02 Whole

ResearchOnline@JCU
This file is part of the following reference:
Barlow, Christopher G. (1998) Aspects of the biology of

juvenile barramundi Lates calcarifer (Bloch) relevant to
production for recreational fisheries and farming, with a
note on the proposal to introduce Nile perch Lates
niloticus (L.) to Australia. PhD thesis, James Cook
University.
Access to this file is available from:
http://eprints.jcu.edu.au/24097/
The author has certified to JCU that they have made a reasonable effort to gain
permission and acknowledge the owner of any third party copyright material
included in this document. If you believe that this is not the case, please contact
[email protected] and quote http://eprints.jcu.edu.au/24097/
Aspects of the biology of juvenile barramundi
Lates calcarifer (Bloch)
relevant to production for recreational fisheries and
farming, with a note on the proposal to introduce
Nile perch Lates niloticus (L.) to Australia
Thesis submitted by
Christopher G. BARLOW BSc (JCUNQ) MSc (UNSW)
in January 1998
for the degree of Doctor of Philosophy in

the Department of Zoology at qvR cv4.7 vee-
-
James Cook University of North Queensland

STATEMENT ON ACCESS
I, the undersigned, the author of this thesis, understand that James Cook University
of North Queensland will make it available for use within the University Library
and, by microfilm or other means, allow access to users in other approved libraries.
All users consulting this thesis will have to sign the following statement:
`In consulting this thesis I agree not to copy or closely paraphrase it in whole
or in part without the written consent of the author; and to make proper written
acknowledgement for any assistance which I have obtained from it.'
Beyond this, I do not wish to place any restriction on access to this thesis.
Christopher G. Barlow 15 January 1998
11
ABSTRACT
The research covered in this thesis concentrated primarily on improving production
protocols for juvenile barramundi Lates calcarifer through studies on diet and
feeding habits, pond rearing techniques, effects of photoperiod on growth, and
weaning strategies. Juvenile barramundi are produced in northern Australia for two
reasons; to supply seed for the aquaculture industry, and to supply fingerlings for
recreational fisheries enhancement programs. Within this context, two related
studies were undertaken: firstly, an analysis of the proposal to introduce Nile perch
to Australia, which preceded the barramundi production studies; and secondly, an
evaluation of the use of circulus patterns on scales for discriminating wild and
hatchery-produced barramundi.
A review of the historical and present distribution of barramundi (a catadromous
species) in Queensland indicated that barriers (barrages, weirs and dams) built in
river systems, coupled with the inability to negotiate even minor stream barriers,
have restricted access of this fish to much of its original, natural habitat. Further,
while the construction of dams has created vast new aquatic habitats (potentially at
least 100 000 ha in Queensland), these have also been inaccessible to barramundi via
its normal life-cycle movements. To fill the 'niche' made available by the
decreasing distribution of barramundi, it was proposed that a congener of
barramundi, the Nile perch Lates niloticus, be introduced to establish sport fisheries
in tropical impoundments. The principal rationale for this introduction was that,
111
unlike barramundi, the Nile perch reproduces in fresh waters and, hence, once
established would be capable of sustaining breeding populations.
Contrarily, however, three lines of evidence suggested that the introduction of the
Nile perch would have negatively impacted upon Australian aquatic fauna. The
lower temperature tolerance of the species and analysis of water temperatures in
rivers in eastern Australia indicated that its range would have extended to temperate
regions, thus endangering established fisheries for native species in those areas. The
introduction of the Nile perch, an opportunistic predator, to Lakes Victoria and
Kyoga in eastern Africa caused a drastic decrease in species diversity and fish
biomass. L. niloticus is not restricted to lacustrine habitats, and known features of its
biology indicate that it could have colonised and adversely affected the fauna in a
broad range of freshwater habitats in Australia. The risks associated with the
proposed introduction were considered to outweigh the potential benefits, and hence
it was abandoned. As an alternative, attention was given to hatchery production of
barramundi as a means of supplying fingerlings for stocking fresh waters for
enhancement of recreational fisheries in northern Australia.
Hatchery-reared barramundi fry were studied to determine feeding behaviour, diel
feeding patterns, stomach evacuation rates, daily food consumption and growth rates.
At 16-18 mm total length (TL), the feeding behaviour of the fry changed abruptly
from a roving zooplanktivore to that of a lurking predator. A distinct change in
pigmentation accompanied the change in feeding behaviour. Fry reared in hatchery
ponds were obligate zooplanktivores between 10 and about 17 mm. Between about
iv
17 and 50 mm the diet changed progressively from zooplankton to insect larvae to
small vertebrates. The fry were visual feeders, taking food throughout the day, and
showed a peak in feeding activity at dusk. Feeding continued at a reduced level
under moonlight conditions, but ceased in total darkness. Stomach evacuation rates
for 16 mm fry under continuous feeding and non-feeding conditions were 47 and 210
minutes, respectively; for 37 mm fry the rates were 73 and 108 minutes, respectively.
The daily rations for these two size groups were 19-86% and 38-56%, respectively.
Specific growth rates were 13-16% body weight/day for fry reared in ponds.
Laboratory-based experiments were conducted to determine the vulnerability of
different sized barramundi fry to predation by nymphs of the dragonfly Pantala
flavescens. Mortality of 10 mm mean TL fry was significantly greater in the
presence of dragonfly nymphs, whereas 20 mm mean TL fry exhibited
comparatively minor levels of mortalities. The results accorded with feeding
behaviour patterns of the different sized fry and the development of an escape
response in barramundi fry at 16-18 mm TL. An examination of the growth rates of
P. flavescens in newly-filled ponds, the development of the pond fauna on which the
barramundi fry feed, and the growth rates of fry, indicated a rearing strategy to
optimise survival of barramundi fry reared in freshwater ponds.
An experiment was conducted to determine the effect of extended periods of light on
the growth, survival, feeding pattern and daily food consumption of barramundi fry
reared in a freshwater hatchery. There was no significant difference in growth or
survival of fry, initially 11-12 mm TL, in either 12, 18 or 24 hours light. Fish
exposed to 12L/12D photoperiod fed continuously during daylight, and ceased
feeding in darkness. Under continuous daylight conditions, fish fed throughout the
normal daytime period, but ceased feeding at a time corresponding to the normal
onset of darkness; feeding started again near midnight. Daily food consumption for
34 mm fish was approximately 40% more in continuous light than in 12L/12D
photoperiod. The results clearly showed that there was no advantage to be gained by
rearing barramundi fry in extended light regimes.
An experiment was conducted to determine if survival during weaning was affected
by the size of the fry at the initiation of weaning. At the outset of the trials, feeding
of live zooplankton was discontinued and a commercially available salmon starter
crumble was dispensed by automatic feeders every hour for the 12 hours of daylight
(photoperiod 12L/12D). Four trials were undertaken using fish initially 12.8, 13.6,
16.7 and 19.6 mm TL. Survival through the 10-day weaning period averaged 39, 58,
97 and 92%, respectively. An asymptotic curve described the relationship between
initial size and survival, and indicated that survival of greater than 90% could be
expected with fry greater than 16 mm TL at the time of weaning. This is the size at
which barramundi fry change their feeding habit from that of a roving zooplanktivore
to a lurking predator. Cost-benefit analyses indicated a considerable economic
saving in delaying weaning until the fry are 16 mm TL.
A study was conducted to determine if hatchery-reared and wild barramundi could be
distinguished by the patterns of circulus spacing on the scales. Proprietary software
and digitising equipment was used to obtain measurements of circulus spacing

vi
within one millimetre of the focus of the scales. Discriminant analyses separated the
groups with up to 83% accuracy. As the technique utilises innate tags laid down in
response to the rearing environment, it has considerable potential for evaluating the
efficacy of large-scale enhancement programs. However, because scales from fish
larger than 350 mm TL were too thick and heavily pigmented to be reliably read, the
applicability of the technique with barramundi is limited to fish smaller than 350 mm
TL.
vi i
TABLE OF CONTENTS
Statement on Access to Thesis ii
ABSTRACT
TABLE OF CONTENTS viii
LIST OF ILLUSTRATIONS xii
LIST OF TABLES xv
DECLARATION .xviii
ACKNOWLEDGEMENTS xix
CHAPTER 1. General Introduction 1

1.1 Production of Fingerling Fish 1
1.1.1 Reasons for production 1
1.1.2 Production of fingerling fish in Australia 4
1.2 Production of Barramundi 14
1.2.1 History of barramundi aquaculture 14
1.2.2 Types of farming systems used in Australia 16
1.2.3 Barramundi production 20
1.3 Aims of the Study 22
CHAPTER 2. The Biology and Exploitation of Barramundi and the Proposal to

Introduce Nile Perch Lates niloticus to Australia 25
2.1 A Brief Review of the Biology and Exploitation of Barramundi and the
Effect of Stream Modification on its Local Distribution 25
2.1.1 Taxonomy 25
2.1.2 Distribution 28
2.1.3 Reproduction and associated movements 30
2.1.4 Capture fishery 35
2.1.5 Recreational fishery 36
2.1.6 Stream modification and its effect on the distribution of barramundi
viii
in Queensland 38
2.2 Evaluation of the Proposal to Introduce Nile Perch to Australia 41
2.2.1 The rationale for introducing Nile Perch 41
2.2.2 Temperature tolerance of L. niloticus, with reference to the species'
potential range if introduced into Australia 44
2.2.3 Impact of L. niloticus in Lake Victoria, Lake Kyoga and Lake
Nabugabo 57
2.2.4 Aspects of the biology of L. niloticus relevant to its proposed role
as a sport fish in Australia 66
2.2.5 Conclusions 68
CHAPTER 3. Feeding Habits of Hatchery-Reared Barramundi Fry 69

3.1 Introduction 69
3.2 Materials and Methods 70
3.2.1 Feeding behaviour 70
3.2.2 Diet 71
3.2.3 Diel feeding 72
3.2.4 Stomach evacuation rate 72
3.2.5 Daily ration 73
3.2.6 Morphometric relationships and growth rates 74
3.3 Results 75
3.3.1 Feeding behaviour 75
3.3.2 Diet 78
3.3.3 Diel feeding 81
3.3.4 Stomach evacuation rate 83
3.3.5 Daily ration 83
3.3.6 Morphometric relationships and growth rates 85
3.4 Discussion 90
CHAPTER 4. Predation by Dragonfly Nymphs of the species Pantala

flavescens (Odonata: Libellulidae), on Barramundi 97
4.1 Introduction 97
DC
4.2.1 Predation on fish by odonate nymphs 99
4.2.2 Colonisation of ponds by odonate nymphs 101
4.2.3 Comparison of husbandry techniques 102
4.3 Results 103
4.3.1 Predation on fish by odonate nymphs 103
4.3.2 Colonisation of ponds by odonate nymphs 107
4.3.3 Comparison of husbandry techniques 108
4.4 Discussion 112
CHAPTER 5. Effects of Photoperiod on Growth and Feeding Periodicity

of Barramundi Fry 119
5.1 Introduction 119
5.2.1 The fish 120
5.2.2 Effect of photoperiod on growth 121
5.2.3 Effect of photoperiod on feeding periodicity 123
5.3 Results 124
5.3.1 Effect of photoperiod on growth 124
5.3.2 Effect of photoperiod on feeding periodicity 126
5.4 Discussion 127
CHAPTER 6. Optimal Size for Weaning Barramundi Fry onto Artificial Diets... 133
6.2.1 Fish and facilities 134
6.2.2 Weaning procedure 135
6.2.3 Size and weaning success 136
6.3 Results 139
6.3.1 Weaning procedure 139
6.3.2 Size and weaning success 139
6.4 Discussion 144
CHAPTER 7. Use of Circulus Spacing on Scales to Discriminate Between
Hatchery and Wild Barramundi 148
7.2.1 Scale origin and preparation 149
7.2.2 Data acquisition 150
7.2.3 Circulus formation and fish length 151
7.2.4 Data analyses 152
7.3 Results ... 153
7.4 Discussion 157
CHAPTER 8. Synthesis 162

8.2 Nile Perch Studies (Chapter 2) 163
8.3 Rearing Juvenile Barramundi (Chapters 3-6) 166
8.4 Recreational Fisheries Enhancement with Barramundi (Chapters 1 & 7) 170
References 176
Appendices 206
Appendix 1. Economic comparison of weaning at 13 mm and
17 nun TL 206
Appendix 2. Published papers 209
X1

LIST OF ILLUSTRATIONS
Figure 1.1 Schematic representation of the various procedural steps involved 13

in fish production programs.
Figure 2.1 Distribution of Lates calcarifer 28
Figure 2.2 Historical distribution of barramundi in the Fitzroy River system 40
in Queensland.
Figure 2.3 Average monthly water temperature and average monthly mean air 51
temperature for 12 sites in the Murray-Darling River System.
Figure 2.4 Predicted water isotherms during July (mid-winter) in rivers of the 53
Murray-Darling River system.
Figure 2.5 Contours showing the percentage probability of average July (mid- 54
winter) temperatures exceeding 10°C in rivers in the Murray-
Darling River system.
Figure 2.6 Water temperatures recorded in the Clarence River at Lillydale 56
between 1971 and 1985.
Figure 3.1 Pigmentation patterns of barramundi fry. 77
Figure 3.2 Percentage composition by volume of zooplankton, insects and 80
vertebrates consumed by barramundi 10-56 mm TL reared in
freshwater ponds.
Figure 3.3a Average stomach fullness indices for pond-reared barramundi fry 82
15.8 + 2.5-mm TL.
Figure 3.3b Average stomach fullness indices for pond-reared barramundi fry 82
37.2 + 3.7-mm TL.
Figure 3.4 Stomach evacuation rates for barramundi fry 15.8 + 2.5-mm TL 84
and 37.2 + 3.7-mm. TL, for continuously feeding fish and non-
feeding fish.
Figure 3.5 Relationship between total length (mm) and standard length (mm) 86
for 129 barramundi in the range 10-56.2 mm total length.
Figure 3.6 Relationship between total length (mm) and wet weight (mg) for 87
135 barramundi in the range 11-87 mm total length.
xii
Figure 3.7 Relationship between standard length (mm) and wet weight (mg) for 87
61 barramundi in the range 8.5-46.9 mm standard length.
Figure 3.8 Relationship between total length (mm) and dry weight (mg) for 173 88
barramundi in the range 10-56.2 mm total length.
Figure 3.9 Relationship between standard length (mm) and dry weight (mg) for 88
Figure 3.10 Relationship between dry weight (mg) and wet weight (mg) for 62 89
barramundi in the range 3.4-507 mg dry weight.
Figure 3.11 Relationship between wet weight (mg) and dry weight (mg) for 62 89
barramundi in the range 22.9-2530 mg wet weight.
Figure 4.1 Mean number of mortalities of small barramundi (9.9 mm TL), 107
large barramundi (20.2 mm TL) and sooty grunter (18.7 mm TL)
exposed for 20 hours to predation by nymphs of the dragonfly
Pantala flavescens.
Figure 4.2 Nymph of the anisopteran dragonfly Pantala flavescens. 110
Figure 4.3 A. Density (mean number/m2 of substrate) of Pantala flavescens 111
nymphs in two 0.1 ha freshwater ponds in north-eastern
Queensland during the first 35 days after filling, as determined by
six 15 m tows with a dredge net 0.46 m wide on each sampling
occasion.
Total lengths (mm) of Pantala flavescens nymphs sampled in
freshwater ponds in north-eastern Queensland during the first 35
days after filling.
Growth in length of barramundi fry of two size groups (10 mm
TL and 20 mm TL at time of stocking) in freshwater ponds in
north-eastern Queensland.
Figure 5.1 Stomach fullness indexes for two size-groups of barramundi fry 128
exposed to 12L/12D (A) and 24L/OD (B) light regimes, with food
continuously available.

Figure 6.1 Experimental arrangement, showing the conical rearing containers, 137
automatic feed dispensers and electronic control boxes for each
dispenser.
Figure 6.2 Percentage survival through weaning onto dry diets for four 143
different size-groups of barramundi fry.
Figure 7.1 Diagrammatic representation of a barramundi scale. 152
Figure 7.2 Relationship between the number of circuli on scales and TL of 43 156
fingerling barramundi 25-38 days old.
xiv
LIST OF TABLES
Table 1.1 Native and introduced fish species produced in Australia for 7
recreational fisheries enhancement: tabulation of scientific and
common names, States in which stocked, and key references —
with comments — on production techniques.
Table 1.2 Fish species produced in Australia for conservation purposes: 11
tabulation of scientific and common names, and key references to
production techniques and management strategies.
Table 1.3 Aquacultural production (tonnes) of Lates calcarifer in 1986-92 20
in countries reporting to FAO.
Table 1.4 Production and value of barramundi from aquaculture and 21
capture fisheries in Australia in the financial years 1989-90 to
1995-96.
Table 2.1 Capture fishery landings (tonnes) of Lates calcarifer in 1986-92 36
in countries reporting to FAO.
Table 2.2 The stations and rivers at which air-water temperature 48
regressions were determined, and the number of water
temperature readings (n) taken during the period 1963-83.
Table 2.3 Values for constants a and b and their standard errors from linear 49
regressions AMWT + a + bAMMAT for 12 stations in the
Table 2.4 Percentage contributions of different fish species to the total 63
weight (tonnes) of fish landed from Kenyan waters of Lake
Victoria from 1970 to 1991.
Table 2.5 Percentage contributions of different fish species to the total 64
weight (tonnes) of fish landed from Tanzanian waters of Lake
Victoria between 1988 and 1992.
Table 3.1 Percentage frequency of occurrence (% FO) and percentage 79
composition by number (% CN) of food items in the diet of
barramundi Lates calcarifer fry reared in freshwater ponds.
xv
Table 3.2 Regression equations, intercepts (a), slopes (b) and r2 values for 86
length, wet weight and dry weight relationships for barramundi
Lates calcarifer fry (10.0-56.2 mm TL).
Table 4.1 Number of mortalities (mean + s.d.) of fish (Table. A) and 106
tadpoles (Table B) exposed for 20 hours to predation by nymphs
of the dragonfly Pantala flavescens.
Table 4.2 Number of mortalities (mean + s.d.) of tadpoles exposed for 20 106
hours to predation by nymphs of the dragonfly Pantala
flavescens.
Table 4.3 Mean number (± s.d.) of odonate nymphs sampled per 15 m tow 109
with a dredge net 0.46 m wide (determined from 6 tows per
sampling occasion) in two 0.1 ha freshwater ponds in north-
eastern Queensland during the first 35 days after filling.
Table 4.4 Number of barramundi stocked into six 0.1 ha freshwater ponds, 112
number harvested as 40-50 mm TL fingerlings, density at
harvest and percentage survival for fish of two sizes at time of
stocking.
Table 4.5 Guideline for stocking barramundi fry into freshwater ponds for 117
on-growing to fingerling size (40-50 mm TL).
Table 5.1 Mean lengths (TL, mm), weights (Wt, mg), and percentage 125
survivals (and standard errors) of barramundi fry after being
exposed to various photoperiod and food availability treatments
for 13 days.
Table 5.2 Mean lengths (TL, mm), weights (Wt, mg), and percentage 126
survivals (and standard errors) of barramundi fry after being
exposed for 13 days to photoperiod regimes of 12L/12D, 18L/6D
and 24L/OD, with food continuously available.
Table 5.3 Literature reports on the effect of extended light periods on 131
growth and survival of larvae and juveniles of several species of
fin fishes.
XVI
Table 6.1 Percentage of feeding and non-feeding barramundi, and final 140
lengths (TL, mm) and wet weights (W Wt, mg) of the feeders,
after 12 days exposure to either a gradual or a sudden transition
from frozen zooplankton to an artificial diet.
Table 6.2 Survival and mortalities due to starvation and cannibalism 142
(expressed as percentages) of barramundi fry of various sizes
during 10 day weaning trials.
Table 6.3 Initial and final total lengths (TL, mm) and computed initial and 142
final wet weights (WWt, mg) and daily growth rates (expressed
as a percentage based on weight) of barramundi fry of various
sizes during 10 day weaning trials.
Table 7.1 Percentage of wild and hatchery barramundi correctly identified 155
using data derived from the spacing of circuli on scales and linear
discriminant analysis.
Table 8.1 Tabulation of the contribution of the major elements of this thesis 174
to management, biology of barramundi, ecological theory, and
aspects of future research requirements.
xvii
STATEMENT OF SOURCES
DECLARATION
I declare that this thesis is my own work and has not been submitted in any form for
another degree or diploma at any university or other institution of tertiary education.
Information derived from the published or unpublished work of others has been
acknowledged in the text and a list of references is given.
Christopher G. Barlow 15 January 1998
XVII I
ACKNOWLEDGEMENTS
Many colleagues and friends have assisted me in various ways while I have been
undertaking this Ph.D. program. I wish to acknowledge and thank them here.
My colleagues at the Freshwater Fisheries and Aquaculture Centre, Walkamin. In

particular, I thank Les Rodgers for his excellent technical assistance, fish
husbandry skills and unfailing attention to detail. Others include Alf Hogan,
Clive Jones, Paul Clayton, Paul Palmer, Clare Longhurst, David Bull and Trevor
Marnock. Kim Hodgon, Sandra Lange and Dianne Crellin provided clerical
support.
My supervisor, Associate Professor Norm Milward for his guidance on the project
and thesis preparation, and considerable patience while awaiting submission of
the thesis.
Zena Seliga, the Fisheries Librarian within DPI, has provided a truly excellent
service during the course of the project, as she continually does for all Fisheries
staff within QDPI.
Barbara Gregg, Bill Rutledge and others within the Texas Parks and Wildlife
Department, who provided facilities and assistance to enable the study reported in
Chapter 7 to be undertaken.
Colleagues at the Northern Fisheries Centre, in particular Mike Rimmer, Rod

Garrett and John Russell, for their comments and advice on aspects of the work.
Joanne De Faveri, David Reed and Allan Lisle helped with statistical analyses.
Ian Ruscoe, Phil Hales and Joanne Grady prepared some of the figures.
Administrative backing for the project was provided by many people within
QDPI, but in this regard I particularly wish to thank Jim Gillespie and Bob
Pearson, who as managers have provided support and encouragement for the work
undertaken at the FFAC, Walkamin.
Financial support for the work described herein was variously provided by the
Fisheries section of QDPI, the Fishing Industry Research and Development
Council (Grant 89/67) and the Churchill Fellowship Memorial Trust Fund.
To the many industry colleagues (barramundi farmers, researchers and anglers) on

whom I drew for information and/or assistance, I express my sincere gratitude.
I particularly thank my family for their patience, encouragement and support in

many ways during the course of this study
xix
Chapter 1. Introduction 1
CHAPTER 1. GENERAL INTRODUCTION
1.1 PRODUCTION OF FINGERLING FISH
1.1.1 Reasons for production
Controlled breeding and production of fingerling fish is primarily undertaken for three
reasons:
to provide fish for farming, either for grow-out as food fish or for the
ornamental fish trade;
to provide stocking material for enhancement of recreational and/or
commercial fisheries; and
as part of management programs for endangered species.
The respective importance of these rationales varies with the species and to a lesser
extent with the country in which the activity is undertaken. Of the three, producing fish
for food is the most important activity on a world-wide basis, and has the longest
history. Pillay (1990) and Landau (1992) have reviewed the origins and growth of
aquaculture, and cited the earliest known report on the subject, namely 'Yang Yu
Ching', variously interpreted as the 'Classic of Fish Culture' or 'Treatise on Fish
Culture', written by Fan Lei, a Chinese politician-turned-fish-culturist around 500 BC.
Food production from fish farming is now practised in various forms on all inhabited
continents, although it is a more common and economically important activity in the

Asian region between India and Japan than elsewhere (Shepherd 1988). The most
important families are the Cyprinidae and Cichlidae (primarily `tilapias') in tropical
regions, and the Salmonidae in temperate areas.
Breeding and rearing ornamental fish, particularly goldfish Cyprinus carpio and carp
Cyprinus carpio, is an ancient practice in Japan and China. Nowadays the ornamental
(or 'aquarium') fish trade encompasses numerous species, and world-wide trade in fish
and accessories is believed to be worth more than US$7000 million per annum
(Andrews 1990). Of the estimate of at least 150 million fish traded each year, the great
majority are freshwater species, and around 90% of these are bred on farms, primarily
in South-East Asia, Japan, Israel and the USA (Andrews 1990).
Producing fish for enhancement of recreational fisheries started in Europe and North
America in the mid-19th Century (Landau 1992). Obviously the activity is confined to
highly sought-after angling species and to areas where recreational fishing is a major
activity. The most valued taxonomic group is the family Salmonidae, various species
of which have been translocated from Europe and North America into cool-water
recreational fisheries throughout the world. Native angling species are also produced in
different areas, for example the Percichthyidae (perches and cods) in Australia, the
Centrarchidae (basses and sunfishes) in North America, and the Esocidae (pikes) and
Percidae (perches) in both North America and Europe. An offshoot of recreational
fisheries is the production of forage and bait species (Brown and Gratzek 1980),
although this is negligible compared with the production of sport fish.

3
Chapter 1. Introduction
Enhancement of commercial fisheries for anadromous species has been undertaken in
Europe, Japan and North America since the mid-1800s and continues to the present day
(Moring 1986). The most important of these fishes are the salmonids, and others
include striped bass Morone saxatilis, clupeid fishes Alosa spp., and various sturgeons
Acipenser spp. Enhancement of commercial fisheries for marine species also started in
the 1800s, but these early efforts were constrained by the inability of fish culturists to
rear the larvae beyond the yolk-sac stage (Richards and Edwards 1986). Techniques
for successful production of marine species were developed in the late 1960s—early
1970s, and since then an increasing number of marine fishes has been reared for
aquaculture and restocking purposes (Sorgeloos et al. 1994). Enhancement of
commercial marine fisheries is now a common practice in the Scandinavian countries
(principally with cod Gadus morhua), the Mediterranean (with European sea bass
Dicentrarchus labrax, gilthead sea bream Sparus aurata, mullets Mugil spp., and
flatfishes Solea spp. and Scophthalmus spp.), Japan (with red sea bream or snapper
Pagrus (Chlysophtys) auratus, flatfish Limander spp. and Paralichthys spp., puffers
Fugu spp. and breams Acanthopagrus spp.), and the USA (primarily with red drum
Sciaenops ocellatus) (McCarty et al. 1986, Richards and Edwards 1986, Danielssen et
al. 1994). It is likely that enhancement of commercial fisheries will be increasingly
used as a fisheries management tool; the real challenge is in determining the efficacy of
such programs (Blankenship and Leber 1995).
Management programs entailing controlled breeding of endangered species is a very
recent activity. The aims of breeding programs with endangered species vary
depending on the degree of alteration of the native habitat. In some instances,
amelioration of those factors which caused the original decline may be possible, in
which case captive breeding is to provide fish for restocking in native habitats or
preserves (for example, desert fishes in north America (Pister 1990)). At the other
extreme, the changes may be irreversible; in such cases, breeding is simply to preserve
the gene pool (for example, cichlid species from Lake Victoria in east Africa (Bruton
1990)).
1.1.2 Production of fingerling fish in Australia
In Australia, fmgerling fish are produced for fisheries enhancement, farming, and
conservation purposes.
Traditionally, production for enhancement of recreational fisheries has been the major
activity. This started in the early 1860s when eyed ova of brown trout Salmo trutta
were brought to Tasmania (Cadwallader and Backhouse 1983). Other salmonids,
cyprinids and percids were later brought to Australia by fish acclimatisation societies,
the European members of which wanted familiar species for recreational fishing
purposes (Clements 1988). These and other introductions, their success in establishing
fisheries and some of their concomitant ecological consequences have been reviewed
by Weatherly and Lake (1967), Tilzey (1980), McKay (1984, 1989), Fletcher (1986),
Arthington (1989, 1991) and Pollard (1990).

Production of salmonids (principally brown trout and rainbow trout Oncorhynchus
mykiss) for recreational fisheries enhancement is still a major activity at Government
hatcheries in NSW, Victoria, Tasmania and Western Australia. Many millions of fish
are stocked annually into mountainous streams and large impoundments (see review of
stocking large reservoirs in Australia by Cadwallader (1983)). Stocking sizes vary
from eyed ova, which are placed in artificial redds (nests) in streams, to yearling fish
weighing approximately 500 g.
Experiments on the production of Australian native fishes began in the early 1900s.
H.C. Dannevig conducted experiments in 1905 on artificial production of Murray cod
Maccullochella peelii (Dakin and Kesteven 1938). T.L. Bancroft was rearing lungfish
Neoceratodus forsteri prior to 1914, and in the 1930s set up what was possibly the first
hatchery for production of an Australian fish (Bancroft 1914, 1933). However, it was
not until the 1960s that programs were initiated with the specific aim of producing
native species for fisheries enhancement. The first of these programs was at the Inland
Fisheries Research Station at Narrandera, New South Wales, which was developed
primarily to investigate breeding of native fishes in the Murray—Darling River system.
Part of the impetus for these investigations was the decline in native fish stocks in
rivers and the failure of many high-quality sport fishes to establish large populations in
the numerous impoundments which were being established for water storage. Breeding
of the principal sport fishes was quickly achieved (Lake 1967a, 1967b), but fry
production was problematical, and it was not until the late 1970s that large-scale
production of fry, and subsequently stocking of impoundments, became routine

(Rowland et al. 1983). Following the early work at Narrandera, breeding programs for
native fishes were established in Queensland and Victoria. At least 11 native species
and 5 salmonids have now been produced in Australia for recreational fishing purposes
(Table 1.1), although not all are currently being propagated.
Enhancement of commercial fisheries via hatchery releases has not been undertaken on
a commercial basis in Australia up to the present time, although preliminary
investigations into the supplementary stocking of prawn fisheries operating in relatively
confined estuarine waters are being made (P. Rothlisberg, CSIRO, Cleveland, pers.
comm. 1996).
Ornamental fish production in Australia at a commercial level is primarily confined to
goldfish. The few native species which are commercially traded (principally
rainbowfishes, family Melanotaeniidae) breed naturally in small ponds.
Aquacultural production of fish in Australia for food production began in the early
1960s with the development of rainbow trout farms in Victoria (Clements 1988),
followed by New South Wales and later Tasmania. In the early 1980s Atlantic salmon
Salmo salar were taken from the mainland to Tasmania, where production of fry was
commenced in a freshwater hatchery to supply smolts for a large marine-based farming
industry. Farming native species for food production started with barramundi Lates
calcarifer in Queensland in the mid-1980s (Pearson 1987). Silver perch Bidyanus

Table 1.1 Native and introduced fish species produced in Australia for recreational fisheries enhancement: tabulation of scientific and common
names, States in which stocked, and key references — with comments — on production techniques.
Scientific name Common name State in which Key references

(Family) stocked * Comments
Macquaria ambigua Golden perch Vic., NSW, Qld Rowland (1983a)
(Percichthyidae) * Spawning in captivity requires hormonal induction; larvae and fry
reared in fertilised freshwater ponds.
Rowland (1986)
Description of facilities and techniques for spawning and extensive
larval rearing.
Macquaria novemaculeata Australian bass NSW, Qld Battaglene et al. (1989)
(Percichthyidae) * Review of hormone-induced spawning, larval rearing and stocking in
the period 1979-86.
Battaglene and Talbot (1990)
High salinity, low light and low aeration required for high percentage
swim bladder inflation in intensive larval rearing.
Maccullochella peelii peelii Murray cod Vic., NSW, Qld Rowland (1983b); Cadwallader and Gooley (1985)
(Percichthyidae) * Spawns naturally in ponds; artificial substrates used to facilitate egg
retrieval; larvae and fry reared in freshwater ponds.
Rowland (1988)
Documentation of hormonal induction of spawning.
Gooley et al. (1995)
Maturation and spawning of a population in a small, confined lake.


Vic., NSW, Qld Rowland (1984)
Bidyanus bidyanus Silver perch
(Teraponidae) Spawning in captivity (in temperate areas) requires hormonal induction;
larvae and fry reared in fertilised freshwater ponds.
MacKinnon (1986)
Spawns naturally in ponds following rain in tropical areas.

Hephaestus fuliginosus Sooty grunter Qld MacKinnon (1986)
(Teraponidae) Production methods entail hormone induction and extensive larval
rearing in freshwater ponds.
Vic., NSW, Qld Merrick and Midgley (1981)
Tandanus tandanus Eel-tailed catfish
(Plotosidae) Description of spawning behaviours.
MacKinnon (1986)
Breeds naturally in ponds; currently not being produced for stocking.
Qld Merrick and Midgley (1982)
Oxyeleotris lineolatus Sleepy cod
(Gobiidae) Description of artificial spawning substrate.
MacKinnon (1986)
Production currently relies on collection of eggs, which are attached as a
mat onto substrates; fry produced in fertilized ponds.

Scleropages jardini Saratoga Qld MacKinnon (1986)
(Osteoglossidae) Breeds naturally in small ponds; parental female incubates eggs in her
mouth.


Scleropages leichardti Saratoga Merrick and Green (1982)

(Osteoglossidae) * Breeds naturally in small ponds; female incubates eggs in her mouth.
Lates calcarifer Barramundi Qld Copland and Grey (1987)

(Centropomidae) * Review papers on culture techniques throughout Indo-Pacific region;
larvae reared intensively in hatcheries.
Rutledge and Rimmer (1991)
* First report of extensive rearing using fertilised brackish water ponds.
Salmo trutta Brown trout NSW, Vic., Tas. Cadwallader and Backhouse (1983)
(Salmonidae) * Adults mature in captivity; gametes obtained by hand-stripping;
fertilised eggs incubated in troughs; larvae reared intensively on
artificial dry foods.
Oncorhynchus mykiss Rainbow trout NSW, Vic., Cadwallader and Backhouse (1983)
(Salmonidae) Tas., WA * As for S. trutta.
Oncorhynchus tshawytscha Chinook salmon Vic. Cadwallader and Backhouse (1983)

(Salmonidae) * As for S. trutta; released to land-locked lakes in Victoria.
Salvelinus fontinalis Brook trout NSW, Tas.

(Salmonidae)
Salmo salar Atlantic salmon NSW, Tas.
(Salmonidae)
bidyanus farming is now developing in New South Wales and Queensland (Rowland
1994). These two species are the only ones being produced commercially, although
there is limited experimental work being conducted on marine snapper Pagrus auratus
(Bell et al. 1991), striped trumpeter Latris lineata (Hutchinson 1993) and freshwater
Murray cod Maccullochella peelii (Gooley and Anderson 1992).
Artificial propagation programs have been developed as part of management strategies
for five threatened freshwater species in Australia, these being trout cod
Maccullochella macquariensis, eastern freshwater cod Maccullochella ikei, Mary River
cod Maccullochella peelii mariensis, Macquarie perch Macquaria australasica, and
Lake Eacham rainbowfish Melanotaenia eachamensis (Ingram et al. 1990). Hatchery
production of these species provides fry for restocking, safeguards stocks, and provides
opportunities for research into the biology of threatened fishes (Pollard et al. 1990).
Key references to production techniques and management strategies are listed in Table
1.2.
From the foregoing discussion it is apparent that the type of species and end use of
progeny from fish breeding programs varies enormously. However, many steps in the
production process are inherently similar and independent of the ultimate aim of the
breeding program. This can be seen in Figure 1.1, in which the various steps involved
in fish production are diagrammatically shown.

Table 1.2 Fish species produced in Australia for conservation purposes: tabulation of scientific and common names, and key references to
production techniques and management strategies.

