SEASONALITY AND FISH REPRODUCTION
IN AN INTERMITTENT STREAM
A Thesis
Submitted in fulfillment of the Requirement
for the degree of
Doctor of Philosophy
of
The University of the West Indies
Mary Edith Helda Alkins
1987
Department of Zoology
Faculty of Natural Sciences
St Augustine
i i
ABSTRACT
Seasonality and fish reproduction in an intermittent stream.
Mary Edith Helda Alkins
During the period 1980 to 1986, the adjacent Quarahoon
and Carlisle Rivers in the southwestern peninsula of Trinidad
were studied with reference to seasonal effects on limnology
and reproductive strategies of six fish species: GasteropeZecus
eternicl.a, Commopoma »i.ieei, Astyanax bimaculatue , Hemi.qramnue
uniZineatus, Corydoras aeneus and PoeciZia reticuZata.
Stream flow regimes were intermittent reflecting a main
dry season from January to May and a main rainy season from
June to December with a minor dry period between September and
November. These streams were considered long-flow intermittent
streams with substantial refuge areas capable of maintaining
diverse faunal assemblages. The fauna was characteristic of
lentic or slow-flow conditions and Was well adapted to survive
conditions of stagnation and drought. Seasonal flow influenced -
variation of physical and chemical features and the biota.
Production of plankton increased significantly during prolonged
lentic periods and decreased with floods. Benthic standing
iii
crop and allochthonous input showed wide variation in both dry
and rainy seasons.
Species richness of fish communities was high and may have
resulted from intermediate disturbance levels and proximity to
the mainland. Seasonal flow regimes strongly influenced fish
population sizes, life history characteristics and reproductive
strategies. Population fluctuations appeared to be due to
mortality, dispersal, concentration and input of juveniles during
breeding seasons. Life history traits were characteristically
r-selected. Peaks of reproduction in five species coincided with
the rainy seasons while P. retieuZata bred continuously. Numbers
of breeding seasons each year and their lengths varied from one
short major season with rarely a second each year to long
seasons twice each year. Spawning patterns also varied from
a highly synchronised total spawning pattern to continuous
small brood production. Maximum batch fecundities were
directly proportional to species size but smaller species
achieved high fertil ities by multiple spawning which allowed
high but variable reproductive output. Reproductive strategies
were diverse but highly adaptable to fluctuating conditions.
iv
ACKNOWLEDGEMENTS
Many individuals and organisations assisted in this work
in some way and I would like to thank them all sincerely
although I could not possibly name them all.
In particular, I would like to thank Mr L. James of
TRINTOC for arranging access to private lands and accommodation
in the study area; Mr B. Barclay for supplying rainfall data;
personnel at the Central Laboratory, WASA, for analysis of
water samples; personnel at CAROl Soil Physics Laboratory for
analysis of sediment samples; and Mr B. Lauckner of the UWI
Computer Centre for assistance in data analysis.
Assistance in the field by the technicians of the Zoology
Department as well as Mr G. de Souza and Mr M. Koo, is very
sincerely appreciated especially considering the many hours
spent battling the most vicious mosquitoes encountered to
date.
I would also like to thank the following for identification
of specimens from the study area: Dr P.R. Bacon, Dr M. Boeseman,
Dr E.J. Duncan, Ms J. Gobin, Prof J.S. Kenny, Mr R. Martinez,
The National Herbarium, Dr N. Nieser, Dr W.L. Peters,
v
Dr P.J. Spangler, Dr S. Weitzman and Prof M.J. Westfall, Jr.
For a very professional job of typing the final draft,
I would like to thank Mrs J. Martin. Photography by
Prof J.S. Kenny and Mr M. Koo is also appreciated.
Finally, I am most grateful for the supervision provided
by Prof J.S. Kenny and for advice and encouragement from
Dr R.W. Bruce and Dr D.L. Kramer.
vi
TABLE OF CONTENTS
Page
r"1
ABSTRACT ii
ACKNOWLEDGEMENTS iv
TABLE OF CONTENTS vi
LIST OF FIGURES ix
LIST OF TABLES xv
LIST OF PLATES xviii
GENERAL INTRODUCTION 1
THE STREAM ENVIRONMENT 15
Introduction 15
The study site 22
Methods 36
Hydrology and physico-chemical
characteristics 36
Rainfall 36
Physico-chemical characteristics 36
Stream biota 40
Plankton 40
Benthic macroinvertebrates 41
Allochthonous input 43
Results 44
Hydrology and physico-chemical
characteristics 44
vii
Page
Rainfall 44
Physical parameters 50
Chemical parameters 62
Diurnal variation 87
Stream biota 97
General observations 97
Plankton 101
Benthic macroinvertebrates 106
Allochthonous input 117
Discussion 121
THE FISH 153
Introduction 153
Review of the species 158
Methods 170
General sampling and morphometric studies 170
Population studies and reproductive
seasonality 173
Fecundity and spawning patterns 179
Results 184
Species composition and distribution 184
Population studies and reproductive
seasonality 191
Population fluctuations 192
Morphometric studies 202
I
.I
viii
Page
Population demography and reproductive
seasonality 212
Fecundity and spawning patterns 283
Ovarian fecundity estimates 283
Laboratory spawning experiments 289
Oocyte size distribution analyses 304
Discussion 315
CONCLUSIONS 362
BIBLIOGRAPHY 367
APPENDICES 395
ix
LIST OF FIGURES
Page.
1. Map showing (a) the location of the study
area in Trinidad and in relation to South
America, and (b) the hydrology of the
southwestern peninsula. 23
2. Map of the study area showing sampling
stations. 26
3. Monthly rainfall totals from January 1980
to July 1982. 48
4. (a) Weekly rainfall totals from January
1980 to July 1982, and (b) pentade analysis
of rainfall from March 1978 to July 1982. 49
5. Monthly variation of air and water tempera-
tures at Stations 1 to 4. 53
6. Monthly variation of depth and width at
Stations 1 to 4. 55
7. Monthly variation of current and discharge
at Stations 1 to 4. 56
8. Monthly variation of turbidity and colour
at Stations 1 to 4. 57
9. Monthly variation of (a)-(d) particle size
composition, and (e) carbon content of
substrates at Stations 1 to 4. 58
10. Monthly variation of pH at Stations 1 to 4. 68
11. Monthly variation of (a) total alkalinity,
and (b) bicarbonate alkalinity at Stations
1 to 4. 69
12. Monthly variation of specific conductance
at Stations 1 to 4. 70
13. Monthly variation of total hardness at
Stations 1 to 4. 71
x
Page
14. Monthly variation of (a) calcium, and
(b) magnesium ions at Stations 1 to 4. 72
15. Monthly variation of (a) total iron,
and (b) soluble iron at Stations 1 to 4. 73
16. Monthly variation of (a) chloride, and
(b) sulphate ions at Stations 1 to 4. 74
17. Monthly variation of (a) silicate, and
(b) phosphate at Stations 1 to 4. 75
18. Monthly variation of (a) nitrites, and
(b) nitrates at Stations 1 to 4. 76
19. Monthly variation of (a) total dissolved
solids, and (b) total solids at Stations
1 to 4. 77
20. Monthly variation of (a) permanganate
value, and (b) BOD at Stations 1 to 4. 78
21. Monthly variation of (a) dissolved
oxygen, and (b) percentage saturation
of oxygen at Stations 1 to 4. , 79
22. Hourly variation of physical and chemical
parameters at Station 2 in the rainy
season (80/08/07-08). 88
23. Hourly variation of physical and chemical
parameters at Station 2 in the 1982 dry
season (82/03/31-04/01). 89
24. Hourly variation of temperature and
dissolved oxygen at Station 2 in the
1981 dry season (81/03/25-26). 90
25. Hourly variation of physical and chemical
parameters at Station 4 in the rainy
season (80/08/19-20). 94
26. Hourly variation of physical and chemical
parameters at Station 4 in the dry season
(81/04/01-02). 95
xi
xii
Page
40. Monthly variation of total and somatic
condition forGo sternicla (a) females,
and (b) males. 227
41. The relationship between gonad maturity
stages and size for C. riisei (a) females
and juveniles, and (b) males. 229
42. Monthly variation of population structure
and the occurrence of gonad maturity
stages for C. riisei. 231
43. Growth curve for C. riisei. 232
44. Monthly variation of gonad maturity stages
for C. riisei (a) females, and (b) males. 234
45. Monthly variation of GSI's for (a) females,
and (b) males, and (c) sex ratios for
C. riisei. 236
46. Monthly variation of total and somatic
condition for C. riisei (a) females, and
(b) males. 238
47. The relationship between gonad maturity
stages and size for A. bimaculatus (a)
females and juveniles, and (b) males. 240
48. Monthly variation of population structure
and the occurrence of gonad maturity
stages for A. bimaculatus. 242
49. Growth curve for A. bimaculatus. 243
50. The relationship between gonad maturity
stages and size for H. unilineatus (a)
females and juveniles, and (b) males. 249
51. Monthly variation of population structure
and the occurrence of gonad maturity stages
for H. unilineatus. 251
52. Growth curve for H. unilineatus. 252
xiii
Page
53. Monthly variation of gonad maturity
stages for H. unilineatus (a) females,
and (b) males over the whole study
period. 254
54. Monthly variation of gonad maturity
stages for H. unilineatus (a) females,
and (b) males over the calendar year. 255
55. Monthly variation of GSI's for (a)
females, and (b) males, and (c) sex
ratios for H. unilineatus over the
whole study period. 257
56. Monthly variation of total and somatic
condition for H. unilineatus (a)
females, and (b) males over the whole
study period. 260
57. The relationship between gonad maturity
stages and size for C. aeneus (a) females
and juveniles, and (b) males. 263
58. Monthly variation of population structure
and the occurrence of gonad maturity
stages for C. aeneus. 265
59. Monthly variation of gonad maturity
stages for C. aeneus (a) females, and (b)
males. 267
60. Monthly variation of GSI's for (a) females,
and (b) males, and (c) sex ratios for
C. aeneue . 268
61. Monthly variation of total and somatic
condition for C. aeneus (a) females,
and (b) males. 271
62. The relationship between gonad maturity
stages and size for P. reticulata (a)
females and juveniles, and (b) males. 273
63. Monthly variation of population structure
and the occurrence of gonad maturity
stages for P. reticulata (a) females and
juveniles, and (b) males. 275
xiv
Page
64. Monthly variation of gonad maturity
stages for P. reticulata (a) females,
and (b) males over the calendar year. 277
65. Monthly variation of gonad maturity
stages for P. reticulata (a) females,
and (b) males over the whole study 279
period. .
66. Monthly variation of (a) female GSI's,
and (b) sex ratios for P. reticulata
over the whole study period. 280
67. Monthly variation of (a) female total
and somatic condition, and (b) male
total condition for P. reticulata over
the whole study period. 282
68. Batch fecundity-length relationships
for (a) G. sternicla~ (b) C. riisei,
and (c) H. unilineatus. 285
69. (a) Batch fecundity-length relationship
for C. aeneus~ and (b) brood fecundity-
length: relationship for P. reticulata. 286
70. Laboratory spawning patterns for
C. riisei. 296
71. Laboratory spawning patterns for
H. unilineatus. 300
72. Oocyte size distribution analyses for
G. sternicZa. 305
73. Oocyte size distribution analyses for I
C. riisei. 308
74. Oocyte size distribution analyses for
A. bimaculatus. 310
75. Oocyte size distribution analyses for
H. uniZineatus. 312
76. Oocyte size distribution analyses for
C. aeneus. 314
xv
LIST OF TABLES
Page
1. Rainfall averages for Chatham from 1933 _
1982. 45
2. Rainfall extremes for Chatham from 1953 -
1982. 46
3. Means and ranges of physical parameters
for Stations 1 to 4. 51
4. Average particle size composition and
carbon content of substrates at Stations
1 to 4. 61
5. Ranges and means of chemical parameters
for Stations 1 to 4. 63
6. Summary of aquatic macrofauna (excluding
fishes) collected at Stations 1 to 4
during the study period. 99
7.. General trends in the seasonal variation
of plankton composition and relative
abundance at Stations 1 to 4. 103
8. Comparison of mean densities of planktonic
organisms in dry and rainy seasons at
Stations 1 and 2. 105
9. Monthly variation of benthic macroinver-
tebrates at Station 1. 107
10. Monthly variation of benthic macroinver-
tebrates at Station 2. 108
11. Monthly variation of benthic macroinver-
tebrates at Station 3. 109
12. Monthly variation of benthic macroinver-
tebrates at Station 4. 110
13. List of macrofauna collected from damp
stream bed at Station 1 in May 1980. 112
xvi
Page
14. Comparison of mean densities and biomass
of.benthic macroinvertebrates in dry and
ralny seasons at Stations 1 and 2. 113
15. Comparison of mean numbers and biomass
of organisms caught on glue boards during
dry and rainy seasons at Stations 1 and 2. 118
16. Comparison of selected water quality
parameters of world and South American
'average' river waters and those of the
freshwater study sites. 126
17. Classification of the six fish species
studied. 159
18. Gross morphological characteristics
used in the classification of maturity
stages for the six fish species studied. 174
19. Mature oocyte size ranges used in
classifying female reproductive state and
minimum oocyte sizes used in batch
fecundity estimates for the six fish
species studied. 175
20. Taxonomic list of teleost fish collected
in the Chatham streams during the study
period. 185
21. Range of standard lengths and total
weights recorded during the study period
for the six fish species studied. 203
22. Sexual dimorphism in selected body
proportions for mature specimens of
G. sternicZa~ H. uniZineatus and
C. aeneus. 205
23. Parameters of the length-weight relation-
ship for preserved male and ~emale
specimens belonging to the SlX study
species. 209
xvii
Page
24. Minimum Developing and Mature standard
lengths and Median Length at first
maturity for the six study species. 214
25. Sex ~atios for the six fish species
studled based on collections made
over the study period. 226
26. Monthly variation of GSI's for
A. bimaculatus. 245
27. Monthly variation of mean total and
somatic condition factors for
A. bimaculatus. 247
28. Parameters of the fecundity-length
relationship for five of the species
studied. 287
29. Maximal batch/brood fecundities recorded
for the six fish species studied. 290
30. Summary of results of laboratory
spawning experiments for C. riisei. 294
31. Summary of results of laboratory
spawning experiments for H. unilineatus. 298
32. Comparison of batch fecundity and total
fertility as numbers of eggs and as egg
weight : female body weight ratio for .
C. riisei and H. unilineatus individuals
of different sizes. 302
33. Summary of life history parameters and
reproductive characteristics of the
six fish species studied. 325
xviii
LIST OF PLA.TES
Page
1. View upstream from Station 1 in May
after the first rains. 30
2. View downstream from Station 1 in May
after the first rains. 31
3. Station 2 in the late dry season. 33
4. Upstream view of Station 3 in the late
dry season. 34
5. Upstream view of Station 4 in the late
dry season. 35
6. The six fish species studied: (a)
G. sternicla~ (b) c. r~~se~~ (c) A.
bimaculatus~ (d) H. unilineatus~
(e) C. aeneus~ and (f) P. reticulata. 160
1
GENERAL INTRODUCTION
Seasonal changes in tropical regions were until
recently, widely believed to be absent because of minimal
variation in temperature and the duration and intensity of
sunlight, all important factors determining seasonality in
temperate regions (Schwassmann 1980). However, it is now
well recognised that rainfall is critical in many tropical
areas and the alternating wet and dry seasons have strong
effects on the structure and functioning of biological
communities (Dobzhansky 1950, Farnworth & Golley 1974,
Krebs 1985). In addition, with increasing latitude away
from the equator, the total amount of rainfall per year
decreases and the seasonality of its distribution during the
year varies from being almost non-seasonal at the equator
to having distinct wet and dry seasons in mid-tropical
latitudes (Jackson 1977). Superimposed on this general
climatic scheme are the localised effects of the positions
of the sub-tropical high-pressure cells, oceans, the shape
and extent of land masses or seas, and varying characteristics
of the trade winds and the equatorial trough. Thus, a
climatically heterogenous situation exists within the
tropics with some regions having a more pronounced seasonal
cycle than others with respect to rainfall.
2
The timing of onset of animal reproduction and its duration
are both closely tied to environmental variables in order to
maximise reproductive success. Many tropical species
belonging to a variety of groups show extended breeding
seasons consistent with favourable year-round environmental
conditions (reviews in Kramer 1978a and UNESCO/UNEP/FAO 1978).
In contrast, many other species have well defined annual
breeding cycles (Bullough 1961, review in UNESCO/UNEP/FAO
1978). Such variation in breeding may be related to studies
being carried out in areas of unequal seasonal variation (Kramer
1978a, Schwassmann 1980, Hails & Abdullah 1982) or may be
due to biotic factors such as competition (Bertness 1981)
or varying food availability (Dobzhansky 19~0, Ricklefs
1966, Karr 1976) controlling the timing of reproductive
cycles. It is felt by some biologists that biotic factors play
a dominant role in regulating the functioning of many tropical
communities (Robinson 1978) but the role of abiotic factors
may also be influential in many respects (for example Stout
1982).
The reproductive timing of tropical freshwater fishes
also mirrors the variability seen in other groups. In an
extensive review of the ecology of tropical freshwater
fishes, Lowe-McConnell (1975) documented a range of
reproductive strategies involving, among other aspects,
3
variation in spawning frequency, fecundity, migration
patterns and parental care. Throughout the review, she
stressed the seasonal nature of reproduction and coincidence
of the time of spawning of fish in lotic habitats with the
annual rains and resulting floods. This is supported by
other workers such as de Vlaming (1974), Billard & Breton
(1978) and others reviewed in Lam (1983). Such timing may
be adaptive since rising water levels increase available
habitat, reduce predator densities and favour increased
primary and secondary productivity for juvenile feeding.
Kramer (1978a) pointed out that this conclusion is largely
based on studies carried out in floodplains or large rivers
which are both highly seasonal habitats ..
Schwassmann (1978, 1980) attempted to summarise more
recent studies by distinguishing between seasonal cycles of
breeding in floodplain lakes and rivers and those of small
forest streams. The former he described as highly seasonal
environments where fish have very seasonal breeding cycles.
Most of the fish in these habitats are total spawners (all
eggs ripen synchronously and are spawned in one batch) in
order to take advantage of a rapidly expanding and highly
productive habitat. In contrast, small forest streams are
permanent with no appreciable seasonal differences in water
4
level. As a result, fish living in this habitat are multiple
spawners (eggs ripen and develop in several batches during one
reproductive season) and they may spawn for extended periods.
However, in a detailed study of six fish species in a mildly
seasonal Panamanian forest stream, Kramer (1978a) found
'considerable diversity in reproductive periodicity' (p. 982).
Some species showed breeding periods coincident with the rainy
season, others with the dry season, while the duration of breeding
ranged from one to two days per year to almost continuous breeding.
He proposed a number of hypotheses relating breeding
periodicity to biotic rather than climatic causal factors.
In summary, controlling factors proposed were: adult or
juvenile food availability, competition for food among
juveniles, competition for breeding sites, seasonality
acting as a mechanism for reproductive isolation, and the
phylogeny of the species. Nikolskii (1961) had also
proposed previously that biotic factors could influence
reproductive timing in fishes: year-round food availability
would allow extended breeding periods, but conflicts over
spawni~g grounds or food for young would result in different
periods of reproduction. No data were given in support of
these ideas however. In a review of seasonality in tropical
fishes, Lowe-McConnell (1979) pointed out several instances
where biotic factors may determine the breeding time of
5
riverine fishes.
More recent studies of reproductive cycles in tropical
freshwater fishes do not clarify the situation. Many show
seasonal peaks of breeding activity coincident with the rains
(Tan 1980, Hails & Abdullah 1982, Orr & Milward 1984,
Cambray & Bruton 1984). In contrast however, Beumer's (1979)
study of two species in a seasonally flooding Australian
stream described two contrasting strategies: one species
exhibiting a short well-defined spawning period coincident
with the floods, the other showing a protracted spawning
period over most of the year.
Some studies have elucidated the nature of the factors
controlling gonad development and spawning in tropical
freshwater fishes. These include field observations
suggesting that floods trigger spawning activity of Indian
carp (Khanna 1958), while laboratory experiments simulating
rain in association with decreased pH and specific conductivity,
and increased water levels, show that these variables playa
definitive role in initiating gonad development and spawning
in a gymnotoid fish (Kirschbaum 1975). Social factors such
as pheromones or visual stimuli have also been implicated in
promoting ovulation (review in Lam 1983). Lam (1983)
6
concluded that the environmental control of fish reproductive
activity is much more complex in tropical than temperate
species. In addition, the proximate cues for initiating
gametogenesis, spawning and gonadal regression will not
necessarily be the same because of the different times of
year when these activities occur. Much more information
is therefore needed for tropical fishes to determine the
proximate cues controlling their annual cycles, in addition
to determination of the ultimate factors selecting for
different patterns of seasonality in their reproduction
(Kramer 1978a, UNESCO/UNEP/FAO 1978).
