pp. 187–198. Assoc. 89(7), 66–77 (1997).
 BIOFILTRATION 587

127. C. A. C. Sequeira, P. M. N. A. Carrasquinho, and C. M. BIOFILTRATION

Cebola, Microbial Corrosion Conference, Elsevier Applied
Science, Essex, U.K., 1989, pp. 240–255. HINRICH BOHN
128. Shin-Ho-Lee and J. F. Frank, J. Food Prot. 54, 4–6 (1991). University of Arizona
129. A. Mustapha and M. B. Liewen, J. Food Prot. 52, 306–311 Tucson, Arizona
(1989).
Biofiltration is the removal and destruction of gaseous
130. B. E. Christensen, J. Biotechnol. 10, 181–202 (1989).
contaminants in air as the air flows through porous solids
131. M. Exner, G.-J. Tuschewitzki, and J. Scharnagel, Zbl. Bak- containing degradative microorganisms (1,2). In engineer-
teriol, Mikrobiol. Hygiene, Serie B 183, 549–563 (1987). ing nomenclature, it can accurately be called ambient
132. B. Pietersen, V. S. Brozel, and T. E. Cloete, Water SA 22, temperature catalytic oxidation. Microbial enzymatic cata-
239–244 (1996). lysis at ambient temperatures in biofilters achieves the
133. J. F. Kramer, Mater. Perform. 36(8), 42–50 (1997). same results as precious metal catalysts at high tem-
134. S. Nakayama, M. Tanaka, S. Yamauchi, and N. Tabata, peratures. Biofiltration can effectively and economically
Ozone Sci. Eng. 7, 31–46 (1985). destroy most gaseous air contaminants. A rule of thumb
135. S. H. Lin and K. L. Yeh, Chem. Eng. Technol. 16, 275–278
is that if a molecule can burn in air, it can degrade in a
(1993).
biofilter.
Figure 1 shows a schematic biofilter. Moist-contamina-
136. F. Bourion and O. Cerf, Sci. des aliments 16, 151–166
ted air flows slowly through a porous biofilter medium;
(1996).
microbes in the medium oxidize the contaminants to
137. J. W. Costerton, T. J. Marrie, and K.-J. Cheng, in carbon dioxide, water, and calcium salts. The biofilter
D. C. Savage and M. Fletcher, eds., Bacterial Adhesion, medium can be soil, compost mixed with bulking
Plenum Press, New York, 1985, pp. 3–43. agents, composted wood chips or wood bark, peat, or
138. A. Grasshoff, Trans, I. Chem. E. 70(C2) 69–77 (1992). synthetic media (activated carbon, porous ceramic, and
139. EHEDG, Trends Food Sci. Technol. 3, 325–328, (1992). plastic packing). Because the gaseous contaminants are
molecularly subdivided and intimately mixed with oxygen,
the oxidation rates are many-fold faster than the same
compounds in liquid or solid state.
BIOFILMS, METHODOLOGY. See ACTIVATED Biofiltration is the most common application of this air
SLUDGE — MOLECULAR TECHNIQUES FOR DETERMINING treatment process, which is also used by biotrickling filters
COMMUNITY COMPOSITION; LASER SCANNING MICROSCOPY IN and bioscrubbers. In biofiltration, the microbial medium
COMBINATION WITH FLUORESCENCE TECHNIQUES FOR BIOFILM and water are stationary, and air is the moving phase.
STUDY; MICROBIAL FLOCS SUSPENDED BIOFILMS In biotrickling filters, both air and water move over a
stationary microbial support. In bioscrubbers, microbial
slurry is sprayed into the air stream.
In contrast to other air pollution control techniques,
biofiltration creates or emits no secondary pollutants (NOx ,
CO, Cl2 , O3 , wastewater, or waste solids), consumes no fuel
BIOFILMS, MODELING OF. See MODELING OF or chemicals, and minimizes carbon dioxide loading of the
BIOFILMS atmosphere.
The microbial population adapts to the input gas
composition; the adaptation time depends on the gas
composition and, secondarily, to gas concentration (3).
For rapidly biodegradable gases (Table 1), the initial
destruction/removal efficiency (DRE) is 80 to 95% of the
final, steady state DRE. The final DRE is reached after
BIOFILMS, PATHOGENS IN. See PATHOGENS IN
one to three weeks as the microbial population adapts and
ENVIRONMENTAL BIOFILMS
the medium material settles. The DRE for more slowly
biodegradable gases tends to be low initially, and the

Clean air

BIOFILMS, SORPTION PROPERTIES OF.

See SORPTION PROPERTIES OF BIOFILMS
Biofilter
medium

Water
Contaminated
air
BIOFILMS: SUSPENDED BIOFILMS. See MICROBIAL
FLOCS SUSPENDED BIOFILMS Figure 1. A schematic biofilter.
588 BIOFILTRATION

Table 1. Destruction/Removal (DRE) Rates of Air Contam- Some biofilter designs emphasize high effectiveness
inant Gases by Biofiltration per unit volume or per unit weight of the biofilter,
Rapid: with penalties of higher initial cost, complexity, and
maintenance. Other designs emphasize low initial and
H2 S, HCl, SO2 , NOx , Cl2 , O3 , silanes, alcohols, aldehydes,
operating cost, ruggedness, long life, and low maintenance.
ketones, organic acids, ethers, esters, amines, mercaptans
Slow: (in approximate order of decreasing degradation rates): Their penalty is larger size and mass.
CO > terpenes > aromatic hydrocarbons > In addition to the design problems, biofiltration has the
aliphatic hydrocarbons > CH4 > monohalogenated inherent disadvantages:
and dihalogenated hydrocarbons
Unfeasibly Slow: 1. Biofilters require longer reaction time than other
Siloxanes, tri-halogenated and higher-halogenated hydrocarbons techniques. Reaction rates at room temperature
are slower than those at the temperatures of
catalytic and thermal oxidizers and those with
strong oxidants in chemical scrubbers.
adaptation period may be as long as one month. The DRE, 2. Oxidation rates are compound-specific. Microbes
thereafter, is not affected by downtimes up to a week or generally oxidize gases containing oxygen, nitrogen
two. Longer downtimes may require a short readaptation and sulfur functional groups faster than hydrocar-
period. bon gases. Highly halogenated gases (tri-, tetra-, and
The primary function of biofiltration is to destroy gases, higher) do not oxidize in biofilters. Monohalogenated
but biofilters also filter out particles from air streams, even and dihalogenated compounds oxidize slowly.
less than 2.5 µm particles. Particulates generally should
be avoided in the input air, however, because they can The air residence time in biofilters is seconds to minutes,
plug the small pores of the biofilter. Mists are acceptable depending on the biodegradability of the gas, the kind
if the droplets are water or biodegradable liquids. of medium, and the desired DRE. The residence time is
Although similar in appearance to sorption beds, usually expressed as the empty bed residence time (EBRT).
biofilters function by oxidizing rather than by just The residence time in high-temperature oxidizers and
adsorbing airborne contaminants. Biofilter media have adsorbent (activated carbon and zeolite) beds is fractions
little adsorption capacity for gases. The lack of any of seconds. Biofilters are, therefore, larger in size and area
sorption capacity is an advantage in that, the media do not than other air pollution control devices.
accumulate air contaminants in significant amounts and
any adsorbate is quickly oxidized. Spent biofilter media is Applications
not a hazardous waste. The range of applications for biofilter treatment of contam-
The potential benefits of biofiltration over other air inated air is potentially very broad (1,2,4). Biofiltration is
pollution control techniques are most advantageous when the contaminant concentrations
are low (<1,000 ppmv), at air temperatures of 10 to 55 ° C,
1. low to moderate capital costs, the air is humid and particulate-free, the contaminants are
2. low operating cost — no supplemental fuel or energy rapidly biodegradable, space constraints are absent, and
is needed, and the oxidant is atmospheric oxygen recovering the contaminant is not economic, and when
so no chemical oxidant (chlorine, ozone, and low maintenance is needed. The input air can often be
permanganate) is needed, made humid and particulate-free and brought to optimum
3. no secondary pollutants (CO, NOx , Cl2 , O3 , dioxins, temperature simply by water injection into the air stream.
wastewater, and solid wastes) are created or Biofilters are also advantageous where low maintenance,
released, inherent safety, low visibility, and a green technology are
important.
4. the catalysts (microbial enzymes) are free and
Wastewater treatment, food processing, and pharma-
continuously regenerate themselves,
ceutical, and some chemical manufactures have shown
5. low maintenance, and interest in biofiltration to date. This is largely because
6. inherent safety (low temperature and no com- large airflow rates must be treated and the pollutant
bustibles nor hazardous chemicals). concentrations are low, and because their personnel rec-
ognize that microbes can influence the chemistry of their
In some cases, poor design and too high expectations have products.
led to a lower than desired degree of pollution control, In other industries, the responsibility for air pollution
higher than expected maintenance requirements, and too control is often given to mechanical and chemical engineers
frequent changeout of the biofilter medium. Biofilters, who may be unaware of the possibilities of microbial
unfortunately, almost invite naive design because they degradation. The trend toward size minimization also
are mechanically simple and because the air distribution works to the disadvantage of biofiltration.
system is not difficult. The difficulties of (1) maintaining Potential industrial applications for biofiltration
an active microbial population, (2) maintaining low back- include treatment of ventilation air, process gases, stor-
pressure and uniform air distribution over the long age tank vents, petroleum refining and handling, printing,
term, and (3) moisture management are not apparent in solvents of all kinds, and relief valves. Because no start-
Figure 1. up procedures are necessary and flow capacity is large, a
 BIOFILTRATION 589

biofilter provides simple and useful treatment for storage available space to take advantage of biofiltration. In
tank and relief valve emissions. The Bhopal, India, acci- addition, biofiltration may require more ducting because
dent reminds us that even simple treatment of accidental the biofilter must be farther from the source. Biofiltration
releases or safe containment in a biofilter is better than could find more applications if the constraints of biofilter
direct release into the atmosphere. weight and size were included in the initial plant design.
Because of the low cost of biofiltration and the large Biofilter design has shown little imagination in adapting
number of treatable pollutant gases, and because the to the available space and configuration. Parking lot areas
moisture in the air is a benefit rather than a detriment, and rooftops are open invitations to innovative biofilters.
the extent of potential industrial applications is as wide or In compiling Figure 2, the authors avoided some site-
wider than the physicochemical alternatives — thermal specific cost problems by using data for modular contained
oxidation, chemical scrubbing, or activated carbon or biofilters. Built-in-place, open biofilters tend to be less
zeolite adsorption. expensive.
In an ideal world, air pollutants would be recovered
for reuse rather than emitted. Recovery by adsorption or
Compound Specificity
condensation is expensive, energy-intensive, incomplete,
and sometimes hazardous. Hence, release is inevitable. The oxidation rate varies for each gas. Table 1 shows
Released gases are destroyed by oxidation. The choice common gases grouped according to their relative
of oxidative treatment — thermal, chemical, or biofiltra- biodegradation rate. If the molecule contains oxygen,
tion — is determined by economical and environmental nitrogen, or sulfur functional groups, it generally will
constraints. biodegrade rapidly. A corollary is that if a molecule is
water-soluble, it will biodegrade rapidly. Oxygen, nitrogen,
Economics and sulfur functional groups tend to increase both the
Figure 2 shows the applicability, based primarily on water solubility and the biodegradation rate of organic
cost, of various air pollution control technologies. Over gases.
a wide range of airflow rates and pollutant concentra- Chlorine and bromine functional groups in organic
tions — indeed the range of most air pollutant emis- molecules reduce biodegradation rates. Monohalogenated
sions — biofiltration has a strong economic advantage and dihalogenated molecules oxidize slowly under the
(the conversion factors are 1 cfm = 1.7 m3 and mg m−3 = highly aerobic conditions of biofilters. Biofilter control of
ppmv × molecular weight/24.06). The cost estimates are these gases may be feasible under some conditions. More
probably for the rapidly biodegradable gases as shown in highly halogenated compounds need anaerobic conditions
Table 1. For hydrocarbons and other slowly degradable for the first stage of their biodegradation, reduction by
gases, the economic advantage of biofiltration is slightly substituting hydrogen atoms for halogens in the molecule.
less because the longer residence time requires larger The reaction products are less halogenated and may be
biofilters. oxidizable in a biofilter. The reduction step is very slow,
Of the air pollution technologies, the installed cost of and a successful two-stage anaerobic–aerobic biofilter has
biofilters may be the most site-specific because of their yet to be devised.
size and weight. Urban locations may have too little Inorganic gases such as H2 S, NH3 , SO2 , and HCl react
rapidly in biofilters, whereas NOx (NO and NO2 –N2 O4 )
reacts more slowly. Nitrous oxide (N2 O) reacts even more
Gasflow (m3 h−1) slowly. The first step, the removal of the compound from
the gas phase, is often chemical rather than biological
1,000,000
and is very fast. The sulfur and the nitrogen gases are
later oxidized microbially to their oxyacids, H2 SO4 and
100,000 HNO3 . The biofilter must neutralize this acidity or the
Biofiltration microbes cease to function. The lower limit of acidity is
Bioscrubbing Incineration
Trickling filters conservatively about pH 2.5 for the sulfur gases and pH 4
10,000
for the nitrogen gases.
Regenerative
Table 1 gives no absolute numbers for oxidation rates
Scrubbing Condensation
1,000 adsorption or EBRT because the oxidation rates vary with the
medium, temperature, water content; in some cases with
100 Cryocondensation nutrient conditions; and with the degree of DRE required.
Nonregenerative Laboratory measurements of degradation rates tend to be
adsorption optimistic when compared with the field conditions, but
0 the relative degradation rates obtained in the laboratory
1 10 100
Concentration (g m−3) apply.
Figure 2. Applicability of various air pollution control technolo-
gies based on airflow rates and concentrations to be treated. (From Mechanism
Kosteltz, A. M. et al., in Proceedings of the 89th Annual Meeting
and Exhibition of the Air and Waste Management Association, Under most conditions, where the contaminant gas
Pittsburgh, PA, 1996. With permission of KPMG Management concentration is less than 1,000 ppmv (ppmv is the volume
Consultants, Ottawa, Ontario, Canada.) fraction or mole fraction of the gas × 106 ), the kinetics of
590 BIOFILTRATION

biofilter oxidation are approximately first-order (5): The chromatographic partitioning of gases has not been
measured in organic, compost-based biofilter media. The
−dA/dt = kCA (1) greater organic matter content of these media might
suggest greater chromatographic efficiency and greater
retention of organic gases than in inorganic media.
where A is a pollutant gas, k is a reaction coefficient related
Because operating biofilters are very moist, however,
to the biodegradation rate of A, and CA is its concentration.
the gases actually contact the moisture film on the
The oxidation rate increases with gas concentration, and
particle surfaces rather than on the surfaces themselves.
the fraction of pollutant removed and destroyed increases
The water film greatly decreases the chromatographic
logarithmically with reaction time t.
differences between inorganic-based and organic-based
The DRE is
biofilter media.
Cinput − Coutput The sorption capacity of activated carbon mixed
DRE(%) = × 100 (2) into biofilter media slightly increases chromatographic
Cinput
efficiency and buffers the effects of rapid changes of gas
concentration in the input air (1). The benefit of activated
Under first-order conditions, the DRE is independent of carbon is lessened by its being saturated with water,
gas concentration, depends on reaction time (EBRT), and, which decreases its gas sorptivity. Microbes on the pore
therefore, depends on biofilter volume. walls increase gas partitioning by degrading the gases and
As gas concentrations increase above 1,000 ppmv, the thereby reducing gas concentration in the water film.
reaction order gradually decreases to zero order: The second and limiting step of contaminant removal
and destruction is the rates of microbial oxidation, of
−dA/dt = k (3) chemical reactions, or of the gases on the pore walls.
These rates are compound-specific (Table 1).
Then, the oxidation rate is constant and independent of Biofilter modeling of degradation rates has emphasized
gas concentration. Hence, the DRE decreases as the gas the first step: dissolution of the gases into the water
concentration increases at these high concentrations. film and transport across that film. The shortcoming of
The first step of biofilter removal of air contaminants this approach is shown by biofilter response to moisture
is chromatographic. The pore walls of the media are content. Biofilter performance does indeed decrease when
in equilibrium with clean air. The major components the moisture content increases beyond the optimum
of air — dinitrogen, dioxygen, argon, and carbon diox- range. Biofilter performance, however, also decreases
ide — have already saturated the pore walls and, therefore, sharply when the water content is below this range (7).
flow unimpeded through the biofilter. The contaminant Although mass transfer across a thin water film is much
gases, on the other hand, partition out on the moist pore faster, the important step is the microbial metabolic rate
walls in rough accordance with Henry’s law and lag behind which has a maximum over a narrow range of moisture
the major components of the air. The partitioning of a gas content.
from air to the stationary phase increases with its polarity, The water films in biofilters are very thin. In an
water solubility, and molecular weight. The more tightly active biofilter, the water potential or hydraulic head
the gas is adsorbed, the longer the gas remains on the pore is about −0.2 bar or −20 kPa (equivalent to a water
walls, the slower it flows through the biofilter, and the activity of about 0.99 and equivalent to about 50–60%
more likely it will be absorbed by a microbe and degraded wet weight of water in compost and other organic media).
or oxidized. At this water potential, pores with a diameter larger
The chromatographic efficiency of biofilter media has than about 10-µm in diameter are empty of water except
not been measured under moist operating, conditions. In for a water film <0.1-µm thick. ‘‘Pores’’ in biofilters are
an air-dry sandy soil with low organic matter content, actually the network of interconnected cavities between
the retention volume (an index ratio of chromatographic the particles of the medium. The pores, which transmit
efficiency and the reciprocal of the ratio of flow rate the gas by convective flow, are tortuous and have nominal
of contaminant gas relative to flow rate of air) ranged pore ‘‘diameters’’ ranging from perhaps 50 µm to several
from one for methane to 105 for octane (6). The retention millimeters. Smaller pores may be dead ends or may
volume increased with increasing clay content in the soil. have severely restricted openings so that diffusion is an
The retention volumes of the more polar gases methanol, important gas transfer mode in them.
dimethylketone, and diethyl ether were >106 , too large to Where the medium’s particles contact each other at
measure under the same experimental conditions. a water potential of −20 kPa, the water meniscus has a
The dryness was necessary to prevent appreciable diameter of 10 µm. Between the contact points and on the
biodegradation of the gases and to prevent large-scale surfaces of microbes, the water film is less than 0.1-µm
changes of moisture content during the measurements. thick. Because bacteria are on the order of 1 to 10 µm
Biofilters, however, operate under moist conditions in in diameter, this thin water film is a minor factor in
which the water film on soil particles and microbes gas uptake. The major function of water is to support
probably inhibits the partitioning of water-insoluble gases microbial metabolism rather than to determine microbe-
such as hydrocarbons onto the stationary phase and gas contact.
increases their relative flow rates. The presence of water Gas adsorption and absorption are minor components of
increases the partitioning of water-soluble gases. biofiltration. Like most solids, biofilter media adsorb only
 BIOFILTRATION 591

minute amounts of gases. Porous solids can retain large Biofilter Media
amounts of liquids by capillarity, but this mechanism
is absent in biofiltration. At biofilter temperatures and The essential criteria for biofilter media are that they
conditions, the gas concentrations are far below those that support microbial growth and permit gas flow. Microbes
allow gases to condense to liquids. can grow on many materials and some materials, such as
Activated carbon is unique as a solid because its high soil, compost (alone and in mixtures), and peat, are natural
surface area and nonpolar character help adsorb organic supports for microbial growth. Other porous, synthetic
gases. Although compost and its mixtures also are organic, media, including plastic packing, activated carbon, and
composts have a much lower sorption capacity than carbon porous ceramic, can be seeded to provide the initial
because of higher polarity related to cellulose content microbial population.
and because of lower surface area. Silicates in soils are Since many porous materials can serve in biofilters,
quite polar, a disadvantage for adsorbing nonpolar organic the choice hinges on other features of the media,
gases but an advantage for adsorbing inorganic gases. The including cost, stability, hydrophobicity or hydrophilicity,
organic coating on inorganic soil particles gives soils an density, nutrient supply, pH buffering, and maintenance
appreciable nonpolar character. requirements. The cost differences are large. Synthetic
Overloading of biofilters is evidenced by only partial media cost dollars per kilogram, soil mixtures cost up
oxidation of the input gases. Alcohols, for example, to $50/m3 , and compost mixtures cost from $50 to
partially oxidize to aldehydes and organic acids rather $500/m3 .
than to carbon dioxide. A second indication of overloading The stability of the medium affects the long-term flow
is the buildup of microbial biomass that leads to plugging uniformity through the biofilter bed and the frequency
of the biofilter pores. This problem has been seen in of bed replacement (changeout). Organic media are
laboratory experiments at gas input concentrations of inherently unstable, heterotrophs feed on the media
more than 5,000 ppmv and more than 5 g/m3 . and on the input gases. As the medium is consumed,
it shrinks, forms cracks, and pulls away from its
container walls. In addition, the media particles become
Toxicity and Antagonism
smaller and more mobile. Water flow tends to deposit
Chemical toxicity to biofilter microbes is extremely them in the pores at the interface of the medium
unlikely. Microbes are so much more tolerant of biological and the distribution layer so that this layer becomes
toxins than higher organisms are that complete inhibition impermeable.
of microorganisms does not occur, although pulses of Under careful management and in enclosures, some
high concentrations of some pesticides might temporarily compost mixtures have a 10-year lifetime. Open compost
reduce degradation rates. Gas concentrations at even beds can start developing cracks within several months;
10,000 ppmv are low in terms of mass per unit volume of two years might be a maximum before backpressure builds
biofilter media. Adding large amounts of toxic agents via up and the DRE is no longer satisfactory.
the gas phase, therefore, is physically difficult. Biofilter Inorganic media, such as siliceous soil, plastics, and
slowdown usually is because of high or low moisture ceramics, are not attacked by microbes, and impermeable
content or low temperature, rather than because of layers are less likely to occur. Soils and distribution
chemical toxicity. Herbicides for weed control on the layers rich in CaCO3 , however, can clog severely within
surface of open biofilters are in too low a concentration months.
to affect biofilter performance. Soil and other inorganic media are permanent, that is,
Antagonism, a decrease of gas removal or destruction no replacement is required and DRE and backpressure
related to the presence of other gases, is rarely seen in are stable. Native soils, however, usually are far too
biofilters. Gas mixtures even at more than 1,000 ppmv impermeable for industrial-scale biofiltration, and airflow
concentrations show little evidence of antagonism; the is spatially nonuniform. An appropriate soil mixture is
presence of one gas does not inhibit the oxidation a mixture of several components to provide an acceptable
rate of other gases. Instead, the microbes degrade the trade-off between DRE, air permeability, and uniform flow.
more rapidly biodegradable gases first. More slowly Hydrophobicity and/or hydrophilicity affect mois-
biodegradable components are oxidized further along in ture management in biofilters. The most common
the biofilter column. cause of biofilter failure is inadvertent drying of the
Conditions at higher concentrations are more compli- medium. Organic substances tend to be hydropho-
cated, but antagonism is still small at 5,000 ppmv for bic when dry; inorganic substances tend to be
some gas mixtures. These high concentrations are in the hydrophilic. Rewetting a dry organic-based medium
range where oxidation by synthetic catalysts may be more means taking out the medium and physically mix-
economical. ing in the water, which is a tedious, dirty, time-
Some process gases may be low in oxygen so that, as consuming, and often impossible task. Simply pour-
in other oxidation techniques, dilution air is necessary. ing water on the surface is ineffective because the
Microbes are efficient in their oxygen utilization, so only water flows down through cracks and along wall bound-
a little more than the stoichiometric amount of oxygen aries rather than thoroughly and uniformly wetting the
in air need be added. If the exhaust air contains some medium.
oxygen, the microbes in the biofilter are not oxygen- Hydrophilic substances, on the other hand, rewet
starved. easily and water flows to all parts of the biofilter bed.
592 BIOFILTRATION

Simple surface irrigation restores biofilter activity. Soils, Availability of water can be expressed either as
ceramics, and pearlite materials are hydrophilic. chemical potential (also called chemical activity, which
Synthetic media provide little or no mineral nutrients is numerically equal to the relative humidity (RH) of the
(nitrogen, phosphate, calcium, and so on) and hence air at equilibrium) or as physical potential (also called
require periodic supplementation. Microbes attach weakly hydraulic head, water tension, or suction) (8). The 50 to
to plastic media. Compost and the synthetic media are 60% wet weight moisture content in organic biofilters
supplemented with a CaCO3 source such as crushed is a water potential or water activity around 0.99, with
seashells to maintain acidity at pH 6 to 8. Soil media have 0.95 being a lower limit for good performance. The same
greater pH buffering capacity, but may require additional range of water availability expressed as physical potential
lime. is about −0.2 bars, with perhaps −3 bars as a lower
Media density is important when lightweight is limit. The equivalent SI units are −20 and −300 kPa.
important, and when biofilters are portable or placed Measurement of water availability is, however, difficult
on structures. Organic media have densities of about and slow. Instruments that measure moisture content or
400 kg/m3 when wet and soil and ceramic media have availability that can be linked to irrigation devices are
densities of about 1,400 kg/m3 . being developed.
A lower maintenance requirement generally means a Because drying out is the most common cause of
more successful and effective biofilter, and lower operating biofilter breakdown and because the effluent air is at
and maintenance (O/M) costs. Included in O/M costs are 99 to 100% RH, biofilters should ensure that the input
bed replacement, backpressure, moisture monitoring, and air is at 100% RH. Fogger nozzles in the input air
in some biofilters the cost of preheating the air. Since stream are common, indeed almost a necessity. Oxidation
replacement is a high and recurring cost, good biofilter of the contaminants raises the biofilter temperature
designs make replacement of nonpermanent media as slightly. This decreases the RH of the air and increases
easy as possible. Modular biofilters require replacement of evaporation in the medium. Surface irrigation is needed to
the entire unit. Backpressure is lower initially in compost overcome this internal evaporation and evaporation from
biofilters than in soil biofilters but increases with time. the surface. This irrigation usually is triggered by a timer
The backpressure remains stable in inorganic and plastic rather than by need.
media. The initial backpressure of a large biofilter bed Because water availability is in effect an index of the
usually is 2 to 8-cm water head or 200 to 800 Pa. Small thickness of the water film on the medium’s particles,
modular biofilters may have backpressures of 40 cm-water water availability can be estimated visually. At water
head, or 4,500 Pa. availability at the optimium moisture level, expressed as
Synthetic media are more homogeneous (thereby 0.99 activity, −0.2 bar, −20 kPa water potential, or 50 to
providing more uniform distribution of airflow through 60% wet weight in organic media, the medium appears
the medium), their DRE per unit mass is greater than dark and moist. A visible sheen indicates that the water
that of natural media, and their density is low. As film is too thick and the moisture content is too high. If the
yet, these benefits rarely outweigh their much higher medium is light-colored and darkens when water is added,
cost. the medium is too dry. Also, the water film at optimum
moisture availability has sufficient surface tension to hold
Moisture the medium’s particles in a cast when the particles are
squeezed lightly in the hand. These moisture properties
Biofiltration is a microbial process, and the microbes func- should all be tested at greater than and equal to 15 cm
tion best over a rather narrow range of moisture content. below the surface of open biofilters. Moisture content at the
Moisture content is usually expressed as percentage of wet surface reflects recent climate effects rather than biofilter
weight of the bed medium. For organically based media, moisture content.
the range is 50 to 60%. At greater moisture contents, pores Irrigation and rainfall are sometimes in excess of
become saturated with water, whereas at lower moisture biofilter needs. Drainage must be provided to prevent
contents, microbial activity decreases. At less than 40%, water accumulation. Irrigation should be scheduled
the degradative capacity is severely limited (7). Rewetting so that the amount of drainage water is kept to
the bed to 50 to 60% may not restore performance. The lack a minimum. The water phase may contain soluble
of recovery probably is due to the difficulty of thoroughly gases, but these concentrations are low. Drainage water
rewetting hydrophobic media. from compost biofilters contains brown-colored organic
Wet weight is convenient for measurement, but it compounds similar to natural humic acids, and the
depends on the density of the medium and the particle water may be cloudy. The water is nonhazardous but
size distribution. For organic media, density and particle unsightly. Drainage water from inorganic media is clear
size distribution are similar enough that 50 to 60% suffices. and colorless. The drainage water can be recycled to the
The range for inorganic media (soil or ceramic), however, biofilter, or the drainage water can flow slowly into the
is 10 to 25%, depending on particle size. aerobic soil beneath the bed, where natural processes treat
Worse yet, mass units such as wet weight do not express the water further. The risk of groundwater contamination
the actual criterion — moisture availability — that affects is exceedingly small.
microbial activity. Much of the water in biofilter media is Impermeable liners under biofilter beds can complicate
too difficult for microbes to extract without slowing their water management by inducing water buildup. Liners
metabolic (degradation) rate. can be an additional liability because, despite careful
 BIOFILTRATION AND BIOODORS 593

installation, they leak sooner or later, and all the drainage BIOFILTRATION AND BIOODORS
is concentrated at the leaks rather than spreading out the
water for more effective treatment. Concentrating the flow JOSEPH S. DEVINNY
makes natural water treatment much more difficult. University of Southern
California
Temperature Los Angeles, California
The temperature operating range for biofilters is approx- DEREK E. CHITWOOD
imately 10 to 55 ° C. The optimum is generally considered University of California
37 ° C, but some biofilters successfully utilize thermophilic Riverside, California
bacteria at around 45 ° C. Below 10 ° C biodegradation still
occurs at slower rates; biofilters can function successfully With continuing population growth, it has become ever
at very close to freezing temperatures. Input gases beyond more common that residential developments are situated
the range of 10 to 55 ° C can be warmed by steam injection near industrial and utility facilities. At the same time, the
or can be heated and cooled by water injection or mechan- public has become less tolerant of foul odors and is more
ical cooling. Water and steam injection have the benefit of likely to demand that release of odorous vapors into their
maintaining high RH. Water injection has the additional neighborhoods be prevented.
benefit of washing out airborne particulates that otherwise Biological processes are commonly used for treatment
could plug biofilter pores. of odorous wastes such as wastewater, sludge, compost,
The oxidation rates are less temperature-dependent
or garbage. Treatment systems may become sources of
than the typical Q10 = 2 (the reaction rate doubles with
odors because vapors escape in their original form from
each 10 ° C increase of temperature) of microbial reactions
the waste or because odorous compounds are created in
would indicate. At low temperatures, gas partitioning
the collection or treatment system.
improves because gases are less volatile and are more
Odors typically consist of low concentrations of
soluble in the water film. Heterotrophs are adapted to
contaminants in large volumes of air. This type of
function in natural soils in which the average annual
discharge is expensive to treat by thermal methods
temperature is 0 to 15 ° C, except in very warm climates
because of the large amount of fuel needed to heat the
where soil temperatures may reach 25 ° C.
air. Adsorption often is also inefficient because the low
concentrations mean only small amounts of material are
CONCLUSION adsorbed on a given mass of adsorbent. Biofiltration, in
contrast, is well suited for treating odors.
Biofiltration is an effective means of removing and destroy- There is a substantial body of literature on the genera-
ing air contaminants from many waste airstreams. Biofil- tion and control of bioodors in wastewater and composting
tration is a low-cost technique over most of the range of systems (1–4), and several summary treatments of sulfide
industrial emissions because it requires no fuel or chemical generation in wastewater systems (5–7). There is a mod-
and relatively little maintenance. In addition, biofiltration est but rapidly growing body of literature on biofiltration
creates little or no secondary emissions or wastes, is inher- for air contaminants (8) and its application to wastewater
ently safe, and is environment friendly. Limitations of odors (9,10).
biofiltration are its compound specificity and relatively low
oxidation rates, requiring relatively large reactor volumes.
Good biofilter designs maintain high microbial activity
ODORS IN BIOLOGICAL TREATMENT SYSTEMS
over a wide range of moisture conditions and utilize the
most appropriate biofilter medium.
Biological treatment systems are designed to transform
waste materials to benign products. Organic compounds
BIBLIOGRAPHY
are converted to carbon dioxide and water, sulfides are
1. J. S. Devinny, M. A. Deshusses, and T. S. Webster, Biofiltra- oxidized to sulfate, amines are oxidized to nitrates,
tion for Air Pollution Control, Lewis, Boca Raton, Fla., 1999. wastewater organics are converted to stabilized sludge,
2. G. J. Skladany et al., in Handbook of Odor and VOC Control, and yard waste becomes useful compost (11). However, the
McGraw-Hill, New York, 1998. variable nature of the waste, the complexity of the waste
3. M. A. Deshusses, G. Hamer, and T. Dunn, Biotechnol. Bioeng. collection and treatment systems, and the great diversity
49, 587–598 (1996). of microorganisms lead to the creation of transformation
4. C. van Lith, G. Leson, R. Michelson, and F. E. Reynolds, eds., products that are offensive odorous compounds.
Proceedings of the 1996 Conference on Biofiltration, The It is useful to classify odors as those that are
Reynolds Group, Tustin, Calif., 1996, pp. 102–110. released from aerobic environments and those that come
5. S. P. P. Ottengraf et al., Bioprocess Eng. 1, 67–76 (1986). from anaerobic environments. The odors from aerobic
6. H. L. Bohn, G. K. Prososki, and J. G. Eckhardt, J. Environ. environments generally are less offensive, and often are
Qual. 9, 563–565 (1980). described as ‘‘musty.’’ Odors from anaerobic environments
7. R. Auria, A.-C. Aycaguer, and J. S. Devinny, J. Waste Mgmt. are often extremely offensive. Humans undoubtedly
Assn. 49, 65–70 (1998). evolved to perceive odors from anaerobic biological systems
8. H. L. Bohn and K. H. Bohn, Environ. Prog. 18(3), 156–161 as disgusting because it causes them to avoid spoiled food
(1999). and human waste that may be sources of diseases. The
594 BIOFILTRATION AND BIOODORS

Table 1. Odor Compounds in Wastewater Treatment as the electron acceptor and is added to the reactants.
Hydrocarbons are converted to aldehydes, alcohols, or
Reduced Sulfur Reduced Nitrogen
Compounds Compounds Others
carboxylic acids. These compounds are further oxidized
to carbon dioxide and water that are compounds with
Hydrogen sulfide Ammonia Acetaldehyde lowest energy under aerobic conditions. Although there
Allyl mercaptan Dibutyl amine Chlorine is a multitude of metabolic pathways, depending on the
Amyl mercaptan Diisopropyl amine Ozone starting compound and the microorganism, the ultimate
Benzyl mercaptan Dimethyl amine Sulfur dioxide products usually are carbon dioxide and water, which are
Dimethyl sulfide Ethyl amine odorless.
Diphenyl sulfide Indole Microorganisms that oxidize organic compounds tend to
Ethyl mercaptan Methyl amine
dominate aerobic ecosystems because the energy available
Methyl mercaptan n-Butyl amine
per mole of material processed is far greater than that
Phenyl mercaptan Pyridine
Propyl mercaptan Skatole for reactions possible without oxygen. However, other
Thiocresol Trimethyl amine reactions are possible. Organic nitrogen or nitrite can
be oxidized to nitrate by the nitrifying bacteria. Organic
Source: Adapted from P. L. Schafer, Chair, Joint Task Force of the Water sulfide and hydrogen sulfide are oxidized by bacteria such
Environment Federation and the American Society of Civil Engineers, Odor
as those in the genus Thiobacillus. Nitrate and sulfate
Control in Wastewater Treatment Plants (ASCE Manuals and Reports on
Engineering Practice No. 82 and Water Environment Federation Manual have essentially zero vapor pressures, and so do not
of Practice No. 22), American Society of Civil Engineers, New York, 1995. contribute to odor problems.
Ultimately, oxidative processes tend to convert odorous
compounds to end products that are inoffensive (Fig. 1).
dominant compounds in waste treatment odor problems Although some intermediate compounds such as short-
are sulfides and amines (Table 1). chain fatty acids and amines may be troublesome,
The distinction between aerobic environments and they tend to be rapidly biodegraded, so the general
anaerobic environments is also useful because it is a tendency is for wastewater, compost, or biosolids to
key to some control methods. An adequate supply of become less odorous as aerobic treatment proceeds.
oxygen can stop generation of the odors that arise from Primary sedimentation basins and fresh solid waste must
anaerobic systems. On the other hand, conditions that be covered, but secondary aeration basins and treated
promote vigorous transfer of oxygen from the atmosphere compost are far less offensive.
to the microbial ecosystem also promote release of volatile Oxidative processes will dominate as long as oxygen
compounds from the liquid. In some cases, it is best is plentiful. If the amount of oxygen-consuming material
to suppress gas transfer until treatment of the odor is small or if there is a mechanism by which oxygen
compounds is complete. is continuously supplied, the organic matter will be
degraded and the sulfides and amines will be oxidized.
Disposal of Odorous Compounds to Treatment Systems However, oxygen is not highly soluble in water. Water
in equilibrium with the atmosphere carries perhaps 7 to
Some of the compounds that cause odor problems in 9 mg/L of dissolved oxygen, depending on its temperature.
wastewater or composting systems are those that have A liter of wastewater typically contains material that
been disposed of to the system. Human waste, of course, will consume over 100 mg of oxygen, so the oxygen will
has odors at the time it is discharged, and manures smell soon be depleted in wastewater that is not vigorously
bad when they are added to compost. Industrial wastes
may also include odorous compounds, particularly volatile
solvents.
Odors
Biological Odor Generation
Microorganisms transform chemicals for two purposes:
to synthesize compounds that they use to make growing Aldehydes CO2
alcohols, amines H2O
cells and to provide energy. The reactions that produce Odors NO3− Odorless
fatty acids
energy generally process far more material than the SO4− release
synthetic reactions do. Microorganisms need the energy
Waste: Aerobic Aerobic
to move and drive their metabolism and often to drive hydrocarbons, conditions
synthetic reactions. Because larger amounts of products treatment
carbonhydrates,
are generated, catabolic reactions are more likely to proteins, fats, Anaerobic
produce odor problems. Further, the products of synthetic microorganisms conditions
reactions are rarely gases, and only gases can become
Unaltered waste Odors
odors.
constituents, fatty acids, mercaptans
By definition, energy production requires reactions that
amines, H2S, NH3
are thermodynamically favored under the conditions the
cell is experiencing: the products must contain less energy
than the reactants. Under oxidizing conditions, the usable Figure 1. Pathways for odor production and release in biological
reactions are usually redox reactions in which oxygen acts treatment systems.
 BIOFILTRATION AND BIOODORS 595

aerated. When this occurs, microorganisms that are the water, promoting volatilization. For sulfide
capable of using other compounds as electron acceptors and many other compounds in wastewater flow,
will proliferate. Denitrifying organisms will use nitrate as concentrations in the air above the water typically
the electron acceptor to produce nitrogen gas or nitrous are far below equilibrium. The mixing constantly
oxide. As nitrate becomes less plentiful, bacteria such as brings high concentrations of odor compounds near
Desulfovibrio will oxidize organic matter using sulfate as the surface where volatilization can occur.
the electron acceptor, thereby producing sulfides. 3. Rates of ventilation of the headspace above the
Hydrogen sulfide (H2 S) is produced in wastewater water. If the wastewater is within a pipe or some
collection systems by anaerobic sulfate-reducing bacteria. other enclosed structure, the rate of release of odor
Organic material contains approximately 1% sulfur (dry from the structure will depend on the rate at which
weight, 12). Sulfur as sulfate in wastewater ranges in the air in the structure is discharged. In some cases,
concentrations from a few parts per million (ppm) to it is possible to minimize this discharge by making
hundreds of ppm, depending on the source (5). Sulfate- the structures as airtight as possible. However, in
reducing bacteria can grow only in the absence of oxygen, any space that will be entered by workers, 20 to
when the conversion is thermodynamically favored. The 30 changes per hour are required for safety (5).
reduction rate of sulfate to sulfide depends on the
wastewater BOD or COD, sulfate concentration, dissolved Although Quigley and Corsi (12) described release of
oxygen concentration, and a number of other growth volatile organic pollutants that had been dumped in
factors (5). sewers, the same principles apply to any volatile compound
Without the energy benefits of oxygen as an elec- in a wastewater handling structure. They emphasized that
tron acceptor, degradation of waste organic matter by not all these factors may be present at the point where
microorganisms is much slower and less complete. Under the compounds are discharged or created, and releases are
anaerobic conditions, odorous organic compounds that are highly variable along the path from wastewater discharge
present in the original waste are less likely to be degraded. through treatment to treated effluent discharge. It is quite
Odorous intermediates such as fatty acids, mercaptans, common that odorous compounds generated in one part of
and amines will accumulate. the collection or treatment system can be released from
another part of the system.
Transfer of Odor Compounds to Air Van Durme (1) made the same point with respect to
The general strategies for control of biologically generated odor releases at various points within a treatment plant,
odors follow from Figure 1. The objective is to convert emphasizing that a compound may be produced in one part
the waste to odorless carbon dioxide, water, nitrate, of the plant but dominantly released at some downstream
and sulfate, without releasing odorous intermediate point where there is turbulence and exposure to the
compounds along the way. The best approach is to keep atmosphere. She noted, for example, that wastewater
the waste aerobic and contained until it is converted to flowing over weirs is commonly a major source of odor.
the odorless product. However, in complex collection and
treatment systems, anaerobic zones will develop where
ODOR TREATMENT BY BIOFILTRATION
force mains are required, where flow is unexpectedly slow,
or where sediment or slime layers build up (2,5). Garbage
Principles of Biofiltration
left too long in the truck or compost at the bottom of
a storage pile may consume all of a limited supply of Biofiltration removes air pollutants by passing the
oxygen. Some treatment processes intentionally are kept contaminated air through a damp, porous medium on
anaerobic. If anaerobic conditions do occur, the waste flow which a mixed culture of microorganisms is immobilized.
should be contained until aerobic treatment can complete The contaminants are absorbed from the air phase into
the transformation to oxidized products. For air, this may the water phase, which includes water and a biofilm,
be done in a downstream biofilter. surrounding the medium (Fig. 2). The transfer rate
An odorous compound becomes a nuisance only after between the air phase and water phase is dependent
it is transferred to the air. It becomes a problem for the on the compound characteristics, the concentration of the
surrounding community only if it is released in sufficient chemical in the air and water phase, the concentration of
amounts to overcome the dilution that occurs between the other chemicals in the water phase, the type of medium,
point of release and the neighbors and to arrive there in the surface area of the water/air interface, and other
concentrations above the odor threshold. factors (8).
Quigley and Corsi noted that three factors contribute Biofilters may be characterized as completing the
to the release of volatile compounds from wastewater (13): aerobic treatment pathway (Fig. 1). Heterotrophic bacteria
in the water and biofilm that use organic compounds for
1. The amount of compound present in the wastewater. their energy source are chiefly responsible for the oxidation
High concentrations of the odorous substance of volatile organic compounds (VOCs). Hydrogen sulfide
increase the rate of transfer of the compound from and ammonia are transformed by chemoautotrophs.
water to air. Under ideal conditions, carbon dioxide and water
2. The presence of conditions that create turbulence. are the ultimate products of contaminant degradation.
Drops in collection system piping and treatment Treatment of certain VOCs or compounds containing
plant equipment such as weirs and flumes mix nitrogen, chlorine, or sulfur may produce acids (14–19).
596 BIOFILTRATION AND BIOODORS

Air

Air flow

Air

Water Dissolution

Biofilm

Biodegradation By-products

Support
particle

Figure 2. Detail of biofilter mechanisms.

Acid production is a common problem in Publicly Owned amount of wash water is needed to stabilize the pH. The
Treatment Works (POTW) biofilters because hydrogen amount of acid discharged in biofilter leachate increases
sulfide usually is the dominant odor-causing compound, as pH falls because the concentration of protons in solution
with typical concentrations ranging from the low ppm increases. Thus, the pH falls until acid production is
to the hundreds of ppm. The production of acid causes equaled by the discharge of acid in biofilter drainage.
difficulties for POTW biofilters because most known Uncontrolled systems typically stabilize at a pH between
heterotrophic bacteria capable of consuming VOCs prefer 1 and 3, where treatment can continue indefinitely.
neutral pH (20). Researchers have assumed that removal Shinabe and coworkers (24) studied the removal of H2 S
efficiencies decrease with decreasing pH, and early data by Thiobacillus thiooxidans. They found that greater
supported this (14,15) assumption. A second problem of than 99% removal of H2 S was possible in their two-
acid production is that acid degrades organic packing stage full-scale biofilter using a ceramic medium. The
such as compost, opening channels in some areas and removal efficiency was not pH-dependent. Webster and
compacting others. coworkers (25) conducted a bench- and pilot-scale study at
Buffers, such as oyster shells (14) or dolomite (21,22), a POTW for 18 months. They used a variety of packing
have been added to the packing to avoid pH decrease. materials and compared pH-controlled and uncontrolled
However, the rates of acid production commonly seen conditions. They found that their biofilters were effective
in wastewater treatment plants consume packing buffer at simultaneously removing hydrogen sulfide and VOCs
relatively rapidly. The spent minerals and acid-degraded at both bench scale and pilot scale irrespective of pH
compost often form small particles and contribute to control. Chitwood and coworkers (26) operated a pilot-
biofilter clogging. Head loss rises, increasing operation scale biofilter with a 1-cm diameter lava rock as the
costs, and efficiency declines. Elemental sulfur also may medium. The system was operated with an average
accumulate and may be inhibitory. This commonly limits hydrogen sulfide inlet concentration of 5 ppm and the
compost to a lifetime of three or four years (15). The bed at pH 4. They observed greater than 90% removal
packing also may be washed with water to help control of hydrogen sulfide and 75% removal of total VOC at an
the pH (15). However, it is difficult to ensure uniform empty bed residence time (EBRT) of 13 seconds. Each
treatment because the water flows down through the of these studies used inorganic medium that did not
packing in irregular patterns. In practical applications, deteriorate at low pH conditions.
it is difficult and expensive to control the pH, and A practical limitation of biofilters is the required
the medium is usually replaced when its alkalinity is footprint. Traditionally, biofilters have been designed with
completely exhausted (20,23) and the pH falls below 6. an EBRT (calculated as the volume of packing divided by
However, several recent studies have shown that airflow) of 60 seconds. A loading rate of 15–25 × 10−4 m3 /s-
biofilters are effective at lower pH where only a small m2 is typical. However, biofilters designed for removal
 BIOFILTRATION AND BIOODORS 597

of H2 S from air can accept higher loads. Greater than This is especially beneficial when the biofilter is located
90% removal of H2 S has been achieved in several inside a building or when the contaminants are strongly
biofilters operated with an EBRT of 15 seconds or odorous or toxic. It also means that the blower will not
less (24,26,27). Under practical conditions for POTWs, be in contact with untreated air that may be corrosive.)
transfer of hydrogen sulfide from the air phase to the A humidification chamber is used to increase the rela-
liquid phase and maintenance of uniform airflow may be tive humidity of the air to approximately 100% before it
more important limitations than contaminant degradation enters the biofilter. This is necessary to prevent drying of
rates. the bed. Air distribution and ventilation piping provides
homogeneous distribution of the air and then collects the
Biofilter Designs treated air to be released. The biofilter bed can be single
Conventional biofilters are single-stage, up-flow, and open- or multiple layers, depending on the treatment require-
bed systems with an organic medium (Fig. 3). These ments and available space. An irrigation system is used to
biofilters require periodic watering with sprinklers but maintain proper water content in the bed. It can also be
no nutrient addition, and require little maintenance. used to add nutrients and pH buffers. Although biofilters
They are often constructed by excavating a pit, lining produce only a small amount of leachate, a leachate collec-
it with a plastic sheet, placing a perforated pipe, tion system is also required. Leachate is often recycled to
and adding gravel and compost. They are attractive the inlet of the irrigation system so that the nutrients in
because of their simplicity and low cost. However, it can be reused. Periodic wasting of leachate is necessary
closed systems (Fig. 4) that allow better system control to prevent buildup of salts or toxins in the system.
and performance are largely replacing open systems in
Application to Treatment Plant Odor Control
industrial applications (28).
A biofilter system generally consists of eight compo- Application of biofiltration to odor-emitting facilities such
nents. Ductwork collects the gases from the source. An air as POTWs and compost plants is appropriate because
blower is provided to move the air through the system. biofilters are most suited for treating large volumes of air
The blower is usually located upstream of the biofilter with low concentrations of contaminants characteristic of
vessel, but it may be placed after the biofilter if a closed these facilities. Application of biofiltration may be more
bed is used. (With this arrangement, the biofilter operates complicated at facilities that have a tendency to emit lower
under pressure slightly less than that of the surrounding volumes with higher concentration, such as bakeries and
atmosphere, so that any leaks will draw ambient air into rendering facilities. This problem can be solved in part by
the system rather than allow contaminated air to escape. simple dilution of the inlet air.

Nutrients, buffer
(Discontinuous)

Water influent Discontinuous water addition

Blower
Waste
air
Humidifier

Air distribution piping

Medium

Gravel

Figure 3. Conventional open bed biofilter design.

Discontinuous
Water influent water addition

Nutrients, buffer
(discontinuous)

Blower
Waste Clean
air air
Humidifier Figure 4. Closed bed system with downward flow.
598 BIOFILTRATION AND BIOODORS

Emissions from POTWs are characterized by high whereas aromatic VOC removal was consistently greater
airflow rates with high moisture content. They are than 90%. However, they also controlled the pH at
corrosive and may contain aerosols. Emission points neutrality. Shinabe and coworkers (24) found greater than
include plant headworks (bar screens and grit chambers), 99% removal of H2 S by Thiobacillus thiooxidans in their
primary clarifiers, aeration basins, and solids-handling two-stage full-scale biofilter using a ceramic medium.
processes (anaerobic digestion, biosolids dewatering, and The removal efficiency was not pH-dependent. They did
composting facilities) (29). Emissions consist of various not study simultaneous removal of VOCs. Webster and
odor and organic compounds. Concentrations are low and coworkers (25) conducted a 1.5-year bench- and pilot-scale
vary greatly with time (30). study at a POTW. They used a variety of media and
compared pH-controlled and uncontrolled media. They
Treatment of Hydrogen Sulfide in Biofilters. The effective- found that their biofilters were effective at simultaneously
ness of biofilters for removing hydrogen sulfide was inves- removing hydrogen sulfide and VOCs at both bench scale
tigated in numerous studies (15,19,23,31–46). Yang and and pilot scale, irrespective of pH control. Recent work has
Allen (15) conducted a comprehensive laboratory study confirmed that cotreatment of H2 S and aromatics at low
of compost biofilters treating hydrogen sulfide. The rela- pH is possible (60).
tionship between H2 S removal efficiency and temperature,
medium pH, water content, air retention time, sulfur load-
MICROBIOLOGICAL ASPECTS OF BIOFILTRATION
ing rate, and sulfate concentration in the medium were
experimentally determined. More than 99% removal effi-
ciencies under a wide range of conditions were achieved Microbial Ecology of Biofilters
with a maximum elimination capacity of 150 g/m3 /hr. Microorganisms in biofilters grow on the surface of the
However, lower pH, tested by the addition of hydrochloric packing material. They first form colonies that spread to
acid, negatively affected removal efficiency, and the opti- become patches, and the patches finally merge to create
mum pH was above 3.2. In long-term experiments, water a continuous biofilm over the packing. To be active in
was periodically added to the system for pH control. Addi- the removal of contaminants, they must have adequate
tion of 0.57 liters of deionized water per liter of medium supplies of water, oxygen, and nutrients, as well as
caused an increase in 0.2 to 0.5 pH units. adequate contact with the contaminant.
Biofilter start-up requires an inoculum of microorgan-
Treatment of Ammonia in Biofilters. Research has been isms that are capable of growing under the conditions in
conducted to determine the effectiveness of biofilters the biofilter and obtaining their energy from the degra-
for removing ammonia from waste gas streams (47–50). dation of the contaminant. Many biofilters use partially
Ammonia is a common odor problem from composting composted materials, such as yard waste, as the pack-
facilities at POTWs. The studies demonstrated that biofil- ing material. These have the advantage that they carry
ters with an established nitrifying bacteria population a diverse and active ecosystem of microorganisms with a
could effectively oxidize ammonia to nitrite, nitrate, and broad capability for degradation of organic compounds. It
organic nitrogen compounds. However, the major (68%) is very likely that microorganisms capable of degrading
end product was nitrite (48). The optimum medium pH was the contaminant will be present and ready to serve as the
approximately 8. Hartikainen and coworkers (47) demon- seed for rapid development of a dense culture for treat-
strated adequate removal of ammonia at pH 6. However, ment. Biofilters that use inorganic media such as plastic
no removal occurred in their test at pH 4. shapes or stones must be provided with an appropriate
inoculum. Biofilter operators commonly choose notably
Treatment of VOCs in Biofilters. A wide range of studies diverse inocula such as sewage sludge or extracts from
has been conducted to determine the applicability of biofil- previously active biofilters. In a few cases, pure cultures
tration to the removal of organic compounds from waste of chosen species known to degrade the contaminant have
air. Applications have included bakeries (16), soil vapor been used.
extraction emissions (51,52), the print industry (53,54), In any case, the microbial ecosystem will undergo sharp
pulp mills (55), plastics manufacturing (56), and various change and rapid development when it is exposed to the
other industrial applications (8). In general, it was shown contaminant. The species that are capable of utilizing
that chlorinated alkanes display the poorest removal effi- it will grow vigorously. Because of the large flow of
ciencies, followed by aldehydes and ketones, with the best air through the biofilter, new species are continually
removal for aromatics (25). Actual removal is compound- being introduced. In time, predators (e.g., protozoa) and
specific. A summary of elimination capacities of various parasites of the successful species will proliferate and
compounds in past biofilter studies has been compiled (8). then predators of the predators (e.g., nematodes) will
grow, creating elaborate food chains. Insect larvae may
POTW Biofiltration Research. Biofilters are an effective grow. In one case, the headspace above a compost biofilter
means of controlling odors and low concentrations of was colonized by thousands of spiders that were eating
mixed VOCs from POTWs (14,23,24,31,37,57–59). The the insects. This process of ecosystem development may
first biofilters constructed were used at POTWs to control proceed for some time. Although a biofilter typically
odors (31). Ergas and coworkers (14) conducted a 10- reaches a steady state with respect to treatment success
month pilot-scale compost biofilter study at a POTW. in weeks, there have been reports of continuing ecosystem
They found that H2 S removal was greater than 99%, change after more than a year of operation (61,62). That
 BIOFILTRATION AND BIOODORS 599

is, new species are still being introduced and the relative as a sink, so that the concentration of the contaminant
abundances of the species present are changing. declines with depth. If degradation kinetics are first order,
degradation rates will decline where the concentration
Factors Affecting Biological Activity is low, and regions where concentration of contaminant
is zero will contribute nothing to treatment success. In
Water content is a key environmental variable for the
many ways, the challenge of providing contaminants
microbial ecosystem in a biofilter. Drying is commonly
to the microorganisms is much like that for oxygen.
the cause of declines in biofilter performance (63). The
The major difference is that oxygen is present in high
microorganisms do best when the packing is covered
concentrations and has moderate solubility, whereas
with a film of water. Indeed the microbial community
contaminants are present in relatively low concentration,
helps hold this water by creating a biofilm composed
and some are only slightly soluble. This means that it is
of polysaccharides that has a significant water-holding
much more likely that degradation rates will be limited
capacity. Chitwood (26) showed, for example, that the
by contaminant concentrations in the biofilm. Because
water-holding capacity of an experimental biofilter using
contaminant concentrations fall as the air passes through
inorganic packing increased by 35% during the first three
the biofilter, this is especially true near the outlet.
months of its operation.
Nutrients such as nitrate and phosphate must be
Maintaining an ideal water content in the biofilter,
supplied to the microorganisms. Biofilters that use
however, is difficult. The porous medium water content at
compost or other organic packing may not require further
equilibrium with the atmosphere is a very sharp function
addition of nutrients because the slow decomposition of
of the RH of the air. Relative humidity values just a percent
the organic matter releases the necessary compounds to
or two below saturation will cause the biofilter to gradually
the biofilm.
dry (8). Such slight declines in RH can in turn be caused
Inorganic packing such as stones or plastic shapes does
by an increase in temperature by a few degrees, resulting
not provide nutrients, and these must be added from an
from compression in the blower or heat gain through duct
external source. It is particularly important to add these
walls. For this reason, biofilters are often equipped with
with the inoculum, to ensure that an initial biofilm is
humidifiers on the incoming air and irrigation systems
rapidly established.
that spray the packing directly. Desiccation-resistant
Although the general concepts of nutrient requirements
degraders, particularly fungi, have been investigated for
are well established, the detailed dynamics in biofilters
use in biofilters (64), and may have valuable applications.
are not. The composition of biofilter biomass is probably
However, even these systems require that the air be near
much like other organic material, so that the mass of
100% humidity.
nutrient required for growth of a given amount of biomass
An adequate oxygen supply is necessary for the
can be calculated. However, the amount of biomass
great majority of biofilters that rely on oxygen as the
in biofilters, particularly compost biofilters, is poorly
electron acceptor in the degradation process. Unlike water
known. Furthermore, significant amounts of nutrients are
treatment systems, biofilters for air are rarely short of
presumably recycled. As cells die or are consumed by
oxygen. The gas being treated is air, and the contaminants
predators, these nutrients are released for reuse by other
are present at concentrations reaching at most a few
cells. Some nutrient is probably sequestered in dead or
thousand ppm, so the oxygen concentration is far above
inactive biomass. Some will be lost in irrigation water that
stoichiometric requirements. Oxygen availability within
drains from the bottom of the bed. In compost biofilters,
the biofilm is far less certain. Oxygen must diffuse from the
the rate at which nutrients are released by decomposition
surface of the biofilm to its depths. Oxygen concentrations
of the packing is not known and it is not certain that it is
and diffusion rates are far lower in water, especially if it
always sufficient to avoid nutrient limitation.
contains a thick gel of cells and exudates. Biodegradation
Given these uncertainties, it is not surprising that
within the biofilm is constantly consuming oxygen, and
recommendations for nutrient addition vary widely in
in the deeper portions oxygen depletion is common. This
the literature. Although the observation that poorly
is probable in biofilters treating high concentrations of
performing biofilters can be improved by nutrient addition
contaminants that will cause development of a thick
is common, the amounts necessary are not certain.
biofilm and rapid consumption of oxygen. The development
Because nutrient addition is inexpensive, the common
of anaerobic regions at the base of the biofilm, or in the
solution is to add them in excess.
smaller and more remote pores of the packing are not
desirable. First, the biomass in the anaerobic regions
Microbial Ecology of Sulfide Biofilters
will not be active, reducing the effectiveness of the
biofilter. Second, anaerobic activity may generate odorous Biofilters are quite effective for the control of hydrogen
compounds, contributing to the very problem the biofilter sulfide or organic sulfide odors, but special problems arise
is intended to solve. Problems of inadequate distribution because the ultimate product of microbiological sulfide
of oxygen will be more prevalent in deteriorating packing, oxidation is sulfuric acid. As treatment proceeds, the pH
where the airflow becomes less uniform and local volumes of the biofilter packing declines. Rapid decline may inhibit
of packing are cut off from the oxygen supply. the activity of sulfide-oxidizing microorganisms, and a
Biofilter success requires that the contaminant be sufficiently low pH will hasten the deterioration of compost
brought into close contact with the microorganisms. Like packing (17).
oxygen, it must dissolve in the water phase and diffuse One approach is to neutralize the acid and maintain
into the biofilm. Degradation within the biofilm serves the pH at 7. This can be done by adding buffer material to
600 BIOFILTRATION AND BIOODORS

the packing or adding base to the irrigation water. While Odor Control in Wastewater Treatment Plants, ASCE Manu-
this approach encounters some operational problems, as als and Reports on Engineering Practice No. 82 and Water
previously described, sulfide oxidation will proceed well. Environment Federation Manual of Practice No. 22, American
In addition, a neutral pH is ideal for a wide variety of Society of Civil Engineers, New York, 1995.
heterotrophs, so that VOCs in the gas stream are also 3. Odor and Corrosion Control in Sanitary Sewerage Systems
likely to be degraded. and Treatment Plants, EPA/625/1–85/018, U.S. Environmen-
tal Protection Agency, 1985.
A second approach is to allow the pH to decline. As
4. R. P. G. Bowker, J. M. Smith, and N. A. Webster, Design
pH falls, a successional change in ecosystem composition
Manual, Odor and Corrosion Control in Sanitary Sewerage
occurs, utilizing the sulfide and producing additional
Systems and Treatment Plants, EPA 625/1–85/018, 1985.
acid. This succession is better known than most in
5. K. K. Kienow et al., Sulfide in Wastewater Collection and
microbiology because it has been studied extensively
Treatment Systems, ASCE Manuals and Reports on Engi-
in efforts to control sulfide corrosion in sewers (5,59). neering Practice No. 69, American Society of Civil Engineers,
Ultimately, sulfide-oxidizing organisms tolerant of low New York, 1989.
pH dominate the ecosystem. These are presumed to be 6. R. D. Pomeroy, J. D. Parkhurst, J. Livingston, and H. H.
of the genus Thiobacillus, and the species Thiobacillus Bailey, Sulfide Occurrence and Control in Sewage Collection
thiooxidans is probably common, but the species is Systems, EPA 600/X-85-052, U.S. Environmental Protection
rarely identified in practice. Thiobacillus produces short- Agency, 1975.
chain fatty acids as metabolic by-products and these 7. R. D. Pomeroy, Process Design Manual for Sulfide Control
are self-inhibitory. Acidophilic heterotrophs that live in in Sanitary Sewerage Systems, EPA 625/1-74-005, U.S.
a symbiotic relationship with the Thiobacillus consume Environmental Protection Agency, 1974.
these acids. Some experiments have shown that many 8. J. S. Devinny, M. A. Deshusses, and T. S. Webster, Biofiltra-
organic compounds can also be degraded in low-pH tion for Air Pollution Control, CRC Lewis Publishers, New
biofilters (60) and that some of the same acidophiles may York, 1999.
be involved. 9. D. E. Chitwood, J. S. Devinny, and F. E. Reynolds, Jr., Envi-
Most sulfide oxidizers are autotrophic and so obtain ron. Prog. 18(3), 212–221 (1999).
their carbon from carbon dioxide, which is a costly 10. G. J. Skladany et al., in H. J. Rafson, ed., Odor and
process in terms of energy. This may be the reason VOC Control Handbook, McGraw Hill, New York, 1998,
these microorganisms tend to release little polysaccharide pp. 8.150–8.191.
and thick biofilms generally are not seen in sulfide 11. W. Stumm and J. J. Morgan, Aquatic Chemistry, Wiley
biofilters. Interscience, New York, 1995, p. 1022.
Overall, biofilters are an excellent technology for 12. O. J. Hao, J. M. Chen, L. Huang, and R. L. Buglass, Crit. Rev.
removal of hydrogen sulfide from airstreams. Systems Environ. Sci. Technol. 26(2), 155–187 (1996).
have removed up to 99% of sulfide in the tens of parts per 13. C. J. Quigley and R. L. Corsi, J. Air Waste Manage. Assoc. 45,
million with an EBRT as low as 15 seconds (65). 395–403 (1995).
14. S. J. Ergas, E. D. Schroeder, D. P. Y. Chang, and R. L. Mor-
ton, Water Environ. Res. 67(5), 816–821 (1995).
CONCLUSION 15. Y. H. Yang and E. R. Allen, J. Air Waste Manage. Assoc. 44(7),
863–868 (1994).
An ideal biological treatment system uses microorganisms 16. D. S. Hodge and J. S. Devinny, Environ. Prog. 13(3), 167–173
to convert waste materials to benign products. In most (1994).
cases, the most efficient approach is to use dominantly 17. A. B. Jensen and C. Webb, Enzyme Microb. Technol. 17(1),
aerobic systems to produce carbon dioxide, water, sulfate, 2–10 (1995).
and nitrate. Odors arise if waste materials or partially 18. A. B. Pincince, Toxic Air Emission from Wastewater Treat-
transformed intermediates escape to the atmosphere ment Facilities, Water Environment Federation and the
before oxidation is complete. Odorous compounds are American Society of Civil Engineers, New York, 1995.
generated profusely in anaerobic conditions that may arise 19. J. R. Degorcedumas, S. Kowal, and P. Lecloirec, Can. J.
from inadvertent stagnation in the waste stream or from Microbiol. 43(3), 264–271 (1997).
anaerobic treatment processes. 20. C. Van Lith, G. Leson, and R. Michelsen, J. Air Waste
Passing discharged air through a biofilter that com- Manage. Assoc. 47(1), 37–48 (1997).
pletes the transformation to oxidized products can pre- 21. E. D. Schroeder, S. J. Ergas, D. P. Y. Chang, and R. Morton,
vent the release of odors. Operation of biofilters requires Proceedings of the Water Environment Federation 65th
Annual Conference Exposition, New Orleans, La., 1992.
attention to the needs of the microbial ecosystem, but well-
operated units can be effective and economical solutions 22. R. A. Jager, M. E. Lang, and K. J. Saltmarsh, Proceedings of
the 66th Annual Conference and Exposition of the Water
to the odor problems at treatment facilities.
Environment Federation, 1993.
23. T. S. Webster, Control of Air Emissions from Publicly Owned
BIBLIOGRAPHY Treatment Works Using Biological Filtration, Ph.D. Thesis,
University of Southern California, Los Angeles, Calif., 1996.
1. G. P. Van Durme, in H. J. Rafson, ed., Odor and VOC Control 24. K. Shinabe, S. Oketani, T. Ochi, and M. Matsumura, J.
Handbook, McGraw-Hill, New York, 1998, pp. 6.36–6.55. Ferment. Bioeng. 80(6), 592–598 (1995).
2. P. L. Schafer, Chair, Joint Task Force of the Water Environ- 25. T. S. Webster, J. S. Devinny, E. M. Torres, and S. S. Basrai,
ment Federation and the American Society of Civil Engineers, Environ. Prog. 15(3), 141–147 (1996).
 BIOFOULING: CHEMICAL CONTROL OF BIOFOULING IN WATER SYSTEMS 601

26. D. E. Chitwood and J. S. Devinny, Environ. Prog. 18(3), Presented at the Emerging Technologies for Hazardous Waste
212–221 (1999). Management, American Chemical Society I EC Division
27. D. Dechant, P. Ball, and C. Hatch, in Proceedings of the Special Symposium, Atlanta, GA, 1992.
Water Environment Federation 72nd Annual Conference and 52. P. A. Togna, G. J. Skladany, and S. A. J. M. Caratura, Treat-
Exposition, Water Environment Federation, 1999. ment of BTEX and Petroleum Hydrocarbon Vapors Using a
28. M. A. Deshusses, Biodegradation of Mixtures of Ketone Field-Pilot Biofilter, Presented at the 49th Annual Purdue
Vapours in Biofilters for the Treatment of Waste Air, Ph.D. University Industrial Waste Conference, 1994.
Thesis, Department of Chemical Engineering, Swiss Federal 53. K. Fouhy, Chem. Eng. 99(12), 41–46 (1992).
Institute of Technology, Zurich, Switzerland, 1994. 54. D. E. Chitwood, D. S. Hodge, and J. S. Devinny, Proceedings
29. W. Cleveland, Water Environ. Technol. 8, (April), 46–52 of the USC-TRG 1996 Conference on Biofiltration, The
(1996). Reynolds Group, Tustin, CA, 1996.
30. J. R. Witherspoon, W. J. Bishop, and M. J. Wallis, Environ. 55. E. C. Dombroski, I. D. Gaudet, and R. N. Coleman, Proceed-
Protection 4(5), 59–66 (1993). ings of the USC-TRG 1995 Conference on Biofiltration, The
31. D. A. Carlson and C. P. Leiser, J. Water Pollut. Control Assoc. Reynolds Group, Tustin, CA, 1995, pp. 164–173.
38(5), 829–840 (1966). 56. D. Rho, N. Matte, and M. Utiger, Proceedings of the USC-
32. H. L. Bohn, J. Air Pollut. Control Assoc. 25, 953–955 (1975). TRG 1996 Conference on Biofiltration, The Reynolds Group,
Tustin, CA, 1996.
33. H. Van Langenhove, E. Wuyts, and N. Schamp, Water Res.
20(12), 1471–1476 (1986). 57. Y. Hwang, T. Matsuo, K. Hanaki, and N. Suzuki, Water Res.
28(11), 2309–2319 (1994).
34. E. R. Allen and Y. Yang, Proceedings of the 84th Annual
Meeting Exhibition of the Air Waste Management Association, 58. K. Vaith, M. Cannon, D. Milligan, and J. Heydorn, Water
Air Waste Management Association, Cincinnati, Ohio, 1991. Environ. Technol. 8(6), 35–38 (1996).
35. H. L. Bohn, Recent Developments and Current Practices in 59. R. L. surIslander et al., J. Environ. Eng., 117(6), 761–770
Odor Regulations, Controls and Technology, 1991. (1991).
36. K. S. Cho, M. Hirai, and M. Shoda, J. Ferment. Bioeng. 73(3), 60. J. S. Devinny, D. E. Chitwood, and D. S. Choi, Proceedings of
219–224 (1992). the 92nd Annual Meeting Exhibition of the Air Waste Man-
agement Association, Air Waste Management Association,
37. M. Joyce, N. Wadsworth, B. Howard, and J. Bewley, Control Cincinnati, Ohio, 1999.
of Sulfide and VOC Emissions with a Soil Bed Scrubber, Pre-
61. T. S. Webster, J. S. Devinny, E. M. Torres, and S. S. Basrai,
sented at the California Water Pollution Control Association
Biotechnol. Bioeng., 53(3), 296–303 (1997).
Conference, Sacramento, Calif., 1992.
62. A. De Castro and D. G. Allen, Proceedings of the USC-TRG
38. T. O. Williams and F. C. Miller, Biocycle 33(10), 72–77
1996 Conference on Biofiltration, The Reynolds Group, Tustin,
(1992).
CA, 1996, pp. 164–173.
39. S. J. Park, K. S. Cho, M. Hirai, and M. Shoda, J. Ferment.
63. C. Van Lith and G. Leson, Proceedings of the USC-TRG 1996
Bioeng. 76(1), 55–59 (1993).
Conference on Biofiltration, The Reynolds Group, Tustin, CA,
40. P. Wolstenholme and R. Finger, Long-Term Odor and VOC 1996, pp. 102–127.
Performance Tests on Biofilters, Presented at the 67th Annual
64. J. Van Groenestijn and M. Harkes, Proceedings of the USC-
Conference Exposition of the Water Environment Federation,
TRG 1995 Conference on Biofiltration, The Reynolds Group,
Chicago, IL, 1994.
Tustin, CA, 1995, pp. 317–324.
41. B. M. Brennan, M. Donlon, and E. Bolton, J. Chart. Inst.
65. D. E. Chitwood, Two-stage Biofiltration for Treatment of
Water Environ. Manage. 10(3), 190–198 (1996).
POTW Off-gases, Ph.D. Thesis, University of Southern
42. Y. C. Chung, C. P. Huang, and C. P. Tseng, J. Biotechnol. California, Los Angeles, CA, 1999.
52(1), 31–38 (1996).
43. P. Amirhor, J. D. Gould, F. D. Arnold, and H. J. Gracia,
Biocycle 38(9), 69–70 (1997).
44. L. Finn and R. Spencer, Biocycle 38(1), 40 (1997). BIOFOULING: CHEMICAL CONTROL OF
45. R. L. Morton and R. C. Caballero, Proceedings of the 90th BIOFOULING IN WATER SYSTEMS
Annual Meeting Exhibition of the Air Waste Management
Association, Air Waste Management Association, Cincinnati, T. EUGENE CLOETE
Ohio, 1997. VOLKER S. BRÖZEL
46. J. S. Devinny, D. Chitwood, and F. E. Reynolds, Jr., Proceed- University of Pretoria
ings, 91st Annual Meeting and Exhibition of the Air Waste Pretoria, South Africa
Management Association, Air Waste Management Associa-
tion, Cincinnati, Ohio, 1998. Microbial biofilms are problematic in a range of industrial
47. T. Hartikainen, J. Ruuskanen, M. Vanhatalo, and P. J. Mar- environments in which large areas of submerged surfaces
tikainen, Environ. Technol. 17(1), 45–53 (1996). are exposed to relatively high nutrient fluxes, providing
48. Y. C. Chung, C. P. Huang, and C. P. Tseng, Biotechnol. Prog. niches for the formation of copious surface-associated
13(6), 794–798 (1997). growth (1,2). Bacterial colonization of surfaces in an
49. Y. C. Chung and C. P. Huang, Environ. Prog. 17(2), 70–75 aqueous environment is a basic strategy for survival
(1998). in nature as nutrients are more available at the
50. M. Yani, M. Hiral, and M. Shoda, Environ. Technol. 19, solid–liquid interface (3,4). The resulting aggregates form
709–715 (1998). microcolonies that develop into biofilms (5). Biofilms
51. V. F. Medina, Treatment of Soil Vapor Extraction Off-Gases increase fluid frictional resistance (5) and decrease
by GAC Based Biological Filtration: Bench and Pilot Studies, the rate of heat energy transfer, collectively termed
602 BIOFOULING: CHEMICAL CONTROL OF BIOFOULING IN WATER SYSTEMS

biofouling. These biofilms also promote corrosion of readily in nature (termed antiseptics, disinfectants, bio-
ferrous and other metals by the concerted metabolic cides, bactericides, sanitizers, and preservatives). Mem-
activity of a number of biofilm-associated bacterial bers of the second group are classified depending on their
types (6–8), a process collectively termed microbially- chemical nature, but more often on their specific field of
influenced corrosion (MIC). MIC encompasses a number of application.
specific mechanisms relating either directly or indirectly The use of biocides to control biofouling in water
to the metabolic activity of a variety of microorganisms, systems is still an accepted practice, although higher levels
notably the action of sulfidogenic bacteria (9,10). As of environmental awareness and tighter legislation have
the costs attributable to MIC and biofouling are high, placed increased pressure on the water treatment industry
effective control of bacterial numbers in industrial aqueous to seek alternative means of control (16,17). Nevertheless,
environments is essential. biocides are still indispensable components for effective
A range of bactericidal substances, commonly termed control of biofouling as they remain the core technology for
biocides or microbicides, are available, all of which are decimating viable numbers whereas all other techniques
claimed by their producers to kill bacteria occurring in merely aid in their efficacy.
aqueous systems quantitatively. Biocides target a range The modes of action of a plethora of antibiotics have
of cellular loci, from the cytoplasmic membrane to res- been investigated in detail. Much less is known about the
piratory functions, enzymes, and genetic material (11). mechanisms of action of the many biocides available for
However, different bacteria react differently to bacteri- biofouling control. Biocides for water treatment target a
cides, either owing to inherent differences such as unique range of specific cellular components and functions, from
cell envelope composition and nonsusceptible proteins (12) membrane permeability and electron transport to enzyme
or to the development of resistance, either by adaptation function.
or by genetic exchange (13). Bactericides should therefore Bactericides attack functional cell components, placing
be evaluated against the organisms that they are chosen the bacterium under stress (18). At low concentrations,
to control, that is, the dominant ones in the system to bactericides often act bacteristatically, and are only
be treated. The composition of microbial populations in bactericidal at higher concentrations (19). Targets of
systems varies with the type of water used, and changes bactericide action are components of the cytoplasmic
considerably following treatment with various biocides membrane or of the cytoplasm (11). For bactericides to be
by selection for resistant strains (14). Bacteria growing effective, they must attain a sufficiently high concentration
as biofilms are also significantly more resistant to most at the target site in order to exert their antibacterial action.
antimicrobial agents known currently, so that methods for In order to reach their target site(s), they must traverse the
their control pose an ongoing challenge (2,15). outer membrane of the gram-negative bacteria. Therefore,
Successful biofouling control depends on rationally different bacteria react differently to bactericides owing to
developed treatment strategies that are based on informa- differing permeabilities of their respective cell walls (20).
tion about the specific system. The primary target should Bacteria with effective penetration barriers to biocides
always be the biofilm-associated flora as it is the catalyst generally display a higher inherent resistance than
for MIC and impacts negatively on system operation. those bacteria that are readily penetrated. The rate of
Five approaches are currently available and may be penetration is linked to concentration so that a sufficiently
used in combination: high biocide concentration will kill bacteria with enhanced
penetration barriers (11).
1. Bacteria are chemically killed by application of Water treatment bactericides fall into two categories,
biocides; oxidizing (e.g., chlorine and hydrogen peroxide) and
nonoxidizing. Nonoxidizing bactericides can be divided
2. Cells are dislodged from surfaces by dispersants;
into five groups based on their chemical nature or mode of
3. The biofilm structure is weakened by enzymes or action, and these are discussed later.
chelating agents of divalent cations;
4. Biofilms are removed physically by a variety of Oxidizing Biocides
processes; and
5. Biocide efficacy is enhanced by applying an alternat- Oxidizing biocides are general chemical oxidants. They
ing current or ultrasonic sound across the biofilm. are not selective for living organisms, but react with
any oxidizable matter. However, they are bactericidal
This entry focuses on the application of biocides to because certain bacterial cell components can react
biofouling control. readily with them, having a higher oxidation potential
than most other chemicals present in water. Three
classes of oxidizing biocides are available for bactericidal
BIOCIDES applications; oxidizing halogens, peroxides, and ozone.