Scientific name (Family) Common name Key references
* Comments
Maccullochella Trout cod Ingram and Rimmer (1992)

macquariensis * Fish mature in ponds, but hormonal induction necessary for spawning; egg
(Percichthyidae) incubation and larval rearing as for M peelii.
Douglas et al. (1995)
* Hybridisation between trout cod and Murray cod in natural populations in the Murray
River.
Macullochella ikei Eastern freshwater Rowland (1993)
(Percichthyidae) cod * Bred for restocking native habitat.
Maccullochella peelii Mary River cod
mariensis
(Percichthyidae)

Scientific name (Family) Common name Key references
* Comments

Macquaria australasica Macquarie perch Ingram et al. (1994)

(Percichthyidae) Breeding protocols have been investigated at NSW and Victorian
Government hatcheries; induced spawning of captive fish generally unsuccessful.

et al. (1990); Moritz et al. (1995).
Melanotaenia eachamensis Lake Eacham Barlow et al. (1987); Ingram
(Melanotaeniidae) rainbowfish Disappearance from Lake Eacham, Queensland, associated with translocation
of other Australian native fishes into the lake; stocks being maintained in ponds at
Walkamin Research Station; re-introduction to Lake Eacham has not been successful.
Taxonomic status of this species and congenerics under review.
Access breeders
on spawning grounds Collection of broodstock
Broodstock maturation
Spawning — 3 methods
Manual Natural spawning Induced with exogenous Induced with exogenous
and fertilisation of hormones, followed by hormones, followed by
stripping and gametes natural spawning and manual stripping and
fertilisation of fertilisation of gametes fertilisation of gametes
gametes
Egg incubation
Larval an --production
- —*yr -
xtetisive; ggfertifiTairchrtanOr -:4;14ny, Intensive, in controlled
organitlikaye osel asPeets.ofiii4e two conditions; food organisms
e -treines cultured separately from larval
rearing tanks
ws....fitata,r.'',Asi
:ass
STO—alrellritkv—gious Endangered species
sizes:ta enhance: Food trade on- Aquarium trade retained as source of
recreational or grown to marketable genetic material or
commercial fisheries size stocked into
rehabilitated areas
Figure 1.1 Schematic representation of the various procedural steps involved in

fish production programs. The shaded section highlights the area of
research covered within the present study.
1.2 PRODUCTION OF BARRAMUNDI
1.2.1 History of barramundi aquaculture
There are early reports of farming of barramundi using wild-caught fingerlings in India
(Menon 1948), and pond and cage rearing of wild progeny were traditionally practised
in other south-east Asian countries. However, aquaculture of the species was limited
by the difficulty and unreliability of obtaining fry from natural waters. In an attempt to
solve this problem, the Thai Government commenced research in the early 1970s on
artificial propagation of barramundi at the Songkhla Marine Fisheries Station.
Hatchery production of fingerlings was achieved for the first time in 1973 using eggs
stripped from wild spawners (Wongsomnuk and Manevonk 1973). By 1975, the
station had successfully produced the first batch of fry from eggs obtained from
induced breeding of captive broodstock (Sirikul 1982). The supply of hatchery-
produced fry facilitated large-scale farming in Thailand, which, coupled with
international training programs offered by the Thais, provided the impetus for other
countries in the region to investigate hatchery production of barramundi. Within a
decade, hatchery production was achieved in the Philippines (Harvey et al. 1985),
Taiwan (Lin et al. 1985), Singapore (Lim et al. 1986) and Malaysia (Ali 1987).
Barramundi now forms the basis of well-developed farming operations in these
countries.
In Australia, detailed work on barramundi culture commenced in 1983, in two separate
projects. The first was a project on development of barramundi hatchery and farming
techniques, funded by the Fishing Industry Research Trust Account (FIRTA Project
83/38). This project was conducted in the Cairns—Innisfail region in north Queensland,
and led to the formation in 1985 of the publicly listed aquaculture company 'Sea
Hatcheries'.
The second was a program conducted by the Queensland Department of Primary
Industries (QDPI) at its Cairns and Walkamin laboratories. The initial aims of the
work were to investigate controlled breeding and production of fingerlings for stocking
freshwater impoundments in north Queensland. The program was given high priority
within QDPI because, if successful, the use of barramundi for enhancing recreational
fisheries would provide an alternative to the exotic Nile perch, which was concurrently
being investigated with a view to importation and stocking in northern Australian
waters.
Both the FIRTA and QDPI projects were successful in establishing controlled breeding
of barramundi (Heasman et al. 1985, MacKinnon 1987a), and the resultant supply of
fingerlings led to the establishment of grow-out farms in northern Queensland in the
late 1980s. About the same time, several private hatcheries commenced operation.
The capability of the industry to produce fingerling barramundi was dramatically
improved with the development of pond larval-rearing techniques (Rimmer and
Rutledge 1991), and with the sale of fertilised eggs by QDPI which commenced in
1990. These advances allowed barramundi farmers to produce their own fingerlings,
rather than relying on production by the larger hatcheries. In turn, this has resulted in
increased output of product for market.
Outside Queensland, the Northern Territory Department of Primary Industry and
Fisheries started a program on barramundi aquaculture in 1987. In 1993, six grow-out
farms were licensed in the Northern Territory. In South Australia, West Beach
Aquaculture Pty Ltd has established a hatchery operation, which provides fingerlings,
expertise and facilities to a network of grow-out farms (in 1998, these were located in
New South Wales and South Australia). There is also one farm in South Australia
producing barramundi in geothermal water; it uses fingerlings from farms in north
Queensland.
1.2.2 Types of farming systems used in Australia
Broodstock maintenance
Early work on culture techniques for barramundi relied heavily on obtaining fertilised
eggs by stripping running-ripe male and female barramundi caught on spawning
grounds. This approach is expensive, unreliable and can potentially be in conflict with
commercial fisheries' interests. It has now largely been replaced by the development
of controlled breeding techniques for captive broodstock. Fertilised eggs are obtained
from captive broodstock in two ways (Palmer et al. 1993). The most reliable and
commonly used method is hormone-induced tank spawning, in which the fish spawn
naturally following the injection of reproductive hormones (Garrett and Connell 1991).
The other is hormonal induction of broodstock followed by manual stripping of semen

and ova soon after ovulation. The semen and ova are manually mixed to enable
fertilisation. This method is not preferred because of the labour requirement, the
difficulties in predicting the time of ovulation, and the fact that fertilisation rates are
generally not as good as with natural spawning.
Purpose-built systems are required for broodstock maintenance. The fish are usually
held indoors, in either flow-through or recirculating systems. Barramundi broodfish
may be kept in either fresh or salt water but must be placed in salt water prior to the
breeding season to enable gonads to fully mature. At the Northern Fisheries Centre,
Cairns, barramundi have recently been bred throughout the year using controlled
environment systems to manipulate temperature—photoperiod cycles, thus enabling
spawning outside the normal spawning season (Garrett and O'Brien 1994).
Larval—fingerling rearing
Barramundi larvae have a physiological requirement for salt water up to about 8-10
days old, but thereafter they have the ability to survive in both salt and fresh water
(MacKinnon 1987a). Consequently, larval production systems can be based entirely on
salt water, or on salt water followed by fresh water after about 10 days.
Production techniques are broadly classified as either 'intensive' or 'extensive'.
Intensive larval rearing involves the culture of larvae in a controlled environment, such
as a hatchery, where the fish larvae are supplied with prey organisms which are also
cultured under controlled conditions (Ruangpanit 1987). The intensive system requires
dedicated facilities and a high degree of technical skill. In contrast, extensive larval
rearing involves the culture of larvae in a largely uncontrolled environment (a pond)
and the culturist has relatively little direct control over factors such as water quality,
prey organism density and disease (Rimmer and Rutledge 1991, Rutledge and Rimmer
1991). The intensive technique also requires a higher labour input than the extensive
technique.
In South-East Asia, barramundi larvae are reared intensively and it was these larval
rearing techniques which were introduced to Australia, where they are still used in a
few hatcheries. The QDPI barramundi rearing project originally used the intensive
technique for early larval production, but modified it to incorporate a freshwater pond
rearing phase for larvae older than 15-20 days. In the early 1990s the intensive system
was dropped in favour of extensive rearing in brackishwater ponds for all larval and
early juvenile stages. The majority of barramundi fry now produced in northern
Australia is produced using extensive larval rearing procedures. More recently,
barramundi farmers have begun using the 'green water' culture technique, which
entails the phytoplankton, zooplankton and larval fish being grown together in large
(5-20 tonne) tanks.
Grow-out systems
There are three quite different methods currently used in Australia for growing weaned
fingerlings to market size. One is cage culture in estuarine waters or marine waters,
which has considerable advantages over other systems, particularly where large-scale
production (several hundred tonnes or more per annum) is envisaged. There are,
however, problems with biofouling and, to a lesser extent, predators. Consequently,
relatively few companies are using this technique.
The most common grow-out system currently utilised in Australia is pond culture, in
either brackish or fresh water. Fish are usually maintained in cages, although recently
cage culture of fish less than 120-150 mm total length and free-ranging for larger fish
have been sometimes combined. Pond rearing of free-ranging fish does not require the
labour associated with cage culture, and produces fish with a better appearance and
colour (silver rather than dark grey to black). The major disadvantages of these
methods are difficulties in stock management and harvesting.
The third method of farming barramundi is intensive production in an indoors,
controlled environment building, using underground water (i.e. pathogen free) and a
high level of recirculation through biological filters. This facility and technology have
been developed and patented by a South Australian company (West Beach Aquaculture
Pty Ltd), which has established -plants in South Australia, New South Wales and
Queensland. The facility is generic, in that it is suitable for any species which can be
intensively farmed. Because of the controlled environment, it allows for year-round
production virtually anywhere that underground water is available, regardless of the
temperature requirements of the species. By the mid 1990s, many small recirculation
units had been established in southern Queensland for production of barramundi.

1.2.3 Barramundi production
Aquacultural production of barramundi in South-East Asia is presently more than
10 000 tonnes per annum (Table 1.3) (FAO 1994a).
Table 1.3 Aquacultural production (tonnes) of Lates cakarifer in 1986-92 in

countries reporting to FAO (FAO 1994a). Data for the Philippines in
1991 and 1992 are from Ferdouse (1995); she also lists 'other Asia' as
Taiwan.
COUNTRY 1986 1987 1988 1989 1990 1991 1992

Australia 2 1 22 18 33 94 145
Brunei Darus 1 1 0 0 0 8
Hong Kong 105 34 23 170 167 224 187
Indonesia 798 1 384 1 356 2 645 2 267 1 539 1 670
Malaysia 819 1 067 1 267 3 999 1 608 1 954 2 784
Other Asia 2 708 4 673 4 674 6 063
Philippines 779 698 636
Singapore 203 219 235 201 307 239 393
Thailand 823 1 183 1 034 1 290 1 005 1 650 1 600
TOTAL 2 750 3 889 3 938 11 031 10 839 11 072 13 486
In Australia, production of barramundi from aquaculture commenced in 1986, with 2
tonnes of fish coming from Sea Hatcheries' farm in that year. Since then, farms have
been developed in the Northern Territory, South Australia and New South Wales.
There is also one pilot-scale operation in Lake Argyle, Western Australia. Production
has increased significantly since then, with approximately 460 tonnes (live weight)
being sold from farms Australia-wide in 1995-96 (Table 1.4). The value of the
barramundi aquaculture industry now equals that of the capture fishery (Table 1.4).

Table 1.4 Production and value of barramundi from aquaculture and capture fisheries in Australia in the financial years 1989-90 to 1996-97.
Production figures are tonnes live weight (t).
1989-90 1990-91 1991-92 1992-93 1993-94 1994-95 1995-96 1996-97
FISHERY STATE t $'000s t $'000s t $'000s t $'000s t $'000s t $'000s t $'000s t $'000
AQUACUL- Qlda 37 429 103 1100 152 1538 232 2313 248 2419 200 2192 328 3332 350 3439
TURE
NV' 0 0 0 0 5 51 10 100 50 500 100 1000 20 200 10 100
SA/NSW 0 0 2 22 5 55 10 110 40 420 55 633 116 1469 150 1762
TOTAL 37 429 105 1122 162 1644 252 2523 338 3339 355 3825 464 5001 510 5301
CAPTURE Qldd 733 5864 783 3913 471 2035 582 3055 478 2870 520 2705 667 2567 601 2103
FISHERY
N'Td 550 2052 459 1819 457 1689 451 2073 472 2111 482 2263 552 2956 577 2839
Wac 56 346f 61 282f 46 212f 46 228` 43 225f 40 197f 39 280 46 319
TOTAL 1 1339 8262 1 1303 6014 1974 3936 11079 5356 1993 5206 11042 5165 11258 5803 11224 5261
Information sources:
Aquaculture Production Surveys 1989-90, 1990-91, 1991-92. Reports published by Queensland Department of Primary Industries.
Personal communication from Dr. C. Shelley, N.T. Department of Primary Industries and Fisheries. The data are estimates.
Personal communication from Dr. J. Trendall, West Beach Aquaculture Pty Ltd; and L. Gray, SA Fisheries.
Australian Fisheries Statistics 1992, 1995 and 1998. ABARE, Canberra. (The value of Queensland production reported in Aust. Fish. Stat. 1995 and
1998 were twice the actual value, due to confusion between the value of fillet and the reporting of tonnes live weight).
Personal communication from Ms H. Brayford, W.A. Fisheries Department
Estimated from value of QLD and NT product.
1.3 AIMS OF THE STUDY
The overall goal of this study was to investigate the biology of juvenile barramundi to
determine effective means of producing the species for recreational fisheries
enhancement and farming purposes. Several aims were identified, which in effect
comprise Chapters 2-7 of the thesis.
Ch. 2 The aim of this section was i) to review the sport fishery for barramundi and
the effect of catadromy and stream modification on the species' distribution;
and ii) to analyse the proposal to introduce the exotic Nile perch Lates niloticus
for establishment of sport fisheries in northern Australia. In the context of
developing sport fisheries in impoundments through the stocking of
barramundi, it is helpful to appreciate why the alternative proposal of
introducing Nile perch was not, in an ecological sense, soundly based.
Ch. 3 At the commencement of this project, there was no information available on
pond-rearing of juvenile barramundi. Consequently, the feeding habits of
juvenile barramundi reared in a hatchery and pond complex were studied, with
the aim of developing guidelines on husbandry techniques to maximise survival
and growth ofjuvenile barramundi reared in fresh water.
Ch. 4 Predation (particularly by dragonfly nymphs) is a serious problem in the rearing
of fish larvae and fry in freshwater ponds. Experiments were undertaken i) to

determine if barramundi fry are susceptible to predation by dragonfly nymphs;
ii) to compare predator avoidance responses of barramundi fry and similar
sized sooty grunter, Hephaestus fuliginosus (a freshwater breeding species);
and iii) to determine procedures to minimise predation by dragonfly nymphs on
juvenile barramundi in ponds.
Ch. 5 Barramundi are visual feeders and as such it was hypothesised that extended
periods of light may influence feeding and growth patterns. Consequently an
experiment was conducted to determine growth, survival and feeding
periodicity ofjuvenile barramundi reared in extended light regimes.
Ch. 6 Weaning juvenile fish from live (or natural) food onto dry, manufactured diets
is a critical phase in an aquacultural production process. Size is an important
biological factor influencing a fish's amenability to weaning. Literature reports
indicate that barramundi fry are generally weaned at about 10 mm TL
(effectively at the time of metamorphosis), but with considerable associated
mortality. Thus an experiment was performed to determine the optimum size
for weaning barramundi onto formulated feed.
Ch. 7 The ability to recognise hatchery fish after stocking into open systems is a
fundamental requirement for determining the efficacy of enhancement
programs. Fingerling fish, because of their small size, are particularly hard to
tag, and traditional methods are subject to error (tag loss and mortality) and are
labour intensive. Studies overseas have shown that growth patterns laid down
on scales can be a function of the rearing environment. A study was undertaken,
using facilities provided by the Marine Fish Hatchery at Corpus Christi, Texas,
USA, with the aim of determining if patterns of circulus spacing on
barramundi scales could be used to distinguish hatchery-reared and wild fish.

Chapter 2. Barramundi biology and the Nile Perch proposal 25
CHAPTER 2. THE BIOLOGY AND EXPLOITATION OF BARRAMUNDI
AND THE PROPOSAL TO INTRODUCE NILE PERCH Lates
niloticus TO AUSTRALIA
2.1 A BRIEF REVIEW OF THE BIOLOGY AND EXPLOITATION
OF BARRAMUNDI AND THE EFFECT OF STREAM
MODIFICATION ON ITS LOCAL DISTRIBUTION.
2.1.1 Taxonomy
The barramundi was originally described by Bloch (1790) using specimens collected
from the Indo-Pacific region. The exact location of collection of the type specimen is
not known. Bloch (1790) listed it as Japan, but Cuvier and Valenciennes (1828)
considered that Bloch was mistaken and that it was Java. They based this on the facts
that the specimen was collected by Dutch traders (whose centre of operations was the
Indonesian Archipelago), and that Bloch had mistakenly listed Japan as the locality for
several other Javanese species.
Bloch (1790) named the species Holocentrus calcarifer (`calcarifer' meaning 'thorn
carrier', a reference to the spines on the operculum and pre-operculum). The genus
Lates was erected by Cuvier and Valenciennes in 1828 to encompass barramundi and
several related African species. (According to Cuvier and Valenciennes (1828), Tates'
was the name used by 'des anciens' for the Nile perch, Lates niloticus, and it was still
in use in parts of Egypt up to the 17th Century). Congeners of barramundi are seven
African species (Greenwood 1976) and the akame Lates japonicus from southern Japan
(Katayama and Taki 1984). Lates has at various times been placed in several families
in the Order Perciformes (see listing by Grey 1987), and is now in the family
Centropomidae. Thus, the current taxonomic status of the barramundi is as follows.
Phylum Chordata
Subphylum Vertebrata
Class Osteichthyes
Subclass Actinopterygii
Order Perciformes
Suborder Percoidei
Family Centropomidae
Genus Lates
Species calcarifer (Bloch, 1790)
Lates calcarifer is known by a variety of common names throughout its range. In
Australia the accepted common name is barramundi (derived from an aboriginal word
'burramundi', meaning large scales), but in the past it has also been called palmer, cock-
up and giant perch (Grant 1985). Throughout South-East Asia the accepted common
name in English is sea bass (or Asian sea bass). As expected for a fish with such a
wide distribution, there are numerous local names for the species, such as `bhekti' in
India (James and Marichamy 1987), `ka-ka-dif in Burma (Htin 1987) and 'pia kapong
kao' in Thailand (Ruangpanit 1987).
The taxonomic description of Lates calcarifer from FAO (1974) is given below. It
should be noted that the colour description is incomplete as it applies only to subadults
and adults. Juveniles up to approximately 100 mm TL exhibit various patterns of
black/dark-grey/silver (see section 3.3.1. of this thesis; also Russell (1987)).
Description. Body elongate, compressed, with a deep caudal peduncle. Head pointed,
with concave dorsal profile becoming convex in front of dorsal fin. Mouth large,
slightly oblique, upper jaw reaching to behind eye; teeth villiform, no canines present.
Lower edge of pre-operculum with a strong spine; operculum with a small spine and
with a serrated flap above origin of lateral line. Dorsal fin with 7-9 spines and 10-11
soft rays; a very deep notch almost dividing spiny from soft part of fin; pectoral fin
short and rounded, several short, strong serrations above its base; dorsal and anal fins
both have scaly sheaths; anal fm rounded. Scales large, ctenoid (rough to touch).
Colour. Two phases, either olive-brown above with silver sides and belly (usually
juveniles) or green-blue above and silver below. No spots or bars present on fins or
body.
2.1.2 Distribution
Lates calcarifer is widely distributed in rivers and coastal areas throughout the tropical
and subtropical Indo-West Pacific region (Fig. 2.1). Its western limit is the Persian Gulf.
To the north it ranges to Xiamen (24° 30'N) in southern China. Its eastern and southern
limits are the eastern tip of Papua New Guinea and the east coast of Australia south to the
Noosa River (26° 30'S) (Dunstan 1962, Greenwood 1976).
Figure 2.1 Distribution of Lates calcarifer .
In Australia, the barramundi is found in all the coastal waters and river systems north of
the Noosa River on the east coast and the Ashburton River (22° 30'S) on the west coast
(Morrissy 1985). It has also been recorded from the Abrolhos Islands (28° 40'S) off the
coast of Western Australia (Whitley 1959), but this record results from misidentification
of the similar-looking sand bass, Psammoperca waigiensis (Morrissy 1987). It is rarely
collected at any distance from the coast, and is generally found only in waters subject to
freshwater influence (Grey 1987).
In Papua New Guinea and Irian Jaya Lates calcarifer is found only on the southern side
of the island. Moore (1980) proposed that its distribution in PNG is linked to the
occurrence of river systems with associated large deltaic swamps combined with large
tidal variations. On the southern (Papuan) side of the island the abundance of L.
calcarifer is greatest around the Fly River, form where it decreases in numbers to both
the east and west. Its absence from the northern (New Guinea) side of the island is
intriguing, as the Sepik River provides habitat very similar to the Fly River. Reynolds
(1978) referred to three possible reasons for L. calcarifer not being found there. The
first, postulated by Dunstan (1961), was the small tidal rise in this area (0.6 m as
compared with 2.5 m at the mouth of the Fly River). The second, suggested by Rapson
(1968), was that the south equatorial current sweeping the north coast would remove
eggs and larvae from the breeding area. The third, pointed out by Reynolds (1978),
was that there is very little continental shelf at the mouth of the Sepik, with a sharp
drop off within 800 m of the shore. He concluded that further information was required
before any firm explanation could be given for the absence of barramundi from New
Guinea.
The limits to the distribution of barramundi have been reported by Grey (1987). He
considered that the lack of large rivers and fresh water to the west and the lack of
continental land masses (and thus rivers) to the east would limit further distribution
past these extremes. Latitudinal distribution is most likely controlled by temperature
(note that southern China and southern Queensland are at similar latitude (24-26° N
and S, respectively)). At the southern extreme of the barramundi's range in
Queensland it is effectively dormant through winter at which time riverine water
temperatures drop to 15-16°C. Captive broodstock held at these temperatures exhibit
stress-related mortalities. Breeding during summer does not start until January in
southern Queensland, when water temperatures reach 27-28°C. The seasonal
temperature cycle is such that 0+ fish would have a very short growing season prior to
the cessation of growth due to low temperatures (less than about 20°C) (J. Burke,
Bribie Island Aquaculture Research Centre, pers. comm. 1996).
2.1.3 Reproduction and associated movements
Lates calcarifer in Australia and Papua New Guinea is a protandrous hermaphrodite.
Individuals start life as a male, and after spawning one or more times, become female
(Moore 1979, 1982; Davis 1982). Thus, the occurrence of females becomes more
common with increasing size. Sex inversion takes place rapidly (probably in less than
two months) immediately after spawning (Moore 1979). The size and age at which this
occurs varies markedly between populations. For instance, stocks of L. calcarifer in
Papua New Guinea and areas of the Northern Territory and southern Gulf of
Carpentaria have a 1:1 sex ratio at about 80-100 cm total length (6-7 years old)
(Moore 1979; Davis 1982, 1984). However, there have been precocious populations
identified from northern Cape York Peninsula which mature much earlier (functional
males at 1-2 years versus 3-5 years in 'normal' populations), and in which a 1:1 sex
ratio occurs at about 45 cm total length (3 years old) (Davis 1984).
Primary females (that is, fish which mature initially as females and do not undergo sex
reversal) are sometimes found, but their proportion in populations studied to date is
very low (apparently less than about 1%) (Moore 1979, Davis 1985).
Researchers in the Philippines (Parazo et al. 1990) and Malaysia (Ali 1987) consider
that stocks of L. calcarifer in those countries change sex, but there are no records of sex
inversion in other areas of the fish's distribution. Detecting sex change in wild
populations is very difficult because of the rapidity with which barramundi undergo sex
inversion and the difficulty in detecting transitional fish macroscopically (Moore
1979). Nevertheless, the protandrous nature of the barramundi in Australian
populations is very evident from the disparity in size between males and females. It is
noteworthy that researchers in the Songkhla Lakes region in Thailand, where wild and
captive populations of L. calcarifer have been studied in great detail since the early
1970s, do not consider it to be hermaphroditic (S. Maneewong, Songkhla Marine
Fisheries Station, Thailand, pers. comm. 1983). Thus, while it appears that most L.
calcarifer populations are protandrous hermaphrodites, some populations may not
undergo sex reversal.
Barramundi is generally classed as catadromous; that is, a species which spends most
of its life in freshwater and migrates to the sea to breed (Myers 1949, cited in
McDowall 1987). However, movement and migration associated with spawning is a
particularly complex process in barramundi, and shows considerable regional variation.
In Australia, colonisation of upstream freshwater reaches of rivers is by 0+ fish, which
leave estuarine nurseries when only a few months old (Dunstan 1959, Davis 1985,
1987, Russell and Garrett 1985, Griffin 1987). The fish mature in freshwaters, then
travel downstream at first maturity to spawn in estuarine or coastal waters adjacent to
the rivers from which the fish emigrated (Shaklee and Salini 1985). It seems that most
adult fish then remain in estuarine waters or lower river reaches, with recolonisation of
the upper riverine reaches again being by 0+ juveniles (Davis 1987, Griffm 1987),
although there is a small percentage of fish which apparently spend their entire lives in
estuarine or coastal waters.
In contrast, in Papua New Guinea maturing barramundi move out of inland rivers and
swamps and then undergo extensive coastal migrations to specific spawning sites. The
coastal migration is apparently a consequence of inadequate salinities for spawning and
larval recruitment at the mouths of many river systems. After spawning, most adults
leave the coastal waters and migrate back to the inland swamps and rivers from which
they originated, although there are some adult fish which remain in coastal waters.
Juvenile fish remain in coastal nursery swamps until about six months old, when the
drying swamps force them to leave and distribute along coastal and estuarine regions.
They later colonise the inland waters as 1+ or 2+ fish (Moore 1982, Moore and
Reynolds 1982).
Chapter 2. Barrarnundi biology and the Nile Perch proposal 33
Spawning takes place in areas of high salinity (27-36 ppt) close to nursery grounds
(Garrett 1987). Generally the spawning sites are in sheltered waters near river mouths,
although barramundi will breed in the upper reaches of large estuarine systems if the
salinities are suitable (Garrett et al. 1987). The spawning season in Australia is linked
to seasonal temperature cycles, and only takes place when water temperatures exceed
27°C (Garrett 1987). This coincides with the summer monsoons in Australia, when
there is maximum availability of nursery swamp habitat.
In equatorial regions the spawning season for barramundi is considerably extended; for
instance in Thailand it breeds all year round (Ruangpanit 1987) and in India it breeds
for 8 months of the year (James and Marichamy 1987).
Movement of young barramundi into freshwaters is an integral part of their life history.
Juvenile fish have been recorded up to 800 km upstream in the Fly River in Papua
New Guinea (Moore 1980), and up to 130 km upstream in India (James and
Marichamy 1987). In Australia it is regularly found more than 100 km upstream from
the mouths of rivers traversing flat country.
Natural or artificial structures across streams present barriers to fish migration. The
capacity of percoid fishes to negotiate the barriers depends largely on the swimming
ability of the particular species. Hogan and Graham (1994) documented flood gates on
streams preventing barramundi moving upstream in the Herbert River in Queensland.

Kowarsky and Ross (1981) and Russell (1991) have studied two tidal barrages in
central Queensland that effectively stopped upstream migration of barramundi (even
though both were equipped with fish ladders). Research on swimming ability has
shown that under experimental conditions burst swimming speeds of barramundi are
considerably less than those exhibited by Australian bass M novemaculeata (Mallen-
Cooper 1989), another catadromous species from eastern Australia with similar body
shape and ecology to barramundi. Field observations tend to confirm this finding, as
Australian bass is found at higher altitudes and over steeper stream inclinations than is
barramundi. Similarly, in tropical eastern Australia the catadromous jungle perch
Kuhlia rupestris ascends streams to considerably greater altitudes than barramundi.
The comparatively poor swimming ability of barramundi was further confirmed by
Clague (1991), who showed that under laboratory conditions small barramundi (30-70
mm total length) exhibited slower burst and sustained swimming speeds than did
similar-sized sooty grunter Hephaestus fuliginosus.
Obviously, the barramundi is not a strong swimmer, and even minor barriers across
streams will prevent the species from moving further upstream.
From this brief review it is apparent that the reproductive biology of L. calcarifer varies
geographically. This flexibility has enabled it to colonise a range of habitats within an
extremely wide natural distribution. Despite this flexibility, there are several key
factors which control the occurrence of barramundi within its natural range. These are:
the presence of major freshwater influence (barramundi is not found on small
islands without major river systems);
high salinity for spawning;
nursery swamps for recruitment; and
unimpeded passage for upstream colonisation of river systems.
2.1.4 Capture fishery
Barramundi is a highly esteemed fish and consequently is exploited by commercial
fishermen throughout its range. FAO (1994b) lists that the capture fishery for the
species worldwide was 34 479 tonnes in 1992, but the actual catch was presumably
considerably more, as no data were provided for Taiwan, China, Vietnam, Kampuchea,
Thailand, Singapore, Burma, India and Sri Lanka (Table 2.1). Its relative importance
in regional fisheries in unclear. The only pertinent reports are those of Sodikin (1987)
who considered barramundi an important marine species in Indonesia, accounting for
about 1% by weight of the commercial marine catch; Kasim and Jones (1987) who
reported that in certain estuarine fisheries in India it comprises 4-6% of the catch.
The statistics for the capture fishery in Australia indicate landings of approximately
1000 tonnes annually (Table 1.4). Production is static (or slowly declining) due to
stringent controls aimed at preventing over-exploitation and, in the case of the Northern
Territory, redistribution of the resource between the commercial and the recreational
fisheries (Lea et al. 1987, Anon. 1991). The number of fishers involved is difficult to
determine as not all licence holders are active, and in many cases barramundi fishing
provides only a minor part of their income. Nevertheless, in 1996-97 licence
endorsements for barramundi fishing totalled 386 in Queensland (Queensland Fish
Management Authority, pers. comm. 1996), 28 in the Northern Territory (Roland
Griffm, N.T. Department of Primary Industries and Fisheries, pers. comm. 1996) and
14 in Western Australia (Tim Bray, WA Fisheries, pers. comm. 1996).
Table 2.1 Capture fishery landings (tonnes) of Lates calcarifer in 1986-92 in

countries reporting to FAO (FAO 1994). It is possible that some of
these figures (e.g. Singapore, Hong Kong) include aquaculture product;
cf. Table 1.3.
Country 1986 1987 1988 1989 1990 1991 1992
Pakistan 605 500 420 325 312 250 237
Malaysia 944 1 067 1 267 3 999 1 608 1 954 2 010
Singapore 203 219 235 230 345 274 424
Indonesia 17 443 18 872 19 029 26 746 27 503 24 059 30 160
Australia 1 034 1 320 1 385 1 295 1 417 1 372 1 051
Papua New Guinea 442 473 331 526 520 400 410
Hong Kong 105 34 23 170 167 224 187
Total 20 776 22 485 22 690 33 291 31 872 28 533 34 479
2.1.5 Recreational fishery
The recreational fishery for barramundi in Australia is considered an important fishery
in terms of harvest of fish, participation rate and economic activity. Although this is
widely accepted, quantitative data on the extent of the fishery are limited because the
collection of reliable information is inherently difficult due to the dispersed effort in the
fishery.
Most surveys on recreational fishing for barramundi have been carried out in the
Northern Territory. Griffin (1982, 1988) estimated that recreational fishermen
accounted for 23% and 31% of the total barramundi catch landed in the eastern part of
the Northern Territory in 1979-80 and 1986 respectively. Commercial fishing in the
Mary River (the principal fishing site in the area) was totally banned in 1988, as the
Government considered that on an economic and social basis the resource was more
productive if exploited by the recreational fishery. Participation in the Mary River
fishery was estimated at 17 000 and 11 000 angler days in 1991 and 1992 respectively,
but this was only about half the participation level in 1986 (Griffin 1993).
There is no direct assessment of the extent of the recreational fishery for barramundi in
Queensland. However, Rutledge et al. (1990) estimated the recreational harvest to be
in the range of 50 000-100 000 fish, and that this was worth $8-15 million per annum.
Their data were in part derived from a tag-recapture study which indicated that in the
east coast fishery in Queensland the ratio of commercial to recreational caught fish was
3:1 (Russell 1988). However, Russell (1988) warned that commercial fishermen may
have under-reported the return of tagged fish; if this was the case the estimates of
Rutledge et al. (1990) would have been too high.

The only specific study on recreational fishing for barramundi in Queensland was
conducted in the Lakefield National Park in the period 1986-91 (Russell and Hales
1993). They estimated that each year 1500 anglers landed 4.4-9.4 tonnes of
barramundi. Direct expenditure (consumables used during the fishing trip) amounted
to $49 per trip, and capital investment per fishing party was $32 500.
There are no data available on the recreational fishery for barramundi in Western
Australia, or in other areas within the range of barramundi.
2.1.6 Stream modification and its effect on the distribution of barramundi
in Queensland.
Construction of barriers across streams to impound water for domestic, irrigation,
industrial, flood- and salinity-control purposes has been widespread in Queensland,
particularly in areas of major population centres or agricultural activity. Russell (1990)
reported that all the major rivers in eastern Queensland have dams or weirs on them.
The Water Resources section of the Queensland Department of Primary Industries
manages 27 dams impounding 72 781 ha and 79 weirs impounding 8327 ha of stored
water (Anon 1993). The South-east Queensland Water Board manages 3 dams
impounding 17 300 ha. Significantly more stored water is contained in large dams
controlled by various other water boards, local councils, mining companies and private
land holders. Thus, although there is no complete listing of the dams and weirs in
Queensland, it is clear that there are more than 100 000 ha of impounded water
(excluding farm dams) at full supply level in the State.

The largest storages are built in the headwaters of river systems, outside the natural
range of the barramundi. However, there are numerous weirs and tidal barrages built on
the lower reaches of rivers and creeks which have effectively excluded the barramundi
(and other diadromous fishes) from much of their former range. The extent of habitat
lost to the barramundi is not known, as a complete listing of all barriers is not available.
One documented example of the impact of man-made barriers on the distribution of
barramundi is the tidal barrage on the Fitzroy River at Rockhampton (Midgley 1987).
Historically, barramundi accessed many hundreds of kilometres of streams in the
Fitzroy, Don, Dawson, Comet, Nogoa, MacKenzie, Isaac and Connors Rivers (Fig.
2.2). Construction of the 5 m barrage on the Fitzroy River in 1970 effectively
precluded barramundi from all the freshwater reaches of the system. The barrage was
constructed with a fish ladder, but it is ineffective for upstream movement of
barramundi (Kowarsky and Ross 1981; Peter Long, Fisheries Inspector, QDPI,
Rockhampton, pers. comm. 1996). Several other weirs were constructed on the system
after the barrage, but they are comparatively low-level and are submerged in medium
floods, and so would not prevent movement of barramundi during floods.

Chapter 2. Barrarnundi biology and the Nile Perch proposal 40
Mackay
.••\
-) -.1;6.
_ ,c).0 r"' 1
Soi. , f 1
..-
Nebo I
J 1. Coral Sea •
WE I
-7
:1
v, IA,
I Moranbah ■1,...
(
% % :1
\
Peak Dowiia Mine Darn
41. Ii
-.. l
) Sanni Mine -•
..,--..-, ., ./. ..,
410 Dysart
1
ft
Sandy Creek Weir
• -%
Clermont
• 14_,
Theresa Creek)
A Tartrus Weir - -
Bundoora Dam River
f Eden Bann Weir
'fitzroy River Barrage
Emerald Weir ROckhampion
weir Blackwater
lifool Weir
Fairbarm
etcher Creek Wyiii
River
Gladstone
Neville Hewitt WE
Callide'Dam
•
r -Rolleston bit ISam

Mourn Karibo reek We
k
/
t
t Theodore Wei p
Theodore /
Legend
■
Major stream Orange Greek Weir
Gyandm Weir
Urban Centres
Glebe Weir
A Existing storage
k
ESA Existing storage with fish ladder ')
Barramundi distribution prior ( Injune
to barrage construction (1970) (
1. • - •
Fitzroy River catchment boundary Wandoan
f
0 100 200 •••••■ •
Kilometres
Figure 2.2 Historical distribution of barramundi in the Fitzroy River system in

Queensland. The tidal barrage at Rockhampton now effectively
prevents upstream colonisation by barramundi.