In summarising the ecological characteristics of fish
communities in different tropical freshwater habitats,
Lowe-McConnell (1975) characterised the African Great Lakes
as being very stable and savanna floodplain rivers as very
seasonal with upper layers of lakes and equatorial forest
rivers lying between on a continuum of seasonality from
stable to seasonal. She described forest rivers as showing
twice yearly variation of two high and two low periods per
year but rain generally occurs all year whereas floodplain
rivers show one annual flood. Both show lateral flooding
into the surrounding forest or savanna to increase habitat
size during the wet season. Although intermittent streams
7
were not included in Lowe-McConnell IS seasonal spectrum, it
can be expected that they should lie at the most seasonal
end of the continuum because of their drastic flood/drought
characteristics (Williams & Hynes 1977). However, compared
with floodplain and forest rivers, the nature of habitat
variation is quite different. For example, although the
available aquatic habitat expands in the wet season, it
also changes from previously lentic conditions (dry season
pools) to lotic conditions. Current velocity is a major
factor affecting aquatic communities (Hynes 1970) and the
sudden increase in current velocity and water levels in the
wet season can result in reduced fish and invertebrate
populations (Stehr & Branson 1938, Paloumpis 1958, John 1964).
In comparison, floodplain rivers expand laterally in the
wet season onto surrounding low-lying savanna or forest
resulting in an increased habitat with lentic characteristics
which fish then utilise for feeding or spawning (Lowe-
McConnell 1975, 1979, Welcomme 1979, 1985). In the dry
season, intermittent streams are affected by drought
conditions causing a rapid decrease in water level leading
either to the formation of remnant pools or complete drying
out. It is well documented that in these conditions
increased mortality of both fish and invertebrate fauna may
occur, especla. 11y l·f complete drying takes place (Paloumpis
8
1958, Larimore et al 1959, Williams & Hynes 1977) or when fish
become concentrated into remnant pools and suffer heavy losses
from aquatic and terrestrial predators (Larimore et al 1959,
John 1964), a situation not unlike floodplain savanna pools.
Lentic conditions develop at this time however, and crops of
plankton, algae and other organisms adapted to stagnant
conditions, as well as opportunistic species may colonise
these remnant pools (Larimore et al 1959, Williams & Hynes
1977, Extence 1981) and secondary productivity may increase.
The increase in food availability and reduction in current
velocity during the dry period may be favourable for
breeding for some species as compared to the wet flood
season (Lowe-McConnell 1979). Therefore, although intermittent
streams may show a similar range of seasonal variability
as floodplain rivers, the differences outlined above may
impose different selective pressures on the breeding biology
of the component species.
In temperate intermittent streams, the timing of fish
reproduction varies: some species spawn with the floods
(normally in spring), others do not seem to have any
correlation with floods, yet others spawn in the stagnant summer
pools (Paloumpis 1958, Deacon & Minckley 1974, Williams &
Coad 1979). Certain advantages exist for reproduction and
9
juvenile survival in intermittent streams as compared to per-
manent ones including early spring breeding being possible,
plentiful food and reduced predation in summer pools (Williams
& Coad 1979). However John (1964) noted that prolonged
drought prevented reproduction because of reduced availability
of food. The duration of the breeding cycle may also be
unpredictably long for such a seasonal environment. Deacon &
Minckley (1974) and Williams & Coad (1979) recorded for
seasonal desert and intermittent streams, that protracted
breeding seasons may occur, in some cases leading to a rather
complex situation. For example in the Sonoran and Mohave
Deserts, stream minnows show a predominance of reproduction
in spring, first by older females and later by younger ones;
old females may then develop an additional complement of
eggs, in some cases throughout the summer, resulting in a
highly seasonal cycle for younger individuals and a protracted
cycle for older individuals (Deacon & Minckley 1974).
Studies on the reproductive biology of fishes in
tropical intermittent streams are few. The reproduction and
development of two catfishes were studied by Orr & Milward
(1984) in North Queensland, Australia. Both species showed
timing of spawning coincident with the floods and there
were two migrations and spawnings about a month apart for
---------------------"
10
each species during the flood season.
Timing of onset and the duration of reproductive activity
are only two aspects of an animal's reproductive strategy.
Other aspects include reproductive effort, age of onset of
reproduction, fecundity and frequency of reproduction (Moyle
& Cech 1982, p. 127). Reproductive effort represents the
amount of energy or time invested in the production of off-
spring and includes investment in gonads and secondary
sexual characters as well as courtship, parental care and
other reproductive behaviour. Some of the above aspects
of reproductive strategies have been reviewed in some
detail for tropical freshwater species by Lowe-McConnell (1975)
but primarily for floodplain species. Considerable variation
seems to exist in the reproductive strategies of these fishes
even within the same habitat. Beumer (1979) further
exemplifies this in showing that two contrasting strategies
can be equally successful in the same habitat even though the
species differ with respect to the timing of onset and
duration of spawning, fecundity, egg size, mode of dispersal
of eggs, developmental rate and stage of development at
hatching.
11
In order to generalise however, life history theories
predict that in a very variable environment such as expected
in an intermittent stream, species will show characteristics
of r-selection (Pianka 1970). They should have a small
body size, short life span, rapid development, early
reproduction, high intrinsic rate of population increase,
semelparity (one period of reproduction during its lifetime),
low energetic investment per offspring, and no parental care.
Stearns (1976) suggested that in variable seasonal environments,
the predictability of such variation would affect the life
history pattern. It has been suggested that an important
component of tropical seasonality is its low degree of
predictability (Farnworth & Golley 1974). Although a dry season
may occur every year, the time of its onset and its duration
may be highly variable (Jackson 1977) and this variance will
exert a powerful influence on the evolution of population
characteristics (Levins 1968). In unpredictable environments,
Stearns (1976) suggested that the optimal tactic would be to
spread the risk of hatching too early or too late in any given
cycle by generating a distribution of hatching times in the
clutch that matches the historical probability distribution
of the optimal date for reproduction. This concurs with ideas
put forward by Nikolsky (1963) and Lowe-McConnell (1975) that
multiple spawning (including partial and small-brood spawning)
12
is advantageous as a strategy in unpredictable habitats in
order to decrease the chances of one or more generations being
lost due to unfavourable environmental conditions. If
predictability of the conditions during the cycle is also low,
the reproductive tactic should be modified further. For
animals with a generation time less than the period of the
cycle, the amount of expended reproductive effort should be
variable and clutch sizes should be small. If the generation
time is relatively long compared to the period of the cycle
of ~nvironmental fluctuation, then iteroparous reproduction
(more than one period of reproduction during a lifetime),
and not semelparity as earlier predicted, may be expected with,
again, small clutch sizes. Variable reproductive effort may
theoretically be achieved by multiple spawning fishes which
may vary the number of times spawning takes place during
the breeding season depending on the favourability of
environmental conditions (Macer 1974). Multiple spawners
are quite commonly found in seasonal rivers (Lowe-McConnell
1975, Welcomme 1979, Cambray & Bruton 1984). In addition,
multiple spawners have lower fecundities than total spawners
in the same habitat (Lowe-McConnell 1975, Welcomme 1979),
presumably concurring with small ·clutchl sizes as predicted.
In summary, certain predictions may be made on the basis
13
of previous studies as to the nature of seasonality and its
effects on the reproductive strategies of fish in an
intermittent stream:
(a) Tropical intermittent streams should be highly
seasonal, variable and unpredictable habitats.
(b) Seasonal variation of hydrology and biological
productivity should not be of the same nature as that
seen in other tropical lotic habitats.
(c) The reproductive strategies of fishes should reflect
this seasonality and degree of unpredictability but
may be highly variable in expression.
(i) Life history parameters of fish should be
characteristic of those of r-selected species.
(ii) The timing of onset and duration of the
breeding seasons should be cued to the most
favourable period for the species and this
may not be the flood season.
(iii) Multiple spawning should occur in a high
frequency of fishes due to its advantageous
nature in unpredictable habitats.
In order to test these predictions, this study has the
following objectives:
1. To document the physical and chemical characteristics
t
14
of the stream environment.
2. To document the differences in biological productivity
between dry and wet seasons for the benthos and
plankton within the stream, and the influence of
allochthonous material from the terrestrial environment.
3. To determine the diversity of fish species within
the stream.
4. To monitor the population demography and dynamics of
six species of fish, including population fluctuations
and structure, growth rates and life spans.
5. To investigate the reproductive strategies of six
fish species, including size at maturation, timing
of onset and duration of breeding activity, fecundity,
and spawning pattern.
15
THE STREA!\1 ENVIRONMENT
INTRODUCTI ON
Studies of tropical freshwater systems have concentrated
primarily on lakes (Beadle 1974) or on large floodplain river
systems such as those of the Amazon, Parana, Niger and
Senegal Rivers (Welcomme 1979). Floodplain rivers are
especially interesting because of the unique interdependence
of their lotic and lentic components and high fisheries
production; features not readily seen in temperate rivers
in their present forms (Bayley 1980). Lowe-McConnell (1975)
and Welcomme (1979, 1985) thoroughly reviewed the nature of
floodplain river systems with particular reference to their
fisheries ecology.
Small tropical rivers and streams have not received the
same attention as lakes or floodplain rivers despite the
fact that they may make up a significant proportion of the
aquatic habitat of important river systems such as the
Amazon, in addition to being individually unique and
surprisingly diverse in faunal composition (Junk 1983). Data
on these small aquatic systems are scattered and pertain to
isolated, often limited studies. Hynes (1970) reviewed some
16
of the work carried out in tropical rivers and streams and
attempted to integrate it into general ideas on the
functioning of lotic systems. More recent work on tropical
streams includes a very detailed and comprehensive study of
a small Malayan river, Sungai Gombak (Bishop 1973). Other
studies are geographically widespread and include those in
Australia (review in Williams 1981), Africa (reviews in
Davies 1979, Davies & Hart 1981, Awachie 1981, Davies et al
1982), Central and South America (Templeton et al 1969,
reviews in Ziesler & Ardizzone 1979, Sioli 1984).
Hydrobiological studies in the West Indies are lacking
compared to those from other Neotropical regions. Many
studies are related to the taxonomy of aquatic groups (reviews
in Hurlbert et al: 1981, Hurlbert & Villalobos 1982). A few
are of an ecological nature. For example in Puerto Rico
Padgett (1976) investigated the processes of leaf decomposition
in a small stream and Covich (pers. comm.) has continued work
in this area. Hunte (1976) conducted a thorough study on the
biology and ecology of three freshwater prawns in high gradient
streams in Jamaica.
In the Lesser Antilles, published studies on lotic
systems and their fauna have concentrated on certain islands.
17
The Bredin-Archbold-Smithsonian Survey of Dominica included
studies on the physico-chemical nature of the aquatic habitats
over a limited period of time (Hart & Hart 1969). Other
lotic systems have been studied in some detail in St Vincent
and St Lucia. Emphasis in these studies was largely on
comparisons of different types of aquatic habitats and the
'"
zonation and distribution of aquatic fauna (Harrison &
Rankin 1975, 1976 a, b, McKillop & Harrison 1980) and the
population dynamics of molluscs of medical importance
(Harrison & Rankin 1978, Rankin & Harrison 1979, McKillop
et al 1981, McKillop & Harrison 1982).
In Trinidad, several studies of varying detail and
emphasis have been carried out on lotic systems. These
include Thornhill et aZ (1967) on the general ecology of a
small portion of the Maracas River; Hynes (1971) on the
longitudinal zonation of fauna in the Arima River; the
Caroni River Basin Survey (Trinidad & Tobago 1976 a, b)
which included analyses of the hydrology and water chemistry
of the Caroni River and some of its tributaries; and Bacon
et aZ (1979) who documented the physico-chemical characteristics
of streams and rivers of the Nariva Swamp catchment.
Unpublished studies include Carrington (1980) giving a
detailed description of the physico-chemical features of the
18
Maracas River, and Caesar (1985) documenting the effect of
siltation on the benthic fauna of a tributary of the Maracas
River.
From the above review, it is clear that few studies in
the Neotropical region have focused on the general ecology
of an entire stream system and in particular, seasonal
variation is not dealt with in many studies. Also, most of
the studies conducted have concentrated on upland streams
originating from high altitudes and only Bacon et al (1979)
and McKillop & Harrison (1980) dealt with aspects of lowland
streams in east central Trinidad and St Lucia respectively.
Since stream flow characteristics are affected by a variety
of factors such as rainfall seasonality and variability,
catchment size, slope, geology, soil and vegetation (Beaumont
1975, Jackson 1977), lowland stream hydrology can be expected
to vary markedly from that of upland streams and consequently
affect the nature of the biota differently. In addition,
the chemical composition of river water and its seasonal
variation are largely determined by factors such as the
amount and variability of precipitation, the nature of the
bedrock and the evaporation-crystallisation process
(Welcomme 1979). These factors will also vary between upland
and lowland regions as well as between streams on soils of
19
volcanic origin (the Lesser Antilles) and those of alluvial
origin (Trinidad). In particular, some of the above
factors may account for the phenomenon of intermittency seen
in the stream under study.
Intermittent streams are defined as those which flow
during the wet season but dry up during the season of
drought (Ward 1975, p. 243). They are fed mainly by
quickflow (channel precipitation, surface runoff and rapid
interflow) but during the wet season baseflow (groundwater
runoff and delayed interflow) makes some contribution when
the water table rises above the bed of the stream. Such
streams can be distinguished from ephemeral streams which
are temporary lotic habitats fed entirely by precipitation,
and perennial streams which flow continuously because of
adequate year-round groundwater flow. However, many streams
may be ephemeral or intermittent in their upper reaches and
perennial in their lower, making it difficult to allocate
them to a specific category (Ward 1975).
Intermittent streams are considered special lotic
habitats having much in common with temporary lentic habitats
such as temporary ponds and other small water bodies (Hynes
1970, Williams & Hynes 1976). They exhibit many features
20
of ecological interest including the wide variety of life
history strategies which may be utilised for survival by flora
and fauna. Also of importance is the process of biological
succession which occurs seasonally as conditions in the
stream change from being lotic to lentic and eventually
terrestrial (Williams & Hynes 1977). Even when the stream
bed is completely dry, organisms characteristic of the soil
fauna may be present (Moon 1956). Williams & Hynes (1977)
also documented situations in temporary streams where there
may be coexistence of closely related species with similar
ecological requirements. They attributed this to be the
result of either abundant food or a lack of selective
pressures to promote specialisation. Intermittent streams
therefore, are specialised aquatic systems bearing
characteristics of both lotic and 1entic environments and
providing opportunities for studies of life history
strategies, ecological succession and community interactions.
Both intermittent and temporary streams have received
some attention in temperate moist and arid regions (reviews
in Hynes 1970, Deacon & Minckley 1974, also Williams &
Hynes 1976, 1977, Williams 1977, Iversen et aZ 1978).
Several studies have also been conducted in the tropics,
predominantly in Africa. Thornton (1980) made a preimpound-
21
ment study of the nutrient loadings and water chemistry of the
Kwe Kwe River, Zimbabwe. Adebisi (1981 a, b) gave details on
the hydrology and water chemistry as well as the feeding
biology of the fishes of the Upper Ogun River, Nigeria.
Studies on the Bandama River system, Ivory Coast, with
particular reference to the taxonomy, zonation and population
dynamics of the Hydropsychidae have been reported by Statzner
(1982) and Gibon & Statzner (1985). A further study of the
Bandama River by Leveque et al (1983) thoroughly documented
the hydrology and seasonal variation in phytoplankton and
benthic invertebrate production in addition to the general
biology and population dynamics of some fish species as part
of a surveillance programme of aquatic habitats regularly
treated with insecticides for the control of blackflies. In
the Caribbean, McKillop & Harrison (1980) gave brief descrip-
tions and water chemistry data for both high and low
altitude temporary streams in St Lucia. However, Rzoska (1961)
in a review of temporary aquatic habitats in Africa, stressed
that even within the tropics climatic variation is suffi-
ciently large to produce temporary aquatic habitats of
varying permanence. This factor he asserted, is a major
influence on faunal composition of such habitats.
In consequence, the objectives of this part of the
22
study are:
(1) To determine the range of physical and chemical
variation in the study site in relation to influencing
factors such as rainfall and watershed characteristics,
and as compared to other streams studied locally and
regionally.
(2) To document the diversity of the stream fauna and the
range of strategies for survival in the habitat.
(3) To determine the nature of seasonal variation and
succession of stream faunas with particular reference
to those aspects important in influencing the
reproductive seasonality of the fishes.
THE STUDY SITE
Trinidad is the most southerly located of the Eastern
Caribbean Islands lying approximately 20 km off the coast
of northeastern Venezuela (Fig.1). Seasonal variation of
climate generally fluctuates between a dry season (January
to May and a rainy season (June to December) with a short
dry period in September/October, the 'petit careme'.
Rainfall is unevenly distributed over the island with annual
rainfall totals ranging from over 3000 mm in the northeast
23
I I I I I I •
o 20 40 60 km
N
TRINIDAD
i
(0)
Gulf of Poria
o 1 2 3 4 5 km
I , , , I J
~~ roads
k watercourses
swamps
Erin Bay N
i
61"50'
(b)
FIGURE 1: Map showing (a) the location of the study area in
Trinidad and in relation to South America, and (b)
the hydrology of the southwestern peninsula.
24
of the island to about 1500 mm in the northwest and southwest
peninsulas (Berridge 1981). Differences in rainfall between
wet and dry seasons are large with wet season totals
ranging from 2000 to 600 mm in different parts of the island,
and dry season totals from 1000 to 250 mm. Variation of
mean monthly temperatures ranges from 24.5°C in January to
26.6°C in May (Berridge 1981). However, diurnal variation
may amount to 10 to 15°C between night and day. Relative
humidity varies in response to precipitation and temperature,
being lowest in the cool dry season (April mean R.H.: 77%)
and highest in the rainy season (August and November mean
R.H.: 86%) (Berridge 1981).
The southwestern peninsula of Trinidad is an ecologically
diverse area including large tracts of natural forest,
herbaceous and mangrove swamps, as well as agricultural
estates cultivated with coconuts or cocoa,and small
vegetable gardens. For a number of years this region of
Trinidad has been of some biological interest because of its
ecological diversity and proximity to the mainland of South
America from which source several recent colonisations of
the peninsula have been documented (Kenny 1977, Sturm & de
Souza 1984, Alkins & de Souza 1983/84).
25
A number of small watercourses drain the peninsula,
running northward into the Gulf of Paria or southward into
the Columbus Channel (Fig.1). According to Ordnance Survey
maps (1977), the Quarahoon River in conjunction with its
major tributaries, drains an area of approximately 16 km2
along the Chatham Road (South) and enters the sea at two
points in Erin Bay. This river has not been mapped
adequately and local residents maintain that two watercourses
exist: a larger more westerly one being the Carlisle River,
and the smaller eastern stream being the Quarahoon River.
Field checks confirm the presence of two separate water-
courses and in this study they have been named according to
the local residents (Fig.2). However, they are connected
in their lower reaches by a series of artificial canals and
ditches and are therefore not entirely independent of each
other.
The catchment area of the Quarahoon and Carlisle Rivers
has a gently undulating physiography with maximum elevations
not greater than 60 m. Geologically, the area is formed of
Pliocene sedimentary deposits, the Erin Formation, which is
characterised by thick, predominantly regularly bedded
sands with interbeds of grey silty clay (Barr & Saunders
1965). The clay layers may often be mixed with lignite
26
N
r
:-0_'_'-'-"
I .
I
L.._. -' _'1
I
Cedros 5 t?tion 2
Fores t
I
Reser ve
I
'"
........................
<,
Los .........
Blanquizales
Lagoon
Erin Bay
I I I I I
o 0.5 1.0km A Rain gauge
FIGURE 2: Map of the study area showing sampling stations
(key as for Figure 1).
27
deposits. In addition, there are occurrences of naturally
burnt clays called 'porcellanitel, which are brick-red to
orange in colour and often contain well preserved plant leaf
impressions. Porcellanite is quarried in the Chatham area
for use as road metal. The soils of the area include
Moruga fine sandy clays and La Retraite clays with imperfect
drainage on the higher areas, and Cromarty clays characterised
by impeded drainage, in the valleys (Land Capability Survey
Map 1971).