Bactericides are antimicrobial agents employed in var- Peroxides. Peroxides are unstable oxygen compounds
ious spheres of human activity to prevent, inhibit, or that decompose to form free hydroxyl radicals. These react
eliminate microbial growth. They can be divided into two oxidatively with organic matter. The peroxides include
groups — those derived from naturally occurring antimi- hydrogen peroxide, peracetic acid, aromatic peroxyacids,
crobial agents (termed antibiotics) and those not occurring persulfates, and calcium peroxide.
 BIOFOULING: CHEMICAL CONTROL OF BIOFOULING IN WATER SYSTEMS 603

Hydrogen Peroxide. Peroxide has good antimicrobial acid is dependent on pH. It dissociates at pH greater
properties and decomposes to water and oxygen, leaving than 7, and the undissociated moiety is the antibacterial
no toxic waste. Hydrogen peroxide penetrates cells one (23). Above pH 7.5 it loses its antibacterial activity.
causing site-directed damage on account of metal- It is excellent for biofouling control as it weakens
dependent OH formation (21). It causes DNA strand the extracellular polysaccharide structure, leading to
breaks and base hydroxylation. Guanine and thymine are sloughing and removal of sections of the biofilm.
the two main targets of peroxide-generated free radical
attack. The resulting 7,8-dihydro-8-oxoguanine mispairs Bromine Compounds. Hypobromous acid works simi-
with adenine whereas thymine oxidation products stop larly to hypochloric acid. It is, however, stable at a pH
DNA polymerase, halting replication (21). Most bacterial up to 8.5. This makes it more suitable for application in
mutants cannot survive because of incoherent metabolism, cooling waters, which are often maintained at a slightly
so that peroxide treatment at low concentration leads to alkaline pH. Certain organic compounds release hypo-
slow death. Hydrogen peroxide also inhibits mitochondrial bromic and hypochloric acid slowly when in solution. An
ADP-phosphorylation. example is 3-bromo-1-chloro-5,5–dimethylhydantoin (28).
The development of resistance to oxidizing bactericides Such compounds maintain a longer antibacterial level of
has not been reported in the biofouling control literature. hypohalous acid in the system treated.
However, a variety of bacteria, mostly fermentative,
exhibit oxidizing stress response by producing oxidant-
Ozone. Ozone is a strong oxidizing material capa-
degrading and repair enzymes. Stress response means
ble of killing bacteria and algae and of inactivating
that cells become more resistant to a deleterious
viruses. It is an unstable gas with a pungent odor. It
factor within minutes of exposure to subinhibitory
further degrades the extracellular polysaccharide, holding
quantities of the factor. A variety of defense genes have
biofilms together, so that treatment results in loosening
been characterized in Escherichia coli, encoding various
of the biofilm. This leads to loosening of scale from the
superoxide dismutases, catalases, alkyl hydroperoxide
surface. Ozone has a very short half-life and therefore has
reductases and glutathione reductases, as well as DNA
to be generated on-site. In distilled water, the half-life of
repair enzymes (22). In addition, various regulatory genes
ozone at 20 ° C is 25 minutes. Its solubility in water is 13
have been characterized, including oxyR and soxR. These
times that of oxygen. Upon reaction with organic mate-
regulators determine intracellular redox potential and
rial it decomposes to oxygen. It does, however, react with
activate stress response when cells are exposed to oxidizing
several cations and anions such as Fe2 +, Mn2 +, MnO4 2− ,
agents.
N02− , and CN− . Ozone is toxic to humans, and detec-
tors should be installed together with ozone generators.
Organic Peroxides. Peracetic acid is the best known of Furthermore, ozone is very costly to produce when com-
the organic peroxides. Like hydrogen peroxide, it forms pared with chlorine. However, treated water is perfectly
free hydroxyl radicals, which react with various protein safe as ozone degenerates to oxygen whereas chlorine
structures and DNA. In addition, the dissociation of may react with humic substances in water to yield toxic
peracetic acid leads to formation of acetic acid, which trihalomethanes.
is mildly antibacterial. Application of peracetic acid to
systems does not leave any toxic waste behind. The
Electrochemically Activated Water
antibacterial activity of peracetic acid is not affected by
water hardness or organic contamination. Principles of Electrochemically Activated Water Tech-
nology. Electrochemically activated water has recently
Oxidizing Halogens. Hypochloric and hypobromic acids become available, and has been found effective in bio-
pose excellent antibacterial activity, although within a fouling control (29). Water of varying mineralization is
defined pH range. passed through an electrochemical cell, the specific design
of which permits the harnessing of two distinct and elec-
Chlorine Compounds. Chlorine, chlorine dioxide, and trically opposite streams of activated water.
hypochlorous acid (HOCl) are the most widely used Aside from its distinctive attributes, the negatively
biocides worldwide. Hypochlorite was first employed as a charged antioxidant solution (catholyte) can also be chan-
wound disinfectant by Hüter in 1831, and its bactericidal neled back into the anode chamber, thereby modulat-
activity was confirmed by Koch in 1881 (23). Hypochlorite ing the quality of the positively charged oxidant solu-
is used in industrial water systems to control biofouling. tion (anolyte) that is produced. Depending on the spec-
The antibacterial mechanism of action of hypochlorite ifications of the required application, variations in the
is not clear to date although much work has been done on design of the hydraulic systems can be effected to meet the
the mechanism of action in eukaryotic cells (8). HOCl requisite objectives.
is a powerful oxidizing agent, oxidizing thiol groups The design of the cell is such as to ensure a
and halogenating amino groups in proteins (24), as well uniformly high voltage electrical field through which
as oxidizing lipids and proteins (25). Specific bacterial each microvolume of water must pass. This unipolar
targets are cytochromes, nucleotides, and iron-sulfur electrochemical activation created by potential gradients
proteins (26). Protein synthesis is impaired (data not of millions of volts per cm2 between the anode and
published) and the uptake of nutrients is also affected (27). cathode terminals results in the creation of solutions
The stability and antimicrobial activity of hypochloric whose pH, oxidation reduction potentials (ORP) and other
604 BIOFOULING: CHEMICAL CONTROL OF BIOFOULING IN WATER SYSTEMS

physicochemical properties lie outside of the range, which This heightened biocidal capacity relative to traditional
can be achieved by conventional chemical means. chemical solutions permits the incorporation of ECA
solutions at substantially lower dose rates, therein
Properties of Activated Water. The properties of the obviating the risk of intoxication, adverse environmental
activated solutions are dependent on a number of factors. impact, while providing cost-effective resolutions.
These comprise the solution flow rate through the reactor
cell, the current being applied, temperature, the degree Nonoxidizing Biocides
of feedback of catholyte into the anolyte chamber and the
degree of mineralization of the water. Nonoxidizing biocides include a variety of organic chemical
During electrochemical activation, three categories of compounds that have antimicrobial activity. Their modes
products within the solution are generated. They comprise: of action differ vastly, and their only common denominator
is that they are nonoxidizing organic molecules. Most
1. Stable products that include acids and bases that currently used biocides fall into five distinct categories,
influence the pH of the solutions, although a number of miscellaneous compounds are also
2. Highly active, unstable products including free rad- useful.
icals and electrolytic gases in the form of microbub-
Detergent-Type Biocides. Three groups of surface-active
bles that influence the ORP of the solutions, and
antimicrobial agents have been documented to date;
3. Quasi-stable products comprising complexes of
anionic, cationic, and amphoteric (23). Anionic antimi-
hydrated membranes that form clusters of water
crobials are only effective at pH < 3.0 and include the
molecules, which impart the catalytic activity of the
aliphatic acids such as sodium dodecyl sulfate. The
solutions.
cationic antimicrobial agents are the generally organic
Without maintenance of the activated state, these ammonium salts, commonly termed quaternary ammo-
diverse products degrade to the relaxed state of benign nium compounds (QAC). The best known is benzalkonium
water and the anomalous attributes of the activated chloride (N-alkyl-N,N-dimethyl benzylammonium chlo-
solutions such as altered conductivity and surface tension ride) (23). Recently, a number of novel organic ammonium
similarly revert to preactivation status. salts of general structure alkyl trimethyl and dialkyl
It is important to note that the level of mineralization dimethyl ammonium bromide were synthesized and eval-
of input water required to generate optimally metastable uated (30). The authors reported extremely low minimum
solutions is not significantly different from the composition inhibitory concentrations to Pseudomonas aeruginosa of
of benign potable water. However, the heightened 12.5µg/ml where the second alkyl group was C6 or longer.
electrical activity and altered physicochemical attributes Benzalkonium chloride adsorbs to the cell surface of
of the solutions differ significantly from the benign negatively charged cells (pH > 7.0) in an irreversible
state, yet remain nontoxic to mammalian tissue and the way (11,31). The pH minimum for antimicrobial activity
environment. is 3.0. It is membrane active and induces leakage of cyto-
plasmic constituents (23). Upon exposure to benzalkonium
Biocidal Properties of Anolyte. Earlier technologies that chloride, membranes of Pseudomonas cepacia appeared
have employed electrochemical activation to generate irregular, indicating membrane damage (32). At 37 ° C it
biocidal solutions have not been capable of separating is twice as active as at 20 ° C. It is active against gram-
the output of the anolyte and catholyte solutions. In these positive as well as against gram-negative cells, but not
cases the two opposing solutions have neutralized each against spores. Cations such as Ca2 + and Fe3 + decrease
other with regard to potential electrical activity. its activity, as does NaCl (23).
The advantages of the current ECA technology has been
confirmed, wherein the biocidal activity of hypochlorous Biguanides. Biguanides are polymer derivatives of
acid generated by the current ECA technology is 300 times a general guanidine structure. Two biguanides are
more active than the sodium hypochlorite generated by currently used as industrial bactericides. These are
earlier systems. Additionally, a comparison of neutral polyhexamethylene biguanide (PHMB) and 1,6-di-(4-
anolyte (pH = 7), with alkaline glutaraldehyde (pH = 8.5), chlorophenyldiguanido)-hexane, better known as chlor-
showed that the latter required a concentration of hexidine (23). Both are not corrosive and all are well-suited
2% versus 0.05% of the former, in order to achieve the for application in cooling water (19).
same biocidal efficacy. Similarly, it has been shown that Biguanides are bacteriostatic at low concentrations and
a 5% solution of sodium hypochlorite can only be used for bactericidal at higher concentrations, and have a wide
purposes of disinfection whilst a 0.03% solution of neutral spectrum of activity, especially against gram-negative
anolyte has both disinfectant and sterilizing properties. bacteria (23). They are membrane active agents and
In general, the biocidal activity of nonactivated neutral attach rapidly to negatively charged cell surfaces (pH
anolyte (only stable products and no electrical charge) is neutral or alkaline). By making use of 14 C-radiolabeled
80 times the potential activity of the hypochlorite solution, polyhexamethylene biguanide, it has been shown that
but still exhibits only one-third of the full biocidal potential polyhexamethylene biguanide absorbed into cells of E. coli
of the optimally activated ECA solution. within 20 seconds after exposure (33). Bactericidal action,
Thus, activated solutions have been conclusively shown however, requires a few minutes. Biguanides compete
to exceed chemically derived equivalents both in low- with divalent cations for negative sites at LPS, displacing
dosage effectiveness as well as physicochemical purity. these. PHMB then interacts by electrostatic interactions
 BIOFOULING: CHEMICAL CONTROL OF BIOFOULING IN WATER SYSTEMS 605

with the charged headgroups of phosphatidyl glycerol membrane, inducing the leakage of cytoplasmic con-
and diphosphatidyl glycerol (negative), but not with the stituents (11). 3- and 4-chlorophenol uncouple oxidative
neutral phosphatidyl ethanolamine (34). By binding to phosphorylation from respiration by increasing the per-
phospholipids of the inner leaflet of the outer membrane meability of the cytoplasmic membrane to protons (37).
and of the outer leaflet of the inner membrane, the two
membranes attain net positive charges and are repelled Thiol-Oxidizing Biocides. Thiols on amino acids such
from each other, causing membrane damage by distortion. as cysteine are important groups that influence the
This is supported by TEM studies on P. cepacia in which tertiary structure of proteins by forming disulfide bridges.
both membranes acquired a distinct irregular appearance Three groups of antimicrobial agents, isothiazolones,
after treatment with chlorhexidine (32). Cytoplasmic bronopol (2-bromo-2-nitropropane-1,3-diol), and mercury,
constituents start leaking out of the cell because of the and other heavy-metal compounds, react with accessible
rupture of the membranes, and the cell loses its viability. thiols, altering the three-dimensional structure of enzymes
and structural proteins (38). Mercury interacts with
Aldehyde-Based Biocides. Two aldehydes are commonly sulfhydryl groups by complexing with sulfur (23). Bronopol
used as antimicrobial agents, that is, formaldehyde oxidizes thiols to disulfides, reacting especially with the
and glutaraldehyde. Further, there is a range of bac- active center of hydrogenase enzymes (23).
tericides such as the hydroxyethyl- and ethyltriazine- Four water-soluble isothiazolones possess antibacterial
bactericides available, all of which release formalde- activity; 5-chloro-2-methyl-3-(2H)-isothiazolinone (CMIT),
hyde (35). Formaldehyde has a high polarity and high 2-methyl-3-(2H)-isothiazolinone (MIT), 1,2-benzisothiazo-
nucleophilic reactivity, so that it reacts primarily with lin-3-one (BIT) and 2 methyl 4,5-trimethylene-4-isothiazo-
free primary amino groups, but also reacts with amines, lin-3-one (MTI) (23). MIT and CMIT are often supplied
amides, sulfides, purines, and pyrimidines (35). In water in a 3 : 1 ratio. Isothiazolones react oxidatively with
it hydrates to methylene glycol. Reaction with primary accessible thiols such as cysteine and glutathione (39).
amino groups leads to the formation of methyloamines that These thiols are oxidized to their disulfide adjuncts,
react further with cellular components. Formaldehyde which, in the case of cysteine, leads to an alteration of
damages the transport properties of membrane porins, protein conformation and functionality. Isothiazolone is
decreasing the rate of proline uptake and of enzyme syn- thus reduced to mercaptoacrylamide, which in the case
thesis. It is active over a wide pH spectrum (3.0–10.0) and of CMIT tautomerizes to thioacyl chloride, the latter
is sporicidal (23). reacting with amines such as histidine and valine (38).
Glutaraldehyde also reacts with amino and sulfhydryl Isothiazolones are primarily bacteriostatic and are only
groups (11). It is stable in acid solution but is only bactericidal at high concentrations.
active at pH 7.5 to 8.5, so it must be alkalinified before
application (23). A 2% solution at the correct pH is 10 times Miscellaneous Biocides. The mechanisms of action
more bactericidal than a 4% solution of formaldehyde (23). of various antimicrobial agents, employed to con-
Its reactivity is related to temperature; a 2% solution trol bacterial growth in cooling water sys-
kills spores of Bacillus anthracis in 15 minutes at 20 ° C, tems, have not been formally published to date.
whereas it requires only 2 min at 40 ° C. In gram-positive These include tetra-alkyl phosphonium chloride,
bacteria it reacts with, and binds to, peptidoglycan and sodium diethyldithiocarbamate, methylene bisthio-
teichoic acid and is also sporicidal (11). In gram-negative cyanate (MBT), 2-thiocyanomethylthiobenzothiazole, 2-
bacteria it reacts primarily with lipoproteins of the outer thiocyanomethylthiobenzothiazole, hexachlorodimethyl-
membrane, preventing the release of membrane-bound sulfone, tetrakishydroxymethylphosphonium sulfate
enzymes. (THPS) and a range of 2-arylthio-N-alkylmaleimides (6).
Phosphonium chloride probably has surfactant prop-
Phenol Derivatives. Phenol was the antimicrobial agent erties, damaging the bacterial cell envelope. It is active
that revolutionized invasive surgery, and was pioneered over an extremely wide pH range (2–12) and can be used
by Lister in 1870 (36). It enters the cell by dissolving in in conjunction with oxidizing biocides. MBT is readily
the membrane, and upon entry into the cytoplasm, precip- miscible with dispersants to aid the penetration into and
itates proteins. It is, however, harmful to humans, and its the dispersion of biofilms. The mode of action of this
antibacterial activity is not very high (23). A range of halo- nonmetallic organosulfur compound is as yet unknown.
genated phenols, cresols, diphenyls, and bisphenols have Thiocarbamates are used as agents for the extraction
been developed from phenol, and have excellent antimi- of trace metals such as Fe, Cd, Co, Cu, Mn, Ni, Pb,
crobial activity, many being applied in the preservation and Zn (40). This would imply that they chelate iron,
of pharmaceutical products. Halogenation increases the a vital trace element of most bacteria. The nucleophilic
antimicrobial activity of phenol, as does the addition of sulfur atom indicates potential reactivity with accessi-
aliphatic and aromatic groups (23). Bisphenols have the ble thiols. Thiocarbamates do react with accessible thiols
highest antimicrobial activity of the phenol derivatives, and amines. Therefore, their antibacterial mechanism of
especially the halogen substituted ones. Hexachlorophen action would rest partially on denaturation of surface pro-
and 2,2 -methylenebis(4-chlorophenol) (dichlorophen) fall teins. We have found that the antibacterial mechanism
into this group (13). of action depends on the alkyl chain length of the thio-
Phenol derivatives are membrane-active agents. They carbamate. Sodium diethyldithiocarbamate is inactivated
penetrate into the lipid phase of the cytoplasmic by free ferrous iron, indicating that it removes iron (a
606 BIOFOULING: CHEMICAL CONTROL OF BIOFOULING IN WATER SYSTEMS

growth factor) from the cell. Sodium dimethyl dithiocarba- that mixtures be applied (47,48). Recently, Johansen
mate is not inactivated to the same degree, showing that and coworkers (49) demonstrated that the addition of
its antibacterial activity does not rest as much on iron lactoperoxidase to saccharolytic cocktails contributed not
removal. 2-Thiocyanomethylthiobenzothiazole is also sta- only to biofilm removal, but also to a decrease in the
ble over a wide pH range and can be used synergistically culturable count.
with surfactants to aid in dispersion of and/or penetration Dispersants (Surface-Active Compounds). More recen-
into biofilms. Hexachlorodimethylsulfone is stable at high tly, surface-active compounds (surfactants) have been
temperature and is an effective slimicide over a wide pH employed to prevent bacterial adhesion to surfaces. It
range, although it is not water-soluble and must be dosed is unlikely that surfactants will have any mutagenic
as an emulsion. THPS is an amphoteric surfactant miscible effects on bacteria or that microorganisms would be
with various other surfactants to improve action against able to become resistant to the action of surfactants,
biofilms. It is active in both acidic and alkaline conditions as can be the case with biocides (11,14). Unfortunately
and at high temperatures, and remains water-soluble in little published information is available on the efficacy
oil-water mixtures. of different biodispersants (surfactants) against bacterial
The 2-arylthio-N-alkylmaleimides are extremely bac- attachment (50). According to Paul and Jeffrey (51),
teriostatic, and many derivatives demonstrate an dilute surfactants completely inhibited the attachment
MIC against various bacteria except P. aeruginosa of of estuarine and marine bacteria. Surfactants result in
6.25 µg/ml (6). both uniform wetting of the surface to be treated and
have an additional cleaning effect (14,50). Whitekettle (52)
Alternative Approaches of Chemical Control found a correlation between the ability of a surface-
active compound to lower surface tension and its ability
Surface-Embedded Biocides and Catalysts. Various ter-
to prevent microbial adhesion. White and Russel (53)
tiary amines have been covalently linked to polystyrene
classified surfactants according to the ionic nature of the
(41), and quaternary amine compounds have been
hydrophilic group, viz. anionic, cationic, nonionic, and
complexed into polyvinyl chloride and ethylene vinyl
zwitterionic.
acetate latex (42). The broad-spectrum antimicrobial Tri-
closan (2,4,4 -trichloro-2 -hydroxy-diphenylether-5-chloro-
2-(2,4-dichlorophenoxy)-phenol) has been impregnated FACTORS AFFECTING EFFICACY OF TREATMENT
into various polymers and shown to suppress surface- PROGRAMS
associated growth. It is active against most bacteria,
excepting P. aeruginosa. The antibacterial activity of bactericides is determined
Recently, Wood and coworkers (43) reported a strategy by their chemical reactivity with certain organic groups.
for circumventing the resistance of bacteria in biofilms Bactericides do not select between free and cell-bound
by generating biocidal species at the biofilm-substratum groups. Therefore, oxidizing bactericides react with any
interface from less active agents using surface-embedded readily oxidizable organic compound, and not only with live
catalysts. They embedded the catalysts cobalt- and copper- cells. Bactericide activity is influenced by the chemistry
sulfonated phthalocyanine in surface-associated resin. of the environment it is used in (23). Factors affecting
Both hydrogen peroxide and potassium monopersulfate bactericide effectivity are the following:
are broken down into highly reactive oxygen species
and significantly enhanced the killing of P. aeruginosa • pH
biofilms (44). • Water hardness
• Organic compounds such as proteins or saccharides
Enhancement of Biocidal Activity by an Electrical
• Additives such as antiscaling agents or corrosion
Field. Bacteria in biofilms are well known to be more resis-
inhibitors
tant to biocides than are their planktonic counterparts. A
select number of studies have reported that application These factors affect different bactericides to different
of low-intensity electric fields in the range 15 µA/cm2 to degrees. Some bactericides are not very stable in
2.1 mA/cm2 and with a field strength of 1.5 to 20 V/cm concentrated form and undergo changes. Formaldehyde
can override the biofilm-associated resistance (45,46). All polymerizes when exposed to polar compounds (acids or
areas within the electric field are affected as the antimi- alkalis) or high temperature and oxidizes to formic acid
crobial agent is moved by an electrophoretic effect. when exposed to air (23). Isothiazolones are unstable at
temperatures above 40 ° C and chlorhexidine is unstable
Saccharolytic Enzymes. The biofilm is stabilized by an above 70 ° C (23). A decrease in the efficacy of a bactericide
extracellular polysaccharide matrix, often interspersed treatment program can be due to a decrease in bactericide
by other extracellular polymers such as proteins. The activity or due to inactivation by adverse conditions and
extracellular polysaccharides are composed of homo- does not always indicate bacterial resistance (1).
and heteropolysaccharides of mannose, glucose, fucose,
galactose, mannuronic acid, guluronic acid, and pyruvate,
Chemistry of the Water
giving rise to a complex array of polymeric structures (47).
Saccharolytic enzymes have been used to degrade the Chemicals that inhibit scaling and corrosion are also added
polysaccharide matrix, thereby destabilizing the biofilm to industrial water systems, and some of these chemicals
structure, but the heterogeneity of polymers dictates interact with certain biocides. Chromates are used to
 BIOFOULING: CHEMICAL CONTROL OF BIOFOULING IN WATER SYSTEMS 607

inhibit corrosion and also suppress microbial growth, thiocarbamate killed 99.87% of P. stutzeri at l74 ppm and
acting synergistically with the biocide used. Glycolic left it the dominant planktonic survivor (62.5%) in the
acid secreted by algae can, however, reduce chromate, treated system. QAC-tin killed 100% of Acinetobacter
rendering it inactive. Dithiocarbamates reduce chromate, calcoaceticus and left Acinetobacter spp. dominant (40%)
so the two substances are incompatible (40). QACs form in the system. Although the surviving strains could be
insoluble chromate precipitates at high concentrations, so different, the correlation is striking.
the two should not be added simultaneously to water. The Two reasons explain why the efficacy of bactericide
careless application of chloride can lower the pH to a point treatment programs can decrease at times. One is a
at which the protective chromate film is solubilized. Na- decrease in the activity of the bactericide and the other
2-mercaptobenzothiazole is a corrosion inhibitor that is is a decrease in the bacterial susceptibility toward the
oxidized by chlorine dioxide. Methylene bis-thiocyanate is bactericide. Three mechanisms of resistance have been
hydrolyzed under slightly alkaline conditions (pH 7.5). reported in the field of antibiotic study:
Where the chlorine demand of water is high, the
large quantity of chlorine added leads to a high chloride • Inaccessibility of the antimicrobial agent to its site of
level that increases the corrosion potential of the water. action,
Chlorine and quaternary ammonium compounds increase • Absence of the susceptible site, or alteration to an
the corrosion rate of copper alloys. insusceptible form, and
• Inactivation of the antibacterial agent.
Bacterial Resistance
Bactericide treatment regimes for cooling water systems Whereas antibiotics used in the medical field are rather
often fail, posing the question of bacterial resistance specific in their modes of action, targeting specific cellular
to the bactericide. Some authors have argued that components or processes, biocides are far less specific so
the failure of treatment programs was on account that the alteration of a reactive site or the substitution
of selection for resistant strains. We have, however, of an amino acid in a protein will not render bacterial
shown that susceptible bacterial isolates do acquire cells resistant. Therefore, inaccessibility and inactivation
increased tolerance to bactericides following serial transfer are the two classes of possible mechanisms of resistance.
in subinhibitory concentrations. Resistance has been Additionally, active removal of biocide is a third class of
defined as the temporary or permanent ability of an resistance.
organism and its progeny to remain viable and/or multiply
under conditions that would destroy or inhibit other Decreased Permeability. The initial stage of bactericide
members of the strain (31,54). Bacteria may be defined as action is binding to the bacterial cell surface. It must
resistant when they are not susceptible to a concentration then traverse the cell wall (gram-positive) or outer
of antibacterial agent used in practice. Traditionally, membrane (gram-negative) to reach its site of action
resistance refers to instances in which the basis of at the cytoplasmic membrane or cytoplasm. In gram-
increased tolerance is a genetic change and in which the positive bacteria there are no specific receptor molecules
biochemical basis is known. Whereas the basis of bacterial or permeases to assist or block bactericide penetration.
resistance to antibiotics is well known, that of resistance The cell wall of Bacillus megaterium is permeable to
to antiseptics, disinfectants, and food preservatives is less molecules up to 30 kDa. Intrinsic resistance of gram-
well understood. The basis of bacterial development of positive bacteria to bactericides is therefore low. The
resistance to water treatment bactericides is little known. gram-negative cell envelope has, however, evolved to
As biocides are selective in their action, application of regulate the passage of substances into and out of the cell
any one could result in selection for resistant bacteria. to a remarkable degree of specificity. All the components of
As cells in biofilms and planktonic communities are in the cell envelope, except peptidoglycan, play a role in the
continuous exchange, death of cells in the planktonic phase barrier mechanisms because peptidoglycan is spongy and
would influence the equilibrium and shifts would occur in therefore permeable. Pseudomonas aeruginosa is the most
both the planktonic and the sessile populations. Biocides resistant nonsporeforming bacteria to most bactericides,
attack targets of cell function, placing the bacterium under as a result of the superior barrier properties of its
stress. It is well recognized that communities under stress outer membrane. In a recent study, the antimicrobial
have a lower species diversity and select for fitter species. activity of a series of new 2-arylthio-N-alkylmaleimides
Therefore, a more resistant community could develop. were compared and many were found active against
The concentration of a biocide is not related linearly Staphylococcus aureus, Bacillus subtilis, and E. coli. Only
to its activity; a concentration exponent is involved in one of the 51 derivatives tested was marginally active
the relationship. In many cases a small decrease in against P. aeruginosa.
concentration will result in a notable decrease in activity. The physiological state of cells and the nature of
In an in situ biocide evaluation study we found that the habitat can lead to considerable variation in the
the dominant planktonic survivor after 48 hours was a susceptibility of bacteria to bactericides. The composition
species most effectively killed by the relevant biocide of the bacterial cell envelope does change as a response
under laboratory pure culture conditions. An example to available or limiting nutrients, so that the barrier
is dichlorophen, which killed 99.94% of Pseudomonas properties of the envelope are affected. Exposure to
stutzeri at 50 ppm and yet left this species the dominant subinhibitory concentrations of bactericides can lead
isolate after 48 hours (43%) in the cooling system. Also, to phenotypic adaptation, resulting in a resistant cell
608 BIOFOULING: CHEMICAL CONTROL OF BIOFOULING IN WATER SYSTEMS

population. In E. coli, certain proteins induced by heat 3,000-fold more resistant to hypochlorous acid and 2 to
or starvation stress also confer resistance to H2 O2 and to 100 times more to monochloramine than were unattached
UV light. Most bactericide-resistance is due to adaptation, cells. Pseudomonas aeruginosa growing as a biofilm has
and the resistant phenotype is mostly lost upon removal been found 20 times as resistant to tobramycin as are
of the bactericide. planktonic cells. A number of reasons for the increased
biocide resistance of biofilm bacteria have been put
Efflux Systems. Bacteria can actively pump compounds forward, but none adequately explains the phenomenon.
out of the cell via membrane efflux systems. Only one type The current arguments are addressed adequately by
of bactericide-efflux system has been described to date, the Stewart (56) and Lewis (10).
QAC efflux system of S. aureus. This efflux system is coded
by two gene systems. The genes qacA and qacB encode for
CONCLUSION
a high level of resistance, and qacC and qacD encode for a
low level of resistance. qacC and qacD are further identical
to the ebr gene encoding for resistance to ethidium bromide Biofilms are problematic in a range of industrial
in S. aureus, explaining why resistance to QAC is often environments in which large areas of submerged surfaces
concurrent with resistance to ethidium bromide. The are exposed to relatively high nutrient fluxes, providing
qacA gene codes for a 50-kDa protein, which mediates niches for the formation of biofilms. These biofilms also
energy-dependent efflux of both benzalkonium chloride promote corrosion of ferrous metals by the concerted
and ethidium bromide. The qacC gene also mediates metabolic activity of a number of biofilm-associated
energy-dependent efflux of benzalkonium chloride and bacterial types, a process collectively termed microbial
ethidium bromide. Two different but isofunctional gene influenced corrosion (MIC). The use of biocides to control
systems appear to have evolved in S. aureus. biofouling in industrial water systems is still an
accepted practice, although higher levels of environmental
Enzymatic Degradation of Biocides. Resistance to awareness and tighter legislation have placed increased
antimicrobial agents can be caused by enzymes trans- pressure on the water treatment industry to seek
forming the bactericide to nontoxic form. The phenomenon alternative means of control. As the costs attributable to
is usually investigated from the biodegradation point of MIC and biofouling are high, effective control of bacterial
view, that is, the biodegradation of toxic pollutants. A host activity and numbers in industrial aqueous environments
of aromatic, phenolic, and other compounds toxic to many is essential.
bacteria (some of which are employed as bactericides) The primary target of a biofouling control strategy
can be degraded by certain bacteria. The topic has been should always be the biofilm-associated microorganisms
reviewed by various authors and the current literature as these are the catalysts of MIC and impact negatively on
reports extensively on biodegradative pathways (55). system operation. The most common approaches currently
Two types of enzyme-mediated resistance mechanisms available to control biofouling include the use of:
have been documented, that is, heavy-metal resistance
and formaldehyde resistance. Resistance to heavy metals • Biocide to kill biofilm-associated microorganisms;
includes resistance to the following: mercury, antimony, • Alternating currents to enhance the efficacy of
nickel, cadmium, arsenate, cobalt, zinc, lead, tellurite, biocides against biofilms;
copper, chromate, and silver. Detoxification is usually • Dispersants to dislodge cells from the surface;
by enzymatic reduction of the cation to the metal. • Enzymes to hydrolyze biofilm-associated polymers;
Whereas some heavy-metal resistance genes are carried
• Physical processes for biofilm removal; and
on plasmids, others are chromosomal. The resistant
phenotype is usually inducible by the presence of the • Ultrasonic sound to dislodge and kill microorganisms.
heavy metal. Some heavy metals induce resistance to a
It is clear that an integrated biofouling management
broader spectrum of heavy metals. Arsenate, arsenite, and
strategy is required to control biofilms and associated
antimony, for example, induce resistance to one another
problems in industrial water systems.
in E. coli.
The detoxification of formaldehyde by P. aeruginosa
and P. putida has been studied extensively. Formalde- BIBLIOGRAPHY
hyde is reduced by an NAD+gluthathione-dependant
dehydrogenase, giving formaldehyde NAD+ oxidoreduc- 1. T. E. Cloete, V. S. Brözel, and A. Von Holy, Int. Biodeteriora-
tase. This enzyme is probably plasmid-encoded, and tion Biodegradation 29, 299–341 (1992).
appears to be constitutively expressed. Resistance to most 2. J. W. Costerton et al., Annu. Rev. Microbiol. 49, 711–745
formaldehyde-releasing condensates of formaldehyde is (1995).
therefore also caused by formaldehyde dehydrogenase 3. H. G. Hoppe, Attachment of bacteria: advantage or dis-
activity. advantage for survival in the aquatic environment, in
K. C. Marchall ed., Microbial Adhesion and Aggregation,
Springer-Verlag, Berlin, Germany, 1984, pp. 283–301.
Biofilm-Associated Resistance
4. J. R. Lawrence, P. J. Delaquis, D. R. Korber, and D. E. Cald-
Bacteria in biofilms are much more protected from well, Microb. Ecol. 14, 1–14 (1987).
bactericidal action than are planktonic bacteria. In a 5. W. F. McCoy, J. D. Bryers, J. Robbins, and J. W. Costerton,
recent study, biofilm bacteria were found to be 150- to Can. J. Microbiol. 27, 910–917 (1981).
 BIOFOULING: CHEMICAL CONTROL OF BIOFOULING IN WATER SYSTEMS 609

6. Y. Igarashi and S. Watanabe, J. Ind. Microbiol. 9, 91–96 35. H. W. Rossmoore and M. Sondossi, Adv. Appl. Microbiol. 33,
(1992). 223–275 (1988).
7. E. S. McLeod et al., Syst. Appl. Microbiol. 21, 297–305 (1998). 36. T. J. Franklin and G. A. Snow, Biochemistry of Antimicrobial
8. I. U. Schraufstätter et al., J. Clin. Invest. 85, 554–562 (1990). Action, 3rd ed. Chapman & Hall, London, 1982.
9. Z. Dawood and V. S. Brözel, J. Appl. Microbiol. 84, 929–936 37. P. Gilbert and M. R. W. Brown, J. Bacteriol. 133, 1066–1072
(1998). (1978).
10. W. Lee, Z. L. Lewandowski, P. H. Nielsen, and W. A. Hamil- 38. P. J. Collier, P. Austin, and P. Gilbert, Int. J. Pharm. 74,
ton, Rev. Biofouling. 8, 165–194 (1995). 195–206 (1991).
11. A. D. Russell and I. Chopra, Understanding Antibacterial 39. P. J. Collier et al., J. Appl. Bacteriol. 69, 578–584 (1990).
Action and Resistance, Ellis Horwood, New York, 1990. 40. J. M. Lo, J. C. Yu, F. I. Hutchison, and C. M. Wai, Anal.
12. V. S. Brözel and T. E. Cloete, Water SA. 17, 57–66 (1991). Chem. 54, pp. 2536–2539 (1982).
13. V. S. Brözel and T. E. Cloete, CAB Biodeterioration Abstracts. 41. Y. Endo, T. Tani, and M. Kodama, Appl. Environ. Microbiol.
7, 387–395 (1993). 53, 2050–2055 (1987).
14. V. S. Brözel and T. E. Cloete, Water SA. 18, 87–92 (1992). 42. D. L. Price, A. D. Sawant, and D. G. Ahearn, J. Ind. Micro-
15. W. L. Cochran, G. A. McFeters, and P. S. Stewart, J. Appl. biol. 8, 83–90 (1991).
Microbiol. 88, 22–30 (2000). 43. P. M. Woodcock, Biguanides as industrial biocides, in
16. G. Bognolo, G. R. John, and J. G. Evans, Surfactants and the K. R. Payne, ed., Industrial Biocides, John Wiley & Sons,
environment — some recent developments, in D. R. Karsa, Chichester, U.K., 1988, 19–36.
ed., Industrial Applications of Surfactants 3rd ed., The Royal 44. P. Wood, M. Jones, M. Bhakoo, and P. Gilbert, Appl. Environ.
Society of Chemistry, Cambridge, U.K., 1992, pp. 23–28. Microbiol. 62, 2598–2602 (1996).
17. W. Paulus, Development of microbiocides for the protec- 45. S. A. Blenkinsopp, A. E. Khoury, and J. W. Costerton, Appl.
tion of materials, in D. R. Houghton, R. N. Smith, and Environ. Microbiol. 58, 3770–3773 (1992).
H. O. W. Eggins eds., Biodeterioration, Elsevier Applied Sci- 46. J. W. Costerton et al., Antimicrob. Agents Chemother. 38,
ence London, U.K., 1987, pp. 1–19. 2803–2809 (1994).
18. M. Wainwright, Structure and biology of bacteria relevant to 47. J. F. Brisou, Biofilms, Methods for Enzymatic Release of
the action of disinfectants, in K. R. Payne ed., Industrial Microorganisms, CRC Press, Boca Raton, Fla., 1995.
Biocides, John Wiley & Sons, Chichester, U.K., 1988, 48. I. W. Sutherland, FEMS Microbiol. Rev. 16, 323–347 (1995).
pp. 1–18.
49. C. Johansen, P. Falholt, and L. Gram, Appl. Environ. Micro-
19. P. M. Woodcock, Biguanides as industrial biocides, in biol. 63, 3724–3728 (1997).
K. R. Payne ed., Industrial Biocides, John Wiley & Sons,
50. R. W. Lutey, Process cooling water, in H. W. Rossmoore, ed.,
Chichester, U.K., 1988, pp. 19–36.
Handbook of Biocides and Preservative Use, Blackie Academic
20. H. Nikaido, Nonspecific and specific permeation channels of and Professional, London, U.K, 1995, pp. 50–82.
the Pseudomonas aeruginosa outer membrane, in E. Galli,
51. J. H. Paul and W. H. Jeffrey Can. J. Microbiol. 31, 224–228
S. Silver, and B. Witholt, eds., Pseudomonas: Molecular
(1985).
Biology and Biotechnology, American Society of Microbiology,
Washington, D.C., 1992, pp. 146–153. 52. G. F. White and N. J. Russel Biodegradation of anionic
surfactants, in J. Tyman, H. P. Cambridge ed., Surfactants
21. H. Ahern, ASM News 59, 119–122 (1993).
in Lipid Chemistry: Recent Synthetic and Biodegradative
22. S. B. Farr and T. Kogoma, Microbiol. Rev. 55, 1309–1316 Studies, The Royal Society of Chemistry, London, U.K., 1992,
(1991). pp. 99–122.
23. K. H. Wallhäuser, J. Ind. Microbiol. 7, 105–116 (1991). 53. P. Wood et al., J. Appl. Microbiol. 84, 1092–1098 (1988).
24. A. D. Russell, J. A. Furr, and J. Y. Maillard, ASM News 63, 54. P. Gilbert and N. Wright, Non-plasmidic resistance towards
481–487 (1997). preservatives of pharmaceutical products, in R. G. Board,
25. P. A. Hyslop et al., J. Biol. Chem. 263, 1665–1675 (1988). M. C. Allwood, and J. G. Banks, eds., Preservatives in the
26. J. M. Albrich, C. A. McCarthy, and J. K. Hurst, Proc. Nat. Food, Pharmaceutical and Environmental Industries, Black-
Acad. Sci. 78, 210–214 (1988). well Scientific Publications, Oxford, U.K., 1987 pp. 255–279.
27. W. R. Barrette, Jr., D. M. Hannum, W. D. Wheeler, and 55. L. P. Wackett and C. D. Hershberger, Biocatalysis and
J. K. Hurst, Biochemistry 28, 9172–9178 (1989). Biodegradation: Microbial Transformation of Organic Com-
28. B. Pietersen, V. S. Brözel, and T. E. Cloete, Water SA. 21, pounds, ASM Press, Washington, D.C., 2001.
173–176 (1995). 56. P. S. Stewart, Antimicrob. Agents Chemother. 40, 2517–
29. J. T. Marais and V. S. Brözel, Br. Dent. J. 187, 154–158 2522 (1996).
(1999). 57. M. Heinzel, The phenomena of resistance to disinfectants and
30. V. Majtán, L. Majtánová, and D. Mlynarcı́k, Drug Res. 45, preservatives, in K R Payne ed., Industrial Biocides, John
198–199 (1995). Wiley & Sons, New York, 1988, pp. 52–67.
31. L. Guérin-Méchin, F. Dubois-Brissonnet, B. Heyd, and J. Y. 58. W. P. Iverson, Adv. Appl. Microbiol. 32, 1–36 (1987).
Leveau, J. Appl. Microbiol. 87, 735–742 (1999). 59. K. Lewis, Antimicrob. Agents Chemother. 45, 999–1007
32. R. M. E. Richards and R. H. Cavill, Microbios 29, 23–31 (2001).
(1980). 60. Von Holy and T. E. Cloete, S. Afr. J. Science 84, 17–19 (1988).
33. K. A. Fitzgerald, A. Davies, and A. D. Russell, Microbios 70, 61. D. Allsop and K. J. Seal, Introduction to Biodeterioration,
77–91 (1992). Edward Arnold, London, U.K., 1986.
34. P. Broxton, P. M. Woodcock, F. Heatley, and P. Gilbert, J. 62. V. S. Brözel and T. E. Cloete, Biodeterioration and Biodegra-
Appl. Bacteriol. 57, 115–124 (1984). dation 9, 108–113 (1993).
610 BIOFOULING IN THE MARINE ENVIRONMENT

BIOFOULING IN THE MARINE ENVIRONMENT their ability to attach to surfaces and their interactions
with other organisms are examined.
JAMES S. MAKI
Marquette University
THE BIOFOULING ‘‘SEQUENCE’’
Milwaukee, Wisconsin
RALPH MITCHELL Fouling has been traditionally separated into three types:
Laboratory of Microbial Ecology molecular, microbial, and macrofouling (3). The classical
Harvard University and often reported view of biofouling has been described
Cambridge, Massachusetts
as a successional process (Fig. 1a; 4,5): first, involving a
rapid (instantaneous) adsorption of organic and inorganic
Biofouling has been defined as the modification of molecules to form a conditioning film (molecular fouling);
an immersed, artificial structure by organisms, which second, a rapid formation of a biofilm (microbial fouling)
results in a deterioration in its performance (1). Its involving formation of a bacterial film followed by algal
occurrence is a complex mixture of physical, inorganic, and and protozoan attachment; and third, invertebrate and
organic chemical and biological phenomena (1). Basically, macroalgal fouling (macrofouling). The idea of a succession
biofouling means that organisms (both microorganisms in biofouling implies some sort of causality between the
and macroorganisms) are acting out their normal life levels of fouling. On the one hand, unless causality can be
cycles, which include a sessile/attached phase, and doing it demonstrated, it cannot be assumed (6), and on the other
where humans do not want them to do so. Importantly, it hand, it should not be assumed that causality between
has been observed that biofouling communities on artificial stages does not exist (7).
structures can be quite different from communities on The aforementioned sequence of events in biofouling
natural surfaces (2). The consequences of biofouling are has been observed and reported many times largely due to
ultimately realized financially. In practical terms, they the numbers of the components in the different levels and
can be measured as increased fuel consumption by a thus, their availability to attach to a surface (1,7). Thus,
fouled ship or submarine, deterioration of navigational as an alternative to causal succession, biofouling can be
buoys, blockage of pipes transferring seawater, length considered to be seasonal or probabilistic (3). Probabilistic
of downtime it takes to clean a fouled surface, or some fouling (Fig. 1b) depends on the number of molecules and
other loss of performance. In addition, because fouling and organisms that could attach to or colonize a surface.
corrosion are intimately related, the consequences of the There are billions of organic and inorganic molecules
former go beyond the simple accumulation of organisms per milliliter, around a million bacteria per milliliter,
on a surface to include the deterioration of the surface thousands of algae per milliliter, and possibly only a few
itself. Microorganisms play a variety of important roles larvae or spores per liter. For the most part, the number
in biofouling. They are both a component of a fouling of molecules and bacteria per milliliter are ‘‘relatively’’
community and may stimulate, inhibit, or have no effect constant over time and so are readily available to attach
on the subsequent attachment of other organisms. Both to any new surface or one cleaned by a disturbance at

(a) 3 Invertebrate larvae (b)

Marine snow

2b Unicellular algae
Algae Particulates

2a Bacteria Bacteria Invertebrate larvae

? ? ?
1

Conditioning film 1

Surface Conditioning film

Surface
Figure 1. Diagrams of two models for biofouling. (a) The classical view of successional fouling
on a substratum. Each fouling stage is a step that precedes the following stage and causality
between levels is implied. (b) A diagrammatic representation of substratum fouling based on the
probability that a particular fouling component will encounter the substratum. After formation
of the conditioning film, further colonization will depend on which type of organism encounters
the substratum. In the absence of a substratum, molecules and organisms may attach to each
other and participate in the formation of marine snow. Adapted from A. S. Clare, D. Rittschof,
D. J. Gerhart, and J. S. Maki, Invertebr. Reprod. Dev. 22, 67–76 (1992).
 BIOFOULING IN THE MARINE ENVIRONMENT 611

any time of the year. Although there should also be MICROBIAL FOULING-BIOFILM FORMATION
unicellular algae present throughout the year they will
probably undergo seasonal quantitative and qualitative Bacterial biofilms have been defined as matrix-enclosed
fluctuations. On the macrofouling level, some organisms bacterial populations adherent to each other and/or
reproduce throughout the year, but many reproduce surfaces or interfaces (24). In the marine environment,
seasonally (8). Thus, their arrival late in the fouling biofilm communities contain a variety of other populations
‘‘sequence’’ may be more due to seasonal reproduction, of microorganisms, which also colonize solid substrata
and therefore a probabilistic relationship, than to any including unicellular algae, fungi, and protozoa. Thus, the
causal succession. definition of a marine biofilm will be modified in this report
to be matrix-enclosed populations of microorganisms
MOLECULAR FOULING adherent to each other and/or surfaces or interfaces. A
biofilm should be considered to be more than a layer of cells
Molecular fouling forms what is called a conditioning film and slime coating a solid substratum. It possesses dynamic
on a surface exposed to an aquatic environment (See structure and function. It is the nature of the various
CONDITIONING FILMS IN AQUATIC ENVIRONMENTS, this Encyclo- structures and functions and their changes, which may be
pedia). Macromolecules that adsorb to the surface include involved in the attachment of other microorganisms and
proteins, glycoproteins, proteoglycans, and polysaccha- macrofoulers. Examples of some of the microorganisms
rides (9,10). Specificity between the nature of the substra- found in marine biofilms follows.
tum and the chemical identity and quantity of adsorbed
materials in the early stages of macromolecular adsorp- Bacterial Colonization
tion also appears to occur (11–13). The surface free energy Bacteria will be among the initial colonizers of a new
of the substratum can also affect the molecules. On high- surface for reasons stated earlier (Fig. 2a). Arrival at the
energy surfaces the organic molecules are flattened and vicinity of a surface will usually be by hydrodynamic
bind strongly, while on low energy surfaces the molecules means. Once within the viscous boundary layer of
are more weakly bound, but may have greater thick- still water near the surface, positive and negative
ness (9). As pointed out in the preceding text, there are
billions of organic and inorganic molecules in a milliliter
of water. In the formation of the conditioning film on a
(a) Attachment of cells
surface exposed to the bulk phase water, not all organic
species present in the water column initially adsorb (11).
However, after a certain period, differences disappear.
Because the air/water interface already contains a layer
of surface-active organic molecules (14), molecular fouling Spores Bacteria
that occurs on surfaces exposed to wave action is different
and more complex (3).
The presence of a conditioning film can change
both the charge and the free energy characteristics
of the substratum (15,16). After the adsorption of the (b) Sporulation, reproduction
conditioning film, the substratum will assume the net & additional colonization
surface charge and the surface fee energy characteristics
of the outermost molecular groups of the film. However, Production of exopolymers &
other secondary metabolites
the surface energy may not change much because the
outermost molecules of the conditioning film reflect the
nature of the underlying substratum (17).
The presence of a conditioning film can have two
potential effects on attaching microorganisms. First,
facilitation between adsorbed molecules and attaching
microorganisms has been reported (18–20). However, (c) Formation of microcolonies
adsorbed molecules have also been reported to inhibit
attachment by microorganisms (12,21). Second, adsorbed
molecules may provide attaching microorganisms (e.g.,
bacteria) with a carbon source for growth. Even molecules
that are toxic in solution may become nontoxic when
they are immobilized at a surface and provide a suitable
substrate for bacteria (22). Attached bacteria on a surface
may be able to metabolize adsorbing organic molecules
very rapidly (23). In summary, the possibility exists that
the nature of the substratum will influence the initial Figure 2. Steps involved in the formation of microbial biofilms in
adsorption of components of the conditioning film, and the marine environment. (a) Attachment of cells. (b) Sporulation,
these in turn may influence the types of microorganisms reproduction, and additional colonization. (c) Formation of
that initially attach. microcolonies. See text for details.
612 BIOFOULING IN THE MARINE ENVIRONMENT

chemotaxis may be important to motile bacteria for these regulatory systems may not function with the same
their transport to the surface. For nonmotile bacteria, efficacy in all habitats as they do in the laboratory.
transport through the boundary layer may involve Marine bacteria have been shown to produce a wide
Brownian motion or the hydrophobicity of the cell range of organic molecules that have antibacterial, antial-
surface. As described by Zobell (25) and later defined gal, antifungal, antiviral, and anti-inflammatory activity,
by Marshall and coworkers (26), bacteria approaching a as well as enzymes, enzyme inhibitors, toxins, and poly-
solid substratum have two phases of adhesion — reversible mers (52–54). Understanding how marine bacteria are
and irreversible or firm. Reversibly attached bacteria adapted to life in the various habitats of the sea (includ-
do not come in direct contact with the substratum ing biofilms), and how these may affect the production of
because of electrostatic repulsion, but are capable of secondary metabolites, may help understand their biosyn-
scavenging nutrients from the surface (27). The bacteria thetic potential (55). These secondary metabolites may
must overcome the repulsion to be firmly attached. have a wide range of effects on other potential fouling
Firm adhesion of a bacterium to a solid substratum is micro and macroorganisms.
mediated by cell surface and extracellular polymers (28),
including polysaccharides (29,30), proteins (31,32), and Diatoms and Other Unicellular Algae
lipopolysaccharides (33,34). It is also likely that bacteria Of the unicellular algae that may be involved in biofilm
can utilize more than one mechanism to attach to different communities, diatoms are usually among the primary
surfaces (31,32). algal colonizers of new surfaces (56–59). Because of their
When a planktonic bacterium arrives in the vicinity of a size and mass, diatoms (and many other unicellular
solid substratum it is faced with a microenvironment that algae) in aquatic systems are transported to surfaces
has different conditions. These conditions can act to stim- through hydrodynamic means, except under completely
ulate changes in morphology (35), and turn on previously calm conditions. Among the diatoms are a number of
unexpressed genes (36–38). A bacterium may respond to different attachment systems (Table 1), but all have
more than one type of signal when near a surface (39). one thing in common — the presence of mucilaginous
This may induce the production of new proteins (40) and material (56,60). The polymers have a high molecular
stimulate the synthesis of exopolysaccharides (Fig. 2b; 41). weight, and are largely acidic, with lesser amounts of
Conditions that are associated with a surface and may be neutral polysaccharides and small amounts of protein
responsible for inducing genetic responses by an attach- in certain species (57,60). The extracellular polymers
ing bacterium, either individually or in concert, include are involved in both gliding of diatoms and their
changes in gas diffusion, pH, cell density, metabolite adhesion (61,62).
accumulation, nutrients, surfactants, viscosity, osmolar- Once it has arrived at a surface, a motile diatom (e.g.,
ity, water activity, and inorganic ions (42). Amphora coffeaeformis) can move across the surface by
The formation of a bacterial biofilm is a complex means of its raphe and can respond chemotactically to
process and can be influenced by a number of factors (43). nutrient gradients (58,59,63). Ca+2 plays a role in both the
Once on the surface, factors including motility mediated initial adhesion and the motility of diatoms (59,64,65).
by flagella and twitching, outer membrane proteins, The role of Ca+2 is probably dual as it is involved
and exopolysaccharides have been demonstrated to be intracellularly in transport and secretion of the adhesive
important in establishing cell-cell interactions for biofilm and extracellularly by cross-linking the adhesive (63).
formation by several bacteria in laboratory studies (44). Adhesion of A. coffeaeformis can occur under either
Microscopic evidence has indicated that the three- light or dark conditions and is not inhibited by the
dimensional structure of a bacterial biofilm can be
quite complex (Fig. 2c; 24,43). In addition, bacteria may
Table 1. Characterization of Adhesion Mechanisms Used
communicate between their own species and perhaps
by Diatomsa
others through the use of N-acyl-L-homoserine lactones
(AHL, 45–48). These small molecules act as extracellular Adhesive Type Description Examples
signals that, when accumulated in the presence of
increased cell density, activate transcription and modulate Extracellular Pads peripherally Synedra affinis
physiological processes. They have been demonstrated to pads located or distinct
pads at polar end or
influence the three-dimensional structure of a bacterial
at corners of cells
biofilm (49). A number of bacteria have been demonstrated
Extracellular Adhesive secreted Achnanthes sp.
to use this type of cell-to-cell communication, which stalks through slits in Licmophora sp.
raises the possibility of one bacterium receiving analogous valve region
signals produced by another species. These signals may Extracellular Mucilage secreted Amphora sp.
actually be directed toward a different physiological encapsula- through raphe
response. AHL activity has been demonstrated to occur in tion/Raphe fissures and large
natural biofilms (50). Thus, some biofilm physiology and secretions punctae
phenotypic expression may be the result of intercellular Extracellular Sheath mucilage Navicula sp.
communication between bacteria of same and/or different sheaths released all over cell
through small
species. However, it is also apparent that some marine
punctae
algae produce compounds, which interfere with AHL
regulatory systems (51). Thus, in the natural environment, a
Adapted from Chamberlain (56) and Jones and coworkers (57).
 BIOFOULING IN THE MARINE ENVIRONMENT 613

photosystem II inhibitor, DCMU, but does require energy MACROFOULING: INVERTEBRATE LARVAE
and protein synthesis (63). Once attached, the strength of
the adhesion of diatoms to a surface is species-dependent Sessile invertebrates involved in fouling generally have a
and will increase over time (66). Stalked diatoms appear free-living planktonic stage for dispersal and development
to stay attached to substrata at higher shear stress while back into the adult stage (Fig. 3). The planktonic larvae
diatoms attached by means of an adhesive pad do not (66). of many sessile invertebrates are recruited to surfaces by
The adhesion of the unicellular alga Chlorella has various biochemical stimuli (78). For the purposes of this
been reported to be stimulated by polymers recovered discussion, recruitment will be considered to be a two-
from Chlorella exudate, bacterial cultures, and seawater stage process — first, an exploratory phase of repeatable
that were adsorbed to glass surfaces (18). This interaction behavior (settlement), and second, an irreversible develop-
appeared to be mediated by lectins or lectin-like mental event that occurs once and includes secretion of a
activity (67). The growth of a number of diatoms and permanent adhesive (metamorphosis) (Fig. 3; 79–81). The
other unicellular algae was demonstrated to be inhibited stimuli can be broadly divided into two types — conspecific
by marine bacteria in laboratory studies (68,69). Thus, (or gregarious) and associative (79,82). Conspecific cues
both positive and negative interactions between bacteria are those that induce the larvae to settle and metamor-
and algae may occur in biofilms. phose on or near their own species, while associative
cues are those produced by other organisms (e.g., those
Fungal Colonization in biofilms) and provide information to the larvae on the
Fungi are widely distributed in the marine environment choice of habitat. Both positive and negative cues or stim-
and include saprophytic and parasitic types (70,71). uli are probably important for habitat choice. Although
Marine fungi can also be members of biofilm communities. there is increasing evidence that some larvae can respond
Colonization of surfaces by fungi usually takes place to soluble biochemical cues in either settlement or meta-
through attachment of their spores. Many marine morphosis (83–91), many larvae will only metamorphose
fungal spores have appendages (72,73). The presence of in response to specific stimuli adsorbed onto surfaces (92).
appendages assists the ability of the spores to remain The length of time larvae remain in the planktonic stage
in the water column for longer periods of time than of development varies among species. Some, including
those without appendages (74). Because the spores are spirorbid polychaetes, are nonfeeding and generally only
nonmotile, they are transported to surfaces through remain planktonic for some hours before settlement and
hydrodynamic forces. Once in the vicinity of a surface, metamorphosis (93). Others, including barnacles, have a
spores may initially become entrapped (which may be number of larval feeding stages (naupliar) followed by a
assisted by the presence of appendages) or attach to nonfeeding larval stage (cyprid) that undergoes settlement
the surface (75). Attachment is mediated by several and metamorphosis, and can be in the planktonic stage
methods (Table 2), all of which employ some mucilaginous for a considerable length of time (94). Once the larval
adhesive (75,76). The strength of the initial adhesion is stage responsible for settlement and metamorphosis is
variable between species and can (1) remain constant, produced, hydrodynamics can affect its availability to
(2) increase and then level off, or (3) show variable increase arrive at a given substratum (95) and the settlement
with time (77). Eventually the spores will germinate and and attachment of larvae at a surface (96,97). For sessile
adhesion will be further improved by production of a organisms, such as fouling invertebrates, space is a
mucilaginous sheath around both the spore and the germ resource (defined as something that can be consumed or
tube (76). made less available to other organisms, 98). As a result, it
is something that can be competed for by organisms in the
fouling community (98). Additionally, predation also plays
Table 2. Mechanisms of Spore Attachment by Marine
Fungia

Type Examples
Dispersal
Release of drop of Kohlmeyeriella tubulata
mucilage from polar Lulworthia medusa Planktonic
appendages larvae
Disc or padlike Ondiniella torquata
attachment Ceriosporopsis circumvestita
Coiling of sticky Halosarpheia retorquents
appendages around
objects Sessile
Straplike sticky Haligena elaterophora adult
appendages Remispora maritima Metamorphosis Settlement
Remispora pilleata cue detection
substratum Selection
Hairlike sticky Nereiospora comata
appendages or pads Crinigera maritima
Substratum
composed of fibrillar
components Figure 3. A generalized life history of a sessile marine
invertebrate. Adapted from J. R. Pawlik, Oceanogr. Mar. Biol.
a
Adapted from Rees and Jones (75). Annu. Rev. 30, 273–335 (1992).
614 BIOFOULING IN THE MARINE ENVIRONMENT

a role in the community structure (98). Rittschof (3) has and/or novel bacterial compounds that inhibit or stimulate
pointed out that it is important to remember that many larval attachment. However, it should be pointed out that
invertebrate foulers have short life cycles. In fact, after stimulation or inhibition of larvae by a biofilm in the field
they attach, the animals reach sexual maturity in about a may turn out to be the result of actions by a consortium
month, and after four months the whole cohort is dead (3). of microorganisms, and focusing on films composed of a
After that, their remains on the substratum may act as a single bacteria may probably not duplicate that result.
surface for other fouling organisms. There is no way that the film produced by a single species
can mimic the complex structure and function of a natural
Biofilms and Invertebrates biofilm consisting of a number of microbial populations.
Natural films composed of a variety of microorgan-
The role of biofilms in invertebrate settlement and/or isms in laboratory tests have been reported to pro-
metamorphosis is complex (see reviews 99–101). Although
mote settlement and metamorphosis of spirorbid poly-
a larva may not necessarily require a film of microor-
chaetes (93,102,113), bryozoans (107,114–116), and bar-
ganisms to settle and metamorphose (102), it is likely to
nacles (117,118). Factors such as composition of the
encounter a film on almost every surface that it comes
biofilm (117) and its age appear to be of importance
in contact with during settlement. This occurs because
to some types of larvae (118). Natural biofilms have
although larvae may only be produced seasonally as
been demonstrated to have very different effects on the
stated earlier, microorganisms will attach continuously.
attachment of different types of larvae. For example,
As the previous information indicates, a marine microbial
Crisp and Ryland (102) reported that while the bry-
biofilm is quite complex, with potentially a wide variety
ozoan Bugula flabellata was inhibited by the presence
of organic molecules that may act as associative cues for
of biofilms, the spirorbid polychaete Spirorbis borealis
larval attachment. However, the abundance of any one cue
was stimulated. Similar results were reported for larvae
will probably vary over space and time in any biofilm. The
of B. flabellata and larvae of the ascidian Ciona intesti-
biofilm is particularly important for those organisms that
nalis (119).
respond to surface-associated stimuli.
Films composed of individual strains of bacteria can
Field studies have demonstrated that invertebrate
have varying effects on larval attachment (108,120–127).
attachment to substrata, with or without biofilms, is
In fact, individual strains of bacteria isolated from surfaces
complex and potentially governed by a number of vari-
may have an effect on larval settlement and metamor-
ables (103–106). The nature and beginning wettability of
phosis, which was not expressed by the community it
the substratum is important for some of the initial macro-
was isolated from. For example, the spirorbid polychaete,
foulers (105,106). As time passes, the wettability of the
Janua brasiliensis, attaches to a variety of different sub-
surface changes and is probably not a major influence on
strata, and biofilms are an important component to the
continued recruitment (106). Films of microorganisms can
polychaetes’ response to a surface (120,128). One surface
change the wettability of a substratum (107–109) and can
that J. brasiliensis is commonly found on is the green
modify the response of some larvae to surface energy (e.g.,
alga Ulva lobata. When a number of bacteria were iso-
bryozoans; 107,108).
lated from Ulva and films of these bacteria were tested
Recent field studies used natural films prepared by
for their effect on attachment of J. brasiliensis larvae,
enclosing panels in a mesh that allowed development
it was found that not all bacteria were effective induc-
of a biofilm but prevented attachment of macroorgan-
ers of larval attachment (120). This suggests that (1) the
isms (110–112). Comparisons were made of filmed and
effect of a single strain of bacteria can be masked by
unfilmed substrata, age of films, and films from different
the biofilm as a whole and that (2) just because a natu-
locales on the fouling of panels by macroinvertebrates and
ral film does not show stimulation or inhibition of larval
results showed that (1) filming generally enhanced larval
attachment, it does not follow that it does not contain
attachment, although not for all type of organisms at all
strains of bacteria that could do either. Laboratory tests
times (110); (2) some larvae responded positively to film
have also shown that a single bacterium can have a
age, some did not, and some preferred less filmed sur-
variety of effects on the settlement and metamorpho-
faces (111); and (3) generally, larvae could not distinguish
sis of different invertebrate larvae (129). For example,
between films developed in different areas, but were more
the marine bacterium Halomonas marina (formerly, Pseu-
attracted to more heavily-filmed substrata (112). From
domonas marina and Deleya marina) has been reported
these data, it would appear that each type of invertebrate
to stimulate the settlement and metamorphosis of the
has its own response to biofilms, that there may also be
variability in that response within any one group of larvae, spirorbid polychaete J. brasiliensis (120,130), inhibit or
and that biofilm density and/or age may play a role in the stimulate (see later) attachment of the barnacles Bal-
larval response. anus amphitrite (109,123,131,132) and Balanus improvi-
sus (127), inhibit the temporary adhesion of the barnacle
Elminius modestus (133), and inhibit the attachment of
Laboratory Observations
the larvae of the bryozoan Bugula neritina (108). This
Many laboratory observations of larval interactions with implies that a single species of bacterium can possess
biofilms have been performed without flowing water different biochemical cues for larval settlement and meta-
(although not always) to examine single strains of bacteria morphosis. Some of these cues could be stimulatory to
and their effects on the type of larva under study. certain larvae, whereas other cues could be inhibitory to
This is particularly useful if the goal is to find bacteria other larvae.
 BIOFOULING IN THE MARINE ENVIRONMENT 615

It has been suggested that there is an interaction MACROFOULING: ALGAL SPORES

between the substratum and the bacterium, which influ-
ences the response of the larvae (127,131–134). Films Macroalgae are also major fouling organisms in the marine
of single species of bacteria (e.g., H. marina) or their environment. The macroalgae that contribute to biofouling
extracellular products attached or adsorbed onto one belong to Chlorophyta (green algae), Phaeophyta (brown
substratum are inhibitory, while on another substra- algae), and Rhodophyta (red algae). These may develop
tum they are not inhibitory or may even be stimula- sequentially in the environment with green and brown
tory compared to no-film controls. Bacteria have been algae occurring first followed by red (150). Mechanisms
demonstrated to have the capability of using separate for the dispersion and colonization of new substrata by
mechanisms to attach to different surfaces (31,32). This macroalgae in the marine environment include vegetative
may cause changes in the surface polymers that are fragmentation (151) and young sporophytes (152). How-
expressed by an attached bacterium (i.e., production of ever, most of the dispersal and colonization of macroalgae
a new extracellular polymer or rearrangement of an occurs through the production of reproductive spore bod-
existing one to mediate adhesion). Polymers with dis- ies formed in sporangia (153). There are two major types
tinct characteristics may be produced by a bacterium in of spore bodies: (1) sexual spores (known as gametes)
response to attachment to different substrata (134). It and (2) asexual spores (known as zoospores or swarm-
is also important to remember that the substratum can ers for green and brown macroalgae and carpospores and
cause new genes to be expressed (42). Thus, the response tetraspores in red algae). The occurrence of two major
of the bacterium to the substratum before, during, or types of spore bodies is the result of life cycles that
after adhesion may bring about a change in extracellu- contain both diploid (usually sporophytes) and haploid
lar polymers, which in turn is detectable by the larval (usually gametophytes) vegetative phases. The distribu-
receptors. tion of sporophytes and gametophytes may depend on how
In most cases, the stimulatory or inhibitory compo- rigorous the environmental conditions are (154), and their
nent of the individual strain of bacteria for larval set- respective release of reproductive spore bodies is both
tlement and metamorphosis has not been elucidated. A seasonal and periodic (154,155). Both sexual and asexual
number of invertebrates, some of which are considered spores are potential foulers, the latter as fused gametes or
fouling organisms, have been shown to be stimulated parthenospores (156).
to metamorphose by marine bacteria, and some of the The swarmers of green and brown algae are flagellated
inducing molecules have been isolated (Table 3). The and motile and the length of time they exhibit motility
marine bacterium Pseudoalteromonas tunicata has also varies among groups. In some cases they may swim
been demonstrated to produce a number of compounds for days (157–161). The asexual spores of red algae are
that are inhibitory to a variety of micro- and macroor- nonmotile and colonize by sinking onto substrata (162).
ganisms (125,135,136). In addition to the production of Their sinking rates are dependent on size, density, the
specific molecules, biofilms can influence larval settle- amount of mucilage present on the spore surface, and
ment and metamorphosis in other ways. The extracellular hydrodynamic conditions. Water motion is a major force in
polymers of bacteria in films provide sites for sorption the distribution of spores because the velocity of currents
or binding of metals or organic pollutants (28,137–139). is generally greater than their sinking rates (163). The
For example, tributyl tin, a very toxic compound, has motility of green and brown algal swarmers may only be of
been demonstrated to be more concentrated in a bacterial value once the swarmer is out of turbulent hydrodynamic
biofilm and inhibits attachment and metamorphosis of conditions and enters the boundary layer associated with
oyster larvae (140). Thus, the cause of inhibition of larval the surface (164).
attachment may be much more subtle than the presence Motile spores go through a selection process in
or absence of a biofilm alone. Production of extracellu- the search for substrata, which are suitable for colo-
lar polymers can also been demonstrated to cause some nization (160). Spores will settle on substrata of both
larval types (e.g., ascidian) to become ‘‘trapped’’ in films, high (hydrophilic, wettable) and low (hydrophobic, low-
and these underwent metamorphosis and contributed to wettable) surface free energy (164,165). However, sub-
recruitment (124,119). strata that are more hydrophobic appear to be more

Table 3. Examples of Larvae That Have Metamorphosis Induced by Marine Bacteria

Invertebrate Bacteria Involved Nature of Inducer Reference

Hydractinia echinata (hydrozoan) Alteromonas espejiana lipid material 141,142

Lytechinus pictus (sea urchin) unidentified unknown, <5, 000 MW 143
Cassiopea andromeda (scyphozoan) Vibrio sp. oligopeptides 144,145
Acanthaster planci (starfish) unidentified unknown 146
Aurelia aurita (scyphozoan) Micrococcaceae glycolipid 147,148
Janua brasiliensis (spirorbid polychaete) Halomonas marina unknown glycoconjugate 130
Crassostrea gigas Alteromonas colwelliana L-dopa mimetic 122
Crassostrea virginica (oysters)
Hydroides elegans (polychaete) unidentified unknown, <7, 000 MW 149
616 BIOFOULING IN THE MARINE ENVIRONMENT

attractive to some zoospores (e.g., Enteromorpha, 167). participants in the biofouling process and may be sources
Before settlement, the spores switch from a random and of cues to other organisms arriving at a surface. Biofoul-
erratic motion to a searching mode (166,167). This is fol- ing communities are dynamic both in their structure and
lowed by a spinning motion with the anterior of the function. A better understanding of these processes will
zoospore down on the substratum (167,168). Initial con- be required in order to prevent or limit biofouling and its
tact with the surface was believed to be mediated by the consequent deleterious effects on structures in the marine
flagella (153) but recent information using high-resolution environment.
video microscopy found no evidence to support this hypoth-
esis (167). Once the zoospore is on the surface the flagella BIBLIOGRAPHY
are adsorbed (159,169,170) and the spore flattens and con-
tracts against the surface (167), followed by the secretion 1. A. S. Clare, D. Rittschof, D. J. Gerhart, and J. S. Maki,
of a protein-polysaccharide adhesive (159,169,171,172). Invertebr. Reprod. Dev. 22, 67–76 (1992).
There also appeared to be gregarious settlement by Entero- 2. T. M. Glasby and S. D. Connell, Ambio 28, 595–598 (1999).
morpha zoospores although no specific mechanism was 3. D. Rittschof, Biofilms, bioadhesion, corrosion, and bio-
elucidated (167). fouling, in M. Fingerman, R. Nagabhushanam, and M.-
Nonmotile spores generally have a coating of polysac- F. Thompson, eds., Recent Advances in Marine Biotech-
charide mucilage (164,173) and this mucilage is proba- nology, vol. 3, Science Publishers, Enfield, N.H., 1999,
pp. 245–257.
bly responsible for the initial spore attachment (153).
4. R. Mitchell and D. Kirchman, in J. D. Costlow and R. C. Tip-
After initial attachment the cell components undergo
per, eds., Marine Biodeterioration: An Interdisciplinary
a rearrangement (174). Large quantities of a secondary
Study, Naval Institute Press, Annapolis, Md., 1984,
adhesive is secreted and forms a pad beneath the pp. 49–56.
spore (153,164,173).
5. L. V. Evans, in C. A. Lembi and J. R. Waaland, eds., Algae
Once the initial spore adhesion is completed, the and Human Affairs, Cambridge University Press, New York,
spores germinate and primary and secondary rhizoid 1988, pp. 433–453.
formation occurs. Primary rhizoids function to secure 6. D. J. Crisp, in J. D. Costlow and R. C. Tipper, eds., Marine
attachment during early development and secondary Biodeterioration: An Interdisciplinary Study, Naval Insti-
rhizoids grow in response to enlargement of the shoot. tute Press, Annapolis, Md., 1984, pp. 103–126.
The attachment of the rhizoids is also mediated by 7. M. Wahl, Mar. Ecol.: Prog. Ser. 58, 175–189 (1989).
secreted adhesives (175). The adhesion of at least one 8. G. W. Smeddes, in J. D. Costlow and R. C. Tipper, eds.,
type of macroalgae (Enteromorpha) that were grown on Marine Biodeterioration: An Interdisciplinary Study, Naval
substrata with a high surface free energy was found to be Institute Press, Annapolis, Md., 1984, pp. 155–160.
stronger than those grown on a low surface free energy 9. R. E. Baier, in G. Bitton and K. C. Marshall, eds., Adsorp-
substratum (176). tion of Microorganisms to Surfaces, John Wiley & Sons, New
York, 1980, pp. 59–104.
Algal Spores and Biofilms 10. M. Fletcher and K. C. Marshall, Adv. Microb. Ecol. 6,
Similar to invertebrate larvae because of the periodic and 199–236 (1982).
seasonal production of spores by macroalgae, the spores 11. B. J. Little and A. Zsolnay, J. Colloid Interface Sci. 104,
will usually encounter substrata, which possess a biofilm. 79–86 (1985).
However, little is known about how macroalgae develop 12. B. J. Little et al., J. Adhes. 20, 187–210 (1986).
on substrata with biofilms (101). Enteromorpha zoospores 13. J. S. Maki, B. J. Little, P. Wagner, and R. Mitchell, Biofoul-
attached in greater numbers to surfaces with a natural ing 2, 27–38 (1990).
multispecies film than to unfilmed surfaces (177,178). 14. J. S. Maki and M. Hermansson, in R. S. Wotton, ed., The
It was hypothesized that the presence of the film Biology of Particles in Aquatic Systems, 2nd ed., CRC Press,
increased the surface free energy, which stabilized the Boca Raton, Fla., 1994, pp. 161–182.
adhesive of the zoospore (177). However, other studies 15. P. R. Vincent and B. Vincent, in edsK. C. Marshall, ed.,
have shown that zoospores will attach in greater numbers Microbial Adhesion and Aggregation, Springer-Verlag,
Berlin, Germany, 1984, pp. 21–38.
to hydrophobic surfaces (166), suggesting that a higher
surface free energy is not the answer. Studies using 16. G. Loeb, in D. C. Savage and M. Fletcher, eds., Bacterial
Adhesion: Mechanisms and Physiological Significance,
individual strains of bacteria have demonstrated that
Plenum Press, New York, 1985, pp. 111–129.
some can be inhibitory to macroalgal spores while others
17. A. E. Meyer, R. E. Baier, and R. W. King, Can. J. Chem.
may be stimulatory (179). The exact mechanisms for these
Eng. 66, 55–62 (1987).
results have not been elucidated.
18. T. R. Tosteson and W. A. Corpe, Can. J. Microbiol. 21,
1025–1031 (1975).
CONCLUSION 19. B. R. Zaidi, R. F. Bard, and T. R. Tosteson, Appl. Environ.
Microbiol. 48, 519–524 (1984).
The preceding discussion indicates the complex nature of 20. T. R. Tosteson et al., J. Colloid Interface Sci. 104, 60–71
biofouling. Marine biofouling occurs because humans place (1985).
structures in this environment and the life cycle of many 21. A. H. L. Chamberlain and R. van Woerkom, in S. Barry,
marine organisms includes a sessile phase, which ends up D. R. Houghton, G. C. Llewellyn, and C. E. O’Rear, eds.,
on these structures. Both microorganisms and macroor- Biodeterioration VI, C.A.B. International, Farnham Royal,
ganisms are involved in this process. Microorganisms are U.K., 1986, pp. 556–561.
 BIOFOULING IN THE MARINE ENVIRONMENT 617

22. S. Kjelleberg, in D. C. Savage and M. Fletcher, eds., Bacte- 53. V. S. Bernan, M. Greenstein, and W. M. Maiese, Adv. Appl.
rial Adhesion: Mechanisms and Physiological Significance, Microbiol. 43, 57–90 (1997).
Plenum Press, New York, 1985, pp. 163–194. 54. R. M. Weiner, Trends Biotechnol. 15, 390–394 (1997).
23. M. Fletcher and K. C. Marshall, Adv. Microb. Ecol. 6, 55. P. R. Jensen and W. Fenical, J. Ind. Microbiol. 17, 346–351
199–236 (1982). (1996).
24. J. W. Costerton et al., Annu. Rev. Microbiol. 49, 711–745 56. A. H. L. Chamberlain, in J. M. Sharpley and A. M. Kaplan,
(1995). eds., Proceedings of the 3rd International Biodegradation
25. C. E. Zobell, J. Bacteriol. 46, 39–56 (1943). Symposium, Applied Science Publishers, Ltd., London, U.K.,
26. K. C. Marshall, R. Stout, and R. Mitchell, J. Gen. Microbiol. 1976, pp. 417–432.
68, 337–348 (1971). 57. A. M. Jones, R. L. Fletcher, G. F. Daniel, and E. B. G. Jones,
27. M. Hermansson and K. C. Marshall, Microb. Ecol. 11, in J. Mauchline, ed., Fouling and Corrosion of Metals in
91–105 (1985). Seawater, Scottish Marine Biological Association, 1982,
pp. 31–77.
28. A. W. Deccho, Oceanogr. Mar. Biol. Annu. Rev. 28, 73–153
58. B. Cooksey et al., in J. D. Costlow and R. C. Tipper, eds.,
(1990).
Marine Biodeterioration: An Interdisciplinary Study, Naval
29. C. Shea, J. W. Nunley, J. C. Williamson, and H. E. Smith- Institute Press, Annapolis, Md., 1984, pp. 167–171.
Somerville, Appl. Environ. Microbiol. 57, 3107–3113 (1991).
59. K. E. Cooksey and B. Wigglesworth-Cooksey, Aquat. Microb.
30. E. J. Quintero and R. M. Weiner, Appl. Environ. Microbiol. Ecol. 9, 87–96 (1995).
61, 1897–1903 (1995).
60. K. D. Hoagland, J. R. Rosowski, M. R. Gretz, and S. C. Roe-
31. M. Fletcher and K. C. Marshall, Appl. Environ. Microbiol. mer, J. Phycol. 29, 537–566 (1993).
44, 184–192 (1982).
61. J. L. Lind et al., Planta 203, 213–221 (1997).
32. J. H. Paul and W. H. Jeffrey, Appl. Environ. Microbiol. 50,
62. Y. Wang et al., Plant Physiol. 113, 1071–1080 (1997).
431–437 (1985).
63. K. E. Cooksey and B. Cooksey, in L. V. Evans and K. D.
33. S. A. Makin and T. J. Beveridge, Microbiology 142, 299–307
Hoagland, eds., Algal Biofouling, Elsevier, New York, 1986,
(1996).
pp. 41–53.
34. V. Williams and M. Fletcher, Appl. Environ. Microbiol. 62, 64. K. E. Cooksey, Appl. Environ. Microbiol. 41, 1378–1382
100–104 (1996). (1981).
35. H. M. Dalton et al., J. Bacteriol. 176, 6900–6906 (1994). 65. B. Cooksey and K. E. Cooksey, Plant Physiol. 65, 129–131
36. L. Dagostino, A. E. Goodman, and K. C. Marshall, Biofoul- (1980).
ing 4, 113–119 (1990). 66. S. Pyne, R. L. Fletcher, and E. B. G. Jones, 6th International
37. D. G. Davies, A. M. Chakrabarty, and G. G. Geesey, Appl. Congress of Marine Corrosion and Fouling Marine Biology,
Environ. Microbiol. 59, 1181–1186 (1993). Athens, Greece, 1984, pp. 99–112.
38. D. G. Davies and G. G. Geesey, Appl. Environ. Microbiol. 61, 67. T. R. Tosteson and B. R. Zaidi, in H. W. Weber and
860–867 (1995). G. D. Ruggieri, eds., Food and Drugs from the Sea Sympo-
39. L. L. McCarter, R. E. Showalter, and M. R. Silverman, Bio- sium, Marine Technology Society, Washington, D.C., 1974,
fouling 5, 163–175 (1992). pp. 457–466.
40. V. S. Brözel, G. M. Strydom, and T. E. Cloete, Biofouling 8, 68. B. R. Berland, D. J. Bonin, and S. Y. Maestrini, Mar. Biol.
195–201 (1995). 12, 189–193 (1972).
41. P. Vandevivere and D. L. Kirchman, Appl. Environ. Micro- 69. C. Holmström, S. James, S. Egan, and S. Kjelleberg, Bio-
biol. 59, 3280–3286 (1993). fouling 10, 251–259 (1996).
42. A. E. Goodman and K. C. Marshall, in H. Lappin-Scott 70. G. Rheinheimer, Aquatic Microbiology, 4th ed., John Wiley
and J. W. Costerton, eds., Microbial Biofilms, Cambridge & Sons, New York, 1992.
University Press, Cambridge, U.K., 1995, pp. 80–98. 71. K. D. Hyde et al., Biodiversity Conservation 7, 1147–1161
43. J. Wimpenny, W. Manz, and U. Szewzyk, FEMS Microbiol. (1998).
Rev. 24, 661–671 (2000). 72. E. B. G. Jones, in R. F. Acker, B. F. Brown, J. R. DePalma,
44. L. A. Pratt and R. Kolter, Curr. Opin. Microbiol. 2, 598–603 and W. P. Iverson, eds., Proceedings of the Third Interna-
(1999). tional Congress on Marine Corrosion and Fouling, North-
western University Press, Evanston, Ill., 1973, pp. 640–647.
45. G. P. C. Salmond, B. W. Bycroft, G. S. A. B. Stewart, and
P. Williams, Mol. Microbiol. 16, 615–624 (1995). 73. E. B. G. Jones, in E. B. G. Jones, ed., Recent Advances in
Aquatic Mycology, John Wiley & Sons, New York, 1976,
46. C. Fuqua, S. C. Winans, and E. P. Greenburg, Annu. Rev. pp. 1–49.
Microbiol. 50, 727–751 (1996).
74. G. Rees, Botanica Marina 23, 375–385 (1980).
47. S. Swift et al., Trends Biol. Sci. 21, 214–219 (1996).
75. G. Rees and E. B. G. Jones, Botanica Marina 27, 145–160
48. C. Fuqua and A. Eberhard, in G. M. Dunny and (1984).
S. C. Winans, eds., Cell-Cell Signaling in Bacteria,
76. K. Hyde, E. B. G. Jones, and S. T. Moss, in S. Barry,
American Society for Microbiology, Washington, D.C., 1999,
D. R. Houghton, G. C. Llewellyn, and C. E. O’Rear, eds.,
pp. 211–230.
Biodeterioration 6, C.A.B. International, Farnham Royal,
49. D. G. Davies et al., Science 280, 295–298 (1998). U.K., 1986, pp. 584–589.
50. R. J. C. McLean, M. Whiteley, D. J. Stickler, and W. C. 77. K. Hyde, S. T. Moss, and E. B. G. Jones, Biofouling 1,
Fuqua, FEMS Microbiol. Lett. 154, 259–263 (1997). 287–298 (1989).
51. S. Kjelleberg et al., Aquat. Microb. Ecol. 13, 85–93 (1997). 78. D. E. Morse, in J. D. Costlow and R. C. Tipper, eds., Marine
52. P. R. Jensen and W. Fenical, Annu. Rev. Microbiol. 48, Biodeterioration: An Interdisciplinary Study, Naval Insti-
559–584 (1994). tute Press, Annapolis, Md., 1984, pp. 134–140.
618 BIOFOULING IN THE MARINE ENVIRONMENT