2.2. EVALUATION OF THE PROPOSAL TO INTRODUCE NILE
PERCH TO AUSTRALIA
2.2.1 The rationale for introducing Nile perch
There have been vast new aquatic habitats created in Queensland as a consequence of
stream impoundment. As a general rule, these impoundments do not naturally contain
either a wide variety or a large number of good angling fishes. Obviously, the primary
reason for this is the comparatively impoverished Australian freshwater fish fauna,
which is to a very large extent comprised of families of marine origin (Allen 1989).
The breeding biology of the majority of Australian freshwater species is generally
geared to riverine environments. Since rivers and their running waters obviously
provide very different habitats from that found in impoundments, our native fishes are
not necessarily capable of successfully colonising the latter.
There are two ways of developing the population of angling fishes in these water
storages. The first is to develop fisheries based on continual restocking. The
impoundments are stocked with a large number of hatchery-reared fingerlings, a
proportion of which then grow to a size suitable for catching and eating. When the
number of fish available declines to low levels due to fishing or natural mortality, the
impoundments are restocked. This is a common practice overseas and also in southern
Australia, where the best example is the chinook salmon fishery in Lake Purrumbete in
Victoria (Barnham 1977). Obviously the costs of maintaining hatcheries to supply
fingerlings make this system expensive.
The second method is to stock the impoundments with species of fish that will breed in
the confined waters, resulting in self-sustaining fisheries. In general, most native
species that qualify as sport fishes are unsuitable for this purpose, due to their riverine
requirements for breeding. However, there are some storages in which the inflowing
rivers provide suitable habitats for a few native sport fishes to breed, and from there to
recruit to the impoundments. In addition, there are some less-favoured sport fishes,
such as the eel-tailed catfish Tandanus tandanus, the saratoga Scleropages spp., the
archer fish Toxotes chatareus, and the sleepy cod Oxyeleotris lineolatus, that will breed
in most Queensland storages. In this context, the quest to obtain populations of high-
quality angling fish able to breed successfully in the majority of impoundments led to
arguments for the introduction of a 'suitable' tropical foreign species.
It was on this basis that S. H. Midgley (a private fisheries consultant) proposed in 1968
that the Queensland Government introduce the Nile perch Lates niloticus from Africa
for stocking impoundments in North Queensland (Midgley 1968, 1969). He
considered that the Nile perch was the most appropriate fish for the purpose because of
the following purported attributes: it is an excellent angling and eating fish; it breeds in
lakes and rivers; its temperature requirements would prevent it colonising waters in
southern Australia; and it inhabits the open waters as well as margins of lakes.
Midgley (1981) argued that it was unlikely there would be serious ecological
consequences resulting from the introduction of the Nile perch, on the grounds that it
was closely related taxonomically, was physically similar, and had similar feeding
habits to the barramundi. He considered that the only major biological difference
between them was that the barramundi bred in salt water, while the Nile perch bred in
fresh water.
Arguments against the proposal stressed the unpredictable consequences of faunal
introductions (Williams 1970) and also questioned the basic assumption that an exotic
fish was necessary, given the possibility that hatchery-bred barramundi could be used
for creating put-and-take fisheries (Reynolds and Moore undated, Williams 1982).
In order to resolve the debate on the Nile perch issue, a pre-release study was
conducted to determine the likely impact of the Nile perch on native fauna should it be
introduced to Australia. The experimental protocol was divided into three phases,
namely determining the lower temperature limiting the fish's distribution, determining
its salinity tolerance, and finally investigating its interaction with native fishes (Barlow
1984). Quarantine facilities were built for this purpose. However, before the fish was
imported, data on its lower temperature tolerance (Hashem and Hussein 1973) and
information from Africa on the impact of the Nile perch in lakes into which it had been
introduced (Barel et al. 1985), indicated that the Nile perch might not be suitable for
Australia. Consequently, the program was halted pending evaluation of the
information.
In the remainder of this chapter, I present data on the lower temperature tolerance of the
Nile perch and its potential distribution in Australia. I then summarise findings on its
effect on the native fish and fisheries of Lake Victoria and Lake Kyoga in eastern
Africa, as information from these exotic environments provides valuable background
against which the Nile perch proposal can be evaluated. Finally, I discuss other aspects
of the biology of the species relevant to its proposed role as a sport fish in Australia.
2.2.2 Temperature tolerance of L. niloticus, with reference to the species'
potential range if introduced into Australia.
Importance of the potential range of Nile perch if introduced into Australia
The potential range of the Nile perch if introduced into Australia was an important
evaluation criterion, as it was essential that the species not extend as far south as the
Murray-Darling River system. The Murray-Darling is the biggest river system in
Australia, in terms of both flow and catchment area. It originates in Queensland, and
extends through New South Wales into Victoria and South Australia. The requirement
not to colonise the Murray-Darling system stemmed from the possible impact of the
Nile perch — a large piscivore which often eats prey about 25% of its own body length
(Gee 1969, Okedi 1971, Bishai 1975, Ogutu-Ohwayo 1984) — on native fishes in the
system. Two of the most important recreational and commercial species in this system
are the golden perch Macquaria ambigua and the silver perch Bidyanus bidyanus.
Both species migrate upstream to breed, and recruitment downstream is by egg and
larval drift (Reynolds 1983). Thus it was conceivable that Nile perch living in the
upper reaches of the Darling River (southern Queensland) could prey on adult golden
perch and silver perch, and consequently affect their recruitment into New South Wales
waters, and perhaps even into South Australia.
Lower temperature tolerance of Nile perch
It was postulated that the southern extension of the Nile perch's potential range in
inland Australia would be controlled by a combination of biotic and abiotic factors
(Arthington and Mitchell 1986), with the most important probably being temperature.
The only experimental work on temperature tolerance of L. niloticus was conducted by
Midgley (1968), who suggested that the fish should not be subjected to temperatures
below 15°C but warned that his uncontrolled experiments could not be regarded as
conclusive and that future work could prove a lower thermal minimum.
More ecologically-meaningful information on the lowest temperature tolerated by the
Nile perch can be inferred from temperature data for waters in the Nile delta, which is
the coldest part of the species' natural range. Hashem and Hussein (1973) reported that
some mortalities of L. niloticus were usually observed in the Nozha Hydrodome near
Alexandria and in other shallow Nile delta lakes ... when the water temperature
dropped to less than 10°C, especially when this relatively low temperature prolonged
for some days'. The temperature in these lakes occasionally drops to 8°C overnight
(M. T. Hashem, pers. comm. 1985). At the Manzalah Fish Farm, where the Nile perch
was cultured in ponds with no reported temperature-related mortalities, the average
water temperatures recorded at 1100h in December (the coldest month) in 1969 and
1970 were 12.0°C and 11.2°C respectively (Eisawy et al. 1974). Considering that the
quoted temperatures were averages for the month, and were recorded after the
temperature had started to rise in the morning, one can reasonably assume that the
overnight minimum would have occasionally been less than 10°C. Thus, the
conclusion of Hashem and Hussein (1973) that 10°C may be considered a limiting
factor in the geographical distribution of L. niloticus into higher latitudes is reasonable.
Water temperatures in the Murray-Darling River system
To relate temperature tolerance information to the potential distribution of the Nile
perch in the Murray-Darling system it is necessary to define the temperature regimes of
these rivers. Unfortunately, there are insufficient long-term, continuous records of
water temperature over the whole river system to enable this to be done directly from
water temperatures. However, a study by Crisp and Howson (1982) showed that the
relationship between air and water temperatures in streams in northern England was
approximately linear when air temperature was above 0°C. If a similar relationship
could be shown to exist in the Murray-Darling system, use could be made of the
extensive air temperature data available from the Australian Bureau of Meteorology to
predict water temperatures over the whole catchment.
Relationship between water and air temperatures in the Murray-Darling River system
Water temperature data were obtained for 12 stations situated in New South Wales,
Victoria and South Australia. For convenience I refer to these sites as 'calibration
stations'. These stations were all in large rivers sufficiently removed from the snowline
to avoid the influence of melting snow on water temperatures.
All water temperature records were taken in the top 50 cm of the water column.
Temperatures were usually taken in the morning with mercury-in-glass thermometers,
although bimetallic thermometers were also used in Victoria. No allowance was made
for diurnal variation in temperatures, as this is usually negligible in large rivers (N.
McKay, River Murray Commission, pers. comm. 1985).
The New South Wales (Bourke, Menindee, Buronga, Bathurst, Yass, Condobolin,
Bairanald and Deniliquin) data are from Llewellyn (1978a,b,c, 1979a,b, 1980a,b,c,d,
1981). The Victorian (Torrumbarry and Swan Hill) temperatures were supplied by the
State Rivers and Water Supply Department (unpublished data). The periods during
which water temperatures were recorded and the number of records taken at each
station are listed in Table 2.2.

Table 2.2 The stations and rivers at which air-water temperature regressions were
determined, and the number of water temperature readings (n) taken
during the period 1963-83. (Air temperatures for Torrumbarry,
Buronga and Cadell were taken at Echuca, Mildura and Waikere,
respectively.)
Stations River Period monitored

Bourke Darling 1963-80 680
Menindee Darling 1965-80 593
Torrumbarry Murray 1977-82 189
Swan Hill Murray 1977-83 197
Buronga Murray 1963-71,1973-80 690
Cadell Murray 1 June 1981-31 May 1983 500
Murray Bridge Murray 1 June 1981-30 September 1982,
1 November 1982-30 November 1982
1 February 1983-31 March 1983 403
Bathurst Macquarie 1963-66,1969-80 686
Yass Yass 1966-71,1974-80 425
Condobolin Lachlan 1965-79 550
Balranald Murrumbidgee 1965-68,1970-80 512
Deniliquin Edwards 1963-80 592
Air temperature data corresponding to the periods of water temperature measurement
were obtained for the meteorological stations nearest to the calibration stations. The
only information readily available was average monthly mean air temperature (mean
air temp. = 1/2 (maximum + minimum)), in future referred to as AMMAT. Water
temperature records were available on a daily basis, so these were condensed to form
average monthly water temperatures (AMWT). The relationship between AMWT and
AMMAT was then determined by simple linear regression for each of the 12
calibration stations. This relationship was found to be statistically adequate and highly
significant (P < 0.01) in all cases. The results of these analyses are summarised in
Table 2.3.
Table 2.3 Values for constants a and b and their standard errors from linear
regressions AMWT = a + bAMMAT for 12 stations in the Murray-
Darling River system. (n is the number of monthly data points.)
Station a (SE) b (SE) R2

Bourke 1.48 (0.448) 0.96 (0.021) 0.92 176
Menindee 1.51 (0.664) 0.69 (0.034) 0.73 154
Torrumbarry 0.54 (0.586) 1.06 (0.034) 0.94 62
Swan Hill 0.68 (0.541) 1.02 (0.031) 0.94 75
Buronga 1.12 (0.479) 1.06 (0.027) 0.90 171
Cadell 1.33 (0.821) 0.94 (0.045) 0.92 24
Murray Bridge 1.17 (1.446) 0.98 (0.084) 0.89 19
Bathurst 0.59 (0.445) 1.12 (0.031) 0.89 169
Yass 3.09 (0.414) 0.99 (0.029) 0.91 115
Condobolin 1.31 (0.456) 0.95 (0.025) 0.91 149
Balranald 1.79 (0.668) 0.96 (0.038) 0.83 132
Deniliquin 1.95 (0.556) 0.96 (0.032) 0.84 168
An analysis of parallelism and displacement was then carried out to determine whether
a single regression relationship could be used for all stations. Results of this analysis
indicated that this was not possible. Thus, although water temperature was strongly
related to air temperature, the precise nature of the relationship varied according to
location.
Prediction of mid-winter water temperature in the Murray-Darling River system
Plots of long-term averages of air and water temperatures showed the same pattern over
all the calibration sites (Fig. 2.3) with both reaching minima during July. Thus,
conditions during this month would be most likely to determine the limits to the
distribution of Nile perch in the Murray-Darling system.
In order to estimate the positions of midwinter water isotherms, long-term mean air
temperatures for July were calculated using Bureau of Meteorology data from 142 sites
distributed throughout the catchment. These air temperatures were then used to predict
the average July water temperature for each site. Since it was not possible to obtain a
single regression equation for the entire catchment it was decided that, for any given
site, the equation obtained for the nearest calibration station would be used to predict
water temperature. These predicted water temperatures, along with the latitude and
longitude of each site, were processed using the program 'CONTOUR' (Briggs 1981)
to produce a map of the region showing midwinter water isotherms (Fig. 2.4).
The values used to plot the midwinter water isotherms are long-term means, around
which the average July water temperature in any year would fluctuate. Consequently, a
further map was constructed showing the approximate percentage probabilities of the
occurrence of average July water temperatures greater than 10°C (Fig. 2.5). This was
done using a pooled estimate of the variance in air temperatures of 0.94 and the
estimates of variance in water temperatures at each of the calibration stations.

Bark,
J F MA MJJ A S OND J F MA MJJ A S 0 ND
Figure 2.3 Average monthly water temperature ( ) and average monthly mean
air temperature (- - - ) for 12 sites (refer Table 2.2) in the Murray-
Darling River system. Data points are averages ± one standard
deviation.
JFMAMJJASOND JFMAMJJASOND
Figure 2.3 (Continued from previous page)

Figure 2.4 Predicted water isotherms during July (mid-winter) in rivers of the
Figure 2.5 Contours showing the percentage probability of the average July (mid-
winter) temperature exceeding 10°C in rivers in the Murray-Darling
River system.
Predicted range of Nile perch in eastern Australia
Definitive statements cannot be made regarding the potential range of the Nile perch
because many factors may influence its establishment and spread in Australia
(Arthington and Mitchell 1986). Nevertheless, generalisations based on the data
presented herein and the temperature information from northern Egypt can be
considered. Taking 10°C as the limiting temperature for the Nile perch, and cognizant
that the species can tolerate lower temperatures for at least short periods of time
(Hashem and Hussein 1973), it seems clear that this species could form permanent
populations in the headwaters of the Darling River system, where more than 9 out of
every 10 years can be expected to have average July temperatures greater than 10°C
(Fig. 2.5). In fact, the 80 percentile line (Fig. 2.5) takes in all of the Darling River and
its northern inflowing rivers. Spawning in the headwaters of the system would result in
recruitment downstream, which would be facilitated by the pelagic nature of the Nile
perch's eggs. Consequently, during a succession of wanner years it is quite probable
that the Nile perch could survive throughout the Darling River.
L. niloticus ceases feeding at temperatures lower than 15°C (Hashem and Hussein
1973). In the Nozha Hydrodome, in which a permanent population of L. niloticus
exists, temperatures lower than 15°C are experienced for 2-3 months each year
(Hashem and Hussein 1973). This thermal regime is similar to that at Bourke, where
the water temperature is lower than 15°C for approximately 3 months (Fig. 2.3).
Therefore, decreased feeding during the colder months would probably not be as
influential as the minimum temperature of 10°C in controlling the southern extension
of the Nile perch in inland Queensland and New South Wales.
Analysis of water temperatures in the coastal rivers of northern New South Wales was
not undertaken. Nevertheless, the water temperature in the Clarence River in winter is
higher than 12°C (Fig. 2.6), indicating that temperature would not restrict Nile perch
from colonising the river.
34
A
A A
30 AA
0 0 a A
A
A CA Le■
(°C )
28 g AA
A A
AA
a A
Te mp e ra tu re
A
22 A
A A
A I 'S a it
18 ACM.
&AA C. 41
as
14 A
10
J F M A M J J A S 0 ND
Date
Figure 2.6 Water temperatures recorded in the Clarence River at Lillydale between
1971 and 1985. (Data supplied by NSW Water Resources Commission).
In conclusion, the temperature data presented here indicate that the Nile perch, if
introduced into Australia, would be capable of forming permanent populations in the
headwaters of the Darling River and in the coastal rivers of northern New South Wales,
if all other environmental requirements were met. Consequently, if the Nile perch were
released in Australia the native fishes in these rivers could be adversely affected.
2.2.3 Impact of L. niloticus' in Lake Victoria, Lake Kyoga and Lake
Nabugabo
Range of L. niloticus in Africa
In western Africa the Nile perch is found in all the large river basins between the
Senegal and Zaire Rivers, and in the internal drainage of Lake Chad. In the Nile River
its natural distribution extends from Alexandria (32°N) to Lake Albert (2°N) on the
White Nile and to the headwaters of the Blue Nile in Ethiopia. It is also found in the
closed drainages of Lakes Tana, Chamo and Abaya in Ethiopia and Lake Turkana in
Kenya (Hopson 1972, Vanderpuye and Ocansey 1972, Greenwood 1976, Moreau
1982).
The range of the Nile perch was extended during the 1950s and early 1960s by
translocations within Uganda to Lake Victoria, Lake Kyoga, Lake Nabugabo, Lake
Kijanebolola, Lake Saka, Kabaka's Private Lake and Lake Awajo (Gee 1964 2). The
species subsequently thrived in Lake Victoria, Lake Kyoga and Lake Nabugabo
`Harrison (1991) concluded that the characters currently used in taxonomy of Lates are inappropriate.
He found that the Lake Victoria Lates showed significant morphological differences from all possible
parental stocks. Nevertheless, the stock is considered herein to be Lates niloticus, in accordance with all
other reports on Lates in Lake Victoria.
2 Gee (1964) also lists the stocking of Lake Kwania and Lake Bisina (formerly Salisbury) with Nile
perch. These (and numerous other small lakes) are discrete during dry years, but during years of high
water level (e.g. 1965-71) they form a confluent sheet of water comprising the Kyoga Lakes complex
(Vanden Bossche and Bemacsek 1990).
(Ogutu-Ohwayo 1993). Considerable information has been published on Nile perch in
Lake Victoria and to a lesser extent Lake Kyoga, but not on Lake Nabugabo other than
mentions by Ogutu-Ohwayo (1993, 1995). There are no literature reports documenting
the Nile perch's fate in Kabaka's Private Lake, Lake Awajo, Lake Saka and Lake
Kiganebolola; it is possible it failed to establish in these lakes.
Stocking of Lake Victoria and Lake Kyoga
The introduction of the Nile perch to Lake Victoria was originally suggested by keen
anglers (see Worthington 1973), but later proponents argued that the highly esteemed
Nile perch would eat the abundant haplochromine fishes (Cichlidae), of which more
than 300 species were present in the lake (Oijen et al 1981, Witte et al. 1992a,b), and
which were said to be under-exploited (Anderson 1961). Opponents of the idea argued
that it was indefensible on scientific grounds (introduction of a predator would reduce
the efficiency of production by 80%) and that since the diet of the Nile perch was
unpredictable the introduction would jeopardise the existing commercial fisheries
(Fryer 1960).
Lake Kyoga was stocked on eight occasions between 1954 and 1960 by the Uganda
Game and Fisheries Department, purportedly to obtain information on the effects of the
Nile perch on the haplochromine-dominated fauna, which was similar to that in Lake
Victoria (Anderson 1961). The stocking of Lake Nabugabo in 1959 and 1960, and
again in 1963, was similarly justified (Anderson 1961, Gee 1964). Unfortunately,
before these trials were evaluated the first Nile perch was caught in Lake Victoria in
May 1960 (Gee 1964). There is no official record of the original release of Nile perch
into Lake Victoria, but it probably took place in 1959 with fish brought from Lake
Albert (A. Achieng, Lake Basin Development Authority, pers. comm. 1985). Further
stockings were carried out in 1962 by the Uganda Fisheries Department with fish from
Lake Albert, and in 1963 by the Kenya Fisheries Department with fish from Lake
Turkana (Gee 1964, Achieng 1990).
Despite the documented justification for stocking the Nile perch in Lake Victoria and
Lake Kyoga, there is evidence that the real reason was to create sport fisheries. Firstly,
the proposed rationale does not equate with the fact that the haplochromines were eaten
by many of the numerous tribes bordering the lakes (but not by Europeans who
preferred the larger fishes) (Bwathondi 1984; G. Fryer, Windemere Research Station,
U.K., pers. comm. 1985). Indeed, the haplochromines traditionally (that is, pre-Nile
perch irruption) constituted 30-40% of the weight of all fish landed in the Kenyan and
Tanzanian waters of Lake Victoria (Ssentongo and Welcomme 1984; see also Table
2.4). Secondly, avid sport fishermen within the Uganda Game and Fisheries
Department advocated its dispersal (Achieng 1990), and Gee (1964) noted that Lake
Saka was stocked before any analyses of the stockings in Lake Kyoga or Lake
Nabugabo were undertaken. Finally, the interest in transport procedures (Hamblyn
1961), angling techniques (Hamblyn 1962), and the widescale translocation of the Nile
perch (Gee 1964) indicate that the development of sport fisheries based on this species
was a concern of the colonial administrators at the time.

Nile perch in Lake Victoria
The wave of colonisation of Lake Victoria by the Nile perch spread from the north to
the south. Small numbers of fish were caught in the Kenyan and Ugandan regions
from the early 1960s onwards (Gee 1965), but the species did not form a major portion
of the commercial catch until the mid-1970s. Thereafter, the contribution of the Nile
perch to the commercial catch increased dramatically, and by 1983 it constituted almost
70% of the catch (Table 2.4). By 1988, it formed over 90% of the fish biomass in the
northern part of the lake, as determined by experimental fishing using a variety of gears
(gillnets, large and small seines, trawls) (Ogutu-Ohwayo 1990a). In the south of the
lake (Tanzania) the Nile perch was caught as early as 1972 but catches in experimental
trawls did not increase substantially until 1983 (Goudswaard and Witte 1984).
Along with the massive increase in the catch of the Nile perch there was a marked drop
in the catch of virtually all other species (Acere 1984, Goudswaard and Witte 1984,
Okemwa 1984, Hughes 1986, Achieng 1990, Ogutu-Ohwayo 1990b, Ligtvoet and
Witte 1991). The original multi-species fishery (see Table 2.4) changed to one
dominated by Nile perch, the sardine-like Rastrineobola argentea and Oreochromis
niloticus (another species introduced from the Nile River). The catch composition is
partly due to target fishing for Nile perch, but stock assessment studies have indicated
that the catch statistics are indicative of the relative composition of the fish stock
(Ogutu-Ohwayo 1990a, Witte et al. 1992a). Fishing practices have contributed to these
changes, but there is no doubt that the major influence has been predation by Nile perch
(Hughes 1986, Ogari and Dadzie 1988, Ogutu-Ohwayo 1990a,b, Witte et al. 1992a).
The impact of the Nile perch on the haplochromine (Cichlidae) stock has been
catastrophic. The endemic fauna comprised more than 300 species, and virtually all
were restricted to Lake Victoria and Lake Kyoga (which receives its water from Lake
Victoria). Prior to the stocking of Nile perch, the haplochromines constituted about
80% of the demersal fish biomass, estimated at 200 kg/ha in the sublittoral areas of the
lake (Kunhongania and Cordone 1974, Witte et al. 1992b). The haplochromines were
the preferred prey of the Nile perch (Hamblyn 1962, Gee 1964, 1965, 1969), and their
numbers were rapidly depleted. Research in one area in Tanzania has shown that of the
123+ species present prior to 1980, about 80 had disappeared by 1990 (Witte et al.
1992). Extrapolation of these data indicates that approximately 200 of the 300+
endemic haplochromine species have disappeared, or are threatened with extinction
(Witte et al. 1992a,b).
The establishment of Nile perch ( and the tilapia Oreochromis niloticus) resulted in a
marked increase in the commercial fishery landings from Lake Victoria. Prior to 1975,
the commercial fishery in the Kenyan region was about 16 000 tonnes per annum; with
the introduction of Nile perch, the yield rose to 135 000 tonnes by 1989 (Table 2.4).
Similar trends were recorded in the Ugandan and Tanzanian sections of the lake
(Ogutu-Ohwayo 1990b), and the total annual yield from the lake was estimated to have
increased three- to fourfold (Ligtvoet and Witte 1991).

It is possible that the high yields and associated economic activity (Achieng 1990) will
not be sustained. Indeed, recent data indicate that the yield from the Tanzanian section
of the lake is declining (Table 2.5), and this is accompanied by a reduction in the
average size of Nile perch being caught (P.C. Goudswaard, Institute of Evolutionary
and Ecological Sciences, Leiden University, the Netherlands, pers. comm. 1996).
Ecological theory suggests that the introduction of a top predator (in this case a
piscivore) will result in the loss of approximately 80% of the biomass from the system
(Fryer 1960). However, the food web in the lake is changing, and as the present
situation is unstable (Hughes 1986, Ligtvoet and Witte 1991, Goldschmidt et al. 1993),
the future cannot be predicted. Ogutu-Ohwayo (1990b), using data from Lake Kyoga,
suggested that the population of L. niloticus is likely to decrease to a level
commensurate with oscillations in the abundance of its prey, and that the major primary
converter 0. niloticus (which is eaten by Nile perch, but apparently is not overly
susceptible) may increase and dominate the fishery.

Table 2.4 Percentage contributions of different fish species to the total weight (tonnes) of fish landed from Kenyan waters of Lake Victoria
from 1970 to 1991*.
Genus 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991
14.0 16.0 59.8 54.4 67.7 57.5 56.5 63.5 69.1 59.3 54.3 56.7 57.3
',ales 0.2 0.3 0.2 0.9 0.5 0.13 0.5 1.1 4.5
27.8 21.6 13.5 2.1 4.2 0.8 0 0 0.6 0.3 0.3 1.5 1.1 0.5
Haplochromis 32.7 32.0 29.0 33.2 35.0 27.9 34.0 32.4
Ragrineobola 3.2 5.1 7.8 10.5 21.8 27.4 30.3 34.7 36.5 30.5 35.1 20.4 17.1 21.3 27.1 29.2 30.5 24.5 36.5 38.5 39.6 39.2
27.5 21.1 14.8 10.1 5.6 3.9 5.4 7.4 10.9 9.0 18.6 10.2 7.3 5.5 10.4 10.7 2.7 2.8 2.4 2.3 2.2 2.4
Tilapiine
7.2 10.0 4.5 2.6 3.4 2.7 1.1 0.6 0.7 0.3 0.6 1.4 0.2 0.2
Clarias 9.7 12.5 17.0 15.7 12.9 15.6 13.0 9.1
Bagrus 6.7 7.1 5.4 8.6 6.4 8.4 5.5 6.0 5.9 5.8 2.4 1.1 4.2 3.1 0.1 0.1 0.1 0 0 0.1 0.1 0
2.6 1.5 1.4 0.5 0.4 0.3 0.1 0.2 0.2 0.1 0.1 0.1 0.1 0
Protoplents 11.0 12.8 12.7 13.0 8.6 1.1 5.0 4.0
0.3 0.7 0.5 1.0 0.4 0.1 0.1 0 0 0 0 0 0 - -
Schilbe 0.4 0.4 0.4 0.9 0.2 0.3
0.08 - - 0 0 0 0 0 "i 0 0 - -
Aleues 0.1 0.1 0.01 0.02 0.01
1.7 1.0 1.0 0.8 1.4 1.6 0.8 1.1 0.1 0.1 0.1 0.1 0.1 0.1 - - -
Barbus 1.3 1.6 1.7 1.1 0.7
Labeo 1.8 1.5 2.0 0.8 0.3 0.7 0.7 0.3 0.6 1.5 1.8 0.3 1.5 0.1 0.1 0 0 0
0.5 1.2 1.2 0.5 4.4 0.3 0.1 0 0.1 0.1 0 -

Mormyrus 0.5 0.5 0.5 1.1 0.8 0.3 0.5 0.6
Synodomis 1.1 0.7 1.3 1.3 1.1 0.8 1.0 1.6 0.6 1.6 1.4 1.3 0.4 0.1 0.1 0 0 0
5.0 5.2 7.9 3.8 1.9 3.8 - - - - - 3.3 2.6 1.0 0.1 0.7 - - -
Small mixed
16400 14918 15989 16797 17175 16581 18680 19332 23856 30592 26914 45667 60958 77327 71854 89589 103000 113000 123000 225000 185000 175000
Tonnes
* Source: Fisheries Department of Kenya, Statistical Bulletin. Also Ochumba et al. 1995.
Table 2.5 Percentage contributions of different fish species to the total weight
(tonnes) of fish landed from Tanzanian waters of Lake Victoria between
1988 and 1992*.
Genus 1988 1989 1990 1991 1992
Lates 68.0 78.6 75.4 65.8 56.7

Haplochromis 0.5 0.3 0.6 0.7 2.2
Rastrineobola 6.0 4.8 12.5 20.1 28.5
Tilapiine 9.7 8.5 7.1 11.2 9.1
Clarias 4.8 2.6 2.6 1.0 1.5
Bagrus 4.5 2.4 0.2 0.1 0.1
Protopterus 1.6 1.2 0.6 0.6 1.0
Schilbe 1.3 0.6 0.6 0.4 0.3
Alestes 0.1 0 0 0 0
Barbus 0.6 0 0 0 0.1
Labeo 0.7 0.3 0.1 0 0
Mormyrus 0.7 0.2 0.1 0.1 0.2
Synodontis 1.4 0.5 0.1 0.1 0.1
Tonnes 218 3 205 4 232 5 146 3 132 1

*Source: P.C. Goudswaard, Institute of Evolutionary and Ecological Sciences, Leiden University, The
Netherlands, pers. comm. 1996
Nile perch in Lake Kyoga
The Nile perch spread quickly throughout the lake, with the first fish being caught in
1958 (Anderson 1961). Records show that the species flourished in the 1960s and that
the fishery yield increased dramatically after the introduction of Nile perch (Gee 1969).
Prior to its introduction the yield was about 4500 tonnes; by 1963 it had increased to
20 000 tonnes and peaked at 167 000 tonnes in 1977 (Ogutu-Ohwayo 1990c).
However, in 1986-87 the yield was only 57 000 tonnes (Marriot et al. 1988 cited in
Ogutu-Ohwayo 1990c), and the fishery is now in decline.

As in Lake Victoria, the introduction of Nile perch to Lake Kyoga has had a
devastating impact on the indigenous fishes. Even in the late 1960s, it was suspected
that the decreasing abundance of two commercially important species, the catfish
Bagrus dogmac and the lungfish Protopterus aethiopicus, was due to competition from
the Nile perch (Stoneman and Rogers 1970). By the early 1980s the haplochromines
were virtually absent from the lake and other originally abundant species rare (Ogutu-
Ohwayo 1984). Evaluation of the diet of the Nile perch through time has shown that it
has changed as preferred species have been depleted (Ogutu-Ohwayo 1990c).
Overfishing is also implicated in the change in community composition, but
unquestionably the major impact has been predation by Nile perch (Ogutu-Ohwayo
1988, 1990a,b,c, Twongo 1988).
Nile perch in Lake Nabugabo
The native fish fauna of Lake Nabugabo was similar to that in Lakes Victoria and
Kyoga. Commercial catch statistics are not available for the lake, but experimental
fishing in 1991-92 showed that the Nile perch constituted by weight 79-87% of the
catch in seine nets and 68% of the catch in gill nets. The next most abundant species
was the Nile tilapia 0. niloticus which was recorded at 8-15% of the catch (Ogari
1995). The Nile perch has 'apparently had a similar effect in Lake Nabugabo to that in
Lakes Victoria and Kyoga.

2.2.4 Aspects of the biology of L. niloticus relevant to its proposed role as a
sport fish in Australia
In light of the preceding discussion, it is appropriate to examine other aspects of the
biology of the Nile perch to see how they equate with its proposed role as a sport fish in
Australia.
The Nile perch grows to approximately 170 cm and 70 kg (based on more than 500 000
specimens taken from Lake Victoria by the Haplochromis Ecology Survey Team: P.C.
Goudswaard, Institute of Evolutionary and Ecological Sciences, Leiden University, The
Netherlands, pers. comm. 1996). It is primarily a piscivore (Moreau 1982). Large fish
commonly eat prey up to 25% of their own body length, and on occasion will eat prey
up to 50% of their length (Hopson 1972, Bishai 1975, Ogari 1984, Ogutu-Ohwayo
1990a). Thus the Nile perch if introduced into Australia would have the potential to eat
the juvenile and breeding stock of every endemic freshwater fish, with perhaps the
exception of the Murray cod Maccullochella peeli peeli, the lungfish Neocertatodus
forsteri and the barramundi Lates calcarifer. Moreover, the Nile perch is opportunistic,
with the capacity to shift to different types of prey depending on availability (Ogari
1984, Ogutu-Ohwayo 1984, 1990b,c, Okemwa 1984, Hughes 1986, Ogari and Dadzie
1988). Consequently, the foraging activity of the Nile perch is not restricted to any
particular trophic level, but directly impacts on fish at all levels in the food chain.
The only documented requirement for breeding of the Nile perch is water temperature
rising to about 24°C, and thereafter the breeding season extends for as long as the
temperature exceeds that value (Hopson 1972). Thus, in Australia the Nile perch could
potentially breed in any body of water in which it could live. In fact, in Africa it has
colonised a broad range of habitats from large rivers to lakes and small ponds (Pruginin
and Kanyike 1967, Gee 1969, Planquette 1974, Moreau 1982, Bedawi 1985). There is
little doubt that the Nile perch would be capable of occupying the same range of
habitats in Australia, rather than being restricted to impoundments.
The Nile perch is clearly an r-strategist (that is, an animal that maximises its
reproductive potential) with large fish producing up to 16 million eggs per spawning
(Kenchington 1939, Okedi 1971, Loubens 1974, Ogutu-Ohwayo 1988). High
fecundities predispose invading species to build up large populations rapidly, as has
recently happened in Australia with the highly fecund carp Cyprinus carpio
(Arthington and Mitchell 1986). An r-strategy generally results from selection pressure
caused by highly variable environments (MacArthur 1972). In Australian rivers, flow
rates (and thus overall riverine environments) are characteristically highly variable,
both seasonally and between years (Lake et al. 1985).
The Nile perch is also a long-lived fish. Specimens from Lake Chad have been reliably
aged at 15 years (Hopson 1972) and there is evidence that in Lake Victoria it may live
for up to 20 years (Ligtvoet 1989, cited in Achieng 1990). The longevity of the species
endows its populations with resistance to, or the ability to survive, environmental
perturbation.
2.2.5 Conclusions
The biological characteristics of the Nile perch, and its success in colonising lakes
outside its natural range in Africa, indicate that it would be well suited to widespread
colonisation of Australian aquatic environments.. It is quite possible that flourishing
sport fisheries would develop, at least temporarily, soon after the release of the fish in
lakes. However, the temperature tolerance of the Nile perch is such that it would be
capable of forming resident populations in the headwaters of the Murray-Darling
system and in coastal rivers in northern New South Wales, thus endangering valuable
fish stocks in those regions. Furthermore, the Nile perch's dietary habits and its impact
on the fishes of Lake Victoria and Lake Kyoga suggest that its introduction into
Australia would almost certainly be detrimental to native aquatic fauna, not only in
man-made lakes but also in rivers and other water bodies.
On the basis of the analyses and information presented herein, fisheries agencies
abandoned the concept of introducing the Nile perch to Australia. In its stead,
emphasis was placed on investigations into the breeding of the barramundi to provide
juveniles for stocking freshwaters for enhancement of recreational fisheries in northern
Australia.
Chapter 3. Feeding habits 69
CHAPTER 3. FEEDING HABITS OF HATCHERY-REARED BARRAMUNDI
FRY
3.1 INTRODUCTION
The feeding habits of barramundi in the wild have been documented by Menon (1948),
Dunstan (1959), Patnaik and Jena (1976), Tait (1981), Davis (1985) and Russell and
Garrett (1985). At the time this study was undertaken, however, few papers had been
published on the rearing of barramundi fry in nursery ponds: MacKinnon (1987b) had
mentioned growth rates and survival in freshwater ponds, and De (1971) and Ghosh
and Pandit (1979) had discussed the food of barramundi fry reared in brackish water
ponds.
The scant information available was identified as a possible limiting factor in the
production of fingerlings and, consequently, the feeding habits of barramundi reared in
a freshwater hatchery between metamorphosis and stocking as advanced fry were
investigated. (Metamorphosis is defined herein as the time at which scales first form,
which in barramundi is at 11 mm TL). This entailed observations on feeding behaviour
in aquaria, analysis of stomach contents of pond-reared fish, investigation of diel
feeding patterns and stomach evacuation rates (and hence daily food consumption), and
description of morphometric relationships and growth rates. Such information was
considered to be essential for developing guidelines on husbandry techniques, for

70
Chapter 3. Feeding habits
maximising survival and growth and, in turn, large scale production of high quality
barramundi fry.
3.2 MATERIALS AND METHODS
Larvae were reared at the Northern Fisheries Centre, Cairns, Queensland. At 16-20
days after hatching and 8-10 mm total length (TL), the larvae were acclimated to fresh
water (Russell et al. 1987) and transferred to the Freshwater Fisheries and Aquaculture
Centre at Walkamin, Queensland. The fry were then either stocked directly into 0.1 ha
earthen nursery ponds which had been fertilised to promote the algal-zooplankton food
chain, or on-grown in the laboratory to 20 mm TL before stocking into ponds.
3.2.1 Feeding behaviour
Observations on feeding behaviour were conducted for two weeks in two glass aquaria
900*350*300 mm, each stocked with 50 fish. Both aquaria had a fine-sand substrate
and external filtration. Cover was provided in one aquarium in the form of 30 strips,
each 30*400 mm, of black plastic mesh (1 mm2) suspended vertically from polystyrene
floats. Several lengths of opaque plastic tubing of various diameters were also placed
on the substrate. There was no cover in the other aquarium. The fish were fed with
mixed live zooplankton obtained from freshwater ponds, and the temperature
maintained at 28°C. Night-time observations were made using dim red light. During
the study, the fish grew from 9 mm TL to 15-25 nun TL. Pigmentation patterns were
recorded and correlated with feeding behaviour. Complementary observations on the
behaviour of similar-sized fish in ponds were also made. This was done by observing
the areas in the ponds occupied by the fish, as well as their feeding activity.
Observations were made during the day, and at night-time with the aid of spot-lights.
3.2.2 Diet
Dietary analyses were conducted on 163 fish, ranging in TL from 10.0 mm to 56.2 mm,
taken from three stocked ponds at regular intervals through the 15-30 day rearing
period. Samples were collected at 0900 hours by seine netting from these ponds.
Immediately after collection the fish were killed by plunging them into an ice-water
slurry (which prevented regurgitation) and the stomach contents removed. Food items
in each stomach were identified to the lowest possible taxon and counted. The
zooplankton, insects, and vertebrates were individually grouped, and the percentage
volumes of these three groupings in each stomach were visually estimated. Estimates
were made independently by the author and a second person, and then averaged. Diet
was analysed as frequency of occurrence (number of fish containing a given food type
as a percentage of the total number of fish examined) and percentage composition by
number (number of food items of a given food type as a percentage of the total number
of food items) for individual taxa, and as percentage composition by volume for the
zooplankton, insect, and vertebrate groups.

3.2.3 Diel feeding
Diel feeding patterns were determined by sampling 8 fish from a pond-reared
population every 3 hours for 24 hours. This was done with small fish (average 15.8
mm TL, standard deviation 2.5 mm) and repeated with larger fish (37.2 ± 3.7 mm TL).
Specimens were processed immediately after collection. Dry weights of all fish and of
the stomach contents of the larger fish were recorded after drying at 60°C for 24 hours.
The weights of the stomach contents of the small fish were determined by counting the
zooplankters (copepods and Moina sp.) in each sample and interpolating from the
weight of a known number of the particular taxon (average weight of copepods 4.19
mg (n = 215), Moina sp. 3.44 mg (n = 465)). The stomach fullness index was
calculated from the formula
Stomach fullness index = dry weight of stomach contents *100
dry weight of fish
Values for stomach fullness index at different times were compared by analysis of
variance (ln transformed data) and Duncan's multiple range test.
3.2.4 Stomach evacuation rate
Stomach evacuation rates of feeding and starved fish were determined for two groups,
corresponding to the small and large fish used in the diel feeding study. Approximately
100 fish were held in an aquarium and fed to excess with wild zooplankton (copepods,
cladocerans and rotifers) for 24 hours. All fish were then transferred to another
aquarium and given a short meal (20 minutes duration) of Artemia nauplii. At the end
of Artemia feeding (to), 50 fish were removed to an aquarium without food and the
remainder to an aquarium with abundant wild zooplankton. Eight fish from each group
were then dissected at regular intervals (starting at a time — 20-150 minutes after to —
determined from preliminary trials) and the presence or absence of Artemia in the
stomachs recorded. The temperature in all aquaria was maintained at 28.0 ± 0.1 °C.
This temperature was chosen because it is close to the optimum temperature for growth
of barramundi. The median evacuation time was taken as the period between to and the
point when 50% of the fish contained no Artemia in their stomachs.
3.2.5 Daily ration
Several methods have been proposed for estimation of evacuation rates and daily ration
of fishes (see, for example, Jobling (1986) and Persson (1986)), but the method of
Bajkov (1935), and a variant of it by Eggers (1977), has recently been shown to be
particularly robust (Amundsen and Klemetsen 1988, Boisclair and Leggett 1988).
Hence, the Bajkov/Eggers method was used, where
Daily ration = 24 hour average stomach fullness index * 24
number of hours to evacuate food from stomach
The daily ration was calculated using evacuation times for both feeding and non-
feeding fish. (In this case, 'non-feeding' means complete evacuation of the stomach
before consumption of another meal.)