The natural vegetation of the study area is seasonal
evergreen forest, specifically the blackheart (Clathrotropis
brachypetala) - cocorite (Maximiliana eZegans) faciation of
the crappo-guatecare association (Beard 1946). To the west
of the catchment area is the Los Blanquizales Lagoon, a
large coastal herbaceous sedge and mangrove swamp. Beard
(1946) described evergreen seasonal forest as being primarily
influenced by the rainfall regime which always features a
regular seasonal drought. The blackheart-cocorite faciation
is considered by Beard (1946) to be characteristic of the
sandy-clay soils of the southwestern part of Trinidad.
Commonest species in this faciation are guatecare, wild
chataigne, crappo, mahoe, blackheart, malbalata, cocorite,
timite, bois pois and bois tatou. In wet areas especially
28
along stream edges, dominants are crappo, wild chataigne,
mahoe, wild nutmeg, bloodwood and olivier mangue with an
understorey of timite (Beard 1946). Large areas of the
natural forest east of the Carlisle River have been replaced
by cocoa/coffee/banana or coconut plantations. During the
study period, about 80 ha in the northern part of the
catchment area were cleared for dairy pasture. In addition,
some secondary forest and logged areas are present. However,
west of the Carlisle much of the forest remains in a natural
state being part of the Cedros Forest Reserve.
Both the Carlisle and Quarahoon Rivers are generally
between 1 and 8 m wide with depths up to 3 m in large pools
during the rainy season. The slopes of both rivers are very
slight, being about 1 : 300 for the Carlisle and 1 : 200 for
the Quarahoon, thus producing a sinuous meandering course along
most of their length. Where they pass through cultivated
areas, they receive many small drainage channels which contain
water only during the rainy season. The banks of the rivers
rise steeply about 2 to 3 m from the water1s edge. Flow is
intermittent in most years, ceasing completely during the
height of the dry season when shallow stretches dry up, leaving
isolated pools or chains of pools. Most of the upper and
middle regions of the Quarahoon River dry up during severe dry
29
seasons while only the upper half of the Carlisle does so.
Four sampling stations were chosen for this study: two
on the Carlisle and two on the Quarahoon (Fig.2 ) and were
representative of the range of conditions found. Only
Stations 1 and 2 on the Carlisle River were sampled for the
second year of the study period.
Station 1 was a shallow pool on the Carlisle River
with maximum dimensions of about 10 m in length, 4.5 m width
and 0.65 m depth during the rainy season. It dried out
completely during severe dry seasons. Vegetation cover
included the swamp immortelle (Erythrina glauca), fineleaf
(Pentaclethra macroZoba) and Bactris among other species
characteristic of seasonal evergreen forest. Exposure was
minimal because of fairly dense overhanging vegetation
(Plates 1 and 2).
Station 2 was a deeper larger pool on the Carlisle about
25 m long in the dry season, but with maximal rainy season
width and depth of 5.2 m and 2.4 m respectively. It did
not dry out even during the most severe dry periods during
the study and seemed to be the most northern refuge pool in
the Carlisle River. This portion of the river meandered
- <
. . . . . .
m
~
. . . . . .
o t :
<
. . . . .
r o
: : e :
0 -
o
: : e :
~
V I
M -
- s
r o
Q J "
3
- I )
- S
o
3
( / )
M -
Q J
M -
. . . . .
o
~
. . . . .
- s
V I
M -
- S
Q J
. . . . .
~
V I
u :
32
through a cocoa estate and the banks were vegetated with
cocoa (Theobroma cacao), immortelle (Erythrina erythrina)
and fineleaf. Most of this stretch of stream was shaded by
overhanging vegetation (Plate 3).
Station 3 was a fairly deep pool on the Quarahoon
River immediately downstream of a road culvert. Rainy
season dimensions were about 15 m in length, 7.3 m width
and 1.1 m depth. It was reported to have been deepened for
use as a cattle watering hole. During one very severe dry
season in 1985, this pool dried out completely. Its banks
were grassy and no large trees grew nearby thus allowing
complete exposure (Plate 4).
Station 4 was brackish, located about 300 m from the
sea on the Quarahoon River. At this stretch of the river,
maximal width was 7.3 m and depths 1.5 m. During most dry
seasons flow ceased completely and tidal influence was
minimised as the river mouth became blocked by a sand bank.
Vegetation cover included coconut trees, Bactris~ swamp fern
(Acrostichum), red mangrove (Rhizophora) and smaller shrubs.
Banks were high and steep and because of the width at this
point, the stream was relatively exposed (Plate 5).
V I
r - t -
P >
r l "
- - ' .
o
: : : s
N
- - ' .
: : : s
r l "
: : : s -
( D
P >
r l "
( D
a .
- S
' <
V l
( D
P >
V l
o
z s
( D
- + >
r - t -
<
. . . . . .
. . . . . .
o
~
w
. . . . . .
~
r t
: : : : r
r o
U l
r o
Q I
U l
o
~
c :
" 0
( / l
r t
- s
r o
O J
3
<
. . . . . .
: : s
r t
: : 5 ' "
r o
O J
r t
r o
0 -
~
( / l
r o
O J
( / l
o
: : s
S f
36
METHODS
Hydrology and physico-chemical characteristics
Rainfall:
Daily rainfall measurements were obtained from a guage
located within the catchment area approximately 1.5 km from
Stations 1 and 2, and 3 to 4 km from Stations 3 and 4.
Records were obtained for the period March 1978 to July 1982
from this source. Longterm rainfall records were taken from
published data (Wehekind 1955, Trinidad & Tobago 1979) based
on measurements from the above and a more northerly located
guage (Fig.2). Pentade analysis of daily rainfall data was
carried out according to Griffiths (1959, quoted in Jackson
1977, p. 60) (Appendix 1).
Physico-chemical characteristics:
During the period from March 1980 to June 1981, certain
physical parameters were measured at all four stations on a
monthly basis. Each station was visited at approximately
the same time of day between 0900 and 1300 hours. Radiation
intensity was measured in an unshaded area at each station
37
using a LI-COR Integrating Radiometer with pyranometer
sensor. Readings were taken for three consecutive 100-
second periods and a mean value calculated for the overall
five minute interval. Data for the complete study period were
not obtained due to malfunctioning of equipment. Air
temperature, surface and bottom water temperatures were
measured to the nearest 0.5°C. Depth of the water column and
width of the stream channel were measured at specific points
for each station. Surface current was measured in mid-
channel using a flat plastic disc timed over a two to four
metre distance (depending on relative current speed) and
replicated five times to obtain mean current speed. Stations
1 and 2 were mapped in detail for cross-sectional shape and
this was used to calculate cross-sectional area (A) as a
function of width (W) and depth (D). For Station 1, A was
found to be 67% (W x D), while for Station 2, A was 75%
(W x D). It was assumed that because of similarities in
stream bed shape, cross-sectional areas for Stations 3 and
4 bore the same relationship to width and depth as for
Station 2. Discharge rates were calculated every month for
which data were available using the following formula:
Q = A.v
where Q = discharge (m~s-l), A = cross-sectional area (m2)
and v = current speed (m.s-l (Wetzel & Likens 1979).
38
Three surface water samples were taken at each station
for determinations of (i) dissolved oxygen, (ii) biochemical
oxygen demand (BOD), and (iii) general physical and
chemical analysis. Dissolved oxygen samples were chemically
fixed immediately in the field for a standard Winkler
determination. BOD was determined from a 7-day incubation at
20°C. All water samples were submitted to the Central
Laboratory of the Water and Sewerage Authority for analysis
the same day or the following day after overnight refrigeration.
General physical and chemical determinations included
turbidity, colour, pH, specific conductance, alkalinity,
hardness, nutrient and ionic composition, total dissolved
solids, total solids, permanganate value and total organic
carbon. Methods used were according to standard methods
of the American Public Health Association except in the case
of iron where an ammonium thiocyanate method was used
(Lue Chin, pers. comm.).
Substrate samples were taken monthly from each station
from April 1980 to June 1981 using a 5 cm diameter plastic
corer to a depth of 10 cm. These were submitted to the
Central Analytical Laboratory, Soils Department, U.W.I. for
analysis of particle size composition and carbon content.
39
From July 1981 to July 1982 only Stations 1 and 2 were
sampled for the above parameters with the exception of
substrate samples.
Stations 2 and 4 were sampled over 24-hour periods to
determine diurnal variation in physical and chemical
parameters. In the dry season, samples were taken on
81/03/25-26 and 82/03/31-04/01 (Station 2) and 81/04/01-02
(Station 4), while rainy season samples were taken on
80/08/07-08 (Station 2) and 80/08/19-20 (Station 4). On
each occasion the following were measured on an hourly
basis: irradiance, air temperature, surface and bottom
water temperatures, depth of water column and current
speed.· Surface water samples were taken for determinations
of dissolved oxygen, pH, specific conductance and total
alkalinity using the same methods as those used for monthly
samples.
On 82/03/30-31 Stations 1 and 2 were mapped in detail
and dissolved oxygen profiles were taken using a YSI Oxygen
Meter with probe calibrated in saturated air.
40
Stream biota
Plankton:
Non-quantitative plankton samples were taken at each
station on a monthly basis from July 1980 to June 1981 using
a hand-held plankton net with a mesh size of 110 ~m. Samples
were returned to the lab and preserved by addition of an
equal volume of buffered 4% formalin. Individual elements
of the plankton were subsequently identified as far as
possible and a taxonomic list drawn up.
Quantitative samples were collected from Stations 1 and
2 in order to contrast dry and rainy season plankton
communities on 83/05/01 and 83/08/09 respectively. On each
occasion a one litre sample of water was taken from the upper
10 cm of the water column. In the laboratory the sample
was agitated to mix the contents thoroughly and four 5 ml
subsamples were removed and each mixed with an equal volume
of 2% buffered formalin. Preliminary tests showed that no
rupturing of algal or other cells occurred on addition of the
formalin. From each of the preserved subsamples in turn, a
1 ml aliquot was removed after thorough mixing and introduced
into a Sedgewick-Rafter Counting Cell. Under 100x
41
magnification the number of organisms .ml-1 or colonies
1
.ml- for each taxon was determined according to the methods
of Wetzel & Likens (1979). Numbers of colonial rotifers
were estimated separately from other plankton because of their
larger size and numbers of individuals .1-1 were estimated
initially. From these data the mean number of organisms
-1
.ml for each taxonomic group and the total number of
organisms .ml-1 were calculated for each station for each
season. These were compared using Student's t-tests
(Appendix 2).
Benthic macroinvertebrates:
Benthic invertebrate samples were taken monthly from
all stations from August 1980 to July 1981. At each station,
one sample of bottom detritus and sediment covering an area
625 cm2 and to a depth of 5 cm was collected by hand. In
September 1981 an Ekman grab became available and was used to
collect a sample (225 cm2 to a depth of 10 cm) from each of
Stations 1 and 2 every month until July 1982. Preliminary
sampling showed that benthic macroinvertebrates were
confined to the uppermost layers of the substate. On return
to the lab, samples were sieved through a series of sieves
with mesh sizes 2.0, 1.0 and 0.5 mm. Samples were sorted
42
visually in white enamel dishes. Live invertebrates were
identified as far as possible, counted and preserved in
ethanol. Time constraints prevented the replication of
monthly samples for each station and these data were
intended to give only a qualitative view of the composition
of benthic invertebrate communities.
More quantitative sampling was conducted to contrast
dry and rainy season benthic invertebrate populations. On
83/03/17 and 83/08/09, four Ekman grab samples were taken
at each of Stations 1 and 2 and they were sorted in the same
way as monthly samples. In addition, blotted wet weights of
all live macroinvertebrates were obtained. Mean densities
and biomass of benthic macroinvertebrates were calculated
for each of these two stations for each season and were
compared using the Student's t-test.
Larger benthic invertebrates were collected with a two-
man push seine, mesh size 6 mm. In particular, Dilocarcinu8
dent.atue and Pomace a ql.auca were collected monthly by seining
and were preserved in buffered 4% formalin. Carapace widths
and widths across the body whorl were measured for
D. dentatu8 and P. glauca respectively for population
structure analysis. P. glauca egg masses were also located
43
within a constant area at each station and were counted each
month.
Allochthonous input:
In order to estimate the variability in amount of
allochthonous animal input falling onto the stream ~urface
as a source of food for fishes, four 'glue boards' (Geisler
et aZ 1975) were set out at each of Stations 1 and 2 for
24-hour periods on 83/04/30-05/01 and 83/08/08-09. Each
tray measured 0.25 m2, was covered with 'Tangle Trap'
adhesive and was supported no more than 10 cm above the
water surface to prevent inundation. On return to the lab,
all animal material was removed and excess adhesive dissolved
in benzene. Plant material and frass were not taken into
account. All animal material was sorted to taxonomic group,
counted and weighed after blotting and air drying for 1 hour.
Occasional animal fragments were included with their
respective taxa for weighing; only whole individuals were
counted. Mean values of number and biomass of animals
caught on the glue boards for each station in each season
were calculated and compared using Student's t-tests.
44
RESULTS
Hydrology and physico-chemical characteristics
Rainfall:
Long term rainfall records for the Chatham area are
shown in Table 1 while Table 2 summarises extremes of rain-
fall for the area. Long term monthly averages reflected a
dry season from January to May with February being the driest
month, and a wet season from June to December with July and
August being the wettest months. The petit careme, a short
dry period in the middle of the rainy season occurred between
September and November. Although long term annual totals
averaged 1735 mm, variability was high, ranging from 1489 to
2136 mm annual totals in two consecutive years (1974 and 1975
respectively). Monthly averages for the period during the
study, indicated the occurrence of somewhat wetter than
average dry seasons (110.1% 1953-77 average) with earlier and
wetter than average rainy seasons at this time (110.6% 1953 -
77 average). Overall, the average annual total rainfall for
1980 to 1982 was slightly higher than those values derived
from long term records (108.4% 1953-77 average).
; p o
e . . . .
3 :
e . . . . - n
; p o e . . . . 3 :
V I 3 :
0
: z : P
C J
t v > - ' C J : E
. .
( 1 )
0
0 . . >
O J
0 . . > u
( 1 ) e
r o e e
0 n
r o
- s
<
: : : : l
: : : : l
- s c r
e . . c - s
- 0 : : : : l
M -
r o n
H -
< ' <
" <
. . . . . . .
( 1 ) e M -
( 1 ) ( 1 ) n
e " " " i
0 M -
" " " i
' <
~
: : : r
( 1 ) : : : : : r e
0 . . >
V 1
V l 0 . . > c r
U ' l
~ N
e . . . . e . . . . 3 3
( 1 )
0 . . > - s
( 1 )
M -
< . 0
r o c r 3
~ c r
0
e 0 . . >
( 1 )
- s
c r
O J r o r o - s
0 . . > ' <
: : : : l
: : : : l
( 1 )
( 1 ) U ' l V l
" " " i " " " i
e " <
" <
" <
0 . . > - s
0 0
r o r o
0 . . >
: : : : l : : : : l
0 . . > 0 . . > : : : : l
M - - s
: : : : l
- s
0 - s
" <
0 . . > O J e
. . . . . . - 0 0 . . >
- 0
C J <
<
. . . . . . .
( 1 )
( 1 )
r o r o
~ r o
( X )
n - s " " " i - s - s
. . . . . . .
( 1 )
0 . . > O J H -
0
0 < . 0 < . 0
0 0
3
0 - 0 - r o r o H -
c r
( 1 )
U 1 . .
0 . . >
~
c . . . .
- s
- s
- s
( 1 ) ( 1 )
e
~
n
n
0
0
" <
" " " i " " " i
~ 0 - 0 -
( 1 )
< . 0 r o
( X )
0 - 0 -
N - l
0 . . >
0 . . > : P
c o
M - H -
" " " i
( 1 )
r
. . . . . .
~
r r 1
n V I
. . . . . .
. . . . . .
< . 0
0 0 . . >
" " " i
. p .
~ ~
~ ~ ~ ~ ~
0 N - . . . . J ~ N W
: : : : l N
" " " i
. . . . . .
. p .
Q )
( X ) ( X ) ( J )
' - J W W W
0 - M - ' - J 0 N < . 0 0 - . . . . J W
( 1 ) . p .
~
U 1 U ' l ~ ' - J N I
0 . . > U 1 ' - J N ' - J < . 0 ~ ' - J ' - J
W W
~
0 -
. p . . p .
G ) ( J )
. . . . . . ( J )
~ ( X )
' - J
W < . 0 0 W C J 1 < . 0
~ ' - J
3 :
. p .
( X ) ( J ) ( J )
. . . . . .
e . . . . ~
0 . . > 0 . . > 0 w N N N C J 1
C J 1 ' - J C J 1 0
0
; : : 0
0 . . > N
H - " " " i
<
( 1 )
> - ' 0 . . >
: : : : l
e V I 0 . . >
" " " i
0 . . >
0 . . > : : : : l
: : : : l
r r 1
: : : : l
" " " i 3 : P
- n
( 1 )
V 1
H - 0 . . >
<
" <
( 1 )
: : : : l - - - '
0 . . > M -
0 . . > H - - - - '
M -
" " " i
0 H -
3 : 0 . . >
( 1 )
0 . . > V I
0 . . >
n
" " " i
3 :
<
( 1 )
: : : r
0 . . >
0
n
0 . . >
" " " i
" <
: : : r
0
0 . . >
~
l Q
r r 1 0 . . > ~
. . . . . . . . . . . .
V l ( 1 )
M -
< . 0
. p .
~
: : : r N ' - J ~ ~ ~ ( J l
N . . . . . . ~
M - N
N V 1
. p . . p .
( J ) ( J )
0 . . > 0 . . > n C J 1 ' - J C J 1
e x > W C J 1
0 - . . . . J W
0 0 W
. p .
( J )
: : : r
N ( X ) ( X )
H - w U 1 W N
0 ' - J ' - J I
3 < . 0 - . . . . J 0
- n
( 1 )
0 . . >
~
0
~ ( J )
~
H - ~ N ~ ( X ) . . . . . .
e x > < . 0
0 C J 1 ~ 0 W
< . 0
" " " i " " " i
. p .
- j Q ) . p .
: : : r N ( X )
N N W ~
e x >
< . 0 ' - J
0 e x > 0 ' - J 0 . . >
. . . . . .
n O J
" " " i
' - J n
. . . . . .
: : : r
t v
3 : : : r
: : : : l
0 . . > : : : : l
0 . . >
- n
. . . . . .
H -
0 . . > M -
~
: : : r 0 -
: E :
: : : r
( 1 )
~
0 . . >
0 . . >
0 . . >
: : : r
0 -
3
3
( 1 )
Q O
7 '
- n
3
n
3 " " " i
- j
: : : : l
0
0
e 0 0 -
3
" " " i c r
. . . . . .
0 . . >
H -
~
l Q
r o < . 0
~
< . 0
. . . . . .
. . . . . .
U ' l
0 U 1
W
< . 0
~
U 1 C J 1 ~ ~
~ e x > N N ~
N ~ ~ ~ ( X )
W
" <
( J ) . p .
. . . . . .
~ ( X )
( X )
0 W ( X ) . . . . . .
e x > N
C J 1 0 W
< . 0
0
. p .
~ . p .
< . 0 0 W N ~ ( J )
0 N
3 : N ~
< . 0 - . . . . J I
0
' - J
" " S
~
. p .
. p .
. p .
< . 0 0 0 N W ( X )
< . 0 C J 1 W
W
< . 0 0
W ~
< . 0
. p .
( J )
. p .
U 1 ' - J
l J : J 0 < . 0 C J 1
< . 0
' - J W ( X )
< . 0 N W N
< . 0
( X )
N
w
N
l J : J
0 . . >
" " " i
n
~
0 . . >
" <
~ . . . . . . , a
< . 0
( X )
0
. . . . . . . . . . . .
. . . . . .
~ . . . . . .
~
N ~
- n 0
~ . . . . . . ( X )
0 < . 0 N
C J 1 . p . ( J )
< . 0
W N
C O
< . 0 ' - J I
( X )
N
0 0 . p .
0 W < . 0
O J ' - J
< . 0
' - J N ~ ( X )
N
e x >
1 . 0 N
( J )
. . . . . .
~ ( X )
~ . p .
W < . 0 ( J )
O J N ( X )
O J C J 1
0
C J 1
W . O . . >
I V 1
' - J
' - J 0 . . >
S t l
46
TABLE 2: Rainfall extremes for Chatham from 1953-1982.
1953-1977 1978-1982
Date Rainfa 11 Date Rai nfa11
(mm) (mm)
Wettest year 1975 2136.39 1979 2040.64
Driest year 1974 1489.64 1981 1811.53
Wettest month 77/06 496.57 80/10 447.55
Driest month 73/03 7.11 78/04 3.56
Mean wettest month August 246.13 June 304.29
Mean driest month February 68.07 March 40.39
Greatest in 24 hours 77/08/20 107.70 80/10/27 251. 71
Sources of data as for Table 1.