79. J. R. Pawlik, Oceanogr. Mar. Biol. Annu. Rev. 30, 273–335 108. J. S. Maki et al., Biol. Bull. 177, 295–302 (1989).
(1992). 109. J. S. Maki, D. Rittschof, and R. Mitchell, Microb. Ecol. 23,
80. R. D. Burke, Can. J. Zool. 61, 1701–1719 (1983). 97–106 (1992).
81. F.-S. Chia in J. S. Ryland and P. A. Tyler, eds., Reproduc- 110. C. D. Todd and M. J. Keough, J. Exp. Mar. Biol. Ecol. 181,
tion, Genetics, and Distributions of Marine Organisms, Olsen 159–187 (1994).
and Olsen, Fredensborg, Denmark, 1989, pp. 3–12. 111. M. J. Keough and P. T. Raimondi, J. Exp. Mar. Biol. Ecol.
82. D. J. Crisp, in P. T. Grant and A. M. Mackie, eds., Chemore- 185, 235–253 (1995).
ception in Marine Organisms, Academic Press, New York, 112. M. J. Keough and P. T. Raimondi, J. Exp. Mar. Biol. Ecol.
1974, pp. 177–265. 207, 59–68 (1996).
83. M. G. Hadfield and D. Scheuer, Bull. Mar. Sci. 37, 556–566
113. P. S. Meadows and G. B. Williams, Nature 198, 610–611
(1985).
(1963).
84. S. L. Coon et al., Biol. Bull. 179, 297–303 (1990).
114. M. A. Miller, J. C. Rapean, and W. F. Whedon, Biol. Bull.
85. M. G. Hadfield and J. T. Pennington, Bull. Mar. Sci. 46, 94, 143–157 (1948).
455–464 (1990).
115. B. Wisely, Aust. J. Mar. Freshwater Res. 9, 362–371 (1958).
86. W. K. Fitt and S. L. Coon, Biol. Bull. 182, 401–408 (1992).
116. M. S. Brancato and R. M. Woollacott, Mar. Biol. 71, 51–56
87. M. N. Tamburri, R. K. Zimmer-Faust, and M. L. Tamplin, (1982).
Biol. Bull. 183, 327–338 (1992).
117. A. Daniel, in Proceedings of the First Summer School of
88. D. S. Stoner, Marine Biol. 121, 319–326 (1994).
Zoology (Simla 1961), Government of India, New Delhi,
89. E. J. Turner et al., Limnol. Oceanogr. 39, 1579–1593 (1994). India, 1963, pp. 363–382.
90. R. K. Zimmer-Faust and M. N. Tamburri, Limnol. Oceanogr. 118. D. J. Tighe-Ford, M. J. D. Power, and D. C. Vaile, Helgolan-
39, 1075–1087 (1994). der Meeresuntersuchungen 20, 393–405 (1970).
91. L. J. Walters, M. G. Hadfield, and C. M. Smith, Mar. Biol. 119. S. Wieczorek and C. D. Todd, Mar. Biol. 128, 463–473
126, 383–393 (1996). (1997).
92. D. J. Crisp, in J. D. Costlow and R. C. Tipper, eds., Marine 120. D. Kirchman, S. Graham, D. Reish, and R. Mitchell, J. Exp.
Biodeterioration: An Interdisciplinary Study, Naval Insti- Mar. Biol. Ecol. 56, 153–163 (1982).
tute Press, Annapolis, Md., 1984, pp. 103–126.
121. R. M. Weiner, A. M. Segall, and R. R. Colwell, Appl. Envi-
93. E. W. Knight-Jones, J. Mar. Biol. Assoc., U.K. 30, 201–222
ron. Microbiol. 49, 83–90 (1985).
(1951).
122. R. M. Weiner et al., J. Shellfish Res. 8, 117–123 (1989).
94. J. D. Costlow and C. G. Bookhout, Biol. Bull. 114, 284–295
(1958). 123. J. S. Maki, D. Rittschof, J. D. Costlow, and R. Mitchell, Mar.
Biol. 97, 199–206 (1988).
95. J. Roughgarden, S. Gaines, and H. Possingham, Science
241, 1460–1466 (1988). 124. U. Szewzyk et al., Mar. Ecol.: Prog. Ser. 75, 259–265 (1991).
96. L. S. Mullineaux and C. A. Butman, Mar. Biol. 110, 93–103 125. C. Holmström, D. Rittschof, and S. Kjelleberg, Appl. Envi-
(1991). ron. Microbiol. 58, 2111–2115 (1992).
97. A. Abelson and M. Denny, Annu. Rev. Ecol. Systematics 28, 126. A. Mary, V. Mary, D. Rittschof, and R. Nagabhushanam, J.
317–339 (1997). Chem. Ecol. 19, 2155–2167 (1993).
98. J. H. Connell, Ecology 42, 710–723 (1961). 127. N. J. O’Conner and D. L. Richardson, J. Exp. Mar. Biol.
99. S. K. Wieczorek and C. D. Todd, Biofouling 12, 81–118 Ecol. 206, 69–81 (1996).
(1998). 128. J. S. Maki and R. Mitchell, Bull. Mar. Sci. 37, 675–683
100. J. S. Maki, Biofilms, bioadhesion, corrosion, and bio- (1985).
fouling in M. Fingerman, R. Nagabhushanam, and M.- 129. R. Mitchell and J. S. Maki, in M.-F. Thompson, R. Sarojini,
F. Thompson, eds., Recent Advances in Marine Biotech- and R. Nagabhushanam, eds., Marine Biodeterioration:
nology,, vol. 3, Science Publishers, Enfield, N.H., 1999, Advanced Techniques Applicable to the Indian Ocean, Oxford
pp. 147–171. & IBH Publishing Co. Pvt. Ltd., New Delhi, India, 1988,
101. C. Holmström and S. Kjelleberg, Biofilms, bioadhe- pp. 489–497.
sion, corrosion, and biofouling in M. Fingerman, 130. D. Kirchman, S. Graham, D. Reish, and R. Mitchell, Mar.
R. Nagabhushanam, and M.-F. Thompson, eds., Recent Biol. Lett. 3, 131–142 (1982).
Advances in Marine Biotechnology, vol. 3, Science Publish- 131. J. S. Maki et al., Bull. Mar. Sci. 46, 499–511 (1990).
ers, Enfield, N.H., 1999, pp. 173–201.
132. J. S. Maki, A. B. Yule, D. Rittschof, and R. Mitchell, Biofoul-
102. D. J. Crisp and J. S. Ryland, Nature 185, 119 (1960). ing 8, 121–131 (1994).
103. R. R. Strathmann and E. S. Branscomb, in S. E. Stancyk, 133. A. L. Neal and A. B. Yule, J. Exp. Mar. Biol. Ecol. 176,
ed., Reproductive Ecology of Marine Invertebrates, Uni- 127–139 (1994).
versity of South Carolina Press, Columbia, S.C., 1979,
pp. 77–89. 134. J. S. Maki et al., Biofouling 16, 159–170 (2000).
104. C. Hudon, E. Bourget, and P. Legendre, Can. J. Fish. Aquat. 135. C. Holmström and S. Kjelleberg, Biofouling 7, 329–337
Sci. 40, 1186–1194 (1983). (1993).
105. D. Roberts, D. Rittschof, E. Holm, and A. R. Schmidt, J. Exp. 136. C. Holmström et al., Int. J. Systematic Bacteriol. 48,
Mar. Biol. Ecol. 150, 203–211 (1991). 1205–1212 (1998).
106. E. R. Holm et al., J. Exp. Mar. Biol. Ecol. 215, 189–203 137. G. M. Wolfaardt et al., Microb. Ecol. 27, 279–291 (1994).
(1997). 138. H.-C. Flemming, Water Sci. Technol. 32(8), 27–33 (1995).
107. J. W. Mihm, W. C. Banta, and G. I. Loeb, J. Exp. Mar. Biol. 139. T. J. Beveridge, S. A. Makin, J. L. Kadurugamuwa, and
Ecol. 54, 167–179 (1981). Z. Li, FEMS Microbiol. Rev. 20, 291–303 (1997).
 BIOFOULING OF INDUSTRIAL SYSTEMS 619

140. M. L. Labare, S. L. Coon, C. Matthias, and R. M. Weiner, 170. R. Toth, J. Phycol. 12, 222–233 (1976).
Appl. Environ. Microbiol. 63, 4107–4110 (1997). 171. A. O. Christie, L. V. Evans, and M. Shaw, Ann. Bot. 34,
141. K.-D. Spindler and W. A. Müller, Wilhelm Roux’ Archive of 467–482 (1970).
Developmental Biology 169, 271–280 (1972). 172. M. E. Callow and L. V. Evans, Protoplasma 80, 15–27
142. E. Leitz and T. Wagner, Mar. Biol. 115, 173–178 (1993). (1974).
143. R. A. Cameron and R. T. Hinegardner, Biol. Bull. 146, 173. A. H. L. Chamberlain and L. V. Evans, Protoplasma 76,
335–342 (1974). 139–159 (1973).
144. R. Neumann, Mar. Ecol.: Prog. Ser. 1, 21–28 (1979). 174. A. M. Jones and E. B. G. Jones, in S. Barry, D. R. Houghton,
145. D. K. Hofmann and U. Brand, Symbiosis 4, 99–116 (1987). G. C. Llewellyn, and C. E. O’Rear, eds., Biodeterioration
146. C. R. Johnson and D. C. Sutton, Mar. Biol. 120, 305–310 VI, C.A.B. International, Farnham Royal, U.K., 1986,
(1994). pp. 590–595.
147. G. Schmahl, Helgolander Meeresuntersuchungen 39, 33–42 175. R. L. Fletcher, in J. M. Sharpley and A. M. Kaplan, eds.,
(1985). Proceedings of the 3rd International Biodegradation Sym-
posium, Applied Sciences Publishers, Ltd., London, U.K.,
148. G. Schmahl, Helgolander Meeresuntersuchungen 39, 117–
1976, pp. 443–464.
127 (1985).
176. R. L. Fletcher and R. E. Baier, Mar. Biol. Lett. 5, 251–254
149. C. R. C. Unabia and M. G. Hadfield, Mar. Biol. 133, 55–64
(1984).
(1999).
177. P. S. Dillon, J. S. Maki, and R. Mitchell, Microb. Ecol. 17,
150. L. V. Evans, Botanica Marina 24, 167–171 (1981).
39–47 (1989).
151. B. Moss and A. Marsland, Br. Phycological J. 11, 309–313
178. I. Joint, M. E. Callow, J. A. Callow, and K. R. Clarke, Bio-
(1976).
fouling 16, 151–158 (2000).
152. B. L. Moss, D. Tovey, and P. Court, Botanica Marina 24,
179. R. W. S. P. Thomas and D. Alsopp, in T. A. Oxley and
207–209 (1981).
S. Barry, eds., Biodeterioration 5, 348–357 (1983).
153. R. L. Fletcher, A. M. Jones, and E. B. G. Jones, in J. D. Cost-
180. M. Begon, J. L. Harper, and C. R. Townsend, Ecology, 3rd
low and R. C. Tipper, eds., Marine Biodeterioration: An
ed., Blackwell Science, Ltd., Oxford, U.K., 1996.
Interdisciplinary Study, Naval Institute Press, Annapolis,
Md., 1984, pp. 172–182.
154. A. O. Christie and L. V. Evans, Nature 193, 193–194 (1962).
155. J. D. Pringle, J. Exp. Mar. Biol. Ecol. 100, 97–111 (1986). BIOFOULING OF INDUSTRIAL SYSTEMS
156. A. O. Christie, in R. F. Acker, B. F. Brown, J. R. DePalma,
and W. P. Iverson, eds., Proceedings 3rd International HANS-CURT FLEMMING
Congress of Marine Corrosion and Fouling, Northwestern University of Duisburg and IW
University Press, Evanston, Ill., 1973, pp. 674–681. Duisburg, Germany
157. J. M. Kain, J. Mar. Biol. Assoc., U.K. 44, 415–433 (1964).
158. W. E. Jones and M. S. Babb, Br. Phycol. Bull. 3, 525–528 Biofouling is referred to as the undesired development of
(1968). microbial layers on surfaces. Such microbial layers are
159. J. R. J. Baker and L. V. Evans, Protoplasma 77, 1–13 termed biofilms (1) and the effect is called biofouling.
(1973).
The term has been adapted from the heat exchanger
160. R. L. Fletcher and A. H. L. Chamberlain, in R. J. Gilbert technology, wherein fouling is defined generally as
and D. W. Lovelock, eds., Microbial Aspects of the Dete-
the undesired deposition of material on surfaces (2). A
rioration of Materials, Academic Press, New York, 1975,
distinction is made between different types of fouling,
pp. 59–81.
including the following:
161. R. L. Fletcher, Botanica Marina 24, 211–221 (1981).
Scaling, mineral fouling: Deposition of inorganic
162. H. P. Lin, M. R. Sommerfield, and J. R. Swafford, J. Phycol.
material precipitating on a surface
11, 452–457 (1975).
Organic fouling: Deposition of organic substances such
163. D. A. Coon, M. Neushul, and A. C. Charters, in K. Nishi-
as oil, fat, proteins and so forth.
zawa, ed., Proceedings of the VII International Seaweed
Symposium, University of Tokyo Press, Tokyo, Japan, 1972,
Particle fouling: Deposition of particles from the water
pp. 237–242. phase, such as humic substances, clay, silica, debris and
164. A. H. L. Chamberlain, in J. M. Sharpley and A. M. Kaplan,
so forth.
eds., Proceedings of the 3rd International Biodegradation Biofouling: Adhesion and growth of microorganisms on
Symposium, Applied Sciences Publishers, Ltd., London, surfaces.
U.K., 1976, pp. 417–432. In the first three kinds of fouling, the increase of a fouling
165. R. L. Fletcher and M. E. Callow, Br. Phycol. J. 27, 303–329 layer arises from the transport and abiotic accumulation
(1992). of the material from the water phase on the surface. What
166. M. E. Callow et al., Appl. Environ. Microbiol. 66, 3249–3254 is deposited on the surface originates quantitatively from
(2000). the water. In these cases, fouling can be controlled by
167. M. E. Callow, J. A. Callow, J. D. Pickett-Heaps, and R. eliminating the foulants from the liquid phase. However,
Wetherbee, J. Phycol. 33, 938–947 (1997). this is different from biofouling — microorganisms are
168. A. O. Christie and M. Shaw, Br. Phycol. Bull. 3, 529–534 pseudo ‘‘particles’’ that can multiply.
(1968). Even if 99 to 99.9% of all bacteria are eliminated by
169. L. V. Evans and A. O. Christie, Ann. Bot. 34, 451–466 pretreatment, there may be a large enough inoculum,
(1970). which will enter the system to become protected, adhere to
620 BIOFOULING OF INDUSTRIAL SYSTEMS

surfaces, and multiply at the expense of biodegradable Bolsena (13), which is attributed to the growth of Serratia
substances. Thus, microorganisms convert dissolved marcescens on sacramental bread and polenta.
organic material into biomass locally. This is the same Not surprisingly, biofouling occurs in toilet bowls (14)
mechanism that supports biofilm technology — biofouling but is commonly observed in water systems in general (15)
can be considered as a ‘‘biofilm reactor in the wrong place and during the filtration of seawater (16). It also
and at the wrong time.’’ Substances suitable as nutrients, represents a serious problem in fish farms where
which would not act as foulants per se, will support the cage netting fouls rapidly (17). The submerged
fouling indirectly. As most antifouling measures target structural surfaces of offshore oil and gas production
the microorganisms, the role of nutrients as a potential platforms are covered by biofilms (18), potentially giving
source of biomass is overseen and biocides tend not to rise to microbially influenced corrosion (19). Little (20)
decrease the nutrient level. On the contrary, nutrients has given an excellent overview on marine biofouling.
supplied by the oxidation of recalcitrant organics can Microorganisms can contribute to calcareous deposits,
support rapid aftergrowth (3). As it is virtually impossible adding scaling to biofouling (21). Massive deposition
to keep a common industrial system completely sterile, of manganese and iron minerals is frequently due to
microorganisms on surfaces will always be present, microbial activity (22). Boreholes and aquifers can be
‘‘waiting’’ for nutrients. clogged by excessive biofilm growth (23), leading to
Usually, the different kinds of fouling occur together (4). considerable technical problems. A Thiothrix species
The proportion of biofouling can be considerable. An resistant to 200-ppm free chlorine was identified as the
example is the development of dental plaque, namely, cause of biofouling in groundwater systems in Florida (24).
mineral depositions on teeth that are favored by Desiccation of biofilms can also lead to the deposition of
biofilms (5). In algal biofilms, precipitation of calcium minerals (25). In some instances, unexpected biofouling
carbonate is increased, mainly because of the rise in pH has led to problems such as the deterioration of pH
resulting from photosynthesis (6). electrode response due to biofilm formation on the
Generally, biofouling has to be considered as a biofilm- glass membrane (26). Moored spectroradiometers were
based problem. To understand the effects and dynamics of stricken by biofouling (27), and the list of other water-
biofouling and to design appropriate countermeasures, it exposed sensors that may yield erratic data because
is important to understand the natural processes of biofilm they are affected by biofilms could easily be expanded.
formation and development. Biofilms growing on the walls of houses can influence
In biofilms, microorganisms live in an environment that the surface temperature of the building walls and thus
is entirely different from that experienced by planktonic increase mechanical weathering process due to differential
organisms (7). They are more or less immobilized, thermal expansion. These biofilms can increase the heat
embedded in a highly hydrated matrix of extracellular uptake, leading to an increase of energy demand for air
polymeric substances (EPS), which is commonly described conditioning (28).
as ‘‘slime.’’ The cell density is some orders of magnitude In medicine, implant devices such as catheters are
higher than that in suspension. Owing to the long prone to biofouling (29). Dental waterlines can be seriously
retention time of the cells next to each other, the contaminated by pathogens (30).
development of interacting aggregates is possible, leading
to so-called microconsortia. Biofouling in Drinking Water
A biofilm consists of 80 to 95% water. The organic
proportion is dominated by the EPS, comprising mainly The maintenance of the hygienic and aesthetic quality
polysaccharides and proteins (8), whereas the cell material of water during transport in the distribution systems
usually makes up only a minor fraction. Additionally, par- still challenges drinking water technology (31). Massive
ticulate matter such as clay, humic substances, corrosion growth of cells and the occurrence of pathogenic organisms
products, and so forth can be included. Chemical analysis are observed (32). Both can originate from growth in the
shows that this collected abiotic material can represent water phase but also from biofilms. Thus, all surfaces
the major part of the dry matter, although it is deposited in contact with water have to be considered as potential
mainly because of the adhesive effect of the biofilm. Thus, supports for biofilm growth and as contamination sources.
the analysis of unwanted deposits can be misleading when In particular, the surface to volume ratio in distribution
the cause of a fouling layer development is investigated. systems leads to the fact that more than 95% of the entire
From an ecological point of view, life in a biofilm biomass is located on the walls and less than 5% in the
may offer important advantages to the cells (9). Among water phase. Fouling and corrosion in water filtration and
these are (1) the possibility to form microconsortia; transportation systems can therefore represent a serious
(2) the facilitated exchange of genetic material; (3) the problem (33).
accumulation of nutrients from the bulk water phase (10); Water that leaves the water works in a hygienically
and (4) protection against toxic substances and against optimum state can contain substantial numbers of
desiccation. microorganisms after passing the distribution system.
This is mainly due to contamination by biofilms (34,35).
OCCURRENCE OF BIOFOULING Biofilms release cells both continuously by erosion and
discontinuously by sloughing events (36) and by swarmer
Biofouling, by definition, can occur in extremely diverse cells (37), which deliberately leave the biofilm. The
situations ranging from space stations (11,12) to pro- contamination rate, however, is unpredictable. Thus, there
fane explanations for religious miracles like that of is no correlation between the cell concentration in the
 BIOFOULING OF INDUSTRIAL SYSTEMS 621

water and in the biofilm. Data from the water phase give no coworkers (56) could demonstrate that in the ultrapure
information about the extent or the site of biofilm growth. water system (18 MOhm water) of the IBM plant in
In the same system, the cell concentration in a biofilm — as Burlington, Vermont, United States, all surfaces in contact
compared with the water phase and expressed in relation with water carried biofilms. The piping material consisted
to volume — ranges from 107 to 1010 colony forming units of poly(vinylidene dichloride) (PVDF). The authors found
(cfu) mL−1 (38,39), whereas the cell concentration in the that the contamination of the water was due to the
water phase can be as low as 105 cells mL−1 and below biofilms. By sequential sampling, they showed that the
10 cfu mL−1 . cell numbers in water close to the walls were significantly
In drinking water technology, it is assumed that a higher than those in the bulk liquid phase.
massive development of microorganisms on surfaces only What can these organisms live on? There are pioneer
occurs if the support material releases biodegradable mat- bacteria, which can multiply in water with less than
ter (40). Certain paints and coatings have led to increased 10 µg L−1 glucose equivalents (57). If they form biofilms,
cell numbers in drinking water (41) for this reason. How- they can sequester nutrients from the flowing water phase
ever, similar incidents have been observed even when the and accumulate them in the EPS matrix. This allows the
support material did not leach nutrients (34,36,42). organisms to live with an even lower concentration of
biodegradable material, which is why biofilm formation is
Pathogenic and Potentially Pathogenic Organisms in Biofilms considered an ecological strategy in oligotrophic environ-
ments (58). These pioneers convert the organic materials
It is now acknowledged that biofilms can provide a
into biomass, which represents spots of high concentra-
habitat for pathogenic microorganisms. In medicine,
tions of organic material. This offers a suitable habitat to
mycobacteria have been known for a long time as
other organisms, which may have invaded the system.
‘‘trumpet bacteria’’ occurring in the mouthpieces of
A common strategy to prevent microorganisms from
musical instruments of tuberculosic performers, and,
entering a system is the use of so-called ‘‘sterile’’
less artistically, in telephone handles (43). Klebsiellae,
filters. Nevertheless, ‘‘breakthroughs’’ occur repeatedly.
Mycobacteriae, Legionellae, Escherichia coli, and coliform
An important cause is ultramicrobacteria (59). During
organisms have been found in biofilms (44–48). In the
starvation, the morphology of cells can change drastically,
presence of corrosion products, the organisms seem to be
resulting in cells of less than 0.2 µm in diameter. This is
particularly protected. This is the conclusion of the study
demonstrated with Pseudomonas fluorescens rods within
conducted by Emde and coworkers (49), which identified
three weeks of nutrient depletion. Cells of that size have
a much higher variety of species in corrosion product
been shown to migrate through sterile filters (60).
deposits, called ‘‘tubercles,’’ compared with the free water
phase, even after extended periods of chlorination. The
fate of viruses in biofilms is still in question. Schmitt and Biofouling in Water Treatment Plants and Membrane Systems
coworkers (50) showed that enteroviruses could be isolated
from biofilms on sand filters of drinking water plants fed Water treatment plants frequently include granular filters
with surface water. with a large surface area, which are predestined for biofilm
growth. Biofilms can grow on piping material, insulations,
fittings, elastic filling materials and so forth and develop
Biofouling in Purified Water
significant tolerance against disinfectants. These biofilms
One of the major problems in the production of contribute considerably to the overall purification process
purified and ultrapure water is contamination from because they degrade diluted organic matter. A change
microorganisms. Processes that depend on ultrapure water in nutrient concentration, temperature, or other factors
suffer accordingly. In the production of microelectronic can cause either mass production or sloughing of biofilms,
devices, in which ultrapure water is often used to wash which leads to an increasing contamination of the water.
printed circuit boards, microorganisms can cause severe In ion exchangers, biofilms clog the material and lead to an
quality losses. Dial and Chu (51) found a direct correlation increased pressure drop, and in membrane processes such
between the cell content in ultrapure water and the as reverse osmosis, ultrafiltration, and others, biofouling
defects in the microchips. Craven and coworkers (52) represents a serious problem (61–63).
attributed failures in the metal coating process to the Biofilms form a gel layer on top of the membrane and
adhesion of microorganisms. Eisenman and Ebel (53) prevent direct tangential flow adjacent to the membrane
describe horizontal and vertical dislocations in the crystal surface. This leads to concentration polarization at the feed
lattice during the photoresist process, which are due waterside, which increases the transmembrane pressure
to microbial contamination. Microorganisms can bridge drop ( pmembrane ). Thus, both permeate production and
electrical circuits where the distances are below 1 salt rejection will decrease. The costs of biofouling of
µm. Electromigration and corrosion of the oxide layers membranes can be summarized as follows (64):
have also been observed. Sodium and calcium (Ca2+ )
ions from lysed bacteria or from bacterial metabolism
• Loss of product quantity
can diffuse into the surface of the silicon wafer and
influence field effects (54). Bacteria are carbon-containing — Flux decline due to increased pmembrane (see
particles, and, thus, can increase the conductivity of Fig. 1)
semiconductors (55). These microorganisms are mainly • Loss of product quality
released from biofilms. In a detailed study, Patterson and — Possible microbial contamination of permeate
622 BIOFOULING OF INDUSTRIAL SYSTEMS

Figure 1. Biofilm on the feed side of a reverse osmosis

membrane; bacteria are embedded in the slime matrix (39).

— Decrease of salt rejection because of concentra-

tion polarization
• Higher energy demand
— Increased pmembrane to overcome
— Increased concentration polarization
• Higher pretreatment and cleaning demand
— Chemicals (biocides, cleaners)
— Labor- Shutdown time
— Eventual treatment of effluent to meet legal
regulations Figure 2. Permeate side of a reverse osmosis membrane;
• Higher replacement costs bacteria are attached and can contaminate the permeate
— Damage by cleaning (Courtesy of H. F. Ridgway, in B. S. Parekh, ed., Reverse Osmosis
Technology, Marcel Dekker, New York, Basel, 1988, pp. 429–481).
— Damage by microbial attack
• Frustration and demotivation of the operating
personnel The costs of biofouling are increased by some common
mistakes, which originate from the fact that biofouling is
Figure 1 shows a scanning electron micrograph of the not understood as a biofilm process. Inadequate methods
surface of an irreversibly biofouled RO membrane (39) of detection and monitoring of countermeasures lead
with a biofilm that has survived many cleaning cycles. to delayed recognition and to suboptimal or ineffective
Reverse osmosis (RO) membranes are theoretically biofouling control.
impermeable for microorganisms as these membranes Investigations about the beginning of the fouling
act as diffusion membranes and do not contain pores processes revealed that the microorganisms settle on the
at all. However, microorganisms have been found at the membrane during the first hours of operation (66). This
permeate side of RO membranes (65). Figure 2 shows an means that every operating plant possesses a biofilm from
example. its beginning. Only the extent of biofilm growth will decide
Even today, there has been no explanation for their whether biofouling occurs: only if a certain threshold of
occurrence at this site, although it has been observed interference is exceeded does biofouling take place; thus,
in various cases. Permeate-side bound biofilm cells may it is an operationally defined expression.
seed the permeate, which will partially attach to the
surfaces and form biofilms. Traces of nutrients will support
Biofouling on Heat Exchangers
their growth and, thus, further the contamination of the
purified water. Indeed, this is also of great importance The efficiency of a heat exchanger can be seriously
in the pharmaceutical industry. A detailed assessment decreased by biofilms (4,67). Forming a gel layer, they
of the costs of biofouling has been carried out for the ‘‘insulate’’ the heat exchanger surface from the liquid
reverse osmosis (RO) plant at Water Factory 21 in Orange phase. Although the heat-transfer resistance of biofilms
County, California (64). The ‘‘bottom line’’ is US $730,000 is similar to that of water, the gel allows only diffusive
spent each year to control membrane biofouling. This heat transport. Thus, heat transfer by convective transport
represents about 30% of the total operating costs for Water is inhibited. In addition, friction resistance is increased,
Factory 21. which increases energy consumption; this aspect is
 BIOFOULING OF INDUSTRIAL SYSTEMS 623

considered to be an even more costly problem in The installations can be overgrown so heavily that
heat exchange (15). Similar to membrane systems, heat they physically break down (4). In ocean thermal energy
exchangers in aqueous systems operating in a temperature conversion (OTEC), biofouling is addressed as the most
range between 5 and 60 ° C will always carry biofilms, difficult problem, seriously limiting the use of an otherwise
which remain unnoticed because their effects stay below promising technology (68).
the threshold level of interference. However, sudden
biofouling problems do not arise from microorganisms Biofouling in Oil Industry
that have suddenly invaded the system but are much The production, transport, storage, and processing of oil is
more likely to be caused by an increase of nutrient stricken by severe biofouling problems (69,70). Sulfate-
concentration or an absence of inhibiting factors. The reducing bacteria (SRB) are of particular importance
additives can represent such factors. As an example, because they are responsible for souring of oil and for
chromate has been replaced by other corrosion inhibitors microbially influenced corrosion (MIC). A water content
because of its environmentally undesirable properties. of only one percent is sufficient to support substantial
A new organic inhibitor that had excellent properties, development of SRB within only a few weeks (71).
including full biological degradability, was introduced.
Microbial growth at the production sites can cause
The system protected by this substance failed after three
clogging, especially if water is used as replacement
months. Biodegradation had started during operation and
agent (70).
turned the exchanger into an activated biofilm sludge
reactor (Fig. 3). Thus, an environmentally reasonable step Biofouling on Ship Hulls
has to be harmonized with all other effects in the system.
If light has access, substantial algal growth can A classical case of surfaces prone to biofouling are
occur. This is frequently the case in cooling towers. ship hulls. As indicated earlier, the speed of ships can
be significantly reduced even by thin biofilms. Before
higher organisms such as mussels or others, a microbial
biofilm develops (72). D.C. White (Inst. Appl. Microbiol.,
Univ. of Knoxville, Personal Communication) has assessed
that the United States Navy has to spend more than
US$ 500 million in additional fuel that is required to
overcome the additional friction resistance caused by
biofilms: a biofilm of only 100 µm thickness leads to a
10% increase in friction resistance (4). The attempts for
effective antifouling coatings go back to the medieval ages
when copper plating was introduced, without long-term
effect. After a certain time, some organisms can tolerate
the toxicity of copper, colonize it, complex the ions, and
shield other organisms for which copper would be toxic.
Later colonizing species do not encounter the copper
surface but encounter the primary biofilm. The problem
has been transiently solved by modern antifouling paints
containing tributyl tin compounds. Unfortunately, they
are so toxic and recalcitrant to biological degradation (73)
that many countries have resorted to a partial ban. Thus,
the problem remains unresolved.

Biofouling and Microbially Influenced Corrosion

In biofouling, the biofilm colonizes a given surface and
creates problems by contamination of the adjacent medium
or by impeding the function of the surface. However, it
can also influence corrosion, as corrosion is an interfacial
process, which is dependent on the physicochemical
conditions at the interface such as pH value, redox
potential, oxygen concentration, conductivity, and others.
All of these factors are influenced by biofilms. The effect
is called microbially influenced corrosion (MIC) and is
basically a biofilm problem. Thus, biofouling and MIC are
closely related (19).

COUNTERMEASURES
Figure 3. Heat exchanger in a cooling water cycle that was
‘‘protected’’ by a biologically degradable corrosion inhibitor after As biofouling often occurs far from the focus of attention,
six weeks of operation. namely a product, a common practice is to ignore the
624 BIOFOULING OF INDUSTRIAL SYSTEMS

problem as long as possible. The next phase is to Even with low numbers in the water phase, consid-
look for causes other than microorganisms and to test erable biofilm growth can occur, as has been shown
countermeasures. If none of the common causes proves with microbially contaminated ion exchangers: while 5
to be valid and none of the countermeasures is effective, to 8 cFu mL−1 were found in the process water, sampling
it is concluded that this must be a microbial problem. of the exchanger bed revealed cell numbers as high as
Certainly, this conclusion is true in many cases. Thus, an 104 cFu mL−1 (75). The concentration of cells released from
expert in the field of microbial deterioration is consulted. biofilms is usually diluted in the water body. However, in
Unfortunately, this person encounters the consequences situations of stagnation, these cells can become concen-
of all former attempts, and symptoms that would help trated — a frequently observed phenomenon. It does not
identify the microbial origin of the problem are obscured arise because cells grow better under stagnant conditions.
by provisional repair actions. This can be a serious On the contrary, cells grow better under agitated con-
obstacle for the design and application of an effective ditions (all other conditions being favorable); this is the
counterstrategy. A checklist that links the symptoms to reason fermenters are vigorously mixed. The comparison of
eventual microbially influenced corrosion (MIC) has been two ion exchanger columns, one left standing idle and the
produced for this purpose. other operated with circulating water, revealed an iden-
If a problem is suspected to be of microbial origin, a tical extent of microbial growth (76). Similar observations
three-step program has to be developed according to the were reported by other authors, who showed that the num-
following questions: ber of sulfate-reducing bacteria in the water phase was
completely unrelated to the number of biofilm cells (71).
1. How can biofilms and biofilm-related damage be Thus, proper sampling of biofilms must be carried out
identified? on the surfaces in question or on representative surfaces
and by adequate methods (77). These include either the
2. How can the system be sanitized?
removal of biofilms, or of biofilm-carrying surfaces, or new
3. What can be done to prevent future problems? optical and spectroscopic methods. Special candidates for
the occurrence of biofilms are crevices, holes, curves, and
Detection of Biofilms other areas protected from high shear stress.
The presence of biofilms in a technical system is
Analysis of Biofilms
usually detected only indirectly. Operational parameters
represent only the symptoms. In heat exchangers, heat Biofilms in engineered systems have usually reached the
transfer decreases while drag resistance increases. In plateau phase and are in most cases visible to the naked
membrane technology, both membrane resistance and eye. Thus, optical inspection gives a first hint. In technical
drag resistance increase. In practice, identification of environments, biofilms tend to display a slimy consistency,
biofouling generally follows this pattern: a problem in which can be detected by wiping; thin layers can be made
plant performance occurs; all common countermeasures visible if a white tissue is used for wiping. In case of doubt
are applied; if they do not work, and there is no better about the biological origin of a deposit or layer, it is useful
explanation, the problem is attributed to biology, and in to take a small amount of the material and put it over
most cases, this is correct. a lighter until it smolders. A smell of burnt protein is
characteristic of biological material. Presently, a test kit
The Problem of Sampling with simple dye reactions is under development to deliver
reliable information about the presence or absence of
As soon as biofouling is suspected, water samples biofilms and to semi-quantify them. The ultimate detection
are taken and investigated microbiologically. Different of biofilms is carried out in the laboratory. In Table 1,
methods are applied and they give different results.
Cultivation methods detect only those organisms that are Table 1. Examples of Biofilm Parameters
capable of multiplying on the supplied growth media.
According to the cultivation temperature and time, the Parameter Detection Method
numbers of colony-forming units can vary by some orders
Water content 24 hours, 110 ° C
of magnitude. The majority of bacteria occurring in
Organic carbon TOC, COD, incineration loss
natural and technical waters do not grow on common
Protein 78
microbiological media (74). Microscopical methods give the Carbohydrates 79
number of all cells present in a sample but usually cannot DNA 80
discriminate between viable and nonviable cells. Lipids 81
A more fundamental problem in attempting to detect Muramic acid 81
biofouling by sampling of the water phase is that there Polyhydroxybutyrate 81
is no correlation between numbers of suspended cells Total cell number 82
detected by water sampling and the location or extent of Colony-forming units Various standard methods
biofilms. Although biofilms contaminate the water phase, ATP 83
Hydrolase activity 84
they do so discontinuously. Cells can be eroded or leave the
Respiratory activity 85
biofilm actively (37), and parts of the biofilm can slough Indolacetic acetic acid production 86
off, leading to episodes of high cell numbers in the bulk Catalase activity 87
liquid phase.
 BIOFOULING OF INDUSTRIAL SYSTEMS 625

some parameters suitable for the detection of biofilms are Table 2. Requirements for an Ideal Biocide
summarized.
Efficacy against the organisms in the system if they occur in
For screening, the contents of water and organic carbon biofilms
are indicative; if these are high, there is a high probability Compatibility with other conditioning chemicals
that the material is a biofilm. The presence of ATP or Satisfactory time, concentration, and temperature of action
respiratory activity indicates living organisms. It is wise Lack of corrosive effects on the system
to use more than one parameter for the characterization of Safety and health aspects concerning handling and storage
a biofilm. If the occurrence of a problem can be related to Stability
the actual presence of biofilms, the diagnosis ‘‘biofouling’’ Biodegradability
is justified. Minimal effect in effluent water
Analytical detection
Low cost
Sanitation
The most common practice to eliminate unwanted biofilms
is the application of biocides. They are used to ‘‘disinfect’’
a system. This strategy is historically rooted in medicine: [CFU cm−2] [CFU ml−1]
the germs that cause the disease are to be killed. Then,
the immune system will dispose them. However, killing
the organisms in an engineered system will not solve the 106
108
problem because dead biomass remains. Unlike a living
organism, an engineered system has no mechanism to
discard dead microbes. This leads to an accumulation 106
of readily biodegradable material and gives rise to 104
rapid microbial regrowth. In addition, engineered systems
generally cannot be run under sterile conditions (except 104
for pharmaceutical or electronics water supply systems 102
that are kept aseptic with great effort). Reinfection of the
102
‘‘disinfected’’ system is already assured by the water that
is used to rinse the biocide. New cells encounter dead
biomass that is readily biodegradable and will support 100
rapid microbial regrowth. Thus, removal of biomass 0 10 30 60 90 [min] 120
may be more important than killing. Rational cleaning Figure 4. Survival of P. fluorescens in suspension (•) and
measures include a step in which the physical stability in a biofilm (π ) (after M. C. Mittelman, in H.-C. Flemming
of the biofilm matrix is weakened and a subsequent and G. G. Geesey, eds., Biofouling and Biocorrosion in Indus-
step in which it is removed from the system. The use trial Water Systems, Springer, Heidelberg, Germany, 1991,
of biocides can be justified in the framework of a combined pp. 113–134).
strategy to minimize the biofouling potential. However,
the considerations as discussed earlier must be kept in
mind, particularly because ‘‘disinfection’’ and ‘‘cleaning’’ The efficacy of biocides against biofilm organisms is
tend to be mixed up. known to be reduced compared with where the same
The variety of biocides has grown to such an organisms are suspended in the water phase (95,96).
extent (88–91) that it justifies the need for ‘‘biocide muse- This is demonstrated by the different resistance of
ums’’ (92) to avoid the reinvention of forgotten biocides P. fluorescens shown in Figure 4 (97).
and to collect experience and data. In the United States, Organisms in biofilms are well known for increased
Canada, England, and Australia, chlorine and chlorine- tolerance toward biocides (95). Keevil and coworkers (96)
containing compounds make up most of the spectrum of reported that coliform organisms in a biofilm survived
biocides in industrial applications (4). Peracetic acid, the prolonged exposure to 12-ppm free chlorine. Living
peroxide of acetic acid, has been widely applied in the biofilms have been found on the inner walls of disinfectant
former eastern block; its excellent disinfection proper- pipes (98). In many cases, not all organisms in a biofilm
ties were summarized by Flemming (93). Peracetic acid is are killed. They recover and metabolize the dead biomass.
an example of a disinfectant that was discovered twice, This leads to a rapid regrowth after a disinfection measure.
namely, in 1905 and 1947. The selection of a biocide suit- Unfortunately, most data on biocides relate to suspensions
able for a given system requires experience and should of test organisms. Although this evaluation of biocide
be supported by preceding laboratory screening. Natu- efficacy is reproducible, it does not reflect the effectiveness
ral product antifoulants seem to represent an interesting against biofilms. It is very difficult to obtain biofilms that
alternative (94). Some general criteria for the selection of can be used as reference systems with sufficient similarity
biocides are summarized in Table 2. It is noticeable that to ‘‘natural’’ biofilms. Kinniment and Wimpenny (99)
they are contradictory. Economic advantages may cause developed the ‘‘Cardiff constant depth biofilm fermenter’’
increasing environmental damage or result in insufficient using Pseudomonas aeruginosa as the test organism
efficacy. Thus, a biocide that meets all of the criteria listed and keeping the biofilm thickness constant by accurate
hereunder is not expected. It is impossible to discuss the abrasion with a razor blade. They could demonstrate that
vast range of biocides presently on the market. 200-ppm formaldehyde killed all planktonic cells while the
626 BIOFOULING OF INDUSTRIAL SYSTEMS

viable cell number in the biofilm was only reduced from easily from a surface. In this case, it is irrelevant whether
108 cells per square centimeter to 106 cells per square the cells are dead — the ‘‘side effect’’ of decreasing matrix
centimeter. Griebe and coworkers (100) successfully used stability is much more important. Some biocides, however,
an annular biofilm reactor to compare the efficacy of can have the opposite effect. An example is formaldehyde
different biocides against biofilms. It was shown that that is used in microscopy as a preserving reagent because
chloramine was much more effective against biofilms than it cross-links proteins. In such a case, ‘‘disinfection’’ can
chlorine. However, in no case was biomass removed from make the problem worse, as has been reported many times
the surfaces. Keevil and coworkers (96) suspected that in practice (105).
the high nitrite levels in London tap water were caused If a system is to be cleaned, the mechanical resistance
by incomplete nitrification of monochloramine in biofilms, of the slimy biofilm layer must be overcome. Although
where ammonia oxidizers metabolized the amino groups biofilms usually can be wiped off easily, they are
only to nitrite. surprisingly resistant to the shear forces of running
The use of chlorine will be increasingly limited water. This is the reason for the observation of biofilms
for reasons of sewage water pollution and tighter in water pipelines with very high water velocity (4).
environmental regulations. Thus, ozone offers some They become thinner but more stable. Interestingly,
advantages. Toxic by-products are formed to a much lesser an increase in performance after the application of
extent and the method has been shown to be effective cleaning substances does not necessarily have to be due
against biofilms (101,102). Ozone weakens the biofilm to the removal of a biofilm. As an example: reverse
matrix and, thus, facilitates the removal of biomass by osmosis membranes were cleaned with a commercial
shear forces (103). A drawback is that the costs for ozone cleaning formulation (Ultrasil U 53, Henkel, Germany).
generation are about four times as high as that for chlorine Permeation performance increased, but biomass quantity
treatment. on the membrane surface remained almost the same.
A potential problem in the application of oxidizing The cleaning agent simply increased the permeability
biocides is their effect on the bioavailability of recal- of the fouling layer (106), thus mitigating the original
citrant organic matter. Refractory organic substances, hydrodynamic resistance of the biofilm.
for example, humic substances, have been shown to Time plays an important role in the mechanical
become biodegradable by chlorine and ozone treatment (3). resistance of a biofilm. ‘‘Young’’ biofilms are much easier
The raw water was treated with 2.5 mg L−1 ozone for to remove than ‘‘old’’ biofilms. Table 3 shows the rinsing
15 minutes. Then, the water was filtered using granular efficacy of surface-active substances such as sodium
activated carbon (GAC), and a concentration of 2.0 mg L−1 dodecyl sulfate (SDS) and Tween 20. Cells were allowed
of free chlorine was maintained. The assimilable organic to attach to a polyethersulfone membrane for either four
carbon (AOC) increased significantly after the application hours or three days.
of ozone. Subsequently, membrane samples were agitated for one
The essence of these considerations is that concentra- hour with the surface-active agent. The number of cells
tion, time of action, temperature, and interaction with per surface area was determined after this procedure. The
dissolved substances and with biofilms must be increased results demonstrate that early action against an unwanted
if biocides are applied against biofilms. biofilm is more effective.
For an effective cleaning, the bond between the biofilm
Cleaning: Removal of Biofilms from Surfaces and the underlying surface must be broken. Although
biofilms can develop on practically any surface, the binding
For cleaning, the mechanical properties of biofilms are of
force varies. A reasonable approach to better cleanability
fundamental importance. Although the biofilm matrix is
is the choice of materials or coatings with low adhesion
kept together by weak physicochemical interactions (104),
force to biofilms (107). Low surface energy materials tend
it is very difficult to remove it only by shear forces.
to display low adhesion forces to biofilms; this is exploited
Hydrogen bonds, electrostatic interactions, and dispersion
in Teflon coatings on ship hulls.
forces add to an overall binding force between the EPS
Another choice for cleaning of surfaces is ultrasound.
molecules, which is equivalent to several covalent chemical
It has been applied successfully in cleaning of medical
bonds per biopolymer. Oxidizing or dispersing chemicals
devices, dental equipment, heat exchangers, and other
are useful for the first step, while flushing, brushing, or
surfaces. A critical appraisal of published material reveals
similar physical techniques are useful for the second step.
that in almost all cases there is no specification of the
Of particular importance is an efficiency control in order
ultrasound intensity to which the object is exposed.
to optimize the process.
However, ultrasonic methods seem to be an attractive
Prevention of further microbial problems usually
cannot be achieved by single measures but by a combined
strategy, which can be called ‘‘good housekeeping.’’ A very Table 3. Influence of Biofilm Age on the Detachment
important element of this strategy will be the elimination of Cells, Expressed as Percentage of Detached Cells
of biodegradable material, as this represents potential Compared to the Control (76)
biomass.
Age of Biofilm Control 1% Tween 20 1% SDS
If an oxidizing biocide is used, it will weaken the biofilm
matrix as a side effect. This is the case for chlorine, which 4 hours 0 93 98
is consumed by reaction with extracellular polysaccharides 3 days 0 42 57
and proteins. As a result, the biofilm can be removed more
 BIOFOULING OF INDUSTRIAL SYSTEMS 627

Table 4. Critical Ultrasonic Parameters for the Detach- Table 5. Physical Methods for the Removal of Biofilms
ment of Pseudomonas diminuta from a Polyether Sulfone
Membrane (95% Detachment of Cells)(108) Method Remarks

Variable Value Boundary Conditions Rinsing Simplest method; limited

efficacy; thin biofilms are not
Intensity uptake (I) 4W t = 300 s, D = 2 cm affected; very effective if
Transducer distance (D) 4 cm t = 300 s, I = 4 W biofilm stability is first
Exposure time (t) 60 s D = 3 cm, I = 4 W decreased, for example, by
biodispersants
Backwashing Similar to rinsing
alternative to chemical cleaning. Thus, an ultrasonic bath Sponge balls, abrasive Reasonable efficacy; may
was used to remove biofilms from separation membrane damage supporting material
material. The sonification field was mapped with a Sponge balls, nonabrasive Less effective than abrasive
hydrophone; this technique allowed an assessment of the sponge balls; not suitable for
energy, which arrived at the object. The critical values are biofilms, which display
greasing effects
summarized in Table 4 (108).
The application of ultrasonic energy seems to be feasible Sandblasting Very effective, but difficult to
for the cleaning of surfaces from unwanted biofilms. focus the abrasive effect on
the biofilm and not on the
However, it has to be considered that ultrasound may
supporting material
damage the surface on which the biofilm is growing. This
was the case with membrane materials. In other cases, Brushing Very effective but limited to
accessible surfaces; can shift
ultrasound may still offer a suitable method.
biofilm population to species,
which form extremely stable
Cleaning Strategy EPS
It seems wise not to rely on a single substance or method Hot water, steam Especially suitable for purified
when cleaning a surface. Killing the cells, as effectively water systems; efficacy
against thick biofilms not yet
as this may be performed, and leaving them at the place
evaluated; energy intensive
will not solve the problem. Removal of biomass seems to
be much more important. An effective strategy must be Ultrasonic energy Good efficacy; restricted to
surfaces accessible to
defined according to the system to be cleaned. If possible,
ultrasonic energy; may
facilitated cleaning should be integrated in the design of damage underlying material
a system. In general, a cleaning strategy will include the
Ice crystallization
following two steps:
Application of ethyleneglycol
(-12 C), destabilizes biofilm
1. Weakening of the biofilm matrix, mostly by chemi-
matrix (Costerton, 1983). No
cals such as oxidants such as chlorine, ozone, hydro- practical experience reported
gen peroxide, peracetic acid, or others; by alkaline
UV irradiation
treatment, tensides, enzymes (109), complexing sub-
Poor efficacy against biofilms;
stances (110), or by biodispersants (111). The latter abiotic particles may protect
are based on polyethyleneglycol and are supposed to microorganisms; does not
weaken the interactions in the EPS matrix as well as remove biofilm material
the interaction between biofilm and support mate-
rial. A combination of various agents may increase
the efficacy. However, it is important to establish a
system to quantify the success. Control of Efficacy
2. Removal of the biofilm by mechanical forces such In practice, the efficacy of a cleaning measure is assessed
as rinsing with water, air, steam, or a combination by the recovery of process parameters. However, this is
thereof; application of sponge balls, brushing, or an indirect method and not very well-suited to assess the
ultrasound. Some methods are listed in Table 5. success. As an example, the biofouling layer on a reverse
osmosis membrane was removed by about 80% by chemical
The experience of the authors in the application of enzymes treatment. This led to a substantial improvement of the
was not encouraging. This may be due to the specificity process parameters. However, the remaining 20% of the
of enzymes, which are directed against certain bonds layer offered optimum conditions for the regrowth of the
that may not be present in all matrix-forming molecules. biofilm bacteria. Thus, after a very short time, the old
Thus, microbial strains that form EPS against which situation was reestablished, which led to the well-known
the enzymes are not effective gain a selection advantage. ‘‘saw-toothed curve’’ (39).
Mechanical cleaning can also select those species, which As biofouling is a biofilm problem, which in turn is
form the most resistant EPS matrix, as has been reported related to surfaces, the latter have to be sampled for
in biofouling problems with the technique to recover ocean a proper effectiveness control. This can be generally
thermal energy (112). achieved in several ways:
628 BIOFOULING OF INDUSTRIAL SYSTEMS

1. Integration of test surfaces, which are analyzed been shown that adhesion to polysulfone membranes was
before and after the cleaning procedure, or, equally rapid both with living and dead cells of P. diminuta
2. Establishment of on-line, sensitive monitoring [‘‘passive adhesion,’’ Flemming and Schaule (119)]. In
devices such as the continuous measurement of drag other cases, cells excrete EPS in response to surface
resistance, heat-transfer resistance, hydrodynamic contact [‘‘active adhesion’’ (114)]. In a biocoenosis of water,
resistance, weight of deposit, light transmittance, or there will be various species present, some of them
reflectance. adhering even when dead and others developing contact-
specific adhering substances.
Efficacy control allows the optimization of countermea- Roughness of the surface will support colonization.
sures against biofouling. Quite ineffective but widely However, smoothing of a surface will not completely
practiced is the measurement of cell numbers in the water prevent microbial adhesion. Electropolishing of metal
phase; these are relatively easy to measure but lead only heat exchanger surfaces only slightly prolonged the
to numerous data, which can never be interpreted. induction phase of biofilm development (4); EPS provide
a ‘‘glue’’ that can stick to the smoothest surfaces. A
PREVENTION OF BIOFOULING certain success has been reported with anionic block
copolymers (114,120). An interesting approach is coating
Disinfection still remains the most common ‘‘prevention’’ with biocidal compounds (121). However, the coating
practice. After considering the points discussed earlier, can be dissolved or covered with dead microorganisms,
it is not surprising that disinfection tend to display which could nullify the effect. Practical applications of
transient effects. The disinfectant is usually rinsed out these approaches have not been reported. A similar
with nonsterile water, which leads to reinfection of case may be the use of plaque inhibitors, namely,
the system. This explains why disinfection should not substances that prevent microbial adhesion to teeth.
be expected to offer a suitable choice in prevention of Among these are quaternary ammonium compounds,
biofouling. biguanidines such as chlorhexidine, uncharged phenolic
As biofouling is a biofilm problem, prevention of compounds (‘‘Listerine’’), enzymes, peroxides, and surface-
primary adhesion seems to be the most reasonable active substances (122). All of these substances are
strategy. The first steps in microbial adhesion are candidates for the extension of the induction phase of
dominated by physicochemical forces rather than by colonization in closed systems.
physiological activity of microorganisms (114). Among the Open systems such as ships in seawater are protected
abiotic parameters, surface energy is considered to be by antifouling paints. Success is always transient (123),
of particular importance (115). Surfaces used in water implying that frequent cleaning is necessary. Tributyltin
technology will display different surface energies. When compounds are the only really effective substances known.
in contact with water, a conditioning film arises, formed However, they are perhaps too successful, as they belong
by traces of macromolecules present in water. Among to the most toxic materials (73) and are going to be
these are proteins, polysaccharides, humic substances, banned worldwide because of their negative impact on
and others. Although they do not provide specific binding the biosphere. Copper is widely believed to be bactericidal
groups, many parts of a macromolecule can interact by and copper plating belongs to the first antifouling
weak forces such as van der Waals and electrostatic measurements in naval history. However, colonization of
interactions and hydrogen bonds. However, the molecules copper surfaces is a well-known phenomenon (98,105) and
possess many binding sites and will bind to a given surface. once covered with a primary biofilm, even copper-sensitive
Thus, they will adhere irreversibly and result in a slightly organisms can survive in a biofilm on a copper surface.
negative overall surface charge (116). The conditioning Copper still plays an important role in marine antifouling
film is not a prerequisite for microbial adhesion but as an additive to Teflon coatings. Biofilm growth is not
only a result of the fact that macromolecules reach the prevented but slowed down, and cleaning is easy.
surface faster than microorganisms. In some cases, the Another so-called biocidal element is silver, to which
conditioning film may decrease the adhesion rate (117). an ‘‘oligodynamic effect’’ is attributed. In fact, most
Hydrophobic surfaces will become more hydrophilic, microorganisms are killed by very low concentrations of
leading to a surface tension in the range of 30 to silver ions (<10 µg L−1 ), and silver is not known to be
40 mJ m−2 (115). Slowest microbial adhesion is reported toxic to humans. Unfortunately, microorganisms develop
between 23 and 27 mJ m−2 ; below these values, the a tolerance within a few weeks, allowing them to multiply
adhesion rate is higher (118). Thus, efforts to use surfaces in the presence of more than 1 mg L−1 of Ag+ (76).
with very low surface energy will not succeed in preventing Highlighted by these observations, inhibition of pri-
microbial adhesion. An example is given by leaves in mary adhesion is not a very promising strategy to prevent
autumn. The wax layer provides a high hydrophobicity, biofouling. Microorganisms are ubiquitous, and among
which allows the leaves to float on water. After a these, there will always be some that can colonize in spite
while, specifically hydrophobic microorganisms such as of the inhibitor. If UV light is used, the prospects are not
Rhodococcus sp. will colonize leaf surfaces, resulting in a much better. Only on surfaces that are directly irradi-
hydrophilic biofilm, which causes the leaves to sink. ated, biofilm growth can possibly be prevented. As soon as
Primary adhesion occurs very rapidly, within the first abiotic particles adhere, microorganisms can be protected
hours of contact between water and surface. It is not and can grow (Huber and Flemming, unpublished obs.).
necessary that the microorganisms are viable; it has Additionally, UV light may oxidize humic acids and make
 BIOFOULING OF INDUSTRIAL SYSTEMS 629

them more bioavailable, therefore creating more problems

than what it solves.
The most common technical measure to prevent biofoul-
ing is the permanent dosage of biocides. This is usually
performed with chlorine. Bott (102) showed that a concen-
tration of at least 0.2 ppm had to be maintained in a techni-
cal system. Below this concentration, rapid biofilm growth
was observed. In many cases, permanent dosage of biocides
is undesirable for environmental reasons. Biocides and
their toxic by-products contaminate the sewage water and
require special attention in terms of storage and handling.
Induction Log. accumulation Plateau Time
Living with Biofilms
Figure 5. Schematic depiction of biofilm development. Dotted
Prevention of biofouling can only be achieved by a line: biofilm below the threshold of interference.
combined strategy. It must be based on a ‘‘clean system
philosophy’’ that includes all surfaces in contact with
water. It includes the minimization of biodegradable Table 6. Biofilm, Water, and Performance Parameters as
organic substances, good access of surfaces that must be Determined by Experimental Membrane Test Cells Before
kept clean, surfaces that are easy to clean, and no or very and After Sand Biofilter (After 125)
few crevices, fittings, edges, or other niches predestined for Before Sandfilter After Sandfilter
biofilm growth. Ultrapure water technology provides many
experiences and solutions for less pure water systems. Biofilm
The elements of good housekeeping can be summarized as Biofilm thickness (27.3 ± 3.1) µm (3.0 ± 0.5) µm
follows: Protein content 77.7 µg cm−2 3.6 µg cm−2
Carbohydrate content 22.5 µg cm−2 2.6 µg cm−2
Uronic acid content 10.6 µg cm−2 2.3 µg cm−2
• Minimized nutrient concentration
Water
• Monitoring of biofilm development on representative BDOC of water 0.325 mg L−1 0.125 mg L−1
surfaces Microbial content 1 × 107 cfu mL−1 1.2 × 106 cfu mL−1
• Strong shear forces Performance
Flux decline 40% in 30 days <2% in 30 days
• Frequent cleaning (if possible, mechanical)
• Easy-cleaning design
• Effectiveness control and optimization of cleaning
measures encouraging the same phenomenon until the extent of
biofilm growth exceeds the level of interference. This is
The importance of effective monitoring systems must why biofouling can be considered as a ‘‘biofilm reactor
be pointed out clearly. Such systems allow for early in the wrong place.’’ The main factors, which govern
intervention in a biofouling process and optimization of biofilm growth, and biofouling potential, are the ubiquitous
countermeasures. biofilms, the nutrients available, and the shear forces.
All approaches to prevent biofouling as discussed ear- As the occurrence of biofilms cannot be avoided without
lier are based on killing and cleaning. A completely extreme efforts (e.g., as performed by the pharmaceutical
different strategy can be developed from some basic con- and microelectronics industries), only nutrients and shear
siderations. Almost all surfaces in contact with water carry forces remain as variables. Shear forces can be increased
biofilms, but not all water systems suffer from biofouling. quite easily in some systems but not in all. Thus, nutrient
It has been shown that biofilms are involved in the separa- concentration represents a candidate for limiting biofilm
tion process on reverse osmosis membranes right from the growth: the lower the nutrient concentration, the less
beginning: after a few hours of operation, a biofilm devel- biofilm will form. This effect is well known in biofilters,
ops and participates in separation (64). Biofouling occurs where most biomass accumulates in the area with the
only when biofilm development exceeds a certain ‘‘thresh- highest nutrient concentration. The biofilm reactor can be
old of interference.’’ This is the case when performance ‘‘put in the right place’’ if positioned in front of the system
parameters remain below certain limits (Fig. 5). to be protected from biofouling: the biofilm will develop
In technical systems, biofilms usually reach the plateau mainly in the reactor. In the subsequent areas, biofilm
phase relatively rapidly. What happens is a natural growth will occur, but below the threshold of interference
phenomenon: microorganisms settle on surfaces and if the system is run properly. For example, river water
convert diluted nutrients into metabolic products and was run through a sand bed as a biofilter. Before and after
locally immobilized biomass. This is exploited in the the filter, a reverse osmosis test cell had been installed,
design of a biofilm reactor. Biomass accumulates in the one operating with raw water and the other with water
reactor, and the effluent water will support microbial after the biofilter. Permeate production was taken as
growth, including the development of biofilms, to a lesser performance parameter (125). A BDOC of 0.125 mg L−1
extent. Separation membranes, filter beds, pipe surfaces, was sufficient to stay below the threshold of interference.
and other technical environments offer large surfaces, Such a value can be reached easily with common biofilters.
630 BIOFOULING OF INDUSTRIAL SYSTEMS

Biofilm thickness was reduced to 10%. The microbiological 19. W. H.Dickinson, F. Caccavo, and Z. Lewandowski, Corros.
data of the water reveal that the biofilter does not cause Sci. 8, 1407–1422 (1996).
additional microbial contamination but rather removes 20. B. J. Little, Treat. Mar. Sci. Technol. 28, 89–119 (1988).
cells from the process stream. 21. S. Dexter and S.-H. Lin, Mater. Perform April, 16–21
The results indicate that an antifouling strategy, (1991).
which refrains from using chemicals and relies simply on 22. L. Tuhela, L. Carlson, and O. H. Tuovinen, J. Environ. Sci.
nutrient depletion, can be successfully applied to technical Health, Part A — Environ. Sci. Eng. Toxicol. Haz. Subst.
water systems. Obviously, it is not necessary to kill all Control 32, 407–426 (1997).
biofilms — it is only necessary to keep their development 23. S. F. Tyrrel and P. Howsam, Q. J. Eng. Geol. 30, 161–169
below the threshold of interference. Thus, it is possible (1997).
to live with biofilms without biofouling. Of course, this 24. R. L. Brigmon, H. W. Martin, and H. C. Aldrich, Curr.
is not feasible in all cases, but if it is applied where it is Microbiol. 35, 169–174 (1997).
appropriate, unnecessary usage of biocides can be avoided. 25. R. J. C. McLean, H. E. Jamieson, and D. R. Cullimore,
It may be pointed out that this approach is only one of a World J. Microb. Biotechnol. 13, 25–28 (1997).
few that are based on the laws of biofilm development. We 26. W. A. Munro et al., J. Sens. Actuat. B-Chem. 37, 187–194
expect that in the near future more biocide-free strategies (1996).
will be developed. 27. F. P. Chavez et al., J. Atmos. Oc. Technol. 17, 215–219.
28. J. Garty, Can. J. Bot. 68, 1349–1353 (1989).
29. E. P. J. M. Everaert et al., Arch. Otolaryngol. Head Neck
BIBLIOGRAPHY Surg. 125, 139–1332 (1998).
30. J. Barbeau, C. Gauthier, and P. Payment, Can. J. Microbiol.
1. W. G. Characklis and K. C. Marshall, eds., Biofilms, John 44, 1019–1028 (1998).
Wiley & Sons, New York, 1990. 31. U. Szewzyk, R. Szewzyk, W. Manz, and K.-H. Schleifer,
2. N. Epstein, in E. F. C. Somerscales and J. G. Knudsen, Annu. Rev. Microbiol. 54, 1–47 (2000).
eds., Fouling of Heat Transfer Equipment, Hemisphere, 32. S. L. Percival, J. S. Knapp, D. S. Wales, and R. G. J. Edy-
Washington, D. C., 1981, pp. 31–53. vean, Biofouling 13, 259–270 (1999).
3. M. W. LeChevallier, in H.-C. Flemming and G. G. Geesey, 33. R. G. J. Edyvean, in P. Howsam, ed., Microbiology in Civil
eds., Biofouling and Biocorrosion in Industrial Water Engineering, Chapman & Hall, London, U.K., 1990,
Systems, Springer-Verlag, Heidelberg, Germany, 1991, pp. 62–74.
pp. 113–132.
34. E. v. d. Wende and W. G. Characklis, in G. A. McFeters, ed.,
4. W. G.Characklis, in W. G. Characklis and K. C. Marshall, Drinking Water Microbiology, Springer International, New
eds., Biofilms, John Wiley & Sons, Chichester, U.K., 1990, York, 1990, pp. 249–268.
pp. 523–633.
36. G. J. Tuschewitzki, P. Werner, and W. Dott, Gas Wasser
5. D. J. White, Biofouling 4, 209–218 (1991). Fach 124, 521–526 (1983).
6. M. E. Callow et al., in J. Wimpenny, P. Handley, P. Gilbert, 37. U. Szewzyk and B. Schink, J. Gen. Microbiol. 134, 183–190
and H. Lappin-Scott, eds., The Life and Death of Biofilm, (1988).
(ISBN 0-9520432-3-8), BioLine, Cardiff, Glamorgan, Wales,
38. W. Dott and D. Schoenen, Zbl. Bakt. Hyg. I., Abt. Orig. B
1995, pp. 113–117.
180, 436–447 (1985).
7. J. Wimpenny, in P. Handley, P. Gilbert, and H. Lappin-
39. H.-C. Flemming and G. Schaule, Vom Wasser 73, 287–301
Scott, eds., The Life and Death of Biofilm, (ISBN 0-
(1989).
9520432-3-8), BioLine, Cardiff, Glamorgan, Wales, 1995,
40. DVGW, Techn. Reg. Arbeitsbl. W 270, DVGW-Regelwerk
pp. 113–117.
1990.
8. J.Wingender, T. Neu, and H.-C. Flemming, in J. Wingender,
41. D. Schoenen, in P. Howsam, ed., Microbiology in Civil
T.Neu, and H.-C. Flemming, eds., Bacterial Extracellular
Engineering, Chapman & Hall, London, U.K., 1990,
Polymer Substances, Springer-Verlag, Heidelberg, Berlin,
pp. 121–145.
1999, pp. 1–19.
42. M. W. LeChevallier, T. M. Babcock, and R. G. Lee, Appl.
9. K. C. Marshall, Microb. Sci. 2, 321–326 (1985).
Environ. Microbiol. 53, 2714–2724 (1987).
10. A. W. Decho, Oceanogr. Mar. Biol. Ann. Rev. 28, 73–153
43. R. C. Müller, Medizinische Mikrobiologie, Urban & Schwar-
(1990).
zenberg, Berlin, Germany, 1946, p. 246.
11. V. Guarnieri et al., Acta Astronautica 40, 195–201 (1997).
44. R. J. Seidler, J. E. Morow, and S. T. Bagley, Appl. Environ.
12. D. W. Koenig and D. L.Pierson, Water Sci. Technol. 35, Microbiol. 33, 893–900 (1977).
59–64 (1997).
45. R. Schulze-Röbbecke, B. Jannig, and R. Fischeder, Tuberc.
13. Cullen et al., ASM News (1997). Lung Dis. 73, 141–144 (1992).
14. B. Pitts, Biofouling 13, 19–30 (1998). 46. J. B. Wright et al., Infect. Control Hosp. Epidemiol. 10,
15. L. F.Melo and T. R. Bott, Exp. Therm. Fluid Sci. 14, 408–415 (1989).
375–381 (1997). 47. M. W. LeChevallier, J. Am. Water Works Assoc. 92(11),
16. C. J. Williams and R. G. J. Edyvean, Water Sci. Technol. 38, 74–86 (1990).
309–316 (1998). 48. R. Szewzyk and W. Manz, 6th International Conference
17. S. L. Hodson, C. M. Burke, and A. P. Bisset, Aquaculture of Microbial Ecology, Barcelona, Spain, Sep. 7–11, Poster
184, 277–290 (2000). Nr. P2-04-21, 1992.
18. A. Y. Zvyagintsev and V. V. Ivin, Oceanogr. Lit. Rev. 43, 473 49. K. M. E. Emde, D. W. Smith, and R. Facey, Water Res. 26,
(1996). 169–175 (1992).
 BIOFOULING OF INDUSTRIAL SYSTEMS 631

50. J. Schmitt, S. Krietemeyer, G. van den Bosche, and 78. M. M. Bradford, Anal. Biochem. 72, 248–254 (1976).
H.-C. Flemming, DVGW Schriftenreihe 110, 259–279 (1995). 79. M. Dubois et al., Anal. Chem. 28, 350–356 (1956).
51. F. Dial and T. Chu, Proceedings of the 7th Annual Semicond. 80. E. Thomanetz, A. Sperandio, and D. Bardtke, Gas Wasser
Pure Water Conference, San Jose, Calif., 1987 pp. 178–193. Fach 124, 8–16 (1983).
52. R. A. Craven, A. J. Ackerman, and P. L. Tremong, Microcon- 81. G. G. Geesey and D. C. White, Annu. Rev. Microbiol. 44,
tamination 4(99), 1421 (1986). 579–602 (1990).
53. D. E. Eisenmann and C. J. Ebel, Proceedings of the 9th 82. L. A. Meyer-Reil, Appl. Environ. Microbiol. 36, 506–512
Annual Meeting Institute of Environmental Science, (ICCS), (1978).
Los Angeles, Calif., 1988 pp. 547–559. 83. L. Young-Bandala and R. J.Kajdasz, Proceedings of the
54. D. C. White and M. W. Mittelman, 9th Annual Semi- International Water Conference, vol. 44, Engineering Society
cond. Pure Water Conference, Santa Clara, Calif., 1990, of West Pennsylvania, 1983, 442–446.
pp. 150–171. 84. U. Obst, A. Holzapfel-Pschorn, and R. Oldenbourg Verlag,
55. S. J. Poirier, Proceedings of the 4th Annual Semicon- München, Germany, 1988.
duct Pure Water Conference, San Francisco, Calif., 1985, 85. G. Schaule, H.-C. Flemming, and H. F. Ridgway, Appl.
pp. 197–210. Environ. Microbiol. 59, 3850–3857 (1993).
56. M. K. Patterson, G. R. Husted, A. Rutkowski, and D. C. Ma- 86. J. M. Bric, R. Rostock, and S. E. Silverstone, Appl. Environ.
yette, Ultrapure Water, 1991 pp. 18–24. Microbiol. 57, 535–538 (1991).
57. R. Y. Morita, in M. Fletcher and G. Floodgate, eds., Bacteria 87. M. A. Line, in T. A. Oxley and S. Barry, eds., Biodeteriora-
in Their Natural Environments: The Effect of Nutrient tion, vol. 5, John Wiley & Sons, New York, 1983, pp. 38–43.
Conditions, Society of General Microbiology, London, U.K.,
88. K. R. Payne, Critical Review of Applied Chemistry, vol. 23,
1985, pp. 111–130.
John Wiley & Sons, New York, p. 118.
58. M. M. Fletcher, Adv. Microb. Physiol. 32, 53–85 (1991).
89. W. Paulus, Microbicides, Chapman & Hall, London, U.K.,
59. J. A. Novitsky and R. Y. Morita, Appl. Environ. Microbiol. 1993.
32, 617–622 (1976).
90. W. Cloete, L. Jacobs, and V. S. Broezel, Biodegradation 9,
60. M. W. Mittelman and D. C. White, Proceedings of the 23–37 (1998).
Pharmaceutical Technical Meeting, Philadelphia, Pa., 1989,
91. B. Buecker and R. Post, Chem. Eng. Proc. 94, 45–50 (1998).
211–221.
92. R. M. Organ, in T. A. Oxley and S. Barry, eds., Biodeteriora-
61. H.-C. Flemming, Exp. Therm. Fluid Sci. 14, 382–391 (1997).
tion vol. 5, John Wiley & Sons, New York, 1983, pp. 409–415.
62. J. S. Baker and L. Y. Dudley, Desalination 118, 81–89
93. H.-C. Flemming, Zbl. Bakt. Hyg. B 179, 97–111 (1984).
(1998).
94. A. S. Clare, Biofouling 9, 211–229 (1996).
63. H. S. Vrouwenvelder et al., Desalination 118, 157–166
(1998). 95. M. W. LeChevallier, in H.-C. Flemming and G. G. Geesey,
eds., Biofouling and Biocorrosion in Industrial Water
64. H.-C. Flemming, G. Schaule, R. McDonogh, and H. F. Ridg-
Systems, Springer-Verlag, Heidelberg, Germany, 1991,
way, in G. G. Geesey, Z. Lewandowski, and H.-C. Flemming,
pp. 113–132.
eds., Biofouling and Biocorrosion in Industrial Water Sys-
tems, Lewis Publishers, Chelsea, Mich., 1994, pp. 63–89. 96. C. W. Keevil, C. W. Mackerness, and J. S. Colbourne, Int.
Biodet. 26, 169–179 (1990).
65. H. F. Ridgway, in B. S. Parekh, ed., Reverse Osmosis Tech-
nology, Marcel Dekker, New York, Basel, 1988, pp. 429–481. 97. M. C. Mittelman, in H.-C. Flemming and G. G. Geesey, eds.,
Biofouling and Biocorrosion in Industrial Water Systems,
66. H.-C. Flemming, G. Schaule, and R. McDonogh, Vom Wasser
Springer-Verlag, Heidelberg, Germany, 1991, pp. 113–134.
80, 177–186 (1993).
98. M. Exner, G.-J. Tuschewitzki, and E. Thofern, Zbl. Bakt.
67. K. K. Satpathy, Bull. Electrochem. 15, 143–147 (1999).
Hyg., I. Abt. Orig. B 177, 170–181 (1983).
68. J. B. Darby, J. Mater. Energy Syst. 6, 130–137 (1984).
99. S. Kinniment and J. W. T. Wimpenny, Int. Biodet. 26,
69. R. G. J. Edyvean, Int. Biodet. 23, 199–231 (1987). 181–194 (1990).
70. J. L. Lynch and R. G. J. Edyvean, Biofouling 1, 147–162 100. T. Griebe et al., Technical Report, Center for Wastewater
(1988). Treatment, University of Munich, Munich, Germany, 2000.
71. J. W. Costerton and J. Boivin, in M. W. Mittelman and 101. R. N. Nesaratnam and T. R. Bott, Heat. Vent. Eng. Feb./
G. G. Geesey, eds., Biological Fouling of Industrial Water March, 5–8 (1984).
Systems. A Problem Solving Approach, Water Micro
102. T. R. Bott, in M. Bohnet, ed., Fouling of Heat Exchanger Sur-
Associates, San Diego, Calif., 1987, pp. 56–76.
faces. Conference Proceedings, vol. 1, Düsseldorf, Germany,
72. B. J. Little and J. R. DePalma, Treat. Mater. Sci. Technol. 1990, pp. 5.1–5.20.
28, 89–119 (1988).
103. K. Kaur, T. R. Bott, and B. S. C. Leadbeater, in L. F. Melo,
73. J. J. Cooney and S. Wuertz, J. Ind. Microbiol. 4, 375–402 M. M. Fletcher, T. R. Bott, and B. Capdeville, eds.,
(1989). Biofilms — Science and Technology, Kluwer Academic,
74. R. R. Colwell, Culture 12, 2–4 (1991). Dordrecht, The Netherlands, 1992, pp. 589–594.
75. H.-C. Flemming, Water Res. 21, 745–756 (1987). 104. H.-C. Flemming et al., in H. Lappin-Scott, P. Gilbert,
76. H.-C. Flemming, in H.-C. Flemming and G. G. Geesey, eds., M. Wilson, and D. Allison, Community Structure and Coop-
Biofouling and Biocorrosion in Industrial Water Systems, eration in Biofilms, SGM symposium 59, Cambridge Univer-
Springer-Verlag, Heidelberg, Germany, 1991, pp. 47–80. sity Press, 2000, pp. 87–105.
77. G. Schaule, T. Griebe, and H.-C. Flemming, in H.-C. Flemm- 105. M. Exner, G. J. Tuschewitzki, and J. Scharnagel, Zbl. Bakt.
ing, T. Griebe, and U. Szewzyk, eds., Investigation of Hyg. B 183, 549–563 (1987).
Biofilms, Technomics Publishers, Lancaster, Pa., 2000, 106. R. McDonogh, G. Schaule, and H.-C. Flemming, J. Membr.
pp. 1–21. Sci. 87, 199–217 (1994).
632 BIOGENIC TRACE GASES

107. H. J. Busscher, R. Bos, and H. C. van der Mei, FEMS the world’s largest bioreactors have been built. Stirred
Microbiol. Lett. 128, 229–234 (1995). tank reactors of 900 m3 have recently been constructed
108. A. Zips, G. Schaule, and H.-C. Flemming, Biofouling 2, in Ghana (2) (Fig. 1) for treatment of sulfide concentrates
323–333 (1990). containing a considerable amount of precious metals. In
109. Eur. Pat. 0,388,115 v. 12.3.90, 1990, C. L. Wiatr. Uganda, reactors with a volume of 1,350 m3 for cobalt
110. M. H. Turakhia, K. E. Cooksey, and W. G. Characklis, Appl. leaching went into operation in 1998 (3). The leaching
Environ. Microbiol. 46, 1236–1238 (1983). microorganisms dissolve the usually poorly water-soluble,
111. H. Schmidt, Wasser, Luft u. Betrieb 27, 12–17 (1983). metal-containing minerals due to an attack by biogenic
112. J. Nickels et al., Appl. Environ. Microbiol. 41, 1442–1453 compounds. The dissolution is achieved by complex inter-
(1981). actions of biological, chemical (especially electrochemical),
113. Biofilm Removal, U.S. Pat. 4.419.248, 1983, J. W. Costerton. and physical processes. Nevertheless, the processes are
114. K. C. Marshall and B. Blainey, in H.-C. Flemming and not yet fully understood nor is their interdependence
G. G. Geesey, eds., Biofouling and Biocorrosion in Industrial elucidated. Bioleaching of nonsulfidic minerals, that is,
Water Systems, Springer-Verlag, Heidelberg, Germany, oxides, carbonates, and silicates, occurs by simple acid
1991, pp. 28–45. attack and/or complexation, whereas for dissolution of
115. C. J. van Oss, Biofouling 4, 25–35 (1991). metal sulfides a combined acid and oxidative attack is
116. R. E. Baier, in G. Bitton and K. C. Marshall, eds., Adsorp- needed. The products of these processes are metal cations
tion of Microorganisms to Surfaces, John Wiley & Sons, New dissolved in aqueous acidic solutions. The protons orig-
York, 1980, pp. 59–104. inate from biogenic acids. The most important acid is
117. K. C. Marshall et al., in G. G. Geesey, Z. Lewandowski, and sulfuric acid, whereas relevant organic acids with good
H.-C. Flemming, eds., Biofouling and Biocorrosion, Lewis metal-complexing properties are oxalic and citric acid.
Publishers, New York, 1992, pp. 15–26. Bacterially produced iron(III) ions are responsible for
118. S. Dexter, J. Coll. Interferon 70, 346–354 (1976). primary oxidation of the sulfur moiety of sulfides. In
119. H.-C. Flemming and G. Schaule, Vom Wasser 71, 207–233 addition, intermediary sulfur compounds are oxidized
(1988). by bacteria. Therefore, metal sulfide leaching bacteria
120. M. Humphries, J. F. Jaworzyn, J. B. Cantwell, and A. are iron and sulfur compound–oxidizing species. The
Eakin, FEMS Microbiol. Lett. 42, 91–101 (1987). most important ones belong to the genera Acidithiobacil-
121. K. J. Hüttinger, NATO ASI Ser., Ser. E 145, 233–239 (1988). lus, Leptospirillum, Metallogenium, Acidianus, and Sul-
122. P. P. Glantz and R. Attström, in H. Loe and D. Kleinman, folobus.
eds., Dental Plaque Control Measures and Oral Hygiene Bioleaching occurs as a natural process in almost
Practices, IRL Press, Oxford, U.K., 1986, pp. 185–194. all climates, as long as water is available and the
temperature does not exceed 114 ° C, which is the the
upper limit for microbial life (4). Bioleaching is applied
in technical conditions for metal winning from raw
BIOGENIC TRACE GASES. See TRACE GASES SOIL
materials that are not processable by conventional
(usually pyrometallurgical) techniques. In most cases, low-
grade ores are concerned, but increasingly even complex
concentrates are treated by bioleaching techniques, such
BIOGEOCHEMICAL CYCLES. See PHOSPHORUS
CYCLING IN AQUATIC ENVIRONMENTS: ROLE OF BACTERIA

BIOHYDROMETALLURGY. See BIOLEACHING

BIOLEACHING

THORE ROHWERDER
PETER-GEORG JOZSA
TILMAN GEHRKE
WOLFGANG SAND
University of Hamburg
Hamburg, Germany

Microorganisms are used worldwide for the extraction of Figure 1. Stirred biooxidation tanks of the Ashanti’s gold
valuable metals. The technique established in the last recovery plant in Obuasi, Ghana, for the bioleaching of refractory
decades, is known as bioleaching (1). For this purpose, ores before cyanidation.
 BIOLEACHING 633

Classical mineralogy is a descriptive science, which

succeeded to identify and classify numerous miner-
als. Relationships between genesis, chemical composi-
tion, structures, and properties have been thoroughly
described (6). Oxides and silicates are the most frequently
encountered minerals but only few of them bear valuable
metals that can or need to be extracted by (bio)leaching
techniques. However, metal sulfides are a very important
class of compounds because some of the base metals occur
mainly in sulfidic ore deposits. Therefore, examples will
focus on this class of minerals. The possible number of
minerals having different compositions and properties is
enormous because of traits classified as:

Figure 2. Dump bioleaching of copper ore in Peru. The dump is • Polymorphism (the same composition but a dif-
irrigated by sprinklers with acidified water. ferent crystallization system: pyrite-marcasite/FeS2 ,
sphalerite-wurtzite/ZnS),
• Polytypism (the same composition and the same
as heap and dump leaching (Fig. 2) or tank leaching. crystallization system but different periodicity of
However, anticipated bioleaching operations need to be crystal layers of the same arrangement: wurtzite-2H,
empirically tested in the lab before realization, thereby -4H, -6H, -8H, -10H, -3R, -15R, and so on),
requiring expensive scale-up. Especially the mineral itself, • Isomorphism (reciprocally unlimited exchangeable
because of the occurrence of minor elements, plays an elements: sphalerite-hawleyite/ZnS-CdS, guadalca-
unpredictable role and causes many uncertainties in zarite Hg, Zn/S, Se; villamaninite Cu, Ni, Co, Fe/S2 ,
the design of the process and plant. It is well known, Se2 , irarsite Ir, Ru, Rh, Pt/AsS), and
for example, that in heap leaching operations treating • Diadochism (limited, nonreciprocal replacement of
chalcopyrite ore, rarely more than 15% of the copper an element by another: up to 20% of zinc atoms
will be dissolved within a ‘‘reasonable’’ time of one to replaceable by iron in sphalerite, but iron in
two years (5). The remaining part has to be deposited pyrrhotite not replaceable by zinc atoms).
on waste dumps. However, the bioleaching process
continues slowly and tends to pollute the environment Additionally, chemical analysis revealed the existence
with acidic solutions containing heavy metal ions. Similar of apparently nonstoichiometric compounds, that is, when
uncontrolled bioleaching is encountered in abandoned calculating the chemical formula; the stoichiometry does
mines, shafts, mine waste dumps, and so on and is not fit with simple valence states of the elements involved.
known as ‘‘acid mine/rock drainage’’ (AMD/ARD). It is Besides minerals, where the involved elements have quite
obviously caused by human activities in most cases but well-defined valence states, such as FeS, FeS2 , ZnS, CuS,
some natural bioleaching sites exist, too. In the past, Cu2 S, CuFeS2 , and so on, minerals are known in which
spontaneous leaching has been used as an indicator one element exists in several different valence states,
of ore bodies because the colored waters attracted for example, Cu9 S5 or Ag3 Fe7 S11 . Moreover, different
attention. minerals may crystallize jointly to form intergrown
crystals. Hence, obviously mineralogically approximately
similar minerals present in ores from different sites
may show differences in their physicochemical properties
MINERALS
and a different behavior in technical processes such
as flotation (7). Some minerals were also (re)produced
Technical applications of leaching processes for the extrac- experimentally (synthetically, in pure or deliberately
tion of valuable metals from natural raw materials must impure state), showing that major differences in specific
consider the quantitative description of the dissolution physicochemical properties may result from only minor
process. Physicochemical properties of the minerals sub- compositional differences (8). Semiconducting properties
jected to dissolution and of the reaction products and (n- or p-conductivity) depend on the nature of minor
reaction kinetics are of great importance for the economy amounts of impurities included in the crystal lattice,
of a process. Because pure minerals are not available for example, n-pyrite contains Co and Ni ions, p-pyrite
in large quantities, an appropriate process design for contains As-ions. These pyrite types show different
technical applications has to consider the presence of behavior in the leaching process. Leaching of n-pyrite
invaluable constituents. More and more low-grade ores starts rapidly but slows down after a certain degree of
must be processed as high-grade ores are used up. Hence, dissolution is reached. On the contrary, leaching of p-
because of impurities and/or gangue material, huge quan- pyrite starts slowly, but subsequently continues at an
tities of ore are mined, which consequently need large enhanced rate (9). Even in homeotypic minerals (different
amounts of reagents for the extractive processes with composition but the same crystallization system: pyrite-
all the inherent technical, economical, and environmental hauerite/FeS2 -MnS2 , molybdenite-tungstenite/MoS2 -WS2 ,
consequences. sphalerite-chalcopyrite/ZnS-CuFeS2 ) electronic structures
634 BIOLEACHING

are not identical. Different energy levels of the atomic acidophilus, and Acidianus brierleyi (14). These species
and molecular orbitals cause different stability, reactivity, can live chemolithoautotrophically with reduced sulfur or
and even different reaction mechanisms. Hence, under iron compounds and carbon dioxide as the sole electron,
certain conditions one mineral may dissolve and the energy, and carbon sources, respectively. Furthermore,
other may not, or perhaps they may dissolve but by some strains can grow by oxidizing molecular hydro-
different pathways (10). Electrochemical properties like gen and are able to fix nitrogen from air. In leaching
electrode potential and polarization also play an important biotopes, these physiologically versatile species are often
role for the specific leaching behavior of a mineral accompanied by less versatile bacteria. Acidithiobacillus
complex. Intergrown particles of different minerals may thiooxidans (formerly Thiobacillus thiooxidans, 13) can
form galvanic cells, causing preferential reaction sites only oxidize sulfur compounds, whereas for Leptospir-
at the solid interface (9). In modern materials sciences, illum ferrooxidans, iron(II) ions are the sole electron
sophisticated procedures have been established to obtain donors. External organic compounds are, if at all, used
special materials with unique, useful properties. From only to a minor extent and surprisingly, many organic
this experience, it became understandable why naturally substances are toxic to these bacteria (14). Therefore they
occurring mineral samples in many cases are also unique. are classified as strictly chemolithoautotrophic organisms.
Hence, it should also be clear that each ore deposit, in many Besides these bacteria, mixotrophic species able to grow on
cases each mineral vein, must be tested for technologically both inorganic and organic substrates have been isolated
reliable process data. For practical purposes, a scale-up from leaching biotopes. For instance, Acidiphilium aci-
also must be done. dophilum (formerly Thiobacillus acidophilus, 15) oxidizes
sulfur and organic compounds such as sugars. Acidimi-
crobium ferrooxidans cannot only live autotrophically on
MICROORGANISMS
iron(II) ion oxidation but also grow mixotrophically or
even chemoorganotrophically, if yeast extract is added
The bacterially catalyzed oxidation of metal sulfides
to the cultures. The oxidation of sulfur compounds and
(mediated by iron(III) ions) was discovered only half a
iron(II) ions is not limited to organisms growing in an
century ago (11,12). Because of the rod-shaped morphology
aerobic environment but has also been observed in the
of the cell and its ability to oxidize both reduced iron
presence of alternative electron acceptors others than
and sulfur compounds, this first isolate of a bioleaching
oxygen. Some species of the genus Acidithiobacillus and
bacterium was named Thiobacillus ferrooxidans. This
Acidiphilium have been found to oxidize elemental sul-
systematic name is no longer valid and the species is
fur and other sulfur compounds with iron(III) under
now called Acidithiobacillus ferrooxidans (13). Leaching
anaerobic conditions (14). The reduction of nitrate can
bacteria have been detected all over the world and
be coupled to bacterial oxidation of both reduced sulfur
representatives are found in both bacterial domains
compounds and iron(II) ions (16,17), however, these latter
(i.e., the Archaea and the Bacteria) (14). Bioleaching
processes have so far not been demonstrated in acidophilic
research and application focus on the metal and sulfur
leaching bacteria. Generally, for the characterization of
compound–oxidizing bacteria. In the following, the biology
the bacterially catalyzed reactions, concerning the sulfur
of these classic leaching bacteria is described. A short
and iron cycles in leaching biotopes, aerobic and anaer-
characterization of the only recently investigated metal-
obic processes have to be considered. In addition to the
complexing chemoorganoheterotrophic microorganisms is
anaerobic reactions mentioned earlier, the reduction of
also given.
sulfate, tetrathionate, thiosulfate, and elemental sulfur
may play an important role in anoxic zones. Further-
Biology of Metal Sulfide Oxidizing Bacteria
more, the disproportionation of elemental sulfur and other
At first sight, the biology of the acidophilic metal and sul- sulfur compounds is discussed as energy-yielding path-
fur compound oxidizing bacteria seems to be complicated. ways to support bacterial growth. Bioleaching occurs in
The bacteria can use the oxidation of metal sulfides as arctic regions with temperatures below 0 ° C and in hot
an energy source and thrive in heavy metal–containing springs at temperatures around 100 ° C. Although, leach-
solutions at pH 2 (14). However, the main metabolic path- ing sites with cold and moderate temperatures (up to
ways do not differ significantly from those found in 40 or 50 ° C) are dominated by the mesophilic, eubacte-
other bacteria and the most investigated leaching bac- rial genera Acidithiobacillus and Leptospirillum, other
terium, A. ferrooxidans, is a close relative of the most leaching bacteria thrive at elevated temperatures. Moder-
thoroughly characterized living organism, that is, the ately thermophilic genera are the eubacteria Sulfobacillus
γ -proteobacterium and gut inhabitant Escherichia coli. and Acidimicrobium that oxidize metal sulfides at tem-
Nevertheless, the adaptation to environments, such as peratures up to 60 ° C. The upper temperature range of
natural, sulfur-rich hydrothermal springs and mining bioleaching is exclusively occupied by the extremely ther-
sites created by humans, gives the leaching bacteria mophilic archaebacterial genera Sulfolobus, Acidianus,
a combination of several biotechnologically interesting Metallosphaera, and Sulfurococcus (14). The latter two
physiological abilities. One of the major characteristics groups of bacteria are of great interest for the design of
of these bacteria is their ability to oxidize inorganic sul- new bioleaching plants and, therefore current research
fur compounds to sulfuric acid and/or iron(II) to iron(III) is focused on the metal sulfide oxidation potential of
ions. In several bacteria, these abilities are found in these thermophiles (18). If the operating temperature in
combination, for example, A. ferrooxidans, Sulfobacillus stirred tank reactors could be permitted to increase to
 BIOLEACHING 635