3.2.6 Morphometric relationships and growth rates
Regression relationships for TL, standard length (SL), wet weight (WIN/0 and dry
weight (DWt) were determined for fish 10.0-56.2 mm TL. All analyses were
performed on untransformed data, and F-tests were used to determine significance
levels of the correlation coefficients.
Growth rates were calculated for fish stocked into a 0.1 ha earthen pond at an initial TL
of 8.7 mm and density of 3.7 fish/m2; food for these fish was zooplankton produced
naturally in the ponds, and the water temperature was 25-33°C. On another occasion,
growth rates were determined for a batch of fish reared from 9.9 to 20.3 mm TL
indoors in a 1500 L concrete tank at a density of 2 fish/L; food was wild zooplankton
harvested from ponds, fed ad libitum; the temperature was maintained at 29-31°C, and
the total volume of tank water exchanged twice daily. After attaining an average TL of
20.3 mm, these fish were then stocked into a pond at a density of 2 fish/m2 and reared
to an average TL of 43.2 mm; water temperature during the pond-rearing phase was
23-28°C.
3.3 RESULTS
3.3.1 Feeding behaviour
Barramundi less than 17-18 mm TL swam actively throughout the water column,
especially when searching for food. The fish targeted zooplankton up to 10 mm away.
When eating, they lunged forward to within about 3 mm of the prey, apparently
ingesting food items by a sucking action associated with expansion of the gill covers.
Sated fish often lay on the sand substrate, while others remained comparatively still in
the water column, generally with the body angled upward. Fish of this size showed
neither territoriality nor affinity for cover. At night, the fish rested on or within 20 mm
of the substrate. Behaviour in aquaria with and without shelters was identical. A
descriptor termed for this behavioural stage was 'roving zooplanktivore'.
Fish larger than 17-18 mm TL adopted a comparatively sedentary habit, positioning
themselves near the mesh strips but not in the tube shelters on the bottom. The fish
oriented in a head-down position, usually in the lower one-third of the aquarium. They
often fed on zooplankton or insect larvae near or on the substrate, in addition to prey in
the water column. These fish would occasionally chase smaller fry if they came within
about 60 mm of the larger fish's station. In the aquarium without shelter, the fish were
more mobile and did not exhibit any form of territoriality. Consequently, aggressive
interactions were common, and resulted in some smaller fish being eaten. A term given
to this behavioural stage, applying to fish larger than about 18 mm TL, was 'lurking
predator'.
These two modes of feeding were characterised by strikingly different pigmentation
patterns. Fish from 9 mm TL to about 16 mm TL were heavily pigmented, with
distinct pale bands running transversely from the posterior half of the spiny dorsal fin
to the anal region, and between the posterior limits of the soft dorsal and anal fins
(Figure 3.1a). There was also a very prominent pale stripe on the uppermost part of the
head, extending from the snout to the dorsal fin. These roving zooplanktivores could
be readily seen in daytime in the surface waters of stocked ponds, but not at night, even
with the use of spot-lights.
Beyond about 16 mm TL, the pigmentation became less dark (Figure 3.1b). At about
18 mm TL (i.e. when the fish adopted the lurking predator strategy), the pigmentation
was paler and the patterning irregular, although the transverse pale bands were still
evident (Figure 3.1c). These fish were never seen in stocked ponds, presumably
because they occupied the deeper waters.
In aquaria, the change from roving zooplanktivore to lurking predator and the
associated change in pigmentation was generally completed in less than 2 days.

Figure 3.1. Pigmentation patterns of barramundi fry. (a) Barramundi 9-16 mm TL

are heavily pigmented with distinct transverse pale bands. (b) At
approximately 16-18 mm TL, the pigmentation becomes less
prominent. (c) At 18-50 mm TL, the pigmentation is comparatively
light and the patterning irregular.
3.3.2 Diet
The food items eaten by barramundi fry grouped into 10 mm TL intervals are presented
in terms of frequency of occurrence and percentage composition by number (Table
3.1). The most commonly consumed taxa were copepods (100-3001.i) and the
cladoceran Moina sp (400-8000. Both became progressively less important with
increasing size of fish, although Moina sp. continued to be a common element of the
diet throughout the study. Larval insects, especially the ephemopteran Centroptilium
sp. mm), were important food items for fish larger than 20 mm TL, whereas
tadpoles (>8 mm) were prominent in the diet of fish larger than 40 mm TL. Only one
fish was found eaten in the course of the study; this was a rainbowfish Melanotaenia
splendida 17.2 mm TL, consumed by a barramundi of 39.6 mm TL.
The change in diet with increasing size of fish is demonstrated by plotting the
percentage volumes of the zooplankton, insect and vertebrate groups against size of fish
(Figure 3.2). The diet of fish less than 18 mm TL was entirely zooplankton. As the
fish became larger, insects became more important, until in fish 33-48 TL they made
up about the same percentage volume as zooplankton. Vertebrates, in this case
predominantly frog larvae, were first eaten by fish 38 mm TL, and appeared to replace
the zooplankton and insects in fish larger than 48 mm TL.

Table 3.1. Percentage frequency of occurrence (% FO) and percentage composition by

number (% CN) of food items in the diet of barramundi Lates calcarifer fry reared in
freshwater ponds. Data are grouped into 10 mm TL size intervals. (n = number of fish
in each size class.)
Food item Size class
10-19 mm 20-29 mm 30-39 mm 4049 mm 50-59 mm

n = 20 n = 47 n = 47 n = 45 n=4
% FO %CN % FO %CN % FO %CN % FO %CN % FO %CN
Rotifera
Brachionus sp. 10.0 0.2 14.9 0.3 29.8 0.4 42.2 1.4 0 0
Asplanchna sp. 5.0 0.1 0 0 0 0 2.2 <0.1 0 0
Crustacea
Copepods 85.0 18.0 78.7 29.4 53.2 6.1 37.8 1.2 0 0
Ostracods 0 0 0 0 4.3 <0.1 2.2 <0.1 0 0
Moina sp. 100.0 81.7 74.5 66.8 74.5 90.2 77.8 93.3 25.0 98.2
Daphnia sp. 0 0 4.3 0.2 2.1 <0.1 6.7 1.4 0 0
Insecta
Ephemeroptera 0 0 40.4 1.6 72.3 2.4 64.4 2.2 25.0 0.9
Odonata 0 0 14.9 0.6 8.5 0.1 20.0 0.2 0 0
Notonectidae 0 0 21.3 0.6 10.3 0.1 6.8 0.1 0 0
Trichoptera 0 0 0 0 4.3 0.1 2.2 <0.1 0 0
Chironomidae 0 0 2.1 <0.1 4.3 0.2 13.3 0.1 0 0
Culicidae 0 0 4.3 0.3 6.4 0.2 4.4 0.2 0 0
Vertebrata 0 0 0 0 6.4 <0.1 11.11 0.1 75.0 0.9
Unidentified 0 0 17.0 0.2 19.2 0.1 6.7 0.1 0 0

100
zooplankton
insects
vertebrates
80 - ••
••• "s,
A \
\
i
20 -
• ''''' ..... \
I ..***
0 1111111Ii I
10 15 20 25 30 35 40 45 50 55
Total length (mm)
Figure 3.2. Percentage composition by volume of zooplankton, insects and

vertebrates consumed by barramundi 10-56 mm TL reared in freshwater
ponds. Five-point moving average data points were used to generate the
curves.
3.3.3 Diel feeding
Small barramundi (15.8 ± 2.5 mm TL) fed continuously during daylight, showing a
distinct peak in feeding activity at dusk, but during the night ceased feeding altogether
(Figure 3.3a). Food remaining in the stomach at the 2100-hour sampling occasion was
well digested, indicating that feeding had ceased soon after sunset. The average
stomach fullness index between dawn and mid-afternoon was 3.5 (n = 32), while at
dusk it was 7.0 (n = 8). At the time of the trial, there was a new moon which set 39
minutes after sunset.
The larger fry (37.2 ± 3.7 mm TL) showed a similar pattern, although this group did
continue feeding at a reduced level in the first half of the night (Figure 3.3b). There
were freshly eaten prey items in the fish sampled at 2200 hours, but food remaining at
0100 hours was well digested. At the time, the moon was in its first quarter, and set at
2315 hours. The average stomach fullness index for the times 1000, 1300, 1600 and
2200 hours (at which times the indices were not significantly different) was 3.5 (n =
32), but at dusk it was 6.3 (n = 8). That is, both the small and larger fish ate virtually
twice as much at dusk as at other times during the period of feeding.

10
9 (A)
8
7
id
Fu llne ss in dex
6
5
cd
4 i
3
ic ic
is
2
sunrise sunset ib
1
0 ! a I I „ I Ta
0 300 600 900 1200 1500 1800 2100 2400
Time (hrs)
Figure 3.3a. Average stomach fullness indices for pond-reared barramundi fry 15.8 ±
2.5 mm TL. The horizontal bars are means; vertical bars are ± one
standard error. Indices marked with the same letter are not significantly
different (P > 0.05, ANOVA, In transformed data, Duncan's Multiple
Range Test).
Fullne ss in dex
6- ie
5-
4-
3 - id
2 -
is
1-
1b sunrise sunset
0— M a 4 '
i 1 1 I I i i i
0 300 600 900 1200 1500 1800 2100 2400
Time (hrs)
Figure 3.3b. Average stomach fullness indices for pond reared barramundi fry 37.2 ±
3.7 mm TL. The horizontal bars are means; vertical bars are ± one
standard error. Indices marked with the same letter are not significantly
different (P > 0.05, ANOVA, In transformed data, Duncan's Multiple
Range Test).

3.3.4 Stomach evacuation rate
Gastric evacuation rates at 28°C for the small and larger fish under conditions of ad
libitum feeding and no feeding are plotted in Figure 3.4. With the small fish, the
median evacuation time for continuously feeding fish was 47 minutes, whereas for non-
feeding fish it was 210 minutes. For the larger fish, the median evacuation time for
feeding fish was 73 minutes, although under a non-feeding regime it took only 108
minutes for the same degree of evacuation to be achieved.
Within all groups the evacuation time was quite similar for individual fish, as
evidenced by the steep slopes of the percentage evacuation versus time plots (Figure
3.4).
3.3.5 Daily ration
The 24-hour average stomach fullness index for the small fish was 2.79 (n = 64) and
for the larger fish it was 2.86 (n = 64). Using these data and the median evacuation
times shown in Figure 3.4, the daily ration for the smaller fish under conditions of ad
libitum feeding and 'non-feeding' were 85.8% and 19.1% body weight respectively; for
the larger fish they were 56.3% and 38.1% body weight respectively.

100 - ■ small fish , continuously feeding

• small fish , non—feeding
■ large fish , continuously feeding
• large fish , non—feeding
,
e vac ua te d
80 -
I
,r
47 min
/
,
I
/ 210 min
60 - /
/
\ / /
i
/
/ / \
/ 73 min / 108 min
/ /
I /
I /
0 II IIIIIIIIII1111111j1111111111
0 25 50 75 100 125 150 175 200 225 250
Time (mins)
Figure 3.4. Stomach evacuation rates for barramundi fry 15.8 ± 2.5 mm TL (solid
line) and 37.2 ± 3.7 mm TL (dashed line), for continuously feeding fish
and non-feeding fish. Median evacuation times for each group are
indicated.
3.3.6 Morphometric relationships and growth rates
The regression equations describing relationships between TL, SL, WWt and DWt are
given in Table 3.2. The correlation coefficients for all relationships were highly
significant (P < 0.001). The relationship between SL and TL was linear (Figure 3.5),
whereas power equations described all length—weight and dry weight—wet weight
relationships (Figures 3.6, 3.7, 3.8, 3.9, 3.10 and 3.11).
Fish stocked prior to metamorphosis in an earthen pond grew from 8.7 to 44.4 mm TL
in 28 days; the daily growth rate based on weight was 16.4%, and density in the pond at
harvest was 0.5 fish/m2 (14% survival). Fish reared intensively in the laboratory grew
from 9.9 to 20.3 mm TL in 13 days; the daily growth rate was 15.5%, and survival
through this period was 98%. After stocking into a pond at 2/m 2, they grew to 43.2 mm
TL in 17 days, achieving a daily growth rate of 12.5% and survival of 99.5%.

Table 3.2. Regression equations, intercepts (a), slopes (b) and 1-2 values for length,
wet weight and dry weight relationships for barramundi Lates calcariftr fry (10.0-56.2
mm TL). (n = number of data pairs; CL = confidence limits; TL = total length (mm);
SL = standard length (mm); WWt = wet weight (mg); DWt = dry weight (mg)).
X n Equation a b(*95%CL) r2
TL SL 129 Y = a+bX -0.359 0.822 (0.006) 0.998
TL WWt 135 Y = cab 0.028 2.81 (0.087) 0.984
SL WWt 61 Y = cab 0.068 2.73 (0.078) 0.991
TL DWt 173 Y = at' 0.004 2.91 (0.031) 0.995
SL DWt 119 Y = a.14' 0.007 2.92 (0.038) 0.995
DWt WWt 62 Y = aXb 7.302 0.93 (0.018) 0.995
WWt DWt 62 Y = aX° 0.121 1.07 (0.021) 0.995
50
Stan dardlength (mm)
40
30
20
10
0

0 10 20 30 40 50 60
Total length (mm)
Figure 3.5. Relationship between total length (mm) and standard length (mm) for
129 barramundi in the range 10-56.2 mm total length.
87
9000
8000
7000
Wet we ig ht (mg)
6000
5000
4000
3000
2000
1000
0
0 10 20 30 40 50 60 70 80 90
Total length (mm)
Figure 3.6. Relationship between total length (mm) and wet weight (mg) for 135
barramundi in the range 11-87 mm total length.
Wet we ig ht (mg )
Standard length (mm)
Figure 3.7. Relationship between standard length (mm) and wet weight (mg) for 61
barramundi in the range 8.5-46.9 mm standard length.
88
E 400
4.1
a) 300
t, 200
100
10 20 30 40
Total length (mm)
Figure 3.8. Relationship between total length (mm) and dry weight (mg) for 173
barramundi in the range 10-56.2 mm total length.
Dry weig ht ( mg )
Standard length (mm)
Figure 3.9. Relationship between standard length (mm) and dry weight (mg) for

Wet weig ht (mg )

100 200 300 400 500 600
Dry weight (mg)
Figure 3.10. Relationship between dry weight (mg) and wet weight (mg) for 62
barramundi in the range 3.4-507 mg dry weight
CD 400
•
ca 300
Z" 200

500 1000 1500 2000 2500 3000
Wet weight (g)
Figure 3.11. Relationship between wet weight (mg) and dry weight (mg) for 62
barramundi in the range 22.9-2530 mg wet weight.
3.4 DISCUSSION
This study has revealed the relationship between ontogenetic changes in pigmentation,
feeding behaviour and diet of barramundi in the size range 10-50 mm TL. The
changes in pigmentation described here had previously been documented by Moore
(1982) and Mukhopadhyay et al. (1983), but, as their specimens were wild-caught fish,
they did not have the data to relate pigmentation changes to the feeding ecology of the
species. Field observation would be required to elucidate the nature of the association
between environmental cues and ontogenetic changes, but the pigmentation changes
may be associated with camouflage or predator avoidance. Post-larval barramundi
inhabit brackish or freshwater swamp systems, typically comprising plant communities
of mangroves, Melaleuca spp. or Eleocharis grasses (Moore 1982, Russell and Garrett
1985). Such systems are characterised by low intensity light, due to filtering by
vegetation and occasionally the presence of tannins in the water. The dark pigmentation
and distinct pale vertical bands on the 9-16 mm TL barramundi would camouflage fish
swimming through the water column in swampy habitats. The irregular but paler
pigmentation, which first develops at about 17 mm TL and occurs at the same time as a
change in feeding behaviour and diet, may provide camouflage for the fish while
utilising the lurking predator feeding habit, particularly when stationed near submerged
vegetation. The change to the adult form of pale silvery colour proceeds gradually in
fish 50-100 mm TL (Moore 1982), and is presumably associated with movement from
swamps to nearshore or riverine systems (Moore 1982, Russell and Garrett 1985).
The ecological significance of the night-time benthic habit of 10-16 mm TL
barramundi is difficult to explain without observation of vertical migration patterns of
co-existing fauna in natural habitats. An alternating benthic-pelagic habit has been
reported in flatfish approaching metamorphosis, and has also been associated with
circatidal rhythms as a means of recruiting to and maintaining position in estuaries
(Boehlert and Mundy 1988). For barramundi being reared in freshwater ponds,
however, this habit could be detrimental because of the presence of predators (for
example, dragonfly nymphs) on the substrate.
The shift in diet of barramundi from planktonic crustacea (the sole food category in fish
smaller than 18 mm TL) to insect larvae (first eaten at 18 mm TL) paralleled the
change from a roving zooplanktivore to a lurking predator feeding habit. The general
trend for the diet to change with increasing fish size from planktonic crustacea to a
mixture of insect larvae and crustacea, to larger arthropods and vertebrates, is similar to
that reported by Patnaik and Jena (1976) and Davis (1985) for wild fish in India and
northern Australia respectively, and by De (1971) for fish reared in a brackish water
pond in India. It is obvious from these studies that to maximise the growth potential
and survival of post-larval barramundi reared in ponds, it is necessary to have abundant
zooplankton while the fish are 10 to 40 mm TL, and to have macroscopic animals of
increasing size as prey for barramundi about 20 mm TL and larger.

Cannibalism was not recorded in the fish sampled from the ponds for the dietary
studies, although it was evident in the feeding behaviour studies conducted in aquaria.
De (1971) found cannibalism in barramundi at the upper end of the 40-200 mm TL
size class, and Ghosh and Pandit (1979) did not consider the piscivorous habit of L.
calcarifer to be developed until the fish attained 125 mm. This apparent lack of
cannibalism during extensive rearing in ponds is in contrast to intensive rearing in
tanks, where it is necessary to constantly grade the population according to size after
metamorphosis to prevent cannibalism (Awang 1987, Maneewong 1987, Fermin 1991,
Trendall 1991).
Diel feeding patterns have not previously been documented for barramundi fry. This
study revealed a distinct pattern of feeding commencing at dawn and remaining
constant through the day, but increasing significantly at dusk. Feeding ceased during
darkness, although there was some minor feeding activity under moonlit conditions, at
least by the larger fish. The enhanced feeding at dusk is presumably an adaptation to
store food for the ensuing non-feeding period. Daytime feeding with a peak at dusk has
been shown for larvae of American shad Alosa sapidissima (Wiggins et al. 1985) and
largemouth bass Micropterus salmoides (Laurence 1971). Daytime feeders can show
other patterns, however. For instance, the rabbitfish Siganus guttatus feeds continually
through the day with no apparent peaks in activity (Hara et al. 1986); the rainbow trout
Oncorhynchus mykiss has been recorded as a daytime feeder with a peak at midday
(Angradi and Griffith 1990); juvenile chinook salmon Oncorhynchus tshawytscha

exhibits maximum feed intake at dawn (Sagar and Glova 1988); and juvenile haddock
Melanogrammus aeglefinus are daytime feeders, with peaks in feeding activity at dusk
and an inferred peak in the morning (Hall 1987).
The evacuation rates of the small and large barramundi fry under feeding and non-
feeding conditions appeared to be consistent with their feeding ecology. The small fish
(about 16 mm TL) were roving zooplanktivores, whereas the larger fish (about 37 mm
TL) had adopted the lurking predator habit. Under conditions of continuous feeding
the smaller fish evacuated food more quickly than the larger fish, but under non-
feeding conditions the smaller fish took considerably longer to evacuate their stomachs
(Figs. 3.3a and 3.3b). It is postulated that the variability in evacuation rates for the
small fish is an adaptation to prey availability. For fish consuming zooplankton, which
is usually patchily distributed, it would be advantageous to have a comparatively fast
evacuation rate when food is abundant and a slow rate when food is scarce. Larval
herring Clupea harengus pallasi exhibit this strategy of increasing zooplankton intake
and evacuation when food concentrations are high. The increased intake is associated
with decreased carbon assimilation efficiencies, but experimental studies have shown
that the magnitude of increasing ingestion more than compensates for the decreased
carbon assimilation, and herring larvae thus gain greater total energy under conditions
of high food concentration (Boehlert and Yoklavich 1984).

The strategy of maximising food intake apparently becomes less important as the
barramundi grows from an obligate zooplanktivore to a lurking predator. Whereas the
smaller fish can be classified as number maximisers, the larger fish, which have a wider
range of prey available and greater energy reserves, appear to be energy maximisers,
foraging and processing prey to maximise net energy. This pattern supports the
conclusion of Griffiths (1975) that many invertebrates and larval vertebrates eat prey as
they are encountered while adult vertebrates feed as energy maximisers.
The energy maximiser strategy would also be advantageous for carnivorous fish which
can consume large prey. Bromley (1987) has shown that for turbot Scophthalmus
maximus the rate of gastric digestion is slower for larger meal sizes, but the amount of
chyme released to the intestine remains constant, or independent of, meal size. The
passage of chyme to the intestine is controlled by the powerful duodenal sphincter,
which apparently lets chyme pass freely but retains solids very effectively. It is quite
possible that digestive processes for juvenile and adult barramundi are similar, given
that, like turbot, they have a powerful duodenal sphincter and the ability to consume
large prey.
Both the diel feeding pattern and evacuation rates demonstrated in this study have
implications for fish culturists rearing fingerling-sized barramundi. The results indicate
that feeding should take place throughout the day with more food being provided at
dusk. If feeding live food, it may be more efficient to feed less food more regularly
rather than large amounts on fewer occasions. If dry diets are being used, feeding
frequency should probably be at least every two hours. However, it is recognised that
optimising both the ration size and frequency would depend on a complex interplay
between evacuation rates (which are temperature-dependent), digestibility and gross
biochemical composition of food, and assimilation efficiency (Boehlert and Yoklavich
1984, Bromley 1987).
The daily growth rates (13-16%) of barramundi fry are high, although not surprisingly
so for fry reared at temperatures in excess of 25°C. Arumugam and Geddes (1987)
reported that another Australian species, the golden perch Macquaria ambigua, reared
from first feeding larval stage to fry in fertilised ponds under temperature regimes
similar to those in the present study, showed daily growth rates of 15%. Similarly,
Houde (1989) has shown a positive correlation between weight-specific daily growth
rates and temperature, and that daily growth rates greater than 20% are common for
larvae reared at temperatures higher than 25°C. Fermin (1991) and Fermin and Bolivar
(1994) recorded daily growth rates of 2.6-18.8% for barramundi larvae reared at 27-
30°C on either live or frozen Moina macrocopa, Artemia salina or minced fish flesh
under laboratory conditions.
The daily rations determined in this study (19-86% for 16 mm TL and 38-56% for 37
mm TL fish) are high, which concurs with the rapid growth rates of, and high rearing
temperatures employed for, barramundi. The daily ration figures also indicate the
amounts of live feeds required for optimal growth of barramundi in intensive nursery
culture. Extrapolation to optimum levels of dry diets, however, would require
comparison of the energy content of the zooplankton used in this study and the dry
diets.
Morphometric relationships for barramundi fry have been recorded by De (1971), but
previous documentation of wet and dry weight relationships for barramundi fiy has not
been located. These relationships, along with information on feeding behaviour, diel
feeding patterns and stomach evacuation rates, are valuable indicators for optimising
feeding regimes for barramundi fiy.

Chapter 4. Predation by dragonfly nymphs 97
CHAPTER 4. PREDATION BY DRAGONFLY NYMPHS OF THE SPECIES
Pantala flavescens (ODONATA: LIBELLULIDAE), ON
BARRAMUNDI
4.1 INTRODUCTION
Barramundi, Lates calcariftr, is being reared in Australia for aquaculture and for
stocking to enhance recreational fisheries. Barramundi is a catadromous species, and
has a physiological requirement for salt water up to about 8-10 days old (approximately
5-6 mm total length, TL), but thereafter it can be grown in either salt or fresh water.
One technique for producing fish for stocking impoundments is to rear them in
freshwater ponds from 8-9 mm TL (that is, prior to metamorphosis) to the commonly
accepted stocking size of 40-50 mm TL. This technique was employed at the
Freshwater Fisheries and Aquaculture Centre, Walkamin, but survival through the
pond-rearing phase was always less than 50%, despite low stocking rates (2-16/m 2) and
good pond management practices.
One possible explanation for the lower-than-expected survival became evident from
studies on the feeding behaviour of small fish, documented in Chapter 3. Barramundi
undergo a distinct change in diet and feeding behaviour at about 18 mm TL. While less
than 18 mm TL (corresponding to about the first 10 days after stocking into ponds) the
fish are obligate zooplanktivores and show no affinity for cover. Fish of this size when
hungry swim throughout the water column, but when sated remain still in the water
column or even lie on the substrate. Fish larger than 18 mm TL feed on insects and
zooplankton, orienting near shelter and adopting a predatory mode of feeding.
It was postulated that the behaviour of fish less than 18 mm TL may make them
susceptible to benthic predators, such as the nymphs of dragonflies (Odonata).
Odonate nymphs are common in freshwater ponds in the tropics and have often been
implicated in mortality of larval fish (e.g. by Huet 1970 and by Piper et al. 1982).
However, the only quantitative report located is that of McGinty (1980) (referred to in
Tave et al. 1990) who found that when dragonfly nymphs were uncontrolled they
consumed all channel catfish, Ictalurus punctatus, fry in nursery ponds. McGinty
stocked earthen ponds with 50 000 and 100 000 fry/ha. Prior to stocking, some ponds
were treated with methyl parathion to kill dragonfly nymphs, while other ponds were
untreated. Survival of fry in the treated ponds ranged from 79-98%, whereas in the
untreated ponds dragonfly nymphs consumed all fry within 47 days. Horn et al. (1994)
conducted laboratory experiments on predation by odonate nymphs (Enallagma sp. and
Tramea sp.) on larval razorback suckers (Xyrauchen texanus). The fish larvae (11-15
mm TL) were highly susceptible to both odonate species and, based on observations of
nymph densities in natal rearing environments and the ecology of the water bodies
(water depth, vegetation, absence of predatory fish), Horn et al. (1994) concluded that
predation by odonate nymphs may be a severe constraint to recruitment of razorback
suckers.
In the present study, laboratory-based experiments were conducted to determine the
vulnerability of different-sized barramundi fry to predation by dragonfly nymphs. For
comparative purposes, the experiment was also conducted on larvae of the sooty
grunter, Hephaestus fuliginosus, a tropical freshwater sportfish reared at the Walkamin
Research Station for stocking impoundments. The colonisation and development of
odonate faunas in two ponds were also monitored. Based on the results of this work, an
alternative rearing strategy which precludes predation by odonate nymphs was devised
and tested.
4.2.1 Predation on fish by odonate nymphs
The impact of predation by nymphs of the anisopteran dragonfly, Pantala flavescens,
on barramundi and sooty grunter larvae was determined in 16 aquaria of dimensions
900*350*300 mm, set up in a 4*4 randomised block design. Temperature was
maintained at 28.0 ± 0.5°C. Each aquarium contained 20 fish, together with abundant
zooplankton (captured using air-lifts and plankton nets in ponds) and 6 tadpoles
(approximately 20 mm total length). The zooplankton was eaten by both the fish and
dragonfly nymphs, and was supplied in excess of the feeding requirements of the fish
and nymphs. The tadpoles were included because they are a common component of
the macrofauna of ponds, and it was thought that their exclusion might artificially
increase predation on fish due to the lack of alternative prey for the dragonflies. The
treatments were 0 (control), 3, 6 and 12 dragonfly nymphs per aquarium, with the
numbers chosen to represent low, medium and high densities of nymphs typically seen
in ponds (see Section 4.3.2). The nymphs, zooplankton and tadpoles were introduced
to the aquaria at least one hour before adding the fish. This ensured that the nymphs
had the opportunity to feed prior to adding the fish. The nymphs had an average TL of
23 mm (standard deviation 2.5 mm, range 17-25 mm, n = 14).
Two size-groups of barramundi were tested on separate occasions. The two size-
groups exhibited different feeding strategies, being categorised as roving
zooplanktivores, applying to fish less than about 18 mm TL, and lurking predators,
applying to fish larger than about 18 mm TL. In this experiment, the roving
zooplanktivores averaged 9.9 mm TL (s.d. 0.7 mm, range 9.1-11.9 mm, n = 30), and
the lurking predators 20.2 mm TL (s.d. 1.6 mm, range 17.9-24.3 mm, n = 30). The
sooty grunter larvae, which do not undergo a similar change in feeding habit, averaged
18.7 mm TL (s.d. 0.7 nun, range 17.2-20.2 mm, n = 30). The narrow size ranges of the
fish within each group effectively precluded cannibalism. The respective sizes of the
fish, dragonfly nymphs and tadpoles prevented predation by the fish on either the
dragonfly nymphs or the tadpoles.
The number of mortalities of fish over a 20 hour exposure period (10 h light/10 h dark)
was recorded, along with observations on the behaviour of the fish and the dragonfly
nymphs. Night-time observations were made using dim red light. Mortality data
(arcsin transformation) were analysed using analysis of variance and means were
compared using the LSD test. In analysing the tadpole mortality data, Bartlett's test
was used to test for homogeneity of variances for the three trials (i.e. small barramundi,
large barramundi and sooty grunter) prior to combining the data to examine the
influence of dragonfly density on mortality.
4.2.2 Colonisation of ponds by odonate nymphs
Colonisation and development of odonate fauna was investigated in two 0.1 ha earthen
ponds, situated at the Freshwater Fisheries and Aquaculture Centre, Walkamin. Black
plastic, extending 2 m above and below full supply level, was laid around the perimeter
of one pond (Pond A). This effectively eliminated ovipositing habitat for odonate
species that lay eggs into soil or the tissues of plants near the water (endophytic species,
see Williams 1980). Both ponds were filled on 16 February 1988 with channel water
originating from Lake Tinaroo. Maximum and minimum temperatures at the bottom of
the pond (approximately 1.3 m depth) were recorded every 3 days during the 5 weeks
of sampling. Fish were not present in either pond.
Epibenthic fauna was collected with a dredge net (Williams 1980, Barlow et al. 1982)
made of 425 pin mesh net. The open end was 460*310 mm, and the net tapered for 1
m to a collecting jar 60 nun in diameter. The net was supported 60 mm off the
substrate by a frame attached to a ski on each side. A "kick chain" joined the skis 130
mm in front of the mouth of the net.

Six samples were collected twice weekly from each pond for 5 weeks, starting 3 days
after filling. The net was positioned randomly in the ponds, and then towed over 15 m
of substrate at approximately 1 m/sec. The samples were preserved in 70% alcohol.
Odonate nymphs were separated, identified and enumerated in the laboratory.
Development of the population of the most abundant species, Pantala flavescens, was
monitored by measuring head-widths (HW) of up to 100 individuals on each sampling
occasion. HW was converted to total length (TL) using the formula
TL = 4. 035H10.872
derived from measurements taken from 100 fresh specimens.
4.2.3 Comparison of husbandry techniques
Newly metamorphosed barramundi were transferred to freshwater and held at 2 fish/1 in
two 1500 L indoor tanks, in a flow-through system with water exchanged every 12
hours. The fish were fed on freshwater zooplankton harvested from ponds. After 14
days, when they had attained an average TL of 20 mm, the fish were introduced into 2
ponds which had been filled 7 and 11 days prior to the stocking. Survival to fingerling
size (40-50 mm TL) and density at harvest under this rearing system were compared
with that under the traditional system, which was to place the fish prior to
metamorphosis into nursery ponds, filled 14-16 days previously, and rear them to
fingerling size entirely in the ponds.

4.3 RESULTS
4.3.1 Predation on fish by odonate nymphs
The impact of predation by different densities of dragonfly nymphs on two size-groups
of barramundi and one group of sooty grunter larvae is shown in Table 4.1 and Figure
4.1. Comparing the different groups of fish, the mortality of the large barramundi and
sooty grunter was not significantly different, and both groups had significantly less
mortality (P < 0.05) than did the small barramundi.
For the small barramundi, there was a significant difference in the mortality of fish at
each density of nymphs (P < 0.05). One fish died in one control aquarium, and there
were average mortalities of 46%, 63% and 84% in the 3, 6 and 12 nymphs per
aquarium treatments, respectively.
In contrast, predation on the larger barramundi was much less severe. There were no
mortalities in the control aquaria, and averages of 4%, 11% and 16% mortality in the 3,
6 and 12 nymphs per aquarium treatments, respectively. The numbers of mortalities at
the 0, 3 and 6 nymphs per aquarium treatments were not significantly different, but the
number of mortalities at the 12 nymphs per aquarium treatment was significantly
different from that at the 0 and 3 nymphs per aquarium treatments (P < 0.05).
The impact of dragonfly nymphs on the sooty grunter larvae was similar to that on the
larger barramundi. In the controls there were no mortalities, and averages of 11%, 13%
and 30% mortality in the 3, 6, and 12 nymphs per aquarium treatments, respectively.
There was no significant difference in the number of mortalities at 0, 3 and 6 nymphs
per aquarium, but the number of mortalities at 12 nymphs per aquarium was
significantly different from that in the other three treatments (P < 0.05).
The response of the two groups of barramundi to the dragonfly nymphs was markedly
different. The small barramundi showed no avoidance or escape response even when
being attacked. During the day, the fish continually swam into the zone within 50 mm
of the substrate, where they were easily captured by nymphs jumping off the substrate.
At night, nearly all the fish were near or on the substrate. Behaviour of fish in the
control and treatment aquaria was similar. Selective predation on fish near the
substrate meant that, at the end of the exposure period, all survivors in the treatment
aquaria were in the upper reaches of the water column, whereas in the controls most
remained near the substrate.
The larger barramundi generally remained more than 50 mm above the substrate, in
controls and treatments, and during light and dark conditions. When a nymph lunged
at a barramundi, the fish darted away, easily escaping the predator.
During the daytime, the sooty grunter larvae swam continuously in a loose school in
the upper half of the water column in both control and treatment aquaria. At night, in
the control aquaria all fish were resting motionless within 1 cm of the sand substrate,
dispersed evenly around the aquaria. In contrast, in the treatment aquaria, only 0-3 fish
were near the sand, with the remainder in a school near the surface.
Analysis of the mortality data for the tadpoles indicated that within each trial there was
no significant effect (P > 0.05) of dragonfly density on tadpole mortality (Table 4.1).
Inspection of the data indicated that this negative result was a consequence of the low
power of the test (Searcy-Bernal 1994). Further analysis employed Bartlett's test for
equality of variance between the trials, which indicated that the variances were not
unequal (X22a= 3.504 (P > 0.1)). Thereafter, grouping the data from all three trials
showed that overall there was a significant effect (P < 0.05) of increasing mortality of
tadpoles with increasing density of dragonflies (Table 4.2).

Table 4.1. Number of mortalities (mean ± s.d.) of fish (Table A) and tadpoles
(Table B) exposed for 20 hours to predation by nymphs of the dragonfly
Pantala flavescens. At each density of dragonflies there were four
replicates, initially stocked with 20 fish and 6 tadpoles. Different
superscript letters within the same column indicate significantly
different (P < 0.05) means.
Table A - Fish mortality

Treatment Small barramundi Large barramundi Sooty grunter
0 dragonflies 0.3 ± 0.5a Oa Oa
3 dragonflies 9.3 ± 1.7b 0.8 ± 0.5a 2.3 ± 1.0a
6 dragonflies 12.5 ± 2.4c 2.3 ± 2.2ab 2.5 ± 1.3a
12 dragonflies 17.3 ± 2.2d 3.3 ± 1.3b 6.0 ± 2.6b
Table B - Tadpole mortality

Treatment Small barramundi Large barramundi Sooty grunter
0 dragonflies Oa 0.3 ± 0.5a Oa
3 dragonflies 1.3 ± 1.3a 0.8 ± 0.5a 0.8 ± 0.5a
6 dragonflies 1.3 ± 1.3a 1.0 ± 0.8a 1.3 ± 1.5a
12 dragonflies 1.8 ± 1.3a 1.3 ± 1.0a 2.5 ± 1.7b
Table 4.2. Number of mortalities (mean ± s.d.) of tadpoles exposed for 20 hours
to predation by nymphs of the dragonfly Pantala flavescens. The data
are combined for the three trials (namely, small barramundi, large
barramundi and sooty grunter). At each density of dragonflies there
were 12 replicates, each initially stocked with 6 tadpoles. Different
superscript letters within the same column indicate significantly
different (P < 0.05) means.
Treatment Mortalities
0 dragonflies 0.08 ± 0.29a
3 dragonflies 0.92 ± 0.79ab
6 dragonflies 1.17 ± 1.11 be
12 dragonflies 1.83 ± 1.34c
107
Chapter 4. Predation by dragonfly nymphs
Small barramundi
Large barramundi
• Sooty grunter
Figure 4.1. Mean number of mortalities of small barramundi (9.9 mm TL), large
barramundi (20.2 mm TL) and sooty grunter (18.7 mm TL) exposed
for 20 hours to predation by nymphs of the dragonfly Pantala
flavescens. At each density of dragonflies there were four replicates,
initially stocked with 20 fish.
4.3.2 Colonisation of ponds by odonate nymphs
Maximum and minimum temperature ranges in both ponds were 27-30°C and 23-
27°C, respectively.
The number of dragonfly larvae sampled from the two ponds (A with the black plastic
perimeter, B without plastic) in the first 35 days after filling is presented in Table 4.3.
By far the most abundant species in both ponds was Pantala flavescens (Fig. 4.2),
which generally made up more than 90% of the odonate fauna sampled.
The density of P. flavescens determined from the data in Table 4.3 is plotted in Figure
4.3a. The density appeared to peak in both ponds at day 14 after stocking, at 6
larvae/m2 in Pond A and 12/m2 in Pond B. The nymphs continued to increase in size
throughout the 35 days of sampling (Fig. 4.3b).
4.3.3 Comparison of husbandry techniques
The barramundi stocked into the two indoor flow-through tanks grew from an average
of 9.9 mm to 20.2 mm TL in 16 days. Survival through this period was 94.5% and
97.6% in the two tanks. These fish were then stocked into two 0.1 ha outdoor ponds
and on-grown to fingerling size (40-50 mm TL). Mean (±s.e.) survival during the
pond phase was 98.2±1.3% (Table 4.4). In comparison, mean (±s.e.) survival of
barramundi stocked at an initial average size of 9.9 mm TL into four 0.1 ha ponds and
reared to 40-50 mm TL was 31.2±9.1% (Table 4.4). Mean (±s.e.) density at harvest
for the larger fish at time of stocking was 2.8±0.75 fish/m 2, while for the smaller fish it
was 0.71±0.06 fish/m2.