47
More detailed illustration of rainfall just prior to
and during the study period shows the range of variability
in the distribution and intensity of rainfall (Figs. 3
and 4). Monthly and weekly totals for the period indicated
that most of the annual rainfall occurred during the months
of May to December with a drier period from January to
April. However, the time of onset and duration of each
season was extremely variable. An extreme situation was
recorded in 1981 to 1982 where a particularly marked petit
careme occurred in late 1981 and no appreciable dry season
can be recognised in 1982 from these data.
Pentade analysis of rainfall data from 1978 to 1982
also illustrates the unpredictability of timing of onset and
duration of each season ( Fig.4). However, according to
the criteria defining dry and rainy pentades (Appendix 1),
a dry period did occur in early 1982. The beginning of each
rainy season varied by up to three weeks from early to late
May and was frequently followed by dry periods. Precipitation
was also not consistent in its distribution over time with
very high weekly totals alternating with drier weeks (Fig.4).
Thus although wet months could be fairly clearly defined, on
a more detailed analysis most of the rain fa11 ing duri ng
those months fell during one week or even on one or two days.
R a i n f a l l ( m m )
. t - .
w
I V
. . . .
o
w
o
o
o
o
o
o
o o
I
I "
. . .
T
c . . . .
1
1
"
' <
~ H
~
1
c . . . .
I
I
r i -
» 1
o
r i -
l J )
1
O J
1
o
z
- + >
I
- s
o
o
l
3
1
c . . . . .
O J
: : : s
I
c
O J
I
- - s
' <
1
I - '
\ D
1
c o
o
1
r i -
o
1
c . . . . .
»
c
I ,
l f l
' <
I - '
o
I
\ D
c o
Z
I
N
o
I
I
1
1
1
I
l
I ~
I
8 1 7
- n
. . . . . . . .
G )
. . . . . -
c
0
~ . . . . . . . . . . . .
; ; 0
x r
\ . D \ . D \ . D
r n
- . . . . : I ( » ( »
-
e : > O N
. j : : >
R a i n f a l l ( m m )
I V
w
~
•
0
0
0
O J
I -
0
0
O J
0
: : : l 0
O J
: ; ; : :
•
' <
l / l
r o
: : 0
- 0 .
I l
r o
. . .
0
l / l 7 '
= '
0
' <
- e
- t )
- s
- 0
3 :
O J
- s
t t l
- 0 .
O J
- 0 . = '
: : : l
~
- t )
: : : l
0
- -
\ . D I -
- t )
O J
a .
e : >
O J
: t >
t t l
0 1 -
: 1
M -
- t )
0
M -
- s
O J
0
' 3
o
l / l
3 :
- t ,
O J
0
- ,
- s
- s
n 0
- <
: Y
' 3
- 0
. . . . . ,
c . . . .
t t l
O J
< . 0
= '
: : : l
- . . . . J
0
c
0 0 - -
a .
O J
t t l
- s
M -
0
' <
. . . . . .
c . . . .
~
c
0 0
0
' <
. . . . . ,
M -
< . 0 0
0 0
c . . . .
N
C
. . . . . . .
' <
. . . . . .
~
l -
0 0
N
I I r > - - ~
3 :
O J
: : : l
U ; : t >
0 . .
~ 3 : ' ' ' ' ' ' ~
c : r
I ~
6 1 7
50
Physical parameters:
Table 3 summarises the ranges and means of physical
characteristics of each station during the study period.
Mean irradiance was highest for the most exposed site
(Station 3) while Stations 1 and 4 show intermediate levels
of intensity and Station 2 the lowest. Although each
radiation reading was averaged over a five minute period,
large fluctuations occurred both over the five minute
interval and from month to month due to variability in
cloud cover. Clear seasonal differences in radiation
intensity were not observed since conditions during each
season were not consistent. It was as common to find sunny
or mixed conditions during the rainy season as it was to
record overcast days during the dry.
Average air temperatures and hence water temperatures
also reflected the degree of exposure at each site. Generally,
Stations 3 and 4 showed the highest air and water temperature~,
possibly due not only to exposure but to generally slower
current speeds as compared to Stations 1 and 2. The annual
range of temperatures measured was large: Station 1 showed
a range of 8.5°C between maximum and minimum air temperatures
- i
n o n ~ C J
r o
- s
"
o c r t >
- s
- s
3
' "
V l - s 0 . - 0
- 0
: E :
o n " " " S c - t - c +
Q .
r o 3
' "
' "
c : : : : : r r o : : : : r : : : r
r t >
- s
. . . . .
' "
- s ' " : : >
r t
r t >
- s . . . . .
r t >
e - e : : >
' "
- s ' "
. . . . . u : > 3 3
C n - s
: c r t > ~
' <
r t >
- s
c r
' " n r o
o
N 3
. . . . .
- r t r t > 3
: E :
. . . . .
: : > w V I
o
o
o
I
c - s
n
3
C V I . . . . . . 3
: : > 3
I
: : > I
~ . ' "
N
. . . . . .
. . .N . .
V I ~ .
. . . . .
~ : : >
V I
~
. . . . . . .
3
N
. . . . . . N
< J 1
~ .
N
o
o
o o
o o
< J 1 o
a
o
V I
r t
. . . . . . .
< . n r t
3
W ' "
0 ' \
N w
N
< J 1
o
x
e x >
. . . . w ' "
< J 1 0 0 ' \
o o
: : >
a
a
< J 1
o 0 ' \
< J 1
< J 1
W
w
3
N o r t >
N
. ; : . w
e x >
w 0
o
: : >
' "
w
. ; : . w
e x >
o
I . D
e x >
: : >
' "
Q .
. . . . . .
N
N N
< J 1
. ; : .
- s
. . . . N 0 o o
o o o
: : >
o N o o o
' "
u : >
o
r t >
V I
V I
r t o
. . . .
N U " 1
0 ' \
< J 1 N N W o r t
o 3
' "
- 0
o N e x >
o o
e x > e x >
: : r
x e
' "
. . . . . . W : : >
N a
o o
' <
V I
< J 1
N
n
' "
N
N
3
0 ' \ W N o r t > - 0
. . . . .
o I . D
o w 0 < J 1
: : > - s
' " ' "
c , . , . . . .
o . . . .
< J 1
< J 1 w
3
' "
r t >
r t
r t >
- s
V I
. . . .
o
. . . .
N N N
3 - s
~ .
. ; : .
o o w 0 U " 1
o < J 1
V I
. . . . . . N
o o a r t
o
. . . . .
' "
~ .
V I
o
r t
: : >
. . . . .
. . . . . . .
V I
. ; : .
< J 1 N
3 r t
' "
. . . . . .
~ .
o
N W W . . . . . . . . . .
o
o o W I . D . . . . . . . I . D N W
0 ' \ x e
' "
: : >
r t
. . . . . . .
N . . . . . . .
< J 1 o U " 1
o
o
o o
w
. ; : .
. . . . . .
0 ' \
3
w
< J 1
N N N
r t >
. . . . . .
W
o o < J 1 0 0 ' \
e x > I . D
: : >
' "
o . . . . . . .
0 0 < J 1 0 ' \ < J 1 I . D
N . . . .
. . . . .
N N
N 0 ' \
3
N
~ .
o
o w 0
0 ' \ 0 ' \ N
0 ' \
: : >
0 ' \ . ; : .
o
a < J 1 I . D
o
V I
r t
. . . . .
< J 1
C l " ' >
N
o
w
w
w W
r t
3
' "
o
o . ; : .
o
c o o
o W
x
o
' "
w
o < J 1
o o
o
: : >
. . . . . .
o
N
3
< J 1 N
N
N w
r t >
N
. . . .
o
e x > . . . .
o
: : >
' "
o
o
0 0 w
I S
I . D
52
recorded; Station 3 a variation of surface water temperatures
of 11.5°C. Bottom water temperatures fluctuated much less
than surface temperatures, the greatest range being 8.0°C
at Station 2. Unlike other stations, there was a large
difference between mean surface and bottom temperatures at
Station 3, indicating a possible tendency towards stratifica-
tion of the water column as a result of rapid heating of
upper layers, slow current speeds and greater depths (Table
3) .
Monthly variation of temperatures is shown for each
station in Fig. 5. Generally, water temperatures tracked
air temperatures closely. Air temperatures, and to a
lesser extent, water temperatures tended to show a slight
increase during the middle to late rainy season consistent
with higher mean monthly air temperatures at this time for
the island as a whole (Berridge 1981). During the dry season
months, temperatures were on occasion lower than wet season
temperatures and showed large fluctuations from one month to
the next. Despite this general pattern of seasonal variation
in temperatures,periods of heavy rainfall were often associa-
ted with low air temperatures for example June 1980 and June
1982. Occasionally bottom water temperatures were higher
53
Stn 1
30
20
Stn 2
30
20
u
o
M A M J J A SON 0 J F M A M J J A SON 0 J F M A M J J
C1I ---1980 1981 1982-
L.
.:..J.
o
L.
C1I Stn 3
a.
E
C1I 30
t-
Stn 4
30
0--0 Air
x- -xWater surface
n- ····0Water bottom
2 0 ..L.,-.....--..-.:......----r--r-..--r--r--1r--T--r--,.-----r-~
M A M J J A SON 0 J F M A M J
---1980 -1981-
FIGURE 5: Monthly variation of air and water temperatures at
Stations 1 to 4.
54
than surface water temperatures by up to 2°C. At Stations 1,
2 and 3 this may have been due to normal diurnal variation
(ct Results section 'Diurnal variation'), while at Station 4
some warm salt water intrusion may have been taking place.
Each station possessed distinct characteristics with
respect to other physical parameters (Table 3). Station
1 was shallower than all others and therefore attained
highest average current velocities. Both depth and current
velocity and hence discharge, increased significantly with
spates at all stations. In June 1982, Station 2 was observed
in flood when the maximum depth of 2.40 m was measured.
Unfortunately it was not possible to measure current velocity
at this time. In July 1980, current velocity at Station 3
during a flood was roughly estimated at about 25 cm.s-I,
the highest value attained for any of the deeper stations.
Station 3 showed both the highest colour values and degree
of turbidity. All other stations were roughly equivalent with
respect to these parameters.
Monthly variation of physical parameters is illustrated
in Figs. 6 to 9. Generally, at Stations 1, 2 and 3 high
levels of precipitation during the wet season resulted in
increased habitat size, current velocities and discharge.
55
Stn 6
/
1.0 " 4\ , x
x
2
"U
Q) 0
"U
0
.9 0-
8-5
~ Stn 2 ..J.:.:.
2.0 I \ J\ 6~
I \
I \ f / \ 3
,. / \ -"f\ 4I x-x- x_x- x-x \ "...x. x x-x / "/.. x- "- x x-x_x_x~
I x ..... -, / ~x
1.0 x-x x
o
MAM J J AS 0 N D J FMAM j j AS 0 N D j FMAM j j
..J.:.:. 1980 1981 --1982 --
0-
C1T
o 10
Stn 3
2.0 8
~-x
-,
/ x x-x 6
x...... / '\ / "-
x-x x-x / -x,
1.0 I.-E
2:=
"U
03=
8
/\ Stn 4
2.0
r / \\ x 6x/ \
/ x .....
x \ /
\/ 4
x 0--<:> Depth
1.0 x- -x Width
2
o
o MAMJJASONDJFMAMJ
---1980 --1981--
FIGURE 6: Monthly variation of depth and width at Stations
1 to 4.
56
Stn 0.6
0.4
0.2_
30 ....
I
X VI
20 x 0 M
0..
::::l "'0 E
Q)
10 "'0 "'0 Q)
Q) 0 0 C
.... 0 ...'lo
O ..........-..-o-~"'0.,..--,r-r-,..-.--~-:---.-4=~.,....:;~---.-....--.~-c?~~L.-~~~ - s:u
VI
x x x 0
/\ /\ 0.2
/ \ / \
Stn 2
/ \ / \
/ \ ~ \
x,,,,1t. k /
x-x-x x <, x-I o10
-0 "'0
Q)
Q)
-0
"'0
o o
o o
~ 0...L-.--o--<>--<>-.--,_.--r--r----.-~)___(~._~.,....~__r__r__r--.::::e:>__a::::.:::;:::::::o__o___o.::::::;:::::::.o--::;::_ ..o...
MAMJJASONDJ FMAMJJASONDJFMAMJJ
E ---1980 1981 --1982-
-u 002.... _
c Stn 3
Q)
I.. x x ~
I..
::::l / "<, / 'xV!
U ."'g0 / x....... / <, t5 x-x-x x-x x-x-X x ° C"'lE
o
Q)
C\
I..
o
s:
u
~...x.... V!
o Stn 41 "-
I x,
0.2
I '.,
I '.
I '
/X-"x-x I '\
5J X-------~ ° 0--0 Current
x- ~ Discharge
o MA~JJASON DJFMAM J
___ 1980 - 1981-
Monthly variation of current and discharge at
FIGURE 7:
Stations 1 to 4.
57
Stn 600
400-;
c
200 ::::J
C
N
a c
E
"-
600 CCStn 2
>-
400 OO·~"0
..0
"-
200 200 ::::JI-
..I..I.I. a ....L...,---,.--,---.-_,__..,.-T-T-r---T-T----r--r-..:;-~:r--,----.-~--;..-..:.:-~:c..,.-4--,--....-~~ 0
C MAMJ J ASONDJ FMAMJ J ASONDJ FMAMJ J
::::J ---1980 1981 - 1982--
C
<Ll
N
e
I 600
Stn 3
"-
g1QOO 400 -;......
o c
u 200 ::::J
C
N
a 0
E
"-
0
600 !::.
>- Stn
400 .~
"0
.£l
"-
200 ::::J
I-
,x_x- - -- - --x. 0--0 Colour
'.
a x-~ Turbidity6jFMAMJ
-1981-
Monthly variation of turbidity and colour at
FIGURE 8:
Stations 1 to 4.
58
10~J ~~t~~~~~~~~~
1::JI;II~~~~~~~
1::Ji~~~I~~~~~
Coarse sand
!1~Ji~I~~~~.~~Fine~sandSilt ~~~
~ Clay
{; AMJ JASONDJFMAMJ
~ 198-0- -1981-
(1)
>
o (e)
~ 4 ~
,JI;"' A I'I I I [J
3 It: If' .: •• A
I I ., '. I' "Q>~~ "A- ''.A\ I • \ -/ \ I
IAI ' ? ,.i.".." , ,'l=:f '.~~.... ,".0." ,.. '" Stn 12 [J ...,......--v• - \ r » ... I·
..0"-0: '. '1/.: (.~: [J.· ..cStn 2
.' .' • ""t' 'Q, , .
[J "'0' ... .[J ~':'d: • _. Stn 3
A- - -.A St n "
FIGURE 9: Monthly variation of (a)-(d) particle size composition,
and (e) carbon content of substrates at Stations 1 to 4.
59
However, such increase was not sustained for long periods
and values of these parameters quickly decreased creating
large fluctuations over an annual period. During the drier
periods of the year (the petit car&me and the main dry
season) water levels decreased along with current velocity,
often creating lentic conditions. The extent to which the
habitat decreased in size at each station depended on the
severity of the dry period. For example, by the end of
the 1980 dry season, depths of Stations 2 and 3 were the
lowest recorded and Station 1 had dried out completely.
Lentic conditions ensued for about four months during this
year. During other dry seasons decreases in depth and width
were small except during the uncharacteristically dry petit
careme of 1981.
Station 4 showed a somewhat different picture from the
other stations. Generally the size of the river at this
point increased during the dry season and decreased durlng
the wet. Increased width and depth appeared to increase
discharge but in fact, throughflow to the sea did not take
place. This may be attributed to the development of a sand
bar across the mouth of the river during periods of low
flow when discharge was insufficient to prevent the accumu-
lation of sand moved by longshore drift into the river mouth.
60
This sand bar presumably prevented throughflow of water into
the sea during the early rainy season when accumulation of
water upstream and from runoff increased. Only after
sufficient flow developed or after several spates occurred
to break through the sand bar, did water levels decrease at
this station.
Largely because of infrequent analyses of turbidity and
colour, it is not possible to make definitive comments on
the seasonal variation of these parameters. However, there
appeared to be an increase in both properties with the
rainy season. Although some spate events were associated
with increased turbidity, this was not a consistent
correlation and other factors such as increased plankton
densities in the late dry season may have been contributory
(c~ Results section 'Plankton').
Analysis of particle size composition of sediments
allowed characterisation of sediments as follows: Station
1 - sandy clay loam, Station 2 - sandy clay, Station 3 and
4 - clay (Table 4). Mean carbon content was highest at
Station 4 and next highest at Station 2 whereas Stations 1
and 3 showed substantially lower carbon contents of
sediments. Such a pattern could be correlated with high
61
TABLE 4: Average particle size composition and carbon
content of substrates at Stations 1 to 4.
Stations
1 2 3 4
% oven dried sediment
Coarse sand (>0.2 mm) 36.2 18.9 3.7 5.9
Fine sand (0.02-0.2 mm) 37.6 32.5 38.1 38.5
Silt (0.002-0.02 mm) 2.9 9.2 10.7 13.3
Clay «0.002 mm) 23.3 39.4 47.5 ,42.3
Carbon content 0.7 1.9 0.4 2.6
62
inputs and rates of accumulation of organic material in
Stations 2 and 4 from overhanging vegetation or upstream
respectively. Lower carbon levels occurred in shallow high
velocity areas (Station 1) or areas lacking in input of
allochthonous material from overhanging vegetation (Station
3). Substrate analyses showed little seasonal variation in
particle size composition at all stations (Fig.9). However
carbon content showed some variability particularly in
Stations 2 and 4. At Station 2 carbon content seemed to
increase in rainy periods and may reflect the increased
input of organic matter from surface runoff and from upstream.
Somewhat irregular variation of carbon content at Station 4
may have been due to tidal influence and accumulation of
material due to blockage of the river mouth during the dry
season.
Chemical parameters:
Mean values and ranges of chemical parameters measured
over the study period are shown in Table 5. Generally,
Stations 1 and 2 were very similar in water quality
reflecting their proximity and common location on the
Carlisle River as compared with Station 3 which was both
physically different and located on the Quarahoon River.
- 1 - 0 - 0
- I
: z - 0 V l V l
- I - I
o : I : ' "
o
o 0 : T C
r t > - ' " " ' 5
r t 0 ~
r t
r t r t
~ 3
. . . v > " 0
' " ' "
. . . . . " " C l n : : r
r o
' "
' " ' "
C
r t : : r Q . I O J
: T
3
o
v > 0 .
' " ~
I I ) ' " r t
" " ' 5 C
. . . w r t I I )
o
' " n
7 ' " " ' 5
. . .
o 3
~ v >
l C I I ) V 1
0 . n
: : : >
v >
~
: : > 0
' "
: : >
0 . 0 o
~ .
' "
I I ) : : : >
v >
N
: : : >
v > 0 -
( " )
<
v > C
I I )
n
( " )
0 .
e - e
. . .
v >
n ' "
' "
t r
o
' " : : : >
n n
~
o
o r o
: : >
W t o
0 .
v >
N
3
N
o
O O O O l C O
~ .
: : : >
o
N
N W < . n
w . . . . .
N
. . . . . . . . . .
N N . . . . . . , N. . f : -
N . . . I . D
3
o N
N 0 0 O N . . . . . . . . . 0
x
' "
0 0 0 ; , . , ; , . ,
C o N
o
N . . . . .
' "
w
3
. . . . .
W . . . . . . . . . .
o N
N
: : >
( J ' I
r o
' "
W C O . . . . . .
N < . n O l < . n C O O l
c o
I I )
: ': ": >
N . . . . . . O l
C o O W C o . . . o
v >
: : : >
' "
a .
3
. . . . .
I I )
N
o O l
( J ' I
3
- I : ' >
C O W O O O < . n O
: : >
' "
v >
\ 0
N
o ~
W
O l
o
. . . . ,
o
. . . . . . . . : T
. . .
( J ' I
W I . D r o
' - W O O W W . . . . . . . . . .
W
o 0 C J " l N O " l 0 ' \ . . . . . . ,
o . . . . .
N
C O 3
3
~ .
( " )
x
~ O O O l O ' \ o N 0
o
' "
' "
" 0
. . . . . . . . . .
. . .
' "
I . D N
N
3
C O . . . . . .
N
r o
3
' "
I I )
I . D < . n
N
: : : >
' " r t
I I )
. . .
v >
. . . . ,
o
" " ' 5
N . . . . .
I " \ ) 0 U 1 N
N
V 1
~ ~ 0 0 a 0 ~ 0 ~ ~ ~
o U 1
r t
3
~ .
U 1 N . . . . . ~. 0
. . . . . .
\ 0 : : : >
r t
' "
~ .
N
o
: : : >
v >
. . . .
