Book 5 Complete PDF

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/348124630
The Andean Cloud Forest
Book · January 2021

DOI: 10.1007/978-3-030-57344-7
CITATION READS
1 216
1 author:
Randall W. Myster
Oklahoma State University - Oklahoma City
137 PUBLICATIONS 2,869 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
book 3 View project
All content following this page was uploaded by Randall W. Myster on 15 March 2021.
The user has requested enhancement of the downloaded file.

Randall W. Myster Editor
The Andean
Cloud Forest
Randall W. Myster
Editor

Editor
Randall W. Myster
Department of Biology
Oklahoma State University
Oklahoma City, OK, USA
ISBN 978-3-030-57343-0 ISBN 978-3-030-57344-7 (eBook)

https://doi.org/10.1007/978-3-030-57344-7
© Springer Nature Switzerland AG 2021

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Dedicated to the memory of my Norwegian
grandmother, Dora.
“Det som er elsket er alltid vakkert”.

Contents
1 Introduction�� 1
Randall W. Myster
2 Dynamics of Andean Treeline Ecotones: Between Cloud
Forest and Páramo Geocritical Tropes�� 25
Fausto O. Sarmiento
3 Análisis Regional En Ecosistemas De Montaña En Colombia:
Una mirada desde la funcionalidad del paisaje y los servicios
ecosistémicos�� 43
Paola Isaacs-Cubides, Julián Díaz, and Tobias Leyva-Pinto
4 Ecohydrology of Tropical Andean Cloud Forests�� 61
Conrado Tobón
5 Litterfall in Andean Forests: Quantity, Composition, and
Environmental Drivers�� 89
Wolfgang Wilcke
6 Arbuscular Mycorrhizal Fungi and Ectomycorrhizas in the
Andean Cloud Forest of South Ecuador�� 111
Ingeborg Haug, Sabrina Setaro, and Juan Pablo Suárez
7 Nesting Ecology of the Tucuman Amazon (Amazona tucumana)
in the Cloud Forest of Northwestern Argentina�� 131
Luis Rivera and Natalia Politi
8 Adaptive Strategies of Frugivore Bats to Andean Cloud Forests�� 151
Adriana Ruiz and Pascual J. Soriano
9 Neotropical Biodiversity: Hypotheses of Species Diversification
and Dispersal in the Andean Mountain Forests�� 177
Angela M. Mendoza-Henao and Juan C. Garcia-R
vii
viii Contents
10 Mapping Hydrological Ecosystem Services and Impacts of

Scenarios for Deforestation and Conservation of Lowland,
Montane and Cloud-Affected Forests�� 189
Mark Mulligan
11 Conclusions, Synthesis, and Future Directions �� 219
Randall W. Myster
Contributors
Julián Díaz Instituto de Investigación de Recursos Biológicos Alexander von

Humboldt, Bogota, Colombia
Juan C. Garcia-R Molecular Epidemiology and Public Health Laboratory,
Hopkirk Research Institute, School of Veterinary Science, Massey University,
Palmerston North, New Zealand
Ingeborg Haug Universitat Tubingen, Tubingen, Germany
Paola Isaacs-Cubides Instituto de Investigación de Recursos Biológicos Alexander
von Humboldt, Bogota, Colombia
Tobias Leyva-Pinto Universidad Nacional de Colombia, Bogota, Colombia
Conrado Tobon Universidad Nacional de Colombia, Medellin, Colombia
Angela M. Mendoza-Henao Departamento de Zoología, Instituto de Biología,
Universidad Nacional Autónoma de México, Mexico City, México
Mark Mulligan Department of Geography, King’s College London, London, UK
Randall W. Myster Department of Biology, Oklahoma State University, Oklahoma
City, OK, USA
Natalia Politi Insituto de Ecorregiones Andinas CONICET/Universidad Nacional
de Jujuy, S.S. de Jujuy, Jujuy, Argentina
Luis Rivera Insituto de Ecorregiones Andinas CONICET/Universidad Nacional
de Jujuy, S.S. de Jujuy, Jujuy, Argentina
Adriana Ruiz Postgrado en Ecología Tropical (ICAE), Facultad de Ciencias,
Universidad de Los Andes, Mérida, Venezuela
Fausto O. Sarmiento University of Georgia, Athens, GA, USA
Sabrina Setaro Wake Forest University, Winston-Salem, USA
ix
x Contributors
Pascual J. Soriano Laboratorio de Ecología Animal, Departamento de Biología,

Facultad de Ciencias, Universidad de Los Andes, Mérida, Venezuela
Juan Pablo Suárez Universidad Técnica Particular de Loja, Loja, Ecuador
Wolfgang Wilcke Institute of Geography and Geoecology (IfGG), Karlsruhe
Institute of Technology (KIT), Karlsruhe, Germany
Chapter 1
Introduction
Randall W. Myster
1.1 The Andes
The Andes or Andean Mountains (“Cordillera de los Andes” in Spanish) are located
along the western edge of South America (Fig. 1.1). They are the longest continuous
mountain range in the world—at least 7000 km—and have more volcanoes than any
other mountain range. These volcanoes include Mount Chimborazo (Fig. 1.2),
which has the point on Earth closest to the moon, and Mount Aconcagua, which is
the highest peak in the Andes at 6962 m a.s.l. The northern end of the Andes begins
in Venezuela (~10° N latitude), continues through Colombia, Ecuador, Peru, Bolivia,
Argentina and ends in Chile (~57° S latitude). The longitude of the Andes is between
~70° W and ~80° W (Knapp 1991).
The Andes are not just one mountain range, but rather a succession of parallel
and transverse mountain ranges—called cordilleras—along with their intervening
plateaus and depressions. The distinct eastern ranges are referred to collectively as
the Cordillera Oriental, and the distinct western ranges are referred to collectively
as the Cordillera Occidental. The directional trend of both the Cordillera Oriental
and the Cordillera Occidental are north-to-south, but the Cordillera Oriental bulges
eastward to form isolated peninsula-like ranges and high inter-montane plateau
regions. Researchers usually divide the Andes into the Southern Andes (the Chilean,
Fuegian, and Patagonian cordilleras), the Central Andes (the Peruvian cordilleras),
and the Northern Andes (the Ecuadorian, Colombian, and Venezuelan [or Caribbean]
cordilleras: Oncken et al. 2006).
Although temperature generally increases northward from Tierra del Fuego in
Chile (and southward from Venezuela) towards the equator, elevation, proximity to
the sea, rainfall, and topographic barriers to the wind create much variety in tem-
perature and other climatic conditions. Indeed, the Andes is a place of extremes
R. W. Myster (*)
Department of Biology, Oklahoma State University,
© Springer Nature Switzerland AG 2021 1

R. W. Myster (ed.), The Andean Cloud Forest,
https://doi.org/10.1007/978-3-030-57344-7_1
2 R. W. Myster
Fig. 1.1 The Andean Mountains, located along the Western edge of South America and marked in
dark gray
Fig. 1.2 “Alexander von Humboldt and Aimé Bonpland at the foot of Chimborazo” by Friedrich
Georg Weitsch, 1810
1 Introduction 3
where temperatures range from below freezing on the tops of mountains to very hot
in the lower elevation tropical forests, and precipitation ranges from dry in deserts
to very wet in those same tropical forests. While such climatic variation obviously
occurs among elevations, there are large differences among the aspects of slopes as
well, for example, those that face the Pacific, those that face the Amazon basin, or
those that face different directions within the Andes.
Given this large variation in environmental conditions and thus in potential
“niches” (Bazzaz 1996), it is not surprising that the Andes are an area of great bio-
diversity. For example, about 30,000 species of vascular plants live in the Andes—
roughly half of which are endemic—surpassing the diversity of any other “hotspot”
on earth (Hoorn et al. 2010). Moreover, many of these species have wide distribu-
tions, for example, animals such as bear and deer which are normally found in
North America exist at high elevations in the Andes, while animals such as monkeys
which are normally found in the Amazon exist at low elevations (overlap: Cardelús
et al. 2006). Indeed, the Andes can seem to be a microcosm of the entire Western
hemisphere! Adding to this biodiversity are species introduced by human activities
such as farming (e.g., horse, cow, pig, chicken). Indeed, crops commonly include
potato, maize, and beans raised for local consumption, and tobacco, cotton, and cof-
fee raised for export (see chapters in Myster 2007a). After cultivation is no longer
profitable, these areas often become pastures. Finally, mining for copper, gold, and
silver, as well as iron and tin, is common in the Andes of Peru, Chile, and Bolivia.
1.2 The Andean Cloud Forest
Cloud forests exist in mountainous regions all over the earth (Bruijnzeel et al. 2010).
A commonly adapted definition of cloud forests are “forests that are frequently
covered in cloud or mist” (Stadtmüller 1987; Hamilton et al. 1995) or “forests
affected by frequent and/or persistent ground-level cloud” (Grubb 1977). The
ground vegetation in these forests receive most of their moisture from fog and wind-
driven precipitation (Nadkarni and Wheelwright 2000) but also get some from rain-
fall. These low-level clouds profoundly affect the temperature and light regime of
Cloud forests as well (Mulligan and Burke 2005). Researchers have categorized
cloud forests (1) at low elevation (usually starting around 1200 m a.s.l. with 25–50%
moss cover on stems) where there is incipient and intermittent cloud formation, (2)
at high elevation (~2500 m a.s.l with 70–80% moss cover on stems) where cloud
condensation becomes most persistent, and (3) at the highest elevation (possibly
~3500 m a.s.l with more than 80% moss cover on stems) where the forest has an
“elfin” physiognomy created by exposure to wind-driven fog and rain (Bruijnzeel
et al. 2010). Not counting this elfin cloud forest, cloud forests are two-layered can-
opy systems with abundant mosses, lichens and epiphytes (Veneklaas and Van Ek
1990). Moreover, these cloud forest types blend into each other along gradients of
temperature, moisture, and elevation with average rainfalls ranging from 500 to
10,000 mm/year (Hamilton et al. 1995) and an average temperature of 17.7° C
(Bruijnzeel et al. 2010).
4 R. W. Myster
Cloud forests have the capacity to capture or strip water from clouds which may
result in an increased catchment water yield compared to other vegetation types.
They also have a high proportion of biomass in the form of epiphytes, fewer woody
climbers than in lower altitude tropical forests, high local biodiversity in terms of
shrubs, herbs, and epiphytes, high proportion of endemic species and wet, fre-
quently water-logged soils that typically have high organic matter contents
(Histosols). These volcanic soils are generally shallow with weakly developed hori-
zons (Leptosols) or they may be drought sensitive with low water holding capacity
(Regosols: Hamilton et al. 1995). Soils tend to be anaerobic, high in soil organic
matter (SOM), highly acidic, and may be both nitrogen (N) and phosphorus (P)
limited (Bruijnzeel et al. 2010).
Perhaps the most obvious way to investigate cloud forests is to focus on their
elevational gradient (Beck et al. 2008). Since this gradient is known, direct gradient
analysis can be used to explore how both abiotic (e.g., temperature, precipitation)
and biotic (e.g., plant, animal, and fungi taxa) parameters vary with either increas-
ing or decreasing altitude. For example, we know that with an increase in elevation,
both net primary productivity and precipitation (crown drip) increases as does
humidity, wind speed and cloudiness, while overall water use (evapotranspiration)
decreases as temperature declines. In addition, litter decomposition and soil organic
matter (SOM) turnover both slow down with increasing elevation but amount of
SOM and the carbon nitrogen ratio (C/N) increases, limiting nutrient availability.
Mineralization rates and concentrations of all macronutrients decrease with increas-
ing elevation. With increasing elevation, height of emergent trees, tree canopy
height, and number of strata decreases as does stem diameter and leaf area, while
moss, epiphytes (e.g., ferns and orchids) and tree fern cover all increase. With
increasing elevation, tree stems increasingly take on a crooked and gnarled physiog-
nomy, and bamboos often replace palms as the dominant undergrowth species.
Finally, biodiversity, turnover of biodiversity and endemism may reach its peak at
mid-elevation due to overlap of species distributions commonly found at higher and
lower elevations (Cardelús et al. 2006).
In the Andes as elsewhere, plant and animal species form statistically significant
associations (Myster and Pickett 1992; Myster 2012a). While these include species
associations found in deserts and grasslands (see chapters on the Andean grasslands
páramo, puna, and steppe in Myster 2012b), this book is mainly concerned with
forests, and forests within the Andes range from lowland forests at low elevations
(bordering, and similar to, Amazonian terra firme forest: Myster 2016) to, pre-
montane forests, montane forests and finally cloud forests at the higher elevations
(von Humboldt and Bonpland 1807). At the highest elevations, due to low atmo-
spheric pressure, it can take 5-min to boil a 3-min egg (author, pers. trekking experi-
ence)! Not just elevation determines the kind of forest found in the Andes, however,
but also other landscape characteristics such as aspect, slope, and the nature of the
substrate (Myster et al. 1997). Andean cloud forests are also defined by disturbance
including natural tree-fall (Myster 2014, 2015, 2017, 2018a) and landslides (Myster
and Sarmiento 1998), and human activity such as conversion to agriculture and
1 Introduction 5
pasture after slash and burn forest clearing techniques, and the building of roads and
urban areas (Myster 2004a, 2007b, 2009, 2010a, 2012a, 2013).
The focus of this book is Andean cloud forests, which are among the most diverse
ecosystems in the world (having 1/6 of all its’ plant species) and are important for
the hydrological and other biogeochemical cycles of large parts of the Neotropics.
This includes the carbon cycle and Andean Cloud forests may show reduced deposi-
tion and other effects of global warming/climate change in the future (Deutsch et al.
2008; Chen et al. 2009). These evergreen Andean cloud forests exist in the same
elevational zones as outlined above for cloud forests found elsewhere in the world,
and thus can also be characterized as low elevation cloud forests, high-elevation
cloud forests, and “elfin” cloud forests (Gould et al. 2006) exposed to wind-driven
fog and rain at the highest elevation. In addition to biodiversity and biogeochemical
cycling, Andean cloud forests are unique and important because they exist not just
over a large elevational (altitudinal) gradient, but also over a large latitudinal gradi-
ent. The interaction of these two gradients (Holdrige 1967) creates incredible varia-
tion and thus allows a more complete understanding of cloud forests than was
previously possible (Beck et al. 2008; Bruijnzeel et al. 2010; Hamilton et al. 1995;
Nadkarni and Wheelwright 2000; Stadtmüller 1987). Chapter authors will use this
framework to explore Andean cloud forest structure, function, and dynamics.
1.3 C
ase Study: Primary Cloud Forest at the Reserva
Biologica San Francisco
My first study site was the Reserva Biologica San Francisco (RBSF: 3° 58’ 30” S,
79° 4’ 25” W, Bussmann 2001; Beck et al. 2008, Fig. 1.3). The reserve covers
1000 ha of the northern slopes of Cordillera de Consuelo in Zamora-Chinchipe
Province, Ecuador adjacent to Podocarpus National Park. RBSF extends between
1800 m and 3150 m a.s.l and is covered by cloud forest. Soils are Dystrudepts and
Haplosaprists at lower elevations, and Petraquepts and Epiaquents at higher eleva-
tions (Bussmann 2003). Temperatures range from 15° to 17° C at lower elevations
to 9–11 ° C at upper elevations, and annual precipitation from 2200 mm per year at
lower elevations to 5000 mm per year at upper elevations (Bussmann 2003).
In January 2019, my field assistants and I first set up and then sampled all trees
at least 10 cm diameter at breast height (dbh) in ¼ ha plots at 10 different elevations
(1900 m, 2000 m, 2100 m, 2200 m, 2300 m, 2400 m, 2500 m, 2600 m, 2700 m,
2800 m) across from the Rio San Francisco, next to the RBSF field station. The dbh
measurement was taken at the lowest point where the stem was cylindrical, and for
buttressed trees above the buttresses. Trees were identified to family, genus, and
species using Martinez (2005) and the Missouri Botanical Garden website www.
mobot.org as taxonomic sources. In addition, voucher samples, kept at the RBSF
on-site herbarium, were compared to field samples to facilitate identification.
6 R. W. Myster
Fig. 1.3 A map of Ecuador showing the location of my four study sites, all in Andean cloud forest
(1) Reserva Biologica San Francisco, (2) Maquipucuna Reserve, (3) Guandera Biological Station,
and (4) Yanacocha Reserve. The Andean Mountains are indicated in brown and yellow
1.3.1 Curve-Fitting Floristic Patterns
Using this collected data set, the number of stems in each family was first complied
at each sampling point on the elevational gradient. The familial data was then sub-
jected to curve-fitting analysis (Wilson 1991; Guest 2013) for these four response
patterns (1) a symmetric unimodal pattern, (2) a skewed unimodal pattern, (3) a
linear pattern, or (4) a plateau pattern (Austin et al. 1994; Oksanen and Minchin
2002; Rydgren et al. 2003) using least-squares regression analysis after the appro-
priate transformation (Hill 1977; SAS 1985; Bongers et al. 1999; McCain 2005;
1 Introduction 7
Marini et al. 2010), as was done to generate dominance-diversity curves (Myster

2010a). Elevation was the independent variable and number of stems was the depen-
dent variable in all cases. In addition, two ordinations were performed on the famil-
ial data (R-type) and two ordinations on the elevational data (Q-type) (1) an R-type
principal components analysis (PCA: Pielou 1984) ordination using the correlation
matrix, (2) an R-type non-metric multidimensional scaling ordination (NMDS: SAS
1985), (3) an Q-type principal components analysis (PCA: Pielou 1984) ordination
using the correlation matrix, and (4) an Q-type non-metric multidimensional scaling
ordination (NMDS: SAS 1985; Rydgren et al. 2003). By computing both PCA and
NMDS, I could compare the results to examine possible distortion of the data by
PCA (a horseshoe effect: Legendre and Gallagher 2001).
There were 29 families in the data set where the most common were
Melastomataceae, Lauraceae, Clusiaceae, and Rubiaceae, and the rarest were
Solanaceae, Malvaceae, Annonaceae, Cyatheaceae, and Hypericaceae. In addition,
the most common species were Clusia sp., Nectandra membranaceae, and Miconia
punctata. Statistics revealed significant symmetric unimodal response curves for
Aquifoliaceae, Clusiaceae, Euphorbiaceae, Primulaceae, and Podocarpaceae,
skewed unimodal response curves for Lauraceae and Melastomataceae, a linear
response curve for Myrtaceae, and plateau response curves for Primulaceae and
Rubiaceae. The most important families defining ordination axes were Rubiaceae,
Primulaceae, Euphorbiaceae, and Anacardiaceae, and within those ordination
spaces, families Melastomataceae and Rubiaceae were most different. The most
important elevations defining ordination axis were 1900 m, 2400 m, 2700 m, and
2800 m, and within those elevations 1900 m, 2300 m, and 2400 m were most differ-
ent. The elevation pairs 2200 m/2500 m, 1900 m/2000 m, and 2600 m/2800 m were
most similar, and PCA distortions were minimal. In conclusion (1) families sampled
along this gradient had an individualistic pattern, (2) all four major kinds of response
pattern had significant curves among these families with a symmetric unimodal pat-
tern most common, and (3) ordinations suggested that species in the families
Melastomataceae and Rubiaceae have the most dissimilar ecological roles and tol-
erances (niches: Bazzaz 1996) while species in most of the families may be similar.
1.3.2 Curve-Fitting Physical Structure Patterns
Using the same data set, these structural parameters were computed for each plot at
each elevation: (1) the total number of tree stems, the total number of tree stems
divided into four size classes: 10 ≤ 19 cm dbh, 20 ≤ 29 cm dbh, 30 ≤ 39 cm dbh and
≥40 cm dbh, and mean dbh, (2) tree family, genus, and species richness, (3) fishers
α diversity using the formula in Fisher et al. (1943) as realized by the JavaScript
program in http://groundvegetationdb-web.com/ground_veg/home/diversity_index,
(4) the sum of the basal areas of all individual tree stems (∑πr2; where r = the dbh
of the individual stem/2), (5) above-ground biomass using the formula in Nascimento
and Lawrance (2002) suggested for tropical trees of these stem sizes, and (6) canopy
8 R. W. Myster
closure using the formula in Buchholz et al. (2004) for tropical trees. These param-
eters were subjected to curve-fitting analysis (Wilson 1991; Guest 2013) for these
four response patterns (1) a symmetric unimodal pattern, (2) a skewed unimodal
pattern, (3) a linear pattern, or (4) a plateau pattern (Austin et al. 1994; Oksanen and
Minchin 2002; Rydgren et al. 2003) using least-squares regression analysis after the
appropriate transformation (Hill 1977; SAS 1985; Bongers et al. 1999; McCain
2005; Marini et al. 2010). Elevation was the independent variable and number of
stems was the dependent variable in all cases. Curves with zero slope were not con-
sidered ecologically meaningful and not reported, even if significant.
The total tree stems increased with elevation from 772/ha (1900 m) to almost
1000/ha (2800 m), but not montonically decreasing from 2100 m to 2200 m, the
smallest stem size was always most common reaching 100% of stems at 2800 m,
and the next smallest size took almost all of the remaining stems. There were very
few stems over 29 cm dbh, four elevations had no tree stems over 29 cm dbh and
there were no tree stems larger than 39 cm dbh in any elevation above 2100 m. No
elevation was family, genus or species rich and fisher’s α diversity was always low.
Finally, total stems and % stems 20 ≤ 29 cm dbh had significant linear curves and
% stems ≥40 cm dbh had a significant plateau curve. In conclusion, as number of
stems increased with elevation size of stems decreased, so structural parameters
such as basal area and above-ground biomass (AGB) remained relatively constant.
Richness and diversity were low and also relatively constant regardless of elevation.
1.4 C
ase Study: Primary Cloud Forest at Maquipucuna
Reserve
The second study site was the Maquipucuna Reserve (MR: 0° 05’ N, 78° 37’ W;
www.maqui.org; Sarimento 1997; Myster and Sarmiento 1998, Fig. 1.3) located
20 km from the town of Nanegalito, Ecuador. The reserve lies between 1200 m and
1800 m a.s.l. and is classified as tropical lower montane wet/cloud forest (Edmisten
1970). It has deeply dissected drainages with steep slopes and has an annual precipi-
tation of 3198 mm (measured from Nanegal: Sarimento 1997). The temperature
ranges yearly between 14° and 25 °C, with an average temperature of 18 ° C. The
reserve’s fertile andisol soil is developed from recent volcanic ash deposits.
1.4.1 One ha Plot: Floristics and Physical Structure Sampling
In May 2012, my field assistants and I set up a 1 ha permanent plot in a primary
cloud forest at MR at an elevation of 1200 m a.s.l. We tagged and measured the
diameter at breast height (dbh) of all trees at least 10 cm dbh in the 1 ha plot. The
dbh measurement was taken at the nearest lower point where the stem was cylindrical
1 Introduction 9
and for buttressed trees it was taken above the buttresses. The tagged trees were
identified to species, or to genus in a few cases, using Romoleroux et al. (1997) and
Gentry (1993) as taxonomic sources. We also consulted the website of the Missouri
Botanical Garden (http://www.missouribotanicalgarden.org).
From this data set floristic tables of family, genus, and species were first com-
plied. There were a total of 18 families—Actinidaceae (24 stems), Asteraceae (1),
Brunneliaceae (1), Cecropiacea (7), Fabaceae (49), Lauraceae (118),
Melastomataceae (10), Mimosaceae (7), Monimiaceae (1), Moraceae (2),
Myristicaceae (9), Myrtaceae (4), Piperaceae (5), Rubiaceae (11), Solaneceae (5),
Tiliaceae (2), Urticaceae (12), Verbenaceae (24)—in the plot. Lauraceae was by far
the most common family and all families were dominated by a small number of
genera and species. The families Actinidaceae, Fabaceae, and Verbenaceae were
also common, but three families (Asteraceae, Brunneliaceae, Monimiaceae) had
only one stem. The families Asteraceae, Brunneliaceae, Fabaceae, Melastomataceae,
Moraceae, Myrtaceae, Piperaceae, Rubiaceae, Solaneceae, Tiliaceae, Urticaceae,
and Verbenaceae had a monotonic decline in stem number as stems get thicker. This
was not true, however, of the families Actinidaceae, Cecropiacea, Lauraceae,
Mimosaceae, Monimiaceae, and Myristicaceae. The most common species were
Erythrina megistophylla and Nectandra acutifolia (Myster 2004a, 2013, 2017).
These structural parameters were generated for the plot (1) the total number of
stems in the 1 ha plot, the mean and maximum among those stems, and the total
number of stems divided into four size classes: 10 ≤ 19 dbh, 20 ≤ 29 dbh, 30 ≤ 39
dbh and ≥40 dbh, (2) the stem dispersion pattern (random, uniform, clumped) com-
puted by comparing plot data to Poisson and negative binomial distributions using
Chi-square analysis and, if clumped, greens index was also computed to access
degree of clumping (Ludwig and Reynolds 1988), (3) canopy closure using the
formula in Buchholz et al. (2004) for tropical trees with the resulting percentage of
the 1 ha plot area closed, (4) total basal area as the sum of the basal areas of all
individual stems (Πr2; where r = the dbh of the individual stem/2), and (5) above-
ground biomass (AGB) using the formula in Nascimento and Lawrance (2002) sug-
gested for tropical trees of these stem sizes. There were 294 total tree stems in the
plot, with 153 stems between 10 and 19 cm dbh, 82 stems between 20 and 29 cm
dbh, 46 stems between 30 and 39 cm dbh, and 13 stems 40 cm dbh or greater. The
mean stem size was 23.2 cm dbh. Family richness was 18, genus richness 24, and
species richness 25. Basal area was 11.2 m2, above-ground biomass (AGB)
198.4 Mg, and canopy closure was 46.3%.
1.4.2 One ha Plot: Plant-Plant Replacements over Six Years
The 1 ha plot was resampled in 2014 and in 2016, using the same protocol. The data
was examined for plant–plant (here tree–tree) replacements happening over time,
which define plant community dynamics (Myster 2010b, 2012c, 2018b). Each liv-
ing tree had a location within the plot, a species, a dbh, and a neighborhood space
10 R. W. Myster
defined as a circular area with the tree at the center and the radius = 25 × dbh/2
(Canham et al. 2004), making the minimum neighborhood space for each sampling
a circle of diameter 1.25 m. This definition of neighborhood space captured the
effect of shading/canopy cover of each tree. The neighborhood space is graphed for
each living tree in the 2012 sampling, in the 2014 sampling, and in the 2016 sam-
pling (Fig. 1.4).
Comparing the 2014 sampling to the 2012 sampling, (1) 12 trees died of which
10 were in the first size class and 2 were in the second size class, (2) there were 14
new trees all in the first size class, and (3) for the rest of the trees 20 went from size
class 1 to size class 2, 5 went from size class 2 to size class 3, and 2 went from size
class 3 to size class 4. There were no trends due to species identity or spatial loca-
tion. Changes in the neighborhood spaces suggest that there were 3 none → one
replacements (Table 1.1), 4 one → none replacements, 3 one → one replacements,
1 one → many replacement, and 1 many → one replacement. Theory suggests that
there were probably many none → none replacements (Myster 2018b).
Comparing the 2016 sampling to the 2014 sampling, (1) 13 trees died of which
9 were in the first size class, 3 were in the second size class and 1 was in the third
size class, (2) there were 13 new trees all in the first size class, and (3) for the rest
of the trees 22 went from size class 1 to size class 2, 7 went from size class 2 to size
class 3, and 2 went from size class 3 to size class 4. There were no tends due to spe-
cies identity or spatial location. Changes in the neighborhood spaces suggest that
there were 4 none → one replacements (Table 1.1), 1 none → many replacement, 4
one → none replacements, 2 one → one replacements, and 2 many → one replace-
ments. Theory suggests that there were probably many none → none replacements
(Myster 2018b).
In the replacements found in both temporal sequences, size of stem seemed to be
a more important factor than species. Species replacement patterns followed the
ranking of species due to abundance—that is, the more common a species is the
more likely it would be in replacements. This is supportive of the neutral theory
(Hubbell 2001) of plant community dynamics that would preserve species ranking
of abundance, and thus species richness over time. Such rankings may be, however,
the result of niche-specific recruitment mechanisms (Myster 2018b). More sam-
pling is needed to explore these patterns and so I have replicated this plot and sam-
pling protocol in igapó forest (Myster 2019).
1.4.3 C
losed-Canopy Cloud Forest vs. Tree-Fall Gaps:
Recruitment Experiments
Tree seeds put out in closed-canopy cloud forest and in tree-fall gaps showed that
seed predation took more seeds than seed pathogens or that germinated for all tree
seed species and in both closed-canopy forest and in tree-fall gaps. In addition, seed
predation was greater in the closed-canopy forest than in the tree-fall gaps, but this
1 Introduction 11
a:2012 sampling
b:2014 sampling
c:2016 sampling
Fig. 1.4 Map of the 1 ha cloud forest sampled at Maquipucuna Reserve in (a) 2012, (b) 2014, and
(c) 2016. Circles indicate the location and size of the neighborhood spaces of all living trees. For
some trees, their neighborhood spaces extended outside of the plot
12 R. W. Myster
Table 1.1 The nine classes of plant–plant replacement (from Myster 2018b)
Replacement
class Description and example
None → none No plant died and no new plant joined the neighbourhood. New phyto-space
and/or new neighborhood space may, however, have been released by a
pre-existing plant and reoccupied by the same or another pre-existing plant. For
example, when a branch falls off a tree and another tree grows a branch into that
space(s)
None → one No plant died and one new plant joined the neighborhood. The new plant may
reoccupy space(s) that was released by a pre-existing plant. For example, when
a branch falls off a tree and a tree seedling recruits into that space(s)
None → many No plant died and more than one new plant joined the neighborhood. These new
plants may reoccupy space(s) that was released by a pre-existing plant. For
example, when a branch falls off a tree and more than one tree seedling recruits
into that space(s)
One → none One plant died and no new plant joined the neighborhood. The space(s) released
by the now dead plant may be reoccupied by a pre-existing plant. For example,
when a tree dies and another tree grows a branch into that space(s)
One → one One plant died and one new plant joined the neighborhood. The space(s)
released by the now dead plant may be reoccupied by the new plant. For
example, when a tree dies and another tree recruits into that space(s)
One → many One plant died and more than one plant joined the neighborhood. The space(s)
released by the now dead plant may be reoccupied by the new plants. For
example, when a tree dies and more than one tree recruits into that space(s)
Many → none More than one plant died and no new plant joined the neighborhood. The
combined or collective space released by the now dead plants may be
reoccupied by a pre-existing plant. For example, when a tree dies and as it falls
it pulls down another tree with it. Then another tree grows a branch into that
collective space
Many → one More than one plant died and one new plant joined the neighborhood. The
combined or collective space released by the now dead plants may be
reoccupied by this new plant. For example, when a tree dies and as it falls it
pulls down another tree with it. Then another tree recruits into that collective
space
Many → many Can be decomposed into the other replacement classes
trend was reversed for seed pathogens. Most seeds that survived, germinated.
Cecropia sp. seeds in the tree-fall gaps and Otoba gordoniifolia seeds in both
closed-canopy and tree-fall gap were most significantly different among all treat-
ments (Myster 2015, 2018a).
1 Introduction 13
1.5 C
ase Study: Secondary Cloud Forest at Maquipucuna
Reserve
Within MR are secondary cloud forests of recovering sugarcane plantations, banana

plantations, and pastures seeded with the grass Setaria sphacelata (Myster 2014;
Sarimento 1997). In June of 1996, I started a long-term study by selecting (as sug-
gested by local field assistants) six just abandoned agricultural fields: two recent
sugarcane (Saccharum officinarum) plantations, two recent banana (Musa sp.) plan-
tations, and two recent pastures seeded in Setaria sphacelata all at an elevation of
1300 m a.s.l. In each field, 25 5 m × 2 m contiguous plots were laid out making all
six fields 250 m2 rectangles. They were located within a few hundred meters of each
other and had the 25 m plot border next to primary cloud forest. The plots did not
have any remnant trees or sprouting tree roots at the beginning of the study, and
their tree seed bank was very small. Starting in 1997 each of the 5 m × 2 m subplots
of each of the six plots were sampled annually to identify each plant species and
sample its percent cover—an indication of a species’ ability to capture light and,
therefore, to dominate these areas in the process of becoming forested communi-
ties—estimated visually in relation to each plot’s area. We also sampled the diam-
eter at breast height (dbh) of each tree stem at least 1 cm. Maquipucuna plant
taxonomists, trained at the University of Georgia, USA, where voucher specimens
are kept on file, assisted in the identification of species by using specimens located
on site. They also used Romoleroux et al. (1997) and Gentry (1993) as taxonomic
sources, as well as consulting the website of the Missouri Botanical Garden (http://
www.missouribotanicalgarden.org).
These plots are part of the longest and largest old field study in the Neotropics
(Myster 2004a, b, 2007a, b, 2009, 2010a, b, 2012a, 2014, 2015, 2017, 2018a)
funded by the US National Science Foundation as part of the long-term ecological
research (LTER) program at the University of Puerto Rico (see Myster 2013 for
details). All the data is archived as LTERDATB #97, LTERDATB #100,
LTERDATB#101, and LTERDBAS#109 at the LTER website (http://luq.lternet.
edu). The LTER program has been a continuous part of the University of Puerto
Rico and the U.S. Forest Service in Puerto Rico since 1988.
1.5.1 P
ermanent Plot Sampling After Sugarcane Cultivation,
Banana Cultivation and Use as Pasture
Permanent plots in sugarcane fields showed grass dominating early (mostly sugar-
cane itself, Saccharum officinarum) and then declining to 50% cover levels, with
Panicum spp. and Brachiaria subquadripara also present, ferns (e.g., Nephrolepis
sp., Thelypteris deltoidea) and forbs (e.g., Desmodium sp., Bidens pilosa, Lantana
camara) as a small part of the flora that begins to decline immediately with cover of
woody species increasing steadily, trees showing this gradual increase with members
14 R. W. Myster
of the family Melastomataceae most common, and Piper aduncum more common
here than in other fields (Myster 2007b). In sugarcane fields, Acalypha platyphylla,
Piper aduncum, and Vernonia pallens were common. Further, P. aduncum was more
common in sugarcane fields than in other fields, and there were 18 tree species in
sugarcane fields and 21 in banana fields. For the sugarcane fields, total cover, an
indicator of stratification and developing structure, was greater compared with the
other fields, and banana greater than pasture, starting at 200% cover and gradually
declining with time. Species richness was also greater in the sugarcane fields com-
pared with the other fields, and banana had more species than did pasture.
In the plots located in recovering banana fields, grass and Musa sp. dominate
early and then decline, but to greater cover levels than in the sugarcane. Ferns attain
greater cover levels in banana than in sugarcane but forbs are a small part of the
flora. Also, woody species increase, but at smaller cover levels than in sugarcane.
There were many of the same tree species in banana fields as in sugarcane fields, but
with additional species (e.g., Cecropia monostachya) present. Piper aduncum was
frequent in the permanent plots. Total stems were slightly more in the banana fields
compared with the sugarcane fields, and pastures had few stems. Compared with
banana, mean stem height was also greater in the sugarcane and increased through
time, while pasture stems remained small. Basal area stabilized at 500 cm2 for sug-
arcane and banana fields.
Finally, in the seeded pastures, grass dominated almost completely (mainly the
planted grass Setaria sphacelata), with a few woody plants showing in year five.
Fern and forb cover was small, and pastures had few trees, but the species they did
have were common to the sugarcane and banana fields (Myster 2007b). There were
life-form changes such as a domination by grass (e.g., Panicum sp., Axonopus com-
pressus) that peaks in cover in year three and then declines, fem (Nephrolepis sp.,
Thelypteris deltoidea) and herbaceous (e.g., Commelina sp., Desmodium adscen-
dens, Clidemia hirta) cover that also peaks in year three but at lower cover levels
than grass, woody vines (e.g., Ipomea sp.) that continue to increase in cover over the
first 5 years, and trees and shrubs that enter, peak, and decline in an individualistic
pattern. Community parameters in pasture plots show after 5 years (1) two or three
strata or horizontal levels that developed quickly, (2) total stems at 200/250 m2 plot,
(3) tree richness constant at 20 species, (4) tree-stem evenness approximately 0.7,
(5) productivity peaking at year two at 0.5 kg/m2/year and then declining slightly,
(6) turnover decreasing over the first 5 years to 50%, and (7) mean height at 250 cm
and total basal area of stems at 1000 cm2.
Data in each of the 25 subplots were then pooled to tabulate a total percent cover
for each species for the entire plantation or pasture per sampling year. Those total
percent cover data were first sorted in decreasing order, then log10-transformed,
and finally plotted for the first 20, most common, species to create dominance-
diversity curves for the sugarcane and banana plantations and pastures, for each
sampling year. Sampling data were then fitted to Preston’s log-normal model,
MacArthur’s broken stick model, and both the geometric and the harmonic series
using least-squares regression after the appropriate transformation. I found that the
sugarcane and banana plantations had a significant log-normal pattern for the first
1 Introduction 15
5 years and then flattened out. The plots in pasture, however, showed a significant
geometric pattern over the first 7 years of succession. Pastures lag behind planta-
tions and do not resemble them within this sampling time frame (Myster 2010a).
Total cover, an indicator of stratification and developing structure, was greater in
the pastures than in the other fields. The cover started at 200% in pastures and
gradually declined. Species richness was greater in sugarcane fields than in the
other fields, with more species in pastures than in banana fields. The total number of
stems was greatest in sugarcane fields and pastures, whereas mean stem height was
greatest in the pastures. Basal area increased to 500 cm2 in sugarcane and banana
fields but remained low in pastures.
Further analysis of the plot data showed that the plots in Ecuador plantations
show a significant log-normal pattern for the first 5 years with a flattening out after
that. The plots in Ecuador pasture, however, show a significant geometric pattern
over the first 7 years of succession. Pastures lag behind plantations and do not
resemble them within this sampling time frame. I also found that species richness
had a significant positive relationship with productivity, where the slopes of the
regression lines decreased over time for all fields taken together suggested the early
upswing, leveling off, and later downswing of a unimodal curve; and this unimodal
pattern held true both in the recovering sugarcane (Saccharum officinarum L.) and
in the recovering Banana (Musa sp.) plantations. There was a delay, however, in the
development of the unimodal pattern in the seeded pasture, perhaps due to root
competition between the residual tussock grass (Setaria sphacelata) and Neotropical
tree seedlings (Myster 2009).
Finally I used the plot data to investigate significant positive associations between
pairs of old field species were first computed and then clustered together into larger
and larger species groups. I found that no pasture or plantation had more than 5% of
the possible significant positive associations, clustering metrics showed groups of
species participating in similar clusters among the five pasture/plantations over a
gradient of decreasing association strength, and there was evidence for repeatable
communities—especially after Banana cultivation—suggesting that past crops not
only persist after abandonment but also form significant associations with invading
plants (Myster 2013).
1.5.2 R
ecruitment Experiments in Recovering Sugarcane
Fields, Banana Fields and Pastures
Seed rain sampling showed sugarcane fields had twice the number of dispersed
seeds as banana fields, and 20 times the number in pasture. Seedlings germinating
from soil samples were low in all fields and pastures, with a few species found in
multiple fields/pastures. Seeds put out in these same six fields showed that most
seeds were lost to predators, most of the remaining seeds were destroyed by seed
pathogens, and most of the seeds left germinated (Myster 2004b). There was more
16 R. W. Myster
predation in the forest compared to the old field/forest border and more seed disease
away from the forest compared to in it. Tree seeds put out in sugarcane fields,
banana fields and pastures showed that seed predation took more seeds that seed
pathogens or that germinated in all fields and pastures, but this was less that in pri-
mary closed-canopy forest and tree-fall gaps (see Sect. 1.4.3). Pathogens took most
of the remaining seeds and most seeds that survived, germinated. Solanum ovalifo-
lia seeds (only species that had seeds germinate without any losses to pathogens) in
banana fields and Piper aduncum in all fields and pastures were the most signifi-
cantly different among all treatments. Insect predation was lower in the forest bor-
der microsite for P. aduncum, seed disease greater at 10 m from border. All planted
seedlings died in the pasture, 25% survived in banana and 15% survived in sugar-
cane, and there was less growth away from the forest border (Myster 2004b). Results
taken together with primary forest suggest that forests may recover faster after
human disturbance (here agriculture) than after natural disturbance (here tree-fall:
Myster 2015).
1.5.3 P
ermanent Plot Sampling and Recruitment Experiments
in a Recovering Landslide
Plant cover in a permanent landslide plot showed four fern families and 20 vascular
plant families, with species in Asteraceae, Melastomataceae, and Poaceae most
common. PCA showed that plots separated best on axes defined by the families
Cecropiaceae, Urticaceae, Melastomataceae, Papilionaceae, Asteraceae, and
Araceae and clumping of families in PCA space suggesting common successional
strategies. In two landslides, 1304 seeds were trapped from the seed rain, mainly in
the family Asteraceae. Germinated soil samples from the same landslides had 475
seedlings including nonvascular and vascular families where species in Asteraceae
and Piperaceae dominated. Further, ordination showed that (1) for seed rain, seed-
propagule pool, and plant cover, spatial variation was dominated by differences
between the two landslides rather than within-landslide plot differences and (2) the
combined seed rain and seed pool data could predict the percent cover of the family
Verbenaceae and that the current plant cover families could predict Asteraceae seeds
and seedlings (Myster and Sarmiento 1998).
1.6 C
ase Study: Primary Cloud Forest at Guandera
Biological Station
The third study site was the Guandera biological station (GBS: 0° 36’ N, 77° 42’ W:
www.jatunsacha.org/guandera-reserve-andbiological-station: Bader et al. 2007;
Nierop et al. 2007, Fig. 1.3) situated on the inner flank of the eastern Cordillera at
1 Introduction 17
approximately 11 km from the town of San Gabriel in northern Ecuador. The area is
of volcanic origin and has deep dark humic Andosols developed in old volcanic
ashes. Annual precipitation is 1700 mm. Diurnal temperature fluctuations range
from 4 to 15 °C but annual temperature fluctuations are low (monthly means of
maximum temperature vary between 12 and 15 °C: Bader et al. 2007). Within the
GBS, primary cloud forest occurs between the agricultural areas, mainly potato
(Solanum tuberosum L.) cultivation below 3300 m a.s.l., and the páramo grasslands
above 3640 m a.s.l. The GBS consists of upper montane cloud forest at lower eleva-
tions and sub-alpine dwarf forest at higher elevations. The upper montane cloud
forest is dominated by the trees Clusia flaviflora Engl. (Clusiaceae), Ilex colombi-
ana Cuatrec. (Aquifoliaceae), Weinmannia cochensis Hieron. (Cunoniaceae),
Miconia tinifolia Naudin (Melastomataceae) and Gaiadendron punctatum (Ruiz
and Pavón) G.Don. (Loranthaceae) and the epiphytes Tillandsia sp. (Bromeliaceae)
and Blechnum schomburgkii (Klotzsch) C.Chr. (Blechnaceae). The sub-alpine dwarf
forest is dominated by the trees Miconia tinifolia and Weinmannia cochensis, a
shrub Gaiadendron punctatum, a fern Blechnum schomburgkii, and a bamboo
Neurolepis aristata (Munro) Hitchc. (Poaceae).
1.6.1 C
losed-Canopy Forest vs. Tree-Fall Gaps: Recruitment
Experiments
Ten primary cloud forest areas were selected randomly at GBS in May 2015. Five
had new tree-fall gaps between the normal range of 100–300 m2 in area (Brokaw
1982) and five were in closed-canopy forest. Ripe fruits were collected by hand
(using gloves to reduce transfer of infections or human odor) from one local indi-
vidual of Solanum stenophyllum Bitter (Solanaceae: bird dispersed), one local indi-
vidual of Clusia flaviflora Engl. (Clusiaceae: mammal dispersed), and one local
individual of Palicourea amethystina (Ruiz & Pav.) DC. (Rubiaceae: bird dis-
persed). Seed species are ordered by increasing seed mass (www.data.kew.org/sid/).
Seeds were examined for damage using a dissecting microscope and by floating
them in water. Visually damaged seeds and those that floated were discarded. Ten
viable seeds of each of the three test species were then placed in three separate plas-
tic 9-cm-diameter Petri dishes spaced 50 cm apart and placed in the center of each
closed-canopy forest area and in the center of each tree-fall gap. Three dishes per
gap × ten gives a total of 30 dishes. After 2 weeks in the field, losses due to seed
predation or pathogens were counted and the remaining seeds were tested for
germination.
In both closed-canopy forest and tree-fall gaps, S. stenophyllum Dunal seeds suf-
fered the greatest losses to predators, P. amethystina seeds had the greatest germina-
tion and C. flaviflora seeds had the greatest losses to pathogens. Comparison with
data from MR primary cloud forest (Sect. 1.4.3) and secondary cloud forest (Sect.
1.5.2) showed the following: (1) Solanum sp. suffered the greatest losses for seeds
18 R. W. Myster
lost to seed predators in general but Cecropia sp. and Ficus sp. also had high losses
at MR primary cloud forest; (2) for seeds lost to pathogens, species that lost the
most seeds were unique to each study site: Clusia flaviflora seeds at Guandera pri-
mary cloud forest, Cecropia sp. seeds at MR primary cloud forest, and Piper adun-
cum L. seeds at MR secondary cloud forest; (3) for seeds that germinated the most,
species were again unique to the study site: Palicourea amethystina seeds at
Guandera primary cloud forest, Otoba gordoniifolia (A. DC.) A.H. Gentry seeds at
MR primary cloud forest, and Solanum ovalifolium Dunal seeds at MR secondary
cloud forest; and (4) in general, forest types differed significantly for both seed
predation and seed pathogens.
Within the primary cloud forest at MR, there was a significant difference among
tree seed species for pathogens and a significant difference among the tree seed spe-
cies for germination, and within the secondary cloud forest at MR, there was a sig-
nificant difference among tree seed species for pathogens. As elevation increases in
primary cloud forests, the proportion of seed that germinates remain largely con-
stant, but the major seed loss shifts from being due to predators to being due to
pathogens. Conversion to agriculture also leads to seeds mainly lost to predators,
but individual species loss levels depended on what crop had been planted previ-
ously (Myster 2018a).
1.7 Case Study: Primary Cloud Forest at Yanacocha Reserve
The last study site was Yanacocha Reserve (YR: 0° 07’ S, 78° 35’ W: http://fjo-
cotoco.org/reserves-yanacocha, Fig. 1.3) managed by Fundacion Jocotoco and sup-
ported by a World Land Trust land purchase and Carbon Balanced funding is located
on the northeastern slope of the Pichincha Volcano about 45 min northwest of Quito
along the old Nono-Mindo Road on route to the Mindo Valley. The reserve was
established in 2001 to protect the black-breasted Puffleg (Eriocnemis nigrivestis)
whose known range is restricted to the Pichincha Volcano. The YR is mainly high-
elevation elfin Polylepis sp. forest.
1.7.1 One ha Plot: Floristics and Physical Structure Sampling
In May 2015, one 2500 m2 (50 m × 50 m) plot was established in YR primary cloud
forest at 3400 m a.s.l., in a random location suggested by local field assistants, and
measured using the exact protocol as in previous study sites. Voucher specimens,
kept on file at the University of Georgia USA, assisted plant taxonomists in species
identification. They also used Romoleroux et al. (1997) and Gentry (1993) and con-
sulted the web site of the Missouri Botanical Garden (www.mobot.org). Data col-
lected from all plots were used to compile floristic tables of family, genus, and
1 Introduction 19
species. Also calculated were (1) the total number of stems, the mean dbh among
those stems, and the total number of stems divided into four size classes: 10 ≤ 19 cm
dbh, 20 ≤ 29 cm dbh, 30 ≤ 39 cm dbh, and ≥40 cm dbh; (2) total basal area as the
sum of the basal areas of all individual stems (π*r2; where r = the dbh of the indi-
vidual stem/2); (3) above-ground biomass (AGB) using the formula in Nascimento
and Lawrance (2002) and suggested for tropical trees of these stem sizes; and (4)
canopy closure using the formula in Buchholz et al. (2004) for tropical trees.
Palicourea sp. was the only species found at YR primary cloud forest and at MR
primary cloud forests (see Sect. 1.4.1); Vernonia pallens Sch.Bip., Erythrina megis-
tophylla Diels, Nectandra sp., and Miconia sp. were found in both primary and
secondary plots at MR (see Sect. 1.5.1) and Miconia sp. was the only species in
common between the MR secondary plots and the primary plot at YR. The mean
stem size was similar between the primary MR plots and the YR plot, but the YR
plot had more total stems and more stems in each size category, which lead to more
basal area, above-ground biomass, and canopy closure at YR compared to MR. In
the secondary plots, there were no stems larger than 29 cm dbh at breast height
which lead to a much smaller mean stem size and lower basal area, above-ground
biomass, and canopy closure compared to the primary plots at both sites. For the
primary cloud forest at MR, an increase in elevation changed the species-level flo-
ristics more than conversion to and then abandonment from agriculture; however,
while a rise in elevation increased the number of stems, agriculture reduced stem
size structure. The plot at YR had a stem density of 193 stems/ha, with 91 stems
between 10 ≤ 19 cm dbh, 56 stems between 20 ≤ 29 cm dbh, 24 stems between
30 ≤ 39 cm dbh, and 22 stems >40 cm dbh, Mean dbh was 23.1 cm, family richness
was 11, genus richness was 13, species richness was 13, basal area was 5.2 m2,
above-ground biomass was 87.2 Mg, and canopy closure was 65.2% (Myster 2017).
1.8 Conclusions
In primary Andean cloud forest, as elevation increased the plant families Actinidacaea
and Fabaceae were gained and the plant family Clusiaceae was lost. This was also
seen when latitude increased. In primary Andean cloud forest, as elevation increased
the number of tree stems increased while the size of the stems decreased. This lead
to other structural parameters being fairly constant. Richness and diversity were low
everywhere. This was also seen when latitude increased. Andean cloud forest
recruitment was dominated by seed predation and its sources of variation. Seed
pathogens could play a secondary part in some cases.
Secondary Andean cloud forest starts different from primary Andean cloud for-
est in both floristics and physical structure, but with time begins to resemble it, first
floristically then structurally. This process is slower after landslides, compared to
sugarcane and banana fields, pasture is between these two cases. Recruitment mech-
anisms are very similar to primary cloud forest in relative importance.
20 R. W. Myster
About this Book

Because cloud forests in the Andes are among the most important ecosystems of the
Neotropics, they demand further study in order to better prepare for our shared
human future. Here I take advantage of my 25+ years in the Andean cloud forests of
Ecuador working with professors, researchers, local students, and technicians at
four difference study sites (Maquipucuna Reserve, Guandera Biological Station,
Yanacocha Reserve, Reserva Biologica San Francisco), with assistance from the
LTER program at the University of Puerto Rico (http://luq.lter.network), to edit this
book. One may check my personal website (www.researchgate.net/profile/Randall_
Myster) for a complete listing and hard copies of those Andean cloud forest publica-
tions. Although this will be the first research book focused exclusively on cloud
forests in the Andean mountains—and thus its chapter scope and detail is unique—
there have been several books written on cloud forests in the past (e.g., Stadtmüller
1987; Hamilton et al. 1995). Among these books, three recent books on cloud for-
ests in the Neotropics (Beck et al. 2008; Bruijnzeel et al. 2010; Nadkarni and
Wheelwright 2000) provide background and context for this book.
Acknowledgements I thank Arcenio Barras, Michael Dilger, Bernardo Castro, Jorge Reascos,
Bert Wittenberg Rebeca Justicia, and Rodrigo Ontaneda (Maquipucuna Reserve), Jose Cando and
Geovanna Coello (Guandera Biological Station), and Jessica Paccha, Edgar Dario Ramon Castillo,
and Pedro Paladines (Reserva Biologica San Francisco) for their help in facilitating my research in
cloud forest in Ecuador. I also thank N. V. L. Brokaw and the LTER program at the University of
Puerto Rico for their support at Maquipucuna Reserve. Finally, I thank Eda Meléndez and her staff
for managing the LTER data sets. Some of the research was performed under grants BSR-8811902
and DEB-9411973 from the National Science Foundation to the Institute for Tropical Ecosystem
Studies, University of Puerto Rico, and to the USDA Forest Service International Institute of
Tropical Forestry as part of the Long-Term Ecological Research Program in the Luquillo
Experimental Forest. Additional support was provided by the Forest Service and the University of
Puerto Rico.
References
Austin MP, Nicholls AO, Doherty MD, Meyers JA (1994) Determining species response functions
to an environmental gradient by means of a β-function. J Veg Sci 5:215–228
Bader MY, Geloof I, Rietkerk M (2007) High solar radiation hinders tree establishment above the
alpine treeline in northern Ecuador. Plant Ecol 191:33–45
Bazzaz FA (1996) Plants in changing environments: linking physiological, population and com-
munity ecology. Cambridge University Press, Cambridge
Beck E, Bendix J, Kottke I, Makeschin F, Mosandl R (2008) Gradients in a tropical mountain
ecosystem of Ecuador. Springer-Verlag, Berlin
Bongers F, Poorter L, Van Rompaey RSAR, Parren MPE (1999) Distribution of twelve moist forest
canopy tree species in Liberia and Cote d’Ivoire: response curves to a climatic gradient. J Veg
Sci 10:371–382
Brokaw NVL (1982) The definition of treefall gap and its effect on measures of forest dynamics.
Biotropica 11:158–160
Bruijnzeel LA, Scatena FN, Hamilton LS (2010) Tropical montane cloud forests: science for con-
servation and management. Cambridge University Press, Cambridge
1 Introduction 21
Buchholz T, Tennigkeit T, Weinreich A (2004) Maesopsis eminii—a challenging timber tree spe-
cies in Uganda—a production model for commercial forestry and smallholders. Proceedings of
the international union of forestry research organizations (IUFRO) conference on the econom-
ics and management of high productivity plantations, Lugo, Spain
Bussmann RW (2001) The montane forests of Reserve Biologica San Francisco. Erde 132:9–25
Bussmann RW (2003) The vegetation of Reserva Biológica San Francisco, Zamora-Chinchipe,
Southern Ecuador—a phytosociological synthesis. Lyonia 12:71–177
Canham CD, LePage PT, Coates KD (2004) A neighborhood analysis of canopy tree competition:
effects of shading versus crowding. Can J For Res 34:778–787
Cardelús CL, Colwell RK, Watkins JE (2006) Vascular epiphyte distribution patterns: explaining
the mid-elevation richness peak. J Ecol 94:144–156
Chen IC, Shiu HJ, Benedick S, Holloway JD, Cheye VK, Barlow HS (2009) Elevation increases
in moth assemblages over 42 years on a tropical mountain. Proc Natl Acad Sci 106:1479–1483
Deutsch CA, Tewksbury JJ, Huey RB, Sheldon KS, Ghalambor CK, Haak DC, Martin PR (2008)
Impacts of climate warming on terrestrial ectotherms across latitude. Proc Natl Acad Sci
105:6668–6672
Edmisten J (1970) Some autoecological studies of Ormosia krugii. In: Odum HT, Pigeon RF (eds)
A tropical rain forest. National Technical Information Service, Springfield. Chapter D-8
Fisher RA, Corbet AS, Williams CB (1943) The relation between the number of species and the
number of individuals in a random sample of an animal population. J Anim Ecol 12:42–58
Gentry A (1993) A field guide to woody plants of Northwest South America (Colombia, Ecuador,
Peru). Conservation International, Washington, DC
Gould WA, González G, Carrero-Rivera G (2006) Structure and composition of vegetation along
an elevational gradient in Puerto Rico. J Veg Sci 17:653–664
Grubb PJ (1977) Control of forest growth and distribution on wet tropical mountains: with special
reference to mineral nutrition. Annu Rev Ecol Syst 8:83–107
Guest RG (2013) Numerical methods of curve fitting. Cambridge University Press, Cambridge
Hamilton LS, Juvik JO, Scatena FN (1995) Tropical 25 montane cloud forests. Springer-Verlag,
New York
Hill MO (1977) Use of simple discriminant functions to classify quantitative phytosociological
data. In: Diday E, Lebart L, Pages JP, Tomassone R (eds) First Inern. Symposium on data anal-
ysis and informatics, Versailles, 7–9 Sept. 1977. Vol. 1. Institute de Recherche d’informatique
et d’automatique, Domaine de Voluceau, Rocqnencourt B. P. 105, 78150 le Chesney
Holdrige LR (1967) Life zone ecology. Tropical Science Center, San Jose
Hoorn C, Wesselingh FP, ter Steege H, Bermudez MA, Mora A, Sevink J, Sanmartín I, Sanchez-
Meseguer A, Anderson CL, Figueiredo JP, Jaramillo C, Riff D, Negri F, Hooghiemstra H,
Lundberg J, Stadler T, Särkinen T, Antonelli A (2010) Amazonia through time: Andean uplift,
climate change, landscape evolution, and biodiversity. Science 330:927–931
Hubbell SP (2001) The unified neutral theory of biodiversity and biogeography. Princeton
University Press, Princeton, NJ
Knapp G (1991) Andean ecology: adaptive dynamics in Ecuador. Westview Press, Boulder
Legendre P, Gallagher ED (2001) Ecologically meaningful transformations for ordination of spe-
cies data. Oecologia 129:271–280
Ludwig JA, Reynolds JF (1988) Statistical ecology. Wiley, New York
Marini L, Bona E, Kunin WE, Gaston KJ (2010) Exploring anthropogenic and natural processes
shaping fern species richness along elevational gradients. J Biogeogr 38:78–88
Martinez CEC (2005) Manual de Botanica: Sistematica, Etnobotanica y Metodos de Estudio en el
Ecuador. Ditorial Universitatia, Quito
McCain CM (2005) Elevational gradients in diversity of small mammals. Ecology 86:366–372
Mulligan M, Burke S (2005) DFID FRP Project ZF0216 Global cloud forests and environmental
change in a hydrological context
Myster RW (2004a) Post-agricultural invasion, establishment and growth of Neotropical trees. Bot
Rev 70:381–402
22 R. W. Myster
Myster RW (2004b) Regeneration filters in post-agricultural fields of Puerto Rico and Ecuador.
Plant Ecol 172:199–209
Myster RW (2007a) Post-agricultural succession in the Neotropics. Springer-Verlag, Berlin
Myster RW (2007b) Early successional pattern and process after sugarcane, banana and pasture
cultivation in Ecuador. N Z J Bot 46:101–110
Myster RW (2009) Are productivity and richness consistently related after different crops in the
Neotropics? Botany 87:357–362
Myster RW (2010a) Testing dominance-diversity hypotheses using data from abandoned planta-
tions and pastures in Puerto Rico and Ecuador. J Trop Ecol 26:247–250
Myster RW (2010b) A comparison of tree replacement models in oldfields at Hutchenson Memorial
Forest. J Torrey Bot Soc 137:113–119
Myster RW (2012a) A refined methodology for defining plant communities using data after sug-
arcane, banana and pasture cultivation in the Neotropics. Sci World J 2012:9. https://doi.
org/10.1100/2012/365409
Myster RW (2012b) Ecotones between forest and grassland. Springer-Verlag, Berlin
Myster RW (2012c) Plants replacing plants: the future of community modeling and research. Bot
Rev 78:2–9
Myster RW (2013) Long-term data from fields recovering after sugarcane, banana and pasture
cultivation in Ecuador. Dataset Pap Ecol 2013:468973. https://doi.org/10.7167/2013/46873.
10 pages
Myster RW (2014) Primary vs. secondary forests in the Neotropics: two case studies after agri-
culture. In: Forest ecosystems: biodiversity, management and conservation. Nova Publishers,
New York, pp 1–42
Myster RW (2015) Seed predation, pathogens and germination in primary vs. secondary Cloud
forest at Maquipucuna Reserve, Ecuador. J Trop Ecol 31:375–378
Myster RW (2016) Forest structure, function and dynamics in Western Amazonia. Wiley-
Blackwell, Oxford
Myster RW (2017) Gradient (elevation) vs. disturbance (agriculture) effects on primary cloud for-
est in Ecuador: floristics and physical structure. N Z J For Sci 47:1–7
Myster RW (2018a) Gradient (elevation) vs. disturbance (agriculture) effects on primary cloud for-
est in Ecuador: seed predation, seed pathogens and germination. N Z J For Sci 48:4
Myster RW (2018b) The nine classes of plant-plant replacement. Ideas Ecol Evol 11:29–34
Myster RW (2019) Igapó (black-water) forests of the Amazon Basin. Springer-Verlag, Berlin
Myster RW, Pickett STA (1992) Dynamics of associations between plants in ten old fields through
31 years of succession. J Ecol 80:291–302
Myster RW, Sarmiento FO (1998) Seed inputs to microsite patch recovery on tropandean land-
slides in Ecuador. Restor Ecol 6:35–43
Myster RW, Thomlinson JR, Larsen MC (1997) Predicting landslide vegetation in patches on
landscape gradients. Landsc Ecol 12:299–307
Nadkarni NM, Wheelwright NT (2000) Monteverde: ecology and conservation of a Tropical Cloud
forest. Oxford University Press, New York
Nascimento HEM, Lawrance WF (2002) Total aboveground biomass in central Amazonian rain-
forest: a landscape-scale study. For Ecol Manag 68:311–321
Nierop K, Tonneijck GJ, Jansen FH, Verstraten JM (2007) Organic matter in volcanic ash soils
under forest and paramo along an Ecuadorian altitudinal transect. Soil Sci Am J 71:1119–1127
Oksanen J, Minchin PR (2002) Continuum theory revisited: what shape are species responses
along ecological gradients? Ecol Model 157:119–129
Oncken O, Chong G, Franz G, Giese P, Götze H, Ramos VA, Strecker MR, Wigger P (2006) The
Andes: active subduction orogeny. Frontiers in earth sciences. Springer-Verlag, Berlin
Pielou EC (1984) The interpretation of ecological data: a primer on classification and ordination.
Wiley, New York
Romoleroux K, Foster R, Valencia R, Condit R, Balslev H, Losos E (1997) Especies Lenosas (dap
→ 1 cm) encontradas en dos hectareas de un bosque de la Amazonia ecuatoriana. In: Valencia
1 Introduction 23
R, Balslev HR (eds) Estudios Sobre Diversidad y Ecologia de Plantas. Pontificia Universidad

Catolica del Ecuador, Quito, pp 189–215
Rydgren K, Okland RH, Okland T (2003) Species response curves along environmental gradients.
A case study from SE Norwegian swamp forests. J Veg Sci 14:869–880
Sarimento FO (1997) Arrested succession in pastures hinders regeneration of tropandean forests
and shreds mountain landscapes. Environ Conserv 24:14–23
SAS User’s Guide: Statistics (1985) SAS Institute, Cary, NC
Stadtmüller T (1987) Cloud forests in the humid tropics. United Nations University, Turrialba
Veneklaas EJ, Van Ek R (1990) Rainfall interception in 2 tropical montane rain-forests, Colombia.
Hydrol Process 4:311–326
Von Humboldt A, Bonpland A (1807) Essai sur la geographie des plantes – accompagne´ d’un
tableau physique des re´gions e´quinoxiales, fonde´ sur des mesures exećuteés, depuis le
dixie`me degre´ de latitude boreále jusqu’au dixie`me degre´ de latitude australe, pendant les
anneés 1799, 1800, 1801, 1802 et 1803. – Levrault et Schoell, Paris
Wilson JB (1991) Methods for fitting dominance/diversity curves. J Veg Sci 2:35–46
Chapter 2
Dynamics of Andean Treeline Ecotones:
Between Cloud Forest and Páramo
Geocritical Tropes
Fausto O. Sarmiento
2.1 Andean Critical Biogeography of Scale
In the new critical biogeography framework, Cloud Forests are conceived as scalar
artifacts of both, historicity’s (c.f. temporal) and spatiality’s (c.f. areal) interdigita-
tions of the socio-ecological production mountainscapes. Particularly in the tropical
Andes, where the greatest concentration of Tropical Montane Cloud Forest (TMCF)
ecosystems exists, most of the research endeavor has been devoted to spatio-
temporal “landscape characterization” of many functional traits or ecological niches
available along the verdant gradient (see Chap. 1), and to create a “sense of place”
that could make them unique; they persist at present appealing to the eye and to the
mind. Because of my affinity with the neotropical realm, I include most examples
or case studies from the tropical and subtropical Andes, yet extreme cases of cloud
forests can be pointed from the Patagonian or even the Magallanic Andes of the
Terra Australis, where the mystical fog shrouding is as spectacular as in the Andean
crescent. When possible, I use vernacular descriptors of pueblos originarios to
describe the terms used regionally for the cloud forest ecosystem, wherever they
might be located.
2.1.1 Annotations of Narrative Framework
I use the phonetic alphabet to write “Kichwa” words with cursive, and I place them
within quotation marks; other foreign words, including scientific notations are also
italicized but without marks. I use double quotation marks to denote “alternative
F. O. Sarmiento, PhD (*)

University of Georgia, Athens, GA, USA
e-mail: [email protected]

https://doi.org/10.1007/978-3-030-57344-7_2
26 F. O. Sarmiento
meaning” or to directly cite quotes from other authors verbatim. Geocritical texts
are written in active voice, in the first person.
This chapter intersects unorthodox angles in the geoecological study of cloud
forests, one that requires readers to familiarize themselves not only with TMCF
terminology and morphology of the mountain landscape (a-la Carl O. Sauer), but
also with the alternative explanations of geocriticism of decolonial ecologies of the
Andes, and the need to assume a transdisciplinary approach to understand mountain
studies, namely Montology (a-la Jack D. Ives). With this integrative and holistic
view, I attempt to guide the reader through an intellectual expedition that might
conduce either to: (a) the discovery of new narratives; (b) the affirmation of biogeo-
graphical dogmas; or, (c) deter curiosity and ignite activism in favor of the preserva-
tion, conservation, restoration, and regeneration of these jungles in the clouds (or
“nuboselva”) of the tropandean landscapes.
At any rate, this chapter to a book on Andean Cloud Forests triggers geographic
and ecological inquiry of the ecotones that is fully developed later in individual
chapters. In this quest of Cloud Forests borders (a-la Daniel W. Gade), we shall
become explorers of the mountain lines in the landscape (Fig. 2.1), where the unex-
pected turn, the slippery rock, the thorny pole, the mossy carpet, the hanging gar-
dens, the misty and shrouded canopy, the waterlogged ponds, the myriad of sounds,
the ubiquitous rainbows, the exquisite flavors, the poisonous stings, and the psyche-
delic elixirs will all conform a collection of discursive mysteries that ultimately
inform and elucidate what hidden dynamics can be seen at the Andean Treeline
Ecotone Region (ATER). The so-called alpine treelines (Malanson et al. 2011), fol-
lowing European traditions of seeing them as in the Alps (a-la Alexander
V. Humboldt), is a line that separates forests from grasslands, a line that divides
cleavages and hills, a line that draws agriculture and livestock, a line that mimics
firewalls and mowers, a line, that is neither theoretical nor practical, but imagined…
A line that is al-Barzakh! So, seat back, relax, grab your poncho, and rubber boots,
clear your sunglasses, refill the insect repellent and sunscreen… and be prepared to
be wowed!
2.2 Why We Call them Cloud Forests? Onomastics at Play
As the only Commandment of Geography dictates, “Know Thy Word” will prompt us
to start the exploration of place naming with instances where scientists have used
misnomers. By using etymology and onomastics, the real meaning of cloud forests
words will be unveiled to better grasp the complexities of the adaptive system of the
verdant (or “Yunga”). During the 1980s, late professor Larry Hamilton led an effort to
coalesce a global network of mountain scholars interested in TMCFs, spearheaded by
the institutional concerns of the East-West Center in Hawai’i, the US Forest Service
and the Luquillo Forest scientists led by Fred Scatena in Puerto Rico, and James Juvik
of the University of Hawai’i, Hilo, as well as catalyst and foundational research on
forest hydrology by L.A. Bruijnzeel and collaborators in European and North
2 Dynamics of Andean Treeline Ecotones: Between Cloud Forest and Páramo… 27
LINEAR DEFINITIONS
TREE (EDGES, NO ECOTONES) LINES
SNOW LINE NIVAL
TREELINE Glacial
CLOUD BASE LINE