Table 4.3. Mean number (± s.d.) of odonate nymphs sampled per 15 m tow with a dredge net 0.46 m wide (determined from 6 tows per
sampling occasion) in two 0.1 ha freshwater ponds in north-eastern Queensland during the first 35 days after filling.
Days after filling

Taxon Pond

3 7 10 14 17 21 24 28 31 35
Anisoptera
Libellulidae
Pantalaflavescens A - 0.2 (0.4) 0.14 (0.8) 39 (25) 27 (8) 34 (19) 24 (11) 12 (5) 25 (20) 17 (5)
- 5 (3) 3 (3) 82 (19) 59 (19) 52 (23) 52 (17) 36 (6) 47 (14) 38 (24)
Other A - - 0.2 (0.4) 0.4 (0.8)
2 (2) 2 (3) 10 (7) 0.8 (1.0) 0.3
(0.8)
Aeshnidae sp. 1 A - - -
- 0.2 (0.4)
Aeshnidae sp. 2 A 0.5 (0.5) -
- 0.2 (0.4) 2 (2) 2 (3) 8 (10) 5 (10) 2 (2) 1 (1.3) 1 (1)
Zygoptera
Coenagrionidae A - 0.2 (0.4) 0.2 (0.4) - -
0.2 (0.4) - 1 (1) 3 (5) 5 (7) 1 (1) 0.8 (1.2) 0.8 (1.3) 1 (2) 2 (3)

Figure 4.2. Nymph of the anisopteran dragonfly Pantalaflavescens.

12 Pond A
E 10 A Pond B
0 8
5 10 15 20 25 30 35
25
E 20
15
0.
E 10
5
0
0 5 10 15 20 25 30 35
50
T. L. Barra mu nd i (m m)
C
40
30
20
10
0 5 10 15 20 25 30 35
Days after filling
Figure 4.3. A. Density (mean number/in' of substrate) of Pantala flavescens

nymphs in two 0.1 ha freshwater ponds in north-eastern Queensland
during the first 35 days after filling, as determined by six 15 m tows
with a dredge net 0.46 m wide on each sampling occasion.
Total lengths (mm) of Pantala flavescens nymphs sampled in
freshwater ponds in north-eastern Queensland during the first 35 days
after filling. Data points are means and ranges; 28 specimens were
measured on day 7, 16 on day 10, and 100 on days 14 to 35.
Growth in length of barramundi fry of two size groups (10 mm
TL and 20 mm TL at time of stocking) in freshwater ponds in north-
eastern Queensland. The dashed line indicates the size (17-18 mm
TL) at which barramundi develop an escape response.

Table 4.4 Number of barramundi stocked into six 0.1 ha freshwater ponds,
number harvested as 40-50 mm TL fingerlings, density at harvest and
percentage survival for fish of two sizes at time of stocking. The
smaller fish averaged 9.9 mm TL at stocking and were harvested 27-
31 days later; the larger fish averaged 20.2 mm TL at stocking and
were harvested 14-17 days later.
9.9 mm TL at stocking 20.2 mm TL at stocking

Number stocked 4028 3889 1737 1737 3613 2172
Number harvested 577 646 793 835 3500 2162
Density at harvest 0.57 0.65 0.79 0.84 3.5 2.0
(No./m2)
Percentage survival 14.3 16.6 45.7 48.1 96.9 99.5
4.4. DISCUSSION
The different levels of predation by dragonflies on the two size-groups of barramundi
in this experiment concur with previously identified aspects of the behaviour of
juvenile barramundi (Chapter 3). The lack of an escape response in the small
barramundi (10 mm TL) and their night-time habit of resting on the substrate resulted
in considerable mortality at all densities of dragonflies, even in a comparatively short
exposure period of 20 hours. The impact of predation was less severe on the larger
barramundi (20 mm TL), which had a well-developed escape response to an attack by a
dragonfly nymph. The development of the escape response is associated with a range
of other dietary, behavioural and pigmentation changes which occur in barramundi at
17-18 mm TL (discussed in detail in Chapter 3).

Interestingly, the area occupied within the experimental aquaria did not differ for either
group of barramundi in the presence or absence of dragonfly nymphs. This is in stark
contrast to the sooty grunter larvae, which at night-time rested on the substrate in
control (zero dragonfly) aquaria, but schooled near the surface in aquaria containing
dragonflies. The sooty grunter is a freshwater fish and has obviously evolved in
habitats containing dragonfly nymphs, whereas the barramundi has a life history in
which post-larvae inhabit brackish waters (see Section 3.4), which are generally devoid
of dragonfly nymphs. It seems reasonable to assume that sooty grunter larvae have
developed a flexible antipredator avoidance behaviour, enabling them to recognise
dragonfly nymphs as predators and avoid the benthic areas occupied by the nymphs. I
have been unable to locate references to larval and juvenile fish exhibiting avoidance
responses to dragonfly nymphs; however, there is a wealth of literature on the converse,
namely odonate fauna exhibiting behaviours appropriate for occupying habitats
characterised by the presence or absence of insectivorous fish (see review by Johnson
1991).
Little can be inferred from the comparatively minor dragonfly predation on the
tadpoles. It is possible that the fish were preferred prey for the nymphs, or that the
tadpoles did not generally exhibit movements that elicited an attack response from the
nymphs. Other factors, such as predator—prey size interactions and density of tadpoles
(Sherratt and Harvey 1989), could also have contributed to the result.
There were more dragonfly nymphs of all species collected in Pond B than in Pond A
(Table 4.3). This was most likely because the substrate in Pond B was covered with a
dense growth of grass, providing a complex three-dimensional habitat and associated
fauna, whereas Pond A had a bare clay substrate. The effect of the black plastic placed
around the edges of Pond A in eliminating ovipositing habitat for endophytic species
was unclear. Only five specimens of endophytic species (Aeshnidae and
Coenagrionidae) (Williams 1980) were collected in Pond A, but they were also
comparatively uncommon in Pond B. Therefore, in the ponds at Walkamin at least, the
application of the plastic is essentially irrelevant in controlling dragonfly nymphs,
because the numerically dominant species (family Libellulidae) are exophytic (that is,
their eggs are dropped freely into the water).
Escapement or avoidance of the dredge net used to collect the samples was not
measured. Thus, the density of P. flavescens shown in Figure 4.3a is almost certainly
an underestimate of the real abundance in the ponds. The apparent decline in density
after day 14 could have been due to cannibalism, which has been shown to be a factor
influencing population densities in other dragonfly species (van-Buskirk 1989);
alternatively, the larger nymphs may have been better able to avoid the sampling
device.
Pantala flavescens was by far the most numerous of the odonate species sampled from
both ponds. P. flavescens is circumtropical, and is the most widely distributed of all
the odonate species (Lamb 1925). It is an opportunist and is known to colonise small,
temporary ponds (R. Rowe, James Cook University, pers. comm. 1989). In the present
study, it appears that only one cohort may have colonised the pond, as early instars
were never found in the samples collected during days 21-35 (Fig. 4.3b).
Alternatively, nymphs from the initial colonisation may have cannibalised larvae
arising from later hatchings, although if this was the case one would reasonably
anticipate that a small number of early instars would still have been collected in the
later samples.
The nymphs grew to an average size of 7 mm TL within 14 days and 20 mm TL within
35 days of filling the ponds (Fig 4.3b). Lamb (1929) reported the final instar of P.
flavescens to be 16.7-18.7 mm TL. This suggests that the larval stage in these tropical
waters was close to completion after 35 days, which is considerably faster than the 51
days reported for the same species in summertime in Victoria (Hawking and Ingram
1994).
Gonzalez and Leal (1995) reported that P. flavescens nymphs as small as 5 mm TL
could consume common carp fry, and that nymphs 5-20 mm TL were capable of
consuming fry slightly in excess of their own body length. Thus, both the size and
density of P. flavescens nymphs 14 days after filling the ponds would have been
sufficient to impact on barramundi if stocked at an initial size of about 10 mm TL.

The results of the trial on comparison of husbandry techniques clearly demonstrated
that survival of barramundi is enhanced if the fry are stocked at 20 mm TL rather than
10 mm TL. However, the conditions of this trial were such that it was not possible to
isolate the factor(s) responsible for the better survival of the larger fish. This is
demonstrated by the growth curves for the fish and development of the dragonfly
population (Fig. 4.3a, b, c). The dragonfly nymphs would have been abundant at the
time of stocking the small fish and for the subsequent 8-10 days during which the fish
would have grown to 17-18 mm TL. The barramundi would have been vulnerable to
predation by the dragonfly nymphs throughout this period, according to the results of
the aquarium trials (Fig. 4.1) (see also Gonzalez and Leal 1995). On the other hand,
the larger barramundi were stocked earlier in the pond cycle, and they were already
beyond the vulnerable stage (determined from the aquarium studies) at the time of
stocking. Also, the nymphs were also extremely small (<3 mm TL) and comparatively
few at the day 7-11 period (Figs. 4.3a, b). Thus, the influence of size at stocking
cannot be isolated from the influence of the different time of stocking. To obtain more
definitive information of the effect of dragonfly predation in ponds on the different-
sized barramundi it would be necessary to stock the two size groups into ponds filled
for the same preparatory period, say 14 days.
Nevertheless, it is still possible to devise pond-rearing strategies for barramundi fry,
based on ontogenetic changes in diet and behaviour of the fry and the development of
pond fauna (both food and predators). Zooplankton, in particular copepods and
cladocerans (Table 3.1), is the sole food for fry smaller than 18 mm TL. Insect larvae

are first eaten at 18 mm TL and become increasingly more important food items as the
fish grow; at about 30-45 mm TL zooplankton and insect larvae are equally
represented volumetrically in the diet (Fig. 3.2). Pond fauna studies at Walkamin
(unpublished data) have shown that there is a time lag of 10-14 days after filling a
pond before microcrustacea become abundant. The insect larvae, particularly
chironomids, are reasonably common 8-10 days after filling. Considering these
factors, as well as the growth rate of the fry (1-1.5 mm per day, Chapter 3), the
schedule depicted in Table 4.5 is suggested as a guideline for stocking barramundi fry
of various sizes into freshwater ponds for on-growing to fingerling size.
Table 4.5 Guideline for stocking barramundi fry into freshwater ponds for on-
growing to fingerling size (40-50 mm TL).

TL of fry Days after pond Rationale
filled
15 mm Not to be stocked into freshwater ponds.
Fry dependent on zooplankton, susceptible
to predation by dragonfly nymphs.
16-17 mm 10 Approximately 2 days before escape
response developed. Dragonfly nymphs
very small (2-3 mm TL).
18-19 mm 8-12 Escape response developed. Fry can eat
small insect larvae. Dragonfly nymphs
small (2-5 mm TL).
> 20 mm > 8 Diversity of food for fry. Fry not vulnerable
to dragonfly nymphs.
Control of air-breathing insects in fish-rearing ponds has been achieved using diesel or
oil to create a surface film through which the insects cannot penetrate (Brown and
Gratzek 1980). In the case of gill-breathing insects such as dragonfly nymphs,
insecticides (organophosphates and pyrethroids) have been used (e.g. Burleigh et al.
1993), but such chemicals have not been approved for use in aquaculture in Australia.
Moreover, the use of insecticides would be detrimental to the fish, as insecticides are
also toxic to microcrustacea, which are the major component of the food base for fish
larvae reared in ponds
In summary, the experimental results have clearly shown that barramundi up to 10 mm
TL are extremely susceptible to predation by nymphs of the dragonfly P. flavescens.
By 20 mm TL the barramundi are no longer susceptible, as at that size they have a
well-developed escape response. The escape response first becomes apparent at about
18 mm TL, in association with changes in behaviour, diet and pigmentation (Chapter
3). Under the trial conditions, the dragonfly populations in ponds at the Walkamin
Research Station peaked at about 14 days after the ponds were filled, and growth of the
most abundant species P. flavescens was rapid. Barramundi fry 20 mm TL show
significantly better post-stocking survival in freshwater rearing ponds than do 10 mm
TL fry. On-growing barramundi fry to fingerling size in freshwater ponds requires
management strategies to preclude predation by dragonfly nymphs.

Chapter 5. Photoperiod 119
CHAPTER 5. EFFECTS OF PHOTOPERIOD ON GROWTH AND
FEEDING PERIODICITY OF BARRAMUNDI FRY
5.1 INTRODUCTION
Barramundi have been routinely reared in hatcheries in several South-East Asian
countries since the mid-1970s, and consequently there is considerable published
information on rearing procedures (Chomdej 1986, Copland and Grey 1987, Parazo et
al. 1990). However, there is surprisingly little literature on the biology of larvae and
juveniles as it relates to survival and growth in hatcheries.
Several authors have shown that fish biology and behaviour are influenced by light.
Extending the photoperiod under which teleost larvae are reared affects growth and
survival, although the nature and extent of the effect varies between species (see, for
example, Marliave 1977, Tandler and Helps 1985, Duray and Kohno 1988). Pearce
(1991) reported that artificially extending daylight during larval rearing resulted in
increased growth due to greater food consumption, but that it had no effect on survival.
In Chapter 3 it was shown that barramundi fry are visual feeders, taking food
throughout the day, with a peak in feeding activity at dusk. The fry continue feeding at
a reduced level under moonlit conditions, but cease feeding in total darkness.
Based on this information, I hypothesised that artificially increasing the day length
during hatchery rearing may increase the growth rate of barramundi fry. Consequently,
trials were conducted to determine if photoperiod had any effect on the growth,
survival, feeding pattern and daily food consumption of barramundi fry reared in a
freshwater hatchery.
5.2.1 The fish
The fish used in this experiment were bred from eggs stripped from wild fish captured
on the spawning grounds in the Hay River, Weipa (12°34'S, 142°53'E). The larvae
were reared at the Northern Fisheries Research Centre, Cairns. At about 18 days old,
the fish were acclimated to fresh water over a 24 hour period and transferred to the
Walkamin Research Station. They were maintained in fresh water for at least 5 days
prior to starting the growth and survival experiment, and at least 25 days prior to
starting the feeding periodicity experiment. Two trials were conducted within each
experiment, using fish from separate spawnings.

5.2.2 Effect of photoperiod on growth
The first trial was designed to test the effect on growth of 12 hours light and 12 hours
dark (12L/12D) and continuous lighting (24L/OD) in combination with food being
available for either 12 or 24 hours. Treatments were as follows:
Treatment 1 12L/12D, food available for the 12 daylight hours
Treatment 2 : 12L/12D, food available 24 hours
Treatment 3 : 24L/OD, food available 12 hours
Treatment 4 : 24L/OD, food available 24 hours
The second trial was designed to test the effect on growth of 12L/12D (treatment 1),
18L/6D (treatment 2) and 24L/OD (treatment 3) light regimes, with food constantly
available.
All treatments were replicated 5 times, with 20 fish per replicate. Each replicate
consisted of an aquarium tank 900*350*300 mm 3, fitted with an independent biological
filter and aeration. Artificial shelter, in the form of 12 strips (30*400 mm 2) of black
plastic mesh (1 mm2) was suspended in each aquarium. Temperature was maintained
at 29.0 ± 1.0°C.
Lighting was controlled by covering each aquarium with black material and positioning
one 15 watt incandescent globe 25 cm above the surface of the water at the centre of
the aquarium. Automatic time switches turned lights on and off at either 0700 hours
and 1900 hours (Trial 1, treatments 1 and 2; Trial 2, treatment 1) or 0700 hours and

0100 hours (Trial 2, treatment 2) respectively. Light intensity at the water surface
varied from 300 lux at the ends of the aquaria to approximately 1100 lux directly under
the globes.
Equal aliquots of live zooplankton were added to each aquarium at 0700 hours and
1600 hours. The zooplankton was harvested from a pond using airlifts and a 250 mm
net. The amount fed was sufficient for excess food to be available continuously during
the feeding period. To preclude zooplankton being extracted from the aquaria by the
biological filters, inlet water to the filters was strained through 200 mm filter boxes. In
the 12 hour food-availability treatments in Trial 1, the inlets to the biological filters
were removed from the filter boxes at 1900 hours, which effectively removed all
zooplankton from the aquaria within one hour.
Fish were allocated randomly at the start of each trial, with a subsample retained for
weighing and measuring. Total lengths (TL) and wet weights (Wt) at the start of each
trial were as follows:
Trial 1 TL (meant s.d.) 11.9 ± 1.4 mm, range 10.0-16.1 mm, n = 46
Wt (mean f s.d.) 23.0 ± 8.7 mg, range 10.0-50.0 mg, n = 46
Trial 2 TL (mean ± s.d.) 11.0 ± 1.1 mm, range 9.1-13.0 mm, n = 50
Wt (mean ± s.d.) 21.3 ± 7.1 mg, range 9.0-36.0 mg, n = 50

Each trial was run for 13 days, after which all fish were weighed and measured. Two-
way analysis of variance (ANOVA) was used to analyse growth data in Trial 1, and
one-way ANOVA was used for growth comparisons in Trial 2.
In Trial 1, 10 fish from each treatment were killed by immersion in 0°C water and fixed
in 70% alcohol. Longitudinal section mounts stained with haematoxylin and eosin
were made and examined to determine if histological abnormalities were induced by
continuous lighting.
5.2.3 Effect of photoperiod on feeding periodicity
Two trials were conducted with different-sized fish to determine the patterns of food
consumption over a 24 hour period for fish exposed to 12L/12D and 24L/OD light
regimes. Approximately 80 fish for each treatment were placed into aquaria set up as
described above, and acclimated to the experimental conditions for 3 days prior to the
trials. The TLs of the fish used in each trial were as follows:
Trial 1 TL (meant s.d.) 33.5 ± 3.3 mm, range 24.7-39.9 mm, n = 64
Trial 2 TL (mean ± s.d.) 51.7 ± 5.0 mm, range 41.3-62.3 mm, n = 64
Excess live zooplankton was fed one hour prior to sampling, which was every three
hours for 24 hours. Eight fish per treatment were sampled on each occasion.
Immediately after sampling the fish were killed by immersion in 0°C water, and
preserved in 70% alcohol. Within one week, the stomach contents were removed, and
the dry weights of fish and stomach contents were recorded after drying at 60°C for 24
hours. The stomach fullness indexes for each sampling period were calculated from the
formula
Stomach fullness index = dry weight of stomach contents * 100

dry weight of fish
For the fish in Trial 1, the percentage body weight eaten per day was determined from
the formula
Daily ration = 24 hours average stomach fullness index * 24

number of hours to evacuate food from stomach
The number of hours for stomach evacuation for barramundi with a mean total length
of 34 mm was taken as 1.5 hours (results reported in Chapter 3). (The percentage body
weight eaten per day was not calculated for fish in Trial 2, because the stomach
evacuation rate for 52 mm barramundi is not known.)
5.3 RESULTS
5.3.1 Effect ofphotoperiod on growth
Final lengths, weights and survival of barramundi fry in Trials 1 and 2 are listed in
Tables 5.1 and 5.2 respectively. In Trial 1, analysis involved the partitioning of
treatment effects into photoperiod, food availability and interaction effects. The only
effect which was statistically significant was that of food availability. Having food
available for 24 hours increased length by 0.7 mm (P = 0.032) and increased weight by
19 mg (P = 0.052). Note that the latter falls just short of the conventional 5%
significance level. The lack of any interaction indicated that this increase was unrelated
to photoperiod. This was confirmed in Trial 2, in which there was no significant
difference in the length or weight of fish in the three photoperiod treatments.
Mean survival in Trial 1 was in the range 93-97%; in Trial 2 it was 66-76%. There
was no significant difference between the survivals in the various treatments within
either trial.
Histological examination of specimens from Trial 1 revealed no abnormalities or
lesions on any fish.
Table 5.1. Mean lengths (TL, mm), weights (Wt, mg) and percentage survivals
(and standard errors) of barramundi fry after being exposed to various
photoperiod and food availability treatments for 13 days. Initial size
was TL = 11.9 ± 1.4 mm, Wt = 23.0 ± 8.7 mg.
Treatment Mean TL ± s.e. Mean Wt ± s.e. % Survival

(mm) (mg) Mean ± s.e.
T1 = 12L/12D, food 12 hours 30.5 ± 3.2 396.5 ± 109.9 93 ± 5.7
T2 = 12L/12D, food 24 hours 31.0 ± 3.1 412.2 ± 113.6 95 ± 0.0
T3 = 24L/0D, food 12 hours 30.7 ± 3.0 400.6 ± 106.4 95 ± 6.1
T4 = 24L/OD, food 24 hours 31.6 ± 2.9 423.7 ± 107.8 97 ± 2.7

Table 5.2. Mean lengths (TL, mm), weights (Wt, mg) and percentage survivals
(and standard errors) of barramundi fry after being exposed for 13 days
to photoperiod regimes of 12L/12D, 18L/6D and 24L/OD, with food
continuously available. Initial size was TL = 11.0 ± 1.1 mm, WT = 21.3
± 7.1 mg.
Treatment Mean TL ± s.e. Mean Wt ± s.e. % Survival

(mm) (mg) Mean ± s.e.
Ti = 12L/12D 30.3 ± 4.5 385.8 ± 175.3 66 ± 11.4
T2 = 18L/6D 29.4 ± 4.1 337.7 ± 142.6 73 ± 10.4
T3 = 24L/OD 30.4 ± 4.2 374.6 ± 166.7 76 ± 2.2
5.3.2 Effect of photoperiod on feeding periodicity
The patterns of food consumption over 24 hours were different between the two
treatments, with similar patterns being exhibited by both small (Trial 1) and large (Trial
2) fish. Fish subject to a 12L/12D regime commenced feeding upon first exposure to
light and continued to feed at a high rate over the next 6 hours of the light period.
Following this, on the basis of stomach fullness, feeding slowed and tapered off during
the latter part of the light period (ie., after about 1500 hours). The advanced stage of
digestion of food in the gut after about 3 hours of darkness indicated that feeding had
ceased completely at the switch from light to dark conditions.
Fish exposed to continuous light also showed the same trend of decreased feeding after
1500 hours, and at 2100 hours all had empty stomachs. Thereafter, the fish started
feeding again (Fig. 5.1B), in marked contrast to those in the 12L/12D light regime.
The increase in feeding was particularly marked in the smaller fish, which consumed
more at the 2400 hour feed than at other times during the 24 hour cycle (Fig. 1B).
The 24 hour average stomach fullness index for the small fish (Trial 1) was 1.65 for the
12L/12D treatment and 2.28 for the 24L/OD treatment. Using these data, the daily
ration in the 12L/12D treatment was 26.4%; in the 24L/OD it was 36.5%. That is, the
fish exposed to continuous lighting consumed approximately 1.4 times more food than
the fish in the 12L/12D regime.
5.4 DISCUSSION
The results of this study clearly show that there is no advantage to be gained by rearing
barramundi fry under extended light regimes. While fry in extended daylight did not
show any adverse effects (in either survival or morphological development), they grew
no faster than those in normal daylength light regimes. This was despite the fact that
fry in 24 hours light consumed approximately 40% more food than did those in 12
hours light (with food continuously available in both cases). Presumably, the extra
intake under continuous light was expended as non-productive energy associated with
increased activity of the fish.
Barramundi fry reared in outdoor ponds have a distinct feeding pattern, with food being
consumed throughout the day and with a distinct peak in food intake at dusk. The fry

4
Mea n fu llness index
0300 0600 0900 1200 1500 1800 2100 2400 0300

Time of day (hours)
5
Mean fu llness index
0300 0600 0900 1200 1500 1800 2100 2400 0300
Time of day (hours)
Figure 5.1. Stomach fullness indexes for two size-groups of barramundi fry
exposed to 12L/12D (A) and 24L/OD (B) light regimes, with food
continuously available. Each datum is the mean ± s.e. (n = 8).
The solid bar indicates the period of darkness. ■ = fry TL
33 ± 3.3 mm (mean ± s.d.); o = fry TL 51.7 ± 5.0 mm
(mean ± s.d.).
do not feed in total darkness (results presented in Chapter 3). The experimental
protocol in the present study did not simulate dawn/dusk conditions; lighting was
turned on abruptly at 7.00 am and off at 7.00 pm. Nevertheless, it is instructive to
review the results in terms of a diel cycle, particularly with respect to the patterns
reported in Chapter 3. Peak feeding in the 12L/12D regime occurred about 6 hours
after first light (in the 'early afternoon'), with intake decreasing after that time and
ceasing in darkness. A similar 'afternoon' pattern was maintained by fish in
continuous light, with all fish having empty stomachs at 2100 hours. Prior to 2400
hours, however, feeding resumed in continuous light (Fig. 5.1B), in contrast to fish in
the 12L/12D cycle. The maintenance of a modified feeding pattern in continuous light
and in the absence of other external cues indicates that the feeding pattern is only partly
controlled by light, and that other, presumably innate or genetic, controls are involved.
The data presented in Figure 1B appear to contain an anomalous result, namely the
surprisingly high level of intake (stomach fullness index) exhibited by the smaller fish
in continuous light at 2400 hours, that is, immediately after the non-feeding period.
The larger fish resumed feeding at the same time, but did not show the same magnitude
of increase in feeding. Consequently, in the absence of further experimentation, the
datum point for 2400 hours for the smaller fish should be considered indicative of a
trend — resumption of feeding — rather than an indicator of a particular level of food
consumption.
The lack of growth response to the different light regimes contrasts with the results of
earlier work with barramundi larvae (Pearce 1991). He reported that during the rotifer
feeding stage (2-10 days old), larval growth was progressively faster under conditions
of 8, 16 and 24 hours light each day. The effect was less pronounced during the brine
shrimp feeding stage (8-20 days old), when larvae in 8 hours light grew more slowly
than in 16 and 24 hours light, but growth rates in 16 and 24 hours light were the same.
Thus, although continuous light may be advantageous for growth of larval barramundi
during their first eight days, as shown by Pearce (1991), the experiments conducted in
this study show that it becomes less important thereafter and has no effect after
metamorphosis (which is at 11 mm TL, or about 20-25 days of age). These findings
are similar to those for several other species reported upon, which also tend to show a
diminishing effect of continuous light with age (see Table 5.3). It appears that
extended periods of light during larval rearing are generally beneficial for growth, but
not necessarily survival. After metamorphosis, however, it seems that growth of
juvenile fishes is usually the same in both normal and extended light regimes.
Table 5.3. Literature reports on the effect of extended light periods on the growth and survival of larvae and juveniles of several species of fin
fishes. (Salmonid fishes are not included, because with these anadromous species the effect of photoperiod interacts with the state of
physiological development and time of year.)
Species Developmental Stage Treatment Response Reference
Siganus guttatus First-feeding larvae, 0- Continuous light vs natural Growth rate and survival better in continuous Duray and Kohno 1988
7 days light/dark cycle light.
Lates calcarifer Larvae, 2-20 days 12,18 and 24 hour light Days 2-10 (rotifer feeding) — growth Pearce 1991
periods significantly better in 24 hours light, survival not
significantly different.
Days 8-20 (Anemia feeding) — growth
significantly better in 16 and 24 hour light,
survival not significantly different.
Dicentrarchus labrax Larvae, 0-30 days 12,18 and 24 hour light Maximum growth at 18 hours light, maximum Barahona-Fernandes 1979
periods survival at 12 hours light.
Dicentrarchus labrax Larvae, 1-30 days 9 and 24 hour light periods Growth significantly better in 24 hour light; Cerqueira and Chatain 1991
survival and swim bladder inflation rate
significantly better in 9 hour light
Nautichthys oculofasciatus Larvae, 0-38 days 13 and 24 hour light periods; Survival in 24 hours light and simulated natural Marliave 1977
simulated natural photoperiod significantly better than in 13L/11D.
photoperiod (dawn/dusk plus
low intensity light at night).
Species Developmental Stage Treatment Response Reference
Sparus aurata Larvae, 0-70 days . 12 and 24 hour light periods Survival and growth best in continuous light. Tandler and Helps 1985
Solea solea Larvae and juveniles to 12,18 and 24 hour light Larvae — survival not significantly different, Fuchs 1978
3 months old periods growth better in 18 and 24 hours light.
Juveniles — survival and growth not significantly
different.
Mylio macrocephalus Larvae and juveniles 13,18 and 24 hour light Larvae — growth best under continuous light. Kiyono and Hirano 1981
0-108 days periods Juveniles — growth not significantly different.
Sebastes diploproa Juveniles, 30-55 mm 12 and 16 hour light periods. Growth rates better in 16 hour light. Boehlert 1981
standard length.
Chapter 6. Weaning 133
CHAPTER 6. OPTIMAL SIZE FOR WEANING BARRAMUNDI FRY
ONTO ARTIFICIAL DIETS
6.1 INTRODUCTION
The transition from live food to an artificial diet is an important period in the rearing of
fish larvae or fry. In general terms, the transition should be accomplished as early as
possible, because of the costs involved in live food production, while at the same time
avoiding excessive mortalities due to the fish being unable to adapt to the artificial
food. In this context, failure to adapt can be due to either a physiological problem (the
food is unpalatable or nutritionally unsuitable), a behavioural problem (the fish require
live prey to stimulate a feeding response) or a physical problem (the food is rejected
because of particle size or texture).
Literature reports have indicated that weaning barramundi onto inert foods (formulated
dry diets, or minced fish, prawns or meat) is generally commenced when the fish are
about 20 days old, or 10-12 mm total length (TL) (Awang 1987, MacKinnon 1987a,
Maneewong 1987, Tucker et al. 1988, Walford and Lam 1993). At this length,
barramundi have not completed metamorphosis (scales are first apparent when the fish
are 11 mm TL). None of these reports have indicated the degree of mortality
associated with weaning barramundi at this size.
Studies detailed in Chapter 3 have shown that barramundi undergo an abrupt change in
feeding habit at about 16-18 mm TL. At sizes less than 17 mm TL the fish are roving
zooplanktivores. When larger than 17 mm TL they change rapidly to a lurking predator
mode of feeding, eating small insect larvae in addition to zooplankton. The change in
diet is associated with changes in pigmentation and feeding behaviour. I hypothesised
that survival of fry through the weaning period would be enhanced if weaning is
commenced once the fish have adopted the lurking predator mode of feeding.
Consequently, an experiment was conducted to determine the ease of weaning and
survival of barramundi fry at various sizes between 12 and 20 mm TL when fed a
formulated dry diet.
6.2.1 Fish and facilities
The fish were obtained from hormone-induced spawnings of captive broodstock held at
the Northern Fisheries Centre, Cairns. The larvae were reared in salt water until
approximately 10 mm TL, when they were transferred to fresh water and transported to
the Freshwater Fisheries and Aquaculture Centre, Walkamin, where they were
maintained in 1500 L tanks at 26-30°C. Prior to the commencement of each trial, the
fish were fed live Artemia nauplii and zooplankton harvested from freshwater ponds.
The experiment was conducted indoors in steep-sided, conical-based tanks, 800 mm
deep and of 130 L capacity (Fig. 6.1). Water was exchanged at the rate of 4 L/min via
a central stand pipe which was screened to prevent the escape of fish. The
experimental temperature was maintained at 28.0 ± 0.5°C. Air was supplied via an air-
stone in the bottom of each tank. Photoperiod was 12L/12D. Artificial feed was
delivered by automatic feeders positioned over each tank.
6.2.2 Weaning procedure
Literature reports (Awang 1987, Maneewong 1987) have indicated that a gradual
transition from live to formulated feeds is appropriate for weaning barramundi. In the
absence of information to the contrary, it was assumed that a gradual transition would
be the procedure adopted for the size-at-weaning trial. Nevertheless, a preliminary
trial, conducted primarily to test facilities and procedures, was undertaken with the
secondary aim of determining whether a gradual or sudden transition was a more
appropriate procedure for weaning barramundi fry onto dry diets. Because it was
undertaken as a preliminary trial, replication of the sudden transition treatment was not
considered a priority.
The gradual transition treatment (T1) was frozen zooplankton (mixed taxa, obtained
from a freshwater pond) offered in conjunction with a dry diet (Ski -et-tinge 0.3-0.6 mm
salmon starter diet) for 5 days; during the 5th to 7th days of the trial the volume and
frequency of zooplankton offered was decreased; and from the 8th day the zooplankton
was discontinued and the dry diet only dispensed for a further 5 days. The sudden
transition treatment (T2) was the dry diet only fed from day 1 and maintained for 12
days. T1 was replicated 4 times, but T2 was not replicated (for reasons as explained
above). Each tank was stocked with 105 fish, TL 15.8 ± 0.7 mm, at the start of the
experiment. Feed was offered every hour for 12 hours (darkness one hour after the last
feed). The tanks were cleaned each morning by scrubbing the walls, turning off the air
and water and siphoning out the settled debris. Dead fish were collected and counted.
At Day 8 five fish were sampled from each tank and dissected to determine whether
they had commenced feeding on the dry diet. The fish were counted and measured
when the trial was terminated. Any difference between the number of missing fish at
the end of the trial and the number of dead fish recorded during the trial was assumed
to be attributable to cannibalism.
6.2.3 Size and weaning success

The experimental design consisted of 4 treatments replicated 4 times. The treatments
were fish of different initial TLs (meant s.d.) as follows (CV = coefficient of
variation);
Ti: 12.8 ± 0.94 'run, n= 53, CV = 7.3%
13.6± 1.36 mm, n= 71, CV = 10.0%
16.7 ± 1.42 mm, n= 50, CV = 8.5%
19.6±1.42 mm, n = 60, CV = 7.3%

Because of a shortage of facilities, it was not possible to test the four treatments
simultaneously on the same batch of fish. Thus, four trials (corresponding to the four
treatments) were conducted over a two-month period, using fish from different
spawnings.
Figure 6.1 Experimental arrangement, showing the conical rearing containers,

automatic feed dispensers and electronic control boxes for each
dispenser.
At the start of each trial, 200 healthy fish were placed into each of the four replicate
tanks and maintained on zooplankton for one day. Any mortalities during this period,
assumed to be due to handling, were replaced. Thereafter, feeding of zooplankton was
stopped and the dry diet (Skretting 0.3-0.6 mm salmon starter diet) was dispensed
every hour for 12 hours (darkness one hour after the last feed). At each feeding, 5 g of
food was dispensed, which was in excess of the requirements of all sizes of fish. The
trials were conducted for 10 days, which was sufficient time for all surviving fish to be
fully adapted to the dry diet. Tank maintenance and data collection were as above.
Survival data were analysed by analysis of variance (arcsin transformation of
percentage data), and means compared using the LSD(T) test (Siegel 1992). Non-
linear regression analysis was used to determine the relationship between percentage
survival through the weaning period and initial size. Total lengths were transformed to
equivalent wet weights (WWt) using the equation derived in Chapter 3 (WWt =
0.028TL2.82) and daily growth rates (G), expressed as percentages, determined from the
equation
G = (1nWWtf — hiWWt)/(tf —
where WW; and WWtf are the initial and final wet weight of fry and (tf — t) is the
duration of the experiment in days.