I . D . . .
. . . . c o
U 1
. . .0 < . n c o N W O l
t - - 6 N . . . . . . N 0
\ 0
3
\ D
o 0 ~ 0 a ~ w ~ ~
O N < . n . . . . . . O
c o
x
' "
o
W O O ( J l ~ W O
. . . . . . . ~, N ~
o
N
3
W
. . . . . 0 ' \ W
r o
. . . . . .
N 0 1 l . . 0 0 N W
: : : >
' "
N
~ " " " O O O I . D
W . . . . .
o I . D
. . . . . O l
N
o . . . < . n
0 0 0 0 ' \ U 1 0
o
. . . . . .
0 " 1
3
N O ' l U l O
o
o
W
N
o
W W
. . . . .
. . . . W
N N
. . . . .
\ 0
. . . . . < n . . . . . .
. . . . . . . .
o
W . , l : : ; a ( J 1 < . n
. . . . .
. . . . .
I . D W
3
o
t - - ' a 0 0
. . . . . . w
o 0
o
W
. . , .
N 0 0 0
x
' "
U 1
I . D
N
o
. . . . . . . . . . .
N . . . . .
. . .
3
. j : > c o r o
N
. . . . . N W
. . . . . . . j : > N
0 " 1
N l C . . .
U 1
~ 1 . 0 0 ' l ~ t - - '
: : : >
' "
I . D
. . . N . . . . .
f : 9
( X l o . . . . .. ;.w:. :., . ,
N
W
- 0
- 0
c o
C l
r o
0
: : :
. . ,
" "
- s
' "
V I V I C J
3
V I - . J
' "
3
0
. . . . .
0 ' "
r o
~
: : : >
c :
' "
" 0
t t
e - e -
. . ,
. . . . .
l D
<
: I :
< . 0
: : r 3
r o
( 0
z r
r o
- s
: : : >
. . , . . . . . 0 -
' "
c :
0
' "
I V I : : : >
. . . . .
0 0
' "
. . . . . .
0 . .
r o
x
: : : >
. . . . .
V I
' <
. . . . .
' "
< 0 <
0
( 0
N
' "
: : : >
' "
c :
- -
: : : >
r o
3
< . 0
I
. . . . . .
3
w
C >
a
0
: : : >
W
W
0
0 >
U " 1
V ' l
. . . . .
. . . . . .
. . . . . .
. . . . . .
C X >
. p . . . . . .
. p .
- . J 3
' "
0
0
X
- . J U " 1
N ~
' "
: : : >
. . . . . .
3
0 >
r o
N
C X >
w U " 1
. p .
: : : >
- . J 1 . 0
I D
' "
N
. . . . . .
C >
C >
3
w
- . J
0 " >
~
: : : >
V ' l
r - t -
. . . . . .
3
. . . . . ,
. . . . . .
. . . . . .
. . . . .
' "
0
w
0 >
X
' "
"
0
. . . . . ,
~
w
U l
: : : >
. / 0 >
. / 0 >
3
< . 0
W
N
0 0
r o
P 9
0 ' \
1 . 0
0 0
0 0
: : : >
' "
65
Station 4 showed the influence of the sea due to its proximity
to the coast.
The freshwater stations were characterised by water of
slight acidity, in all cases mean pH was less than 7.0 and
values ranged from a minimum of 5.39 at Station 1 to 7.54
at Station 3. Mean total alkalinity values ranged from
1
35.1 mg.l- at Station 1 to 45.6 mg.l-1 at Station 3 with
Station 1 showing the greatest range. Bicarbonate
alkalinity values were similar to those for total alkalinity.
Total hardness, measured as CaC03, was roughly equivalent
for Stations 1 and 2 but was somewhat higher at Station 3.
No carbonates were detected in any water samples during
the study period. These waters were also of high specific
conductance, means ranging between 139.4 ~mhos at Station 2
and 233.9 ~mhos at Station 3. Station 3 showed generally
higher specific conductance attaining a maximum of 800 ~mhos.
Individual ions, in particular chloride and iron, were found
in fairly high concentrations at all sites. Maximal chloride
concentration was attained at Station 1 (85.9 mg.l-1) while
maximal total iron concentrations at all stations were
greater than 10 mg.l-1 Inorganic nutrients such as phosphate,
nitrite-nitrogen, and nitrate-nitrogen occurred in moderately
high concentrations. Average values ranged from 30.0 -
66
54.4 ~g.l -1 (phosphates), 5.0 - 17.6 ~g.l-1 '(nitrites) and
1.5 - 2.8 mg.l -1 (nitrates). At Station 3 nitrite-nitrogen
was more than twice as high and nitrate-nitrogen was lower
than at the other two stations. Another major constituent
included silica (maximal concentration 36.6 mg.l-1).
Dissolved and total solids attained maximum concentrations of
400 and 940 mg.1-1 respectively at Station 3. Dissolved
solids made up more than 50% of total solids measured, the
remainder presumably being attributable to suspended
organic and inorganic solids. Both mean chemical and
biochemical oxygen demands (permanganate value and BOD
respectively) were highest at Station 3 (maximal values
18.5 and 8.3 mg.l-1 respectively). Relatively high oxygen
demands at all stations could partially have accounted for the
relatively low mean oxygen concentrations, in addition to
other factors such as lack of mixing'due to low current
velocities. Nevertheless supersaturation did occur
occasionally. Stations 1 and 3 showed highest mean oxygen
saturation values (63.9 and 75.5% respectively) as compared
to Station 2 (45.0%).
Water quality at Station 4 indicated the influence of
the sea with an almost neutral mean pH and higher buffering
capacity (mean total alkalinity 64.1 mg. 1-1) as compared
67
to the freshwater stations. Specific conductance was high
(up to 37,500 ~mhos) and this was supported by correspondingly
high concentrations of individual ions. In particular
calcium, magnesium, chloride and sulphate ions were found in
high concentrations indicating the direct influence of sea
water. Unlike the freshwater stations mean magnesium
concentrations were greater than mean calcium concentrations
at this brackish water site. Total and soluble iron and
inorganic nutrients were found in lower concentrations than
at the freshwater stations. Silica was roughly within the
same range as at other stations upstream. Total and
dissolved solids were found in substantially higher quanti-
ties as was expected due to higher concentrations of
dissolved salts. Dissolved solids accounted for more than
90% of total solids indicating minimal quantities of
suspended solids, a feature which might have contributed to
generally lower turbidity values at this site. Total
organic carbon was lower than upstream. Chemical and
biochemical oxygen demands and oxygen saturation values were
roughly within the same range as for other stations.
Monthly variation of chemical parameters for all
stations is illustrated in Figs.10 to 21 .
68
8
7
6 /;
L..
-0
5
8 Stn 2
7
6
5
I MAM J J A SON D J FMAMJJASO NDJFMAMJJ
0.. 1980 1981 --1982--
Stn 3
8 Stn 4
7
!=- Lentic conditions
6
5~=~,...---r~~:-:"::::-~~-:--~
MAMJ J ASONDJ FMAMJ
---1980 -1981-
FIGURE 10: Monthly variation of pH at Stations 1 to 4.
69
Stn 1
I
2]/i O'lE
..>...-
2l
.-
I o : : : : : i • •
M A M J J A 5 0 bJFMAMJJAS DJFMAMJ
O'l
1980 ---- 1981----- --1982--E
M
0 [00 Stn 3u
I
o .-
-I200J O'l
E
o .... ..>...-i
c
t~;Stn 4
o I-
(a) 0--0 Total alkalinity
2]::::: _:.____'. (b)....-- BicarbonateF==='" Lentic conditionsI 40 • i : ::
MAMJ J ASONDJ FMAMJ
___ 1980 - 1981-
FIGURE 11: Monthly variation of (a) total alkalinity, and (b)
bicarbonate alkalinity at Stations 1 to 4.
400 70o
Stn
200
a. ~'"
:::J "
""Cl '
C1J
L.
""Cl
O~===+=-'---'r--r--,---r--.---r--,r--r--r---r-,.--r--r--r-~=--T'"""""'T"-=--'--""""""'.....-r.......,
400
Stn 2
VI
~ 200
E
::l.
C1J
U
.g... °~==F=~r-,--,---r--r--r--:~:--r--r--'-- ..... ----,---r-""+==+-;r-==~~::--r...,
.ucg -
M--AMJJASONDJFMAMJJASONDJFMAMJJ1980 1981 - 1982-_
e 800
u Stn 3
u
C1J
a.
lJ) 400
40,000
Stn 4
20,000
~ Lentic conditions
°~=:=--r----.---~~~~~r:-?"~
M A M J J A SON D J F M A M J
---1980 --1981-
FIGURE 12: Monthly variation of specific conductance at
Stations 1 to 4.
71
Stn
, , .. i •
1001 Stn 2
~ J~----
E 0 MA' MJJ'"1A9S8O0N:::D:::JFMA'M
I
Ji JI AI----1981 S
i ON DI J 1+' i--1F9M82A- M J J
III
III
Q)
C
".0.. Stn 3
o
s:
o 100
~ ~---- -o---':>-"--./-'
20,000 Stn 4
.occo
~ Lentic conditions
M A M J J A SON 0 J F M A M J
---1980 -1981-
Monthly variation of total hardness at Stations 1 to 4.
FIGURE 13:
72
Stn 1 20~
/ ~---- [
20] O~C1l
E
.>i.r-: E
o :::JI I I u
Stn 2
E
:::J
VI Stn 3
Q1
C
01
o
Z [~
20 01
rv---- E---E:::J
~
o
U
Stn 4
n--<>--<l COO
1000 (0) 0---0 Calcium
(b) ......... Magnesium
~Lentic conditions
u ......... ~,.
o MAMJ J AS ONDJ FMAMJ
___ 1980---- -1981-
Monthly variation of (a) calcium, and (b) magnesium
FIGURE 14:
ions at Stations 1 to 4.
73
Stn
10
..~...
o Cl. ~---" I
::J
01
5 .~~ .".----•. 0 E
oJ ~ -0 • .-- •• -------------------..----~----~ c0
iii Iii I iii I iii I ~. ~
Stn 2
lID ~
7~ ] ~-~--~--:-~--;--.=:-~--:--,--~ ...~. 0
g MAMJJASONDJFMAMJJASONDJFMAMJJ
~ 1980 1981 --1982 --
Stn 3
01
-E
c
o
~
Stn 4
~----
(a) 0--0 Total iron
(b) ~ Soluble iron
= Lentic conditions
Monthly variation of (a) total iron, and (b) soluble
FIGURE 15:
iron at Stations 1 to 4.
80
40

.-
40 a. ..-
:J ...-
I
,,, o -
01
L.. " E
"'0 '
o-,-::;:::::+==:::;:---r-~--,---,----.--r--o--T~~~r-=~:-;:~~'· C1l
"'0
L..
o
Stn 2
~-.- [~
,...,,,
...- 0 '.
-I FMAMJJ A S 0 NDJFMAMJJ
01 1981 --1982--
E
..C.1..l Stn 3
0
.s::
a. A.... 40-:J 40 --Vl --I
C'l
3~..:.·.~:: EC1l"'0
L..
20,000 -0x: Stn 4
u
10,000
(a) 0--0 Chloride
5'j -. o (b)'-" Sulpha te= Lentic conditions
o ....... ---MAMJ --.
- ......-...-
u •• .-.- • ...-.-~~. •
JASONDJFMAMJ
___ 1980 - 1981 -
Monthly variation of (a) chloride, and (b) sulphate
FIGURE 16:
ions at Stations 1 to 4.
75

 __- SIn 1 [40
0- - ~----~-------~
2 ~ 0
r-
I
"0
0\
~ " E
.-. "e0ll J '",
o .-------------- . -----------~
400 ~- ~ ~I~[O~
--~----
o
r----
I o ~------------~----- ---------~
MAMJ JASONDJFMAMJJAS·ONDJFMAMJJ
1980 1981 1982-
Stn 3
40 rz Stn 4
(a) 0--0 Silicate
\ /----~----~ [ V)
(b)""'-'" phosphate
100 0 ~ Lentic conditions
01••• ~--;-.-.-.-r-'--~ .--,--~
M A M J J A SON 0 J F M A M J
_--1980 -1981-
Monthly variation of (a) silicate, and (b) phosphate
FIGURE 17:
at Stations 1 to 4.
76
Stn 40
" 0--_ _ ---0..... [
20 ;- - \J --- - - - - - -- - - 0- -0- -0-0- ~ <.T --0 0
~
I
"'0
<II
O--'--:l==::;:::~-.--r--,--,- r=- -~- .:..:- -.;..- -:..:;-~- -:.,:-:..:.- ~• .:...:- -:,..:.- ;:.-~ 'l:.- .:;=-,-'-=-- f'-'r -..... ,- ... -..lI.!:.!.::.s.:-.~..--..'~...~~~~-:":"-i-:':-:t~'~.
..-..
Stn 2
Stn 3
z
40
-_- ~-,0 ,1 U__hu u -_m-<f- 0 01
::l.
o dt-- t "'~U;';':'--,---,--;-i--l-=f- I .... ..C..1.l
.\..-..
Stn 4
~ ~ __ -- ~-- __ -o [Z (a) ~ Nitrite
(b)..-. Nitrate
01, , , I==I==l Lentic conditions"---:--=u,--;--,--:;o-,- ..u, .. ~
MAMJ JASONDJFMAMJ
___ 1980 - 1981 --
Monthly variation of (a) nitrites, and (b) nitrates
FIGURE 18:
at Stations 1 to 4.
77
5 t n 1 500.::-
o a. 0-0-----0-
:::l ----0--0-----0--0--------0--0 <>-_---0---- ..........<> [.!..:<>--0-"""" -----0 0 E
"0
VI
~ 0 ~ ----__ ..-.- ~ _ ..0----_ .. -0____........_ "0
oJL-===,.......---._ ..-.._.--,---,-..---.----,.-.;.....-.._-..._.-:,- o_-:;''''-'--'---'---::':==--",,,- -_-..;.~.~..-.-.:.:.::I=--"'-r-- -...:;. VI
. Iii iii iii iii iii I ............i... i '+ Ii ......i I
-0
Q)
>
o
VI
Stn 2 500~
<>--0-<>-<>- ----0--_ - -0-<>-----0--0- - - - ---0--0-----<> - ---<>-------~ [0 .s
..-
~, o~
~'"J0 I : : : :--,--~/:,' ~i .--=-..~--._-~--~--~--~-=:--.--~--~
"0 MAMJJASONDJFMAMJJASONDJFMAMJJ
o --- 1980 1981 -- 1982--
VI
.o...
o OO Stn 3
~ ..-
~ - -- -<>-- - - - --0- --- -0--0- ----- _-0C ...-
O I
0'1
E
l°]l...:z:=A=--r-r-, --='~_=.-~'--.'--~,-'-.-. VI-""+'~-==I-~-~..',.-,--_ "0~I_- -:
o
VI
Stn 4
40.000] (0)0--0 Total dissolved solids
"\ ---------- (b)e---e Total solids
o ------, -....-..,.-.-~/.... -0, == Lentic conditionsI I I :
MAMJJASONDJFMAMJ
___ 1980 -- 1981--
FIGURE 19: Monthly variation of (a) total dissolved solids, and
(b) total solids at Stations 1 to 4_
78
Stn
10 ,,-,-0 [ ~0 ~ /-----'-~/'"
J!]~/~/~~ Q)::Jo>
~ .Q..)
o
Stn 2 20 §01
C
_ 0
----0 [ 0 ~
.--
I
0~~"f"T~~~~~~~.....-=:-'--=::z:--r-~
01 MAMJ J ASONDJ FMAMJJ ASONDJ FMAMJJ
E 1980 1981 --1982 --
r---
o 20 Stn 3
o
III
~//~ T\ A ;
o , , , , , .:::-~ ~ ~o
01
Stn 4
~---~!/'\ [ J
1 (0) c>--o Permongonate value
(b)"-' BOD7
.-----~. f\:: ~ Lentic conditions
o MAMJ J AS 0 NDJ FMAMJ
---1980 -1981-
Monthly variation of (a) permanganate value, and
FIGURE 20:
(b) SOD at Stations 1 to 4.
79
c:
cv
Cl
>-
X
o
200
100
c:
0
......
~0 0
..:..:.l. ONDJ FMAMJ JASO NDJFMAMJ J
0 1981 1982--
VI
•-e 10 Stn 3
[
O.=-
I
Cl
E
c:
cv
Cl
>-
X
o
"'0
cv
>
10 .~ Stn 4
[
0°
(a) 0--0 Dissolved oxygen
(b) .............'10 saturation
= Lentic conditions
MAMJ J ASONDJ FMAMJ
---1980 -1981-
Monthly variation of (a) dissolved oxygen, and (b)
FIGURE 21:
percentage saturation of oxygen at Stations 1 to 4.
80
Variations in pH were similar for all stations but did
not show any consistent trends (Fig.l0). Maximum variation
was seen at Station 1 (range of 2.19 pH units) as compared
to Station 4 (range 1.28 pH units). While some dry periods
resulted in an increase in pH of the water followed by a
decrease with the onset of the floods (Stations 1 and 2:
October/November 1981, Station 3: April 1981), at other
times increased runoff caused higher pH values (Stations 1,
2 and 3: July 1980, December 1980). At Station 4 however,
higher pH values were associated with times of low discharge
and increased salt water influence.
Total alkalinity, also expressed as bicarbonate
alkalinity reflected gross variation in pH (Fig.Il ).
Variation at Station 2 was minimal compared to other stations.
At these other stations both total and bicarbonate alkalinity
tended to increase during lentic conditions and periods of
low flow.
Specific conductance showed the largest seasonal
variation at Station 4 as a result of salt water influence
during the dry season and increased freshwater discharge
during the rainy perl.o d (F'19. 12) . These values however
refer to surface samples and do not preclude the possibility
81
of salinity stratification. Station 3 attained a maximum
conductance of 800 ~mhos during the severe 1980 dry season
with lower but still marked increases during subsequent
dry periods. Although Stations 1 and 2 also showed increased
conductance with lentic conditions, values were never
higher than 400 ~mhos which was also recorded in the 1980
dry season.
Variation in total hardness, a measure of calcium and
magnesium ions expressed as CaC03, is shown in Fig. 13, while
concentrations of calcium and magnesium ions are shown
separately (Fig. 14). At all stations there was some tendency
for hardness to increase with low flow periods. This was
supported by trends shown for individual calcium and magnesium
ion concentrations. Highest magnesium concentrations seemed
to occur later in the dry season than calcium. While distinct
increases in both calcium and magnesium ions occurred at
Station 4 in the dry season, magnesium ion levels were often
more than twice those of calcium.
Both total and soluble iron occurred in such
concentrations for this element to be considered a fairly
important chemical component of the freshwater sites. There
was a slight tendency for increased concentration of total
82
iron (predominantly insoluble Fe3+) in the dry periods at
these freshwater stations while at the same time decreases in
soluble iron (predominantly Fe2+) were observed (Fig. 15).
At Station 4 iron was found only in very low concentrations
except during the rainy season when discharge of freshwater
with high levels of iron increased. During the dry season,
most of the iron present at this brackish water station was
in the insoluble form.
Both chloride and sulphate anions showed fairly clear
seasonal variation in concentrations (Fig. 16). Chloride
concentrations increased consistently with the existence
of dry conditions attaining maximal values for freshwater
sites at Station 1 during the 1981 dry season. Marked
increases also occurred during low flow periods at Station
4. Of all the sites sampled Station 2 tended to be the most
stable with respect to chloride concentrations. Sulphate
ion levels generally increased during the dry season at
Station 1 but at Stations 2 and 3 distinctly low concentra-
tions were evident during some lentic periods (Station 2:
October/November 1981), January-March 1982, Station 3:
September 1980, February/March 1981).
Silica, measured as silicate, was relatively constant for
83
Inorganic nutrients measured included phosphate, nitrite-
nitrogen and nitrate-nitrogen (Figs. 17, 18). However
analyses were irregular and as a result only tentative
generalisations may be proposed. While phosphate levels
increased early in the dry periods and subsequently decreased
markedly at the freshwater stations, one very high value
was recorded in July 1981 for Stations 1 and 2. It is
unclear whether this is characteristic for flood periods
since only one other flood period was analysed for phosphates
(June/July 1982) or whether this was due to extrinsic
factors such as fertiliser application. The latter
possibility could not be assessed since analyses for
associated nutrient elements were not carried out on these
water samples.
Analyses of nitrites and nitrates were made too
sporadically to make definitive comments on the seasonal
variation of these parameters. However there seemed to be
84
an inverse relationship between concentrations of nitrite and
nitrate ions. Nitrite levels were generally highest during
the 1980 dry season at all stations and decreased with the
onset of the rainy season. Nitrate levels decreased during
the same dry period.