FOREST LINE
Andino
ALPINE
BORDER LINE
TIMBER LINE Ceja
RELIEF LINE
CONTOUR LINE Montano
SKYLINE MONTANE
DIVIDE LINE Loma
Colina
COLLINE
Pie de monte
ALTITUDINAL BELTS
Nieve perpetua Snowcap
Páramo andino Highland Andean grassland
Ceja de montaña Elfin forest
Bosque altoandino Highland Andean forest
Nuboselva alta Upper Montane
Bosque nublado Nuboselva baja Lower Montane Andean cloud forest
Ceja de Selva (pie de monte) Andean piedmont
Bosque húmedo Selva de bajío Tropical moist forest Tropical rain forest
Tropical Selva de llanura Tropical lowland forest
Bosque seco tropical Terra firme forest
Matorral desértico Varzea forest. Hilea
Manglar Igapó forest
TROPICAL EASTERN
WESTERN
SLOPE MOUNTAINS SLOPE
Fig. 2.1 A display of different lines found in the tropical mountainscape

28 F. O. Sarmiento
American Universities (Hamilton et al. 1993). The conference proceedings, published

first as a book from the East/West Center and then as special issue of Ecological
Studies, prompted a frenzy of articles published in ensuing years. The introductory
chapter of the Ecological Studies volume on the subject left a clear impression that it
was the subject matter for botanists, forest ecologists, and hydrologists (Hamilton
et al. 1994). There was a dearth of contributions from the social sciences, particularly
political ecology, human geography or environmental anthropology.
That failure was partially corrected when another international conference was
held at the New York Botanical Garden, mostly energized by E. Forero, S. Churchill,
H. Balslev, and J. Luteyn, botanists wishing to highlight the use of the extensive yet
old botanical collections (Churchill et al. 1995). This international conference in the
Bronx’s Gardens highlighted floristic and cladistic emphasis but it was also deco-
rated with a few ethnobotanical works and some socioeconomic analyses of the
useful cloud forest plants and the concern for the human impact on those ecosystems
(Sarmiento 1995a). It was not until a global meeting was held again, this time in the
big island of Hawai’i, (Bruijnzeel et al. 2011) that the human dimension and the
need of management was at the core of concerned cloud forest scientists, pushing the
need of integration and multidisciplinary research (Martin and Bellingham 2016).
Perhaps ignited by the imminent deforestation frontier of these orobiomes in the
developing world, the plight of shrouded mountainscapes became the target of inter-
national programs, not only of biodiversity conservation and water capture, but also
on other ecosystem services, hunger prevention, poverty alleviation, and the nega-
tive prospects of climate change (Foster 2001). There were programs on Cloud
Forests at the IUCN, at the WCMC, at the FAO’s Mountain Partnership, ICIMOD,
GMBA, GLORIA, MRI, and other multinational initiatives for pushing the scien-
tific needs of TMCFs to the next level. Furthermore, diverse NGOs were created
with specific targets of working with cloud forests issues (Bubb et al. 2004).
However, prolific writers on the social sciences have delved notions of manufac-
tured landscapes, landscape archeology, hybrid socio-ecological systems, coupled
nature-society relations at the forest edges with alpine treelines, and published their
research in a number of professional journals. Recently, a renewed effort to bring
traditional ecological knowledge (TEK) and ethnoecology makes the study of
mountains and of cloud forests, a vivid front of scholarship (Taylor 2010; Sarmiento
and Hitchner 2017). A special call for action is formulated (Frascaroli and Fjeldted
2018) and special panels within international conferences on Mountains (e.g. Perth
III in Scotland, Mountains 2018 in Nova Friburgo, Brazil, an International confer-
ence on Past Plant Diversity, Climate Change and Mountain Conservation held in
Cuenca, Ecuador 2019, and the global IMC 2019 in Innsbruck, Austria) are held.
2.2.1 Landscape Phenology and Threatening Treeline Myths
By switching the rhetoric order, we highlight these cloud forests environments

because of their main characteristic: they are “forests in the clouds.” Literally, they
become islands in the sky, where the circadian rhythm of the incoming atmospheric
tides bathing the mountain slopes at least once a day. The higher the cloud base, the
more difficult to identify the treeline which is useful to separate the forest and the
grassland, whether páramo proper or highland pasture. Indeed, so clear is the effect
of the high tide of clouds, that several plants show the adaptations of aerial roots,
even pneumatophora, just like their distant cousins in the mangrove swamps by the
seashore. Horizontal precipitation affects the slope depending on the aspect of the
mountain. In some cases, such as in the Atacama Desert, this phenomenon locally
known as the “kamanchaka” provides the only source of water to support the
“Lomas” vegetation and the xeric fauna therein. At present, with the advent of tech-
nological advances (e.g., netting with hydrophobic fabrics and drip irrigation), cap-
turing the “kamanchaka” fog with “cloud stripping” technique has transformed
those once-dry-valleys into the bread-baskets and fresh-fruit-providers of the global
economies. Nowadays, avocado, cantaloupes, grapes, cranberries, and a myriad of
other foodstuff have switched the economies of former abandoned mining towns in
the desert, because of the trapping of the mist with physical obstacles serving as
sponges to collect irrigation water, just like the leaves and branches of cloud forest
trees do.
This is exemplified in the literature of the drier areas with the myth of Garoé, the
“fountain tree” or “rain tree” (Oreodaphne sp.) of the Canary Islands (Gioda et al.
1995) and the Argan bush (Argania spinosa) with tree-climbing goats of the
Moroccan mountains (Delibes et al. 2017). In the wetter confines of the Tilarán
mountains of Costa Rica, the epiphytic garden (including mosses, lichens, liver-
worts, ferns, even tank-bromelias, and small berries) increase the capacity of the
leaves and branches of the trees to act likewise in cloud stripping (Lyons 2019). It
was precisely at the Pico del Teide in the Canary Islands that Alexander von
Humboldt in 1799 first provided a mountain profile sketch of the treeline location,
something that he later was able to expand to explain the different vegetation zones
along the slopes of Chimburasu in his famous Tableau Physique of 1807, establish-
ing the dogma of altitudinal zonation in mountain geoecology (Appenzeller 2019).
I recall when influenced by the Humboldtian law, mountain scientists including
Lauer, Acosta Solis, Budowsky, and specially Troll, developed a nuanced approach
to study Andean heights under the light of mainly German scholarship that followed
the altitudinal layering school (Appenzeller 2019). I also perceived the mountains
segmented, determined by the climatic envelope of temperature, precipitation, and
evapotranspiration. As a clear example, I include how I represented the highlands in
Santa Cruz Island of the Galápagos with treelines separating the “pampa insular” of
bracken ferns (Pteridium aquilinum aracnoides) of the higher reaches from the
lower scrub of “cacaotillo” (Miconia robinsoniana) and la “selva insular” mainly
of Scalesia pedunculata (Fig. 2.2).
These descriptive views were soon challenged by a cadre of biogeographers
included Hans Ellenberg (1958, 1979) who famously asked ecologists to inquire
deeper whether forest or grasslands in the Andes exist in the upper reaches; Knapp
(1991) who recalled the Andes as dynamic systems adapted through time; Baeid
(1999) who brought the attention of the human driver for Polylepis woodlands; and
Luteyn and Balslev (1992) who compiled evidence to determine the burning of
páramo ecosystems as the influence of community structure and composition, and
30 F. O. Sarmiento
Fig. 2.2 Comparing the illustration of El Teide and Santa Cruz to illustrate the influence of the
Humboldtian law in the concept of treeline
Sarmiento and Frolich (2002) who contested the naturalness of the Andean treeline,
arguing for the importance of agroecology and the human dimension as drivers of
the treeline dynamics in transformed farmscapes.
2.2.2 Atmospheric Tides and Fluvial Rhythms
The diurnal adiabatic warming pushes de warm water vapor towards the higher
reaches, making coalescence of minute condensation nuclei, creating a mist of con-
densation fronts that never finish cooling enough (or “puyu”) as to produce rain (or
“para”). The continual push of trade winds moves ever slowly the big mass of
cloudiness towards the terrain, which makes it ideal for the presence of many bryo-
phytes that decorate every square inch of the contact surface, whether roots, stems,
branches, leaves, even flowers or fruits, could harbor a massive presence of com-
mensals amidst the small droplets of the cold, windswept mist (or “paramuna”).
Sphagnum-rich ponds, bogs, fens, and rocky substrate actually act as a giant sponge
Fig. 2.3 Representation of the rain shadow effect as observed in northwestern Ecuador
sucking up more than twice their weight in water! (Nadkarni and Solano 2002).
Tillandsia-rich branches support the increased weight of the wetted foliar mass and
other epiphytic flora, contributing to the sponge-like function of the Cloud Forest
“belt” (or “yunga”) (Fig. 2.3). However, there is also the nocturnal flow of the kata-
batic winds that bring the load of heavy cold fog towards the valley bottom (or
“tutapuyu”). Throughout the tropandean mountains it is very common to see the
clouds accumulated down below at dawn, covering low laying cities, crops, tree
plantations, and remnant forests, submerged under a “sea of clouds” (or “puyu-
manta”) that disappear as soon the morning warmth returns.
2.2.3 Tide, Gravity and the Will of People
With this “intertidal” movement of fog it is obvious that the water content of both
atmospheric and terrestrial interphases is saturated and remain waterlogged several
days, even during dry spells of lacking rains. The arboreal life becomes stressed by
the assured presence of the clouds, the limitation of nutrients and other factors asso-
ciated with the fast growth, lixiviation and decay of organic matter, such that break-
ing of the enormous weight of the epiphytic garden is one of the main factors of
ecological succession, creating canopy gaps and triggering patch dynamics and
matrix regeneration on the slopes by small clearings and mini-landslides (Myster
and Sarmiento 1998). This common breakage of the branches or complete falling of
twisted trees, also triggers rockslides, mudslides, and other erosional processes. For
instance, the “Waiku” of the central Andes is the phenomenon of destructive forces
of lahars, rockslides, landslides, and slope failures that runs down the river bed cre-
ating havoc downstream. Excess water is not the only culprit, as sporadic or
32 F. O. Sarmiento
episodic it might be, but frequent tremors, even earthquakes and glacial lake out-
burst flows also contribute to this geodiverse mosaic of rugged topography. However,
as it has been demonstrated profusely elsewhere, the zigzagging road construction
(or “kingu”) and the switchback hiking trails (or “kulunku”) with faulty design are
among the most pernicious factors of TMCFs destruction and change (Sarmiento
1995b; Laimer 2017). Landscape architects now are reviving ancient techniques to
build mountain roads without the detrimental effect of destruction, such as the
“Inka” did to cross precipitous cliffs and forested slopes when implementing the
Great Inka road (or Kapak Ñan) (Penney and Oschendorf 2015).
2.3 Lines in the Longer Cyclicity of Mountainscapes
As circadian rhythms mark the daily routine of back and forth cloudiness, the dif-
ferential temperature is as significant as prompting extremes cold and hot measure-
ments. It is common for equatorial TMCFs to exhibit the four seasons in one day.
Throughout the tropics, seasonality is expressed also in the circannual rhythm but
instead of having winter and summer as the opposite thermal drivers, it is the abun-
dant rainfall and its complete lack thereof that pulls phenology. The Dry Season
“época seca” could last several months as well as the Rainy season “época lluviosa”
could be prolonged and severe. World record annual precipitation measurements
have been registered near Puyo pluviometric station towards the Amazon, and also
near Lita pluviometric station towards the Choco, two of the rainiest places, where
people often mentioned that there are only two seasons in the year: rain and deluge
(Sarmiento 1987). In these wettest places, the vigorous growth of the forest at high
elevation makes it difficult to find a treeline, and one has to be inferred only from
proxies or actual evidence of occurrence of “indicator” plants, as for the presence of
silver-leafed Cecropia sp. in the lower limits, or the presence of tree fern Cyathea
sp. in the actual belt, and the presence of Blechnum in the upper limits.
Conversely, in the driest extremes, cloud forest formations on the semi-desertic
watersheds of the Atacama are forming the “Lomas” vegetation, with green cano-
pies that form the oasis following the riverbeds. Treelines of the xerophytic pacific
mountains are definitely constrained by the continuous browsing of llamas and gua-
nacos, and more recently of goats and sheep mowing the slopes. On the other flank,
drier areas are found at extreme northern ends whereby herbivory is also an impor-
tant factor controlling the vegetation, absent of conifers but abundant of Cactacea,
Leguminosae, and Burseraceae.
The curious effect of the yearly fluctuation in temperature but a rather uniform
precipitation (horizontal and otherwise) has been observed as the climatic envelope
that identify TMCFs with circannual rhythms; however, having virtually no change
in photoperiodicity, the abundant 12 h of daily radiant energy throughout the year
(either with sunlight in open blue skies, with filtered UV-ß shrouded mist or with a
heightened incidence of infrared light), triggers a huge diversification in flora and
fauna (Kapelle and Brown 2001) due to the proneness of mutation generated by the
electromagnetic conditions in the shortest wavelength of TMCFs. This mutation-

prone community shows the fastest rate of evolution due to transiliency between
short-lived extinction rates, such as those observed in orchids at the Centinela farm-
stead in Western Ecuador (Dodson and Gentry 2001). The Centinelan extinctions
have been identified as the most severe threat to biodiversity conservation in TMCFs
areas, where a bulge in the distribution of life on terrestrial ecosystems makes them
hot-spots for biodiversity (Myers 2003); hence the meteorological conditions of
cloudiness and photoperiodicity have far-reaching biological consequences in the
mountainous terrain.
Ultimately, we shall not forget the quinquennial, decadal or centennial cyclicity
driven by either planetary forces, such as forming atmospheric rivers’ deltas every
5 years or so; or the El Niño Southern Oscillation (ENSO) every 10 years or so; or
biological forces, such as the population emergence of cycads and other insects
every 20 years or so; the masting of mountain bamboo (Chusquea spp.) every
40 years or so; the algae bloom of cyanobacteria related to freshwater lake toxicity
every century or so; or the massive hemoglobin saturation of some lakes that are
dyed red (e.g., “Yawarkucha”) every 300 years or so due to either algae bloom
(Dunaliella c.f. salina) or the eutrophication by Halobacteria and microphytic
Dinoflagellata. These larger temporal cycles create a mosaic of life in the TMCFs
that is made of differential spatial components and of various time periods, where
the tessellation (or how the tessellar microsites—ecotopes—are arranged) in func-
tional traits (or cellularity of the landscape) determining either the actual or realized
fabric of the Andean treeline ecotone region.
2.4 T
reeline Differentiation as Response of Latitudinal
Variation
It is clear that the location and extent of the Andean treeline ecotone region (ATER)
changes with the latitude, due to three main factors: firstly, the incidence of direct
sunlight in northern or southern exposures after the tropics of Cancer or Capricornio
will determine the presence/absence of a growing season in Spring and Summer and
of a withering season for the Autumn and Winter towards the subtropical or temper-
ate mountains. While trees from the Neartic could be distributed into these TMCFs,
such as Quercus sp. into Costa Rica, the southern ecological equivalents are present
from the Antarctic, such as Podocarpus sp. into Colombia. Secondly, the extent of
the ATER is directly related to altitudinal verdant prompted by the location of the
slope itself in relation to the mass of the mountain edifice. Throughout the southern
Andes, for instance, tree line appears to be just a sharp boundary around 1800 m as
many of the tallest volcanoes do not reach the 2500 m a.s.l. altitude. In fact, it is
possible to walk from the glacier directly to the tall Araucaria forest with no appar-
ent ecotone present at any side of the glacier mass, while in the equatorial moun-
tains the forest composition and structure remain robust but with noticeable
reduction in stature. Finally, the long extent of the Andes cordillera allows for the
34 F. O. Sarmiento
influence of meteorological conditions to affect the manifestation of a continuous

belt of cloud forest, as it happens in the Andean crescent, or as patchily distributed
among the tallest mountain peaks isolated from the longer cordillera. There is much
more ruggedness towards the higher latitudes, which creates atmospheric condi-
tions that could change violently. Microclimatic conditions generated by the pres-
ence of tall cliffs, or isolated ledgers are important consideration in the study of the
distribution of TMCFs and the robustness of a distinctive ecotone in the treeline
region (ATER).
The curious effect of the yearly fluctuation in temperature but a rather uniform
precipitation (horizontal and otherwise) has been observed as the climatic envelope
that identify TMCFs with circannual rhythms; however, having virtually no change
in photoperiodicity, the abundant 12 h of daily radiant energy throughout the year
(either with sunlight in open blue skies, with filtered UV-ß shrouded mist or with a
heightened incidence of infrared light, triggers a huge diversification in flora and
fauna (Kapelle and Brown 2001) due to the proneness of mutation generated by the
electromagnetic conditions in the shortest wavelength of TMCFs. This mutation-
prone community shows the fastest rate of evolution due to “transiliency” between
short-lived extinction rates, such as those observed in orchids at the Centinela farm-
stead in Western Ecuador (Dodson and Gentry 2001). The Centinelan extinctions
have been identified as the most severe threat to biodiversity conservation in TMCFs
areas, where a bulge in the distribution of life on terrestrial ecosystems makes them
hot-spots for biodiversity (Myers 2003); hence the meteorological conditions of
cloudiness and photoperiodicity have far-reaching biological consequences in the
mountainous terrain. Ultimately, we shall not forget the quinquennial, decadal or
centennial cyclicity driven by either planetary forces, such as forming atmospheric
rivers’ deltas every 5 years or so; or the El Niño Southern Oscillation (ENSO) every
10 years or so; or biological forces, such as the population emergence of cycads and
other insects every 20 years or so; the masting of mountain bamboo (Chusquea
spp.) every 40 years or so; the algae bloom of cyanobacteria related to freshwater
lake toxicity every century or so; or the massive hemoglobin saturation of some
lakes that are dyed red (e.g., “Yawarkucha”) every 300 years or so due to either
algae bloom (Dunaliella c.f. salina) or the eutrophication by Halobacteria and
microphytic Dinoflagellata. These larger temporal cycles create a mosaic of life in
the TMCFs that is made of differential spatial components and of various time peri-
ods, where the tessellation (or how the tessellar microsites—ecotopes—are
arranged) in functional traits (or cellularity of the landscape) determining either the
actual or realized fabric of the treeline ecotone region.
2.5 Mountain Effects and the Andean Treeline Ecotone Region
Whether in América, Africa or Asia, the timing of the intertropical convection zone
(ITCZ) drives the permanence of the cloud forest belt, yet the boundary of forest/
pasture received the name from Europe: Alpine treeline. In this chapter, I will
follow Acosta-Solís’s advice to emphasize the use of “Andean” treeline instead of

“Alpine” treeline for the importance of highlighting unorthodox models followed in
the Neotropics that contest the Humboldtian paradigm of altitudinal belts (Acosta-
Solis 1976; Sarmiento 2000, 2002; Varela 2008). This is reflected in the orientation
of the forested slopes, as the rainshadow effect is critical in the presence/absence of
the biota exposed to horizontal precipitation along the longitudinal extend of the
Andes. An obvious case is that people live on high mountains in the tropical Andes
and they do not in the Alps. Similarly, depending on where in the tropics the TMCF
exists, you will find significant differences in the location of the cloud base, and
thus, the treeline ecotone in the Central American or Caribbean páramos, the
Mexican or Guatemalan zacatonales, the Venezuelan, Colombian and northern
Ecuadorian páramos, the southern Ecuadorian and northern Peruvian jalca, the
southern Peruvian, Bolivian, northern Chilean and Argentinian punas, and southern
Chilean and Argentinian meadows.
Slope and aspect determine the location for Alpine treeline in north/south dif-
ferential exposure, while in the Andean treeline is west/east exposure that deter-
mines the ATER location. For instance, when exposed to the rugged topography,
plants of the exposed slopes can experience strong lateral winds, in the Venturi
effect, that prompt canyon lands and cliff faces, including ledges, to have flagged
trees with obvious morphology impacted by lateral winds along the brooks, whether
coves “rinconadas” or meadow-like “esteros”; however, if the strong winds are adia-
batic and move vertically from the bottom of the hill towards the summit, the
Bernoulli effect takes places, pushing the canopies closer to the ground, and the
wind speed exerts this continual forcing towards the summits to “comb” the leaves
and twist the branches, typical of the elfin forests. Depending on the force of the
wind due to its velocity, some extreme cases are observed, generating the Krumholtz
of short stature and ground hugging stems or “chaupicaspi” typical of the mountain
bamboo (Chusquea sp.), the mountain palms (Geonoma sp.), and the broadleaf
Gumnera sp. Further up, plants adopt the strategy of clumping to grow as rosette
(e.g., “Frailejón,” “almohadilla,” or “romerillo”). In both cases, due to intense wind
flow, desiccation of the foliar tissue is prevented by developing coriaceous leaves
and spines, so that the system is faked to function as in desert physiology, despite
being waterlogged.
Known also as the telescopic effect, the “massenerhebung” effect correlates the
size of an island and the distance from continental areas to the lower/higher altitu-
dinal belt of TMCFs. It is not surprising that we find insular jungles of TMCFs on
isolated islands in the Galápagos archipelago at 500 m a.s.l., and the equivalent
formation at 2500 m a.s.l. in the continental cordillera more than 600 nautical miles
away. Whether be in the Western flank (c.f., Cisandean domain) or in the Eastern
flank (c.f., Transandean domain) both effects determine the harsh xerophytic semi-
desertic slope on the leeward, while the abundant forest cover brings the hydro-
phytic hilea on the windward. Countless records of this differential humidity due to
the Phön effect in the “rainshadow” and the mass/energy effect created the differ-
ence between the enigmatic layering of mountains (Rahbek et al. 2019) and the
altitudinal differentiation of the ATER on outer slopes of tropandean landscapes,
36 F. O. Sarmiento
hanging valleys, and “ceja de selva” (or “yungas”) being saturated of water and
plant life, while the inner slopes toward the drier Interandean valleys, harbor “bol-
sones,” “hoyas,” “vegas,” and “pampas” (c.f., Interandean domain) where little
“natural” vegetation could be used to draw a treeline.
Other effects are also found in TCMFs, including the little studied Cascading
Effect. I am not referring to the actual physical phenomenon of downslope move-
ment, sometimes catastrophic, related to rocksiles, mudslides, landslides or other
forms of mass movement. The trophic cascades generated by introduction of some
animal species have to be better understood. For instance, prior to the declaration of
the Chimborazo Faunal Reserve in 1976, no vicuña existed in the area. After a suc-
cessful pilot project and then a concerted effort to promote valuable wool trade with
the communities nearby, the image of the “natural” paramo with the iconic camelids
is now used to promote ecotourism. An important cascade effect detected with the
introduction of the black fly and the ubiquitous effect of voracious grazers such as
sheep, swain, cattle. Free roaming cattle of the past decades have allowed the pres-
ence of feral bulls that are traditionally sough after the entertainment of the villages
festivities with bullfighting.
2.6 C
onclusion: Andean Treeline Ecotones: Neither Trees
nor Lines
Much research has taken place to explain why the plant formations in the upper
basins of mountains show the physiognomy of elfin forests, chaparral, páramos or
balds (Allen et al. 2004; Young and Leon 2006; Malanson et al. 2007; Körner 2012;
Mathisen et al. 2014; Wang et al. 2016; Malanson and Resler 2016; Möhl et al. 2019
and elsewhere). Most available references relate to the study of physical parameters
and positivistic explanations. In the last decade, thou, a renewed effort of under-
standing gradients of many types, altitudinal and otherwise, from alternative points
of view; social sciences, humanities, and even the arts are increasingly available in
both scholarly and gray literatures (Naess 1989; Chepstow-Lusty et al. 1996;
Bowman et al. 2002; Sarmiento and Frolich 2002; Varela 2008; Sarmiento 2012;
Sarmiento et al. 2017; Huisman et al. 2019 and elsewhere).
There are two emergent themes that coincide with both fronts. Firstly, that there
is no such a thing as a “line” in the treeline; the boundary mosaic tends to be an
ecotone (Myster 2012) that shows different structures within, such as the distur-
bance edge, the saum, the drip shadow, the mantel, and the veil, that it can hardly be
represented with a line. These are mountains with Andean treeline ecotone region
(ATER) extending several 100 m down the of the Andean flank, exhibiting a marked
ecocline in functional ecology, while others tend to be localized around cliff faces,
promontories or ledges (Resler 2006; Elliott and Kipfmueller 2010; Sylvester et al.
2014). Secondly, that there is no doubt that the human impact has caused the loca-
tion of the forest/pasture edge; mostly as a response to fire management and grazing
intensity, the ensuing boundary mosaic in the ATER resembles the intended purpose
of agriculture or livestock rearing instead of a natural formation following climatic
envelopes (Miehe and Miehe 2000; Gehrig-Fasel et al. 2007; White 2013).
As per the structure, composition and changes of the TMCF ecosystems, compo-
sition and change, the notion of Andean treeline is often problematic, as most spe-
cies in the highlands are short, woody vegetation that could fall under the typology
of forest, traditionally applied to the areas where significant timber biomass and tall
poles could be readily harvested. There are notable exceptions, such as the Andean
pine (Podocarpus—cf Retrophyllum or Decussocarpus—rospigliossii) that shows
its tall stature and yellowish cylindric shaft well into the upper reaches of most pri-
mary watersheds, due to its phenotypic plasticity that allows it to survive both in the
lowland Amazon and the Andean highlands. However, most of ATER species exhibit
asymmetrical architecture to achieve anchorage (Chlatante et al. 2003); they are not
tall and their stems are not vertical, but twisted and turned, following the contour
and sustained by buttresses and aerial roots. Notwithstanding physiological adapta-
tions, most vegetation of the ATER is also affected by both topographic and meteo-
rological stresses, showing pubescent, tomentose or spinose stems that add to the
potential of capturing horizontal precipitation. Even more impressive, the display of
prostrate architecture with decumbent and procumbent stems, climbers, and lianas
provide a physical tension that affects overall growth and offers the short stature,
twisted appearance of “krummholz” or elfin forests in the areas or chaparral, or in
generally dwarf physiognomy of the ATER, as well creating a “false ground” that
often makes it difficult to walk straight, as you are stepping on a fabric of superficial
structures that not only capture water but also create appropriate soil biome in
ephemeral knoll, hummock, dune, or tumulus on the slopes. For instance, these elfin
forests and hillocks are territory for one of the largest earthworms (or “kuika”)
(Thamnodrilus sp.).
2.7 Discussion Remarks: Treelines of the Mindscape
The Andean treeline is a good example of the P.T. Barnum effect: most people think
they know what it is, but only few can grasp the actual meaning of the word of this
reified, situated, and partialized views of the ATER. The areas where Andean
treeline ecotone regions are located have been described as socio-ecological, pro-
duction landscapes (SEPLS). I have argued that these areas fit very well in the
Satoyama Initiative, as they exemplify the hybridity of the natural and cultural fac-
tors in “manufactured landscapes” (Fig. 2.4). A short grapple with the human
dimension follows to emphasize the need of Montology to really capture the essence
of place of the Andean treeline (Fig. 2.5).
Environmental cognition based on Western science afford us the possibility to
characterize the landscape with conventional tools of vegetation science, taxonomy,
and allied disciplines. However, there is a large chapter that is present in every zone
of the TMCFs that relates to the intangible properties of the mountainscape. It is not
38 F. O. Sarmiento
Fig. 2.4 Representation of the socio-ecological production landscape of the Imbabura volcano, as
exemplifying the nature/culture hybrid of the Andean treeline, hosting ancestral practices of the
Utawallu runakuna and the modernity of mestizos of Ibarra
Fig. 2.5 The tendency of Montology as a transdisciplinary mountain science, shifting paradigms
of the past to construct a new vision of tropandean mountains whereby the Andean treeline is con-
tested, appropriated, and adapted to several environmental cognition pathways
only the higher elevation, the low temperature, and the excruciating humidity of the
climatic envelope of cloud forests, but the appearance of the shrouded slopes, full of
epiphytic growth, and udic overtones that portrait a mysterious, even dangerous
perception of the ombrophilous environment of TMCFs. Phosgene, along with
methane emanating from highly putrid fens due to organic decomposition in the
forest floor, cause autoignition’s subtle flares that prompt people to think they are
seeing ghosts or mysterious lights within the canopy. Sounds coming from the
undergrowth and hummock add to the magic envelope to which residents have asso-
ciated metaphysical explanations.
These qualitative observations make these sites prone to superstitious and magic
as the wherewithal to survival in these extreme conditions. The myths of the Cloud
Forests are yet told anew, although for centuries people of the region have known
about them. For instance, in areas of the Pululahua Geobotanical Reserve in
Ecuador, near the town of Nieblí (apropos name for a town that submerges in cloud-
iness every day at 2:30 pm without fault) the existence of the duende who owns the
seepage basin of el infiernillo that could take the unsuspecting visitor and kidnap
her/him towards mysterious waterfalls and caves. There are myths that associate the
presence of wild people (“sacharuna”) that walk the boundary of the forest in the
ATER, but most likely they refer to the Andean spectacled bear (Tremarctos orna-
tus) (or “Ukumari”), the black Andean tapir (Tapirus pinchaque) (or “wagra
pinchaki”), or the Andean small deer (Masama rufina) (or “sochi”) that tend to be
vagrant along the crepuscular treeline.
Also, in the TMCFs of the western flank, the aluvillo tree (Toxicodendron stria-
tum) is revered for the properties of generating rashes and violent dermatitis if the
distracted pedestrian crosses under its canopy without waving the hat to salute:
“Buenos días, compadre aluvillo”; it seems that taking the hat off the head, moves
the colloidal aerosol underneath the branch of this Anacardiaceae so that the person
could walk without facing this invisible barrier. In other cases, original people who
inhabit these environs, seek the help of curanderos (or “yachay”) so they could
clean off the negative energy loaded when walking those mountain trails. All resi-
dents relate to the elements of the cloud forests with reverence, given animistic
interpretation to thin air (mal aire), dense clouds (espanto), heavy downpours (or
“washishka”), small ponds and lakes, and, of course, ice, rain, and waterfalls
(Sarmiento 2016). This is still experienced by the Kogi people in Colombia, inhabit-
ing the Sierra Nevada de Santa Marta National Park who are the custodians of the
Lost City or “Ciudad Perdida.” Just like the Pemon nation in Auyantepuy massif of
Venezuela, who held sacred the tallest waterfall, Kerepakupai Meru (formerly
Angel’s fall).
Moreover, in Ecuador, the Shwar people define themselves as the people of the
waterfall and hold therein initiation rites that include hallucinogenic chants and
astral travels. Further south in Peru, along the Ceja de Selva near Chachapoyas, the
Catarata de Gocta (the third tallest waterfall on Earth) is just one of the many flow-
ing from the Andean heights towards the Amazon, in what is now considered the
“route of the waterfalls” as a tourist destination, adding to the impressive Kuelap
40 F. O. Sarmiento
citadel (or “llakta”) ruins and the cliff hanging sarcophagi of the Chachapoya mum-
mies or pinchudos.
Myth and superstition are still defining factors in the understanding of these
little-known places. Underneath the closed canopy of the cloud forest is still possi-
ble to find ruins of monumental architecture that remains hidden to the current afflu-
ence of tourist who fledge to see the many hummingbirds or myriad exotic flowers
of this biodiversity hotspot. When tourists see Machu Picchu now, only few of them
realize that this marvel of the world is really located amidst the most magnificent
wonder of nature, the tropical montane cloud forest landscapes, that is likely to have
been highly modified in ancestral time, abandoned for centuries, and now “redis-
covered” by gold miners, timber companies, cattle ranchers, hungry villagers, and
curious ecotourists. Where will they draw the lines in the mountains next? How
many more citadels (or “Llakta”) we must find to accept that the ATER are hybrid
cultural landscapes?
References
Allen TR, Walsh SJ, Cairns DM, Messina JP, Butler DR, Malanson GP (2004) Geostatistics and
spatial analysis: characterizing form and pattern at the alpine treeline. In: Geographic informa-
tion science and mountain geomorphology. Springer, Berlin, pp 189–218
Appenzeller T (2019) Global warming has made iconic Andean peak unrecognizable. Science
365(6458):1094–1097
Acosta-Solis M (1976) Vocabulario basico de fitoecologia: rectificacion terminologica de la obra
Holdridge. Instituto Ecuatoriano de Ciencias Naturales. Quito, EC.
Baeid C (1999) Polylepis spp. en los Andes centrales: un análisis preliminar sobre cambios
climáticos y el impacto de la actividad humana en su distribución. In: Desarrollo Sustentable de
Montañas: Entendiendo las Interfaces Ecológicas para la Gestión de Paisajes Culturales en los
Andes. Sarmiento F, Hidalgo J (eds) Memorias del III Simposio Internacional de la Asociación
de Montañas Andinas, Quito
Bowman WD, Cairns DM, Baron JS, Seastedt TR (2002) Islands in the sky: alpine and treeline
ecosystems of the Rockies. Rocky Mountain futures: an ecological perspective. Island Press,
Washington, DC, pp 183–202
Bruijnzeel LA, Scatena FN, Hamilton LS (2011) Tropical montane cloud forests: science for con-
servation and management. Cambridge University Press, Cambridge
Bubb PI, May I, Miles L (2004) Cloud forest agenda. UNEP-WCMC biodiversity series 20.
UNEP-WCMC, Cambridge
Chepstow-Lusty AJ, Bennett KD, Switsur VR, Kendall A (1996) 4000 years of human impact and
vegetation change in the central Peruvian Andes—with events parallelling the Maya record?
Antiquity 70(270):824–833
Churchill S, Balslev H, Forero E, Luteyn J (1995) Biodiversity and conservation of Neotropical
montane forests. The New York Botanical Garden, Bronx. 702pp
Chlatante D, Scippa G, DiLorio A, Sarnataro M (2003) The influence of steep slopes on root sys-
tem development. J Plant Growth Regul 4:247–260
Delibes M, Castañeda I, Fedriani JM (2017) Tree-climbing goats disperse seeds during rumina-
tion. Front Ecol Environ 15(4):222–223
Dodson CH, Gentry AH (1991) Biological extinction in western ecuador. Annals of the Missouri
Botanical Garden 78(2):273
Ellenberg H (1979) Man’s influence on tropical mountain ecosystems in South America. J Ecol
67:401–416
Ellenberg H (1958) Wald oder Steppe? Die natürliche Pflanzendecke der Anden Perus. Umschau
1958:645–681
Elliott GP, Kipfmueller KF (2010) Multi-scale influences of slope aspect and spatial pattern on
ecotonal dynamics at upper treeline in the southern Rocky Mountains, USA. Arct Antarct Alp
Res 42(1):45–56
Foster P (2001) The potential negative impacts of global climate change on tropical montane cloud
forests. Earth Sci Rev 55(1–2):73–106
Frascaroli F, Fjeldted T (2018) From abstractions to actions: re-embodying the religion and con-
servation nexus. J Stud Relig Nat Cult 11(4):511–534
Gehrig-Fasel J, Guisan A, Zimmermann NE (2007) Tree line shifts in the Swiss Alps: climate
change or land abandonment? J Veg Sci 18(4):571–582
Gioda A, Hernández Z, Gonzáles E, Espejo R (1995) Fountain trees in the Canary Islands: legend
and reality. Adv Hortic Sci 9(3):112–118
Hamilton L, Juvik J, Scatena F (1993) Tropical montane cloud forests. The East West Center,
Honolulu. 284pp
Hamilton L, Juvik J, Scatena F (1994) The Puerto Rico tropical montane cloud forests sym-
posium: introduction and workshop synthesis. In: Ecological studies, vol 110. Springer,
New York, pp 1–23
Huisman SN, Bush MB, McMichael CN (2019) Four centuries of vegetation change in the mid-
elevation Andean forests of Ecuador. Veg Hist Archaeobotany 1:1–11
Knapp G (1991) Andean ecology: adaptive dynamics in Ecuador. Westview Press, Boulder
Kapelle M, Brown A (2001) Bosques Nublados del Neotrópico. National Institute of Biodiversity
(InBIO), San José. 698pp
Körner C (2012) Alpine treelines: functional ecology of the global high elevation tree limits.
Springer Science & Business Media, Basel
Laimer HJ (2017) Anthropogenically induced landslides—a challenge for railway infrastructure in
mountainous regions. Eng Geol 222(1):92–101
Luteyn JL, Balslev H (1992) Páramo: an Andean ecosystem under human influence. Academic
Press, London
Lyons W (2019) Cloud forests of Costa Rica: ecosystems in Peril. Weatherwise 72(3):32–37
Malanson GP, Butler DR, Fagre DB, Walsh SJ, Tomback DF, Daniels LD, Resler LM, Smith WK,
Weiss DJ, Peterson DL, Bunn AG (2007) Alpine treeline of western North America: linking
organism-to-landscape dynamics. Phys Geogr 28(5):378–396
Malanson GP, Resler LM, Bader MY, Holtmeier FK, Butler DR, Weiss DJ, Daniels LD, Fagre
DB (2011) Mountain treelines: a roadmap for research orientation. Arct Antarct Alp Res
43(2):167–177
Malanson GP, Resler LM (2016) A size-gradient hypothesis for alpine treeline ecotones. J Mt Sci
13(7):1154–1116
Martin PH, Bellingham PJ (2016) Towards integrated ecological research in tropical montane
cloud forests. J Trop Ecol 32(5):345–354
Mathisen IE, Mikheeva A, Tutubalina OV, Aune S, Hofgaard A (2014) Fifty years of tree line
change in the Khibiny Mountains, Russia: advantages of combined remote sensing and dendro-
ecological approaches. Appl Veg Sci 17(1):6–16
Miehe G, Miehe S (2000) Comparative high mountain research on the treeline ecotone under
human impact: carl troll’s “asymmetrical zonation of the humid vegetation types of the world”
of 1948 reconsidered. Erdkunde, pp 34–50
Möhl P, Mörsdorf MA, Dawes MA, Hagedorn F, Bebi P, Viglietti D, Freppaz M, Wipf S, Körner
C, Thomas FM, Rixen C (2019) Twelve years of low nutrient input stimulates growth of trees
and dwarf shrubs in the treeline ecotone. J Ecol 107(2):768–780
Myers N (2003) Biodiversity hotspots revisited. Bioscience 53(10):916–917
Myster RW (2012) Ecotones between forest and grasslands. Springer-Verlag, Berlin
42 F. O. Sarmiento
Myster RW, Sarmiento FO (1998) Seed inputs to microsite patch recovery on two Tropandean
landslides in Ecuador. Restor Ecol 6(1):1–10
Nadkarni NM, Solano R (2002) Potential effects of climate change on canopy communities in a
tropical cloud forest: an experimental approach. Oecologia 131(4):580–586
Naess A (1989) Metaphysics of the treeline. Trumpeter 6(2):45
Penney D, Oschendorf J (2015) The Great Inka Road: Engineering an Empire. Smithsonian
Institution.
Rahbek C, Borregaard M, Coldwell R, Dalsgaard B, Holt B, Morueta-Holme N, Nogues-Bravo D,
Whottaker R, Fjeldsa J (2019) Humboldt’s enigma: what causes global patterns of mountain
biodiversity? Science 365:1108–1113. (special issue on Mountain Life)
Resler LM (2006) Geomorphic controls of spatial pattern and process at alpine treeline. Prof
Geogr 58(2):124–138
Sarmiento FO (2016) Neotropical mountains beyond water supply: environmental services as a
trifecta of sustainable mountain development. In: Greenwood G, Shroder J (eds) Mountain ice
and water: investigations of the hydrologic cycle in alpine environments. Elsevier, New York,
pp 309–324
Sarmiento FO (2012) Contesting Páramo: critical biogeography of the northern Andean Highlands.
Kona Publishing. Higher Education Division, Charlotte, NC. 150pp
Sarmiento FO (2002) Anthropogenic change in the landscapes of highland Ecuador. Geogr Rev
92(2):213–234
Sarmiento FO (2000) Breaking mountain paradigms: ecological effects on human impacts in man-
aged tropandean landscapes. AMBIO J Hum Environ 29(7):423–432
Sarmiento FO (1995a) Restoration of equatorial Andes: the challenge for conservation of trop-
Andean landscapes. In: Churchill SH, Balslev H, Forero E, Luteyn J (eds) Biodiversity
and conservation of Neotropical montane forests. The New York Botanical Garden, Bronx,
pp 637–651. 702pp
Sarmiento FO (1995b) Naming and knowing an Ecuadorian landscape: A case study of the
Maquipucuna Reserve. The George Wright Forum 12(1):15–22
Sarmiento FO (1987) Antología Ecológica del Ecuador: Desde la Selva hasta el Mar. Editorial
Casa de la Cultura Ecuatoriana. Museo Ecuatoriano de Ciencias Naturales, Quito
Sarmiento FO, Hitchner S (2017) Indigeneity and the sacred: indigenous revival and the conserva-
tion of sacred natural sites in the Americas. Berghahn Books, New York. 266pp
Sarmiento FO, Ibarra JT, Barreau A, Pizarro JC, Rozzi R, González JA, Frolich LM (2017) Applied
montology using critical biogeography in the Andes. Ann Am Assoc Geogr 107(2):416–428
Sarmiento FO, Frolich LM (2002) Andean cloud forest tree lines. Mt Res Dev 22(3):278–288
Sylvester SP, Sylvester MD, Kessler M (2014) Inaccessible ledges as refuges for the natural veg-
etation of the high Andes. J Veg Sci 25(5):1225–1234
Taylor BR (2010) Dark green religion: nature spirituality and the planetary future. University of
California Press, Berkeley
Varela L (2008) La alta montaña del norte de los Andes: El páramo, un ecosistema antropogénico.
Pirineos 163:85–95
Wang Y, Zhu H, Liang E, Camarero JJ (2016) Impact of plot shape and size on the evaluation of
treeline dynamics in the Tibetan Plateau. Trees 30(4):1045–1056
White S (2013) Grass páramo as hunter-gatherer landscape. The Holocene 23(6):898–915
Young KR, Leon B (2006) Tree-line changes along the Andes: implications of spatial patterns and
dynamics. Philos Trans R Soc B 362(1478):263–272
Chapter 3
Análisis Regional En Ecosistemas De
Montaña En Colombia:Una mirada desde
la funcionalidad del paisaje y los servicios
ecosistémicos
Paola Isaacs-Cubides, Julián Díaz, and Tobias Leyva-Pinto
3.1 Introduction
Colombia tiene una superficie de 1.141.000 ha, y es altamente heterogénea en tér-

minos geográficos con cinco regiones biogeográficas que cubren una amplia gama
de elevaciones (0–5800 m), tiene una gran variedad de ecosistemas, como páramos,
bosque andino, bosques húmedos, bosques secos, manglares, etc.), moderada pre-
cipitación anual (300–1000 mm) y diversas características geológicas (Etter et al.
2006). Esta variabilidad ambiental en relación con el tamaño geográfico de Colombia
ha resultado en altas tasas de endemismo y notable riqueza de especies (51.330
especies), convirtiendo a Colombia en el segundo país con mayor diversidad (prim-
ero en aves y orquídeas, segundo en plantas, anfibios, mariposas y peces de agua
dulce, tercero en palmas y reptiles y cuarto en mamíferos; Etter et al. 2011; SIB
Colombia 2019).
Esta variedad se debe en gran parte a la presencia de un sistema de cordilleras,
que facilita la creación de condiciones microclimáticas que propiciaron la elevada
diversificación. En estos sistemas de montaña, en buena parte se presentan bosque
nublados, pero que sobretodo, albergan ecosistemas de alta montaña y páramo úni-
cos en el mundo.
En Colombia, al menos el 40% del territorio continental está degradado (Etter
et al. 2008), con una tasa de deforestación cercana a las 220.000 ha/año (IDEAM
2014). Además, la calidad y cantidad de los servicios ecosistémicos se han visto
gravemente afectadas, al igual que el capital social y las relaciones entre comuni-
dades y ecosistemas naturales (Murcia and Guariguata 2014). La continua
P. Isaacs-Cubides (*) · J. Díaz

Instituto de Investigación de Recursos Biológicos Alexander von Humboldt,
Bogota, Colombia
T. Leyva-Pinto
Universidad Nacional de Colombia, Bogota, Colombia

https://doi.org/10.1007/978-3-030-57344-7_3
44 P. Isaacs-Cubides et al.
deforestación y degradación de los ecosistemas tiene un impacto negativo en los

suelos y el abastecimiento de agua, afecta a los sistemas agrícolas y otros sistemas
productivos y amenaza los servicios ecosistémicos requeridos por millones de
personas.
Esta degradación es el resultado de un desarrollo económico con base en la
sobreexplotación de los recursos naturales y una industria extractiva, que no con-
sidera la dependencia que tiene el crecimiento económico de la capacidad que tenga
el ambiente natural para tolerar todos los procesos económicos, sociales, tecnológi-
cos y culturales (Aguilar-Garavito and Ramírez 2015).
En Colombia los ecosistemas degradados se localizan principalmente en las
regiones andina y caribe, lo cual coincide con la ubicación que históricamente ha
presentado la mayoría de los asentamientos humanos en el territorio (Etter and
Wyngaarden 2000; MADS 2015; Ramírez et al. 2015).
En el país, alrededor de 6.000.000 ha de bosques presentan fragmentación o
representan vegetación secundaria (IDEAM 2014; Ramírez et al. 2015). Esto es de
gran preocupación porque después de la degradación viene la deforestación. Por
otro lado, los páramos ocupan 2,9 millones de hectáreas (IAvH 2012), de las cuales
un 15.9% presenta pérdida en áreas que no es permitido su uso (Isaacs 2014;
Cadena-Vargas and Sarmiento 2016).
Entender estos patrones de transformación del territorio es clave para la toma de
decisiones, es por ello que se requiere desarrollar diversos tipos de insumos que
permitan evaluar las condiciones de intervención en el territorio.
En la actualidad se cuenta con diversas herramientas de análisis a nivel espacial,
que son necesarias para determinar qué zonas son apropiadas para la conservación
y cuales son prioritarias para iniciar acciones de restauración, las cuales adicional-
mente facilitan la evaluación del comportamiento de las dinámicas de los ecosiste-
mas y sus coberturas en determinada zona.
La selección de dichas áreas implica un análisis comparativo entre las diferentes
variables posibles, mediante la identificación y definición de criterios apropiados
para resolver y abordar el objetivo del estudio de acuerdo a las características de
la zona.
Por otro lado, para evaluar cuál es el comportamiento de las coberturas en una
región o paisaje, y determinar las dinámicas de ocurrencia de estos eventos, se
emplean métricas o estadísticos bajo la teoría de patrones espaciales, que incorpo-
ran el paisaje representado como un mosaico de parches o coberturas discretas. Esta
teoría busca explicar la distribución de los objetos geográficos, sus patrones y pro-
cesos a través del tiempo y en otras regiones (Legendre and Legendre 1998).
Evaluar estos elementos espaciales es insumo para realizar una zonificación del
área y determinar las necesidades de restauración, de preservación o de uso, siendo
una buena herramienta para la planificación del territorio (GREUNAL 2010).
Estas métricas se enmarcan en dos categorías generales: La composición se
refiere a los atributos asociados con la variedad y abundancia de tipos de parches sin
considerar el carácter espacial o ubicación de estos. La segunda categoría evalúa la
configuración espacial, la cual se refiere al arreglo, posición, orientación y carácter
espacial (forma, área núcleo) de los parches en el paisaje (McGarigal et al. 2012).
3 Análisis Regional En Ecosistemas De Montaña En Colombia… 45
3.2 Methods
Adicionalmente, es necesario considerar un aspecto desde la funcionalidad del

paisaje, en donde se incluyen necesidades propias de las especies como por el ejem-
plo la conectividad de su hábitat.
Las zonas de montaña tienen como particularidad, que la conectividad la da el
gradiente de altitud, por lo que en el presente capítulo, evaluamos las condiciones
de las zonas de montaña en términos de composición de bosques nublados monta-
nos, andinos y altoandinos y su transición hacia el páramo, como un conjunto insep-
arable de condiciones para las especies (Fig. 3.1).
Se realizó un análisis de las condiciones de estos ecosistemas, teniendo en cuenta
una cota mayor a los 1200 m que abarca hasta alturas de 5775 m, en donde se pre-
senta gran cantidad de bosques nublados. Se evaluó la composición de coberturas
usando el mapa Corine Landcover para Colombia más reciente a escala 1:100.000
y se realiza el ejercicio de conocer las diferentes condiciones del paisaje y su estado
de preservación especialmente considerando el tipo de cobertura, el tamaño y la
forma (relación perímetro-área) de los fragmentos de bosque, y el grado de frag-
mentación entre ellos (Mcgarigal et al. 2012). La obtención de la capa del índice de
fragmentación se realizó a partir de la capa de uso de la tierra, en este caso se toman
los sitios pertenecientes a áreas naturales los cuales se intersectan con una malla de
puntos distanciados cada 300 metros. Estos puntos resultantes se procesan mediante
el cálculo de densidad Kernel, a partir del cual se obtiene valores de densidad altos
para zonas naturales y densidad baja para zonas transformadas (Correa-Ayram et al.
2017). Del raster obtenido, se realizó una reclasificación por cuartiles de los valores
para obtener umbrales de 1 a 5, siendo 5 las áreas de mayor fragmentación por dis-
tancia entre parches. Se construyó un índice mediante la suma de los atributos, que
relaciona el tamaño, la forma y el grado de fragmentación (actúa como distancia),
para definir zonas de áreas naturales más degradadas (tamaños pequeños, formas
pequeñas y alta fragmentación).
Adicionalmente, se realizó un ejercicio para identificar áreas con mayor valor de
preservación, a partir de modelos de prestación de servicios ecosistémicos de regu-
lación, disponible en análisis cartográficos.
Se mapearon los servicios de carbono a través del cálculo de biomasa aérea de
acuerdo al tipo de cobertura siguiendo lo propuesto por Yépes y colaboradores
(IDEAM 2013).
Se evaluó la oferta hídrica, a través de la interpolación de datos de estaciones
climáticas para la estimación de la evapotranspiración potencial y la escorrentía
(Anexo 1). Esta capa se cruzó con los tipos de cobertura, el material parental y la
pendiente, para definir zonas donde hay mayor regulación hídrica por el efecto de la
presencia de la vegetación, la permeabilidad y la infiltración.
Por último, se reconocieron aquellas zonas donde aún se mantiene control de
erosión e inundaciones al mantener la vegetación natural, empleando los mapas de
erosión nacional (2015) y susceptibilidad a inundaciones (IDEAM 2013). Del mapa
Fig. 3.1 Zonas de montaña que albergan bosques nublados, a partir de los 1200 m
de servicios, se obtuvo un mapa final acumulado, para definir zonas de importancia

para la preservación.
3.3 Resultados and Discussion
De acuerdo al mapa de coberturas de la tierra, de las 18.368.518 ha evaluadas, el

45% corresponden a pastos y cultivos y el 53% a áreas naturales (Fig. 3.2).
Esto de entrada denota un estado elevado de intervención desde el inicio. También
se presentan 5.900 ha de humedales de montaña, 55.000 ha de cuerpos de agua y
135.000 ha de infraestructura y ciudades (Fig. 3.3). En este rango altitudinal, se
encuentra un total de 900 pueblos y ciudades lo que también evidencia la alta densi-
dad de ocupación humana allí presente.
En cuanto a las áreas naturales presentes en ese 53%, el 65% corresponde a
bosques, el 18% arbustales y el 8% pastizales naturales y vegetación secundaria en
transición de crecimiento (Fig. 3.4).
De acuerdo a la distribución de los tamaños de los parches, solo existe una zona
continua de coberturas naturales de más de 1.700.000 ha, hacia las zonas de
Putumayo, Caquetá y Huila, al sur del país (en verde). También se destacan los
bosques entre Huila y Tolima de más de 240.000 ha, los bosques entre Cauca y Valle
(217.000 ha), al norte de Antioquia y noroccidente del Meta con más de 2 millones
60.00
53.33
50.00 45.28
Percentage
40.00
30.00
20.00
10.00
0.74 0.03 0.30
0.00
Urban Agricultural Natural Wetlands Water bodies
and livestock
Fig. 3.2 Porcentaje de cobertura presente en las zonas de montaña

Fig. 3.3 Distribución de las coberturas presentes en las zonas de montaña
de hectáreas (en amarillo todos). Si bien aún existe una proporción alta de parches
de más de 200.000 ha, se observa una gran cantidad de parches de menos de
10.000 ha, distribuidos en todo el país (Fig. 3.5).
Por su parte la relación perímetro y tamaño (métrica de forma Shape index,
Mcgarigal et al. 2012), presenta una alta proporción de las formas debido al gran
tamaño entre estas, con algunos parches cuyo tamaño evidencia un alto estado de
70.00 65.08
60.00
50.00
40.00
30.00
20.00 17.80
8.25 8.86
10.00
0.00
Forest Shrubs Grasses Secondary
vegetation
Fig. 3.4 Porcentaje de cada cobertura natural presente en las zonas de montaña
remanencia y transformación (Fig. 3.6). Esto se debe a que los parches empiezan a

adquirir formas más cuadradas resultado del uso de la tierra para cultivos y ganadería
y su parcelación (Mcgarigal et al. 2012).
De acuerdo a la distribución de frecuencias de los valores de forma, la mayoría
se encuentra en valores de 2.7 (S.D: 1.47), seguido de valores por debajo de 1.47,
que son los que corresponden a los parches más pequeños (Fig. 3.7).
Ya teniendo en cuenta la composición de las demás coberturas, al momento de
evaluar el tamaño, la relación perímetro-área de las áreas naturales, y su estado de
fragmentación, cómo indicador de distancia entre parches, los bosques en Antioquia,
Santander y Cundinamarca, presentan mayor estado de fragmentación al también
coincidir con las áreas de mayor ocupación humana, históricamente. Las zonas de
la vertiente occidental de la cordillera occidental y al sur de la cordillera central y
oriental, presentan mayor estado de conectividad. Sin embargo, es importante con-
siderar que la cobertura en tierras bajas se ha perdido pero aún se mantiene la may-
oría de la conectividad en la alta montaña, en especial en las zonas de páramo
(Fig. 3.8).
Considerando el análisis desde los servicios ecosistémicos, el carbono es el más
conocido y empleado en análisis de estimaciones de captura y almacenamiento en
biomasa. En este caso, los bosques juegan el papel fundamental como elemento
para cuantificar este servicio, siendo el que mayor aporte realiza. Los bosques son
grandes almacenes de carbono en esta zona, por lo que la destrucción de las cober-
turas y el cambio de uso de la tierra representan un incremento en las emisiones de
carbono como contribuyente de gases de efecto invernadero y su huella de carbono
(Yepes et al. 2011). En este caso de montaña, las zonas altas pueden no ser tan rep-
resentativas para este servicio, ya que los páramos por ejemplo, corresponden a
coberturas de pastizales naturales, cuyo aporte de carbono en biomasa es menor
Fig. 3.5 Distribución por tamaños de las coberturas naturales

Fig. 3.6 Distribución por índice de forma de las coberturas naturales
(Fig. 3.9). Por su parte, las coberturas intervenidas presentan un menor aporte por
su contenido de biomasa, pero sin incluidas en este análisis. Es por ello que emplea-
mos otros insumos para evidenciar oferta de servicios, en especial en las áreas que
aún se mantienen en pie.
Es claro que las zonas de montaña desempeñan un papel fundamental en la oferta
de agua y su disponibilidad. Esto se refleja en los altos valores que pueden llegar a
albergar las zonas evaluadas, con valores altos de casi 12.000 mm. Zonas del noroc-
cidente del Meta, vertiente occidental de la cordillera occidental en Chocó, Cauca y
Fig. 3.7 Distribución de frecuencias de los parches naturales en la zona de montaña
Fig. 3.8 Estado de fragmentación de las áreas naturales. En verde se observan las coberturas con
mayor conectividad y en rojo, las más fragmentadas
Fig. 3.9 Acumulación de carbono en biomasa, de acuerdo a los tipos de cobertura
Nariño, cordillera central en Antioquia, son lugares de elevada importancia por la

acumulación de agua, en especial aquella que viene de las mayores altitudes
(Fig. 3.10).
Una vez incluido el tipo de cobertura, la pendiente y el material parental, es
posible evidenciar las zonas donde hay mayor regulación de esa oferta de agua, al
mantener la vegetación natural aún en pie. En especial las zonas al pie de la montaña
son las que albergan mayor cantidad de servicios (Fig. 3.11).
Fig. 3.10 Oferta hídrica calculada para las zonas de montaña
Es importante destacar, que para las zonas de montaña adicionalmente el servicio

de control de erosión es de gran preponderancia, ya que actualmente la zona se ve
muy afectada por eventos de remoción en masa por el mal manejo del suelo (Servicio
Geológico Colombiano 2013). En estas zonas de montaña, se presenta control de
erosión en 1.688.375 ha, sin embargo existe una alta cantidad de zonas ganaderas,
lo que aumenta aún más la susceptibilidad de erosión (Fig. 3.12).
Fig. 3.11 Mapa de hotspots de servicios de carbono, oferta y regulación hídrica
La modelación de los servicios ecosistémicos permite darle valor monetario, a

un valor casi intangible que poseen las áreas naturales y esto ha podido ser traducido
a cifras. Por nombrar algunos, de acuerdo con Petley (Dinero 2017), cerca de 100
colombianos mueren al año a causa de deslizamientos de tierra, siendo el 50% de
las muertes asociadas a desastres naturales, seguida de las inundaciones (42%) y los
Fig. 3.12 Áreas de control de erosión en coberturas vegetales en la montaña
terremotos (8%). Asimismo, en términos de costos por ejemplo, en el año 2017

ocurrió una gran avalancha en la ciudad de Mocoa, la cual tuvo un costo inicial de
cerca de 13 millones de dólares para sólo atender la emergencia (Dinero 2017), con
afectación a cerca de 10.000 viviendas que aproximadamente tienen un costo de 27
millones de dólares para su reemplazamiento. Las donaciones nacionales llegaron a

cerca de 3.5 millones, mientras que las internacionales ya alcanzaron los 9 millones,
según cifras de la Unidad Nacional de la Gestión del Riesgo (Razón pública 2018).
Según el Plan Nacional de Desarrollo 2010–2014, hubo 269 acueductos y 751
vías afectadas, se estimó que murieron 600.000 aves y 115.000 bovinos, aparte.
del desplazamiento de 1.430.200 animales y la pérdida de 2.601 toneladas de
carne, así como la pérdida de 1.080.000 ha de área cultivable (Sánchez 2011). Por
su parte, la atención de la emergencia invernal fue estimada en $8.5 billones de
dólares a precios en 2010, además de los sobrecostos que se generaron por la oferta
de alimento en las poblaciones que se abastecen de productos de la región (Sánchez
2011). Estos costos pudieron ser evitados o invertidos en una adecuada gestión del
territorio, con miras a la prevención y mitigación de los desastres.
Sin tener en cuenta el valor intrínseco de las áreas naturales, el mal manejo del
suelo y su depreciación en términos de servicios ecosistémicos acarrea costos
humanos y económicos altos. Su rehabilitación podría disminuir estos gastos y gen-
erar economías más rentables con una mejor productividad.
En este sentido la presencia de áreas protegidas es una de las opciones que los
gobiernos adoptan para garantizar la preservación de la biodiversidad, pero también
para garantizar la prestación de servicios. Para las zonas de montaña se han real-
izado grandes esfuerzos para aumentar la declaratoria de áreas de preservación
(Fig. 3.13). En su mayoría estas áreas buscan proteger ecosistemas de páramos, por
su oferta de agua, pero también ha aumentado la representatividad de estas áreas en
tierras más bajas.
6.264.498 ha se encuentran en estado de protección, sin embargo es importante
garantizar su real preservación, teniendo en cuenta condiciones sociales desfavora-
bles en Colombia y poco acceso a la tierra. Es necesario proponer y apoyar tipos de
uso más acordes con la capacidad productiva de los suelos, las cuales mejoren las
condiciones de vida de los habitantes. Muchas de las zonas modeladas y que con-
tienen alta oferta se servicios ecosistémicos, cómo elevada integridad en sus áreas
naturales, aún están por fuera de alguna categoría de protección.
Así mismo, se debe mejorar en el establecimiento de áreas para la conectividad
entre estas áreas protegidas, las cuales en su mayoría están separadas. Esta conec-
tividad entre páramos y la alta montaña, es clave para seguir manteniendo las condi-
ciones naturales. Dichas estrategias pueden incluir sistemas de producción hacía
una transición similar a la de los bosques y sus productos derivados.
Fig. 3.13 Sistema Nacional de Áreas protegidas en zonas de montaña en Colombia
References
Aguilar-Garavito M, Ramírez W (2015) Monitoreo a procesos de restauración ecológica, apli-

cado a ecosistemas terrestres. Instituto de Investigación de Recursos Biológicos Alexander von
Humboldt (IAvH), Bogotá D.C., 250 pp
Camilo A. Correa A, Manuel E. Mendoza, Andrés E, Diego R. Pérez S (2017) Anthropogenic
impact on habitat connectivity: A multidimensional human footprint index evaluated in a
highly biodiverse landscape of Mexico. Ecological Indicators 72:895–909
Cadena-Vargas CE, y Sarmiento CE (2016) Cambios en las coberturas paramunas. En Gómez,
MF, Moreno, LA, Andrade GI, y Rueda C (Eds.). Biodiversidad 2015. Estado y tendencias de
la biodiversidad continental de Colombia. Instituto Alexander von Humboldt. Bogotá, D.C.,