60 or even 80 ° C, cooling costs would be significantly enzymatically, to sulfate and protons. There is a long-
reduced (see Section on Applications). Furthermore, recal- standing debate in the literature about the mechanism
citrant metal sulfides such as chalcopyrite become more of bioleaching, whether cells are able to oxidize the
readily oxidizable at elevated temperatures. To date, com- sulfur moiety at the metal sulfide surface enzymatically
mercial bioleaching processes operate at temperatures (‘‘direct mechanism’’) or whether the primary oxidant are
below 50 ° C (2). The bacterial communities in these reac- iron(III) ions and the role of bacteria is to reoxidize the
tors are dominated by Leptospirillum-like bacteria and resulting iron(II) ions and sulfur compounds (‘‘indirect
Acidithiobacillus caldus (formerly Thiobacillus caldus, 13) mechanism’’) (23). The differentiation between direct and
rather than by A. ferrooxidans and A. thiooxidans (19). For indirect bioleaching was developed at a time when
the latter, the conditions in the continuous-flow bioreac- only A. ferrooxidans and A. thiooxidans were known
tors with constantly high redox potentials above 700 mV as leaching bacteria. As soon as L. ferrooxidans (24)
(SHE), pH around 1.5, and temperatures above 40 ° C are was discovered, a serious contradiction became evident:
not conducive for growth. Defined inocula of pure or mixed the cells of this bacterium dissolve metal sulfides
bacterial cultures are not applied to industrial bioleach- exclusively by regeneration of the iron(III) ions; no sulfur
ing processes. Mostly, poorly characterized enrichment compounds can be catabolized by this organism. The direct
cultures or endemic bacteria are used. More thoroughly mechanism is considered disproved because of these and
selected or even genetically engineered organisms may other contradictions (10,23) and, especially, because the
have a potential to optimize leaching operations; however, required enzyme system has never been demonstrated.
the industrial applicability of these bacteria still remains The most widely agreed scheme of bioleaching integrates
to be demonstrated. Besides the bacteria described pre- findings from microbiological, chemical, electrochemical,
viously, leaching biotopes are inhabited by acidophilic, and mineralogical studies. Together with images obtained
chemoorganoheterotrophic bacteria and fungi, as well as by light microscopy, and scanning and transmission
by protozoa, grazing on bacteria, and by photoautotrophic electron microscopy (SEM, TEM), as well as atomic force
algae (14,20). The primary (biomass) producers in these microscopy (AFM), a holistic theory on bioleaching of
generally oligotrophic biotopes are the chemolithoau- metal sulfides has been developed (10,25,26). Accordingly,
totrophs because, in most cases, the biomass produc- observations regarding bacterial attachment, reaction
tion of phototrophs is negligible. Interactions between pathways, and kinetics fit together.
chemolithoautotrophs and chemoorganoheterotrophs may
beneficially affect leaching activity. As leaching bacteria Attachment of Bacteria to Mineral Surfaces
usually excrete low molecular weight organic compounds,
which inhibit their own growth, the presence of organ- Leaching bacteria attach to metal sulfide surfaces (Fig. 3).
otrophs can prevent the accumulation of these compounds From biofilm research it is known that attachment occurs
and enhance leaching performance (14). via diffusion, convection, and chemotaxis. Although the
first two are mostly random processes, chemotaxis is the
Biology of Chemoorganoheterotrophic Leaching active orientation of the bacterium in a chemical gradient.
Microorganisms The bacterium is able to detect the site of active metal

Bacteria, algae, fungi, and consortia such as lichens have

the potential to extract metals of interest from nonsulfide
ores or metal-containing industrial wastes such as fly ash
and slag. A general characteristic of these organisms is the
production and excretion of high amounts of organic acids
and other metal-complexing compounds, basically because
of an imbalance of their metabolism. Mainly organic
acids are relevant, such as oxalic, citric, malic, succinic,
pyruvic, acetic, glucuronic, and amino acids. Furthermore,
they exhibit a rather high heavy metal resistance. These
abilities are widespread among chemoorganoheterotrophic
species and such diverse genera as Bacillus, Pseudomonas,
Penicillium, and Aspergillus, are investigated in this
context (21,22).
2
MECHANISMS OF MICROBIOLOGICAL METAL SULFIDE
DISSOLUTION 4

Dissolution of metal sulfides is achieved by acid attack 6

and the oxidation of the sulfur moiety. The first oxidation
step is carried out by an electrochemical surface reaction 8 µm
of iron(III) ions with the sulfide species. Subsequently, Figure 3. Atomic force microscopic image of a cell of
the released intermediary sulfur compounds are further Acidithiobacillus ferrooxidans attached to the surface of a pyrite
oxidized by iron(III) ions or oxygen, abiotically or crystal.
636 BIOLEACHING

sulfide dissolution by sensing the resulting gradients 2FeS 2

(e.g., dissolved iron(II) ions and sulfate) and actively 18H2O 12Fe(H 2O)63+
swimming in the direction of an increased concentration + 14Fe(H 2O)62+
12H
of the attracting ions (27). There, the bacterial cell 2− 2Fe 3+
2S2O3
attaches to the metal sulfide surface by mostly unknown 2Fe 2+
processes. The extracellular polymeric substances (EPS),
SO42 − + 2H +
surrounding the cells play an important role in the (2x)
attachment of the leaching bacteria. In the cases of
A. ferrooxidans and L. ferrooxidans grown on pyrite, the
2− H2O 2−
exopolymers contain, besides neutral sugars and lipids S 3O 6 S4O6
as their major quantitative constituents, some glucuronic H2O
acid residues and complexed iron(III) ions (26,28). The +
6Fe 2+ 6H
stoichiometry of glucuronic acid to iron(III) ions amounts 3H2O
to a molar ratio of 2 : 1 for the strains of A. ferrooxidans 6Fe 3+
and L. ferrooxidans tested so far (26,28). This indicates (or 1.5O2) SO42 − + 2H +
2−
that regardless of the phylogenetic class of the leaching S3O3
bacteria, the physicochemical characteristics of the
+ S2O32− + 2Fe 3 + + S4O62−
environment enforced a similar adaptation. According to
the stoichiometry, a net positive charge can be attributed (or 0.5O2 + 2H +)
to the EPS of the leaching bacteria. At the extremely acidic S5O62− + 2Fe 2+ S2O32− + S5O62−
(or H 2O) 0.25S 8 + SO 32−
pH of the leaching environment, pyrite has a net negative
charge. Thus, an attachment of a positively charged
bacterium to a negatively charged surface may take place 2Fe 3+ + H 2O
(or 0.5O2)
because of an electrostatic interaction between the two
2Fe 2 + + 2H +
surfaces. Once the bacterium has attached, additional
processes cause an intimate and tight adhesion to the SO 42−
respective mineral surface. These processes have not yet
Figure 4. Cycle of oxidative pyrite degradation by bacterial
been elucidated. Besides, nothing is known about the and/or chemical leaching (29).
specificity of the EPS for various metal sulfides. To date,
only data for EPS of pyrite- and of sulfur-grown cells
are available. They indicate that the bacteria change the and elemental sulfur, as indicated in Figure 4. Some
composition of their EPS according to the substrate or of the reactions occurring in the cyclic thiosulfate
substratum, which they use. In the EPS of A. ferrooxidans degradation have been described to be mediated by
grown on sulfur, more lipids, less neutral sugars, and only enzymes. For example, the enzyme for the oxidation of
traces of glucuronic acid were present as compared to those thiosulfate to tetrathionate (thiosulfate dehydrogenase)
of pyrite-grown cells (26). has been isolated from cells of A. ferrooxidans (30), as
has been the enzyme for the hydrolysis of tetrathionate
Reaction Pathways (tetrathionate hydrolase) from cells of A. ferrooxidans and
Experimental results indicated that reaction pathways A. thiooxidans (31–33). The extent to which enzymatic
of metal sulfide dissolution are not determined by the catalysis contributes to thiosulfate degradation has to
crystal structure (10,25). The reactivity of the minerals be clarified. Yield calculations, using the amount of
with protons is the only relevant criterion. For instance, biomass produced in the course of pyrite dissolution,
acid-nonsoluble monosulfides such as molybdenite and indicated that some of these reactions must release
tungstenite are degraded by the same mechanism as electrons at a level sufficient for cytochrome b/flavine
the acid-nonsoluble disulfide pyrite. In contrast, the reduction (34). This would consequently allow for two
structurally similar disulfides pyrite and the acid-soluble oxidative phosphorylations, whereas electrons at the
hauerite are dissolved by different mechanisms. cytochrome c level, as provided by iron(II) ions, would
give only one ATP per mol of oxidized substrate. The
Thiosulfate Pathway. The following reaction mechanism increased growth yields, as measured, are explainable
describes the degradation of acid-nonsoluble metal sulfides only with this assumption. The question whether the
such as pyrite, molybdenite, and tungstenite (25,29). tetrathionate degradation is dominated by abiotic or by
Pyrite is dissolved via electron extraction by hydrated enzymatically catalyzed reactions is important because
iron(III) ions from the crystal lattice according to these reactions determine the end products: whether a
equation (1): considerable amount of elemental sulfur is formed or
mainly sulfate and protons. In the first case, acidification
FeS2 + 6[Fe(H2 O)6 ]3+ + 9H2 O → S2 O3 2− + 7[Fe(H2 O)6 ]2+ in ARD could be reduced. In the second situation, full
sulfuric acid production would enhance ARD problems. A
+ 6H+ (1) regulation and control of this reaction is, furthermore, of
high value for industrial application of pyrite bioleaching
Thiosulfate subsequently yields via a series of reactions before gold extraction by cyanidation. The presence of
either sulfate and protons or higher polythionates elemental sulfur in the leach residue causes unnecessary
 BIOLEACHING 637

cyanide consumption as a result of the formation of bacteria because enzymatic catalysis is 106 times faster
isothiocyanates. than abiotic oxidation (40). These two subprocesses are
linked by the redox potential (41). Because in leaching
Polysulfide Pathway. The mechanism for metal sul- experiments only a control of the bulk solution is
fide degradation with the main intermediate polysulfide possible, differences existing between the conditions
is valid for acid-soluble sulfide minerals, for example, in the bulk compared to the solid–liquid interface,
hauerite, sphalerite, galena, chalcopyrite, or arsenopy- for example, concentrations of iron species, pH, and
rite (25). Here the attack on the crystal lattice (MS) is oxygen, could be kinetically relevant. Such differences
carried out by a combined action of protons and iron(III) might occur if a delimited reaction space is assumed
ions. Protons induce polarization of a surface sulfide ion either because of solid reaction products remaining
and enables its liberation from the crystal lattice concomi- at the mineral surface (e.g., formation of a layer of
tantly with an electron transfer to an iron(III) ion (25). elemental sulfur or iron precipitates) or because of the
The first free intermediate of this attack is a H2 S+ cation defined space, created by attached bacteria and the
radical, which after a series of reactions ends in a polysul- mineral surface (25,26). Solid–liquid, solid–solid–liquid,
fide, H2 Sn (35). The latter compound decomposes to H2 S and solid–biofilm–liquid interfaces should be considered.
and elemental sulfur (S8 in the case of H2 S9 ). The reactions Mainly concentration gradients of reactants and/or
may be summarized as shown in Eq. (2). reaction products, as well as reduction of diffusion rates
Subsequently, bacteria regenerate the protons by are the consequence but also other mechanisms of diffusion
further oxidizing the elemental sulfur to sulfuric acid. control can be imagined. Furthermore, in the case of
intergrown sulfide minerals galvanic effects may cause
Hydrogen Sulfide Pathway. In the absence of iron(III) preferential dissolution. Hence, dissolution kinetics is
ions, readily acid-soluble metal sulfides such as sphalerite strongly dependent on the purity of the investigated
can be dissolved by simple acid attack, liberating hydrogen materials.
sulfide. Sulfur compound oxidizing bacteria oxidize it to Recent work focused on the comparison of abiotic and
sulfate and release protons, which can enter the cycle bioleaching of sphalerite and pyrite at electrolytically con-
again to dissolve additional metal sulfide. Hydrogen trolled redox potentials. Abiotic leaching was performed
sulfide is oxidized stoichiometrically to elemental sulfur at similar high iron(III) to iron(II) ratios as are typical
in A. thiooxidans (36). The reaction is probably catalyzed for bioleaching, where recycling of the iron(III) ions is
by a sulfide : quinone oxidoreductase because the gene enzymatically catalyzed. Comparative experiments with-
sequence of this enzyme has been detected in the genome out redox control showed that bioleaching rates were
of A. ferrooxidans (37). up to 20 times higher than in sterile controls started
at the same initial amounts of iron(III) ions (42). How-
Galvanic Influences ever, in the absence of bacterial iron(III) reoxidation,
the redox potential decreased rapidly as a result of
Besides the direct attack of iron(III) ions, protons, and the reduction of the iron(III) ions by the metal sul-
complexing agents, in the case of intergrown crystals
fide. Therefore, solution conditions in bioleaching exper-
of different minerals, galvanic interactions have to
iments and sterile controls were not identical during
be considered (9). Because of their different electrode
the entire experiments and the data obtained are not
potentials, a polarization can occur between different
comparable.
metal sulfides in an ore, if these are in direct electrical
contact. The sulfide with the lower potential functions
Leaching of Sphalerite in Acidic Iron(III) Solutions. The
as the anode, whereas the metal sulfide with the higher
kinetics of sphalerite dissolution under potentiostatic con-
potential represents the cathode. At the anode, the metal
ditions is determined by two processes. High leaching rates
sulfide is dissolved by electron transfer to the cathode
are achieved at high redox potentials where no signifi-
where iron(III) ions are then reduced without oxidation of
cant difference between bacterial and abiotic leaching is
the cathode itself. As long as the electrical contact between
observed. In this case, the dissolution kinetics is controlled
the different sulfides exists, the sulfide with the higher
only by the electrochemical reaction at the sphalerite sur-
electrode potential is protected against oxidant attack and
face (43). According to the polysulfide mechanism, this
is dissolved at slower rates than the isolated (galvanically
reaction consists of the proton-mediated electron extrac-
unprotected) sulfide and vice versa. For example, pyrite
tion by iron(III) ions, which yields elemental sulfur as the
could be protected by intergrown chalcopyrite or sphalerite
first stable intermediate (Eq. (2), (25)). This could form
that have lower electrode potentials (38,39). Hence, these
a layer on the solid surface if no bacterial sulfur oxi-
electrochemical effects may influence considerably the
dation occurs. However, only at low leaching rates (low
dissolution rates.
redox potentials) the kinetics becomes diffusion-controlled
in the absence of sulfur-oxidizing bacteria (44). Con-
Kinetics of Metal Sulfide (Bio)leaching
sequently, sulfur compound–oxidizing bacteria enhance
The kinetics of metal sulfide dissolution in the presence sphalerite dissolution at low redox potentials even under
of iron(III) ions is controlled by the charge-transfer potentiostatic conditions by preventing the accumulation
reactions at the mineral surface (the shrinking-core of elemental sulfur on the sphalerite (44). Generally,
model has to be considered). Reoxidation of the iron(II) the formation of a diffusion barrier of elemental sulfur
ions in acidic solutions is primarily performed by has to be taken into account in all cases in which the
638 BIOLEACHING

polysulfide mechanism is involved (e.g., for the leaching mined and processed besides concentrates. With the easy-
of chalcopyrite). to-mine rich surface ore deposits becoming exhausted,
bioleaching is the alternative for metal winning from pre-
viously uneconomic ores. Secondly, in comparison with
2 H+ M2+ H+ H+
the nonbiological ore-processing techniques, bioleaching
stands for low energy–consuming processes with reduced
MS [H2S* + HS* H 2 Sn ] S8
soil, water, and air pollution problems. Environmental
laws increasingly regulate the emission of sulfur dioxide
Fe3+ Fe2+ Fe3+ Fe2+
and heavy metals in the flue gas from smelters in many
(2)
countries. Consequently, bioleaching or biohydrometal-
lurgy turn out to be an important alternative to the
Leaching of Pyrite in Acidic Iron(III) Solutions. In
conventional pyrometallurgy (1). In addition, no proper
contrast to the dissolution kinetics of sphalerite, the
pyrometallurgical technique is available for processing
rate-determining step of pyrite leaching is always the
of complex concentrates (Cu−Zn−Pb−Ni−Co), therefore
electrochemical reaction at the mineral surface. Because
hydrometallurgical alternatives to smelting are increas-
pyrite is oxidized via the thiosulfate pathway with water-
ingly investigated. It is well known that zinc and lead
soluble tetrathionate as the first detectable intermediate,
contained in copper concentrates and copper contained
no diffusion-controlling sulfur layer can develop on the
in lead-zinc concentrates are undesirable because they
mineral surface. The surface reaction is dependent not
generate major problems in the specific pyrometallurgical
only on the redox potential, as it is for sphalerite
processes.
oxidation in acidic solutions, but it is also influenced by
the pH (45) because the oxygen incorporated into sulfate Current Biomining Techniques
originates from water (46), probably via hydroxyl ions (46).
In potentiostatic and pH-static experiments, bacterial Microbiological leaching is used for two different purposes.
pyrite leaching was enhanced over sterile controls by First, it is applied to mobilize the metal species of
a factor of about 1.3 (48). As the reaction orders with interest from insoluble compounds. For metal sulfides
respect to iron(III) ions are identical in biotic and abiotic and uranium(IV) oxides, the formation of dissolved metal
experiments but differ with respect to pH, it was concluded ions is achieved by oxidation processes. Furthermore,
that bacteria enhance pyrite dissolution by increasing the metals can be extracted from nonsulfide ores and
the pH at the mineral surface, whereas the bulk pH is industrial wastes containing metal oxides, by microbially
kept constant. This pH increase could be caused by the produced inorganic and organic acids. To date, the
combined actions of bacteria and their EPS in the biofilm, latter technique is not applied on an industrial scale,
that is, the proton consumption during bacterial iron(II) whereas the solubilization of base metals such as copper
oxidation and the buffering properties of the EPS (48). by bacterial oxidation of the corresponding sulfides
Additionally, the positively charged EPS of the leaching plays an increasingly important role in commercial
bacteria (26) may obstruct influx of protons from the more mining operations. Furthermore, bioleaching is applied to
acidic bulk solution. remove interfering metal sulfides before the conventional
extraction of gold and silver from refractory ores by
cyanidation or other solubilizing reagents. In refractory
APPLICATIONS ores, the gold and silver are entrapped within the metal
sulfide matrix such as pyrite and arsenopyrite. Because
The industrial application of bioleaching for the winning the precious metals must be reached by the leaching
of metals is often called biomining. At the end of the twen- solution to become oxidized and subsequently complexed
tieth century, it has become one of the main processes by cyanide, they need to be at least partially liberated from
used in biotechnology (1). The gigantic dimensions of the the surrounding compounds. As a result of the small size
bioreactors at the Ashanti plant in Ghana for the pre- of the encased particles, even intensive grinding is often
treatment of gold ores (about 1,000 t ore concentrate are insufficient, and the interfering metal sulfides have to
processed per day in 24 stirred tanks of 900 m3 each be oxidized by conventional pyrometallurgy (i.e., roasting,
since 1995, Figure 1), mentioned earlier (2), even sur- smelting, or pressure leaching), nitric acid oxidation, or
passed the biological steps of sewage treatment plants bioleaching. In the last decade, the latter technique has
as the largest bioreactors on Earth. It has been estimated turned out to be economically and environmentally the
that bioleaching is yielding up to 25% of the world’s cop- most beneficial option (2). The pretreatment of gold ores
per and about 10% of the uranium production (49,50). by bioleaching even allows to process ores from deposits
The total value of the products generated by this tech- that were previously considered uneconomical.
nology was assessed at U.S. $10 billion in 1998 (51). One Several simple, low-cost techniques are applied for
of the major reasons for the increasing use of biomining the winning of metals from low-grade sulfidic ores by
is that conventional means such as ore dressing, followed bioleaching (1,9,52). Generally, an acid leaching solution
by smelting techniques are uneconomical for low-grade with a pH around 2, containing the appropriate bacteria
ores. The oxidation efficacy of metal sulfides by bioleach- is pumped through an ore-bearing material. The metal-
ing bacteria is nearly independent of the metal content, containing (pregnant) leachate is collected for recirculation
therefore extremely low-grade ores, in the case of copper, or further processing. In underground and in situ leaching,
typically containing less than 0.3% of the metal, can be the metal sulfide oxidation takes place directly in the
 BIOLEACHING 639

natural ore body. Leaching solutions are pumped into a decompose the metal sulfide matrix of recalcitrant gold and
fractured deposit. The pregnant leachate is collected at silver ores to facilitate the contact of the encased precious
impermeable sites below the ore body or pumped from metal particles with the cyanide solution. If the interfering
new boreholes back to the surface. This technique is often metal sulfides contain chalcopyrite, the tank leaching
used for the solubilization and winning of uranium(VI) process could combine both the winning of the base metal
ions from underground uranium ore deposits. In dump (copper) and the extraction of the precious metals.
leaching, the ore material is piled in heaps 10 to 20 m
high on an impermeable base (liner, clay, and so on) and Research and Future Applications
irrigated by sprinklers or flooded temporarily (Fig. 2). The
drainage is collected and recirculated several times. Dump Today, bioleaching research focuses on the optimization of
leaching is the oldest bioleaching technique. It is generally the biooxidation processes in stirred tank reactors. With
used to oxidize metal sulfides in waste rock (run-of-mine) tank bioleaching becoming an accepted alternative for the
that was deposited in the course of exploiting ore bodies pretreatment of recalcitrant gold ores in countries such
for high-grade deposits. Furthermore, it was uneconomical as South Africa, Ghana, Brazil, and Australia, the mining
to process this rock by conventional techniques. Both the industry also intends to use bioleaching tanks for the
underground and in situ and the dump leaching methods recovery of base metals such as copper, cobalt, nickel,
are rather inefficient processes with low oxidation rates and zinc (1). For instance, the Kasese Cobalt Project in
and low total metal yields. The leaching solution often Uganda represents the first large-scale cobalt bioleaching
cannot uniformly percolate the underground deposit or plant (3). The development of bioleaching plants, as with
the dump because of the low degree of fragmentation every large-scale biotechnological process, has scale-up
of the ore body or waste material. The conditions for problems. Laboratory experiments in shake flasks and
bioleaching cannot be rigorously controlled, especially percolation columns are often used to determine optimum
insufficient oxygen supply will limit the dissolution. A particle size, ore composition, and other bioleaching
more sophisticated low-cost bioleaching technique is called parameters but cannot simulate the real conditions in
heap leaching. The process is similar to dump leaching a stirred reactor with 1,000 m3 ore pulp. The conditions in
but the dimensions of the pile are limited to about 2 m the final design still have to be conducive for the bacteria,
in height, to facilitate oxygen diffusion. Furthermore, for example, agitation and aeration should balance the
usually fine-ground ores or even concentrates are used. oxygen demand without killing the bacteria by detrimental
Before piling, the ore material is mixed with sulfuric shear forces. In addition, the bacterial process has to be
acid to form an agglomerate. This process prevents the integrated and linked to the other nonbiological units of
segregation of fine and coarse particles and causes a the plant, for example, the upstream and downstream
uniform permeability of the heap. For the circulation processing. Bacterial demands on the feed and the
of the leaching solution, drip-type irrigation above and composition of the leaching product can have a significant
below the heap surface instead of sprinklers are installed influence on the total metal recovery process. For instance,
to prevent cooling of the heaps by evaporation. The if the leaching solutions are to be recycled to reduce process
leachate is collected by a drainage tubing system above water demand, toxic substances added (e.g., in the course
an impermeable liner. Heap leaching is mainly used of solvent extraction) have to be sufficiently removed.
for the winning of copper from chalcocite and covellite. Besides reactor design and total flow sheet development,
Once concentrations of 1 g/L for copper ions have been one of the major current research topics deals with
achieved in this pregnant solution by continuous cycling, the operating temperature in bioleaching plants. To
techniques such as solvent extraction and electrowinning date, because of the generally applied mesophilic and/or
are used to produce cathode copper. In contrast to the moderately thermophilic leaching bacteria, bioleaching
above-mentioned low-cost techniques for the winning of in stirred tanks is limited to temperatures of 40 to
base metals from low-grade ores, more expensive methods 50 ° C. As a result of the high sulfide oxidation rates,
can be applied where high-grade ores and/or precious and consequently, high heat evolution obtained in these
metals such as gold and silver are concerned. For this reactors, cooling contributes significantly to the total
purpose, stirred-tank reactors are used (1,9,52). Tank or operating costs. Therefore, the development of bioleaching
reactor leaching operate with a sophisticated process processes that can operate at 60 to 80 ° C are of
control for optimum leaching conditions. In the newest great interest. Besides the reduction of cooling costs,
tank bioleaching plants, several huge reactors with single bioleaching activities would be enhanced at these elevated
volumes of about 1,000 m3 are continuously fed with temperatures and metal sulfides difficult to leach, such
fine-ground, high-grade concentrates at pulp densities as the commercially important chalcopyrite, could be
of up to 20% (w/v). The main operating costs result biooxidized with sufficient efficacy (53). In addition to
from the highly energy-consuming stirring and aeration the extraction of base and precious metals, bacterial
of the leaching suspension. In addition, because of the metal sulfide oxidation could also be applied for the
high oxidation rates that are achieved with the sulfide- depyritization of coal before combustion, thus drastically
rich concentrates (>20%, w/w), cooling systems have to reducing sulfur dioxide emission from power plants,
remove the evolving heat of the exothermic oxidation especially in cases where expensive flue gas cleaning is
reactions to maintain a temperature below 50 ° C. This not available. Despite several positive feasibility studies,
is the temperature limit of the usually applied moderately this interesting technique has not been tested yet beyond
thermophilic bacteria (2). Tank leaching is mainly used to pilot-scale dimensions (54).
640 BIOLEACHING

In contrast to the biooxidation of metal sulfides, Radon, formed as an intermediate of uranium (238 U) decay
other bioleaching methods for the winning of metals are to lead (206 Pb), usually remains in the mineral in which the
at present not applied commercially (21). Metals from uranium is located. However, if the uranium is enclosed
nonsulfide ores, that is, metal oxides, carbonates, and into pyrite or other metal sulfides, bioleaching may result
silicates, and from industrial wastes, containing metal in the liberation of this gas to the atmosphere. Because
oxides such as fly ash, incineration slag, and contaminated radon (222 Rn, half-life time of about 4 d) may be taken up
soil, can be extracted by inorganic and organic acids and for example, by humans and in the worst-case decay to a
alcohols. For instance, sulfuric acid produced from the solid radioactive isotope of lead (210 Pb, half-life time 21 a),
bacterial oxidation of inexpensive elemental sulfur has it may contaminate the lung for decades with an increased
been used successfully to solubilize copper, cadmium, level of radioactivity. Besides these processes, which are
and zinc from solid-waste incineration fly ashes (55). directly and/or indirectly related to oxidative metal sulfide
Furthermore, the potential of heterotrophic bacteria and dissolution, reductive microbiological processes may also
fungi to produce organic compounds with good metal contribute to pollution with heavy metals. Manganese(IV)
complexation properties is under investigation (22,56). and iron(III) ion–reducing bacteria thrive under anaerobic
conditions, oxidizing organic substances present and
ENVIRONMENTAL PROBLEMS CONNECTED WITH reducing the inorganic heavy metal ions to their soluble
BIOLEACHING forms, manganese(II) and iron(II) ions, respectively. These
tend to migrate. In addition, other coprecipitated metals
The discovery of bioleaching resulted from research become mobile too, causing a spreading of heavy metal
conducted on an environmental problem: the acidic pollution, in fact due to the availability of degradable
drainage, containing heavy metals, which resulted from a organic matter. This natural process is believed to be the
waste heap of a coal mine (11). Consequently, bioleaching major process for mineralizing organic matter in anaerobic
does not only help in an economic production of sediments (23,58,59,60) and thus cannot be avoided.
valuable and precious metals but also causes serious Countermeasures for problems connected with ARD/
environmental problems known as AMD/ARD. Leaching AMD deal mainly with a reduction of oxygen and/or
bacteria are ubiquitous whenever metal sulfides become water access to the metal sulfides because these are the
exposed to air and water. Most of these bacteria are two essential compounds needed for (bio)degradation. For
chemolithoautotrophs, hence, only carbon dioxide from the this purpose, sophisticated systems have been developed,
atmosphere is needed for cell carbon generation (fixation). including multilayered covers, geotextiles, clay, limestone,
The degradation products of bioleaching, sulfuric acid and organic substrates, antimicrobials, or underwater storage
dissolved heavy metals, if not neutralized and precipitated (also artificial dams). The use of reactive, neutralizing
by endogenous buffer material such as carbonates, will walls only stops the underground migration of unavoidably
enter into the soil and/or the water paths and cause released acidity and heavy metals. Often, these systems
environmental pollution. In the case of a sufficiently function well for a while but fail unpredictably some
high amount, all neutrophilic life will vanish, leaving time in the future. Too often, microbiologists have not
an acidophilic ecosystem. In addition, the acidity may been consulted in the design of these systems to improve
seriously reduce the fertility of the affected soils and their reliability. The interested reader can find plentiful
waters. This has happened to the river Rio Tinto (Spain) information on these subjects (61–63).
as a result of mining activities for more than two thousand
years, or naturally in the geothermal areas of Yellowstone BIBLIOGRAPHY
National Park (U.S.A.) and similar environments. Soils,
rivers, and lakes polluted in this manner serve as a source 1. D. E. Rawlings, ed., Biomining, Springer-Verlag, Berlin,
of heavy metals for incorporation into the food chain, either Germany, 1997, p. 302.
via plant uptake (and accumulation) or via predation, 2. D. W. Dew, E. N. Lawson, and J. L. Broadhurst, in Biomin-
grazing, and so on. Finally, the heavy metals contaminate ing, D. E. Rawlings, ed., Springer-Verlag, Berlin, Germany,
food products intended for consumption by humans. 1997, pp. 45–80.
Another connected problem is the generation of 3. A. P. Briggs and M. Millard, International Biohydrometal-
dust and fine particles. Because of the biological lurgy Symposium/Biomine 97, Australian Mineral Founda-
degradation of metal sulfides, the rock containing these tion, Glenside, 1997, pp. 6.4.1–6.4.12.
compounds becomes fragmented and breaks down to small 4. Bertoldo et al., this volume, pp. ??–??.
(microscopic) particles, especially under a combined action 5. H. A. Schnell, in D. E. Rawlings, ed., Biomining, Springer-
with water and freeze-thaw attack (57). These become Verlag, Berlin, Germany, 1997, pp. 21–43.
wind-borne and because they may still contain some heavy 6. H. Strunz, Mineralogische Tabellen, Akademische Verlagsge-
metals, they may pollute the environment as dust. Often, sellschaft Geest & Portig K. G., Leipzig, Germany, 1978.
tailings that have not properly been protected against 7. K. Laajalehto, E. Suoninen, and S. Heimala, Int. J. Miner.
wind and water erosion are located within inhabited areas Process 33, 95–102 (1990).
or even within cities (e.g., Johannesburg, South Africa). 8. P. J. Harris and K. Richter, in K. S. E. Forssberg, ed., Flota-
The dust causes heavy metal pollution, in addition to tion of Sulphide Minerals, vol. 6, Elsevier, Amsterdam, The
respiratory diseases like false (mild) croup. In the case of Netherlands, 1985, pp. 141–157.
radioactive wastes, another problem occurs: the liberation 9. G. Rossi, Biohydrometallurgy, McGraw-Hill, Hamburg, Ger-
of usually entrapped radon gas into the atmosphere. many, 1990, p. 609.
 BIOLEACHING OF METALS 641

10. W. Sand, T. Gehrke, P.-G. Jozsa, and A. Schippers, Hydro- 45. P. R. Holmes and F. K. Crundwell, Geochim. Cosmochim.
metallurgy 59, 159–175 (2001). Acta 64, 263–274 (2000).
11. A. R. Colmer and M. E. Hinkle, Science 106, 253–256 (1947). 46. B. E. Taylor, M. C. Wheeler, and D. K. Nordstrom, Nature
12. K. L. Temple and A. R. Colmer, J. Bacteriol. 62, 605–611 308, 538–541 (1984).
(1951). 47. C. O. Moses, D. K. Nordstrom, J. S. Herman, and A. L. Mills,
13. D. P. Kelly, Int. J. Syst. Evol. Microbiol. 50, 511–516 (2000). Geochim. Cosmochim. Acta 51, 1561–1571 (1987).
14. D. B. Johnson, FEMS Microbiol. Ecol. 27, 307–317 (1998). 48. T. A. Fowler, P. R. Holmes, and F. K. Crundwell, Appl. Envi-
15. A. Hiraishi et al., Int. J. Syst. Bacteriol. 48, 1389–1398 ron. Microbiol. 65, 2987–2993 (1999).
(1998). 49. D. E. Rawlings and S. Silver, BioTechnology 13, 773–778
16. K. Baalsrud and K. S. Baalsrud, Arch. Microbiol. 20, 34–62 (1995).
(1954). 50. C. L. Brierley, Sci. Am. 246, 44–53 (1982).
17. K. L. Straub, M. Benz, B. Schink, and F. Widdel, Appl. 51. D. B. Johnson, in R. Amils and A. Ballester, eds., Biohy-
Environ. Microbiol. 62, 1458–1460 (1996). drometallurgy and the Environment Toward the Mining of the
18. P. R. Norris, N. P. Burton, and N. A. M. Foulis, Extremo- 21st Century, Part A, Elsevier, Amsterdam, The Netherlands,
philes 4, 71–76 (2000). 1999, pp. 645–656.
19. D. E. Rawlings, H. Tributsch, G. S. Hansford, Microbiology 52. H. L. Ehrlich and C. L. Brierley, Microbial Mineral Recovery,
145, 5–13 (1999). McGraw-Hill, New York, 1990, p. 454.
20. A. P. Harrison, Annu. Rev. Microbiol. 38, 265–292 (1984). 53. D. W. Dew, C. van Buuren, K. McEwan, and C. Bowker, in
R. Amils and A. Ballester, eds., Biohydrometallurgy and the
21. K. Bosecker, FEMS Microbiol. Rev. 20, 591–604 (1997).
Environment Toward the Mining of the 21st Century, Part A,
22. P. P. Bosshard, R. Bachofen, and H. Brandl, Environ. Sci. Elsevier, Amsterdam, The Netherlands, 1999, pp. 229–238.
Technol. 30, 3066–3070 (1996).
54. G. Rossi, Fuel 72, 1581–1592 (1993).
23. H. L. Ehrlich, Geomicrobiology, Marcel Dekker, New York,
55. C. Brombacher, R. Bachofen, and H. Brandl, Appl. Environ.
1996, p. 719.
Microbiol. 64, 1237–1241 (1998).
24. G. E. Markosyan, Biol. J. Armenia 25, 26–29 (1972).
56. G. M. Gadd, Adv. Microb. Physiol. 41, 47–92 (1999).
25. A. Schippers and W. Sand, Appl. Environ. Microbiol. 65,
57. W. Sand, P.-G. Jozsa, and R. Mansch, this volume, pp. ??–??.
319–321 (1999).
58. D. R. Lovley, Microbiol. Rev. 55, 259–287 (1991).
26. T. Gehrke, J. Telegdi, D. Thierry, and W. Sand, Appl. Envi-
ron. Microbiol. 64, 2743–2747 (1998). 59. D. R. Lovley, Annu. Rev. Microbiol. 47, 263–290 (1993).
27. J. Acuna et al., FEMS Microbiol. Lett. 75, 37–42 (1992). 60. K. L. Straub, M. Benz, and B. Schink, FEMS Microbiol. Ecol.
34, 181–186 (2001).
28. T. Gehrke, R. Hallmann, and W. Sand, in T. Vargas,
C. A. Jerez, J. V. Wiertz, and H. Toledo, eds., Biohydromet- 61. Bureau of Mines, Proceedings from the Third International
allugical Processing, vol. 1, University of Chile, Santiago, Conference on the Abatement of Acidic Drainage, vol. 1–4.
Chile, 1997, pp. 1–11. U.S. Department of the Interior, Washington, D.C., 1994.
29. A. Schippers, P.-G. Jozsa, and W. Sand, Appl. Environ. 62. Society for Mining, Metallurgy, and Exploration, Proceedings
Microbiol. 62, 3424–3431 (1996). from the Fourth International Conference on Acid Rock
30. M. Silver and D. G. Lundgren, Can. J. Biochem. 46, 1215– Drainage/ICARD 1997, vol. 1–4, Littleton, Col., 1997,
1220 (1968). p. 1934.
31. T. Sugio et al., J. Ferment. Bioeng. 82, 233–238 (1996). 63. Society for Mining, Metallurgy, and Exploration, Proceedings
from the Fifth International Conference on Acid Rock
32. G. A. H. de Jong, W. Hazeu, P. Bos, and J. G. Kuenen,
Drainage/ICARD 2000, vol. 1–2, Littleton, Col., 2000,
Microbiology 143, 499–504 (1997).
p. 1658.
33. T. Tano et al., Biosci. Biotech. Biochem. 60, 224–227 (1996).
34. A. Schippers, T. Rohwerder, and W. Sand, Appl. Biotechnol.
Microbiol. 52, 104–110 (1999). See also METALS: MICROBIAL PROCESSES AFFECTING METALS.
35. R. Steudel, Ind. Eng. Chem. Res. 35, 1417–1423 (1996).
36. C. W. Chan and I. Suzuki, Can. J. Microbiol. 39, 1166–1168
(1993). BIOLEACHING OF METALS
37 T. C. Nübel et al., Arch. Microbiol. 173, 233–244 (2000).
38. V. K. Berry and L. E. Murr, in L. E. Murr, A. E. Torma, and FRANCISCO F. ROBERTO
J. A. Brierley, eds., Metallurgical Applications of Bacterial Idaho National Engineering and
Leaching and Related Microbiological Phenomena, Academic Environmental Laboratory
Press, New York, 1978, pp. 103–136. Idaho Falls, Idaho
39. M. Sato, Geochim. Cosmochim. Acta 56, 1202–1231 (1992).
40. P. C. Singer and W. Stumm, Science 167, 1121–1123 (1970). Bioleaching is the term used to describe the microbial dis-
41. G. S. Hansford and T. Vargas, Hydrometallurgy 59, 135–145 solution of metals from minerals. The commercial bioleach-
(2001). ing of metals, particularly those hosted in sulfide minerals,
42. M. Boon and J. J. Heijnen, Hydrometallurgy 48, 27–41 is supported by the technical disciplines of biohydrometal-
(1998). lurgy, hydrometallurgy, pyrometallurgy, chemistry, elec-
43. T. A. Fowler and F. K. Crundwell, Appl. Environ. Microbiol. trochemistry, and chemical engineering. The study of the
64, 3570–3575 (1998). natural weathering of the same minerals, above and below
44. T. A. Fowler and F. K. Crundwell, Appl. Environ. Microbiol. ground, is also linked to the fields of geomicrobiology and
65, 5285–5292 (1999). biogeochemistry. Studies of abandoned and disused mines
642 BIOLEACHING OF METALS

indicate that the alterations of the natural environment

caused by human activities leave microbiological activity
as remnants, which continues the biologically mediated
release of metals from the host rock known as acid-rock
drainage (ARD). A significant fraction of the world’s cop-
per, gold, and uranium is now recovered by exploiting
native or introduced microbial communities. Although
some members of these unique communities have been
extensively studied over the past 50 years, our knowledge
of the composition of these communities and the function
of the individual species present remain relatively limited.
Nevertheless, bioleaching represents a major strategy in
mineral resource recovery, and its importance will increase
as ore reserves decline in quality, become more difficult to
process (because of increased depth and increased need for
comminution), and as environmental considerations elimi- Figure 1. The Rio Tinto, Spain, passing through the historic
nate traditional physical processes such as smelting, which mining area. Note the ochre-colored water, which is reported to
have served the mining industry for hundreds of years. have dissolved ferric iron concentrations exceeding 2 g/L (1). See
color insert.

HISTORY OF BIOLEACHING

Microbial oxidation of iron and sulfur present in sulfide

minerals is an important part of the natural iron and
sulfur geochemical cycles. This biological weathering of
sulfides, whether it occurs above or below ground, is fre-
quently evident as surface seeps of acidic water, primarily
sulfuric acid that is generated by microorganisms acting
on these deposits.
Humans have exploited these biological phenomena for
centuries. For example, cementation of copper with iron
from naturally occurring acidic leachate in the Rio Tinto
(so-called because of the deep red color resulting from
dissolved ferric iron; Fig. 1) of Spain appears to have been
practiced since the time of the Tartessians 5,000 years
ago (1). The Chinese emperor Liu-An (177 to 120 B.C. (2))
describes this process in detail. Recent analyses of Figure 2. A view of the Roman era iron mines at Parys Mountain,
Greenland ice cores showed increased atmospheric copper Anglesey, Wales. See color insert.
deposition dating back 2,500 years, which is attributable
to crude smelting technologies used by the Romans and
Chinese (3). Similarly, vestiges of iron and copper mining What began as a process used largely for the recovery of
during Roman times can be observed throughout Wales in copper from low-grade mine refuse, or as a means of getting
the United Kingdom (e.g., Parys Mountain on Anglesey, value from old tailings, is now a significant operation at
North Wales; Fig. 2) and across Europe. virtually all open pit mines in the United States where
The Rio Tinto mine has played a prominent role in the billions of tons of low-grade ore are treated (8).
history of hydro- and biohydrometallurgy. It was the first Heap leaching, in which ores are crushed to a defined
documented site to perform heap leaching on an industrial size range, pretreated with acid, acidic ferric iron solutions
scale in the 1700s and the first large-scale open pit copper containing bacteria, or other materials, and then stacked
mine was excavated here. Until the connection between the onto impervious pads, is employed to leach copper and
chemolithotrophic bacterium, Acidithiobacillus ferrooxi- recover gold from refractory ores. This strategy is selected
dans (formerly known as Thiobacillus ferrooxidans; (4)) to overcome a variety of economical considerations, which
and acid mine drainage in Pennsylvania (5,6) was estab- can include the low-grade nature of the ore, long distances
lished, it was generally thought that the Rio Tinto ore was from smelter facilities that would lead to prohibitive
uniquely suited to heap leaching, and this approach was shipping costs, the small size of the projected mine,
not applied to low-grade copper ores in the United States. which restricts the total capital that may be invested
Today it is estimated that 25% of the world’s copper is at the site, and higher smelter costs (penalties) caused by
derived from dump and heap bioleaching. contaminants in the ore (such as mercury). In some cases,
With the association of At. ferrooxidans with sulfide the value of the metal recovered from low-grade reserves
mineral leaching, commercial exploitation of bioleaching can be used to provide additional financial justification for
was quick to follow. Researchers at the Kennecott a project (8).
Bingham Canyon mine near Salt Lake City, Utah, Perhaps the most important development in the
patented a process for biological dump leaching in 1958 (7). bioleaching of sulfide minerals since 1985 has been the
 BIOLEACHING OF METALS 643

deployment and increasing scale of stirred tank biooxida- The acidithiobacilli are gram-negative, obligately aci-
tion plants for the pretreatment of refractory gold ores. dophilic, chemolithotrophic proteobacteria in the gamma
Biooxidation is distinguished from bioleaching in that it subdivision of the proteobacteria. Genetically, they resem-
pertains primarily to the release of particulate gold and ble other γ -proteobacteria such as Escherichia coli in
silver through the partial microbial oxidation of sulfides terms of genetic organization and regulation. The physiol-
in pyrite or arsenopyrite ores hosting the precious metals. ogy (10) and genetics (10,11) of At. ferrooxidans have been
The first biooxidation plant for treatment of refractory reviewed in detail recently, and the reader is directed to
gold ore was brought on-line at the Fairview site in South those excellent reviews for additional details. Although
Africa by GENCOR Process Research (now Billiton Pro- they must inhabit environments in which the pH is below
cess Research; commercial operations are marketed by 3.5, it has been determined that these bacteria maintain
Gold Fields Ltd.) to treat 10 metric tons per day of flota- a circumneutral internal pH by actively pumping pro-
tion concentrate. The success of the initial operation led tons against the steep pH gradient between the cytoplasm
to the plant being expanded to treat the entire mine and the external environment (12). Although these organ-
concentrate production of 40 metric tons in 1991, and the isms have been the focus of intense study in relation to
elimination of the previous roaster technology, with sub- the microbial dissolution of sulfides, another phylogeneti-
stantial environmental and economical benefit (9). Today cally unrelated eubacterium, Leptospirillum ferrooxidans,
one of the world’s largest contained bioprocesses oper- is perhaps even more important in natural and engineered
ates at the Ashanti Gold Fields Obuasi mine in Ghana, leaching environments.
where 960 tons of ore concentrate per day have been pro- In 1964, Silverman and Ehrlich (13) published a
cessed since 1996 (9). The solid commercial success of comprehensive analysis of microbial metal interactions,
these operations has established a momentum in the field and suggested for the first time that there must be both
of biohydrometallurgy, which is evidenced by the current an indirect mechanism of microbial leaching of sulfide
consideration of bioleaching/biooxidation for a wide range minerals (in which ferric ion plays a primary role, and
of base metals, including copper, zinc, nickel, and lead. where microorganisms accelerate the observed rate of
The benefits of the current commercial biooxidation plants mineral dissolution by regenerating ferric ion) and a direct
include significantly reduced capital investment in phys- mechanism, in which the microorganism attaches to the
ical plant, reduced operating costs when compared with mineral surface, and enzymatically attacks the sulfide.
smelting (10 to 15% estimated reduction), and stabilization These two mechanisms have till recently been universally
of arsenic-containing waste (in the case of arsenopyrite accepted as the means by which sulfide minerals are
ores). Knowledge gained from the long-term operating leached, and many kinetic and mechanistic studies have
experience with these plants has also raised interest- explained observed results in terms of these mechanisms.
ing scientific questions about the microbiology associated In the indirect mechanism, a metal sulfide, MS, is
with these engineered environments such as what the attacked by ferric ion:
actual microbial composition of the bacterial populations
is during operation, whether thermophilic bacteria are 2Fe3+ + MS → 2Fe2+ + M2+ + S0
more appropriate to use because of practical considera-
tions such as cooling costs and many others. Most exciting The resulting ferrous ions could be oxidized by
to researchers in the field is the serious commercial con- At. ferrooxidans (sulfur could also be oxidized by
sideration being given to bioleaching/biooxidation as an At. thiooxidans):
alternative to smelting in the recovery of metals besides
gold from ore concentrates. 2Fe2+ + 0.5O2 + 2H+ → 2Fe3+ + H2 O
(At. ferrooxidans, L. ferrooxidans, others)
THE MICROBIOLOGY OF SULFIDE MINERAL LEACHING S0 + 1.5O2 + H2 O → H2 SO4

Since the discovery of T. ferrooxidans (now At. ferrooxi- (At. ferroooxidans and At. thiooxidans)
dans) and its role in the generation of acid mine
drainage from pyrite in abandoned coal piles, exten- According to the direct mechanism, after attachment to
sive research has focused on the metabolism, physiology, the mineral surface, enzymatic oxidation of a metal sulfide
and genetics of this microorganism. The recent reclassi- occurs as follows:
fication of T. ferrooxidans, Thiobacillus thiooxidans, and
Thiobacillus caldus into the new genus Acidithiobacil- MS + 1.5O2 + H2 O → M2+ + 2H+ + SO4 2−
lus (4) recognized sulfur-utilizing species of the genus
Thiobacillus; these particular species inhabited a unique Bacterial attachment to the sulfide mineral occurs
environmental niche of low pH and high dissolved rapidly, and several recent studies have provided
metal concentrations. Much of our understanding of how ultramicroscopic examples of such interactions between
inorganic substrates such as Fe2+ and sulfur are oxi- At. ferrooxidans and synthetic pyrite (14,15) using scan-
dized by chemolithotrophs has come from the study of ning electron microscopy (SEM) or atomic force micro-
At. ferrooxidans and At thioxidans. It is important to note scopy, and sphalerite (16) using SEM. In the case of
that although At. ferrooxidans is able to oxidize both iron sphalerite (ZnS), the SEM micrographs also show quite
and sulfur, At. thiooxidans can only oxidize sulfur species. dramatically the difference in mineral surface appearance
644 BIOLEACHING OF METALS

when the bacteria are present, as opposed to when the min- mechanism of Silverman and Ehrlich, however, and ulti-
eral is leached chemically (substantial deposits of material mately, any postulated mechanisms must support the
‘‘cladding’’ the mineral surface that were shown to be pri- observed kinetics and energetics of microbial mineral
marily sulfur, when examined by electron dispersive x-ray leaching.
analysis, EDAX; 16). Attachment has been shown to be a A schematic depiction of sulfide leaching, including key
function of the extracellular polysaccharide (EPS) layer of features of the direct and indirect mechanisms is shown
the bacterial cell, and ferric ion has been demonstrated to in Figure 3.
be bound within the EPS (15). In the direct mechanism, Only since the 1990s has there been a widespread
this ferric ion would be reduced to ferrous ion through acknowledgment that the acidithiobacilli are merely
interaction with the sulfide at the mineral surface, and part of a more complex acidophilic microbial commu-
then recycled by respiratory enzymes at the cell surface nity associated with the oxidation and weathering of
(rusticyanin is believed to be a key enzyme in the respira- pyrites, in which they may in fact only be minor play-
tory chain of At. ferrooxidans responsible for the extraction ers. From studies of acidic mine drainage (25,26), heap
of energy from the oxidation of Fe2+ → Fe3+ ; 17,18). leaches (27–30), and commercial stirred tank bioreac-
Since 1996, a major debate has emerged concerning tors (31), it has become clear that the iron-oxidizing
the validity of the direct mechanism of sulfide mineral bacterium, L. ferrooxidans (32), is a dominant and impor-
bioleaching (16,19–21). Sand’s group in Hamburg and tant microbe that survives under conditions in which
other proponents have performed in-depth studies of sul- At. ferrooxidans is less active. In fact, there is certainly
fur intermediates formed during microbial attack of pyrite a succession of microbial populations that occur during
and other reduced sulfur minerals, and concluded that the leaching of sulfide minerals. Acidophilic, heterotrophic
the sulfur compounds observed during microbial attack bacteria of the genera Acidiphilium and Acidocella are
on sulfides substantiate only an indirect leaching mech- often found in close association with At. ferrooxidans,
anism that involves Fe3+ and the extracellular polysac- At. thiooxidans, and L. ferrooxidans. The affinity of these
charide (EPS) layer of the microorganism. The nature organisms for one another is remarkable, and it has been
of the sulfur compounds produced during this process noted several times in the literature that At. ferrooxidans
are a function of the type of sulfide and whether the cultures are often contaminated with these heterotrophic
sulfide is soluble in acid (19,20). For acid-insoluble sul- microorganisms (33–35). The first such contaminant, orig-
fides, including FeS2 , MoS2 , and WS2 , ferric ion extraction inally described as Thiobacillus acidophilus (33), has since
of electrons from the sulfide requires water to induce been reclassified on the basis of 16S rDNA sequence
corrosion and release sulfur intermediates upon dis- and a variety of biochemical markers as another mem-
ruption of the metal-sulfur bond. Thiosulfate has been ber of the genus Acidiphilium, A. acidophilus (36). It has
proposed as the initial sulfur compound released from long been thought that these heterotrophic species scav-
the sulfide, which is rapidly oxidized further to sulfate enge organic molecules that are metabolic by-products of
(thiosulfate mechanism). Acid-soluble sulfides, such as the chemolithotrophic microbes in acidic environments,
sphalerite (ZnS), galena (PbS), hauerite (MnS2 ), chalcopy- which have been shown to be detrimental to the growth
rite (CuFeS2 ), and realgar (As4 S4 ), are argued to be readily of the chemolithotrophs (especially L. ferrooxidans; 35).
susceptible to metal-sulfur bond dissociation by protons, More recently, it has been suggested that increased rates
with the resulting sulfur reacting with ferric ion to produce of iron oxidation observed when L. ferrooxidans and an
a reactive sulfhydryl radical that leads to the generation of Acidiphilium sp. are coinoculated on pyrite are due to the
polysulfides and ultimately, elemental sulfur (polysulfide heterotroph consuming EPS on the pyrite surface, pro-
mechanism). They propose that direct leaching does not viding more sites for the chemolithotroph to attach and
actually occur, but rather, the indirect action of ferric ion dissolve the sulfide (20,29). Although the acidophilic het-
facilitated by contact of the EPS of the microbe with the erotrophic bacteria are nutritionally quite versatile as a
mineral surface accounts for the sulfide dissolution his- group (37), it is still not clear what natural form(s) of
torically ascribed to the direct mechanism. Tributsch (22) organic carbon these microorganisms use.
has suggested that direct leaching should more accurately As the temperature of acidic environments increase,
be termed contact leaching. Other researchers employing At. ferrooxidans activity has been noted to decline
a suite of microscopic techniques to study the interac- (>45 ° C). In this case, L. ferrooxidans, At. thiooxidans,
tions of At. ferrooxidans and At. thiooxidans with colloidal and Acidithiobacillus caldus are the dominant leach-
sulfur and thin films of synthetic pyrite concluded that ing organisms. The impact of environmental condi-
the chemolithotrophs present at the surface of the sul- tions, including temperature, on the dominance of
fides accumulate sulfur colloids within the EPS layer, and L. ferrooxidans in commercial leaching operations has
release excess sulfur species into the surrounding media recently been discussed in depth (24). At. caldus (38)
that could be used by unattached bacteria (23). This has had been known for some years as a moderately ther-
been termed cooperative leaching. Tributsch (22) and col- mophilic sulfur-oxidizing species, which could comple-
leagues (24) provide a discussion of these various proposed ment L. ferrooxidans (iron oxidation capacity only) in
mechanisms, integrated with electrochemical mechanisms the bioleaching of chalcopyrite concentrate (39). A sim-
associated with chemical and microbial interactions with ilar synergistic effect has been seen between At. caldus
the sulfide, and environmental conditions to explain the and Sulfobacillus thermosulfidooxidans (40).
interactions leading to dissolution and the resulting prod- As the temperature increases beyond 50 ° C, moderately
ucts. There is still much support for the original direct thermophilic, gram-positive and gram-variable bacteria
 BIOLEACHING OF METALS 645

Figure 3. A schematic depiction of the mechanisms of microbial sulfide dissolution, including

direct and indirect mechanisms. Cells attached to the surface of pyrite (shaded surface) are
meant to depict not only the direct mechanism of sulfide leaching (consuming oxygen through
the metabolic activity of the attached cell), but indirect leaching by attached and planktonic cells
(such as At. ferrooxidans, L. ferrooxidans can also oxidize ferrous iron as a planktonic cell, but is
not shown for clarity). Attack is either by attached cells, or bulk ferric ion. The lightly shaded
cell walls of the various cells depict EPS embedded with polythionates and ferric ion. Note that
although thiosulfate is indicated as an intermediate species of pyrite oxidation, it is unstable at
pH less than 5, breaking down to sulfite and elemental sulfur.

are observed, including a range of gram-positive organ- organism (44) as a consequence of its own more rapid
isms comprising the Alicyclobacillus group of the high growth. There is currently substantial interest in the use
G+C% gram-positive bacteria. These organisms were first of moderately thermophilic bacteria in heap and stirred
isolated from hot pools in Iceland and thought to be tank leaching, as these organisms appear to be more
thermophilic relatives of Thiobacillus because they used efficient in the dissolution of complex sulfides such as
sulfur (41). The original TH1 isolate has since been deter- chalcopyrite (CuFeS2 ). Because cooling of stirred tank
mined to be a strain of Sb. thermosulfidooxidans (42). biooxidation reactors is one of the major costs in the oper-
Sulfobacillus acidophilus is another iron-oxidizing species ation of such plants, use of moderate thermophiles would
in this group. The iron-oxidizing sulfobacilli are capable also have an economic benefit, and this alternative is
of growing autotrophically on pyrite, but require carbon being pursued aggressively (see discussion later in this
dioxide supplementation to achieve reasonable growth article).
rates. They can also grow mixotrophically using both car- Beyond 70 ° C, chemolithotrophic and heterotrophic
bon dioxide and glucose, and heterotrophically (43). Other archaea dominate acidic mineral environments. Typi-
heterotrophic thermophiles appear to be phylogenetically cal of these microorganisms are Crenarchaeota, such as
closer to the species Alicyclobacillus acidoterrestris and Acidianus brierleyi (45; originally identified as Sulfolobus
Alicyclobacillus acidocaldarius, first isolated by Brock in acidocaldarius; 46). There was immediate interest in eval-
Yellowstone in the early 1970s. Acidimicrobium ferrooxi- uating the utility of these thermophilic microorganisms
dans is an unrelated eubacterial thermophile that grows in bioleaching (47). Since that time, other thermophilic
readily on iron or pyrite without carbon dioxide supple- species have been identified and studied in the context of
mentation, and which has been shown to promote the mineral leaching, including Sulfolobus metallicus (48,49)
growth of Sb. acidophilus when both are grown together, and Metallosphaera sedula (50). The application of these
presumably by providing organic nutrients to the latter organisms, particularly in stirred tank bioreactors, has
646 BIOLEACHING OF METALS

been hampered by their inherent physiology. Because they [Fe3+ ] (At. ferrooxidans is inhibited at ferric concentra-
lack a conventional cell wall, they are especially suscep- tions more than 10 times lower than that which inhibits
tible to physical damage from shearing by the mineral L. ferrooxidans), and low [Fe2+ ] (L. ferrooxidans demon-
particles as the solutions are stirred, effectively limiting strates a Km for ferrous iron six times lower than
the amount of solids that can be introduced. At. ferrooxidans).
Figure 4 depicts the spectrum of acidophilic bacteria Recently, Banfield and coworkers (26,51) studying the
when considering changes in pH and temperature. There massive Iron Mountain site in California, have discovered
is considerable overlap in populations as environmental that L. ferrooxidans is distributed widely in waters
conditions change, but in both natural and artificial and sediments at the site, but were unable to find
environments, the shifts in active iron- and sulfur- significant numbers of At. ferrooxidans. They concluded
oxidizing species have been well documented. Organic that At. ferrooxidans did not play a significant role
carbon is indicated in Figure 4 as playing a role in in pyrite oxidation in this environment. It would be
the interactions of leaching microorganisms, whereby interesting to consider the microbial ecology of this
autotrophic species may provide substrates for the growth extremely acid environment in light of the discussion of
of strictly heterotrophic species. As mentioned earlier, Rawlings and coworkers (24). Further, they have isolated
iron oxidation by more rapidly growing species, as in a hitherto unknown species of Archaea, Ferroplasma
the case of At. ferrooxidans, may also stimulate slower acidarmanus (52), which appears to be a dominant
growing thermophilic iron-oxidizing species (such as microorganism at the extremely low pH (<0.5) found in the
Sb. acidophilus) when organic carbon is released into the Iron Mountain environment. Ferroplasma acidarmanus
environment. As was noted before, it is still not clear is not an obligate chemolithotroph, however, because
what form(s) of organic carbon may be involved in these comparable growth rates were observed when the
microbial interactions. organism was grown on 0.02% yeast extract without
Additional dimensions can be included in the con- added ferrous sulfate. Interestingly, although the total
sideration of how environmental conditions affect the iron concentration measured at Iron Mountain is more
microbial population of iron and sulfur-oxidizing species. than 20 g/L (some locations within the study site have
For example, L. ferrooxidans has been noted to be measured iron concentrations exceeding 100 g/L), more
more tolerant of high [Fe3+ ] and lower pH (<1.5) than than 90% is ferrous iron, which is quite different from
At. ferrooxidans, which explains the larger numbers of many ARD sites, where most of the iron exists as dissolved
L. ferrooxidans found in the commercial BIOX process, ferric ion.
heap leaches, and in acidic mine drainage (24,26,27–30). The preceding discussion has focused on the dissolution
In their comprehensive analysis of the factors that of sulfide minerals by chemolithotrophic acidophilic bac-
might influence the selection of L. ferrooxidans over teria and archaea, and associated heterotrophic species.
At. ferrooxidans in commercial processes, Rawlings and However, the field of biohydrometallurgy has not ignored
coworkers (24) took into account the effects of redox oxide, silicate, and carbonate ores where heterotrophic
potential (L. ferrooxidans favored at > +700 mV), pH microorganisms are likely to be most important to
(L. ferrooxidans is more tolerant of acidic pH), tem- any eventual commercial application. Ehrlich (53,54) has
perature (L. ferrooxidans is active beyond 35 ° C), high recently provided excellent overviews of our state of

Acidophilic heterotrophs
Acidiphilium spp., Acidocella spp.
3 Ferromicroblum acidophilus

pH
Organic carbon
2 Moderate thermophiles Extreme thermophiles
Mesophilic autotrophs
Sulfobacillus spp.
Alicyclobacillus spp. Acidianus brierleyi
Acidithiobacillus ferrooxidans Acidimicrobium spp. Metallosphaera sedula
At. thiooxidans Acidithiobacillus caldus Sulfolobus metallicus
1
Leptospirillum ferrooxidans

Ferroplasma spp.

35 60 75+
Temperature, °C

Figure 4. The microbial ecology of acidic environments as a function of temperature and pH.
 BIOLEACHING OF METALS 647

understanding with respect to microbial attack of these are made to provide essential nutrients to the microbial
minerals. In many cases, fungi producing organic acids, consortium. The temperature of the process is controlled
such as oxalate and citrate, have been demonstrated to between 30 to 45 ° C, with a desired average ceiling
leach aluminum ores (55) and lateritic nickel ores (56). temperature of 40 ° C to provide some cushion in the case
Anaerobic bacteria may also be useful in the release of of loss of cooling. Operating pH is maintained between
metals from metal oxides (57,58). Heterotrophic leaching 1.2 to 2.0. Aeration (maintained near saturation; 2.2 kg
is likely to be a more complex process because the ores O2 /kg sulfide consumed) and agitation to maintain the ore
cannot be sterilized, and environmental conditions are not particles in suspension account for 30 to 40% of the overall
highly selective as is the case in the leaching of sulfides power consumption of the process. After biooxidation, the
by obligately acidophilic microorganisms. Inexpensive car- treated ore concentrate is washed extensively to remove
bon sources will also be required to facilitate economically iron and other cyanide consuming materials, and the
feasible microbial oxide leaching (53). dried ore is then removed for cyanidation to recover
gold. Liquid effluents from the process, containing iron,
arsenic, and other undesired metals, are treated with lime
COMMERCIAL PRACTICE to neutralize, precipitate, and stabilize the waste. The solid
precipitates are deposited in a tailings impoundment. The
BIOX overall process is presented in a simplified process diagram
Dew and coworkers (9) recently provided a comprehensive in Figure 5.
picture of the development of Billiton Process Research’s
BIOX process. From a pilot study handling 10 metric tons BacTech
per day (tpd) to the current full-scale operation in Ghana A similar process description has been made for the mod-
handling 960 tpd, considerable process knowledge has erately thermophilic BacTech process (59). The BacTech
been obtained to characterize the operating parameters process is distinguished from the BIOX process in oper-
of these plants that have become an economically viable ating between 45 and 55 ° C, where a different microbial
alternative to pressure oxidation and smelting. consortium, dominated by sulfobacilli, At. ferrooxidans,
The typical plant design is composed of a series and At. caldus have been described. The overall layout of
of stirred tanks, with up to three being operated such a plant is very similar to that of the BIOX pro-
as primary reactors in parallel, and the downstream cess, although special considerations must be made to the
reactors functioning in series. Total residence time of the materials from which the reactors are constructed, owing
arsenopyrite flotation concentrate in the reactor train is to the elevated temperature and low pH of the operat-
four days. The ore concentrate is ground to 75 µm particle ing environment. Special grades of stainless steel have
size (>80%, 100% <150µm) and the solids loading is 20 to been described to be more appropriate for construction of
30%. Phosphorus, potassium, and nitrogen amendments these systems, as compared with the 304-L stainless steel

Flotation concentrate in

Primary reactor train

(2 day residence time typical)

Secondary reactor train

(2 day residence time typical)

Solids Liquid effluent

Wash/Fe removal Lime/precipitation

Acidic water

Cyanidation Tailings impoundment

Figure 5. Schematic representation of commercial biooxidation plants in operation today.

648 BIOLEACHING OF METALS

commonly used for mesophilic processes. Corrosion result- nonexistent, and internal heap temperatures are known
ing from elevated chloride ion appears to be a particular to exceed 65 ° C.
concern. As in the BIOX process, aeration, mixing, and
temperature control are major contributors to the overall
THE FUTURE
power requirements of these plants. It is estimated that 3
to 12 MW of heat is produced in a typical 100-tpd plant.
Interestingly, oxygen requirement does not appear to be Since the first association of the acidophilic chemolithotro-
higher in the BacTech plant, although the solubility of oxy- phic bacterium, At. ferrooxidans, with sulfide mineral
gen is lower in water at the higher operating temperatures leaching some 50 years ago, there has been a tremendous
growth in our understanding of the microbes associated
of this process.
with natural and engineered bioleaching. Commercial
practice of bioleaching has grown from a peripheral
Heap Leaching
operation treating marginal ores and waste rock to highly
Heap leaching is without question the largest scale optimized heap and stirred tank processes that handle
application of microbial mineral leaching now practiced thousands of tons of valuable ores and concentrates
(dump leaching, in which no attempt is made to optimize each day. The efforts of many scientists, engineers, and
particle size is included in this statement). It has been mine operators have resulted in the largest bioprocesses
noted that although the principles used in the commercial currently practiced. There is a great promise that
operation of these plants were developed in North America, bioleaching and biooxidation of minerals will increase
virtually all commercial plants have been constructed in the efficiency with which the Earth’s metal resources are
the Southern Hemisphere (60). Copper sulfide bioleach utilized, reduce the impacts associated with the extraction
plants range from just over 1,000 tpd to over 17,300 tpd of metals from often deeply buried ores, and replace
at Quebrada Blanca in Chile. At Lo Aguirre, Chile, more energy intensive, polluting technologies. However,
16,000 tpd were processed continuously from 1980 to there is still much to be learned. Brierley (62) has
1996. The use of heap leaching, in which finely ground noted that many of the environmental factors associated
ore is agglomerated and placed on highly engineered with bioleach operations that may influence microbial
heaps, is now being considered for the biooxidation of populations, including acid conditioning of ores, sulfide
refractory gold ores. Brierley (61) has provided extensive concentration, internal heat generation, dissolved solutes
details of such a plant that has been designed and tested (Cl− , NO3 − , heavy metals, total dissolved solids), aeration,
by Newmont Gold Company. This state-of-the-art gold nutrients, and organic carbon, have rarely been considered
heap leach operation, with an idea of the scale of such from the context of the microorganisms operating in
operations, is depicted in Figure 6. The sheer scale of just the commercial environment. Some modern heap leach
a single such operation, in which millions of tons of ore operations now incorporate design features, such as
are treated, substantiate claims that bioleaching of metals aeration systems, to accommodate microbial activity, but
represents the largest bioprocessing application extant. In there is still much to be learned with respect to optimizing
optimizing these heap leach operations in the future, more these artificial environments for optimum activity of the
understanding of thermophilic bacteria and archaea and biological catalysts present. Even our understanding of
their ability to oxidize sulfides will be necessary because, the mechanisms involved, let alone the genetic regulation
unlike stirred tank reactors for the biooxidation of valuable of those mechanisms, is incomplete. The recent partial
gold concentrates, temperature control will be limited or sequencing of the genome of At. ferrooxidans (63) raises
the hope that our future understanding and control of
these important biogeochemical and commercial processes
will be extensive enough to permit the utilization of
genomics and bioinformatics in the control of commercial
operations.

TERMINOLOGY AND DEFINITIONS

Bioleaching
The liberation of metal values into solution from minerals
and ores through microbiological action.