As expected, variation in total dissolved and total
solids reflected that of specific conductance (Fig.19).
However readings were too irregular for the full picture
of events to be obtained. There appeared to be an increase
in concentration of both parameters during the dry season
and this was especially marked at Stations 3 and 4. The
other stations were relatively stable with respect to these
parameters.
Oxygen demands were recorded as chemical oxygen demand
(COD or permanganate value) and biochemical oxygen demand
(BOD) (Fig.20 ). BOD values showed distinct increases
coincident with the dry seasons and lentic conditions while
COD values showed slightly different fluctuations. It
appears that increased COD values occurred both towards the
end of the dry periods and also during the flood period for
all stations.
85
Monthly variation in dissolved oxygen concentrations
and percentage saturation are shown in Fig. 21. Although
many readings appear to be in a normal range for most
freshwaters, some cases of extreme supersaturation up to
more than 190% were recorded. Such readings for Stations 1,
2 and 4 were abrupt and seemed unrelated to ensuing stagnant
conditions. It may be assumed that errors in methodology may
have caused such aberrant values. However at Station 3,
supersaturation was attained gradually over several months
and seemed to follow a clear trend of seasonal variation
with maximal oxygen concentrations occurring during the late
dry and early rainy periods and minimal values at the onset
of the dry season and during short lentic periods. This
pattern coincided with the seasonal development of phyto-
plankton communities (cf Results section ·Plankton') and
may explain such trends in oxygen concentrations but some
errors in methodology can not be excluded altogether.
Although some field measurements of oxygen were made with
an oxygen meter and probe, this method was unreliable at low
oxygen concentrations. However, at normal oxygen levels,
differences between field meter readings and Winkler analyses
were relatively small. Measurements of dissolved oxygen
at several points within Stations 1 and 2 showed no major
spatial variation in oxygen concentrations although minor
86
decrease with depth and increase with current speed were
observed.
On one occasion on 82/06/16 the opportunity arose to
make observations and obtain water samples during a flood
at Stations 1 and 2. Water temperatures were low and
turbidity very high as compared to values obtained at other
times. Notable features of the water quality at this time
were the very low specific conductance and dilution of ions
such as chloride, iron and magnesium. Oxygen levels were
only 50.3 and 44.1% saturation for Stations 1 and 2
respectively. Other parameters were found to lie within
the range of other monthly samples taken.
In summary, the range of monthly variation of chemical
parameters differed for the four stations. Variation at Stations
1 and 2 was relatively slight compared to Stations 3 and 4.
Station 4 was most subject to fluctuations in water quality
due to the varying seasonal influences of seawater in the
dry periods and increased stream discharge in the rainy
periods. Station 3 showed large fluctuations based on the
alternation of lentic and lotic conditions during the year.
87
Diurnal variation:
Diurnal variation of physical and chemical characteristics
was studied over 24-hour periods during both rainy and dry
seasons at Stations 2 and 4. After sampling Station 2 on
81/03/25-26 to represent dry season conditions, inadequate
chemical analyses were carried out and hourly sampling was
repeated on 82/03/31-04/01. Differences in weather
conditions existed between 1981 and 1982 dry seasons with
1982 being somewhat wetter, therefore data for both sampling
periods are presented. Hourly variation in irradiance and
surface current speed is only shown for rainy season sampling
periods.
Variation in air, water surface and bottom temperatures,
pH, specific conductance, total alkalinity and dissolved
oxygen are illustrated for Station 2 in the rainy season
sampling period (Fig. 22) and the 1982 dry season sample
(Fig. 23). Air and water temperatures and dissolved oxygen
content only are illustrated for the 1981 dry season sampling
period (Fig. 24).
Depths at Station 2 did not change during the sampling
periods being 0.5 m in the dry season and 0.7 m in the rainy
88
200
N
I lrradiance
E
3:
0
..-
I Current
VI
E 1:]
u
30 Tempe ra tures
28
u 0--0 Air
• ...·.11·,0· .'0."
26 x-x...JI!I ...·.II.... x-x Water surface
CJo •• O Water bottom
24
pH
Specific conductance
Total alkalinity
Dissolved oxygen
0\
E :J
1] Oxygen saturation
o==F"""O Night
i ,
I 4 6 8 10 12 14 h
16 18 20 22 24 2
Hourly variation of physical and chemical parameters
FIGURE 22:
at Station 2 in the rainy season (80/08/07-08).
89
28
Temperatures
26
u ....x..-x-x-x-x 'J<
o X /x-x . ......x- -x x- :P o--oAir
24 0 :-:-·1lD··'·00"·'0· " '0" '0·· '0" '0" '0"'0"'0" '0' "0 .. '""",".'.0..x.;.:.:IlD.'=0;"-X'II.!.I..'.-GX......:1....l·.D.0...x.,_0x"-.~l x-x Water surface
o...oWater bottom
6,8] pH
6.6
6.4 ~-,---,r--r--r--r--.---.-.,.--,--.,.--,--r--.---.....----..---.-.--.,.--,--r--'---r--T"---'
Specific conductance
i i
Total alkalinity
Dissolved oxygen
~
I
01 2 ~-----
E
10 Oxygen saturation
= Night
~-----
10 12 14 16
Hourly variation of physical and chemical parameters
FIGURE 23:
at Station 2 in the 1982 dry season (82/03/31-04/01).
90
3 Temperatures
0--<> Air
x-x-x x-x Water surface
/ ....x......
..0"'0"'111=
/ a' 0...0Water bottom... [J•.•[J.
/ 0'" :-"'x-x "'[J
u 24 x .'.' ..... '.,/
o III···[J
22
20
Dissolved oxygen
10 Oxygen saturation
~ F=F=='O Night
0
0--<;>--0----
0
12 14 16 18 20 22 24 2 4 6 8
10 h
10
FIGURE 24: Hourly variation of temperature and dissolved oxygen
at Station 2 in the 1981 dry season (81/03/25-26).
91
season. There was some temporal variability of irradiance over
the daylight period due to changing cloud cover during the rainy
sampling period. During the dry season sampling period no
current flow was detected. However current speed decreased
slowly over the August sampling period, increasing slightly
after light rains at mid-morning.
During all three 24-hour samples at Station 2, air and
water temperatures showed diurnal variation with highest
temperatures recorded during the late afternoon and lowest
in the early morning. Diurnal variation in temperatures was
greatest over the period 81/03/25-26 with air temperatures
varying by 10oe, surface water temperatures varying by 7.5°C
and bottom temperatures by 6.5°C. Temperature variation on
82/03/31-04/01 was relatively small and was similar to that
recorded in the rainy season. Surface water temperatures
generally tracked air temperatures and bottom water
temperatures tracked those at the surface with a lag period
of up to two hours. During the August sample current speed
seemed sufficient to ensure some degree of mixing so that
surface and bottom water temperatures were very similar and
equilibrated quickly. In both dry seasons some stratification
92
in the water column occurred especially during the day
when rapid heating of the upper layers occurred. During
the night however stratification was less marked and on
several occasions reverse temperature gradients were
established with bottom temperatures higher than those at
the surface. Such temperature inversions may have resulted
in 'overturns' of the water column and promoted mixing of
layers hence reduced stratification at night.
Specific conductance showed minimal changes over
both sampling periods although some decreases occurred which
could not be explained by rainfall or increased runoff
(Figs. 22 and 23). Variation in pH and total alkalinity
were minimal over both dry and rainy sampling periods.
However dissolved oxygen content fluctuated to some extent
during all three sampling periods. During the rainy season
and 1981 dry season samples dissolved oxygen increased
slightly during daylight hours indicating the influence of
photosynthetic production of oxygen. However during the
night some fluctuations occurred which may be related to
overturn of the water column or equilibration of surface
and bottom water temperatures, leading to a decrease of
surface oxygen levels by mixing with hypoxic deeper layers
(for example Fig. 23, 82/04/01:0500 h).
93
Diurnal variation of physical and chemical parameters at
Station 4 is illustrated in Figs. 25 and 26 for rainy and
dry seasons respectively. Stages of the tide as determined
from Tide Tables are also indicated.
Depths remained constant during each sample period at
0.7 m and 0.8 m for rainy and dry season samples respectively.
No current flow was detected during the dry season sample
and only slight currents (less than 3 cm.s-I) due to tidal
influence were measured during the rainy season sampling
period. Irradiance was measured only during the rainy
season sample period and was maximal at midday to early
afternoon decreasing drastically in mid afternoon due to
overcast conditions and subsequent sunset.
Air and consequently water temperatures changed less
over the 24-hour period during the rainy season sample. At
this time air, surface and bottom water temperatures varied
by only 8.5°C, 5.0°C and 2.0°C respectively. In contrast
during the dry season diurnal variation of the same parameters
was II.5°C, 7.0°C and 7.0°C respectively. Maximal temperatures
were measured during the early afternoon decreasing steadily
to a minimum between 0300 and 0600 h. As at Station 2 water
temperatures followed air temperatures closely with a lag
94
Irradiance
N
I
E
3
Temperatures
32
x ....x........ 0-0 Air
30 ,/ xx x-x Water surface
/
x· -x 0..·0Water bottom
.u 28 -,.,0. x
..0' ". .O··{J. 'x-Ill!. ll!I
..·o"'0'-cr '0'" .0..·0'..'-""'3"/.. ...'.1.lI"'Ill",
0 ··ll.!.I.l.l.I=ll!I,O···O...
26 x-x-ll!I:-:"
24
pH
600j Specific conductance'B
E 400 o-v-...n--rY
:::J
200 -.--......-.,--.,,...---,.--.--.----r--r--r--.---.--r-.-----.--,--..---.....--.-----.---.--~~
~ I.°l Total alkalinity
E J'-.---T---r-.,---,---,---.--......-.,..--,r-T--.--.,---r--,--..-...----.---r--r---r-~~
Dissolved oxygen
Oxygen saturation
Night
h
tide
FIGURE 25: Hourly variation of physical and chemical parameters
at Station 4 in the rainy season (80/08/19-20).
95
32 Tempera tures
30
0-<> Air
x-x Water surface
x 0"'0 Water bottomI ....x.
I .:0
u
o
. ~.:.r;J'
24
C
22
20
80] pH
7.0
Specific conductance
t~J
15,000
Total alkalinity
.- 14]
.!.-
CTI
E
120 i
i
Dissolved oxygen
.-
I
CTI
E ']
o i
Oxygen saturation
100]
= Night
~
0
o i '"' 24 2 4 6 8 i 1'0 hii i 14 16 18 20 2210 12
high low
high low tide
Hourly variation of physical and chemical parameters
FIGURE 26:
at Station 4 in the dry season (81/04/01-02).
96
period. During the rainy season sampling period stratification
of the water column occurred during the day, there being a
4.0°C difference between surface and bottom water temperatures.
In the dry season stratification was not seen to develop to
this extent. As at Station 2 temperature inversions also
occurred at night, possibly resulting in overturns and mixing
of the water column (for example Fig. 26, 81/04/02:0300-
0400 h).
Specific conductance values were much lower during the
rainy season sampling period than in the dry. Slight
salinity stratification was recorded on one occasion at
rising tide during the dry season period. Surface conductance
was 20,000 ~mhos while bottom conductance was 22,000 ~mhos.
Fairly marked diurnal changes in conductance occurred
at both times of the year and may have been due to sea water
inflow during high tides (for example Figs. 25 and 26,
80/08/19: 2300 h, 81/04/01-02: 1400 - 1500 h, 0400 - 0600 h).
Mixing of the water column by overturns or surface-bottom
temperature equilibration may also have promoted small
increases in surface conductivities. pH values were higher
in the dry season but tended generally to change little over
both 24-hour periods. pH was not constant with depth; on
one occasion during the dry season sample period surface
97
pH was 7.48 while bottom pH was 7.05. During this same period
a sudden decrease in surface pH might have been due to mixing
of surface and bottom water layers. Alkalinity was low and
remained fairly constant over the rainy season sample period
but was higher and fluctuated more widely during the dry
season peri od. There was a tendency for a1ka1inity to increase
generally during the dry season sampling period and might
reflect increased inputs of carbon dioxide of respiration
during the night. Dissolved oxygen levels showed diurnal
variation at both times of the year with higher values during
daylight hours possibly due to photosynthetic activity. Daytime
oxygen saturation values were higher during the dry season
sampling period.
Stream biota
General observations:
Aquatic macrophytes were not found within the study area.
However algal mats did develop during slow flowing or lentic
conditions in well illuminated parts of the stream and were
mainly composed of Oedogonium and Spirogyra with numerous
attached diatoms.
98
A list of benthic and other macrofauna (excluding fishes)
collected during the study period at the four stations is
given in Table 6 and Appendix 3. Each station supported
quite rich faunas with Station 4 having a distinct brackish
water fauna. In terms of the numbers of taxa represented,
Station 1 showed greater diversity than all other stations,
while Stations 2 and 3 were roughly equivalent and Station 4
was the least diverse. Station 1 was especially rich
with respect to aquatic arthropods, particularly insects,
presumably due to the more favourable flow conditions and
the sandy substrate at this site.
The most commonly collected taxa at all freshwater
stations included oligochaetes, particularly tubificids and
occasionally naidids, the trichodactylid crab Dilocarcinus
dentatus~ aquatic insects such as the dragonflies Perithemis
mooma and Dythemis spp, the water scorpion Ranatra mixta,
the gerrid Brachymetra albinervis~ and chironomid larvae,
planorbid gastropods, Pomacea glauca~ and sphaeriid bivalves.
Ancylid limpets were common at Stations 2 and 3 but were not
found at Station 1 while nematodes were always abundant at
Station 3 but not elsewhere. Vertebrate and other invertebrate
taxa were collected only occasionally.
99
TABLE 6: Summary of aquatic macrofauna (excluding fishes)
collected at Stations 1 to 4 during the study
period.
Stations
Taxon 1 2 3 4
Nematoda x
Nemertea x
Oligochaeta - Tubificidae x x x
Naididae x x x
Enchytraeidae x x
Hirudinea - Glossiphonidae x x x
Polychaeta - Nereidae x
Cap ite11idae x
Cladocera x x
Ostracoda x
Copepoda x x
Isopoda x
Amphipoda x
Decapoda - Penaeidae x
Palaemonidae x x
A 1phaeidae x
Portunidae x
Trichodactylidae x x x
Arachnoidea Hydracarina x x
Ephemeroptera Leptophlebiidae x x
Odonata Coenagri onidae x x
Calopterygidae x
Aeshnidae x
Gomphidae x x
Libellulidae x x x
100
TABLE 6: (continued)
Stations
Taxon 1 2 3 4
.Hemiptera - Belostomatidae x x
Nepidae x x x
Hydrometri dae x x
Notonectidae x x
Gerri dae x x x
Vel iidae x
Coleoptera - Dytiscidae x
Gyrinidae x
Hydrophil idae x
Carabidae x
Diptera - Chi ronomidae x x x
Heleidae x x x
Gastropoda - Planorbidae x x x
Ampullariidae x x x
Hydrobi idae x
Ancyl idae x x
Bivalvia - Sphaeri idae x x x
My tel 1idae x
Amphibia - Bufonidae x
Hylidae x
Leptodactylidae x
Reptil ia - Chelidae x x
Emydidae x
Kinosternidae x x
Crocodyl idae x x x
101
Station 4 supported a brackish water faunal assemblage
with groups such as nereid polychaetes, isopods, amphipods,
hydrobiid gastropods and juvenile mussels being well
represented. Other crustacean fauna included penaeid and
snapping shrimp and swimming crabs CaZZinectes sapidus.
Terrestrial crabs on the stream banks and vegetation included
Goniopsis cruentata~ Sesarma sp~ Aratus pisonii and Cardisoma
guanhumi. No aquatic insects were found in this part of the
stream. Caiman were occasionally seen as far as these lower
stream reaches.
Plankton:
The groups represented most commonly in the plankton
were algal groups with desmids, some filamentous chlorophytes
and cyanophytes, diatoms and euglenoids being prominent in
most monthly samples at all stations (Appendix 4). However
Station 1 supported a somewhat poorer planktonic community
when compared with other stations. Crustaceans such as
cladocerans, ostracods, copepods (mainly cyclopoid) and
amphipods were restricted in distribution being most prevalent
at Stations 3 and 4. Only copepods and occasionally ostracods
were found at Stations 1 and 2 and even then only rarely.
While certain taxa were represented at most stations, notable
102
exceptions were the diatoms Coseinodiseus and Synedra and
the medusa of the freshwater hydroid Craspedaeusta which
were found only at the brackish water Station 4. Other
restricted taxa included the colonial rotifer ConochiZus
and the previously mentioned crustaceans which were found
mainly in the slower flowing deeper Stations 2 and 3.
From the monthly samples some gross variation in
seasonal abundance could be recognised (Table 7). For all
freshwater stations (and especially at Station 2) numbers
of taxa appeared to increase in the dry season. Stations
1 and 3 showed less marked seasonal changes in terms of number
of taxa coll ected. Certain groups occurred only in the dry
season or increased in abundance at this time, for example
blue green algae, desmids, euglenoids, rotifers, copepods
and hydracarina. During the dry season, desmids in
particular increased in abundance at Station 1 while blue
green algae, euglenoids, rotifers and hydracarina predominated
at Station 2. At Station 3 euglenoids were abundant through-
out the dry season and also in August, while rotifers peaked
in numbers in the dry season only. Also at Station 3 peaks
in abundance of diatoms occurred late in the rainy season
(December) and ostracods were found in large numbers in
August 1980. Station 4 showed a different trend with fewer
103
TABLE 7: General trends in the seasonal variation of plankton
composition and relative abundance at Station 1 to 4.
Stn 1 Stn 2 Stn 3 Stn 4
Taxon R D R D R D R D
Myxophyceae xx x x
Chlorophyceae x XXI x X X X X x
Bacillariophyceae x x x x xx x xx
Euglenidae x x xx xx xx x
Zoomastigophorea x x x
Sarcodina x x x x x x x
Ciliata x x x x x
Hydrozoa x
Rotifera x xx x xx
Cladocera x x xx
Ostracoda x xx x x x
Copepoda x x x xx xx
Amphipoda x x x
Arachnida x xx x
x present
xx common or abundant
R Rainy season (July-December 1980, June 1981)
o dry season (January-May 1981)
I primarily Desmidaceae
104
planktonic taxa in the dry season. Increased abundance
of diatoms occurred in the early rainy season in June and
July while dense crustacean-dominated plankton samples were
collected in August/September and again in January/February.
Results of quantitative sampling carried out in the
1983 dry and wet seasons for Stations 1 and 2 are shown in
Table 8. Dry season samples collected at Stations 1 and 2
were not exactly comparable since Station 1 had apparently
dried out and subsequently refilled with recent rains and
therefore represented a newly developing planktonic
community. Nevertheless both stations supported more
planktonic taxa in the dry season than in the rainy
season with diatoms, actinopods, rhizopods and rotifers
present in addition to flagellates and ciliates found all
year round. The composition of the flagellates varied from
one season to the next, being almost entirely composed of
euglenoids in the dry season but containing none of these
in the rainy season. The rotifers, collected only in the dry
season samples, were solitary species at Station 1 whereas at
Station 2 colonies of ConoehiZus were common. Plankton
densities varied markedly with Station 2 samples having more
than 13,000 organisms.ml-1 in the dry season and decreasing
to 27.5 organisms.ml-1 in the wet. Station 1 mean plankton
105
TABLE 8: Comparison of mean densities of planktonic organisms
in dry and rainy seasons at Stations 1 and 2.
Mean number of organisms.ml -1
Dry season Rainy season
Stn 1 Stn 2 Stn 1 Stn 2
Bacillariophyceae 2.50 2.50 0 0
.Mast igophorea 1.251 12907.501 48.752 26.252
Sarcodina 17.50 20.00 0 0
Ciliata 2.50 242.50 2.50 1.25
Rotifera 3.75 7.59 0 0
Copepoda 0 0 1.25 0
Tota 1 ± Standard Error 27.50 13180.09 52.50 27.50
±7.2 ±211.9 ±7.5 ±9.2
1 predominantly euglenoids
2 no euglenoids
106
densities increased slightly from dry to wet season from
27.5 to 52.5 organisms.ml-1 respectively. Comparisons of
total numbers of organisms in samples showed that mean dry
season densities of planktonic organisms at Station 2 were
significantly higher than mean densities at the same station
in the rainy season (t = 62.02, df = 3, p<O.OOl) and were
also significantly higher than mean densities attained at Station
1 in both the dry (t = 62.05, df = 3, p<O.OOl) and rainy season
samples (t = 61.93, df = 3, p<O.OOl). No significant
differences were found between dry and rainy season mean
densities at Station 1 (t = 2.40, df = 6, p>O.05) or between
rainy season mean densities at Stations 1 and 2 (t = 2.10, df =
6, p>O.05).