Colombia
Dinero (2017) https://www.dinero.com/empresas/confidencias-on-line/articulo/tragedia-de-
mocoa-tiene-un-costo-no-calculado/243678
Etter A, van Wyngaarden W (2000) Patterns of landscape transformation in Colombia, with empha-
sis in the Andean region. Ambio 29:432–439. https://doi.org/10.1579/0044-7447-29.7.432
Etter A, McAlpine C, Wilson K, Phinn S & Possingham H (2006) Regional patterns of agricultural
land use and deforestation in Colombia. Agricultural Ecosystems Environment 114, 369–386.
https://doi.org/10.1016/j.agee.2005.11.013
Etter A, McAlpine C, Possingham H (2008) Historical patterns and drivers of landscape change in
Colombia since 1500: a regionalized spatial approach. Ann Assoc Am Geogr 98:2–23. https://
doi.org/10.1080/00045600701733911
Etter A, McAlpine C, Seabrook L, Wilson K (2011) Incorporating temporality and biophysical vul-
nerability to quantify the human spatial footprint on ecosystems. Biol Conserv 144:1585–1594.
https://doi.org/10.1016/j.biocon.2011.02.004
GREUNAL (Grupo De Restauración Ecológica) (2010) Guías técnicas para la restauración
ecológica de ecosistemas. Convenio de asociación no. 22 entre Ministerio de Ambiente,
Vivienda y Desarrollo Territorial (MAVDT) y la Academia de Ciencias Exactas, Físicas
y Naturales (ACCEFYN). Departamento de Biología, Facultad de Ciencias, Universidad
Nacional de Colombia
IAvH (2012) Cartografía de Páramos de Colombia Esc. 1:100.000. Proyecto: Actualización del
Atlas de Páramos de Colombia. Convenio Interadministrativo de Asociación 11–103, Instituto
Humboldt y Ministerio de Ambiente y Desarrollo Sostenible. Bogotá D.C. Colombia
IDEAM (2013) Mapa de susceptibilidad de inundación escala 1:100.000
IDEAM (2014) Subdirección de Ecosistemas e Información Ambiental. Grupo de Bosques.
Proyecto Sistema de Monitoreo de Bosques y Carbono. Bogotá, D.C., Colombia
Isaacs P (2014) Composición y configuración de los páramos de Colombia. En: Restauración
ecológica de los páramos de Colombia. Transformación y herramientas para su conservación.
Instituto de Investigación de Recursos Biológicos Alexander von Humboldt (IAvH), Bogota,
D.C., 296 pp
Legendre P, Legendre L (1998) Numerical ecology. Elsevier, Ámsterdam
McGarigal K, Cushman SA, Neel MC, Ene E (2012) FRAGSTATS: spatial pattern analysis program
for categorical maps. Computer software program produced by the authors at the University of
Massachusetts, Amherst. www.umass.edu/landeco/research/fragstats/fragstats.html
Ministerio de Ambiente y Desarrollo Sostenible de Colombia (MADS) (2015) Plan Nacional
de Restauración: restauración ecológica, rehabilitación y recuperación de áreas disturbadas.
Bogotá, D.C., Colombia. ISBN: 978-958-8901-02-2
Murcia C, Guariguata MR (2014) La restauración ecológica en Colombia: Tendencias, necesidades
y oportunidades. Occasional paper 107. CIFOR, Bogor. https://doi.org/10.17528/cifor/004519
Ramírez W, Murcia C, Guariguata M, Thomas E, Aguilar M, Isaacs-Cubides P (2015) Restauración
Ecológica, los retos para Colombia. En: Biodiversidad 2015. Estado y tendencias de la biodiver-
sidad continental De Colombia. Instituto de Investigación de Recursos Biológicos Alexander
von Humboldt (IAvH), Bogota, D.C.
Razón pública (2018) Los costos de las lluvias torrenciales: el caso de Mocoa. https://www.
razonpublica.com/index.php/econom-y-sociedad-temas-29/11345-los-costos-de-las-lluvias-
torrenciales-el-caso-de-mocoa.html
Sánchez (2011) Documentos de trabajo sobre economía regional. Banco de la República, Centro
de Estudios Económicos Regionales. ISSN 1692–3715 https://www.banrep.gov.co/sites/
default/files/publicaciones/archivos/DTSER_150_0.pdf
Servicio Geológico Colombiano (2013) Mapa de susceptibilidad de erosión escala 1:100.000
SIB Colombia (2019) https://sibcolombia.net/biodiversidad-en-cifras-2019/
Yepes AP, Navarrete D, Duque aJ, Phillips JF, Cabrera KR, Alvarez E, García MC, Ordoñez MF
(2011) Protocolo para la estimación nacional y subnacional de biomasa - carbono en Colombia
Chapter 4
Ecohydrology of Tropical Andean Cloud
Forests
Conrado Tobón
4.1 Introduction
Montane ecosystems around the world are found from the equator to the poles and
occupy approximately one fifth of the surface of continents and islands (Ives et al.
1997). In South America, the Andes, as the longest ridge in America, extends over
approximately 1.5 million km2, running from 11° N to 23° S, with altitudes up to
6000 masl. The tropical part runs mainly from Venezuela through Colombia,
Ecuador, and Peru, comprising area of approximately 35,824 km2 (Cuesta et al.
2009). Main ecosystems in these environments are the glaciers, páramos, and mon-
tane forests, including cloud forests; however, montane forests cover most of the
region, whereas páramos are insular formations around the highest peaks (Smith
and Cleef 1988).
The Tropical Andean Cloud Forests (TACF) are high elevation forests that appear
on the Andean ridge flanks, between 2000 and 3200 masl, altitudes that change
according to the specific site exposition, thus cloud or fog presence (Bruijnzeel
et al. 2011), and it is characterized by middle size trees (around 12 m high), abun-
dant epiphytes hanging from tree trunks and branches, which are embedded fre-
quently in fog, and low intensity and long duration rainfall events (Tobón 2009).
Associated with the altitude and fog conditions, the temperature is low, air humidity
is high, and solar radiation is patchy and predominantly low.
Several attributes make these ecosystems of special importance as the high bio-
logical diversity (Bruijnzeel et al. 2010), with endemic plants (Sklenár and Ramsay
2001; Myers et al. 2000), large water supply (Tobón 2009), and dominant climate
conditions of high air humidity and low temperatures. Consequently Andean eco-
systems are important not only for local people, but for the millions living in the mid
and lowlands, as main rivers bring fresh water from those ecosystems, which,
C. Tobón (*)
Universidad Nacional de Colombia, Medellín, Colombia

https://doi.org/10.1007/978-3-030-57344-7_4
62 C. Tobón
through the vegetation, including the bryophytes hanging from trees and standing
on soil surface (Tobón et al. 2010a), capture moisture from air masses and together
with rainfall falling throughout the year, infiltrates into the soil, and gets further
available as baseflow (Tobón 2009), during least rainfall periods.
Ecohydrology became the emergent science dealing with the study of the relation-
ships and mechanisms among climate, soils, and vegetation with the hydrological
processes at different scales, and the dynamics beyond these interactions seems to
differ between ecosystems (Rodríguez-Iturbe 2000). The ecohydrological signifi-
cance of cloud forest ecosystems lies on the interaction between the specific climate
conditions, which includes the presence and frequency of fog and low clouds, the
vegetation, including epiphytes (Bruijnzeel et al. 2010), and the reach organic matter
soils (Tobón et al. 2010b), playing an important role on water infiltration and storage
capacity, thus on water regulation. The provision and regulation of hydrological ser-
vices of these ecosystems are explained by a combination of combined inputs of
water, through rainfall and fog deposition, low evapotranspiration, the high infiltration
capacity of litter and soils, and moss and soil water storage (Tobón 2009). Overall,
these processes need to be characterized in TACF, and provide a unique opportunity
to understand the hydrological processes linked to the ecology of Andean cloud
forests.
Worldwide native forest is changing rapidly (Condit et al. 2005), and Andean
cloud forests are not the exception (Wassenaar et al. 2007; Bruinsma 2003).
Contradictory, while growing population demands larger amounts of ecosystems
services, as water, conversion of TACF to crop plantation and pastures affects, at
different rates, ecosystem water yield and water regulation (Bruijnzeel et al. 2011).
Moreover, these ecosystems are also particularly vulnerable to climate change
(Pounds et al. 1999; Barradas et al. 2010), thus the potential loss of actual specific
ecohydrological functioning. Overall, under the contemporary conditions of wide-
spread land use changes and climate change, water supply to population depending
on water from Andean cloud forest is at risk.
Recently there has been much concern about the hydrological importance of par-
amo ecosystems, under the consideration that they provide water for cities in Andean
region (Tobón 2009; Bruijnzeel 2004). This has enhanced by those authors that con-
sider that disappearance of Andean glaciers, as it is happening at high rates in the last
decades, will affect water supply for population in the region (Kaser et al. 2005). All
this concern left out the Andean cloud forests, which may be as important, or even
more important, for water supply that the previous ecosystems (Tobón 2009) consid-
ering that the extension of the area occupy by the TACF is much larger than the two
foregoing ecosystems, and annual rainfall amounts are also higher (Tobón 2009).
Although the tropical ecosystems offer a great opportunity to understand the
hydrological processes linked to the ecology of the native and transformed ecosys-
tems, the number of ecohydrological research projects remain limited. This was
demonstrated in a recent paper, which shows that the number of ecohydrological
studies carried out in the tropics remain very low, as compared from nontropical
regions (Wright et al. 2017). The number of these ecohydrological studies decrease
considerably for the Neotropics (Bendix et al. 2008), and in spite of the ecological
4 Ecohydrology of Tropical Andean Cloud Forests 63
and hydrological importance, the studies of the ecohydrology of the Andean cloud
forests have received only marginal attention (Garreaud 2009).
Interest in understanding ecohydrological processes of forest ecosystems and the
role of forests in affecting water supply and ecosystem services has been triggered
by recent worldwide water crisis and the ongoing climate change debate. Although
some processes governing the ecohydrological functioning of ecosystems, as tran-
spiration, rainfall interception, groundwater flows were early understood (e.g.
Hursch and Brater 1941) and remains up today (Bruijnzeel et al. 2011), other vari-
ables as fog water inputs and soil water dynamics have not been widely considered.
Moreover, the actual increased pressure on water resources, as water supply to
increase planet population, but also for crop production demanded (Jackson et al.
2001), makes ecohydrology a timely science.
It should be stated here that in this century, among several papers, some compre-
hensive books have been published concerning cloud forest (Bruijnzeel and
Hamilton 2000; Nadkarni and Wheelwright 2000; Bruijnzeel 2001; Bruijnzeel et al.
2010; Hamilton et al. 1995), and some of those included chapters regarding cloud
forest hydrology (Bruijnzeel et al. 2010), however, little information was included
in those, specifically related to Tropical Andean cloud forest, which agrees with the
scarce ecohydrological investigations carried out in these ecosystems in last decades.
Understanding the complex ecohydrological functioning of TACF and their role
in providing hydrological services is a key issue for policy makers, institutions
focused on conservation, and restoration of millennium ecosystems providing envi-
ronmental services, and for population depending on water flowing from these eco-
systems. Therefore, the objective in this paper is to present the current understanding
of the ecohydrological functioning of Tropical Andean Cloud Forests, through
determining and discussing the critical parameters that control this functioning. It is
also proposed the future research needed for more comprehensive understanding of
ecohydrology of these ecosystems, and the water supply regimes.
4.2 Methodology
In this paper we present the current understanding of the ecohydrological function-

ing of Tropical Andean Cloud Forest, which is, results from ecohydrological studies
made in cloud forests in Venezuela, Colombia, Ecuador, and Peru. The results pre-
sented in this paper come from two different sources: the revision of peer-review
articles in English and Spanish about ecohydrology and/or hydrology of Andean
cloud forests published in Science Direct, ISI Web of Science, Scopus, Redalyc,
Google Scholar, and SciELO. Database was scrutinized through 2019. Some docu-
ments of a more local nature have also been reviewed, such as thesis and project
reports (see Table 4.2). Further within the revised papers, we search for cited litera-
ture focused on hydrology or ecohydrology of the TACF, including Master and
PhD thesis.
The second source of data consisted in unpublished information generated by the
research group “Hydrology and modelling of ecosystems, Universidad Nacional de
64 C. Tobón
Table 4.1 Location and main characteristics of TACF studied in Colombia

Site 1. Cloud Forest Pantano Redondo—Zipaquirá
Location 5° 02′30.30” Altitude 3160 masl
N 74° 02′02.19”W
Average annual Average temperature: diurnal 14.7 °C and night 7,7 °C
rainfall 1615 mm
Exposition: Exposed to the Cundiboyacense Montane Plateau, with air masses coming from the
dry savannah
Main soil classes Humitropets, Dystrandepts, Cryumbrepts e Histosoles Typic
(Soil Survey Staff Placudands, Typic Hapludands, Lithic Hapludands, Pachic Melanudands
2004) y Typic Endoaquand (USDA 2014)
Main plant species Weinmannia tomentosa (encenillo), Drimys granadensis (canelo de
páramo), Clusia multiflora (gaque), Hedyosmum bonplandianum
(granizo), Cervantesia (santalaceas), Ilex (acebos), Vallea (raques),
Escallonia (tobos) y Myrica (laurel de cera) (Gentry 1991; CAR 2001)
Predominant land use Cloud forest, surrounded by pastures and crops
Site 2. Parque Nacional Natural Chingaza, Mundo Nuevo
N 73° 50′36.01”W
Average annual Average temperature. Diurnal 13.2 °C, night 9.3 °C
rainfall 3280 mm
Exposition: Exposed to the East, with air masses coming from the Amazonia and Orinoquia
(Colombia)
Main soil classes Typic Hapludands, Humic Dystrudepts, Humic Lithic Dystrudepts,
(Soil Survey Staff Typic Humudepts, Lithic Hapludands, Pachic Melanudands y Lithic
2004) Melanudands (USDA 2014)
Main plant species Weinmannia spp., Weinmannia microphylla, Drimys granadensis, Clusia
cf. Multiflora y Lepechinia conferta (este estudio)
Predominant land use Cloud forest and some small areas with pastures
Site 3. Cloud Forest Estrella de agua—Salento
N 73° 25′44.73” W
Average annual Average temperature. Diurnal 12.6 °C and night 7,2 °C
rainfall 2164 mm
Exposition: Exposed to the West, air masses coming from the Pacific
Main soil classes Andic Humicryepts, Typic Melanocryands, Typic Humudepts y Typic
(Soil Survey Staff Hapludands. (IGAC 2004)
2004)
Main plant species Ceroxylon quindiuense, Poulsenia armata, Ficus sp., Clusia lineata,
Clusia sp., Lippia hirsuta, Hyeronima oblonga y Cuatresia riparia (este
estudio)
Predominant land use Undisturbed cloud forest
Colombia,” through the research projects developed in three Andean cloud forest in
Colombia, since 2007 (Table 4.1). The climate of each site was characterized by
installing an automatic weather station (CampbellSci Ltd) to measure variables as
precipitation, temperature, solar radiation, air humidity, wind speed, and direction.
Table 4.2 The magnitude of the ecohydrological variables in Tropical Andean cloud forests (NI = No Information). F means fog inputs measured with a
cylinder and F2 is fog inputs measured as net precipitation under forest, only during fog events
Precipitation (P) Temperature (°C) Evapotranspiration
and fog inputs (F) and relative air forest interception Average soil water
Authors Location (coordinates and altitude) (mm y−1) humidity (%) (mm y−1) content
This study—site 1 Colombia P = 1615 14.7 434 0,58 cm3·cm−3
(Tobón and Arroyave 5° 04′56.02” N 74° 06′12” W F = 334 83.5 511
2007) 3074 masl F2 = 65
This study—Site 2 Colombia P = 3280 13.2 389 0,63 cm3·cm−3
4° 40′13.13” N 73° 50′36.01”W F = 465 89.0 575
3040 masl F2 = 154
This study—site 3 Colombia P = 2164 12.6 407 0,65 cm3·cm−3
4° 37′17.59” N 73° 25′44.73” W F = 421 91.0 485
3210 masl F2 = 128
Ramírez et al. (2017) Colombia P = 4588 14.5 NI 46%
72.900 E F = 80 96.3 NI
5.243 N
4 Ecohydrology of Tropical Andean Cloud Forests
2048 masl
Vásquez (2016) Colombia P = 2472 15.0 929 35%
4° 49′ 10” N, 75° 33′ 50” W F = ND 85.0 NI
2000 masl
Burbano-Garcés et al. Colombia P = 1488 20 NI NI
(2014) 2° 29′ N and 76° 32′ W F = ND 72 256
1870 masl
Clark et al. (2014) Peru P = 3112 14.2 688 NI
13° 3′ 37” S, 71° 32′ 40” W F = 316 NI 226
2805 masl
Jarvis and Mulligan (2010) Colombia Andes P = 2000 17.7 NI NI
2° 21′ 47” N and 76° 24′ 28” W F = NI NI
2000–2600 masl
(continued)
65
Table 4.2 (continued)
66
Precipitation (P) Temperature (°C) Evapotranspiration

and fog inputs (F) and relative air forest interception Average soil water
Authors Location (coordinates and altitude) (mm y−1) humidity (%) (mm y−1) content
Oesker et al. (2010) Ecuador P = 2737 16.2 NI NI
3° 58′ S and 79° 04′ W F = NI NI 247
2275 masl
León Peláez et al. (2010) Colombia P = 1725 15.2 NI 46%a
06° 18” N 75° 30” W F = NI 84.0 250
2490 masl
Bendix et al. (2008) Ecuador P = 2193 15.8 570 44.7%b
3° 58′ 30” S and 79° 4′ 25” W F = 210 1010
2270 masl
Ecuador P = 4779 13.4
2660 masl F = 527 86.3
Ecuador P = 4743 9.4 49.1%b
04° 06′ 71” S, 79° 10′ 581” W F = 1958 93
3180 masl
Gómez-Peralta et al. Peru P = 2753 14.5 NI NI
(2008) 10° 31′ 47” S, 75° 21′ 23” W F = 221 ND 211
2815 masl
Garcia (2007) Colombia P at 3370 = 1453 12.8 NI NI
4° 50’ N and 75° 30’ W F = 0 95 265
2550 masl
3370 masl
Fleischbein et al. Ecuador P = 2592 15.2 471 NI
(2006, 2005) 4° 00′ S and 79° 12′ W F = ND NI 985
2000 masl
Colombia P = 1313 13.2 465 70%
04° 30′ 42” N 74° 01′ 41” O F = ND 685
3000 masl
C. Tobón
Fonseca and Ataroff Colombia P = 3153 14.5 NI NI
(2005) 5° 25′ 13” N, 75° 45′ 17” W F = 438 91 1580
2350 masl
Ataroff (2002, 2005) Venezuela – La Mucuy P = 3125 14.0 498 26%
8° 38′ N 70° 02′ W F = 300 1751
2300 masl
Ataroff and Rada (2000) Venezuela P = 3124 14.0 558 23%
(8° 38′ N and 71° 02′ W) F = 309 90.3 1687
2350 masl
González (2000) Colombia P = 4120 13.0 NI NI
2° 30′ N, 76° 60′ W F = 371 94.0 NI
2050 masl
Jarvis (2000) Colombia P = 3900 19.0 NI NI
2.30 N, 76.59 E F = 2854 90.0 NI
1525 masl
Ataroff (1998) Venezuela P = 2959 12.0 NI NI
08° 36′ 59.95” S, 71°03′ 00.31” W F = 91 NI 1331
4 Ecohydrology of Tropical Andean Cloud Forests
2300 masl
Veneklaas and Van Ek Colombia P = 2115 15.4 NI NI
(1990) 4° 50′ N and 75° 30′ W. F = 0 91 262
2550 masl
Cavalier and Goldstein Venezuela P = 1983 8.0 NI NI
(1989) 2500 masl F = 72 NI NI
Steinhardt (1979) Venezuela P = 1575 NI 675 NI
2300 masl F = NI NI 305
a
Tobón (2009)
b
Moser et al. (2008)
67
68 C. Tobón
Total and average values were registered each 15 min. At each site, rainfall was also
measured nearby the forests, but outside the forest, by installing two automatic rain
gauges (Texas instruments) at different sites, which were programmed to record
total rainfall each 15 min. To measure fog inputs, we used Juvik-type fog gauges,
consisting of a louvered cylindrical aluminum shade screen (Juvik and Ekern 1978).
The fog gauges have a diameter of 12.8 cm and 42 cm in height, with a conical
aluminum cover on top, of 65 cm diameter to minimize the contribution of rainfall.
To quantify inputs, fog gauges were installed at 1.5 m height and connected to a
tipping bucket rain gauge (Texas instruments), recording data every 15 min. Fog
inputs were also determined through net precipitation measurements, which is fur-
ther explained.
At each cloud forest, two plots of 20x50 m were randomly selected inside each
forest site, to measure throughfall. This was measured by installing 4 V-shaped
stainless steel throughfall gutters of 4 m length and 0.32 m width, at each plot.
These troughs were placed at inclination of about 15° angle to facilitate drainage of
the collected water into a pre-calibrated tipping bucket of 51 mL (Vrije Universiteit
Amsterdam), which gives a resolution of about 0.051 mm per tip. Each tipping
bucket was connected to Tinytag data logger (Gemini data logger), which measured
water inputs each minute and register total values each 15 min, coupled with rain
gauges outside the forest. To characterize the effects of forest structure on through-
fall, gutters were randomly moved to a new site (within the 20 × 50m), each
3 months. Troughs were also used to measure water inputs generated by drips fall-
ing from the canopy during fog events. In this case, we consider fog inputs those
values registered, either in isolated fog events (events without rainfall) or the drip-
ping occurring sometime after rainfall had ceased when normally fog comes after
the rain (based on field observations inside the forests, for the studied sites in
Colombia, we consider a lapse of 15 min, as the time required for the forests to drain
all drops from the rain). For better precision on fog inputs, we also calculated the
amount of water required to wet the inner part of the troughs, allowing us for the
real determination of water inputs in fog events, but also for throughfall, therefore a
value of 0.4 mm was added to all data registered, either to throughfall or to fog.
To determine the wetting of leaves during rainfall and fog events, leaf wetness
was measured using a Decagon dielectric wetness sensor (LWS-L Decagon
Devices), connected to a CR1000 datalogger (Campbell Sci). LWS were up-facing
installed at three different heights, from the tree crown: one on top, the second 2 m
below, and the third one, at 5 m below the treetop. Sensor outputs are mV, where the
higher the mV registered on the datalogger, the larger the leaf wetness. Measurements
were made at intervals of 30 s with averages each 15 min. Sensors were calibrated
at the laboratory, to interpret, first the value at which the sensor, as the canopy, is
dry, second, the rate at which canopy is getting wet, and third, the percentage of the
canopy that got wet. To this purpose, before taking the sensors to the field, in the
laboratory we wet part by part the sensors (from around 5% till 100%), and made
visual observations of the fraction of the sensor which was wet, and compare with
measurements (mV) registered in the datalogger, at time intervals of 10 s.
The second component of net precipitation is stemflow. Some authors indicate

that in cloud forests the amounts of stemflow are insignificant (<2% of gross pre-
cipitation) when compared to throughfall (Bruijnzeel 2004; Dietz et al. 2006;
Gómez-Peralta et al. 2008; Holwerda et al. 2010¸ Bruijnzeel et al. 2011; Muñoz-
Villers et al. 2012), however, in studied cloud forest in Colombia, stemflow amounts
were measured at each plot and each forest, according to Tobón et al. (2000).
Therefore amounts of water flowing down the tree trunks were measured using flex-
ible transparent tubing that was cut longitudinally and wrapped in a downward spi-
ral around tree trunks. The tubing was nailed to the tree trunk, and sealed with
silicone sealant. Tubing were channeled into a tipping-bucket guage (Texas instru-
ments), connected to a Tinytag data logger (Gemini dataloggers). Stemflow was
computed (mm), by using the projected area of each tree, as the input exposed area
to rain. Net precipitation was computed as the sum of throughfall plus stemflow for
each rainfall event, and rainfall interception was calculated as the difference between
gross rainfall outside the forests and net precipitation.
At each specific site (plots of 20x50m), volumetric soil water content (θ, cm3
cm−3) was automatically measured. To this purpose, we used Time Domain
Reflectometry (TDR) sensors connected to a CR1000 datalogger (Campbell
Scientific Ltd., Shepshed, UK). At each site, a soil pit of 2.0 m × 1.5 m × 1.5 m was
excavated and TDR sensors were horizontally installed at five different depths rep-
resenting the respective soil horizons (5, 15, 35, 50, 80, or 100 cm depth). The
probes were dug into the upslope face of the pit with the depth of the hole being
equal to the measuring depth to minimize the effect of soil disturbance created by
digging into the soil to install sensors. The pits were backfilled after installing the
TDR probes, making sure to return the respective soil layers to their original posi-
tions. Measurements were made on a minute basis, registering average values each
15 min, over the total measured period at each forest (Table 4.1). Field measure-
ments were calibrated according to Tobón et al. (2010b).
For most investigated sites in TACF no discharge was measured, except for two
out of the three sites in Colombia and sites in Ecuador (Bendix et al. 2008).
Consequently water yield from these ecosystems was calculated using the water
balance approach, applying the general equation Q = P + F – ETa – I – dS/dt, where
Q is discharge, P is rainfall, F is the net water input by fog, I is the forest intercep-
tion of rainfall, and dS/dt is the storage changes per time step, all in units of mm y−1.
4.3 Results and Discussion
4.3.1 The Ecohydrology of TACF
The Andes mountain range is the result of the tectonic activity and continental drift
in the South American continent (Brown and Lomolino 1998; Pielou 1979), which
originated from the collision of the Nazca plate with the South American plate
70 C. Tobón
(Fittkau et al. 1968), event that gave rise to a complex chain of mountains that
extends from the south of the continent to Venezuela and includes a variety of eco-
systems, including the tropical Andean cloud forests (TACF). Although TACF do
not have a defined altitude range, as they are distributed at different altitudes depend-
ing on the environmental conditions existing at each site, the conditions of tempera-
ture, mainly dependent on the altitude, their exposure to currents of humidified air
masses, and the condition of inversion of the trade winds (Stadtmüller 1987), in
South America these ecosystems are mostly located above 2000 and below
3200 masl., that is, below the limit of the páramos, where those exists (Beck et al.
2008; Young 2006; Rada 2002; Fontúrbel 2002; Föster 2001; Hamilton et al. 1995;
Stadtmüller 1987), and main differences between them, concern to the amount and
frequency of fog (Tobón 2009; Lawton et al. 2001), which normally can be seeing
by the amounts of mosses on tree branches and trunks.
Unlike other Andean forests, the Tropical Andean cloud forests have specific
vegetation composition, that is, the abundance of epiphytes, mainly mosses, liver-
worts, lichens, bromeliads, and orchids attached to the tree trunks or hanging from
the branches (Fig. 4.1), which largely constitute the lower strata or undergrowth of
these ecosystems. While some authors reported biomass values of around 16 ton/h
for Costa Rican cloud forests (Köhler et al. 2007), Hofstede et al. (1993) reported a
value of 44 ton/h for TACF in Colombia. Values may range from site to site (Köhler
Fig. 4.1 Tropical Andean Cloud forest with epiphytes hanging from tree trunk and branches
et al. 2011), which seems to be dependent on the permanence and frequency of fog,
among other variables, as disturbance (Fig. 4.1).
Connected to this vegetation, the TACF have a specific water dynamics, mainly
connected to the high frequency of low density rainfall events and frequency of fog,
the last generating and additional water input to these ecosystems (Tobón et al.
2008; Tobón and Arroyave 2007; Bruijnzeel 2001; González 2000), through the fog
water interception by vegetation and falling down into the forest floor (Tobón et al.
2010b, Richardson et al. 2000) and the fog control on plant transpiration, as the
result of decrease on solar radiation and increases on air humidity during fog events
(Ferwerda et al. 2000). Additionally, climate conditions of TACF are mainly con-
trolled by altitude and winds from the Pacific and the Atlantic, maintaining constant
humidity on both external slopes of the Cordilleras, while conditions are more vari-
able on the inner flanks. Condensation is significant on the upper portions of the
inner flanks, and the middle and lower portions of the valleys have a marked bimodal
dry–wet pattern resulting from the rain shadow effect (Herzog et al. 2011).
4.3.2 K
ey Variables that Control Ecohydrological
Processes of TACF
Not different from other terrestrial ecosystems, the TACF receive different amounts
of precipitation (Espinoza et al. 2009; Emck et al. 2006; DeAngelis et al. 2004;
Liebmann et al. 2004; Vuille et al. 2000; Póveda and Mesa 1997; Grubb and
Whitmore 1996), ranging from 500 to 5000 mm y−1 (Schawe et al. 2008; Sarmiento
2001), depending on their exposure and position within the Andean mountains
(Martínez et al. 2011). A long series of data (1998–2004) indicates that precipitation
on the eastern side of the Andes ranges from 2060 mm y−1 at an altitude of 1960 m
to 4400 mm y−1 at 3200 m (Oesker et al. 2008), indicating that variability depends
on the altitudinal position of each site. A different tendency is found in Colombia,
when studying sites located at similar altitude, but exposed to different provenance
of air masses. Those TACF exposed to the Pacific air masses, but located far from
the ocean (see Table 4.1, site 3), receive intermediate values of precipitation, as
compared to those directly exposed to the Amazonia (Table 4.1, site 2), while eco-
systems exposed to the plateau valleys receive the least precipitation (Table 4.1, site
1). This implies that when considering data from different sites at similar altitudes,
rainfall variability is related to differences in air masses ascending the Andean
mountains: some very humid such as those facing the Pacific basin (Póveda et al.
2005; González 2000), and those from the Amazon basin (Espinoza et al. 2009;
Tobón 1999).
In this study it was found that rainfall in TACF is characterized by low magnitude
and intensity (Tobón 2009; Ataroff 2005; Ataroff 2002), values ranging from to
1575 to 4588 mm y−1, with an average of 2337 mm y−1. Although there is not a clear
tendency, rainfall seems to increase with altitude, with some differences between
72 C. Tobón
those TACF facing the east or the west, the last receiving more precipitation; con-
trary to that found by some authors (Espinoza et al. 2009; Bendix et al. 2006; Lara
et al. 2003), who indicated that Andean cloud forest facing the east receive greater
precipitation than those oriented to the west. Moreover, yearly amounts of precipita-
tion decrease with altitude towards the interior valleys, which is, those TACF
exposed to the inter-Andean valleys are relatively dry. This was also found in other
cloud forests (DeAngelis et al. 2004; Liebmann et al. 2004; Lara et al. 2003; Vuille
et al. 2000). In brief, depending on their position in the altitudinal gradient
(Rollenbeck et al. 2008; Harden 2006; Rollenbeck 2006; Bendix et al. 2004; Lauer
1981), and degree of exposure to these cloud masses (Espinoza et al. 2009; Póveda
and Mesa 1997), TACF receive certain amounts of vertical and horizontal precipita-
tion, including fog water (Table 4.2).
Fog water seems to contribute significantly to the ecohydrology of TACF ecosys-
tems, directly or indirectly (Tobón et al. 2008; Tobón and Arroyave 2007; Dengel
and Rollenbeck 2003; Bruijnzeel and Hamilton 2000; Hamilton et al. 1995;
Stadtmüller 1987, 2003). Directly, fog water intercepted by TACF may fall into the
forest floor (Bruijnzeel et al. 2010), therefore contributes to net water inputs to the
ecosystem. This interception of fog water by TACF is not only an action by forest
canopy (Holder 2004), but also by epiphytes, mainly mosses and lichens attached to
tree branches and trunks are involved in this process, since they act like a sponge
capable of retaining moisture from fog (Tobón et al. 2010a; Köhler et al. 2007;
Walker and Ataroff 2005; Mulligan and Jarvis 2000; Cavelier 1996; Veneklaas and
Van Ek 1990), to subsequently release it drip to the soil surface (Coxson and
Nadkarni 1995; Lovett et al. 1985). Water contribution made by fog also depends on
fog density and frequency, wind speed, and the presence of dense vegetation that
captures the water present in the fog (Cárdenas, Tobón and Buytaert 2017; Frumau
et al. 2011; Villegas et al. 2008; Holder 2004; González 2000; Cavelier 1991).
Average cloud frequency in studied sites in Colombia (2007–2016) was
0.93 ± 0.12 events per day, with the lowest value at site 1 and the highest at site 2
(Table 4.2). This value is higher than that presented by Mulligan et al. (2010), using
MODIS cloud climatology, which means that TACF is wrapped at least once a day,
being more frequent during the dry periods (January to March and July to
September); however those TACF facing a relative dry montane plateau, as the one
in Guerrero (site 1), experience much less events, and most of them occur during the
wet periods. Here most of the fog events (58%) occurred early in the morning
(between 5:00 and 9:00 am), and late in the afternoon and early night (36%).
The Andean altitudinal gradient of precipitation is accompanied by a gradient of
fog water inputs, increasing with the altitude, for studied sites (Table 4.2). However
values found by Bendix et al. (2008) and Jarvis (2000) in Ecuador and Colombian
cloud forests, respectively, are much larger than all values reported by most TACF
(see Table 4.2) and elsewhere, in cloud forests (Bruijnzeel et al. 2011). Average
value for water inputs through fog deposition to TACF is 295 mm y−1 without con-
sidering the above-mentioned data. Mostly, large values of fog water inputs can be
the result of mixing fog inputs with horizontal precipitation, under events with high
wind speed conditions (Frumau et al. 2010), or erroneous potential values predicted
from cloud deposition (Jarvis 2000).
Moreover, based on measurements in the three sites in Colombia, the amounts of
fog inputs registered through the Juvik method were always higher than those mea-
sured with gutters, as net precipitation during fog events (Table 4.2), the last being
considering as real fog water inputs to the ecosystems (Bruijnzeel et al. 2011). From
the three sites in Colombia, annual inputs of fog water measured with Juvik cylin-
ders averaged 407 mm, while net fog inputs to troughs installed under the forest
were only 116 mm y−1, in average (Table 4.2). From field observations these differ-
ences mostly relate to total surface area exposed to fog, acting as barriers trapping
water from fog, but also the texture of measurements devices and texture of vegeta-
tion. Cylinders seem to be more efficient on capturing the very small fog drops,
while in the forests, fog mostly gets in contact with treetops, but most of it passes
over, thus, very small amounts of fog water can be captured by exposed tree foliage
and branches, generating low values of throughfall. Moreover, it was observed in
the field, that very little of the fog that comes in, enters the forests, therefore little or
almost no water is trapped by undergrowth or leaves and branches below the tree-
tops. This was confirmed by data from the leaf wetness sensors, when analyzing
results only from fog events. Measurements show that those installed on top of the
three, got wet during most fog events, but only during prolonged and dense fog
events, those installed 2 m and 5 m below the tree crown, get wet (Fig. 4.2).
400 0.00
0.10
380
0.20
360 0.30
Leaf wetness (mV)
0.40
340
Milimeters
0.50
320
0.60
300 0.70
0.80
280
0.90
LW (Top) LW (-2m)
260 1.00
4:15:00
9:30:00
1:15:00
6:30:00
3:30:00
8:45:00
0:30:00
5:45:00
2:45:00
8:00:00
5:00:00
2:00:00
7:15:00
4:15:00
9:30:00
23:00:00
14:45:00
20:00:00
11:45:00
17:00:00
22:15:00
14:00:00
19:15:00
11:00:00
16:15:00
21:30:00
13:15:00
18:30:00
23:45:00
10:15:00
15:30:00
20:45:00
12:30:00
17:45:00
23:00:00
Time (hours, 2009)
Fig. 4.2 Leaf wetness through forest canopy during fog events in Chingaza, TACF
74 C. Tobón
0.820 0.00
Volumetric water content (cm3 cm-3)
0.780 0.20
0.740 0.40
Milimeters
0.700 0.60
0.660 0.80
0.620 1.00
0.580 1.20
8:00:00
23:00:00
2:00:00
5:00:00
17:00:00
8:00:00
2:00:00
11:00:00
14:00:00
23:00:00
11:00:00
5:00:00
20:00:00
8:00:00
2:00:00
8:00:00
20:00:00
23:00:00
5:00:00
2:00:00
17:00:00
11:00:00
17:00:00
5:00:00
8:00:00
14:00:00
20:00:00
17:00:00
14:00:00
23:00:00
11:00:00
14:00:00
23:00:00
Time (hours, 2009-2010)
5 cm 15 cm 35 cm
Fig. 4.3 Temporal and spatial dynamics of soil water content during two dry periods (2009–2010),
but with water inputs through fog interception (Chingaza, TACF)
Moreover when analyzing data on soil moisture it is observed that during very
long duration and dense fog events, soil moisture slightly increased at soil surface
(5 and 35 cm depth) (Fig. 4.3); however during most fog events, although data was
registered as leaf wetness, and fog water inputs, soil moisture sensors did not show
any increase in soil humidity. This is mostly related to the fact that soils in studied
TACF in Colombia (Table 4.1) remained either close to field capacity or saturated,
almost during the entire studied period, thus small water inputs into the soil surface
could not be registered, given the sensibility of TDR sensors (Campell Sci).
Indirectly, the presence of fog covering the TACF intercepts part of the solar
radiation, thus reduces direct radiation into the forest, which in turn reduces tem-
perature and increases air humidity (Tobón 2009). These combinations of specific
environmental conditions during the fog events reduce plant transpiration and water
evaporation from the trees crown (Jarvis and Mulligan 2010). Studies carried out in
TACF in Colombia indicated that evapotranspiration during foggy days was reduced
in average, 63%, as compared to the value found for open sky day, which is 1.47 mm/
day in average for measured events. Noteworthy that, for the same days with fog,
inputs by fog drip into the gutters were, in average, of the order of 1.75 mm. It
should be stated that during short fog events (shorter than 30 min) or under light fog
events (those events were objects can be seen and recognized at a distance shorter
than 10 m), evapotranspiration was slightly reduced (lower than 10% as compared
0.800
0.750
Volumetric water content (cm3 cm-3)
0.700
0.650
0.600
0.550 (5 cm) (15 cm) (35 cm) (50 cm)
(100 cm) FC 5 cm FC 15 cm FC 35 cm
0.500
3 3 4 4 4 5 5 5 6 6 6 7 7 7 8 8 8 9 9 9 1010101111121212 1 1 1 2 2 2 3 3 3
Time (months, 2009-2010)
Fig. 4.4 Spatial and temporal dynamics of volumetric water content in soils from Chingaza
TACF. Values of field capacity (FC) are also shown for the three first soil layers
to that at open sky) and no inputs of water were registered on the troughs installed
under the TACF (unpublished data).
Although the reduction on plant transpiration during foggy events was not
directly measured in studied cloud forests (e.g. through sapflow measurements),
available data on volumetric soil water content shows that during fog events, soil
moisture within the fine root zone remains constant, which implies that there were
no plant transpiration during such foggy periods. This could be proved in a foggy
day during the dry period (e.g. February 13, 2010) when soil moisture remained
constant at all soil depths (Fig. 4.4), while reduction on volumetric water content
was observed in those open sky days, values been in the order of 2.89 ± 0.67 mm/
day, in the first 50 cm of the soil profile.
Rainfall interception by forests is defined as the amount of rain that is intercepted
by vegetation and drains into the soil surface or evaporates (Fleischbein et al. 2005;
Tobón and Arroyave 2008; Tobón et al. 2000a; Veneklaas and Ek 1990; Vis 1986).
The fraction reaching the soil surface is called net precipitation and the fraction that
evaporates becomes the net loss of water from forest ecosystems together with plant
transpiration. Moreover, under forests, the litter lying on the forest floor intercepts
part of the water entering forest ecosystems, as forest floor interception (Tobón et al.
2000b). These have been considered as the most important ecohydrological pro-
cesses in forest ecosystems (Savenije 2004; Gerrits et al. 2007).
76 C. Tobón
Reported values of rainfall intercepted by TACF range from 7 to 52% of gross

rainfall and fog inputs measured with devices outside the forests (Table 4.2), with an
average value of 28%. Large values of rainfall interception reported for TACF are rare
and mostly related to the specific methodologies used to measured net precipitation
and the number of rain gauges located under the forest. When low number of gauges
are used to measure net precipitation (Fonseca and Ataroff 2005; Ataroff 2002; Ataroff
and Rada 2000; Wolf 1993), the amount of water intercepted by the forest can be
either, very low or very high, depending on rainfall amounts and characteristics, and
canopy characteristics as the fraction of crown cover or leaf area index (Tobón et al.
2000a). Values can also be dependent on fog frequency and drop density in fog, forest
disturbance, and abundance of epiphytes (Hofstede et al. 1993; Bruijnzeel 2001;
Köhler et al. 2007, 2011; Tobón et al. 2010a). For those cloud forests were fog fre-
quency and density is high, amounts of rainfall interception seem to be low (Mulligan
et al. 2010; Hölscher et al. 2004). Contrary, the presence of large amounts of epiphytes
biomass in some cloud forests seems to be capable of interception large amounts of
rainfall (Oesker et al. 2010; Tobón et al. 2010a; Tobón and Arroyave 2007; Hölscher
et al. 2004) although the real ecohydrological contribution of epiphytes in montane
tropical forests is still under discussion (Veneklaas and Van Ek 1990; Coxson and
Nadkarni 1995; Clark et al. 2014; Hafkenscheid 2000; Köhler et al. 2007, 2011),
given their capacity to capture water from fog (Tobón et al. 2010a).
Unpublished data from Colombian sites (Table 4.1) has shown that, when using
same methodology (e.g. same devices and same number of devices, with reloca-
tion), amounts of rainfall interception (including fog) are much lower (13% of gross
precipitation plus fog inputs) than reported values for all TACF (Table 4.2) and
depend on leaf area index and fog incidence. As indicated above, during fog events,
the canopy gets wet (Fig. 4.2), therefore less water from rain is required to saturate
the forest canopy in these TACF. Consequently, low values reported for these sites
(Table 4.2) will clearly change if the fog was not present.
Soils from TACF have received little attention, including poor soil classification
and when those exists, they are general soil studies, to the point that most studies
refer to them with the general words “tropical soils” (IGAC 2004). This becomes
worse for studies on the hydro-physical properties and soil water dynamics (Tobón
et al. 2010b), as compared with other terrestrial ecosystems, including tropical
montane ecosystems (Tobón et al. 2010b; Tobón 2009; Podwojewski et al. 2002).
This is surprising considering the importance of TACF as water supplier for a large
Andean population and the potential threat posed to this function by land use
changes and climate change (Eller et al. 2015; Beninston 2003; Walther et al. 2002).
In spite of this fact, the few studies carried out in TACF that included soil descrip-
tion, coincide that these have high organic matter content, which is connected to
vegetation, including the abundance of epiphytes (Köhler et al. 2007, 2011; Nadkarni
1984), the low temperatures, high soil moisture values throughout the year (Tobón
2009), and the stable organo-metallic complexes that reduce the biological activity
and therefore decrease the rate of organic matter mineralization (Bohlman et al.
1995; Vance and Nadkarni 1990).
Soil water content not only determines the excess of water that becomes dis-
charge or groundwater recharge, but also the water availability for plants and trans-
fer processes between soil–plant–atmosphere (Famiglietti et al. 1998), however
temporal and spatial variability may constraint a proper role of soil moisture on
those processes (Dobriyal et al. 2012). Moreover, soil moisture varies considerably
with depth, mainly on those heterogeneous soils with contrasting textures between
successive soil layers (Hincapie and Germann 2009; Hincapié and Tobón 2012;
Neary et al. 2009; Lazarev et al. 2005), while soil moisture shows more stability on
high organic matter soils.
Data from studies listed in Table 4.2 indicate that average soil water content var-
ies considerably between sites. Average value for those sites where soil moisture
was measured is around 45%, implying that almost half of the soil components is
water. However this percentage may vary depending on the rainfall temporal condi-
tions. During the dry seasons, the lowest values observed were around 23%, while
during the wet seasons, soil remains near field capacity or saturated, depending on
the frequency of rainfall/fog events.
From the studies carried out by this author in the three sites in Colombia, soil
moisture showed to be persistently high, remaining close to field capacity through-
out the year, except for those short dry periods when soil water content gently
decreases below field capacity (Fig. 4.3). Contrary, during the very wet periods
(March to June and October to December), studied TACF soils remain between field
capacity and saturation. This unique behavior seems to be related, first to the con-
tinuous low intensity rainfall and the frequency of fog events, and second, to the
large amounts of organic matter content in these soils, the presence of volcanic
ashes or both. At Guerrero and Chingaza cloud forests, soils present large amounts
of soil organic matter (between 17.8 and 25%), which partly controlled their soil
moisture dynamics, while in Frontino cloud forests, the water dynamics are con-
trolled by both, soil organic matter (26.5%) and volcanic ashes (unpublished data).
Contrary to that indicated by Jung et al. (2010) on limited moisture environ-
ments, the high soil water content in TACF indicates that plant transpiration is not
limited by soil moisture, at least not for water deficit. However, no studies carried
out in TACF have shown the effects of water stress under saturation condition. A
study from different forest types showed that water excess in the soil root zone lim-
ited plant transpiration (Purdy et al. 2018), therefore affecting plant growing and
physiological forest dynamics (Fig. 4.3).
In studied TACF in Colombia, changes in soil water content were larger for the
first 0.4 m, but only during the short dry periods, when soil water content slightly
decreased from field capacity, but never reached the wilting point, at any soil depth
(Fig. 4.4). This was in line with fine root distribution, whose results indicated that
most fine roots in studied TACF concentrate in the topsoil. Smallest variability was
observed during wet periods, when soil moisture remained close to field capacity,
and to some extent, close to saturation, mostly controlled by rainfall and possibly
decreasing evapotranspiration due to the presence of low clouds and fog, less radia-
tion, high air humidity, and permanent low temperatures. Additionally, as indicated
above, during some fog events, soil moisture slightly increased, which was clear
78 C. Tobón
during the short dry periods (Fig. 4.3); however it was observed through the through-
fall measurements that fog water inputs to TACF occur throughout the year, but
increases on soil water content cannot be visible, as soil moisture remains very high,
close, either to field capacity or saturation, therefore small water inputs are not
reflected on soil water content increases. This phenomenon may occur in most
organic or volcanic soils, where moisture remains very close to, either field capacity
or saturation, thus, small water inputs through fog deposition cannot be observed by
measuring soil moisture.
Finally, data from studied TACF in Colombia shows that although small varia-
tions on soil moisture occur with soil depth and through the soil profiles (mainly
during the short dry periods), an analysis of the root mean square error (RMSE) of
the relative differences in volumetric soil water content measurements at the differ-
ent depths ad sites showed that there were some temporal stability of soil water
content, throughout the soil profile and time. However, soil water content was tem-
porarily more stable at those soil horizons richer on organic matter than on those
volcanic layers. This can be explained by the large amounts of soil organic matter,
with large water storage capacity (Tobón et al. 2010b; Hincapié and Tobón 2012)
and the specific climatic conditions of TACF, notably, the frequency of fog, which,
as indicated above, reduces plant transpiration, and by dripping into the forest floor,
add some extra water to the soil surface, thus, any extra water inputs may compen-
sate for the outputs by forest transpiration.
Although the low number of studies that included calculations of evapotranspira-
tion (ETa), values in Table 4.2 indicated that most TACF have low rates of evapo-
transpiration, which are compared with values from most cloud forests (Silva et al.
2017; Rittera and Regalado 2017; Bruijnzeel 2001, 2004), but differ from other
forest types (Tanaka et al. 2008; Tobón 1999). ETa in TACF averages 553 mm y−1,
with larger values for sites below 2300 masl. Highest values are mostly related to
differences on climatic conditions within sites, but also on fog frequency. In studied
sites in Colombia, values of ETa were lower for Frontino and Chingaza sites, which
mostly relates to fog frequency. The low temperature and moist conditions provided
by fog may reduce leaf temperature and vapor pressure deficit thus enhancing sto-
mata conductance (Smith and McClean 1989).
Data on evapotranspiration of TACF ecosystems shows that there are two main
tendencies on the magnitude of ET. The first is related to the occurrence of humid
and dry periods in the tropics. Larger values are observed during the dry periods or
less-humid periods, and on open sky days. However when there were some fog
events in these periods, ET was reduced considerably, due to that fog reduced vapor
pressure deficit. The second relates to fog presence and frequency. Leaf wetness,
connected to either rainfall or fog events (Fig. 4.5), suppress plant transpiration
(Gotsch et al. 2014). This implies that in sites where either rainfall events are fre-
quent or fog events are persistent and of long duration, tree transpiration must be
reduced by both, the increase in leaf wetness and the changes in meteorological
conditions connected to the presence of fog. The last was observed in the studied
sites in Colombia, showing that during rainfall events and/or dense fog events last-
ing for at least 1 h, forest canopy exhibited some leaf wetness, being these larger as
larger and dense was the event (Fig. 4.2).
950 0
2
850
4
750 6
Leaf wetness (mV)
8
650
Milimeters
10
550
12
450 14
16
350
18
250 20
10:45:00
13:15:00
15:45:00
18:15:00
20:45:00
04:15:00
11:45:00
14:15:00
16:45:00
19:15:00
21:45:00
05:15:00
10:15:00
20:15:00
22:45:00
23:15:00
01:45:00
06:45:00
09:15:00
00:15:00
02:45:00
07:45:00
12:45:00
15:15:00
17:45:00
Time (Hours, 2009)
LW (Top) LW (-2m) LW (-5m) Rainfall Fog
Fig. 4.5 Spatial and temporal dynamics of leaf wetness during rainfall and fog events in
Chingaza TACF
Moreover, provenance of air masses and their moisture seems to control evapo-
transpiration from TACF ecosystems. When air masses come from areas relatively
wet (Pacific and Amazon basins), ET is lower (see sites 2 and 3) than when these air
provenance is from a relatively dry valley, as the case site 1 (Guerrero). In this case,
saturated vapor pressure restricts evapotranspiration, no matter if soil water is avail-
able for plant transpiration. This indicates that contrary to the global tendency of
land evapotranspiration decline found by (Jung et al. 2010), ETa data from TACF
shows that actual evapotranspiration rates are not controlled by soil moisture, which
remains near to field capacity, throughout the year, but for atmospheric control
parameters as fog, the frequency of rainfall events, and wet air masses, reducing
incident solar radiation and temperature, increasing air humidity, and wetting tree
crowns (Fig. 4.5).
Water balance for a given ecosystem allows us for a better understanding on the
relationships between water inputs and outputs, specifically in cloud forests, where
no discharge has been measured, where its calculation renders the water yields.
Water cycle within tropical cloud forests is a very dynamic and complex process,
both spatially and temporally, therefore accuracy of water balance is highly depen-
dent on precision of measured hydrometeorological variables, as rainfall and fog
inputs, rainfall water interception, evapotranspiration, and soil moisture temporal
dynamics, preferably on annual or multiannual basis.
80 C. Tobón
Applying the general equation for the water balance in those sites cited in
Table 4.2, water yield from TACF ecosystems ranges between 0.34 and 0.73, with
an average value of 0,48 ± 0.16, implying that for each millimeter of water entering
to these ecosystems, either though rainfall or fog, 0.48 mm of water leaves the eco-
system through discharge. This value is lower than the average value presented by
Tobón (2009) for páramo ecosystem, but higher than values found for tropical rain
forests (Jones et al. 2017; van Dijk et al. 2012; McJannet et al. 2007) and much
higher than those from tropical dry forest (Allen et al. 2017; Caldwell et al. 2016).
It should be stated here that in most hydrological studies reported for TACF
(Table 4.2), no account has been taken for fog inputs, and in most cases evapotrans-
piration has been evaluated as potential evapotranspiration, which may generate
errors on the final calculation of water yield. In those studies where actual evapo-
transpiration and fog inputs were considered, the average water yield by TACF is
0.56, which is similar to that value found for tropical cloud forests (Tobón 2009).
4.4 Conclusions
Within the framework of this study it is clear that vegetation of TACF, including the
epiphytes, modifies water inputs to the ecosystems: 28% of total water inputs is
intercepted and evaporated from the forest canopy, which becomes in a net loss
from the ecosystem, while fog water is also intercepted and new inputs of water are
added to the forest floor, whose proportion is 560 mm y−1, in average. Moreover,
authors indicate that during fog events, transpiration by plants is depressed.
Ecohydrologically, TACF are specifically influenced by four important factors:
Specific climate conditions, as low temperatures and high air humidity, relatively
high and continuous rainfall inputs, with almost no presence of dry periods, fog pres-
ence, in some sites frequently, and consequently low evapotranspiration rates. These
variables significantly control the ecohydrological functioning of undisturbed TACF.
Finally, a better understanding of TACF ecohydrology can be reached if future
studies include real fog inputs to these ecosystems, preferably measured through net
precipitation methods, and soil moisture can be also measured. Given the large spa-
tial and temporal variability, it will be preferred if studies can be made for long
periods on different TACF sites.
References
Allen K, Dupuy JM, Gei MG, Hulshof C, Medvigy D, Pizano C, Salgado-Negret B, Smith CM,
Trierweiler A, Van Bloem SJ, Waring BG, Xu X, Powers JS (2017) Will seasonally dry tropi-
cal forests be sensitive or resistant to future changes in rainfall regimes? Environ Res Lett
12(2):023001
Ataroff M (1998) Importance of cloud water in Venezuelan Andean cloud forest water dynamics.
In: Schemenauer RS, Bridgman HA (eds) Proceedings of the first international conference on
fog and fog collection. International Development Research Centre, Ottawa, pp 25–28
Ataroff M (2002) Precipitación e intercepción en ecosistemas boscosos de los andes Venezolanos.

Ecotropicos 15(2):195–202. Sociedad Venezolana de Ecología
Ataroff M (2005) Estudios de dinámica hídrica en la Selva Nublada de La Mucuy, Andes de
Venezuela. In: Ataroff M, Silva JF (eds) Dinámica Hídrica en Sistemas Neotropicales.
ICAE. Univ. Los Andes, Mérida
Ataroff M, Rada F (2000) Deforestation impact on water dynamics in a Venezuelan Andean cloud
forest. Ambio 29:440–444
Barradas VL, Cervantes-Pérez J, Ramos-Palacios R, Puchet-Anyul C, Vázquez-Rodriguez P,
Granados-Ramírez R (2010) Meso-scale climate change in the central mountain region of
Veracruz State, Mexico. In: Bruijnzeel LA, Scatena FN, Hamilton LS (eds) Tropical montane
cloud forests. Science for conservation and management. Cambridge University Press,
Cambridge, pp 549–556
Beck E, Bendix J, Kottke I, Makeschin F, Mosandl R (2008) Gradients in a tropical mountain eco-
system of Ecuador. Ecological studies (analysis and synthesis), vol 198. Springer, Berlin, 543 p
Bendix J, Fabiany P, Rollenbeck R (2004) Gradients of fog and rain in a tropical montane cloud
forest of southern Ecuador and its chemical composition. In: Proceedings 3rd Int. Conf. on
Fog. Fog Collection and Dew, 11–15 Oct. 2004. Cape Town
Bendix J, Rollenbeck R, Reudenbach C (2006) Diurnal patterns of rainfall in a tropical Andean
valley of southern Ecuador as seen by a vertically pointing K-band Doppler radar. Int J Climatol
26(6):829–846
Bendix J, Rollenbeck R, Richter M, Fabian P, Emck P (2008) Gradual changes along the altitu-
dinal gradients. The climate. In: Beck E, Bendix J, Kottke I, Makeschin F, Mosandl R (eds)
Gradients in a tropical mountain ecosystem of Ecuador. Ecological studies, vol 198. Springer,
Berlin, pp 63–73
Beninston M (2003) Climatic change in mountain regions: a review of possible impacts. Climate
Change 59:5–31
Bohlman SA, Matelson TJ, Nadkarni NM (1995) Moisture and temperature patterns of canopy
humus and forest floor soil of a Montane Cloud Forest, Costa Rica. Biotropica 27(1):13–19
Brown JH, Lomolino MV (1998) Biogeography, 2nd edn. Sinauer Associates, Sunderland, MA,
xii + 691 pp
Bruijnzeel LA (2001) Hydrology of tropical montane cloud forests: a reassessment. Land Use
Water Resour Res 1:1D1–1D18
Bruijnzeel LA (2004) Hydrological functions of tropical forests: not seeing the soil for the trees?
Agric Ecosyst Environ 104:185–228
Bruijnzeel LA, Hamilton LS (2000) Decision time for cloud forests, IHP humid tropics pro-
gramme series, vol 13. Paris, UNESCO Division of Water Sciences. http://sea.unepwcmc.org/
forest/cloudforest/index.cfm. Accessed Oct 2019
Bruijnzeel LA, Scatena FN, Hamilton LS (2010) Tropical montane cloud forests. Science for con-
servation and management. Cambridge University Press, Cambridge, 342 pp
Bruijnzeel LA, Mulligan M, Scatena F (2011) Hydrometeorology of tropical montane cloud for-
ests: emerging patterns. Hydrol Process 25:465–498
Bruinsma J (2003) World agriculture: towards 2015/2030. A FAO perspective. Earthscan,
London, 432 pp
Burbano-Garcés ML, Figueroa-Casas A, Peña M (2014) Bulk precipitation, throughfall and
stemflow deposition of N-NH4+, N-NH3 and N-NO3 - in an Andean forest. J Trop For Sci
26(4):446–457
Caldwell PV, Minia FC, Elliot KJ, Swank WT, Brantley ST, Laseter SH (2016) Declining water
yield from forested mountain watersheds in response to climate change and forest mesophica-
tion. Glob Chang Biol 22:2997–3012
Cárdenas MF, Tobón C, Buytaert W (2017) Contribution of occult precipitation to the water bal-
ance of páramo ecosystems in the Colombian Andes. Hydrol Process 31:4440–4449
Cavalier J, Goldstein G (1989) Mist and fog interception in Elfin cloud forests in Colombia and
Venezuela. J Trop Ecol 5:309–322
82 C. Tobón
Cavelier J (1996) Environmental factors and ecophysiological processes along altitudinal gradients
in wet tropical mountains. In: Mulkey SS, Chazdon RL, Smith AP (eds) Tropical forest plant
ecophysiology. Chapman and Hall, New York, pp 399–439
Clark K, Torres MA, West A, Hilton R, New M, Horwath A, Fisher J, Rapp J, Robles A, Caceres
A, Malhi Y (2014) The hydrological regime of a forested tropical Andean catchment.
Hydrological. Earth Syst Sci 18:5377–5397
Condit R, Aguilar S, Hernandez A, Perez R, Lao S, Pyke C (2005) Spatial changes in tree com-
position of high-diversity forests: how much is predictable? In: Bermingham E, Dick CW,
Moritz C (eds) Tropical forests: past, present and future. University of Chicago Press, Chicago,
pp 271–294
Coxson DS, Nadkarni NM (1995) Ecological role of epiphytes in nutrient cycles. In: Lowma MD,
Nadkarni NM (eds) Forest canopies. Academic Press, Nueva York
Cuesta F, Peralvo M, Valerazo N (2009) Los bosques montanos de los Andes Tropicales. Una
evaluación regional de su estado de conservación y de su vulnerabilidad a efectos del cambio
climático. Programa Regional ECOBONA-INTERCOOPERATION, 74 pp. www.bosquesan-
dinos.info
Cavelier J, Solis D, & Jaramillo M (1996) Fog interception in montane forests across the cen-
tral cordillera of panama. Journal of Tropical Ecology 12(3):357–369. Retrieved September 3,
2020, from http://www.jstor.org/stable/2560056
CAR - Corporación Autónoma Regional de Cundinamarca (2001). Atlas Ambiental. http://hdl.
handle.net/20.500.11786/36139
DeAngelis CF, McGregor GR, Kidd C (2004) A 3-year climatology of rainfall characteristics over
tropical and subtropical South America based on tropical rainfall measuring mission precipita-
tion radar data. Int J Climatol 24:385–399
Dengel S, Rollenbeck R (2003) Methods of fog quantification in a tropical mountain forest of
southern Ecuador. Fog Newsletter, 15
Dobriyal P, Qureshi A, Badola R, Hussain SA (2012) A review of the methods available for esti-
mating soil moisture and its implications for water resource management. J Hydrol 4:110–117
Dietz J, Hölscher D, Leuschner C, Hendrayanto H (2006) Rainfall partitioning in relation to forest
structure in differently managed montane forest stands in Central Sulawesi, Indonesia. Forest
Ecology and Management 237:170–178. https://doi.org/10.1016/j.foreco.2006.09.044
Eller CE, Burgess SSO, Oliveira RS (2015) Environmental controls in the water use patterns of a
tropical cloud forest tree species: Drimys brasiliensis (Winteraceae). Tree Physiol 35:387–399
Emck P, Moreira-Muñoz A, Richter M (2006) El clima y sus efectos en la vegetación. In: Moraes
M et al (eds) Botánica Económica de los Andes Centrales. Universidad Mayor de San Andrés,
La Paz, pp 11–36
Espinoza JC, Ronchail J, Guyot JL, Filizola N, Noriega L, Ordoñez JJ, Pombosa R, Romero H
(2009) Spatio – temporal rainfall variability in the Amazon Basin countries (Brazil, Peru,
Bolivia, Colombia and Ecuador). Int J Climatol 29:1574–1594
Famiglietti JS, Rudnicki JW, Rodell M (1998) Variability in surface soil moisture content along a
hillslope transect: Rattlesnake Hill, Texas. J Hydrol 210:259–281
Ferwerda W, Hadeed L, McShane T, Rietbergen S, Stolton S, Dudley D (2000) Bosques Nublados
Tropicales Montanos. WWF International/IUCN. The World Conservation Union, 63 p
Fittkau E, Illies J, Klinge H, Schwab G, Sioli H (1968) Biogeography and ecology in South
America. Junk Publishers, The Hague
Fleischbein K, Wilcke W, Goller R, Boy J, Valarezo C, Zech W, Knoblich K (2005) Rainfall inter-
ception in a lower montane forest in Ecuador: effects of canopy properties. Hydrol Process
19:1355–1371
Fleischbein K, Wilcke W, Valarezo C, Zech W, Knoblich K (2006) Water budgets of three small
catchments under montane forest in Ecuador: experimental and modelling approach. Hydrol
Process 20:2491–2507
Fonseca H, Ataroff M (2005) Dinámica Hídrica en la selva nublada de la cuenca alta del Rio
Cusiana y un pastizal de reemplazo, Cordillera Oriental, Colombia. In: Ataroff M, Silva JF
(eds) Dinámica Hídrica en Sistemas Neotropicales. ICAE, Univ. Los Andes, Mérida
Fontúrbel F (2002) Los bosques andinos: reseña biogeográfica y elementos representativos. Long
Rev. 10:12–19. biologia.org
Föster P (2001) The potential negative impacts of global climate change on tropical montane cloud
forests. Earth Sci Rev 55:73–106
Frumau A, Schmid S, Burkard R, Bruijnzeel LA, Tobón C, Calvo J (2010) Fog gauge performance
as a function of wind speed in northern Costa Rica. In: Bruijnzeel LA, Juvik J, Scatena FN,
Hamilton LS, Bubb P (eds) Forests in the mist: science for conservation and management of
tropical montane cloud forests. Cambridge University Press, Cambridge, pp 294–301
Frumau KFA, Burkard R, Schmid S, Bruijnzeel LA, Tobón, C, and Calvo-Alvarado, J (2011) A
comparison of the performance of three types of passive fog gauges under conditions of wind-
driven fog and precipitation. Hydrological Processes 25(3):374–383
Garcia CF (2007) Regulación hídrica bajo tres coberturas vegetales en la cuenca del rio San
Cristóbal, Bogotá D.C. Rev Colomb For 10(20):127–147
Garreaud RD (2009) The Andes climate and weather. Adv Geosci 7(1–9):2009
Gerrits AMJ, Savenije G, Hoffmann L, Pfister L (2007) New technique to measure forest floor inter-
ception – an application in a beech forest in Luxembourg. Hydrol Earth Syst Sci 11:695–701
Gómez-Peralta D, Oberbauer SF, McClain ME, Philippi TE (2008) Rainfall and cloud-water inter-
ception in tropical montane forests in the eastern Andes of Central Peru. For Ecol Manag
255:1315–1325
González J (2000) Monitoring cloud interception in a tropical montane forest of the southwestern
Colombian Andes. Adv Environ Monit Model 1:97–117
Gotsch SG, Asbjornsen H, Holwerda F, Goldsmith GR, Weintraub AE, Dawson TE (2014) Foggy
days and dry nights determine crown-level water balance in a seasonal tropical montane cloud
forest. Plant Cell Environ 37:261–272
Grubb PJ, Whitmore TC (1996) A comparison of montane and lowland rain forest in ecuador II: the
climate and its effects on the distribution and physiognomy of the forests. J Ecol 54(2):303–333
Gentry AH (1991) The distribution and evolution of climbing plants. In: The Biology of Vines (eds
F.E. Putz & H.A. Mooney). Cambridge University Press, Cambridge, pp. 3– 49.
Hafkenscheid RRLJ (2000) Hydrology and biogeochemistry of tropical montane rain forests
of contrasting stature in the Blue Mountains, Jamaica. PhD Dissertation, Vrije Universiteit
Amsterdam, Ámsterdam
Hamilton LS, Juvik JO, Scatena FN (1995) The Puerto Rico tropical cloud forest symposium:
introduction and workshop synthesis. In: Hamilton LS, Juvik JO, Scatena FN (eds) Tropical
montane cloud forests. Springer-Verlag, New York, pp 1–23
Harden C (2006) Human impacts on headwater fluvial systems in the northern and Central Andes.
Geomorphology 79:249–263
Herzog SK, Martínez R, Jørgensen PM, Tiessen H (2011) Climate change and biodiversity in
the Tropical Andes. Inter-American Institute for Global Change Research (IAI) and Scientific
Committee on Problems of the Environment (SCOPE), Brazil
Hincapie I, Germann PF (2009) Impact of initial and boundary conditions on preferential flow. J
Contam Hydrol 104:67–73
Hincapié E, Tobón C (2012) Dinámica del Agua en Andisoles Bajo Condiciones de Ladera.
Revista de la Facultad Agronómica de Medellín, Universidad Nacional de Colombia. Review
of the Facilative National. Agric Medellín 65(2):6771–6783
Hofstede RGM, Wolf JH, Benzing DH (1993) Epiphyte biomass and nutrient status of a Colombian
upper montane rain forest. Selbyana 14:37–45
Holder CD (2004) Rainfall interception and fog precipitation in a tropical montane cloud forest of
Guatemala. For Ecol Manag 190:373–384
Hölscher D, Köhler L, Van Dijk AIJM, Bruijnzeel LA (2004) The importance of epiphytes to total
rainfall interception by a tropical montane rain forest in Costa Rica. J Hydrol 292:308–322
Holwerda F, Bruijnzeel LA, Muñoz-Villers LE, Equihua M, Asbjornsen H (2010) Rainfall and
cloud water interception in mature and secondary lower montane cloud forests of Central
Veracruz, Mexico. J Hydrol 384:84–96
Hursch CR, Brater EF (1941) Separating storm-hydrographs from small drainage-areas into sur-
face- and subsurface-flow. Trans Am Geophys Union Part 3(22):863–871
84 C. Tobón
Ives JD, Messerli B, Spiess E (1997) Mountains of the world: a global priority. In: Messerli B, Ives
JD (eds) Mountains of the world: a global priority. Estados Unidos y Carnforth, Reino Unido,
Parthenon Publishing Group, Nueva York, pp 1–15
IGAC – Instituto Geográfico Agustín Codazzi (2004) Levantamiento de suelos y zonificación de
tierras. Estudio de suelos regionales. Bogotá, D.E. 537 p. http://www2.igac.gov.co/igac_web/
contenidos/plantilla_general_titulo_contenido.jsp?idMenu=129
Jackson RB, Carpenter SR, Dahm CN, McKnight DM, Naiman RJ, Postel SL, Running SW (2001)
Water in a changing world. Ecol Appl 11:1027–1045
Jarvis A (2000) Quantifying the hydrological role of cloud deposition onto epiphytes in a tropical
montane cloud forest, Colombia. http://www.ambiotek.com/herb/hydjar.pdf
Jarvis A, Mulligan M (2010) The climate of tropical montane cloud forests. Hydrol Process
25(3):327–343. https://doi.org/10.1002/hyp.7847
Jones J, Almeida A, Cisneros F, Iroumé A, Jobbágy E, Lara A, Lima WDP, Little C, Llerena C,
Silveira L, Villegas C (2017) Forests and water in South America. Hydrol Process 31:972–980
Jung M, Reichstein M, Ciais P, Seneviratne SI, Sheffield J, Goulden ML, Bonan G, Cescatti A,
Chen J, de Jeu R, Dolman AJ, Eugster W, Gerten D, Gianelle D, Gobron N, Heinke J, Kimball
J, Law BE, Montagnani L, Mu Q, Mueller B, Oleson K, Papale D, Richardson AD, Roupsard
O, Running S, Tomelleri E, Viovy N, Weber U, Williams C, Wood E, Zaehle S, Zhang K
(2010) Recent decline in the global land evapotranspiration trend due to limited moisture sup-
ply. Nature 467:951–954
Juvik JO, Ekern PC (1978) A climatology of mountain fog on Mauna Loa, Hawaiï Island, technical
report, vol 118. Water Resources Research Center, University of Hawaiï, Honolulu
Kaser G, Georges C, Juen I, Moelg T (2005) Low latitude glaciers: unique global climate indicators
and essential contributors to regional fresh water supply. A conceptual approach. Global change
and mountain regions: a state of knowledge overview, advances in global change research, vol
23, Huber U, Bugmann HKM, Reasoner MA. Kluwer Publishers: New York; 185–196
Köhler L, Tobón C, Frumau A, Bruijnzeel L (2007) Biomass and water storage dynamics of epi-
phytes in old-growth and secondary montane cloud forest stands in Costa Rica. Plant Ecol
193(2):171–184
Köhler L, Hölscher D, Bruijnzeel LA, Leuschner C (2011) Epiphyte biomass in Costa Rican old-
growth and secondary montane rain forests and its hydrological significance. In: Tropical mon-
tane cloud forests: science for conservation and management. Cambridge University Press,
Cambridge, pp 268–274. https://doi.org/10.1017/CBO9780511778384.029
Lara M, Rollenbeck R, Fabian P, Bendix J (2003) Relaciones entre precipitación y vegetación
en el bosque tropical de montaña. 2da conferencia. Ecología de bosques tropicales. Loja,
Ecuador, 2002
Lauer W (1981) Ecoclimatological conditions of the Páramo belt in the tropical high mountains.
Mt Res Dev 1:209–221
Lawton RO, Nair US, Pielke RA Sr, Welch RM (2001) Climatic impact of tropical lowland defor-
estation on nearby montane cloud forests. Science 294:584–587
Lazarev YN, Petrov PV, Tartakovsky D (2005) Interface dynamics in randomly heterogeneous
porous media. Adv Water Resour 28:393–403
León Peláez JD, González Hernández MI, Gallardo Lancho JF (2010) Distribución del agua lluvia
en tres bosques altoandinos de la cordillera Central de Antioquia, Colombia. Rev Fac Natl
Agron 63:5319–5336
Liebmann B, Kliadis G, Vera C, Saulo A, Carvalho L (2004) Subseasonal variations of rainfall in
South America in the vicinity of the low-level jet east of the Andes and comparison to those in
the South Atlantic Convergence Zone. J Climatol 17:3829–3842
Lovett GM, Lindberg SE, Richter DD, Johnson DW (1985) The effects of acidic deposition on
cation leaching from three deciduous forest canopies. Can J For Res 15:1055–1060
Martínez R, Ruiz D, Andrade M, Blacutt L, Pabón D, Jaimes E, León G, Villacís M, Quintana J,
Montealegre E, Euscátegui C (2011) Synthesis of the climate of the Tropical Andes. In: Herzog
S, Martinez R, Jørgensen P, Tiessen H (eds) Climate change and biodiversity in the tropical
Andes. Inter-American Institute for Global Change Research (IAI) and Scientific Committee
on Problems of the Environment (SCOPE). https://doi.org/10.13140/2.1.3718.4969
McJannet D, Wallace J, Fitch P, Disher M, Reddell P (2007) Water balance of tropical rainforest
canopies in north Queensland. Aust Hydrol Process 25:3473–3484
Moser G, Röderstein M, Soethe N, Hertel D, Leuschner C (2008) Altitudinal changes in stand
structure and biomass allocation of tropical mountain forests in relation to microclimate and
soil chemistry. In: Beck E et al (eds) Gradients in a tropical mountain ecosystem of Ecuador.
Ecological studies, vol 198. Springer, Berlin, pp 229–242
Mulligan M, Jarvis A (2000) Laboratory simulation of cloud interception by epiphytes and implica-
tion for hydrology of the Tambito experimental cloud Forest, Colombia. J Hydrol 403:853–858
Mulligan M, Jarvis A, González J, Bruijnzeel LA (2010) Using ‘biosensors’ to elucidate rates
and mechanisms of cloud water interception by epiphytes, leaves, and branches in a sheltered
Colombian cloud forest. In: Bruijnzeel LA, Scatena FN, Hamilton LS (eds) Tropical mon-
tane cloud forests. Science for conservation and management. Cambridge University Press,
Muñoz-Villers LE, Holwerda F, Gómez-Cárdenas M, Equihua M, Asbjornsen H, Bruijnzeel LA,
Marín-Castro BE, Tobón C (2012) Water balances of old-growth and regenerating montane
cloud forests in Central Veracruz, Mexico. J Hydrol 462–463:53–66. https://doi.org/10.1016/j.
jhydrol.2011.01.062
Myers N, Mittermeier R, Mittermeier C, da Fonseca G, Kent J (2000) Biodiversity hotspots for
conservation priorities. Nature 403:853–858
Nadkarni NM (1984) Epiphyte biomass and nutrient capital of a Neotropical Elfin Forest.
Nadkarni NM, Wheelwright NT (2000) Ecology and natural history of a tropical montane cloud
forest, Monteverde, Costa Rica. Oxford University Press, New York
Neary DG, Ice GG, Jackson CR (2009) Linkages between forest soils and water quality and quan-
tity. For Ecol Manag 258:2269–2281
Oesker M, Homeier J, Dalitz H (2008) Spatial heterogeneity of throughfall quantity and qual-
ity in tropical montane forests in southern Ecuador. Second international symposium moun-
tains in the mist: science for conserving and managing tropical montane cloud forest. Hawaii
Preparatory Academy (HPA), Waimea (July 27–August 2, 2004)
Oesker M, Homeier J, Dalitz H, Bruijnzeel LA (2010) Spatial heterogeneity of throughfall quan-
tity and quality in tropical montane forests in southern Ecuador. In: Bruijnzeel LA, Scatena FN,
Hamilton LS (eds) Tropical montane cloud forests. Science for conservation and management.
Cambridge University Press, Cambridge, pp 393–401
Pielou EC (1979) Biogeography. Wiley, Hoboken, 351 p
Podwojewski P, Poulenard J, Zambrana T, Hofstede R (2002) Overgrazing effects on vegeta-
tion cover and properties of volcanic ash soil in the páramo of Llangahua and La Esperanza
(Tungurahua, Ecuador). Soil Use Manag 18:45–55
Pounds JA, Fogden MPL, Campbell JH (1999) Biological response to climate change on a tropical
mountain. Nature 398:611–615
Póveda G, Mesa OJ (1997) Feedbacks between hydrological processes in tropical South America
and large-scale ocean–atmospheric phenomena. J Clim 10:2690–2702
Póveda G, Mesa OJ, Salazar LF, Arias PA, Moreno HA, Vieira SC, Agudelo PA, Toro VG, Álvarez
JF (2005) The diurnal cycle of precipitation in the tropical Andes of Columbia. Mon Weather
Rev 133:228–240
Purdy A, Fisher JB, Goulden ML, Colliander A, Halverson G, Tuc K, Famigliettia J (2018) SMAP
soil moisture improves global evapotranspiration. Remote Sens Environ 249:1–14
Rada F (2002) Los bosques andinos: reseña biogeográfica y elementos representativos.
Biologia 10:1–16
Ramírez B, Teuling A, Ganzeveld L, Hegger Z, Leemans R (2017) Tropical montane cloud forests:
hydrometeorological variability in three neighbouring catchments with different forest cover.
J Hydrol 552:151–167
86 C. Tobón
Richardson BA, Richardson MJ, Scatena FN, McDowell WH (2000) Effects of nutrient availabil-
ity and other elevational changes on bromeliad populations and their invertebrate communities
in a humid tropical forest in Puerto Rico. J Trop Ecol 16:167–188
Rittera A, Regalado C (2017) Tree stomata conductance estimates of a wax myrtle-tree heath
(fayal-brezal) cloud forest as affected by fog. Agric For Meteorol 247:116–130
Rodríguez-Iturbe I (2000) Ecohydrology: a hydrologic perspective of climate-soil-vegetation
dynamics. Water Resour Res 36:3–9
Rollenbeck R (2006) Variability of precipitation in the Reserva Biólogica San Francisco, Southern
Ecuador. Lyonia 9:43–51
Rollenbeck R, Bendix J, Fabian P (2008) Spatial and temporal dynamics of atmospheric water and
nutrient inputs in tropical mountain forests of Southern Ecuador. Second international sym-
posium mountains in the mist: science for conserving and managing tropical montane cloud
forest. Hawaii Preparatory Academy (HPA), Waimea. July 27–August 2, 2004
Sarmiento FO (2001) Ecuador. In: Kapelle M, Brown AD (eds) Bosques Nublados del Neotrópico.
Instituto Nacional de Biodiversidad, INBio, Heredia, 698 p
Savenije HHG (2004) The importance of interception and why we should delete the term evapo-
transpiration from our vocabulary. Hydrol Process 18:1507–1511
Schawe M, Gerold G, Bach K, Gradstein SR (2008) Hydrometeorologic patterns in relation to
montane forest types along an elevational gradient in the Yungas, Bolivia. Second international
symposium mountains in the mist: science for conserving and managing tropical montane
cloud forest. Hawaii Preparatory Academy (HPA). Waimea (July 27–August 2, 2004)
Silva B, Álava-Núñez P, Strobl S, Beck E, Bendix J (2017) Area-wide evapotranspiration monitor-
ing at the crown level of a tropical mountain rain forest. Remote Sens Environ 194:219–229
Sklenár P, Ramsay PM (2001) Blackwell Science, Ltd diversity of zonal páramo plant communi-
ties in Ecuador. Divers Distrib 7:113–124
Smith JM, Cleef A (1988) Composition and origins of the world’s tropicalpine floras. J Biogeogr
15:631–645
Smith WK, McClean TM (1989) Adaptive relationship between leaf water repellency, stomatal
distribution, and gas exchange. Am J Bot 76:465–469
Stadtmüller T (1987) Cloud forests in the humid tropics. A bibliographic review. United Nations
University, Tokyo y CATIE, Turrialba
Stadtmüller T (2003) Forests in watershed management as a means to reduce flood risks–the exam-
ple of the PROMIC project. Presentation at the conference on “Multifunctional forestry and
sustainable water management in development cooperation”, 26 February 2003. Berna
Steinhardt U (1979) Studies on the water and nutrient balance of an Andean cloud forest in
Venezuela. Gött Bodenkd Ber 56:1–185
Soil Survey Staff (2004) Official soil series descriptions. Available at http://soils.usda.gov/techni-
cal/classification/osd/index.html (accessed 10 Feb. 2004, verified 17 Dec. 2007). USDA-NRCS.
Tanaka N, Kume T, Yoshifuji N, Tanaka K, Takizawa H, Shiraki K, Tantasirin C, Tangtham N,
Suzuki M (2008) A review of evapotranspiration estimates from tropical forests in Thailand
and adjacent regions. Agric For Ecol 148:807–819
Tobón C (1999) Monitoring and modelling hydrological fluxes in support of nutrient cycling
studies in Amazonian rain forest ecosystems. Tropenbos series 17, Wageningen, the
Netherlands, 169 pp
Tobón C (2009) Los bosques andinos y el agua. Serie Investigación y Sistematización 4. Programa
Regional ECOBONA—INTERCOOPERATION— CONDESAN: Quito, Ecuador, 122 pp
Tobón C, Arroyave F (2007) Hidrología de los bosques alto andinos. In: León Peláez JD (ed)
Ecología de Bosques Andinos, Universidad Nacional de Colombia. Universidad Nacional de
Colombia, Bogota, 213 p
Tobón C, Arroyave F (2008) Hidrología de los bosques Alto Andinos. In: León Peláez JD (ed)
Ecología de Bosques Andinos. Universidad Nacional de Colombia, Bogota, 260 p
Tobón M, Bouten C, Sevink W (2000a) Gross rainfall and its partitioning into throughfall, stem-
flow and evaporative loss in four forest ecosystems in western Amazonia. J Hydrol 237:40–57
Tobón MC, Bouten C, Dekker S (2000b) Forest floor water dynamics and root water uptake in four
forest ecosystems in northwest Amazonian. J Hydrol 237:169–183
Tobón C, Gil G, Villegas C (2008) Aportes de la niebla al balance hídrico de los bosques alto-
andinos. In: León Peláez JD (ed) Ecología de Bosques Andinos. Universidad Nacional de
Colombia, Bogota, 213 p
Tobón C, Köhler L, Schmid S, Burkard R, Frumau KFA, Bruijnzeel LA (2010a) Water dynamics
of epiphytic vegetation in a lower montane cloud forest: interception, storage and evaporation
of horizontal precipitation. In: Bruijnzeel LA, Scatena FN, Hamilton LS (eds) Tropical mon-
tane cloud forests: science for conservation and management. Cambridge University Press,
Tobón C, Bruijnzeel LA, Frumau KFA, Calvo JC (2010b) Changes in soil hydraulic properties and
soil water status after conversion of tropical montane cloud forest to pasture in northern Costa
Rica. In: Bruijnzeel LA, Scatena FN, Hamilton LS (eds) Tropical montane cloud forests: sci-
ence for conservation and management. Cambridge University Press, Cambridge, pp 765–778
USDA - Soil Survey Staff (2014) Keys to soil taxonomy, Twelfth Edition, 2014. United States
Department of Agriculture Natural Resources Conservation Service. USDA-Natural Resources
Conservation Service, Washington, DC. 372 p.
van Dijk IJM, Peña-Arancibia JL, Bruijnzeel LA (2012) Land cover and water yield: infer-
ence problems when comparing catchments with mixed land cover. Hydrol Earth Syst Sci
16:461–3473
Vance ED, Nadkarni NM (1990) Microbial biomass and activity in canopy organic matter and the
forest floor of a tropical cloud forest. Soil Biol Biochem 22:677–684
Vásquez VG (2016) Influencia del uso de la tierra en la respuesta hidrológica de cuencas de cabecera
en los Andes Centrales de Colombia. Disertación doctoral. Tesis, Universidad Nacional de
Colombia, Medellín. 146 pp
Veneklaas EJ, Van Ek R (1990) Rainfall interception in two tropical montane rain forests,
Colombia. Hydrol Process 4:311–326
Veneklaas EJ, Zagt RJ, van Leerdam A, van Ek R, Broekhoven AJ, van Genderen M (1990)
Hydrological properties of the epiphyte mass of a montane tropical rain forest, Colombia.
Vegetatio 89:183–192
Vis M (1986) Interception, drop size distributions and rainfall kinetic energy in four Colombian
forest ecosystems. Earth Surf Process Landf 11:591–570
Vuille M, Bradley RS, Keimig F (2000) Interannual climate variability in the Central Andes and its
relation to tropical Pacific and Atlantic forcing. J Geophys Res 105:12447–12460
Villegas JC, Tobón C, Breshears DD (2008) Fog Interception by non-vascular epiphytes in the trop-
ical cloud forests: dependencies on gauge type and meteorological conditions. Hydrological
Processes 22:2484–2492
Walker R, Ataroff M (2005) Intercepción y drenaje en las epífitas de dosel de una selva Nublada
andina venezolana. In: Ataroff M, Silva JF (eds) Dinámica Hídrica en Sistemas Neotropicales.
ICAE, Univ. Los Andes, Mérida
Walther GR, Post E, Convey P, Menzel A, Parmesan C, Beebee TJC, Fromentin J, Hoegh-Guldberg
O, Bairlein F (2002) Ecological responses to recent climate change. Nature 416:389–339
Wassenaar T, Gerber P, Verburg PH, Rosales M, Ibrahim M, Steinfeld H (2007) Projecting land
use changes in the Neotropics: the geography of pasture expansion into forest. Glob Environ
Chang 17:86–104
Wolf JHD (1993) Diversity patterns and biomass of epiphytic bryophytes and lichens along an
altitudinal gradient in the northern Andes. Ann Mo Bot Gard 80:928–960
Wright C, Kagawa-Viviani A, Gerlein-Safdi C, Mosquera G, Poca M, Tseng H, Chun KP (2017)
Advancing ecohydrology in the changing tropics: perspectives from early career scientists.
Ecohydrology 106(17):e1918. https://doi.org/10.1002/eco.1918
Young KR (2006) Bosques húmedos. In: Moraes R et al (eds) Botánica Económica de los Andes
Centrales. Universidad Mayor de San Andrés, La Paz
Chapter 5
Litterfall in Andean Forests: Quantity,
Composition, and Environmental Drivers
Wolfgang Wilcke
5.1 Introduction
The carbon storage of an ecosystem is driven by the balance between the production
of organic matter and its decomposition. The currently on-going environmental
changes in the Andes including increasing temperatures and locally variable posi-
tive and negative changes of rainfall (Vuille et al. 2003; Urrutia and Vuille 2009;
Peters et al. 2013) and increasing N deposition (Galloway et al. 2004; Wilcke et al.
2013) will likely affect both productivity and decomposition. This together with the
increasing atmospheric CO2 concentrations will possibly change the equilibrium
between organic matter production and decomposition. To assess the direction of
these changes requires the knowledge of the current productivity and decomposi-
tion and their drivers. Net primary production can be estimated with the help of the
comparatively easily measured fine litterfall because it was reported that fine litter-
fall accounted for ca. one-third of total net primary production (Bray and Gorham
1964; Chave et al. 2010). Moreover, litterfall carries nutrients from the forest can-
opy to the soil and is thus also an important part of the nutrient cycling (Vogt et al.
1986; Veneklaas 1991; Wilcke et al. 2002).
Fine litterfall consists of leaves/needles, twigs, reproductive materials (flowers,
fruits, seeds), and other small dead organic material dropping from the forest cano-
pies to the soil and is usually collected with several at least 0.25 m2 large traps. To
estimate decomposition, e.g., in situ litter bag experiments or the ratio of annual fine
litterfall to organic matter storage in the soil organic layer can be used (Vogt
et al. 1986).
The comprehensive early review of Vogt et al. (1986) has shown that worldwide
litterfall correlates negatively with latitude because it is driven by light availability.
W. Wilcke (*)
Institute of Geography and Geoecology (IfGG), Karlsruhe Institute of Technology (KIT),
Karlsruhe, Germany