Biooxidation
The liberation of metal values (typically as microscopic
particles) from ores. In this process, microbial oxidation of
Figure 6. One of the three biooxidation pretreatment heaps
sulfides in the ore leads to deterioration of the rock matrix,
measuring approximately 1,000 by 500 feet, with ore stacked and release of the bound metal (e.g., gold), which can
to 30 feet, containing 709,000 tons of low-grade refractory ore. then be recovered with increased efficiency through other
Courtesy of Dr. J. Brierley, Newmont Mining Corporation. See physical and chemical treatments such as cyanidation
color insert. and flotation. Often only partial oxidation of the sulfides
 BIOLEACHING OF METALS 649

present is required to achieve efficient liberation of the 10. D. E. Rawlings, in D. E. Rawlings, ed., Biomining: Theory,
metal values. Microbes and Industrial Processes, Springer-Verlag, Berlin,
Germany, 1997, pp. 229–245.
11. D. E. Rawlings and T. Kusano, Microbiol. Rev. 58, 39–55
Dump Leaching
(1994).
The leaching of run-of-mine ore placed in large piles. 12. J. C. Cox et al., Biochem. J. 178, 195–200 (1979).
Because of environmental regulations, this frequently is 13. M. P. Silverman and H. L. Ehrlich, Adv. Appl. Microbiol. 6,
done on impervious pads today to prevent escape of the 153–206 (1964).
leachate into the environment, but always involves run- 14. M. Rodriguez-Leiva and H. Tributsch, Arch. Microbiol. 149,
of-mine ore. 401–405 (1988).
15. T. Gehrke, J. Telegdi, D. Thierry, and W. Sand, Appl. Envi-
ron. Microbiol. 64, 2743–2747 (1998).
Heap Leaching
16. T. A. Fowler and F. K. Crundwell, Appl. Environ. Microbiol.
The leaching of ores on prepared pads (geotextile 65, 5285–5292 (1999).
membrane, concrete pad, sometimes including provisions 17. C. F. Kulpa, M. T. Roskey, and N. Mjoli, Biotechnol. Appl.
for aeration, fluid collection, etc.) where the ore has been Biochem. 8, 330–341 (1986).
crushed to a particular size range, and may have been 18. E. Jedlicki et al., Biotechnol. Appl. Biochem. 8, 342–350
coated or mixed with microorganisms and/or nutrients to (1986).
promote more rapid leaching. 19. A. Schippers and W. Sand, Appl. Environ. Microbiol. 65,
319–321 (1999).
20. W. Sand, T. Gehrke, P.-G. Jozsa and A. Schippers, in R. Amils
Run-of-Mine
and A. Ballester, eds., Biohydrometallurgy and the Environ-
Ore as it is extracted from the mine, which can contain rock ment Toward the Mining of the 21st Century, Part A, Elsevier
BV, Amsterdam, The Netherlands, 1999, pp. 27–49.
ranging from fine particles to hunks approaching 1 ton in
weight, and perhaps a meter or more in linear dimension. 21. T. A. Fowler, P. R. Holmes, and F. K. Crundwell, Appl. Envi-
No attempt is made to fractionate the ore further before ron. Microbiol. 65, 2981–2997 (1999).
other operations (such as dump leaching). 22. H. Tributsch, in R. Amils and A. Ballester, eds., Biohydromet-
allurgy and the Environment Toward the Mining of the 21st
century, Part A, Elsevier BV, Amsterdam, The Netherlands,
Acknowledgments 1999, pp. 51–60.
I am indebted to Dr. James Brierley for the use of the photo-
23. J. A. Rojas-Chapana, C. C. Bartels, L. Pohlmann, and H. Tri-
graph appearing in Figure 6, and to Dr. Henry Ehrlich for his
butsch, Proc. Biochem. 33, 239–248 (1998).
willingness to read a draft of this manuscript and provide sug-
gestions that have strengthened this contribution. The reader 24. D. S. Rawlings, H. Tributsch, and G. S. Hansford, Microbiol-
is directed to Dr. Ehrlich’s authoritative book, Geomicrobiology, ogy 145, 5–13 (1999).
(4th edition, Marcel Dekker, in preparation) for a comprehensive 25. M. O. Schrenk et al., Science 279, 1519–1522 (1998).
treatment of this and other topics relating to microbe-mineral 26. P. L. Bond et al., Appl. Environ. Microbiol. 66, 4962–4971
interactions. (2000).
27. J. Pizarro et al., Appl. Environ. Microbiol. 62, 1323–1328
(1996).
BIBLIOGRAPHY
28. W. Sand, K. Rohde, B. Sobotke, and C. Zenneck, Appl.
Environ. Microbiol. 58, 85–92 (1992).
1. A. I. Lopez-Archilla, I. Marin, and R. Amils, in C. A. Jerez,
29. R. Hallmann et al., Geomicrobiol. J. 10, 193–206 (1992).
T. Vargas, H. Toledo, and J. V. Wiertz, eds., Biohydrometal-
lurgical Processing, vol. II, University of Chile Press, Santi- 30. B. M. Goebel and E. Stackebrandt, Appl. Environ. Microbiol.
ago, Chile, 1995, pp. 63–73. 60, 1614–1621 (1994).
2. G. Rossi in G. Rossi, ed., Biohydrometallurgy, McGraw-Hill, 31. D. E. Rawlings, in C. A. Jerez, T. Vargas, H. Toledo, and
Hamburg, Germany, 1990, pp. 1–13. J. V. Wiertz, eds., Biohydrometallurgical Processing, vol.II,
3. S. M. Hong et al., Science 272, 246–249 (1996). University of Chile Press, Santiago, Chile, 1995, pp. 9–17.
4. D. P. Kelly and A. P. Wood, Int. J. Syst. Evol. Microbiol. 50, 32. H. Hippe, Int. J. Syst. Evol. Microbiol. 50, 501–503 (2000).
511–516 (2000). 33. R. Guay and M. Silver, Can. J. Microbiol. 21, 281–288
5. A. R.Colmer and M. E. Hinkle, Science 106, 253–256 (1947). (1974).
6. A. R. Colmer, K. L. Temple, and M. E. Hinkle, J. Bacteriol. 34. A. P. Harrison, Jr. Int. J. Syst. Bacteriol. 31, 327–332
59, 317–328 (1950). (1981).
7. U.S. Patent No. 2,829,964 1958, S. R. Zimmerley, D. G. Wil- 35. D. B. Johnson and W. I. Kelso, J. Gen. Microbiol. 123,
son, and J. D. Prater. 2969–2972 (1983).
8. C. L. Brierley, in D. E. Rawlings, ed., Biomining: Theory, 36. A. Hiraishi et al., Int. J. Syst. Bacteriol. 48, 1389–1398
Microbes and Industrial Processes, Springer-Verlag, Berlin, (1998).
Germany, 1997, pp. 3–17. 37. D. B. Johnson and F. F. Roberto, in D. E. Rawlings, ed.,
9. D. W. Dew, E. N. Lawson, and J. L. Broadhurst, in D. E. Biomining: Theory, Microbes and Industrial Processes,
Rawlings, ed., Biomining: Theory, Microbes and Indus- Springer-Verlag, Berlin, Germany, 1997, pp. 259–279.
trial Processes, Springer-Verlag, Berlin, Germany, 1997, 38. K. B. Hallberg and E. B. Lindstrom, Microbiology 140,
pp. 45–80. 3451–3456 (1994).
650 BIOLOGICAL CONTROL, USE OF BIOSURFACTANTS IN

39. P. R.Norris, in H. L. Ehrlich, and C. L. Brierley, eds., Micro- BIOLOGICAL CONTROL, USE OF
bial Mineral Recovery, McGraw-Hill, New York, 1990, BIOSURFACTANTS IN. See BIOSURFACTANTS: TYPES,
pp. 3–27. SCREENING METHODS, AND APPLICATIONS
40. M. Dopson and E. B. Lindstrom, Appl. Environ. Microbiol. 65,
36–40 (1999).
41. N. Le Roux, D. S. Wakerley, and S. D. Hunt, J. Gen. Micro-
biol. 100, 197–201 (1977).
42. R. S. Golovacheva and G. I. Karavaiko, Mikrobiologiya 48, BIOLOGICAL WARFARE. See BIOTERRORISM
658–665 (1979).
43. P. R. Norris, D. A. Clark, J. P. Owen, and S. Waterhouse,
Microbiology 142, 775–783 (1996).
44. D. A. Clark and P. R. Norris, Microbiology 142, 785–790
(1996). BIOLOGY OF CRYPTOSPORIDIUM
45. A. Segerer et al., Int. J. Syst. Bacteriol. 36, 559–564
(1986). DEBRA E. HUFFMAN
46. T. D. Brock, K. M. Brock, R. T. Belly, and R. L. Weiss, Arch. University of South Florida
Microbiol. 84, 54–68 (1972). St. Petersburg, Florida
47. C. L. Brierley and L. E. Murr, Science 179, 488–490
(1973).
TAXONOMIC CLASSIFICATION
48. P. R. Norris and L. Parrott, in R. W. Lawrence, R. M. R. Bra-
nion, and H. G. Ebner, eds., Fundamental and Applied
Cryptosporidium is an obligate intracellular parasite
Biohydrometallurgy, Elsevier, Amsterdam, The Netherlands,
belonging to the phylum Apicomplexa, order Eucoccidiida,
1986, pp. 355–365.
suborder Eimeriina, and family Cryptosporidiida. Cryp-
49. G. Huber and K. O. Stetter, Syst. Appl. Microbiol. 14,
tosporidium was first diagnosed as an agent of disease in
372–378 (1991).
humans in 1976 (1,2). Since that time, it has been well
50. G. Huber et al., Syst. Appl. Microbiol. 12, 38–47 (1989).
recognized as a cause of diarrheal illness throughout the
51. K. J. Edwards, T. M. Gihring, and J. F. Banfield, Appl. Envi- world (3). Currently, there are approximately 23 named
ron. Microbiol. 65, 3627–3632 (1999). species of Cryptosporidium infecting a wide range of verte-
52. K. J. Edwards, P. L. Bond, T. M. Gihring, and J. F. Banfield, brates, including humans. However, many of these species
Science 287, 1796–1799 (2000). are considered to be misidentified. There are 10 species
53. H. L. Ehrlich, in R. Amils and A. Ballester, eds., Biohy- commonly recognized today (Table 1; 4–6). These species
drometallurgy and the Environment Toward the Mining of the include five mammalian (C. parvum, C. muris, C. wrairi,
21st Century, Part A, Elsevier BV, Amsterdam, The Nether- C. andersoni, C. felis), two avian (C. meleagridis and
lands, 1999, pp. 3–12.
C. baileyi), two reptilian (C. serpentis, C. saurophilum),
54. H. L. Ehrlich, in D. E. Rawlings, ed., Biomining: Theory, and one fish (C. nasorum). Speciation of isolates, in accor-
Microbes and Industrial Processes, Springer-Verlag, Berlin, dance with the practice for other coccidians, has primarily
Germany, 1997, pp. 129–150.
been based on host specificity, oocyst morphology, and site
55. L. V. Ogurtsova et al., Mikrobiologiya 58, 774–780 (1989). of infection. However, these traits have been shown to
56. K. Boseker, in R. W. Lawrence, R. M. R. Branion, and H. G. be insufficient to accurately identify separate species of
Ebner, eds., Fundamental and Applied Biohydrometallurgy, Cryptosporidium. Much of the information, regarding host
Elsevier, Amsterdam, The Netherlands, 1986, pp. 367–382. specificity has been based on a limited number of studies
57. D. R. Lovley, Microbiol. Rev. 55, 259–287 (1991).
58. H. L. Ehrlich, in R. W. Smith and M. Misra, eds., Mineral
Bioprocessing, The Minerals, Metals and Materials Society, Table 1. Named Species of Cryptosporidium and Their
Warrendale, Pa., 1991, pp. 27–41. Hosts
59. P. C. Miller, in D. E. Rawlings, ed., Biomining: Theory,
Species Hosts
Microbes and Industrial Processes, Springer-Verlag, Berlin,
Germany, 1997, pp. 81–102.
Cryptosporidium baileyi Chicken
60. J. A. Brierley and C. L. Brierley, in R. Amils and A. Ballester, Cryptosporidium meleagridis Turkeys, immunocompromised
eds., Biohydrometallurgy and the Environment Toward the human
Mining of the 21st century, Part A, Elsevier BV, Amsterdam, Cryptosporidium muris Mice
The Netherlands, 1999, pp. 81–89. Cryptosporidium felis Cats, immunocompromised
61. J. A.Brierley, in D. E. Rawlings, ed., Biomining: Theory, human
Microbes and Industrial Processes, Springer-Verlag, Berlin, Cryptosporidium serpentis Reptiles
Germany, 1997, pp. 103–115. Cryptosporidium saurophilum Skunk
62. C. L. Brierley, in R. Amils and A. Ballester, eds., Biohy- Cryptosporidium nasorum Fish
drometallurgy and the Environment Toward the Mining of the Cryptosporidium wrairi Guinea pig, immunocom-
promised human
21st Century, Part A, Elsevier BV, Amsterdam, The Nether-
Cryptosporidium andersoni Cattle
lands, 1999, 91–97.
Cryptosporidium parvum∗ Cow, mouse, human
63. E. Selkov et al., Proc. Natl. Acad. Sci. U.S.A. 97, 3509–3514
∗
(2000). Appears to be infectious for 152 species of mammals (6).
 BIOLOGY OF CRYPTOSPORIDIUM 651

using a small number of Cryptosporidium isolates that

render this information inconclusive (7,8). Finally, Cryp- Intestinal
tosporidium has been shown to infect numerous sites in epithelial cells
the body including the upper and lower respiratory tracts,
the uterus, bladder, and also the liver and pancreatic sys- Sporozoites
tems, rendering the site of infection a relatively useless Trophozoite
tool to describe a species (9).
Schiz
Two different genotypes of Cryptosporidium parvum ogon
Reinfection y
have been shown to be infectious in humans (10).
Genotype-1 isolates have been shown to be infectious only
in humans, while genotype-2 isolates have been shown Ingestion
to be infectious in mice, calves, lambs, goats, and horses, Schizont with
as well as humans. This suggests the possibility that 8 merozoites
there are two distinct populations of oocysts cycling in yte
ametoc
humans with distinct transmission cycles; (1) zoonotic Macrog
Microgametes Microgametocyte
transmission from animal to human with subsequent Resistant Autoinfection
oocyst
human to human and human to animal transmission and
Faeces
(2) a transmission cycle exclusively in humans.
Oocyst
Zygote

BASIC BIOLOGY

The environmentally stable stage of this organism is

known as an oocyst. The oocysts of Cryptosporidium
parvum, the species infecting humans, are 4 to 6 µm in
size and have been shown to survive in the environment Figure 1. Reproductive life cycle of Cryptosporidium parvum.
for extended periods of time, depending on temperature
and water quality conditions because of the robust nature
Two morphologically different types of schizonts have
of the oocyst wall (11). The oocyst wall of C. parvum and
been observed. Type-I schizonts contain six to eight
related species appears to be composed of three layers
nuclei. When the schizont matures, each nucleus becomes
when viewed by electron microscopy (7,12–14). The inner
incorporated into a merozoite. Each merozoite can either
layer is relatively thick and is thought to be a filamentous
develop into another Type-I schizont or a Type-II schizont,
glycoprotein (11,15). The middle layer is thought to be a
which will produce 4 merozoites. Merozoites formed from
more rigid layer composed of a complex lipid, whereas
Type-II schizonts initiate sexual reproduction known
the outer layer is thought to be composed of acidic
as gametogony. During gametogony, the merozoite can
glycoproteins (11). Spanning one-third to half of the oocyst
invade other host cells where they may differentiate
wall is a linear suture, which under the proper conditions
into either microgametes or macrogamonts (male and
opens to release the four internal sporozoites that are
female stages). When the microgamont develops, it
capable of initiating infection within the host.
becomes multinucleate. Each nucleus, eventually, becomes
incorporated into the equivalent of a sperm, which
REPRODUCTION will fertilize the macrogamont. After fertilization, the
macrogamont develops into an oocyst. After sexual
The reproductive life cycle of Cryptosporidium is complex, reproduction is complete, the oocysts will be released in
with both sexual and asexual reproduction cycles being excreted feces. Approximately 80% of the oocysts excreted
completed within a single host (16, Fig. 1). Infection begins in the feces are thick-walled oocysts. These oocysts have
with the ingestion of an oocyst. Once ingested, excystation multilayered walls that allow them to remain infective
(release) of the four sporozoites from the suture line outside of the host. The remaining 20% of the oocysts
located along the side of an oocyst begins. Sporozoites produced are thin-walled oocysts that can rupture within
once released from an oocyst may penetrate individual the host, releasing their sporozoites and initiating an auto-
epithelial cells located in the region of the ileum and begin infectious life cycle. Infected hosts may excrete between
reproduction. 109 and 1010 oocysts per day during the course of the
The initial phase of reproduction begins when the disease (17).
parasite attaches itself to the host cell by means of a
feeder or attachment organelle that appears to be used CLINICAL ASPECTS
to facilitate the exchange of materials. The sporozoites
and all subsequent lifecycle stages are intracellular The incubation (or prepatent period) time between ingest-
beneath the host cell membrane, but are extracytoplasmic. ing oocysts and shedding of oocysts varies among species.
Once the sporozoite enters the host cell, it begins to In humans, the average incubation period ranges from
differentiate into a spherical trophozoite with a prominent 5 to 10 days with a mean of 7.2 days (18–20). Oocyst
nucleus. The division of this nucleus signals the start of shedding can last anywhere from 3 to 15 days and may
asexual reproduction known as merogony or schizogony. continue well past the cessation of symptoms (18,19,21).
652 BIOLOGY OF CRYPTOSPORIDIUM

The severity and duration of the disease will vary assessment of the prevalence or incidence of the disease,
greatly, depending on the status of the host’s immune and they should be viewed with caution. Further analysis
system (22). In immunocompetent individuals, symptoms of data compiled from 40 countries between 1983 and 1990,
including fever, malaise, nausea, and diarrhea typically excluding outbreaks, suggests that the prevalence rates in
lasts 10–14 days. Immunocompromised patients, such as industrialized countries of North America and Europe are
patients with AIDS, patients undergoing immunosuppres- between 1 and 3% and the rates in developing countries
sive therapy, those with immunoglobulin deficiencies, or are approximately 5% in Asia and 10% in Africa (34).
with concurrent viral infections affecting immunity may It has been estimated that Cryptosporidium accounts
suffer a prolonged, life-threatening illness either because for 13 to 16% of diarrheal cases in AIDS patients in
of dehydration or secondary complications or both. developed nations, with that number climbing to 24 to
Chemotherapy is currently not available for treatment 50% in developing nations (35,36,26,24,4). A prospective
of this disease, although more that 100 therapeutic agents long-term study in Europe suggests that 3 to 4% of person
have been studied. Supportive therapy, including the use with HIV will have cryptosporidiosis when diagnosed with
of intravenous fluids is considered to be the treatment of HIV, and that an equal number of them will develop it,
choice. Treatment of HIV-infected patients may include during the course of their disease (37). A study in Los
the use of antiretroviral therapy, including the use of Angeles in 1994 suggests that 3.4% of HIV patients in the
protease inhibitors to improve the immune status of the USA will become infected with cryptosporidiosis during
patient. This treatment regime has been shown to result their symptomatic period (38).
in both clinical and parasitological improvement in the Several studies have noted a temporal or seasonal
infected host (23). peak in levels of cryptosporidiosis. However, these tend
Studies to determine the number of oocysts necessary to to vary from country to country, including summer in
cause infection in humans have produced differing results. Australia, rainy season in Central America and India,
The first human-feeding study in healthy volunteers spring or late summer in North America, and late
suggested that the infectious dose ID50 or the dose that summer in Germany (26). These trends may reflect rainfall
is required to establish infection in 50% of the persons patterns, farming events such as calving and lambing, or
exposed was approximately 132 oocysts, although one fertilization practices. However, outbreaks still remain
recipient was infected with as few as 30 oocysts (24). unpredictable.
A similar study using three geographically diverse
isolates of genotype 2 showed a variance in the ID50 of
OCCURRENCE IN THE ENVIRONMENT
9–1042 oocysts (25). The variability between infectious
doses is most likely because of the susceptibility of the host
Cryptosporidium oocysts have been detected in 4–100%
and the virulence of the particular isolate. Interestingly,
of surface waters sampled at levels of 0.0012–5800/L,
the isolate with the lowest ID50 also had the highest illness
depending on the impact from sewage and animals (39–41,
attack rate (86% versus 50–55%) (25).
Table 2). A survey performed by LeChevalier in 1991
found 87.1% of the surface water sites sampled positive for
PREVALENCE OF CRYPTOSPORIDIOSIS IN HUMANS Cryptosporidium with maximum concentrations ranging
from 10 to 484 oocysts/L, with the highest levels being
Because cryptosporidiosis infections are generally self- found in the Mississippi, Ohio, and Missouri Rivers (43).
limiting and often symptomatically similar to other A similar study by Rose in 1991 noted a tenfold difference
diarrheal diseases, the disease may often be undiagnosed in oocyst densities between urbanized and pristine
or misdiagnosed in the absence of a recognized outbreak. watersheds (43). In pristine watersheds, concentrations
Incidence of Cryptosporidium infections in the population ranging from 0.003 to 0.29/L have been detected (39).
range from 0.6 to 20%, depending on the geographical Groundwater, previously thought to be a more pro-
locale with a peak incidence of the disease in children one tected water source, has shown between 9.5 and 22% of
to five years of age with no sex related disposition (26). samples positive for Cryptosporidium (44). A similar sur-
Approximately 60% of all positive stool samples come from vey of groundwater in the United Kingdom (258 samples)
children with 30% coming from adults less than 45 years revealed a slightly lower percentage of positive sam-
of age (26). Although clinical infection is uncommon over ples (5.8%) (40). A U.S. survey of wells that were shown
the age of 40, it does occur from time to time, and there is to be coliform-positive was also shown to be contami-
currently no evidence of increased incidence in the elderly. nated with Cryptosporidium oocysts at concentrations of
Most early reports of human cryptosporidiosis were 0.004–0.922/L (39). Outbreaks of cryptosporidiosis caused
single sporadic cases or small clusters of cases. In by the contamination of a drinking water well have been
1983, population surveys began to be reported from documented on four occasions in various geographical
Australia, Finland, the United Kingdom, and numerous areas (Texas, Pennsylvania, Yakima, Washington, and
other countries (23,27–33). Unfortunately, many of these Walla Walla, Washington). Two of the outbreaks (Penn-
surveys were not well controlled with data being derived sylvania and Yakima, Washington) were associated with
from selected populations, using specimens that had shallow wells where the ground water was suspected of
been routinely submitted to the laboratory. Furthermore, being under the influence of surface water. The outbreaks
accurate identification of the parasite could not be in Texas and Walla Walla, Washington, were due to
assumed in all studies. For these reasons, the information the direct contamination from wastewater and irrigation
produced by these surveys may not provide an accurate water, respectively (45).
 BIOLOGY OF CRYPTOSPORIDIUM 653

Table 2. Occurrence of Cryptosporidium in the Aquatic

Environment. Adapted from H.V. Smith and J.B. Rose,
Parasitol. Today 14, 14–22 (1998).

Number of % Positive Sporozoite

Country Samples Samples Oocysts/L

In surface waters Empty

USA 181 55 0.0025–44 oocyst
Oocyst
USA 85 87 0.007–484
USA 35 97.1 0.18–63.5
USA 11 100 2–112 400 × magnification DIC
USA 101 24 0.005–252.7 Cryptosporidium muris
UK 262 40.5 0.006–2.3
UK 403 15 0.0012–0.12 Sporozoite
UK 375 4.5 0.07–2.75
UK 691 52.2 0.04–3
UK 375 4.4 0.07–2.75
UK 196 39.9 0.006–15.6
Germany 9 78 not stated
Spain 8 50 <0.01–0.31 Empty
Netherlands 52 11.6 0.002–0.018 oocyst
Australia 114 46.5 0.1–14.3
Israel 16 68.8 0.006–0.52
Malaysia 76 10.5 2–245.6 Oocyst

In drinking water
USA 36 17 0.005–0.017 400 × magnification DIC
USA 82 26.8 not stated Cryptosporidium parvum
UK 15 7 0.006 Figure 2
UK 209 37 0.007–1.36
Spain 9 33 <0.01–0.02
Brazil 18 22.2 not stated
In wastewater effluents
USA 11 100 4–3,960
USA 130 not stated up to 0.05
A similar level of inactivation was shown at extremely
USA 60 67 <0.6–120
low temperatures. When oocysts were held at −15 ° C,
UK 50 74 1–321
UK 70 37 0.03–2.3 they gradually became noninfectious and after a period of
UK 117 65 5–60 7 days no infectious oocysts could be detected. At −20 ° C
UK 94 25.5 10–60 and −70 ° C a >99.99% inactivation could be achieved after
8 h and 1 h, respectively (48).

TRANSMISSION
Cryptosporidium oocysts by nature are very robust
and have been shown to survive for extended periods
Several factors related to the genus Cryptosporidium
of time in the environment. They have also been shown
have contributed to their high incidence of disease in
to survive in both human and cattle fecal material, which
the population. These factors include the organism’s
can protect them from desiccation. Therefore, the practice
ability to complete its lifecycle in a single host, the
of disposal of both human and animal waste, by land
robust, environmentally stable infective oocyst stage
application may lead to indirect contamination of water
and the large number of infectious oocysts excreted
supplies or food crops. Large amounts of contaminated
from infected hosts. These factors along with the broad
fecal material can also provide a source of viable oocysts
host range for this organism increases the potential for
that may contaminate surface or groundwater through
waterborne, foodborne, person-to-person, and also zoonotic
infiltration or run-off during periods of peak rainfall. Once
transmission.
deposited in receiving waters, oocysts have been shown to
survive in river water up to six months with a die-off rate
Waterborne Outbreaks
of 94% after 176 days (46).
The effects of both high and low temperatures on the Numerous waterborne outbreaks of Cryptosporidium both
survival of Cryptosporidium have been investigated using in the United States and throughout the world have
animal infectivity analysis (47). Inactivation greater than been documented (6,40). Outbreaks have been attributed
99.9% can be achieved by raising the temperature to 70 ° C to all water types including contaminated rivers, lakes,
for 1 min. These findings suggest that Cryptosporidium springs, and groundwater sources. The first reported
oocysts can be rendered noninfectious when held for outbreak in the United States occurred in 1984, in
relatively short times at temperatures far below boiling. Braun Station, Texas, with 368 cases identified (49).
654 BIOLOGY OF CRYPTOSPORIDIUM

This was followed by a larger outbreak in 1987 in Table 3. Outbreaks of Cryptosporidium Attri-
Carrollton, Georgia, with an estimated 13,000 cases buted to Recreational Exposure
of cryptosporidiosis (50). The largest reported drinking Date Number of Outbreaks/Facility Reference
water outbreak of cryptosporidiosis in the United States
took place in Milwaukee, Wisconsin, in 1993, with 1988 2/pool (58,59)
an estimated 403,000 cases — 4,000 hospitalization and 1990 1/pool (56)
several deaths (51). The outbreak was caused by oocysts 1992 2/pool, 1/wave pool (60,61)
that had passed through the filtration system of the city’s 1993 4/pool (62,63)
southern water treatment plant. It has been hypothesized 1994 2/ pool, 1/lake (64–66)
that heavy spring rains and runoff from melting snowfall 1995 1/pool, 2/water park (67)
1996 1/water park, 2/pool, 1/lake (68)
led to the contamination of the city’s source water (Lake
1997 1/river, 2/pool, 1/fountain (69)
Michigan) with agricultural wastes (51,52). This outbreak
1998 4/pool (70,71)
was estimated to have cost the community $53 million in 1999 1/fountain (57)
lost wages, lost productivity, medical bills, and emergency
room visits. Further, the outbreak is said to have cost
upward of $100 million in claims for loss of life (53).
Although treatment failures have been documented of oocysts into a confined body of water, causing an
to have occurred during outbreaks situation, numerous immediate risk to the pools occupants. Further, the
waterborne outbreaks of Cryptosporidium have occurred oocysts are extremely resistant to chlorine levels used
in communities within the United States where evaluation in conventional swimming pools and their small size
of the water treatment plant during outbreak periods enables them to pass through filter systems, which are
revealed no violations for turbidity, chlorine concentration, not operating optimally.
or coliform levels (54). Because of the small size of
this organism (4–6 µm), the optimization of filtration Foodborne Outbreaks
for the removal of oocysts during drinking water Although Cryptosporidium appears to be spread more
treatment is extremely important. Similarly, the use often by a waterborne route, contamination of food
of alternative disinfectants that have been shown to products either by infected food handlers or through the
inactivate Cryptosporidium oocysts, such as ozone and use of contaminated processing or irrigation waters has
ultraviolet light, are also important factors to be been documented (72–75). It has been suggested that the
considered for the prevention of drinking water outbreaks. high prevalence of Cryptosporidium species in some food
Although the source of contamination during an animals increase the risk for contamination of other food
outbreak often goes undetected, suspected sources of products.
contamination include wastewater and heavy rainfall In nonindustrialized countries, the use of excreta as
events, which can lead to increasing agricultural and other fertilizer or the use of irrigation water that has been
nonpoint source inputs. Although the agricultural runoff contaminated with industrial, domestic, or agricultural
from cattle and sheep raising areas have been repeatedly
wastes can also lead to increased risk of foodborne
implicated as sources of contamination during waterborne
cryptosporidiosis, especially from consumption of raw
outbreaks both in the United States and abroad, only
vegetables (76). In a study in Costa Rica, oocysts from
once has the bovine genotype been identified as being
Cryptosporidium species were detected in 5.0% of cilantro
present, during an outbreak within North America.
leaves, 8.7% of cilantro roots, and 2.5% of lettuce
This outbreak occurred in Cranbrook, British Columbia,
samples (77). In Peru, a survey of vegetables collected
with approximately 2,000 cases of cryptosporidiosis. The
from small rural markets detected the presence of C.
bovine isolate (genotype 2) was detected in human fecal
parvum oocysts in 14.5% of the samples that were
specimens, in cattle manure specimens found near the
collected (78). Although these studies reveal the presence
watershed, and in water samples collected from the
of oocysts in raw produce, it is often impossible to detect
reservoir (10).
the presence of Cryptosporidium oocysts in food products
because of the lack of sensitive detection methods and the
Recreational Outbreaks
potential for inhibition of detection when using molecular
Several recreational outbreaks of gastrointestinal illness methods. Similarly, the long incubation period and
from Cryptosporidium have been reported during recent potential for person-to-person transmission of the disease
years, both in the United States and abroad (Table 3.) make epidemiological investigation of potential foodborne
Fecal accidents from a person with current or recent outbreaks of cryptosporidiosis extremely difficult.
diarrhea were suspected in most of the outbreaks. Infected The first foodborne outbreak of Cryptosporidium in
individuals may continue to excrete large numbers of the United States was reported in 1993. The outbreak
infectious oocysts for weeks after symptoms cease (55). was associated with fresh-pressed apple cider (72). An
Attack rates during outbreaks have been estimated to be environmental survey revealed that the apples had been
as high as 78% (56). Although there have been documented collected from uncultivated trees on the edge of a pasture
outbreaks of cryptosporidiosis associated with freshwater where cows had recently been grazing. The apples were
sources, many more cases have been associated with harvested both from trees and from the ground. The
swimming pool exposures (57). This is not surprising apples were then stored overnight and sprayed with
because a single fecal accident can input high number municipal water from a hose on the morning that the
 BIOLOGY OF CRYPTOSPORIDIUM 655

cider was prepared. Frozen samples of the cider revealed families, in day care centers, and also in other institutional
the presence of Cryptosporidium oocysts at levels of 375 settings (88–91). Day care centers have also been linked
to 750 oocysts per liter. There were 160 primary cases of to transmission of cryptosporidiosis through contact
cryptosporidiosis with a 15% rate of secondary household with contaminated toys, surfaces, and other fomites (89).
illness transmission. Although the main route of infection remains the fecal-
Three other outbreaks have been reported since oral route, Cryptosporidium oocysts have been detected in
then, one was associated with chicken salad, one sputum and vomit, and these may provide alternate modes
with green onions, and the third was also associated of transmission.
with apple cider (73–75). Suspected foodborne transmis-
sion of Cryptosporidium has also been reported from CONCLUSION
travelers visiting Mexico, the United Kingdom, and
Australia. Suspected foods include salad, raw milk, Cryptosporidium has been endowed with a number
sausage, and tripe. It has been proposed that live- of physical and physiological characteristics that have
stock infection and water contamination are the two enabled it to become an important cause of waterborne
major causes of food contamination with Cryptosporidium disease throughout the world. Although our body of
oocysts (26). knowledge, concerning the biology and epidemiology of this
Cryptosporidium oocysts have currently been detected organism continues to grow, we have only recently gone
in shellfish such as oysters, mussels, clams, and other beyond the tip of the iceberg with regard to understanding
bivalves that were harvested from both U.S. waters this organism’s place in the microcosm of waterborne
and abroad (79,80). Studies have shown that shellfish disease.
are capable of concentrating pathogenic bacteria, virus,
and parasites from the large volumes of water that
BIBLIOGRAPHY
they pass over their gills. Because many of these filter
feeders are consumed raw, they can present a significant
1. J. L. Meisel, D. R. Perera, C. Meligro, and C. E. Rubin,
public health risk to consumers. Currently, no outbreak Gastroenterology 70, 1156–1160 (1976).
of cryptosporidiosis caused by the consumption of raw
2. F. A. Nime et al., Gastroenterology 70, 592–598 (1976).
shellfish has been reported.
3. J. P. Dubey, C. A. Speer, and R. Fayer, Cryptosporidiosis of
Man and Animals, CRC Press, Boca Raton, Fla., 1990.
Zoonotic Transmission
4. C. Franzen and A. Muller, Diag. Microbiol. Infect. Dis. 34,
Zoonotic transmission of the organism is well recognized, 245–262 (1999).
and domestic cattle have been implicated in both direct 5. L. Xiao et al., Parasitol. Today 16(7), 287–292 (2000).
(fecal–oral) and indirect (via water) outbreaks of human 6. R. Fayer, U. Morgan, and S. J. Upton, Intern. J. Parasit. 30,
infection (24,27). Seasonal peaks in the detection of 1305–1322 (2000).
Cryptosporidium oocysts in the environment tend to 7. R. Fayer, C. A. Speer, and J. P. Dubey, in R. Fayer, ed.,
coincide with lambing and calving, in spring and autumn. Cryptosporidium and Cryptosporidiosis, CRC Press, Boca
Although symptomatic infection in animals generally Raton, Fla., 1997.
occurs only in the very young, adult animals such as 8. T. K. Graczyk, R. Fayer, and M. R. Cranfield, J. Parasitol.
deer, horses, pigs, and sheep may excrete low number of 82, 748–751 (1996).
oocysts (81). These adult animals may provide a reservoir 9. S. Tzipori and J. K. Griffiths, Adv. Parasitol. 40, 5–36
for transmission of this organism to the young of their (1998).
own species and also to humans. Studies conducted by the 10. M. M. Peng, L. Xiao, and A. R. Freeman, Emerg. Infect. Dis.
U.S. Department of Agriculture, Animal Plant Health 3, 567–573 (1997).
Inspection Service, and Veterinary Services (82) have 11. J. R. Harris and F. Petry, J. Parasitol. 85(5), 839–849
found Cryptosporidium species commonly in dairy and (1999).
beef calves in the United States. It has been estimated 12. D. W. Reduker, C. A. Speer, and J. A. Blixt, Can. J. Zool. 63,
that this parasite is present in more than 90% of the dairy 1892–1896 (1985).
farms and nearly 40% of the beef operations. 13. F. Bouchet and Y. Boulard, Paratiol., Res. 77, 585–589
Companion animals, such as cats and dogs, have also (1991).
been implicated in the transmission of cryptosporidio- 14. R. D. Pinckney et al., Vet. Parasitol. 47, 225–233 (1993).
sis to humans on rare occasion (5,26). However, they 15. A. Bonnin, J. F. Dubremetz, and P. Camerlynck, Infect.
may represent an important reservoir for human infec- Immun. 59, 1703–1708 (1991).
tion (83). A recent study using genetic sequencing of 16. W. L. Current and N. C. Reese, J. Protozool. 33, 98–108
the small subunit ribosomal RNA (SSU–rRNA) has (1986).
identified the first human cases of infection by a 17. C. L. Chappel et al., J. Inf. Dis. 173, 232–236 (1996).
canine and a feline species of Cryptosporidium in AIDS 18. B. L. P. Ungar, in J. P. Dubey, C. A. Speer, and R. Fayer,
patients (84). Transmission by arthropods and by house- eds., Cryptosporidiosis of Man and Animals, CRC Press,
flies and wild-filth flies that were exposed to feces con- Boca Raton, Florida, 1990, p. 59.
taining Cryptosporidium oocysts has been documented 19. F. G. Crawford and S. H. Vermund, CRC Crit. Rev. Micro-
(85–87). biol. 16, 113 (1988).
Secondary spread of cryptosporidiosis through person- 20. L. Jokipii and A. M. M. Jokipii, N. Engl. J. Med. 315,
to-person transmission has been well documented within 1643–1647 (1986).
656 BIOLOGY OF CRYPTOSPORIDIUM

21. L. Jokipii, S. Polijola, and A. Jokipii, Lancet 2, 358–361 57. Center for Disease Control, Morb. Mortal. Wkly. Rep. 2000.
(1983). 58. R. E. Joce et al., Epidemiol. Infect. Dis. 107, 496–508 (1991).
22. K. A. Adal, C. R. Sterling, and R. L. Guerrant, in M. J. 59. F. J. Sorvillo, K. Fujioka, and B. Nahlen, Am. J. Public
Blaser et al., eds., Infections of the Gastrointestinal Tract, Health 82, 742–744 (1992).
Raven Press, New York, 1995, pp. 1107–1128.
60. D. A. Hunt et al., Commun. Dis. Rep. CDR Rev. 4, R20–R22
23. A. Carr et al., Lancet 351, 256–261 (1998). (1994).
24. H. L. Dupont et al., N. Engl. J. Med. 332, 855–859 (1995). 61. J. M. McAnulty, D. W. Fleming, and A. H. Gonzalez, J. Am.
25. P. C. Okhuysen et al., J. Infect. Dis. 180, 1275–1281 (1999). Med. Assoc. 272, 19597–19600 (1994).
26. D. P. Casemore, Epidemiol. Infect. 104, 1–28 (1990). 62. J. Bongard et al., JAMA 272, 914 (1994).
27. S. Tzipori, Microbiol. Rev. 47, 84–96 (1983). 63. W. R. MacKenzie, J. J. Kzmierczak, and J. P. Davis, Epid-
28. D. P. Casemore and F. B. Jackson, Lancet 2, 679 (1983). miol. Infect. 115, 545–553 (1995).
29. D. P. Casemore, R. L. Sands, and A. Curry, J. Clin. Pathol. 64. L. A. Wilberschied, Kansas Med., 67–68 (1995).
38, 1321–1336 (1985). 65. J. M. Lemmon, J. M. McAnulty, and J. Bawden-Smith, Med.
30. R. Soave and D. Armstrong, Rev. Infect. Dis. 8, 1012–1023 J. Aust. 165, 613–616 (1996).
(1986). 66. M. H. Kramer et al., Clin. Infect. Dis. 26, 27–33 (1998).
31. R. Fayer and L. P. Unger, Microbiol. Rev. 50, 458–483 67. D. A. Levy et al., Morb. Mortal. Wkly. Rep. 47(SS-5), 1–34
(1986). (1998).
32. P. Ma, in J. S. Remmington and M. N. Swartz, eds., Current 68. T. Sundkvist et al., CDR Rev. 12, R190–R192 (1997).
Clinical Topics in Infectious Diseases, McGraw-Hill, New
69. Centers for Disease Control, Morb. Mortal. Wkly. Rep. 1997.
York, 1987, pp. 99–149.
70. M. Baker et al., N. Zeal. Pub. Hlth. Rpt. 5, 41–45 (1998).
33. W. L. Current, CRC Crit. Rev. Envir. Contr. 17, 21–51, 1986.
71. M. E. Hellard et al., Aust. N. Zeal. J. Publ. Hlth. 24, 272–275
34. W. L. Current and L. S. Garcia, Clin. Microbiol. Rev. 4,
(2000).
325–358 (1991).
72. P. Millard et al., JAMA 272, 1592–1596 (1994).
35. H. B. Mayer and C. A. Wanke, AIDS 8, 1639–1648 (1994).
73. Center for Disease Control, Morb. Mortal. Wkly. Rep. 45,
36. K. W. Angus, in J. P. Dubey, C. A. Speer, and R. Fayer, eds.,
Cryptosporidiosis of Man and Animals, CRC Press, Boca 783 (1996).
Raton, Fla., 1990, pp. 83–103. 74. Center for Disease Control, Morb. Mortal. Wkly. Rep. 46,
37. C. Pedersen et al., Genitourin. Med. 72, 128–131 (1996). 4–8 (1997).
38. F. J. Sorvillo et al., Epidemiol. Inf. 113, 313–320 (1994). 75. Center for Disease Control, Morb. Mortal. Wkly. Rep. 47, 27
(1998).
39. J. T. Lisle and J. B. Rose, J. Water Supply Res. Technol. 44,
103–117 (1995). 76. R. Feachment et al., in Sanitation and Disease Health
Aspects of Excreta and Wastewater Management, John Wiley
40. H. V. Smith and J. B. Rose, Parasitol. Today 14, 14–22
& Sons, New York, 1983, p. 501.
(1998).
77. R. Monge and M. Chinchilla, J. Food Protect. 59, 202–203
41. C. Chauret et al., J. Am. Water Works Assoc. 87, 76–84
(1995).
(1995).
78. Y. R. Ortega et al., Am. Society Trop. Med. Hyg. 57(6),
42. M. W. LeChevalier, W. D. Norton, and R. G. Lee, Appl.
Environ. Microbiol. 57, 2617–2621 (1991). 683–686 (1997).

43. J. B. Rose, C. P. Gerba, and W. Jakubowski, Environ. Sci. 79. R. Fayer et al., Appl. Environ. Microbiol. 64, 1070–1074
Technol. 25, 1393–1400 (1991). (1998).
44. C. M. Hancock, J. B. Rose, and M. Callahan, J. AWWA 90(3), 80. F. Freire-Santos et al., J. Parasit. 86, 853–854 (2000).
58–61 (1998). 81. K. W. Angus, in K. W. Angus and D. A. Blewett, eds., Cryp-
45. H. Solo-Gabriele and S. Neumeister, J. Am. Water Works tosporidiosis. Proceedings of the First International Work-
Assoc. 88, 67–86 (1996). shop, Moredun Research Institute, Edinburgh, Scotland,
1988, pp. 43–53.
46. L. J. Robertson, A. T. Campbell, and H. V. Smith, Appl.
Environ. Microbiol. 58, 3494–3500 (1992). 82. I. Laberge, M. W. Griffiths, and M. W. Griffiths, Int. J. Food
Microbio. 31, 1–26 (1996).
47. R. Fayer, Appl. Environ. Microbiol. 60, 2732–2735 (1994).
83. I. D. Robertson et al., Int. J. Parasit. 30, 1369–1377 (2000).
48. R. Fayer and T. Nerad, Microbiol. Rev. 50, 458–483 (1996).
84. N. J. Pieniazek et al., Emerg. Infect. Dis. 5, 444–449
49. R. G. D’Antonio et al., Ann. Intern. Med. 103, 886–888
(1985). (1999).

50. E. B. Hayes et al., N. Engl. J. Med. 320, 1372–1376 (1988). 85. R. Zerpa and L. Huicho, Ped. Inf. Dis. J. 13, 546–548 (1994).
51. W. MacKenzie et al., N. Engl. J. Med. 331(3), 161–167 86. T. K. Graczyk et al., Am. J. Trop. Med. Hyg. 61, 500–504
(1994). (1999).
52. H. G. Cicirello et al., Epidemiol. Infect. 119, 53–60 (1997). 87. T. K. Graczyk et al., Emerg. Inf. Dis. 5, 726–727 (1999).
53. M. H. Kramer et al., Morb. Mortal. Wkly. Rep. 45, 1–33 88. D. P. Casemore, PHLS Microbiol. Digest 4, 6–7 (1987).
(1996). 89. H. Heijbel et al., Pediatr. Infect. Dis. J. 6, 532–535 (1987).
54. H. V. Smith and J. B. Rose, Parasit. Today 6, 8–12 (1990). 90. J. Hannah and T. Riorden, Publ. Hlth. 102, 539–544 (1988).
55. P. L. Meinhardt, D. P. Casemore, and D. Miller, Epidemiol. 91. M. A. Thompson, J. W. T. Benson, and P. A. Wright, Arch.
Rev. 18, 118–136 (1996). Dis. Child 62, 559–563 (1987).
56. A. Bell et al., Canadian J. of Public Health 84, 334–337 92. B. Rush, P. Chapman, and R. Ineson, Lancet ii, 632–633
(1993). (1987).
 BIOLUMINESCENCE, METHODOLOGY 657

93. Anonymous, in Cryptosporidium in Water Supplies. Report Bioluminescence has proven important for a wide range
of the Group of Experts, Dept. of the Env., Dept. of Health, of disciplines central to the study of human disease, includ-
London, U.K., HMSO., 1990. ing hematology, immunology, bacteriology, and pharma-
94. A. Richardson et al., Epidemiol. Infect. 107, 485–495 (1991). cology, and clinical chemistry (8). More generally, it has
95. C. Joseph et al., Epidemiol. Infect. 107, 509–519 (1991). stimulated studies in analytical biochemistry, molecu-
96. S. A. Bridgman et al., Epidemiol. Infect. 115, 55–566 (1995). lar biology, cell-based assays, and various environmental
applications (10).
Specific examples of bioluminescent-facilitated re-
search include work with pathogens such as the Mycobac-
BIOLUMINESCENCE, METHODOLOGY terium tuberculosis, which not only pose biosafety hazards
but grow very slowly in laboratory conditions. An enhanced
PAMELA CONTAG luciferase-expressing mycobacterial strain has been used
Xenogen Corporation to evaluate antimycobacterial activity in mice (11), permit-
Alameda, California ting evaluation of drugs, with significant savings in time,
labor, and expense. A pharmaceutical application, in vivo
Bioluminescence is the capacity of certain living organisms tracking of cells and monitoring of gene expression (12) has
to emit visible light. The species that exhibit biolumines- broad applications to the study of infectious disease. Cur-
cence are found in many environments, but especially in
rently, this technology has developed into widely applied
the deep sea. They have achieved bioluminescence through
methods for monitoring biological events noninvasively in
different evolutionary processes. The responsible genes in
living animals.
bacteria, algae, coelenterates, beetles (fireflies), and fish
are all unrelated to one another (4). The fundamental bio-
chemistry of light production in these organisms consists IN VIVO BIOPHOTONIC IMAGING
of the oxidation of a luciferin, a cellular substrate, under
the catalysis of luciferase enzymes, which produces, as a In vivo biophotonic imaging is a noninvasive means of
decay product of the chemical reaction, an almost heatless, tracking pathogens, tumor cells, or molecular events in live
bluish green light until the excited-state molecules return animal subjects that exploits the light-emitting properties
to the ground state. The luciferin substrates and the struc- of photoproteins such as luciferase enzymes. Typically,
tures of the luciferase enzymes vary by organism, as do animal models for biological assessment represent late
the mechanisms controlling the speed and intensity of the or end stages of a disease, usually require sacrifice of
luminescence. Certain cofactors, such as the nucleotide the animal, and employ cumbersome ex vivo assays to
ATP, may need to be present for the conversion of the reveal the disease process. Moreover, much contextual
substrate to take place (5). information is lost that an in vivo analysis might
Because these reactions can be replicated outside the yield, because late stage protocols, more distressful
organisms to which they are native, bioluminescence has to experimental animals, are likely to miss important
proven extremely useful for research at the cellular and information about events in the early stages of a disease
molecular levels (6). The availability of instrumentation and from intact systems.
that can measure the light emitted in these reactions Biophotonic imaging holds great promise for the accel-
with great sensitivity and dynamic range has made eration of data generation and analysis for the devel-
them powerful tools for biochemical and clinical analysis opment of new drug candidates by the pharmaceutical
because the involved components can be detected at industry.
a very low level (7). Compared with other techniques, This is because 80% of drug candidates fail to satisfy
such as colorimetric or spectrophotometric indicators, safety and efficacy requirements in clinical trials, and
bioluminescent analysis offers the advantages of high the need to conduct biological assessments in animal
sensitivity, wide linear range, low cost per test, and models to determine safety and efficacy is the principal
relatively simple and inexpensive equipment (8). Because obstacle in making the development process more efficient
of their sensitivity, immobilized bioluminescent reagents and effective. In vivo biophotonic imaging offers a rapid
have come into widespread use in biochemical analysis means of conducting in vivo analysis to improve the
over the last 20 years. predictive quality of data used in the assessment of
Firefly and bacterial luciferases are the two most new drug candidates. The technology provides data sets
commonly used bioluminescent systems in research (5). from relevant, intact animal systems, increasing both the
The latter is the light-producing reaction of the North efficiency and the effectiveness of selecting new drug
American firefly Photinus pyralis, which is also the candidates for clinical development. Because it is an in
most extensively studied bioluminescent system. Firefly vivo technology, fewer animals are needed, whereas time
bioluminescence, for example, allows the presence of and costs are reduced and more data per protocol are
the ATP nucleotide, which supplies energy to cells for obtained.
many important biochemical processes, to be determined Before the advent of biophotonic imaging, noninvasive
in reactions with extreme sensitivity (7). Although most in vivo assays were based on magnetic resonance imaging
enzyme assays yield either a product or the disappearance (MRI), positron-emission tomography (PET), and X-ray
of a substrate, firefly luciferase acts as a quantifiable (computerized tomography) technologies. Although these
reactant rather than as a catalyst, whose most significant methods are useful for monitoring structural changes due
and easily measured product is light (9). to disease processes and, on occasion, the response to
658 BIOLUMINESCENCE, METHODOLOGY

therapy, they required relatively expensive devices that in vivo processes, and then in vivo to test predictions
are not practical for use with most animal models and very made in vitro. In this way, a predictable animal model
long scan times. Moreover, specific contrast agents that with highly correlative data can offer validation of an in
could reveal the cellular and molecular changes relating vitro assay approach. Several parameters of drug efficacy
to the disease processes were not readily available. and pharmacokinetics can be monitored in the same
In vivo biophotonic imaging technology is based on the animal over time, yielding benefits for the discovery and
observation that light is transmitted through living tissue development stages of drug evaluation.
with efficiencies suitable for use in monitoring both struc- In animal studies, the number of subjects required
ture and function. The external application of light as a is reduced, and the issues of animal-to-animal variation
biological monitor currently has both routine clinical uses are overcome by the zero time point functioning as
(e.g., pulse oximetry) and novel experimental applications an internal control, improving biostatistics. In bacterial
in medicine. Biophotonic imaging, however, employs pho- models, bioluminescent tags are useful for distinguishing
toproteins, internally emitted biological sources of light, infecting pathogens from normal flora.
to tag such biological functions as infection and gene
expression (12–16). METHODOLOGY
Use of in vivo luciferase monitoring in living mam-
mals was first described in 1995 with an infectious Labeling Bacterial Cells
disease model, gastrointestinal infections by Salmonella
typhimurium (12). This study indicated that the technique Lux operons are ideal for labeling bacterial pathogens
was not only feasible but that it yielded quantitative data because they not only encode the luciferase genes but also
that provided more information in less time than could the genes for the biosynthetic enzymes (fatty aldehyde
be obtained from conventional assays. Because a large synthases) for the substrate decanal. Use of these operons
variety of bacterial species can be labeled with luciferase, obviates the need for an exogenous supply of substrate.
this approach is generally applicable to microbiological Few deleterious effects of substrate biosynthesis on
investigations (see Fig. 1, Gastrointestinal infection with bacterial metabolism or virulence have been observed.
S. typhimurium and treatment with ciprofloxacin). Ongo- Among the lux operons from bioluminescent bacteria,
ing research is showing that mammalian cells can be which from Photorhabdus luminescens, previously known
labeled in the same way as bacterial species, indicating as Xenorhabdus luminescens, appears to be ideally
that the method is applicable to the evaluation of genetic- suited for use in mammalian models (17,18), given that
and cell-based therapies, anticancer drugs, studies of gene mammalian body temperatures lie within the temperature
expression and development. optimum for this enzyme (19–21), unlike beetle luciferases
A significant advantage of in vivo luciferase monitoring (Luc) and other characterized bacterial luciferases (from
is that bioluminescent tags allow for an integrated the genus Vibrio) that have low-temperature optimum.
approach in which the same label can be used for in Well-characterized vectors are used for the expression
vitro, for quantitation and in cell culture correlates of of the lux operons in gram-negative organisms. The
original pUC-based plasmid used to clone the lux
operon from P. luminescens, including the promoter region
(vector designated pCGLS1) is well suited for expressing
luciferase from common gram-negative organisms such
as Escherichia coli and Salmonella (19,21,22). Standard
methods of bacterial cell transformation for these
organisms are used to introduce the lux-encoding vector
into cells of these bacterial species.
Bacteria may be labeled with luciferases from either
a bacterial or eucaryotic source (as shown in Fig. 2).
Several arrangements for the expression of luciferase have
been used, namely, the lux AB genes from firefly with
luciferin as the substrate, the lux AB genes from Vibrio
or Photorhabdus with decanal as substrate, or finally the
entire lux operon from Photorhabdus with no exogenous
substrate requirement.
Transposon-based vectors have been used to introduce
Figure 1. Gastrointestinal infection/Salmonella. The pseudo- lux genes modified for optimum expression in both gram-
color image is a color depiction of the intensity and quantity positive bacteria, such as Staphylococci and Streptococci,
of photons with red being most intense and violet least intense. and gram-negative organisms outside the Enterobacte-
Untreated and treated mice are depicted at time zero (eight
riaceae. The in vitro sensitivity for the labeled bacte-
days after oral inoculation) with SL1344 lux S. typhimurium, but
rial strains is approximately 10 to 100 bacterial cells
before treatment with ciprofloxacin. The same mice were imaged
five and a half hours after treatment with ciprofloxacin (untreated (Salmonella, LB5000 lux) that yields a signal in culture
control-received injection of phosphate-buffered saline). The over background, regardless of whether an absorbing and
graph shows relative bioluminescence intensity measured from scattering biological medium such as blood is present.
the abdominal area at the time after treatment for both treated The lower limit of detection in some in vivo models is
and untreated animals. See color insert. 1,000 bacterial cells.
 BIOLUMINESCENCE, METHODOLOGY 659

Figure 3. Eukaryotic expression. Expression of light from

Figure 2. Prokaryotic expression. For expression of light from Eukaryotic cells may be accomplished with the luc gene of the
prokaryotic cells, a construct is used that includes the lux North American firefly regulated by an SV40 viral promoter.
operon from P. luminescens in an appropriate gram-negative or Luciferin is administered introperitoneally as the substrate for
gram-positive plasmid vehicle. More recent constructs include
the enzymatic reaction. See color insert.
transposons for chromosomal integration. Expression of the luc
gene from the North American firefly may also be used in
prokaryotes as long as the luciferin substrate is administered. and acetyl CoA is added to samples and read in a standard
See color insert.
luminometer.
For cells in culture, 10 µl of a 15 mg/mL stock solution
Cell culture correlates for pathogenesis and expression of luciferin (per 1 mL of medium) is added to cells for
can provide a rapid means of evaluating tagged events a final concentration of 150 µg/mL. Cells are imaged
before introduction into living animals. Although corre- immediately after substrate addition. Bioluminescent
signals are apparent in these cultures hours later.
lates have not been established for some systems (e.g.,
Increased bioluminescent signals are obtainable by
propagation of human papillomavirus), when they are
increasing the substrate concentrations. An ultrasensitive
available, they can accelerate the development of in vivo
camera is used to measure the signals in living cells. The
assays and, in combination with luciferase tags, the same
cells can be lysed and luciferase activity measured in a
reporter can be used in vitro, in vivo, and finally as an ex
luminometer with the standard assay procedure if the
vivo assay to confirm in vivo observations. In the case of
signals produced, are too weak to be detected by camera.
bacterial infections, such assays have been described for
many organisms, and some have been adopted for use with
Substrate Delivery to Tissues
bioluminescent reporters, including for microscopic detec-
tion. In the case of mycobacterium, use of luciferase as a An aqueous solution of the substrate, luciferin (50 mM),
reporter can significantly accelerate in vitro assays and is injected into the peritoneal cavity 20 minutes prior to
such labeled strains have also begun to be used in vivo. imaging for systemic delivery at a dose of 126 mg/kg.
Alternatively, substrate can be applied topically (50 mM)
Cultured Mammalian Cells in 100% DMSO for transdermal delivery to monitor
luciferase expression in the skin. Luciferin has also been
Reporter genes, such as firefly luciferase or variants of electroporated (ionophoresis) into the skin of the ear, using
green fluorescent protein (GFP), in transformed cells a caliper electrode connected to an electroporator. The
can be employed to reveal the molecular and cellular duration of detectable luciferase activity in tissues varied
features of cancer. In a recent study (23), tumor cells with the different modes of delivery. With introperitoneal
were labeled with pGL3, a luciferase gene driven by the (IP) delivery, peak activity is detected at 20 minutes in the
SV40 viral promoter. Eucaryotic expression of the firefly brain and at five minutes in other tissues with proximity
luciferase in both cultured tumor cells and transgenic to the peritoneal cavity.
animals is accomplished by placing a constitutive or
inducible promoter in front of luciferase, transfection of Anesthesia
the vector into the chromosome, and then finally, the light
For imaging, mice and rats were anesthetized using
from luciferase expression in stable transfectants is seen
pentobarbital (35–70 mg/kg body weight) (12). Low doses
following the addition of luciferin (see Fig. 3). are used with neonatal mice and rats sensitive to the
effects of pentobarbital. Animals are observed carefully
Determination of Luciferase Activity Ex Vivo and In Vitro while under anesthetic and their respiration is monitored.
A commercially available assay kit that includes acetyl Following the imaging, the animals are kept warm and
CoA to prolong light emission can be used to determine allowed to recover.
luciferase activity for the firefly enzyme expressed in
Imaging
mammalian cells for tissue homogenates. Tissues are
removed and homogenized with a tissue disrupter in Anesthetized animals need to be placed in a light-
minimal volumes. A lysis buffer containing luciferin, ATP, tight box. A gray-scale body surface image is collected
660 BIOLUMINESCENCE, METHODOLOGY

for reference in low light. Even the smallest light Applications and Discussion
leaks can present significant background when using
In vivo monitoring of luciferase is widely applicable for
sensitive imaging devices and long integration times.
tracking pathogens and other disease processes, yielding
The potential sites for light leaks are the door closure,
more information in less time than conventional assays.
camera mount (if camera is mounted externally), and
Noninvasive monitoring of light emitted from within
the entry point for the cables (if camera is mounted
a living mammal, in which the light is constitutively
internally).
expressed or is reporting fluctuations in gene expression,
Photons emitted from luciferase within the animal, is a platform technology that results in increased
and then transmitted through the tissue, are collected information about specific physiological processes as well
and integrated for 5 to 30 minutes. A pseudocolor image as integrated, whole biological systems. The monitoring
representing light intensity (generally blue representing of biological light requires less time and fewer animals,
least intense and red representing most intense light) reducing the cost of analyses. This is especially relevant
is generated on an image processor. The images are to the costly processes of drug discovery in which
transferred using a plug-in module for image process- significant advances have been made in generating large
ing software to a microcomputer. Gray-scale reference numbers of potentially useful therapeutic compounds
images and pseudocolor images are superimposed using (e.g., combinatorial chemistries, genomics, and high-
the image software, and saved as PICT files. Superimposed throughput screening).
images are opened within a graphics software package In the absence of significant advances in in vivo
and composites made and annotations added. An intensi- analyses, the bottleneck in drug development is at the
fied charge coupled device (CCD) camera, described in the animal model step, and improvements in the earlier
following section, outfitted with a 50 mm f 1.2 lens and an steps only accentuates this problem. In vivo luciferase
image processor, has been successfully used in research monitoring addresses this limitation (16) with broad
studies. applications.
In studies of infectious disease, for example, the in vivo
Use of the CCD Camera biophotonic monitoring technology can be superimposed
Low light imaging systems employ technologies that over current methodology to perform faster experiments,
either reduce the background noise or increase the and use fewer animals, while still extracting relevant data
signal. By cooling the CCD chip in a video camera to that provides more information relating to early stages of
temperatures of about −30 ° C the background caused disease and patterns of disease.
by infrared irradiation (so-called thermal noise) can be In Figure 4, a lung model illustrates a high level of
reduced permitting detection of low light signals. An bacteria in the lung and nasopharangeal area that can be
alternative is to specifically increase the signal using specifically characterized as pneumonia, with the lungs
an image intensifier. Intensifiers are designed around a clearly outlined. In Figure 5, a typical model used to
high voltage microchannel plate technology that amplifies evaluate antibiotics, the soft tissue thigh model, can
signals in the form of electrons. A photocathode is used now be used as a fast screen for effective antibiotics.
to convert photons to electrons that are then amplified Biophotonic imaging allows this model to be run in half
over the microchannel plate; the increased number of the time of the traditional thigh model. In Figure 6, labeled
electrons contact a phosphor screen generating photons tumor cells allow researchers to follow the expansion
that are then focused on a CCD video camera. The and regression of tumors in response to therapy. This is
materials used in the photocathode determine the spectral especially important as tumor diagnostics leap ahead and
sensitivity of the intensifier. To reduce thermal noise
blue-sensitive materials can be employed. Alternatively,
thermal noise on the intensifier is reduced by cooling
intensifiers with photocathodes made of materials that are
sensitive to longer wavelengths of light-cooled intensified
cameras.
In a comparison between a cooled camera and an
intensified camera, essentially equivalent signals are
obtained from neonatal Tg mice expressing luciferase in
their eyes and skin. In adult mice with bioluminescence
originating from deeper tissues, differences have been
apparent between cooled intensified systems (sensitive to
light ≤850 nm) and the C2,400-32 ICCD camera (sensitive
to light ≤600 nm). This is likely because of a suspected
filter effect of the tissues in which shorter wavelengths of
light are more readily absorbed by mammalian tissues,
and the longer wavelengths of light (luciferases have Figure 4. Lung model. Neonatal mice (under two weeks of
rather broad spectral peaks of up to 50 mm bands) age) are susceptible to lung infection following oral inoculation
may preferentially pass through tissues and more red- with S. typhimurium SL1344 lux. A nasopharyngeal infection
sensitive instruments may be well suited for collecting is common in both adults and neonates from this mode of
this light. inoculation. See color insert.
 BIOLUMINESCENCE, METHODOLOGY 661

200
150
100
50
0

Figure 7. Transgenic animal exposed to heavy cadmium. A

mouse line transgenic for a cadmium-regulated element, the
hemooxygenase promoter, demonstrates in vivo monitoring of
Figure 5. Soft tissue thigh model Lux tagged E. coli are injected
cadmium induction of luc in front of the luc gene. 10 mM cadmium
at 104 bacteria/mL (right thigh) and 105 bacteria/mL (left thigh).
was delivered IP and the mouse was imaged at zero and three
Bacteria can then be quantified noninvasively with and without
hours. The three-hour time point with a camera sensitivity set at
antibiotic treatment. See color insert.
a bit range of zero to three is shown here. See color insert.

affect gene expression can be monitored noninvasively.

This transgenic animal was exposed to the heavy metal
600
cadmium, which is known to induce gene expression, and
three hours after exposure, the liver and testes show
400
a specific increase in gene expression (see Fig. 7). The
changes in gene expression were monitored and it was
200
demonstrated that they occurred in a dose-dependent
fashion. By Northern analysis it was also demonstrated
0
that HO protein was present.
0 7 14
CONCLUSION
Figure 6. Labeled tumor cells. Labeled NIH 3T3 cells were
It has been demonstrated that in vivo biophotonic imaging
injected with matrigel at zero time [104 cells (right thigh) and
105 cells (left thigh)] and mice were imaged between 7 and 14 days can deliver higher throughput animal models for a variety
to monitor the increase in cell number, as demonstrated by the of pharmacological screens. The variety of animal models
increase in red pixels of the pseudocolor image. See color insert. that this technology can be applied to is very broad
and includes bacteria (gram-negative and gram-positive)
and tumor models. Increasing the efficiency of the drug
we need tumor therapies that provide earlier intervention development process, although laudable, is only a small
in the disease process. Virtually the first steps in the part of the story.
disease can be monitored, including metastatic disease. Gene sequencing has made it possible to assess the
Light-producing transgenic animals have allowed relatedness of species. More recently, there have been
scientists to follow gene expression spatially and over increases in our ability to group genes into families and to
time. The power of this technology rests in the ability to draw possible conclusions about relatedness of function.
measure from baseline to maximum gene expression in But what remains elusive is a true understanding of
one animal over time and during the changing conditions pathways, system connections, and the multifunctionality
of disease state, therapy, and toxic events. We can now of genes in metabolism and development. This is because
begin to understand the relationship between the level our in vitro assays are essentially two-dimensional;
at which a gene is expressed in any given animal and pathways may be connected but total systems, such as the
the concomitant change in phenotype. Thus, it is possible central nervous system, and the immune system are not
to have surrogate markers of gene expression for specific part of the assay. Current animal models are less robust
biological events. than desired because of artifacts of ex vivo analysis and
In a recent study, a transgenic animal was created using end points that do not reveal the biological process. In vivo
the promoter for hemooxygenase (also known as heat shock biophotonic imaging, however, allows real-time data to be
protein (HSP) (32)) placed in front of luciferase. Thus, collected noninvasively; thus, the animal subject functions
because the endogenous gene was not altered, events that as its own internal control. This means that investigators
662 BIOLUMINESCENCE, METHODOLOGY

are no longer limited to a snapshot of biological data 17. N. Boemare, R. Akhurst, and R. Mourant, Int. J. Syst.
but now can watch the actual process of metabolism and Bacteriol. 43, 249–255 (1993).
disease. 18. F. Rainey et al., Int. J. Syst. Bacteriol. 45, 379–381 (1995).
Bioluminescent-based assays have also shown promise 19. S. Frackman, M. Anhalt, and K. H. Nealson, J. Bacteriol. 172,
as a means of rapidly and reliably measuring bio- 5767–5773 (1990).
logical and chemical contamination of the environ- 20. E. Meighen, FASEB J. 7, 1016–1022 (1993).
ment and for detecting pathogens that are unde- 21. R. Szittner and E. Meighen, J. Biol. Chem. 265, 16581–
tectable by standard techniques, yet still threatening 16587 (1990).
to human health (5). Other recent environmental appli- 22. L. Xi, K. W. Cho, and S. C. Tu, J. Bacteriol. 173, 1399–1405
cations include luciferase-based measurement of water (1991).
contaminants (24), luminescence-facilitated detection of 23. C. H. Contag et al., Neoplasia 2, 41–52 (2000).
mercury in natural waters (25), and firefly luciferase- 24. M. A. Costanzo, J. Guzzo, and M. S. DuBow, Methods Mol.
tagged bacteria in environmental samples (26). Biol. 102, 201–217 (1998).
The ability to collect quantitative data that preserves 25. T. Barkay et al., Methods Mol. Biol. 102, 231–246 (1998).
spatial and temporal context with extraordinary sensi- 26. R. Tombolini and J. K. Jansson, Methods Mol. Biol. 102,
tivity opens new avenues for exploring in situ various 285–298 (1998).
environmental mechanisms of growth, colonization, and 27. J. M. Strong-Gunderson et al., Biosurfactants and Increased
community structure and organization. Using a bacterial Bioavailability of Sorbed Organic Contaminants: Measure-
biosensor (27,28) that produces light in the presence of a ments Using a Biosensor, Presented at the Conference on Gas,
particular signal would be very helpful in isolating cru- Oil, and Environmental Biotechnology, Colorado Springs,
cial molecular communication among bacteria. We may Colo., 1993.
speculate that this knowledge, along with two to three- 28. A. Heitzer et al., Appl. Environ. Microbiol. 60, 1487–1494
dimensional organizational structure, would give us a (1994).
clearer picture of microbial communities in nature.
ADDITIONAL READING
Acknowledgments
The author thanks Darlene Jenkins, Carole Bellinger-Kawahara, Aguilar-Cordova E. et al., AIDS Res. Hum. Retroviruses 10,
Weisheng Zhang, and Christopher H. Contag for sharing images 295–301 (1994).
of their work. Benaron D., Cheong W.-F., and Stevenson D. , Tissue Opt. Science
276, 2002–2003 (1997).
BIBLIOGRAPHY Campbell A. K., Chemiluminescence. Principles and Applications
in Biology and Medicine, Ellis Horwood Ltd. and VCH
Verlagsgesellschaft mbH, Chichester, U.K., 1988.
1. C. Brandes et al., Neuron 16, 687–692 (1996).
Finlay B. B., and Falkow S., Mol. Microbiol. 3, 1833–1841 (1989).
2. A. J. Millar et al., Science 267, 1161–1163 (1995).
Hastings J. W., Gene 173, 5–11 (1996).
3. A. J. Millar, S. R. Short, N. H. Chua, and S. A. Kay, Plant
Cell 4, 1075–1087 (1992). Hickey M. J. et al., Antimicrob. Agents Chemother. 40, 400–407
(1996).
4. T. Wilson and J. W. Hastings, Annu. Rev. Cell Dev. Biol. 14,
197–230 (1998). Hoiseth S. K. and Stocker B. A. D., Nature 291, 238–239 (1981).
5. P. Billard and M. S. DuBow, Clin. Biochem. 31, 1–14 (1998). Jobsis F. F., Science 198, 1264–(1977).
6. J. M. Kendall and M. N. Badminton, Trends Biotechnol. 16, Jones B. D. and Falkow S., Infect. Immun. 62, 3745–3752 (1994).
216–224 (1998). Lee C. A., Jones B. D., and Falkow S., Proc. Natl. Acad. Sci.
7. L. J. Blum and P. R. Coulet, in O. S. Wolfbeis, ed., Fiber Optic U.S.A. 89, 1847–1851 (1992).
Chemical Sensors and Biosensors, CRC Press, Ann Arbor, Liu Y. et al., J. Biol. Chem. 270, 24864–24870 (1995).
Mich., 1991, pp. 301–313. Magallanes M. et al., Antimicrob. Agents Chemother. 37,
8. T. P. Whitehead et al., Clin. Chem. 25, 1531–1546 (1979). 2293–(1993).
9. S. R. Ford and F. R. Leach, in R. La Rossa, ed., Biolumines- Morrey J. D. et al., J. Virol. 65, 5045–5051 (1991).
cence Methods and Protocols, Humana Press, Totowa, N.J., Morrey J. D. et al., J. Acquir. Immune Defic. Syndr. 5, 1195–1203
1998, pp. 3–20. (1992).
10. See the two dozen studies in R. La Rossa, as cited in Morrey J. D. et al., Intervirology 36, 65–71 (1993).
Reference 9. Morrey J. D. et al., Intervirology 37, 315–320 (1994).
11. M. J. Hickey, Antimicrob. Agents Chemother. 40, 400–407 Petterson J. et al., Science 273, 1231–1233 (1996).
(1996).
Sambrook J. et al., Molecular Cloning: A Laboratory Manual,
12. C. H. Contag et al., Mol. Microbiol. 18(4), 593–603 (1995). Cold Spring Harbor Laboratory Press, Cold Spring Harbor,
13. C. Contag et al., Photochem. Photobiol. 66, 523–531 (1997). N.Y., 1989.
14. C. H. Contag and P. R. Contag, Opt. Photonics News, 7(1), Sanderson K. E. and Stocker B. A. D., in Neidhardt F. C. et al.,
22–23, (1996). eds., Escherichia coli and S. typhimurium. Cellular and
15. C. H. Contag et al., in E. Sevick-Muraca and D. Benaron, Molecular Biology, vol. 2. American Society of Microbiology,
eds., Biomedical Optical Spectroscopy and Diagnostics, Washington, D.C., 1987.
vol. 3., Optical Society of America, Washington, D.C., 1996, Widder E., Latz M., and Herring P., Science 255, 512–514 (1984).
pp. 220–224. Wilson B., Jeeves W., and Lowe D., Photochem. Photobiol. 42,
16. P. I. Contag et al., Nat. Med. 4, 245–247 (1998). 153–162 (1985).
 BIOMASS: SOIL MICROBIAL BIOMASS 663

BIOMARKERS, LIPID. See LIPID BIOMARKERS IN uptake. Depending on the quality of litter, the micro-
ENVIRONMENTAL MICROBIOLOGY bial biomass either utilizes the nutrients (immobilize) or
if there are excess of nutrients, they become available
(mineralize) for other organisms or plant uptake. In this
respect, the soil microbial biomass acts as a source and
sink for nutrients in the soil. For these reasons, the soil
BIOMASS, ALGAL. See PLANKTONIC ALGAE IN THE microbial biomass is the main regulator of nutrient cycling
MARINE ENVIRONMENT in the soil, and therefore, regulates net primary production
of ecosystems.
The importance of the soil microbial biomass goes
beyond nutrient cycling. One of the important by-products
of decomposition is the formation of stable organic
BIOMASS, BACTERIOPLANKTON. See PLANKTONIC matter. The soil organic matter, through its interaction
MICROORGANISMS: BACTERIOPLANKTON with minerals, serves many functions that increase soil
quality through enhancement of physical, chemical, and
biological characteristics of the soil matrix. An important
consequence of an increase in soil organic matter is
the storage of vital plant nutrients, such as nitrogen,
BIOMASS: SOIL MICROBIAL BIOMASS phosphorus, sulfur, and trace metal elements. The
diversity of soil microbial biomass leads to other important
WILLIAM R. HORWATH biogeochemical processes, which regulate gaseous flux
University of California of carbon, nitrogen, and other nutrients. In contrast,
Davis, California
other functions of the soil microbial biomass, such as
the production of growth regulators, can serve to be
Soil microbial biomass is an important component of soil detrimental or enhance plant growth. The amazing
that regulates many processes associated with energy diversity of the soil microbial biomass is the foundation
transfers and nutrient cycling. These functions are crit- of a complex ecosystem component that regulates the
ical to maintaining ecosystem productivity at all levels productivity of the earth’s biomes. This article will
of the food web. The soil biomass is composed of a concentrate on the soil microbial biomass, which is
wide range of microorganisms including viruses, bacte- composed mainly of bacteria and fungi and will emphasize
ria, fungi, microfauna, and macrofauna. The soil microbial carbon and nitrogen cycling in soil.
biomass expresses functions to take advantage of the
multitude of soil niches composed of different habitats HABITAT OF THE SOIL MICROBIAL BIOMASS
and substrates. These functions range from pathogenesis,
symbiosis, and heterotrophic to chemoautotrophic activi- Soils are formed through the dissolution of primary min-
ties. erals and subsequent reformation of secondary minerals,
The soil microbial biomass is a component of the such as clays and sesquioxides. The secondary minerals
soil that is considered to be part of the active fraction. are rich in ion exchange activity and form stable com-
The active fraction includes the microbial biomass, plexes with organic matter. The interaction of minerals
recently deposited plant residues, root exudates, and and soil organic matter lead to the formation of soil struc-
easily degradable portions of the soil organic matter ture through the creation of aggregates and porosity. This
such as light fraction, which are thought to play a complex three-dimensional matrix produces a wide vari-
prominent role in nutrient cycling and major energy ety of habitats in which the soil microbial biomass exist.
transfers (1). In most soils, the soil microbial biomass The three-dimensional matrix is composed of solid, liq-
comprises about 5% of total soil carbon and about 1% uid, and gas phases. Characteristically, a well-developed
of total soil nitrogen (2). The microbial biomass is most soil contains 50% solids and 50% pores. The pores in the
active in the surface soil where most of the recent plant soil matrix contain the soil solution and air. The ionic
and litter inputs, mainly from above and belowground exchange capacity of the minerals act to adsorb nutrients
production and turnover, provide substrate (food) for required for the growth of the soil microbial biomass and
microbial activity. Deeper soil horizons contain less plant act as a surface to exist on. The wide array of habits in
input, and therefore a corresponding decrease in microbial soil creates a complex predator-prey interaction, which is
population size and activity. However, in most soils, unrivaled compared to aboveground ecosystems (Fig. 1).
microbial biomass is present at all soils depths to the depth Soil microorganisms inhabiting larger pore areas are
of the bedrock including deep sediments of 1,000 feet or subject to grazing by larger organisms, such as protozoa
more (3). and nematodes. However, soil organisms living in larger
The most important function of the soil microbial spaces normally have access to a greater food supply
biomass is decomposition of organic material. During through the movement of the soil solution and exploration
decomposition, the microbial biomass releases nutrients by plant roots. Microorganisms inhabiting smaller areas,
from plant litter and gains energy for metabolic processes. such as capillary spaces, are protected from predation
Without this function, dead plant material would accu- by larger organisms. However in these confined spaces,
mulate and limited nutrients would be available for plant microorganisms are often subject to oxygen limitations
664 BIOMASS: SOIL MICROBIAL BIOMASS

Figure 1. Depiction of the soil habitat

showing soil structure and the interaction
of organisms inhabiting various niches. The
trophic level interactions depict protected
and nonprotected regions of the soil matrix.

and substrate availability. The variety of habitats Table 1. The Mass of Soil Biomass Components in
is directly responsible for influencing the immense Soil
diversity of the soil microbial biomass. Through long-term Soil Biomass Component Tonnes Per Hectare
adaptive strategies microorganisms have adapted to these
specialized niches in the soil. Bacteria 1 to 2
The ecological theory that examines the distribution Fungi 2 to 5
of species based on substrate availability and growth Actinomycetes 1 to 2
involves the concept of r and K selection (4). A soil microor- Protozoa Up to 0.5
ganism adapted to bountiful energy and nutrient sources Nematodes Up to 0.2
Earthworms 0 to 2.5
are designated as r-selected. Microorganisms existing
Other fauna (collembola, mites, Up to 0.5
under low energy and nutrient deprived conditions are arthopods etc.)
termed K-selected. Selection pressures and physical envi-
ronment would be unique to r- and K-selected organisms Source: K. Killham, Soil Ecology, Cambridge University Press,
leading to diversity of function (5). The K-selected organ- New York, 1994.
isms would strive to produce a high growth- rate per unit
of substrate because food supply would be erratic in the
smaller capillary pores or protected spaces in the soil. The debate among scientists. The size of soil microbial biomass
r-selected organisms on the other hand would put much and all of its major components is described in Table 1.
energy into competitiveness to be able to survive the pre- (see SOIL BACTERIA and SOIL FUNGI: NATURE’S NUTRITIONAL
dation pressures, which exist in the larger soil pore spaces. NETWORK, this Encyclopedia, for more information on these
This interesting ecological hierarchy has led to the notion soil organisms).
of protected and nonprotected soil microorganisms (1). Because they represent the majority of the soil micro-
The concept has been widely used in simulation efforts bial biomass, bacteria, and fungi are considered responsi-
to describe microbial activity in soil. It is believed that ble for the majority of the energy flow and nutrient cycling
K-selected or protected organisms live mainly dormant that occur in soil. However, the faunal component of the
existences until exposed to a new food supply. Predation soil biomass is responsible for influencing the size of the
pressure, substrate availability, and specialized habitats soil microbial biomass through predation thus influencing
lead to selection of microorganisms with specific functions their ability to process plant litter and soil organic matter.
creating a complex nutrient cycle for all of the essential Soil fauna also increases microbial substrate availability
elements in soil. by physically burying plant litter in the soil. Larger soil
fauna are also responsible for reducing the size of plant
THE COMPOSITION OF THE SOIL MICROBIAL BIOMASS litter thus increasing its surface area, making it more
accessible to the soil microbial biomass.
The soil microbial biomass is composed of a large number The functional diversity of the soil microbial biomass
species that vary widely in their function. The soil is required to be able to take advantage of the wide array
microbial biomass is generally considered to be composed of plant materials and habitats found in the soil. The soil
primarily of bacteria and fungi. Bacteria and fungi represents an oligotrophic environment, becoming more
normally compose over 70% of the total soil biomass nutritionally limited as the depth of soil increases. In the
(Table 1). Faunal components make up a significant soil surface, soil aggregates also represent oligotrophic
portion of the total soil biomass and also contribute to environments. For example, the interior of an aggregate
carbon cycling and other biogeochemical processes. The may have extremes in pH, aeration, and redox potential.
importance of the microbial and faunal components of the The soil has a wide variety of these niches, sometimes
soil biomass to soil processes can often lead to a contentious called hot spots, producing an array of microhabitats,
 BIOMASS: SOIL MICROBIAL BIOMASS 665

which maintain the immense diversity of the soil microbial climates the landscape often contains areas with frequent
biomass. or sustained high water tables leading to lower aeration
and thus a decrease in the activity of microbial biomass.
The rhizosphere effect or plant root effect dramati-
DISTRIBUTION OF THE SOIL MICROBIAL BIOMASS
cally affects the number of microorganisms distributed
The soil microbial biomass requires energy, nutrients, and throughout the soil profile. In the vicinity of plant roots,
habitat to exist. The soil horizon containing the most high microbial activity exists due to increased deposition
organic matter, nutrients, and plant influence normally of compounds, such as amino acids and carbohydrates.
contains the most soil microbial biomass. Soils are The rhizosphere affect exists wherever plant roots explore
typically composed of distinct layers or horizons formed different soil horizons. The rhizosphere effect is tempo-
from the depositional, eluvial, and illuvial processes. The rary and is maintained as long as the plant root is alive
surface soil horizon, called the A horizon, is generally rich and immediately following its death during decomposition
processes.
in organic matter as a result of plant litter deposition
and root turnover. Eluvial and illuvial processes and plant
root exploration deposits organic matter and nutrients METHODS TO ASSESS MICROBIAL BIOMASS IN SOIL
in deeper soil horizons. Deep sediments and buried soils
can also contain significant amounts of organic matter, The size of the soil microbial biomass is primarily
however the bulk of the organic matter is usually located dependent on soil type and ecosystems productivity.
in the A horizon. Beneath the A horizon is the B horizon, Finer textured soils and soils with more silt and clay
which often contains appreciable amounts of eluviated tend to have a larger soil microbial biomass. Finer
clay or other amorphous minerals, depending on soil textured soils have more surface area and structure
age. Beneath the B horizon is the C horizon containing leading to greater number of niches for the soil microbial
unweathered parent material. Having additional or not biomass to inhabit. Plant detrital input from aboveground
well-defined horizons can often complicate soil horizon production and root turnover regulates the size of the soil
determination. The A horizon, because of the large organic microbial biomass through substrate (food) availability.
matter input from the above- and belowground plant Other factors, such as soil temperature and moisture,
production, forms a friable structure because of a wider also regulate the soil microbial biomass but manifest
range of aggregates sizes promoting good aeration and themselves more in regulating microbial activity and
moisture holding capacity. For these reasons, the greatest turnover. The following discussion will focus on methods
number of microorganisms exist in the A horizon (Fig. 2). to determine soil microbial biomass carbon and nitrogen.
There are a number of exceptions in which the number These two elements have been examined extensively
of microorganisms can be influenced by soil characteristics especially in their relationship to nutrient cycling in soil.
such as the type of mineralogy, ecosystem, and water A multitude of other methods exist to describe specific
regime. A young or highly weathered old soil may influence microbial components and are described elsewhere in
the number of microorganisms because of low organic QUANTIFICATION OF MICROBIAL BIOMASS, this Encyclopedia)
matter content. An example of this type of soil is a
tropical soil, which is highly weathered and contains Collection of Soil Samples
most of its organic matter in a quickly decomposed litter The appropriate selection of soil samples dictates the kind
layer. Conifers forests and bogs, often produce acidic soil of information needed to assess soil microbial biomass
affecting the number of microorganisms in the upper soil size and activity. Soils are not uniform and often vary
layers. Extreme climatic conditions produce soils with dramatically on scales of less than a meter. Table 1
prolonged dry (arid) or wet (aquic) regimes. Arid soils can lists the physical, chemical, and biological factors, which
accumulate salt and/or sometimes sodium causing high influence the distribution in size of the soil microbial
electrolyte potential in the soil solution thus affecting the biomass.
number of microorganisms. On the other hand, in wet One of the most important considerations when taking
soil samples is the representativeness of the sample.
Nontypical areas, such as low spots, steep slopes, eroded
Number of soil microrganisms per gram soil areas and so on should be avoided if they represent only
Soil
106 107 108 109 horizon a small fraction of the landscape of interest. Statistically
0 speaking, soil sample collection should strive for reducing
A
the error of the measurement to less than 20% and
Soil depth (meters)

0.5
B
preferably 10% of the sample population mean. Depending
on the uniformity of a landscape, 4 to 10 samples
1 C are required to determine statistical significance when
comparing treatments or other experimental variables.
Compositing a number of soil samples to produce up to
1.5
five individual samples may often reduce the error of the
measurement. Heterogeneous landscapes may require a
2 form of geostatistics called ‘‘Kriging’’ to determine the
dependence of soil properties on microbial biomass size or
Figure 2. The carbon and nitrogen cycle in soil. activity.
666 BIOMASS: SOIL MICROBIAL BIOMASS

Soil samples taken under field conditions should be Table 2. Soil, Environmental, and Organismal Factors
stored at 4 ° C until analyzed. If samples cannot be Affecting the Distribution and Activity of the Soil
immediately transported to the laboratory, they should Microbial Biomass
be stored at field soil temperatures taking care to avoid Environmental Organismal
exposure to direct sunlight. Soils samples should be Soil Factor Factor Factor
analyzed within one week after sampling and preferably
in less than three days (6). Mineralogy Rainfall Plant cover
Parent material Temperature Net primary
production
Microscopy
Soil age Rain shadow Vegetation history
The oldest method used to examine soil microbial biomass Topography Exposure (north Animal (grazing)
size is microscopic examination. Dutch merchant A. Van vs. south)
Leeuwenhock first saw microorganisms through a micro- Soil pH Elevation Human influence
scope in the seventeenth century. The examination of Water-holding Landscape Animal
capacity depression (burrowing)
a dispersed soil suspension is the most common proce-
Water infiltration Water springs
dure for counting soil microorganisms with a microscope.
Particle size(sand, History of burning
The advent of fluorescence microscopy utilizing ultraviolet silt, and clay)
light combined with flurogenic vital stains, which fluo- Bulk density (g/cc) Cultivation
resce under ultraviolet light enables the observation of Soil organic matter Riparian influence
living microorganisms in a soil dispersion preparation (7). content
Many stains have been developed to assist in identifying Fertility Pollution
microorganisms in the dark background of a soil disper- Erosion
sion. Dyes that bind DNA (acridine orange) sulfhydryl
groups in protein (5-(4,6-dichlorotriazin-2-yl) aminofluo-
rescein) and helical DNA (4,6-diamidino-2-phenylindole-
Chloroform Fumigation Incubation Method to Estimate Soil
2-HCl) are available. Other dyes can probe metabolic
Microbial Biomass
activities, such as electron transfer, with the reduction
of sensitive dye tetrazolium chloride. Immunofluorescent A number of chemical techniques have been used to
techniques using specific antigens are also available to estimate soil microbial biomass and activity. The most
count specific bacteria or fungal groups or species. The popular approach to estimate soil microbial biomass
use of dyes and microscopic counting permits counting of is the use of chloroform vapor followed by incubation.
both bacteria and fungi to estimate their mass in soil. Other methods utilizing chloroform with direct extraction
Many of the methods to determine microbial biomass size techniques and metabolic approaches have also been
and carbon and nitrogen content are calibrated using the utilized (6). Exposing the soil microbial biomass to
microscopic counting techniques. chloroform vapor dissolves lipids in their cell walls
The volume of carbon and nitrogen and their content allowing the cytoplasmic constituents to leak into the
in the soil microbial biomass can be estimated using soil. For the determination of microbial carbon and
microscopic counts of bacteria and fungal hyphal length. nitrogen, the soil can be incubated for 10 days following
Bacteria and fungal volumes and mass can be derived exposure to chloroform vapor and the amount of carbon
using average cell lengths and diameter. The number dioxide and ammonium determined (8). Because not
of microorganisms and their approximate carbon and all microorganisms are killed during the chloroform-
nitrogen content can be determined as shown in the fumigation procedure, the surviving biomass quickly
formulas presented in Table 2. consumes the labile components of the dead biomass
The carbon content of bacteria and fungi is assumed mineralizing both carbon and ammonium. The fraction
to be constant using the factors shown in Table 2. The of the mineralized carbon and ammonium is used to
nitrogen content of bacteria and fungi can be estimated estimate the original standing microbial biomass (9,10).
assuming a carbon to nitrogen ratio of six to one for Soil microbial biomass carbon is calculated as follows:
bacteria and eight to one for fungal biomass. The carbon
to nitrogen ratio for bacteria and fungi were developed FC − UFC
BC =
from pure culture studies. The carbon to nitrogen ratio 0.41
under field conditions may vary substantially depending
on substrate availability and species, leading to some Where BC is soil microbial biomass carbon, FC is the
uncertainty in the exact estimation of microbial carbon carbon dioxide produced from the chloroform treated
and nitrogen. Besides estimating true carbon and nitrogen soils after 10 days of incubation and UFC is the carbon
contents, other limitations of microscopic counting include dioxide produced in an untreated control sample during a
the tedious nature of counting microorganisms and soil period of 10 days following a preincubation of 10 days (8).
colloidal interference. These shortcomings often lead The proportion (0.41) used to convert the measured
to nonstandard approaches among users of microscopic carbon dioxide into standing microbial biomass carbon
counting techniques. Chemical analysis of soil microbial was determined using radioactively labeled microbial
biomass carbon and nitrogen can avoid some of these biomass (9). The use of the 10- to 20-day control often
problems; however, it must be stressed that these methods leads to small or negative microbial biomass carbon
are calibrated with the microscopic counting technique. values leading to the following equation, which subtracts
 BIOMASS: SOIL MICROBIAL BIOMASS 667

a partial control: in its ability to store nutrients and alter soil properties,
which affects net ecosystem productivity. In addition, the
(FC − FC K1 ) − (UFC K2 ) large store of soil carbon acts as a buffer in regulating
BC =
0.41 atmospheric carbon dioxide. Manipulating the soil carbon
pool through increased storage of organic matter is thought
Values for K1 and K2 are 0.29 and 0.23, respectively. to be a viable soil management approach to mitigate rising
The use of this formula strongly correlated to soil levels of atmospheric carbon dioxide and associated global
microbial biomass estimations obtained by microscopic warming effects. Therefore, the importance of the soil
counting (11). The preceding equation corrects for control microbial biomass cannot be understated in its ability to
soils having high background carbon dioxide production influence ecosystem carbon and nutrient budgets.
rates. These types of soils often have high organic matter
content or recently added plant residues. Soil Carbon Turnover
The determination of soil microbial biomass nitrogen is
calculated with the following equation: Carbon inputs to soil come primarily from plant produc-
tion. The heterotrophic soil microbial biomass and fauna
FN − UFN consume plant production according to the following:
BN =
0.54
CH2 O + O2 → CO2 + H2 O + energy
Where BN is soil microbial biomass N, FN is the ammonium
mineralized in the chloroform treated soils and UFN is the The oxidation of carbohydrates and other plant carbon
ammonium mineralized in an untreated control sample provides the soil microbial biomass with an energy source
incubated for 10 days in conjunction with the chloroform- for metabolic activity and growth. The effect of soil and
fumigated sample. The proportion (0.54) used to convert crop management can greatly impact standing soil carbon
the measured ammonium into standing microbial biomass pools. For example, soil tillage and reduced plant carbon
nitrogen was proposed for samples with a FC to FN ratio input under conventional agriculture is suspected to have
of <6.7 (9). As with microbial carbon, the use of the 0- led to a 50% decrease in soil carbon over a 50-year
to 10-day control often leads to small or sometimes large period (14). Soil tillage is thought to affect soil microbial
microbial biomass nitrogen values, which can be corrected biomass activity through aeration and liberation of soil
using the following equation: carbon from protected areas, such as within aggregates.