Benthic macroinvertebrates:
Monthly samples of benthic macroinvertebrates were non-
quantitative but gave a general indication of the nature of
the communities and gross seasonal changes at each station
(Tables 9 to 12). Certain groups such as oligochaetes
(primarily tubificids), chironomids, planorbid gastropods,
ancylid limpets and sphaeriid bivalves were dominant in the
benthos at the freshwater stations. No clear seasonal
changes in the nature of the fauna were observed at these
stations. Fauna at Station 4 included primarily nereid
polychaetes, isopods, amphipods, hydrobiid gastropods and
- - - i
: z
0
: r :
r r l : r :
C l
0
c ; : : J
- l G > ( D
~
+
- 0
0 -
' <
n
: r :
0 . "
~ . > <
" " '
3
- s
0 -
: : T
0
- 0
: : T
r t v > r o 0
( D
<
. . . ,
< D
c
r o
: : >
r t
r - t -
. . . . .
0
. . . , 0 -
. "
. . . ,
r o 3
r o
: z
" "
( ) ' "
" " r t
. . . , ( )
r o
r t
0
" " '
0 0
( D
<
. . . ,
: : T
: : >
. "
. "
: : >
0 -
- 0
r o
- s
0
" "
0
v > . "
0
" "
r o
" " '
- 0
r o
. " 3
0 -
" " '
r - t -
: : >
r t
" "
. "
3
. "
r o
0 -
- 0
" "
- s
~
. "
r" o" '
r o
V l
r t
0 " 1 W
~
c o
> - '
. " N
o - m
- " "
- " "
0
m
C O
C O
C O
- " " - " "
r o
' "
: : >
N
V >
W
. . . . ,
> - '
C O
< o - " "
N
m C O
0
N c r .
> - '
W
U 1 < . D
C l
m
> - '
o v
N
C O
C O
c o
m
0 0 C l C l
- " "
a
: z
- - - i
~
+ 0
c ; : : J
r
r r l
< . D
c . . .
W
W
N
N
3 :
0
N > - '
: : >
. . . . . .
. . . . . .
. . . . ,
. . . . . .
m
m
r t
0 ' 1
m
N ' m
C l
: : T
- " "
~
' <
<
a
. "
. . . ,
. "
r t
. . . . . .
. . . . .
W
< 0 - " "
- " "
- " " 0
c o
m
N c o
c o
: : >
' "
0
. . . . .
. . . . . .
N
. . . . ,
. . . . .
. . . . .
. . . . . .
- " "
- " "
0 -
m c o
m
m c o
N
c o
' "
l D
: : >
r t
: : T
c . . .
. . . . .
. . . . .
( )
m
0 " 1
. . . . .
< . D
3
. "
c o
( )
. . . . .
. . . . . .
N
. . . ,
c . . .
. . . . .
w
. . . . . .
a
- " "
0
N
m
m
~
0 0
: : >
<
I D
. . . . .
+ : : t : o
r t
I D
0 -
. . . ,
. "
. . . . . .
N
. . . . ,
r t
0 " 1 ~
- " "
I D
~ c o
0 " 1
- " "
V >
. "
. . . . . .
. . . . . . r t
W
. . . . ,
W
U 1
- " "
V l
0 0
W
U 1
- " "
r t
. "
r t
. . . . . .
N
. . . . ,
0
~
N
c o
~ : : >
N
> - '
. . . . . .
w
. . . . .
. . . . . .
w
: : >
- " "
W
w c
0 " 1
W
0 " 1
3
0 -
l D
. . . ,
V >
a
3
I
. . . . . .
. . . . .
N
. . . . . .
C l
U 1
. . . . ,
. . . . ,
~ ~
N c o c o c o
U 1
~
~
c o
U 1
' "
' " ' "
' "
> - '
N
U 1 > - '
C l W
a
. . . . ,
. . . . . .
w
3 :
- " "
W . . . . . .
N c o
' "
- " "
> - '
. . . . . .
N
. . . . . .
< 0
> - '
- " "
. . . . ,
N
. . . . . .
m
~ < . D
- " "
. . . . ,
N
. . . . .
c o
- " " c o
N
N
N
0
0 0 m
c o ~ c o N
0 0
- " " 3 :
< . D
~
N
0 0
- " "
' "
a c . . . . .
. . . . .
. . . . . .
. . . . ,
~ w
c . . . . .
L O T
. . . . ,
w
- " "
- j
n
: I e
Q 0
C J
- j
c o
+
S
o ,
n
I ' <
0
o ,
> <
' "
: r u 0
V I r o
r t
<
' "
0
o ,
t . O
~
- S
e - t - : >
r t ' "
: : >
0
0
Z r o
r o - S
- S
' "
' "
( " )
n
n
0 0 M -
- S
0
' " ' "
<
: : Y
r o
o ,
: : : >
u
- S
V I - S
0 ' "
0
' " ' "
~ .
r o
o ,
r o ' "
3
' "
' "
M -
: : >
' "
3
' "
U
a .
' "
" "
' "
" "
r o
r o
V I
M - . . . . . .
: P
. . . . . .
C J ' I
. . . . . .
. . . . . .
~ W
C J ' I ~
A N
C J ' I
~ m
' "
r o
: : >
N
w
V J
N
. j : >
W
C J ' I m
~
N
0 0
O J ~
> - - '
W N
i l l
0
W
w W
O J O J
C O
N
N N
~ O J
0
N
N
; Z
. . . . . .
. . . . . .
U l N
- j
m
. j : > o
m
m
: P
C O
r
r n
C J
+
. . . . . .
C l
. . . . . . . . . . . .
W
W
. . . . . .
. . . . . . . . . . . .
w m
3 :
N
N
C J ' I 0 C J ' I m
0
: : >
M -
: ; , -
. . . . . .
N
~
N
l D
N
O J ' <
. j : > o
O l
<
0 >
- S
~ .
O J
O J
0
0
M -
' "
0
. . . . . . : : >
c o
N
0
N
. j : >
. . . . . .
U l
. . . . . .
C J ' I m 0 0
0 -
l D
. . . . . .
. j : > o
: : >
. . . . . .
N
W W
M -
N O J
N
O J m
: ; , -
n
c . . .
3
0
0 >
> - - '
n
i l l
- S
0 0 0
> - - '
c . . .
: : >
0
<
r o
- S
M -
r o
: P 0 -
+
- S
0 >
M -
I D
V I
. j : >
. j : > o
O J
. j : >
O J ~ 0 >
r t
V l
r t
0
0 >
r t
0
: : >
. . . . . .
U l W
. . . . . .
w
O J W
N
. . . . . .
W
t . O
W
: : >
< :
w
N
3
. j : > o
. j : > o
t . O
0 0 N
C J
0 -
. j : > o . j : > o
t . O
l D N
r o
. . . ,
V I
N
O J
~
3
. j : >
O J O J O J O l
c . . .
I
t . O < . D t . O . . . . . . . j : >
N
N
. j : > o
N
O J
O J
. j : > o
N
t . O
t . O
. . . . . . . . . . . .
. . . . . .
U l
. . . . . .
. . . . . . . . . . . .
. . . . . .
. j : >
3 :
O J O J
O J 0 0 . j : >
. . . . . .
0
: P i l l
0 0
N
W
N
. j : >
. j : > o
U l . j : >
N
3 :
. j : > . j : > o
. j : > o
. j : >
N
0
c . . .
0
c . . . .
8 0 1
- j
0
: : c
- i
c o n
+ G J 0
0
O J
- - '
- ' .
0
0
O J n V l
: : c
X
r t - V l - S
- 0
r o : : r ' " 0
c - t -
<
0
u : : l
O J C
- '
O J
c - t - r o
e - t - - S
: : l
0
0 -
r o ' " 0
- S 0 >
- S r o
- ' .
n
0 0
0 n
<
- s
: : z
: : : >
: J
' " 0 0 - : J 0 -
0 > 0
0
0 >
O J r o
0 O J
0 0 >
0 -
l 1 >
O J
0 - r o
0 >
: 3
V l
r - t -
0 >
O >
0 >
0 -
- j
3
O J
: P o
' " 0
r o
C O
r
r o
V l r r l
. . . . . . .
. . . . . . . .
. . . . . .
c - t -
- . . . J
. . . . . . . 0 " 1
. . . . . .
C J )
0 >
C J )
0
A
r o
: J
3 :
0
0 V ' l
: J
r I -
: : r
- '
N
. . . . . . .
' <
U 1
. . . . . . .
0 " \
- + : : >
- + : : >
C J )
l . O
0 C O 0
C O
<
O J
C O
0
- S
O J
r l -
O
: : z
0
: J
0
- t o ,
0 -
+ 0
m
: J
r I -
: : r
W . . . . . . .
. . . . . .
o
U 1 . . . . . . .
. . . . . . .
W . . . . . . .
~
- + : : >
N
N
N
0 " \
~ 0 " 1
C O
: 3
O J
( )
- S
0
W
. . . . . . . .
: J
. . . . . . .
0
W
~
0 " 1
<
~
0 " \ N
O " l
0
r o
- S
r t -
r o
0 -
~
~
- S
C J )
C J )
3 : O J
r - t -
r o
U l
. . . . . .
O J
C J )
. . . . . . .
~
- + : : >
r l -
C J )
O 0 " \
C O
. . . . . . .
>
l . O
V l
C O
r t -
. . . . . . .
O J
r I -
0 0
O " l
. . . . . . .
0
0
~
0 " 1
3 :
: J
W
: J
C
0
c . . . . .
: 3
0 -
r o
- S
U l
0
: 3
c . . . . .
I
N
6 0 1
- - I
- , : )
. . . . . . .
) : = 0
G " l
O J
O J
- I
0
+
V 1
: 3
O J
0
> <
- - - '
0
- 0
V 1
< 0
r - t -
- 0 ' <
: : : r
r - t -
O J
O J ~
n
0
. . . . . . . - - - ' - S
; : : ' l " "
0 -
- 0
0
<
: z
O J
0 >
0
- 0
0
r o
0 -
0
O J
c - T
O J
0 -
V 1
O J
O J
O J
- I
: 3
) : >
- 0
C O
. . . . . . .
r
r o
r n
V I
u . >
. . . . . .
w
c - +
u . >
N
W
O J
0 ' >
Q )
7 "
r o
~
3 :
~
0
. . . . . .
V ' l
0 ' >
~
0 ' 1
r t
: : : r
. . . . . . .
' <
~
~
Q )
<
I . D
0
0 ' >
-
0 0
O J
+ : >
0 0
- S
0
O J
c - +
- J .
~
+ : >
. . . . . .
\ . 0 0
: z
Q ) 0 " 1
C O
Q )
0 " 1
~
-
0
- t ,
0 -
r o
+ 0
: : : : : l
r t
: : : r
o
: 3
0
O J
n
- S
0
- J .
: : : : : l
u . >
W
<
. "
N
N
r o
- S
r t
r o
0 -
- S
O J
0
r t
r o
V I
W
W
0 >
f - '
f - '
W
r t
f - '
0 ' >
~
0
. . . . . .
) : : >
I . D
~
~
~
N
V 1
0 0
f - ' r t
0 >
r t
. . . . . .
f - '
f - '
W
0
u . >
W
f - '
+ : >
0 " 1
C O
N : : : : : l
N
N 3 :
0 0
0
+ : >
+ : >
: : : : : l
c
: 3
0
0 -
r o
- S
V 1
: 3
I
c . . . .
0
N
0 1 1
111
immature mussels. The latter were found in large numbers
only during dry periods.
While the data are not quantitative they do show that
at certain times of the year numbers of benthic invertebrates
decreased to relatively low levels, for example after spates
at the beginning of the rainy season (June/July) and during
some periods of low flow. At other times, numbers of
invertebrates rose to fairly high densities, especially during
the 1982 dry season when moderate flow was maintained at
Stations 1 and 2.
Late in the 1980 dry season Station 1 dried out completely
leaving only moist substrate covered with a thick layer of
leaves and other detritus. A sample of this substrate and
detritus contained oligochaetes, mites, hydrophilid and dytiscid
beetles, dipteran larvae (including heleid midge larvae) and
small planorbid snails (Table 13).
A more quantitative analysis of the benthic macroinver-
tebrates was made to compare dry and rainy season densities
and biomass between Stations 1 and 2 (Table 14).
btal densities of benthic macroinvertebrates were 1066 and
112
TABLE 13: List of macrofauna collected from damp
stream bed at Station 1 in May 1980.
Taxon Density (per m2)
Oligochaeta 20
Hydracarina 176
Coleoptera 68
Diptera
Heleidae 96
Other 400
Gastropoda 16
Total 776
113
TABLE 14: Comparison of mean densities and biomass of benthic
macroinvertebrates in dry and rainy seasons at
Stations 1 and 2.
Mean densities (per m- )
Dry season Rainy season
Stn 1 Stn 2 Stn 1 Stn 2
Oligochaeta 444 544 56
Hirudinea 11 78 22
Hydracarina 11
Ostracoda 11 33
Chironomidae 267 100 122
Odonata 11
Gastropoda 56 44 967
Bivalvia 67 356 2100
Total ± Standard Error 845 78 1066 3311
±110.3 ±77 .8 ±184.2 ±2971.5
-2)
Biomass (mg.m 1082.1 46.7 1295.4 4243.1
± Standard Error ±451.5 ±46.7 ±305.3 ±3468.9
114
3311 individuals.m-2 for Stations 1 and 2 respectively in the
rainy season. Large numbers of tubificids, chironomid larvae,
planorbid snails, limpets and bivalves were present in the
rainy season. In comparison oligochaetes were represented
by tUbificids as well as lumbriculid and naidid worms in
the dry season. Dry season samples at Station 2 were
practically depauperate and may have been the result of
anoxic conditions developing at that time. Mean densities
of dry season samples at Station 2 were significantly
lower than those at the same station in the rainy season
(t = 4.15, df = 3, p<0.05) in addition to being lower than
those at Station 1 in the dry season (t = 6.56, df = 5,
p<o.ol). No significant differences in mean densities existed
between dry and rainy season samples for Station 1 (t = 1.20,
df=5, p>O.1). During the rainy season, samples from
Stations 1 and 2 showed no significant difference in mean
densities (t = 2.83, df = 3, p>0.05).
Although densities of benthic macroinvertebrates varied
somewhat, no significant differences were found in benthic
biomass either between stations or between seasons (t-test,
p>O.05 for all comparisons). Benthic standing crop varied
from as low as 46.7 mg.m-2 in the dry season to 4243.1 mg.m-2
in the wet season at Station 2.
115
Two prominent benthic invertebrates which were also
studied were the trichodacty1id crab Dilocarcinus dentatus
and the prosobranch gastropod Pomace a glauca. The former
is unusual in that the eggs and juvenile crabs are brooded
in the abdomen of the female and the latter in that it is
both gi11- and lung-breathing.
Monthly samples of D. dentatus and P. glauca were
pooled for all the freshwater stations sampled. Ranges and
means of carapace widths for D. dentatus are shown in Fig.27
Juvenile crabs were present in the population at the beginning
of the 1980 and 1981 rainy seasons (June-August) and also
in December 1981 when particularly heavy rains occurred after
a pronounced petit careme. Mean carapace width decreased at
this time but increased rapidly afterwards possibly indicating
rapid growth of individuals and maturation to reproduce by
the following dry season. Maximum adult size attained was
51.7 mm carapace width for a male individual; maximum size
of females measured was 49.6 mm carapace width. Large
individuals were present in the population throughout the
year but maximal size showed a slight decrease in the mid to
late rainy season (October/November) possibly due to death
of larger individuals at this time.
116
(0)
60 0 7 21 17 4 5 10 5 11 1 3 1 1 4 5 2 9 9 11 32 14 4 6 3 0 0 4
Vl
....c..~..C
:20 40
~~
"U
Q.olC
""""-1
u
°oV..lc
eooQ.l
uE 2
---
O--&--,r-T---r---r---r-..,....--r-,.--,.---T---r----r---r---.--.....--..-..--.--.---.-....--..---.....-.--.--~
(b) 0 0 2 0 0 0 15 9 8 4 6 0 11 11 30 5 29 18 5 27 19 14 6 1 5 1 3
O-L-..--..--..--..---r-.....---.---.----'----'----'----'----'---'--"--'---'----'---'---'---'---'----'-""'-""'-""--'
(c)
I
I
_Vl .
o ~ 20 II
V I~loVl ..~E ..
I
EO' I
I
z~<Ov'
FIGURE 27: Monthly variation of size of (a) Q. dentatus, and (b)
£. glauca, and (c) numbers of egg masses of~. glauca
at Stations 1 to 3 (numbers at top indicate sample
sizes) .
117
Variation in size of P. ql-auca individuals and numbers
of egg masses counted over specified areas are shown in
Fig. 27. Mean body whorl size remained relatively constant
over the study period and did not show distinct seasonal
variation like D. dentatus. Small individuals were generally
collected in samples mainly during rainy periods. Peaks in
egg-laying generally coincided with the appearance of small
individuals in the population, eggs hatching in approximately
two weeks. Maximum adult size attained was 43.0 mm body whorl
width for a female individual; maximum male size was 38.3 mm.
Allochthonous input:
Only allochthonous animal material falling onto glue
boards was investigated at Stations 1 and 2 during dry and
wet seasons to determine the amount of potential food
material available to fish and other aquatic organisms. Table
15 summarises the composition and quantity of material from
dry and rainy season samples in 1983. It was noted during
the 24-hour sampling period that flying insects tended to use
the glue boards as a resting area thus producing over-
estimates of actual 'f'all in", Nevertheless these samples
generally reflect terrestrial insect abundance in the vicinity
} : >
. . . . . .
0
0 n
- 0 0
r : : r : f T 1 - i
- i n
0
- S
0
V l V l 0 . .
: : T
- s
0 - 0
I 1 l - s r o
~
O J
" ' 0
0
0 0 M -
" ' 0 : : T
- 0
3
' <
~ ( )
( )
- 0 : : : : r : : : : l
V l
r o
r o
M - 0
: : s
0
O J r o
0 0
z r - 0 O J M -
0
r o 0 .
3
: : l
" ' 0
- 0 M -
" ' 0 I 1 l
r o 3
0 " ' 0 M - : : : : l
0
- s
0
C T "
O J
M - - S M -
" ' 0 - 0 r o 0
O J M - - s
0 .
0 0 .
r o
r o O J
I 1 l 0 " ' 0
M -
M - - S
~
. . . . . ,
O J
O J
O J - 0 - s - s
r o I 1 l M -
O J
0 >
O J
I 1 l
O J M -
- S
- S
~
r o
O J O J - S
: t : : >
O J
- s
C O
0 >
I
r n
' - '
U " l
: : : : l
( X )
0
m
. . . . . . .
~
W
" - J
~
. p o
. ' - '
( X )
t - '
N
. . . . . .
N N
t - '
N U 1 U " 1 0
' - '
m
( / )
n
e r
M -
0
0
: : : : l
O J 3
- 0
- S
' - '
t - '
N ' - '
0 >
0 .
: E :
w
W
W 1 . O
V l - s
0
. . . . . .
M -
0
w
I - - ' 0 W m
0 0
\ . 0 1 . O W
0
~
• l ' - ' - s
V I
0 .
I - - ' ' <
N
0
( X ) W
- . . . J \ . 0 m
N ~
t - ' 1 . O
0 0
c 0
: : : : l
- s
V l
r o
0
: : : : l
0 >
- - t ,
t o
V l
0
z s
m
0 .
3
: : : : l
0
t - '
U " 1
U " 1
C D
- S
. p o
t - ' 0
0 W
. . . . . . I - - ' - . . . J 0
U " 1 0
~
w 0
0 >
' <
( / ) : : : : l
O J
M -
: : : l
: : l
: : : : l
c :
0 .
r - o 3
t - '
C T
- s
: E :
~
W
N
( 0
O J
N M -
0
( X ) I - - ' W 0
0
0
U " 1 0 0
W
U " 1 0
- s
.
V I
: : : : l
t - '
. . . . . . . 0
( X ) 0
" - J ~
W 0
' <
0 >
V l : : : : l
r o 0 -
O J
V l
C T
: : : : l
" - J
0
0
N W
~
. p o
0
t - ' : : : : l
t - ' 0
C J ) I - - ' 0
t - ' N ~
~ m
W 0
V l
3
( / )
0 >
V I
O J
M -
V I
M -
: : : : l
( / )
0
I - - ' ' - '
- - t ,
l ' - '
I - - ' M -
N : E :
U " 1
; ; 0
O J
0 0 m 0 0 W M -
N " - J 0 0
1 . O
0 \ . 0
~
O J
M - 0
. . . . . ,
C J )
- ' .