https://doi.org/10.1007/978-3-030-57344-7_5
90 W. Wilcke
The correlation is closer for broad-leaved than for needle-leaved forests. The N flux
with litterfall in broad-leaved forests (but not in needle-leaved forests) is also nega-
tively correlated with latitude. The N and P fluxes with litterfall are positively cor-
related in broad-leaved but not in needle-leaved forests. The N concentrations in
litterfall decrease from the equator to the pole. The N concentration in litterfall is
positively correlated with the mass of litterfall, while there is no consistent relation-
ship between P concentration in litterfall and mass of litterfall. This suggests a
higher importance of N than of P availability for litterfall (Vogt et al. 1986).
Several local studies in the humid forests on the Eastern Cordillera in Ecuador
and Peru, which forms the rim of the Amazon basin, have found that litterfall
changes with topographical elevation. Girardin et al. (2010) observed along an ele-
vational transect from 194 to 3025 m a.s.l. in Peru decreasing litterfall in the order
lowland > premontane > upper montane > lower montane forests, but the difference
between upper and lower montane forests was small (on average 165 vs.
144 g m−2 year−1 C, while the lowland forest produced 510 g m−2 year−1 C). Pinos
et al. (2017) determined a surprisingly high leaf litterfall in an upper montane
Polylepis reticulata Hieron. forest near the treeline at 3735–3930 m a.s.l. in the
Cajas National Park in Ecuador of 279 to 465 g m−2 year−1. Röderstein et al. (2005)
reported that leaf litter mass decreased to less than a third (from 862 to
263 g m−2 year−1) with increasing elevation from 1890 m to 3060 m a.s.l. in south
Ecuador. Similarly, Wilcke et al. (2008) observed that forest productivity reflected
by forest stature, tree basal area, and tree growth decreased from 1960 to 2450 m
a.s.l. Wolf et al. (2011) found a maximum of the litterfall at ca. 2000 m a.s.l. along
an elevational gradient from 990 to 3000 m a.s.l. again in south Ecuador. In their
study, the forests at ca. 2000 m a.s.l. showed a high productivity at the upper end of
the range of published litterfall in tropical forests. A similar high litterfall of
850–970 g m−2 year−1 was reported by Wilcke et al. (2002) for five study sites
between 1900 and 2010 m a.s.l. Wolf et al. (2011) furthermore observed that litter-
fall in the North Andes in south Ecuador decreased from lower slope to upper slope
positions at the scale of a few 10s of meters.
Compared with tropical lowland forests and old-world temperate forests, the
montane forests of the Andes have been up to now much less studied. Of the 81
study sites included in the comprehensive review of the litterfall in tropical South
America (including Panama) of Chave et al. (2010) only five were located in the
Andes. The knowledge about litterfall in the southern beech forests (Nothofagus
sp.) of the south Andes seems even more limited, because I only found a single
study from the Chilean Andes including four differently managed and unmanaged
stands in the international journal literature (Staelens et al. 2011), although there is
some more information about the productivity of Nothofagus sp. forests from the
south Chilean island of Chiloë outside the Andes (Pérez et al. 1998, 2003).
In this review, I focus on fine litterfall, its fractions (leaves/needles, twigs, repro-
ductive materials, and miscellaneous), its most important macronutrient concentra-
tions N, P, and K, and its C:nutrient ratios along with soil and forest stand properties.
My objectives are (1) to compile the published data about litterfall and its fractions
in the Andean forests accessible via the Web of Science, (2) to analyze the
5 Litterfall in Andean Forests: Quantity, Composition, and Environmental Drivers 91
relationship of geographic location (latitude, longitude, elevation) and climate (pre-

cipitation, temperature) with litterfall quantity and quality, and (3) to investigate the
role of soil, forest stand, and chemical litter quality for litterfall mass.
5.2 Study Sites and Methods
Using the Web of Science in September 2019, I found published reports of litterfall
with varying additional information on soil and litter properties from 44 forest
stands, which I combined with my own group’s results from another 12 sites in a
lower montane forest in south Ecuador (Table 5.1). The reviewed studies used at
least five samplers per studied forest plot of varying size with a minimum surface
area of 0.09 m2. In some work, only leaf litterfall was reported (León et al. 2011;
Pinos et al. 2017), which I extrapolated to total fine litterfall assuming that leaf lit-
terfall accounted for 70% of total litterfall.
The study area in south Ecuador, from which I included results of 12 study sites
which are unpublished except for the first measurement year of five of these sites
(Wilcke et al. 2002), is located on the eastern slope of the “Cordillera Real,” the
Eastern Andean Cordillera in south Ecuador facing the Amazon basin at 4° 00‘S and
79° 05‘W. From 1998 to 2012, 12 measurement sites were run in native old-growth
tropical montane rain forest between 1900 and 2130 m a.s.l. for varying durations.
One measurement site was located at each of the microcatchments (MC) 1 and 3 at
1900 m a.s.l. (years of 1998–2003), three in MC 2 at 1900, 1950, and 2000 m a.s.l.
(1998–2007) (Wilcke et al. 2002), three in MC 5 at 2050–2090 m a.s.l. (Plots F, J,
and M, 2004–2009, Wilcke et al. 2009), and four in the Nutrient Management
Experiment (NUMEX) between 2060 and 2130 m a.s.l. (the unfertilized control
plots of NUMEX, 2007–2009, Homeier et al. 2012).
From 1999 to 2010, annual incident precipitation at the Ecuadorian study site
ranged 1900–2700 mm. June was the wettest month with approx. 310 mm and
November the driest with approx. 130 mm. Mean annual precipitation is nearly
constant between 1950 and 2270 m a.s.l. (Bendix et al. 2008). The mean annual
temperature at 1950 m a.s.l. between 1999 and 2010 was 14.9 °C. The coldest
month was July, with a mean temperature of 13.7 °C, the warmest November with
a mean temperature of 15.8 °C. The average gradient of air temperature between
1950 m a.s.l. and 3180 m a.s.l. in the study area is 0.61 °C/100 m−1 (Bendix et al.
2008). Thus, the study area covered an altitudinal gradient of 1.4 °C but received
similar annual precipitation.
Recent soils have mainly developed from surface sediments on steep slopes
(Wilcke et al. 2001, 2003). The underlying bedrock consists of interbedding of
paleozoic phyllites, quartzites, and metasandstones but in the study area phyllites
dominate. Dominating soils included Humic Eutrudepts, Humic Dystrudepts,
Oxyaquic Eutrudepts, and Histic Humaquepts (Soil Survey Staff 2014). All soils are
shallow, loamy-skeletal with high mica contents. The organic layer consisted of Oi,
Table 5.1 Overview of the collected literature results about location, rock and forest type, litterfall, and litterfall components ordered from N to S
92
Mean Reproductive
Mean annual annual Leaves materials
Latitude Longitude Elevation precipitation temperature Forest Litterfall (L) (R) Twigs Rest
Country [°] [m a.s.l.] [mm] [°C] Rock type typea [g m−2 year−1] [%] R/L [%] Source
1 Colombia 10.739 −72.833 267 777 27.3 DRY 208 66.6 15.9 0.24 11.1 6.4 Fuentes-
Molina et al.
(2018)
2 Venezuela 8.617 −71.350 2350 1500 12.6 Schist/ UMF 697 48.5 15.6 0.32 32.6 3.3 Steinhardt
sandstone (1979)
3 Venezuela 8.000 −72.000 2550 2500 13 Gneiss UMF 430 Tanner et al.
(1992)
4 Colombia 6.760 −75.108 1200 2078 23 Granite PRE 905 Sierra et al.
(2007)
5 Colombia 6.541 −75.802 542 1594 26.6 Amphibolite PLA 56 65.0 24.0 0.37 9.0 2.0 Flórez-Flórez
et al. (2013)
6 Colombia 6.300 −75.500 2490 1948 14.9 Amphibolite UMF 748b 71.0 14.3 14.7 León et al.
(2011) and
Ramírez
et al. (2014)
7 Colombia 6.300 −75.500 2490 1948 14.9 Amphibolite PLA 777b 63.0 12.5 24.5 León et al.
(2011) and
Ramírez
et al. (2014)
8 Colombia 6.300 −75.500 2490 1948 14.9 Amphibolite PLA 349b 71.0 20.6 8.4 León et al.
(2011) and
Ramírez
et al. (2014)
9 Colombia 5.000 −75.000 2550 2115 12.2 Volcanic ash UMF 703 65.6 9.4 0.14 15.1 9.9 Veneklaas
(1991)
10 Colombia 5.000 −75.000 3370 1453 7.7 Volcanic ash UMF 431 65.4 6.3 0.10 17.6 10.7 Veneklaas
(1991)
W. Wilcke
Mean Reproductive
11 Ecuador −2.773 −79.220 3833 876 5.44 Volcanic ash UMF 596b Pinos et al.
(2017)
(2017)
(2017)
(2017)
(2017)
(2017)
17 Ecuador −3.967 −79.067 1950 1950 15.7 Schist/ LMF 950 Wolf et al.
sandstone (2011)
sandstone (2011)
sandstone (2011)
20 Ecuador −3.975 −79.074 1900 2500 15 Schist/ LMF 1084 This study
sandstone
sandstone
5 Litterfall in Andean Forests: Quantity, Composition, and Environmental Drivers
sandstone
sandstone
93
(continued)
94
Mean Reproductive
sandstone
sandstone
sandstone
sandstone
sandstone
sandstone
sandstone
sandstone
32 Ecuador −3.976 −79.085 1890 1950 15.7 Schist/ LMF 834 59.5 4.3 0.07 10.6 25.7 Moser et al.
sandstone (2011) and
Leuschner
et al. (2013)
33 Ecuador −3.989 −79.082 2380 5000 13.2 Schist/ UMF 365 72.3 3.0 0.04 10.7 14.0 Moser et al.
sandstone (2011) and
Leuschner
et al. (2013)
34 Ecuador −4.100 −78.967 1095 2230 19.4 Granodiorite PRE 700 Wolf et al.
(2011)
W. Wilcke
Mean Reproductive
(2011)
(2011)
37 Ecuador −4.100 −79.183 2900 4500 9.4 Schist/ UMF 470 Wolf et al.
sandstone 2011
sandstone (2011)
sandstone (2011)
40 Ecuador −4.112 −78.972 1540 2300 17.5 Granodiorite PRE 734 68.9 5.9 0.08 12.7 12.5 Moser et al.
(2011) and
Leuschner
et al. (2013)
41 Ecuador −4.115 −78.967 1050 2230 19.4 Granodiorite PRE 778 64.9 11.7 0.18 14.5 8.9 Moser et al.
(2011);
Leuschner
et al. (2013)
42 Ecuador −4.120 −79.183 3060 4500 9.4 Schist/ UMF 270 66.3 2.6 0.04 14.4 16.7 Moser et al.
sandstone (2011);
Leuschner
et al. (2013)
43 Peru −12.959 −71.566 1000 3087 20.7 Alluvial PRE 476a Girardin
sediments et al. (2010)
and
5 Litterfall in Andean Forests: Quantity, Composition, and Environmental Drivers
Zimmermann
et al. (2009)
(continued)
95
Mean Reproductive
96
44 Peru −13.049 −71.537 1500 2631 18.8 Granite PRE 528a Girardin
et al. (2010)
and
Zimmermann
et al. (2009)
45 Peru −13.071 −71.558 1855 2472 18 Granite LMF 275a Girardin
et al. (2010)
46 Peru −13.073 −71.559 2020 1827 17.4 Schist LMF 275a Girardin
et al. (2010)
47 Peru −13.075 −71.589 2720 2318 13.5 Schist UMF 350a Girardin
et al. (2010)
48 Peru −13.109 −71.600 3020 1776 11.8 Schist UMF 227a Girardin
et al. (2010)
49 Peru −13.190 −71.587 3025 1706 12.5 Schist UMF 367 69.0 9.0 0.13 18.0 3.9 Girardin
et al. (2010,
2014) and
Zimmermann
et al. (2009)
50 Peru −13.190 −71.597 2825 1560 13.1 Schist UMF 269 73.0 10.0 0.14 16.0 0.9 Girardin
et al. (2010)
51 Chile −39.583 −73.117 805 4500 9 Volcanic ash SBE 476 66.4 11.1 0.17 21.5 1.0 Staelens
et al. (2011)
et al. (2011)
et al. (2011)
et al. (2011)
a
Litterfall estimated from reported C fluxes with litterfall by assuming a factor of 1.92 taken from the comparison of the report of the C flux with litterfall in
Girardin et al. (2010) and the litterfall in Girardin et al. (2014) of one identical site
W. Wilcke
b
DRY is dry forest, PLA is plantation, PRE is tropical premontane rain forest, LMF is tropical lower montane forests, UMF is tropical upper montane forest,
SBE is southern beech (Nothofagus sp.) forest
c
Litterfall estimated from reported leaf litterfall by assuming a leaf litter contribution of 70%
Oe, and frequently also Oa horizons. The thickness increased with increasing eleva-
tion giving Histosols (mainly Terric Haplosaprists) above ca. 2100 m.
The study forest is a Lower Montane Rain Forest (Bruijnzeel and Hamilton
2000). All measurement sites were below old-growth forest. The forest was nearly
undisturbed except for MC 5 where in 2004 a slight forest intervention removing
10.2% of the initial basal area was conducted (Günter et al. 2008; Wilcke et al.
2009). More information on the composition of the forest can be found in the work
of Homeier (2004).
At each measurement station, litterfall was collected with three to five 0.09
(April 1998–October 2005) or 0.25 m2 large traps (after October 2005). I analyzed
all available monthly litterfall samples from MC1, MC3, and NUMEX, while the
data set for MC2 covered April 1998–March 2005 and September 2009–August
2010 and that for MC5 June 2008–May 2009. The thickness of the organic layer
was measured on the wall of a soil pit. The lower end of the organic layer was
located at the position in the soil profile where the bulk density increased abruptly
from >0.3 g cm−3 to >1 g cm−3.
Mineral topsoil pH was measured in water (soil:solution ratio 1:2.5 v/v for the
mineral soil and 1:10 for the organic horizons) with a glass electrode (Orion
U402-S7, Thermo Fisher Scientific, Waltham, MA, USA), total C and N concentra-
tions in soil and litterfall with an Elemental Analyzer (vario EL, Elementar
Analysensysteme, Hanau, Germany) on finely ground samples, and effective cation-
exchange capacity (ECEC) by extraction with 1 M NH4NO3. To determine P, K, Ca,
and Mg concentrations in litterfall, finely ground samples were digested with con-
centrated HNO3 under pressure (Heinrichs et al. 1986). The concentrations of Al, K,
Ca, Mg, and Na were determined with flame atomic absorption spectrometry (AAS;
Varian SpectrAA 400) and those of P with inductively-coupled plasma atomic-
emission spectrometry (ICP-AES; GBC Integra XMP). Base saturation (BS) was
calculated by summing up the proportion of charge equivalents of extractable Ca, K,
Mg, and Na and dividing the sum by the ECEC.
5.3 Results
While there is already an ample literature about the litterfall in the tropical forests
of the North Andes, the southern part of the Andes (Bolivia, Chile, Argentina) is
hardly covered (Table 5.1). I only found reports from four Andean sites under south-
ern beech (Nothofagus sp.). The most intensively studied country is Ecuador with
32 sites. Moreover, I only found a single report about litterfall in the dry forests
which dominate the inner Andean depressions and the tropical part of the Western
Cordillera (Table 5.1; Fuentes-Molina et al. 2018). Most of the available publica-
tions referred to old-growth native rain and cloud forests except three which included
tree plantations in the rain and cloud forest area (all monocultures; León et al. 2011;
Flórez-Flórez et al. 2013).
98 W. Wilcke
1400
Old-growth
1200 Plantations
Means
Litterfall [g m-2 yr-1]
1000
800 LMF
UMF
600 PRE
400
SBE
200
0
500 1000 1500 2000 2500 3000 3500 4000
Elevation [m a.s.l.]
Fig. 5.1 Relationship between elevation and litterfall in 50 old-growth native rain forest stands
(open circles) and three 5–43-year-old tree plantations (open triangles); a 5-year-old Azadirachta
indica A.Juss., a 43-year-old Pinus patula Schltdl. & Cham. and a 43-year-old Cupressus lusitan-
ica Mill. plantation in the Andes (for the data sources see Table 1). Black rectangles illustrate mean
litterfall of southern beech (SBE), premontane (PRE), lower montane (LMF), and upper montane
forests (UMF)
Litterfall of the old-growth native rain forests increased from higher and lower
latitudes to the equator and was independent from longitudinal location (Table 5.1).
In these old-growth native rain and cloud forests, litterfall showed a hump-shaped
elevational distribution with maximal productivity around 2000 m a.s.l. (Fig. 5.1).
The litterfall of the two 43-year-old tree plantations (León et al. 2011) fell well
within the range of litterfall of the old-growth forests, while the 5-year-old tree
plantation (Flórez-Flórez et al. 2013) showed the lowest litterfall of the whole data
set (triangles in Fig. 5.1). When grouped into forest types, mean litterfall increased
from premontane (1000–1540 m a.s.l., n = 8) to lower montane forests (including
lower montane rain and cloud forests; Bruijnzeel and Hamilton 2000; 1855–2130 m
a.s.l., n = 18) and decreased to upper montane forests (including upper montane
cloud and stunted subalpine forests/elfin forests; Bruijnzeel and Hamilton 2000;
Bruijnzeel et al. 2011; 2350–3950 a.s.l., n = 20). The given elevational range refers
to the particular study sites included in this review and should not be considered as
general elevation ranges of the considered forest types, particularly because these
elevational ranges can locally vary. The temperate southern beech forests (630–805,
n = 4) showed a similar productivity as the tropical upper montane forests (Fig. 5.1).
At the same time the variation in litterfall was highest in the lower montane forest
ranging from 270 to 1280 g m−2 year−1 with a coefficient of variation of 42% as
compared to 21%, 32%, and 6% for premontane, upper montane, and southern
beech forests, respectively. The four southern beech forests showed the smallest
variation in litterfall in spite of the fact that these stands included variable forest
properties such as evergreen and deciduous and native, modified, and managed
stands (Staelens et al. 2011).
Litterfall consisted of 68 ± s.d. 7.2% leaves, 17 ± 5.6% woody material,
7.6 ± 3.5% reproductive material, and 8.3 ± 7.3% unidentifiable rest (n = 15;
Table 5.1). The composition of the litterfall did not show a discernible spatial pat-
tern, which was also true for the ratio of reproductive material to leaves in litterfall
(0.12 ± 0.07). However, the 5-year-old tree plantation produced with 24% a higher
contribution of reproductive materials than the old-growth native rain forests, fol-
lowed with 16% by the dry forest (Table 5.1). There were no data on the contribu-
tion of reproductive materials to litterfall for the other two 43-year-old tree
plantations. The mineral topsoils were consistently acidic (Table 5.2). The pH of the
mineral topsoil did not change systematically with elevation and was not related
with litterfall. The mineral topsoil of the 5-year-old tree plantation was the only one
with a slightly alkaline pH of 7.3 (Flórez-Flórez et al. 2013). There was also no
relationship between clay concentration, effective cation-exchange capacity, base
saturation, and N concentration of the mineral topsoil with elevation or litterfall.
Organic layer thickness correlated positively with elevation (r = 0.57, p < 0.001,
n = 34) and litterfall correlated negatively with organic layer thickness (Fig. 5.2) in
the tropical old-growth native rain and cloud forests. There were no data of organic
layer thickness in the dry and southern beech forests and in the plantations. The C/N
ratios of the mineral soil correlated negatively with litterfall (r = 0.70, p < 0.001,
n = 23).
Mean stem density was highest in the upper montane forest and similar in the
three other forest types (Table 5.3). Mean basal area and mean tree height were
highest in southern beech forests. In the tropical rain forests, mean basal area was
similar and mean tree height decreased from premontane to lower montane forests
but did not change with further increasing elevation. Leaf area index (LAI) increased
from premontane to lower montane and decreased to upper montane forests in line
with litterfall although LAI did not correlate with litterfall, when all 18 individual
study sites, for which LAI data were available, were considered. The study of lit-
terfall in the southern beech forests did not report LAI (Staelens et al. 2011). There
was also no relationship between the number of stems with a dbh > 10 cm ha−1,
basal area, and mean tree height with litterfall.
The N and P concentrations in litterfall tended to be lowest in the upper montane
< southern beech < premontane ≈ lower montane forests (Table 5.3). The K concen-
trations were lowest in the southern beech forests < upper montane < lower mon-
tane, while there were no K data for premontane forests. There was no relationship
between the N, P, and K concentrations in litterfall of the old-growth rain and cloud
forests and elevation. However, C/N (n = 25) and C/P ratios decreased marginally
significantly (p < 0.1) with increasing elevation, while N/P ratios (p = 0.437) did not
change (Fig. 5.3). There was also no relationship between elevation and N/K and
P/K ratios. There were significant positive correlations of N (r = 0.82, p < 0.001,
100
Table 5.2 Mean soil properties (standard deviations; number of measurements) and pH ranges (number of measurements) of the four old-growth native rain
forest types. For the data sources see Table 5.1
Cation-
Organic layer exchange Base
thickness Sand Silt Clay capacity saturation C N P C/N concentration
a −1
Forest type [mm] pH in H2O [%] [nmolc g−1] [%] concentration [mg g ] ratio
Premontane 81 3.9–5.1 49 41 19 No data 43 180 10 0.70 18
(76; 7) (7) (19; 5) (17; 5) (16; 5) (35; 3) (100; 2) (4; 2) (0.10; 3) (2.9; 5)
Lower montane 18 3.6–5.3 31 52 15 66 31 29 2.4 0.62 16
(110; 18) (13) (6.6; 7) (4.9; 7) (3.4; 7) (42; 9) (36; 12) (19; 12) (1.7; 9) (0.19; 8) (2.0; 12)
Upper montane 270 3.3–6.3 25 57 17 148 22 84 4.9 0.22 18
(140; 9) (11) (11; 4) (10; 4) (2.9; 4) (−; 1) (−; 1) (55; 3) (2.7; 3) (0.20; 5) (3.7; 6)
Southern beech No data 4.7–5.4 No data No data No No data No data 140 No data No data No data
(4) data (190; 4)
a
pH measurements in CaCl2 were converted to pH in H2O by multiplying with 1.15 based on a regression of pH(H2O) on pH(CaCl2) (r = 0.94, p = 0.019, n = 5)
W. Wilcke
500
y = -0.24x + 340
Organic layer thickness [mm]
450
r = 0.59
400 p < 0.001
350
300
250
200
150
100
50
0
0 200 400 600 800 1000 1200 1400
Litterfall [g m-2 yr-1]
Fig. 5.2 Relationship between litterfall and organic layer thickness in the old-growth native rain
and cloud forests (n = 34)
n = 30), P (r = 0.65, p < 0.001, n = 30), and K concentrations (r = 0.75, p < 0.001,

n = 20) in litterfall with annual litterfall mass.
As a consequence of the small variation in nutrient concentrations, the fluxes of
N, P, and K with litterfall in the tropical rain forests followed the same elevational
distribution as litterfall, i.e., they showed a hump-shaped elevational distribution
(Fig. 5.1; Table 5.3). The N and P fluxes with litterfall in the southern beech forests
were again similar to those in the upper montane forests, but the K fluxes were con-
siderably lower. The fluxes of N with litterfall correlated closely with those of P and
K (Fig. 5.4).
5.4 Discussion
5.4.1 Litterfall and Litterfall Fractions in Andean Forests
Based on the available data, it is hardly possible to draw a consistent and compre-
hensive picture of litterfall in the Andes, because there are large regions that are
almost not covered by data including Bolivia, Chile, and Argentina (Table 5.1).
Moreover, I only found one reference reporting litterfall in dry forests (Fuentes-
Molina et al. 2018), which is with ca. 200 g m−2 year−1 lower than in all other forest
types. Dry forests cover large parts of the Pacific-exposed slope and the inner
Andean depressions in the tropical part of the Andes. The Tumbesian dry forest in
Ecuador and Peru as an example, which has been comparatively intensively studied,
covers 87,000 km2 (Espinosa et al. 2011). Therefore, the productivity of the Andean
102
Table 5.3 Mean forest stand properties, chemical litter quality, and nutrient fluxes (standard deviations; number of measurements) of the four old-growth
native rain and cloud forest types. For the data sources see Table 5.1
Number of Mean
trees >10 cm tree Leaf area
dbh Basal area height index C N P K N P K
Forest type [ha−1] [m2 ha−1] [m] [m2 m−2] Concentration [mg g−1] C/N C/P N/P Flux [g m−2 year−1]
Premontane 1080 36.7 26.8 5.70 530 15.9 0.77 No 33.4 695 39.5 11.7 0.57 No
(624; 5) (14.6; 5) (7.14; 2) (0.42; 2) (159; 3) (0.60; (0.06; data (2.6; (76.2; (1.90; (1.94; (0.12; data
3) 3) 3) 3) 3) 3) 3)
Lower 1010 32.7 15.4 6.9 468 15.3 0.85 6.23 33.3 725 39.8 14.4 0.82 6.11
montane (418; 14) (9.34; 14) (2.99; 10) (0.79; 6) (73.8, 15) (4.08; (0.45; (2.63; (11.6; (387; (13.6; (7.08; (0.55; (3.78;
15) 15) 12) 15) 15) 15) 15) 15) 12)
Upper 2730 41.0 16.8 3.70 543 9.86 0.50 4.37 59.1 1360 43.3 5.54 0.28 2.89
montane (2560; 15) (16.5; 15) (5.78; 8) (1.19; (14.9; 3) (1.84; (0.19; (3.04; (17.3; (484; (4.73; (1.83; (0.15; (2.27;
10) 8) 8) 4) 3) 3) 3) 8) 8) 4)
Southern 1180 75.7 34.5 No data 448 12.2 0.61 2.84 36.6 744 38.7 5.93 0.29 1.45
beech (677; 4) (38.0; 4) (9.53; 4) (18.4;4) (0.08; (0.07; (2.35; (1.43; (68.6; (3.96; (1.02; (0.06; (1.22;
4) 4) 4) 4) 4) 4) 4) 4) 4)
W. Wilcke
Fig. 5.3 Relationship between elevation and (a) C/N, (b) C/P, and (c) N/P concentration ratios in
litterfall of old-growth rain and cloud forests (n = 25 for C/N and C/P and n = 27 for N/P)
dry forests and their response to changing environmental conditions cannot be

neglected when considering the C budget and the mountain risks of the Andes,
which depend on the forest cover. Consequently, more work about the productivity
of the Andean dry forests is needed.
The three studies on litterfall in plantations (Léon et al. 2011; Flórez-Flórez et al.
2013) suggest that the plantations produce considerably less litterfall in the first few
104 W. Wilcke
12
P or K flux with litterfall [g m-2 yr-1]
P K
10
8
y = 0.50x - 0.84
r = 0.943
p < 0.001
6
n = 20
y = 0.062x - 0.080
4
r = 0.872
p < 0.001
2
n = 30
0
2 4 6 8 10 12 14 16 18 20 22
N flux with litterfall [g m-2 yr-1]
Fig. 5.4 Relationship between the fluxes of N in litterfall and those of P and K in litterfall
years but can reach a similar productivity as the native forests in a few decades. This
might indicate that the afforestation of the wide-spread abandoned low-fertility pas-
tures is a promising way to restore the C stocks and budget of the native forest and
to reduce mountain risks, including enhanced erosion and less buffered water levels
of the headwater streams. However, currently most of the plantations are monocul-
tures of exotic species including those of the genera Pinus, Cupressus, and
Eucalyptus. There have only been few attempts to establish plantations of native
species, mostly because of limited silvicultural knowledge of these species (Weber
et al. 2008). To avoid ecological problems associated with the establishment of
monocultures of exotic tree species, more research into the silvicultural manage-
ment and the economic value chains of native tree species with a potential to be
grown in plantations, possibly even in mixed-species stands is needed, if ecosystem
C storage and biomass productivity on degraded lands are to be restored in a sus-
tainable way.
The premontane forests had on average a similar litterfall as the Andean premon-
tane forest included in the early evaluation of Vitousek (1984) of 780 g m−2 year−1
and the range of the lower montane forests was with 275–1285 g m−2 year−1 slightly
wider than spanned by the two Andean lower montane forests in Vitousek (1984)
with 780–1160 g m−2 year−1 (Table 5.1; Fig. 5.1). The Andean forests with the high-
est productivity, the tropical lower montane forests of the Eastern Cordillera of the
North Andes, showed an only slightly lower mean productivity as the old-growth
tropical lowland forests in South America (Table 5.1; 861 ± 191 g m−2 year−1; Chave
et al. 2010). The Andean southern beech forests had a higher litterfall than a broad-
leaved Nothofagus-dominated and a coniferous Fitzroya-dominated forest of the
Chilean island of Chiloë with 304 ± 62 and 174 ± 69 g m−2 year−1 of litterfall,
respectively (Pérez et al. 1998, 2003).
The mean contribution of leaf litterfall to total litterfall in the data set evaluated
here is close to the mean value of 70.8 ± 8.5% of the data set from tropical South
America of Chave et al. (2010) suggesting that the contribution of leaf litterfall to
total litterfall varies little in old-growth forests (Table 5.1). The same seems true for
the contribution of reproductive materials to total litterfall for which Chave et al.
(2010) reported 8.9 ± 5.6%. Consequently, the mean R/L ratio of 0.118 ± 0.071 in
my data set was also similar to that in the forests of tropical South America studied
by Chave et al. (2010) with a mean R/L ratio of 0.135 ± 0.119. However, Chave
et al. (2010) pointed at the fact that this R/L ratio may be an underestimation,
because concerted fruiting during mast years might not be appropriately included in
this measure, which is also the case in my data set.
5.4.2 R
elationship of Geographical Location and Climate
with Litterfall
The litterfall increased towards the equator as already reported by Bray and Gorham
(1964) and Vogt et al. (1986) because Jordan (1971) had shown that light availabil-
ity during the growing season, which increases from the higher latitudes to the equa-
tor, was strongly correlated with litterfall (Table 5.1). Vogt et al. (1986) found that
particularly broad-leaved forests like most of the here included forests showed a
close correlation with latitude, while needle-leaved forests correlated less strongly
with latitude. There was no relationship between longitude and litterfall likely
owing to the small longitudinal spread of the Andes and possibly also because lon-
gitude is confounded with elevation. The hump-shaped elevational distribution of
litterfall in the humid tropical part of the Andes with a maximum around 2000 m a.s.l.
was also reported by Wolf et al. (2011) from Ecuador (Fig. 5.1). The data of the
study of Wolf et al. (2011) is included here (Table 5.1).
The finding that there was no relationship between rainfall and litterfall was also
reported in the review of Chave et al. (2010) covering tropical South America,
which included five tropical montane forest sites, the data of which are also included
here. Although increasing temperature generally stimulates biological activity and
thus should increase litterfall as was observed by Vogt et al. (1986) for a global data
set, temperature and litterfall were unrelated in the Andes. This finding might be
attributable to the fact that in a mountainous area temperature and precipitation
interact to influence nutrient supply and water stress in a complex way.
106 W. Wilcke
5.4.3 R
ole of Soil, Forest Stand, and Chemical Litter Quality
for Litterfall Mass
The lacking relationship between the pH of the mineral topsoil and litterfall can be
attributed to the small variation in pH values, which fell mostly in the aluminum
oxide buffer range (Table 5.2), except for the soil under the young plantation.
Similarly, Schawe et al. (2007) observed no consistent change in soil pH along an
elevational gradient in Peru from 1700 to 3300 m a.s.l. In contrast, Wilcke et al.
(2008) reported that pH of A and B horizons decreased with increasing elevation
between 1960 and 2450 m a.s.l. in Ecuador. The variation in other soil properties
such as texture, cation-exchange capacity, and base saturation, for which again no
relationship with litterfall was detected, was similarly small in the included studies.
Organic layer thickness correlated negatively with litterfall although increasing
litterfall provides more substrate for the formation of an organic layer on top of the
mineral soil. This illustrates that other factors than litterfall are more important for
the thickness of the organic layer (Fig. 5.2) including other organic matter sources
such as root litter and abiotic conditions (climate, nutrient supply). Vogt et al. (1986)
reported in their review that there was no general relationship between litterfall and
organic layer thickness and attributed this mainly to the role of root litter. Leuschner
et al. (2013) have shown that total belowground biomass increased with increasing
elevation under humid tropical forest in Ecuador from 1050 to 3060 m a.s.l., while
aboveground biomass decreased. For a partly overlapping transect from 1960 to
2450 m a.s.l. in Ecuador, Wilcke et al. (2008) observed that the nutrient concentra-
tions in O and A horizons decreased with increasing elevation resulting in decreas-
ing nutrient supply which hampers organic matter degradation and results in
increasing organic layer thickness.
The finding that none of the forest stand properties listed in Table 5.3 correlated
with litterfall is unexpected. In the literature, it was, for instance, shown for several
coniferous forest plantations that the tree basal area correlated with litterfall (Bueis
et al. 2017; Erkan et al. 2018). Apparently, the static forest stand properties do not
sufficiently reflect the dynamics of the litterfall in the structurally and species-rich
old-growth Andean forests.
The finding that the C/N and C/P ratios in litterfall increased with increasing
elevation (Fig. 5.3; Table 5.3) confirms an increasing nutrient-use efficiency already
reported in Vitousek (1984). However, the fact that no ratio of the nutrient elements
(i.e., N/P, N/K, and P/K) showed a systematic change with elevation suggests that
there is no general change in the kind of nutrient limitation along the considered
elevational range, although it must be borne in mind that all three major nutrients
can be retranslocated to variable degrees during leaf senescence (Marschner 2012).
In the literature, contrasting results with respect to the relationship between eleva-
tion and nutrient limitation have been reported. The classical view based on the
work of Vitousek (1984) is that lowland forests are mainly P and montane forests
mainly N-limited. Fisher et al. (2013), however, reported an N + P limitation of
lowland forests in Peru, while the Andean forests were more N-limited and
particularly so with increasing elevation. Similarly, Tanner et al. (1992) found a
positive response of litterfall to N addition in a Venezuelan tropical montane forest
confirming the classical view. In contrast, Homeier et al. (2012, 2013) reported a
significant increase in litterfall after low-level addition of N + P at 2000 m a.s.l. in
an Ecuadorian lower montane rain forest after only one year and also found the
strongest response of litterfall to fertilization with N + P at 3000 m.s.l., while lit-
terfall in the premontane forest at 1000 m a.s.l. hardly responded to any nutrient
addition. Thus, the kind of nutrient limitation seems to vary among different loca-
tions and as a consequence no general kind of nutrient limitation appears in the
overall data set. The lack of a general kind of nutrient limitation is also reflected by
the fact that the fluxes of all considered nutrients correlated closely with each other
illustrating that the concentrations of all nutrients in litterfall responded in a similar
way (Fig. 5.4). However, similar to the findings of Vogt et al. (1986) for N and P, the
correlations between the concentrations of P and K in litterfall and litterfall
mass were slightly less close than that of N in litterfall with litterfall mass, support-
ing a generally more important role of N for litterfall than of P and K.
5.5 Conclusions
The currently available litterfall data from Andean forests has major gaps.
Particularly, the south Andes including Bolivia, Chile, and Argentina are hardly
covered. Moreover, there is almost no litterfall data from the dry forest, which cov-
ers large parts of the inner Andes and there is little data about the litterfall in the
far-spread tree plantations of mostly exotic Pinus, Cupressus, and Eucalyptus spe-
cies. The composition of litterfall showed little variation with ca. 70% leaf litterfall
and ca. 10% flower and fruit litterfall.
Litterfall increased towards the equator and showed a hump-shaped elevational
distribution. The most productive forests of the Andes are the lower montane forests
in the tropical north part. The temperate forest in the south Andes has a similar pro-
ductivity as the tropical upper montane forests. Litterfall and organic layer thickness
are negatively related illustrating different controls of litterfall and organic layer
thickness.
Forests became increasingly nutrient-efficient with increasing elevation but there
was no general change in nutrient relationships suggesting that the kind of nutrient
limitation varies among different locations. Nevertheless, concentrations of N in
litterfall showed a closer correlation with litterfall mass than those of P and K con-
firming a generally slightly more important role of N for the productivity of Andean
forests than of other nutrients.
The observed elevational influence of litterfall in the humid tropical Andes sug-
gests that the forest productivity will likely respond to climate change driving the
vegetation belts to higher elevation with an unknown overall effect on C sequestra-
tion by net primary production of these forests.
108 W. Wilcke
Acknowledgments I thank E. Beck, K. Müller-Hohenstein, M. Richter, and W. Zech for co-
initiating the long-term study; C. Valarezo and the National University of Loja for their long-term
cooperation, J. Boy, K. Fleischbein, R. Goller, M. Meyer-Grünefeldt, M. Sequeira, A. Velescu,
H. Wullaert, S. Yasin, and numerous Ecuadorian and German graduate and undergraduate students
for data acquisition during parts of the observation period; the Ecuadorian Environmental Ministry
for the research permits; Naturaleza y Cultura Internacional (NCI) in Loja for providing the study
area and the research station; and the Deutsche Forschungsgemeinschaft (DFG) for funding (FOR
402 and 816).
References
Bray JR, Gorham E (1964) Litter production in forests of the world, Adv Ecol Res 2:101–157
Bendix J, Rollenbeck R, Richter M, Fabian P, Emck P (2008) Chapter 8: Climate. In: Beck E,
Bendix J, Kottke I, Makeschin F, Mosandl R (eds) Gradients in a Tropical Mountain Ecosystem
of Ecuador. Ecol Stud 198, pp 63–73, Springer-Verlag, Heidelberg
Bruijnzeel LA, Hamilton LS (2000) Decision time for cloud forests. In: IHP Humid Tropics
Programme Series 13, IHP-UNESCO, Paris
Bruijnzeel LA, Mulligan M, Scatena FM (2011) Hydrometeorology of tropical montane cloud
forests: emerging patterns. Hydrol Process 25:465–498
Bueis T, Bravo F, Pando V, Belén-Turrión M (2017) Influencia de la densidad del arbolado sobre
el desfronde y su reciclado en pinares de repoblación del norte de España. Bosque 38:401–407
Chave J, Navarrete D, Almeida S, Álvarez E, Aragão LEOC, Bonal D, Châtelet P, Silva -Espejo
JE, Goret J-Y, von Hildebrand P, Jiménez E, Patiño S, Peñuela MC, Pillips OL, Stevenson
P, Malhi Y (2010) Regional and seasonal patterns of litterfall in tropical South America.
Biogeosciences 7:43–55
Erkan N, Comez A, Aydin AC, Denli O, Erkan S (2018) Litterfall in relation to stand parameters
and climatic factors in Pinus brutia forests in Turkey. Scand J For Res 33:338–346
Espinosa CI, Cabrera O, Luzuriaga AL, Escudero A (2011) What factors affect diversity and
species composition of endangered Tumbesian dry forests in southern Ecuador? Biotropica
43:15–22
Fisher JB, Malhi Y, Cuba Torres I, Metcalfe DB, van de Weg MJ, Meir P, Silva-Espejo JE, Huaraca
Huasco W (2013) Nutrient limitation in rainforests and cloud forests along a 3,000-m elevation
gradient in the Peruvian Andes. Oecologia 172:889–902
Flórez-Flórez CP, León-Peláez JD, Osorio-Vega NW, Restrepo-Llano MF (2013) Dinámica de
nutrientes en plantaciones forestales de Azadirachta indica (Meliaceae) establecidas para res-
tauración de tierras degradadas en Colombia. Rev Biol Trop 61:515–529
Fuentes-Molina N, Rodriguez-Barrios J, Isenia-Leon S (2018) Caída y descomposición de hoja-
rasca en los bosques ribereños del Manantial de Cañaverales, Guajira, Colombia. Acta Biol
Colomb 23:115–123
Galloway JN, Dentener FJ, Capone DG, Boyer EW, Howarth RW, Seitzinger SP, Asner GP,
Cleveland CC, Green PA, Holland EA, Karl DM, Michaels AF, Porter JH, Townsend AR,
Vörösmarty CJ (2004) Nitrogen cycles: past, present, and future. Biogeochemistry 70:153–226
Girardin CAJ, Malhi Y, Aragão LEOC, Mamani-Solórzano M, Huaraca-Huasco W, Durand L,
Feeley KJ, Rapp J, Silva-Espejo JE, Silman MR, Salinas N, Whittaker RJ (2010) Net primary
productivity allocation and cycling of carbon along a tropical forest elevational transect in the
Peruvian Andes. Glob Change Biol 16:3176–3192
Girardin CAJ, Silva-Espejo JE, Doughty CE, Huaraca-Huasco W, Metcalfe DB, Durand-Baca
L, Marthews TR, Aragão LEOC, Farfán-Rios W, García-Cabrera K, Halladay K, Fisher JB,
Galiano-Cabrera DF, Huaraca-Quispe LP, Alzamora-Taype I, Eguiluz-Mora L, Salinas -Revilla
N, Silman MR, Meir P, Malhi Y (2014) Productivity and carbon allocation in a tropical mon-
tane cloud forest in the Peruvian Andes. Plant Ecol Div 7:107–123
Günter S, Cabrera O, Weber M, Stimm B, Zimmermann M, Fiedler K, Knuth J, Boy J, Wilcke
W, Iost S, Makeschin F, Werner F, Gradstein R, Mosandl R (2008) Chapter 26: Natural for-
est management in neotropical mountain rain forests – an ecological experiment. In: Beck E,
Bendix J, Kottke I, Makeschin F, Mosandl R (eds) Gradients in a Tropical Mountain Ecosystem
Heinrichs H, Brumsack H-J, Loftfield N, König N (1986) Verbessertes Druckaufschlußsystem für
biologische und anorganische Materialien. Z Pflanzenernaehr Bodenkd 149:350–353
Homeier J (2004) Baumdiversität, Waldstruktur und Wachstumsdynamik zweier tropischer
Bergregenwälder in Ecuador und Costa Rica. Dissertationes Botanicae 391. J. Cramer, Berlin
Homeier J, Hertel D, Camenzind T, Cumbicus NL, Maraun M, Martinson GO, Poma LN, Rillig
MC, Sandmann D, Scheu S, Veldkamp E, Wilcke W, Wullaert H, Leuschner C (2012) Tropical
Andean forests are highly susceptible to nutrient inputs – rapid effects of experimental N and
P addition to an Ecuadorian montane forest. PLoS One 7:e47128
Homeier J, Leuschner C, Bräuning A, Cumbicus NL, Hertel D, Martinson GO, Spannl S, Veldkamp
E (2013) Chapter 23: Effects of nutrient addition on the productivity of montane forests and
implications for the carbon cycle. In: Bendix J, Beck E, Bräuning A, Makeschin F, Mosandl
R, Scheu S, Wilcke W (eds) Ecosystem Services, Biodiversity and Environmental Change
in a Tropical Mountain Ecosystem of South Ecuador. Ecol Stud 221, pp 315–329, Springer-
Verlag, Berlin
Jordan CF (1971) A world pattern in plant energetics. Am Sci 59:426–433
León JD, González MI, Gallardo JF (2011) Ciclos biogeoquímicos en bosques naturales y planta-
ciones de coníferas en ecosistemas de alta montaña de Colombia. Rev Biol Trop 59:1883–1894
Leuschner C, Zach A, Moser G, Homeier J, Graefe S, Hertel D, Wittich B, Soethe N, Iost S,
Röderstein M, Horna V, Wolf K (2013) Chapter 10: The carbon balance of tropical mountain
forests along an altitudinal transect. In: Bendix J, Beck E, Bräuning A, Makeschin F, Mosandl
R, Scheu S, Wilcke W (eds) Ecosystem Services, Biodiversity and Environmental Change
in a Tropical Mountain Ecosystem of South Ecuador. Ecol Stud 221, pp 117–139, Springer-
Verlag, Berlin
Marschner P (ed) (2012) Marschner’s Mineral Nutrition of Higher Plants, 3rd edn. Academic
Press/Elsevier, Amsterdam
Moser G, Leuschner C, Hertel D, Graefe S, Soethe N, Iost S (2011) Elevation effects on the carbon
budget of tropical mountain forests (S Ecuador): the role of the belowground compartment.
Glob Change Biol 17:2211–2226
Pérez CA, Hedin LO, Armesto JJ (1998) Nitrogen mineralization in two unpolluted old-growth
forests of contrasting biodiversity and dynamics. Ecosystems 1:361–373
Pérez CA, Armesto JJ, Torrealba C, Carmona MR (2003) Litterfall dynamics and nitrogen use
efficiency in two evergreen temperate rainforests of southern Chile. Austral Ecol 28:591–600
Peters T, Drobnik T, Meyer H, Rankl M, Richter M, Rollenbeck R, Thies B, Bendix J (2013) Chapter
2: Environmental changes affecting the Andes of Ecuador. In: Bendix J, Beck E, Bräuning
A, Makeschin F, Mosandl R, Scheu S, Wilcke W (eds) Ecosystem Services, Biodiversity and
Environmental Change in a Tropical Mountain Ecosystem of South Ecuador. Ecol Stud 221,
pp 19-29, Springer-Verlag, Berlin
Pinos J, Studholme A, Carabajo A, Gracia C (2017) Leaf litterfall and decomposition of Polylepis
reticulata in the treeline of the Ecuadorian Andes. Mount Res Develop 37:87–96
Ramírez JA, León-Peláez JD, Craven D, Herrera DA, Zapata CM, González-Hernández MI,
Gallardo-Lancho J, Osorio W (2014) Effects on nutrient cycling of conifer restoration in a
degraded tropical montane forest. Plant Soil 378:215–226
Röderstein M, Hertel D, Leuschner C (2005) Above- and belowground litter production in three
topical montane forests in southern Ecuador. J Trop Ecol 21:483–492
Schawe M, Glatzel S, Gerold G (2007) Soil development along an altitudinal transect in a Bolivian
tropical montane rainforest: Podzolization vs. hydromorphy. Catena 69:83–90
110 W. Wilcke
Sierra CA, Harmon ME, Moreno FH, Orrego SA, Del Valle JI (2007) Spatial and temporal vari-
ability of net ecosystem production in a tropical forest: testing the hypothesis of a significant
carbon sink. Glob Change Biol 13:838–853
Soil Survey Staff (2014): Keys to Soil Taxonomy, 12th ed. United States Department of Agriculture,
Natural Resources Conservation Service, Washington DC, USA
Staelens J, Ameloot N, Amonacid L, Padilla E, Boeckx P, Huygens D, Verheyen K, Oyarzún C,
Godoy R (2011) Litterfall, litter decomposition and nitrogen mineralization in old-growth ever-
green and secondary deciduous Nothofagus forests in south-Central Chile. Rev Chil Hist Nat
84:125–141
Steinhardt U (1979) Untersuchungen über den Wasser- und Nährstoffhaushalt eines andinen
Wolkenwaldes in Venezuela. In: Göttinger Bodenkundliche Berichte, vol 56. University of
Göttingen, Göttingen, pp 1–146
Tanner EVJ, Kapos V, Franco W (1992) Nitrogen and phosphorus fertilization effects on
Venezuelan montane forest trunk growth and litterfall. Ecology 73:78–86
Urrutia R, Vuille M (2009) Climate change projections for the tropical Andes using a regional
climate model: temperature and precipitation simulations for the end of the 21st century. J
Geophys Res 114:D02108
Veneklaas EJ (1991) Litterfall and nutrient fluxes in two montane tropical rain forests, Colombia.
J Trop Ecol 7:319–336
Vitousek PM (1984) Litterfall, nutrient cycling, and nutrient limitation in tropical forests. Ecology
65:285–298
Vogt KA, Grier CC, Vogt DJ (1986) Production, turnover, and nutrient dynamics of above- and
belowground detritus of world forests. Adv Ecol Res 15:303–377
Vuille M, Bradley RS, Werner M, Keimig F (2003) 20th century climate change in the tropical
Andes: observations and model results. Clim Chang 59:75–99
Weber M, Günter S, Aguirre N, Stimm B, Mosandl R (2008) Chapter 34: Reforestation of aban-
doned pastures: silvicultural means to accelerate forest recovery and biodiversity. In: Beck E,
Bendix J, Kottke I, Makeschin F, Mosandl R (eds) Gradients in a tropical mountain ecosystem
Wilcke W, Yasin S, Valarezo C, Zech W (2001) Change in water quality during the passage through
a tropical montane rain forest in Ecuador. Biogeochemistry 55:45–72
Wilcke W, Yasin S, Abramowski U, Valarezo C, Zech W (2002) Nutrient storage and turnover in
organic layers under tropical montane rain forest in Ecuador. Eur J Soil Sci 53:15–27
Wilcke W, Valladarez H, Stoyan R, Yasin S, Valarezo C, Zech W (2003) Soil properties on a chro-
nosequence of landslides in montane rain forest, Ecuador. Catena 53:79–95
Wilcke W, Oelmann Y, Schmitt A, Valarezo C, Zech W, Homeier J (2008) Soil properties and tree
growth along an altitudinal transect in Ecuadorian tropical montane forest. J Plant Nutr Soil
Sci 171:220–230
Wilcke W, Günter S, Alt F, Geißler C, Boy J, Knuth J, Oelmann Y, Weber M, Valarezo C, Mosandl
R (2009) Response of water and nutrient fluxes to improvement fellings in a tropical montane
forest in Ecuador. For Ecol Manag 257:1292–1304
Wilcke W, Leimer S, Peters T, Emck P, Rollenbeck R, Trachte K, Valarezo C, Bendix J (2013)
The nitrogen cycle of tropical montane forest in Ecuador turns inorganic under environmental
change. Glob Biogeochem Cycle 27:1194–1120
Wolf K, Veldkamp E, Homeier J, Martinson GO (2011) Nitrogen availability links forest pro-
ductivity, soil nitrous oxide and nitric oxide fluxes of a tropical montane forest in southern
Ecuador. Glob Biogeochem Cycle 25:GB4009
Zimmermann M, Meir P, Bird MI, Malhi Y, Ccahuana AJQ (2009) Climate dependence of het-
erotrophic soil respiration from a soil-translocation experiment along a 3000 m tropical forest
altitudinal gradient. Eur J Soil Sci 60:895–906
Chapter 6
Arbuscular Mycorrhizal Fungi
and Ectomycorrhizas in the Andean
Cloud Forest of South Ecuador
Ingeborg Haug, Sabrina Setaro, and Juan Pablo Suárez
6.1 Introduction
First recognized by A. B. Frank (1885) in forests around Berlin and later corrobo-
rated by thousands of scientific studies, it was shown that most plants depend on
fungi for nutrient uptake and—by a symbiotic interaction called mycorrhiza—sup-
port these root-inhabiting fungi by delivering carbohydrates and amino acids. While
ectomycorrhizas in the boreal forests dominate with the fungal partners Ascomycetes
and Basidiomycetes, in tropical forests most trees have arbuscular mycorrhizas with
Glomeromycotina. The detection of ectomycorrhizal fungi is standardized, while
the description of environmental communities of arbuscular mycorrhizal fungi at
the species level remains challenging: no sexual structures of the Glomeromycotina
are known, thus the microscopic spores are the only morphological structures to
describe species. Isolation of spores—especially from humus soil—is difficult, and
reliable determinations can only be done by specialists. Added to this is the problem
that not all AM fungi form spores in equal measure, and thus only a part of the com-
munity is recorded. The structures inside the mycorrhizas (hyphae, vesicles, arbus-
cles) are not suitable for differentiating species. So remains the molecular approach,
which also has difficulties. Several “AM-specific” primers have been published, but
the question is how well they work and whether all groups of Glomeromycotina are
covered. Furthermore, the sequences have to be split into OTUs (Operational
Taxonomic Units = surrogate for species)—what is the “right” threshold?!
I. Haug (*)
Universitat Tübingen, Tübingen, Germany
S. Setaro
Wake Forest University, Winston-Salem, USA
J. P. Suárez
Universidad Técnica Particular de Loja, Loja, Ecuador

https://doi.org/10.1007/978-3-030-57344-7_6
112 I. Haug et al.
As far as we know, there are only some mycorrhizal investigations in the Andean
Cloud Forest. Camenzind et al. (2014) analysed the AMF community of ECSF
(South Ecuador) by 454-pyrosequencing and identified 74 OTUs, among which 26
were placed in the Diversisporales and 48 in the Glomerales. Regarding their abun-
dances, 63% of sequences represented Diversisporales. Marín et al. (2016, 2017)
investigated the spore communities of AMF in Nothofagus and Araucaria forests.
They proved 18 respectively 14 species, mainly members of the genus Acaulospora.
In the Andean Yungas Forests, Geml et al. (2014) studied the community composi-
tion of all fungi in three elevation belts (400–3000 masl) based on Ion Torrent
sequencing of the ITS2 rDNA from soil samples. Glomeromycota comprised only
a small part of the entire soil fungal community—2.91% equalling 409 OTUs, of
which 321 OTUs were Glomerales. AMF richness was negatively correlated with
elevation.
In our study (Haug et al. 2019), we worked with hundreds of samples in different
regions of the South Ecuadorian Andean cloud forest with the same method (nested
PCR with primers AM1–AM2 [Lee et al. 2008], Sanger sequencing, threshold of
99% sequence similarity), thus providing a good database and offering the possibil-
ity to compare the communities at the different sites and elevation levels.
6.2 The Study Region
Samples were taken at four elevation belts in Southern Ecuador: 1000 masl,
2000 masl, 3000 masl and 4000 masl (Fig. 6.1, Table 6.1). The 1000 m sites are
located in the Bombuscaro area in Parque Nacional Podocarpus (4°11’S, 78°96’W);
the 2000 m sites in the Reserva Biológica San Francisco (RBSF; 3°58’S, 79°04’W)
on the eastern slope of the Cordillera El Consuelo, Zamora-Chinchipe Province; the
3000 m sites in Cajanuma area (4°12’S, 79°17’W) in Parque Nacional Podocarpus
and in the Nero area (2°95’S, 79°10’W) at Parque Nacional Cajas; and the 4000 m
sites in Tutupali (3°03’S, 79°15’W), Soldados (2°98’S, 79°31’W) and the Toreadora
area (2°47’S, 79°11’W) at Parque Nacional Cajas. The vegetation at 1000 masl con-
sists of evergreen premontane rainforest, whereas the forest at 2000 masl is charac-
terized by evergreen lower montane forest. Upper montane forest occurs at the
3000 m sites; Polylepis forest and grass páramo can be found at the 4000 m sites.
Details are summarized in Table 6.1.
6.3 Materials and Methods (For Details See Haug et al. 2019)
At each elevation belt, we sampled three to four sites (Table 6.2). We collected
arbuscular mycorrhizas directly from the soil without regarding the identity of the
plant partner. Each fine-root system was placed in a separate PCR tube to ensure
6 Arbuscular Mycorrhizal Fungi and Ectomycorrhizas in the Andean Cloud Forest… 113
Fig. 6.1 Map of the study region in Southern Ecuador. Location of the study areas: Bombuscaro,
Reserva Biologíca San Francisco (RBSF), Cajanuma, Cuenca region with Nero, Tutupali, Soldados
and Cajas Toreadora. http://vhrz669.hrz.uni-marburg.de/tmf_respect/data_pre.do?citid=1745
that each mycorrhizal sample is from a single host plant. Mycorrhizas were dried at
about 50˚ C for 24 h by placing open sample tubes on an electric dryer. After the
drying step, silica gel was added and the tubes were closed for long-term storage.
In total, we successfully worked with 646 root samples: 211 at 1000 masl, 184 at
2000 masl, 128 at 3000 masl and 123 at 4000 masl (Table 6.2). Total DNA was
114 I. Haug et al.
Table 6.1 Site characteristics for the four elevation levels. Data from (Homeier et al. 2008)
(Moser et al. 2007), (Homeier et al. 2013a, b), (Martinson et al. 2013), (Crespo et al. 2011), (Pinos
et al. 2017)
Elevation level 1000 m site 2000 m site 3000 m site 4000 m site
Sites PNP— RBSF PNP—Cajanuma Cajas
Bombuscaro NP—Toreadora
Elevation 1050 masl 1890 masl 3060 masl 3930 masl
Vegetation Evergreen Evergreen lower Upper montane Grass Páramo,
premontane forest montane forest forest, shrub Polylepis forest
Páramo
Annual mean air 19.4 °C 15.7 °C 9.4 °C 5.4 °C
temperature
Annual rainfall c. 2230 mm c. 1950 mm c. 4500 mm c. 876 mm
Soil type (FAO) Dystric Cambisol Stagnic Cambisol Stagnic Histosol Histosol,
andosol
Thickness of 0 cm 10–30 cm 10–40 cm 10–40 cm
organic layer
Organic layer 3.9 3.9 4.2–4.8
pH-H2O
Stand height 20–25 m, up to 18–22 m 8–10 m
40 m
Table 6.2 Sampling subsites with number of samples taken and sequences obtained per site.
1000 m 2000 m evergreen 4000 m grass
evergreen lower montane 3000 m upper montane páramo/Polylepis
premontane forest forest/shrub páramo forest
forest
1000–1140 m 1900–2500 m 2880–3250 m 3500–4000 sm
B1 B2 B3 Q2 Q5 T1 T2 Cc Cm Cl Clp N Tu So Cpo Cpa
Samples 103 82 26 10 63 6 105 36 21 46 17 8 20 15 32 56
analysed
Number of 246 193 55 13 173 15 152 55 28 67 20 11 30 29 41 81
sequences
Abbreviations: B1, B2, B3 Bombuscaro plot 1, 2, 3, Q2, Q5 ravines, T1, T2 ridges, Cc Cajanuma
PNP Upper Montane Forest 2700 m, Cm Cajanuma PNP Upper Montane Forest 3000 m, Cl
Cajanuma PNP Upper Montane Forest 2800 m, Clp Cajanuma PNP Shrub Páramo 3100 m, N
Cajas NP Nero Shrub Páramo 3250 m, Tu Cajas NP Tutupali Grass Páramo 3500 m, So Cajas NP
Soldados Grass Páramo 3750 m, Cpo Cajas NP Toreadora Polylepis 4000 m, Cpa Cajas NP
Toreadora Grass Páramo 4000 m
isolated from the mycorrhizal samples with the innuPREP Plant DNA Kit (Analytik
Jena, Germany). Part of the SSU region of the nuclear ribosomal rDNA repeat was
amplified by PCR using a volume of 0.5 μL of the DNA template. A nested PCR
approach was applied, amplifying the larger outer fragment with the primer combi-
nation NS1/NS4 and the smaller, inner fragment with the primer pair AML1/AML2.
Amplified PCR products were cloned with the Invitrogen TA Cloning Kit (Life
Technologies). The final dataset consisted of 1209 Glomeromycotan sequences