  For these reasons, it is thought that practices such
FN 1 − KC as no-till and off-season cover crops can be used as
BN = BC KC +
FC L management techniques to store more soil carbon and
to mitigate rising atmospheric carbon dioxide levels.
BC is soil microbial biomass carbon, FC is carbon These management techniques may make it is possible
dioxide produced in the chloroform treated soils, KC to manipulate soil carbon storage through management of
is the proportion of microbial carbon mineralized after heterotrophic microbial biomass size and activity leading
fumigation and L is the carbon to nitrogen ratio of the to the formation of stable soil carbon.
post chloroform-fumigated biomass (12). The chloroform The turnover of soil organic carbon is often related to
fumigation of soils to estimate soil microbial biomass ecosystem productivity. Soil carbon normally turns over
carbon and nitrogen has been used extensively to rapidly in productive ecosystems. The rapid turnover of
determine microbial biomass size and processes associated carbon can often be attributed to higher temperatures
with nutrient cycling. and/or nutrient availability. In tropical rainforests, the
soil carbon pool is smaller compared with temperate
forests or grasslands yet ecosystem productivity is four
CARBON AND NITROGEN CYCLING
times higher in the tropical forest as a result of
the rapid turnover of nutrients from plant litter (5).
The soil microbial biomass is the ‘‘waste management’’
Temperate grasslands store more soil carbon compared
crew of an ecosystem. The enormous amount of plant
with other ecosystems in temperate and tropical biomes.
detrital material produced annually is processed by the
In grasslands, the increase in soil carbon storage can be
soil microbial biomass and other faunal components of
attributed to large belowground productivity leading to a
soil. The processes that degrade and consumes plant
rhizosphere effect that deposits more soil carbon leading
litter, both leaves and roots, play an important role
to increased microbial biomass production and activity.
in maintaining the global carbon budget by balancing
Other ecosystems that can store more soil carbon than
the carbon dioxide fixed through photosynthesis and
grasslands are peat bogs, tundra, and boreal forest. The
released through decomposition activities. A by-product
storage of soil carbon in these ecosystems can often be
of the consumption and respiration of plant litter is the
attributed to anaerobic conditions and low temperature.
production of waste materials. These waste products,
consisting of unpalatable plant litter components and
Rates of Soil Carbon Turnover
microbial products, form soil organic matter, a relatively
stable pool of soil carbon. Soil organic matter represents The rate of soil carbon turnover is intimately tied to
approximately two-thirds of the total terrestrial carbon the activity and turnover of the soil microbial biomass.
pool (13). The importance of this large carbon pool lies The fate of soil carbon has become an important issue
668 BIOMASS: SOIL MICROBIAL BIOMASS

in determining the role of terrestrial soil carbon in Table 3. Factors Used in Estimating the Carbon and
influencing the global carbon budget. Therefore, the Nitrogen Content of the Soil Microbial Biomass Using
importance of the soil carbon cycle and the factors affecting Microscopy
its stability are important global warming research topics.
The competitive and synergistic interactions among Bacterial Numbers
components of the soil microbial biomass may affect A v W
the fate of soil carbon by affecting the decomposer N = nf d
F V D
community composition and/or its activity. However, the N = number of bacteria per gram soil
inoculum potential or diversity of the soil microbial n = bacteria per field of view
biomass is assumed to be rarely limiting unless the soil A = smear of filter area
environment has been drastically disturbed (1). Examples F = counting field area
of soil disturbance include erosion, fire, and volcanic v = volume of sample applied to smear or filter
V = volume of dispersed soil
deposition, which either remove surface soil or bury it.
d = dilution factor
Other disturbance events that affect certain components of W = wet weight of soil
the soil microbial biomass such as pollution or flooding may D = dry weight of soil
have the same effect. Besides all the biological interactions
of soil, soil physical properties, particularly those that Bacterial Carbon Content
affect temperature and moisture, will greatly influence B = NUeSC (10−6 )
the activity of the soil microbial biomass. B = bacterial biomass carbon (microgram (µg) per gram soil)
The quantity and quality of plant litter will determine N = number of bacteria per gram soil
how carbon becomes metabolized and stabilized in soil. U = average bacterial volume (µm3 )r2 L; r =
Similarly, the potential to utilize and mineralize nitrogen bacterial radius, L=bacterial length
in plant components is linked to the metabolism of carbon. e = bacterial density (1.1 × 10−3 mm−3 )
The forms of plant litter entering the decomposition S = bacterial solids content (0.4 in soil)
stream are a complex mixture of products with various C = % bacterial carbon (0.45)
degrees of biodegradability. Plants are composed of
Fungal Biomass Carbon
various labile (cytoplasm) and structural (cell wall)
chemical components listed in Table 3. Each component F= π r2 LeSC (106 )
has an inherent degradability based on its chemical F = fungal biomass carbon (microgram (µg) per gram soil)
structure. For example, proteins are much easier to r = average hyphal radius (1.13µm)
degrade than polymeric substances such as cellulose and L = hyphal length cm/g soil
e = density in soil (1.1)
lignin. In addition, to the chemical recalcitrance of plant
S = solids content (0.3 in soil)
components, the nitrogen content of each component is
also a good indicator of its biogradability. The lower the Source: (E. A. Paul and F. E. Clark, Soil Microbiology and Biochemistry,
carbon to nitrogen ratio of the plant material the more Academic Press, New York, 1996.)
likely it is easier to degrade. The carbon to nitrogen
ratio or the nitrogen to lignin ratio has often been used
to predict the degradability of plant litters. However, availability during decomposition. The soil microbial
the carbon and nitrogen ratio of specific plant litter biomass contains significant amounts of nitrogen often
components is not a reliable index of decomposition (15). equivalent to the entire needs of plants during one growing
For example, because lignin contains no nitrogen it season. The soil microbial biomass nitrogen content and
would be very difficult to degrade without an external that nitrogen becoming available during decomposition of
nitrogen source. Therefore, nitrogen availability during plant residues and organic matter are the primary sources
degradation ultimately controls the decomposition of plant of available soil nitrogen for plants. As stated earlier, for
litter. this reason the soil microbial biomass is often considered
the source and sink of nitrogen in soil.
Factors Controlling Nitrogen Availability During
The soil nitrogen cycle is complex and often the different
Decomposition
components of the nitrogen cycle compete for different
Nitrogen is an essential component of many cellular forms of nitrogen. The nitrogen cycle is a complex set of
building blocks including nucleotides, protein, and cell processes converting elemental nitrogen to organic forms
walls. The competition for nitrogen in soil by plants and the and back to elemental nitrogen or other gaseous inorganic
soil microbial biomass is fierce. The large surface area and forms of nitrogen. The forms and availability of plant
distribution of the soil microbial biomass compared with residue and soil organic carbon influence the formation of
plant roots often gives them an advantage in competing for different nitrogen forms in the nitrogen cycle (Fig. 3).
soil nitrogen. In general, plant roots explore approximately Nitrogen mineralization is the primary nitrogen cycling
1% of the total surface area of the soil. Because the soil process determining the availability of nitrogen for plants
microbial biomass is distributed widely in soil, it has and decomposition processes. The availability of organic
a competitive advantage over plant roots in capturing nitrogen to be mineralized to ammonium is the limiting
nutrients in the soil solution moving by both mass flow process that controls nitrogen available to plants and
and diffusion. The competition for nitrogen in soil among the soil microbial biomass. Sources of organic nitrogen
plants and the soil microbial biomass controls nitrogen include soil organic matter, plant litter residues, and
 BIOMASS: SOIL MICROBIAL BIOMASS 669

CO2 Table 4. Mass of Cytoplasmic and Cell

N2 fixation
C flow Wall Components in Plants
Plant uptake
N flow
Plant Component % of Total
C pool
Waxes and pigment 1
Amino acids, sugars, 5
N pool nucleotides and so on
Plant residues
Protein 5–7
Denitrification Hemicellulose 15–20
CO2 Cellulose 4–50
Volatilization
Lignin 8–20
CO2
−
NO3
Soil biomass

CO2
NH4+
Metabolites
CO2 Low carbon to nitrogen ratio
Mineralized carbon
Active organic
matter 15 parts C
CO2 Plant residue
Leaching
20 parts corbon Decomposition
1 part N
5 parts carbon
Stable organic Soil microbial biomass
matter
1 part N

CO2 Old organic

matter
CO2

High C to N ratio
Figure 3 Mineralized carbon
30 parts carbon
Plant residue
40 parts carbon Decomposition
1 part N
10 parts carbon
root exudation. Because extra nitrogen is often required Soil microbial biomass
for the decomposition of plant residues, other sources of 1 part nitrogen + 1 part nitrogen
nitrogen primarily from soil organic matter are utilized for from soil organic matter
the decomposition process. The amount of nitrogen made Figure 4. The influence of carbon to nitrogen ratio on the
available from the mineralization of soil organic matter decomposition of plant litter. The scenarios assume that the soil
for degrading plant residues depends on the amount of microbial biomass will utilize 25% of the plant carbon and produce
nitrogen already present in the plant residue. When the additional biomass at a carbon to nitrogen ratio of 5 to 1. Under
plant litter has a carbon to nitrogen ratio of 20 to 25 low carbon to nitrogen ratio plant litter, 15 parts of carbon are
to one the decomposition process has sufficient nitrogen mineralized leaving five parts of carbon and one part of nitrogen,
and no nitrogen is required from the soil organic matter thus no additional soil nitrogen is required. Under high carbon to
or other sources (Fig. 4). Plant litter carbon to nitrogen nitrogen ratio, 10 parts of carbon and one part of nitrogen are used
to produce new biomass thus requiring an additional one part of
ratios above 25 to 1 require additional nitrogen from other
nitrogen from the soil. This is termed nitrogen immobilization.
sources to complete the degradation process leading to Nitrogen mineralization would occur when the carbon to nitrogen
nitrogen immobilization in the soil. Carbon to nitrogen ratio is less than 20 to 1.
ratios of less than 20 to 1 often lead to mineralization of
nitrogen to the soil from the plant litter. Although Figure 4
is simplistic it does show how the carbon to nitrogen ratio
control decomposition process in soil and the availability plant residue components disappear during decomposi-
of nitrogen for plant and microbial uptake. tion and microbial production increases. As decomposition
proceeds more microbial production occurs. Microbial pro-
duction is the living soil microbial biomass and metabolic
Conversion of Plant Carbon into Soil Organic Matter
and turnover products that are difficult (recalcitrant) to
During the decomposition process, plant residues are process any further. The metabolic products contain car-
converted into microbial biomass energy, microbial pro- bon and nitrogen from plant, microbial, and faunal origin.
duction, carbon dioxide, and water. Figure 5 shows how The recalcitrant products are probably difficult to degrade
670 BIOMINERALIZATION BY BACTERIA

100 2. J. L. Smith and E. A. Paul, in J. M. Bollag and G. Stotzky,

eds., Soil Biochemistry, vol. 6 Marcel Dekker, New York,
pp. 357–396, 1990.
80
% C total remaining

Total C
3. R. M. Maier, I. L. Pepper, and C. P. Gerba, Environmental
60 Microbiology, Academic Press, New York, 2000.
Cellulose Straw C 4. J. H. Andrews, Current Perspectives in Microbial Ecology,
40 Plant American Society of Microbiology, Washington, D.C., 1984.
labile 5. K. Killham, Soil Ecology, Cambridge University Press, New
Microbial production York, 1994.
20
Lignin 6. W. R. Horwath and E. A. Paul, in R. W. Weaver, S. Angle,
0 and P. Bottomley, eds., Biochemical and Microbiological
0 10 20 30 40 50 Properties of Soil, Soil Science Society of America, Madison,
Wis., 1994, pp. 753–773.
Time (days)
7. P. J. Bottemley, in R. W. Weaver, S. Angle, and P. Bottomley,
Figure 5. Loss of total carbon and plant components during
eds., Biochemical and Microbiological Properties of Soil, Soil
decomposition. Microbial production increases during the decom-
Science Society of America, Madison, Wis., 1994, pp. 81–106.
position process leading to the production of soil microbial biomass
and soil organic matter (11). 8. D. S. Jenkinson and D. S. Powlson, Soil Biol. Biochem. 8,
209–213 (1976).
9. J. P. E. Anderson and K. H. Domsch, Soil Biol. Biochem. 10,
or they are protected from degradation by their association 207–213 (1978).
with soil mineral colloids. The formation of soil aggregates 10. D. S. Jenkinson, in J. R. Wilson, ed., Advances in Nitrogen
is an example of a soil structure containing protected Cycling in Agricultural Systems, C.A.B. International,
carbon. Wallingford, Conn., pp. 368–386.
The microbial production and resistant plant litter 11. W. R. Horwath et al., Can. J. Soil Sci. 76, 459–467 (1996).
components are the source of soil organic matter. Soil 12. D. Harris, R. P. Voroney, and E. A. Paul, Can. J. Soil Sci. 77,
organic matter formation is the result of life in soil. The 507–514 (1997).
formation and maintenance of soil organic matter creates 13. H. Eswaran, E. Van den Berg, and P. Reich, Soil Sci. Soc.
habitats through the formation of soil structure and Am. J. 57, 192–194 (1993).
provides a source of nutrients for plants and decomposition 14. R. Lal, J. M. Kimble, R. F. Follett, and C. V. Cole, The
processes. The formation of soil organic matter is a Potential of U.S. Cropland to Sequester Carbon and Mitigate
beginning and an end in itself because it creates a the Greenhouse Effect, Ann Arbor Press, Chelsea, Mich., 1998.
favorable habitat for both the soil microbial biomass and 15. W. R. Horwath and L. F. Elliott, Biol. Fertil. Soils 22, 1–9
plants. (1996).
16. K. A. Vogt, C. C. Grier, C. E. Meier, and M. R. Keyes, Ecol.
CONCLUSION Monogr. 63, 370–380 (1982).
17. D. C. Coleman and D. A. Crossley, Fundamentals of Soil
The soil microbial biomass plays an important role in Ecology, Academic Press, San Diego, Calif., 1996.
ecosystem function primarily through the regulation of 18. J. G. Holt et al., Bergey’s Manual of Determinative Bacteriol-
nutrient cycles. In their search for energy to grow, the soil ogy, 9th ed., Williams & Wilkins, Baltimore, Md., 1994.
microbial biomass relies on nutrients in plant litter and 19. D. Redecker, R. Kodner, and L. E. Graham, Science 289,
soil. Because of its competitive ability, the soil microbial 1920–1921 (2000).
biomass is often considered the source and sink of essential
plant nutrients. As result of the heterotrophic activities,
the soil microbial biomass greatly influences the terrestrial
BIOMINERALIZATION BY BACTERIA
carbon cycle. The disturbance of ecosystems leads to net
carbon losses to the atmosphere by exposing protected
TERRY J. BEVERIDGE
soil carbon and affecting the activity and turnover of
ANTON KORENEVSKY
the soil microbial biomass accentuating the greenhouse
SUSAN GLASAUER
effect. Future management of ecosystems will need to
University of Guelph
consider the role of the microbial biomass in sequestering
Guelph, Ontario, Canada
and producing stable soil carbon. The wide diversity and
function of the soil microbial biomass exemplifies its
role in a wide variety of processes including parasitism, Biomineralization is the process by which organisms form
pathogenesis, and symbiosis. This wide functionality has minerals; this ability is widespread among all taxa (1).
made the study of soil microbial biomass problematic The diversity of biogenic minerals is remarkable: more
and often leads to misinterpretation of its importance than 60 different minerals have been identified and
in ecosystem function. the number continues to increase (1,2). Bacteria produce
practically all types of biominerals including amorphous
BIBLIOGRAPHY and crystalline varieties (e.g., carbonates, phosphates,
silicates, sulfates, sulfides, oxalates, oxides, hydroxides,
1. E. A. Paul and F. E. Clark, Soil Microbiology and Biochem- etc.). Moreover, because bacteria can inhabit extreme
istry, Academic Press, New York, 1996. environments enriched with heavy metals, minerals
 BIOMINERALIZATION BY BACTERIA 671

uncommon to eukaryotes are formed from elements such thermodynamically possible, certain energy barriers must
as Ag, As, Au, Cr, Me, Sb, Tc, Ti, V, and U. Microorganisms be overcome. The level of supersaturation must be
play principal roles in biogeochemical cycling and form high enough to promote spontaneous crystallization
some biominerals on a huge scale. Many sediments, (homogenous nucleation) or centers of crystallization must
sedimentary rocks, and ore deposits are formed by the be introduced (heterogeneous nucleation). Particulate
direct participation of microorganisms. surfaces make excellent crystallization centers because
their interfaces reduce the free energy necessary for
mineralization at lower ion concentrations.
MECHANISMS OF BIOMINERALIZATION
For this reason, bacteria, with their high surface area-
All biomineral-forming processes can be divided into to-volume ratio, make exceedingly good foci for mineral
two types, ‘‘biologically induced’’ and ‘‘biologically con- nucleation; their highly reactive surfaces are good tem-
trolled’’ (3,4). When an organism influences solution chem- plates to initiate and promote further mineral growth
istry and, thus, creates favorable conditions for mineral (Fig. 2). These surfaces consist of biopolymers (lipopolysac-
formation, it is biologically induced mineralization (BIM). charides, phospholipids, peptidoglycan, polysaccharides,
Bacteria release different types of metabolites or ions, proteins, etc.; see Reference 8 for more structural details)
which change the oxidation state of metals and intro- bearing important ionizable functional groups at physio-
duce charged surfaces (cell walls, capsules, etc.) that can logical pH (e.g., amine, amide, carboxyl, carbonyl, phos-
induce mineral formation. The minerals produced, if they phate, sulfhydryl, and hydroxyl groups). However, the
are crystalline, have a wide range of sizes and are usu- most abundant are carboxyl, hydroxyl, phosphate, and
ally randomly aggregated (1). For biologically controlled amino groups (9). The ratio of acidic to basic groups in
mineralization (BCM), the mineral nucleation site is sepa- these biopolymers is generally greater than 1 and, overall,
rated from the external environment. Bilayer lipid vesicles bacterial surfaces have an isoelectric point of 2 : 3. Above
(e.g., magnetosomes; Fig. 1) and proteins (e.g., ferritin) this point, bacteria are usually negatively charged and
can serve as such sites and allow better regulation of the below it they are positively charged (10,11).
physicochemistry for mineralization (1). BIM is the pre- Because of this, the interaction of metal ions with bac-
dominant process among bacteria. Only a few examples terial surface polymers is strongly pH-dependent; pH not
of bacterial BCM are known; that is, the formation of only dictates metal ion speciation but also activates surface
magnetite (Fe3 O4 ) or greigite (Fe3 S4 ) in magnetotactic functional groups for metal ion–polymer complexation. All
bacteria (5,6), magnetite and ferrihydrite in ferritin (7), metallic ions, as well as the lanthanides and actinides,
and sulfide clusters in metalloproteins such as metalloth- have been implicated in these interactions (12,13). Gener-
ioneins. ally, the sorption of the metal cations increases with pH,
A necessary precondition for mineral formation is the whereas that of anionic forms decreases. The sorption of
development of supersaturation in the fluid phase. At Ag+ , Hg2+ and [AuCl4 ]− does not, however, change much
this point, ions begin to associate into small unstable over a wide range of pH (14).
clusters. The rate of cluster dissociation prevails, however, Metal sorption is facilitated by cation hydrolysis
over the rate of mineral growth. To exceed the critical (hydroxo-complex formation) or protonation of oxyanions
nucleus size, above which further crystal growth is and negatively charged complexes. If ion concentration
exceeds 10−3 –10−7 M, hydrolyzed and protonated metal
ion species will polymerize. Compared to ‘‘free’’ metal
ions these polymeric species are less hydrated and
more hydrophobic. This should reduce the amount of
free energy required for sorption, which appears to be
noncoulombic (mainly via hydrogen-bonding and van der
Waals cohesive forces). The amount sorbed exceeds the
number of potential functional groups (13,15–17).

Figure 1 Figure 2
672 BIOMINERALIZATION BY BACTERIA

The strength of the metal–polymer bond is important was impossible (29). Moreover, bacteria are able to release
because loosely bound metals can be more easily displaced. metal ions (Ca, Fe, and Mn) from insoluble forms thereby
However, these can also attract anions to form ion pairs promoting carbonate deposition of these metals. For
at the nucleation site. The anions at these sites can example, rhodochrosite (MnCO3 ) and siderite (FeCO3 )
coordinate with secondary cations so that more and more deposition in the oligotrophic Punnus Yarvi Lake (Karelia,
metal is deposited. This is called ionotrophy and both high- Russia) was attributed to the activity of manganese-
binding capacity and low-binding affinity are necessary for reducing bacteria (30).
this type of nucleation. Usually amorphous phases result, Cyanobacteria can produce significant carbonate
which are transformed (over time) through dehydration to deposits in carbonate-saturated waters (31). The giant
crystalline phases (18,19). Strongly bound metals are not towerlike microbialites (or bioherms) of Lake Van (Turkey)
easily displaced and will frequently react with bacterial are a dramatic example of recent thrombolite formation,
metabolites (e.g., HS− , HPO4 2− , and HCO3 − ), or undergo being up to 40 m high (32). Here, at sites where calcium-
hydrolysis, or redox transformations to form mineral rich groundwater seeps into the alkaline (pH∼9.7) soda
phases. lake water, inorganic calcite precipitation occurs. Micro-
In addition to the de novo development of fine-grained bial colonization of these crusts results in the formation
minerals by surface polymer–metal ion interaction, bacte- of 10–40 µm biofilms, which are dominated by small coc-
rial surfaces also sorb particulate, finely dispersed mineral coid cyanobacteria (Pleurocapsa spp.). Scanning electron
phases such as colloidal hydroxides and clays (20–22). microscopy (SEM) revealed that their gelatinous sheaths
These serve as new reactive surfaces for subsequent metal were heavily encrusted by aragonite that cemented and
ion interaction (21,23) but they do not bind the ions as stabilized the soft porous calcite crust. Because groundwa-
strongly as the biopolymers (24). ter is forced up and through this matrix, the tower grows.
When the towerhead rises 8–10 m in height, noncalcifying
Carbonate Minerals cyanobacteria (predominantly Oscillatoria) begin to domi-
nate the biofilm. This replacement makes further upward
Bacteria are active agents of carbonate mineral forma-
growth of the microbialites impossible (32).
tion because they release bicarbonate during respira-
Extensive studies by Schultze-Lam and coworkers and
tion while at the same time increasing pH in their
Thompson and coworkers on carbonate mineralization in
immediate surroundings (25). These factors, together
Fayetteville Green Lake provided unequivocal evidence
with the processes mentioned in the preceding section,
of this biogenic process (33–37). This lake contains
account for several biocarbonate formations. Hydromag-
Ca2+ -, Mg2+ - and SO4 2− -rich waters with a pH of 7.9.
nesite (Mg5 [CO3 ]4 [OH]2 4H2 O) and needlelike aragonite
Synechococcus GL24, a small cyanobacterium, dominates
(orthorhombic CaCO3 ) deposits have been found in asso-
the mixolimnion because of high oligotrophic conditions.
ciation with mucous biofilms in a karstic cave (Altamira,
This cyanobacterium can be planktonic or it can form
Spain). Such mineral crystals are usually embedded in
biofilms. In both cases, it contributes to an active
mucilaginous (producing hydromagnesite) or filamentous
biomineralization of calcium carbonate. As planktonic
(aragonite) microbial films. In this case, abiotic hydromag-
cells, it precipitates calcite so that a gentle but
nesite precipitation seems improbable because the cave
continuous rain of calcite contributes to the marl sediment
waters are oversaturated with respect to calcite, aragonite,
of the lake. When it grows as a biofilm, it also
and dolomite (CaMgCO3 ) but are strongly undersaturated
precipitates calcite, which contributes to the size of the
with respect to hydromagnesite and other magnesium
microbialites. Transmission electron microscopic (TEM)
carbonates (26).
analysis showed that the surface of GL24 consisted
The crystal type and mineral composition are obviously
of an S-layer of hexagonally arranged proteinaceous
dependent on solution chemistry and rate of precipitation.
subunits. Strategically placed carboxyl groups on the
Laboratory simulations with pure cultures of different bac-
S-layer interact with Ca2+ , thereby nucleating mineral
terial and cyanobacterial strains demonstrated that either
formation; the mineral initially mimics the large lattice
calcite or aragonite can develop, but not simultaneously.
structure of the S-layer (Fig. 3). Here, there is a
In all cases, calcite was precipitated in the viscous medium
remarkable dependence on pH for mineral type. Because
close to cells, whereas aragonite was precipitated in the
the solid field of the circumneutral pH lake waters is
fluid phase. This is in good agreement with observations
for gypsum, photosynthesis and respiration increase the
of bacterially mediated calcium carbonate deposition in
pH immediately surrounding the cells so that a calcite
natural environments (27,28).
solid field is produced. Laboratory studies in which Ca2+
Accordingly, slime-producing bacteria not only influ-
is replaced by Sr2+ , Mg2+ , or a mixture of the two has
ence the viscosity of their immediate surroundings but
revealed that strontianite, hydromagnesite, and celestite
also control the mineral composition of their precipitates.
can be formed (33,34).
Heterotrophic anaerobes frequently employ electron
acceptors such as sulfates, nitrates, ferric iron, and
Silicate Minerals
manganite compounds, resulting in an increase in pCO2
and pH. Deposition of a dolomite-ankerite (CaFeCO3 ) Bacteria play a role in the formation of silicate
phase was observed in the presence of sulfate-reducing minerals because their sorption of monosilicic acid at
bacteria (SRB; mostly Desulfovibrio) in Lagoa Vermella, low Si concentrations (∼0.5–3.0 mM) promotes silica
Brazil. Geochemical solid field analysis showed that under polymerization (autopolymerization of monosilicic acid
the given conditions, inorganic precipitation of dolomite occurs at concentrations greater than 3.5 mM). Moreover,
 BIOMINERALIZATION BY BACTERIA 673

Figure 3

silica mobilized from quartz by Bacillus mucilaginosus

or Thiobacillus thiooxidans was found to be mainly
bound to cells and their exopolysaccharides (38). There
are several ways in which silicate ions can bind to a
bacterial surface, that is, through ionic condensation
with phosphoryl, carboxyl, or hydroxyl groups (39), via
H-bonding with =O, −NH2 , −COOH, −OH, and other
groups (40,41), and by electrostatic interaction of silicate
anions with positively charged groups (e.g., amines; 42,43).
Covalent binding appears to have minor significance.
Another mechanism, adsorption of multivalent metal
cations on bacterial surfaces, results in reduced or
reversed surface electronegativity, and facilitates silica
sorption; silica-binding occurs through cationic bridging
by preexisting wall-bound metal cations or via binding to
surface hydroxyl groups of metal hydroxides adsorbed to
the cell surface (42–44).
Numerous electron microscopy (EM) studies of environ-
mental samples from river sediments, hot springs, mine
tailings, and so forth have revealed bacterial cells to be
partly or completely encrusted with different silicate min-
erals. Living cells have silicates associated with their Figure 4
surfaces or slimes, but when they die, silicates can be
found within the cytoplasm (Fig. 4). These silicates exhibit
a wide range of size (15 nm to >1 m), morphology, and
chemical composition, and greatly differ from suspended
minerals found in water samples. The mineral phases most
frequently associated with bacteria are complex polymor- 2 : 1 phyllosilicates have been detected as well (42,43). It
phic silicates ranging from amorphous silica or poorly seems that (Fe-Al) silicates, like other biominerals, mature
ordered chamositic clays [(Fe5 Al)(Si3 Al)O10 (OH)8 ] to well- over time from amorphous phases into crystalline forms
crystallized kaolinitic [Al4 (Si3 O10 )(OH)4 ] or glauconitic by dehydration during which more Fe and Si are added
[K(Al0.38 Fe1.28 Mg0.34 )(Si3.7 Al0.3 )O10 (OH)2 ] clays. Thus the (42–44,46,47,53).
most abundant elements of these minerals are Fe, Al, and
Si, although Mn, Mg, K, and Ti are often present (44–52). Mineralization Coupled to Redox Transformations
Interestingly, there seems to be no relationship Bacteria are able to oxidize or reduce metals such as
between different taxonomic groups of bacteria and their Fe, Mn, As, Cr, U, and Au via enzymes (oxidases
silicate mineral type (46,47) and this has been verified and reductases), metabolites (peroxides and thiols), inor-
by laboratory simulations using Bacillus subtilis cells ganic substances (H2 S), and cellular redox potential
to nucleate poorly ordered allophanes and serpentines. (25,54,55). This can impact biomineralization processes
However, more crystalline kaolinite phases and (possibly) because when a metal valency is changed, an insoluble
674 BIOMINERALIZATION BY BACTERIA

product can be formed by hydrolysis or anion inter-

action (e.g., with phosphates, sulfides, or carbonates).
For instance, Pseudomonas vanadiumreductans and P.
isatchenkovii can produce V2 O3 from vanadate ions (56),
Desulfovibrio desulfuricans and Escherichia coli deposit
low-valence hydroxides and oxides (e.g., TcO2 , Tc2 O5 ) from
Tc(VII) (57–59), and several bacteria, for example, Desul-
fovibrio spp., Geobacter metallireducens, and Shewanella
spp., reduce UO2 2+ to form insoluble uraninite (UO2 •nH2 O)
(60–63).
Reduction of Fe(III) (e.g., from iron oxides) by
dissimilatory iron-reducing bacteria serves as a good
example of complex interactions occurring between
the solution and a solid phase. For example, simple
solution chemistry dictates that as the concentration
of Fe(II) increases, siderite (FeCO3 ) and vivianite XXX (×102)
5
(Fe3 (PO4 )2 •8H2 O) phases should result. Yet, because
bacteria can modulate the Fe(II/III) couple as they grow 4
with an iron oxide, ferrous ions often interact with
amorphous ferric hydroxides resulting in the additional 3
Cu
formation of crystalline magnetite (Fe3 O4 ) and green
2 P
rust [FeII III x+ 2−
(6−x) Fex (OH)12 ] (A )x/2 •yH2 O (64–67). If Cr(VI)
Au
is also in the system, its reduction by bacteria leads 1
to mixed Cr(III)-Fe(oxy)hydroxide phases (Fig. 5) (68,69). Fe
Such systems show that the dominant mineral phase 0
2 4 6 8 10
depends on a complex interplay between bacterial Range (heV)
PAG1 Au
reduction processes, the aqueous geochemistry, and the
surface chemistry of an oxide (66,70). Figure 6
Metal ions with high electrochemical potentials
(>700 mV), such as those formed by Ag, Au, Hg, and
Pt, can be readily reduced by bacteria. The resulting
precipitates may grow on the cell surface and intracel-
lularly as metal colloids. The metal ions easily damage
the normal permeability of the plasma membrane; the
of Ag, Au, Hg, Pt, and so forth is a relatively quick
membrane is breached, killing the cell and allowing intra-
process, occurring within 15 minutes in the B. subtilis
cellular mineral growth (Fig. 6) (71–73). The biosorption
system of Beveridge and Murray (74,75). Interestingly,
once small metallic colloids are formed, they coalesce to
form macroscopic crystalline phases (pseudotrigonal and
hexagonal-octahedral forms) similar to placer gold (76).
Similar results have been reported by Russian researchers
Cu
using Bacillus cereus and Micrococcus luteus systems
Cr in which approximately 80% of the colloidal particles
annealed into 0.2–0.8-mm-sized crystals after 8 to 10
Fe
P K months (77). The M. luteus culture was isolated from a
O Cr
gold ore deposit and is capable of sorbing up to 45% of
its weight in gold, producing a crystal size similar to that
found in the deposit. For these systems, the production of
the gold crystals depended on the physiological activity of
growing cells because metabolically inactive and dead cells
sorbed the gold poorly (77–81). In field studies, bacterial
numbers coincide well with the number of gold particles in
oxic zones of gold deposits and in placers of Far East Rus-
sia (77). Sometimes gold deposits contain microfossils and
these prokaryotic remnants possess golden castes in the
shape of bacteria, suggesting that lithification developed
over living cells (82).
If protein sulfhydryl groups are available (−SH groups
Figure 5
are rare on bacterial surfaces but can be present in
the periplasm), they will dominate the binding of gold
ions, presumably via ion–ion interaction or complexation.
It is also possible that gold and its analogs can bind
 BIOMINERALIZATION BY BACTERIA 675

to oxygen- and nitrogen-containing functional groups of shown that some anaerobic phototrophic purple and
surface biopolymers (83,84). green bacteria (e.g., Rhodobacter, Rhodomicrobium, and
Chlorobium) are able to oxidize Fe2+ , FeCO3 , and FeS,
Oxidation of Iron and Manganese with concomitant reduction of carbon dioxide to produce
cell material (92,93).
These oxidation processes deposit Fe and Mn oxides,
have been found in gram-positive bacteria, gram-negative 2. Oxidation of Mn bound to a manganese(IV) oxide:
bacteria, and mycoplasmas, and are well studied. Fe–Mn-
oxidizing bacteria are widespread in nature, being found in Mn•MnO2 + 1/2O2 + H2 O ⇒ 2MnO2 + 2H+
marinewaters, freshwaters, and groundwaters, sediments,
and soils (85,86). The mechanism of Fe–Mn deposition Some bacteria (e.g., Arthrobacter, Vibrio, and Oceanospir-
involves both metabolic activity and sorption. As with illum) are unable to oxidize free Mn2+ but oxidize the
other metal ions, anionic polymers of capsules, sheaths, ion when it is bound to manganese(IV) oxides, ferroman-
and cell walls readily sorb ionic and colloidal forms of these ganese, or some clays (e.g., montmorillonite and kaolinite).
metals. For example, distribution coefficients (Kd values), For a more complete summary of this work, see Refer-
calculated as the ratio between sorbed Fe and Mn (on ence 25.
marine bacteria) and dissolved metal concentration, were 3. Oxidation catalyzed by catalase during hydrolysis of
as high as 107 (87). Metal accumulation on the bacterial metabolic H2 O2 :
surface structures is poised for subsequent oxidation and
mineralization. Mn2+ + H2 O2 ⇒ MnO2 + 2H+
Ferrous iron is unstable at pH higher than 5.5 because
of autooxidation, and it is difficult to distinguish between 2Fe2+ + 3H2 O2 ⇒ 2Fe(OH)3
abiotic and microbial precipitates formed during the oxi-
dation of iron. Microbial surfaces lower the free energy This mechanism is widespread in various heterotrophic
barrier for precipitation, but physiological processes seem bacteria (85,94,95).
to be necessary for large depositions. In this way although
nonliving cells of Leptothrix, Sphaerotilus, and Pedomi- In acidic environments, chemolithoautotrophic and
crobium accumulate iron via sorption at autooxidation mixotrophic bacteria play a principal role in Fe(II) oxi-
pHs (88,89), the amounts are low when compared to liv- dation by enzymatically increasing chemical oxidation as
ing cells because active oxidative processes occur at rates much as six orders of magnitude. These acidophilic bacte-
that are ∼8 to 10 times higher than for dead cells (85). ria can be mesophilic or thermophilic and usually belong
At neutral pH, Mn2+ is more stable than Fe2+ and the to the genera Thiobacillus, Leptospirillum, Sulfobacillus,
rate of its microbial oxidation can be ∼500 times higher and Sulfolobus. To obtain energy and reducing power for
than abiotic oxidation. The processes of Fe–Mn oxidation CO2 fixation they can use as electron donors not only
are complex and are complicated by the fact that different Fe(II) but also Cu(I), U(IV), Se(II), reduced sulfur species
bacteria can have distinctly different oxidative pathways; and sulfide minerals (96). It has recently been shown that
in some instances, oxidation occurs via enzymatic routes, acidophilic filamentous bacteria (similar to the Leptothrix-
whereas in others no enzymes are required. Sphaerotilus group) and acidophilic–heterotrophic uni-
cellular bacteria can oxidize Fe2+ and pyrite (96–98).
Nonenzymatic Reactions. During their metabolism, Bacterial Fe(II) oxidation results in extensive formation
bacteria will frequently excrete oxidants or bases while at of jarosites, (crystalline basic iron sulfates), amorphous
the same time consuming organic anions. These processes ferric hydroxides, ferrihydrite, lepidocrocite (γ -FeOOH),
will increase the pH and redox potential of the ambient and magnetite. In the case of arsenopyrite oxidation, the
solution, thereby promoting Fe–Mn autooxidation. It deposition of iron arsenate takes place as well (99,100).
has also been shown that hydrocarboxylic acids, acidic Depending on the conditions, Mn(II) oxidized by bac-
exopolymers, and some proteins produced by the cells teria can be deposited as manganese(IV) oxides or as
enhance this same reaction (25,86,90). mixed manganese(II,III)/manganese(IV) oxides and oxy-
Enzymatic Reactions. The enzymatic mechanisms of hydroxides. Thus in these precipitates the following min-
Fe–Mn oxidation can be separated into three major erals have been detected: vernadite, δ-MnO2 ; todorokite,
groups: [Ca,Na,K][Mg,Mn]Mn5 O12 •H2 O; buserite, [Ca,Na,K][Mg,
Mn]Mn6 O14 •5H2 O; feitknechtite, β-MnOOH; manganite,
1. Oxidation of free Mn–Fe ions by constitutive or γ -MnOOH; hausmannite, Mn3 O4 (101,102–104).
inducible oxidases that transfer electrons to oxygen Localization of deposited Fe and Mn usually occurs on
via cytochrome systems: bacterial surface structures such as cell walls, capsules,
sheaths and, in the case of mycoplasmas, on the plasma
Mn2+ + 1/2O2 + H2 O ⇒ MnO2 + 2H+ membrane. The most striking example is the sheath
of Leptothrix discophora. Here, manganese is oxidized
Fe2+ + 1/2O2 + 2H+ ⇒ Fe3+ + H2 O enzymatically on this outermost structure (105). The cells
depart from old sheaths leaving them totally encrusted
This pathway for Mn oxidation is found in Pseudomonas in manganese oxides (Fig. 7). Sometimes cells become so
sp. S-36 (91), Pseudomonas putida MnB1 (90), Citrobacter completely encrusted with metal oxides that the weight
freundii E4 , and Pseudomonas sp. E1 (25). It has been of the deposited mineral can exceed the cell weight
676 BIOMINERALIZATION BY BACTERIA

ancient sedimentary horizons for biogenic mineral for-

mation is harder to find. Sometimes, the most that can
be concluded is that biomineralization is the likeliest
explanation for what is observed, given the difficulties in
identifying remnants of bacteria in situ and relating their
activities to a specific mineral or geological structure. The
existence of an unusual mineral (in an environment in
which it would not normally form) is a first clue that
microbes have been active. Stronger evidence comes from
electron microscopy studies of more recent horizons where
visual observations have been combined with element
analysis (energy dispersive X-ray spectroscopy; EDS) and
selected area electron diffraction (SAED) (Fig. 6). These
allow precipitates associated with cells to be characterized,
and make it possible to identify those minerals around the
cell perimeter, which have been mentioned in our previous
sections. Isotope fractionation is another in situ method
that has been used to identify past geological activity
of microbes; it is based on the fact that life-forms (in
this case early microorganisms) prefer to metabolize the
lighter isotopes of elements such as C, H, O, N, S (25).
Radioisotopes can be used for in situ studies of present
microbial action (25).

Mineral Deposits

Iron Ores. Iron in ore deposits may be in the

form of oxides, sulfides, or carbonates, each of which
can be formed by bacteria. Clues that bacteria have
been active include the presence of fossilized or living
microbes possessing probable iron-metabolizing or iron-
accumulating potential, or the existence of environmental
conditions (at the time of formation) that enabled biogenic
iron deposition (25). An example of modern iron ore
deposits is found in some bogs overlying sediments
rich in ferrous minerals such as glauconite or pyrite.
Groundwater brings Fe2+ to the surface, where the rapid
Figure 7 rate of oxidation is attributed to biocatalysis (see previous
section on redox transformations); bacteria involved in
the oxidation include Leptothrix, Crenothrix, Siderocapsa,
Metallogenium, Thiobacillus, and Gallionella. Rivers and
swamps in the Pine Barrens in southern New Jersey
contain such oxidized iron deposits (108). The iron there
by an order of magnitude. However, deposition can was mined until it became economically unfavorable in
also take place in the surrounding medium, especially the mid-1850s.
when hydrogen peroxide is involved in the oxidation Banded iron formations are remarkable ancient strat-
reaction (25,86). ified deposits that were formed throughout the world in
Deposited Fe and Mn biominerals are finely dispersed the Precambrian era; most deposits formed around 2.2 to
and possess a high specific surface area. It has been 1.9 billion years ago. The bands occur as continuous layers
reported that the specific surface area of Leptothrix- of alternating iron-rich and iron-poor minerals, with the
oxidized Mn is about four times the specific surface area of total iron of a quality and quantity in some places to be a
fresh abiotic precipitates and more crystalline manganese valuable ore. The microbial origin of these deposits is still
oxides. Biogenic manganese oxides have a high affinity for controversial; one theory, based in part on fossil evidence,
trace metals; the adsorption of Pb is more than four times favors oxidation of ferrous iron by anaerobic photosynthe-
that of abiotic crystalline oxides (106). sizing bacteria and algae (109). Another theory considers
that bacteria oxidized carbon to produce the carbonate
GEOLOGICAL EVIDENCE OF BIM minerals characteristic of these formations, particulate
Fe(III) serving as the electron acceptor (110).
That bacteria can biomineralize under laboratory condi- The role of bacteria in forming ore deposits of
tions has been shown for many combinations of bacte- other metals is still debatable but the evidence points
ria and minerals (107); however, conclusive evidence in increasingly toward biogenesis. The formation of uranium
 BIOMINERALIZATION BY BACTERIA 677

ores is one example for which bacterial activity seems activity of bacteria dominates tailings, it is quite possible
likely. The ore bodies develop when mobile hexavalent that Thiobacilli and SRB continuously recycle the sul-
uranyl ions are reduced to insoluble tetravalent uranium fur in these minerals from soluble to insoluble forms. In
forms. A likely reductant is hydrogen sulfide, given the another example in the oxic zone of tailings derived from
close proximity of pyrite (FeS2 ) to tetravalent ore minerals, a copper mine in Quebec, Canada, diagenetic iron phos-
and (at the generally low temperatures of sedimentary phates and iron chlorides were detected associated with
ore-forming environments) bacteria are the most likely thiobacilli cell surfaces (76). The formation of minerals in
source of H2 S. Indeed, sulfate-reducing bacteria (SRB) the tailings piles shows the potential for the genesis of a
have been found in uranium ores in Hungary (111). mineral soil.
Further support for biogenic formation of ores comes from
laboratory studies showing that both gram-negative and Freshwater Lakes and Rivers
gram-positive bacteria can concentrate silver and gold as
Microbial populations in lakes tend to be much greater
particles in the cell wall (see previous section on redox
than those found in the open waters of seas; freshwater
transformations; 76,112).
systems have populations approximately 102 to 105
per milliliter of water and approximately 106 per
Coal, Peat, and Oil. These share steps in their
gram of sediment (25). The range reflects the wide
initial formation. Both coal and peat begin with large
variation in conditions that can be found in these
accumulations of biomatter that were most likely initially
environments, which are particularly susceptible to
degraded by bacteria and fungi. Microbial activity stopped
fluctuations in oxygen, temperature, and nutrient status.
once all of the easily metabolizable compounds and some of
As mentioned previously (in the carbonate section),
the more refractory substances were consumed, the latter
calcium mineralization at Fayetteville Green Lake,
probably by anaerobic bacteria (25). Microbial evidence
New York is mediated by Synechococcus GL24. Here
comes from remains of microorganisms identified in coal
the mineral phase alternates between gypsum and
and from signature isoprenoid substances found in ancient
calcite, each being nucleated by the S-layer on the
oils. Further physical and chemical processes, such as heat
cyanobacterium. Because photosynthesis and respiration
and pressure, converted peat to coal or oil.
determine the pH of each cell’s microenvironment,
Synechococcus precipitates gypsum during the cold winter
Mine Tailings and Their Leachates
months (poor growth and circumneutral pH promotes
Base metals (Cu, Zn, Fe, Ni, etc.) are generally mined gypsum) and calcite during the warmer summer season
from metal sulfide ores, most often pyrite. Exposure of (good growth and a more alkaline pH promotes calcite).
sulfidic rock to air begins a slow oxidation process that is For more details on this study, see Reference 113.
accelerated by native bacteria grouped together under the Biomineralization can be a way to neutralize the toxi-
term acidophilic chemolithotrophs. The oxidation process city of metals introduced to a lake. Lower Moose Lake
generates acidity and dissolves the metal, creating acid in Ontario, Canada receives mine tailings drainage con-
mine drainage in the process. This natural process is taining sulfate and other sulfur- and metal-containing
exploited by mining companies to harvest metals from compounds (44). Studies by TEM-EDS-SAED revealed
low-grade ores; however, higher-grade ores are extracted that bacteria in the lake sediment served as nucleation
by more efficient means and the finely ground waste sites for metal sulfides and (Fe,Al)-silicates. The forma-
rock is deposited in tailings dumps (25). This mineral tion of the minerals was explained by a combination of
waste represents an oxidizable energy source to the conditions including circumneutral pH, high metal con-
bacteria, as exemplified by the activity of Thiobacillus centrations, and low levels of sulfate reduction by bacteria,
ferrooxidans. These bacteria oxidize Fe2+ and sulfide, demonstrating again the specificity of the minerals to the
thereby generating Fe3+ , sulfuric acid, pH values as low as environment that forms them.
1–2, and solutions with high metal concentrations (113). In contrast to lakes, rivers tend to be uniform in
Drainage of acid leachates from the tailings represents temperature with depth and are usually well aerated.
some of the worst environmental problems worldwide. In addition, rivers vary in water velocity and turbulence.
Hence, there is increasing pressure to understand the As with lakes, the nutritional status of rivers is variable
ecology and microbiology of these environments. and depends on the environmental setting. Researchers
In addition to the Thiobacilli, researchers have studying two rivers in the Amazon region of Brazil
shown that SRB play a significant role in tailings found that bacteria collected from the solute-rich Rio
environments (107,114). These anaerobic, generally het- Solimões were encrusted with mineral precipitates,
erotrophic, bacteria require neutral to alkaline pH. Their whereas bacteria sampled from the solute-deficient Rio
numbers increase with depth in tailings deposits, although Negro were not mineralized (46). The bacteria grew as
they are also found in the oxic zone where Thiobacilli biofilms on plant and sediment surfaces in the river and
are active. The H2 S generated by SRB reacts strongly sequestered iron, as well as traces of Mn and Ni, in their
with metal ions and the resulting precipitates are found capsules, suggesting a selective accumulation of these
both on the cell surface and throughout the environment elements on the cell surface. Among the minerals found in
surrounding the cell. The diagenetic minerals associated these biofilms was a complex claylike (Fe,Al) silicate that
with SRB in Cu-Zn ore tailings from Kidd Creek Mine in has also been detected in other freshwater environments
Ontario, Canada, included amorphous FeS, mackinawite on other bacteria (115). Several characteristics point to
(FeS1−x ) and pyrite (114). Although the sulfate-oxidizing a biogenic origin for these minerals, that is, the size of
678 BIOMINERALIZATION BY BACTERIA

the precipitates (∼100 nm), their crystal structure, the carbonate-rich bands that have been cemented together
tendency to form a crust around bacteria, and their over time. In form and shape they are similar to present-
chemical composition (which differed significantly from day microbial mats. Because they appeared at a time when
that found in other suspended material in the water). As the earth’s atmosphere was reducing, it is thought that
discussed with respect to other minerals in the previous the organisms may have been lithoautotrophs or anaerobic
sections, an amorphous Al–Fe–Si phase forms, which photoautotrophs (120). These may have induced carbonate
dehydrates slowly and rearranges into a more crystalline precipitation by creating an alkaline environment around
form. In time, this process could form large deposits of the cell, although abiotic precipitation may also have
minerals. contributed. For more information about stromatolites,
see Reference 121.
Hydrothermal Environments
The term hydrothermal encompasses a variety of natural Magnetotactic Bacteria
terrestrial or marine environments that have exposure to The term magnetotactic bacteria has no taxonomic
water heated by geological activity. Often these regions significance and encompasses a diverse group of bacteria
have high concentrations of soluble silica and reduced that display magnetotaxis. They are found throughout
metals. Although these are hostile environments, the aquatic habitats, most often at the transition zone
metal-rich waters can provide energy — from reduced sul- between oxic and anoxic conditions (6). Two regions
fur compounds or metals — to bacteria and large microbial where magnetobacteria have been studied in detail are
populations may be found. Because silica and metals Pettaquamscutt Estuary (Rhode Island, U.S.A.), and
can embalm cells, preserving their natural arrange- Salt Pond (Massachusetts, U.S.A.). Researchers studying
ments in biogenic minerals, a fossil record is formed. magnetic properties within the water column at Salt
Because of the extreme conditions [high temperature, high Pond found that the peak in magnetic remanence (the
reduced metal concentrations, sulfide, salinity, and (some- magnetization recorded in the mineral) occurred in the
times) high pressure], these microorganisms often possess oxic-anoxic transition zone. At this depth a great number
unique adaptations (116). They frequently belong to the of bacteria and the highest available Fe2+ (needed to form
Archaea. Three examples of biomineralization in different magnetite or greigite) were found (6).
hydrothermal environments will be described — an ocean Magnetotactic bacteria form unique structures (mag-
vent, a hot spring, and sediments from a shallow water netosomes) that are aligned in rows to the cell axis
hydrothermal system. For a review of microbiology in and that contain membrane-bound single-domain mag-
hydrothermal environments, see References 116 and 117. netic particles of either magnetite or greigite (Fig. 1; 6).
Bacteria living near hydrothermal vents of the Magnetosomes can be preserved in sediments after these
Southern Explorer Ridge in the Pacific Ocean provide bacteria die and thus contribute to the paleomagnetic
nucleation sites for the growth of poorly ordered iron record if the concentration is sufficient. Some limitations
oxide, manganese oxide, and iron silicate particles, which are the detrital input of magnetic material, secondary for-
coat cell surfaces and cell-associated filaments (118). A mation of magnetic minerals, and mineral transformation
direct metabolic role for bacteria in mineral formation of magnetosomes at depth. For example, Snowball (1994)
at ocean vent sites has not been shown. Researchers found that the high concentration of biogenic magnetite
observed a similar preservation of bacterial cells in clay in the upper portion of a Swedish lake sediment declined
and mud sediments collected from an acidic hot spring with depth (122). Bacteria can mediate chemical reactions
at Yellowstone National Park (45). Gallionella ferruginea leading to dissolution of magnetite or they can dissolve
associated with hydrothermal brines in shallow waters the mineral directly. The discovery that a Mars mete-
around the Greek island of Santorini have a direct role orite (ALH84001) contained magnetite particles similar
in oxidizing Fe(III) bubbling up from vents, forming at in size and morphology to those produced by magnetotac-
least two meters of gellike sediments that contain poorly tic bacteria on earth has researchers speculating about
crystalline hydrous ferric oxide (HFO) precipitated around interplanetary life (124). Such a discovery, if true, would
bacterial stalks (119). Transformations to more crystalline considerably widen the geographic distribution of bacterial
phases such as goethite (iron oxyhydroxide), siderite, and mineral formation. For a detailed review of magnetotactic
pyrite take place at greater depths, and pyrite formation bacteria, see Reference 6.
is probably mediated by the activity of SRB.

Stromatolites SIGNIFICANCE OF BIOMINERALIZATION

Stromatolites are ancient microbiolites that have been
Iron Where It Is Not Wanted
preserved by the minerals fixed in their biofilm commu-
nities as these communities grew. They are interesting Oxidation of reduced iron by bacteria is a worldwide
because they record the presence of complex microbial nuisance. It results in plugged water pipes, wells,
ecosystems living up to 3.5 billion years ago during the bore holes and field drains, as well as contaminated
Middle Archean period (120). The oldest known forma- lakes and rivers. The slimy orange (although color
tions are the Onverwacht group in South Africa and the and texture may vary) precipitates, consisting of iron
Warawoona group in Western Australia. oxyhydroxide or HFO, can be produced in copious amounts.
In cross section, stromatolites are finely laminated, Although the significance of Fe-oxidizing microbes in
with darker organic-rich cherts alternating with lighter producing the deposits is controversial, researchers have
 BIOMINERALIZATION BY BACTERIA 679

shown that both autotrophic and heterotrophic bacteria Broad Implications

are responsible; the species depends on environmental
conditions. G. ferruginea is found at neutral pH and low Biomineralization is a remarkably young science and
dissolved O2 concentration. The characteristic twisted has taken great strides in maturity over the last two
stalks of Gallionella, which extend from one side to three decades. Who would have believed that the
of the bean-shaped cells, become coated with HFO. world’s smallest life forms could have such a great impact
Thiobacillus ferrooxidans is an example of a Fe-oxidizing on the earth’s sedimentary fine-grained minerals? It is
chemoautotroph that also grows at the oxic–anoxic clear that one of the fundamental traits of microbial
interface, although it prefers pH ∼1 to 4. The bacterium, cells is to interact with the metal ions that surround
as previously described, produces acidity and ferric iron them and, in so doing, profoundly affect geochemical
from pyritic ores. This iron leaches from the tailings to conditions. Eh, pH, metal speciation, and interfacial
precipitate when it encounters higher pH, such as in a river chemistry are all affected, and sometimes enzymes act
or lake. See References 115 and 124 for more information as catalysts for geochemical processes. It is imagined that
on Fe-oxidizing bacteria. since the dawn of life almost four Ga ago, prokaryotes
have been reworking the earth, immobilizing metals
Bioremediation with appropriate counter-ions and (presumably) even
altering the primitive recipe of the oceans. Global carbon,
Bioremediation strategies that use bacteria exploit the hydrogen, nitrogen, oxygen, sulfur, phosphorus, and (now)
affinity of bacteria for metals as well as their ability metal cycling are all tied, at least partially, to microbes
to reduce and oxidize metals. The latter requires living and their growth. With respect to mineralogy, bacteria
cells, whereas the former can use both living and have added their own particular flavor by instigating
dead biomass. Living systems include natural settings the production of a wide range of fine-grained minerals
in which existing landforms such as ponds or streams and by making possible the formation of some unusual
are used to remediate metal contamination. These may minerals under geochemically unfavorable conditions.
already be in place, but are often simulated and have Exploring deep subsurface and (even) extraterrestrial
some kind of containment layer (e.g., a mine tailings microbiology may broaden the significance of bacteria to
leachate pond). In this case, immobilization of the metals biomineralization.
is based on the interactions of an entire ecosystem,
and contamination levels must be low enough that the
microbes are not poisoned, although through natural BIBLIOGRAPHY
selection heavy metal-resistant microorganisms often
evolve. Over time, in aqueous systems, metals become 1. H. A. Lowenstam and S. Weiner, On Biomineralization,
part of the anaerobic sediments when the metal-loaded Oxford University Press, New York, 1989.
microorganisms die and sink to the bottom. Other kinds of 2. K. Simkiss and K. M. Wilbur, Biomineralization. Cell Bio-
living systems are those constructed to use organisms logy and Mineral Deposition, Academic Press, San Diego,
in an artificial environment for metals removal. One Calif., 1989.
example is publicly owned treatment plants, which may 3. H. A. Lowenstam, Science 211, 1126–1131 (1981).
incorporate a biological system to remove metals that 4. S. Mann, Struct. Bonding 54, 125–174 (1983).
remain after primary sedimentation; usually this is an 5. S. Spring and K.-H. Schleifer, System. Appl. Microbiol. 18,
activated sludge or trickling filter system. The metals are 147–153 (1995).
most often accumulated in the biopolymers produced by 6. D. A. Bazylinski and B. M. Moskowitz, in J. G. Banfield
floc-forming microorganisms (usually bacteria) and not and K. H. Nealson, eds., Geomicrobiology: Interactions
within the cells themselves (125). Another type of system Between Microbes and Minerals, Reviews in Mineralogy 35,
uses cells colonized and immobilized (as a biofilm) on Mineralogical Society of America, Washington, D.C., 1997,
support material with the metal-loaded water flowing past. pp. 181–218.
Good removals may be achieved with such systems, which 7. P. M. Harrison and T. H. Lilley, in T. M. Loehred, ed., Iron
can be tailored to immobilize specific metals (12). Carriers and Iron Proteins, VCH Publishers, New York,
Methods using sorption of metals by dead bacterial 1989, pp. 123–238.
biomass are viable technologies for scouring metals from 8. T. J. Beveridge, Annu. Rev. Microbiol. 43, 147–171 (1989).
wastewaters and recovering valuable metals from ore- 9. J. B. Fein, C. J. Daughney, N. Yee, and T. A. Davis, Geo-
processing (126). Systems using dead biomass have the chim. Cosmochim. Acta 61, 3319–3328 (1997).
advantage that they are not susceptible to transient 10. S. Aaronson, in Chemical Communication at the Microbial
highly toxic concentrations of metals or other conditions; Level, vol. 1, CRC Press, Boca Raton, Fla., 1981, pp. 55–56.
most waste streams are variable and can be toxic 11. A. M. James, Adv. Colloid Interface Sci 37, 266–273 (1982).
to living biomass. Furthermore, dead microorganisms 12. C. L. Brierley, J. A. Brierley, and M. S. Davidson, in
can have higher metal-sorption capacities than living T. J. Beveridge and R. J. Doyle, eds., Metal Ions and
material (127). The dead biomass may be processed into Bacteria, John Wiley & Sons, New York, 1989, pp. 359–382.
a form in which it acts as an ion-exchange resin so that 13. B. Volesky and Z. R. Holan, Biotechnol. Prog. 11, 235–250
it can be regenerated by simple acid or alkali treatment. (1995).
This makes such systems more economically competitive 14. D. W. Darnall et al., in R. Thompson, ed., Trace Metal
because the sorbed metals are easily recovered and the Removal from Aqueous Solutions, Royal Society of Chem-
biomass can be reused (126). istry, Burlington House, London, U.K., 1986, pp. 1–23.
680 BIOMINERALIZATION BY BACTERIA

15. Y. E. Collins and G. Stozky, in T. J. Beveridge and R. J. 45. F. G. Ferris, T. J. Beveridge, and W. S. Fyfe, Nature 320,
Doyle, eds., Metal Ions and Bacteria, John Wiley & Sons, 609–611 (1986).
New York, 1989, pp. 31–90. 46. K. O. Konhauser, W. S. Fyfe, F. G. Ferris, and T. J. Beve-
16. A. A. Korenevskii, G. I. Karavaiko, and V. V. Sorokin, ridge, Geology 21, 1103–1106 (1993).
Microbiology 62, 1085–1092 (1993). 47. K. O. Konhauser et al., Appl. Environ. Microbiol. 60,
17. A. A. Korenevsky, V. V. Sorokin, and G. I. Karavaiko, Min. 549–553 (1994).
Pro. Ext. Met. Rev. 19, 341–353 (1998). 48. K. O. Konhauser et al., Geomicrobiology 15, 209–222 (1998).
18. E. M. Greenfield, D. C. Wilson, and M. A. Crenshaw, Am. 49. K. O. Konhauser and F. G. Ferris, Geology 24, 323–326
Zool. 24, 925–932 (1984). (1996).
19. S. Mann, Nature 332, 119–124 (1988). 50. S. Schultze-Lam, F. G. Ferris, K. O. Konhauser, and
20. G. Stozky, in D. C. Savage and M. Fletcher, eds., Bacterial R. G. Wiese, Can. J. Earth Sci. 32, 2021–2026 (1995).
Adhesion, Plenum Press, New York, 1985, pp. 195–253. 51. D. Fortin and T. J. Beveridge, Chem. Geol. 141, 235–250
21. S. G. Walker et al., Appl. Environ. Microbiol. 55, 2976–2984 (1997).
(1989). 52. K. Tazaki, Clays Clay Miner. 45, 203–212 (1997).
22. L. A. Warren and F. G. Ferris, Environ. Sci. Technol. 32, 53. T. J. Beveridge, J. T. Meloche, W. S. Fyfe, and R. J. E. Mur-
2331–2337 (1998). ray, Appl. Environ. Microbiol. 45, 1094–1108 (1983).
23. F. G. Ferris, K. O. Konhauser, B. Lyvén, and K. Pedersen, 54. H. L. Ehrlich, Geomicrobiol. J. 16, 135–153 (1999).
Geomicrobiol. J. 16, 181–192 (1999). 55. D. R. Lovley, in L.Donald and L. Sparks, eds., Advances
24. C. A. Flemming, F. G. Ferris, T. J. Beveridge, and G. W. Ba- in Agronomy, vol. 54, Academic Press, New York, 1995,
iley, Appl. Environ. Microbiol. 56, 3191–3203 (1990). pp. 175–231.
25. H. Ehrlich, Geomicrobiology, Marcel Dekker, New York, 56. N. A. Yurkova and N. N. Lyalikova, Microbiology 59,
1996. 672–677 (1990).
26. J. C. Cañaveras et al., Geomicrobiol. J. 16, 9–25 (1999). 57. J. R. Lloyd, J. A. Cole, and L. E. Macaskie, J. Bacteriol. 179,
27. H. S. Chafetz, P. F. Rush, and N. M. Utech, A.A.P.G. Bull. 2014–2021 (1996).
73, 421 (1989). 58. J. R. Lloyd et al., Geomicrobiol. J. 15, 45–58 (1998).
28. C. Buczynski and H. S. Chafetz, J. Sediment. Petrol. 61, 59. J. R. Lloyd et al., Appl. Environ. Microbiol. 65, 2691–2696
226–233 (1991). (1999).
29. C. Vasconcelos et al., Nature 377, 220–222 (1995). 60. D. R. Lovley and E. J. P. Phillips, Appl. Environ. Microbiol.
30. G. A. Sokolova-Dubinina and Z. P. Derugina, Microbiology 58, 850–856 (1992).
36, 445–451 (1967). 61. Y. A. Gorby and D. R. Lovley, Environ. Sci. Technol. 26,
205–207 (1992).
31. S. Golubic, in N. G. Carr and B. A. Witton, eds., The Biology
of Blue-Green Algae, Blackwell Scientific Publications, 62. F. Caccavo, Jr., R. P. Blakemore, and D. R. Lovley, Appl.
Oxford, U.K., 1973, pp. 334–472. Environ. Microbiol. 58, 3211–3216 (1992).
32. S. Kempe et al., Nature 349, 605–608 (1991). 63. D. R. Lovley, P. K. Widman, J. C. Woodward, and
E. J. P. Phillips, Appl. Environ. Microbiol. 59, 3572–3576
33. S. Schultze-Lam, G. Harauz, and T. J. Beveridge, J. Bacte-
(1993).
riol. 174, 7971–7981 (1992).
64. D. R. Lovley and E. J. P. Phillips, Appl. Environ. Microbiol.
34. S. Schultze-Lam and T. J. Beveridge, Appl. Environ. Micro-
54, 1472–1480 (1986).
biol. 60, 447–453 (1994).
65. D. R. Lovley, J. F. Stolz, G. L. Nord, and E. J. P. Phillips,
35. J. B. Thompson and F. G. Ferris, Geology 18, 995–998
Nature 330, 252–254 (1987).
(1990).
66. J. K. Fredrickson et al., Geochim. Cosmochim. Acta 62,
36. J. B. Thompson, F. G. Ferris, and D. A. Smith, Palaios 5,
3239–3257 (1998).
52–75 (1990).
67. H. Dong et al., Chem. Geol. 169, 299–318 (2000).
37. J. B. Thompson, S. Schultze-Lam, T. J. Beveridge, and
J. Des Marais, Limnol. Oceanogr. 42, 133–141 (1997). 68. J. McLean, T. J. Beveridge, and D. Phipps, in A. Leeson and
B. C. Alleman, eds., Bioremediation of Metals and Inorganic
38. Z. A. Avakyan, N. P. Belkanova, G. I. Karavaiko, and
Compounds, Proceedings of the Fifth International In
V. P. Piskunov, Microbiology 54, 250–256 (1984). Situ and On-Site Bioremediation Symposium, San Diego,
39. W. E. Krumbein and D. Werner, in W. E. Krumbein, ed., Calif., April 19–22, Batelle Press, Columbus, Ohio, 1999,
Microbial Geochemistry, Blackwell Scientific Publications, pp. 121–126.
New York, 1983. 69. R. Gerlach, A. Cunningham, and F. Caccavo, Jr., in
40. J. D. Birchall, in R. J. P. Williams and J. R. R. F. Da Silva, A. Leeson and B. C. Alleman, eds., Bioremediation of
eds., New Trends in Bioinorganic Chemistry, Academic Metals and Inorganic Compounds, Proceedings of the
Press, New York, 1978, pp. 209–252. Fifth International In Situ and On-Site Bioremediation
41. C. W. Sullivan, Silicon Biochemistry, Ciba Foundation Symposium, San Diego, Calif., April 19–22, Batelle Press,
Symposium 121, John Wiley & Sons, Chichester, U.K., 1986, Columbus, Ohio, 1999, pp. 13–18.
pp. 59–89. 70. M. M. Urrutia, E. E. Roden, J. K. Fredrickson, J. M. Zach-
42. MM. Urrutia and T. J. Beveridge, J. Bacteriol. 175, ara, Geomicrobiol. J. 15, 269–291 (1999).
1936–1945 (1993). 71. T. J. Beveridge, Can. J. Microbiol. 24, 89–104 (1978).
43. MM. Urrutia and T. J. Beveridge, Chem. Geol. 116, 261–280 72. S. Langley and T. J. Beveridge, Appl. Environ. Microbiol.
(1994). 65, 489–498 (1999).
44. F. G. Ferris, W. S. Fyfe, and T. J. Beveridge, Chem. Geol. 73. A. A. Korenevskii, V. V. Sorokin, and G. I. Karavaiko,
63, 225–232 (1987). Microbiology 62, 630–633 (1993).
 BIOMINERALIZATION BY BACTERIA 681

74. T. J. Beveridge and R. G. E. Murray, J. Bacteriol. 127, 101. F. J. Ferris, W. S. Fyfe, T. J. Beveridge, Geomicrobiol. J. 5,
1502–1518 (1976). 3342 (1987).
75. T. J. Beveridge and R. G. E. Murray, J. Bacteriol. 141, 102. D. Hastings and S. Emerson, Geochim. Cosmochim. Acta 50,
876–887 (1980). 1819–1824 (1986).
76. G. Southam and T. J. Beveridge, Geochim. Cosmochim. Acta 103. K. W. Mandernack, J. Post, and B. M. Tebo, Geochim. Cos-
58, 4527–4530 (1994). mochim. Acta 59, 4393–4408 (1995).
77. S. A. Marakushev, Geomicrobiology and Biochemistry of 104. B. Wehrli, G. Friedl, and A. Manceau, in C. P. Huang,
Gold, Nauka, Moscow, Russia, 1991 (in Russian). C. R. O’Melia, and J. J. Morgan, eds., Aquatic Chemistry:
78. D. I. Nikitin, M. S. Oranskaya, A. S. Savvichev, and P. V. Interfacial Interspecies Processes, American Chemical Soci-
Mikheev, Izv. AN SSSR, Ser. Biol. 284, 302–306 (1986) (in ety, Washington, D.C., 1995.
Russian). 105. L. F. Adams and W. C. Ghiorse, J. Bacteriol. 169, 1279–
79. F. D. Ovcharenko, Z. R. Ulberg, S. V. Grabara, Izv. AN 1285 (1987).
SSSR, Ser. Biol. 283, 711–713 (1985) (in Russian). 106. Y. M. Nelson, L. W. Lion, W. C. Ghiorse, and L. M. Shuler,
80. A. S. Savvichev, D. I. Nikitin, and M. S. Oranskaya, Geo- Appl. Environ. Microbiol. 65, 175–180 (1999).
khimiya 1, 1529–1531 (1986) (in Russian). 107. D. Fortin, F. G. Ferris and T. J. Beveridge, in J. G. Banfield
81. A. S. Savvichev, D. I. Nikitin, G. G. Mineev, and M. S. Oran- and K. H. Nealson, eds., Geomicrobiology: Interactions
skaya, Geokhimiya 10, 117–119 (1985) (in Russian). Between Microbes and Minerals, Reviews in Mineralogy 35,
82. R. A. Amosov, S. L. Vasin, and E. B. Kolesnikova, Ecosystem Mineralogical Society of America, Washington, D.C., 1997,
Reconstruction and Evolution of Biosphere, vol. 2, The pp. 161–180.
Paleontological Institute of the Russian Academy if Sciences, 108. D. A. Crerar, G. W. Knox, and J. L. Means, Chem. Geol. 24,
Moscow, Russia, 1995, pp. 147–152 (in Russian). 111–135 (1979).
83. D. W. Darnall, B. Green, and J. Gardea-Torresdey, in 109. P. Cloud, Econ. Geol. 68, 1135–1143 (1973).
P. R. Norris and D. P. Kelly, eds., Biohydrometallurgy, Pro- 110. J. C. J. Walker, Nature 309, 340 341 (1984).
ceedings of the International Symposium, Kew, Surrey, 111. H. Mann, in B. Volesky, ed., Biosorption of Heavy Metals,
U.K., July 12–16, 1987, Sci. Technol. Lett. 487–498 CRC Press, Boca Raton, Fla., 1990, pp. 93–138.
(1988).
112. T. Klaus, R. Joerger, E. Olsson, and C. G. Granqvist, PNAS
84. D. W. Darnall, B. Green, and J. Gardea-Torresdey Sci. 96, 13611–13614 (1999).
Technol. Lett. 487–498 (1988).
113. S. Douglas and T. J. Beveridge, FEMS Microbiol. Ecol. 26,
84. N. Kuyucak and B. Volessky, Biorecovery 1, 219–235 (1989). 79–88 (1998).
85. V. V. Balashova and G. A. Dubinina, in M. I. Ivanov, ed., 114. D. Fortin and T. J. Beveridge, Geomicrobiol. J. 14, 1–21
Chemosynthesis, Nauka, Moscow, 1989, pp. 101–122 (in (1997).
Russian).
115. K. O. Konhauser, Earth Sci. Rev. 43, 91–121 (1998).
86. W. C. Ghiorse, Annu. Rev. Microbiol. 38, 515–550 (1984).
116. K. O. Stetter, in K. Horikoshi and W. D. Grant, eds.,
87. T. P. Stupakova, G. A. Dubinina, L. L. Demina, Microbiol- Extremophiles, Microbial Life in Extreme Environments,
ogy 60, 108–115 (1991). Wiley-Liss, New York, 1998, pp. 1–24.
88. V. O. Kalinenko, Mikrobiologiya 15, 364–371 (1946) (in 117. H. W. Jannasch, in S. E. Hamphris, R. A. Zierenberg,
Russian). L. S. Taullineaux and R. E. Thomson, eds., Sea Floor
89. P. Hirsh, Arch. Microbiol. 60, 201–216 (1968). Hydrothermal Systems: Physical, Chemical, Biological
90. B. M. Tebo et al., in J. F. Banfield and K. H. Nealson, and Geological Interactions, Geophysical Monograph 91,
eds., Geomicrobiology: Interaction Between Microbes and American Geophysical Union, Washington, D.C., 1995,
Minerals, Reviews in Mineralogy 35, Mineralogical Society pp. 273–296.
of America, Washington, D.C., 1997, pp. 225–266. 118. D. Fortin, F. G. Ferris, and S. D. Scott, Am. Mineral. 83,
91. P. E. Kepkay and K. H. Nealson, Arch. Microbiol. 148, 1399–1408 (1998).
63–67 (1987). 119. N. G. Holm, Chem. Geol. 63, 45–57 (1987).
92. S. Heising and B. Schink, Microbiology 144, 2263–2269 120. F. G. Ferris, W. Shotyk and W. S. Fyfe, in T. J. Beveridge
(1998). and R. J. Doyle, eds., Metal Ions and Bacteria, John Wiley
93. S. Heising, L. Richter, and B. Schink, Arch. Microbiol. 172, & Sons, New York, 1989, pp. 413–441.
116–124 (1999). 121. G. Gerdes and W. E. Krumbein, Biolaminated Deposits, Lec-
94. V. V. Balashova and G. A. Zavarzin, Microbiology 41, ture Notes in Earth Sciences 9, Springer Verlag, Weinheim,
808–810 (1972). Germany, 1987.
95. G. A. Dubinina, Microbiology 47, 471–478 (1978). 122. F. Snowball, Earth Planet. Sci. Lett. 126, 129–142
96. J. Barrett, M. N. Hughes, G. I. Karavaiko, and P. A. Spen- (1994).
cer, Metal Extraction by Bacterial Oxidation, Ellis Horwood, 123. D. S. McKay et al., Science 273, 924–930 (1996).
New York, 1993. 124. S. F. Tyrrel and P. Howsam, Q. J. Eng. Geol. 30, 161–169
97. D. B. Johnson, M. A. Ghauri, and M. F. Daid, Appl. Environ. (1997).
Microbiol. 58, 1423–1428 (1992). 125. P. R. Dugan, EPA Report 600/2-75-032, National Technical
98. P. Bacelar-Nicolau and D. B. Johnson, Appl. Environ. Information Service, Springfield, VA, 1975.
Microbiol. 65, 585–590 (1999). 126. B. Volesky, in B. Volesky, ed., Biosorption of Heavy Metals,
99. H. L. Ehrlich, Econ. Geol. 59, 1306–1312 (1964). CRC Press, Boca Raton, Fla., 1990, p. 343.
100. L. Carlson, E. B. Lindstroem, K. B. Hollberg, and O. H. Tuo- 127. M. M. Urrutia, M. Kemper, R. Doyle, and T. J. Beveridge,
vinen, Appl. Environ. Microbiol. 58, 1046–1049 (1992). Appl. Environ. Microbiol. 58, 3837–3844 (1992).
682 BIOMINING

BIOMINING. See BIOLEACHING

R O

O (CH2)x
n
BIOMINING USING ARCHAEA. See ARCHAEA IN Figure 1. Chemical structure of PHAs. R represents a side chain
BIOTECHNOLOGY having from 0 to 14 carbon atoms, x may range from 0 to 4, and n is
an integer such that the molecular mass of the chain is higher than
about 100,000 Da, typical of biologically produced PHAs. PHB:
R = CH3 , x = 1. P(4HB): R = H, x = 2. PHO: R = (CH2 )4 CH3 ,
x = 1.
BIOODORS CONTROL IN WASTE TREATMENT
FACILITIES. See BIOFILTRATION AND BIOODORS
the availability of some other element required for growth
(such as nitrogen or phosphorus) is low. The polymer can
accumulate to as much as 90% of the dry cell weight
under controlled fermentation conditions, but generally
BIOPESTICIDES. See INSECTICIDES, MICROBIAL much less under natural conditions. Table 1 shows some
of the PHA-producing organisms and typical composition
ranges for their PHAs. It is by no means an exhaustive list;
more than 300 organisms are known to produce PHAs (1).
Note that the absence of a composition in Table 1 does not
BIOPLASTICS necessarily signify the inability of the organism to produce
that composition; these are only compositions that have
FRANK SKRALY typically been reported.
Metabolix, Inc.
Cambridge, Massachusetts
Table 1. Natural PHA Producers and Chain Lengths
Every living organism produces a tremendous variety
Found in Polymers
of biopolymers. Each organism contains DNA, RNA,
and proteins, but other important macromolecules found Monomer Chain Length
in living systems include polysaccharides (such as (Carbon Atoms)
starch, cellulose, and xanthan gum), polyphosphates, Organism 3 4 5 6 7+
polysulfates, polyphenols (such as lignin), isoprenoids
(such as natural rubber), and polyhydroxyalkanoates Acinetobacter sp. ● ●
(PHAs). PHAs are unique among the biopolymer groups Aeromonas caviae ● ●
in that they are all genuine thermoplastics — plastic Agrobacterium sp. ● ●
Alcaligenes latus ● ●
materials that can be melted, formed, and reset. PHAs
Azotobacter vinelandii ● ●
are unique among the plastics in that they are made Bacillus megaterium ● ●
inside living organisms, and thus they may be called Chromatium vinosum ● ●
‘‘bioplastics.’’ Sometimes materials such as polylactide Chromobacterium violaceum ● ●
are called ‘‘bioplastics’’ as well, but the polymerization Clostridium kluyveri ●
is synthetic, using lactic acid produced by bacteria. This Comomonas acidovorans ● ●
article focuses on PHAs because they are plastics made Corynebacterium hydrocarboxydans ● ●
entirely by biological systems. Haloferax mediterranei ● ●
PHAs are water-insoluble polyesters that accumulate Hydrogenophaga pseudoflava ●
in the form of inclusions or granules inside microbial Legionella pneumophila ●
Methylobacterium extorquens ● ● ●
cells. The purified polymers have properties ranging
Methylobacterium organophilum ●
from hard and brittle to soft and rubbery, depending
Nocardia corallina ● ● ●
on the composition. Many different kinds of monomers Paracoccus denitrificans ● ●
can be incorporated into PHAs, but the backbone of all Pseudomonas oleovorans ● ● ● ●
of these monomers is essentially the same; they are all Pseudomonas putida ● ●
hydroxyalkanoic acids activated by linkage to coenzyme A. Ralstonia eutropha ● ● ●
The polymerization is a process by which an enzyme called Rhizobium meliloti ● ●
PHA synthase, or PHA polymerase, adds the activated Rhodobacter capsulatus ● ●
monomer to the end of the growing polymer chain, yielding Rhodococcus sp. ● ● ●
free coenzyme A and a polymer chain that has become one Rhodospirillum rubrum ● ● ● ●
Sphaerotilus natans ● ●
unit longer. The basic structure of PHAs is shown in
Spirulina platensis ●
Figure 1.
Synechococcus sp. ●
PHAs are made naturally by many species of bacteria Syntrophomonas wolfei ● ● ●
as a carbon and energy storage compound, often under Zoogloea ramigera ● ●
conditions where the carbon source is present in excess and
 BIOPLASTICS 683

The occurrence of PHAs in living cells has been known compounds. W. R. Grace & Co. were the first to rec-
since the 1920s. Maurice Lemoigne of Institut Pasteur ognize the commercial potential of PHB. They used
first discovered in 1925 that Bacillus spp. can synthesize B. megaterium and Rhodospirillum rubrum for produc-
3-hydroxybutyrate (3HB) (2), the first reported production tion, and in 1960 the company filed the first patent
of 3HB by a microbe. In the two years that followed, he concerning efficient production of PHB (12). W. R. Grace
definitively identified a polymer of 3HB units having the eventually abandoned the project, but Imperial Chemical
formula (C4 H6 O2 )n and melting point 157 ° C to be present Industries (ICI), later Zeneca, revived it in the 1970s,
as a reserve material in certain Bacillus spp. (3,4). He with an important modification: the efficient introduction
also observed that the polymer could be degraded by the of monomers other than 3HB. They found that the crys-
organism under autolytic conditions to a material having a tallinity of PHB could be reduced in this way, thus making
lower molecular weight, lower melting point, and different the polymer more flexible and commercially of greater
solubility characteristics. potential (13). This led to the development of Biopol, a
The notion that PHAs are a carbon and energy storage copolymer of 3HB and 3HV units in about a 4 : 1 ratio.
material was solidified by the work of Macrae and Biopol first became commercially available in 1990 in the
Wilkinson, who set out to determine the physiological form of shampoo bottles in Germany (14). The Monsanto
role of poly(3-hydroxybutyrate) (P(3HB), or simply PHB) Corporation acquired the technology in 1996, but it dis-
formation in Bacillus spp. in 1958 (5,6). They observed continued production of Biopol, which had reached about
the necessary conditions for demonstrating that PHB 800 tons per year (15), blaming lagging market interest
is a reserve carbon and/or energy source: that the in renewable materials (16). The Biopol composition will
accumulation of PHB increased as the carbon-to-nitrogen become public domain in 2001 on the expiration of the
ratio increased, that PHB was rapidly degraded in original ICI patent.
the absence of a carbon/energy source, and that the The remainder of this article explores in more
degradation of PHB in the absence of a carbon/energy detail the following aspects of PHAs: metabolism (both
source could prevent cell autolysis and death. This naturally occurring and via metabolic engineering),
degradation makes PHB not only useful to the cells as material properties and applications, production and
an energy source, but it also makes PHB useful to humans recovery, and market and economics.
for its biodegradability.
Direct confirmation that PHB was the main constituent
PHA METABOLISM
of the insoluble granules observed in bacilli occurred in
1958, when Williamson and Wilkinson (7) demonstrated
There are three main modes of PHA production in natu-
that PHB within cells could be isolated and quantified
rally occurring organisms: (1) condensation of acetyl-CoA,
by treating whole cells with sodium hypochlorite. Among
(2) fatty acid b-oxidation, and (3) fatty acid biosynthesis.
their many analyses of the isolated material was
It is possible, of course, for these organisms to derive
the precipitation of a film of PHB from chloroform
hydroxyalkanoyl-CoAs from closely related carbon sources
by evaporation, and they described it as ‘‘somewhat
such as the corresponding hydroxyacids, but the three
reminiscent of plastic sheeting.’’ They also showed that
modes discussed here are concerned primarily with PHA
the material underwent thermal decomposition, giving
production from common carbon sources such as sugars
crotonic acid as a main product.
and oils. The final PHA composition is affected strongly
PHB was the only known naturally occurring
by the mode or modes employed for its synthesis. Conden-
PHA until 1974, when Wallen and Rohwedder (8)
sation of acetyl-CoA generally leads to short-chain PHAs,
reported the isolation of a polyester containing 3HB,
commonly defined as having monomer units of up to 5 car-
3-hydroxyvalerate (3HV), and 3-hydroxyhexanoate (3HH)
bon atoms, whereas fatty acid b-oxidation generally leads
units from activated sludge. Nine years later, Findlay
to medium-chain PHAs having units of 6 to 14 carbon
and White (9) found PHAs with monomer units hav-
atoms. Long-chain PHAs have 16 or more carbon atoms
ing four to eight carbons in marine sediments and in
per unit, and these have appeared in PHAs only in small
Bacillus megaterium extracts, and in the same year, de
proportions from b-oxidation of long-chain fatty acids (17).
Smet and coworkers (10) isolated a polymer composed
Metabolic engineering of PHA pathways is also discussed
predominantly of 3-hydroxyoctanoate (3HO) units from
in the following sections because it provides both flexibil-
Pseudomonas oleovorans grown on octane. Presently, at
ity and novelty in polymer compositions, and it also can
least 100 different monomer types have been found in
enable PHA production in nearly any organism desired.
PHAs, ranging from three to fourteen carbons in length,
and the position of the hydroxyl group has ranged from
Condensation of Acetyl-CoA
C-2 to C-6 (the α to ε carbon). The constituents found thus
far have invariably been (R)-hydroxyacids; the (R) configu- Macrae and Wilkinson (5) speculated in 1958 that PHB
ration, the hydroxyl group, and activation with CoA are is synthesized via acetoacetate by the condensation of
characteristics to which no exceptions have yet been found. acetate and some other two-carbon compound, and that
For a comprehensive review of the diversity of monomer the active compound for polymerization is 3HB-CoA. A
types found in PHAs, the reader is encouraged to consult direct demonstration that 3HB-CoA is the species used
Steinbüchel and Valentin (11). for polymerization was given by Merrick and Doudoroff
The thermoplasticity and biodegradability of PHAs in 1961 (18). They also provided evidence that 3HB
has led to more than just academic interest in these was not directly activated with CoA. On the basis of
684 BIOPLASTICS

assays of 3-ketothiolase and 3HB-CoA dehydrogenase seems to be at the enzymatic level (25). The intracellular
in cell extracts, Schindler (19) correctly concluded in free CoA level does appear to be significantly lower during
1964 that the pathway shown in Figure 2 is the one PHB-producing conditions (29).
employed by Ralstonia eutropha. The PHB pathway The reduction of acetoacetyl-CoA to (R)-3HB-CoA is
proceeds by condensation of two acetyl-CoA molecules to accomplished in most cases by a single stereospecific
form acetoacetyl-CoA, followed by a reduction to 3HB-CoA reductase. Most of the reductases characterized to date
and polymerization. The most common implementation have shown a preference for NADPH over NADH as
of this pathway is that found in organisms such as cofactor (30–32), but there is at least one case where
R. eutropha and Zoogloea ramigera, where the reduction NADH is preferred (33). This second step of the pathway
specifically produces the (R) isomer of 3HB-CoA. A more exerts some thermodynamic control over PHB synthe-
complex pathway exists in R. rubrum, however; the overall sis. The 3-ketothiolase reaction is thermodynamically
reduction to (R)-3HB-CoA appears to be accomplished more favorable in the direction of cleavage, not con-
first with a reduction of acetoacetyl-CoA to (S)-3HB-CoA, densation (34). Therefore, for PHB synthesis to proceed,
followed by the action of two enoyl-CoA hydratases with the acetoacetyl-CoA concentration must remain low. This
the intermediate crotonyl-CoA (20). can be accomplished with a high NADPH/NADP+ ratio,
The first step in the PHB pathway, the reversible which will thermodynamically favor (R)-3HB-CoA forma-
condensation of two acetyl-CoA molecules to acetoacetyl- tion. Provided that the PHA synthase activity is sufficient,
CoA with the release of free CoA, is catalyzed by the NADPH/NADP+ ratio is instrumental in fueling the
a 3-ketothiolase. The substrate specificity of these pathway (35).
enzymes dictates that one reactant is always acetyl- PHA synthase is the final step of all the pathways to
CoA, but the other reactant can vary in length. Many PHA, so the substrate specificity of this enzyme controls
organisms possess more than one of this type of enzyme. the polymer composition to some degree regardless
Ralstonia eutropha, for example, has at least two 3- of the pathway employed. Most synthases have a
ketothiolases that can participate in PHA synthesis. One preference for either short-chain or medium-chain PHAs,
will condense acetyl-CoA with acetyl-CoA or propionyl- although some recent exceptions have been reported.
CoA, and the other appears to accept all acyl-CoAs The Aeromonas caviae synthase accepts both 3HB-
from acetyl-CoA to octanoyl-CoA (21). Condensation of CoA and 3-hydroxyhexanoyl-CoA (3HH-CoA) (36), and the
acetyl-CoA with propionyl-CoA yields 3-ketovaleryl-CoA, Thiocapsa pfenigii synthase accepts a broad range of
which is reduced to 3HV-CoA. Propionyl-CoA, not nearly substrates (37,38). The majority of synthases have only
as common as acetyl-CoA in biological systems, is an one type of subunit, but of the first 30 to be sequenced,
end product of the catabolism of fatty acids with odd four have two subunits (39).
numbers of carbon atoms, but it is also derived in
some organisms from succinyl-CoA via methylmalonyl- Fatty Acid b-Oxidation
CoA (22,23). 3-Ketothiolase has a regulatory role in PHA The b-oxidation pathway (Fig. 3) is the most common mode
synthesis by this route. In many cases, the condensation of fatty acid catabolism. Some bacteria, most notably the
reaction is strongly inhibited by free CoA (21,24–27). pseudomonads, can divert carbon flow out of b-oxidation
In R. eutropha, the PHB pathway is constitutively to store the carbon as PHA. Fatty acids are first activated
expressed (28), so the regulation of 3-ketothiolase activity with CoA, then undergo repeated rounds of removal
of two-carbon units, ultimately in the form of acetyl-
CoA, by a thiolase. This is the same type of reaction
O O O catalyzed by 3-ketothiolase, but the reaction proceeds
phaA
predominantly in the direction of cleavage. Hydroxyacyl-
O
CoAs are intermediates in b-oxidation, but they appear
R SCoA R SCoA
SCoA
in this pathway as the (S) stereoisomer. The diversion of
these intermediates therefore requires the conversion of
phaB
the (S) form to the (R) form. Alternatively, the (R) form
can be formed directly by the action of an enzyme that is
R O OH O not part of the b-oxidation pathway.
phaC The pathways that organisms actually use to divert
carbon flow from b-oxidation to PHA synthesis have only
O R SCoA recently begun to be explored. It has been known since
n
1983 (10) that Pseudomonas spp. can effectively synthesize
Figure 2. Pathway from acetyl-CoA and other acyl-CoAs to medium-chain PHAs from fatty acid feeds. Precisely how
PHAs. When R = CH3 , two acetyl-CoA molecules are condensed they do this is still not known, although it is presumed
by 3-ketothiolase (EC 2.3.1.9), the product of the phaA gene. The
to be via one of the three possible activities for the
resulting acetoacetyl-CoA is reduced to (R)-3-hydroxybutyryl-CoA
generation of (R)-hydroxyacyl-CoAs shown in Figure 3.
by 3-ketoacyl-CoA reductase (EC 1.1.1.36), the product of the
phaB gene, and finally PHA synthase (no EC number), the Expression of Pseudomonas aeruginosa PHA synthase
product of the phaC gene, incorporates an (R)-3-hydroxybutyrate in E. coli fatty acid degradation mutants resulted in
unit into PHA with the release of CoA. When R = C2 H5 , medium-chain PHA production (40). This suggests that
propionyl-CoA is condensed with acetyl-CoA, and ultimately an Escherichia coli possesses at least one of these three
(R)-3-hydroxyvalerate unit is incorporated into PHA. activities, although it does not normally synthesize PHAs
 BIOPLASTICS 685

R OH
Fatty acid

R SCoA
acyl-CoA

R SCoA
enoyl-CoA
Hydratase (phaJ )

OH O OH O

Epimerase
R SCoA R SCoA
(S)-hydroxyacyl-CoA (R)-hydroxyacyl-CoA
Figure 3. The derivation of PHA precursors from
the β-oxidation pathway. The solid arrows represent
O O O reactions that are known to occur widely in biological
Reductase systems. The dashed arrows represent reactions that
can generate (R)-3-hydroxyacyl-CoAs, the precursors of
SCoA R SCoA PHA. In each round of β-oxidation, the alkyl group R
acetyl-CoA 3-ketoacyl-CoA becomes shorter by two carbons.

at all, and it also suggests that these additional activities PHA side chains decreased with temperature, as is gene-
are common in nature. Only one of the three activities rally also true of membrane lipid composition, offering
has actually been found, but not in Pseudomonas: an (R)- more evidence that the fatty acid biosynthesis pathway
specific enoyl-CoA hydratase. The first example of this can be used for PHA synthesis (45).
activity was found in A. caviae, which can synthesize PHA synthesis using the fatty acid biosynthesis
P(3HB-co-3HH) from fatty acids or oils (36,41). Another pathway, just as from the b-oxidation pathway, requires
example of the activity was found in R. rubrum (42). at least one additional activity for the synthesis of
Both genes have been cloned and are denoted as phaJ. (R)-hydroxyacyl-CoAs. The intermediates of fatty acid
Both of these hydratases accept enoyl-CoAs of four to biosynthesis are all of the (R) form, but the prosthetic
six carbons, but not eight. The R. rubrum enzyme, when group is the acyl carrier protein (ACP). Therefore,
expressed in E. coli with the promiscuous PHA synthase ACP must be exchanged for CoA, either directly by a
of Thiocapsa pfenigii, produced P(3HB-co-3HH) when the transacylase or by two sequential activities, a thioesterase
carbon source was oleic acid. This shows how enzymes to remove ACP and a transferase or synthetase to activate
other than PHA synthase can at least partially dictate with CoA. One development in this area was the discovery
polymer composition. of the phaG gene in P. putida and the demonstration that
it is a 3-hydroxyacyl-ACP : CoA transferase (46). The gene
Fatty Acid Biosynthesis was also found in P. aeruginosa (47).