6.3 RESULTS
6.3.1 Weaning procedure
The survival and growth of barramundi fry in the gradual and sudden transition
weaning regimes are listed in Table 6.1. Survival was similar in all tanks, with the
exception of one replicate in the gradual transition treatment, in which cannibalism
accounted for most losses. The percentage of surviving fish which adapted to the
artificial diet varied from 90-100% in both treatments. In the sudden transition
treatment it was apparent from the behaviour and distended stomachs of the fish that
feeding on the artificial diet was established by Days 2-3. In contrast, in the gradual
transition treatment only 8 of the 20 fish sampled at Day 8 had started feeding on the
artificial diet. Growth was markedly superior in the sudden transition treatment, in
which the fry were three times heavier than those in the gradual transition treatment at
the end of the 12 day trial.
6.3.2 Size and weaning success
The percentage survival and the percentage of fish which died due to starvation and
cannibalism are listed in Table 6.2. Analysis of variance of the percentage survival
data indicated that there was a significant difference (P < 0.001) in survival between
the treatments. Least significant difference tests showed that the treatments separated
into two groups with significantly different (P < 0.05) mean survivals, namely T4 and
Table 6.1 Percentage of feeding and non-feeding barramundi, and final lengths
(TL, mm) and wet weights (WWt, mg) of the feeders, after 12 days
exposure to either a gradual or a sudden transition from frozen
zooplankton to an artificial diet. The figures are means ± standard
deviations.
Treatment Survivors Mortalities

Feeders Final TL Final WWt Non-feeders (%)
(%) (mm) (mg) (%)
Gradual
transition
Rep. 1 49 21.4 ± 3.4 170 ± 88 1 50
Rep. 2 88 20.4 ± 3.5 149 ± 70 2 10
Rep. 3 90 19.6 ± 4.3 140 ± 95 10 0
Rep. 4 93 19.8 ± 2.5 133 ± 50 7 0
Sudden
transition
Rep. 1 85 30.2 ± 5.0 448 ± 201 0 15
T3 being different from T2 and Ti. The relationship between TL and percentage
survival was described by the equation
y = 96.85 - 155981(0.540L) (r2 = 0.88, P < 0.01)
where y = percentage survival and L = TL in mm (Figure 6.2).
Survival within treatments was uniform, with the exception of the fourth replicate in
T2, wherein the survival was 79% compared with 46%, 51% and 55% for the other
replicates (Figure 6.2). There was no obvious explanation for the higher survival in
this one replicate. Daily growth rates varied between 11.1% and 17.7% in the four
treatments (Table 6.3), with the lowest rates being for T1 and T4 (the smallest and
largest fish, respectively) and the highest for T2 and T3 (the intermediate sized fish).
The initiation of feeding was evident by day 3 in all treatments. At this stage visual
observation indicated that about 5-10% of the fry had started feeding in T1, about 30-
40% in T2, and the majority in T3 and T4. Fish in each tank congregated as a loose
school near the water surface awaiting food, but once it was dispensed they fed
throughout the water column.
Weaker, non-feeding fish were darkly pigmented. In T1, there was a marked increase
in mortality of these fish during days 5-7, peaking at day 6. There was little mortality
thereafter, as virtually all remaining fish were feeding. There was no similar peak in
mortality of non-feeding fish in the other treatments. Cannibalism was significantly
different (P < 0.001) in T2 (26%) to that recorded in the other treatments (2-7%).
The aeration systems were effective in lifting the sinking food particles back into
suspension, thus giving the fry longer exposure to the food at each feeding period than
would have been the case if the food was allowed to fall through the water column to
the base of the tanks. A slimy, organic film developed each day on the walls of the
tanks, as a consequence of the high organic load. The circular current swept food
particles onto the film, where they attached, effectively becoming unavailable to the
fry. This effect became more pronounced as the film redeveloped after each daily
cleaning.
' Chapter 6. Weaning 142
Table 6.2 Survival and mortalities due to starvation and cannibalism (expressed as
percentages) of barramundi fry of various sizes during 10 day weaning
trials. The figures are means ± standard errors.
Treatment Initial TL % Survival % Mortalities due to

(mm) Starvation Cannibalism
T1 12.8 ± 0.9 38.9 ± 5.8 59.8 ± 5.4 2.8 ± 2.6
T2 13.6 ± 1.4 57.6 ± 14.6 16.8 ± 11.0 25.6 ± 7.7
T3 16.7± 1.4 97.0± 3.8 0.8 ± 0.8 2.3 ± 3.8
T4 19.6 ± 1.4 92.3 ± 4.0 1.0 ± 0.6 6.9 ± 4.6
Table 6.3 Initial and final total lengths (TL, mm) and computed initial and final
wet weights (WWt, mg) and daily growth rates (expressed as a
percentage based on weight) of barramundi fry of various sizes during
10 day weaning trials. The figures are means ± standard errors.
Treatment Initial TL Final TL Initial WWt Final WWt Daily

growth rate
(mm) (mm) (mg) (mg) (%)
T1 12.8 ± 0.9 19.9 ± 0.5 36 ± 9 128 ± 10 12.7 ± 0.7
T2 13.6± 1.4 25.1 ± 1.3 42± 10 250 ± 36 17.7± 1.4
T3 16.7 ± 1.4 31.3 ± 0.3 80 ± 18 463 ± 12 17.6 ± 0.3
T4 19.6± 1.4 29.1 ±0.2 123 ± 20 375±9 11.1± 0.2
143
Chapter 6. Weaning
100
Perce ntag e survival
80
60
40

12 14 16 18 20
Total length (mm)
Figure 6.2 Percentage survival through weaning onto dry diets for four different
size-groups of barramundi fry.
6.4 DISCUSSION
There are a variety of protocols that can be used for weaning fish onto artificial diets,
but generally a gradual transition from live food to formulated feed or minced fish has
been advocated for barramundi (Awang 1987, Maneewong1987). However, the
appropriateness of this procedure was not vindicated by the results of the weaning
procedure trial. Growth was markedly superior in the fish exposed to the sudden
transition to artificial food, although the percentage of fish which were successfully
weaned did not differ between the sudden and gradual transition treatments. In the
sudden transition treatment, in which there was no alternative food source, the fish had
commenced feeding on the artificial diet within 2-3 days of starting the trial. In
contrast, in the gradual transition treatment the fish apparently consumed the frozen
zooplankton in preference to the artificial food when the two were supplied
simultaneously. This observation was supported by the low percentage (40%) of fish
feeding on the day after the zooplankton feeding was discontinued. That is, weaning
effectively commenced only after the zooplankton feeding was stopped and the
artificial diet was the sole food being offered.
The experimental design used in the size and weaning success trial was compromised
by the fact that the treatments were tested on batches of fish from different spawnings.
It is a common phenomenon in fish culture that larval fish originating from different
parents, or even from the same parents but different spawning events, may exhibit
different levels of fitness (often referred to as 'larval quality'). Quality is usually

appraised in terms of vigour (swimming strength, ability to orientate, response to
stimuli) and health (absence of mortalities, disease and deformities). To avoid
`experimental noise' due to variation in fitness, researchers generally aim to test
treatments on high-quality fish from the same spawning. In the present trial, it was not
possible to use fish from the same spawning because of a shortage of experimental
tanks and automatic feeders. Nevertheless, I am confident that the survivals determined
in the different treatments faithfully reflect the influence of initial size, because of the
consistency within treatments and similar survivals shown in other weaning events
which were not part of this experiment.
The survival curve for barramundi fry presented in Figure 6.2 indicates that greater
than 90% survival during weaning onto dry diets can be expected when using fish
larger than 16.2 mm TL at the initiation of weaning. The shape of the curve confirmed
that survival through the weaning period is enhanced if weaning is delayed until the
fish have adopted the lurking predator mode of feeding. This is also supported by the
increasing rapidity with initial size with which feeding on the dry diet was commenced
by the majority of fish. The asymptotic shape of the curve indicates that there is
negligible gain in survival if weaning is initiated with fish larger than about 16-17 mm
TL. The results of this experiment are similar to those of Verreth and van Tongeren
(1989) who showed that larvae of the African catfish, Clarias gariepinus, could be
successfully adapted onto dry diets after two days feeding on live foods, and that
delaying weaning beyond that stage did not improve survival.

Cannibalism was more severe in T2 than in the other treatments. This was possibly
due to the comparatively wide size range of fish in this treatment at the start of the trial
(coefficient of variation for T2 was 10%, while for T1, T3 and T4 it was 7.3%, 8.5%
and 7.2%, respectively). Moreover, the rapid adaptation of fish in T3 and T4 onto the
dry diet would have militated against the establishment of cannibalism.
The daily growth rates of 11-18% for fish in the size and weaning success trial are
comparable with the 13-15% reported in Chapter 3 for fry reared intensively in the
laboratory on live zooplankton. This indicates that with respect to growth the fish were
not compromised by the change to, or nutrient composition of, the dry diet. The
comparatively slow growth rate of the largest fish (T4) could not be explained by any
obvious physical factors; one possible explanation is that it was a consequence of this
batch of fish being genetically less fit than the others.
Rearing of barramundi fry on live foods is a costly process, as it requires major
commitment of both manpower and facilities. Hence, hatchery operators generally
attempt to wean the fry onto artificial diets as soon as practicable, which is at about 10
mm TL (Awang 1987, MacKinnon 1987a, Maneewong 1987, Tucker et al. 1988). This
study has shown, however, that survival through the weaning period is considerably
enhanced if the initiation of weaning is delayed until the fry are 16-17 mm TL.
Hatchery operators need to interpret these results from a commercial perspective.
However, it is apparent that the economic advantage of survival of say 90% with 16-
17 mm TL fry would outweigh the reduced rearing costs associated with weaning fish
at 13 mm TL and concomitant survival of say 50%. Indeed, a steady-state comparison
of hatcheries producing 530 000 weaned fingerlings per season indicates that there is
an overall benefit of approximately $70 000 per annum in weaning at 17 mm compared
with 13 mm TL (Appendix 1). The disparity would be even greater if weaning is
initiated when the fry are 10 mm TL.

Chapter 7. Circulus spacing 148
CHAPTER 7. USE OF CIRCULUS SPACING ON SCALES TO
DISCRIMINATE BETWEEN HATCHERY AND WILD
BARRAMUNDI
7.1 INTRODUCTION
Stocking hatchery-reared fingerling fish is a widely practised management technique
for enhancing recreational fisheries' stocks and restoring populations of endangered
species. Determining the efficacy of such programs depends on recognising the
hatchery fish after stocking. In the past, fingerling fish have been marked by
mutilation or excision of fins, by insertion of micro-wire tags, and with chemicals
(for instance, stains, dyes and tetracycline antibiotics) applied either externally or
internally (Wydoski and Emery 1983). These techniques are subject to error from
mortality and tag loss and are quite labour intensive.
As early as 1913 it was realized that growth patterns on scales could be used to
identify races of salmonids (Gilbert 1913, referenced in Henry 1961). Subsequent
studies showed that the rearing environment (analogous to the racial history)
effectively induced an innate tag, as a consequence of the high correlation between
environmental conditions, growth rates of fish and scale growth (Henry 1961, Major
et al. 1972, Doyle et al. 1987). The features on the scales reflecting growth rates are
circuli, which are fine ridges laid down in a circular pattern around the centre or
focus of scales (Jearld 1983). Later developments led to methods for distinguishing
between populations of various salmonid species by analysing patterns of scale shape
and circulus spacing using discriminant function techniques (Amos et al. 1963, Cook
and Lord 1978). These methods have recently been shown to be capable of
distinguishing hatchery and wild striped bass Morone saxatilis stocks, and even
assigning cultured fish to their hatchery of origin (Ross and Pickard 1990). The
development of specific software applications has facilitated use of the methods,
which require extremely sensitive measurements of shape and distance.
The aim of the present study was to determine if patterns of circulus spacing on
barramundi scales could be used to distinguish hatchery and wild fish. If valid, this
discrimination technique would be of considerable benefit to the management of
barramundi in areas where hatchery fish are released to supplement wild stocks.
7.2.1 Scale origin and preparation
The wild fish were obtained from the Cairns region (17°S, 146°E) northern
Queensland. A scale set was obtained from approximately 120 fish collected
between August 1979 and May 1980. The total lengths (TL) of the fish ranged from
150 to 450 mm.
The hatchery fish were bred at the Northern Fisheries Centre (NFC), Cairns, in
November 1988. Larval rearing was conducted in salt-water tanks at the NFC. When
20 days old (spawning = Day 0, hatching = Day 1), the fish (about 10 mm TL) were
transferred to freshwater at the Walkamin Research Station. They were reared
indoors until approximately 20 mm TL, then stocked into a pond. Scale samples
were collected from 100 fish (range 120— 220 mm TL) in March 1989.
Scales from both wild and hatchery fish were taken from the region immediately
posterior to the pectoral fin. Scales were washed in water and 4-6 non-regenerated
scales from each fish were mounted between glass microscope slides for examination
and analysis.
7.2.2 Data acquisition
Data were acquired using the Optical Pattern Recognition System (OPRS) software,
microcomputer, video camera and monitor, frame grabber, digitiser pad and mouse
(Biosonics 1989). The video camera was attached to a microscope (*2 objective).
The frame grabber transformed video images of scales to digital images, which were
displayed on the monitor. Linear distance on the image display was calibrated with a
stage micrometer. All measurements were obtained using the mouse and menu-
driven software.
The cleanest scale from each set was chosen for image analysis. Measurements were
conducted along two 1 mm lines (Fig. 7.1), creating two data sets. The first set,
herein termed straight-line data, was derived along a line located adjacent to the
radius, where the circuli curve and are most widely spaced. The other, herein termed
450 reference-line data, was from a line at an angle of 45° to the first, or at
approximately 90° to the anterior-posterior axis. Because the distance from the
centre of the focus to the first complete circulus varied between scales, the origin of
the measuring line was located just inside the first circulus. The program
automatically located the circuli by differences in luminance, although manual
control was necessary to correctly mark circuli with inadequate contrast between
light and dark zones, such as at overlapping or broken circuli.
7.2.3 Circulus formation and fish length
To examine the relationship among circulus formation, TL and age, 10 scales from
each of 43 hatchery-reared fish were examined. Circuli were counted on each scale
from the focus to the margin adjacent to the radius. The fish were 25 to 38 days old
and from 11 to 40 mm TL (scales had not formed in fish smaller than 11.0 mm TL).
Because of the small size of these fish, it was not possible to take the scales from a
confined area on the body.

Figure 7.1. Diagrammatic representation of a barramundi scale. Straight-line data

were derived along line a and 45° reference-line data along line b.
7.2.4 Data analyses
Discriminant function analysis developed by Cook (1982) and Cook and Guthrie
(1987), and available as part of the OPRS software package (Biosonics 1989), was
conducted to separate hatchery from wild barramundi. The procedure uses a
jackknifing technique, termed 'leaving-one-out' by Lachenbruch (1967, 1975), to
calculate the discriminant function. This iterative procedure uses the total data set as
the training and test sets and produces better estimates with less bias than other
commonly used techniques. Discriminant function coefficients calculated for all
variables in each type of analysis were the basis for selecting variables to enter into
the discriminant function analysis. Results of the analysis were displayed as stock
separation estimates in an error rate classification table.
The relationship between the number of circuli, TL and age was evaluated using
regression analysis. The number of circuli used in the analysis was taken as the
mean of the number of circuli on the 10 scales taken from each fish.
7.3 RESULTS
Initial screening of the scales from the wild fish showed that scales from fish larger
than approximately 350 mm TL were unreadable with the equipment available. The
thickness of the scales and the irregularly positioned pigment spots interfered with
light transmittance, effectively disrupting the pattern of circulus spacing. After
rejection of the unreadable scales, analysis was conducted on a set of 88 0+ wild fish
(150-350 mm TL). The set comprised of 68 fish from the 1978-79 spawning season
and 20 fish from the 1979-80 spawning season. Determining the position of circuli
on scales from wild fish smaller than 350 mm TL and from all the hatchery fish
(120-220 mm TL) was readily accomplished.
The basic measures analysed for these scales were various sets of single, double and
triple inter-circulus distances. For the paired circulus measures, distances between
circuli were combined two at a time beginning at the centre of the scale (e.g. pair 1 =
(1+2), pair 2 = (3+4), and so on). With triplet measures, the circulus distances were
combined three at a time, from the centre.
Separation of hatchery from wild barramundi scales was examined using
combinations of variables that displayed high negative or positive unstandardised
discriminant coefficients. The following combinations of circulus distances were
tested.
Straight-line data
A/ Singles 13, 14, 15, 16, 18
B/ Pairs (7-8), (9-10), (13-14), (15-16)
C/ Triplets (1-3), (4-6), (7-9), (13-15), (16-18)
45° Reference-line data
D1 Pairs (1-2), (9-10), (13-14)
E/ Pairs (1-2), (3-4), (9-10), (13-14)
F1 Triplets (1-3), (7-9), (13-15), (16-18)
The straight-line luminance data provided better separation than did the 45°
reference-line data, particularly when using triplets and pairs (Table 7.1). The
straight-line data were also easier to derive, as the positioning of the line, adjacent to
the radius, was simple, and the spacing of the circuli was clearer and with less
overlapping than along the 45° reference-line.
Using the triplet combination, 83% of the hatchery scales and 82% of the wild scales
were correctly classified (17 and 18% errors, respectively). The paired circulus
combinations were separated with 83% correct for hatchery scales and 77% correct
for wild scales. The classification rates for the 45° reference-line luminance data
ranged from 74-77% of hatchery scales classified correctly to 74-76% of the wild
scales correctly scored.
Table 7.1. Percentage of wild and hatchery barramundi correctly identified using
data derived from the spacing of circuli on scales and linear discriminant
analysis.

Test Wild Hatchery Composite
(n = 88) (n =100)
Straight-line data
Test A (Singles) 68 68 68
Test B (Pairs) 77 83 80
Test C (Triplets) 82 83 83
45° reference-line data
Test D (Pairs) 74 77 76
Test E (Pairs) 75 77 76
Test F (Triplets) 76 74 75
Multiple linear regression analysis revealed that 94% of the variation in the number
of circuli was attributable to TL (P < 0.001, partial r2 = 0.9413). Adding age to the
equation did not improve its descriptive power (P > 0.05, partial r2 = 0.0047). The
relationship between the number of circuli and TL was described by the linear
equation
y = —2.64 + 0.63x,
where y = number of circuli and x = TL (mm) (Fig. 7.2).
Figure 7.2. Relationship between the number of circuli on scales and TL of 43

fingerling barramundi 25-38 days old. Each datum is the mean of 10
scales.
7.4 DISCUSSION
The separation of hatchery from wild scales achieved with these data (83% for
hatchery and 82% for wild) was encouraging, especially since only luminance line
data were used. This compares favourably with other stock discrimination
investigations using the same technique. For instance, Schwartzberg and Fryer
(1993) separated natural and hatchery stocks of chinook salmon Oncorhynchus
tshawytscha with 80-94% accuracy; Cook and Lord (1978) separated three riverine
stocks of sockeye salmon Oncorhynchus nerka with 64-80% accuracy; Whaley
(1988) separated a lake strain and riverine strain of brown trout Salmo trutta reared
under virtually identical hatchery conditions with 94-98% accuracy. Ross and
Pickard (1990) used scale shape information analysed as Fourier transformations in
addition to variables associated with circulus spacing to obtain 87-91% accuracy in
separating yearling hatchery and wild striped bass stocks.
It is noteworthy that certain of the circulus distances repeatedly occurred in the sets
of variables tested. For example, circuli 13 and 14 occurred in all of the 6 variable
sets analysed and circuli 15 and 16 were found in four of the six. Circuli 13-16
would have been formed when the fish were 25-30 mm TL, which would have
coincided with the first 5-10 days after the hatchery fish were transferred from the
laboratory to the ponds. This suggests that hatchery fish might have been
incidentally marked during that portion of the rearing procedure.

Initial formation of scales in barramundi occurred at about 11 mm TL, and thereafter
circulus formation was related to length. In walleye Stizostedion vitreum squamation
occurs at about 20 mm TL (Glenn and Mathias 1985), while for three species of
salmonids it occurs at 37-42 mm TL (Bilton 1988). The comparatively small size at
which barramundi forms scales, coupled with the fact that juveniles can be reared
over a wide range of temperatures (at least 24-32 0C), indicates that barramundi may
be particularly amenable to batch marking of scales.
One obvious problem with barramundi, however, is that circuli on scales from fish
larger than 350 mm TL could not be read with the present methods. Only minimal
scale preparation — washing in water — was used in this study. Whaley (1991) has
shown that the use of pancreatin (an enzyme that breaks down protein substances
adhering to scale surfaces) in conjunction with sonic agitators increased the
percentage of readable scales in a rainbow trout Oncorhynchus mykiss population
from 4% for uncleaned scales to 70% for cleaned scales. Similar techniques may
also improve the readability of scales from barramundi larger than 350 mm TL,
although it is likely that the absolute thickness of barramundi scales will remain a
problem for analytical techniques dependent on light transmittance close to the focus
of scales. These older scales are thicker through the focus than towards the margins,
and this made it difficult to focus the microscope on all circuli at once. Other
workers have avoided similar problems with salmonids by making impressions of

scales on preheated plastic or acetate slides and reading intercirculus measures
directly from the impressions (e.g. Lear and Misra 1978, Fisher and Pearcy 1990).
The difficulty experienced in reading circuli on scales from large barramundi does
not apply to similar-sized fishes with smaller scales, such as salmonids. In Australia,
species with suitably small scales include silver perch Bidyanus bidyanus, golden
perch Macquaria ambigua, Australian bass Macquaria novemaculeata, sand whiting
Sillago ciliata and dusky flathead Platycephalus fuscus. All of these species are
currently used to enhance recreational fisheries, and patterns of circulus spacing are
being employed for monitoring stocking success (Willett 1994a, 1995, 1996; Palmer
1995)
If it proved possible to induce permanent tags in the patterning of the circuli on
scales of hatchery-reared fish, many of the questions associated with the assessment
of stocking programs (in particular, management techniques, estimates of post-
stocking survivals and population sizes) could be more readily answered. Overseas
researchers have shown that manipulating the rearing environment can induce
specific marks on fish scales and otoliths. For instance, Skurdal and Andersen
(1985) have demonstrated that scales of brown trout can be marked by varying the
temperature during egg incubation and rearing of the alevins, and Volk et al. (1987)
showed that sudden temperature shifts can induce banding on otoliths of juvenile
chum salmon Oncorhynchus keta.

Similar research has recently been taken up in Australia as part of the Queensland
Government's recreational fisheries enhancement programs. Willett (1993) showed
that pond-reared silver perch from three hatcheries in south-east Queensland could be
distinguished with accuracies in the range 56-63%. Removing the geographically
intermediate hatchery increased the accuracy to 85-91%, leading him to conclude
that better discrimination is achieved if the environmental differences (in this case,
temperatures) are well defined. He followed this up with a laboratory-based study
(Willett 1994b) in which he grew silver perch fry for 4 weeks at temperatures of 20°,
25° and 30°C. Classification of the scales based on circuli spacing was 54-58%, 58-
72%, and 94-96% accurate at each of the respective temperatures. The reduced
classification accuracy at 20°C and 25°C was a consequence of similar growth rates,
which in turn resulted in some overlap of lengths of fish from the two treatments. He
speculated that simultaneously altering other variables (for example, feed ration) to
induce disparate growth rates would further enhance the ability to batch-mark
hatchery-reared fishes (Willett 1994b).
Since the present study was restricted to investigation of scale circulus pattern
analysis as a means of discriminating hatchery from wild barramundi, only early
growth zones were examined. Other workers have shown that analysis of the shapes
of whole otoliths or scales can also be used to differentiate stocks or genetic races
(Jarvis et al. 1978, Riley and Margraf 1983, Bird et al. 1986, de Pontual and Prouzet
1987, Maceina and Murphy 1989, Campana and Casselman 1993, Margraf and Riley
1993). In Australia, the barramundi resource consists of a series of genetically

distinct stocks (Shaklee and Salini 1985, Russell and Garrett 1988, Keenan 1994).
Scale or otolith shape analysis (also a function within OPRS software) may provide
an alternative means of separating these stocks or delineating the composition of
other multi-stock assemblages.
In summary, hatchery-reared and wild barramundi could be distinguished by the
pattern of circulus spacing near the origin of the scales. The technique, which relies
on luminance data generated from the light and dark zones created by the circuli, was
suitable for barramundi less than 350 mm TL. However, with the technique
employed herein, scales from fish larger than 350 mm TL were too thick and
pigmented to provide reliable luminance data. Consequently, analysis of circulus
patterns on scales has only limited applicability in assessment of fisheries stocked
with barramundi, unless methods are developed to clean larger scales and thus
improve readability. On the other hand, it is an extremely versatile tool for
separating mixed stocks of species that have small scales as adults.

Chapter 8. Synthesis 162
CHAPTER 8 SYNTHESIS
8.1 INTRODUCTION
This study evaluated the proposal to introduce the Nile perch, Lates niloticus, from
Africa, and examined aspects of rearing juvenile barramundi, Lates calcarifer, for
aquaculture and enhancement of recreational fisheries in Australia. It is now
appropriate to review the results in terms of their overall contribution in the
following fields:
management of the barramundi resource, aquaculture development and
recreational fisheries — because of the applied nature of the study, most
emphasis is placed on this area;
knowledge of the biology of barramundi; and
theoretical aspects of fish conservation, fish propagation and fisheries
enhancement.
Finally, within the context of the above, some future areas for research are identified.
Much of the following discussion is summarised in Table 8.1, in which the outcomes
of the study are tabulated within the generic headings of management, biology,
theory and future research.

8.2 NILE PERCH STUDIES (CHAPTER 2)
The introduction of Nile perch into Lake Victoria has had dramatic consequences for
the ecology of the lake and its riparian inhabitants. Prior to the introduction, Lake
Victoria (and Lake Malawi) contained the most species-rich lacustrine fish faunas in
the world. As a result of the introduction, over 200 of the original 300+ (possibly
400+, see Seehausen and Witte 1995) endemic haplochromine species have
disappeared from Lake Victoria in less than a decade (Witte et al. 1992a).
Goldschmidt et al. (1993) described this as the largest extinction event amongst
vertebrates this century. The massive reduction in biodiversity has led to
international campaigns to conserve the native fishes of Lake Victoria (Ribbink
1987, Bruton 1990, Kaufman 1992), but such activities cannot realistically be
expected to contribute to maintenance of biodiversity within the lake because of the
scale of the problem and the difficulties associated with reintroductions (Craig 1992).
While the impact on the fish fauna was to a large extent predicted by early opponents
of the introduction (e.g. Fryer 1960, Anderson 1961), there have been flow-on effects
which were not foreseen (Witte et al. 1995). Diets and feeding behaviour of
piscivorous birds (Goudswaard and Wanink 1993, Wanink and Goudswaard 1994)
and otters (Kruuk and Goudswaard 1990) have changed. Drastic shifts in the food
web within the lake have taken place and continue to do so (Ligtvoet and Witte
1991), and may in fact be linked to apparent water quality changes, particularly
increasing algal blooms (Goldschmidt et al. 1993, Ochumba 1995). The fish harvest
from the lake has increased 3-4 times since the establishment of the Nile perch
(Ligtvoet and Witte 1991), but it is currently not clear if it is sustainable at the
present level. Indeed, recently available information indicates that the yield may
already be decreasing (see Tables 2.4 and 2.5, also Kunhongania and Chitamwebwa
1995). Pitcher and Bundy (1995) analysed historical catch-per-unit-effort data for
the lake and concluded that if the fishery continues to expand in the 1990s as it did in
the 1980s it will collapse as a consequence of overfishing rather than ecosystem
instability. Obviously, research on trends in the fishery and the continuing shifts in
trophic relationships is required to fully assess the impact of the introduction.
Socio-economic changes have also been far-reaching, although the pros and cons of
these are still being debated. The early views on disruption to riparian communities
and deforestation associated with fuel-wood demand for curing Nile perch (Barel et
al. 1985) were countered by others arguing the economic benefits of development of
the fishery (Acere 1988), which are considerable. Reynolds et al. (1995) calculated
that the net economic benefit to the riparian states of the post-Nile perch fishery was
US$200 million up to 1989, and that it would accrue to a billion dollars by the early
2000s. On the other hand, Harris et al. (1995) pointed out that much of the foreign
exchange earnings actually went to foreign companies supplying the gear and
infrastructure to support the fishery. In addition, they identified many other negative
impacts at the household, community and national levels.

Extinction of fish species in lakes as a result of the introduction of exotic fish has
been recorded in other areas, such as Lake Kyoga in Uganda, Lake Lanao in the
Philippines, Lake Atitlan in Guatamala, Lake Luhondo in Ruanda, and Gatun Lake in
Panama (see review in Witte et al. 1992a). In Australia, local exterminations of
fishes (and possibly crayfish) have occurred in Lake Eacham as a consequence of the
introduction of non-endemic Australian fishes (Barlow et al. 1987). In Lake Pedder,
Tasmania, two indigenous galaxiid fishes have been extirpated as a result of
interaction with brown trout (Salmo trutta) and the introduced Galaxias brevipennis
(S. Chilcott, Inland Fisheries Commission, Tasmania, pers. comm. 1996). These
scenarios, while not as dramatic as the Lake Victoria situation, further illustrate the
possible consequences of introductions of fish to exotic environments.
From a management perspective, the approach taken in this study of marrying the
biology of a species proposed for introduction with information on the ecology of
recipient environments enabled determination of the probable range and possible
impacts of the Nile perch had it been released in Australia. Based on the information
presented, it is unequivocal that the release of Nile perch in Australia posed an
unacceptable risk to aquatic ecosystems over a large part of the country. The
alternative approach of enhancing recreational fisheries through the breeding and
release of fingerling barramundi was clearly a better management option.

8.3 REARING JUVENILE BARRAMUNDI (CHAPTERS 3-6)
The trials on feeding behaviour and diet of barramundi fry undertaken in this study
have provided valuable new information relevant to hatchery management, as well as
contributing to knowledge on the biology of the species. In management terms, the
most important result was the identification of the change in diet and behaviour at
about 17 mm TL, as this change had implications for pond-rearing and weaning
procedures. To reiterate, barramundi smaller than 17 mm TL are zooplanktivores,
exhibit no escape response, and rest on the substrate at night. Larger than 17 mm TL,
they consume larval insects in addition to zooplankton, and they have a well-
developed escape response. The vulnerability of fish smaller than 17 mm TL to
predation by dragonfly nymphs was demonstrated in aquarium trials, as was the
ability of fish larger than 17 mm TL to avoid predation. The responses of both size
groups were mirrored in pond-rearing trials, in which fish stocked at 10 mm TL
showed very poor survival (14%), while those stocked at 20 mm TL exhibited 99%
survival.
The feeding behaviour of barramundi fry also had practical implications for weaning
onto dry diets. The change in feeding behaviour at about 17 mm TL indicated that
weaning may be more successful if initiated at that stage, rather than with fish of
approximately 10 mm TL (the usual size for initiation of weaning reported in the
literature). Trials conclusively demonstrated an asymptotic response in weaning
success with respect to size, with the asymptote at 16-17 mm TL.

The information on diet and behaviour of the barramundi fry, and its practical
application in hatchery management, highlight the importance of basic biological
studies for determining appropriate techniques for aquaculture of a particular species.
In the present case, the susceptibility of barramundi fry to predation by dragonfly
larvae meant that traditional freshwater pond rearing techniques did not work.
Research was required to elucidate the necessary modifications to rearing procedures
to successfully marry the biology of the species with the ecology of the rearing
environment. Similarly, species-specific larval production requirements have been
shown in the aquaculture of other species. For instance, initial swim-bladder
inflation in Australian bass Macquaria novemaculeata is about 70% under conditions
of darkness, salinity greater than 25 ppt and low to zero aeration, but only
approximately 15% in a 12:12 light:dark photoperiod; inflation is also reduced at
salinity of 10 ppt and under high aeration (Battaglene and Talbot 1990). Murray cod
Maccullochella peeli ovulates within a very narrow, defined period of time after
hormonal injection, and fertilisation must then be accomplished within 1-2 hours of
ovulation (Rowland 1988). These examples further illustrate the need for basic and
species-specific research in aquaculture, particularly for species for which little
biological information is available.
The gut evacuation rates for barramundi fry indicate a functional relationship
between feeding strategy and prey availability. The small, obligate zooplanktivores
have a fast rate of passage under conditions of continuous feeding, but a

comparatively slow rate when not feeding (i.e. they are numbers maximisers).
Larger fry, at a size capable of eating larval insects, have a more constant rate of gut
passage whether feeding continuously or on single meals (i.e. they are energy
maximisers). It would be beneficial to extend the gut passage and energetics studies
to larger fish, as this would facilitate definition of appropriate feeding frequencies for
barramundi being reared on pelleted diets.
It was also interesting that extending the photoperiod for barramundi fry resulted in
increased food intake, but that it did not increase growth rate. A review of the
literature showed that extending the photoperiod during nursery rearing of fry (i.e.
fish which have metamorphosed) usually has no effect on growth rate (although for
larval fish prior to metamorphosis it often does increase growth rate). The increased
food intake at dusk is of particular significance for farming: barramundi farmers have
observed that juvenile barramundi exhibit a more vigorous feeding response at dusk,
and consequently feeding at dusk is now the normal procedure at most production
facilities. Furthermore, it is possible that juvenile barramundi may assimilate food
more efficiently at night when the fish have discontinued feeding (slower gut passage
rate) and are comparatively inactive (at least at the fry stage). It has been shown that
the South American fish Piaractus brachypomus grows faster and converts food
more efficiently when fed at night, compared with feeding during the daytime or over
24 hours (Baras et al. 1996). The different performance was proposed as being a
result of either variable energy assimilation at different periods of the day or a
decrease in energy expenditure when food was given at night, possibly due to the
correspondence between the nocturnal feeding schedule and the activity rhythm
pattern of P. brachypomus. Studies have shown that feeding schedule significantly
influences growth patterns and food conversion rates in many species of fishes (see
review by Boujard and Leatherland 1992). These examples, and the information
generated on barramundi in the present study, emphasise the need for fish culturists
to optimise on-farm management procedures in accordance with the particular
biological properties of the species being farmed.
It should be noted that during the course of this study other researchers within
Australia were investigating methods of improving larval rearing procedures for
barramundi (Palmer et al. 1992, Rimmer 1996). The technique of intensive larval
rearing in indoor hatcheries followed by a freshwater fry production stage is now of
minor importance. Nowadays, the majority of barramundi fingerlings produced in
Australia are grown extensively in brackishwater (about 18-25 ppt salinity) larval
rearing ponds. The ponds are stocked with 1-2 day old larvae, and harvested 20-25
days later at which time the fry are about 25-35 mm TL (Rimmer 1996). The fry are
then transferred to nurseries for weaning onto artificial diets prior to further grow-
out, or stocked into impoundments and rivers for fisheries enhancement purposes
(Russell and Rimmer 1997). The use of brackishwater larval rearing ponds has
greatly reduced the costs associated with fry production, and contributed significantly
to the expansion of barramundi farming in the country (Rimmer 1996).

8.4 RECREATIONAL FISHERIES ENHANCEMENT WITH
BARRAMUNDI (CHAPTERS 1 AND 7)
One of the overall aims of the barramundi rearing studies undertaken by the QDPI
has been to provide fish for stocking of water impoundments and rivers for
enhancement of recreational fisheries. Production techniques covering controlled
spawning of captive broodstock (Garrett and Connell 1991, Garrett and O'Brien
1994), larval rearing (Rimmer 1996) and juvenile rearing (this thesis) have been
documented and taken up by industry. Juvenile barramundi are now available for
most of the year, opening up the possibility for large-scale fishery-enhancement
activities.
Stocking to enhance fisheries can be in closed systems (e.g. natural lakes, freshwater
impoundments) or open systems (e.g. rivers, estuaries or the sea). In both systems,
one of the fundamental questions concerning enhancement techniques is the effect of
size at stocking on survival. Obviously, the larger the fish at the time of stocking the
more expensive it is to produce, so management agencies are keen to stock fish at the
smallest size possible while still getting satisfactory post-stocking survival. Many
agencies maintain that the larger fish show better survival, which is certainly evident
in investigations conducted with various species released in marine waters (e.g.,
Lundqvist et al. 1994, Yamashita et al. 1994, Hoff and Newman 1995, Leber 1995).
On the other hand, Willett (1996) showed that post-stocking survival of silver perch
stocked at 34 mm or 50 mm TL in two Queensland impoundments was the same.

Even in cases where larger fish show better survival, there is still a trade-off between
post-stocking survival and cost which needs to be assessed in determining the
optimum size for release (Hume and Parkinson 1988).
An assessment of the effect of size of barramundi at release on post-stocking survival
was conducted in Copperlode Dam, north Queensland, as an ancillary project to the
present study. Size groups were distinguished by stocking at different times. Fry 10
mm TL at stocking showed zero survival, due to predation by schools of rainbowfish
Melanotaenia splendida which was observed at the time of stocking. However, fish
20 mm TL (too large to be consumed by rainbowfish) showed 80-90% survival
when monitored 6-12 months after stocking (A. Hogan, Freshwater Fisheries and
Aquaculture Centre, Walkamin, pers. comm. 1993).
Further studies are now being undertaken on stocking of barramundi in an open
system, the Johnstone River, in north Queensland. The aims are to determine the
optimal stocking strategy with respect to size and release site, and whether stocked
barramundi enhances the recreational fishery (Russell 1995). The tagging system
being used is micro-wire tags injected into the cheek musculature, which has been
shown to be the most suitable technique for physically marking multiple batches of
barramundi (Russell and Hales 1992). Analyses of data for three years of stocking
indicated that the size of fish at stocking (30-40 mm TL and 50-60 mm TL) did not
affect survival (Russell and Rimmer 1997). Furthermore, estuarine, freshwater and
upper tidal habitats all appear to be suitable locations for stocking, although more
data are required to determine if any one site is superior (Russell and Rimmer 1997).
The use of circulus spacing as an innate tag to distinguish different batches or stocks
of fish proved reliable for barramundi, as it has for many other species (see Chapter
7). Its application with barramundi is, however, limited to fish smaller than 350 mm
TL, as beyond that size the scales are too thick and pigmented to be reliably read
using light transmittance techniques. In view of this finding, it would be beneficial
to evaluate the use of pancreatin for cleaning barramundi scales. This was not done
in the present study, as Australian quarantine authorities currently prohibit the import
of pancreatin because of its porcine origin (D. Willett, Southern Fisheries Centre,
Deception Bay, pers. comm. 1996).
On the other hand, the scale-circulus pattern-analysis technique has proved
particularly useful for a range of other Australian freshwater and marine fish which
have small scales as adults. These include silver perch Bidyanus bidyanus, golden
perch Macquaria ambigua (Willett 1995), summer whiting (Sillago ciliata) and
dusky flathead (Platycephalus fuscus) (D. Willett, Southern Fisheries Centre,
Deception Bay, pers. comm. 1996).
In conclusion, enhancement of recreational fisheries in closed systems will continue
to be an important part of fisheries management. In open systems, it is likely to
become more prominent given the initial success of stockings with barramundi in
northern Queensland (Russell 1995, Russell and Rimmer 1997), and the increasing
use of marine stocking overseas. Tagging of fish to facilitate monitoring will
continue to be based on a variety of techniques (innate tags, internal or external
physical tags) depending on the species and information requirements of the
program.

Table 8.1 Tabulation of the contribution of the major elements of this thesis to management, biology of barramundi and ecological
theory, and aspects of future research requirements.