(Table 6.2).
All sequences were ~ 800 bp long and were aligned with MAFFT version 7
(https://mafft.cbrc.jp/alignment/server [Katoh et al. 2005], MAFFT-L-INS-i) and a
distance matrix based on p-distances was created with PAUP for OTU delimitation.
Operational Taxonomic Units (OTUs) were defined as surrogates for species based
on 99% sequence similarity and intermediate linkage clustering with OPTSIL
(Goker et al. 2010).
To explore differences in AMF community compositions among sampling sites,
we carried out a non-metric multidimensional scaling (NMDS) ordination with
metaMDS from the R package VEGAN (Oksanen et al. 2016).
6.4 Results
6.4.1 A
MF Community in Evergreen Premontane Rain Forest
at 1000 masl
Samples (n = 211) from the evergreen premontane forest in Bombuscaro, Podocarpus

National Park, revealed 57 AMF OTUs (494 sequences). The genus Glomus showed
the highest proportion of the overall AMF community in number of sequences
(89%, Fig. 6.2a) as well as in number of OTUs (39 OTUs, 68%, Fig. 6.2d). Two
Glomus OTUs (OTU17, OTU22) were especially abundant occurring in 39%
(OTU17) and 34% (OTU22) of all samples (Fig. 6.6). The second most common
genus was Acaulospora comprising 19.3% of all OTUs (11 total) and 8.1% of all
sequences (Fig. 6.2a, d). All other genera had low OTU and low sequence numbers
(Fig. 6.2a, d). Many OTUs occurred in low numbers with 78.9% of all OTUs
(n = 45) occurring in less than 5% of all samples.
6.4.2 E
ctomycorrhizas and AMF Communities in Evergreen
Lower Montane Forest at 2000 masl
Three members of the Nyctaginaceae, two Neea species and one Guapira species,
occurred scattered within a very species-rich evergreen lower montane forest. The
three species were found to form ectomycorrhizas of very distinctive characters
(Haug et al. 2005), while all other tree species examined formed arbuscular mycor-
rhizas. Neea species 1 was found to form typical ectomycorrhizas with five different
fungal species, Russula puiggarii, Lactarius sp., two Tomentella/Thelephora spe-
cies, and one ascomycete. Neea species 2 and Guapira species were associated with
only one fungus each, a Tomentella/Thelephora species clustering closely together
in an ITS-neighbour-joining tree (Haug et al. 2005). The long and fine rootlets of the
116 I. Haug et al.
1000 m Glomus
Claroideoglomus
Acaulospora
Diversispora
Gigaspora
Scutellospora
Paraglomus
Archaeospora
Ambispora
1000 m Glomus
Claroideoglomus
Acaulospora
Diversispora
Gigaspora
Scutellospora
Paraglomus
Archaeospora
Ambispora
Fig. 6.2 (a) Relative abundance of Glomeromycotina genera in 1000 masl in number of sequences;

(b, c) evergreen premontane rain forest; (d) relative abundance of Glomeromycotina genera in
1000 masl in number of OTUs
Guapira species showed proximally a hyphal mantle and a Hartig net, but distally
intracellular fungal colonization of the epidermis and root hair development. The
ectomycorrhizal segments of the long roots of Neea species 2 displayed a hyphal
mantle and a Hartig net around alive root-hair-like outgrowths of the epidermal cells
(Haug et al. 2005).
The arbuscular mycorrhizal trees at 2000 m elevation showed 66 OTUs (184
samples, 353 sequences). Seventy percent of the sequences belonged to the genus
Glomus (Fig. 6.3a), with the most frequent Glomus OTUs being OTU16, OTU49
and OTU1 (Fig. 6.6). However, these OTUs occurred only in 10–13% of all samples
(Fig. 6.6). Glomus was also the most diverse genus displaying 62% of all OTUs
(Fig. 6.3d), followed by Acaulospora (17%) and Archaeospora (11%). The genera
Claroideoglomus, Diversispora, Gigaspora and Scutellospora had low OTU and
sequence (4%) numbers (Fig. 6.3a, d).
6.4.3 A
MF Communities in Upper Montane Forest
at 3000 masl
The analysis of mycorrhizal samples at 3000 m elevation revealed 37 OTUs (181

sequences in 128 samples). This number is lower than the total number of estimated
OTUs occurring in the samples, as shown by the rarefaction not reaching a stable
phase. Richness indices estimated 49–51 OTUs in total (Chao2, Jack1). More than
half of the sequences (55%) belonged to the genus Acaulospora and 38% to the
genus Glomus (Fig. 6.4a). With 16 respectively 15 OTUs, the genera Acaulospora
and Glomus represented high proportions (43 + 40 = 83%) of the overall AMF com-
munity (Fig. 6.4d). All other genera showed very low OTU and sequence numbers
(Fig. 6.4a, d). The most frequent OTUs were Acaulospora OTU87 and OTU80
occurring in 12.7% respectively 12% of all samples and Glomus OTU43 in 11% of
samples (Fig. 6.6). As is the case for the other elevations, many OTUs were rare,
with 83% of all OTUs (n = 31) occurring in less than 5% of all samples.
6.4.4 A
MF Communities in Polylepis Forest and Páramo
at 4000 masl
The samples from 4000 m sites at Cajas National Park (n = 123) revealed 32 OTUs
(181 sequences). As is the case for the sampling at 3000 m, the rarefaction curve
showed that the area is still under-sampled and richness indices estimated a total of
48–45 OTUs (Chao2, Jack1). Acaulospora and Glomus were, again, the most domi-
nant AMF.
With 49% respectively 44%, the Acaulospora and Glomus sequences were most
frequent (Fig. 6.5a). With 16 OTUs, the genus Glomus represented a high propor-
118 I. Haug et al.
2000 m Glomus
Claroideoglomus
Acaulospora
Diversispora
Gigaspora
Scutellospora
Paraglomus
Archaeospora
Ambispora
2000 m Glomus
Claroideoglomus
Acaulospora
Diversispora
Gigaspora
Scutellospora
Paraglomus
Archaeospora
Ambispora

(b, c) lower montane rain forest; (d) relative abundance of Glomeromycotina genera in 2000 masl
in number of OTUs
3000 m Glomus
Claroideoglomus
Acaulospora
Diversispora
Gigaspora
Scutellospora
Paraglomus
Archaeospora
Ambispora
3000 m Glomus
Claroideoglomus
Acaulospora
Diversispora
Gigaspora
Scutellospora
Paraglomus
Archaeospora
Ambispora

(b, c) upper montane rain forest; (d) relative abundance of Glomeromycotina genera in 3000 masl
in number of OTUs
120 I. Haug et al.
4000 m Glomus
Claroideoglomus
Acaulospora
Diversispora
Gigaspora
Scutellospora
Paraglomus
Archaeospora
Ambispora
4000 m Glomus
Claroideoglomus
Acaulospora
Diversispora
Gigaspora
Scutellospora
Paraglomus
Archaeospora
Ambispora

(b, c) Polylepis forest and páramo; (d) relative abundance of Glomeromycotina genera in 4000 masl
in number of OTUs
Fig. 6.6 Relative abundance of most frequent OTUs (occurrence in >10% of samples) of each
altitudinal belt
tion (50%) of the overall AMF community (Fig. 6.5d). The second most common
genus Acaulospora comprised 10 OTUs (31%, Fig. 6.5a). There was also a surpris-
ingly high proportion of Claroideoglomus sequences (6,1%, Fig. 6.5a) at 4000 masl.
All other genera showed low OTU and low sequence numbers (Fig. 6.5a, d). Most
OTUs occurred in low numbers too, with 75% only occurring in up to 5% of the
samples. The exceptions are four Acaulospora OTUs (82, 84, 87, 92) and four
Glomus OTUs (23, 25, 36, 41) with occurrences ranging from 6 to 16% of all sam-
ples (Fig. 6.6).
6.4.5 A
MF Community Composition Along the Elevational
Gradient
Most OTUs (50%) occurred at only one altitudinal level. The highest number of
these unique OTUs both in richness and frequency was in the 1000 m belt (Fig. 6.7)
with a downward trend towards higher elevations (Fig. 6.7).
Three OTUs (2.6%) were found at all altitudinal levels: OTU1 (Glomus), OTU41
(Glomus) and OTU82 (Acaulospora); and thirteen OTUs (11%) occurred in three
elevation belts.
The composition of the arbuscular mycorrhizal communities changed gradually
with elevation: among the lowest (1000 masl) and highest (4000 masl) sites are only
five OTUs shared and all similarity indices are low (Table 6.3, Fig. 6.7). The neigh-
bouring altitudinal belts 1000/2000 masl and 3000/4000 masl showed a high over-
lap of OTUs (32%, Table 6.3, Fig. 6.7). The overlap between 2000 and 3000 m was
much lower (23%), though, and was associated with a decrease in Glomus OTUs
122 I. Haug et al.
Fig. 6.7 Shared and unique OTUs of each altitudinal belt. The thickness of connector lines repre-
sents the percentage of OTUs shared
Table 6.3 Shared OTUs between elevational belts

First sample Second sample OTUs first sample OTUs second sample Shared OTUs observed
1000 m 2000 m 57 66 30 (32%)
1000 m 3000 m 57 37 11 (13%)
1000 m 4000 m 57 32 5 (6%)
2000 m 3000 m 66 37 19 (23%)
2000 m 4000 m 66 32 15 (18%)
3000 m 4000 m 37 32 19 (32%)
and an increase in Acaulospora OTUs on 3000 masl (Table 6.3). All similarity indi-

ces showed highest values for the neighbouring 3000 and 4000 masl communities
(Table 6.4). The gradual change in OTU composition with altitude is also shown by
the NMDS plot (Fig. 6.8a), which shows a high percentage of explained deviance
(93%) and a mixed linearity indicated by the estimated degrees of freedom (edf) of
5.45 (edf = 1 indicates a linear relationship).
Table 6.4 Similarity indices

Chao- Chao-
Chao- Chao- Sorensen- Sorensen-
Jaccard-Raw Jaccard-Est Raw Est
First Second Jaccard Sorensen Abundance- Abundance- Abundance- Abundance-
Sample Sample Classic Classic Based Based Based Based
1000 m 2000 m 0,32 0,49 0,23 0,28 0,38 0,44
1000 m 3000 m 0,13 0,23 0,16 0,45 0,27 0,63
1000 m 4000 m 0,06 0,112 0,073 0,13 0,136 0,23
2000 m 3000 m 0,226 0,369 0,31 0,534 0,473 0,696
2000 m 4000 m 0,181 0,306 0,211 0,322 0,348 0,487
3000 m 4000 m 0,38 0,551 0,55 0,74 0,71 0,85
Fig. 6.8 (a, b) NMDS plot for AM fungal communities at 1000 m (blue), 2000 m (violet), 3000 m
(green) and 4000 m (red) level with modelled elevation (a) and rainfall (b) mapped (pink lines) on
to it. Stress value: 0.12 Abbreviations for the figures: B1, B2, B3 Bombuscaro plot 1, 2, 3, 1000 m
Q2, Q5 ravines 2000 m, T1 2100 m, T2 2150 m ridges, Cc Cajanuma PNP Upper Montane Forest
2700 m, Cm Cajanuma PNP Upper Montane Forest 3000 m, Cl Cajanuma PNP Upper Montane
Forest 2800 m, Clp Cajanuma PNP Shrub Páramo 3100 m, N Cajas NP Nero Shrub Páramo
3250 m, Tu Cajas NP Tutupali Grass Páramo 3500 m, So Cajas NP Soldados Grass Páramo 3750 m,
Cpo Cajas NP Toreadora Polylepis 4000 m, Cpa Cajas NP Toreadora Grass Páramo 4000 m
124 I. Haug et al.
6.5 Discussion
6.5.1 T
he Influence of Elevation on AMF Community
Composition
AMF community composition is strongly influenced by elevation, which is shown

by the similarity indices (Table 6.4), ordination results (Fig. 6.8) and phylogenetic
community analysis. Even though few OTUs were distributed across all elevations,
half of the OTUs were restricted to one height level—suggesting high community
turnover among the altitudinal belts. This finding is important because it indicates
that AMF communities are structured in a similar way as plant communities
(Homeier et al. 2008; Homeier et al. 2013a, b).
Overall, our results showed a gradual change in AMF community composition
(Figs. 6.2, 6.3, 6.4, 6.5, and 6.7). However, there was a pronounced switch from
2000 masl and below to 3000 masl and above in respect to the dominance of
Glomeraceae and Acaulospora. Glomeraceae were dominant in lower elevations,
whereas Acaulospora took over in higher elevations. This result was surprising
because we suspected that if a shift in dominance occurred, it would be from
3000 masl to 4000 masl as this marks the switch from forest vegetation to páramo.
Many frequent OTUs from our study sites were “generalists”, meaning they were
widespread along the altitudinal gradient but there were just as many frequent OTUs
with a narrow distribution range referred here as “specialists”. Contrary to our
expectations, the number and frequency of “specialist” OTUs decreased with eleva-
tion, indicating that more “generalists” and fewer “specialists” are occurring under
stressful environmental conditions. This was also found by Kawahara et al. (2016)
who hypothesized that AM fungi in acidic soils are specifically adapted to this soil
type. Yet what they observed was that fungi in strongly acidic soils were pH general-
ists occurring in soils with a wide range of pH values. Thus, the conceptual frame-
work of Gostincar et al. (2010) that fungi in extreme environments may evolve
towards niche specialists unable to compete or survive in moderate environments
does not seem to apply to AM fungi.
6.5.2 Species Richness Along the Elevational Gradient
Differences in observed number of OTUs may be a consequence of variable sam-

pling intensity. OTU accumulation curves indicated that additional sampling might
have resulted in the detection of more additional OTUs at 3000 and 4000 masl.
6.5.3 A
bundant AMF Taxa in Our Study Sites Put in a Global
Context
To put the most abundant OTUs we found in this study into a global perspective, we
report here on habitat and distribution of the closest hits from the MaarjAM data-
base. Some AMF were found to have a worldwide distribution, but some are reported
here for the first time. The most abundant OTUs at the 1000 m level (OTUs 17 and
22) had no close match, neither in the MaarjAM nor in the NCBI database. One of
the most widespread OTUs with global distribution and occurring in different
biomes belong to the Rhizophagus intraradices/irregularis/vesiculiferus group
(MaarjAM ID: VT113/114/115—Opik et al. 2006; Opik et al. 2010) here called
OTU1. This generalist OTU was present in low numbers at all altitudinal belts but
was only frequent at the 2000 m level. Another generalist and globally distributed is
OTU43 (MaarjAM ID: VT191), belonging to Glomeraceae. Also found worldwide
but restricted to forests is OTU16 (MaarjAM ID: VT183). It belongs to the family
Glomeraceae and was only found at the 2000 m belt in our study. Another frequent
2000 m OTU was OTU49 (Glomeraceae), the type sequence for VT183 (MaarjAM
ID). OTU49 might be restricted to tropical rain forests as it is so far only known
from tropical rain forests in French Guiana, Gabon and Ecuador. As for Acaulospora,
three OTUs were frequent in our study area: OTU80 (Maarjam ID: VT12), occur-
ring from 1000 m to 3000 m but with main abundance in 3000 m, is otherwise only
known from South America (Argentina and French Guiana); OTU 82 (Maarjam ID:
VT14), a generalist that also occurs in Europe and Asia; and OTU87 (Maarjam ID:
VT30), in our study only at 3000 m but with worldwide distribution and in
many biomes.
The dominance of Glomeraceae at 1000 and 2000 masl is in accordance with
many other studies using spore-, Sanger-sequence- or NGS-based sampling tech-
niques from different ecosystems (e.g. Gai et al. 2006, Camenzind et al. 2014, Leal
et al. 2013, da Silva et al. 2015, Rodriguez-Echeverria et al. 2017, Geoffroy et al.
2017). For Acaulospora, however, the literature is controversial. Dominance of
Acaulosporaceae in higher elevations, as our results show, is supported by Egan
et al. (2017), who observed an increase in the number of Acaulospora species along
a high elevation gradient in Montana (USA) and in three studies who found the
same pattern in the Himalaya (Yang et al. 2016; Li et al. 2014; Liu et al. 2015). Also,
two unknown Acaulospora species were the dominant colonizers of Andean pota-
toes growing between 2658 and 4075 masl (Senes-Guerrero and Schussler 2016).
Other studies in the Himalaya did not find Acaulospora to be dominant, but
Glomeraceae instead (Li et al. 2015; Yang et al. 2016; Kotilinek et al. 2017). The
reason behind the frequency patterns of Glomeraceae and Acaulosporaceae is not
clear and remains to be studied.
126 I. Haug et al.
6.5.4 R
are OTUs in AMF Communities and Its Potential Role
in Ecosystem Functioning
Like previous studies of AM fungal community structure (Rosendahl 2008;

Dumbrell et al. 2010), we found a high proportion of rare OTUs at all altitudes:
79–85% of OTUs occurred in less than 5% of the samples. However, abundant taxa
representing on average 40% of total abundance within the community as shown in
previous studies (Dumbrell et al. 2010) were only present at the 1000 m level. At
higher altitudes, the most common OTUs were detected in only 13–15% of the
samples, which is not very frequent. This mirrors the situation for plants in tropical
forests (Wright 2002). The processes behind these patterns are not fully understood,
not even for plants (Prada and Stevenson 2016). Niche differentiation is considered
to be an important factor in the structuring of communities, for plants (Prada and
Stevenson 2016; Wright 2002) and AM fungi (Zobel and Opik 2014), but stochastic
processes also play a role (Lekberg and Waller 2016; Encinas-Viso et al. 2016). The
study sites we investigated had a broad range of edaphic variability at each altitudi-
nal belt, which could provide many microhabitats for AMF to specialize in.
Examples are ridges vs. ravine habitats at the 2000 m belt and grassland páramo vs.
Polylepis forest at 4000 masl.
The importance of rare plant species and their distributions across ecosystems is
discussed in several studies (e.g. Gaston 2012; Mi et al. 2012). Determining the
ecological similarity of rare and common species is not easy even in plants, particu-
larly when a large number of species are involved. We can assume, though, that the
more the species and the higher the phylogenetic diversity, the more traits are avail-
able in a community. The multitude of rare fungi in our study with a broad phyloge-
netic level at each elevation thus offers a wide range of traits with potential benefits
to their plant partners. Therefore, rare species might be important for maintaining
ecosystem functioning in a changing environment (Bachelot et al. 2015). We can
see evidence for this in our data as some OTUs (e.g. OTU1, OTU43, OTU80,
OTU82, OTU87) are rare at a certain altitude but become dominant in another alti-
tude. The ability to maintain many rare OTUs in combination with low specializa-
tion might be a reason why AM fungi have been one of the most successful organisms
on the earth, being around since 450 million years and occurring in all terrestrial
habitats.
Acknowledgements We thank Jutta Bloschies for laboratory work. We appreciate the German
Research Foundation (DFG) for financing the study (Research Units 402 and 816, PAK825),
Naturaleza y Cultura Internacional (Loja, San Diego) for support, our Ecuadorian partner
Universidad Técnica Particular de Loja (UTPL) for outstanding cooperation, the Ecuadorian
Ministerio del Ambiente for granting research permits, Patrick Hildebrandt and Carlos Quiroz for
the possibility to take samples in Nero, Soldados and Tutupali, and the staff of the Estación
Científica San Francisco—especially Felix Matt and Jörg Zeilinger—for their great assistance.
References
Bachelot B, Kobe RK, Vriesendorp C (2015) Negative density-dependent mortality varies

over time in a wet tropical forest, advantaging rare species, common species, or no species.
Oecologia 179(3):853–861
Camenzind T, Hempel S, Homeier J, Horn S, Velescu A, Wilcke W, Rillig MC (2014) Nitrogen and
phosphorus additions impact arbuscular mycorrhizal abundance and molecular diversity in a
tropical montane forest. Glob Chang Biol 20(12):3646–3659
Crespo PJ, Feyen J, Buytaert W, Bucker A, Breuer L, Frede HG, Ramirez M (2011) Identifying
controls of the rainfall-runoff response of small catchments in the tropical Andes (Ecuador). J
Hydrol 407(1–4):164–174
da Silva DKA, Coutinho FP, Escobar IEC, de Souza RG, Oehl F, Silva GA, Cavalcante UMT, Maia
LC (2015) The community of arbuscular mycorrhizal fungi in natural and revegetated coastal
areas (Atlantic Forest) in northeastern Brazil. Biodivers Conserv 24(9):2213–2226
Dumbrell AJ, Nelson M, Helgason T, Dytham C, Fitter AH (2010) Idiosyncrasy and overdomi-
nance in the structure of natural communities of arbuscular mycorrhizal fungi: is there a role
for stochastic processes? J Ecol 98(2):419–428
Egan CP, Callaway RM, Hart MM, Pither J, Klironomos J (2017) Phylogenetic structure of arbus-
cular mycorrhizal fungal communities along an elevation gradient. Mycorrhiza 27(3):273–282
Encinas-Viso F, Alonso D, Klironomos JN, Etienne RS, Chang ER (2016) Plant-mycorrhizal fun-
gus co-occurrence network lacks substantial structure. Oikos 125(4):457–467
Frank B (1885) Über die auf Wurzelsymbiose beruhende Ernährung gewisser Bäume durch
unterirdische Pilze. Ber. Deutsch. Bot Ges 3:128–145
Gai JP, Christie P, Feng G, Li XL (2006) Twenty years of research on community composition
and species distribution of arbuscular mycorrhizal fungi in China: a review. Mycorrhiza
16(4):229–239
Gaston KJ (2012) ECOLOGY the importance of being rare. Nature 487(7405):46–47
Geml J, Pastor L, Fernandez S, Pacheco TA, Semenova AG, Becerra CY, Wicaksono W, Nouhra
ER (2014) Large-scale fungal diversity assessment in the Andean Yungas forests reveals
strong community turnover among forest types along an altitudinal gradient. Mol Ecol
23(10):2452–2472
Geoffroy A, Sanguin H, Galiana A, Ba A (2017) Molecular Characterization of Arbuscular
Mycorrhizal Fungi in an Agroforestry System Reveals the Predominance of Funneliformis spp
Associated with Colocasia esculenta and Pterocarpus officinalis Adult Trees and Seedlings.
Front Microbiol 8:1426
Goker M, Grimm GW, Auch AF, Aurahs R, Kucera M (2010) A clustering optimization strat-
egy for molecular taxonomy applied to planktonic foraminifera SSU rDNA. Evol Bioinform
Online 6:97–112
Gostincar C, Grube M, de Hoog S, Zalar P, Gunde-Cimerman N (2010) Extremotolerance in fungi:
evolution on the edge. FEMS Microbiol Ecol 71(1):2–11
Haug I, Setaro S, Suarez JP (2019) Species composition of arbuscular mycorrhizal communities
changes with elevation in the Andes of South Ecuador. PLoS One 14(8):e0221091
Haug I, Weiss M, Homeier J, Oberwinkler F, Kottke I (2005) Russulaceae and Thelephoraceae
form ectomycorrhizas with members of the Nyctaginaceae (Caryophyllales) in the tropical
mountain rain forest of southern Ecuador. New Phytol 165(3):923–936
Homeier JL, Bräuning A, Cumbicus NL, Hertel D, Matrinson GO, Spannl S, Veldkamp E (2013a)
Effects of nutrient addition on the productivity of montane forests and implications for the
carbon cycle. In: Bendix JB, Bräuning E, Makeschin A, Mosandl F, Scheu R, Wilcke W (eds)
Ecosystem services, biodiversity and environmental change in a tropical mountain ecosystem
of South Ecuador. Springer Verlag, Heidelberg, pp 315–329
Homeier JW, Gawlik FA, Peters J, Diertl T, Richter M (2013b) Plant diversity and its relevance
for the provision of ecosystem services. In: Bendix JB, Bräuning E, Makeschin A, Mosandl
128 I. Haug et al.
F, Scheu R, Wilcke W (eds) Ecosystem services, biodiversity and environmental change in a

tropical mountain ecosystem of South Ecuador. Springer Verlag, Heidelberg, pp 93–106
Homeier JW, Gradstein FA, Breckle SR, Richter M (2008) Potential vegetation and floristic com-
position of Andean forests in South Ecuador, with a focus on the RBSF. In: Beck EB, Kottke J,
Makeschin I, Mosandl F (eds) Gradients in a tropical mountain ecosystem of Ecuador. Springer
Verlag, Berlin, pp 87–100
Katoh K, Kuma K, Toh H, Miyata T (2005) MAFFT version 5: improvement in accuracy of mul-
tiple sequence alignment. Nucleic Acids Res 33(2):511–518
Kawahara A, An GH, Miyakawa S, Sonoda J, Ezawa T (2016) Nestedness in arbuscular mycor-
rhizal fungal communities along soil pH gradients in early primary succession: acid-tolerant
Fungi are pH generalists. PLoS One 11(10):e0165035
Kotilinek M, Hiiesalu I, Kosnar J, Smilauerov M, Smilauer P, Altman J, Dvorsky M, Kopecky M,
Dolezal J (2017) Fungal root symbionts of high-altitude vascular plants in the Himalayas. Sci
Rep 7:6562
Leal PL, Siqueira JO, Sturmer SL (2013) Switch of tropical Amazon forest to pasture affects taxo-
nomic composition but not species abundance and diversity of arbuscular mycorrhizal fungal
community. Appl Soil Ecol 71:72–80
Lee J, Lee S, Young JPW (2008) Improved PCR primers for the detection and identification of
arbuscular mycorrhizal fungi. FEMS Microbiol Ecol 65(2):339–349
Lekberg Y, Waller LP (2016) What drives differences in arbuscular mycorrhizal fungal communi-
ties among plant species? Fungal Ecol 24:135–138
Li XL, Gai JP, Cai XB, Li XL, Christie P, Zhang FS, Zhang JL (2014) Molecular diversity of arbus-
cular mycorrhizal fungi associated with two co-occurring perennial plant species on a Tibetan
altitudinal gradient. Mycorrhiza 24(2):95–107
Li XL, Zhang JL, Gai JP, Cai XB, Christie P, Li XL (2015) Contribution of arbuscular mycorrhizal
fungi of sedges to soil aggregation along an altitudinal alpine grassland gradient on the Tibetan
plateau. Environ Microbiol 17(8):2841–2857
Liu L, Hart MM, Zhang JL, Cai XB, Gai JP, Christie P, Li XL, Klironomos JN (2015) Altitudinal
distribution patterns of AM fungal assemblages in a Tibetan alpine grassland. FEMS Microbiol
Ecol 91(7):fiv078
Marín C, Aguilera P, Cornejo P, Godoy R, Oehl F, Palfner G, Boy J (2016) Arbuscular mycor-
rhizal assemblages along contrastingAndean forests of southern Chile. J Soil Sci Plant Nutr
16(4):916–929
Marín, C., P. Aguilera, , F. Oehl and R. Godoy (2017). Factors affecting arbuscular mycorrhizal
fungi of Chilean temperate rainforests. J Soil Sci Plant Nutr 17 (4): 966–984
Martinson GO, Corre MD, Veldkamp E (2013) Responses of nitrous oxide fluxes and soil nitro-
gen cycling to nutrient additions in montane forests along an elevation gradient in southern
Ecuador. Biogeochemistry 112(1–3):625–636
Mi XC, Swenson NG, Valencia R, Kress WJ, Erickson DL, Perez AJ, Ren HB, Su S, Gunatilleke
N, Gunatilleke S, Hao ZQ, Ye WH, Cao M, Suresh HS, Dattaraja HS, Sukumar MKP (2012)
The contribution of rare species to community phylogenetic diversity across a global network
of Forest plots. Am Nat 180(1):E17–E30
Moser G, Hertel D, Leuschner C (2007) Altitudinal change in LAI and stand leaf biomass in tropi-
cal montane forests: a transect shady in Ecuador and a pan-tropical meta-analysis. Ecosystems
10(6):924–935
Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, McGlinn D (2016) Vegan: community
Ecology package. R package version 2:4–1
Opik M, Moora M, Liira J, Zobel M (2006) Composition of root-colonizing arbuscular mycor-
rhizal fungal communities in different ecosystems around the globe. J Ecol 94(4):778–790
Opik M, Vanatoa A, Vanatoa E, Moora M, Davison J, Kalwij JM, Reier U, Zobel M (2010) The
online database MaarjAM reveals global and ecosystemic distribution patterns in arbuscular
mycorrhizal fungi (Glomeromycota). New Phytol 188(1):223–241
Pinos J, Studholme A, Carabajo A, Gracia C (2017) Leaf Litterfall and decomposition of Polylepis
reticulata in the Treeline of the Ecuadorian Andes. Mt Res Dev 37(1):87–96
Prada CM, Stevenson PR (2016) Plant composition associated with environmental gradients in
tropical montane forests (Cueva de Los guacharos National Park, Huila, Colombia). Biotropica
48(5):568–576
Rodriguez-Echeverria S, Teixeira M, Correia S, Timoteo R, Heleno M, Opik M, Moora M (2017)
Arbuscular mycorrhizal fungi communities from tropical Africa reveal strong ecological struc-
ture. New Phytol 214(1):487–487
Rosendahl S (2008) Communities, populations and individuals of arbuscular mycorrhizal fungi.
New Phytol 178(2):253–266
Senes-Guerrero C, Schussler A (2016) A conserved arbuscular mycorrhizal fungal core-species
community colonizes potato roots in the Andes. Fungal Divers 77(1):317–333
Wright SJ (2002) Plant diversity in tropical forests: a review of mechanisms of species coexis-
tence. Oecologia 130(1):1–14
Yang W, Zheng Y, Gao C, Duan JC, Wang SP, Guo LD (2016) Arbuscular mycorrhizal fungal com-
munity composition affected by original elevation rather than translocation along an altitudinal
gradient on the Qinghai-Tibet Plateau. Sci Rep 6:36606
Zobel M, Opik M (2014) Plant and arbuscular mycorrhizal fungal (AMF) communities - which
drives which? J Veg Sci 25(5):1133–1140
Chapter 7
Nesting Ecology of the Tucuman Amazon
(Amazona tucumana) in the Cloud Forest
of Northwestern Argentina
Luis Rivera and Natalia Politi
7.1 The Tucuman Amazon
Parrots (Order Psittaciformes) are one of the largest groups of birds, with 398 spe-
cies distributed in tropical and subtropical areas of the world and are among the
most threatened avian orders with 30% of the species endangered at global level
from which 15% are Neotropical (BirdLife International 2018). Parrots of the genus
Amazona are 27–31 species broadly distributed in Central and South America
(México to Argentina) and the Caribbean that inhabit from lowlands to highlands,
and from dry to wet habitats (Juniper and Parr 1998). All parrot species of this genus
were heavily traded in the international pet market, are currently included in the
CITES appendix I and II, and 18 species are categorized as threatened (Russello and
Amato 2004, BirdLife Internationa 2018).
The Tucuman Amazon (A. tucumana) is one of the four Amazona parrot species
from Argentina (Narosky and Yzurieta 2003), which was shown to be sister species
of the Spectacled Amazon (Amazona pretrei) (Russello and Amato 2004; Rocha
et al. 2014). The group formed by Tucuman Amazon, Spectacled Amazon, and
Vinaceous-breasted Amazon probably is a basal group for all the other species from
the genus Amazona (Russello and Amato 2004). Recent genetic studies have allowed
assessing allopatric models to better explain the diversification process of Spectacled
Amazon and Tucuman Amazon, to determine genetic variability, effective popula-
tion size, and divergence time of lineage separation (Rocha et al. 2014). Haplotype
diversity in Tucuman Amazon (DH = 0.41) is smaller than in Spectacled Amazon
(DH = 0.68), the effective population size of females for Tucuman Amazon was
approximately 94,000 individuals, and the divergence time of lineage separation
was estimated to be approximately 1.3 million years ago, which corresponds to the
lower Pleistocene (Rocha et al. 2014). The genetic information suggest allopatry by
L. Rivera (*) · N. Politi

Insituto de Ecorregiones Andinas CONICET/Universidad Nacional de Jujuy,
S.S. de Jujuy, Jujuy, Argentina

https://doi.org/10.1007/978-3-030-57344-7_7
132 L. Rivera and N. Politi
the dispersion/founder effect model (peripheral isolation) as the more likely process
to explain the diversification of these two Amazon species (Rocha et al. 2014).
As most Amazon species Tucuman Amazon sexes are alike (Forshaw 1977).
Tucuman Amazon total body length is 31 cm, weights 280 g, and has a short tail.
The head of mature individuals of Tucuman Amazon shows a front patch of red
feathers, and juveniles show yellow feathers around this patch. Also the primary
covers in the wings form a patch of red feathers (speculum) visible in flight. The tip
of the wing feathers is purple blue and the body feathers are strongly emarginated
with dark grey (Fig. 7.1).
Tucuman Amazon was described by Cabanis in 1885 since then, little informa-
tion has been obtained about its natural history and ecology and anecdotical men-
tions about its biology are known (Hoy 1968; Orfila 1938; Wetmore 1926). Diet of
the Tucuman Amazon includes seeds of Alnus acuminata, Podocarpus parlatorei,
fruits of Cedrela lilloi, and flowers of Erythrina falcata (Fjeldså and Krabbe 1990;
Juniper and Parr 1998; Low 2005). In autumn and winter the species can feed on
groves and crops (Low 2005) when it descend to lower elevations even visiting
fruiting trees in cities (Fjeldså and Krabbe 1990, Low 2005, Juniper and Parr 1998).
Fig. 7.1 Tucuman Amazon and plumage characteristics

7 Nesting Ecology of the Tucuman Amazon (Amazona tucumana) in the Cloud Forest… 133
Fig. 7.2 Distribution of the Southern Yungas or Tucumano-Boliviano forest. Study sites include
El Rey National Park and Santa Barbara
The Tucuman Amazon is a threatened species (BirdLife International 2018),

with a small geographic range restricted to the narrow strip of montane forest on the
eastern slopes of the Andes (the Yungas) from south-eastern Bolivia to northwestern
Argentina (Fjeldså and Krabbe 1990) (Fig. 7.2). The Yungas is a biogeographic
province belonging to the Amazonic domain that spread along the eastern slopes of
the Andes from Venezuela to northwestern Argentina encompassing an elevation
range between 400 and 3500 m a.s.l. (Cabrera and Willink 1980). Climate is highly
seasonal with rainfall concentrated (75–80%) during the summer (i.e., November to
March) (Mendoza 2005). The southernmost limit of Neotropical montane forests is
known as Southern Yungas or Tucumano-Boliviano forest (Fig. 7.2, Tortorelli 1956,
Hueck 1978, Cabrera 1994). In Bolivia, the Southern Yungas goes from 18° S
through Tarija, Chuquisaca, and Santa Cruz de la Sierra departments and in
Argentina it extends to 29° S, through Catamarca, Tucumán, Salta, and Jujuy prov-
inces. Southern Yungas are distributed discontinuously by 1200 km forming iso-
lated patches of 50 km wide covering an area of approx. four million hectares in
Argentina (Brown 1995), and three million hectares in Bolivia (Ibisch and Mérida
2003). Along the Southern Yungas there is a latitudinal gradient (14° S–29° S) with
a marked loss of plant and animal species richness as latitude increases (Brown
et al. 2001). In Argentina, the Southern Yungas represents 2% of the terrestrial land
but holds the higher level of endemism of flora and fauna of the country (Kappelle
and Brown 2001). In the Southern Yungas the breeding habitat of Tucuman Amazon
is the cloud forest found between 1500 and 2200 m asl that has an annual rainfall
averages of 1300 mm and annual temperature averages of 11.7 °C (Brown et al.
2001; Arias and Bianchi 1996).
7.2 T
he Reproductive Habitat of Tucuman Amazon:
The Cloud Forest
Of the eight parrot species that inhabit the Southern Yungas only the Tucuman
Amazon and the Red-mitred parakeet are sympatric in the cloud forest. The remain-
ing six species occupy other forest types along the elevation gradient of the moun-
tain ranges (Politi and Rivera 2005). The forest structure in one old-growth cloud
forest in El Rey National Park (Rivera 2011) is characterized with a density of
406 ± 136 trees/ha >10 cm DBH and a basal area of 45.2 ± 21.4 m2/ha. The size
distribution for trees >10 cm DBH has an inverse J form typical of multispecies
stands with larger numbers of small stems and a diminishing density of larger ones
as DBH increase (Fig. 7.3). Species richness of trees >10 cm DBH is of 19 species
(Table 7.1). Twelve tree species were dominant with 97.3% of the individuals
>10 cm DBH and 7 tree species showed very low densities. Dominant tree species
are Blepharocalyx salicifolius, Podocarpus parlatorei, Cinnamomum porphyria,
250
193
200
stems / ha
150
100 83
47
50 34
22
16
9 8 3 2 3 1 1 0 1
0
DBH
Fig. 7.3 Forest structure in DBH size classes of the stems >10 cm for all tree species. Number
above bar is the total number of stems
Table 7.1 Tree species composition of live trees >10 cm DBH and snags in El Rey National Park,
Argentina. Tree density, DBH, height, and basal area expressed as mean ± SE
Frequency Density (N/ DBH Height Basal area (m2/
Tree species (%) ha) (cm) (m) ha)
Myrcianthes pseudomato 21.7 90.0 ± 15.6 27.9 ± 2.6 13.5 ± 0.4 3.42 ± 0.6
Blepharocalyx 16.6 69.0 ± 12.1 38.0 ± 3.3 19.8 ± 1.0 12.44 ± 3.7
salicifolius
Allophylus edulis 14.0 59.0 ± 12.9 28.7 ± 3.5 12.6 ± 0.4 1.62 ± 0.4
Podocarpus parlatorei 8.9 37.0 ± 15.1 42.8 ± 4.8 17.4 ± 0.6 7.0 ± 2.4
Cinnamomum porphyria 6.5 27.0 ± 7 28.1 ± 3.6 21.7 ± 0.5 4.56 ± 1.3
Myrrhinium 6.3 26.0 ± 7.2 19.1 ± 1.1 12.2 ± 1.1 0.56 ± 0.2
atropurpureum
Snags1 6.0 25.0 ± 5.9 33.1 ± 5.6 – 3.08 ± 1.1
Cedrela lilloi 4.3 18.0 ± 4.8 45.5 ± 4.9 24.4 ± 2.2 8.18 ± 2.8
Prunus tucumanensis 4.1 17.0 ± 6.2 26.0 ± 2.3 12.4 ± 0.9 1.0 ± 0.4
Juglans australis 3.4 14.0 ± 2.6 42.4 ± 4.9 23.0 ± 1.8 2.34 ± 0.6
Sambucus peruviana 1.9 8.0 ± 3.9 17.0 ± 1.4 – 0.18 ± 0.1
Ilex Argentina 1.7 7.0 ± 3 21.7 ± 1.8 14.2 ± 2.1 0.26 ± 0.1
Myrcianthes sp. 1.7 7.0 ± 2.6 15.9 ± 2.0 12.2 ± 1.1 0.14 ± 0.1
1
Snags represent standing dead trees of different tree species, given that after trees die it is difficult
to identify.
Cedrela lilloi, Juglans australis, Myrcianthes pseudomato, Allophylus edulis,

Myrrhinium atropurpureum, Prunus tucumanensis, Sambucus peruviana, Ilex
argentina, and Myrcianthes callicoma. Mean height of trees >10 cm DBH was
16.5 ± 5.9 m (Range = 6–34 m). The forest shows three well defined stratum with a
continuous canopy formed by trees between 10 and 20 m height, some emergent
trees higher than 20 m, and a lower stratum of smaller trees between 0 and 10 m
height forming the subcanopy, and an understory of saplings, ferns, shrubs, and
herbs (Rivera 2011).
Analysis of availability of Tucuman Amazon habitat at the eco-regional level
using distribution models indicates that conditions for supporting Tucuman Amazon
habitat were distributed across the entire north–south gradient of the Southern
Yungas, but less than half of the Southern Yungas (~46,000 km2, 42%) was pre-
dicted to provide suitable non-breeding habitat, with slightly more occurring in
Bolivia (~26,000 km2) than in Argentina (~20,000 km2) (Pidgeon et al. 2015).
Potential breeding habitat is more limited than non-breeding habitat since condi-
tions that could support breeding habitat were found in only ~21,000 km2, or 19%,
of the Southern Yungas with three times as much breeding habitat found in Argentina
(~15,400 km2) than in Bolivia (~5600 km2) (Pidgeon et al. 2015). In Bolivia, poten-
tial breeding habitat existed only in very narrow belts and disjunct patches along
ridgetops, whereas in Argentina potential breeding habitat patches were larger, with
a more regular shape (Pidgeon et al. 2015). Yet only 15% (3134 km2) of the breeding
habitat is currently within protected areas, and only ~6%, all in Argentina, is strictly
protected.
From the conservation side the main factors affecting the Tucuman Amazon are
natural habitat destruction and fragmentation, as well as the illegal trafficking of
nestlings and adults (Martinez and Prestes 2008; Rivera et al. 2007). Tucuman
Amazon has a small population size, with population estimates have identified
approximately 8000 individuals, 80% of which are in Argentina in the northern and
central distribution (Rivera et al. 2007; Rivera et al. 2010). The low genetic vari-
ability observed in this species (DH = 0.41) may be mainly the result of its evolu-
tionary history, due to the founding event in the early formation of this species
(Rocha et al. 2014). Furthermore, perhaps in recent centuries, it could have also
been influenced by the different negative impacts of human actions, making the spe-
cies more genetically vulnerable. Thus, conservation efforts for the Tucuman
Amazon should prioritize the northern and central regions of Southern Yungas in
Argentina, where 80% of the remaining Tucuman Amazon population occurs, with
high priority given to the creation of public and private reserves (Rivera et al. 2007).
Furthermore, this species may require actions aimed at increasing the population
size, such as eliminating nest poaching and retaining nest and food trees in forestry
operations (Rivera et al. 2012; Rivera et al. 2013). Anecdotal evidence of Tucuman
Amazon kept as pets in villages in Bolivia suggests that local customs and attitudes
toward parrots are a challenge to the conservation of this species (Tella et al. 2013).
Finally, the low genetic variability and genetic distinctiveness in Tucuman Amazon
can be seen as very valuable information given the categorization of this species on
the IUCN Red List as Vulnerable (BirdLife International 2018).
Information on habitat requirements allows predictions to be made on the ability
of Tucuman Amazon to adapt to disturbed habitats (Saunders et al. 1982) and to
develop effective strategies for conservation and management of threatened species
(Renton 2000). It is necessary to conduct studies on mature or old-growth forests
(sensu Hunter Jr. and White 1997) that set a reference for future comparisons against
habitat modification. Detailed knowledge of breeding biology is necessary for
understanding variation in avian reproductive strategies, because it provides critical
natural history data that are useful for generating new hypotheses and testing old
ones (Auer et al. 2007). The lack of information on the natural history, ecology, and
demography of many parrot species precludes an assessment of the mechanisms
that regulate population dynamics (Koenig 2001). Prior to this study there were no
specific studies on the reproductive biology of Tucuman Amazon—there was one
record of a nest, found in Chuquisaca, Bolivia, with a female incubating four eggs
in January (Bond and Meyer de Schauensee 1943); a note that Tucuman Amazon
nests in large trunks of Alnus or Podocarpus trees between January and March in
forests located at an elevation of 2600 m (Juniper and Parr 1998); and a comment
that Tucuman Amazon places its nests at elevations between 900 and 2100 m (Low
2005). In captivity, Tucuman Amazon brood mean size was 3.4 eggs with a develop-
ment period of the chicks lasting between 60 and 67 days (Low 2005).
7.3 Nesting Habitat Requirements of Tucuman Amazon
To characterize the nesting habitat of Tucuman Amazon we assessed cavity avail-

ability, reuse, and spatial pattern of nests in an old-growth forest in the central sector
of the Southern Yungas of northwestern Argentina, on the eastern slopes of the
Sierras Subandinas Centrales or Sistema de Santa Bárbara—a mountain
range ~ 100 km long, between the Cordillera Oriental to the west and the Chaco
plain to the east. Within the central sector of the Southern Yungas, we focused on
two areas where we knew of active nests: (1) El Rey National Park, Salta Province
(24° 43’S, 64° 38’W, 44,000 ha) and (2) Portal de Piedra Private Reserve, Jujuy
Province (24° 05’S, 64° 26’W, 400 ha). Both areas are strictly protected and con-
trolled, no people live within the reserves and, to the best of our knowledge, there is
no poaching. The study was conducted within accessible areas only. These com-
prised ~170 ha in Portal de Piedra Private Reserve and 45 ha in El Rey National
Park. Elevation of these areas is between 1450 and 2100 m above sea level.
We carried out fieldwork from 2004 to 2009. Nest searches of Tucuman Amazon
were conducted daily during egg-laying and incubation periods (December to mid-
January). We found nests by following males to the nest area and locating the cavity
when the female left the nest to be fed by the male (González Elizondo 1998). Nest-
cavity characteristics were determined by climbing (Perry 1978) and measuring
cavity dimensions (Fig. 7.4). We recorded the following nest characteristics: (1)
height from the ground to the cavity entrance; (2) size of cavity entrance (horizontal
and vertical); (3) internal diameter at the cavity floor; (4) internal cavity depth from
Fig. 7.4 Nest monitoring of Tucuman Amazon and nest entrance in Podocarpus parlatorei tree
cavity entrance to the floor; (5) trunk or branch diameter at cavity entrance; (6) tree
diameter at the cavity floor; (7) tree diameter at breast height (DBH); (8) tree height;
(9) tree species; (10) cavity origin (excavated or decayed); (11) cavity location (tree
trunk, primary branch, secondary branch, or third branch); (12) tree status (alive or
dead); and (13) tree location (latitude and longitude). Our characterization of nest-
ing requirements is based on 44 active nests found in El Rey National Park. To
assess tree-cavity availability we conducted sampling during the non-breeding sea-
son (April–August 2007 and 2008) when many trees are leafless. We used distance
sampling methodology to estimate the density of suitable cavities. We performed 20
variable-width, random direction, 300-m long transects that were at least 150 m
apart. We measured the perpendicular distance from the central line of the transect
to each detected cavity. We only considered a cavity to be suitable if it had a hollow
chamber surrounded by sound wood (not collapsing wood), accessed by entrance
holes with a floor to support an incubation chamber and a roof to provide overhead
protection, a minimum diameter entrance of 5 cm, an internal diameter of at least
15 cm (minimum cavity dimensions suggested for Amazona species of similar
body-size to Tucuman Amazon; Snyder et al. 1987, Enkerlin-Hoeflich 1995), a min-
imum cavity height from the ground of 2 m, cavity depth from 0 to 200 cm, and a
tree DBH of 30 cm (minimum dimensions observed for Tucuman Amazon in
another site; Rivera 2011). Therefore, the minimum characteristics used to deter-
mine a suitable cavity were in the range of the cavities used for nesting. We used a
tree-peeper (Richardson et al. 1999) to estimate or measure the following cavity and
tree characteristics: (1) height from the ground to the cavity entrance using the grad-
uated metric scale in the telescopic rod of the tree-peeper; (2) cavity entrance diam-
eters (horizontal and vertical); (3) internal diameter at the cavity floor; (4) internal
cavity depth from cavity entrance to the floor. Cavity entrance bearing was mea-
sured with a compass, tree DBH was measured with metric tape and tree height with
a hypsometer. Due to tree-peeper limitations we only inspected suitable cavities
below 15 m (Richardson et al. 1999).
We used Manly’s selection index to compare use of cavities as nest-sites in dif-
ferent tree species with the availability of cavities in those tree species (Krebs 1999;
Manly et al. 2002). We calculated a selection coefficient and the 95% confidence
interval for the categorical nest-site variable (tree species). Coefficients greater than
1.0 indicated preference, while values less than 1.0 indicated avoidance (Krebs
1999; Manly et al. 2002; Aitken and Martin 2004). Selection coefficients were
tested for significance using the log-likelihood ratio (G-test, Manly et al. 2002).
Frequencies of nest cavities in different categories (tree species, origin, and cavity
location) were compared with a χ2 test. We determined the number of cavities reused
by Tucuman Amazon for nesting over several breeding season. We define reuse as
those cases where the same cavity was used in more than 1 year (Berkunsky and
Reboreda 2009) and a cavity was considered to be used if it contained eggs or chicks.
Cavity density were determined following line transect guidelines and modelled
using the software Distance 5.0 (Buckland et al. 2001; Thomas et al. 2006). The
model with the lowest Akaike’s Information Criterion (AIC) was selected (Burnham
and Anderson 2002). The adequacy of the selected model for the perpendicular
distances was assessed using a Kolmogorov–Smirnov test (Buckland et al. 2001).

We determined the average nest density by calculating the mean of the number of
nests found in 45 ha during the four breeding seasons. We used the Spatial Analyst
tool of ArcGIS to determine distances to evaluate spacing among all simultaneously
active nests and using the locations of all trees used as nest-sites over the four-year
study (Salinas-Melgoza et al. 2009). Each nest-tree location was considered only
once for the analysis regardless of how many times the tree was reused as a nest-
site. In addition, for each nest-tree used by parrots in any year we calculated the
distance to the nearest neighbouring tree that had been used as a nest-site in any
year. We compared the nearest neighbour distances for active nests among breeding
seasons with a Kruskal–Wallis test. To determine whether the spacing of breeding
pairs differed from the distribution of all nest-trees we compared distances among
active breeding pairs with distances among nest-tree for all years combined, using a
Mann–Whitney U test. Using a paired Wilcoxon test, we further evaluated the influ-
ence of conspecifics on the spacing of parrot nests to compare the distance to the
nearest active nest vs. the distance to the nearest potential unoccupied nest-tree for
each parrot nest active in the 2008–2009 breeding season. We restricted this analy-
sis to the 2008–2009 datasets, which had the most complete record of potential
nest-trees, to avoid overduplication of distance values between years (Salinas-
Melgoza et al. 2009). Distance values obtained previously among all nests were
used to assess the spatial pattern of nest-bearing trees and active nests (Salinas-
Melgoza et al. 2009) with the Average Nearest Neighbor Distance tool from ArcGIS
(Mitchell 2005). All the values are expressed as mean standard deviation (SD)
unless otherwise specified. We set the significance level of statistical tests at
P = 0.05.
We recorded 44 Tucuman Amazon nesting attempts in 37 nest-trees, 30 during
incubation, and seven during brooding. Most Tucuman Amazon nests occurred in
live trees (95%) of six species, and only 5% were in nest cavities in snags. There
was a significant difference in the frequency of tree species used for nesting
(χ2 = 27.6, P < 0.001), with most nest cavities in Blepharocalyx salicifolius (59.5%,
22 out of 37), followed by Juglans australis (13.5%, 5), Podocarpus parlatorei
(8.5%, 3), Cinnamomum porphyria (5.4%, 2), Cedrela lilloi (5.4%, 2), and
Myrcianthes pseudomato (5.4%, 1), and 5.4% (2) of the nests were found in snags.
Compared to the availability of cavities in different tree species, B. salicifolius,
J. australis, and C. lilloi were used significantly more than expected (G6 = 91.6,
P < 0.01). Most nests were found in decay-originated tree cavities (95%, n 5 35),
compared to nests excavated (5%, n = 2) (χ2 = 21.5, P < 0.001). Cavity location was
predominantly in primary branches (43%, n = 16), followed by main trunk (32%,
n = 12), secondary (16%, n = 6), and tertiary branches (8%, n = 3) (χ2 = 8.9,
P < 0.03). Most nests (92%) were found in trees with a DBH >60 cm. Average nest-
tree DBH was 89.9 ± 26.9 cm, cavities were located on average at 14.4 ± 3.9 m
above the ground, had a horizontal entrance diameter of 13.3 ± 4.5 cm, and a depth
of 38.2 ± 38.6 cm. Nests of Tucuman Amazon were found at an elevation range
between 1470 and 1710 m a.s.l. Six of the 37 cavities (16%) were reused. One nest-
cavity was used in three breeding seasons and five were used twice. One nest-tree
had two nest cavities used in separate years. Annual mean nest density in the study
area was 0.24 ± 0.04 nests ha-1 (n = 4). In two breeding seasons (2006–2007 and
2007–2008) the spatial pattern of active nests was dispersed (Z = 3.8 observed dis-
tance/expected distance = 1.6 and Z = 2.7, observed distance/expected distance = 1.4,
respectively), random in 2005–2006 (Z = −0.3, observed distance/expected dis-
tance = 0.9), and intermediate between random and dispersed in 2008–2009
(Z = 1.8, observed distance/expected distance = 1.3). The spatial pattern of all nest-
trees used in the four breeding season was clustered (Z = 4.2, observed distance/
expected distance = 0.6). There was no significant difference in the nearest mean
distance among active nests among breeding seasons (H = 1.4, P = 0.69). For all
years combined, the distance to the nearest active nest was significantly greater than
the distance between all trees used as nests (W = 520, P < 0.001). Nesting pairs of
Tucuman Amazon were separated by 144.1 ± 152.8 m (range 5 12–674 m), while
potential nest-trees were 66.0 ± 55.4 m apart (range = 12–252.4 m). Furthermore,
for nests located in 2008–2009, the same pattern prevailed, a nesting pair being
significantly farther from the nearest neighbouring pair (138.1 ± 165.3) than from
the nearest available nest-tree (53.5 ± 19.5, Z = 2.29, P = 0.02). The estimation of
the density of available suitable cavities for nesting was 4.6 cavities ha-1 (95%
CI = 3.1–7.0 cavities ha−1), a coefficient of variation of 20%, and an effective detec-
tion width of 9.1 m (95% CI = 7.5–11.1 m).
Nest cavities of Tucuman Amazon were higher, but shallower than nests of other
Amazona species from the lowlands. This can probably be related to a lower rate of
decomposition related to specific tree and sapwood characteristics that compart-
mentalize decaying wood or are very resistant to fungal decay (Shigo 1984) or to a
lower temperature that retards decomposition rates (Politi et al. 2010). As expected,
we found that nest cavities of Tucuman Amazon have a larger internal diameter than
lowland parrots nests which can be an advantage to maximize nest space and ther-
mal insulation (Joy 2000) in a high elevation breeding habitat where low tempera-
tures are reached. Tucuman Amazon selects three tree species for 83% of nests.
Tucuman Amazon sited a high number (95%) of nests in living trees. We reported
0.24 nests ha−1 (i.e. one nest every 4 ha) of Tucuman Amazon. Active Tucuman
Amazon nests have a dispersed distribution at a mesoscale of 45 ha of the study site.
Recently it has been suggested that behavioural spacing requirements of nesting
parrots may limit breeding densities and restrict management strategies to increase
numbers of nesting pairs within protected areas (Salinas-Melgoza et al. 2009).
Tucuman Amazon shows significantly shorter distances among nearest nest-trees
than distance among nearest breeding pairs in a breeding season suggesting that
breeding pairs influence spacing of conspecifics. The fact that we found a stable
number of nests during the four breeding seasons for the fixed area under study, that
the mean distance to the nearest breeding pairs was similar in every breeding sea-
son, and that the spatial pattern of active nests is mainly dispersed, suggests that
spacing due to territorial behaviour could be limiting breeding pair density. This
limitation may occur despite the availability of suitable cavities. We estimated that
approximately 16 suitable cavities were available for each breeding pair (0.25
breeding pair ha−1 and four suitable cavities ha−1). Tucuman Amazon occupied 5%
of the suitable cavities available. The relatively low percentage of reuse of nest cavi-
ties for Tucuman Amazon may show a strategy for avoiding predation, since preda-
tion is the main cause of nest-loss for the species (Rivera 2011). Nests placed in new
cavities have lower predation rates compared to nests in previously used cavities
(Brightsmith 2005). The frequent shift of nesting sites to avoid predation is proba-
bly due to the high density of suitably cavities available for nesting in this
mature forest.
7.4 Reproductive Biology of the Tucuman Amazon
We identified a total of 94 breeding attempts for which we recorded nesting success