Both prokaryotes and eukaryotes synthesize fatty acids.

Metabolic Engineering
Both use them for lipids to construct membranes. Many
eukaryotes store energy by fatty acid synthesis in the Naturally occurring PHA-producing organisms certainly
form of fats and oils, but bacteria generally do not. generate a wide array of polymers. The diversity of
Some bacteria, however, can use PHA synthesis to divert PHAs and their numerous potential applications have
carbon flow from fatty acid synthesis to store carbon and motivated a great deal of effort in metabolic engineering
energy (Fig. 4). In this sense PHAs can be thought of as of PHA production in both academia and industry. The
‘‘bacterial fat.’’ main advantages that metabolic engineering offers in
In 1990 it was found that many pseudomonads can syn- PHA production are that it can generate novel PHA
thesize medium-chain PHAs when fed gluconate (43,44), compositions, allow PHA synthesis from less expensive
suggesting that the fatty acid biosynthesis pathway might substrates, improve the characteristics of the material,
be involved. Later it was shown in Pseudomonas putida give better control over composition, improve the efficiency
grown on glucose that the degree of unsaturation in the of production, and enable PHA synthesis in almost any
686 BIOPLASTICS

O O

R S-ACP
3-ketoacyl-ACP

Transacylase
OH O (phaG) OH O

R S-ACP R SCoA
(R )-hydroxyacyl-ACP (R )-hydroxyacyl-CoA

R S-ACP
enoyl-ACP O O

CO2
HO S-ACP
O malonyl-ACP
Figure 4. The derivation of PHA precursors from the
fatty acid biosynthesis pathway. In each instance of
condensation with malonyl-ACP, the alkyl group R R S-ACP
becomes longer by two carbons. acyl-ACP

desired organism. These advantages are summarized in PHAs or synthesize unwanted PHA constituents. Unlike
Table 2, with an example of each. many natural PHA producers, E. coli grows at 37 ° C and
Metabolic engineering of PHA production became is relatively easily lysed. Furthermore, the use of a single
possible in the late 1980s, when three independent organism to make an array of compositions necessitates
groups cloned the PHA biosynthetic genes (phaCAB) from only one general fermentation procedure, even if feeding
R. eutropha (48–50). Much progress has been made since strategies and operating conditions vary. Good control over
then in assembling a library of cloned genes useful for composition allows the tailoring of material properties
PHA synthesis. For example, at least 30 PHA synthase within a design space, a theme that is explored in the next
genes have been cloned (39). section.
The host organism of choice for metabolic engineering
of PHAs has most often been E. coli. PHAs with monomer
Properties and Applications
units ranging from 3 to at least 12 carbons have been
synthesized in engineered E. coli (42,51–54). The use of
Because more than 100 different constituents have been
E. coli for PHA production is desirable for technical as
incorporated into PHAs, and because these can be mixed
well as economic reasons. The genome of E. coli has been
in the same polymer chain, the number of compositions
sequenced in its entirety, and its metabolism and genetic
that can be generated is quite large. The constituents vary
techniques are well understood. Because E. coli does not
in length, position of the hydroxy group, and presence
naturally produce PHAs, it possesses no ability to degrade
of functional groups and double bonds. As a result, the
material properties and applications of PHAs are quite
Table 2. Some Advantages of Metabolic Engineering in diverse. All of the possible combinations of monomer
PHA Production constituents make up the design space in which the
engineering of PHA material properties can take place.
Advantage Example Ref.
The first conscious exercise of this nature was undertaken
Novel compositions Poly(3-hydroxyhexenoate) with 121 by ICI. They wished to develop PHAs commercially, but
polyketide synthase PHB by itself suffers from some disadvantages. Its melting
Inexpensive P(3HB-co-3HV) from glucose in 122 point (about 180 ° C) is very close to the temperature
substrates E. coli, from CO2 in plants at which it begins to undergo rapid loss of molecular
Characteristics of Molecular weight control 123 weight by thermal degradation, making melt processing
material somewhat difficult. When processed conventionally, it is
Control over P(4HB) with no residual PHB 124 also too brittle a material to be useful in many applications
composition because of its very low nucleation density and high
Efficiency of Reduction of Pseudomonas 125
crystallinity. On adding propionate to an R. eutropha
production putida lysate viscosity
Any desired PHB synthesis in cotton fiber 126 fermentation as a cofeed, it was found that significant
organism amounts of 3HV were incorporated into the polymer. This
increased the flexibility of the material, as measured by
 BIOPLASTICS 687

modulus, elongation to break, and impact strength, and it is translucent. P(4HB) homopolymer is extremely tough
significantly depressed the melting point (13). Thus began but also is very flexible. Medium-chain PHAs such as
the endeavor to explore more fully the PHA design space. poly(3-hydroxyoctanoate-co-20%-3-hydroxydecanoate) are
translucent and rubbery but somewhat stiff, and they can
Material Properties be tacky due to their low melting points.

Many PHAs have been evaluated for basic material Biodegradability and Biocompatibility
properties over the last two decades. The short-chain A distinguishing feature of PHAs is that they are fully
PHAs tend to be more crystalline than the medium-chain biodegradable (can be reduced to carbon dioxide and water
PHAs, and hence generally the short-chain PHAs are more in a natural environment in a relatively short time) and
rigid, whereas the medium-chain PHAs are more rubbery. often biocompatible (do not cause adverse reactions when
Table 3 gives an overview of the material properties of present in living biological systems). Although they are
some PHAs. Its purpose is to illustrate the range of degraded by microorganisms in composting, PHAs are
properties that exists; certainly many more compositions stable in everyday environments; they do not break down
have been characterized. on the shelf.
Biologically produced PHAs can have molecular masses The environmental degradation of PHAs primarily is
of more than 1 million daltons. Although the ring-opening due to the action of microbially secreted depolymerases,
polymerization of lactones to produce PHAs is a well- which hydrolyze the polymer to compounds that are
known procedure, its main drawbacks are that its products taken up by bacteria, fungi, and algae and converted
are generally of lower molecular weight and that, in cases ultimately to carbon dioxide and water. Methane can also
where chirality is an issue (when the hydroxyl group is not result from anaerobic biodegradation. The depolymerases
terminal), its products are not stereoregular, and hence only hydrolyze ester linkages of monomers in the (R)
not isotactic (61). Molecular weight problems become more configuration (63), and thus synthetic nonstereoregular
pronounced for the products of less strained lactones such PHAs are not completely biodegradable in this sense.
as P(4HB) from γ -butyrolactone. The rate of degradation varies with composition and
PHB does not have especially advantageous properties, molecular weight, and it also varies with the type of
as discussed earlier, but it is evident from Table 3 natural environment. A bottle made from P(3HB-co-3HV),
that copolymerization with another monomer type can for example, will degrade completely in 5 to 10 years
improve the situation. In many cases, even a few percent in a freshwater lake (64) but within 12 weeks in sewage
of another constituent lowers the melting point and sludge (65). P(3HB-co-4HB) degrades in seawater at about
apparent crystallinity significantly. Sometimes the other the same rate as P(3HB-co-3HV) (66) and somewhat faster
constituent can cocrystallize with PHB to some extent, than PHB and P(3HB-co-3HV) in soil (67). In sterile
as has been reported for 3HV (56), so a somewhat aqueous environments, polymers such as these hydrolyze
higher percentage may be required to achieve the same very slowly. Doi and coworkers (68) showed that after
plasticizing effect. PHB copolymers of this type are 180 days at 37 ° C in 0.01 M phosphate buffer at pH 7.4,
similar to commodity plastics such as polypropylene and PHB and copolymers with 3HV and 4HB had not lost
high-density polyethylene, hence the first commercial significant mass, but their molecular weights (or average
application of P(3HB-co-3HV) as shampoo bottles (14). polymer chain lengths) began to decrease, apparently due
As the PHB comonomer fraction increases, material to simple hydrolytic chain scission.
properties may change significantly. For example, P(3HB- The biocompatibility of several PHAs has been
co-15% 4HB) stretches and recovers like a rubber band but evaluated. PHAs can perform very well in this capacity,

Table 3. Range of PHA Material Properties in Comparison with Those of Common Plastics

Tensile Elongation
Mw Tm Tg Hm Crystallinity Strength to Break
Composition (kDa) (° C) (° C) (J/g) (%) (MPa) (%) Ref.

P(3HB) 1,020 177 4 97 60 43 5 127

P(4HB) 780 53 −48 9 34 104 1,000 128
P(3HP) 200 77 −19 74 37 129
P(3-hydroxy-4-pentenoate) 95 63 −11 15 130
P(3HB-co-20% 3HV) 130 35 20 50 131
P(3HB-co-25% 3HV) 336 132 0 40 13
P(3HB-co-17% 3HH) 1,122 120 −2 34 26 20 850 127
P(3HB-co-10% 4HB) 1,185 159 −3 54 46 132
P(3HB-co-10% 4HB) 45 24 242 133
P(3HB-co-8% C8 -C12 ) 1,392 133,146 −8 39 17 680 134
P(3HO-co-12% 3HH) 178 59 −37 19 135
Polypropylene 176 −10 50–70 38 400 136
Polyethylene terephthalate 267 69 30–50 70 100 136
High-density polyethylene 135 −80 80–95 30 30–1,000 137,138
688 BIOPLASTICS

provided they are extremely pure. Medium-chain PHAs, P(4HB) and were seeded with mammalian cells, and each
for example, have been shown not to elicit a significant material showed a good level of cell attachment and colla-
inflammatory response after 40 weeks subcutaneously gen formation (82). These materials offer the added benefit
implanted in mice and not to elicit an allergic response of moldability and pliability; in the same study, a work-
on 48-hour exposure to the skin of guinea pigs (69). In ing valve could not be formed from polyglycolic acid, a
many cases, PHAs are also resorbable in animal systems, traditional medical material. PHAs have already been suc-
meaning they can be completely metabolized over time cessfully used to regenerate pericardial tissue and limit
by the host without production of toxins or adverse health postoperative pericardial adhesions (83,84) and to correct
effects. Short-chain PHAs such as PHB have been shown to atrial septal defects (70) in vivo. In drug delivery applica-
be both biocompatible and resorbable in a relatively short tions, PHAs can be used subcutaneously or in pill form in
time (70,71). PHB is actually present as low molecular humans or as a stomach bolus in animals (85). In either
weight oligomers at a few milligrams per liter in human case the composition of the PHA can be tailored so that
blood plasma (72), and 3HB is a commonly encountered the approximate release rate of the drug is appropriate
ketone body (73); significant toxicity of PHB hydrolysis without the need for frequent dosing. In any surgical or
products thus seems unlikely. dressing application, bits of material left in the body are
of no concern as they are simply resorbed (85).
Applications PHA granules produced in biological systems are
usually amorphous, either because of very low nucleation
PHAs can be processed like commonly used thermoplas-
density, as with short-chain PHAs, or to low crystallization
tics. They can be useful in the form of containers and other
rates or lack of inherent crystallinity, as with medium-
molded items, disposable packaging and articles, medical
chain PHAs (62). Aqueous processing of these granules
implants and devices, coatings and paints, and in many
can yield a slurry of granules called latex. On spreading
specialty applications. Different compositions of PHA will,
the latex onto a surface, it dries and can crystallize either
of course, find uses in different areas, especially when the
spontaneously or when stimulated by heat or a nucleating
distinction between the more rigid character of short-chain
agent. Medium-chain PHAs with unsaturated side chains
PHAs and the more rubbery character of medium-chain
can even cross-link during drying (86). Such latexes can
PHAs is considered.
be used to coat paper or plastic and as water-based paint
PHAs can be recycled like many commodity-based
ingredients (86,87) or to coat cheese and other foods (88).
thermoplastics, but because they are biodegradable,
P(3HB-co-3HV) received food contact approval in Europe
they lend themselves especially well to applications in
in 1995 (87).
which recycling is not practical, such as in single-use
Specialty applications of PHAs are too numerous to list
disposable food packaging and garbage bags, in which
here. PHAs can be used in agricultural applications such
the plastic is mixed with food or garbage, and personal
as planters and films, where biodegradability eliminates
care items, in which the plastic is either contaminated
the need for removal, and they can be used as vehicles
or the article is made of different types of plastic
for controlled release of pesticides, with biodegradation
that are too difficult to separate. Biodegradation can
being the means to liberate chemicals embedded in the
be a useful alternative to recycling in some cases.
polymer (85). They are useful as binders, where they
Mixtures of polyethylenes of different molecular weights,
hold powders or small pieces of another material together
for example, often suffer deterioration of mechanical
and are subsequently burned out cleanly (89). Intriguing
properties on repeated recycling and processing to the
food applications include chewing gum base (90) and fat
point when the material becomes unusable (74). P(3HB-
substitute (91). PHAs with unsaturated side chains can
co-3HV) did in fact appear in disposable applications
be cross-linked to form materials that behave similar to
during the 1990s, while Zeneca still produced Biopol. It
rubber, yet are still fully biodegradable (62).
was used not only in bottles in Germany, Japan, and
An often overlooked but potentially very important
the United States, but it also appeared in disposable
application of PHAs is not as polymers, but as chemicals
razors and refrigerated meat trays (75). Compostable
derived from those polymers hydrolyzed chemically or bio-
diapers and sanitary napkins made from PHAs are of
logically. (R)-hydroxyalkanoic acids can be used to synthe-
current industrial interest (74,76–79). These are a good
size antibiotics, vitamins, aromatics, and pheromones (92).
illustration of the utility of PHAs; diapers, for example,
(R)-3HB is used as a starting material for the Merck
may be composed of nonwoven absorbable material,
antiglaucoma drug Trusopt (93). Other chemicals with
backsheet, and pressure-sensitive adhesive tape, made
existing markets can be derived from PHAs by simple
from PHAs and fully biodegradable.
chemical reactions; for example, PHB can yield 1,3-
The biocompatibility and in some cases resorbability of
butanediol, crotonic acid, β-amino acids, butyl esters,
PHAs make them very useful as medical materials. Appli-
lactones, and other compounds (94). Analogous arguments
cations include tissue coatings, stents, sutures, tubing,
can be made for P(4HB), P(3HO), etc.
prostheses, bone and tissue cements, tissue regeneration
devices, wound dressings, drug delivery materials, and
diagnostic and prophylactic uses (80). An active area of Production and Recovery
research with PHAs is tissue engineering (81), in which a
resorbable material is used as a scaffold to ‘‘grow’’ human The two types of organisms that are most favorable for
tissue from the patient’s own cells. Recently, trileaflet industrial production of PHAs are bacteria and plants.
heart valves were formed from P(3HO-co-3HH) and from Bacterial fermentation can be carried out as Zeneca
 BIOPLASTICS 689

developed the process, with natural PHA producers such as plasmid accumulated 157 g/L of PHB at 77% of the dry
R. eutropha. With the advent of routine bacterial genetic cell weight in 49 hours with a glucose feed at 30 ° C (98).
manipulation since that time, recombinant producers such PHA production in plants has been implemented
as E. coli have been developed as well. The engineering of only on a very small scale. The first difficulty to be
PHA production in plants is a longer-term solution because overcome with plants is the achievement of high, stable
plant genetic manipulation is not routine. The reward, expression of multiple foreign genes in the same plant. The
however, would justify the task; the resulting large scale compartmentalization and complexity of metabolism is an
and low cost of PHA production could make them a real added difficulty. However, there is reason to be optimistic;
alternative to commodity plastics derived from fossil fuels. in 1994 it was reported that Arabidopsis thaliana plants
Fermentation and agricultural production of PHAs ide- could produce PHB at up to 14% of the dry cell weight
ally coexist. Plants are suited to large-scale commodity and still develop normally when the R. eutropha PHA
applications without much complexity in composition. synthesis genes were targeted to the plastid (99).
Crops are subject to variations in climatic and soil condi- The two main modes of PHA recovery from biomass
tions, and they must synthesize all polymer constituents are solvent extraction and aqueous processing. Solvent
from carbon dioxide. Therefore, fine control of PHA com- extraction consists of suspending biomass, usually dried,
position and characteristics is likely not possible in plants. in a solvent that can dissolve the PHA of interest,
Bacterial fermentation is more expensive, relatively small- removing solids, then precipitating the polymer either
scale, and rather energy-intensive, making it questionable by the addition of a poor solvent, by evaporation, or
for long-term commodity production. However, bacterial by cooling. Aqueous processing is a general term for
fermentations can be controlled precisely, and any feed- recovery techniques that begin with wet biomass, as from
stock desired may be used. Medical and other specialty a fermentation or wet milling operation, and proceed in
applications require this flexibility and do not require the the presence of water to yield either an aqueous, largely
immense scale of agriculture. amorphous latex or dry polymer with some degree of
The most commonly used organisms in fermentation crystallinity, depending on its composition. An amorphous
have been R. eutropha for short-chain PHAs and Pseu- latex can be reconstituted from a crystalline PHA (100).
domonas spp. for medium-chain PHAs. Both types of cells Large-scale application of solvent extraction suffers
are cultured aerobically in a fed-batch mode. Typically, from the need to use large volumes of solvents; PHA solu-
the natural producers of PHA require limitation in some tions become viscous even at relatively low concentrations.
nutrient other than the carbon source to accumulate PHA The most useful solvents for short-chain PHAs are chlori-
to more than a few percent of the dry cell weight. Limi- nated solvents such as chloroform and 1,2-dichloroethane,
tations in nitrogen and phosphorus are typically used which must be contained carefully. The latter prob-
in practice. This can be implemented by including in lem has been addressed to some degree by the use of
the medium all the nitrogen or phosphorus needed to nonchlorinated solvents at higher temperatures (101,102).
achieve the desired cell density; when this is completely Medium-chain PHAs can readily be extracted with rela-
consumed, growth ceases and PHA accumulation begins. tively benign solvents such as acetone, isopropanol, and
For example, R. eutropha accumulated 137 g/L of P(3HB- hexane (103,104).
co-29% 3HV) at 67% of the dry cell weight in 65 hours There are many variations of the aqueous processing
with a feed of corn oil and propionic acid under phos- theme, but generally aqueous processing consists of one
phorus limitation at 34 ° C (95). An interesting alternative or more of the following steps: separation of solid biomass
to R. eutropha is Alcaligenes latus, which can accumulate from bulk liquid, mechanical or enzymatic cell breakage,
PHA in significant amounts during exponential growth solubilization of non-PHA biomass, and isolation of poly-
and can be cultured at up to 42 ° C (96). Pseudomonas mer. Usually the polymer remains in amorphous granule
putida accumulated 73 g/L of medium-chain PHA, con- form throughout the procedure, although exceptions exist,
taining predominantly 3HO and 3HD at 51% of the dry cell such as melting and coalescence of granules followed by
weight in 38 hours on oleic acid and phosphorus limitation settling of bulk polymer (105). Initial reduction of the liq-
at 30 ° C (97). uid volume is especially useful when bacterial cells are
Transgenic PHA production has been developed pri- taken from a fermentation and the reduction is accom-
marily with E. coli, although genetic manipulations have plished by centrifugation, filtration, or flocculation. Bac-
been made on natural PHA producers as well. There are terial cells can be broken by heat treatment (106), homo-
currently many advantages to the use of transgenic E. coli: genization (107), or chemical treatment with hypochlo-
genetic manipulations are routine, the metabolism is well rite (108,109), alkali (110), surfactants or water-soluble
understood, the genome has been completely sequenced, copolymeric dispersants (111), or oxidants such as hydro-
minimal medium can be used with many types of feed- gen peroxide (112) or ozone (113). Heat treatment may be
stocks, nutrient limitation is not necessary, the cells lyse followed by digestion with enzymes such as proteases,
easily, growth at up to 42 ° C is possible, no depolymerases lysozyme, and DNase, which can become costly (106).
are present to degrade the product, and the choice of Homogenization is rather energy-intensive (114) and does
one organism eliminates the need to develop a markedly not effect any purification. The chemical treatments have
different fermentation process every time a new polymer the advantages of being inexpensive and facilitating purifi-
composition is sought. Escherichia coli is capable of per- cation by solubilizing non-PHA biomass (115). Recovery
formance at least on a par with natural producers; a strain of the polymer granules from the resulting solution is
expressing the R. eutropha PHA synthesis genes from a generally done by washing and centrifugation, but other
690 BIOPLASTICS

methods such as sedimentation fractionation (116) and air to be in the neighborhood of $1/lb (121). These processes
classification (117) have been described. set an optimistic example for the foray of fermentation
technology into the plastics market, and they demonstrate
the legitimacy of the industrial effort to produce
Market and Economics
bioproducts on a large scale. Industry has become
committed to the development of biodegradable plastics
Above all else, the factor that will determine the fate of
PHAs on the market is the price. There is no question on a large scale as well. For example, Cargill Dow is
about the usefulness and favorable properties of these expanding its capacity for polylactic acid (synthetically
materials, but the last sales of any volume to take place polymerized from fermentation-derived lactic acid) from
were of Biopol in the 1990s, for about $8/lb (118). Even 16 million lb/yr to 280 million lb/yr (122), and DuPont is
with all the goodwill of biodegradability and derivation developing Biomax, a polyethylene terephthalate with
from renewable resources, bulk commodity applications monomers vulnerable to hydrolysis incorporated (123).
certainly cannot take hold at this price. With larger scale The most promising sectors for growth of biodegradables
and established markets, come lower costs, but these have in the near future are predicted to be food packaging,
not yet developed because plastic users in many cases have compost bags, paper coatings (milk, juice, frozen food,
little experience with PHAs. At present only very small etc.), dishes and cutlery (of which 100 million lb/yr can
quantities of PHAs are available commercially, so market be replaced with biodegradables), and disposable hygiene
development is crucial. products; forecasts for the 2003 biodegradable plastics
One exception to the price issue is the medical use of market range from 50 million to 200 million lb/yr (122).
PHAs. Here performance is much more important. The The maturation of transgenic PHA plant technology
cost of the polymeric material in, for example, a heart would bring about a sea change in the economics of PHA
valve replacement procedure is of little consequence, but production. This is a long-term goal in comparison with
the specifications on the material are quite exacting. The fermentative production, so predictions are not as easy
main hurdles to market acceptance in this area still to make. Preliminary estimates put the price of plant-
lie in regulatory approvals and in confident acceptance derived PHAs at $0.25 to $0.60/lb (94,124), competitive
by medical associations. The overall yearly demand for with petroleum-based commodity thermoplastics, which
vascular grafts and heart valve replacements worldwide typically sell in this range (125). For PHAs to assume a
corresponds to approximately 1 million procedures (81). volume of billions of lb/yr, however, certain milestones
A few analyses of expected cost of PHA production must be met. Starch or oil carbon flux must be diverted
have been conducted recently by van Wegen and by metabolic engineering to achieve a polymer content
coworkers (107) and by Choi and Lee (118,119). These of about 30 to 40% of the plant dry weight. Aside from
studies estimated the cost of producing short-chain technical issues, regulatory approval must be obtained for
PHAs using technology that is currently available. Both the transgenic plant to be used in the field as well as for the
used a process consisting of fermentation followed by use of the particular polymer in applications such as food
centrifugation, hypochlorite treatment, two more rounds contact. Companies with the ability to develop applications
of centrifugation, and finally drying. The main difference on a large scale must be involved, and further development
between the studies was that one used homogenization of composting infrastructure would be necessary if the
(three passes) and the other simply used a surfactant biodegradability aspect of PHAs is to be very meaningful.
to complement the hypochlorite treatment. The studies If these milestones are met, the stage is set for the
agreed that the cost would be approximately $2.50 to release of commodity plastic production from dependence
$2.80/lb for a scale of 5 to 10 million lb/yr. These studies on feedstocks such as crude oil, which is projected to run
also agreed that improvements to the current pilot-scale out during our century (126).
process could readily be made, and that these could reduce
the cost, especially at a larger scale, to $1.20 to $1.40/lb.
BIBLIOGRAPHY
The most sensitive parameter according to both studies is
the PHA content as a percentage of the dry cell weight
1. S. Y. Lee, Trends Biotechnol. 14, 431–438 (1996).
because fewer impurities means less energy expended on
downstream processing. Other reasonable improvements 2. M. Lemoigne, Ann. Inst. Pasteur (Paris) 39, 144–173 (1925).
cited were reduction of raw material cost (i.e., using 3. M. Lemoigne, Bull. Soc. Chim. Biol. 8, 770–782 (1926).
hydrolyzed corn starch at $0.10/lb instead of glucose 4. M. Lemoigne, Ann. Inst. Pasteur (Paris) 41, 148–165 (1927).
at $0.24/lb), efficiency of centrifugation, optimization of 5. R. M. Macrae and J. F. Wilkinson, J. Gen. Microbiol. 19,
the chemical treatment, and higher productivity during 210–222 (1958).
the fermentation. Less-detailed studies were conducted 6. R. M. Macrae and J. F. Wilkinson, Proc. R. Soc. Edin. A 27,
to determine potential costs of medium-chain PHA 73–78 (1958).
production, and in the best case the estimate was about 7. D. H. Williamson and J. F. Wilkinson, J. Gen. Microbiol. 19,
$2.30/lb, again with PHA content of cells being the most 198–209 (1958).
important parameter (120). 8. L. L. Wallen and W. K. Rohwedder, Environ. Sci. Technol.
Citric acid and glutamate are produced by established 8, 576–579 (1974).
large-scale fermentation processes at nearly 2 billion lb/yr, 9. R. H. Findlay and D. C. White, Appl. Environ. Microbiol. 45,
approaching that of smaller-scale bulk petrochemical 71–78 (1983).
processes. The selling price of these bioproducts tends 10. M. J. de Smet et al., J. Bacteriol. 154, 870–878 (1983).
 BIOPLASTICS 691

11. A. Steinbüchel and H. E. Valentin, FEMS Microbiol. Lett. 45. G. N. M. Huijberts et al., Appl. Environ. Microbiol. 58,
128, 219–228 (1995). 536–544 (1992).
12. U.S. Patent 3, 044,942, 1962, J. N. Baptist. 46. B. H. A. Rehm, N. Krüger, and A. Steinbüchel, J. Biol.
13. U.S. Patent 4, 477,654, 1984, P. A. Holmes, S. H. Collins, Chem. 273, 24044–24051 (1998).
and L. F. Wright. 47. N. Hoffmann, A. Steinbüchel, and B. H. A. Rehm, FEMS
14. Anonymous, New Scientist 126, 36 (1990). Microbiol. Lett. 184, 253–259 (2000).
15. G. Braunegg, G. Lefebvre, and K. F. Genser, J. Biotechnol 48. PCT Pat. Appl. WO 97/22711, 1997, D. H. Sherman,
65, 127–161 (1998). M. D. Williams, and Y. Xue.
16. J. Evans, Chemistry in Britain, February 2000, pp. 41–44. 49. H. E. Valentin et al., Int. J. Biol. Macromol. 25, 303–306
17. V. Mittendorf et al., Plant J. 20, 45–55 (1999). (1999).
18. J. M. Merrick and M. Doudoroff, Nature 189, 890–892 50. S. J. Sim et al., Nat. Biotechnol. 15, 63–67 (1997).
(1961). 51. S. Hein, B. Söhling, G. Gottschalk, and A. Steinbüchel,
19. J. Schindler, Arch. Microbiol. 49, 236–255 (1964). FEMS Microbiol. Lett. 153, 411–418 (1997).
20. G. J. Moskowitz and J. M. Merrick, Biochemistry 8, 2748– 52. Z. L. Boynton et al., Appl. Environ. Microbiol. 65, 1524–
2755 (1969). 1529 (1999).
21. G. W. Haywood, A. J. Anderson, L. Chu, and E. A. Dawes, 53. M. E. John and G. Keller, Proc. Natl. Acad. Sci. U.S.A. 93,
FEMS Microbiol. Lett. 52, 91–96 (1988). 12768–12773 (1996).
22. D. R. Williams, A. J. Anderson, E. A. Dawes, and D. F. 54. S. C. Slater, W. H. Voige, and D. E. Dennis, J. Bacteriol.
Ewing, Appl. Microbiol. Biotechnol. 40, 717–723 (1994). 170, 4431–4436 (1988).
23. H. E. Valentin and D. Dennis, Appl. Environ. Microbiol. 62, 55. P. Schubert, A. Steinbüchel, and H. G. Schlegel, J. Bacteriol.
372–379 (1996). 170, 5837–5847 (1988).
24. T. Nishimura, T. Saito, and K. Tomita, Arch. Microbiol. 116, 56. O. P. Peoples and A. J. Sinskey, J. Biol. Chem. 264,
21–27 (1978). 15298–15303 (1989).
25. P. J. Senior and E. A. Dawes, Biochem. J. 134, 225–238 57. R. V. Antonio, A. Steinbüchel, and B. H. A. Rehm, FEMS
(1973). Microbiol. Lett. 182, 111–117 (2000).
26. B. Maekawa, N. Koyama, and Y. Doi, Biotechnol. Lett. 15, 58. S. J. Liu and A. Steinbüchel, Appl. Environ. Microbiol. 66,
691–696 (1993). 739–743 (2000).
27. V. Oeding and H. G. Schlegel, Biochem. J. 134, 239–248 59. S. Slater, T. Gallaher, and D. Dennis, Appl. Environ.
(1973). Microbiol. 58, 1089–1094 (1992).
28. A. Steinbüchel and H. G. Schlegel, Mol. Microbiol. 5, 60. PCT Pat. Appl. WO 99/61,624, 1999, F. A. Skraly and
535–542 (1991). O. P. Peoples.
29. D. A. Mansfield, A. J. Anderson, and L. A. Naylor, Can. J. 61. Y. Doi, S. Kitamura, and H. Abe, Macromolecules 28,
Microbiol. 41(Suppl. 1), 44–49 (1995). 4822–4828 (1995).
30. G. A. F. Ritchie, P. J. Senior, and E. A. Dawes, Biochem. J. 62. Y. Saito and Y. Doi, Int. J. Biol. Macromol. 16, 99–104
121, 309–316 (1971). (1994).
31. T. Fukui, M. Ito, T. Saito, and K. Tomita, Biochim. Biophys. 63. E. Shimamura, M. Scandola, and Y. Doi, Macromolecules
Acta 917, 365–371 (1987). 27, 4429–4435 (1994).
32. G. W. Haywood, A. J. Anderson, L. Chu, and E. A. Dawes, 64. H. E. Valentin et al., Macromolecules 32, 7389–7395 (1999).
FEMS Microbiol. Lett. 52, 259–264 (1988). 65. W. Luzier, Proc. Natl. Acad. Sci. U.S.A. 89, 839–842 (1992).
33. M. Liebergesell and A. Steinbüchel, Eur. J. Biochem. 209, 66. S. Nakamura, Y. Doi, and M. Scandola, Macromolecules 25,
135–150 (1992). 4237–4241 (1992).
34. S. Masamune, C. T. Walsh, A. J. Sinskey, and O. Peoples, 67. Y. Doi, A. Segawa, and M. Kunioka, Int. J. Biol. Macromol.
Pure Appl. Chem. 61, 303–312 (1989). 12, 106–111 (1990).
35. I. Y. Lee, M. K. Kim, H. N. Chang, and Y. H. Park, FEMS 68. H. Matsusaki, H. Abe, and Y. Doi, Biomacromolecules 1,
Microbiol. Lett. 131, 35–39 (1995). 17–22 (2000).
36. T. Fukui and Y. Doi, J. Bacteriol. 179, 4821–4830 (1997). 69. H. Preusting, A. Nijenhuis, and B. Witholt, Macromolecules
37. M. Liebergesell, F. Mayer, and A. Steinbüchel, Appl. Micro- 23, 4220–4224 (1990).
biol. Biotechnol. 40, 292–300 (1993). 70. Y. Doi, Microbial Polyesters, VCH Publishers, New York,
38. H. E. Valentin, A. Schönebaum, and A. Steinbüchel, Appl. 1990.
Microbiol. Biotechnol. 46, 261–267 (1996). 71. H.-G. Elias, An Introduction to Polymer Science, VCH
39. B. H. A. Rehm and A. Steinbüchel, Int. J. Biol. Macromol. Publishers, Weinheim, Germany, 1997.
25, 3–19 (1999). 72. S. S. Schwartz and S. H. Goodman, Plastics Materials and
40. S. Langenbach, B. H. A. Rehm, and A.Steinbüchel, FEMS Processes, Van Nostrand Reinhold, New York, 1982.
Microbiol. Lett. 150, 303–309 (1997). 73. S. Bloembergen et al., Macromolecules 20, 3086–3089
41. T. Fukui, N. Shiomi, and Y. Doi, J. Bacteriol. 180, 667–673. (1987).
42. S. E. Reiser, T. A. Mitsky, and K. J. Gruys, Appl. Microbiol. 74. G. de Koning, Can. J. Microbiol. 41(Suppl. 1), 303–309
Biotechnol. 53, 209–218 (2000). (1995).
43. G. W. Haywood, A. J. Anderson, D. F. Ewing, and E. A. 75. D. Jendrossek, A. Schirmer, and H. G. Schlegel, Appl.
Dawes, Appl. Environ. Microbiol. 56, 3354–3359 (1990). Microbiol. Biotechnol. 46, 451–463 (1996).
44. A. Timm and A. Steinbüchel, Appl. Environ. Microbiol. 56, 76. H. Brandl and P. Püchner, Biodegradation 2, 237–243
3360–3367 (1990). (1992).
692 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES

77. L. L. Madison and G. W. Huisman, Microbiol. Molec. Biol. 113. U.S. Patent 6,043,063, 2000, D. L. Kurdikar et al.
Rev. 63, 21–53 (1999). 114. U.S. Patent 5,894,062, 1999, J. M. Liddell.
78. T. Hammond and J. J. Liggat, in G. Scott and D. Gilead, 115. PCT Int. Appl. WO 93,11,656 A1, 1993, E. Ohleyer.
eds., Degradable Polymers, Chapman & Hall, London, U.K.,
116. U.S. Patent 6, 083, 729, 2000, D. P. Martin, O. P. Peoples,
1995, pp. 88–111.
and S. F. Williams.
79. M. Kunioka, Y. Kawaguchi, and Y. Doi, Appl. Microbiol.
117. U.S. Patent 5, 952, 460, 1999, J. M. Liddell, N. George, and
Biotechnol. 30, 569–573 (1989).
A. Hall.
80. Y. Doi, Y. Kanesawa, Y. Kawaguchi, and M. Kunioka,
118. U.S. Patent 4, 910, 145, 1990, P. A. Holmes and G. B. Lim.
Macromol. Chem., Rapid. Commun. 10, 227–230 (1989).
119. R. J. van Wegen, Y. Ling, and A. P. J. Middelberg, Trans.
81. S. F. Williams, D. P. Martin, D. M. Horowitz, and O. P. Peo-
IChemE 76 (Part A), 417–426 (1998).
ples, Int. J. Biol. Macromol. 25, 111–121 (1999).
120. E. Berger et al., Biotechnol. Tech. 3, 227–232 (1989).
82. T. Malm et al., Scand. J. Thor. Cardiovasc. Surg. 26, 9–14
(1992). 121. J. A. Ramsay, E. Berger, B. A. Ramsay, and C. Chavarie,
Biotechnol. Tech. 4, 221–226 (1990).
83. A. Hazari et al., J. Hand Surg. [Br.] 24, 291–295 (1999).
122. W. J. Page and A. Comish, Appl. Environ. Microbiol. 59,
84. R. N. Reusch, A. W. Sparrow, and J. Gardiner, Biochim.
4236–4244 (1993).
Biophys. Acta 1,123, 33–40 (1992).
123. U.S. Patent 5, 977, 250, 1999, N. George et al.
85. L. Stryer, Biochemistry, 3rd ed., W.H. Freeman and Com-
pany, New York, 1987. 124. U.S. Patent 5, 691, 174, 1997, J. M. Liddell and T. J. Locke.
86. U.S. Patent 5,753,364, 1998, D. R. Rutherford, W. J. Ham- 125. PCT Int. Appl. WO 99,51,760 A1, 1999, D. M. Horowitz and
mar, and G. N. Babu. E. M. Brennan.
87. D. Byrom, in D. P. Mobley, ed., Plastics from Microbes, 126. T. U. Gerngross, Nature Biotechnol. 17, 541–544 (1999).
Hanser, Munich, 1994, pp. 5–33. 127. J.-I. Choi and S. Y. Lee, Biotechnol. Bioeng. 62, 546–553
88. U.S. Patent 5,747, 584, 1998, I. Noda. (1999).
89. U.S. Patent 5,536, 564, 1996, I. Noda. 128. U.S. Patent 5, 918,747, 1999, I. Noda.
90. U.S. Patent 5,489,470, 1996, I. Noda. 129. U.S. Patent 5, 849, 854, 1998, I. Noda.
91. U.S. Patent 5,292,860, 1994, T. Shiotani and G. Kobayashi. 130. J. I. Choi and S. Y. Lee, Bioproc. Eng. 17, 335–342 (1997).
92. PCT Int. Appl. WO 98,51,812 A2, 1998, S. F. Williams, 131. J. Choi and S. Y. Lee, Appl. Microbiol. Biotechnol. 51, 13–21
D. P. Martin, T. Gerngross, and D. M. Horowitz. (1999).
93. S. F. Williams, D. P. Martin, D. M. Horowitz, and O. P. Peo- 132. G. de Koning, M. Kellerhals, C. van Meurs, and B. Witholt,
ples, Int. J. Biol. Macromol. 25, 111–121 (1999). J. Environ. Polymer Degr. 4, 243–252 (1996).
94. R. Sodian et al., ASAIO J. 46, 107–110 (2000). 133. D. Wilke, Appl. Microbiol. Biotechnol. 52, 135–145 (1999).
95. T. Malm et al., Scand. J. Thor. Cardiovasc. Surg. 26, 15–21 134. A. Warmington, presented at Biodegradable Plastics Confe-
(1992). rence, Frankfurt, Germany, June 6, 2000.
96. O. Duvernoy, T. Malm, J. Ramstrom, and S. Bowald, Tho- 135. K. Atwood, presented at Biodegradable Plastics Conference,
rac. Cardiovasc. Surg. 43, 271–274 (1995). Frankfurt, Germany, June 6, 2000.
97. P. A. Holmes, Phys. Technol. 16, 32–36 (1985). 136. F. R. van der Leij and B. Witholt, Can. J. Microbiol.
98. U.S. Patent 6,024,784, 2000, G. J. H. Buisman, F. P. Cupe- 41(Suppl. 1), 222–238 (1995).
rus, R. A. Weusthuis, and G. Eggink. 137. Chemical Market Reporter, July 3, 2000.
99. S. Williams and O. Peoples, Chemistry in Britain, December, 138. D. L. Klass, Biomass for Renewable Energy, Fuels, and
1997, pp. 29–32. Chemicals, Academic Press, San Diego, Calif., 1998.
100. U.S. Patent 5,958,480, 1997, G. Eggink and M. D. Northolt.
101. U.S. Patent 6,071,998, 2000, E. Muller, D. Horowitz, and
A. Egozy.
BIOREMEDIATION: AN OVERVIEW OF HOW
102. PCT Int. Appl. WO 99,39,588 A1, 1999, W. Li, C. P. Orfan, MICROBIOLOGICAL PROCESSES CAN BE
J. Liu, and J. W. Foster.
APPLIED TO THE CLEANUP OF ORGANIC AND
103. U.S. Patent 5,229,158, 1993, M. Yalpani. INORGANIC ENVIRONMENTAL POLLUTANTS
104. S. Y. Lee, Y. Lee, and F. Wang, Biotechnol. Bioeng. 65,
363–368 (1999).
ROGER C. PRINCE
105. S. C. Stinson, Chem. Eng. News 1995, pp. 10–26. ExxonMobil Research and
106. S. F. Williams and O. P. Peoples, CHEMTECH 26, 38–44 Engineering Co.
(1996). Annandale, New Jersey
107. U.S. Patent 5,871,980, 1999, L. A. Naylor and J. C. Wood.
108. U.S. Patent 4,786,598, 1988, R. M. Lafferty and G.
BIOREMEDIATION, ECOLOGY, TECHNOLOGY
Braunegg.
109. S. Y. Lee et al., Biotechnol. Bioeng. 68, 466–470 (2000).
Bioremediation
110. F. Wang and S. Y. Lee, Appl. Environ. Microbiol. 63,
4765–4769 (1997). Bioremediation relies on encouraging biological processes
111. C. Nawrath, Y. Poirier, and C. Somerville, Proc. Natl. Acad. to minimize an unwanted environmental impact; usually
Sci. U.S.A. 91, 12760–12764 (1994). it is the removal of a contaminant from the biosphere.
112. D. M. Horowitz, E. M. Brennan, J. J. Koon, and T. U. Gern- A narrow definition of bioremediation might focus on the
gross, Macromolecules 32, 3347–3352 (1999). conversion of contaminating organic molecules to carbon
 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES 693

dioxide, water, and inorganic ions, and the oxidation or technologies, such as thermal desorption and destruction
reduction of contaminating inorganic ions. A broader defi- of organic contaminants, are also permanent solutions,
nition would include biological processes for ameliorating but the simplest and most widely used remediation
extremes of pH, concentrating contaminants so that they option, that of removing a contaminant to a dump
can be more easily removed by physical techniques, con- site, merely moves the problem, and may well not
verting toxic species to less toxic or less bioavailable forms eliminate the potential liability. Furthermore, by its
that pose less of a threat to the environment, and restor- very nature bioremediation addresses the bioavailable
ing functional ecosystems in contaminated or disturbed part of any contamination, and when biodegradation or
sites when the contaminants or disturbance cannot be bioaccumulation ceases this probably means that the
removed. This article focuses on biological treatments for bioavailable part of the contamination has been addressed.
contaminants when they have escaped into the environ- Residual concentrations of contaminants, while perhaps
ment, particularly on the initial microbial reactions that detectable by sensitive analytical techniques, may have
remove or convert the contaminant. only minimal residual environmental impact. The same
Many approaches are being used effectively to catalyze cannot necessarily be said for nonbiological technologies,
bioremediation. Sometimes the simple addition of an which may leave bioavailable contaminants at low levels.
otherwise limiting nutrient is appropriate, while at other Bioremediation also has the advantage that it can be
times microbial inocula are needed. In other situations, relatively nonintrusive, and can sometimes be used in
it may be important to add readily degradable substrates situations in which other approaches would be severely
to foster cometabolism of a contaminant or to drive an disruptive. For example, bioremediation has been used to
environment toward anaerobiosis to stimulate anaerobic cleanup hydrocarbon spills under buildings, roads, and
degradation. Alternatively, it might be necessary to airport runways without interfering with the continued
add surfactants or other chemicals to increase the use of these facilities.
bioavailability of a contaminant.
Bioremediation therefore overlaps several older bio- Ecology
technologies. Municipal and industrial wastewater treat- The biosphere plays a major role in the great elemental
ment is a well-established industry and while it can be cycles of our planet (2), and bioremedation must be placed
distinguished from bioremediation in that the pollutants in this context if we are to appreciate its broadest
are under physical control during treatment, many of the ramifications. One of the underlying fundamental truths of
fundamental biological processes are the same. Similarly, biological diversity is that if there is free energy available
composting is a well-established phenomenon, currently in the metabolism of a substrate, there is probably a
gaining popularity in the municipal solid waste treatment guild of organisms that has evolved to make use of it.
industry, and the biofiltration of waste gases is becoming a This is particularly important for the biodegradation
useful technology. Development of these technologies, for of organic molecules. For example, crude oil seeps,
cases in which the contaminant is already under physical to land and water, have occurred for millennia, and
control, will undoubtedly aid the development of bioreme- as a consequence, aerobic oil-degrading microorganisms
diation as an accepted tool for dealing with similar wastes are ubiquitous. If biology does not yet take advantage
when they have escaped our control. of a source of free energy, then we can expect that
Bioremediation is already a commercially viable there will be strong selection pressure in favor of any
technology, with estimates of aggregate bioremediation organism that develops an ability to exploit it. This has
revenues of $2 to $3 billion for the period 1994 to been seen with by-products of nylon manufacture, where
2000 (1). There are significant opportunities to enlarge Pseudomonas aeruginosa has gained the ability to degrade
upon this success. Bioremediation has applications in the the novel compound 6-aminohexanoate linear dimer, a by-
gas phase, in water, and in soils and sediments. For product of nylon-6 manufacture, as the sole source of
water and soils, the process can be carried out in situ carbon and nitrogen (3). The successful bioremediation of
or after the contaminated material has been moved to xenobiotic compounds, such as pesticides and herbicides,
some sort of contained reactor (ex situ). The former is may sometimes represent a similar acquisition of genetic
generally rather less expensive, but the latter may be so traits.
much faster that the additional cost of manipulating the Not all organic molecules provide a source of free
contaminated material is overshadowed by the time saved. energy, however. Some, such as small halogenated
Bioremediation may explicitly exploit bacteria, fungi, solvents, provide no significant source of nutrients or
algae, or higher plants, although our focus here is on the energy, and their aerobic destruction can only occur
microbes. Each, in turn, will be part of a complex food web, cometabolically with the degradation of a more nutritious
and optimizing the local ecosystem may be as important as substrate (4) or anaerobically as a terminal electron
focusing solely on the primary degraders or accumulators. acceptor (5). The white-rot fungi offer another variation
Bioremediation always competes with alternative on this theme. These organisms seem to be unique in
technologies for achieving an environmental cleanup goal. their ability to degrade lignin, the structural polymer of
Bioremediation is typically among the least expensive higher plants. They may not gain any direct energetic
options, with the additional important advantage that benefit from lignin degradation, but it clearly allows them
in many cases bioremediation is a permanent solution access to cellulose, which is a substrate for their growth.
to the contamination problem because the contaminant Lignin degradation is catalyzed by a group of extracellular
is completely destroyed or sequestered. Some alternative peroxidases that generate nonspecific oxidants, and there
694 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES

have been several proposals to use these systems for ratio of approximately 30 : 30 : 30 : 10. Most crude oils
destroying contaminating organic compounds (6). contain hydrocarbons ranging in size from methane to
With successful bioremediation, organic compounds molecules with hundreds of carbons, although the lightest
are eventually converted to carbon dioxide, water, and molecules are usually absent in oils that have been par-
biomass. Similarly, nitrogenous molecules, such as excess tially biodegraded in their reservoir. When crude oils reach
ammonia or nitrate in groundwater, can be converted the surface environment the lighter molecules evaporate,
to gaseous nitrogen. Other nonorganic contaminants and are either destroyed by atmospheric photooxidation or
provide a different challenge for bioremediation. A few, are washed out of the atmosphere by rain, and are biode-
such as mercury and selenium, are volatized by some graded. Some molecules, such as the smaller aromatics
biological processes, but it is not clear that this is always (benzene, toluene, etc.), have significant solubilities and
beneficial. In some cases, such as chromium and arsenic, can be washed out of floating slicks, regardless of whether
there is a dramatic difference in environmental toxicity these are at sea or on terrestrial water tables. Fortunately
depending on the redox state of the contaminant. In the majority of molecules in crude oils, and refined prod-
these cases, bioremediation has sometimes focused on this ucts made from them, are biodegradable, at least under
detoxification by bacterial processes. A more satisfying aerobic conditions.
approach would be to use a biological process to accumulate The other major sources of hydrocarbon contamination
and concentrate the contaminant so that it can be removed were the town gas and wood treatment industries. Town
for safe disposal; fungi, algae, and higher plants have all gas was made locally in many towns and cities by heating
been used in these efforts. coal in the absence of air. Methane was produced and
used, but the residue included coke and tarry material
Technology rich in polycyclic aromatic hydrocarbons. The production
and use of creosote was quite similar, and again it gave rise
Successful bioremediation hinges on the effective appli-
to contamination with polycyclic aromatic hydrocarbons.
cation of the biology discussed above. Sometimes the
The 16 polycyclic aromatic hydrocarbons on the U.S. EPA
contaminant is on the surface, so access to it is reasonably
priority pollutant list are shown in Figures 2 and 3;
simple. In these cases the required technology may be as
all are abundant in creosote and coal tar, but only
simple as broadcast spreaders or sprayers to apply fertil-
naphthalene, fluorene, phenanthrene, and chrysene are
izers, or tilling the soil to allow good aeration. Of course
relatively abundant in petroleum.
this is not necessarily as simple as it sounds, since con-
taminated sites are often very different from agricultural
Biodegradation
fields, and the technology has to be significantly stronger
to ‘‘plow’’ the soil. Frequently the contaminant is below the Methane and the volatile plant terpenes are fully
surface, and applying even simple bioremediation strate- biodegradable by aerobic organisms, and most refined
gies provides quite a challenge. Table 1 lists some of the petroleum products are essentially completely biodegrad-
technologies in use today. able under aerobic conditions. Estimates for crude oil
This article addresses bioremediation in its broadest biodegradability range up to 90% (7), and the least
sense. The main sections address organic and then biodegradable material, principally polar molecules and
inorganic compounds. Key classes of contaminants are asphaltenes, lack the ‘‘oily’’ feel and properties that we
discussed, and for each I focus on the contaminants that associate with ‘‘oil.’’ Indeed they are essentially impossible
can be treated, the underlying biological processes that to distinguish from more recent organic material in soils
can mitigate the contamination, and the technologies that and sediments, such as the humic and fulvic acids, and
have been used, or are being developed, to treat them. appear to be biologically inert.
Numerous bacterial and fungal genera have species
that can degrade hydrocarbons aerobically, and the
ORGANIC CONTAMINANTS
pathways of degradation of representative aliphatic,
naphthenic, and aromatic molecules have been well
Hydrocarbons
characterized in at least some species (7). It is a truism
Constituents. Hydrocarbons are ubiquitous in the envi- that the hallmark of an oil-degrading aerobic organism is
ronment, from fossil, biogenic, and anthropogenic sources, its ability to insert oxygen atoms into the hydrocarbon, and
and they provide readily digestible food for oil-degrading there are many ways in which this is achieved (Figs. 4 and
microbes (7). Bioremediation is thus an important poten- 5). Once a hydrocarbon possesses a carboxylate or alcohol
tial technology for remediating hydrocarbons in oil spills functionality it is almost invariably a readily degradable
and at contaminated creosote and town gas plants. Crude compound that can eventually be attached to coenzyme
oils are very complex mixtures, primarily of hydrocar- A and mineralized in the ß-oxidation pathway. A simple
bons, although some components do have heteroatoms example at the human level is the difference between
such as nitrogen (e.g., carbazole) or sulfur (e.g., dibenzoth- oleic acid, present in olive oil, and octadecane, present in
iophene). Chemically, the principal components of crude mineral oil, which is so inert that it serves as an intestinal
oils and refined products can be classified as aliphatics, lubricant!
aromatics, naphthenics, and asphaltic molecules, and rep- For many years it was assumed that oil biodegradation
resentative examples are shown in Figure 1. The ratio was an exclusively aerobic process as any degradation
of these different classes varies in different oils, but a must involve oxidation. Indeed, the very existence of
‘‘typical crude oil’’ might contain the four classes in a oil reservoirs indicates that anaerobic degradation in
Table 1. Some Technological Definitions Relevant to Bioremediation

Technology Description

Air sparging, Aquifer Injection of air to stimulate aerobic degradation. May also stimulate volatilization.
sparging, Biosparging
Air-stripping Injection of air to stimulate volatilization. Contaminant is usually removed by subsequent adsorption.
Aquifer bioremediation In situ bioremediation in an aquifer, usually by adding nutrients or cosubstrates through injection wells.
Aquifer sparging Injection of air into a contaminated aquifer to stimulate aerobic degradation. May also stimulate
volatilization.
Batch reactor A bioreactor loaded with contaminated material, and run until the contaminant has been consumed. It
is then emptied, and the process is repeated.
Bioactive Barrier, A zone, usually subsurface, where biodegradation of a contaminant occurs so that no contaminant
Bioactive Zone, passes the barrier. Sometimes impermeable walls funnel contaminants to this reactive zone.
Biowall, Wall-and-gate
Bioaugmentation Addition of exogenous bacteria with defined degradation potential (or rarely indigenous bacteria
cultivated in a reactor and reapplied).
Biofilm Reactor A reactor where bacterial communities are encouraged on a high surface area support. Biofilms often
have a redox gradient so that the deepest layer is anaerobic while the outside is aerobic, allowing both
aerobic and anaerobic processes to occur.
Biofiltration Usually, an air filter with degrading organisms supported on a high surface area support, such as
granulated activated carbon or compost.
Biofluffing Augering soil to increase porosity.
Bioleaching Extracting metallic contaminants at acid pH, perhaps while attempting to optimize fungal degradation
of organic contaminants.
Biological fluidized bed, Bioreactor where the fluid phase is moving fast enough to suspend the solid phase as a fluidlike phase.
Fluidized bed
bioreactor
Biological Plug In-ground actively aerated bioreactors containing adapted microbial consortia to degrade contaminants
of concern.
Biopile, Soil heaping An engineered pile of excavated contaminated soil, with engineering to optimize air, water, and nutrient
control.
Bioslurping Vacuum extraction of the floating contaminant and water, and vapor from the vadose zone. The airflow
stimulates biodegradation.
Biostimulation Optimizing conditions for the indigenous biota to degrade the contaminant.
Biotransformation The biological conversion of a contaminant to some other form, but not to carbon dioxide and water.
Biotrickling Filter A reactor where a contaminated gas stream passes up a reactor with immobilized microorganisms on a
solid support, while nutrient liquor trickles down the reactor.
Bioventing Vacuum extraction of contaminant vapors from the vadose zone, thereby drawing in air to stimulate the
biodegradation of the remainder.
Borehole bioreactor The addition of nutrients and electron acceptor to stimulate biodegradation in situ in a contaminated
In well bioreactor aquifer.
Closed-loop Groundwater recovery, a bioreactor, and low-pressure reinjection to maximize nutrient use, and
Bioremediation maintain temperature in cold climates.
Composting Addition of biodegradable bulking agent to stimulate microbial activity. Optimum composting generally
involves self-heating to 50-60° C.
Constructed Wetland Artificial marsh for bioremediation of contaminated water.
Continuous stirred tank A bioreactor that is completely back mixed.
reactor (CSTR)
Digester Usually, an anaerobic bioreactor that generates methane.
Ex-situ bioremediation Usually, the bioremediation of excavated contaminated soil in a biopile, compost system, or bioreactor.
Fixed-bed bioreactor Bioreactor with immobilized cells on a packed column-matrix.
Land-farming, land Application of a biodegradable sludge as a thin layer to a soil to encourage biodegradation. Tilling and
treatment fertilizing is usually required.
Natural Attenuation, Unassisted biodegradation of a contaminant.
Intrinsic
bioremediation
Pulsed bioremediation Alternating injection of a cosubstrate for the contaminant, and oxygen.
Pump and treat Pumping groundwater to the surface, treating, and reinjecting.
Rhizosphere Stimulating the bacteria in the rhizosphere (root-zone) to carry out biodegradation of a soil contaminant.
bioremediation
Rotating biological Bioreactor with rotating device that moves a biofilm through the bulk water phase and the air phase to
contactor stimulate aerobic degradation.
Sequencing Batch Periodically aerated solid phase or slurry bioreactor.
Reactor
Soil vapor extraction Vacuum assisted vapor extraction.

695
 octane iso-octane toluene
2,2,4-trimethylpentane

S N
H
dibenzothiophene carbazole
O

HO

a naphthenic acid

17α(H),21β(H)-hopane

Figure 1. Some petroleum hydrocarbons. As in all figures in

this article, hydrogens are omitted unless on a heteroatom.
Octane is an example of a paraffin, iso-octane an isoparaffin,
toluene is an aromatic. Benzothiophene and carbazole are S
heterocycles, but they are considered aromatics by the tetralin
oil industry, as is tetralin. Hopane is an example of a 1,2,3,4-tetrahydronaphthalene
biomarker that is recognizably derived from bacterial lipids,
in this case from bacteriohopanetetrol. Naphthenic acids
and asphaltenes are hard to characterize, but the structures
shown are likely representatives. an asphaltene

naphthalene acenaphthene acenaphthylene

anthracene phenanthrene fluorene

chrysene pyrene fluoranthene

Figure 2. Some of the 16 polycyclic aromatic hydrocarbons

on the U.S. EPA list of Priority Pollutants. Only naphthalene,
benz[a]anthracene
phenanthrene, and chrysene are abundant in crude oils, the
others are abundant in coal tars and other thermally treated
materials. dibenz[a,h]anthracene

696
 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES 697

HO O −O
O

R R R R
Linear alkanes
benzo[b]fluoranthene benzo[k]fluoranthene HOO −OO O
O O−
O

benzo[a]pyrene R R R R R
benzo[g,h,i]perylene
Linear alkanes
O−
O
O O

O
indeno[1,2,3-cd]pyrene
Figure 3. More of the 16 polycyclic aromatic hydrocarbons on the
U.S. EPA list of priority pollutants. These are present in only tiny O
amounts in crude oils, but they are generated during incomplete Cycloalkanes −O
combustion of carbonaceous materials, and can be abundant at
Figure 4. Schematics of the initial aerobic biodegradation steps
coal gasification sites.
for alkanes and cycloalkanes. The upper pathway is found in
Pseudomonas oleovorans, the middle in Acinetobacter species,
and the lower in Nocardia species.
such environments must be very slow. Nevertheless, in
recent years it has become clear that bacteria under
completely anaerobic conditions can oxidize many hydro- it seemed that the compound was completely resistant to
carbons (7). Hydrocarbon biodegradation has now been biodegradation, but complete mineralization has now been
shown under sulfate-, nitrate-, chlorate-, perchlorate-, reported (11,12).
carbon dioxide- and ferric iron-reducing conditions (8).
Pathways have not been characterized in detail, but cur- Bioremediation
rent research suggests that a common motif in anaerobic
hydrocarbon degradation is the addition of the hydrocar- Crude oil and refined products are readily biodegradable
bon to the double bond in fumarate (Fig. 6) to produce under aerobic conditions, but they are only incomplete
alkyl- or aromatic-succinates (9,10). These are readily substrates since they lack any significant nitrogen,
degraded after the addition of Coenzyme A. The anaer- phosphorus, and essential trace elements. Bioremediation
obic biodegradation of hydrocarbons, particularly of small strategies for removing large quantities of hydrocarbon
soluble aromatics such as toluene, is now recognized as must therefore include supplementation with nutrients in
an important part of the natural attenuation of hydrocar- a bioavailable form.
bon spills, and is being exploited in active bioremediation
protocols on a large scale. Air. Airborne hydrocarbon vapors are readily treated
While the majority of molecules in crude oils and with biofilters. These are typically rather large devices
refined products are hydrocarbons, the U.S. Clean Air Act with a very large surface area provided by bulky
amendment of 1990 mandated the addition of oxygenated material such as shredded compost (see Biofiltration, this
compounds to gasoline in many parts of the United States. Encyclopedia). The contaminated air, perhaps from a soil
The requirement is typically that 2% (w/w) of the fuel be vapor-extraction treatment or a factory using hydrocarbon
oxygen, which requires that 5 to 15% (v/v) of the gasoline solvents, is blown through the filter, and organisms,
be an oxygenated additive (e.g., methanol, ethanol, methyl usually indigenous to the filter material or provided by
tertiary butyl ether (MTBE) etc.). While methanol and a soil or commercial inoculum, grow and consume the
ethanol are readily degraded, the degradability of MTBE hydrocarbons. Adequate moisture and nutrients must be
remains something of an open question. The compound maintained for effective operation. Alternatively, trickling
was previously very rare in the environment, but now biofilters with recycled water can be used. Bacteria and
it is one of the major chemicals in commerce. At first fungi readily colonize such filters, and they can be very
698 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES

OH OH
OH OH OH
O

O−
O
Naphthalene

OH

OH

OH

OH

OH

OH

Toluene
Figure 5. Schematics of the initial biodegradation steps for naphthalene and toluene. The upper
pathway is found in many Pseudomonads, and the lower panel illustrates the many ways that
toluene is activated in different Pseudomonads.

R O− R effective. Nevertheless, biofilters are usually equipped

O with a small granulated activated carbon ‘‘backup’’ filter
to handle any sudden pulse loads that might overwhelm
the biological capacity of the filter. Biofilters compete with
+ granulated activated carbon filters, and are often less
O O expensive because they minimize the cost of the granulated
−O O− activated carbon and the energy required to destroy the
contaminant and the granulated activated carbon when
the latter is saturated (13). Potential problems include
plugging and uneven air or water flow, but successful
O
designs work for many years with minimal maintenance
Alkane O− except for the occasional addition of nutrients and stirring
O of the bed.
O−
O O− Sea. Crude oil spills at sea are perhaps the most
widely covered environmental incidents in the national
O−
+ and international media. Despite their notoriety, catas-
trophic tanker spills and well blowouts are fortunately
O O
rather rare, and their total input into the world’s oceans is
−O approximately equivalent to that from natural seeps; sig-
Toluene nificantly more oil reaches the world’s oceans from munic-
Figure 6. The initial reactions of anaerobic biodegradation of ipal sewers (14). Physical collection of the spilled oil is the
alkanes and toluene. In both cases the substrate is activated by preferred remediation option, but if skimming is unable
addition of fumarate. to collect the oil, biodegradation, and perhaps combustion
 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES 699

or photooxidation, are the only routes for elimination of Areas where there are currently few remediation
the spill. One approach to stimulating biodegradation is options include oiled marshes, mangroves, and coral reefs.
to disperse the oil with chemical dispersants (15). Patents These environments are generally easily damaged by
have been issued for dispersant formulations that specifi- human intrusion, and physical cleaning options may not
cally include nitrogen and phosphorus nutrients, but the provide any net environmental benefit. Bioremediation
products are not currently commercially available. may provide some attractive options, and some success has
Bioremediation by the addition of oil-degrading been claimed, on a small scale, with fertilizer applications.
microbes is often promoted as a treatment option for float- Marshes and mangroves offer the additional complica-
ing spills, but this approach has not yet met with any tion that they are typically anoxic. Perhaps the anaerobic
documented success (16). degradation of oil could be stimulated by inoculation
with anaerobic hydrocarbon-degrading microbes or per-
Shorelines. The successful bioremediation of shorelines haps gentle aeration or the addition of slow-release oxygen
affected by the spill from the Exxon Valdez in Prince compounds, such as some inorganic peroxides, might stim-
William Sound, Alaska, was the largest bioremediation ulate aerobic degradation without significantly changing
project to date; more than 111 km of shoreline was the redox balance of these environments. This is an area
treated (17), usually after bulk oil had been collected. where research is very much in its infancy, and there are
Residual oil had typically penetrated into the shoreline no well-documented success stories to date.
gravel, occasionally getting as deep as 30 cm into the Bioremediation also offers options for dealing with
sediment. Since the gravel was typically very permeable, oiled material, such as seaweed, which gets stranded on
oxygen availability was unlikely to be the limiting factor shorelines; composting has been shown to be effective.
for biodegradation, and indeed this was subsequently
shown to be correct. Bioremediation therefore focused Groundwater. Spills of refined petroleum products on
on the addition of nitrogen and phosphorus fertilizers land and leaking underground storage tanks sometimes
to partially alleviate the nutrient-limitation of oil contaminate groundwater. Bioremediation is becoming an
degradation. Of course this was complicated by the increasingly popular treatment for such situations.
fact that oiled shorelines were washed by tides twice Hydrocarbons typically have a density of less than 1,
a day. These tides would have rapidly removed any and refined products usually float on the water table
soluble fertilizer, so a strategy was sought that would if they penetrate soil that deeply. In the parlance of
provide nutrients for a significant length of time. Various the remediation industry, such floating spills are often
approaches to applying fertilizers were tried, including called NAPLs (nonaqueous phase liquids). Indeed they
standard and slow-release nutrients, oleophilic nutrients, are sometimes known as LNAPLs (light nonaqueous
and solutions of liquid fertilizers. Two fertilizers were phase liquids) to distinguish them from more dense
used in the full-scale applications; one, an oleophilic materials such as halogenated compounds. While stand-
product known as Inipol EAP22 (CECA, Paris, France), alone bioremediation is an option for these situations,
is a microemulsion of a concentrated solution of urea ‘‘pump and treat’’ is the more usual treatment, in which
in an oil phase of oleic acid and trilaurethphosphate, contaminated water is brought to the surface, free product
with butoxyethanol as a cosolvent. This product was is removed by flotation, and the cleaned water is reinjected
designed to adhere to oil and to release its nutrients into the aquifer. Adding a bioremediation component to
to bacteria growing at the oil-water interface. The the treatment, typically by adding oxygen and low levels
other fertilizer is a slow-release formulation of inorganic of nutrients, is an appealing and cost-effective way of
nutrients, primarily ammonium nitrate and ammonium stimulating the degradation of the residual hydrocarbon
phosphate, in a polymerized vegetable oil skin. This that was not extracted by the pumping, and this approach
product, known as Customblen (Grace-Sierra, Milpitas, is becoming widely used (18).
Calif.), released nutrients with every tide, and these were Hydrocarbons are not very soluble in water, but
distributed throughout the oiled zone as the tide fell. the most soluble components will leach out of a spill
Fertilizer application rates were carefully monitored so if there is continual flushing. Typically, only small
that the nutrients would cause no harm, and the rate aromatic molecules, the infamous BTEX (Fig. 7), are
of oil biodegradation was stimulated between two- and soluble enough to contaminate groundwater, although
fivefold (17). with the advent of oxygenated gasolines, these oxygenates
Subsequent work has shown that fertilizer application [ethanol, methanol, MTBE (methyl tertiary butyl ether),
is likely to be a useful remediation approach in many etc.] are being found in groundwater (11,12). Remediation
situations, including in the Arctic. It is unlikely though of such situations usually involves ‘‘pump and treat’’
that the addition of fertilizers will have a dramatic effect methodologies, although these methods are slow and may
on situations in which agricultural or municipal runoff leave behind reservoirs of contaminants in pockets that
maintains elevated levels of nutrients, as happens in some are poorly connected to the main water body. Of course the
estuaries and bays. Here, aeration is likely to be most contaminant is biodegradable, and some biodegradation
effective. is probably already occurring when the contamination
Bioremediation by the addition of oil-degrading is discovered. The simplest approach to remediation is
microbes has been promoted as a treatment for oiled thus to allow this intrinsic process to continue. Evidence
shorelines, but the approach has not yet met with any that it is indeed occurring can be found in the selective
well-documented success (16). disappearance of the most biodegradable compounds in the
700 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES

stimulating the anaerobic biodegradation of hydrocarbons

will become a broadly useful remediation technology.
If there are significant amounts of volatile and
nonvolatile contaminants, remediation may be achieved by
a combination of liquid and vapor extraction of the former,
and bioremediation of the latter. This combination has
benzene toluene ethylbenzene been termed ‘‘bioslurping,’’ where the act of pumping out
the liquid contaminant phase draws in air at other wells to
stimulate aerobic degradation (25). Such bioremediation
requires that there be enough nutrients to allow microbial
growth, and fertilizer nutrients are frequently added at
the air injection wells. Bioslurping has had a number of
well-documented successes (18).
m-xylene p-xylene
Where there are large volumes of contaminated water
o-xylene
under a small site, it is sometimes most convenient to
Figure 7. The principal water-soluble components of refined treat the contaminant in a biological reactor at the surface.
fuels; benzene, toluene, ethylbenzene, and the xylenes, commonly Considerable research has gone into reactor optimization
referred to as BTEX.
for different situations, and a variety of stirred reactors,
fluidized bed reactors, and trickling filters have been
contaminant mixture, and the concomitant disappearance developed (21). Such reactors are usually much more
of electron acceptors from the groundwater. Thus, oxygen efficient than in situ treatments, although correspondingly
is depleted as the preferred terminal electron acceptor for more expensive.
metabolism, followed by nitrate, ferric iron, sulfate, and Of course, the presence of a liquid phase of hydrocarbon
finally carbon dioxide for methanogenesis (19). in soil gives rise to vapor contamination in the vadose
Intrinsic bioremediation is becoming an acceptable zone above the water table. This can be treated by vacuum
option in locations where the contaminated groundwater extraction, and the passage of the exhaust gases through
poses little threat to environmental health (20). Never- a biofilter (mentioned earlier) can be an inexpensive and
theless, while it is appealingly simple, it may not be effective way of destroying the contaminant permanently.
the lowest cost option if extensive monitoring and doc-
umentation is involved. In such cases it may well be Soil. Hydrocarbon contamination of soils runs the
more cost-effective to optimize conditions for biodegrada- gamut from crude oils at production wells and pipeline
tion. One way to do this is to funnel the contaminated spills to refined products at refineries, distribution
water through a reactive zone where biodegradation can centers, service stations and accident sites, and polycyclic
be stimulated and monitored. The simplest intervention aromatic hydrocarbons at wood-treatment facilities, town
is a line of wells for ‘‘pump and treat,’’ but including gas manufacturing sites, railroads, and their terminals,
a biological component may be more cost-effective. This and various military bases. Sometimes the contamination
can range from the installation of a sparge line for aer- is the result of leaking underground storage tanks
ating the contaminated plume (18), to installing some and pipelines, leading to subsurface contamination, but
form of semi-contained bioreactor where nutrients can surface spills also occur. Physical removal of gross
be applied with some modicum of control (21,22). Often contamination is an obvious first step at all locations,
these designs are combined with barriers to ensure that and bioremediation is an appealing option for remediating
the entire contaminated plume passes through the reac- residual contamination in many of these sites (18,30).
tive zone. These designs have a variety of names, including Spills from production facilities and pipelines often
‘‘biowall,’’ ‘‘trench biosparge,’’ ‘‘funnel and gate,’’ ‘‘bubble involve both oil and brine, since most oil reservoirs float
curtain,’’ ‘‘sparge curtain,’’ and ‘‘engineered trenches and on top of concentrated brines and both are produced in
gates’’ (22). Both aerobic and anaerobic designs have been later stages of production. The brine is typically separated
successfully installed. Another approach is to optimize from the oil and reinjected into the reservoir, but some is
the levels of electron acceptors in the contaminant plume. retained in many production pipelines. The environmental
Oxygen can be pumped in as the pure gas (23), or as air, impact of spilled brine can be quite deleterious. Salt is not
although this is relatively energy intensive since oxygen only toxic to most plants and inhibitory to many soil
is so poorly soluble. Hydrogen peroxide has been used in bacteria, but can also affect soil structure by altering
some situations (24), but there have been problems with the physical properties of clays. Successful bioremediation
biomass plugging near the injection wells. Slow-release strategies must therefore include remediating the brine.
formulations of inorganic peroxides, such as magnesium In wet regions, salt is eventually diluted by rainfall, but
peroxide, have recently been used with success (25,26). in arid regions and to speed the process in wetter regions,
Nitrate may be added, although there are sometimes reg- gypsum is often added to restore soil porosity (31).
ulatory limitations on the amount of this material that Many hydrocarbons bind quite tightly to soil compo-
may be added to groundwater (27). Ferric iron availability nents and are thereby less available to microbial degrada-
may be manipulated by adding ligands (28), and recent tion. The kinetics of binding seem to be complex, and the
success at stimulating the biodegradation of benzene in process of ‘‘aging’’ is only poorly understood. Nevertheless,
groundwater by the addition of sulfate (29) suggests that it seems clear that hydrocarbons that have been in contact
 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES 701

with soil for a long time are not as available for biodegra- are usually the most expensive bioremediation option
dation as fresh spills (30). Several groups of researchers because of the large power requirements, but under some
have suggested the addition of surfactants to overcome conditions this cost is offset by the rapid biodegradation
this limitation, but this approach is not yet widely used. A that can occur.
significant potential concern is that the surfactant may be In all these cases, it is important to bear in mind
degraded in preference to the contaminant of concern. that while the majority of hydrocarbons are readily
Intrinsic biodegradation occurs, but it usually only biodegraded, some, such as the steranes and hopanes,
removes the lightest refined products, such as gasoline, are very resistant to microbial attack. Estimates of oil
diesel, and jet fuel (20). Active intervention is typically biodegradation range from 60 to 95% for different crude
required. Usually, the least expensive approach is in oils, so fresh spills of crude oils are readily treated by
situ remediation (18,30), typically with the addition of bioremediation (7). Refined products, such as gasoline,
nutrients and the attempted optimization of moisture diesel, jet fuels, and heating oils, are usually more
and oxygen by tilling. Various approaches to applying biodegradable than typical whole oils, but the various
fertilizers have been tried, including standard and slow- heavy fractions of crudes, such as the asphalts, are far
release nutrients, oleophilic nutrients, and solutions of less biodegradable and are not such attractive targets
liquid fertilizers. Oxygen is likely to be limiting in for bioremediation. Some crude oils have already been
many cases, and soil tilling is widely practiced. This in extensively biodegraded in their reservoirs, and these
situ bioremediation of hydrocarbon-contaminated soils is are also poor targets for bioremediation. An example is
akin to the old practice of ‘‘land-farming’’ (32), wherein Orimulsion, a heavy oil in water emulsion (70% bitumen)
sludges and other refinery wastes were deliberately stabilized by low levels of surfactants, used as a fuel
spread onto the soil and tilled and fertilized to stimulate for electricity generation. Similarly, old spills may have
biodegradation. Although this practice is now discontinued already undergone significant biodegradation and the
in the United States, it was quite widely used. residue may be relatively biologically inert. It is thus
Deeper contamination may be remedied with biovent- important to run laboratory studies to ensure that the
ing, by which air is injected through some wells and contaminant is sufficiently biodegradable and that cleanup
extracted through others to both strip volatiles and provide targets can be met. On the other hand, even very heavy
oxygen to indigenous organisms (18). Fertilizer nutrients fuels contain some biodegradable compounds, and a case
may also be added. This is usually only a viable option can be made that stimulating the biodegradation of this
with lighter refined products. fraction will minimize the environmental impact of these
When soil contamination extends to some depth, it may compounds, even if it will not have a dramatic effect on
be preferable to excavate the contaminated soil and to put the total amount of oil in the environment.
it into ‘‘biopiles’’ in which oxygen, nutrient, and moisture
levels are more easily controlled (21). Biopiles can also be HALOGENATED ALIPHATIC COMPOUNDS
kept warm during winter months, increasing the amount
of time available for biodegradation in colder climates. Constituents. Halogenated organic compounds are
Since the soil is well mixed during the construction of widespread in nature, and some, such as dichloromethane,
the pile, there is an opportunity to add selected microbial are produced on an enormous scale (by fungi) and serve
and fungal strains in an additional attempt to maximize as carbon and energy sources for diverse microbes (33).
biodegradation. Synthetic chlorinated solvents, especially chloroalkanes
Composting by the addition of readily degradable and chloroalkenes, have become widely used in industrial
bulking agents is also a useful option for relatively processes in the second half of the twentieth century,
small volumes of excavated contaminated soil (21). Since and some representative examples are shown in Figure 8.
efficient composting invariably involves self-heating as These solvents have been used for more than 50 years,
biodegradation proceeds, this also offers an option for
extending the bioremediation season into the winter
months in cold climates. A potential drawback of Cl Cl
composting is that it usually increases the volume of
contaminated material, but if fully successful, the finished Cl
Cl Cl
compost can be returned to the site as a positive Cl Cl tetrachloroethylene
contribution to soil quality. (perchloroethylene)
Cl
Slurry bioreactors offer the most aggressive approach carbon tetrachloride
to maximizing contact between the contaminated soil and Cl Cl
the degrading organisms (21). Both lagoons and reactor
vessels have been used, but the former is often not Cl Cl
optimally designed for all the soil to be partially suspended trichloroethylene
Cl
by the mixing impellers. Contained reactor designs
Cl
include mixing tank, airlift, and fluidized-bed aeration.
1,1,1-trichloroethane Cl
A major advantage of contained slurry bioreactors is the
potential to continually optimize nutrients, aeration, and vinyl chloride
degradative inocula as fresh soil is added, and to control
waste materials, including gases. Slurry bioreactors Figure 8. Some halogenated solvents.
702 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES

Cl Cl Cl
Cl Cl Cl Cl Cl CH4
Cl Cl

H2 HCl H2 HCl H2 HCl H2 HCl

Carbon tetrachloride

Cl Cl Cl Cl Cl Cl Cl

Cl Cl Cl

H2 HCl H2 HCl H2 HCl H2 HCl

Figure 9. Sequential dehalogenation of carbon tetrachloride
and tetrachloroethylene under anaerobic conditions. Tetrachloroethylene

and unfortunately they are now fairly widespread con- Cl Cl Cl O Cl

taminants. Unlike the hydrocarbons, which usually float
on water, the halogenated solvents typically have densi- Cl Cl
ties greater than one, and generally sink to the bottom of O2 H2O
any groundwater and float on the bedrock. For this reason, Monooxygenase
they are sometimes known as DNAPLs (dense nonaqueous
Cl Cl OH OH
phase liquids).
Cl Cl
Biodegradation Cl
Cl Cl
O2
Halogenated solvents are degraded under aerobic and
Dioxygenase
anaerobic conditions. The anaerobic process is typically
a reductive dechlorination that progressively removes Figure 10. Comparison of the attack of monooxygenase and
one halide at a time (Fig. 9). For example, carbon dioxygenase enzymes on trichloroethylene.
tetrachloride is sequentially dechlorinated to chloro-
form, dichloromethane, methyl chloride, and methane,
and trichloroethylene is sequentially reduced to ethy- The biodegradation of trichloroethylene (trichloro-
lene (34). The discovery of reductive dechlorination has ethene) has been the most studied since this is probably the
led to halorespiration-based remediation technologies in most widespread halogenated solvent contaminant. Sev-
which carbon sources are added to contaminated sites. eral substrates drive trichlorethylene co-oxidation, includ-
Chloromethane and dichloromethane have been shown to ing methane, propane, propylene, toluene, isopropyl-
be the sole carbon source for several anaerobic organ- benzene, and ammonia (36). The enzymes that metabolize
isms (35), and it seems that we have much to learn about these substrates have subtly different selectivities with
the microbial diversity of anaerobic microorganisms capa- regard to the halogenated solvents, and to date, none
ble of dechlorinating solvents. are capable of co-oxidizing carbon tetrachloride or tetra-
The simplest chlorinated alkanes, alkenes, and alcohols chloroethylene. Complete mineralization of these com-
(e.g., chloromethane, dichloromethane, chloroethane, 1,2- pounds can, however, be achieved by sequential anaerobic
dichloroethane, vinyl chloride, and 2-chloroethanol) serve and aerobic processes.
as substrates for aerobic growth for some bacteria, but the
Bioremediation
majority of halogenated solvents do not seem to be able
to support growth. Nevertheless, these compounds can Air. Biofilters are an effective way of dealing with air
be mineralized under aerobic conditions, albeit with no from industrial processes that use halogenated solvents
apparent benefit to the degrading organism. Indeed, the such as chloromethane, dichloromethane, chloroethane,
oxidation appears to be fortuitous, only occurring during 1,2-dichloroethane, and vinyl chloride, and that support
the metabolism of a growth substrate. The phenomenon aerobic growth (37). Both compost-based ‘‘dry’’ systems
is therefore known as co-oxidation or cometabolism. and trickling filter ‘‘wet’’ systems are in use. Similar filters
Numerous bacteria are able to catalyze the oxidation could be incorporated into ‘‘pump and treat’’ operations.
of trichloroethylene; some use monooxygenases (e.g.,
methane and ammonia oxidizing species) while others Groundwater and soil. Soils and groundwater through-
contain dioxygenases (e.g., some toluene oxidizing species). out the industrialized world have been contaminated with
The difference between these two classes of enzymes is the halogenated solvents, and remediation has a high prior-
fate of the two atoms of molecular oxygen. Monooxygenases ity. Since the solvents are so dense, they are typically
insert one oxygen atom into their substrate and reduce found on the bedrock underlying aquifers. Pumping out
the other to water. Dioxygenases insert both atoms into the liquid phase is an obvious first step if the contami-
their substrate (Fig. 10). In either case, biodegradation nant is likely to be mobile, but in situ bioremediation is a
can proceed to complete mineralization. promising option (38). For example, the U.S. Department
 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES 703

of Energy has used anaerobic in situ degradation of car- OH

bon tetrachloride to treat an aquifer some 76 m below Cl Cl
the Hanford, WA, site, with nitrate as electron acceptor Cln Cln
and acetate as electron donor (39). Trichloroethylene is
the most frequent target of remediation, and as discussed Cl Cl
above, this only seems to be metabolized cometabolically.
Cl
Remediation operations thus incorporate the addition of
pentachlorophenol polychlorinated biphenyl
cometabolized substrate. Methane was used successfully
at the U.S. Department of Energy site at Savannah River,
near Aiken, South Carolina (40). This operation had both O
O OH HO
an air stripping and a biological component. Horizontal
Cl Cl
wells were used to pump methane and air below the con- O− Cl
taminant, and an upper horizontal well in the vadose zone
was used to withdraw these gases through the contami-
nated zone. Optimum biodegradation performance seemed Cl Cl
to come from alternating injection of air and methane in air Cl Cl Cl
and the inclusion of nitrous oxide and triethylphosphate, (2,4-dichlorophenoxy)acetate hexachlorophene
both gases, to give a C : N : P ratio of 100 : 10 : 1. Success
has also been achieved with a bioaugmentation strategy, Cl
using a toluene monooxygenase containing Burkholderia
Cl Cl
cepacia (and no cosubstrate) (41).
‘‘Pump and treat’’ operations can also incorporate
bioremediation; methane has been used as cosubstrate
Cl Cl
in aerobic bioreactors for halohydrocarbons including
chlorofluorocarbons (42) and a sequential methanogenic- DDT
denitrifying bioreactor system has been developed (43).
Figure 11. Some halogenated aromatic hydrocarbons.
Adding a slow release source of reductant to drive the
reductive dechlorination of solvents in groundwater has
also proved to be effective (26). One proprietary product is
Cl Cl
a polylactate ester that slowly hydrolyzes to release lac- Cl O Cl
tate, which in turn is metabolized anaerobically to release
hydrogen. This in turn serves as the reductant of the Cl Cl
chlorinated contaminants. A slow-release form of oxygen, Cl O Cl
O
magnesium peroxide, has also proved effective at stimu- 2,3,7,8-tetrachloro-p-dioxin 2,3,7,8-tetrachlorodibenzofuran
lating the aerobic biodegradation of partially chlorinated
species (26). Reactive barrier technologies, including bio- Figure 12. Dioxin and dibenzofuran.
logical barriers that exploit reductive dechlorination (22),
are widely used for halogenated contaminants.
to biodegradation, so at first glance they may not
HALOGENATED AROMATIC COMPOUNDS seem a good target for bioremediation. Indeed, complex
halogenated organic compounds were widely thought to
Constituents. be almost exclusively anthropogenic in origin so that
Complex halogenated organic compounds have been there would have been little time for biodegradation
widely used in commerce during the second half of pathways to evolve. This view is being corrected, for
the twentieth century. A few representative examples in fact a variety of organisms, particularly marine
are shown in Figure 11; pentachlorophenol has been algae and some fungi, produce significant quantities
widely used as a wood preservative and also for termite of these compounds (44). Furthermore, it is now well
control. (2,4-dichlorophenoxy) acetate (2,4-D) is widely established that polychlorinated dibenzo-p-dioxins and
used as a broadleaf herbicide, DDT is widely used as dibenzofurans (Fig. 12) are produced when biomass (or
an insecticide, and hexachlorophene is widely used as municipal waste) is burnt (45). There is thus good reason
a germicide. Polychlorinated biphenyls (PCBs) were sold to expect that halogenated organic degrading organisms
with varying levels of chlorination for a range of purposes. can be found in the biosphere, and this has been borne out
They ranged from light oily fluids (with 2–4 chlorines) in practice.
through viscous oils (5 chlorines) to greases and waxes (6
or more chlorines) and their names indicated the level Biodegradation
of chlorination. Thus, Aroclor 1242 (Monsanto, U.S.A.),
Kanechlor 300 (Kanegafuchi Chemical Industries, Japan), Adriens presented a thorough description of the microbi-
and Clophen A30 (Farbenfabriken Bayer AG, Germany) ology and technology of the bioremediation of chlorinated
contained 42% chlorine by weight and an average of aromatic compounds in this volume. An important char-
3 chlorines per biphenyl. An important property that led acteristic of degradation is the cleavage of carbon-chlorine
to their broad use is their relative inertness, especially bonds, and the enzymes that catalyze these reactions, the
704 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES

dehalogenases, are being characterized (46). The reduc- but already there are suggestions that there is consider-
tive dechlorination seen with aliphatic chlorides (Fig. 9) able diversity in the enzyme systems able to degrade these
seems to be a general phenomenon, and even compounds compounds.
as persistent as DDT and the polychlorinated biphenyls
are reductively dechlorinated under some conditions, par- Bioremediation
ticularly under methanogenic ones. Some compounds,
Soil. Pentachlorophenol has been the target of bioreme-
such as pentachlorophenol, can be completely mineral- diation at a number of wood-treatment facilities, and good
ized under anaerobic conditions, but the more recalcitrant success has been achieved in several applications (56). It
ones require aerobic degradation after reductive dehalo- is rarely the sole contaminant and is often present with
genation. polynuclear hydrocarbons from creosote. In situ degrada-
Pentachlorophenol can be mineralized aerobically tion has been stimulated by bioventing, wherein air is
by bacteria (47) and fungi (48) and anaerobically by injected through some wells and extracted through others
bacteria (49), and several bacterial isolates are able to to both strip volatiles and provide oxygen to indigenous
grow aerobically using pentachlorophenol as the sole organisms (57).
source of carbon and energy. Many of these grow rather The kinetics of such in situ degradation is rather
well when supplemented with a more nutritious substrate, slow, however, and more active bioremediation is usually
but unfortunately, in many cases it seems that the more attempted. For example, contaminated soil at the
vigorous growth with these substrates does not enhance Champion Superfund site in Libby, MT, was placed
the biodegradation of pentachlorophenol. into 1-acre land treatment units in 6-inch layers, and
(2,4-dichlorophenoxy)acetate (2,4-D) has been one of irrigated, tilled, and fertilized. Under these conditions,
the world’s most popular herbicides, and although it is the half-lives of pentachlorophenol, pyrene, and several
somewhat resistant to biodegradation, it is biodegraded other polynuclear aromatic hydrocarbons, initially present
by several bacterial isolates. It is a general truism at around 100 to 200 ppm, were on the order of
that the more halogens on a molecule, the slower its 40 days (57). This success relied on the indigenous
biodegradation, and this is borne out with the related microbial populations in the soil, but many groups are
herbicide 2,4,5-T ((2,4,5-trichlorophenoxy)acetate). Nev- focusing on the addition of organisms (58). Composting,
ertheless, bacterial degradation has been seen under and bioremediation focusing on the use of white-rot fungi,
both aerobic and anaerobic conditions, the latter involv- has also met with success at the pilot scale. Others have
ing reductive dechlorination via 2,4-D. Aerobic degrada- used fed-batch or fluidized-bed bioreactors to stimulate
tion removes acetate from (2,4-dichlorophenoxy)acetate to the biodegradation of pentachlorophenol. This allows
yield 2,4-dichlorophenol, which is subsequently hydroxy- significant optimization of the process and increases rates
lated to 3,5-dichlorocatechol, followed by ring cleavage and of degradation by 10-fold (57).
complete mineralization (50). There is good evidence that A major concern when remediating wood-treatment
the genes for the degradation of 2,4-D have undergone sites is that pentachlorophenol was often used in com-
interspecies transfer (51). bination with metal salts, and these compounds, such as
DDT (1,1,1-trichloro-2,2-bis(p-chlorophenyl)ethane) is ‘‘chromated-copper-arsenate’’ (CCA), are potent inhibitors
a remarkably resistant molecule, which explains both its of at least some pentachlorophenol-degrading organ-
efficacy as an insecticide and its accumulation at the isms (59). Sites with significant levels of such inorganics
top of the food chain. Nevertheless, there are indications may not be suitable candidates for bioremediation.
that it can be biodegraded, both anaerobically, with The phenoxy-herbicide 2,4-D has been successfully
initial dechlorination (52), and aerobically, with initial bioremediated in soil contaminated with such a high level
ring hydroxylation (50). DDT and its partially degraded of the compound (710 ppm) that it was initially toxic to
congeners are very hydrophobic and biodegradation seems microorganisms (60). There were essentially no indigenous
to be stimulated by adding surfactants. White-rot fungi bacteria in the soil, and success relied on washing a
also degrade DDT under lignolytic conditions, although significant fraction of the contaminant off the soil and
there is little mineralization to carbon dioxide (6). adding bacteria enriched from a less contaminated site.
The polychlorinated biphenyls are quite recalcitrant. Success was achieved in remediating both soil washwater
Some lightly chlorinated biphenyls are readily mineralized and soil in a bioslurry reactor (60). 2,4-D is also effectively
under aerobic conditions (53), and indeed the structure of degraded in composting, with about half being completely
an enzyme that catalyzes the key ring-cleavage oxygena- mineralized, and the other half becoming incorporated
tion has been determined by X-ray crystallography (54). in a nonextractable form in the residual soil organic
More chlorinated congeners are resistant to aerobic degra- matter (61).
dation, but they are reductively dechlorinated under The bioremediation of polychlorinated biphenyls in
anaerobic conditions (55). Complete degradation of the soils is receiving significant attention (53,56) because
commercial mixtures thus generally requires an anaer- these compounds are quite widely distributed in the
obic process, followed by an aerobic one. A major issue environment, either from leaking electrical transformers
for the oxidative process is that biphenyl seems to be or because they were applied as part of road main-
required for significant expression of the biodegradative tenance. In the latter case, the contamination usually
system (53); the chlorinated compounds do not induce the includes petroleum hydrocarbons and unfortunately it
enzymes that would degrade them. The biochemistry of the seems that the two contaminants inhibit each other’s
biodegradative process is only beginning to be unraveled, degradation. Nevertheless, cultures are being found that
 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES 705

can degrade both polychlorinated biphenyls and petroleum 2,5,3’,4’-tetrachlorobiphenyl (65), but whether this is a
hydrocarbons. realistic approach for in situ bioremediation remains to
be seen. Harkness and coworkers (66) have successfully
Groundwater. A successful groundwater bioremedia- stimulated aerobic biodegradation in large caisons in the
tion of pentachlorophenol is being carried out at the Libby Hudson River by adding inorganic nutrients, biphenyl,
Superfund site described earlier. A shallow aquifer is and hydrogen peroxide, but found that repeated addition
present at 5.5 to 21 m below the surface, and a con- of a polychlorinated biphenyl degrading bacterium (Alcali-
taminant plume is nearly 1.6 km in length. Nutrients genes eutrophus H850) had no beneficial effect. Essentially
and hydrogen peroxide were added at the source area no biodegradation occurred in the stirred control cais-
and approximately half way along the plume, and pen- sons, but losses on the order of 40% were seen in the
tachlorophenol concentrations decreased from 420 ppm caissons that received nutrients and peroxide, regardless
to 3 ppm whenever oxygen concentrations were success- of whether the stirring was aggressive or rather gentle.
fully raised. A membrane oxygen dissolution system was Whether this approach can be scaled-up for large scale
installed to replace the initial hydrogen peroxide addi- use, with a net environmental benefit, remains to be seen.
tions, and costs were substantially lowered without an
apparent decrease in remediation performance (62).
NONCHLORINATED PESTICIDES AND HERBICIDES
Pentachlorophenol is readily degraded in biofilm
reactors (63), so bioremediation is a promising option for
Constituents. A vast number of herbicides, pesticides,
the treatment of contaminated groundwater brought to
fungicides, and so on, have been sold, and a few are shown
the surface as part of a ‘‘pump and treat’’ operation.
in Figure 13. In order to be effective, they must have their
effect before they are degraded in the environment, but on
River and Pond Sediments. Much of the work on poly- the other hand they must not be so resistant to degradation
chlorinated biphenyls has focused on the remediation of that they accumulate where they are used or in food chains.
aquatic sediments, particularly from rivers, estuaries, and It is unusual for these compounds to become contaminants
ponds (53). As noted earlier, a few of the most lightly where they are applied correctly, but manufacturing
chlorinated compounds are mineralized under aerobic con- facilities, storage depots, and rural airfields where ‘‘crop-
ditions, but the more chlorinated species seem completely dusters’’ are based have had spills that can lead to
resistant to aerobic degradation, even by white-rot fungi. long-lasting contamination. Bioremediation is a promising
On the other hand, there is extensive dechlorination of technology to remediate such sites. There are also some
highly chlorinated forms under anaerobic, particularly locations where groundwater has become contaminated
methanogenic, conditions. Bioremediation thus requires by these chemicals, and again, bioremediation may be a
anaerobic and aerobic regimes. Intrinsic biodegradation of cost-effective remediation strategy.
polychlorinated biphenyls can be recognized by the chang-
ing ‘‘fingerprint’’ of the individual isomers as biodegrada-
Biodegradation
tion proceeds (53,64). The anaerobic dechlorination of the
most recalcitrant congeners can apparently be ‘‘primed’’ The majority of herbicides, pesticides, fungicides, and
by adding a readily dehalogenated congener, such as insecticides are biodegradable, although the intrinsic

N
H NH N Cl
N
N N
N N
N
O O O
NH2 NH
P
S
Amitrole Atrazine Diazinon

NO2

NO2 O O−

NH

O2N O HO P OH
O O
OH P O

S Figure 13. Some representative nitrogen-containing hydro-

Dinoseb Methyl Parathion Glyphosate carbons.
706 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES

biodegradability of individual compounds is one of the has been achieved with indigenous organisms in labora-
variables used in deciding which compound to use for a tory mesocosms after a lag phase, and activity remained
particular task. Some herbicides are acutely toxic to plants once it was found (73). Interestingly, the degradation was
that absorb them, but are so readily degraded in soil that somewhat slowed by the addition of low concentrations
seeds can be planted at the same time as the herbicide of readily assimilatable carbon, such as lactate, and it is
application. Other herbicides are known to be effective at not clear how biodegradation might be stimulated in the
preventing plant growth for many months, and are used field. Nevertheless, it is clear that intrinsic remediation
for long-term weed control. In some cases, the biodegrada- is likely to lead to the disappearance of atrazine from
tion of a pesticide is so rapid that it loses its efficacy, and groundwaters once atrazine utilizers have become abun-
pesticide rotation strategies must be used. Very few degra- dant, and perhaps inoculation with atrazine-metabolizing
dation pathways of these compounds have been worked out species will be effective.
in detail, but some generalizations can be made. If more active treatment is required, such as ‘‘pump
Compounds with organophosphate moieties, such as and treat,’’ it is possible that biological reactors will be a
diazinon, methyl-parathion, coumaphos, and glyphosate cost-effective replacement for activated carbon filters.
are usually hydrolyzed at the phosphorus atom (50,67).
Indeed, several Flavobacterium isolates are able to grow Soil. Herbicides and pesticides are of course metabo-
using parathion and diazinon as sole sources of carbon. lized in the soil to which they are applied, and there are
Triazines pose rather more of a problem, probably many reports of isolating degrading organisms from such
because the carbons are in an effectively oxidized state sites. Degradation activities are typically much higher at
so that little metabolic energy is obtained by their sites that have seen product application, indicating that
metabolism. Very few pure cultures of microorganisms natural enrichment processes occur. Much current effort
are able to degrade triazines such as atrazine, although is aimed at assessing the diversity of degradative path-
some Pseudomonads are able to use the compound as ways, and in many cases it seems that several different
sole source of nitrogen in the presence of citrate or other natural metabolic pathways can degrade individual pol-
simple carbon substrates. The initial reactions seem to lutants. Little work has yet been presented in which the
be the removal of the ethyl or isopropyl substituents on biodegradation of these compounds has been successfully
stimulated, at field scale, by a bioremediation approach,
the ring (50), followed by complete mineralization of the
but inoculation with active organisms may be a promis-
triazine ring. The nitrogen can be used as nitrogen source
ing approach (74). The addition of genetically engineered
by some organisms (68), and as sole carbon and nitrogen
organisms, albeit a killed recombinant E. coli that overex-
source under anaerobic conditions (69).
presses atrazine chlorohydrolase, significantly increased
Nitroaromatic compounds, such as dinoseb, are
atrazine degradation in a heavily contaminated soil (75).
degraded under aerobic and anaerobic conditions (70).
The nitro group may be cleaved from the molecule as
nitrite or reduced to an amino group under either aero- MILITARY CHEMICALS
bic or anaerobic conditions. Alternatively, the ring may
Constituents. The military use a range of chemicals
be the subject of reductive attack. Thus, while these
as explosives and propellants, ‘‘energetic molecules,’’
molecules are sometimes quite long-lived in the environ-
and by the nature of their use it is no surprise
ment, they can be completely mineralized under appro-
that there are now many areas in need of remedia-
priate conditions (70). Alternatively, inert residues can
tion. Generally speaking, modern explosives are cyclic,
be immobilized; for example, a Clostridium bifermentans
often heterocyclic compounds, composed of carbon, nitro-
anaerobically degrades dinoseb cometabolically in the
gen, and oxygen. Perhaps the most well known is
presence of a fermentable substrate to below detectable
2,4,6-trinitrotoluene (Fig. 14), but RDX (Royal Demoli-
levels, although only a small fraction is actually mineral-
tion explosive; hexahydro-1,3,5-trinitro-1,2,3-triazine, also
ized to carbon dioxide (71).
known as cyclonite) and HMX (High Melting explosive;
Carbamates such as aldicarb undergo degradation
octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine) are even
under both aerobic and anaerobic conditions. In fact, the more powerful. N,N-dimethylhydrazine and perchlorate
oxidation of the sulfur moiety to the sulfoxide and sulfone are used as solid rocket fuels. These compounds are some-
is part of the activation of the compound to its most potent times present at quite high levels in soils and groundwater
form. Subsequent aerobic metabolism can completely min- on military bases and production sites and need to be
eralize the compound, although this process is usually removed. One quite infamous problem at the latter is ‘‘pink
relatively slow so that it is an effective insecticide, acari- water,’’ a relatively undefined mixture of photodegradation
cide, and nematocide. Anaerobically these compounds are products of TNT. Bioremediation is a promising technology
hydrolyzed, and then mineralized by methanogens (72). for treating sites contaminated with such compounds.
Bioremediation may also be an appropriate tool for
Bioremediation dealing with chemical agents such as the mustards and
Groundwater. Atrazine dominated the world herbicide organophosphate neurotoxins (76,77), but little work on
market in the 1980s, and contamination of groundwa- actual bioremediation has been published.
ter has been reported in locations worldwide. There are
several reports that once in groundwater it is very recal- Biodegradation
citrant, suggesting that atrazine-degrading organisms are Natural nitro-substituted organic compounds are quite
not widespread. Nevertheless, successful biodegradation unusual, and it was once thought that their degradation
 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES 707

is proceeding. Rocket fuel-contaminated groundwater

NO2 is also being treated at an experimental scale with
O2N NO2
N bioremediation (83).

N N Soil. Optimum bioremediation of nitro compounds

O2N NO2 seems to require an anaerobic stage to encourage the
NO2
reductive removal of the nitro groups, followed by an
TNT RDX
aerobic stage to encourage the mineralization of the carbon
skeleton (78,56). Composting of soils contaminated by high
O2N
explosives is being carried out at several U.S. Army Depot
N sites, including Umatilla in Oregon, Pueblo in Colorado,
NO2
N and Tooele in Utah (56). Bioslurry reactors are also in
N NH2 use (56).
N
O2N
N
NO2 OTHER ORGANIC COMPOUNDS
HDX N,N-dimethlhydrazine
The majority of organic compounds in commerce are
Figure 14. Some military chemicals.
biodegradable, so bioremedation is a potential option for
cleaning up after industrial and transportation accidents.
was principally by abiotic processes. As we saw above in For example, Flathman and coworkers (84) reported the
the case of dinoseb, however, nitro-substituted compounds successful bioremediation of 23,000 liters of vinyl acetate
are subject to a variety of degradative processes. The spilled from a railroad tank car in Albany, New York, at
biodegradation of TNT is well established (78). Under a cost approximately half that of excavation and disposal
anaerobic conditions it can be reduced to the corresponding of the 1,100 m3 of contaminated soil. They also report that
aromatic amines and subsequently deaminated to toluene; in situ biological treatment has been used to remediate
the latter can be mineralized under anaerobic conditions, spills of other organics, including acrylonitrile, styrene, 2-
leading to the potentially complete mineralization of TNT butoxyethanol, and ethacrylate, at other railroad-accident
in the absence of oxygen. Under aerobic conditions, a sites. Bioremediation is thus already an important tool in
range of bacteria and fungi can mineralize TNT often remediating accidental spills of organic compounds.
cometabolically with the degradation of a more degradable Bioremediation is also an option when spills of
substrate. However, there is also ample evidence that such compounds contaminate groundwater. For example,
under some conditions the TNT is converted to insoluble bioremediation seems a feasible treatment for aquifers
large molecular weight compounds, probably by addition contaminated with phenol (85), alkylpyridines (86), sul-
reactions to soil components such as humic and fulvic folane and diisopropanolamine (87). It has also been used
acids or cellular material in the case of plants. Lignolytic to treat soils contaminated with the plasticizer bis(2-
fungi also yield promising results for the degradation of ethyl-hexyl)phthalate in a bioslurry reactor (88). ‘‘Biolog-
both TNT and ‘‘pink water,’’ particularly if the latter is ical plugs,’’ actively aerated in-ground bioreactors, have
pretreated with UV irradiation. been used to bioremediate a site contaminated with
RDX and HMX are rather more recalcitrant, especially monoethanolamine (89).
under aerobic conditions, but there are promising
indications that biodegradation can occur under some INORGANIC CONTAMINANTS
conditions, especially composting (78). Several strains of
bacteria able to use RDX (and triazine) as the sole source Nitrogen Compounds
of nitrogen for growth have recently been isolated, and
this is an area that is witnessing rapid progress. Constituents. Nitrogen-containing compounds are of
Some work has been reported on the biodegradation concern for several reasons. Nitrate levels are regulated
of dimethylhydrazine (79), and there is a lot of current in groundwater because of concerns for human and
interest in perchlorate as a terminal electron acceptor animal health, while ammonia is regulated in streams
for a range of heterotrophic bacteria (80). Perchlorate and effluents as a potential fish toxicant, and any
remediation is being achieved at the McGregor, Texas, nitrogenous contaminant is a potential problem in water
Weapons plant site with a biological ‘‘reactive barrier because of its stimulatory effect on the growth of algae.
technology’’ (81). Other nitrogenous contaminants include cyanides in
mine waters. Fortunately, all are amenable to biological
Bioremediation treatment (90).
Groundwater. Nitrotoluenes have been detected in
Biodegradation
groundwater in some areas, and intrinsic remediation
seems to occur at some sites by anaerobic degradation. The biological mineralization of fixed nitrogen is well
They are readily degraded in biological fluidized bed studied; ammonia is oxidized to nitrite, and nitrite to
reactors (82), and research into whether this can be nitrate, by autotrophic bacteria, and nitrate is reduced to
stimulated in situ with a net environmental benefit nitrogen by anaerobic bacteria using nitrate as terminal
708 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES

electron acceptor. Urea in sewage and industrial wastes Immobilization and Toxicity Minimization
is readily hydrolyzed to ammonia and carbon dioxide by
Adsorption. Biomass, often agricultural by-products
many bacteria, and cyanides are used as sole sources
but also including microbial products, have been widely
of carbon and nitrogen by many aerobic organisms.
used as adsorbents for metals and other contaminants in
Wastewater treatment facilities utilize these organisms
water (95), but they are outside the scope of this article.
in assuring that municipal and industrial effluents meet
strict water quality standards.
Microbial Processes. Most elements display a range
of solubilities and biological effects depending on their
Bioremediation chemical form. For example, chromium, while it may be
an essential micronutrient for many organisms, is known
Surface Water. One example of exploiting biology to
to be toxic at higher levels. Indeed, chromium has a
handle excess nitrogen in surface water arose at the
range of available redox states that differ significantly
USDoE Weldon Spring Site near St. Louis, MO. This
in their environmental impact. Cr(VI) is generally more
site had been a uranium and thorium processing facility,
soluble than Cr(III), and more toxic, but a wide range
and the processing involved nitric acid. The waste
of organisms, both bacteria and fungi, are able to
stream, known as raffinate, was discharged to surface
reduce the former to the latter under both aerobic and
impoundments and neutralized with lime to precipitate
anaerobic conditions, thereby reducing the environmental
the metals. Two pits had nitrate levels that required
impact of the contaminant. Since Cr(III) precipitates
treatment before discharge, but heavy rains in 1993
as insoluble hydroxides under neutral and alkaline
threatened to cause the pits to overflow. Bioremediation
conditions, microbial reduction is a potential treatment
by the addition of calcium acetate as a carbon source
for soils and waters (96).
successfully treated more than 19 million liters of water at
Similarly, selenium is a micronutrient that is toxic at
a reasonable cost (91). The use of sulfur-based autotrophic
higher levels. It also has quite a different bioavailability
denitrification ponds to remediate nitrate is also under
in its different redox states, with the elemental form being
investigation (92). Here the sulfur acts as electron donor
the least biologically available. Many microorganisms,
to the nitrate, and competing sulfur oxidation by aerobes
both bacteria and fungi, are able to reduce more oxidized
must be controlled by minimizing mixing and aeration.
species, especially selenite (Se(IV)), to the red elemental
form, providing an appealing remediation option for this
Groundwater. One approach to minimizing the envi- element (97). An alternative approach to remediating sele-
ronmental impact of excess nitrogen in groundwater is to nium contamination might be to encourage methylation
add a degradable substrate to the contaminated aquifer, to volatile dimethylselenide. Although the microbiology of
in the absence of oxygen, to stimulate bacterial denitri- this process is not yet very well understood, it seems to be
fication. Soluble compounds such as ethanol and acetate the result of degradation of the selenium containing amino
have been suggested (90), but insoluble substrates such as acid selenomethionine.
vegetable oil (93) or sawdust (94) seem to provide better Arsenic is another element with different bioavailabil-
control, especially for wells. ity in its different redox states. Arsenic is not known
to be an essential nutrient for eukaryotes, but arsen-
ate (As(V)) and arsenite (As(III)) are toxic, with the latter
METALS AND METALLOIDS being rather more so, at least to mammals. Nevertheless,
some microorganisms grow at the expense of reducing
A wide range of metallic and nonmetallic contaminants is arsenate to arsenite (98), while others catalyze the reverse
present at industrial and agricultural sites throughout reaction (99). The element can be immobilized as an insol-
the world, both in ground and surface water and in uble polymetallic sulfide by sulfate-reducing bacteria,
soils. They pose quite a different problem from that of presumably adventitiously because of the production of
organic contaminants, since they cannot be degraded hydrogen sulfide (100). Indeed, many contaminant metal
so that they ‘‘disappear.’’ Some metal and metalloid and metalloid ions can be immobilized as metal sulfides by
elements have radically different bioavailabilities and sulfate-reducing bacteria. Sulfate-reducing bacteria can
toxicities depending on their redox state, so one option immobilize uranium quite effectively, but in this case it is
is stabilization by converting them to their least precipitated as uraninite (UO2 ) (101).
toxic form. This can be a very effective way of A rather more specific mechanism of microbial immobi-
minimizing the environmental impact of a contaminant, lization of metal ions is represented by the accumulation
but if the contaminant is not removed from the of uranium as an extracellular precipitate of hydrogen
environment there is always the possibility that natural uranyl phosphate by a Citrobacter species (102). Stag-
processes, biological or abiological, may reverse the gering amounts of uranium can be precipitated — more
process. Removing the contaminants from water phases is than 900% of the bacterial dry weight! Indeed, even ele-
relatively straightforward, and the wastewater treatment ments that do not readily form insoluble phosphates, such
industry practices this on an enormous scale. ‘‘Pump as nickel and neptunium, may be incorporated into the
and treat’’ systems that mimic wastewater treatments uranyl phosphate crystallites (103). The precipitation is
are already being used for several contaminants and less driven by the production of phosphate ions at the cell
complex systems involving biological mats are a promising surface by an external phosphatase, and while the pro-
solution for less demanding situations. cess requires the addition of a phosphate donor, such as
 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES 709

glycerol-2-phosphate, it may be a valuable tool for cleaning neutral, the water is discharged into an aerobic wetland to
water contaminated with radionuclides. encourage the precipitation of iron and aluminum oxides
Microbial processes can also accumulate and immobi- and the coprecipitation of arsenic, if this is present (113).
lize mercury ions. Escherichia coli engineered to overex- If heavy metals are present in mine water, the iron-
press a Hg2+ transporter and metallothionein accumu- free water can be made to flow into an anaerobic part
late and immobilize mercury ions quite selectively (104). of the constructed wetland, where organic material, such
Alternatively, many microorganisms reduce mercury com- as compost, manure, or sawdust, provides reductants to
pounds to the elemental form, which is volatile (105). sulfate-reducing bacteria that become established therein.
Although this certainly reduces the environmental impact These bacteria reduce sulfate to sulfide, which precipitates
of compounds such as methylmercury, such a bioprocess the heavy metal ions as insoluble sulfides. It is important
would have to include a mercury capture system (106) that this part of the constructed wetland be kept anaerobic
before it could be exploited on a large scale with public to prevent oxidation and remobilization of the precipitated
support. metals. For this reason, this part of the wetland is
typically kept flooded and free of aquatic plants that might
Bioremediation introduce oxygen through their roots. (113). Finally, an
aerobic algal mat can act as a polishing step to complete the
Water. Groundwater can be treated in anaerobic
removal of contaminants, particularly manganese (114).
bioreactors that encourage the growth of sulfate-reducing
Mine drainage that is not acidic may not need such
bacteria, where the metals are reduced to insoluble sulfides
complicated systems, and individual parts of the treatment
and concentrated in the sludge. For example, such a
train described above may suffice in some situations.
system is in use to decontaminate a zinc smelter site
Bacteria have been used to reduce Cr(VI) to Cr(III),
in the Netherlands (107).
thereby immobilizing it in soils and water (115). Reactive
Bacterial remediation of selenium oxyanions in San
barrier technologies, including biological barriers (22),
Joaquin (CA) drainage water is under active investiga-
are being widely used for metal ion and metalloid
tion (108,109), but has not yet been commercialized. Agri-
contaminants.
cultural drainage rich in selenium is also typically rich in
nitrates, so bioremediation must also include conditions
that stimulate denitrification (110). CONCLUSION

Rhizofiltration and Other Forms of Phytoremedia- Bioremediation is being successfully used to treat a wide
tion. Rhizofiltration is the use of plant root systems to range of contaminants in a broad range of environments.
remove contaminating metal ions from water (111), and Contaminants amenable to bioremediation include crude
it now seems clear that microorganisms play an essential oils and refined petroleum products, halogenated solvents,
role on the roots (112). Nevertheless, little is known about pesticides, herbicides, military chemicals, and mine
the specific interactions of microbes with the plants, and waters, and they can be treated in air, soil, and water. In all
this area is very much in its infancy. cases it is clear that aiding and abetting natural selection
Mine Drainage. Natural drainage waters that come is the key to success and that making small adjustments
into contact with active and abandoned metal and coal of the local environment to encourage the growth of
mines can become seriously contaminated with a range of remediating organisms may be all that is needed. For
heavy metal ions and/or often become quite acidic, with example, petroleum is biodegradable, but this degradation
a pH near 2. While underground, the water is typically is typically nutrient limited; partially relieving this
anoxic, and any iron is present as soluble ferrous species. limitation can lead to stimulating the biodegradation
When this mixes with aerobic surface water, the iron severalfold with no adverse environmental impact. On the
precipitates as bright orange ferric hydroxide, and this other hand, organisms growing on other substrates most
can have a serious environmental impact. In recent years, effectively remove several halogenated solvents; adding
it has become clear that the environmental impact of these substrates, for example, methane, significantly
acid mine drainage can be minimized by the construction increases the rate of removal of the contaminant.
of artificial wetlands that combine geochemistry and The centrality of natural selection in designing a
biological treatments. These systems are being designed remediation strategy cannot be overestimated. One
for a range of wastewaters, most of which fall outside the area where it sometimes seems to be ignored is the
scope of this section. consideration of the need to add microbial inoculants
The precipitation of ferric hydroxide is typically to stimulate biodegradation (bioaugmentation). Some
biologically mediated, by iron-oxidizing bacteria, at acid modern contaminants of concern (oil, many halogenated
pH, but is usually rather slow. Abiotic oxidation becomes compounds) have been part of the biosphere for millennia,
more important at pH values above 5, and this is usually albeit usually at low levels. Microbes that can degrade
much faster than the biological process. Most constructed them are probably ubiquitous, so adding organisms to
wetlands for treating acid mine waters thus start with stimulate biodegradation is unlikely to be cost-effective.
a zone designed to raise the pH. A bed of crushed If the environment is made hospitable for growth, the
limestone often suffices to raise the pH significantly, indigenous organisms are likely to prove sufficient for
and it is important that this be kept anoxic to prevent bioremediation. Perhaps the most startling demonstration
rust precipitation on the limestone, which would prevent of this was performed by the USEPA when testing
further production of alkalinity. Once the pH is near potential inoculants for stimulating oil biodegradation
710 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES

in Alaska following the Exxon Valdez oil spill (116). 2. S. S. Butcher, R. J. Charlson, G. H. Orians and G. V.Wolfe,
Eight products were tested in small laboratory reactors Global Biogeochemical Cycles, Academic Press, New York,
that allowed substantial degradation of a test oil by 1992.
the indigenous organisms of Prince William Sound; all 3. S. Negoro, Appl. Microbiol. Biotechnol., 54, 461–466 (2000).
eight microbial inocula had a greater stimulatory effect on 4. P. Adriaens and T. M. Vogel, in L. Y. Young and C. E. Cer-
alkane degradation if they were sterilized by autoclaving niglia, eds., Microbial Transformation and Degradation of
prior to addition. This suggests that the indigenous Toxic Organics, Wiley-Liss, New York, 1995, pp. 435–486.
organisms readily outcompeted the added products, but 5. M. M. Haggblom, V. K. Knight, and L. J. Kerkhof, Environ.
that autoclaving the products released some trace nutrient Poll. 107, 199–207 (2000).
that was able to stimulate the growth of the endogenous 6. M. D. Cameron, S. Timofeevski and S. D., Appl. Microbiol.
organisms. Biotechnol. 54, 751–758 (2000).
Of course, bioaugmentation may prove more effective 7. R. C. Prince This volume (2001).
with contaminants that have only recently entered the 8. J. D. Coates and R. T.Anderson, Trends Biotechnol. 18,
biosphere, such as methyl-t-butyl ether (11,12), or are 408–412 (2000).
toxic (58,60). It may also be effective if material is to be 9. K. G. Kropp, I. A. Davidova and J. M. Suflita, Appl. Envi-
composted, where it is likely that composting organisms ron. Microbiol. 66, 5393–5398 (2000).
may take some time to become established. And modern 10. J. Heider, A. M. Spormann, H. R. Beller and F. Widdel,
molecular biology may offer opportunities for moving FEMS Microbiol. Rev. 22, 459–473 (1999).
effective degradation pathways into organisms native to a 11. J. P. Salanitro This volume (2001).
contaminated site, improving biodegradation pathways by 12. R. C. Prince, Crit. Rev. Microbiol. 26, 163–178 (2000).
broadening the substrate range of degradative enzymes 13. H. Jorio, L. Bibeau, and M. Heitz, Env. Sci. Technol. 34,
or by removing biochemical constraints to maximize 1764–1771 (2000).
degradative activities (75). There are, however, many 14. National Research Council Oil in the Sea: Inputs, Fates and
technical and regulatory hurdles to be surmounted before Effects, National Academy Press, Washington, D.C., 1985.
this is likely to be widely used. 15. T. Lunel, J. Rusin, C. Halliwell and L. Davies, in Proc. 1997
Another area where there is controversy is in the role of International Oil Spill Conference, American Petroleum
surfactants. Many contaminants of concern are hydropho- Institute, Washington, D.C., 1997, pp. 185–194.
bic and are adsorbed to soil material. Many bacteria, 16. K. Lee and S. deMora, Environ. Technol. 20, 783–794
particularly those that degrade hydrocarbons, produce (1999).
surfactants that are thought to help increase the bioavail- 17. R. C. Prince and J. R. Bragg, Bioremediation J. 1, 97–104
ability of the contaminant. Some release the surfactants (1997).
into the medium, others incorporate them into their cell 18. B. C. Alleman and A. Leeson eds., In Situ Bioremediation
exterior, and there have been elegant experiments to show of Petroleum Hydrocarbon and Other Organic Compounds.
that inhibiting the production of these compounds inhibits Battelle Press, Columbus, Ohio, 1999.
the ability of the bacteria to degrade their substrates. Allen 19. G. P. Lu, T. P. Clement, C. M. Zheng, and T. H. Wiedemeier,
and coworkers (117) provide an example of how difficult it Ground Water, 37, 707–717 (1999).
is to make broad generalizations; they examined the ability 20. B. C. Alleman and A. Leeson eds., Natural Attenuation of
of a Pseudomonas and a Sphingomonas to degrade naph- Chlorinated Solvents, Petroleum Hydrocarbons, and other
thalene and phenanthrene in the presence of a nonpolar, Organic Compounds, Battelle Press, Columbus, Ohio, 1999.
nondegradable (by these organisms) solvent (2,2,4,4,6,8,8- 21. B. C. Alleman and A. Leeson eds., Bioreactor and Ex-
heptamethylnonane) or the nonionic surfactant Triton Situ Biological Treatment Technologies. Battelle Press,
X-100. The heptamethylnonane increased the rate of Columbus, Ohio, 1999.
degradation of both substrates for both microorganisms. 22. M. M. Scherer, S. Richter, R. L. Valentine, and P. J. J.
The Triton X-100 increased the rate of degradation by the Alvarez, Crit. Rev. Microbiol. 26, 221–264 (2000).
Pseudomonas, but inhibited that by the Sphingomonas. 23. L. D. Zamojski, J. R. Stachowski, and S. R. Porter, in
There have been many suggestions to add surfactants, B. C. Alleman and A. Leeson eds., In Situ Bioremediation
either microbial or synthetic, to stimulate biodegradation. of Petroleum Hydrocarbon and Other Organic Compounds.
This seems to be beneficial in the case of some oil disper- Battelle Press, Columbus, Ohio, 1999, pp. 65–69.
sants at sea (118), but the situation seems complicated in 24. M. Stieber et al., in B. C. Alleman and A. Leeson eds., In
soils, and probably needs to be assessed on a case-by-case Situ Bioremediation of Petroleum Hydrocarbon and Other
basis (119). Surfactants did not stimulate pentachlorophe- Organic Compounds. Battelle Press, Columbus, Ohio, 1999,
nol degradation in one recent study (120). Nevertheless, pp. 337–343.
surfactants are clearly beneficial in the physical part of 25. F. H. Chapelle, Ground Water 37, 122–132 (1999).
technologies such as bioslurping (18), so it will be impor- 26. S. S. Koenigsberg, R. D. Norris eds., Accelerated Bioremedi-
tant to ensure that, at least, they do not inhibit the ation Using Slow Release Compounds, Battelle Press, Ohio,
biodegradation of residual contaminants. 1999.
27. J. J. Kaluarachchi, V. Cvetkovic and S. Berglund, J. Con-
tam. Hydrol. 4, 335–365 (2000).
BIBLIOGRAPHY
28. D. R. Lovley, J. C. Woodward and F. H. Chapelle, Nature
1. K. Devine, in R. E. Hinchee, J. A. Kittel and H. J. Reisinger, 370, 128–131 (1994).
eds., Applied Bioremediation of Hydrocarbons. Battelle 29. R. T. Anderson, D. R. Lovley, Environ. Sci. Technol. 34,
Press, Columbus, Ohio, 1995, pp. 53–59. 2261–2266 (2000).
 BIOREMEDIATION: MICROBIOLOGICAL PROCESSES 711

30. A. Leeson and B. C. Alleman eds., Bioremediation Tech- R. C. Sims, eds., Bioremediation of Contaminated Soils,
nologies for Polycyclic Aromatic Hydrocarbon Compounds, American Society of Agronomy, Madison, Wis., 1999,
Battelle Press, Columbus, Ohio, 1999. pp. 719–742.
31. J. Ashworth, D. Keyes and J. L. Crepin, Can. J. Soil Sci. 79, 58. G. M. Colores and S. K. Schmidt, J. Ind. Microbiol. Biotech-
449–455 (1999). nol. 23, 326–331 (1999).
32. R. C. Sims and J. L. Sims, in D. C. Adriano, J.-M. Bollag, 59. A. J. Wall and G. W. Stratton, Water Air Soil Poll. 82,
W. T. Frankenberger Jr., and R. C. Sims eds., Bioremedia- 723–737 (1995).
tion of Contaminated Soils, American Society of Agronomy, 60. F. Baud-Grasset and T. M. Vogel, in R. E. Hinchee,
Madison, Wis., 1999 pp. 767–782. J. E. Fredrickson and B. C. Alleman, eds., Bioaugmentation
33. C. Coulter et al., Appl. Environ. Microbiol. 65, 4301–4312 for Site Remediation, Battelle Press, Columbus, Ohio, 1995,
(1999). pp. 39–48.
34. J. F. Ferguson and J. M. H. Pietari, Environ. Poll. 107, 61. F. C. Michel, C. A. Reddy, and L. J. Forney, Appl. Environ.
209–215 (2000). Microbiol. 61, 2566–2571 (1995).
35. D. Kohler-Staub, S. Frank and T. Leisinger, Biodegradation 62. C. J. Gantzer and D. Cosgriff, in R. E. Hinchee, R. M. Miller
6, 229–236 (1995). and P. C. Johnson, eds., In Situ Aeration: Air Sparging,
36. D. J. Arp, Curr. Opin. Biotechnol. 6, 352–358 (1995). Bioventing and Related Remediation Processes, Battelle
37. S. J. Ergas, K. Kinney, M. E. Fuller, and K. M. Scow, Press, Columbus, Ohio, 1995, pp. 543–549.
Biotechnol. Bioengi. 44, 1048–1054 (1994). 63. D. G. Karamanev and R. Samson, Environ. Sci. Technol. 32,
38. A. Leeson and B. C. Alleman eds., Engineered Approaches 994–999 (1998).
for In Situ Bioremediation of Chlorinated Solvent Contami- 64. D. L. Bedard and R. J. May, Environ. Scie. Technol. 30,
nation, Battelle Press, Columbus, Ohio, 1999. 237–245 (1996).
39. J. L. Sherwood, J. N. Petersen, and R. S. Skeen, Biotechnol. 65. D. L. Bedard, S. C. Bunnell and L. A. Smullen, Environ. Sci.
Bioengi. 64, 342–348 (1999). Technol. 30, 687–694 (1996).
40. T. C. Hazen, in D. C. Adriano, J.-M. Bollag, W. T. Franken- 66. M. R. Harkness et al., Science 259, 503–507 (1993).
berger Jr., and R. C. Sims, eds., Bioremediation of Contam- 67. R. E. Dick and J. P. Quinn, Appl. Microbiol. Biotechnol. 43,
inated Soils, American Society of Agronomy, Madison, Wis. 545–550 (1995).
1999, pp. 743–765.
68. F. Bichat, G. K. Sims, and R. L. Mulvaney, Soil Sci. Soc.
41. R. J. Steffan et al., Environ. Sci. Technol. 33, 2771–2781 Am. J. 63, 100–110 (1999).
(1999).
69. J. J. Crawford, G. K. Sims, R. L. Mulvaney, and M. Rado-
42. S. H. Streger, C. W. Condee, A. P. Togna, and M. F. Deflaun, sevich, Appl. Microbiol. Biotechnol. 49, 618–623 (1998).
Environ. Sci. Technol. 33, 4477–4482 (1999).
70. R. H. Kaake, D. L. Crawford, and R. L. Crawford, Biodegra-
43. Z. Yu and G. B. Smith, Appl. Microbiol. Biotechnol. 53,
dation 6, 329–337 (1995).
484–489 (2000).
71. T. B. Hammill and R. L. Crawford, Appl. Environ. Microbiol.
44. G. W. Gribble, J. Chem. Ed. 71, 907–911 (1994).
62, 1842–1846 (1996).
45. V. M. Thomas and T. G. Spiro, Environ. Sci. Technol. 30,
72. J. Kazumi and D. G. Capone, Appl. Environ. Microbiol. 62
82A–85A (1996).
2820–2829 (1995).
46. H. Smidt, A. D. L. Ackermans, J. van der Ost and W. M. de,
73. I. Mirgain, G. Green, and H. Monteil, Environ. Technol. 16,
Vos, Enzyme Microb. Technol. 27, 812–820 (2000).
967–976 (1995).
47. R. L. Crawford and M. M. Ederer, J. Ind. Microbiol. Biotech-
74. D. A. Newcombe and D. E. Crowley, Appl. Microbiol. Bio-
nol. 23, 320–325 (1999).
technol. 51, 877–882 (1999).
48. A. Mendoza Cantu, A. Albores, L. Fernandez Linares and
75. L. C. Strong, H. McTavish, M. J. Sadowsky, and L. P. Wack-
R. RodriguezVazquez, Environ. Toxicol. 15, 107–113 (2000).
ett, Environ. Microbiol. 2, 91–98 (2000).
49. B. Tartakovsky et al., Appl. Environ. Microbiol. 65,
76. K. E. LeJeune and A. J. Russell Biotechnol. Bioengi. 62,
4357–4362 (1999).
659–665 (1999).
50. J. Aislabie and G. Lloyd-Jones. A Review of Bacterial
77. Y. F. Zhang et al., Biotechnol. Bioeng. 64, 221–231 (1999).
Degradation of Pesticides. Aust. J. Soil. Res. 33, 925–942
(1995). 78. M. A. Major, in D. C. Adriano, J.-M. Bollag, W. T. Franken-
51. C. McGowan, R. Fulthorpe, A. Wright, and J. M. Tiedje, berger Jr., and R. C. Sims, eds. Bioremediation of Contami-
Appl. Environ. Microbiol. 64, 4089–4092 (1998). nated Soils, American Society of Agronomy, Madison, Wis.,
1999, pp. 111–132.
52. J. F. Quensen, S. A. Mueller, M. K. Jain, and J. M. Tiedje,
Science, 280, 722–724 (1998). 79. A. U. Nwankwoala, N. O. Egiebor, C. Gilbert, and K. Nyavor,
Biodegradation 10, 105–112 (1999).
53. R. Unterman, in R. L. Crawford, and D. L. Crawford, eds.,
Bioremediation, Principles and Practices, Cambridge Uni- 80. R. A. Bruce, L. A. Achenbach, and J. D. Coates, Environ.
versity Press, Cambridge, U.K., 1996, pp. 209–253. Microbiol. 1, 319–329 (1999).
54. S. Han et al., Science 270, 976–980 (1995). 81. D. Cowan, Soil Sed. Groundwater June/July, 25–26 (2000).
55. J. Wiegel and Q. Z. Wu, FEMS Microbiol. Ecol. 32, 1–15 82. J. C. Spain et al., in B. C. Alleman and A. Leeson, eds.,
(2000). Bioremediation of Nitroaromatic and Haloaromatic Com-
pounds, Battelle Press, Columbus, Ohio, 1999, pp. 7–14.
56. B. C. Alleman and A. Leeson eds., Bioremediation of
Nitroaromatic and Haloaromatic Compounds, Battelle 83. A. Rozkov et al., Water Res. 33, 1303–1313 (1999).
Press, Columbus, Ohio, 1999. 84. E. Flathman et al., Remediation 6, 57–79 (1996).
57. R. C. Sims, J. L. Sims, R. L. Zollinger, and S. G. Huling in 85. A. S. Whiteley and M. J. Bailey, Appl. Environ. Microbiol.
D. C. Adriano, J.-M. Bollag, W. T. Frankenberger Jr., and 66, 2400–2407 (2000).
712 BIOREMEDIATION: AQUATIC ECOSYSTEMS

86. Z. Ronen, J. M. Bollag, C. H. Hsu, and J. C. Young, Ground American Society of Agronomy, Madison, Wis., 1999,
Water 34, 194–199 (1996). pp. 457–508.
87. E. A. Greene, D. L. Coy, and P. M. Fedorak, Biorem. J. 3, 113. G. Robb and J. Robinson, Mining Environ. Manage. Sept,
299–314 (1999). 19–21 (1995).
88. G. H. Brox, W. R. Mahaffay, and B. S. Yare, in B. C. Alle- 114. J. Bender and P. Phillips, Mining Environ. Manag. Sept,
man, and A. Leeson eds., Bioreactor and Ex-Situ Biological 25–27 (1995).
Treatment Technologies, Battelle Press, Columbus, Ohio, 115. J. L. Bader, G. Gonzalez, P. C. Goodell, A. S. Ali, and
1999, pp. 19–24. S. D. Pillai, Biorem. J. 3, 201–212 (1999).
89. D. W. Lee and R. J. Portier, Remediation 9, 117–132. 116. A. D. Venosa, et al., J. Ind. Microbiol. 10, 13–23 (1992).
90. A. Leeson and B. C. Alleman eds., Bioremediation of Metals 117. C. C. R. Allen et al., Appl. Environ. Microbiol. 65, 1335–
and Inorganic Compounds, Battelle Press, Columbus, Ohio, 1339 (1999).
1999. 118. R. P. J. Swannell and F. Daniel, in Proc. 1999 International
91. G. C. Schmidt and M. B. Ballew, in R. E. Hinchee, J. L. Oil Spill Conference, American Petroleum Institute, Wash-
Means and D. R. Burris, eds. Bioremediation of Inorganics, ington, D.C., 1999, pp. 169–176.
Battelle Press, Columbus, Ohio, 1995, pp. 109–116. 119. C. O. Belloso in B. C. Alleman and A. Leeson eds., Bioreactor
92. J. M. Flere and T. C. Zhang, Water Sci. Technol. 38, 15–22 and Ex-Situ Biological Treatment Technologies, Battelle
(1998). Press, Columbus, Ohio, 1999, pp. 177–181.
93. W. J. Hunter, in A. Leeson and B. C. Alleman eds., Bioreme- 120. T. Cort and A. Bielefeldt, J. Environ. Engi.-ASCE 126,
diation of Metals and Inorganic Compounds, Battelle Press, 635–643 (2000).
Columbus, Ohio 1999, pp. 47–51.
94. L. A. Schipper and M. Vojvodic-Vukovic, Ecol. Engi. 14,
269–278 (1999). BIOREMEDIATION: AQUATIC ECOSYSTEMS
95. R. Gupta et al., Curr. Sci. 78, 967–973 (2000).
96. W. L. Smith and G. M. Gadd, J. Appl. Microbiol. 88, K. H. BAKER
983–991 (2000).
Middletown, Pennsylvania
97. R. S. Dungan and W. T. Frankenberger, Biorem. J. 3,
171–188 (1999). Contamination of aquatic environments, both freshwater
98. D. Ahmann, A. L. Roberts, L. R. Krumholz, and F. M. M. and marine, with petroleum hydrocarbons and a variety
Morel, Nature 371, 750 (1994). of xenobiotic compounds is a commonplace occurrence
99. M. Santini, L. I. Sly, R. D. Schnagl, and J. M. Macy, Appl. in the modern world. Oceans and rivers remain major
Environ. Microbiol. 66, 92–97 (2000). avenues of transportation and commerce with tankers
100. D. K. Newman, T. J. Beveridge and F. M. M. Morel, Appl. and supertankers traversing the globe. While spectacular
Environ. Microbiol. 63, 2022–2028 (1997). oil spills, resulting in the contamination of hundreds to
101. A. Abdelouas et al., Sci. Total Environ. 250, 21–35 (2000). thousands of square miles, have left the most vivid impact
102. L. E. Macaskie, R. M. Empson, F. Lin, and M. R. Tolley, J. on the general public, scores of smaller releases and
Chem. Technol. Biotechnol. 63, 1–16 (1995). spills have resulted in extensive contamination of open
103. K. M. Bonthrone, G. Basnakova, F. Lin, and L. E. Macaskie, waters, shores, and sediments. One of the technologies
Nat. Biotechnol. 14, 635–638 (1996). that has found widespread application in the remediation
104. S. L. Chen, E. K. Kim, M. L. Shuler, and D. B. Wilson, and restoration of aquatic ecosystems is bioremediation.
Biotechnol. Prog. 14, 667–671 (1998). This chapter will review the basic principles involved
105. K. Michalke et al., Appl. Environ. Microbiol. 66, 2791– in the use of bioremediation in aquatic (freshwater and
2796 (2000). marine) ecosystems, provide an overview of examples of
106. S. Okino, K. Iwasaki, O. Yagi, and H. Tanaka, Biotechnol. successful bioremediation projects, and examine several
Letts. 22, 783–788 (2000). case studies in more depth to present a snapshot of the
107. L. J. Barnes, P. J. M. Scheeren, and C. J. N. Buisman, in
current state of the art. In addition to other chapters in
J. L. Means and R. E. Hinchee, eds. Emerging Technology this work, the reader is referred to numerous recent books
for Bioremediation of Metals, Lewis Publishers, Boca Raton, and review articles for additional information (1–13).
Fla., 1994, pp. 38–49.
108. A. W. Cantafio et al., Appl. Environ. Microbiol. 62, 3298– OVERVIEW OF BIOREMEDIATION
3303 (1996).
109. L. P. Owens, K. C. Kovac, J. A. L. Kipps, and D. W. J. Bioremediation relies on the abilities of microorganisms to
Hayes, in R. E. Hinchee, J. L. Means and D. R. Burris, eds., degrade and detoxify organic and inorganic compounds. In
Bioremediation of Inorganics, Battelle Press, Columbus, its simplest form, bioremediation involves no intervention
Ohio 1995, pp. 89–94. at all and simply allows natural processes to reduce the
110. R. S. Oremland et al., Appl. Environ. Microbiol. 65, concentration of the contaminant in an environment. This
4385–4392 (1999). is usually referred to as natural attenuation or intrinsic
111. V. Dushenkov, P. B. A. N. Kumar. H. Motto, and I. Raskin, bioremediation. In most instances, alteration or manipu-
Environ. Sci. Technol. 29, 1239–1245 (1995). lation of physical/chemical properties of the contaminant
112. W. W. Wenzel, D. O. Adriano, D. E. Salt, and R. D. Smith, in or factors in the environment is used to enhance the
D. C. Adriano, J.-M. Bollag, W. T. Frankenberger Jr., and natural processes. This is referred to as enhanced biore-
R. C. Sims, eds., Bioremediation of Contaminated Soils, mediation. Fundamentally, there are two approaches to
 BIOREMEDIATION: AQUATIC ECOSYSTEMS 713

this enhanced bioremediation: (1) biostimulation, which increase in this ratio following exposure of the community
depends primarily on the modification of the environ- to crude oil. Delille and Delille (51) studied enclosures on
ment and (2) bioaugmentation, which uses the addition of nine different sub-Antarctic intertidal beaches to which
microbial cultures to increase biodegredation. Frequently, crude oil had been added. Despite the initially pristine
these two approaches are combined and microorganisms condition of all of the sites, indigenous populations of
are introduced in conjunction with such environmental oil-degrading microorganisms were present.
modifications as nutrient supplementation. If environmental conditions prevent the natural
Bacteria capable of at least some transformation of all development of microbial populations capable of degrading
but the most exotic of man-made chemicals are probably a particular compound, or if the development of such
present in all aquatic environments (1,2,5,6,14–16). Isola- populations occurs at too slow a rate, it may be
tion and characterization of organisms capable of degrad- possible to enhance biodegradation through the addition
ing petroleum and related hydrocarbons (17–22), polyaro- of supplemental microorganisms (52,53). This process,
matic hydrocarbons (19,23–28), alkylbenzenes (29–31), termed bioaugmentation, may be accomplished using
chlorinated solvents (31,32), polychlorinated biphenyls, either microbial cultures selected from the contaminated
dioxins, and related compounds (33–36), and other organic site, grown under controlled conditions, and then re-added
contaminants (37,38) under either aerobic or anaerobic at much higher concentrations to the site or, alternatively,
conditions is commonly reported in the microbiological a variety of commercially prepared microbial products
literature. Kimura and coworkers (39) isolated 96 strains each designed for degradation of specific contaminants.
of bacteria from seawater that were capable of degrad- Effective bioremediation depends upon extensive site
ing normal alkanes, branched alkanes, cycloalkanes, characterization with a particular emphasis on factors
and aromatic compounds. A marine bacterium, isolated that control or enhance microbial activity. Although there
from intertidal crude-oil contaminated sediments, showed are a large number of factors that influence microbial
enhanced growth in the presence of added petroleum populations and activities, only a limited number of factors
hydrocarbons compared to growth in unsupplemented are amenable to modification from a technological and
marine medium (40). Oil degrading bacteria were iso-
economic perspective.
lated from 96% of the sites examined in the vicinity
All biological systems are influenced by ambient
of Zhejaing islands. The abundance of oil degrading
temperature. Microorganisms, which can be broadly
microorganisms was correlated with the concentration of
categorized in terms of their temperature range, are no
petroleum hydrocarbons in the water (41). Considering
exception. Within the range of temperatures in which
that most (90–99%) microbial species in natural com-
a microorganism can grow, biological activity generally
munities cannot be detected using standard laboratory
increases with increasing temperature with maximum
culture techniques, it is likely that these reports reflect
activity occurring at a temperature just below the
the barest minimum of the biodegradation capabilities of
maximum temperature the organism can tolerate. As a
microorganisms (42).
general rule of thumb, microbial activity doubles with
Chronic release of low levels of contaminants into
each 10 ° C increase in temperature. Seasonal differences
aquatic environments usually results in an enrichment
in temperature, therefore, can have profound effects
of the microbial community for organisms capable of
degrading the contaminant. Under unfavorable environ- on the rate of biodegradation (43,54,55). Wyndiam and
mental conditions or in the absence of prior exposure to Costerton (56) reported a 40-fold reduction in the rate
the contaminant however, both the numbers and activity of naphthalene degradation in river sediment during the
of specific microbial populations may be extremely low, winter (0 to 4 ° C) compared to the summer (8 to 21 ° C).
accounting for 1% or less of the total bacterial commu- Aquatic environments are frequently at temperatures
nity (43–45). in which microbial activity is low. Manipulation of
In many instances a lag or acclimation period is the temperature of aquatic environments usually is
observed before active degradation of the compound takes not feasible and, therefore, temperature can be a
place. Acclimation is a general term that may refer to sev- significant constraint in the use of bioremediation.
eral changes in microbial populations, including (1) growth Low environmental temperature was a significant factor
of an initially small population of organisms capable of limiting the rate of bioremediation during cleanup of
degrading the contaminant, (2) changes in metabolic reg- the Exxon Valdez spill in Prince William Sound, Alaska.
ulation including induction and/or derepression of key In some instances in which contamination is primarily
enzymes in the necessary metabolic pathways, (3) genetic associated with sediments, ex situ treatment in bioreactors
change in the indigenous populations, and (4) the preferen- may allow for modification of temperature.
tial use of alternative substrates by the degrader popula- The optimal pH for microbial activity is generally
tions (8,46–48). Leahy and coworkers (49) concluded that between 5.5 and 8.5. Environmental pH is not a significant
lack of adaptation by microbial communities accounted for factor controlling bioremediation in open waters, such as
low petroleum hydrocarbon degradation rates in sediment rivers or oceans in which pH is usually controlled by
samples from the Gulf of Mexico. Siron and coworkers (50) natural buffering systems; however, pH may be outside the
assessed the adaptation of an indigenous cold seawa- optimal range within sediments. Incubation of salt marsh
ter microbial community to petroleum hydrocarbons by sediments under slightly acidic conditions (pH 5.0 and 6.5)
monitoring the ratio of oil-degrading bacteria to total het- was found to decrease hydrocarbon mineralization relative
erotrophs within the community. They reported a 10-fold to incubation at higher pH (8.0) (57). As with temperature,
714 BIOREMEDIATION: AQUATIC ECOSYSTEMS

manipulation of sediment pH is usually confined to the biodegradation without environmental modification to

treatment of excavated/dredged material in bioreactors. enhance microorganisms, may be sufficient to remediate a
Microorganisms require a terminal electron acceptor as contaminated site (67,68).
a sink for electrons generated during the metabolism of When degradation is limited by the available concentra-
organic compounds. Under aerobic conditions, oxygen is tions of fixed forms of nitrogen and phosphorus, fertilizer
used as the terminal electron acceptor being reduced to addition can be effective in stimulating microbial activ-
water. As many organic contaminants such as petroleum ity and enhancing biodegradation of the contaminating
hydrocarbons are less dense than water they usually petroleum (69). Care must be taken in determining the
are found on the surface of open waters. Under these type and amount of fertilizer that should be added to
conditions, oxygen concentration is not a limiting factor. the site. The amount of supplemental inorganic nutri-
Degradation of organic compounds is generally, but ents can be estimated on a stoichometric basis. Several
not always, faster under aerobic rather than anaerobic standard textbooks on hazardous waste treatment and
conditions. bioremediation outline the methods for making such an
Contamination of aquatic environments is not limited estimation (6,70). A more accurate appraisal of the opti-
to the surface of open waters. Numerous mechanisms mum amount of inorganic nutrients to be added can be
are responsible for the transport of contaminants to sed- made using microcosm studies. A variety of techniques for
iments. Some contaminants such as chlorinated solvents conducting these types of studies have been described in
are heavier than water and sink in aquatic systems. As oil the literature (1,71,72).
weathers, the formation of droplets that are heavier than The amount of inorganic nutrients is not the only
water may result in the transfer of these compounds to the factor that needs to be considered in designing a
sediments underlying open water. High molecular weight biostimulation system. Equally important is the form of
fractions of petroleum, including polyaromatic hydrocar- nutrients used (73,74). Water-soluble nutrients such as
bons such as benzo(a)pyrene tend to sorb to particles inorganic salts rapidly dissolve and are dispersed and
within the water column and hence accumulate within diluted (75). For example, Foght and coworkers (76) found
sediments (14,58). Thus, contaminants within aquatic that the addition of supplemental nitrogen in the form
environments can be transferred downward in the water of ammonium to a marine enrichment culture resulted
column and under conditions of thermal (or in some estu- in decreased degradation of aromatic compounds while
aries salinity) stratification may enter anoxic/anaerobic the use of nitrate, as a source of nitrogen, enhanced
environments or accumulate in sediments with reduced biodegradation. Oleophilic nutrients dissolve slowly and
or depleted oxygen. Finally, wind and wave action can so their use can result in sustained release of nutrients
carry floating contaminants to shore areas such as beaches necessary to support bacterial growth and activity. Santos
and wetlands where they can be deposited in sediments. and coworkers (77) compared two different oleophilic
Under conditions of oxygen limitation, some populations fertilizers for use in enhanced bioremediation. Both of the
of microorganisms that can use alternative compounds, products increased the biodegradation of Iranian crude
such as nitrate, sulfate, and carbonate as terminal elec- oil. Neither formulation was found to be clearly superior
tron acceptors may be able to degrade the contaminants, to the other; rather, each had its own advantage (faster
including petroleum hydrocarbons and polycyclic aromatic biodegradation of alkenes versus quicker disappearance of
hydrocarbons (26,58–60). Microorganisms from the New the oil slick) and the authors concluded that a mixture of
York/New Jersey estuary, a site with a history of pollu- the two fertilizers would offer the best treatment option.
tion, were able to degrade benzene, toluene, ethylbenzene, Microorganisms are unable to efficiently metabolize
and xylenes components in the absence of molecular oxy- compounds that are not dissolved in water. The avail-
gen. Enrichment cultures from the sediments with this ability of a compound is influenced by both its solubility
ability were established under denitrifying, sulfidogenic, and its partitioning between the aqueous and organic
methanogenic, and iron reducing conditions (61). In the phases (78,79). Sorption of organic compounds by sedi-
case of anoxic/anaerobic sediments, electron acceptors (and ments can have profound effects on the bioavailability
redox potential) can be controlled in bioreactor systems for of the compounds. Generally, these materials are less
the remediation of excavated material. bioavailable than dissolved compounds. In addition, sorp-
Carbon, nitrogen, and phosphorus are essential for tion to sediment particles may result in the transport
microbial maintenance and growth (1,2,5,62,63). During of the sorbed compounds to environments not conducive
biodegradation of organic contaminants, the microbe’s to biodegradation. Microorganisms are known to produce
carbon requirements usually are met by the contaminant surface-active compounds — bioemulsifiers or biosurfac-
itself. Nitrogen and phosphorus, on the other hand, tants — that increase the solubility and availability of
are present at extremely low concentrations in aquatic organic compounds (80). In addition, man-made emulsi-
environments. Microbial growth, therefore, is limited by fiers can be used to enhance bioavailability (5,6).
the low nutrient levels (1,5). Fuentes and coworkers (64) Biosurfactants can increase the bioavailability of non-
reported that ambient phosphorus concentrations were aqueous phase liquids, such as hydrocarbons by either
important in the survival of petroleum degrading isolates increasing the aqueous solubility of the hydrocarbons or by
in tropical bays. In some cases, there are sufficient increasing the hydrophobicity of the bacterial cell surface.
inputs of inorganic nitrogen and phosphorus to support Rosenberg and Ron (81) have recently reviewed the struc-
extensive microbial activity (65–68). In these cases, ture, biosynthesis, and possible uses of biosurfactants.
intrinsic bioremediation, the use of natural microbial Rhamnolipids produced by two strains of Pseudomonas
 BIOREMEDIATION: AQUATIC ECOSYSTEMS 715

aeruginosa grown on glucose and hexadecane were found habitats with coastal areas benefiting from the input of
to cause release of LPS from the outer membrane of the nutrients from the surrounding terrestrial systems.
organisms resulting in lower cell surface hydrophobic- In terms of the global water budget, freshwater environ-
ity even at concentrations of the biosurfactant below the ments make a miniscule contribution. In terms of human
critical micelle concentration (82). Acinetobacter radiore- society, freshwater environments have an unparalleled
sistans KA53, on the other hand produces a high molecular importance. The limited global freshwater reserves serve
weight complex of polysaccharide and protein named as a medium for transportation, support agriculture and
Alasan that functions as a bioemulsifier (83). Passeri and industrial processes, function as a repository for waste
coworkers (84) isolated a glycolipid biosurfactant from a products, provide recreational opportunities, and sustain
marine bacterial isolate (MM1) that increased solubility human life by providing potable water. Because of the
of normal alkanes in water. necessity of water for human existence, much research
Effective bioremediation depends upon containment of has been devoted to developing treatment systems and
the contaminant within a defined area in which envi- other means to safeguard freshwater environments.
ronmental conditions can be modified and controlled to Freshwater environments are as diverse as marine
enhance microbial growth and metabolism. This is not environments. Fundamentally, a distinction can be made
possible within open waters. While a few studies have between standing (lentic) water bodies such as lakes and
been conducted in open water, the effectiveness of biore- ponds and flowing (lotic) systems such as rivers and
mediation remains unproven. In contrast, bioremediation streams.
in coastal regions and sediments has been shown to be In contrast to marine environments, freshwater envi-
effective in the treatment of contaminating organic com- ronments are scarce, accounting for less than 1/10,000 of
pounds (1,9,11,12,85). the total water available on the earth. While sharing many
characteristics with marine and estuarine environments,
there are several key points to consider in addressing
AQUATIC ENVIRONMENTS bioremediation of freshwater environments. First, because
of their small size compared to marine environments,
Marine environments — oceans, estuaries, shorelines, and freshwater environments are less constant, less stable in
tidal marshes — are the most extensive environments terms of physical and chemical characteristics. Tempera-
in the world. Oceans alone occupy a volume of 1.46 × ture fluctuation, diurnal cycles of dissolved oxygen, and
109 km3 , approximately 71% of the Earth’s surface. They sudden changes in silt and sediment load owing to storm
are highly diverse in terms of their chemical, physical, and water are among the variables significantly influencing
biological characteristics. Marine water is characterized the biota of lakes, streams, and wetlands. In addition,
by salinity between 3.3 and 3.7%. If brackish waters are because of the drastically reduced ratio of water to sed-
included within this definition, salinity ranges from 1.0 iments, benthic processes assume a more important role
to 3.7% are possible. In addition to the variability in in freshwater environments than in marine systems. Sed-
chemical composition, marine environments are subject iments rather than becoming a geologic sink for materials
to considerable physical mixing. Winds, waves, tides, and as happens in the profoundal depths of the ocean, become
currents all interact to result in a constantly changing a temporary repository with a constant exchange of mate-
environment. Nor is the variability limited to surface rials occurring between the sediments and the water. This
waters. Vertical stratification of water layers results means that sediments can serve as a source of extended
in the presence of multiple habitats that differ in contamination to the surrounding water and biota by the
terms of salinity, light, temperature, pressure, and other slow desorption of organic contaminants. Furthermore, as
factors even within limited geographic areas. Finally, benthic/detrital food chains dominate in freshwater envi-
different marine environments show profound differences ronments, the transfer and subsequent accumulation of
in the biological communities present. Open ocean waters lipophilic organic compounds into upper levels of the food
have extremely low biological productivity analogous to chain are likely.
terrestrial deserts, while estuaries and tidal marshes
have some of the highest productivity in the world. BIOREMEDIATION OF PETROLEUM HYDROCARBONS IN
Heterotrophic bacterial populations in open oceans are AQUATIC ENVIRONMENTS
typically highest at the surface (neuston) and decrease
rapidly with depth. Thus, while numbers of 107 to Contamination of aquatic environments is the result
108 cells/mL are commonly found in upper ocean waters, of both natural processes and human activities. The
this number may decrease by as much as an order of variety of compounds impacting marine environments
magnitude with each 100 m of depth (86). In oligotrophic is enormous; bioremediation of one category of con-
areas, microbial numbers within the water column may taminants — petroleum hydrocarbons — will be considered
reach as low as 100 organisms/mL. In addition to the here. Oil spills from off-shore drilling, tanker accidents and
differences in microbial numbers within the water column, lightings, accidental releases associated with maritime
there are also differences in the typical number of activity, industrial releases, atmospheric transport and
microorganisms between coastal and open water areas. deposition of products of incomplete combustion, runoff
In general, coastal waters have higher concentrations of from surrounding land, and natural oil seeps are major
microorganisms than open waters (87,88). This difference sources of petroleum hydrocarbons into the marine envi-
primarily reflects the different nutrient status of the two ronment (Table 1; 89–95). While oil spills such as the
716 BIOREMEDIATION: AQUATIC ECOSYSTEMS

Table 1. Global Input of Oil to Marine Environ- aliphatic sulfur compounds are oxidized to sufoxides, sul-
ments fones, sulfonates, and sulfates (102). Partial oxidation of
Amount components of the oil, occurring primarily at the edges of
(estimated 1990) Percentage the spill in which oil, water, and air interface, may result
Source (thousand tons/year) of Total in the formation of tar balls, which sink to the bottom
contaminating sediments. Finally, significant transforma-
Offshore oil 47 2 tion of petroleum hydrocarbons results from the metabolic
production activity of indigenous organisms (1,5,8,9,11,12,102–105).
Natural seeps 259 11
Dispersive forces frequently result in the transport of
Atmospheric 306 13
deposition
petroleum from open water to the shoreline. Because of the
Oil transporta- 564 24 importance of coastal, shoreline environments to human
tion and activities, remediation of these areas has been the focus of
shipping most marine bioremediation. Coastal environments vary
Urban runoff 1,175 50 considerably depending upon climate, topography, and
and discharge other local and regional factors.
Most experiments to evaluate the efficacy of bioreme-
Source: Ref. 89.
diation in the treatment of oil spills have been conducted
either in laboratory micro- or mesocosms or in small
1989 Exxon Valdez spill into Prince William Sound are in situ enclosures (1,8,11,12,106,107). While such experi-
spectacular and immediately grab media attention, they ments provide understanding of the physiology and ecology
are not the major source of oil pollution in the marine of microbial populations involved in the degradation of
environment. Soclo and coworkers (91) documented the petroleum hydrocarbons, they are difficult to extrapolate
importance of waste oils from mechanics shops in con- to real-world situations. Field studies on bioremediation
taminating sediments in coastal regions of Benin while could provide the information needed. Unfortunately, field
petroleum discharges from tankers and other wastes asso- studies are technically difficult, expensive, and usually
ciated with petroleum delivery and processing dominated constrained by regulatory restrictions. Ideally, field stud-
in sediments in Aquitaine (France). Zakaria and cowork- ies would involve the controlled release of petroleum
ers (92) reported that runoff of lubricating oils from road- hydrocarbons into either open oceans or adjacent to shore-
ways was a major source of contamination in sediments in lines. In reality, the vast majority is conducted during
the Straits of Malacca (Malaysia). response to actual oil spills and lacks the control of planned
Upon release into the environment, a variety of physical studies. Despite these limitations, both laboratory stud-
and chemical processes result in the transformation and ies and studies conducted coincidentally with oil spills
dispersion of the oil. Weathering causes the spilled oil to have provided useful insight into bioremediation under
break down and become heavier than water. At the same field conditions. Because of the low nutrient concentra-
time, the action of wind, waves, and currents may result tions and low microbial populations typical of open ocean
in dispersion of the oil spill (96–100). In addition, the environments, studies in this type of environment have
turbulence resulting from wind and waves can result in focused on biostimulation (the addition of nutrients) and
the formation of oil droplets that can be further dispersed bioaugmentation (the addition of microbial cultures).
throughout the water, forming secondary slicks or sinking Bioaugmentation has been shown to have variable and
within the water column (98). The turbulence can also often disappointing results when applied to marine oil
result in emulsification of the oil. Two types of emulsions spills. Tagger and coworkers (107) found that supple-
are possible: (1) water-in-oil emulsions and (2) oil-in- mental bacterial cultures rapidly declined when added to
water emulsions. Water-in-oil emulsions, frequently called oiled seawater. There was no enhancement of hydrocarbon
chocolate mousse, are associated with strong turbulence, removal in the presence of the supplemental microorgan-
which forcefully mixes the spill with the surrounding isms compared to natural conditions. Fayad and cowork-
water resulting in a product in which water becomes ers (74) noted that many of the commercial bioaugmen-
trapped within oil droplets. Chocolate mousse floats on the tation products proposed for use in the aftermath of the
water surface and can be dispersed large distances from Persian Gulf War had no convincing ev