Management Biology of barramundi Ecological theory Future research
Chapters 1-2
Nile perch in Lake Victoria provides a Graphic example of predation being Sustainability of fishery in Lake Victoria — will
stark example of ecological impacts the primary agent responsible for biomass reduction result in a significant
resulting from introduction of exotic extinction of a large number of species decrease in Nile perch catches?
fish; multitude of flow-on effects; socio- and reduction in faunal biodiversity
economics complex Preservation and rehabilitation of endangered
species — maintenance of biodiversity
Description of socio-economic impacts —

lessons for other areas
Examination of Nile perch biology plus

information on proposed recipient
environments in Australia enables
prediction of probable range and
possible impacts
Chapters 3 6 -
Feeding behaviour and diet influences Feeding behaviour and diet of juvenile For new species, feeding behaviour and diet
the optimum size for weaning barramundi (10-50 mm TL) changes (under laboratory, pond and/or natural
barramundi onto artificial diets with size, with a major switch from conditions) should be an integral part of
roving zooplanktivore to lurking aquaculture research
predator at 16-18 mm TL
Management of dragonfly predation Feeding behaviour indicated that Innate behavioural response of Documentation of the impact of dragonfly
necessary for freshwater pond rearing – escape response not developed until barramundi (marine spawner) and predation in pond rearing of freshwater fish
need to stock fish at 16 mm TL and/or approx 16 mm TL; this confirmed in sooty grunter (freshwater spawner) to larvae, and development of control methods
minimise period between filling and trials testing susceptibility to predation dragonfly nymphs – lack of
stocking pond. by dragonfly nymphs recognition and recognition
respectively of a predator
Gut passage rate changes with feed Barramundi fry shift from numbers Define optimum frequency for feeding grow-
availability and size of fry maximisers to energy maximisers as out barramundi, including potential for
they grow – inter-relationship between compensatory growth (i.e. increased food
biology and environment consumption and growth rate following a
period of no feeding)
Prolonging daylength increases food Literature review indicates that

intake of juvenile barramundi, but increasing daylength increases growth
growth rate is not increased rates of larval fish but not juvenile fish
Diel feeding patterns have implication Diel feeding patterns — juvenile Define diurnal activity rhythms for all size
for management of feeding regimes on barramundi feed throughout the day, ranges of farmed barramundi
farms, and possibly lighting in indoor peak intake in the evening, cease
production systems feeding in darkness Is food assimilation efficiency increased at
night in barramundi?
Chapter 7
Circulus spacing provides a mechanism Pattern of circulus spacing near the Influence of size at stocking on survival
for recognising separate stocks of origin of scales is determined by the
barramundi released into common water natal rearing environment Contribution of stocked fish to fishery
bodies — but technique presently only
applicable to barramundi smaller than Carrying capacity of recipient waters
350 mm TL
References 176
REFERENCES
Acere, T. 1984. Observations on the biology, age, growth, maturity and sexuality of
Nile perch Lates niloticus (Linne) and the growth of its fishery in the northern
waters of Lake Victoria. In Committee for Inland Fisheries of Africa, FAO
Fish. Rep. (335): 42-61.
Acere, T.O. 1988. The controversy over Nile perch, Lates niloticus, in Lake Victoria,
East Africa. NAGA — The ICLARM Quarterly, October 1988: 3-5.
Achieng, A.P. 1990. The impact of the introduction of Nile perch, Lates niloticus (L.)
on the fisheries of Lake Victoria. Journal of Fish Biology 37 (Supplement A):
17-23.
Allen, G.R 1989. Freshwater Fishes of Australia. T.F.H. Publications. 240 pp.
Ali, H.M. 1987. Sea bass (Lates calcarifer) spawning in tanks in Malaysia. In: J.W.
Copland and D.L. Grey (Editors), Management of Wild and Cultured Sea
Bass/Barramtmdi (Lates calcarifer): Proceedings of an International Workshop
held at Darwin, N.T., Australia, 24-30 September 1986. ACIAR Proceedings
No. 20: 129-131.
Amos, M.H., Anas, R.E. and Pearson, R.E. 1963. Use of a discriminant function in
the morphological separation of Asian and North American races of pink
salmon, Oncorhynchus gorbuscha (Walbaum). Bull. 11, International North
Pacific Fisheries Commission: 73-100.
Amundsen, P.A. and Klemetsen, A. 1988. Diet, gastric evacuation rates and food
consumption in a stunted population of Arctic charr, Salvelinus alpinus L. in
Takvatu, northern Norway. Journal of Fish Biology 33: 697-709.
Anderson, A.M. 1961. Further observations concerning the proposed introduction of

Nile perch into Lake Victoria. East African Agriculture and Forestry Journal
26: 195-201.
Andrews, C. 1990. The ornamental fish trade and fish conservation. Journal of Fish
Biology 37 (Supplement A): 53-59.
Angradi, T.R. and Griffith, J.S. 1990. Diel feeding chronology and diet selection of
rainbow trout (Oncorhynchus mykiss) in Henry's Fork of the Snake River,
Idaho. Canadian Journal of Fisheries and Aquatic Sciences 47(1): 199-209.
References 177
Anonymous. 1991. Northern Territory Barramundi Fishery Management Plan.

Northern Territory Government Gazette S5.85 pp.
Anonymous. 1993. Annual Statistics DPI Water Resources 1992-1993. Queensland

Government Printer, Brisbane. 53 pp.
Arthington, A.H. 1989. Impacts of introduced and translocated freshwater fishes in

Australia. In: S.S. De Silva (Editor), Exotic Aquatic Organisms in Asia:
Proceedings of a Workshop on Introduction of Exotic Aquatic Organisms in
Asia. Asian Fisheries Society, Manila, Special Publication No. 3: 7-20.
Arthington, A.H. 1991. Ecological and genetic impacts of introduced and translocated
freshwater fishes in Australia. Canadian Journal of Fisheries and Aquatic
Sciences 48(Suppl 1): 33-43.
Arthington, A.H. and Mitchell, D.S. 1986. Aquatic invading species. In: R.H. Groves
and J.J. Burdon (Editors), Ecology of biological invasions: An Australian
perspective. Australian Academy of Science, Canberra: 34-53.
Arumugam, P.T. and Geddes, M.C. 1987. Feeding and growth of golden perch larvae
and fry (Macquaria ambigua Richardson) (sic). Transactions of the Royal
Society of South Australia 111(1): 59-65.
Awang, A. 1987. Sea bass (Lates calcarifer) larvae and fry production in Malaysia.
In: J.W . Copland and D.L. Grey (Editors), Management of Wild and Cultured
Sea Bass/Barramundi (Lates calcarifer): Proceedings of an International
Workshop held at Darwin, N.T., Australia, 24-30 September 1986. ACIAR
Proceedings No. 20: 144-147.
Bajkov, A.D. 1935. How to estimate the daily food consumption of fish under natural
conditions. Transactions of the American Fisheries Society 65: 288-289.
Bancroft, T.L. 1914. On an easy and certain method of hatching Ceratodus ova.
Proceedings of the Royal Society of Queensland 25: 1-3.
Bancroft, T.L. 1933. Some further observations on the rearing of Ceratodus.

Proceedings of the Linnaean Society of NSW 58(6): 467-69.
Barahona-Femandes, M.H. 1979. Some effects of light intensity and photoperiod on

the seabass larvae (Dicentrarchus labrax (L.)) reared at the Centre
Oceanologique de Bretagne. Aquaculture 17: 311-321.
References 178
Baras, E., Melard, C., Grignard, J.C. and Thoreau, X. 1996. Comparison of food
conversion by pirapatinga Piaractus brachypomus under different feeding
times. Progressive Fish-Culturist 58: 59-61.
Barel, C.D.N., Dorit, R., Greenwood, P.H., Fryer, G., Hughes, N., Jackson, P.B.N.,
Kawanabe H., Lowe-McConnell, R.H., Nagoshi, M., Ribbink, A.J., Trewavas,
E., Witte, F. and Yamaoka, U. 1985. Destruction of fisheries in Africa's lakes.
Nature, London., 315: 19-20.
Barlow, C.G. 1984. The Nile perch project: progress and plans. Search 15: 88-91.
Barlow, C.G., Hogan, A.E. and Rodgers, L.J. 1987. Implication of translocated fishes
in the apparent extinction in the wild of the Lake Eacham rainbowfish,
Melanotaenia eachamensis. Australian Journal of Marine and Freshwater
Research 38: 897-202.
Barlow, C.G., Leedow, M. and McLaughlin, R. 1982. Macroinvertebrate sampling

using a dredge net in a farm dam in south-western New South Wales.
Australian Zoologist 21(1): 97-104.
Barnham, C.A. 1977. Ten-year effort to acclimatise quinnat salmon pays off.
Australian Fisheries 36(9): 14-15,19.
Battaglene, S.C. and Talbot, R.B. 1990. Initial swim bladder inflation in intensively
reared Australian bass larvae, Macquaria novemaculeata (Steindachner)
(Perciformes: Percichthyidae). Aquaculture 86(4): 431-442.
Battaglene, S.C., Beevers, P.J. and Talbot, R.B. 1989. A review of research into
artificial propagation of Australian bass (Macquaria novemaculeata) at the
Brackish Water Fish Culture Research Station, Salamander Bay, 1979 to 1986.
Fisheries Bulletin No. 3, New South Wales Department of Agriculture and
Fisheries, 11 pp.
Bedawi, R.M. 1985. Recruitment control and production of market size Oreochromis
niloticus with the predator Lates niloticus L. in the Sudan. Journal of Fish
Biology 26: 459-64.
Bell, J.D., Quartararo, N. and Henry, G.W. 1991. Growth of snapper, Pagrus auratus,
from south-eastern Australia. New Zealand Journal of Marine and Freshwater
Research 25: 117-121.
References 179
Baton, H.T. 1988. The body area and size that chinook, coho, and chum salmon fry
first form their scales. Canadian Technical Reports on Aquatic Sciences No.
1632, 21 pp.
Biosonics, 1989. Optical Pattern Recognition System (OPRS) analysis manual.

Biosonics, Inc., Seattle, Washington.
Bird, J.L., Eppler, D.T. and Checkley, D.M. 1986. Comparisons of herring otoliths
using Fourier Series shape analysis. Canadian Journal of Fisheries and
Aquatic Sciences 43: 1228-1234.
Bishai, R.M. 1975. Food and feeding habits of the Nile perch Lates niloticus (L.) at
Gebel Aulyia reservoir (Sudan). Bulletin of the Zoological Society of Egypt 27:
90-7.
Blankenship, H.L. and Leber, K.M. 1995. A responsible approach to marine stock
enhancement. American Fisheries Society Symposium 15: 167-175.
Bloch, M.E. 1790. Ichtyologie, ou Histoire Naturelle, Generale et Particuliere, des

Poissons. Vol. 7. Berlin and Leipzig. pp. 80-81.
Boehlert, G.W. 1981. The effects of photoperiod and temperatures on laboratory

growth of juvenile Sebastes diploproa and a comparison with growth in the
field. Fishery Bulletin 79(4): 789-794.
Boehlert, G.W. and Mundy, B.C. 1988. Roles of behavioral and physical factors in
larval and juvenile fish recruitment to estuarine nursery areas. In: M.P.
Weinstein (Editor), Larval Fish and Shellfish Transport through Inlets:
Proceedings of a Workshop held in Ocean Springs, Mississippi, 19-20 August
1985. American Fisheries Society Symposiums, 3: 63-67.
Boehlert, G.W. and Yoklavich, M.M., 1984. Carbon assimilation as a function of

ingestion rate in larval Pacific herring, Clupea harengus pallasi Valenciennes.
Journal of Experimental Marine Biology and Ecology 79: 251-262.
Boisclair, D. and Leggett, W.C. 1988. An in situ experimental evaluation of the Elliot
and Persson and Eggers models for estimating fish daily ration. Canadian
Journal of Fisheries and Aquatic Sciences 45: 138-145.
Boujard, T. and Leatherland, J.F. 1992. Circadian rhythms and feeding time in fishes.
Environmental Biology of Fishes 35: 109-131.
180
References
Briggs, I.C. 1981. Contour. CSIRO Division of Computing. Research, Canberra,

Computing Note No. 18.
Bromley, P.J. 1987. The effects of food type, meal size and body weight on digestion
and gastric evacuation in turbot, Scophthalmus maximus L. Journal of Fish
Biology 30: 501-512.
Brown, E.E. and Gratzek, J.B. 1980. Fish Farming Handbook — Food, Bait, Tropicals
and Goldfish. Van Nostrand Reinhold Company, New York. 391 pp.
Bruton, M.N. 1990. The conservation of the fishes of Lake Victoria: an ecological
perspective. Environmental Biology of Fishes 27: 161-175.
Burleigh, J.G., Katayama, R.W. and Elkassabany, N. 1993. Impact of predation by

backswimmers in golden shiner, Notemigonus crysoleucas, production ponds.
Journal of AppliedAquaculture 2(3/4): 243-256.
Bwathondi, P.O.J. 1984. The future of the fisheries of the Tanzanian part of Lake
Victoria, in view of the predominance of Nile perch Lates niloticus. Committee
for Inland Fisheries of Africa, FAO Fish Rep., (335): 143-5.
Cadwallader, P.L. 1983. A review of fish stocking in the larger reservoirs of Australia
and New Zealand. FAO Fisheries Circular No. 757: 38 pp.
Cadwallader, P.L. and Backhouse, G.N. 1983. A Guide to the Freshwater Fish of
Victoria. Government Printer, Melbourne. 249 pp.
Cadwallader, P.L. and Gooley, G.J. 1985. Propagation and Rearing of Murray Cod
Maccullochella peeli at the Warmwater Fisheries Station Pilot Project, Lake
Charlegrark. Government Printer, Melbourne. 189 pp.
Campana, S.E. and Casselman, J.M. 1993. Stock discrimination using otolith shape
analysis. Canadian Journal of Fisheries and Aquatic Sciences 50: 1062-
1083.
Cerqueira, V.R. and Chatain, B. 1991. Photoperiodic effects on the growth and
feeding rhythm of European seabass, Dicentrachus labrax, larvae in intensive
rearing. In: P. Lavens, P. Sorgeloos, E. Jaspers and F. 011evier (Editors.)
Larvi '91 - Fish & Crustacean Larviculture Symposium, European
Aquaculture Society, Special Publication No. 15, Gent, Belgium.
Chomdej, W. 1986. Technical Manual for Seed Production of Sea Bass. National
Institute of Coastal Aquaculture, Songkhla, Thailand. 49 pp.
References 181
Clague, C.I. 1991. The burst and sustained swimming abilities of juvenile barramundi
(Lates calcarifer) and sooty grunter (Hephaestus fuliginosus). B.Sc.(Hons.)
Thesis, James Cook University of North Queensland.
Clements, J. 1988. Salmon at the Antipodes: A History and Review of Trout, Salmon
and Char and Introduced Coarse Fish in Australasia. John Clements, Ballarat.
391 pp.
Cook, R.C. 1982. Stock identification of sockeye salmon (Oncorhynchus nerka)

with scale pattern recognition. Canadian Journal of Fisheries and Aquatic
Sciences 39: 611-617.
Cook, R.C. and Guthrie, I. 1987. In-season stock identification of sockeye salmon
(Oncorhynchus nerka) using scale pattern recognition. Canadian Special
Publication of Fisheries and Aquatic Sciences 96: 327-334.
Cook, R.C. and Lord, G.E. 1978. Identification of stocks of Bristol Bay sockeye
salmon, Oncorhynchus nerka, by evaluating scale patterns with a polynomial
discriminant method. Fisheries Bulletin 76: 415-423.
Copland, J.W. and Grey, D.L. (Editors). 1987. Management of Wild and Cultured Sea
Bass/Barramundi (Lates calcarifer): Proceedings of an International Workshop
No. 20. 210 pp.
Craig, J.F. 1992. Human-induced changes in the composition of fish communities in

the African Great Lakes. Reviews in Fish Biology and Fisheries 2: 93-124.
Crisp, D.T. and Hciwson, G. 1982. Effect of air temperature upon mean water
temperature in streams in the north Pennines and English Lake District.
Freshwater Biology 12: 359-67.
Cuvier, G.L.C.F.D. and Valenciennes, A. 1828. Histoire Naturelle des Poissons. Vol.
2. Paris. pp. 88-101.
Dakin, W.J. and Kesteven, G.L. 1938. The Murray cod (Maccullochella
macquariensis (Cuv. et Val.)). N.S.W. State Fisheries Research Bulletin No. 1.
18 pp.
Danielssen, D.S., Howell, B.R. and Moksness, E. (Editors). 1994. An International

Symposium on Sea Ranching of Cod and other Marine Fish Species.
Aquaculture and Fisheries Management 25 (Supplement 1): 264 pp.
References 182
Davis, T.L.O. 1982. Maturity and sexuality in barramundi, Lates calcarifer (Bloch), in
the Northern Territory and Southeastern Gulf of Carpentaria. Australian
Journal of Marine and Freshwater Research 33: 529-545.
Davis, T.L.O. 1984. A population of sexually precocious barramundi Lates calcarifer

in the Gulf of Carpentaria, Australia. Copeia 1984(1): 1/1/1 /9.
Davis, T.L.O., 1985. The food of barramundi, Lates calcarifer (Bloch), in coastal and
inland waters of van Diemen Gulf and the Gulf of Carpentaria, Australia.
Journal of Fish Biology 26: 369-682.
Davis T.L.O. 1987. Biology of wildstock Lates calcarifer in Northern Australia. In:
J.W. Copland and D.L. Grey (Editors), Management of Wild and Cultured Sea
No. 20: 22-29.
De, G.K. 1971. On the biology of post-larval and juvenile stages of Lates calcarifer
Bloch (sic). Journal of the Indian Fisheries Association 1: 51-64.
de Pontual, H. and Prouzet, P. 1987. Atlantic salmon, Salmo salar L., stock
discrimination by scale shape analysis. Aquaculture and Fisheries
Management 18: 277-289.
Douglas, J.W., Gooley, G.J., Ingram, B.A., Murray, N.D. and Brown, L.D. 1995.
Natural hybridization between Murray cod, Maccullochella peelii peelii
(Mitchell), and trout cod, Maccullochella macquariensis (Cuvier)
(Percichthyidae), in the Murray River, Australia. Marine and Freshwater
Research 46: 729-34.
Doyle, R.W., Talbot, A.J. and Nicholas, R.R. 1987. Statistical interrelation of
length, growth, and scale circulus spacing: appraisal of a growth rate
estimator for fish. Canadian Journal of Fisheries and Aquatic Sciences 44:
1520-1528.
Dunstan, D.J. 1959. The barramundi in Queensland waters. CSIRO Australia,

Division of Fisheries and Oceanography Technical Paper No. 5. 22 pp.
Dunstan, D.J. 1962. The barramundi in New Guinea waters. Papua New Guinea
Agriculture Journal 15: 23-31.
References 183
Duray, M. and Kohno, H. 1988. Effects of continuous lighting on growth and survival
of first-feeding larval rabbitfish, Siganus guttatus. Aquaculture 109: 311-321.
Eggers, D.M. 1977. Factors in interpreting data obtained by diel sampling of fish
stomachs. Journal of the Fisheries Research Board of Canada 34: 290-294.
Eisawy, A.M., Ishak, M.M. & Hamza, A. 1974. Experimental rearing of two mullet
species Mugil cephalus and Mugil capito in Egyptian fish farms. Bulletin of the
Institute of Oceanography and Fisheries, Cairo 4: 57-96.
FAO. 1974. FAO species identification sheets for fishery purposes: Eastern Indian
Ocean (Fishing Area 57) and Western Central Pacific (Fishing Area 71). Vol. 1.
FAO. 1994a. Aquaculture Production 1986-1992. FAO Fisheries Circular No. 815
Revision 6. Rome. 216 pp.
FAO. 1994b. FAO yearbook — Fishery statistics, catches and landings. Vol. 74,
1992. FAO Statistics Series No. 120. Rome. 678 pp.
Ferdouse, F. 1995. Cultured sea bass market. Infofish International 2/95: 20-22.
Fermin, A.C. 1991. Freshwater cladoceran Moina macrocopa (Strauss) as an

alternative live food for rearing sea bass Lates calcarifer (Bloch) fry. Journal
of Applied Ichthyology 7: 8-14.
Fermin, A.C. and Bolivar, Ma.E.C. 1994. Feeding live or frozen Moina macrocopa
• (Strauss) to Asian sea bass, Lates calcarifer (Bloch), larvae. The Israeli
Journal of Aquaculture — Bamidgeh 46(3): 132-139.
Fisher, J.P. and Pearcy, W.G. 1990. Spacing of scale circuli versus growth rate in
young coho salmon. Fishery Bulletin, U.S. 88: 637-643.
Fletcher, A.R. 1986. Effects of introduced fish in Australia. In: P. de Deckker and
W.D. Williams (Editors), Limnology in Australia. CSIRO: Melbourne, and Dr
W. Junk: Dordrecht. 231-238.
Fryer, G. 1960. Concerning the proposed introduction of Nile perch into Lake
Victoria. East African Agriculture Journal 25: 267-70.
Fuchs, J. 1978. Effect of photoperiod on growth and survival during rearing of larvae
and juveniles of sole (Solea solea). Aquaculture 72: 73-79.
References 184
Garrett, R.N. 1987. Reproduction in Queensland barramundi (Lates calcarifer). In:

J.W. Copland and D.L. Grey (Editors), Management of Wild and Cultured Sea
No. 20: 38-43.
Garrett, R.N., MacKinnon, M.R. and Russell, D.J. 1987. Wild barramundi breeding
and its implications for culture. Australian Fisheries 46(7): 4-6.
Garrett, R.N. and Connell, M.R.J. 1991. Induced breeding in barramundi. Austasia
Aquaculture 5(8): 10-12.
Garrett, R.N. and O'Brien, J.J. 1994. All-year-round spawning of hatchery barramundi
in Australia. Austasia Aquaculture 8(2): 40-42.
Gee, J.M. 1964. Nile perch investigation. East African Freshwater Fisheries
Research Organisation Annual Report 1962/63: 14-24.
Gee, J.M. 1965. The spread of Nile perch Lates niloticus in East Africa, with
comparative biological notes. Journal of Applied Ecology 2: 407-8.
Gee, J.M. 1969. A comparison of certain aspects of the biology of Lates niloticus
(Linne) in some East African lakes. Revue de Zoologie et de Botanique
Africaines 80: 244-262.
Ghosh, A.N. and Pandit, P.K. 1979. On the rearing of fry of bhetki Lates calcarifer
(Bloch) in brackishwater ponds. Matsya 5: 50-55.
Glenn, C.L. and Mathias, J.A. 1985. Circuli development on body scales of young
pond-reared walleye. Canadian Journal of Zoology 63: 912-915.
Goldschmidt,T., Witte, F. and Wanink, J. 1993. Cascading effects of the introduced

Nile perch on the detritivorous/phytoplanktivorous species in the sublittoral
areas of Lake Victoria. Conservation Biology 7(3): 686-700.
Gonzalez, A.V. and Leal, J.M. 1995. Predation potential of some aquatic insects
(Pantala, Coenagrion, Tropisternus, Notonecta and Sigara) on common carp
fry. Journal of AppliedAquaculture 5(1): 77-82.
Gooley, G.J. and Anderson, T.A. 1992. Growth and survival of intensively reared
juvenile Murray cod, Maccullochella peelii, fed on artificial diets. In: G.L.
Allan and W. Dall (Editors), Proc. Aquaculture Nutrition Workshop,
References 185
Salamander Bay, 15-17 April 1991. N.S.W. Fisheries, Brackish Water Fish
Culture Research Station, Salamander Bay, Australia: 46-47.
Gooley, G.J., Anderson, T.A. and Appleford, P. 1995. Aspects of the reproductive
cycle and gonadal development of Murray cod, Maccullochella peelii peelii
(Mitchell) (Percichthyidae), in Lake Charlegrark and adjacent farm ponds,
Victoria, Australia. Marine and Freshwater Research 46: 723-728.
Goudswaard, P.C. and Wanink, J.H. 1993. Authropogenic perturbation in Lake

Victoria: effects of fish introductions and fisheries on fish eating birds.
Proceedings VIII Pan-African Ornithological Congress. pp.312-318.
Goudswaard, P.C. & Witte, F. 1984. Observations on Nile perch, Lates niloticus (L),
1758, in the Tanzanian waters of Lake Victoria. Committee for Inland
Fisheries of Africa, FAO Fisheries Reports (335): 62-67.
Grant, E.M. 1985. Guide to Fishes. Department of Harbours and Marine, Brisbane.
896 pp.
Greenwood, P.H. 1976. A review of the family Centropomidae (Pisces, Perciformes).

Bulletin British Museum of Natural History (Zoology) 29: 1-81.
Grey, D.L. 1987. An overview of Lates calcarifer in Australia and Asia. In: J.W.
No. 20: 15-21.
Griffin, R.K. 1982. A survey of amateur angling for barramundi (Lates calcarifer) in
the Northern Territory. Technical Report No. 2. Department of Primary
Production, Northern Territory, 37 pp.
Griffin, R.K. 1987. Life history, distribution, and seasonal migration of barramundi in
the Daly River, Northern Territory, Australia. American Fisheries Society
Symposium 1: 358-363.
Griffin, R.K. 1988. Recreational fishing for barramundi in the Arnhem Highway area
— Report of 1986-87 surveys. Fishery Report No. 17. Department of Primary
Industry and Fisheries, Northern Territory.
Griffin, R.K. 1993. The recreational fishery for barramundi (Lates calcarifer) in the
Mary River, Northern Territory 1986-1992. Fishery Report No. 30.
Department of Primary Industry and Fisheries, Northern Territory.
References 186
Griffiths, D. 1975. Prey availability and the food of predators. Ecology 56:
1209-1214.
Hall, S.J. 1987. Maximum daily ration and the pattern of food consumption in
haddock, Melanogrammus aeglefinus (L), and dab, Limanda limanda (L).
Journal of Fish Biology 3: 479-491.
Hamblyn, E.L. 1961. The Nile perch project. East African Freshwater Fisheries
Research Organisation Annual Report 1960: 26-32.
Hamblyn, E.L. 1962. Nile perch investigation. East African Freshwater Fisheries
Hara, S., Kohno, H., Duray, M., Bagarinao, T., Gallego, A. and Taki, Y., 1986.
Feeding habits of larval rabbitfish, Siganus guttatus in the laboratory. In: J.L.
MacLean, L.B. Fijon and L.V. Hosillos (Editors), The First Asian Fisheries
Forum, Asian Fisheries Society, Manila: 573-576.
Harris, C.K., Wiley, D.S. and Wilson, D.C. 1995. Socio-economic impacts of
introduced species in Lake Victoria fisheries. In: T.J. Pitcher and P.B. Hart
(Editors), The Impact of Species Changes in African Lakes, Chapman and Hall,
London: 215-244.
Harrison, K. 1991. The taxonomy of East African Nile perch, Lates spp. (Perciformes,
Centropomidae). Journal of Fish Biology 38: 175-186.
Harvey, B., Nacario, J., Crim, L.W., Juario, J.V. and Marte, D.L. 1985. Induced
breeding of sea bass, Lates calcarifer, and rabbitfish, Siganus guttatus, after
implantation of LHRH analogue. Aquaculture 47: 53-59.
Hashem, M.T. & Hussein, K.A. 1973. Some biological studies of the Nile perch (Lates
niloticus C. and V.) in the Nozha-Hydrodome. Bulletin. Of the Institute of
Oceanography & Fisheries, Arab Republic of Egypt 3: 363-393.
Hawking, J.H. and Ingram, B.A. 1994. Rate of larval development of Pantala
flavescens (Fabricus) at its southern limit of range in Australia (Anisoptera :
Libellulidae). Odonatologica 23(1): 63-68.
Heasman, M.P., Ryall, J.C. and Hockings, I.R. 1985. Development of barramundi
(Lates calcarifer Bloch) hatchery and farming techniques in Australia, October
1983—June 1985. Final Report, FIRTA Project 83/38.
References 187
Henry, K.A. 1961. Racial identification of Fraser River sockeye salmon by means of
scales and its application to salmon management. International Pacific
Salmon Fisheries Commission, Bulletin 12: 1-97.
Hoff, M.H. and Newman, S.P. 1995. Comparisons of fingerling and yearling lake trout
introductions for establishing an adult population in Pallette Lake, Wisconsin.
North American Journal of Fisheries Management 15: 871-873.
Hogan, A. and Graham, P. 1994. Herbert River flood plain fish distributions and fish
habitat. Interim Report to the Consultants, Herbert River Sugar Industry
Infrastructure Package, Queensland Department of Primary Industries. 12 pp.
Hopson, A.J. 1972. A study of the Nile perch (Lates niloticus (L.), Pisces:
Centropomidae) in Lake Chad. Foreign and Commonwealth Office, Overseas
Development Administration, Research Publication No. 19, London.
Horn, M.J., Marsh, P.C., Mueller, G. and Burke, T. 1994. Predation by odonate
nymphs on larval razorback suckers (Xyrauchen texanus) under laboratory
conditions. The Southwestern Naturalist 39(4): 371-374.
Houde, E.D. 1989. Comparative growth, mortality and energetics of marine fish
larvae: temperature and implied latitudinal effects. Fishery Bulletin 87(3):
471-495.
Htin, W. 1987. Review of the sea bass (Lates calcarifer) fishery in Burma. In: J.W.
No. 20: 72-73.
Huet, M. 1970. Textbook of Fish Culture. Fishing News Books/Blackwell Scientific

Publications, Oxford. 436 pp.
Hughes, N.F. 1986. Changes in the feeding biology of the Nile perch, Lates niloticus
(L.) (Pisces: Centropomidae), in Lake Victoria, east Africa since its
introduction in 1960, and its impact on the native fish community of the
Nyanza Gulf. Journal of Fish Biology 29: 541-548.
Hume, J.M.B. and Parkinson, E.A. 1988. Effects of size at and time of release on the
survival and growth of steelhead fry stocked in streams. North American
Journal of Fisheries Management 8(1): 50-57.
References 188
Hutchinson, W. 1993. A glimmer of success for striped trumpeter. Austasia

Aquaculture 7(4): 50.
Ingram, B.A., Barlow, C.G., Burchmore, J.J., Gooley, G.J., Rowland, S.J. and Sanger,
A.C. 1990. Threatened native freshwater fishes in Australia — some case
histories. Journal of Fish Biology 37(Supplement A): 175-182.
Ingram, B.A. and Rimmer, M.A. 1992. Induced breeding and larval rearing of the
endangered Australian freshwater fish trout cod, Maccullochella macquariensis
(Cuvier) (Percichthyidae). Aquaculture and Fisheries Management 24: 7-12.
Ingram, B.A., Rimmer, M.A. and Rowland, S.J. 1994. Induced spawning trials with
captive Macquarie perch, Macquaria australasica (Percichthyidae).
Proceedings Linnaean Society NS. W.. 114(2): 109-116.
Jackson, P.B.N. 1971. The African Great Lakes Fisheries: past, present and future.
African Journal of Tropical Hydrobiology and Fisheries 1: 35-49.
James, P.S.B.R. and Marichamy, R. 1987. Status of sea bass (Lates calcarifer) culture
in India. In: J.W. Copland and D.L. Grey (Editors), Management of Wild and
Cultured Sea Bass/Barramundi (Lates calcarifer): Proceedings of an
International Workshop held at Darwin, N.T., Australia, 24-30 September
1986. ACIAR Proceedings No. 20: 74-79.
Jarvis, R.S., Kludowski, H.F. and Sheldon, S.P. 1978. New method of quantifying
scale shape and an application to stock identification in walleye (Stizostedion
vitreum vitreum). Transactions of the American Fisheries Society 107(4):
528-534.
Jearld, A. Jr. 1983. Age determination. In: L. A. Neilsen and D. L. Johnson (Editors),
Fisheries Techniques, American Fisheries Society, Bethesda, Maryland: 301-
324.
Jobling, M. 1986. Mythical models of gastric emptying and implications for food
consumption studies. Environmental Biology of Fishes 16: 35-50.
Johnson, D.M. 1991. Behavioural ecology of larval dragonflies and damselflies.

Trends in Ecology and Evolution 6(1): 8-13.
Kasim, M.H. and James, P.S.B.R. 1987. Distribution and fishery of Lates calcarifer in
India. In: J.W. Copland and D.L. Grey (Editors), Management of Wild and
References 189

Katayama, M. and Taki, Y. 1984. Lates japonicus, a new centropomid fish from
Japan. Japanese Journal of Ichthyology 30 (4): 361-67.
Kaufman, L. 1992. Catastrophic change in species-rich freshwater ecosystems: the

lessons of Lake Victoria. BioScience 42(11): 846-858.
Keenan, C.P. 1994. Recent evolution of population structure in Australian

barramundi, Lates calcarifer (Bloch): an example of isolation by distance in
one dimension. Australian Journal of Marine and Freshwater Research 45:
1123-48.
Kenchington, F.E. 1939. Observations on the Nile perch (Lates niloticus) in the Sudan.
Proceedings Zoological Society, Ser. A-1939: 157-68.
Kiyono, M, and Hirano, R. 1981. Effects of light on the feeding and growth of black
porgy, Mylio macrocephalus (Basilewsky), postlarvae and juveniles. Rapports
et Proces-Verbaux de Reunions Conseil International pour l'Exploration de la
Mer 178: 334-336.
Kowarsky, J. and Ross, A.H. 1981. Fish movement upstream through a central
Queensland (Fitzroy River) coastal fishway. Australian Journal of Marine and
Freshwater Research 32: 93-109.
Kruuk, H. and Goudswaard, P.C. 1990. Effects of changes in fish populations in Lake
Victoria on the food of otters (Lutra maculicollis Schinz and Aonyx capensis
Lichtenstein). African Journal of Ecology 28: 322-329.
Kunhongania, A. and Chitamwebwa, D.B.R. 1995. Introduced Nile perch in Lake

Victoria: impacts on biodiversity and evaluation of the fishery. In: T.J. Pitcher
and P.B. Hart (Editors), The Impact of Species Changes in African Lakes,
Chapman and Hall, London: 19-32.
Kunhongania, A. and Cordone, A.J. 1974. Past trends, present stocks and possible
future state of the fisheries of the Tanzanian part of Lake Victoria. African
Journal of Tropical Hydrobiology and Fisheries. 3: 15-31.
Lachenbruch, P.A. 1967. An almost unbiased method of obtaining confidence

intervals for the probability of misclassification in discriminant analysis.
Biometrics 23: 639-645.
References 190
Lachenbruch, P.A. 1975. Discriminant Analysis. Hafner Press, New York.
Lake, J.S. 1967a. Rearing experiments with five species of Australian freshwater
fishes. I. Inducement to spawning. Australian Journal of Marine and
Lake, J.S. 1967b. Rearing experiments with five species of Australian freshwater
fishes. II. Morphogenesis and ontogeny. Australian Journal of Marine and
Lake, P.S., Barmuta, L.A., Boulton, A.J., Campbell, I.C. & St Clair, R.M. 1985.
Australian streams and Northern Hemisphere stream ecology: comparisons and
problems. In: J.R. Dodson and M. Westoby (Editors), Are Australian
ecosystems different?, Ecological Society of Australia, Artarmon: 61-82.
Lamb, L. 1925. A tabular account of the differences between the earlier instars of
Pantala flavescens (Odonata : Libellulidae). Transactions of the American
Entomological Society 50: 289-312.
Lamb, L. 1929. The later larval stages of Pantala (Odonata : Libellulidae).

Transactions of the American Entomological Society 55: 331-334.
Landau, M. 1992. Introduction to Aquaculture, John Wiley and Sons, New York.
440 pp.
Laurence, G.C. 1971. Digestion rate of larval largemouth bass. New York Fish and
Game Journal 18: 52-56.
Lea, R., Grey, D. and Griffin, R. 1987. Utilisation and wildstock management of the
barramundi (Lates calcarifer) in the Northern Territory. In: J.W . Copland and
D.L. Grey (Editors), Management of Wild and Cultured Sea Bass/Barramundi
(Lates calcariftr): Proceedings of an International Workshop held at Darwin,
N.T., Australia, 24-30 September 1986. ACIAR Proceedings No. 20: 82-86.
Lear, W.H. and Misra, R.K. 1978. Clinal variation in scale characters of Atlantic
salmon (Salmo salar) based on discriminant function analysis. Journal of the
Fisheries Research Board of Canada 35: 43-47.
Leber, K.M. 1995. Significance of fish size-at-release on enhancement of striped

mullet fisheries in Hawaii. Journal of the World Aquaculture Society 26(2):
143-153.
References 191
Ligtvoet, W. 1989. Stock assessment of Nile perch (Lates niloticus) in Lake Victoria.
In: Lake Victoria Fish Stocks and Fisheries. Mwanza, Tanzania.
Ligtvoet, W. and Witte, F. 1991. Perturbation through predator introduction: effects on

the food web and fish yields in Lake Victoria (east Africa). In: 0. Ravera
(Editor), Terrestrial and Aquatic Ecosystem Perturbation and Recovery. Ellis
Horwood, New York: 263-268.
Lim, L.C., Heng, H.H. and Lee, H.B. 1986. The induced breeding of sea bass Lates
calcarifer (Bloch) in Singapore. Singapore Journal of Primary Industries 14:
81-95.
Lin, L.T., Chang, L.C. and Lin, H.H. 1985. On the induced breeding and larval rearing
of pond-reared giant perch (Lates calcarifer Bloch). China Fisheries Monthly
394: 25-40.
Llewellyn, L.C. 1978a. Temperature and river height data for fifteen inland rivers of
NSW, 1963-67. Sydney, NSW State Fisheries.
Llewellyn, L.C. 1978b. Temperature and river height data for fifteen inland rivers of
NSW, 1968-72. Sydney, NSW State Fisheries.
Llewellyn, L.C. 1978c. Temperature and river height data for fifteen inland rivers of
New South Wales, 1973. Sydney, NSW State Fisheries.
Llewellyn, L.C. 1979a. Temperature and river height data for fifteen inland rivers of
Llewellyn, L.C. 1979b. Temperature and river height data for fifteen inland rivers of
Llewellyn, L.C. 1980a. Temperature and river height data for sixteen inland rivers of
NSW, 1976. Sydney, NSW State Fisheries.
Llewellyn, L.C. 1980b. Temperature and river height data for sixteen inland rivers of
Llewellyn, L.C. 1980c. Temperature and river height data for sixteen inland rivers of
Llewellyn, L.C. 1980d. Temperature and river height data for sixteen inland rivers of
References 192
Llewellyn, L.C. 1981. Temperature and river height data for sixteen inland rivers of
Loubens, G. 1974. Quelque aspects de la biologie de Lates niloticus du Tchad. Cahier

O.R.S.TO.M (Office de la Recherche Scientifique et Technique Outre-mer)
Serie Hydrobiologie 8: 3-21.
Lundqvist, H., McKinnell, S., Faengstom, H. and Berglund, I. 1994. The effect of
time, size and sex on recapture rates and yield after river releases of Salmo
salar smolts. Aquaculture 121: 245-257.
MacArthur, R.H. 1972. Geographical Ecology: Patterns in the Distribution of Species.