or causes of nesting failure. Data on reproductive output were gathered from 86
nests in which complete clutches were laid (defined as consecutive visits to active
nests during which the number of eggs did not change, but the nest remained active).
Nests >15 m above the ground were visually inspected using climbing equipment to
reach the nest (Perry 1978). During the incubation period nests above 15 m were
inspected once only. We monitored contents of nests when females left the nests to
be fed by males (Fig. 7.5). When possible, we recorded the date clutches that were
initiated, and determined incubation period and length of the entire reproductive
effort for each nest. We estimated the following parameters: clutch-size (i.e. number
of eggs laid per nesting attempt), brood-size (i.e. number of chicks per nest), hatch-
ing success (i.e. proportion of eggs present in the nest at the end of incubation that
hatch), fledging success (i.e. proportion of nestlings that fledge), and nesting suc-
cess (i.e. proportion of nests producing at least one fledgling). To allow comparison
with other studies, reproductive output was expressed as fledglings per successful
nest and fledglings per laying female (including successful and unsuccessful
females). We considered a nest successful when it produced at least one fledgling.
We considered a nest abandoned when it contained eggs but no adults were recorded
in more than 1 h of observation on two successive visits. A nest was considered
depredated when all of the eggs disappeared before hatching, nestlings disappeared
before reaching 47 days of age, or remains of eggs or nestlings were found in an
otherwise empty nest. We used 47 days as the cut-off age because fledglings attain
maximum primary feather length at ~50 days and we estimate that before 47 days
wings are too short to sustain flight (Rivera 2011). Nests were considered lost by
starvation when nestlings were found dead with an empty crop. We examined pro-
ductivity at different stages of the nesting cycle on the basis of all laying pairs.
We used Mayfield’s (1975) method to calculate daily nest-survival rate because
this method avoids the bias introduced when nests are found at different stages of
the nesting cycle. Because intervals between visits to nests varied we used a
maximum-likelihood estimate modification of the Mayfield method (Johnson 1979;
Krebs 1989). We calculated the variance of the Mayfield estimator according to
Johnson (1979) to make comparisons with the program Contrast (see below). We
estimated the daily nest-survival rate (DSR) during the incubation (28 days) and
Fig. 7.5 Nests of Tucuman Amazon inspected containing eggs, hatchlings, chicks, and fledglings
nestling (1–50 days after hatching) stages and multiplied the DSR of nests during
incubation with the DSR of nests during the nestling stage to obtain the finite sur-
vival rate. The program Contrast (Hines and Sauer 1989) was used to compare DSR
among periods of the nesting cycle and among breeding seasons. We were not able
to use dataset from the 2004–2005 breeding season because DSR had no associated
variance. We assumed that a juvenile had fledged when it was absent from a nest at
47 days after hatching. Disappearances at earlier ages were considered to be deaths.
In order to understand whether clutch-size of Tucuman Amazon is higher or lower
than expected for the body-size of the species, we assessed whether adult Tucuman
Amazon body-mass (~280 g, Low 2005) and clutch-size followed the allometric
equation (y = 2.2 + 5.5 exp.[−0.006x]) developed by Masello and Quillfeldt (2002)
for other species of parrots, where y is clutch-size is and x is body-mass. Additionally,
we performed a regression analysis of clutch-size as a function of body-mass for all
species of Amazona parrots for which we could obtain data, to test if clutch-size of
Tucuman Amazon lies inside the 95% confidence interval of the allometric
relationship.
In general, in parrots of the genus Amazona the female brood the eggs and the
male provide food to the incubating female (Low 2005). In Tucuman Amazon the
male provide food to the incubating female twice a day, early in the morning and
late in the afternoon. Females pass less than 10 min outside the nest when they
receive food from males (Rivera et al., unpublished). Females leave the nest 15 days
after hatching of the chicks. After the female stop brooding both parents feed the
nestlings three times per day adding a visit to the nest at noon (Rivera et al., unpub-
lished). Recently hatched chicks weight 12 g, coincident with records in captivity
(Low 2005). Length of wing, tail, tarsus, and bill were lower for fledglings than for
adults if we compare our data with those of Forshaw and Cooper (1989) for 19
adults Tucuman Amazon. Our measurements were done 1–2 days before fledging
suggesting that fledglings are in development and growing when leaving the nests.
Weight of fledglings (about 265 g) is similar to chicks of 70 days raised in captivity
(260–269 g) and lower to the weight of adults (280 g) (Low 2005). Because sexes
are alike in Tucuman Amazon the only form to determine sex is by ADN analysis.
Our analysis using two conserved genes (chromo-helicase-DNA-binding) located in
the sexual chromosome (Griffiths et al. 1998) showed no significant differences
(χ2 = 1.8, gl = 1, P = 0.18) (42.5% males and 57.5% females for 40 chicks of 12
nests). The length of the incubation period (mean 28.33 ± 0.58 days, range = 28–29),
nestling period (49.7 ± 1.1 days, range = 49–51), and overall nesting period of
Tucuman Amazon were determined from data from three nests for which date of
laying of the first egg, date of hatching of the first egg, and date of fledging of the
first chick were known with certainty. Clutch-size ranged from one to five eggs
(mean 3.6 ± 1.0, mode = 4, n = 86). Clutches of one are most likely to be complete
clutches because we did not find any evidence of partial loss or predation. The
smallest overall clutch-size was recorded during the 2008–2009 breeding season,
and differed significantly from mean annual clutch-size of three other breeding sea-
sons (H4 = 9.6, d.f. = 4, P = 0.03). Overall hatching success (proportion of eggs that
hatch) was 0.77, with the lowest rate (0.54), recorded in 2006–2007. The number of
fledglings per successful nest (overall mean 3.2 ± 0.2, n = 51) differed among years
with lowest values in 2005–06 and 2006–07 (H4 = 8.3, d.f. = 4, P = 0.05). The
number of fledglings per laying female was 2.3 ± 0.8 (n = 86) and differed signifi-
cantly among years, with the lowest value during the 2006–2007 breeding season
(H4 = 11.9, d.f. = 4, P = 0.01). The daily survival rate during the incubation period
did not differ significantly among years. The daily survival rate during the nestling
period and the entire nesting cycle differed significantly among breeding seasons.
The highest finite survival rate for the nestling period and the nesting cycle was in
the 2005–2006 breeding season and the lowest finite survival rate was in 2006–2007.
Over the nesting cycle, the primary causes of nest-loss were predation (16%) and
abandonment (12%). Nine nests were depredated during the incubation period and
six during the nestling period. Abandonment was the main cause of failure during
the incubation period, followed by predation, whereas predation was the main cause
of failure during the nestling period. Nesting failure was higher during incubation:
22 nests were lost during this period (63% of all losses), whereas 13 nests were lost
during the nestling period (37%). Eleven nests were abandoned during incubation,
whereas none was abandoned during the nestling period. In two breeding seasons
(2006–2007 and 2008–09) predation was the primary cause of nest-loss, in one
breeding season (2005–2006) the main cause was abandonment, and in one season
(2007–2008) predation and abandonment were equal causes. Predation of four nests
<100 m apart might be attributed to Black-capped Capuchin monkeys (Cebus
apella) that were observed on a tree in which we had confirmed active nests the
previous day and found empty the day after the monkeys were observed. In two
nests with advanced nestlings we found bundles of plucked feathers at the entrance
to the nesting cavity and on the ground, suggesting predation by mammals. Most
nesting failures (n = 13) occurred during the 2006–2007 breeding season, with
fewer failed attempts per year in the other breeding seasons.
With a mean body-mass of 280 g, the predicted clutch-size of Tucuman Amazon
would be 3.2 eggs based on the allometric equation of Masello and Quillfeldt
(2002). The 95% confidence interval for the expected clutch-size of the allometric
equation for 14 other species of Amazona parrots is 2.94–3.43, which is signifi-
cantly lower than the mean clutch-size of 3.6 that we observed for the Tucuman
Amazon. The Tucuman Amazon has high rates of nesting success, large clutches,
and a large number of fledglings per laying female. These results differ from trends
observed in other bird species that tend to shift to a slower life-history strategy with
increased elevation (Sandercock et al. 2005a, b). There are several alternative expla-
nations for our finding that the Tucuman Amazon has the second largest clutch-size
among Amazona parrots. Firstly, the health (or nutritional state) of female birds,
which is strongly related to availability of food, might affect both the number and
quality of the eggs laid (Lack 1954; Martin 1987); secondly, birds with higher rates
of nesting success may be expected to lay larger clutches (Skutch 1985); thirdly,
clutch-size is inversely related to population density (Ricklefs 1980), which, in the
case of nesting Tucuman Amazon, may be low; and lastly, the large clutch-size may
be evidence of a life-history trade-off in which increased productivity is compen-
sated for by lower survival of juveniles or adults (Bears et al. 2009). The overall
mean number of fledglings per breeding female and the overall nesting success for
Tucuman Amazon calculated as the maximum-likelihood estimator are the second
highest values reported for Amazona parrots. Likewise, the Tucuman Amazon has
the lowest loss of initial reproductive investment among Amazona parrots that have
been studied. The probability of nesting success varied significantly among years at
the combined nestling stages and showed significant differences among years in
nesting success over the entire nesting cycle. Inter-annual variability in the avail-
ability of key food items may influence productivity and nesting survival of Tucuman
Amazon, as is the case for the Lilac-crowned Parrot (Renton and Salinas Melgoza
2004), the Austral Parakeet (Enicognathus ferrugineus) (Díaz et al. 2012), and for
many other bird species (Newton 1980).
We recorded 11 nests with intact eggs that were abandoned, although we did not
check for infertility or diseases. Low temperatures (e.g. 8C minimum on rainy days)
in the cloud forest might make Tucuman Amazon more susceptible to loss of eggs
through embryonic chilling (Stoodley and Stoodley 1990), especially when incubat-
ing adults are disturbed. Factors that could explain abandonment are predation of
incubating adults or food scarcity, which might lead females to end incubation
because they cannot fulfil the high energy requirements of incubation. Predation
and abandonment of nests accounted for 28% of all losses of nests in the Tucuman
Amazon. Because Tucuman Amazon inhabit the cloud forest between 1400 and
2200 m above sea level, where richness and abundance of predators are expected to
be lower than at lower elevations (Skutch 1985), we expected to observe a lower rate
of nest predation than for other Amazona species of lowlands. Although there are no
snakes—one of the most important predators of other Amazona parrots (Enkerlin-
Hoeflich 1995; Renton and Salinas Melgoza 2004; Berkunsky et al. 2011)—in
cloud forests of the Southern Yungas, other predators present include Black-capped
Capuchins, at least four species of raptor (Barred Forest-Falcon, Micrastur ruficol-
lis; White-rumped Hawk, Parabuteo leucorrhous; Roadside Hawk, Buteo mag-
nirostris; Bicoloured Hawk, Accipiter bicolor) and three other potential nest
predators: two mammals, the Tayra (Eira barbara) and the Lesser Grison (Galictis
cuja), and the Plush-crested Jay (Cyanocorax chrysops). We frequently observed all
of these species in the breeding habitat of Tucuman Amazon. The Tucuman Amazon
breeds during the rainy season and, therefore, we expected to find a high number of
flooded nests. However, we found few flooded nests in extremely rainy years (i.e.
2006–2007 and 2007–2008).
7.5 Conservation Implications
Cloud forest in the Southern Yungas is subject to intense selective logging, primar-
ily of one species (i.e., Cedrela lilloi). Until recently (late 1990s) Podocarpus par-
latorei was intensively logged for paper pulp production, but now paper pulp is
made from sugar cane. Cloud forests are also frequently replaced by grasslands
through intentional fires set to facilitate extensive cattle grazing (Brown and Grau
1993; Brown et al. 2001). The probability of encountering a cavity increases with
tree age (Newton 1994); therefore, the surplus of suitable cavities found in this
study for each breeding Tucuman Amazon pair in an old-growth forest is not sur-
prising. In managed forests, the availability of suitable cavities for nesting might be
lower (Newton 1994; Politi et al. 2010), and if Tucuman Amazon requires a certain
number of suitable cavities in their home range, it is expected that the species will
be particularly vulnerable to the loss of nesting habitat through the impacts of selec-
tive logging or habitat modification (Monterrubio-Rico et al. 2009). Considering
that most of the Southern Yungas is under timber exploitation it is reasonable to
expect that this might be the reason that Tucuman Amazon has not recovered (Rivera
et al. 2010); i.e. large trees with cavities are probably lacking. To ensure the conser-
vation of Tucuman Amazon outside protected areas it is necessary that forest man-
agement guidelines promote the retention of large B. salicifolius trees, since this
species was used more frequently than would be expected from its abundance. This
might be possible because this species does not have a high timber value. However,
this is probably more difficult with other tree species (such as C. lilloi and J. austra-
lis) that are selected for nesting by Tucuman Amazon but have high timber values.
Finally, the nesting and spatial requirements of Tucuman Amazon could limit man-
agement actions intended to increase the density of nesting pairs. A mean distance
among suitable cavities of at least 150 m could represent the minimum distance to
consider in the spacing of active pairs to avoid exclusion by other pairs especially if
nest box provision is necessary.
7.6 Knowledge Gaps
We identified the following aspects that deserve future attention from a research
perspective to have a complete understanding of the ecology or Tucuman Amazon:
1 . Diet and food availability.
2. Survival of fledglings and adults.
3. Genetic structure of populations in North-south latitudinal gradient.
4. Movement ecology studying daily and seasonal movements and dispersion of
juveniles.
References
Aitken KEH, Martin K (2004) Nest cavity availability and selection in aspen conifer groves in a
grassland landscape. Can J For Res 34:2099–2109
Berkunsky I, Reboreda JC (2009) Nest site fidelity and cavity reoccupation by blue-fronted parrots
Amazona aestiva in the dry Chaco of Argentina. Ibis 151:1–34
BirdLife International (2018) Amazona tucumana. The IUCN Red List of Threatened Species
2018: e.T22686246A93104452. https://doi.org/10.2305/IUCN.UK.2016-3.RLTS.
T22686246A93104452.en. Downloaded on 01 August 2019
Bond J, Meyer de Schauensee R (1943) The birds of Bolivia. Part II. Proceeding of the Academy
of Natural Sciences of Philadelphia 95:167–221
Arias M, Bianchi AR (1996) Estadísticas climatológicas de la provincia de Salta. EEA Salta.
Dirección de Medio Ambiente y Recursos Naturales, Gobierno de Salta, Salta
Auer SK, Bassar RD, Fontaine JJ, Martin TE (2007) Breeding biology of passerines in a subtropi-
cal montane forest in northwestern Argentina. Condor 109:321–333
Bears H, Martin K, White GC (2009) Breeding in high-elevation habitat results in shift to slower
life-history strategy within a single species. J Anim Ecol 78:365–375
Berkunsky I, Kacoliris F, Faegre S, Ruggera R, Carrera J, Aramburú R (2011) Nest predation
by tree-snakes on cavity nesting birds in dry Chaco woodlands. Ornitologia Neotropical
22:459–464
Brightsmith DJ (2005) Competition, predation and nest niche shifts among tropical cavity nesters:
ecological evidence. J Avian Biol 36:74–83
Brown AD, Grau HR, Malizia L, Grau A (2001) Argentina. In: Kappelle M, Brown AD (eds)
Bosques nublados del Neotrópico. InBio, Santo Domingo de Heredia
Brown AD, Grau HR (1993) Investigación, conservación y desarrollo en las selvas subtropicales
de montaña. LIEY, Universidad Nacional de Tucumán, Argentina, Tucumán
Brown AD (1995) Fenología y caída de hojarasca en las selvas montanas del Parque Nacional El
Rey, Argentina. In: Brown AD, Grau HR (eds) Investigación, conservación y desarrollo en las
selvas subtropicales de montaña. LIEY, Universidad Nacional de Tucumán, Argentina, Tucumán
Buckland ST, Anderson DR, Burnham KP, Laake JL, Borchers DL, Thomas L (2001) Introduction
to distance sampling; estimating abundance of biological populations. Chapman and Hall,
New York
Burnham KP, Anderson DR (2002) Model selection and multimodel inference: a practical
information-theoretic approach, 2nd edn. Springer Verlag, New York
Cabrera AL, Willink A (1980) Biogeografía de América Latina. Organization of American States,
Washington
Cabrera AL (1994) Regiones Fitogeográficas Argentinas. Enciclopedia Argentina de Agricultura y
Jardinería. Tomo II. Editorial ACME S.A.C.I, Buenos Aires
Díaz S, Kitzberger T, Peris S (2012) Food resources and reproductive output of the austral parakeet
(Enicognathus ferrugineus) in forests of northern Patagonia. Emu 112:234–243
Enkerlin-Hoeflich EC (1995) Comparative ecology and reproductive biology of three species
of Amazon parrots in northeastern Mexico. Doctoral thesis. College Station, Texas A&M
University
Fjeldså J, Krabbe N (1990) Birds of the high Andes. Apollo Books, Stenstrup
Forshaw JM (1977) Parrots of the world, 3rd edn. Lansdowne Press, Willoughby
Forshaw JM, Cooper WT (1989) Parrots of the world. Lansdowne Editions, Sydney
González Elizondo JJ (1998) Productividad, causas de mortalidad en nidos y dieta de los polluelos
de tres especies de loro del género Amazona en el sur de Tamaulipas. Tesis MSc. Universidad
del Noroeste de México
Griffiths R, Double MC, Orr K, Dawson R (1998) A DNA test to sex most birds. Mol Ecol
7:1071–1075
Hines JE, Sauer JR (1989) Program CONTRAST: a general program for the analysis of several
survival or recovery rate estimates. Fish and Wildlife Technical Report 24:1–7
Hoy G (1968) Uber brutbiologie und eier einiger vogel aus nordwest-Argentina. J Ornithol
109:425–433
Hueck K (1978) Los Bosques de Sudamérica. Ecología, composición e importancia económica.
Agencia Alemana de Cooperación Técnica (GTZ), Berlín
Hunter ML Jr, White AS (1997) Ecological thresholds and the definition of old-growth forest
stands. Nat Areas J 17:292–296
Ibisch PL, Mérida G (2003) Biodiversidad: La riqueza de Bolivia. Estado de conocimiento y con-
servación. Ministerio de Desarrollo SostenibleEditorial FAN, Santa Cruz de la Sierra
Johnson DH (1979) Estimating nest success: the Mayfield method and an alternative. Auk
96:651–661
Joy JB (2000) Characteristics of nest cavities and nest trees of the red-breasted sapsucker in coastal
montane forests. J Field Ornithol 71:525–530
Juniper P, Parr M (1998) Parrots: a guide to parrots of the world. Yale University Press, New Haven
Koenig SE (2001) The breeding biology of black-billed parrot Amazona agilis and yellow-
billed parrot Amazona collaria in cockpit country, Jamaica. Bird Conservation International
11:205–225
Kappelle M, Brown AD (2001) Bosques nublados del neotrópico. INBIO, Costa Rica
Krebs CJ (1989) Ecological methodology. Harper Collins, New York
Krebs CJ (1999) Ecological methodology, 2nd edn. Benjamin Cummings, Menlo Park
Lack D (1954) The natural regulation of animal numbers. Clarendon Press, Oxford
Martin TE (1987) Food as a limit on breeding birds: a life-history perspective. Annu Rev Ecol Syst
18:453–487
Low R (2005) Amazon parrots, aviculture, trade, and conservation. Dona/Insignis
Publications, Prague
Manly BFJ, Mcdonald LL, Thomas DL, Mcdonald TL, Erickson WP (2002) Resource selection by
animals: statistical design and analysis for field studies. Kluwer Academic Publishers, Boston
Martinez J, Prestes NP (2008) Biologia da conservacao: estudo de caso com papagaio-charao e
outros papagaios brasileiros. Editora UPF, Passo Fundo
Masello J, Quillfeldt P (2002) Chick growth and breeding success of the burrowing parrot. Condor
104:574–586
Mayfield H (1975) Suggestions for calculating nest success. Wilson Bulletin 84:456–466
Mendoza EA (2005) El clima y la vegetación natural. In: Minetti JL (ed) El clima del noroeste
Argentino. Ed. Magna, Tucumán
Mitchell A (2005) The ESRI Guide to GIS Analysis. Volume 2. ESRI, Redlands
Monterrubio-Rico TC, Ortega-Rodríguez JM, Maríntogo MC, Salinas-Melgoza A, Renton
K (2009) Nesting habitat of the lilac-crowned parrot in a modified landscape in Mexico.
Newton I (1994) The role of nest sites in limiting the numbers of hole nesting birds: a review. Biol
Conserv 70:265–276
Narosky T, Yzurieta D (2003) Aves de Argentina y Uruguay: guía para la identificación. Edición de
oro, 15a ed. Vasquez-Mazzini, Buenos Aires, Argentina
Newton I (1980) The role of food in limiting bird numbers. Ardea 68:11–30
Orfila RN (1938) Los psittaciformes argentinos (cont). Hornero 7:1–21
Perry DR (1978) A method of access into the crowns of emergent and canopy trees. Biotropica
10:155–156
Pidgeon AM, Rivera L, Martinuzzi S, Politi N, Bateman B (2015) Will representation targets based
on area protect critical resources for the conservation of the Tucuman parrot? The Condor
Ornithological Applications 117:503–517
Politi N, Rivera L (2005) Abundance and distribution of parrots along the elevational gradient of
Calilegua National Park, Argentina. Ornitología Neotropical 16:43–52
Politi N, Hunter M Jr, Rivera L (2010) Availability of cavities for avian cavity nesters in selectively
logged subtropical montane forests of the Andes. For Ecol Manag 260:893–906
Renton K (2000) Scarlet macaw. In: Reading RP, Miller B (eds) Endangered animals: a reference
guide to conflicting issues. Greenwood Press, Westport
Renton K, Salinas Melgoza A (2004) Climatic variability, nest predation, and reproductive out-
put of lilac-crowned parrots (Amazona finschi) in tropical dry forest of western Mexico. Auk
121:1214–1225
Richardson DM, Bradford JW, Range PG, Christensen J (1999) A video probe system to inspect
red-cockaded woodpecker cavities. Wildl Soc Bull 27:353–356
Ricklefs RE (1980) Geographical variation in clutch size among passerines birds: Ashmole’s
hypothesis. Auk 97:38–49
Rivera L (2011) Ecología, biología reproductiva y conservación del Loro alisero (Amazona
tucumana) en Argentina. Tesis Doctoral. Córdoba, Argentina: Universidad Nacional de Córdoba
Rivera L, Politi N, Bucher EH (2007) Decline of the Tucuman parrot Amazona tucumana in
Argentina: present status and conservation needs. Oryx 41:101–105
Rivera L, Politi N, Bucher EH (2012) Nesting habitat of the Tucuma’n parrot Amazona tucumana
in an old-growth cloud forest of Argentina. Bird Conservation International 22:398–410
Rivera L, Politi N, Bucher EH, Pidgeon AM (2013) Nesting success and productivity of Tucuman
parrots (Amazona tucumana) in high-altitude forests of Argentina: do they differ from lowland
Amazona parrots? Emu 114:41–49
Rivera L, Rojas Llanos R, Politi N, Hennessey B, Bucher EH (2010) The near threatened Tucumán
parrot Amazona tucumana in Bolivia: insights for a global assessment. Oryx 44:110–113
Rocha AV, Rivera L, Martinez J, Prestes NP, Caparroz R (2014) Biogeography of speciation of two
sister species of Neotropical Amazona (Aves, Psittaciformes) based on mitochondrial sequence
data. PLoS One 9:1–10
Russello MA, Amato G (2004) A molecular phylogeny of Amazona: implications for Neotropical
parrot biogeography, taxonomy, and conservation. Mol Phylogenet Evol 30:421–437
Salinas-Melgoza A, Salinas-Melgoza V, Renton K (2009) Factors influencing nest spacing of a
secondary cavity-nesting parrot: habitat heterogeneity and proximity of conspecifics. Condor
111:305–313
Sandercock BK, Martin K, Hannon SJ (2005a) Life history strategies in extreme environments:
comparative demography of arctic and alpine ptarmigan. Ecology 86:2176–2186
Sandercock BK, Martin K, Hannon SJ (2005b) Demographic consequences of age-structure
in extreme environments: population models for arctic and alpine ptarmigan. Oecologia
146:13–24
Saunders DA, Smith G, Rowley I (1982) The availability and dimensions of tree hollows that
provide nest sites for cockatoos in Western Australia. Australian Wildlife Research 9:541–556
Shigo AL (1984) Compartmentalization: a conceptual framework for understanding how trees
grow and defend themselves. Annu Rev Phytopathol 22:189–214
Snyder NFR, Wiley JW, Kepler CB (1987) The parrots of Luquillo: natural history and conserva-
tion of the Puerto Rican parrot. Western Foundation of Vertebrate Zoology, Los Angeles
Skutch AE (1985) Clutch size, nesting success, and predation on nests of Neotropical birds,
reviewed. Ornithol Monogr 36:575–594
Stoodley J, Stoodley P (1990) Genus Amazona. Avian Publications, Altoona
Tella JL, Rojas A, Carrete M, Hiraldo F (2013) Simple assessments of age and spatial population
structure can aid conservation of poorly known species. Biol Conserv 167:425–434
Thomas L, Laake JL, Strindberg S, Marques FFC, Buckland ST, Borchers DL, Anderson DR,
Burnham KP, Hedley SL, Pollard JH, Bishop JRB, Marques TA (2006) Distance 5.0 release
2. Research unit for wildlife population assessment. University of St. Andrews, St Andrews
Tortorelli L (1956) Maderas y bosques Argentinos. Acme Agency, Buenos Aires
Wetmore A (1926) Observations on the birds of Argentina, Paraguay, Uruguay and Chile. Bulletin
of the United States National Museum 133:1–448
Chapter 8
Adaptive Strategies of Frugivore Bats
to Andean Cloud Forests
Adriana Ruiz and Pascual J. Soriano
8.1 Some Historic Considerations
It is well known that first feature distinguishing bat assemblages from along an
altitudinal gradient is the drastic reduction in the number of constituent species
(Graham 1983; Patterson et al. 1996). In two previous papers, we examined how
the bat assemblages of tropical cloud forests of Venezuelan Andes were com-
posed and structured (Soriano 2000; Soriano et al. 1999) and we demonstrate that
use of functional groups, such as trophic categories, is more useful for ecological
interpretation of differences in bat assemblage structure than the simple compari-
son of lists of names; thereby, we introduce the concept of “Trophic Equivalent”
(TE) as a way to make functional comparisons among bat assemblages
(Soriano 2000).
Andean cloud forest bat assemblages from Venezuela show two relevant charac-
teristics, distinguishing them from those of lowland rainforests: (1) Trophic simpli-
fication: The importance values for each functional category showed that the
assemblage is structured mainly on the base of frugivore and insectivore diets with
a small contribution from nectarivores. The rest of the categories are either not rep-
resented or their contribution is very small, such as hematophages, whose presence
in the cloud forests is associated with disturbed areas used for cattle farming. (2)
Trophic segregation of body sizes: The species with lowest body mass were largely
insectivores, while the inverse relationship was true for the frugivores. The nomadic
frugivore species were all large (in size), and the species in the sedentary frugivore
A. Ruiz
Postgrado en Ecología Tropical (ICAE), Facultad de Ciencias, Universidad de Los Andes,
Mérida, Venezuela
P. J. Soriano (*)
Laboratorio de Ecología Animal, Departamento de Biología, Facultad de Ciencias,
Universidad de Los Andes, Mérida, Venezuela

https://doi.org/10.1007/978-3-030-57344-7_8
152 A. Ruiz and P. J. Soriano
guild, which were small in size, in reality came from nectarivore guild but which
occasionally ate fruits.
In the bat assemblages of Andean cloud forests, frugivores represent the most
species-rich guild, in contrast with lowland rain forest bat communities, where
insectivores are the dominant guild (Graham 1983; Patterson et al. 1996; Soriano
2000; Soriano et al. 1999). Apparently, insectivores of tropical origin have a limit to
their vertical distribution that prevents them from accessing Andean cloud forests
and only some Vespertilionidae of Neartic origin reach these forests, along with a
few representatives of the Molossidae (Soriano 2000; Soriano et al. 1999). Variations
in metabolism observed in bats are mainly related to body mass and feeding habits:
bats with insectivorous or hematophagous diets have low rates of metabolism; bats
with combined diets (frugivore–carnivore) have low-to-intermediate rates of metab-
olism; frugivores have high rates of metabolism; and metabolic rates for nectariv-
ores are very high. Associated with basal metabolic rate is the capacity to regulate
body temperature; therefore, within the same feeding habit, large bats regulate their
temperature better than small ones. Similarly, for bats of a given body mass, insec-
tivores maintain lower body temperatures and are more dependent on environmental
temperature variations than frugivores (Bonaccorso and McNab 1997; McNab 196
9,1970,1982a,1983,1986,1989). At altitudes between 2000 and 3000 m in the tropi-
cal Andes, the mean temperature is 7–10.8 °C lower than in the lowlands (Sarmiento
1986). By virtue of their lower temperatures, high mountain environments demand
higher energetic expenditures for endotherms to maintain a constant body tempera-
ture. Therefore, for endotherms, the low temperatures of montane environments
may impose important physiological constraints on the possession of an adequate
energetic balance.
Such constraints seem to occur in members of the Emballonuridae, Mormoopidae,
Thyropteridae, Furipteridae, and Natalidae families, which never or rarely are found
at elevations above 1000 m (Graham 1983; Patterson et al. 1996; Soriano 2000;
Soriano et al. 1999). Successful adaptation to cold environments presupposes modi-
fications of physiological responses characterizing high mountain species. Among
such adaptive modifications are one or several of the following physiological fea-
tures: increase of basal metabolic rate, displacement of thermoneutrality zone to a
lower temperature range, decrease of thermal conductance by compensatory
increase of insulation, and daily facultative or obligatory torpor. We tested the
hypothesis that bats of the Andean highlands showed distinctive metabolic responses
compared with bats from lowland forests (Soriano et al. 2002); additionally, we
compared the existing literature with new information on three bat species having
the following food habits: a nectarivore (Anoura latidens), a frugivore (Sturnira
erythromos), and an insectivore (Tadarida brasiliensis). Basal metabolic rate, as
determined by oxygen consumption, thermal conductance, and body temperature
were measured at ambient temperatures of 10–38.8 °C. Some distinctive metabolic
responses of these bat species, although varying with respect to food guild, allowed
us to separate them from counterpart species that are typically found in lowland
forests. A. latidens was characterized by higher basal metabolic rate; however, ther-
mal conductance and lower critical temperature values did not showed an adaptation
8 Adaptive Strategies of Frugivore Bats to Andean Cloud Forests 153
to cool environments, and as expected. S. erythromos also increased its basal meta-
bolic rate, but it maintained thermal conductance as expected, which implied a very
important displacement of thermoneutral zone to lower temperatures. At tempera-
tures below lower critical temperature, in addition to an endothermic response,
S. erythromos sometimes expressed a hypothermic response or facultative torpor,
independent of sex and body mass. T. brasiliensis was a lower basal metabolic rate
and thermal conductance and also was its thermoneutral zone range displaced
toward lower temperatures. Likewise, this species entered obligate torpor when
ambient temperatures were below 22.8 °C.
8.2 Frugivore Bats of Genus Sturnira: A Study Case
The fruit bats (family: Phyllostomidae) inhabit mainly in the Neotropics, and few
species, belonging mainly to the subfamily Stenodermatinae, occur marginally in
the subtropical zones of both in North and South America (McNab 1982a,b;
Simmons 2005). In tropical highlands, fruit bats of the genera Platyrrhinus and
Sturnira usually inhabit elevations above 2000 m (Alberico et al. 2000; Graham
1983; Koopman 1978; Patterson et al. 1996; Soriano et al. 1999; Tuttle 1970).
However, only the second genus is diverse at high altitudes (de la Torre 1961; Iudica
2000). The distribution of frugivorous bats in the tropical region has been consid-
ered a result of the combination of historical (phylogeny) and ecological character-
istics that would explain its current restricted range (Colwell and Lees 2000;
Lomolino 2001; McNab 1982a; Rahbek 1995; Stevens 2006). The altitudinal limits
of some species that inhabit the Andean region do not appear to be correlated with
changes in vegetation and productivity along altitudinal gradients (Graham
1983,1990; Terborgh 1971,1977; Willig et al. 2003). In comparison with other
Neotropical bat families, the fruit-eating phyllostomids show thermoregulatory and
food habit restrictions that affect their distribution limits (McNab 1969,1980a,
b,1988; Stevens 2004,2006).
The physiologic mechanisms of ecological and evolutionary relevance are tools
that allow us to know how some physiological trials can have an effect on the distri-
bution patterns (McNab 1969,1973,1974,1976,1980a, b,1982a,2003,Soriano et al.
2002; Stevens 2004; Willig and Bloch 2006). A high-energy intake has been associ-
ated with species that are good thermoregulators as a consequence of their higher
metabolic rates (McNab 2003), with the exception of small size species, whose high
metabolic rates can be insufficient to compensate their higher heat loss (Audet and
Thomas 1997; McNab 1982a; Soriano et al. 2002). The apparent inability of fruit
bats to enter in torpor (hibernation) would presumably restrict them to occur prefer-
ably in the tropical region (McManus 1977), such as those in which flying foxes
(Pteropodidae) have been registered. Some pteropodid bats that inhabit in lowland
and subtropical zones use daily torpor due to the low resources availability (Bartels
et al. 1998; Bonaccorso and McNab 1997; Coburn and Geiser 1998; Geiser et al.
1996; Law 1994; McNab 1989). In Andean highlands, ambient temperatures
decrease by 0.6 °C with every 100 m increase of elevation (Sarmiento 1986); there-
fore, fresh fruit availability and productivity diminish, both factors should affect bat
species diversity along of the altitudinal gradient (Graham 1983; Rosenzweig 1992;
Terborgh 1971; Tilman 1982).
The members of the genus Sturnira are strict frugivores and show a widespread
distribution in the Western Hemisphere (Simmons 2005). Almost half of the 14 spe-
cies of this genus reach highlands in tropical regions, while others are restricted to
lowlands or they may reach the mountains marginally (Contreras-Vega and Cadena
2000; de la Torre 1961; Simmons 2005). The small differences in body size along
with morphological similarities could explain the use of similar resources in high-
lands (Giannini 1999; Molinari and Soriano 1987; Ruiz 2006). So far, very little is
known concerning the thermoregulation capacity of high mountains species; the
only species in which mechanism has been reported until know is Sturnira erythro-
mos, in which higher basal metabolic rate associated with small body size (15 g)
appears to cause difficulties in metabolic maintenance of body temperatures.
Therefore, these limitations are compensated with daily torpor (Soriano et al. 2002).
Although torpor is not an exclusive feature of high altitude dwellers, it could be
used as an energy-saving trait during resource shortage periods (Geiser and Coburn
1999; Geiser et al. 1996; McNab and Bonaccorso 2001).
In frugivorous highland bats, small body size combined with high mass-specific
energy expenditure could result in a limited ability to store vital resources (McNab
1982a). If physiological limits depend on body mass (McNab 1983), it is feasible
that inter-and intra-specific physiological differences between small and large spe-
cies of Sturnira (15–50 g) could be explained by means of body mass. Other fruit
bats that reach marginally the Andean mountains such as Artibeus jamaicensis,
Carollia perspicillata, and Sturnira lilium show similar heterothermic responses;
yet in these cases, resource shortage must be the decisive factor that determines
their higher distribution limits (Audet and Thomas 1997; Studier and Wilson 1970).
It is difficult to separate the effect of resource shortage from temperature variation
along the altitudinal gradient. The fruits consumed by these species are distributed
along the Andean slopes, and their abundance increases in cloud forests (Giannini
1999; Ruiz 2006; Soriano 1983). In some cloud forests, as many as five different
species of Sturnira can coexist (Soriano 1983). If the fruit availability is not restric-
tive at high elevations for these species, low temperature is the most probable factor
that conditions these species altitudinal distribution, as previous studies suggest
(Soriano 1983). Is it possible that these species are capable of entering torpor as
their congener S. erythromos and as other frugivorous and nectarivorous bats of
old world?
The aim of this research was to answer two questions: (1) what is the effect of
low temperatures on the thermoregulation of highlands fruit bats? and (2) what are
the physiological mechanisms that allow them to colonize and to establish in the
Andean highlands? We selected three species of the genus Sturnira that inhabit
cloud forests of the Venezuelan Andes: Sturnira bidens, S. bogotensis, and S. ludovici
(>2000 m of elevation; Linares 1998). In this ecosystem, these species are sympat-
ric with S. erythromos (Ruiz 2006; Soriano 1983). The physiological traits
( regulation of the body temperature, basal metabolic rate, and humid thermal con-
ductance) were compared with the available information concerning lowlands fru-
givorous bats (Family Phyllostomidae). Finally, we analyzed the ecological factors
associated with the physiological measurements and we discussed the physiological
consequences of thermoregulation on the altitudinal distribution patterns of the
genus Sturnira.
8.3 Materials and Methods
Study animals: The three species of the genus Sturnira (S. bidens, S. bogotensis, and
S. ludovici) were captured in an Andean cloud forest near of the city of Merida,
Merida State, Venezuela, between 2000 and 2500 m altitude. We use mist nets to
capture bats at foraging sites in two Andean localities: (1) Monterrey (71° 7’W; 8°
41’N; to 8 km NE Merida; at 2300 m) and (2) Monte Zerpa (71° 10’W; 8° 39’N; to
4 km NNE Merida; at 2000–2200 m). Once captured, we selected only the adults
bats and confirmed that the females did not show signs of reproductive activity
(pregnancy or lactation); then animals were placed in individual cloth bags and
transported within 0.5–1 h to the Laboratory of Animal Ecology (Universidad de
Los Andes, Merida, Venezuela). We maintained them in captivity for 3 or 4 days,
under controlled conditions of light:darkness (12 L:12D) and ambient temperature
(among 21–23 °C). During this period, we feed the bats with commercial baby food
of tropical fruits 4–6 h prior measurements, with the purpose of assuring that ani-
mals were post-absorptive (Kovtun and Zhukova 1994). We followed guidelines for
procedures related to animal care and use, approved by the American Society of
Mammalogists (Gannon et al. 2007).
Experimental protocol: Our field and laboratory work were conducted between
November 2002 and July 2005. We performed the experiments between 0900 and
1800 h, when the bats were at rest and post-absorptive. We measured experimen-
tally the oxygen consumption (VO2 in mL O2 g−1 h−1) and humid thermal conduc-
tance (C in mL O2 g−1 h−1 °C−1) in an open-circuit respirometry system, modifying
ambient temperature from 10 to 40 °C. We placed each bat in sealed plastic-PVC
chamber (450 and 900 mL) with a wire mesh inside from which the bats could hang
in a natural posture. We submerged the sealed chamber in a thermoregulated water
bath. After an acclimation period of ca. 60 min, the current oxygen levels were
measured for 2–4 h. For individuals at ambient temperature of 34 °C or more, the
test period was reduced to 30 min. Measurements of air temperature inside the
chamber were monitored with a copper–constantan thermocouple that passed
through the stopper and connected to a Microprocessor Digital Thermometer
(Model HH23 Omega, Stanford, Connecticut). The lid chamber had two ports con-
nected to tubes for air entry and exit. Room air was drawn in by a pump at a flow
rate that we regulated between 100 and 140 or 220 and 260 mL/min for the 450 and
900 mL chambers, respectively. These airflow rates inside the chambers were ade-
quate, since metabolic rates were not affected by oxygen flow rates. Exiting from
the chamber, carbon dioxide and water were removed from the air by color-
indicating soda lime and silica gel, respectively, and then flow rate was measured by
a flow meter Cole Parmer (Model P/N: 10130).
We measured the oxygen concentration of the air with an oxygen analyzer
(Applied Electrochemistry, model S3A-II Ametek, Pittsburg, Pennsylvania), which
we calibrated previously. Voltage output of the analyzer was recorded with a per-
sonal computer, using A/D converter card, and acquisition software (created by the
Laboratory of Scientific Instrumentation, Universidad de Los Andes, Merida). Bats
were removed from the experimental chamber after assuring that oxygen concentra-
tion was minimal and stable for at least 30 min. We measured body temperature
(Tb) of the animals immediately before and within 15 s after using rectal insertion
of a fine (0.2 mm diameter) thermocouple probe calibrated to the nearest 0.1 °C,
which was read with a digital thermometer. Bats were considered torpid when body
temperature fell below 30 °C and when those rewarmed reached their initial body
temperature. Animals were weighed before and after the experiments, and a linear
decrease of body mass throughout the experiment was assumed for calculation of
mass-specific metabolic rate. We took care that the individuals did not lose more
than 10% of their body mass.
Estimate of parameters: We tested the effect of ambient temperature (Ta) on
physiological traits using lineal regressions. We corrected the values to standard
pressure and temperature, and we calculated the rate of metabolism from Dépocas
and Hart (1957) and the thermal conductance “humid” (C) with the equation of
McNab (1980b). The lowest temperature at which the animal maintains basal met-
abolic rate (lower limit of thermoneutrality) was determined following methods in
Bonaccorso et al. (1992), i.e., by the intersection of the curve below thermoneu-
trality (TNZ) with the basal metabolic rate (BMR) and the minimal squares
method. We calculate the minimal thermal conductance (C) of the fitted slope of a
curve below thermoneutrality; and another that represented an estimate of minimal
conductance versus ambient temperature, where we identified the point in which
conductance increased. Values for basal metabolic rate and minimal conductance
were compared with values expected from standard allometric scaling for mam-
mals and bats: basal metabolic rate (V02′/m) for mammals was calculated as mL
O2 g−1 h−1 = 3.45 m–0.287 (where m is the body mass in g; McNab 1988) and stan-
dard for bats as mL O2 g−1 h−1 = 2.97 m-0.256 (Speakman and Thomas 2003). The
thermal conductance (C′) for mammals was calculated as: mL O2
g−1 h−1 °C−1 = 1.02 m–0.5 (Herreid and Kessel 1967) and for bats as: mL O2
g−1 h−1 °C−1 = 0.901 m–0.466 (Speakman and Thomas 2003). The observed values of
V02′/m and C′ for the species are the percentage of the expected value of the equa-
tions allometrics.
Statistical analysis: Numerical values are presented as mean ± 1SE for n number
of measurements. N represents the number of individuals. Differences among or
between means were tested using a one-way ANOVA or a student’s t-test in order to
determine differences between physiological variables (Zar 1999). When there were
not differences among sexes, the data were joined. Linear regressions were fitted
using the method of least squares, and differences between regressions were
d etermined using ANCOVA. Significance was accepted at P < 0.05. The statistical

package used was JMP 5.0.1.2 (SAS Institute Inc. Campus Drive, Cary, NC, USA,
1989–2003).
8.4 Results
Sturnira bidens: We used 3 males and 11 females in 175 measurements. The mean
mass for this species was 17.15 g ± 0.29 (N = 14; range between 14.9 and 18.8 g).
The mean mass of the males (17.1 ± 0.18 g; n = 23) was similar to the females
(17.5 ± 0.14 g, n = 11) during the experiments (t = 0.67; P = 0.513). Variations of
body temperature, basal metabolic rate, and thermal conductance were not corre-
lated with sex (F = 0.15; P = 0.69).
S. bidens showed a dichotomous response in Tb to low chamber temperatures. At
temperatures below 25 °C, bats either maintained a lightly constant Tb (normother-
mic state with Tb = 32–34) or dropped Tb below 30 °C to values near Ta (herein
defined as torpor; Tb = 19.4–30.8 °C; Fig. 8.1). Both responses were independent of
body mass (t = 1.49; P = 0.139). Normothermic individuals maintained their body
temperature (32–34 °C) slightly dependent on the ambient temperature (line a in
Fig. 8.1a; Tb = 31.16 + 0.1Ta; the slope of the regression differs significantly from
zero, r2 = 0.259; P = 0.000; n = 72). The mean Tb for all individuals between 19 and
25 °C was 33.5 ± 0.08 °C (r2 = 0.038; P = 0.20; n = 44).
Obvious evidence of torpor was observed in individuals exposed to ambient tem-
peratures below 25 °C (line b in Fig. 8.1a; the slope of the regression differs signifi-
cantly from zero, r2 = 0.68; P = 0.0001; n = 26). Bats that entered in torpor maintained
a minimal body temperature of ca 22 °C. Body temperature fell with decreasing
ambient temperature, yet the body temperature of S. bidens was somewhat 6 °C
above environmental temperature (5.6 ± 0.41); such variations were not correlated
with body mass (r2 = 0.08; P = 0.17; n = 26). Some individuals exposed to ambient
temperatures below 18 °C increased their body temperature at Tb–Ta = 9 °C (white
circles in Fig. 8.1a).
As with Tb, metabolic measurements indicated a dichotomous thermoregulatory
response (normothermy and torpor) by bats at Ta < 25 °C (Fig. 8.1b). The zone of
thermoneutrality extended from 25 to 31 °C (line c in Fig. 8.1b). The mean basal
metabolic rate was 1.47 ± 0.02 mL O2 g−1 h−1, which is equal to 95% of the standard
rate expected for mammals and 102% for bats, from a mean mass equal to
16.8 ± 0.17 g (line c, Fig. 8.1b; the slope of the regression does not differ signifi-
cantly from zero, r2 = 0.002; P = 0.72; n = 54). Normothermic metabolism rate
increased according to the regression VO2/m = 5.99–0.18Ta (r2 = 0.823; P = 0.000;
n = 72) at temperatures below thermoneutrality (black circles around line in
Fig. 8.1b). Slope of curve of metabolic rate was projected to intercept the axis of
ambient temperature at 33.3 °C, with 95% confidence limits covering the range of
body temperatures recorded during these experiments. Whereas torpid maintained a
variable lower metabolic rate between 0.58 ± 0.03 and 0.42 ± 0.02 mL O2 g−1 h−1;
Fig. 8.1 (a) Body temperature, (b) metabolic rate, and (c) humid thermal conductance, as a func-
tion of the ambient temperature in Sturnira bidens (N = 14 individuals). The regression lines are
indicated for (a) body temperature regulation by normothermics, (b) by torpid individuals, (c)
basal metabolic rate, (d) metabolic rate below the thermoneutrality, (e) minimal thermal conduc-
tance for normothermics, and (f) for torpids. The discontinuous line in (a) represents the equality
among ambient temperatures (Ta) and body temperature (Tb). The white circles represent the tor-
pid individuals during the experiments
(t = 1.94; P = 0.07; significant heterogeneity of variances, P = 0.008). The Tb–Ta

differential in thermoconforming torpid bats was not a function of ambient tempera-
ture (Fig. 8.1a,b). At low ambient temperature (below 18 °C), torpid bats increased
their metabolic rates (1.21 ± 0.2 mL O2 g−1 h−1) to maintain the body temperature
above a threshold. Above the upper critical temperature, metabolic rate of all indi-
viduals increased, and their body temperature reached 36 °C.
Differences in minimal thermal conductance were found between normothermic
and torpid bats (t = 14.04; P = 0.000; Fig. 8.1c). Minimal thermal conductance of
the normothermic bats was independent of the ambient temperature below the 26 °C
(line e in Fig. 8.1c; the slope does not differ significantly from zero, r2 = 0.013;
P = 0.31; n = 85), the mean of which is 0.19 ± 0.002 mL O2 g−1 h−1 °C−1, equal to
79% of the mammalian and 80% of the bats standard, with a body mass of
17.4 ± 0.12 g. In torpid bats, the mean was 0.12 ± 0.004 mL O2 g−1 h−1 °C−1(line f
in Fig. 8.1c), which is 50% of both standards. The slope of regression of the meta-
bolic rate in the lower limit of the thermoneutrality is 95% confidence limits cover-
ing the minimal conductance mean (0.18 ± 0.01).
Sturnira bogotensis: Mean body mass of six males and four females ranked from
18.7 to 24.4 g; the mean value was 21.5 ± 0.62 g (N = 10) with no differences among
sexes (t = 0.57; P = 0.587). Also, no differences among sexes were found in body
temperature, basal metabolic rate, and minimal conductance measurements
(ANOVA; F = 0.42; P = 0.47).
Body temperature of S. bogotensis was regulated as normothermic but slightly
dependent on ambient temperature (line a in Fig. 8.2a; Tb = 31.42 + 0.054Ta; the
slope of the regression differs significantly from zero, r2 = 0.104; P = 0.017; n = 54).
Among Ta = 22 and 25 °C, the Tb was independent of the ambient temperature
(r2 = 0.085; P = 0.2) with an average of 32.8 ± 0.14 °C (n = 21). In 6 of 60 experi-
ments, ambient temperatures remained below 21 °C, and the body temperature fell
below 29 °C (circles open in Fig. 8.2a). These individuals were hypothermic and
showed lower metabolic rates during the first day of experimentation, yet they grad-
ually recovered normothermy.
The thermoneutrality zone extended from 25 to 31 °C (line b in Fig. 8.2b; the
slope did not differ significantly from zero, r2 = 0.003; P = 0.7; n = 47). Rate meta-
bolic basal average was 1.62 ± 0.02 mL O2 g−1 h−1, or 113% of the standard for
mammals and 119% for bats, with a mass of 21.36 ± 0.29 g. At the lower limit of
thermoneutrality (25 °C), metabolic rates increased as ambient temperature
decreased. According to the regression, VO2/m = 6.65–0.20Ta (line c in Fig. 8.2b;
the slope differed significantly from zero, r2 = 0.944; P = 0.000; n = 54), and the
projection intercepted the axis of ambient temperature at 33.2 °C. The confidence
limit of 95% included the body temperature registered for this frugivore during the
experiments (31.75 < Tb < 33.3). The individuals with hypothermy showed lower
metabolic rates (white circles in Fig. 8.2b; r2 = 0.69; P = 0.039; n = 6). Above the
thermoneutrality, the metabolic rate increased and the animals reached a maximum
body temperature of 38 °C.
Minimal thermal conductance differed significantly between normothermic and
hypothermic individuals (t = 9.57; P = 0.000; n = 58). The minimal conductance
tion of the ambient temperature in Sturnira bogotensis (N = 10 individuals). The regression lines
are indicated for (a) body temperature regulation for normothermics, (b) basal metabolic rate, (c)
the metabolic rate below the thermoneutrality, and (d) minimal thermal conductance. The discon-
tinuous line in (a) represents the equality between ambient temperatures (Ta) and body tempera-
ture (Tb). The white circles represent individual’s hypothermics during the experiments
mean was 0.21 ± 0.002 mL O2 g−1 h−1 °C−1, or 95% of the values expected from a

standard for mammals and bats, weighing 21.35 ± 0.28 g (line d in Fig. 8.2c; the
slope did not differ significantly from zero, r2 = 0.0052; P = 0.6051; n = 54; white
circles correspond to hypothermic individuals, n = 6). The conductance of the slope
of regression of the metabolic rate in the lower part of the thermoneutrality zone
was inside the interval of confidence of 95%.
Sturnira ludovici: We use four females and seven males in 185 experiments. The
mean mass ranked from 23 to 34.3 g, with an average of 28.6 ± 1.17 g (N = 11). The
mean body mass of males (30.6 ± 1.2 g) was significantly higher than those of
females (25 ± 0.89 g; t = 3.22; P = 0.01); therefore, these data were not combined.
This species showed a normothermic response to environmental temperature
changes (Fig. 8.3). Below 25 °C, the females maintained an independent Tb
(r2 = 0.022; P = 0.4015; n = 34), while males show a dependent Tb (r2 = 0.113;
P = 0.024; n = 45). Although we did not find significant differences among sexes
(ANOVA; sex, F = 1.03; P = 0.31), the body temperatures of S. ludovici were
slightly dependent on the ambient temperature changes (line a in Fig. 8.3a;
Tb = 32.19 + 0.072Ta, the slope differed significantly from zero, r2 = 0.07; F = 5.8;
P = 0.018; n = 79). Mass could explain the differences of thermoregulation; how-
ever, at low temperatures, the females were better thermoregulators than males
(ANCOVA; mass, F = 53.2; P = 0.000; sex, F = 43.67; P = 0.0001; interaction,
F = 0.25; P = 0.62). At ambient temperatures ranging between 18 and 25 °C, the
mean body temperature was 33.8 ± 0.13 °C (n = 43). In six experiments, some indi-
viduals entered in hypothermy at ambient temperatures below 21 °C, with a variable
body temperature between 21 and 30 °C.
The thermoneutral zone extended to 32 °C (Fig. 8.3b). We did not find differ-
ences among sexes concerning basal metabolic rates (ANCOVA; mass, F = 34.59;
P = 0.000; sex F = 0.17; P = 0.67; interaction, F = 1.48; P = 0.23; line b in Fig. 8.3b).
The average of the basal metabolic rate was 1.74 ± 0.02 mL O2 g−1 h−1 (n = 82), or
131% and 137% of the value standard for mammals and bats, respectively, with a
mass average of 27.4 ± 0,4 g. Below thermoneutrality (25.5 °C), the metabolic rate
varied with sex (F = 46.98; P = 0.000), which we can explain by body mass differ-
ences (ANCOVA; mass, F = 56.48; P = 0.000; sex, F = 1.15; P = 0.287; interaction,
F = 2.67; P = 0.11). Therefore, we analyzed the values jointly, and we obtained the
following regression: VO2/m = 7.273–0.216Ta (line c in Fig. 8.3b; the slope differed
significantly from zero, r2 = 0.824; P = 0.000; n = 79). The slope of the line of
regression of the metabolic rate intercepted the axis of ambient temperature at
33.7 °C and the limits of confidence included body temperature values registered
during the experiments (33.4 < Tb < 33.8). A few hypothermic individuals main-
tained lower metabolic rates (white circles in Fig. 8.3b). Above the thermoneutral
zone, the maximum body temperature was 39.4 °C.
We did not find differences among sexes for minimal thermal conductance
(ANCOVA; mass, F = 25.85; P = 0.0001; sex, F = 2.6; P = 0.11; interaction,
F = 2.49; P = 0.12). Minimal conductance determined the lower limit of thermoneu-
trality (line d in Fig. 8.3c; the slope of the regression did not differ significantly
tion of the ambient temperature in Sturnira ludovici (N = 11 individuals). The regression lines are
indicated for (a) body temperature regulation by normothermics, (b) basal metabolic rate, (c) the
metabolic rate below the thermoneutrality, and (d) minimal thermal conductance. The discontinu-
ous line in (a) represents the equality between ambient temperatures (Ta) and body temperature
(Tb). The white circles represent individual’s hypothermics during the experiments
from zero, r2 = 0.034; P = 0.101; n = 79). The mean minimal conductance was
0.22 ± 0.002 mL O2 g−1 h−1 °C−1, or 116% of the value expected for mammals and
bats, with a mass of 29.5 ± 0.5 g. The mean minimal conductance was inside the
interval of confidence of 95%, of the slope of regression of metabolic rate
(0.216 ± 0.01). The hypothermic individuals (white circles in Fig. 8.3c) showed a
mean conductance (0.13 ± 0.007) significantly lowest (t = 9.6; P = 0.000; n = 6).
Comparison among species: Body temperature varied significantly among the
species (ANOVA; F = 15.52; P = 0.000). These differences may be related to body
mass (mass, F = 584.97; P = 0.000; S. bidens < S. bogotensis < S. ludovici). We
found that when ambient temperatures are low, differences in body temperature are
independent of body mass (ANCOVA; mass, F = 1.14; P = 0.29; species, F = 7.15;
P = 0.001; Ta, F = 0.004; P = 0.95; interaction, F = 2.17; P = 0.12). An a posteriori
test of multiple comparisons (Tukey HSD) showed that S. bidens and S. bogotensis
maintain significantly lower body temperatures than S. ludovici.
The metabolic rate in the thermoneutral zone varied significantly among species
(ANCOVA; mass, F = 143.84; P = 0.0001; species, F = 3.63; P = 0.022; Ta, F = 0.21;
P = 0.64; interaction, F = 0.30; P = 0.76); only S. ludovici showed a significantly
higher basal rate than that was expected for its mass (Tukey HSD). The minimal
conductance was significantly different among them (ANCOVA; mass, F = 166.04;
P = 0.0001; species, F = 8.50; P = 0.0003; Ta, F = 1.57; P = 0.21; interaction,
F = 0.51; P = 0.60). The test of Tukey HSD showed that S. bogotensis and S. ludovici
had higher minimal conductance. The length of the fur was significantly different
among the three species (F = 13.23; P = 0.0001; Tukey HSD), yet it failed to explain
the differences. The individuals of S. ludovici had shortest coat (7.59 ± 0.09; N = 21)
than those of S. bidens (8.05 ± 0.1; N = 20) and S. bogotensis (8.3 ± 0.11; N = 20).
8.5 Discussion
Thermoregulation: Sturnira ludovici and S. bogotensis were normothermics below

thermoneutrality, and they maintained a high and constant body temperature over a
wide range of air temperatures. However, their body temperatures were lower than
those registered for other lowland fruit bats (ANOVA; F = 7.45; P = 0.02; N = 16;
Table 8.1). In contrast, S. bidens showed a dichotomous response (heterothermy) to
low air temperatures, one normothermic and the other defined as daily torpor.
Indeed, only one individual of S. bidens showed both responses, but it was not pos-
sible to establish a clear threshold of metabolism between normothermy and torpor.
According to these results, neither selective sex-biased nor body mass affect the
faculty of entering torpor. These differences among individuals of S. bidens suggest
that torpor stage is not an obligate response to lower air temperatures. The minimum
body temperature of S. bidens reported here (ca. 22 °C) is similar to those registered
in torpid tropical nectarivorous and insectivorous bats of equivalent body mass
(Bonaccorso and McNab 1997; Genoud and Bonaccorso 1986). Likewise during
torpor, S. bidens failed to reduce its body temperature below 22 °C, which could
Table 8.1 Energetic comparison among frugivore bats from families Phyllostomidae and Pteropodidae
164
Species Body mass (g) BMR %BMR C %C F Tb Tcl Tcm TS Shelter D E R T TR

Phyllostomidae
Artibeus concolor1 19.7 1.67 113.87 0.21 90.95 1.3 35.3 28 Foliage C L TR N G
A. fimbriatus2 63.9 1.22 116.61 Foliage C L TR N ?
A. jamaincensis1 45.2 1.25 108.17 0.14 92.28 1.2 36.4 25 35.8 Foliage C L TR N G
A. lituratus1,2 69.7 1.24 121.51 0.11 90.03 1.3 37.3 25 37.7 19 Foliage/caves C L TR N G
Carollia perspicillata1,2 13.9 2.10 129.55 0.27 98.69 1.3 36.4 28.2 37.7 24.3 Caves/hollow trees C L TR N G
Chiroderma doriae2 19.9 1.56 106.68 26.4 ? C L TR N ?
Platyrrhinus lineatus1,2 22.3 1.38 97.50 0.19 88.89 1.1 36.4 28 Foliage C L TR N G
Rhinophylla pumilio1 9.5 1.71 94.58 0.31 93.37 1.0 34.7 30 37 ? C L TR N G
Sturnira bidens3 16.8 1.47 95.76 0.19 76.35 1.3 33.5 25 33 15 Hollow trees? C H TR Y P
S. bogotensis3 21.4 1.62 113.12 0.21 95.24 1.2 32.8 25 37 15 Hollow trees? C H TR N I
S. erythromos4 15.9 2.01 128.88 0.26 101.64 1.3 34.4 25.5 36 15 Hollow trees? C H ST Y P
S. lilium1 21.9 1.79 125.82 0.19 87.17 1.4 36.4 28.1 37.1 28.4 Hollow trees C L ST N G
S. ludovici3 27.4 1.74 130.43 0.22 112.90 1.2 33.8 26 35 15 Hollow trees? C A TR N I
S. tildae2 20.5 1.95 134.49 Hollow trees? C L TR N ?
Uroderma bilobatum1 16.2 1.64 105.72 0.25 98.65 1.1 35.1 28 35.6 Foliage C L TR N G
Vampyressa pussila1 8.8 2.11 114.16 27.3 ? C L TR N ?
L
Pteropodidae L
Cynopterus brachyotis5 37.4 1.27 104.09 0.19 113.92 0.9 36.5 30 36 Trees C L TR N G
Dobsonia anderseni6 241.4 0.72 100.78 0.09 131.00 0.8 36.4 28 34 Caves I L N G
D. minor6 73.7 1.01 100.57 0.12 101.00 1.0 36.1 28 35 Trees C L TR N G
D. moluccensis6 404.3 0.91 147.69 0.09 177.42 0.8 36.8 27 36 Caves C L TR N G
D. praedatrix6 179.5 0.79 102.07 0.08 105.08 1.0 37.1 27 32.5 Caves I L TR N G
Nyctimene albiventer6 30.9 0.88 68.28 0.09 50.68 1.3 35.9 28 33.5 Trees C B TR N I
A. Ruiz and P. J. Soriano
Species Body mass (g) BMR %BMR C %C F Tb Tcl Tcm TS Shelter D E R T TR
N. cyclotis6 40.4 1.60 134.07 0.09 53.59 2.5 36 17 34 Trees C A N G
Paranyctimene raptor6 23.6 1.04 74.69 0.15 71.44 1.0 33.8 27 35 Trees C B TR Y I
Pteropus giganteus7 (562.2) 0.52 92.77 0.02 46.49 2.0 36.7 30 Trees C B TR N G
P. hypomelanus7 (520.8) 0.56 97.74 0.03 58.17 1.7 35.7 30 Trees I B TR N G
P. poliocephalus8 598.0 0.53 96.24 0.02 55.14 1.7 37 17 35 Trees C B ST N G
P. pumilus7 194.2 0.65 85.47 0.05 68.31 1.3 36.1 23 Trees I B TR N G
P. rodricensis7 254.5 0.53 75.32 0.05 78.20 1.0 36.5 24 Trees I B TR N G
P. vampyrus7 1024.3 0.78 165.30 0.03 103.54 1.6 36.9 30 Trees C B N G
Rousettus aegyptiacus9 146.0 0.84 101.77 0.10 118.46 0.9 34.8 31 34 Trees C B ST N G
R. amplexicaudatus6 91.5 1.14 120.79 0.11 103.16 1.2 36.5 26 34 Caves C B TR N G
Source of data: 1McNab (1982a, b), 2Cruz-Neto et al.(2001), 3This study, 4Soriano et al. (2002), 5McNab (1989), 6McNab and Bonaccorso (2001), 7McNab and
Armstrong (2001), 8Bartholomew et al. (1964), 9Noll (1979)
Abbreviations: BMR mass specific basal metabolic rate; C mass-specific thermal conductance; %BMR percent of spected value of basal metabolic rate (McNab
1988), %C spected value of thermal conductance (Herreid & Kessel 1967), F quotient between spected value of basal metabolic rate and thermal conductance,
Tb body temperature, Tcl lower critical temperature, Tcm maximal critical temperature, TS temperature of shelter, D distribution, C continental, I insular, E
elevation, L lowlands, H highlands, R region, TR tropics, ST subtropics, T torpor, N not enter in torpor, Y enter in torpor, TR regulation of body temperature, G
good, I intermediate, P poor,? data not available. For species that showed several values in bibliography for a parameter done, average was taken if difference
was ≤5% or the lower value if that difference was >5%. For species with sexual dimorphism, average values were taken and represented in brackets
8 Adaptive Strategies of Frugivore Bats to Andean Cloud Forests
165
represent a limitation in its capacity to occupy temperate zones (Bonaccorso and

McNab 1997; Geiser et al. 1996; Law 1994).
The capacity of entering torpor has not hitherto been demonstrated in the fru-
givorous phyllostomid bats. Our results suggest that it occurs on a daily basis in
S. bidens, as well as S. erythromos (Soriano et al. 2002). Other authors such as
Audet and Thomas (1997), Cruz-Neto and Abe (1997), and Studier and Wilson
(1970) have demonstrated that the phyllostomids can be poor thermoregulators in
the laboratory and also under natural conditions, as a consequence of deficient nutri-
tional state. We could explain the hypothermy during the experiments in all of three
species studied here, as a consequence of absence of brown fat, as well as a poor
nutrition before their capture (A. Ruiz, pers. obs.). As we fed them ad libitum and
some individuals were able to recover during the period of captivity (three or four
days), this allowed us to determine the true use of torpor in the individuals of
S. bidens that recovered their normothermic body temperature when experiments
concluded.
In tropical highland species, we observed a reduction in the lower critical limit
of thermoneutrality (to 25 °C), which could represent a decrease in thermoregula-
tory cost, similar to the decrease observed in birds of temperate zones (Canterbury
2002). Therefore, we correlated the decrease of their lower critical temperature with
altitude, when body mass was deleted from the analysis (ANCOVA; mass, F = 48.13;
P = 0.0001; altitude, F = 37.39; P = 0.0001). Low air temperature thermoregulation
is mass-independent substantially reduced energy expenditure. If we compare the
temperature differentials in thermoneutrality (ΔT = Tbody–Tlimit) of our species of
Sturnira with some of other genera (i.e., small Artibeus, Rhinophylla, Vampyressa;
Table 8.1), it seems to be that Sturnira and other small frugivores that reach margin-
ally or that inhabit the mountain ecosystems (such as Carollia) can maintain similar
temperature differentials than those recorded for lowland species. The differences
among these species were observed at maximum ambient temperatures (Table 8.1);
therefore, animals exposed to higher temperatures (>33 °C) were not able to main-
tain their body temperature below the ambient temperature for long periods. At high
ambient temperatures, these bats depend heavily on evaporative cooling to keep
their body temperature below lethal levels (Bartholomew et al. 1970).
For phyllostomid bats, it has been demonstrated that thermoregulation is influ-
enced by body mass (McNab 1969,1970,1983). In these group of bats, body tem-
peratures were variable (34.9 ± 0.33) and clearly correlated with their body mass
(F = 10.75, P = 0.008, r2 = 0.73; N = 16). When we deleted body mass from the
analysis, the elevation appeared to be significant (ANCOVA; F = 15.96; P = 0.0025;
Test of Tukey HSD), yet basal metabolic rates were not (F = 0.89; P = 0.36; r2 = 0.38;
N = 16). As mentioned earlier, we can interpret that a frugivorous diet does not
imply that these results enable us to affirm that regulation of the body temperature
depends directly on the intake of high-energy food. In these species, temperature
regulation appears to be related to body size rather than diet and could be dependent
on the maximum adjustments of their metabolic rates and thermal conductance.
Rate of metabolism: In this study, we were not able to demonstrate a higher
BMR for all of bat species from highlands (Table 8.1). Only, S. ludovici showed
BMR that exceeds those of lowland fruit bats (137%) with equivalent body mass.
Some authors have suggested that higher basal metabolic rates in Sturnira erythro-
mos are an adaptation to highland environments (McNab 2003; Soriano et al. 2002).
But the short time of experimentation used by Soriano et al. (2002) for S. erythro-
mos (~1.5 h) suggests that average values calculated correspond to the rest meta-
bolic rate (2.51 mL O2 g−1 h−1). Consequently, if the rest metabolic rate is equal to
1.25 times the BMR (Aschoff and Pohl 1970; Grodzinski and Wunder 1975;
Kendeigh 1970), then BMR of S. erythromos should be 2.01 mL O2 g−1 h−1 or 129%
(m = 15.9 g). If we compare its BMR with those observed in S. bidens (96%;
m = 16.8 g), we can obtain a 30% difference among both species, which related the
maximum thermogenic capacity of small fruit bats (Hayes and Garland 1995).
The metabolic rates of daily torpor (TMR) for Sturnira bidens and S. erythromos
(0.42 and 0.63 mL O2 g−1 h−1, respectively) lead to a significant greater energy sav-
ing (between 69% and 71%, respectively). At Ta = 17–18 °C, we calculated a Q10
between the basal metabolic rate (BMR) and torpor rate (TMT) of 2.68 for S. bidens
and of 2.1 for S. erythromos, at body temperature of 22 °C, values that are typical
for biochemical reactions (Q10 = 2–3; Schmidt-Nielsen 1990). Thus, temperature
dependence and not physiological inhibition seems to be the factor involved in low-
ering the metabolic rate of S. bidens and S. erythromos, compared with small hiber-
nating species (Bartels et al. 1998; Geiser 1988; Hosken and Withers 1999). The
body temperature of these torpid bats seems to be related to ambient temperature
during torpor (17–18 °C). In Andean cloud forests, the minimum air temperature
remains below 20 °C, which resulted in energetic savings, but at Ta < 18 °C, Tb and
metabolic rates of S. bidens were more variable and tended to increase, resulting in
smaller energy savings (Fig. 8.1). Therefore, both species will be able to live in
environments in which ambient temperatures do not decrease below the torpor tem-
perature and where the highest energy saving is given. The use of adaptive hetero-
thermy in the smallest species of fruit bats is not related with BMRs, which fall both
above and below the boundary line of endothermy (Cooper and Geiser 2008;
McNab 1983).
It is very hard to predict the impact of food habits on basal metabolic rate, since
the specific food habits of some species are poorly known, and they are both geo-
graphic and seasonally variable (McNab 1986,2003). In the Venezuelan cloud for-
ests, these sympatric species feed strictly on Piper, Vismia, and Solanum species
and Araceae family (Ruiz 2006); the higher BMRs of fruit-eater bats have been
associated with the consumption of highly nutritional fruits (McNab 2003). For
example, the high metabolic rates of S. bogotensis and S. ludovici may be associated
with feeding on fruits of Solanum and Piper, respectively (Ruiz 2006). In contrast,
both S. erythromos and S. bidens, which feed on Solanum and Araceae species, both
use torpor (Ruiz 2006). Therefore, the quality of the fruits is not the only factor
responsible for these differences (Dinerstein 1986; Herbst 1983 citated by Fleming
1988; Herbst 1985). However, if fruits are scarce during certain periods (Fleming
et al. 1993; McNab 2003; Morrison 1980; Thomas 1984), one possible response
would be to decrease the energy expenditure (i.e., decrease BMR and/or enter in
torpor). The fruits of Solanum and Piper are available all year round, while Araceae
fruits are seasonal (Soriano 1983). It is possible that the endothermy in these species
will be associated with resource availability; otherwise, they would enter torpor
(Delorme and Thomas 1996,1999). Indeed, differences in body size between these
three species and also S. erythromos (Soriano et al. 2002) could explain the fast
depletion of energy stores of species with body mass lower than 17 g (such as
S. bidens and S. erythromos), whereas those with body mass larger than 20 g (such
as S. bogotensis and S. ludovici) are able to maintain normothermy for long time.
One of the physiological consequences of a frugivorous diet could be a higher
digestive efficiency to regulate fruit intake in relation to protein intake (Delorme
and Thomas 1996,1999; Dumont et al. 1999; Herrera et al. 2002; Korine et al. 1999;
Schondube and Martínez del Rio 2004;Thomas 1984). Many birds and mammals
increase food ingestion rates when acclimated to cold temperatures (Hammond and
Diamond 1997). However, we do not have any evidence that such changes in the
digestive function are operating in Sturnira.
The dependence of basal metabolic rates on body mass can be observed in all
Neotropical fruit bats (F = 134.6; P = 0.0001; r2 = 0.91; N = 16). This relationship
described by the equation: V02 (mL O2 h−1) = 3.66 m0.74, whose slope slightly differs
from that obtained by McNab (1988) for all the mammals (0.713). Aside from body
mass, other factors such as phylogeny, foraging strategy, food preferences, and
roosting sites, have been used to explain the variations of BMR in phyllostomid bats
(Bonaccorso and McNab 2003; Cruz-Neto and Bozinovic 2004; Cruz-Neto et al.
2001; McNab 2003; McNab and Bonaccorso 2001). We used altitude, torpor, and
taxonomic affinities, yet their basal metabolic rate was not correlated with any of
these factors (ANCOVA; F = 0.31; P = 0.59). Other factors proposed by McNab
(2003) were not analyzed, since we did not consider that they provided accurate
data to carry out intra- and inter-specific comparisons. Our data demonstrated that
members of genus Sturnira seem to evolve different adaptations that allowed them
to face a single environmental factor (such as ambient temperature). An examina-
tion on the composition of their diets, availability, seasonality of resources, and
roost microclimate will be required to explain these differences (i.e., growth and
reproduction).
Thermal conductance. The normothermic species studied showed minimal con-
ductances similar to lowland fruit-eater bats (ANCOVA; mass, F = 7.31; P = 0.022;
altitude, F = 0.173; P = 0.67; N = 13; Table 8.1). The length of the fur could explain
partiality in the variations observed among Sturnira spp. For example, the individu-
als of S. ludovici exhibited the shortest fur, coinciding with a highest minimal con-
ductance (112%), while S. bidens and S. bogotensis showed the longest fur with a
lowest conductance (79 and 95%, respectively). Torpid S. bidens exhibited lower
conductance values than normothermic individuals. Lower conductance during tor-
por (50% of expected) has been observed in insectivorous bat species (Cryan and
Wolf 2003; Genoud 1993; Hosken and Withers 1997,1999). It is unclear why
conductance is sometimes lower during torpor, but possible explanations include
changes in breathing rates, posture, and circulation (Hosken and Withers 1997,1999).
The roosting sites used by Andean Sturnira species are not well known. Until the
present, it has been reported that at altitudes of 1800 m, S. bidens and S. aratatho-
masi possibly use caves (Tamsitt et al. 1986), while lowland forest dwellers such as
S. lilium live in hollow trunks, palms, or lianas (Fenton et al. 2000). In the study
area, we found individuals of Carollia brevicauda living in hollow trunks (A. Ruiz,
pers. obs.). Therefore, we consider probable that Sturnira also is using these kinds
of roosts. We have estimated that average air temperature values inside these even-
tual roosts are 15.4 °C ± 0.4 (A. Ruiz, pers. obs.). Nevertheless, this value is below
the lower limit of thermoneutrality. It is probable that gregarious behavior observed
in captivity is also present under natural conditions as a mechanism that allows them
to reduce heat loss, mainly in smaller species (Ashton et al. 2000; Mayr 1963;
McNab 1971; Studier 1970; Thomas and Cloutier 1992).
The influence of body mass on minimal conductance in Neotropical frugivores is
described by the equation C = 1.72 m0.324, where 42% of the variation in conduc-
tance is explained by their mass (ANCOVA; F = 7.94; P = 0.017; N = 13; Table 8.1).
Most of these fruit-eater bats showed conductance values below that was expected
for mammals and were not associated to the basal metabolic rates (ANCOVA;
F = 0.88; P = 0.37; N = 13), roosts (ANCOVA; F = 0.47; P = 0.79; N = 13; Table 8.1),
and ambient temperature (ANCOVA; F = 4.12; P = 0.21). The lowest conductance
observed in small fruit-eater bats (<50 g) of high- and lowlands could explain their
ability to maintain their body temperature above ambient temperature without mak-
ing adjustments in their metabolic rates. However, in torpid individuals it was not
enough to maintain the normothermy, as a consequence of their high surface/vol-
ume ratio values.
Altitudinal distribution and physiological limits: Sturnira erythromos and
S. ludovici occur throughout Neotropical mountains and penetrate subtropical lati-
tudes (Gannon et al. 1989; Giannini and Barquez 2003; Simmons 2005). In the
northern Andes of Argentina, S. erythromos is sympatric with another species of the
genus, such as S. lilium, which was later restricted to lowlands (<1200 m), while the
former replaced it at higher elevations (Giannini 1999). Seasonality and distribution
of resources in subtropical latitudes should explain the altitudinal segregation of
both species (Giannini 1999); however, if they are eating similar fruits, why are they
primarily restricted to tropical highlands and lowlands, respectively? Ecological or
historical factors may be utilized for answering it, but the similarities in their diet
would be correlated with their physiology. Nevertheless, both species with high
BMR (between 126% and 130% of expected) show differences when the tempera-
ture is below the lower critical temperature. The use of torpor at lower temperatures
by S. erythromos may be a physiological trait that has been evolved mass dependent
in some highland fruit-eater bats.
However, S. ludovici showed metabolic rates above those registered in lowland
fruigivorous bats; in fact, it was the only of the three studied species with a broad
altitudinal distribution (500–3000 m). Although S. ludovici can be considered a
complex species (Timm and LaVal 1998), in the Andes of Venezuela this species is
below 2200 m of elevation, yet we have captured at higher elevations (>2000 m). Its
intermediate size could facilitate altitudinal movements in search of resources that
may satisfy their high-energy and nutritional requirements. In contrast, although

S. bogotensis and S. bidens are restricted mainly to tropical highlands (>2000 m;
Molinari and Soriano 1987; Simmons 2005), their thermoregulatory patterns differ
in relation to their body mass. The absence of daily torpor in S. bogotensis could be
related with its capacity for maintaining high BMR, thus avoiding heat loss due to a
smaller relationship surface to volume ratio (Table 8.1).
The ranges of altitudinal distribution of the Sturnira spp. are wide, and few spe-
cies are exclusive of the highlands (>2000 m). There are contradictory hypotheses
concerning the evolutionary history and ancestral origin of genus Sturnira: one
ancestor of lowland or one ancestor of highland (de la Torre 1961; Iudica 2000;
Koopman 1982; Owen 1990; Pacheco and Patterson 1991; Villalobos and Valerio
2002). If this genus evolved in the mountain environments, the torpor must be a
primitive character, because it first appeared in S. bidens, which is a member of
basal group (Pacheco and Patterson 1991; Villalobos and Valerio 2002). However,
the torpor in genus Sturnira seems to have evolved twice, once within the clade that
belongs to S. bidens (coming from a highland ancestor) and again in the clade that
belongs to S. erythromos (coming from a lowland ancestor). Herein, phylogeny and
food habits cannot be the explanation since S. bogotensis (a brother taxon of
S. erythromos) does not use the torpor (if this were adaptive), under the assumption
that they are living under similar environmental pressures. After all, a species is
unlikely to maintain characteristics that are incompatible with a larger size and
higher BMR.
In addition, both adaptation to low temperatures and restricted altitudinal distri-
bution are not explained by the heterothermy. There are several evidence that show
that the torpor is not associated to the altitude, as we did think. Some studies of
fruit-eater bats of the old world (Pteropodidae) that inhabit in lowland show a simi-
lar pattern with the one observed in Sturnira, in which smaller species (23-40 g) are
also poor thermoregulators and torpor is infrequent. In contrast, intermediate and
larger species (73–1030 g) reach montane ecosystems due to their appropriate ther-
moregulatory capacity (Bartels et al.1998; Bonaccorso and McNab 1997; Law
1994; McNab and Armstrong 2001; McNab and Bonaccorso 2001). The choice of a
body size limit for small fruit bats that enter or not in torpor requires more objectiv-
ity. The recognition of small and large bat body size classes has important implica-
tions in metabolic rates, the evolution of life histories, and ultimately, fitness in
these species (Lovegrove 2005).
The physiologic responses of these species to face environmental temperature
fluctuations depend on their body size. The dichotomous thermoregulatory responses
used by the smallest species with the genus Sturnira as well as the normothermy
observed in the largest ones suggest that its fruit-based diet does not compensate the
energy cost of the thermoregulation at highlands for all of them. Therefore, the ulti-
mate factor that explains the thermoregulatory patterns used by these species is their
body mass. We suggest that the adaptation of these species to Andean highlands
depends on their thermoregulatory capacity, conditioned by their body size and,
probably, by the availability and quality of fruits that they consume. However, the
apparent physiologic inability to reduce the metabolic rates and their body tempera-
ture when they are exposed to low temperatures could explain their restricted distri-
bution to the tropical highlands.
Acknowledgments We thank A. Arends and C. Bosque who provided useful comments about an
earlier version of this manuscript, and G. Alba y C. Santiago helped us with the equipment and
software installation. J. Murillo, M. Machado, and C. Cabrera and all the students of Animal
Ecology Laboratory provided field and laboratory assistance. F. Ely helped us with the English
revision. We received partial financial support through the grant from Red Latinoamericana de
Botánica (RLB-Chile), Scott Neotropical Fund of Cleveland Metroparks Zoo, Latin American
Fellowship of the American Society of Mammalogist, Consejo de Estudios Científicos y
Humanísticos (ULA, Venezuela; Project No. C-1097-01-01-ED), Postgrados Integrados en
Ecología-FONACYT (Venezuela), Idea Wild Assistance Support, and COLCIENCIAS (Colombia).
References
Alberico M, Cadena A, Hernández-Camacho J, Muñoz-Saba Y (2000) Mamíferos (Synapsida:

Theria) de Colombia. Biota Colombiana 1(1):43–75
Aschoff J, Pohl H (1970) Der Ruheumsatz von Vögeln als Funktion der Tageszeit und Körpergröβe.
J Ornithol 111:38–47
Ashton KG, Tracy MC, Queiroz A (2000) Is Bergmann’s rule valid for mammals? Am Nat
156(4):390–415
Audet D, Thomas DW (1997) Facultative hypothermia as a thermoregulatory strategy in the phyl-
lostomid bats, Carollia perspicillata and Sturnira lilium. J Comp Physiol B 167:146–152
Bartels W, Law BS, Geiser F (1998) Daily torpor and energetic in a tropical mammal, the northern
blossom-bat Macroglossus minimus (Megachiroptera). J Comp Physiol B 168:233–239
Bartholomew GA, Dawson WR, Lasiewski RC (1970) Thermoregulation and heterothermy in
some of smaller flying foxes (Megachiroptera) of New Guinea. J Comp Physiol 70:391–404
Bonaccorso FJ, McNab BK (1997) Plasticity of energetic in blossom bats (Pteropodidae): impact
on distribution. J Mammal 78(4):1073–1088
Bonaccorso FJ, McNab BK (2003) Standard energetic of leaf-nosed bats (Hipposideridae): its
relationship to intermittent- and protracted-foraging tactics in bats and birds. J Comp Physiol
B 173:43–53
Bonaccorso FJ, Arends M, Genoud D, Cantoni D, Morton T (1992) Thermal ecology of mous-
tached and ghost-faced bats (Mormoopidae) in Venezuela. J Mammal 73(2):365–378
Canterbury G (2002) Metabolic adaptation and climatic constraints on winter bird distribution.
Ecology 83(4):946–957
Coburn DK, Geiser F (1998) Seasonal changes in energetic and torpor patterns in the subtropical
blossom-bat Syconycteris australis (Megachiroptera). Oecologia (Berlin) 113:467–473
Colwell RK, Lees DC (2000) The mid-domain effect: geometric constrains on the geographic of
species richness. Trends Ecol Evol 15:70–76
Contreras-Vega M, Cadena A (2000) Una nueva especie del género Sturnira (Chiroptera:
Phyllostomidae) de los Andes colombianos. Revista de la Academia Colombiana de
CienciasXXIV 91:285–287
Cooper CE, Geiser F (2008) The “minimal boundary curve for endothermy” as a predictor of het-
erothermy in mammals and birds. Rev J Comparative Physiol 178:1–8
Cruz-Neto AP, Abe AS (1997) Taxa metabólica e termorregulaçâo no morcego nectarívoro,
Glossophaga soricina (Chiroptera, Phyllostomidae). Rev Bras Biol 57(2):203–209
Cruz-Neto AP, Bozinovic F (2004) The relationship between diet quality and basal metabolic rate
in endoterms: insights from intraspecific analysis. Physiol Biochem Zool 77(6):877–889
Cruz-Neto AP, Garland TJ, Abe AS (2001) Diet, phylogeny, and basal metabolic rate in phyllosto-
mid bats. Fortschr Zool 104:49–58
Cryan PM, Wolf BO (2003) Sex differences in the thermoregulation and evaporative water loss of
a heterothermic bat, Lasiurus cinereus, during its spring migration. J Exp Biol 206:3381–3390
De laTorreL (1961) The evolution variation and systematic of the Neotropical bats of the genus
Sturnira. Dissertation, University of Ilinois, Urbana
Delorme M, Thomas DW (1996) Nitrogen and energy requirements of the short-tailed fruit bat
(Carollia perspicillata): fruit bats are not nitrogen constrained. J Comp Physiol B 166:427–434
Delorme M, Thomas DW (1999) Comparative analysis of the digestive efficiency and nitrogen and
energy requirements of the phyllostomid fruit-bat (Artibeus jamaicensis) and the pteropodid
fruit-bat (Rousettus aegyptiacus). J Comp Physiol B 169:123–132
Dépocas F, Hart JS (1957) Use of the Pauling oxygen analyzer for measurement of oxygen con-
sumption of animals in open- circuit systems and in a short-lag, closed-circuit apparatus. J Appl
Physiol 10(3):388–392
Dinerstein E (1986) Reproductive ecology of fruit bats and the seasonality of fruit production in a
Costa Rican cloud forest. Biotropica 18(4):307–318
Dumont ER, Etzel K, Hempel J (1999) Bat salivary proteins segregate according to diet. Mammalia
63:159–166
Fenton MB, Vonhof MJ, Bouchard S, Gill SA, Johnston DS, Reid FA, Riskin DK, Standing KL,
Taylor JR, Wagner R (2000) Roots used by Sturnira lilium (Chiroptera: Phyllostomidae) in
Belize. Biotropica 32(4a):729–733
Fleming TH (1988) The short-tailed fruit bat.A study in plant-animal interactions. The University
of Chicago Press, Chicago
Fleming TH, Nuñez RA, Sternberg L (1993) Seasonal changes in diets of migrant and non-migrant
nectarivorous bats as revealed by carbon stable isotope analysis. Oecologia (Berlin) 94:72–75
Gannon MR, Willig MR, Knox Jones JJ (1989) Sturnira lilium. Mamm Species 333:1–5
Gannon WL, Sikes RS, The Animal Care and Use Committee of the American Society of
Mammalogists (2007) Guidelines of the American Society of Mammalogists for the use of
wild mammals in research. J Mammal 88:809–823
Geiser F (1988) Reduction of metabolism during hibernation and daily torpor in mammals and
birds: temperature effect or physiological inhibition? J Comp Physiol B 158:25–37
Geiser F, Coburn DK (1999) Field metabolic rates and water uptake in the blossom-bat Syconycteris
australis (Megachiroptera). J Comp Physiol B 169:133–138
Geiser F, Coburn DK, Körtner G (1996) Thermoregulation, energy metabolism, and torpor in
blossom-bats, Syconycteris australis (Megachiroptera). J Zool 239:583–590
Genoud M (1993) Temperature regulation in subtropical tree bats. Comparative Biochem Physiol
104A:321–331
Genoud M, Bonaccorso FJ (1986) Temperature regulation rate of metabolism, and roost tem-
perature in the greater white-lined bat Saccopteryx bilineata(Emballonuridae). Physiol Zool
59(1):49–54
Giannini NP (1999) Selection of diet and elevation by sympatric species of Sturnira in an Andean
rainforest. J Mammal 80(4):1186–1195
Giannini NP, Barquez RM (2003) Sturnira erythromos. Mamm Species 729:1–5
Graham GL (1983) Changes in bat species diversity along an elevational gradient up the peruvian
Andes. J Mammal 64(4):559–571
Graham GL (1990) Bats versus birds: comparisons among Peruvian volant vertebrates faunas
along an elevational gradient. J Biogeogr 17:657–668
Grodzinski W, Wunder BA (1975) Ecological energetic of small mammals. In: Golley FB,
Petrusewics K, Ryszkowiski L (eds) Small mammals, their productivity and population dynam-
ics. Cambridge University Press, Cambridge
Hammond KA, Diamond J (1997) Maximum sustained energy budgets in humans and animals.
Nature 386:457–462
Hayes JP, Garland T Jr (1995) The evolution of endothermy: testing the aerobic capacity model.
Evolution 49(5):836–847
HerbstLH (1983) Nutritional analyses of the wet season diet of Carollia perspicillata (Chiroptera:
Phyllostomidae) in Parque Nacional Santa Rosa, Costa Rica. Thesis, University of Miami
Herbst LH (1985) The role of nitrogen from fruit pulp in the nutrition of a frugivorous bat, Carollia
perspicillata. Biotropica 18:39–44
Herreid CF, Kessel B (1967) Thermal conductance in birds and mammals. Comparative Biochem
Physiol 21:405–414
Herrera L, Gutierrez E, Hobson KA, Altube B, Díaz WC, Sánchez-Cordero V (2002) Sources
of assimilated protein in five species of new world frugivorous bats. Oecologia (Berlin)
133:280–287
Hosken DJ, Withers PC (1997) Temperature regulation and metabolism of an Australian bat,
Chalinolobus gouldii (Chiroptera: Vespertilionidae) when euthermic and torpor. J Comp
Physiol B 167:71–80
Hosken DJ, Withers PC (1999) Metabolic physiology of euthermic and torpid lesser long-eared
bats, Nyctophilus geoffroyi (Chiroptera: Vespertilionidae). J Mammal 80(1):42–52
IudicaCA (2000) Systematic revision of the neotropical fruit bats of the genus Sturnira: a molecu-
lar and morphological approach. Dissertation, University of Florida, FL,p 284
Kendeigh S (1970) Energy requirements for existence in relation to size of bird. Condor 72:60–65
Koopman KF (1978) Zoogeography of Peruvian bats with special emphasis of the role Andes. Am
Mus Novit 2651:1–33
Koopman KF (1982) Biogeography of bats of South America. In: Mares MA, Genoways HH (eds)
Mammalian biology in South America, vol 6. Pymatuning Laboratory of Ecology, University
of Pittsburgh, Pittsburgh
Korine C, Zinder O, Arad Z (1999) Diurnal and seasonal changes in blood composition of the free-
living Egyptian fruit bat (Rousettus aegyptiacus). J Comp Physiol B 169:280–286
Kovtun MF, Zhukova NF (1994) Feeding and digestive intensity of chiropterans of different tro-
phic groups. Folia Zool 43:377–386
Law BS (1994) Climatic limitation of the southern distribution of the common blossom bat
Syconycteris australis in New South Wales. Aust J Ecol 19:366–374
Linares OJ (1998) Mamíferos de Venezuela. Sociedad Conservacionista Audubon de Venezuela,
Caracas
Lomolino MV (2001) Elevation gradients of species-density: historical and prospective views.
Glob Ecol Biogeogr 10:3–13
Lovegrove BG (2005) Seasonal thermoregulatory responses in mammals. J Comp Physiol B
175:231–247
Mayr E (1963) Animal species and evolution. Harvard University Press, Cambridge
McManus JJ (1977) Thermoregulation. In: Baker RJ, JKJ J, Carter DC (eds) Biology of bats of the
New World family Phyllostomidae, vol 2. Texas Tech Press, Lubbock
McNab BK (1969) The economics of temperature regulations in Neotropical bats. Comparative
Biochem Physiol 31:227–268
McNab BK (1970) Body weight and the energetic of temperature regulation. J Exp Biol 53:329–348
McNab BK (1971) On the ecological significance of Bergmann’s rule. Ecology 52(5):845–854
McNab BK (1973) Energetic and the distribution of vampires. J Mammal 54:131–144
McNab BK (1974) The behavior of temperate cave bats in a subtropical environment. Ecology
55:943–958
McNab BK (1976) Seasonal fat reserves o bats in two tropical environments. Ecology 57:332–338
McNab BK (1980a) Food habits, energetic, and the population biology of mammals. Am Nat
116(1):106–124
McNab BK (1980b) On estimating thermal conductance in endotherms. Physiol Zool 53(2):145–156
McNab BK (1982a) Evolutionary alternatives in the physiological ecology of bats. In: Kunz TH
(ed) Ecology of bats. Plenum, New York
McNab BK (1982b) The physiological ecology of South American mammals. In: Mares MA,
Genoways HH (eds) Mammalian biology in South America. Pymatuning Laboratory of
Ecology, University of Pittsburgh, Pittsburgh
McNab BK (1983) Energetic, body size, and the limits to endothermy. J Zool 199:1–29
McNab BK (1986) The influence of food habits on the energetic of eutherian mammals. Ecol
Monogr 56(1):1–9
McNab BK (1988) Complications inherent in scaling the basal rate of metabolism in mammals.
Q Rev Biol 63:25–53
McNab BK (1989) Temperature regulation and rate metabolism in three bornean bats. J Mammal
70(1):153–161
McNab BK (2003) Standard energetic of phyllostomid bats: the inadequacies of phylogenetic-
contrast analyses. Comparative Biochem Physiol 135:357–368
McNab BK, Armstrong MI (2001) Sexual dimorphism and the scaling of energetic in flying foxes
of the genus Pteropus. J Mammal 82(3):709–720
McNab BK, Bonaccorso FJ (2001) The metabolism of new Guinean pteropodid bats. J Comp
Physiol B 171:201–214
Molinari J, Soriano PJ (1987) Sturnira bidens. Mamm Species 276:1–4
Morrison DW (1980) Efficiency of food utilization by fruit bats. Oecologia (Berlin) 45:270–273
Owen JG (1990) Patterns of mammalian species richness in relation to temperature, productivity,
and variance in elevation. J Mammal 71(1):1–13
Pacheco V, Patterson BD (1991) Phylogenetic relationships of the New World bat genus Sturnira
(Chiroptera: Phyllostomidae). Bull Am Mus Nat Hist 206:101–121
Patterson BD, Pacheco V, Solari S (1996) Distributions of bats along an elevational gradient in the
Andes of South-Eastern Peru. J Zool 240:637–658
Rahbek C (1995) The elevational gradient of species richness: a uniform pattern? Ecography
18:200–205
Rosenzweig ML (1992) Species diversity gradients: we know more and less than we thought. J
Mammal 73:715–730
RuizA (2006) Termoregulación, recursos y límites altitudinales en murciélagos frugívoros y
nectarívoros andinos. Dissertation, Postgrado en Ecología Tropical, Facultad de Ciencias,
Universidad de Los Andes, Mérida, Venezuela, p 215
Sarmiento G (1986) Ecological features of climate in high tropical mountains. In: Vuilleuimier F,
Monasterio M (eds) High tropical biogeography. Oxford University, New York
Schmidt-Nielsen K (1990) Physiology: adaptation and environment, 5th edn. Cambridge
University Press, Cambridge
Schondube J, Martínez del Rio C (2004) Sugar and protein digestion in flower piercers and hum-
mingbirds: a comparative test of adaptive convergence. J Comp Physiol B 174:263–273
Simmons NB (2005) Order Chiroptera. In: Wilson DE, Reeder DM (eds) Mammals species of
the world: a taxonomic and geographic reference, 3rd edn. Johns Hopkins University Press,
Baltimore
SorianoPJ (1983) La comunidad de quirópteros de las selvas nubladas en los Andes de Mérida.
Patrón reproductivo de los frugívoros y las estrategias fenológicas de las plantas. Tesis de
Magister Scientiae, Facultad de Ciencias, Universidad de Los Andes, Mérida
Soriano PJ (2000) Functional structure of bat communities in tropical rain forests and Andean
cloud forests. Ecotropicos 13(1):1–20
Soriano PJ, Díaz de Pascual A, Ochoa J, Aguilera M (1999) Biogeographic analysis of the mam-
mal communities in the Venezuelan Andes. Dermatol Int 24(1):17–25
Soriano PJ, Ruiz A, Arends A (2002) Physiological responses to ambient temperature manipula-
tion by three species of bats from Andean cloud forest. J Mammal 83(2):445–457
Speakman JR, Thomas DW (2003) Physiological ecology and energetic of bats. In: Kunz TH,
Fenton MB (eds) Bat ecology. University of Chicago Press, Chicago
Stevens RD (2004) Untangling latitudinal richness gradients at higher taxonomic levels: familial
perspectives on the diversity of new world bat communities. J Biogeogr 31:665–674
Stevens RD (2006) Historical processes enhance patterns of diversity along latitudinal gradients.
Proc Roy Soc 273:2283–2289
Studier EH (1970) Evaporative water loss in bats. Comparative Biochem Physiol 35:935–943
Studier EH, Wilson DE (1970) Thermoregulation in some neotropical bats. Comparative Biochem
Physiol 34:251–262
Tamsitt JR, Cadena A, Villarraga E (1986) Records of bats (Sturnira magna and Sturnira aratatho-
masi) form Colombia. J Mammal 67(4):754–757
Terborgh J (1971) Distribution on environmental gradients: theory and a preliminary interpre-
tation of distributional patterns in the avifauna of the cordillera Vilcabamba, Peru. Ecology
52(1):23–40
Terborgh J (1977) Bird species diversity on an Andean elevation gradient. Ecology 58:1007–1019
Thomas DW (1984) Fruit intake and energy budgets of frugivorous bats. Physiol Zool
57(4):457–467
Thomas DW, Cloutier D (1992) Evaporative water loss by hibernating little brown bats, Myotis
lucifugus. Physiol Zool 65(2):443–456
Tilman D (1982) Resource competition and community structure. Princeton University Press,
Princeton
Timm RM, LaVal RK (1998) A field key to the bats of Costa Rica. Occ Publ Ser 22:1–24
Tuttle MD (1970) Distribution and zoogeography of Peruvian bats, with comments on natural his-
tory. Univ Kansas Sci Bull 49:45–86
Villalobos F, Valerio AA (2002) The phylogenetic relationships of the bat genus Sturnira gray,
1842 (Chiroptera: Phyllostomidae). Mamm Biol 67:268–275
Willig MR, Bloch CP (2006) Latitudinal gradients of species richness: a test of the geographic area
hypothesis at two ecological scales. Oikos 112:163–173
Willig MR, Patterson BD, Stevens RD (2003) Patterns of range size, richness, and body size in the
Chiroptera. In: Kunz TH, Fenton MB (eds) Bat ecology. University of Chicago Press, Chicago
Zar JH (1999) Biostatistical analysis, 4th edn. Prentice Hall, New Jersey
Chapter 9
Neotropical Biodiversity: Hypotheses
of Species Diversification and Dispersal
in the Andean Mountain Forests
Angela M. Mendoza-Henao and Juan C. Garcia-R
9.1 Introduction
It is well established that species diversity is unevenly distributed across the globe
(Wallace 1876) and a high proportion of it occurs in the tropics (Gaston 2000), par-
ticularly in the Andean mountain forests. The Andes hosts an exceptional diversity
and endemism of vertebrate biota. More than 3400 vertebrate species (excluding
fishes) are found in this mountain region and almost half of them are endemic
(Myers et al. 2000; Orme et al. 2005; Elsen et al. 2018). The substantial number of
species occurring in the Andes is a reflection of high beta diversity and spatial turn-
over at local and regional scales (Jankowski et al. 2009; Sklenář et al. 2014).
However, this high biodiversity raises questions about the relevance of the different
driving factors associated with species diversification and dispersal events.
Several hypotheses have been suggested to help explain the high diversity in the
Andes (Table 9.1). Species richness and endemism are attributed to a mixture of
historical, evolutionary, and ecological processes (Rahbek and Graves 2001; Rahbek
et al. 2007, 2019a; Fjeldså et al. 2012; Rangel et al. 2018). The processes alone are
not mutually exclusive but instead form a complex dimensional interrelationship for
the observed spatial and temporal patterns of species diversity in the Andes region.
Thus, environmental heterogeneity, in both time and space, and evolutionary his-
tory, seems to be relevant for understanding the origin and distribution of extant
biodiversity (Rull 2011; Antonelli et al. 2018; Perrigo et al. 2020). The mechanisms
of diversification differ based on the relative role of each process and the
A. M. Mendoza-Henao
Departamento de Zoología, Instituto de Biología, Universidad Nacional Autónoma de
México, Mexico City, México
J. C. Garcia-R (*)
Molecular Epidemiology and Public Health Laboratory, Hopkirk Research Institute, School
of Veterinary Science, Massey University, Palmerston North, New Zealand

https://doi.org/10.1007/978-3-030-57344-7_9
178 A. M. Mendoza-Henao and J. C. Garcia-R
Table 9.1 Some of the hypotheses that explain the biodiversity in the Andes region
Category Hypothesis Summary description References
Evolutionary Time-for- Earlier colonization of regions or habitats Stephens and
speciation facilitates higher richness because they Wiens (2003)
(mountain have more time for in situ speciation and
museum accumulation of species
hypothesis)
Diversification Faster diversification rates (i.e., Smith et al.
rate (species speciation minus extinction over time) of (2007)
pump) clades in some regions and habitats can
generate higher species richness
Spatial and Pleistocene Pleistocene oscillations caused Turchetto-Zolet
ecological oscillations fragmentation of species ranges and local et al. (2013)
adaptation of populations
Andean Andean uplift created isolated “sky Hoorn et al.
orogenesis islands” surrounded by drier valleys, (2010)
which act as barriers to the dispersal of
species
Across-Andes Amazonian species episodically “spilled Haffer (1974)
dispersal over” into western South America during
humid periods, across low passes in the
Andean region
Environmental Heterogeneous environments at Rosenzweig
heterogeneity mid-elevations can harbor more species, (1995), Fjeldsaå
enhance species persistence, and promote and Lovett
adaptive radiations because they can have (1997), Sedano
a rich array of suitable conditions and Burns (2010)
Climatic zonation Narrow thermal tolerances of tropical Janzen (1967)
by thermal species and greater climatic stratification
tolerances of tropical mountains create
opportunities for climate-associated
parapatric and allopatric speciation
Climatic-niche Accelerated rates of climatic-niche Kozak and Wiens
divergence evolution in montane areas are associated (2010)
with increased diversification rates
lineage-specific attributes. However, the precise association of processes and traits

is often unclear.
Multiple competing hypotheses attempt to explain the diversity of vertebrates in
the tropical montane forest of the Andes. Proposed drivers of diversification in the
tropical Andes can be grouped into spatial (e.g., species–area relationship, mid-
domain effect), climatic and topographic (e.g., temperature, topographical gradi-
ents), historical (e.g., speciation rates, clade age) and biotic (e.g., competition,
thermal tolerance) hypotheses. In this chapter, we summarize some of the leading
hypotheses explaining the origin and distribution of vertebrate clades in the Andean
montane forests. Nonetheless, these are non-exclusive hypotheses to vertebrates and
can also be used to explain the high diversity among other groups in the tropical
Andes, including invertebrates and plants.
9 Neotropical Biodiversity: Hypotheses of Species Diversification and Dispersal… 179
9.2 Drivers of Species Diversification in the Tropical Andes
9.2.1 Shape
The simplest explanation to the observed diversity in the montane forest of the
Andes, without the need to call for any particular evolutionary nor geologic process,
is the spatial pattern (Stein et al. 2014; Rahbek et al. 2019b). The spatial hypothesis
suggests that a reduction in diversity with altitude is only a consequence of more
extensive areas at low elevations than the available area at high elevations. The
Species–Area Relationships (SAR) assumes that more area can bear more species
(Rosenzweig 1995). For example, Rahbek (1997) found that species richness of
birds in South America is controlled by area and remained nearly constant from
lowlands up to 2600 m. However, more recent studies have shown that these lineal
trends are obtained when the extent of the gradients sampled do not cover all the
mountain system (Nogués-Bravo et al. 2008).
Several lineages of vertebrates in the mountain regions exhibit a hump-shaped or
mid-elevational species richness peak (Kattan and Franco 2004; Meza-Joya and
Torres 2016; Hutter et al. 2017; Quintero and Jetz 2018). The pattern has been
observed in many mountain systems. Grinnell and Storer (1924) showed that sev-
eral groups of terrestrial vertebrates have a unimodal richness pattern with the high-
est species richness about a third of the way up the Yosemite mountain. Nevertheless,
this pattern is less likely to be found for small-ranged species (Colwell and Hurtt
1994; Colwell et al. 2004). An explanation for the observed richness pattern is the
mid-domain effect (MDE) hypothesis. This hypothesis states that any random dis-
tribution of species on a bounded map (such as an elevational gradient) produces a
peak of species richness near the center due to geometric boundary constraints
(Colwell et al. 2004). The model predicts a parabolic curve of species richness, with
a peak of 0.5 times the total number of species at the center of the gradient (Bokma
et al. 2001). Despite several studies supporting the mid-domain effect, others have
found that the model is not a general explanation for the diversity patterns observed
(Hawkins and Diniz-Filho 2002; Dunn et al. 2007) but it is still used as a null
hypothesis to test more complex scenarios.
9.2.2 Climate
Another important component influencing Andean biodiversity is the geographical

location in the tropical region. The tropics are favored by gathering direct solar
energy throughout the year, which provides a stable condition and higher annual
productivity. However, the climate of tropical mountain regions is fundamentally
different in complexity and variety than adjacent lowland regions. The uniquely
heterogeneous environment in mountain systems plays a key role in generating and
maintaining high diversity (Antonelli et al. 2018). The Northern Andes, for
Fig. 9.1 Processes of diversification in Andean biota caused by climate heterogeneity and topographic
unevenness. (a) The elevational zonation of mountains in tropical regions presents more stable climatic
conditions than temperate mountains, enabling small-ranged species and, therefore, more diversity; (b)
The presence of multiple peaks, steep ridges, and inter-valleys act as barriers to dispersal for highland
species, allowing independent speciation events throughout the ranges
example, captures roughly half of the world’s climate types in a relatively small
region, much more than those captured in the entire Amazon (Rahbek et al. 2019b).
Temperature is one of the most studied climatic variables related to biodiversity
patterns because it reflects species’ environmental tolerances. The Janzen’s hypoth-
esis (Janzen 1967) provides an explanation for the effect of yearly temperature
variation in the biodiversity of mountain systems. Tropical mountains show less
climatic variation during the year but a great variety of vegetation belts in each alti-
tudinal band than mountains in template regions (Fig. 9.1a). This variation repre-
sents a significant physiological barrier to dispersal because species would
experience extreme conditions beyond their adaptive or acclimation capabilities
(McCain 2009). Therefore, species living in tropical mountains show narrower
range sizes than temperate species at the same elevation, allowing a high species
turnover (Cadena et al. 2011; Fjeldså et al. 2012).
9.2.3 Topography
Topographic complexity plays a major role in the origin and maintenance of biodi-
versity of montane areas. The Andes show multiple peaks, steep ridges, and inter-
valleys that act as barriers to dispersal (Fig. 9.1b). The topographic unevenness
enables habitat partitioning among species, which favor independent speciation
events at similar range slopes (Hazzi et al. 2018). This might explain the high beta
diversity and endemism observed for low dispersal organisms such as anurans
(Lynch and Duellman 1997; García-R et al. 2012; De la Riva 2020) and small-range
clades of hummingbirds (McGuire et al. 2014). The diversity–environment correla-
tion invokes a “pure ecological” explanation involving population-level processes
tied to dispersal and aggregation.
9.2.4 Geological and Environmental Shifts
The orogeny of the Andes has long been identified as a major driver of speciation,
extinction, and biotic interchange (Hoorn et al. 2010; Kattan et al. 2016; Smith et al.
2014). The uplift of the Andes in the late Eocene around 40 million years ago (Ma)
created new montane environmental conditions that opened up opportunities for
diversification. These new conditions allow the isolation and restriction of gene flow
among populations at both sides of the mountain range (Fig. 9.2a). The relation
between Andes uplift and species formation has been studied in multiple groups of
birds (e.g., parrots, Ribas et al. 2007; tanagers, Sedano and Burns 2010; woodpeck-
ers, Weir and Price 2011; hummingbirds, McGuire et al. 2014) and frogs (e.g., poi-
son frogs, Santos et al. 2009; glassfrogs, Castroviejo-Fisher et al. 2014;
direct-developing frogs, Mendoza et al. 2015). For instance, Mendoza et al. (2015),
using phylogenetic and ancestral range analyses, reconstructed the role of Andean
orogeny in the diversification of direct-developing frogs (Pristimantis) across the
Neotropics. The genus-level phylogeny and ancestral range reconstruction suggest
that the uplift between north and central Andes generated a bridge for dispersal of
species and high speciation in mid-elevational ranges (1000–3000 masl).
Besides the tectonic uplift of the Andes, the mountain system suffered multiple
glacial cycles during the Plio-Pleistocene (Gregory-Wodzicki 2000; Graham 2009;
Rahbek et al. 2019a). The elevational limits of the mountain forest belts moved
upwards and downwards, providing isolated habitats and corridors to populations
distributed across both lowland and pre-Andean landscapes (Fig. 9.2b). The radia-
tion of montane species could have resulted from interchanges between phases of
isolation (prompting local speciation) and connectivity (triggering diversification
through dispersal and settlement in new areas) (Flantua and Hooghiemstra 2018).
As a result, species rapidly diversified (species pump) within a variety of optimum
montane habitats. Shifting climates caused range retractions and expansions of
mountain species with consequences in biotic interactions and diversification.
Fig. 9.2 Processes of diversification of Andean biota by orogenic and climatic changes in the
Andes. (a) The Andean mountain uplift drives speciation (here illustrated by differences in the
color of the frogs) by vicariance, fragmenting species distribution that was formerly continuous (b)
Climatic dynamics during the Pleistocene facilitated temporal dispersal to new habitats and poste-
rior isolation. Biological attributes and interactions (e.g., dispersal ability, thermal tolerance, and
competition) are factors that determine the establishment in new habitats and successive genetic
differentiation during climatic and geological events. Frogs, in this example, have features (e.g.,
dependence on water supply) that might limit species distribution and dispersal
(Ramírez-Barahona and Eguiarte 2013; Beckman and Witt 2015). Climatic cycles
during the Pleistocene comprise one of the main sources of bird diversification in
the Andes (Weir 2006; Gutiérrez-Pinto et al. 2012) and other mountain systems
(Quintero and Jetz 2018).
9.2.5 Evolutionary History
Recent hypotheses have been tested to explain Andean diversity in terms of macro-
evolutionary processes. The time-for-speciation hypothesis suggests that intermedi-
ate elevations have higher richness because they were colonized earlier and had
more time for speciation and species accumulation than those in other elevational
zones (Stephens and Wiens 2003; Hutter et al. 2017). This scenario can occur when
extinction rates are low, due to stable environmental conditions that favor an accu-
mulation of highly adapted species. A second hypothesis, the species pump, states
that higher species richness in mid-elevations is a by-product of faster diversifica-
tion rates i.e., speciation minus extinction over time (Fjeldså 1994; Smith et al.
2007). This scenario is expected when selection for specialization across gradients
(e.g., topographical heterogeneity) occurs, leading to a reduction in gene flow and
subsequent speciation.
The time-for-speciation and species pump hypotheses are suggested as valid

explanations for the distribution of diversity along elevational gradients in the Andes
montane forest but direct evaluations are still required (Stephens and Wiens 2003;
Hawkins et al. 2007, 2012; Smith et al. 2007). Based on a molecular divergence
time estimation, Hutter et al. (2013) suggested that the pattern of mid-elevational
diversity peak in Andean glassfrogs (Centrolenidae) is explained by the time-for-
speciation hypothesis rather than diversification rates (Fig. 9.3). However, Hutter
et al. (2017) found faster diversification rates in some Andean Hyloidea families
when temporal and spatial scales were adjusted. Beckman and Witt (2015) showed
that siskin (Aves: Fringillidae: Spinus) lineages in the Andes had higher diversifica-
tion and dispersal rates than siskin lineages outside the Andes due to the expansion
and contraction of the páramo during the Pleistocene climatic cycles. New
approaches are now providing unprecedented means to quantify the dynamic pro-
cesses of dispersal, speciation, and extinction rates, including phylogenetic infor-
mation and comprehensive databases of geographic species records (e.g.,
López-Aguirre et al. 2018; Garcia-R et al. 2019).
9.2.6 Lineage Life-History Traits
Speciation events are linked to geophysical and climatic changes but the biological
traits of the system under study need also be considered. The ability to disperse,
colonize, and persist in a landscape matrix depends on biotic interactions (e.g.,
competition, predation) and physiological constraints (e.g., thermal tolerance) of
the specific lineage (Hawkins et al. 2007, 2012; Graham et al. 2009; Pintanel et al.
2019). Different approaches have been used to understand the relative roles of life-
history traits in dispersal, diversification and assembly of communities in the Andes
region (Gillespie et al. 2012; Morlon 2014). The existence of greater thermal zona-
tions in the Andes and niche conservatism in some tropical taxa limited the dispersal
and facilitated allopatric isolation that, in turn, results in higher levels of speciation
and accumulation of species (Cadena et al. 2011; Hutter et al. 2013). Niche evolu-
tion has also been inferred as a major contributor to ecological speciation in tropical
mountains after finding major niche shifts between biomes in different Andean
clades (Graham et al. 2004; Kozak and Wiens 2010; Pintanel et al. 2019).
9.3 Conclusion
Biodiversity in the Andes results from a long and complex history mediated by eco-
logical, historical, and evolutionary processes. Mountain building, climate cycling,
and montane environmental gradients played an important part in species richness
Fig. 9.3 Chronogram of glassfrogs (Anura: Centrolenidae). Recent studies suggest that glassfrogs
were ancestrally present in mid-elevation habitats while lower and higher habitats were colonized
more recently. Colored circles at each node represent the ancestral state reconstruction of eleva-
tional distributions. Ultrametric phylogenetic tree (time in Mega-annum or Ma) from Mendoza-
Henao et al. (in prep) and altitudinal ancestral states from Hutter et al. (2013). Species at the right
of the figure from up to down: Hyalinobatrachium esmeralda, Nymphargus grandisonae,
Centrolene hybrida, Teratohyla pulverata, and Espadarana prosoblepon. Photos by AMMH
and spatial patterns in the Andes. Dispersal events are key promoters of speciation,
while intrinsic factors (e.g., thermal tolerance) and biotic interactions determined
differences in species diversity and distribution among clades. Further expansion
and integration of different processes may offer a framework for understanding the
patterns and evolution of diversity in the Andean mountain forests.
Acknowledgments AMMH is supported by the scholarship number CVU416922 from Consejo

Nacional de Ciencia y Tecnología (CONACyT, Mexico), and Posgrado de Ciencias Biológicas of
the Universidad Nacional Autónoma de México (UNAM). We thank Nicolas Hazzi for valuable
comments.
References
Antonelli et al. (2018) Geological and climatic influences on mountain biodiversity. Nature
Geoscience 11(10):718–725. https://www.nature.com/articles/s41561-018-0236-z#citeas
Beckman EJ, Witt CC (2015) phylogeny and biogeography of the new world siskins and gold-
finches: rapid, recent diversification in the Central Andes. Mol Phylogenet Evol 87:28–45
Bokma F, Bokma J, Mönkkönen M (2001) Random processes and geographic species richness
patterns: why so few species in the north? Ecography 24:43–49
Cadena CD, Kozak KH, Gómez JP, Parra JL, McCain CM, Bowie RCK, Carnaval AC, Moritz C,
Rahbek C, Roberts TE, Sanders NJ, Schneider CJ, VanDerWal J, Zamudio KR, Graham CH
(2011) Latitude, elevational climatic zonation and speciation in new world vertebrates. Proc R
Soc B Biol Sci 279:720
Castroviejo-Fisher S, Guayasamin JM, Gonzalez-Voyer A, Vilà C (2014) Neotropical diversifica-
tion seen through glassfrogs. J Biogeogr 41:66–80
Colwell RK, Hurtt GC (1994) Nonbiological gradients in species richness and a spurious rapoport
effect. Am Nat 144:570–595
Colwell R, Xa K, Rahbek C, Gotelli N, Xa J (2004) The mid-domain effect and species richness
patterns:what have we learned so far? Am Nat 163:1–23
De la Riva I (2020) Unexpected beta-diversity radiations in highland clades of andean terrara-
nae frogs. In: Rull V, Carnaval AC (eds) Neotropical diversification: patterns and processes.
Springer, Cham, pp 741–764
Dunn RR, McCain CM, Sanders NJ (2007) When does diversity fit null model predictions? Scale
and range size mediate the mid-domain effect. Glob Ecol Biogeogr 16:305–312
Elsen PR, Monahan WB, Merenlender AM (2018) Global patterns of protection of elevational
gradients in mountain ranges. Proc Natl Acad Sci 115:6004
Fjeldså J (1994) Geographical patterns for relict and young species of birds in Africa and South
America and implications for conservation priorities. Biodivers Conserv 3:207–226
Fjeldså J, Bowie RCK, Rahbek C (2012) The role of mountain ranges in the diversification of
birds. Annu Rev Ecol Evol Syst 43:249–265
Fjeldsaå J, Lovett JC (1997) Biodiversity and environmental stability. Biodivers Conserv 6:315–323
Flantua SGA, Hooghiemstra H (2018) Historical connectivity and mountain biodiversity. In:
Hoorn AC (ed) Mountains, climate and biodiversity. Wiley-Blackwell, Oxford, pp 171–185
García-R JC, Crawford AJ, Mendoza ÁM, Ospina O, Cardenas H, Castro F (2012) Comparative
phylogeography of direct-developing frogs (Anura: Craugastoridae: Pristimantis) in the
Southern Andes of Colombia. PLoS ONE 7:e46077
Garcia-R JC, Gonzalez-Orozco CE, Trewick SA (2019) Contrasting patterns of diversification in a
bird family (Aves: Gruiformes: Rallidae) are revealed by analysis of geospatial distribution of
species and phylogenetic diversity. Ecography 42:500–510
Gaston KJ (2000) Global patterns in biodiversity. Nature 405:220–227
Gillespie RG, Baldwin BG, Waters JM, Fraser CI, Nikula R, Roderick GK (2012) Long-distance
dispersal: a framework for hypothesis testing. Trends Ecol Evol 27:47–56
Graham A (2009) The Andes: a geological overview from a biological perspective. Ann Mo Bot
Gard 96:371–385
Graham CH, Ron SR, Santos JC, Schneider CJ, Moritz C (2004) Integrating phylogenetics and
environmental niche models to explore speciation mechanisms in dendrobatid frogs. Evolution
58:1781–1793
Graham CH, Parra JL, Rahbek C, McGuire JA (2009) Phylogenetic structure in tropical humming-
bird communities. Proc Natl Acad Sci 106:19673
Gregory-Wodzicki KM (2000) Uplift history of the Central and Northern Andes: a review. Geol
Soc Am Bull 112:1091–1105
Grinnell J, Storer TI (1924) Animal life in the yosemite: an account of the mammals, birds, rep-
tiles, and amphibians in a cross-section of the Sierra Nevada. University of California Press,
Berkeley
Gutiérrez-Pinto N, Cuervo AM, Miranda J, Pérez-Emán JL, Brumfield RT, Cadena CD (2012)
Non-monophyly and deep genetic differentiation across low-elevation barriers in a Neotropical
montane bird (Basileuterus tristriatus; Aves: Parulidae). Mol Phylogenet Evol 64:156–165
Haffer J (1974) Avian speciation in tropical South America with a systematic survey of the toucans
(Ramphastidae) and the jacamars (Galbulidae). Nuttall Ornithological Club, Cambridge
Hawkins BA, Diniz-Filho JAF (2002) The mid-domain effect cannot explain the diversity gradient
of Nearctic birds. Glob Ecol Biogeogr 11:419–426
Hawkins BA, Diniz-Filho JAF, Jaramillo CA, Soeller SA (2007) Climate, niche conservatism, and
the global bird diversity gradient. Am Nat 170:S16–S27
Hawkins BA, McCain CM, Davies TJ, Buckley LB, Anacker BL, Cornell HV, Damschen EI,
Grytnes J-A, Harrison S, Holt RD, Kraft NJB, Stephens PR (2012) Different evolutionary
histories underlie congruent species richness gradients of birds and mammals. J Biogeogr
39:825–841
Hazzi NA, Moreno JS, Ortiz-Movliav C, Palacio RD (2018) Biogeographic regions and events
of isolation and diversification of the endemic biota of the tropical Andes. Proc Natl Acad Sci
115:7985
Hoorn C, Wesselingh FP, ter Steege H, Bermudez MA, Mora A, Sevink J, Sanmartín I, Sanchez-
Meseguer A, Anderson CL, Figueiredo JP, Jaramillo C, Riff D, Negri FR, Hooghiemstra H,
Lundberg J, Stadler T, Särkinen T, Antonelli A (2010) Amazonia through time: andean uplift,
climate change, landscape evolution, and biodiversity. Science 330:927–931
Hutter CR, Guayasamin JM, Wiens JJ (2013) Explaining Andean megadiversity: the evolutionary
and ecological causes of glassfrog elevational richness patterns. Ecol Lett 16:1135–1144
Hutter CR, Lambert SM, Wiens JJ (2017) Rapid Diversification and time explain amphibian
richness at different scales in the tropical andes, earth’s most biodiverse hotspot. Am Nat
190(6):828–843
Jankowski JE, Ciecka AL, Meyer NY, Rabenold KN (2009) Beta diversity along environmental
gradients: implications of habitat specialization in tropical montane landscapes. J Anim Ecol
78:315–327
Janzen DH (1967) Why mountain passes are higher in the Tropics. Am Nat 101:233
Kattan GH, Franco P (2004) Bird diversity along elevational gradients in the Andes of Colombia:
area and mass effects. Glob Ecol Biogeogr 13:451–458
Kattan GH, Tello SA, Giraldo M, Cadena CD (2016) Neotropical bird evolution and 100 years of
the enduring ideas of Frank M. Chapman. Biol J Linn Soc 117:407–413
Kozak KH, Wiens JJ (2010) Accelerated rates of climatic-niche evolution underlie rapid species
diversification. Ecol Lett 13:1378–1389
López-Aguirre C, Hand SJ, Laffan SW, Archer M (2018) Phylogenetic diversity, types of ende-
mism and the evolutionary history of New World bats. Ecography 41:1–12
Lynch JD, Duellman WE (1997) Frogs of the genus Eleutherodactylus in western Ecuador: sys-
tematics, ecology, and biogeography. Nat History Museum 23:1–236
McCain CM (2009) Global analysis of bird elevational diversity. Glob Ecol Biogeogr 18:346–360
McGuire JA, Witt CC, Remsen JV, Corl A, Rabosky DL, Altshuler DL (2014) Molecular phyloge-
netics and the diversification of hummingbirds. Curr Biol 24:9
Mendoza ÁM, Ospina OE, Cárdenas-Henao H, García-R JC (2015) A likelihood inference of his-
torical biogeography in the world’s most diverse terrestrial vertebrate genus: diversification of
direct-developing frogs (Craugastoridae: Pristimantis) across the Neotropics. Mol Phylogenet
Evol 85:50–58
Meza-Joya FL, Torres M (2016) Spatial diversity patterns of Pristimantis frogs in the tropical
Andes. Ecol Evol 6(7):1901–1913
Morlon H (2014) Phylogenetic approaches for studying diversification. Ecol Lett 17:508–525
Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB, Kent J (2000) Biodiversity hotspots
for conservation priorities. Nature 403:853–858
Nogués-Bravo D, Araújo MB, Romdal T, Rahbek C (2008) Scale effects and human impact on the
elevational species richness gradients. Nature 453:216–219
Orme CDL, Davies RG, Burgess M, Eigenbrod F, Pickup N, Olson VA, Webster AJ, Ding T-S,
Rasmussen PC, Ridgely RS, Stattersfield AJ, Bennett PM, Blackburn TM, Gaston KJ, Owens
IPF (2005) Global hotspots of species richness are not congruent with endemism or threat.
Nature 436:1016–1019
Perrigo et al. (2020) Why mountains matter for biodiversity. J Biogeogr 47(2):315–325. https://
onlinelibrary.wiley.com/doi/10.1111/jbi.13731
Pintanel P, Tejedo M, Ron SR, Llorente GA, Merino-Viteri A (2019) Elevational and microcli-
matic drivers of thermal tolerance in Andean Pristimantis frogs. J Biogeogr 46:1664–1675
Quintero I, Jetz W (2018) Global elevational diversity and diversification of birds. Nature 555:246
Rahbek C (1997) The relationship among area, elevation, and regional species richness in neo-
tropical birds. Am Nat 149:875–902
Rahbek C, Graves GR (2001) Multiscale assessment of patterns of avian species richness. Proc
Natl Acad Sci U S A 98:4534–4539
Rahbek C, Gotelli NJ, Colwell RK, Entsminger GL, Rangel TFLVB, Graves GR (2007) Predicting
continental-scale patterns of bird species richness with spatially explicit models. Proc R Soc
Lond Ser B Biol Sci 274:165–174
Rahbek C, Borregaard MK, Antonelli A, Colwell RK, Holt BG, Nogues-Bravo D, Rasmussen
CMØ, Richardson K, Rosing MT, Whittaker RJ, Fjeldså J (2019a) Building mountain biodi-
versity: geological and evolutionary processes. Science 365:1114
Rahbek C, Borregaard MK, Colwell RK, Dalsgaard B, Holt BG, Morueta-Holme N, Nogues-
Bravo D, Whittaker RJ, Fjeldså J (2019b) Humboldt’s enigma: what causes global patterns of
mountain biodiversity? Science 365:1108
Ramírez-Barahona S, Eguiarte LE (2013) The role of glacial cycles in promoting genetic diver-
sity in the neotropics: the case of cloud forests during the last glacial maximum. Ecol Evol
3:725–738
Rangel TF, Edwards NR, Holden PB, Diniz-Filho JAF, Gosling WD, Coelho MTP, Cassemiro
FAS, Rahbek C, Colwell RK (2018) Modeling the ecology and evolution of biodiversity: bio-
geographical cradles, museums, and graves. Science 361:5452
Ribas CC, Moyle RG, Miyaki CY, Cracraft J (2007) The assembly of montane biotas: linking
Andean tectonics and climatic oscillations to independent regimes of diversification in Pionus
parrots. Proc R Soc B Biol Sci 274:2399–2408
Rosenzweig ML (1995) Species diversity in space and time. Cambridge University Press,
Cambridge
Rull V (2011) Neotropical biodiversity: timing and potential drivers. Trends Ecol Evol 26:508–513
Santos JC, Coloma LA, Summers K, Caldwell JP, Ree R, Cannatella DC (2009) Amazonian
amphibian diversity is primarily derived from Late Miocene Andean lineages. PLoS Biol
7:e1000056
Sedano RE, Burns KJ (2010) Are the Northern Andes a species pump for Neotropical birds?
Phylogenetics and biogeography of a clade of Neotropical tanagers (Aves: Thraupini). J
Biogeogr 37:325–343
Sklenář P, Hedberg I, Cleef AM (2014) Island biogeography of tropical alpine floras. J Biogeogr
41:287–297
Smith SA, Montes de Oca AN, Reeder TW, Wiens JJ (2007) A phylogenetic perspective on eleva-
tional species richness patterns in Middle American treefrogs: why so few species in lowland
tropical rainforests? Evolution 61:1188–1207
Smith BT, McCormack JE, Cuervo AM, Hickerson MJ, Aleixo A, Cadena CD, Perez-Eman J,
Burney CW, Xie X, Harvey MG, Faircloth BC, Glenn TC, Derryberry EP, Prejean J, Fields S,
Brumfield RT (2014) The drivers of tropical speciation. Nature 515:406–409
Stein A, Gerstner K, Kreft H (2014) Environmental heterogeneity as a universal driver of species
richness across taxa, biomes and spatial scales. Ecol Lett 17:866–880
Stephens PR, Wiens JJ (2003) Explaining species richness from continents to communities: the
time-for-speciation effect in emydid turtles. Am Nat 161:112–128
Turchetto-Zolet AC, Pinheiro F, Salgueiro F, Palma-Silva C (2013) Phylogeographical patterns

shed light on evolutionary process in South America. Mol Ecol 22:1193–1213
Wallace A (1876) The geographical distribution of animals, with a study of the relations of liv-
ing and extinct faunas as elucidating the past changes of the earth’s surface. Macmillan and
Co., London
Weir JT (2006) Divergent timing and patterns of species accumulation in lowland and highland
Neotropical birds. Evolution 60:842–855
Weir JT, Price M (2011) Andean uplift promotes lowland speciation through vicariance and disper-
sal in Dendrocincla woodcreepers. Mol Ecol 20:4550–4563
Chapter 10
Mapping Hydrological Ecosystem Services
and Impacts of Scenarios for Deforestation
and Conservation of Lowland, Montane
and Cloud-Affected Forests
Mark Mulligan
10.1 Introduction
Ecosystem services are the benefits that human populations derive from nature, but
which are not usually accounted for in the economic system that we operate within
(see Fisher et al. 2009). For many of these services the local proximity of beneficia-
ries is important to the benefits received, for example, hazard mitigation benefits
(such as erosion control and flood mitigation) are received downhill of the systems
that provide them; hydrological and sediment retention benefits are received nearby
downstream and aesthetic and nature-based tourism tend to benefit local popula-
tions (through economic opportunities) and tourist-facing local businesses.
Cloud forests, though small in global surface area, comprising between
2.21M km2 (Mulligan 2010a) and 215,000 km2 (Bubb et al. 2004) depending on the
definition used, are found in some important headwater areas with very significant
populations locally and downstream. This context is thought to be very different
from that of the much more extensive lowland forests, which are considered to have
relatively few local and downstream populations. Cloud forests may thus provide
important ecosystem services that are disproportionate to their relatively small
global extent. This paper uses a range of spatial data to examine the key ecosystem
services provided by the global cloud forest estate relative to other forest types, the
benefits provided, who benefits and what is necessary to protect these benefits in the
future, continent by continent. We also examine some important considerations
which determine how accessible these services may be downstream.
Living forests are considered to provide a range of ecosystem services as well as
harbouring much of the world’s diversity and rarity of species. The key services
provided include:
M. Mulligan (*)
Department of Geography, King’s College London, London, UK

https://doi.org/10.1007/978-3-030-57344-7_10
190 M. Mulligan
(a) carbon storage and sequestration which mitigate against climate change (Naidoo
et al. 2008),
(b) the provision and regulation of supplies of clean water through nutrient, con-
taminant and sediment retention (Brauman et al. 2007),
(c) the mitigation of environmental hazards such as landsliding, sediment trans-
port, flooding and low flows (De Groot et al. 2010), as well as
(d) the provision of livelihood opportunities through ecotourism and the availabil-
ity of non-timber forest products such as firewood, fruits, nuts, fungi, building
materials and medicinal products (Jenkins et al. 2004).
Some of these services are realised by beneficiaries throughout the world, for
example, through climate change mitigation provided by carbon storage and seques-
tration. Others are more local in their beneficiaries, such as the downslope benefi-
ciaries of hazard mitigation and downstream beneficiaries of water services. Still
others have local first benefits (for example, improved water production sustaining
agriculture) but then teleconnections through commodity supply chains to second
beneficiaries of this productivity throughout the world (see Mulligan 2015a).
Here we use a range of global datasets to map the ecosystem service provision of
all tropical (ALLF), tropical montane (TMF), cloud-affected tree cover (CAFs) and
cloud-affected forests (CFs). The latter two may be broadly considered cloud for-
ests similar to the definitions adopted by Mulligan (2010a) and Bubb et al. (2004),
respectively. Our objective is to examine the significance of cloud-affected forest
types in the provision of key ecosystem services, how well they are protected and
the tropical forest conservation foci that would be necessary to protect the ecosys-
tem services provided whilst also sustaining the maximum tropical forest biodiver-
sity. We examine both cloud-affected tree cover and cloud-affected forests since the
boundary of the cloud forest condition is unclear and it is important to understand
the extent to which cloud affected systems of different tree cover fractions (and thus
forest/non-forest definitions) impact on ecosystem service delivery.
10.2 Materials and Methods
Here we map the provision of key ecosystem services by all tropical (ALLF), tropi-
cal montane forests (TMF), tropical montane cloud-affected forests (CAFs) and (the
more extensive/intact) tropical montane cloud forests (CFs). ALLF are defined as
all areas in the continents of South America, Central America, Africa and tropical
and subtropical Asia (to 38°N) with >40% tree cover according to (Townshend et al.
2011) and thus include (seasonally) dry forests as well as tropical humid forests.
Montane forests (TMF) are forests on these same continents with elevation >500
masl (according to Farr and Kobrick 2000). This is a simple definition of montane
compared with many (cf. Funnell and Price 2003) but is considered relevant here
because of the strong climatic influence on cloud forests (Jarvis and Mulligan 2009).
CAFs are cloud-affected forests according to Mulligan (2010a) remote-sensing
10 Mapping Hydrological Ecosystem Services and Impacts of Scenarios… 191
based hydroclimatic analysis, that is, all areas with forest cover >10% and annual
mean fog frequency >70%, so including very fragmented forest cover (but still for-
ests according to the FAO definition of tree cover >10% FAO (1998)). Cloud forests
(CFs) are also defined according to Mulligan (2010a) but as areas with forest cover
>40% and annual mean fog frequency >70% so only more extensive and intact for-
est is included commensurate with other analyses of cloud forest such as Bubb
et al. (2004).
The analysis of ecosystem services conducted here is performed using the public
domain Co$ting Nature platform (Mulligan 2015b; Mulligan et al. 2010; Mulligan
and Clifford 2015; http://www.policysupport.org/costingnature) and associated
global datasets as described below. The analysis is carried out continent by conti-
nent for South America, Central America, Africa and tropical and subtropical Asia,
according to the hydrological continents defined by Lehner et al. (2008).
10.2.1 Mapping Potential Beneficiaries
Potential beneficiaries of forest ecosystem services include local and downstream

populations that may benefit from provisioning and local regulating services as well
as downstream water infrastructure in the form of dams that could benefit from
water quantity, quality, flow regulation as well as sediment retention services.
Beneficiaries are calculated based on global datasets for the distribution of popula-
tion (LandScan 2007), dams (Mulligan et al. 2011), accumulated downstream at
1 km resolution using the flow directions provided by Lehner et al. (2008). For each
forest type we calculate the mean downstream population for all cells in that forest
type. This reflects the mean number of people potentially affected hydrologically by
the forested pixels, though we will see later that distance downstream is a critical
factor in determining who actually benefits significantly. As a proxy for infrastruc-
ture that potentially benefits from these forests we count the average number of
downstream dams for each forested area using the 36,000 dams mapped by Mulligan
et al. (2011).
10.2.2 Mapping Ecosystem Service Provision
Co$ting Nature uses a range of global datasets and simple models to map the sites
of production for some key ecosystem services including carbon (storage and
sequestration), water (quantity and quality), hazard mitigation (for landslide,
drought, flood, coastal inundation) and nature-based tourism. It also uses a variety
of species range data to calculate richness for the sampled red list species of mam-
mals, amphibians, reptiles and birds. In each case the realised and potential services
are mapped. A potential service is defined here as a service produced but not (cur-
rently) consumed, whereas a realised service is produced and available to current
192 M. Mulligan
beneficiaries (Mulligan and Clifford 2015). For example, all potential carbon ser-
vices are also realised because all carbon storage and sequestration contribute to
climate change mitigation, with beneficiaries worldwide. However, where forests
provide potential hydrological services, these are only realised as services if there
are human beneficiaries (people, dams, irrigation projects) nearby downstream of
these forests who can benefit from the services provided. The further downstream
the potential beneficiaries from a site of ecosystem service production, the less their
services will be influenced by that site.
10.2.3 Mapping Realised Carbon Services
The Co$ting Nature carbon storage and sequestration service is calculated as the
combination of carbon stocks and carbon sequestration. Carbon sequestration is
calculated here from the dry matter productivity (DMP) analysis of Mulligan
(2009a) in which SPOT-VGT DMP calculated every 10 days at 1 km resolution on
the basis of change in NDVI was averaged over the period 1998–2008, globally.
DMP (t biomass/ha/yr) is multiplied by 0.42 (Ho 1976) to convert to units of tC/ha/
yr. Above-ground carbon stock is calculated from Saatchi et al. (2011) for the areas
in which data are available and Ruesch and Gibbs (2008) elsewhere. This is com-
bined with soil carbon calculated from the map of Scharlemann et al. (2009) to
produce a total above- and below-ground carbon stocks.
10.2.4 Mapping Realised Water Services
Water is considered a provisioning service in Co$ting Nature, though it also plays a

role in the regulating services (see the section on hazard mitigation). Potential water
services are measured as the volume of runoff (rainfall minus evapotranspiration)
whose quality is unaffected by human activity, cumulated downstream. This is an
indicator of the volume of clean water produced by a pixel. The human footprint on
water quality index (HF) (Mulligan 2009b) is used as the indicator of water quality.
The HF considers particular land uses to have the potential to contaminate water
with sediment, agrochemicals, manures, etc. Land uses such as unprotected agricul-
ture and pasture are considered non-point sources and roads, mines, oil and gas
wells and urban areas are considered point sources. Agriculture in protected areas is
considered to have a human footprint of zero as are areas with no human land use.
The HF index multiplies the water balance of a pixel by the fractional cover of point
and non-point sources in that pixel and cumulates this “polluted” water downstream
using a streamflow network. The total volume of water flowing is also cumulated
downstream and the HF in a pixel becomes the volume of polluted water as a per-
centage of the total water flowing. The nonpolluted volume of water expressed as a
fraction is considered the potential water service of a pixel. Realised water services
are calculated considering the use of these potential water services which, in turn,
depends on the distribution of beneficiaries for hydrological services.
Maps of globally normalised downstream population, irrigation area and number
of dams are pre-calculated to represent the distribution of beneficiaries. The popula-
tion is derived from LandScan (2007) and is summed downstream at 1 km resolu-
tion using a drainage network derived from HydroSHEDS (Lehner et al. 2008) with
the downstream total for a pixel assigned to that pixel. For number of dams we use
the global dam census of Mulligan et al. (2011) and for irrigated land we cumulate
the downstream irrigated areas of Siebert et al. (2007) for each point. These provide
pixel-level indicators of the distribution of the beneficiaries of hydrological ecosys-
tem services at a much finer resolution than previous global studies. Potential water
services are multiplied by the normalised sum of all downstream beneficiaries to
calculate the realised water services index.
10.2.5 Mapping Realised Hazard Mitigation Services
Hazard mitigation services are perhaps the most complex to assess since they are a
function of:
(a) the potential for multiple hazards to occur and the human and infrastructural
exposure to those hazards and vulnerability to the negative impacts of hazards.
Exposure and vulnerability together define the risk.
(b) the role of local, upstream or near-coast ecosystems in reducing the potential
impact of hazards (i.e. potential hazard mitigation services).
Hazard mitigation ecosystem services are then realised in those areas in which
ecosystems provide hazard mitigation services but where there is also a risk. Areas
with no risk may receive potential hazard mitigation services but these services are
not realised. Co$ting Nature considers hazard potential as:
(a) the normalised frequency of cyclones according to Dilley (2005) multiplied by
the normalised water balance as an index of high magnitude rainfall event
hazards;
(b) for coastal inundation hazards we calculate distance from coast according to
USGS (2006) and consider all pixels within 2 km as coastal. We also produce
an index of low-lying land as all areas from 0 to 30 m according to the SRTM
digital elevation model post-processed by Lehner et al. (2008). The probability
of (coastal) inundation hazard is considered proportional to the normalised
probability of Tsunami (according to NGDC 2011), cyclones and climatic sea
level rise (considered for simplicity as equally likely everywhere) for all coastal
areas. The probability of inundation hazard is the combination of these effects.
(c) for landslide hazards we consider the probability of landslides to increase with
the normalised mean upstream slope gradient. Upstream slope gradient is
194 M. Mulligan
p re-calculated using the 1 km resolution digital elevation model and flow net-
work of Lehner et al. (2008).
(d) the potential for flooding is considered proportional to normalised water bal-
ance with small potential in dry areas and high potential where water is plenti-
ful. Though many floods are fluvial in nature, we use water balance rather than
runoff in recognition that floods also occur from overwhelmed urban drainage,
groundwater flooding and rainfall intensity greater than infiltration rates. These
latter types of flood can be somewhat more predictable than fluvial floods.
Hazard potential is thus the mean of cyclone, inundation, landslide and flood
probabilities, normalised either locally or globally as defined by the user. In addi-
tion to hazard potential, we consider hazard exposure as the exposure of human
populations, activity and infrastructure. Socio-economic exposure is calculated as
normalised GDP for 1990 (CIESIN 2002), population (LandScan 2007), agriculture
(cropland and pasture fractional areas from Ramankutty et al. 2008) and infrastruc-
ture. Infrastructural exposure is calculated as the sum of the presence of dams
(Mulligan et al. 2011), mines (Mulligan 2010b), oil and gas (Mulligan 2010c),
urban areas (Schneider et al. 2009) and roads (FAO 2010). Exposure is multiplied
by hazard potential to produce the index of exposure to hazards. Vulnerability to
hazards is considered to scale with normalised GDP and infrastructure: the greater
the GDP and infrastructure, the greater the capacity to cope with hazards. Risk is
then exposure multiplied by the vulnerability.
Potential hazard mitigation services are then calculated according to a series of
assumptions, based on knowledge of how ecosystems mitigate these hazards. We
assume that landslide impacts at a point are mitigated according to the proportion of
upstream area that is tree covered (using tree cover data from Hansen et al. 2006) or
is protected (according to WDPA 2014). This is because tree cover reduces potential
soil waterlogging and has been shown to reduce landslide frequency (Dapples et al.
2002) and protected areas will tend to have a lower agricultural and infrastructural
impact—both of which can lead to increased frequency of landslides. Regulation of
drought hazards (for example, reduced dry season flows) is assumed proportional to
tree cover upstream. Although trees evaporate significant volumes of water and
thereby reduce flows, they also encourage infiltration that helps to maintain dry
season flows (Pena-Arancibia et al. 2012). Flood hazards are mitigated according to
the proportion of upstream area that provides flood storage in the form of trees,
wetlands (Lehner and Doll 2004), water bodies (USGS 2006) and floodplains
(Mulligan 2010b). Mitigation from coastal inundation is considered to be provided
by wetlands and mangroves (Spalding et al. 1997) but only in low-lying and coastal
areas. Where these ecosystems occur inland they are assumed to have no coastal
inundation mitigation potential. The total potential hazard mitigation services are
then the mean of coastal flood regulation and landslide mitigation services.
Realised hazard mitigation services are calculated as the minimum of risk and
potential hazard mitigation services for areas where risk is greater than 0. In other
words if hazard mitigation potential is greater than risk, then hazard mitigation
potential equals risk and the remaining hazard mitigation potential is unused. If risk
is greater than hazard mitigation potential, then realised hazard mitigation is equal
to potential hazard mitigation and some risk remains unmitigated.
10.2.6 Mapping Species Richness
We use the IUCN sampled red list extent of occupancy (EOO) data for amphibians
(IUCN et al. 2008), mammals (IUCN et al. 2008), reptiles (IUCN 2010) and birds
(Birdlife 2012) and calculate a measure of richness (total number of sampled spe-
cies within each 10 km cell) combining all sampled species. We would have liked to
incorporate similar data for plants but these data are not in the public domain. We
calculate mean richness per forest type and continent as a simple arithmetic mean
for all relevant pixels.
Having mapped these services we then subset them for continental zones of
interest (ZOIs) representative of the four forest types: LRF, TMF, CAF and CF as
defined above and present statistics for the total (sum) and mean (density) of each
service for each forest cover type by continent. We then examine the distribution of
protected areas within these forest types, by continent, using the WDPA August
2014 update (WDPA 2014) and finally the proportion of the realised ecosystem
services provided by each forest type that are derived from land under pro-
tected status.
10.2.7 Distance Decay of Hydrological Services
Finally we examine the distance decay of hydrological ecosystem service from

cloud forests using a new technique termed the hydrological footprint. Based on the
human footprint on water quality methodology (Mulligan 2009b), the hydrological
footprint (HF) is a measure of the influence of an area on flows downstream of that
area. The HF is calculated using Co$ting Nature as the percentage of water in each
pixel derived from a particular area of interest upstream (in this case ALLF, MF,
CAF, CF or protected CAF). The calculation is performed by accumulating down-
stream the water balance derived from the area of influence (WBi) and accumulat-
ing the water balance of all areas (WBa). The HF in a given pixel is then (WBi/
WBa)*100. The accumulation is done using a D8 accumulation function (see Jones
2002) and the water balance calculated monthly and annually from an interpolated
rainfall dataset (Hijmans et al. 2005) minus a remotely sensed actual evapotranspi-
ration climatology based on Mu et al. (2011). The hydrological footprint is thus a
water-balance weighted measure of the distance decay of hydrological influence of
an upstream area. The downstream influence thus depends on the size, distance
upstream and water balance of the areas of interest in relation to surrounding areas
between the upstream area of interest and the downstream site of interest.
196 M. Mulligan
The hydrological footprint is a simple indicator but does not take into account the
fact that cloud forests provide additional inputs of water in the form of fog intercep-
tion that are not usually available outside of the CAF zone. CAFs will also tend to
have lower evapotranspiration than other forest types because of the frequent cloud
cover (Bruijnzeel et al. 2011). The hydrological footprint may thus underestimate
the influence of CAF. We test this for Costa Rica and Madagascar by running the
WaterWorld model (Mulligan 2013) to calculate a water balance including fog con-
tributions and then calculate the hydrological footprint using these data instead of
the remote sensing data. This is done only for Costa Rica since computing limita-
tions preclude it at the continental scale. We also utilise the WaterWorld “fog con-
tribution to runoff” variable in this analysis as another measure of the decay of
downstream influence, this time for the additional CAF input of fog only. Finally we
calculate the rate of change in downstream decay with distance in order to better
understand the distances within which downstream potential beneficiaries will be
significantly influenced by the upstream CAF.
10.3 Results
10.3.1 T
he Distribution and Coverage
of (Cloud-Affected) Forests
Global maps of the distribution of these four forest types are shown in Figs. 10.1,
10.2, 10.3 and 10.4 and statistics of their coverage by hydrological continent are
given in Table 10.1. Clearly cloud forests (CFs) are generally the most restricted
forest type, with lowland forests being the most extensive and montane and CAFs
being in between. All continents have the presence of CAF and CF but they are
particularly restricted in Africa and particularly extensive in Asia (Figs. 10.1, 10.2,
10.3 and 10.4).
Table 10.1 shows that forests constitute some 26 million square km in these con-
tinents, representing 26.9% of land. Of these forests, some 33% can be considered
montane using the definition adopted here, 31% are considered cloud affected and
18% cloud forests. Proportions vary between continents with South America having
the greatest percentage of territory under ALLF but tropical and subtropical Asia
having the greatest percentage of territory under montane and cloud forest types at
10–15% compared with less than 10% for South and Central America. Cloud
affected and cloud forests are particularly rare in Africa at 4.1% and 1.7% of land,
respectively, though the corresponding land areas covered are significant given the
size of the continent.
ALLF (ALL F) Montane forests (MF)
Cloud-affected forest (CAF) Cloud forest (CF)
Fig. 10.1 The distribution of forest types used for ecosystem service comparison: South America
10.3.2 P
otential Beneficiaries of (Cloud-Affected) Forest
Ecosystem Services
Table 10.2 describes some key indicators of the density of potential beneficiaries by
forest type and continent. These are presented as densities to indicate the mean
number of people locally and downstream that are supported by the different forest
types. For the benefitting population within the forested area, numbers are also mul-
tiplied by the area of each forest type from Table 10.1 to arrive at total number of
beneficiaries. These are considered potential beneficiaries because, although eco-
system services may be provided, the extent to which they are realised will also
depend upon economic, social and cultural influences on access and upon distance
decay functions downstream.
198 M. Mulligan
Fig. 10.2 The distribution of forest types used for ecosystem service comparison: Central America
10.3.3 Local Populations
For all continents population densities in forested areas tend to be lower than the
continental average, though in African and Central American CAF areas, popula-
tions are higher than the continental average (perhaps because of the extensive low-
population deserts in these regions). ALLF have a human population density
between 2 and 32 persons per sq. km combining some 247M of the 6756M in the
study region (3.6%). Population in montane forests are higher in Central America
and South America but lower in Africa and Asia, combining 84.9M (1.3% of the
total). Population densities in CAFs are much higher than ALLF or montane forests
and sum to 246.4M (3.6% of the total) in only a third of the forest area occupied by
ALLF. Indeed for Central America and Africa human population densities in CAFs
are greater than the continental average. Population densities in CFs are more simi-
lar to those across ALLF but still represent 60.5M (0.8% of the study area popula-
tion). Thus the local populations supported by these various forest types, especially
CAFs and CFs are a very small proportion of the pantropical total populations.
Fig. 10.3 The distribution of forest types used for ecosystem service comparison: Africa
Fig. 10.4 The distribution of forest types used for ecosystem service comparison: tropical and
subtropical Asia
200 M. Mulligan
Table 10.1 Coverage of forest types by continent