New York, Harper and Row.
Maceina, M.J. and Murphy, B.R. 1989. Differences in otolith morphology among
two subspecies of largemouth bass and their F 1 hybrid. Transactions of the
American Fisheries Society 118: 573-575.
MacKinnon, M.R. 1986. Production of stocking material in Australia. FAO Fisheries

Report No. 371, Supplement: 199-209.
MacKinnon, M.R. 1987a. Rearing and growth of larval and juvenile barramundi
(Lates calcariftr) in Queensland. In: J.W. Copland and D.L. Grey (Editors),
Management of Wild and Cultured Sea Bass/Barramundi (Lates calcarifer):
Proceedings of an International Workshop held at Darwin, N.T., Australia, 24-
30 September 1986. ACIAR Proceedings No. 20: 148-153.
MacKinnon, M.R. 1987b. Pellet-fed barramundi show potential for commercial

farming. Australian Fisheries 46(7): 42-44.
Mallen-Cooper, M. 1989. Swimming ability of juvenile barramundi Lates calcariftr

(Bloch) in an experimental vertical slot fishway. Internal Report No. 47, NSW
Agriculture and Fisheries.
Margraf, F.J. and Riley, L.M. 1993. Evaluation of scale shape for identifying
spawning stocks of coastal Atlantic striped bass (Morone saxatilis).
Fisheries Research 18: 163-172.
Major, R.L., Mosher, K.H. and Mason, J.E. 1972. Identification of stocks of Pacific
salmon by means of scale features. In: R.C. Simon and P.A. Larkin (Editors),
The Stock Concept in Pacific Salmon. University of British Columbia,
Vancouver, Canada: 209-231.
References 193
Maneewong, S., 1987. Research on the nursery stages of sea bass (Lates calcarifer) in
Thailand. In: J.W. Copland and D.L. Grey (Editors), Management of Wild and
Marliave, J.B. 1977. Effects of three artificial lighting regimes on survival of

laboratory-reared larvae of sailfin sculpin. Progressive Fish-Culturist 39(3):
117-118.
Marriot, S.P., Manacop, P.R. and Twongo, T. 1988. A report on the survey of Lake
Kyoga 1986-88. North-east Uganda Agricultural Development Project. 94 pp.
& app.
McCarty, C.E., Geiger, J.G., Stunner, L.N., Gregg, B.A. and Rutledge, W.P. 1986.
Marine fish culture in Texas: a model for the future. In: R.H. Stroud (Editor),
Fish Culture in Fisheries Management: Proceedings of a Symposium on the
Role of Fish Culture in Fisheries Management, at Lake Ozark, Missouri, March
31—April 3,1985. American Fisheries Society, Bethesda: 249-262.
McDowall, R.M. 1987. The occurrence and distribution of diadromy among fishes.
American Fisheries Society Symposium 1: 1-13.
McGinty, A.S. 1980. Survival, growth and variation in growth of channel catfish fry
and fingerlings. Ph.D. thesis, Auburn University, Alabama.
McKay, R.J. 1984. Exotic fishes in Australia. In: W.R. Courtenay, Jr. and J.R.
Stauffer, Jr. (Editors), Distribution, Biology and Management of Exotic Fishes.
John Hopkins University Press, Baltimore. 177-199.
McKay, R.J. 1989. Exotic and translocated freshwater fishes in Australia. In: S.S. De
Silva (Editor), Exotic Aquatic Organisms in Asia: Proceedings of a Workshop
on Introduction of Exotic Aquatic Organisms in Asia. Asian Fisheries Society,
Manila, Special Publication No. 3: 21-34.
Menon, P.M.G. 1948. On the food of the 'bekti', Lates calcarifer (Bloch), in the cold
season. Current Science 17: 156-157.
Merrick, J.R. and Green, L.C. 1982. Pond culture of the spotted barramundi,
Scleropages leichardti (Pisces:Osteoglossidae). Aquaculture 29: 171-176.
References 194
Merrick, J.R. and Midgley, S.H. 1981. Spawning behaviour of the freshwater catfish,
Tandanus tandanus (Plotosidae). Australian Journal of Marine and
Freshwater Research 32(8): 1003-1006.
Merrick, J.R. and Midgley, S.H. 1982. Techniques for collecting and culturing eggs of
two sleepy cod species in the genus Oxyeleotris (Perciformes:Eleotridae).
Australian Society for Limnology Bulletin 8: 27-30.
Midgley, S.H. 1968. A study of the Nile perch in Africa (and consideration as to its
suitability for Australian tropical inland waters). Winston Churchill Memorial
Trust Fellowship Report, No. 3.
Midgley, S.H. 1969. A study of the Nile perch in Africa. Australian Society for
Limnology Bulletin 1: 5.
Midgley, S.H. 1981. To introduce or not to introduce Nile perch: the case for. Safic 5:
6-10.
Midgley, S.H. 1987. The apparent decline in numbers of some diadromous fish
species in northern and east coast areas of Australia, in particular barramundi
(Lates calcarifer). Unpublished report submitted to Queensland Government.
13 pp.
Moore, R.M. 1979. Natural sex inversion in the giant perch (Lates calcarifer).
Australian Journal of Marine and Freshwater Research 30: 803-813.
Moore, R. 1980. Migration and reproduction in the percoid fish Lates calcarifer
(Bloch). Ph.D. Thesis, University of London. 213 pp.
Moore, R., 1982. Spawning and early life history of barramundi Lates calcarifer
(Bloch) in Papua New Guinea. Australian Journal of Marine and Freshwater
Research 33: 647-661.
Moore, R. and Reynolds, L.F. 1982. Migration patterns of barramundi, Lates

calcarifer (Bloch), in Papua New Guinea. Australian Journal of Marine and
Moreau, J. 1982. Exposé synoptique des donnees biologiques sur la perche du Nil
Lates niloticus (Linnaeus, 1762). FAO, Synopsis sur les Peches (132).
Moring, J.R. 1986. Stocking anadromous species to restore or enhance fisheries. In:
R.H. Stroud (Editor), Fish Culture in Fisheries Management: Proceedings of a
Symposium on the Role of Fish Culture in Fisheries Management, at Lake
References 195
Ozark, Missouri, March 31—April 3, 1985. American Fisheries Society,

Bethesda: 75-80.
Moritz, C., Danqing, Z. and Degnan, S. 1995. Evolutionary distinctiveness and

conservation status of the Lake Eacham rainbow fish, Melanotaenia
eachamensis. Final report to the Wet Tropics Management Agency and the
Queensland Department of Environment and Heritage. 30 pp. + appendices.
Morrissy, N. 1985. The commercial fishery for barramundi (Lates calcarifer) in

Western Australia. Report No. 68, Department of Fisheries and Wildlife,
Western Australia.
Morrissy, N.M. 1987. Status of the barramundi (Lates calcarifer) fishery in Western
Australia. In: J.W. Copland and D.L. Grey (Editors), Management of Wild and
Mukhopadhyay, M.K., Verghese, P.U. and Karmakar, H.C., 1983. Pigmentation in

juvenile bhetki Lates calcarifer. Environment and Ecology 1: 265-267.
Ochumba, P.B.O. 1995. Limnological changes in Lake Victoria since the Nile perch
introduction. In: T. Pitcher and P. Hart (Editors), The Impact of Species
Changes in African Lakes, Chapman and Hall, London: 33-43.
Ogari, J. 1984. Distribution, food and feeding habits of Lates niloticus in Nyanza Gulf
of Lake Victoria (Kenya). Committee for Inland Fisheries of Africa, FAO Fish.
Rep. (335): 68-80.
Ogari, J. and Dadzie, S. 1988. The food of the Nile perch, Lates niloticus (L.), after the
disappearance of the haplochromine cichlids in the Nyanza Gulf of Lake
Victoria (Kenya). Journal of Fish Biology 32: 571-577.
Ogutu-Ohwayo, R. 1984. The effects of predation by Nile perch Lates niloticus

(Linne), introduced into Lake Kyoga (Uganda) in relation to the fisheries of
Lake Kyoga and Lake Victoria. Committee for Inland Fisheries of Africa, FAO
Fish Rep. (335): 18-41.
Ogutu-Ohwayo, R. 1988. Reproductive potential of the Nile perch, Lates niloticus L.

and the establishment of the species in Lakes Kyoga and Victoria (East Africa).
Hydrobiologia 162: 193-200.
References 196
Ogutu-Ohwayo, R. 1990a. The reduction in fish species diversity in Lakes Victoria

and Kyoga (East Africa) following human exploitation and introduction of non-
native fishes. Journal of Fish Biology 37 (Supplement A): 207-208.
Ogutu-Ohwayo, R. 1990b. The decline of the native fishes of Lakes Victoria and
Kyoga (East Africa) and the impact of introduced species, especially the Nile
perch, Lates niloticus, and the Nile tilapia, Oreochromis niloticus.
Ogutu-Ohwayo, R. 1990c. Changes in the prey ingested and the variations in the Nile
perch and other fish stocks of Lake Kyoga and the northern waters of Lake
Victoria. Journal of Fish Biology 37: 55-63.
Ogutu-Ohwayo, R. 1993. Lates spp and other predators as a cause of mortality in

small pelagic fish species in the African Great Lakes. In: B.E. Marshall and R.
Mubamba (Editors), Symposium on Biology, Stock Assessment and
Exploitation of Small Pelagic Fish Species in the African Great Lakes Region,
held at Bujumbura, Burundi, 25-28 November 1992. Committee for Inland
Fisheries of Africa, Occasional Paper No. 19: 18-23.
Ogutu-Ohwayo, R. 1995. Diversity and stability of fish stocks in Lakes Victoria,

Kyoga and Nabugabo after establishment of introduced species. In: T.J. Pitcher
and P.B. Hart (Editors), The Impact of Species Changes in African Lakes,
Chapman and Hall, London: 59-81.
Oijen, M.J.P., van Witte, F. & Witte-Maas, E.L.M. 1981. An introduction to

ecological and taxonomic investigations on the haplochromine cichlids from
the Mwanza Gulf of Lake Victoria. Netherlands Journal of Zoology 31: 149-
174.
Okedi, J. 1971. Further observations on the ecology of the Nile perch (Lates niloticus
Linne) in Lake Victoria and Lake Kyoga. East African Freshwater Fisheries
Okemwa, E.N. 1984. Potential fishery of Nile perch Lates niloticus Linne (Pisces:
Centropomidae) in Nyanza Gulf of Lake Victoria, East Africa. Hydrobiologia
108: 121-126.
Palmer, P. 1995. Developing a responsible approach to fish stocking – the

Maroochy River project. In: P. Cadwallader and B. Kerby (Editors), Fish
stocking in Queensland — Getting it right!, Quensland Fish Management
Authority, Brisbane: 79-81.
References 197
Palmer, P., Burke, J., Willett, D. And Simpson, R. 1992. Development of a low-
maintenance technique for rearing barramundi larvae. QDPI Information
Series No.QI92036, 19 pp.
Palmer, P.J., Blackshaw, A.W. and Garrett, R.N. 1993. Successful fertility
experiments with cryopreserved spermatozoa of barramundi, Lates calcarifer
(Bloch), using dimethylsulfoxide and glycerol as cryoprotectants.
Reproduction, Fertility and Development 5: 285-293.
Parazo, N.M., Garcia, L.Ma.B., Ayson, F.G., Fermin, A.C., Almendras, J.M.E., Reyes,
D.M., Jr., and Avila, E.M. 1990. Sea Bass Hatchery Operations. SEAFDEC
Aquaculture Extension Manual No. 18.38 pp.
Patnaik, S. and Jena, S. 1976. Some aspects of biology of Lates calcarifer (Bloch)
from Chilka Lake. Indian Journal of Fisheries 23: 65-71.
Pearce, M.G. 1991. Improved growth rates of hatchery-reared barramundi. Austasia

Pearson, R.G. 1987. Barramundi breeding research — laying the foundations for
industry. Australian Fisheries 46(7): 2-3.
Persson, L. 1986. Patterns of food evacuation in fishes: a critical review.

Pillay, T.V.R. 1990. Aquaculture Principles and Practices. Fishing News Books,
Oxford. 575 pp.
Piper, R.G., McElwain, I.B., Orme, L.E., McCraren, J.P., Folwer, L.G. and Leonard,
J.R. 1982. Fish Hatchery Management. US Fish and Wildlife Service,
Washington DC. 517 pp.
Pister, E.P. 1990. Desert fishes: an interdisciplinary approach to endangered species

conservation in North America. Journal of Fish Biology 37(Supplement A):
183-187.
Pitcher, T.J. and Bundy, A. 1995. Assessment of the Nile perch fishery in Lake
Victoria. In: T.J. Pitcher and P.B. Hart (Editors), The Impact of Species
Planquette, P. 1974. Recherches sur la biologie de lates niloticus (Poisson

Centropomidae), I. Donnees preliminaires sur la croissance et la reproduction
References 198
de Lates niloticus en etangs de petite superficie en Cote-divoire. Ann.

University Abidjan, series E (Ecologie) 7: 587-598.
Pollard, D.A. (Editor). 1990. Proceedings of the Workshop on Introduced and

Translocated Fishes and their Ecological Effects. Australian Society for Fish
Biology, Magnetic Island, Townsville Queensland, 24-25 August 1989,
Proceedings No. 8. Australian Government Publishing Service, Canberra. 181
pp.
Pollard, D.A., Ingram, B.A., Harris, J.H. and Reynolds, L.F. 1990. Threatened fishes
in Australia — an overview. Journal of Fish Biology 37(Supplement A): 67-
78.
Pruginin, Y. and Kanyike, E.S. 1967. Density control of Tilapia sp. population in
ponds by Lates niloticus (Nile perch). Occasional Papers of the Fisheries
Department, Uganda 1: 5-8.
Reynolds, L.F. 1978. The population dynamics of barramundi, Lates calcarifer

(Pisces: Centropomidae) in Papua New Guinea. M.Sc. Thesis, University of
Papua New Guinea. 239 pp.
Reynolds, L.F. 1983. Migration patterns of five fish species in the Murray-Darling
River system. Australian Journal of Marine and Freshwater Research 34:
857-871.
Reynolds, J.E., Greboval, D.F. and Mannini, P. 1995 Thirty years on: the development
of the Nile perch fishery in Lake Victoria. In: T.J. Pitcher and P.B. Hart
(Editors), The Impact of Species Changes in African Lakes, Chapman and Hall,
London: 181-214.
Reynolds, L.F. and Moore, R. undated. Comments on the proposed introduction of the
Nile perch (Lates niloticus (L.)) to tropical Australian waters. Unpublished
report from the Department of Agriculture, Stock and Fisheries, Papua New
Guinea.
Ribbink, A.J. 1987. African lakes and their fishes: conservation scenarios and
suggestions. Environmental Biology of Fishes 19: 3-26.
Richards, W.J. and Edwards, R.E. 1986. Stocking to restore or enhance marine
fisheries. In: R.H. Stroud (Editor), Fish Culture in Fisheries Management:
Proceedings of a Symposium on the Role of Fish Culture in Fisheries
Management, at Lake Ozark, Missouri, March 31—April 3, 1985. American
Fisheries Society, Bethesda: 59-74.
References 199
Riley, L.M. and Margraf, F.J. 1983. Scale shape as an innate tag for identification of
striped bass stocks. RF Project 762788/713991, Final Report. Ohio State
University.
Rimmer, M.A. 1996. Factors affecting survival and growth of larval barramundi Lates
calcarifer (Bloch) in aquaculture, with particular reference to feeding and
extensive rearing. Ph.D. Thesis, James Cook University of North Queensland,
359 pp.
Rimmer, M. and Rutledge, B. 1991. Extensive rearing of barramundi larvae.

Queensland Department of Primary Industries Information Series QI91012, 6
pp.
Ross, W.R. and Pickard, A. 1990. Use of scale patterns and shape as discriminators
between wild and hatchery striped bass stocks in California. In: N.C. Parker,
A.E. Giorgi, R.C. Heidinger, D.B.Jester Jr., E.D. Prince and G.A. Winans
(Editors), Fish-Marking techniques, American Fisheries Society Symposium
No. 7. American Fisheries Society, Bethesda, Maryland: 71-77.
Rowland, S.J. 1983a. The hormone-induced ovulation and spawning of the Australian
freshwater fish golden perch, Macquaria ambigua (Richardson)
(Percichthyidae). Aquaculture 35: 221-238.
Rowland, S.J. 1983b. Spawning of the Australian freshwater fish Murray cod,
Maccullochella peen (Mitchell), in earthen ponds. Journal of Fish Biology 23:
525-534.
Rowland, S.J. 1984. The hormone-induced spawning of silver perch, Bidyanus

bidyanus (Mitchell) (Teraponidae). Aquaculture 42: 83-86.
Rowland, S.J. 1986. The hormone-induced spawning and larval rearing of Australian
native freshwater fish, with particular emphasis on the golden perch,
Macquaria ambigua. In: Reynolds, L.F. (Editor), Proceedings of the First
Australian Freshwater Aquaculture Workshop, N.S.W Fisheries.
Rowland, S.J. 1988. Hormone-induced spawning of the Australian freshwater fish

Murray cod, Maccullochella peeli (Mitchell) (Percichthyidae). Aquaculture
70(4): 371-389.
Rowland, S.J. 1993. Maccullochella ikei, an endangered species of freshwater cod

(Pisces: Percichthyidea) from the Clarence River system, NSW and M peelii
References 200
mariensis, a new subspecies from the Mary River system, Qld. Records of the
Australian Museum 45: 121-145.
Rowland, S.J. 1994. Preliminary evaluation of the Australian freshwater fish silver
perch, Bidyanus bidyanus, for pond culture. Journal of Applied Aquaculture
4(1): 39-48.
Rowland, S., Dirou, J. and Selosse, P. 1983. Production and stocking of golden and
silver perch in NSW. Australian Fisheries 42(a): 24-28.
Ruangpanit, N. 1987. Developing hatchery techniques for sea bass (Lates calcarifer):
a review. In: J.W. Copland and D.L. Grey (Editors), Management of Wild and
Russell, D.J. 1987. Review of juvenile barramundi (Lates calcarifer) wildstocks in

Australia. In: J.W. Copland and D.L. Grey (Editors), Management of Wild and
1986. ACIAR Proceedings No. 20: 41 19.
Russell, D.J. 1988. An assessment of the east Queensland inshore gill net fishery.
Information Series QI88024, Queensland Department of Primary Industries. 57
pp.
Russell, D.J. 1990. Reproduction, migration and growth in Lates calcarifer (Bloch) in
eastern Queensland. M.Sc. Thesis, Qld. Uni. Technology. 194 pp.
Russell, D.J. 1991. Fish movements through a fishway on a tidal barrage in sub-
tropical Queensland. Proceedings of the Royal Society of Queensland 101:
109-118.
Russell, D.J. 1995. Measuring the success of stock enhancement programs. In: P.
Cadwallader and B. Kerby (Editors), Fish stocking in Queensland — Getting
it right!, Quensland Fish Management Authority, Brisbane: 72-78.
Russell, D.J. and Garrett, R.N. 1985. Early life history of barramundi, Lates
calcarifer (Bloch), in north-eastern Queensland. Australian Journal of Marine
and Freshwater Research 36: 191-201.
References 201
Russell, D.J. and Garrett, R.N. 1988. Movements of juvenile barramundi, Lates
calcarifer (Bloch), in north-eastern Queensland. Australian Journal of
Marine and Freshwater Research 39: 117-123.
Russell, D.J. and Hales, P.W. 1992. Evaluation of techniques for marking juvenile
barramundi, Lates calcarifer (Bloch), for stocking. Aquaculture and Fisheries
Russell, D.J. and Hales, P. 1993. A survey of the Princess Charlotte Bay recreational
barramundi fishery. Information Series Q193049, Queensland Department of
Primary Industries. 18 pp.
Russell, D.J., O'Brien, J.J. and Longhurst, C. 1987. Barramundi egg and larval culture.
Australian Fisheries 46: 26-29.
Russell, D.J. and Rimmer, M.A. 1997. Assessment of stock enhancement of

barramundi Lates calcarifer (Bloch) in a coastal river system in far northern
Queensland, Australia. In: D.A. Hancock, D.C. Smith, A. Grant and J.P.
Beumer (Editors), Developing and Sustaining World Fisheries Resources — The
State of Science and Management: Proceedings of the Second World Fisheries
Congress, Brisbane, Australia: 498-503.
Rutledge, W.P. and Rimmer, M.A. 1991. Culture of larval sea bass, Lates calcarifer
(Bloch), in salt water rearing ponds in Queensland, Australia. Asian Fisheries
Science 4: 345-355.
Rutledge, W., Rimmer, M., Russell, D.J., Garrett, R. and Barlow, C. 1990. Cost
benefit of hatchery-reared barramundi, Lates calcarifer (Bloch), in Queensland.
Aquaculture and Fisheries Management 21: 443-448.
Sagar, P.M. and Glova, G.J. 1988. Diel feeding periodicity, daily ration and prey
selection of a riverine population of juvenile chinook salmon, Oncorhynchus
tshawytscha (Walbaum). Journal of Fish Biology 33: 643-653.
Schwartzberg, M. and Fryer, J. 1993. Identification of hatchery and naturally

spawned stocks of Columbia Basin spring chinook salmon using scale pattern
analyses. North American Journal of Fisheries Management 13: 263-266.
Searcy-Bernal, R. 1994. Statistical power and aquacultural research. Aquaculture 127:

371-388.
References 202
Seehausen, 0. and Witte, F. 1995. Extinction of many, survival of some: the current
situation of the endemic cichlids in southern Lake Victoria. Tropical Fish
Hobbyist 43(7): 96-105.
Shaklee, J. B. and Salini, J.P. 1985. Genetic variation and population subdivision in
Australian barramundi, Lates calcarifer (Bloch). Australian Journal of Marine
and Freshwater Research 36: 302-18.
Shepherd, J. 1988. What is Fish Farming? In: Shepherd, J. and Bromage, N.

(Editors), Intensive Fish Farming. Blackwell Scientific Publications, Oxford:
1-16.
Sherratt, T.N. and Harvey, I.F. 1989. Predation by larvae of Pantala flavescens
(Odonata) on tadpoles of Phyllomedusa trinitatis and Physalaemus pustulosus:
the influence of absolute and relative density of prey on predator choice. Oikos
56:170 176.
-
Siegel, J. (Editor). 1992. Stastistix — Version 4.0. Analytical Software, St. Paul,
Minnesota. 319 pp.
Sirikul, B. 1982. Aquaculture for seabass in Thailand. FAO/UNDP Training Course

on Seabass Spawning and Larval Reusing, held at the National Institute of
Coastal Aquaculture (NICA), Thailand, 1-20 June 1982: 9-10.
Skurdal, J. and Andersen, R. 1985. Influence of temperature on number of circuli of

first year scales of brown trout, Salmo trutta L. Journal of Fish Biology 26:
363-366.
Sorgeloos, P., Dehasque, M., Dhert, P. and Lavens, P. 1994. Larviculture of marine
finfish: the current status. Infofish International, 4/94: 49-54.
Ssentongo, G. and Welcomme, R.L. 1984. Past history and current trends in the
fisheries of Lake Victoria. Committee for Inland Fisheries of Africa, FAO Fish
Rep. (335): 123-138.
Stoneman, J. & Rogers, J.F. 1970. Increase in fish production achieved by stocking
exotic species (Lake Kyoga, Uganda). Occasional Papers of the Fisheries
Department, Uganda 3: 16-19.
Tait, R. 1981. Comparison of the diets of the northern spotted barramundi

(Scleropages jardini) and the giant perch (Lates calcarifer) in northern
Australia. Verhandlungen Internationale Vereinigung fuer theoretische and
angwandte Limnologie 21: 1320-1325.
References 203
Tandler, A. and Helps, S. 1985. The effects of photoperiod and water exchange rate on
growth and survival of gilthead sea bream (Sparus aurata, Linnaeus; Sparidae)
from hatching to metamorphosis in mass rearing systems. Aquaculture 48: 71-
82.
Tave, D., Rezk, M. and Smitherman, R.O. 1990. Effect of body colour of
Oreochromis mossambicus (Peters) on predation by dragonfly nymphs.
Aquaculture and Fisheries Management 21: 157-161.
Tilzey, R.D.J. 1980. Introduced fish. In: W.D. Williams (Editor), An Ecological Basis
for Water Resource Management, Australian National University Press,
Canberra: 271-279.
Trendall, J. 1991. Reduce your chances of being a victim of cannibalism. Barramundi

Tucker, J.W., Jr., MacKinnon, M.R., Russell, D.J., O'Brien, J.J. and Ca7zola, E. 1988.
Growth of juvenile barramundi (Lates cakarifer) on dry feeds. Progressive
Fish-Culturist 50(2): 81-85.
Twongo, T. 1988. Recent trends in the fisheries of Lake Kyoga — Uganda. In: D.
Lewis (Editor), Predator-prey relationships, population dynamics and fisheries
productivities of large African lakes. CIFA Occasional Paper No. 15: 140-
151.
van-Buskirk, J. 1989. Density-dependent cannibalism in larval dragonflies. Ecology

70(5): 1442-1449.
Vanden Bossche, J.-P. and Bernacsek, G.M. 1990. Source book for the inland fishery
resources of Africa — Volume 1. CIFA Technical Paper No. 18/1. 411 pp.
Vanderpuye, C.J. and Ocansey, P. 1972. A preliminary study on the biology of Lates
niloticus (Linnaeus 1762) in Volta Lake. Internal report, Department of
Fisheries, Ghana.
Verreth, J. and van Tongeren, M. 1989. Weaning time in Clarias gariepinus

(Burchell) larvae. Aquaculture 83: 81-88.
Volk, E.C., Schroder, S.L. and Fresh, K.L. 1987. Inducement of banding patterns on
the otoliths of juvenile chum salmon. In: Proceedings of the 1987 Northwest
Pacific Pink and Chum Salmon Workshop, Alaska Department of Fish and
Game, Anchorage, Alaska: 206-217.
References 204
Walford, J. and Lam, T.J. 1993. Development of digestive tract and proteolytic
enzyme activity in seabass (Lates calcarifer) larvae and juveniles. Aquaculture
109: 187-205.
Wanink, J.H. and Goudswaard, P.C. 1994. Effects of Nile perch (Lates niloticus)
introduction into Lake Victoria, East Africa, on the diet of pied kingfishers
(Ceryle rudis). Hydrobiologia 279/280: 367-376.
Weatherly, A.H., and Lake, J.S. 1967. Introduced fish species in Australian waters.
In: A.H. Weatherley (Editor), Australian Inland Waters and their Fauna,
Australian National University Press, Canbena: 217-239.
Whaley, R. 1988. Separation of two brown trout strains using scale patterns.
Administrative Report, Wyoming Game and Fish Department.
Whaley, R.A. 1991. An improved technique for cleaning fish scales. North
American Journal of Fisheries Management 11: 234-236.
Whitley, G.P. 1959. The barramundi, north Australia's finest food fish. Australian
Museum Magazine 13: 55-58.
Wiggins, T.A., Bender, T.R., Mudvak, V.A. and Coll, J.A. 1985. The development,
feeding, growth and survival of cultured American shad larvae through the
transition from endogenous to exogenous nutrition. Progressive Fish-Culturist
47: 87-93.
Willett, D.J. 1993. Discrimination between hatchery stocks of silver perch, Bidyanus
bidyanus (Mitchell), using scale growth patterns. Aquaculture and Fisheries
Willett, D. 1994a. Fish scales — a natural fish tag. Freshwater Fishing Australia
26: 30-32.
Willett, D.J. 1994b. Use of temperature to manipulate patterns on scales of silver

perch, Bidyanus bidyanus (Mitchell), for the purpose of stock discrimination.
Fisheries Management and Ecology 1: 157-163.
Willett, D.J. 1995. Use of a scale pattern recognition system to discriminate between
stocks of fish and its implications for the management of inland recreational
fisheries. M.Sc. Thesis, Queensland University of Technology.
References 205
Willett, D.J. 1996. Use of scale patterns to evaluate stocking success of silver perch,
Bidyanus bidyanus (Mitchell), released at two different sizes. Marine and
Williams, W.D. 1970. On the proposed introduction of Lates niloticus (L.) to

Australia. Australian Society for Limnology Bulletin 3: 33-35.
Williams, W.D. 1980. Australian Freshwater Life — The Invertbrates of Australian

Inland Waters. MacMillan, South Melbourne, 321pp.
Williams, W.D. 1982. The argument against the introduction of the Nile perch to
northern Australian fresh waters. Search 13: 67-70.
Witte, F., Goldschmidt, T. and Wanink, J.H. 1995. Dynamics of the haplochromine
cichlid fauna and other ecological changes in the Mwanza Gulf of Lake
Victoria. In: T.J. Pitcher and P.B. Hart (Editors), The Impact of Species
Witte, F., Goldschimdt, T., Wanink, J., Oijen, M., Goudswaard, P.C., Witte-Maas, E.
and Bouton, N. 1992a. The destruction of an endemic species flock:
quantitative data on the decline of the haplochromine cichlids of Lake Victoria.
Witte, F., Goldschimdt, T., Goudswaard, P.C., Ligtvoet, W., Van Oijen, M., Wanink, J.
1992b. Species extinction and concomitant ecological changes in Lake
Victoria. Netherlands Journal of Zoology 42(2-3): 214-232.
Wongsomnuk, S and Manevonk, S. 1973. Results of experiment on artificial breeding

and larval rearing of seabass (Lates calcarifer Bloch). Songkhla Marine
Fisheries Station, Songkhla, Thailand. Contribution No. 5,20 pp.
Worthington, E.B. 1973. The ecology of introductions — a case study from the
African lakes. Biological Conservation 5: 221-222.
Wydoski, R. and Emery, L. 1983. Tagging and marking. In: L. A. Neilsen and D. L.
Johnson (Editors), Fisheries Techniques, American Fisheries Society,
Bethesda, Maryland: 215-237.
Yamashita, Y., Nagahora, S., Yamada, H. and Kitagawa, D. 1994. Effects of release
size on survival and growth of Japanese flounder Paralichthys olivaceus in
coastal waters off Iwate Prefecture, northeastern Japan. Marine Ecology
Progress Series 105(3): 269-276.
Appendices 206
APPENDIX 1
Economic comparison of weaning at 13 mm and 17 mm TL
A steady-state economic comparison was conducted to determine the difference in
costs associated with weaning at 13 mm and 17 mm T.L. A steady-state comparison
is an analytical technique used in economics to determine the cost differential
associated with two or more different management techniques. It takes into account
only the costs which differ, thus it is not an estimate of the total costs associated with
the technique (in the present case, it is not an estimate of the total costs of fingerling
production).
For the purposes of the comparison, it was assumed that both management regimes
(ie., 13 mm and 17 mm weaning) produced a total of 530 000 fingerlings, average
size 31 mm TL, from five batches of fish (ie., 106 000 fish per run) during one
breeding season. All biological estimates relating to growth rates, feeding rates and
mortalities were derived from data in Chapters 3 and 6. Operational costs (feed,
labour and enrichment diets) and fixed costs (capital items and depreciation rates)
were obtained from equipment suppliers and information provided by commercial
barramundi farmers.
A list of the costs associated with each weaning regime is given in Table A1.
Summation of costs indicates that there is a considerable economic advantage in

Appendices 207
delaying weaning until the barramundi fry are about 17 mm TL. Under the scenario
assumed in this comparison, there is a saving of approximately $70 000 per annum in
initiating weaning at 17 mm rather than 13 mm TL.
Table Al. An estimate of costs associated with producing 530 000 barramundi
per season, with fry being weaned onto dry, formulated diets at either
13 mm or 17 mm TL.
Item Costs ($) associated with weaning at

13 mm TL 17 mm TL
Variable Costs
Mortalities' 84 211 9 755
Live feed - Artemia
Enrichment diets for Artemia 0 1 000
Dry diet (weaning) 3 525 3 045
Dry diet (post-weaning) 482 0
Casual labour @ $11/hr 0 413
Other 0 0
Fixed Costs
Administration 0 0
Electricity 0 0
Sundry fuel & oil 0 0
Permanent hired labour 0 0
Repairs and maintenance 0 0
Depreciation 247 56
Other 0 0
Operating Costs 88 465 19 419

plus interest on capital base (8%) 169 39
plus interest on working capital base (8%) 0 0
plus owner/operator costs 0 0
Adjusted operating costs 88 634 19 458
Net difference 69 176
' Mortality data

13 mm fish, 55% mortality through weaning. Therefore, (530 000 x 100/45) =
1 177 778 fry required at initiation of weaning. Fish are valued at lc/mm. Assuming fish die
at 13 mm (ie., they do not initiate feeding on the dry diet), value of mortalities = (1 177 778 x
55/100 x 0.13) = $84 211.
17mm fish, 8% mortality through weaning plus 2.5% mortality during 13-17 mm Artemia
feeding stage.
Therefore (530 000 x 100/92) = 576 087 fry required at initiation of weaning.
Assuming fish die at 17 mm, value of mortalities through weaning = (576 087 x 8/100 x 0.17)
= $7835.
Appendices 208
At the start of the 13-17 mm growing period, required (576 087 x 100/97.5) = 590 828.
Assuming fish die at 13 mm, value of mortalities = (590 828 x 2.5/100 x 0.13) = $1920.
Therefore, total value of mortalities = $7835 + $1920 = $9755

Appendices 209
APPENDIX 2
Published Papers
This Appendix contains copies of four papers which have been published based on
results reported in this thesis. The papers are as follows.
Barlow, C.G. and Lisle, A. 1987. Biology of Nile perch Lates niloticus
(Pisces:Centropomidae) with reference to its proposed role as a sport fish in
Australia. Biological Conservation 39: 269-289.
Barlow, C.G. and Gregg, B.A. 1991. Use of circuli spacing on scales to discriminate
hatchery and wild barramundi, Lates calcarifer (Bloch). Aquaculture and
Fisheries Management 22: 491-498.
Barlow, C.G., Rodgers, L.J., Palmer, P.J. and Longhurst, C.J. 1993. Feeding habits
of hatchery-reared barramundi Lates calcarifer (Bloch) fry. Aquaculture 109:
131-144.
Barlow, C.G., Pearce, M.G., Rodgers, L.J. and Clayton, P. 1995. Effects of
photoperiod on growth, survival and feeding periodity of larval and juvenile
barramundi Lates calcarifer (Bloch). Aquaculture 138: 159-168.

02 Whole

Uploaded by

Copyright:

Available Formats

02 Whole

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

02 Whole

Uploaded by

Copyright:

Available Formats

ResearchOnline@JCU

This file is part of the following reference:

Barlow, Christopher G. (1998) Aspects of the biology of

Access to this file is available from:

for the degree of Doctor of Philosophy in

James Cook University of North Queensland

Christopher G. Barlow 15 January 1998

The research covered in this thesis concentrated primarily on improving production

feeding habits, pond rearing techniques, effects of photoperiod on growth, and

recreational fisheries enhancement programs. Within this context, two related

to Australia, which preceded the barramundi production studies; and secondly, an

A review of the historical and present distribution of barramundi (a catadromous

decreasing distribution of barramundi, it was proposed that a congener of

established would be capable of sustaining breeding populations.

lower temperature tolerance of the species and analysis of water temperatures in

introduction of the Nile perch, an opportunistic predator, to Lakes Victoria and

it was abandoned. As an alternative, attention was given to hatchery production of

barramundi as a means of supplying fingerlings for stocking fresh waters for

enhancement of recreational fisheries in northern Australia.

Hatchery-reared barramundi fry were studied to determine feeding behaviour, diel

from a roving zooplanktivore to that of a lurking predator. A distinct change in

pigmentation accompanied the change in feeding behaviour. Fry reared in hatchery

showed a peak in feeding activity at dusk. Feeding continued at a reduced level

Laboratory-based experiments were conducted to determine the vulnerability of

different sized barramundi fry to predation by nymphs of the dragonfly Pantala

flavescens. Mortality of 10 mm mean TL fry was significantly greater in the

presence of dragonfly nymphs, whereas 20 mm mean TL fry exhibited

comparatively minor levels of mortalities. The results accorded with feeding

response in barramundi fry at 16-18 mm TL. An examination of the growth rates of

optimise survival of barramundi fry reared in freshwater ponds.

An experiment was conducted to determine the effect of extended periods of light on

reared in a freshwater hatchery. There was no significant difference in growth or

34 mm fish was approximately 40% more in continuous light than in 12L/12D

rearing barramundi fry in extended light regimes.

An experiment was conducted to determine if survival during weaning was affected

of live zooplankton was discontinued and a commercially available salmon starter

97 and 92%, respectively. An asymptotic curve described the relationship between

to a lurking predator. Cost-benefit analyses indicated a considerable economic

saving in delaying weaning until the fry are 16 mm TL.

A study was conducted to determine if hatchery-reared and wild barramundi could be

distinguished by the patterns of circulus spacing on the scales. Proprietary software

and digitising equipment was used to obtain measurements of circulus spacing

efficacy of large-scale enhancement programs. However, because scales from fish

Statement on Access to Thesis ii

CHAPTER 1. General Introduction 1

CHAPTER 2. The Biology and Exploitation of Barramundi and the Proposal to

CHAPTER 3. Feeding Habits of Hatchery-Reared Barramundi Fry 69

CHAPTER 4. Predation by Dragonfly Nymphs of the species Pantala

CHAPTER 5. Effects of Photoperiod on Growth and Feeding Periodicity

CHAPTER 8. Synthesis 162

Figure 1.1 Schematic representation of the various procedural steps involved 13

Christopher G. Barlow 15 January 1998

My colleagues at the Freshwater Fisheries and Aquaculture Centre, Walkamin. In

Colleagues at the Northern Fisheries Centre, in particular Mike Rimmer, Rod

To the many industry colleagues (barramundi farmers, researchers and anglers) on

I particularly thank my family for their patience, encouragement and support in

CHAPTER 1. GENERAL INTRODUCTION

1.1 PRODUCTION OF FINGERLING FISH

1.1.1 Reasons for production

ornamental fish trade;