ALLF MF CAF CF
Mulligan (2010a) Mulligan (2010a)
Tree cover >40% Tree cover >10% Tree cover >40%
Tree cover and elevation and fog frequency and fog frequency
Definition >40% >500 >70% >70%
South 7,037,844 km2 987,069 km 2
909,329 km2 (6.5%) 543,396 km2 (3.9%)
America (SA) (50.2%) (7.4%)
Central 590,472 231208 (7.5%) 171,607 (5.4%) 92,971 (2.9%)
America (CA) (19.3%)
Africa (AF) 2,771,542 km2 1,581,671 km2 585,002 km2 (4.1%) 253,017 km2 (1.7%)
(19.8%) (11.1%)
Asia 6,232,291 km2 2,731,052 km2 3,327,390 km2 1,843,343 km2
(62.5%) (28.6%) (34.8%) (17.2%)
Asia (tropical) 3,655,757 km2 1,259,624 km2 1,169,172 km2 874,366 km2
(43.65%) (15.04%) (13.96%) (10.44%)
Asia (tropics 5,762,456 2,630,744 3,434,530 (20.47%) 2,088,445 (12.44%)
and subtropics (34.35%) (15.68%)
to 38°)
(ASTS)
Total (SA, 16,162,314 9,421,368 9,597,030 (31.6% of 5,695,538 (18.4% of
CA, AF, (26.9% of land) (33.6% of ALLF) ALLF) ALLF)
ASTS)
10.3.4 Downstream Populations
Numbers of downstream populations are an order of magnitude higher for forests in

Africa and Asia compared with Central and South America. In all cases the mean
number of persons downstream of forests increases from ALLF, through montane
forests and peaks for CAFs. In Central America, Africa and Asia these have higher
downstream populations than the all-land average. Forests have, on average, tens of
thousands of downstream persons in Central and South America and hundreds of
thousands in Africa and Asia.
10.3.5 Downstream Dams
Dams can be considered points in the landscape at which hydrological ecosystem

services are monetized through the provision of irrigation, domestic and industrial
water or hydropower. Number of downstream dams is highest in Asia and then Africa,
followed by Central America and South America. Once again the mean number of
dams downstream of forested pixels increases as we go from ALLF to MF with the
peak number of dams downstream of CAFs. Each pixel of CAF serves an average of
<0.4 dams in Central and South America, 0.4 dams in Africa and 1.0 dam in Asia.
Whilst some pixels may have no downstream dams and other pixels may have many,
the average is indicative of the overall influence of these forests on dams (Table 10.2).
Table 10.2 Beneficiaries in and downstream of forest types. Values greater than the all-land value
for the given variable and continent are emboldened
Continent Variable ALLF MF CAF CF ALL
Central Mean population 19.9 21.7 63.2 33.4 48.0
America within forested (11.75M) (5.01M) (10.85M) (3.1M) (146M)
area (person/km2) PROT: 30.6
(5.25M)
Mean downstream 13,890 26,473 32,866 26,582 36,079
population of PROT:
forested area 36,381
(person/km2)
Mean downstream 0.116 0.264 0.325 0.230 0.379
# dams of forested PROT:
area 0.427
South Mean population 2.4 5.7 17.5 5.5 21.8
America within forested (16.89M) (5.62M) (15.91M) (2.98M) (305M)
(4.4M)
Mean downstream 18,518 34,578 33,311 26,244 43,803
population of PROT:
(person/km2)
Mean downstream 0.022 0.099 0.060 0.059 0.108
area 0.126
Africa Mean population 11.6 (32.1 10.0 86.0 24.0 51.2
within forested M) (15.8 M) (50.3M) (6.1M) (716.7M)
(21 M)
Mean downstream 144,566 191,456 882,667 404,480 861,804
population of PROT:
(person/km2)
Mean downstream 0.053 0.088 0.418 0.165 0.284
area 0.446
Tropical and Mean population 32.4 22.2 49.3 23.1 192.0
subtropical within forested (186.7M) (58.4M) (169.3M) (48.2M) (5587M)
Asia area (person/km2) PROT: 11.3
(38.8M)
Mean downstream 323,289 539,555 735,186 609,744 564,889
population of PROT:
(person/km2)
Mean downstream 0.343 0.603 1.00 PROT: 0.69 0.668
# dams of forested 0.86
area
Combined Mean population 247M (3.6% 84.9M 246.4M 60.5M 6756M
within forested of study area (1.3%) (3.6%) (0.8%)
area (person/km2) total) PROT:
(69.5M)
(1.02%)
202 M. Mulligan
10.3.6 Beneficiaries of Protected Cloud-Affected Forests
For CAFs we have also calculated the mean for each variable that occurs within
protected CAFs. Population densities are lower in protected CAFs as might be
expected (and which may reflect the limitations of the population data, in which
protected areas are used as one of the variables in the disaggregation of administra-
tive region level population data). Some 69.5M (1.02%) of the study area population
are within protected CAFs. Downstream populations for protected CAFs vary from
tens of thousands in Central and South American protected CAFs to hundreds of
thousands for Asian and African protected CAFs. These are a small fraction of the
millions to billions of people in these continents, all others are vulnerable to chang-
ing hydrological ecosystem services resulting from poor land management.
Downstream dam densities for protected CAFs tend to be higher than for all CAFs,
except for Asia, suggesting strategic protection of dam watersheds.
10.3.7 M
apping Ecosystem Service Provision by Different
Forest Types
For each continent we now examine a range of ecosystem services (both potential
and realised, sensu Mulligan 2015b) and provide mean provision and total provision
within each forest cover type, alongside information on tree cover, biodiversity,
protected areas and deforestation for context. This is used to examine the natural
capital and ecosystem service provision status of different forest types in order to
place cloud forests within context.
10.3.8 Central America
For Central America (Table 10.3), the mean per-unit-area clean water provision tends
to be highest for CAF and CFs, since these are extensive on the continent. For realised
water provision all types of forest have greater provision per-unit-area than the conti-
nent-wide average (ALL-L) and that is also true for all carbon services. Hazard mitiga-
tion services are greatest for MFs, CAFs and especially CFs. Species richness is also
greatest for MF, CAF and especially CF. Protected area coverage is significant for all
forest types, though the most extensive are for ALLF (72% of all protected areas).
Deforestation rates are greater for ALLF than for montane forests and the lowest rates
are in CFs. The rate in protected CAFs is only a little less than the all CAF rate. The
5.4% of the study area that are CAFs provide 13.8% of realised water, 8.9% of above-
ground carbon, 7.7% of soil carbon, 12% of hazard mitigation and 7.9% of the total
tree cover. Their absolute contribution is low relative to the more extensive ALLF class
but for most services their provision per-unit-area is greater. Some 0.04 of 0.17 Mkm2
of CAF (23.3%) are protected but this only leads to a reduction in deforestation rates
of around 25% relative to unprotected CAFs. Protected CAFs secure less than half of
Table 10.3 Ecosystem service provision by different forest types, Central America
Ecosystem services sum
(mean)
(percent of all-land total) ALLF MF CAF CF ALL-L
Water (potential Mm3 clean water 0.69 0.26 0.19 0.13 1.45
produced) (0.92) (0.89) (0.93) (1.13) (0.37)
[48.03%] [17.94%] [13.49%] [8.82%]
PROT:
[22.7%]
Water (realised Mm3 clean water 0.68 0.26 0.19 0.13 1.42
produced) (0.91) (0.89) (0.93) (1.14) (0.36)
[48.1%] [18.31%] [13.77%] [9.00%]
PROT:
[22.5%]
AG carbon (M tonnes C storage) 7011 2730 1580 1132 17600
(tonnes C/km2) (9308) (9381) (7519) (10052) (4515)
[39.8%] [15.5%] [8.98%] [6.43%]
PROT:
[20.7%]
Soil carbon (M tonnes C storage) 8250 3156 2748 1505 35548
(tonnes C/km2) (10952) (10845) (13074) (13359) (9119)
[23.2%] [8.9%] [7.7%] [4.2%]
PROT:
[14.4%]
Carbon sequestration (M Dg/ha/yr) 12.14 4.9 3.2 1.9 35.4
(Dg/ha/yr) (16.12) (16.7) (15.35) (16.6) (9.0)
[34.2%] [13.7%] [9.1%] [5.3%]
PROT:
[15.2%]
Hazard mitigation index (0-1) 0.003 0.001 0.001 0.0006 0.009
(0.004) (0.005) (0.005) (0.006) (0.002)
[34.1%] [15.9%] [12.06%] [7.09%]
PROT:
[14.1%]
Species richness (species) (418) (425) (422) (436) (357)
[22.7%] [8.9%] [6.4%] [3.5%]
PROT:
[13.68%]
Tree cover (M km2) (%) 584 81.5 53 44 673
(70) (66) (49) (67) (40)
[86.8%] [12.1%] [7.9%] [6.6%]
PROT:
[20.6%]
Protected area (M km2) (%) 3.6 0.45 0.04 0.25 4.97
[72.4%] [9.1%] [6.5%] [5%]
Deforestation (2000–2012) % 21.9 3.56 1.94 1.25 65.5
(2.59) (2.86) (1.78) (1.89) (3.9)
[33.4%] [5.4%] [2.97%] [1.91%]
PROT: (1.44)
[20.8%]
204 M. Mulligan
CAF ecosystem services (Table 10.3), making the remaining services vulnerable in the
high deforestation context of this region (Table 10.4).
10.3.9 South America
For South America (Table 10.4) the mean per-unit-area clean water provision tends
to be highest for ALLF (because of the lack of polluting human influence on the
lowland forests). For realised water provision all forest types have greater provision
per-unit-area than ALL-L and that is also true for the carbon services (except
sequestration in CAFs and CFs which is similar to, or lower than, the mean for
ALL-L). Hazard mitigation services are greatest for CAFs and CFs. Species rich-
ness, on the other hand, is greatest for ALLF since it is dominated by high richness
in extensive lowland forests on this continent. Protected area coverage is significant
for all forest types. ALLF represent 72% of protected areas, whereas CAFs only
6.48%. Deforestation rates are greater in montane forests than lowlands but lowest
in CAFs. The rate in protected CAFs is half that of the all CAF rate. The 6.5% of the
study areas that are CAFs provide 6.94% of realised water, 6.94% of above-ground
carbon, more than 8% of soil carbon, 11.45% of hazard mitigation and 7.92% of the
total tree cover. Their absolute contribution is low relative to the much more exten-
sive ALLF class and for most services their provision per-unit-area is less than this
class, though CAFs and CFs provide higher per-unit-area hazard mitigation services
than ALLF in South America. 0.32 of 0.91 Mkm2 of CAF (35.2%) are protected
leading to a halving of deforestation rates relative to unprotected CAFs and securing
significant proportions of CAF biodiversity, tree cover and water/carbon ecosystem
services in particular (Table 10.4).
10.3.10 Africa
For Africa (Table 10.5) the mean per-unit-area clean water provision tends to be
highest for ALLF and CF (because of the relative lack of human influence on these
forest types). For realised water provision all forest types have greater provision
per-unit-area than ALL-L and that is also true for all carbon services (except soil
carbon in ALLF and MFs, which is similar to or lower than the mean for ALL-L).
Provided hazard mitigation services are greatest for CAFs. Species richness is
greatest for CAFs, CFs and MFs, reflecting the geographical variety of these sys-
tems on the continent. ALLF represents only 11% of Africa’s protected areas and
CAFs only 2.2%, with many protected areas in Africa covering non-forest environ-
ments. Deforestation rates are greatest in ALLF and the rate in protected CAFs is a
third that of the all CAF rate. The 4.1% of the study areas that are CAFs provide
6.4% of realised water, 6.2% of above-ground carbon, but only 5% of soil carbon,
4.3% of hazard mitigation but 7.6% of the total tree cover. Their absolute contribu-
tion is low relative to the much more extensive ALLF class and for most services
their provision per-unit-area is less, though CAFs provide higher per-unit-area
Table 10.4 Ecosystem service provision by different forest types, South America
Ecosystem services
sum
(mean)
[percent of all-land ALLF sum MF sum CAF sum CF sum ALL-L sum
total] (mean) (mean) (mean) (mean) (mean)
Water (potential Mm3 10,956,727 1,204,232 1,160,024 802,928 16,801,766
clean water produced)a (1.301) (0.968) (1.064) (1.226) (1.677)
[64.23%] [7.15%] [6.89%] [4.78%]
PROT:
[35.6%]
Water (realised Mm3 10.778 1.181 1.149 0.794 16.477
clean water produced) (1.280) (0.950) (1.054) (1.214) (0.982)
PROT:
[35.0%]
AG carbon (M tonnes C 114,352 14,281 10,966 8,469 136,048
storage) (tonnes C/km2)b (13,586) (11,482) (10,059) (12,937) (8,109)
[64.32%] [7.13%] [6.94%] [4.80%]
PROT:
[43.6]
Soil carbon (M tonnes C 91,434 15,812 16,564 9,342 199,518
storage) (tonnes C/km2) (10,863) (12,713) (15,194) (14,272) (11,892)
[84.05%] [10.50%] [8.06%] [6.23%]
PROT:
[23.9%]
Carbon sequestration 79,818 11,982 8,896 5,923 152,586
(M Dg/ha/yr) (Dg/ha/yr) (9,483) (9,634) (8,160) (9,048) (9,094)
[46.56%] [7.03%] [5.22%] [3.49%]
PROT:
[24.7%]
Hazard mitigation index 242,496 88,420 118,294 56,384 1,040,962
(0-1) (0.028) (0.071) (0.108) (0.086) (0.062)
[21.58%] [8.00%] [11.45%] [5.19%]
PROT:
[11.3%]
Species richness (653) (552) (510) (559) (626)
(species) [52.3%] [6.5%] [5.3%] [3.5%]
PROT:
[24.7%]
Tree cover (M km2) (%) 58.5 8.1 5.3 4.4 (40.15)
(69.5) (65.5) (48.97) (66.96)
[86.83%] [12.09%] [7.92%] [6.56%]
PROT:
[41.8%]
Protected area (M km2) 3.596 0.45 0.32 0.25 4.96
(%) [72.4%] [9.07%] [6.48%] [5.03%]
Deforestation (2.59) (2.85) (1.78) (1.89) (3.90)
(2000–2012)% PROT:
(0.94)
For areas with a positive water balance
a
For areas with a positive water balance

b
206 M. Mulligan
Table 10.5 Ecosystem service provision by cloud forests, Africa

ALLF sum MF sum CAF sum CF sum ALL-L sum
Ecosystem services (mean) (mean) (mean) (mean) (mean)
Water (potential Mm3 clean 2.8 1.38 0.5 0.3 8.1
water produced) (0.84) (0.74) (0.74) (0.9) (0.48)
[34.8%] [17.1%] [6.4%] [3.3%]
PROT:
[12.8%]
Water (realised Mm3 clean 2.79 1.38 0.51 0.26 8.03
water produced) (0.84) (0.75) (0.74) (0.88) (0.48)
[34.8%] [17.2%] [6.37%] [3.28%]
PROT:
[12.7%]
AG carbon (M tonnes C 41003 20854 5680 3896 91790
storage) (tonnes C/km2) (12344) (11234) (8260) (13128) (5471)
[44.6%] [22.7%] [6.18%] [4.24%]
PROT:
[15.8%]
Soil carbon (M tonnes C 27857 15687 8422 3913 159303
[17.48%] [9.85%] [5.29%] [2.46%]
PROT:
[16.2%]
Carbon sequestration (M 51.81 30.3 10.68 4.87 218
Dg/ha/yr) (Dg/ha/yr) (15.6) (16.3) (15.5) (16.4) (13.01)
[23.7%] [13.87%] [4.89%] [2.23%]
PROT:
[14.9%]
Hazard mitigation index 0.002 0.001 0.0009 0.0002 0.023
(0-1) (0.0006) (0.0007) (0.001) (0.0006) (0.001)
[8.64%] [5.6%] [4.3%] [0.76%]
PROT:
[13.2%]
[14.8%] [8.68%] [3.56%] [1.51%]
PROT:
[16.0%]
Tree cover (M km2) (%) 214 116 28.6 20.2 377
(64.6) (62.8) (41.5) (68) (22.46)
[56.9%] [30.9%] [7.58%] [5.35%]
PROT:
[16.0%]
Protected area (M km2) 0.51 0.31 0.1 0.07 4.73
(%) [10.67%] [6.51%] [2.28%] [1.53%]
Deforestation (2.12) (2.02) (1.95) (1.99) 1.61
(2000–2012)% PROT:
(0.60)
[10.5%]
hazard mitigation and soil carbon storage services than ALLF. 0.1 of 0.58 Mkm2 of
CAF (17%) are protected leading to a two-third reduction of deforestation rates
relative to unprotected CAFs but these protected areas secure relatively little of the
CAF biodiversity, tree cover and ecosystem services (Table 10.5).
10.3.11 Tropical and Subtropical Asia
For tropical and subtropical Asia (Table 10.6) the mean per-unit-area clean water
provision tends to be highest for ALLF and TMF (because of the relative lack of
human influence on these forest types on this continent). For realised water provi-
sion all forest types have greater provision per-unit-area than ALL-L and that is also
true for all carbon services (except soil carbon in MF, CAFs and CFs, which is simi-
lar to, or lower than, the mean for ALL-L). Hazard mitigation services are highest
for CAFs. Species richness is greatest for CAFs, CFs and MFs whilst protected area
coverage is significant for all forest types. ALLF represents only 29% of Asia’s
protected areas and CAFs a significant 17.7%. Deforestation rates are greater in
ALLF. The rate in protected CAFs is 70% that of the all CAF rate so the protected
areas do not seem very effective. The 20.5% of the study areas that are CAFs pro-
vide 22.9% of realised water, 28.9% of above-ground carbon, but 17.3% of soil
carbon, 24.1% of hazard mitigation and 33% of the total tree cover. Their absolute
contribution is significant relative to the ALLF class and for most services their
provision per-unit-area is similar but less, though CAFs provide higher per-unit-
area hazard mitigation services than ALLF. Some 0.1 of 3.43 Mkm2 of CAF (13.9%)
are protected leading to a one-third reduction of deforestation rates relative to
unprotected CAFs, but these protected areas secure little (around 10%) of the total
CAF biodiversity, tree cover and ecosystem services (Table 10.6).
10.3.12 Distance Decay of Hydrological Services
Though CAFs and CFs generally provide a higher density of ecosystem services
than other forest types and in some continents have more local and downstream
beneficiaries than other forest types, they remain relatively small in areal extent and
thus their influence on national and continental natural capital and ecosystem ser-
vice accounts will reflect this. The exception may be for water-related services
where cloud forest types occupy important headwater locations with significant
populations downstream. However, we cannot assume that all of these populations
benefit equally from cloud forest hydrological services as the influence of a cloud
forest (or any other hydrological unit) will decay with distance downstream.
Figure 10.5 shows the hydrological footprint for CAFs in each continent, by pixel
and country. It is clear from Fig. 10.5b, d, f and h that the influence of CAFs is high-
est in those CAFs that have an excess of local rainfall over evapotranspiration in all
months and thus consistently produce runoff such that footprint can be 100% with
such zones. As rivers leave CAFs, however, the CAF runoff is mixed with runoff
Table 10.6 Ecosystem service provision by different forest types, tropical and subtropical Asia
ALLF sum MF sum CAF sum CF sum ALL-L sum
Ecosystem services (mean) (mean) (mean) (mean) (mean)
Water (potential Mm3 clean 7.81 2.97 3.06 2.27 13.5
water produced) (1.35) (1.12) (0.89) (1.08) (0.81)
[57.6%] [21.9%] [22.6%] [16.7%]
PROT:
[10.9%]
Water (realised Mm3 clean 7.64 2.94 3.03 2.25 13.3
water produced) (1.33) (1.12) (0.88) (1.08) (0.79)
[57.6%] [22.2%] [22.9%] [16.9%]
PROT:
[10.6%]
AG carbon (M tonnes C 81066 37441 39683 29079 137441
[58.9%] [27.2%] [28.9%] [21.2%]
PROT:
[14.0%]
Soil carbon (M tonnes C 65966 24148 30404 19246 17602
[35.5%] [13.7%] [17.3%] [10.9%]
PROT:
[12.9%]
Carbon sequestration 69.1 30.1 33.9 23.1 153.7
(M Dg/ha/yr) (Dg/ha/yr) (11.99) (11.45) (9.87) (11.1) (9.2)
[44.9%] [19.6%] [22.1%] [15.1%]
PROT: [9.5
%]
Hazard mitigation index (0-1) 0.017 0.008 0.012 0.006 0.047
(0.0029) (0.0028) (0.0033) (0.0028) (0.0028)
[35.68%] [15.89%] [24.1%] [12.6%]
PROT:
[11.0%]
[32.8%] [16.1%] [20.7%] [12.8%]
PROT:
[10.1%]
Tree cover (M km2) (%) 363 166 163 131 495
(63.1) (63.3) (47.7) (62.8) (29.5)
[73.3%] [33.6%] [33.1%] [26.5%]
PROT:
[12.7%]
Protected area (M km2) (%) 0.77 0.45 0.48 0.35 2.70
[28.8%] [16.8%] [17.7%] [13.1%]
Deforestation (2000–2012) % (4.03) (2.32) (1.83) (2.19) (2.23)
PROT:
(1.23)
[8.8%]
from surrounding areas and so the influence of the CAFs decays quickly. A few riv-
ers maintain influences of >50% throughout their course to the sea, others fall
quickly to less than 10%. Transparent areas on the maps have a value close to zero.
When averaged by country, it is clear that for certain small countries, cloud forests
can influence up to 50% of the river flow. For small, wet countries such as Costa
Rica the contribution is much less. For larger countries the hydrological footprint of
cloud forests is generally less than 20%, though it may be greater in lowland dry
seasons (not shown) if CAFs continue to receive rainfall when the lowlands do not
(Fig. 10.5).
If we calculate the hydrological footprint using WaterWorld outputs that include
the peculiar hydrology of CAFs better than global rainfall and evapotranspiration
datasets are able to, we see that many more of the cloud forests areas show all-year
(a) Annual average hydrological footprint (b) Annual average hydrological footprint
(downstream influence) of CAFs averaged (downstream influence) of CAFs (zoom for
over Country boundaries classes % Costa Rica) %
(c) Annual average hydrological footprint (d) Annual average hydrological footprint
over Country boundaries classes % Ecuador) %
Fig. 10.5 Hydrological footprint of CAFs by country average and per pixel, for each continent.
Background map data: Google, NASA, Terrametrics. (a) Annual average hydrological footprint
(downstream influence) of CAFs averaged over country boundaries classes %, (b) annual average
Fig. 10.5 (continued) hydrological footprint (downstream influence) of CAFs (zoom for Costa
Rica) %, (c) annual average hydrological footprint (downstream influence) of CAFs averaged over
country boundaries classes %, (d) annual average hydrological footprint (downstream influence)
of CAFs (zoom for Ecuador) %, (e) annual average hydrological footprint (downstream influence)
of CAFs averaged over country boundaries classes %, (f) annual average hydrological footprint
(downstream influence) of CAFs (zoom for Kenya) %, (g) annual average hydrological footprint
(downstream influence) of CAFs averaged over country boundaries classes %, (h) annual average
hydrological footprint (downstream influence) of CAFs (zoom for Bhutan) %
210 M. Mulligan
(e) Annual average hydrological footprint (f) Annual average hydrological footprint
over Country boundaries classes % Kenya) %
(g) Annual average hydrological footprint (h) Annual average hydrological footprint
over Country boundaries classes % Bhutan) %
Fig 10.5 (continued)
rainfall excess over evapotranspiration and thus generate continuous runoff and a
100% local hydrological footprint. This inflates the national average for small, wet
countries with a significant CAF area like Costa Rica (Fig. 10.6a) in this case from
around 4% to 30%. For larger, drier countries like Madagascar the national average
calculated with the WaterWorld data (8.5%) is similar to that calculated with the
remote sensing data (9%). This gives some confidence that the remote-sensing
based hydrological footprint results are reasonable pantropically even if they under-
estimate the footprint of CAFs in small, CAF dominated countries (Fig. 10.6).
Fig. 10.6 (continued) of CAFs averaged over country boundaries classes %, (b) annual average
hydrological footprint (downstream influence) of CAFs %, (c) annual average hydrological footprint
(downstream influence) of CAFs averaged over country boundaries classes %, (d) annual average
hydrological footprint (downstream influence) of CAFs %, (e) annual % of runoff generated by fog
averaged over major sub-basins (Hydrosheds) classes (%) and (f) annual % of runoff generated by
fog averaged over major sub-basins (Hydrosheds) classes (%)
(b) Annual average hydrological footprint

(a) Annual average hydrological footprint (downstream influence) of CAFs %
(downstream influence) of CAFs averaged
over Country boundaries classes %
(c) Annual average hydrological footprint (d) Annual average hydrological footprint
(downstream influence) of CAFs averaged (downstream influence) of CAFs %
over Country boundaries classes %
(e) Annual % of runoff generated by fog (f) Annual % of runoff generated by fog
averaged over Major sub-basins (Hydrosheds) averaged over Major sub-basins (Hydrosheds)
classes (%) classes (%)
Fig. 10.6 Hydrological footprint of CAFs in Costa Rica and Madagascar, based on WaterWorld data
and annual % of runoff generated by fog averaged over major sub-basins. Background map data:
Google, NASA, Terrametrics: (a) annual average hydrological footprint (downstream influence)
212 M. Mulligan
Moreover, another of WaterWorld’s outputs, the “annual percentage of runoff

generated by fog” indicates the key additional flux of water associated with the
cloud forest condition. These fog inputs are considered a key ecosystem service of
CAFs (Bennett et al. 2009; Breuer et al. 2013; Bruijnzeel et al. 2011) but fog con-
tributions to runoff decay quickly as distance downstream of cloud forest increases
(Fig. 10.6e, f). This is because rainfall inputs from surrounding lands are much
greater in magnitude and areal extent than fog inputs from CAFs, so that averaged
over major basins fog contributions remain below 5%.
10.4 Discussion and Conclusions
10.4.1 Cloud Forests and their Beneficiaries
Cloud forests (both CAFs and CFs) are much more spatially restricted than lowland
forests in all continents except Asia where CAFs are relatively extensive. Whilst
forests represent nearly 30% of the tropical land area, CAFs and MFs represent only
a third of these forests and CFs represent only 18%. These are significant areas
pantropically but represent a small fraction of global forest cover, which is domi-
nated by lowland tropical and boreal forests. Ecosystem services for all tropical
forest types are relatively poorly known (Brandon 2014).
Only 3.6% of the population of the tropics inhabit CAFs, though densities are
higher than for lowland forests in some continents, e.g. South America and higher
than the continent average for Central America and Africa. However CAFs do
occupy important headwater areas and have much larger downstream populations
than other forest types, especially in Africa and Asia where each square kilometre
of CAF has, on average, hundreds of thousands of downstream people. CAFs also
have more downstream dams than any other forest type, with each square kilome-
tre of CAF providing for an average of around 0.4 dams downstream, or put
another way, one in two square km of CAF being upstream of a dam. Protected
CAFs tend to be upstream of greater population densities and more dams than
unprotected CAFs, but not exclusively so between the continents. However, down-
stream populations for protected CAFs vary from tens of thousands in Central and
South American protected CAFs through to hundreds of thousands for Asian and
African protected CAFs. These are a small fraction of the millions to billions of
people in these continents that depend on ecosystem services from other systems
(Summers et al. 2012).
10.4.2 T
he Ecosystem Services of Cloud Forests, Continent
by Continent
In general, forested land tends to provide greater hydrological, carbon and hazard
mitigation services than the continental averages as well as harbouring more species
and, of course, tree cover. Different forest types have different concentrations of
ecosystem services across the continents but in general CAFs and CFs tend to pro-
vide the highest densities of hydrological services with CFs also providing high
densities of carbon storage, hazard mitigation and species richness in some conti-
nents. Though the concentration of this natural capital and ecosystem services is
high in cloud forest types, their contribution to the total national or continental
stocks and flow is low by virtue of their limited extent compared with other for-
est types.
There are also significant differences between continents. Figure 10.7a shows the
mean for each variable within CAFs by continent. Since the variables have different
units, they are expressed as a percentage of the maximum for that variable by
(a) (b)
(c)
Fig. 10.7 (a) Mean (density) of variables within CAFs: comparison between continents and (b)
total (sum) of variables within CAFs: comparison between continents, (c) total (sum) of variables
within all forests: comparison between continents
214 M. Mulligan
continent. Clearly South America has the “richest” CAFs for many ecosystem ser-
vices, followed by Africa (carbon sequestration and species richness) and Central
America (hazard mitigation). Asia tends to have the lowest density of ecosystem
service provision by CAFs. Mean protected area coverage of CAFs is highest in
Latin America and lowest in Asia, whilst mean deforestation in CAFs is highest in
Central America and lowest in South America. For the sum of these variables within
CAFs (Fig. 10.7b) Asia is highest for all variables (having the greatest extent of
CAFs), followed by South America, Africa and Central America. Though Asia has
the lowest fraction of CAFs protected, it has the highest absolute area of protected
CAFs, followed by South America and also the highest deforestation of CAFs,
again followed by South America.
The situation is profoundly different when one examines all forest types
(Fig. 10.7c) in which case the Amazon forests dominate most ecosystem service
totals, bringing South America well ahead of Asia and Africa for all services except
hazard mitigation and for species richness, despite the continent being smaller than
Asia. Asia tops the continents for hazard mitigation services, reflecting the moun-
tainous nature of the continent and importance of forests in mitigating those haz-
ards. South America has the greatest protected forest area but also close to the
greatest deforestation rate.
Clearly Asia is a very significant continent with respect to ecosystem services
value of CAFs but also with respect to the need to better protect these rapidly disap-
pearing forests. This is in contrast to the historic research effort on cloud forests
with a Google Scholar search for [“cloud forest” central America] returning 17,200
results (1,270 including the additional term “ecosystem services”), [“cloud forest”
south America]: 15,500 (1,140 with “ecosystem services”), [“cloud forest” Africa]:
9,480 (1,010 with “ecosystem services”), but [“cloud forest” Asia] only 6,590 (680
with “ecosystem services”). The body of research, including ecosystem services
research, in cloud forests is highly skewed towards Central and South America, yet
the most important cloud forests for most people locally and downstream are in
Africa and Asia. The next decade should see a much greater focus on these impor-
tant but pressured and rapidly changing forest regions.
10.4.3 T
he Implications of Distance Decay for Ecosystem
Services from Cloud Forests
We have provided data on the ecosystem service contributions of tropical forest

types by continent and forest type. Ecosystem service provision has been calculated
on the basis of both potential and realised services. Some realised services such as
carbon storage and sequestration have global beneficiaries meaning that proximity
to the sites of service production is not necessary to receive the benefits. Others
require beneficiaries to be downhill or downstream of the sites of service produc-
tion. These services are much more challenging to account for because, whilst one
can sum the number of potential beneficiaries (e.g. people, dams) downstream, it is
not necessarily the case that all of these people are significantly affected by the
cloud forest upstream. This is because there is a distance decay function for these
services that is determined by the size of the site of production in relation to all other
sites of production between the site of production and the ecosystem service con-
sumer. We use the hydrological footprint and the fog contribution to runoff to indi-
cate that, by nature of their often distant and spatially restricted nature, cloud forests
rarely have more than 10% influence on the flow of rivers. This fundamental hydro-
logical principle needs to be observed in analyses of the influences of cloud forests
and other spatially restricted areas (such as protected areas) on ecosystem services
downstream.
By calculating the downstream difference in hydrological footprint from each
pixel to its downstream neighbour and then masking all but the negative values (i.e.
downstream decay) we can examine the mean decay in hydrological influence per
cell (sq km) (Fig. 10.8). This is highest at the CAF, non-CAF boundary of course but
also occurs along the full length of rivers exiting CAFs. Values for Asian countries
(Fig. 10.8) vary from around −0.3%/km to −25%/km, indicating a decline to imper-
ceptible influence within around 333–7 km, respectively, downstream, depending
on local conditions. These values are dominated by the higher CAF-non CAF
boundary decay values, with values within river systems being much smaller. For
the “annual percent of runoff generated by fog” variable for Madagascar, the rate of
(b) Downstream differences of negative

(a) Downstream differences of negative downstream influence of CAFs, Asia, average
downstream influence of CAFs, Asia by country
Fig. 10.8 Downstream decay of hydrological footprint: example from Asia. Background map
data: Google, NASA, Terrametrics: (a) downstream differences of negative downstream influence
of CAFs, Asia and (b) downstream differences of negative downstream influence of CAFs, Asia,
average by country
216 M. Mulligan
decay downstream is −0.88% within the CAF area and −0.3% overall, indicating
decay to indiscernible within 133–333 km downstream. The “half-life” of a cloud
forest downstream hydrological influence can thus be considered as 166 km with an
influence drop of 25–50% at the CAF, non-CAF boundary and then a fraction of 1%
per km downstream thereafter. However, there are no rules of thumb. Because of the
significant site-to-site variability and thus range of downstream decay functions,
these values should be calculated on a site-by-site basis and can be so with
WaterWorld. They are fundamental to understanding the relationship between pro-
duction and consumption of hydrological ecosystem services, especially for remote,
isolated systems like cloud forests.
Acknowledgements The author gratefully acknowledges all the institutions that make global
datasets available for the scientific community to use.
References
Bennett EM, Peterson GD, Gordon LJ (2009) Understanding relationships among multiple ecosys-
tem services. Ecol Lett 12:1394–1404
BIRDLIFE (2012) Birdlife International IUCN red list for birds. http://www.iucnredlist.org/
technical-documents/spatial-data#birds
Brandon K (2014) Ecosystem services from tropical forests: review of current science. CGD
working paper 380. Washington, DC: Center for Global Development. http://www.cgdev.org/
publication/ecosystem-services-tropical-forests-review-currentscience-working-paper-380
Brauman KA, Daily GC, Duarte TKE, Mooney HA (2007) The nature and value of ecosystem
services: an overview highlighting hydrologic services. Annu Rev Environ Resour 32:67–98
Breuer L, Windhorst D, Fries A, Wilcke W (2013) Supporting, regulating, and provisioning hydro-
logical services. In: Ecosystem services, biodiversity and environmental change in a tropical
mountain ecosystem of South Ecuador. Springer, Berlin, pp 107–116
Bruijnzeel LA, Mulligan M, Scatena FS (2011) Hydrometeorology of tropical montane cloud for-
ests: emerging patterns. Hydrol Process 25:465–498
Bubb P, May I, Miles L, Sayer J (2004) Cloud forest agenda. UNEP-World Conservation
Monitoring Centre, Cambridge, UK. http://sea.unep-wcmc.org/forest/cloudforest/index.cfm
Center For International Earth Science Information Network (CIESIN) (2002) Country-level GDP
and downscaled projections based on the A1 A2 B1 and B2 Marker Scenarios 1990-2100.
CIESIN Columbia University, Palisades, NY. http://www.ciesin.columbia.edu/datasets/
downscaled
Dapples F, Lotter AF, Van Leeuwen JF, Van Der Knap WO, Dimitriadis S, Oswald D (2002)
Paleolimnological evidence for increased landslide activity due to forest clearing and land-use
since 3600 cal BP in the western Swiss Alps. J Paleolimnol 27:239–248
De Groot RS, Alkemade R, Braat L, Hein L, Willemen L (2010) Challenges in integrating the
concept of ecosystem services and values in landscape planning, management and decision
making. Ecol Complex 7:260–272
Dilley M (2005) Natural disaster hotspots: a global risk analysis. Version 1.0. Disaster risk man-
agement series, No. 5. World Bank, Washington DC. https://openknowledge.worldbank.org/
handle/10986/7376
FAO (1998) FRA 2000. Terms and definitions. FRA working paper 1, FAO Forestry Department.
http://www.fao.org/forestry/fo/fra/index.jsp. Under publications
FAO (2010) GIEWS: world road trails. Whole world’s roads and railways. http://ldvapp07.fao.
org:8030/downloads/layers/world_roadstrail.xml
Farr TG, Kobrick M (2000) Shuttle radar topography mission produces a wealth of data. Eos Trans
81:583–585
Fisher B, Turner RK, Morling P (2009) Defining and classifying ecosystem services for decision
making. Ecol Econ 68:643–653
Funnell DC, Price MF (2003) Mountain geography: a review. Geogr. J 169:183–190
Hansen M, Defries R, Townsend JR, Carroll M, Dimiceli C, Sohlberg R (2006) Vegetation con-
tinuous fields MOD44B, 2001 percent tree cover, collection 4. University of Maryland, College
Park, MDaryland, 2001. http://glcf.umd.edu/data/vcf/
Hijmans RJ, Cameron SE, Para JL, Jones PG, Jarvis A (2005) Very high resolution interpolated
climate surfaces for global land areas. Int J Climatol 25:1965–1978
Ho LC (1976) Variation in the carbon/dry matter ratio in plant material. Ann Bot 40:163–165
IUCN (2010) An analysis of reptiles on the 2010 IUCN red list. http://www.iucnredlist.org/
technical-documents/spatial-data
IUCN, Conservation International, and Nature Reserve (2008) An analysis of amphibians on the
2008 IUCN red list. http://www.iucnredlist.org/amphibians
IUCN, Conservation International, Arizona State University, Texas A&M University, University of
Rome, University OF Virginia, Zoological Society London (2008) An analysis of mammals on
the 2008 IUCN Red List. http://www.iucnredlist.org/mammals
Jarvis A, Mulligan M (2009) The climate of cloud forests. Hydrol Process 25:327–343
Jenkins M, Scherr SJ, Inbar M (2004) Markets for biodiversity services: potential roles and chal-
lenges. Environ Sci Policy Sustain Dev 46:32–42
Jones R (2002) Algorithms for using a DEM for mapping catchment areas of stream sediment
samples. Comput Geosci 28:1051–1060
LandScanTM (2007) LandScanTM global population database. Oak Ridge National Laboratory, Oak
Ridge, TN. http://www.ornl.gov/landscan/
Lehner B, Doll P (2004) Development and validation of a global database of lakes, reservoirs and
wetlands. J Hydrol 296:1–22
Lehner B, Verdin K, Jarvis A (2008) New global hydrography derived from spaceborne elevation
data. Eos Trans AGU 89:93–94
Mu Q, Zhao M, Running SW (2011) Improvements to a MODIS global terrestrial evapotranspira-
tion algorithm. Remote Sens Environ 115:1781–1800
Mulligan M (2009a) Global mean dry matter productivity based on SPOT-VGT (1998–2008).
http://geodata.policysupport.org/dmp
Mulligan M (2009b) The human water quality footprint: agricultural, industrial, and urban impacts
on the quality of available water globally and in the Andean region. In: Proceedings of the
international conference on integrated water resource management and climate change, Cali,
Colombia, p 11.
Mulligan M (2010a) Modeling the tropics-wide extent and distribution of cloud forest and
cloud forest loss, with implications for conservation priority. In: Bruijnzeel LA, Scatena FN,
Hamilton LS (eds) Tropical Montane Cloud forests: science for conservation and management.
Cambridge University Press, Cambridge, pp 14–39
Mulligan M (2010b) A combined global database of mines. http://geodata.policysupport.org/mines
Mulligan M (2010c) A combined global database of oil and gas wells. http://geodata.policysup-
port.org/oilandgas
Mulligan M (2010d) SimTerra: A consistent global gridded database of environmental properties
for spatial modelling. http://www.policysupport.org/simterra
Mulligan M (2013) WaterWorld: a self-parameterising, physically-based model for application in
data-poor but problem-rich environments globally. Hydrol Res 44(5):748–769
Mulligan M (2015a) Climate change and food-water supply from Africa’s drylands: local impacts
and teleconnections through global commodity flows. Int J Water Resour Dev 31:450–460
218 M. Mulligan
Mulligan M (2015b) Trading off agriculture with nature’s other benefits, spatially. In: Zolin CA,
De Rodrigues RAR (eds) Impact of climate change on water resources in agriculture. CRC
Press, Boca Raton, FL, pp 184–205
Mulligan M, Clifford NA (2015) Is managing ecosystem services necessary and sufficient to
ensure sustainable development? In: Springet D, Redclift M (eds) Handbook of sustainable
development. Routledge, Abingdon, UK, pp 179–195
Mulligan MA, Guerry K, Arkema K, Bagstad D, Villa F (2010) Capturing and quantifying the
flow of ecosystem services. In: Silvestri S, Kershaw F (eds) Framing the flow: innovative
approaches to understand, protect and value ecosystem services across linked habitats. UNEP
World Conservation Monitoring Centre, Cambridge, UK
Mulligan M, Saenz L, Van Soesbergen A (2011) Development and validation of a georeferenced
global database of dams. http://geodata.policysupport.org/dams
Naidoo R, Balmford A, Costanza R, Fisher B, Green RE, Lehner B, Malcolm TR, Ricketts TH
(2008) Global mapping of ecosystem services and conservation priorities. Proc Natl Acad Sci
105:9495–9500
National Geophysical Data Center/World Data Center (NGDC/WDC) (2011) Historical tsunami
database, Boulder, CO, USA. http://www.ngdc.noaa.gov/hazard/tsu_db.shtml
Pena-Arancibia JL, Van Dijk AI, Guerschman JP, Mulligan M, Bruijzeel LA, McVicar TR (2012)
Detecting changes in streamflow after partial woodland clearing in two large catchments in the
seasonal tropics. J Hydrol 416:60–71
Ramankutty N, Evan AT, Monfreda C, Foley JA (2008) Farming the planet: 1. Geographic distribu-
tion of global agricultural lands in the year 2000. Glob Biogeochem Cycles 22:1–19
Ruesch A, Gibbs HK (2008) New IPCC tier-1 global biomass carbon map for the year 2000.
Available online from the Carbon Dioxide Information Analysis Center, Oak Ridge National
Laboratory, Oak Ridge, Tennessee. http://cdiac.ornl.gov
Saatchi S, Harris NL, Brown S, Lefsky M, Mitchard ET, Salas W, Zutta BR, Buermann W, Lewis
SL, Hagen S, Petrova S, White L, Silman M, Morel A (2011) Benchmark map of forest carbon
stocks in tropical regions across three continents. Proc Natl Acad Sci 14:9899–9904
Scharlemann JPW, Hiederer R, Kapos V (2009) Global map of terrestrial soil organic carbon
stocks. A 1-km dataset derived from the harmonized World soil database. UNEP-WCMC &
EU-JRC, Cambridge
Schneider A, Friedl MA, Potere D (2009) A new map of global urban extent from MODIS data.
Environ Res Lett 4:1–11
Siebert S, Doll P, Feick S, Fremken K, Hoogeveen J (2007) Global map of irrigation areas version
4.0.1. University of Frankfurt (Main), Germany, and FAO, Rome, Italy. http://www.fao.org/nr/
water/aquastat/irrigationmap/index10.stm
Spalding MD, Blasco F, Fields CD (eds) (1997) World Mangrove Atlas. The International Society
for Mangrove Ecosystems, Okinawa, p 178
Summers JK, Smith LM, Case JL, Linthurst RA (2012) A review of the elements of human well-
being with an emphasis on the contribution of ecosystem services. Ambio 41:327–340
Townshend JRG, Carroll M, Dimiceli C, Sohlberg R, Hansen M, Defries R (2011) Vegetation con-
tinuous fields MOD44B, 2001 percent tree cover, collection 5. University of Maryland, College
Park, Maryland. http://glcf.umd.edu/data/vcf/
USGS (2006) Shuttle radar topography mission water body dataset. http://edc.usgs.gov/products/
elevation/swbd.html
World Database on Protected Areas (WDPA) (2014) Annual release 2014. www.protectedplanet.net
Chapter 11
Conclusions, Synthesis, and Future
Directions
Randall W. Myster
11.1 Conclusions
Like the Amazon the Andes is more than just mountains, grasslands, deserts, and
forests. Its cloud forests in particular (Fig. 11.1) provide important ecosystem ser-
vices for South America and help define various peoples and their cultures. In Chap.
1 I presented those cloud forests as occurring across large and important elevational
and latitudinal gradients, where the edges of those gradients may be stressed by
climatic conditions (see Fig. 11.2 as an artist’s presentation). Physical effects of
stress include low temperature, wind, and relatively little precipitation. Chemical
effects of stress include lack of nutrients and chemical effects of burning. I also
presented my research into cloud forests structure, function, and dynamics and the
difficulties of cloud forest recruitment. In particular, the seed rain is probably not
limiting cloud forest recruitment, regeneration, and plant community dynamics.
What happens after dispersal is more limiting and predation took most of those
seeds, more seeds than was taken by pathogens or that germinated.
The book started with a discussion of the tree line between cloud forest and
paramó grassland suggesting that it is determined not just by climatic (temperature,
precipitation) and local conditions (wind) but also human impacts, there was an
ancient anthropogenic legacy of the different ways that people used tropical mon-
tane cloud forests landscapes in the past. Indeed, Andean treeline ecotone regions
are cultural/socioecological landscapes that require more than traditional conserva-
tion approaches. Next, litterfall in the Andean cloud forest as a proxy for the total
productivity of forests and a major vector of nutrient cycling was discussed. These
forests become increasingly nutrient efficient with increasing elevation, while there
is no indication of a general change in the kind of nutrient limitation. There may be
systematic relationships between abiotic conditions and litterfall, which could be
R. W. Myster (*)
Department of Biology, Oklahoma State University,

https://doi.org/10.1007/978-3-030-57344-7_11
220 R. W. Myster
Fig. 11.1 The Andean cloud forest
Fig. 11.2 Wald Bau

(forest-construction), 1919,
by Paul Klee (1879–1940),
mixed media chalk,
27 × 25 cm—Museo del
Novecento, Milan, Italy
used to predict litterfall in the Andes. The observed elevational influence of litterfall
in the humid tropical Andes suggests that the forest productivity will likely respond
to climate change driving the vegetation belts to higher elevation with an unknown
overall effect on C sequestration of these forests.
We saw that just about all investigated trees were arbuscular mycorrhizal (AM),
each tree individual has an individual composition of AM fungal partners, which
usually consists of one (two) common fungi and 3–8 rarely occurring fungi. The
common species are not specialists, but generalists, from 1000 masl to 4000 masl,
there is a high turnover in terms of both the AM fungi and the plant species. Only
three species of fungi occur at all altitudes. Also, information gathered for the
Tucuman Amazon parrot (Amazona tucumana) for 15 years show it to have high
11 Conclusions, Synthesis, and Future Directions 221
rates of nesting success, large clutches, and many fledglings per laying female.
Nesting and spatial requirements could limit management, and so to ensure the
conservation of this parrot outside protected areas managers should promote the
retention of large B. salicifolius, C. lilloi, and J. australis birds and their nests that
are selected for.
Cloud forests of the Andes are diverse and have a high degree of endemism,
attributed to a mixture of historical, evolutionary, and ecological processes. For
example, the relationship between species diversity and spatial structure of the
Andean mountains, or the role of physical barriers that arose during the late Eocene
(uplift of the Andes) and Neogene and the range shifts is caused by climate during
the quaternary period. In addition, species tolerances to environmental conditions
(e.g., temperature) can affect species dispersal and facilitate allopatric isolation.
And finally, ecosystem services are the benefits that human populations derive from
nature, here from cloud forests which occur in some important headwater areas with
very significant human populations locally and downstream. We need to look at the
benefits which accrue from these services, who benefits, and which of these benefits
are protected for the future, within the context of deforestation trends.
11.2 Synthesis
Within the biology of any terrestrial ecosystem, both abiotic and biotic processes
move, primarily, in and out of the plant phytomass (the biomass [Myster 2003] and
necromass together). In this volume, and in my first four books, I have presented a
view of terrestrial ecosystems as plant-centered, where components of ecosystems
cycle in and out of—or flow through, like energy—the total plant phytomass. No
other component or components of the Andean cloud forest ecosystem, except the
phytomass, can assume this central role as a conduit for physical, chemical, and
biological parts of the ecosystem (Myster 2001). Only the phytomass mediates and
integrates between biogeochemical cycles (including cycles of productivity and
decomposition: Myster 2003), conducting most an ecosystem’s energy and nutrient
processing. The phytomass should, therefore, be put in the center in our conceptual
models of terrestrial ecosystems.
Ecosystem structure, function, and dynamics in the Andean Cloud forest is then
fundamentally about the dynamics of the phytomass (Myster 2001, 2003) where
biomass loss opens a patch of space where resources may be available (Bazzaz
1996). Then because plants are sessile organisms—and thus “seek” space—that
patch of space can be occupied by other plants either by them growing into it them-
selves and/or by new plants recruiting into it. Ranking disturbances, using phyto-
mass loss as the metric, creates a hierarchy which includes phytomass removed
without whole plant death (“none” where no plant dies but some of its biomass was
removed and/or some of its necromass decayed), with one plant dying (“one”), and
with more than one plant dying as a group (“many”). And so, disturbances occur at
222 R. W. Myster
different levels of this hierarchy of biomass loss where each disturbance at each
level creates a discrete patch of space where plant responses may be present.
Those responses are conceptualized as plant-plant replacements (Myster 2018)
from “none” where no new plant recruits into that now available patch of space but
neighboring plants may grow into it, to “one” where one new plant recruits into that
now available patch of space and may grow to occupy additional space in that patch,
and/or to “many” where more than one plant as a group recruits into that now avail-
able patch of space and may grow to occupy additional space in that patch. For
example, during old-field succession plants are often replaced by plants with the
same size or larger seeds (Myster 2007). I posit that while natural disturbances may
lead to plant-plant replacements like none => none, none => one, none => many, for
example: Myster 2018), human-caused disturbances—because they remove rela-
tively larger amounts of biomass—may lead to replacements like many => none,
many => one, many => many. And so, the plant-plant replacements in an area could
indicate the level and kind of past disturbance, just like crops and/or associated spe-
cies still growing in an agricultural field after abandonment indicate the crop that
was growing there before abandonment. There may also be a tendency for responses
to match their disturbances (parallelism) that is none => none, one => one and
many => many replacements are most common. These neighborhoods spaces may
overlap: zero degree no overlap/open space, 1° two trees overlap, 2° three trees
overlap etc.… And the more overlap the less likely other plants can use the space.
In conclusion, (1) biomass loss is the best way to measure disturbance (seen as
severity in the disturbance regime) and to compare between disturbances (in this
review we saw that severity differed fundamentally among all disturbances of the
Andean cloud forest), (2) biomass can be lost as part of a plant, as the whole plant,
or as many plants lost as a group, (3) disturbances can thus be placed on a hierarchy
based on the relative amount of biomass lost, (4) at each point on a hierarchy, a
patch of space is created as biomass is lost, and resources may become available
within that patch of space, (5) living plants can then respond in that newly created
patch of space, and (6) those responses are among nine classes of plant-plant
replacement. These plant-plant replacements were documented among the responses
and disturbances of the Andean cloud forest research papers in this review.
11.3 Future Directions
Chapters suggest the continued sampling of all ecosystem components in large

Andean cloud forest plots with an emphasis on exploration of interactive links
among ecosystem components. This should be followed by field experiments that
need to be designed to find these links and should have a special focus on the early
parts of regeneration (i.e. recruitment of seeds and saplings).
11 Conclusions, Synthesis, and Future Directions 223
References
Bazzaz FA (1996) Plants in changing environments: linking physiological, population, and com-
munity ecology. Cambridge University Press, Cambridge
Myster RW (2001) What is ecosystem structure? Caribb J Sci 37:132–134
Myster RW (2003) Using biomass to model disturbance. Community Ecol 4:101–105
Myster RW (2007) Post-agricultural succession in the Neotropics. Springer-Verlag, Berlin
Myster RW (2018) The nine classes of plant-plant replacement. Ideas Ecol Evol 11:29–34
View publication stats

Book 5 Complete PDF

Uploaded by

Copyright:

Available Formats

Book 5 Complete PDF

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Book 5 Complete PDF

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

The Andean Cloud Forest

Book · January 2021

book 3 View project

The user has requested enhancement of the downloaded file.

The Andean Cloud Forest

ISBN 978-3-030-57343-0 ISBN 978-3-030-57344-7 (eBook)

© Springer Nature Switzerland AG 2021

“Det som er elsket er alltid vakkert”.

10 Mapping Hydrological Ecosystem Services and Impacts of

Julián Díaz Instituto de Investigación de Recursos Biológicos Alexander von

Pascual J. Soriano Laboratorio de Ecología Animal, Departamento de Biología,

1.1 The Andes

© Springer Nature Switzerland AG 2021 1

1.2 The Andean Cloud Forest

1.3.1 Curve-Fitting Floristic Patterns

Marini et al. 2010), as was done to generate dominance-diversity curves (Myster

1.3.2 Curve-Fitting Physical Structure Patterns

1.4.1 One ha Plot: Floristics and Physical Structure Sampling

1.4.2 One ha Plot: Plant-Plant Replacements over Six Years

Within MR are secondary cloud forests of recovering sugarcane plantations, banana

1.7 Case Study: Primary Cloud Forest at Yanacocha Reserve

1.7.1 One ha Plot: Floristics and Physical Structure Sampling

About this Book

R, Balslev HR (eds) Estudios Sobre Diversidad y Ecologia de Plantas. Pontificia Universidad

2.1 Andean Critical Biogeography of Scale

2.1.1 Annotations of Narrative Framework

F. O. Sarmiento, PhD (*)

© Springer Nature Switzerland AG 2021 25

2.2 Why We Call them Cloud Forests? Onomastics at Play

CLOUD BASE LINE

Fig. 2.1 A display of different lines found in the tropical mountainscape

American Universities (Hamilton et al. 1993). The conference proceedings, published

2.2.1 Landscape Phenology and Threatening Treeline Myths

By switching the rhetoric order, we highlight these cloud forests environments

2.2.2 Atmospheric Tides and Fluvial Rhythms

2.2.3 Tide, Gravity and the Will of People

2.3 Lines in the Longer Cyclicity of Mountainscapes

electromagnetic conditions in the shortest wavelength of TMCFs. This mutation-­

influence of meteorological conditions to affect the manifestation of a continuous

2.5 Mountain Effects and the Andean Treeline Ecotone Region

follow Acosta-Solís’s advice to emphasize the use of “Andean” treeline instead of

2.7 Discussion Remarks: Treelines of the Mindscape

Paola Isaacs-Cubides, Julián Díaz, and Tobias Leyva-Pinto

Colombia tiene una superficie de 1.141.000 ha, y es altamente heterogénea en tér-

P. Isaacs-Cubides (*) · J. Díaz

© Springer Nature Switzerland AG 2021 43

deforestación y degradación de los ecosistemas tiene un impacto negativo en los

Adicionalmente, es necesario considerar un aspecto desde la funcionalidad del

de servicios, se obtuvo un mapa final acumulado, para definir zonas de importancia

3.3 Resultados and Discussion

De acuerdo al mapa de coberturas de la tierra, de las 18.368.518 ha evaluadas, el

Fig. 3.2 Porcentaje de cobertura presente en las zonas de montaña

Fig. 3.3 Distribución de las coberturas presentes en las zonas de montaña

remanencia y transformación (Fig. 3.6). Esto se debe a que los parches empiezan a

Fig. 3.5 Distribución por tamaños de las coberturas naturales

Fig. 3.6 Distribución por índice de forma de las coberturas naturales

Fig. 3.9 Acumulación de carbono en biomasa, de acuerdo a los tipos de cobertura

Nariño, cordillera central en Antioquia, son lugares de elevada importancia por la

Fig. 3.10 Oferta hídrica calculada para las zonas de montaña

10 Mapping Hydrological Ecosystem Services and Impacts of

1.1 The Andes

1.2 The Andean Cloud Forest

1.3.1 Curve-Fitting Floristic Patterns

1.3.2 Curve-Fitting Physical Structure Patterns

1.4.1 One ha Plot: Floristics and Physical Structure Sampling

1.4.2 One ha Plot: Plant-Plant Replacements over Six Years

1.7 Case Study: Primary Cloud Forest at Yanacocha Reserve

1.7.1 One ha Plot: Floristics and Physical Structure Sampling

2.1 Andean Critical Biogeography of Scale

2.1.1 Annotations of Narrative Framework

2.2 Why We Call them Cloud Forests? Onomastics at Play

2.2.1 Landscape Phenology and Threatening Treeline Myths

2.2.2 Atmospheric Tides and Fluvial Rhythms

2.2.3 Tide, Gravity and the Will of People

2.3 Lines in the Longer Cyclicity of Mountainscapes

electromagnetic conditions in the shortest wavelength of TMCFs. This mutation-

2.5 Mountain Effects and the Andean Treeline Ecotone Region

2.7 Discussion Remarks: Treelines of the Mindscape

3.3 Resultados and Discussion

4.3 Results and Discussion

4.3.1 The Ecohydrology of TACF

5.2 Study Sites and Methods

5.4.1 Litterfall and Litterfall Fractions in Andean Forests

6.2 The Study Region

6.3 Materials and Methods (For Details See Haug et al. 2019)

6.5.2 Species Richness Along the Elevational Gradient

7.1 The Tucuman Amazon

7.3 Nesting Habitat Requirements of Tucuman Amazon

7.4 Reproductive Biology of the Tucuman Amazon

7.5 Conservation Implications

7.6 Knowledge Gaps

8.1 Some Historic Considerations

8.2 Frugivore Bats of Genus Sturnira: A Study Case

8.3 Materials and Methods

d etermined using ANCOVA. Significance was accepted at P < 0.05. The statistical