I - - ' t - '
t - ' U " 1
N N
C O m ' - '
W
u : : : : l
0
: : : : l
: : : : l 0 >
' <
V l
: : : : l
V l
V I
r o . . . . . . .
: : : : l 0 >
t - '
3
0 V l
t - ' V I
O J
t - '
" - J
m 0 : : l
t - ' l ' - '
. . . . . . 0 0 0 0
. . . . . . . " - J " - J 0 0 0 0
0 : : : : l 0 .
n
( / )
O J
M -
N c :
: : : : l
u : : : : l
: : : : r
N
r l -
t - '
: E :
. . . . . . .
N
M -
1 . O
0
C J )
t - ' 0
" - J N 0 0 0
0 0 0
t - ' 0
0
: : : : l
t - ' W
" - J 1 . O
W - . . . J
t o
u : : : : l
c :
( 0
8 1 1
: c
3 :
- i
f - - '
N - i
Q O
' <
0
0 1 +
1 +
U l : 3
M -
r +
r o
O J
s : : n
: E
O J
V l
O J
V l
: : l
: : l
r o
- ' : : l
- - '
M -
r - t -
- ' 0
: : l M -
O J
O J
U l - 0
- ' < . 0
0 -
: : l
: : l
: : l
: : l
0 -
3
r +
r o
O J
. . . . . . r o
s : : :
0 -
s : : :
< . 0 0 -
r o
0 -
: : l : : l
O J
O J 0 3
3
- S
r o
M -
0 -
- S
0 - - S
: 3
3 - - i
. . . . . . O J
0
I 0 - r o
r o 0 - O J
) : : >
- + , s : :
U l
" " ' S
N
" " ' S
O J
U l
V l
U l f " T l
V l
I " T l
. .
r o
- S
- S
r r t
0 -
- S
- S
3
I 0
0
~
- S
N
- S
U 1
t - '
: : l
N
0 " 1
0
N
t . T 1
W
1 +
1 +
. . . . . . . . .
t - '
V l
N
t . T 1
N 0 " 1
n
r +
0 0
< . D 0
N
0
: : l
< . 0 0
- . . . J
- . . . J
: : : s
r t
. . . . . . .
N
t - ' + = >
w
. . . . . . .
+ = >
: : E :
N < . 0
0
: : : s
0
r t
C
- S
• I ' V
N - . . . J
U 1
( 1 )
' <
0 -
U l
r o
O J
U l
0
: : l
t - '
- . . . J
: : l
0
- . . . J
t . T 1
0
1 +
1 +
V l
t - '
I ' V 0 0
r +
w
0
W
+ = >
: : l
t - '
< . 0
< . D
N
N
t - '
t - '
t . T 1
N U 1
: : E :
< . 0
0
0
r t
U 1 0 " 1
: : l
N
0
U 1 t . T 1
1 +
1 +
V l
t - '
N
0 0
r +
U " 1
< . 0
- . . . J
+ = >
. j : : > : : l
N
- . . . J 0 0
t - ' . . . . . . .
N
0
N : : E :
0 " 1
+ = >
r t
. j : : > ; ; 0
. . . . . . . t - '
O J
: : l
' <
U l
r o
t - '
O J
W : : l
1 +
1 +
N
< . D 0 U l
t . T 1
t - '
0
t - ' W
N
: : l
N
U 1 0 V l
W
t - '
. . . . . . . . W
- . . . J r +
: : l
. . . . . . .
0 " 1
C O
N
0 - . . . J W
: : E :
r t
. . . . . . .
U 1 W
6 1 1
120
of the stream and thus give an indication of potential 'fall
inl to the stream.
The majority of organisms collected on the glue boards
were terrestrial or winged insects. Dipterans were dominant
in the samples both in terms of numbers and by weight, with
coleopterans and hymenopterans being next most important.
While most of the organisms collected were winged forms which
were probably caught on landing, non-winged forms were also
fairly well represented, for example spiders, collembolans,
orthopterans (crickets) and ants, and these made up as much
as 27% of the total sample by weight on one occasion. Some
non-living animal material was also collected, for example
exuviae, parts of insects and frass.
Some seasonal variation in the taxa collected was noted,
for example mayflies were collected only in the rainy season
samples. However, other groups do not show any conclusive
variation from one season to the next. Statistically significant
differences were found only between mean densities of dry and
rainy season samples from Station 2 (t = 3.80, df = 6,
p<O.Ol) with rainy season samples containing more organisms
(as many as 1323 organisms.m-2). No significant differences
121
were found in the mean biomass of samples from different
seasons or stations (t-test. p>O.05 for all comparisons).
Maximum biomass of Ifall in' was 697.4 mg.m-2 at Station 1
in the dry season.
DISCUSSION
The phenomenon of intermittency seen in both the Carlisle
and Quarahoon Rivers under study appeared to be a result of
a combination of factors, in particular precipitation regime
and soil characteristics. In a classification of tropical
rivers, Salek (1983) recognised four major categories of
rivers based on general climatic factors. precipitation
regime and geography: (i) Equatorial rivers of the humid
tropics; (ii) rivers of wet and dry regions; (iii) dry
climate rivers; and (iv) rivers of tropical mountains.
Category (ii) includes four sub-categories based on whether
the rivers are found in relatively dry or moist climates and
whether located on lowlands or uplands. Generally rivers
of this category show pronounced annual fluctuations in
discharge due to seasonality of rainfall. Lowland rivers in
relatively dry conditions may show an intermittent flow
regime due to an extended dry season especially if they have
122
small catchment areas. However, Salek (1983) noted that
these lowland rivers are subject to annual rainfall totals
less than 1000 mm and vegetation characteristically being
dry savanna, both conditions not being met for the present
study area.
Annual rainfall totals however, do not completely reflect
the precipitation regime and rainfall intensity and duration
during the year are important in determining runoff patterns
and hence stream flow (Ward 1975). The study area was
characterised by periods of intense rainfall for one to
several days followed by dry periods resulting in irregulari-
ties in stream conditions. During the wet season, high
water levels and stream flow were maintained and augmented
by floods but during dry spells, water levels receded and
current flow slowed or ceased completely. The overall
pattern seen was one of irregular fluctuations in habitat
size and alternation between lotic and lentic conditions at
least twice per year. Small streams generally tend to be
more unstable than larger ones (Hynes 1970) and in addition,
variability in annual rainfall produced varying conditions
of stream flow from year to year and determined the extent
of drying out of the stream in successive years.
123
Although precipitation and climate generally have a
primary influence on river regimes, the rate of infiltration
through surrounding soils is also of major importance (Ward
1975). This is perhaps the single most important factor in
the regulation of intermittent stream flow since it
determines how precipitation will be split up into the
categories of overland, inter- and ground water flow (Williams
& Hynes 1977). In areas of impeded drainage where infiltration
is reduced, overland or Surface flow is of greater importante
than inter- and ground flow thus promoting intense short term
floods after periods of heavy precipitation. Ground flow
may make some contribution to streamflow during the wet
season but during dry periods, the ground water table recedes,
lowering the level of the stream until it ceases to flow.
Only a few pools remain in the channel which may then vanish
due to evaporation (Williams & Hynes 1977). This situation
is characteristic of areas of clayey soils with impeded
drainage such as those found in the study area.
Physical characteristics of the streams under study
such as temperature also showed some variability with quite
large annual and diurnal surface water temperature ranges.
It is generally quoted that tropical watercourses show
minimal fluctuations in water temperature (Hynes 1970) and
124
Fittkau (1964) determined that in forest-shaded Amazonian
streams, annual and daily fluctuations were roughly 1°C around
a mean of 24.5°C. Other studies have documented water
temperature ranges somwhat larger than this, for example
Bishop (1973) measured annual ranges of up to 10°C in the
lower course of the Sungai Gombak and about 5°C variation in
diurnal water temperatures. Similarly, Harrison & Rankin
(1976 a) recorded a diurnal range of 4.1°C for the Greathead
River in StVincent and Adebisi (1981) working on the inter-
mittent Ogun River in Nigeria reported a range of temperatures
similar to that obtained in the present study.
Factors influencing stream temperatures include substrate,
turbidity, depth, exposure ~nd altitude (Hynes 1970, Welcomme
1979). Of these, varying degrees of exposure contributed to
differences seen between stations with Stations 3 and 4
showing highest water temperatures. It is unlikely that
any variation occurred as a result of distance from the
source of the stream since continuous flow was not always
maintained and altitudinal differences were minimal. Low
night-time air temperatures and the rapid cooling of
surface waters in the dry season may have resulted in over-
turns and mixing of the water column at the stations
studied. Carter (1934) in his work on the streams and swamps
125
Chemical composition of river waters is determined by a
multiplicity of factors which inevitably impart a degree of
uniqueness to each river. Nevertheless, it has been possible
to estimate the I average , chemical composition of the world1s
rivers as well as averages for individual continents
(Livingstone 1963). Table 16 compares average concentrations
for selected constituents of the world's rivers with
concentrations for the freshwater stations in the study area.
. . . - - - 0
. . . . . .
( / ' )
( )
. . . . . . C A l
- - i
V l
3 : n : : z :
. . . . . . O J
. . . . . .
0 C
: : : r
P I P I
- s
. . . . . . . . . . . .
( " ) - S
~
- '
r - t -
c o - '
0
r - t -
. . . . . . O J
P I P I
- 0
: : : : : l
: : : : : l o 0
- s
- - '
( " ) 3
0 3
: : : r - s
C D P I
- s
. . . . . . < . Q r o
P I
P I P I 0 -
V I C
r l "
r l "
r i -
a . O
r i -
C D 0 -
3
. . . . . . . . r o
P I . . . . . . . . . .
C D
C : : : : : l
C D
l - S
V I O J
O J
3
. . . . . . . .
V I
- I l V I r i -
- - l
- s
0 C D
: : x : -
- - '
0
V l
C O
. . . . . .
3
<
r -
C D
r n
r -
a .
( " )
. . . . . . . .
V I
O J
<
O " l
. . . . . .
0
r l "
- - '
: : : : : l
r o
. . . . . .
l O
V I
a .
V I
r i -
0
P I
n
r l "
: : : : : l
: : : r 0
: : : : : l
r o
r o a .
3
- 0
. . . . . . . . . .
( / ' )
. . . . . . . O J
. . . . . .
0
U ) - S
- s
: E :
. . . . . .
C
C D
C " I
0
V l V l
r i -
w
- S
. . . . . .
: : : r
: : : r 0
~
. . . . . .
: : E :
0 -
) : >
O J
N . . . . . . . . . . . . . . . . . .
U 1
r l " 0
3
. . . . . . O J
a . . . . . .
w 0 U 1 - . . . . . s 0 0
+ : >
. . . . . . .
r o C D
<
- ;
. . . . . .
- s
. . . . . . r o
- . . . . J N
a 0 0 0 + : >
V l
. . . . . .
- S
r o
( " )
V l
O J
- - '
P I
< . Q
r l "
r o
c
: : : : : l
r o
( " )
0 -
r l "
' <
r o
P I
0 -
V l
<
. . . . . .
C D
U ' l
: : E :
- ;
O J
r l "
0
P I
t : : C D
r l "
l O
V l
O J C D
r l "
- ;
: : : r
C D
<
. . . . . .
0 ' 1
W r o
. . . . . . . . . . . .
. . . . . .
\ D . . . . . .
- - . J . 0
a + : >
+ : > - S : : x : -
.
- ;
c
O J
3
U )
U )
a
+ : > U 1 N - - . J < . Q
0 P I
0 0 r o
. . . . . .
r o - S
<
. . . . . .
C D
( ) - ;
r i -
O J
' <
~
: : E :
P I
- 0
r i - P I
- ;
C D
- ;
P I
0 ' 1
. . . . . .
N
V I
3
3
. . . . . .
w
U 1 0 . . . . . .
0 N
a
0 0
. . . . . . C D
: : : : : l
r i -
. . . . . .
+ : >
U 1
0
+ : >
- . . . . J ; 0
P I
C D
O J - ;
: : : : : l
V I
- I l ~ 0 -
o < . Q
r o
- S
r i - o
+ : >
. . . . . .
V l
: : : r
- I l
. . . . . .
a . . . . . .
w N
N
0 0
+ : >
0 0
3 U ' l
0
. . . . . .
0 ' 1
a N
0 ' 1 a U 1
0
- . . . . J
P I
r l " O J
V I
: : E :
x
~ ~
C D
0
. . . . . . U )
. . . . . .
0 ' 1
0 U )
a
C " I
V l
- ;
0 -
- '
0
. . . . 3. . .
a .
- I l
1
r o
w O J
. . . . . .
3 ~
. . . . . .
0 ' 1
. . . . . . .
N
+ : >
C D
V l
W 0 ' 1
+ : > C " I - . . . . J
N U )
0 0
+ : >
P I
: : : : : l
U )
U )
U 1 + : >
- . . . . J
N
+ : >
U 1 w
9 2 1
127
Most notably, water samples from the study area were lacking
in carbonates and to a lesser extent calcium ions, when
compared to world averages. Under equilibrating conditions,
weak acidity of the water allows existence of carbon dioxide
of respiration primarily in the form of bicarbonates with
only minimal levels of carbonates below the sensitivity of
measurement (Hutchinson 1957). Low calcium levels indicate
a soft water and could be partly explained by the lack of a
geological source of this element in the catchment area.
Chloride and to a lesser extent magnesium and sulphate
concentrations are greater than world average levels and
presumably indicate the proximity of the sea and influence
due to wind-induced sea spray (Golterman 1975, Cole 1979).
Concentrations of iron in the study streams were particularly
high being seven times greater than average river water and
more than 16 times greater than the levels acceptable for
potable water (0.3 ppm, APHA 1971). Such high levels of iron
may contribute to estimates of total hardness (APHA 1971).
High levels of iron appear to be typical of many tropical
waters (Payne 1986). Moderately high nutrient levels
(especially nitrates) may have resulted from runoff and
leachates from surrounding agricultural regions (Reynolds
1984) as well as from the nitrogen-fixing activities of
leguminose plants nearby or from leachates of le~ves and
other organic matter immediately after falling into the
stream (Bishop 1973).
128
Total organic carbon values at the study site (2.4 _
1
59.0 mg.l- ) were very high compared with data for European
inland waters (1 - 10 mg.l-1) given by Hynes (1970). However,
he acknowledged that the soft blackwaters of tropical rivers
would contain much larger amounts of organic matter. Data
cited by Head (1976) indicate a range of total organic carbon
in rivers from 15 - 30 mg.l-1 with 60 mg.l-1 as an extreme
value. Such organic load is partially and non-quantitatively
measured by determining chemical oxygen demand (permanganate
value) which also measures the oxygen demand of reduced com-
pounds (Bishop 1973). Permanganate values measured at the
study sites (3.3 - 18.5 mg.l-1) fell within the range of
values given by Hynes (1970) for lowland USSR rivers (5.5 -
20.12 mg.l-1). Biochemical oxygen demands (BOD) also reflect
organic loads (in addition to biological activity levels) and
can be indicative of organic pollution: values between
1 - 3 ppm indicate clean water, about 5 ppm doubtful water
quality, 10 ppm and greater severe pollution (Hynes 1960).
BOD values measured at Chatham (0.3 - 8.3 mg.l-1) indicate
moderate to high organic loads. Comparatively higher total
organic carbon, COD and BOD values at Station 3 in particular
could be attributed to high inputs of organic material in
the form of dung as this site was used as a watering hole
for cattle. This is supported by high nitrite levels as
129
compared to nitrate concentrations, a feature indicative of
organic pollution (APHA 1971). Oxygen demands due to moderate
organic loads and reduced mixing due to slow current speeds
could account for generally low average dissolved oxygen
values comparable with those found in similar waters in
Indiana (Schneller 1955) and in the Amazon (Foldats 1982).
When compared with average river water composition for the
South American continent (Table 16), the study site showed
higher concentrations of most constituents especially
sulphate, chloride and magnesium. This could be attributed
to the proximity of the study site to the sea. In addition,
average total dissolved solids for the study site was more 'II'
than twice that for the South American continent. On the
basis of more specific data water chemistry of the study area
contrasted markedly with the nutrient and electrolyte-poor
waters characteristic of the black and clear waters of the
Amazonian region but more closely resembled that of the
white water rivers (Klinge&Ohle 1964, Sioli 1964, 1975 a,b,
Schmidt 1972, Furch & Junk 1980). Other mainland rivers
in Venezuela are also of much lower conductivity than the
streams under investigation (Lewis & Weibezahn 1976, 1981,
Sanchez et aZ 1985). Stronger similarities existed between
water chemistry of the study area and that of small lowland
130
streams of St Vincent (Harrison & Rankin 1976 a) and
St Lucia (McKillop & Harrison 1980). In St Lucia in particular,
ranges of most water quality parameters for slow-flowing
temporary lowland streams coincided extremely closely with
data for the Chatham streams with the exception of higher
sulphate ion concentrations in St Lucia due to the influences
of volcanic fumaroles. Comparable nutrient levels (especially
of nitrates and phosphates) have been recorded in mainland
rivers on the eastern plains of Venezuela (Sanchez et at
1985).
Studies conducted locally on the Maracas River (Carrington
1980) and in the rivers flowing into the Nariva Swamp (Bacon
et aZ 1979) show slight differences in water chemistry from
that found at the Chatham streams. Generally the Chatham
streams were slightly more acidic, lower in alkalinity and
calcium ions but higher in total dissolved solids. While
the Maracas River contained low concentrations of iron, some
rivers entering the Nariva Swamp showed levels of total iron
slightly in excess of those found at Chatham. Such high
levels (especially of soluble Fe2+ in Chatham) may be due to
the slightly acid reducing conditions existing (Cole 1979).
The estuarine nature of Station 4 imparted certain
131
distinct characteristics indicative of marine influence on
water qual ity at this site. In particular, higher pH and
specific conductance and the resemblance of relative ion
concentrations to that of sea water were apparent. In
addition, certain estuarine chemical processes alter the
nature of inflowing freshwater to effect changes in certain
constituents such as iron, phosphates, suspended solids and
total organic concentrations (Burton 1976). Dissolved iron
is precipitated out largely by the influence of increasingly
saline water with high pH and such activity is increased in
the presence of suspended sediment (Liss 1976). Phosphate
concentrations in temperate estuaries are known to be
regulated to constant levels close to 37 flg.l-1 irrespective
of fluctuations in salinity by adsorption and release at the
sediment surface, a process believed to be dependent on the
presence of ferric oxides (Read 1976). Organic and fine
particulate suspended material may be precipitated by the
presence of dissolved salts in estuaries (Hynes 1970, Head
1976). Fine particulate suspended material becomes coagulated
forming a heavy flocculent material (Burton 1976) which may
settle thus allowing increased light penetration due to
lower turbidity levels. Such a process, in addition to
downstream transport of organic matter could account for sedi-
mentary carbon levels up to six times greater at this site
132
than upstream stations.
Seasonal variation in water chemistry of the Chatham
streams was largely dependent on alternating lotic and lentic
conditions and fluctuating levels of discharge. The latter
is particularly important in all lotic environments
(Golterman 1975). In particular, reduced flows and increased
rates of evaporation during the dry season accounted for
increased specific conductance and hardness as well as
increased concentrations of individual ions such as calcium,
magnesium, and chloride ions and as a result, high concentra-
tions of total dissolved solids. Accumulation of organic
matter from dry season leaf fall during times of low flow
may have contributed to the existence of an organic colloid
buffering system (Kleerekoper 1955) which could result in
higher total alkalinities in dry periods.
When lentic conditions persisted over prolonged periods;
at Stations 2 and 3 in particular, increased phytoplankton
populations composed of actively photosynthesising species
developed. This resulted in increased dissolved oxygen
concentrations in the late dry and early rainy seasons. Under
such circumstances, nutrients such as phosphates and nitrates
may show declines with the passage of the dry season possibly
133
as a result of uptake by these organisms (Hynes 1970) and by
decomposing organic matter (H.B.N. Hynes 1975). High rates
of photosynthetic activity could also result in increased
pH values (Prowse & Talling 1958).
Times of reduced discharge were in some cases associated
with increased chemical and biochemical oxygen demands and
reduced dissolved oxygen levels due to accumulated organic
matter, high temperatures and stagnation. As a consequence,
highly reducing conditions may have developed leading to
certain associated changes in water chemistry. Under reducing
conditions and in association with high concentrations of
electrolytes and humic colloids, increases in alkalinity,
development of higher concentrations of total iron and
phosphates (as a result of their release from anoxic sediments) I'
and conversion of insoluble Fe3+ to soluble Fe2+ might occur
(Hutchinson 1957, Hynes 1970, Bishop 1973). Later in the
season, decreased sulphate ion concentrations due to
reduction of this anion and formation of iron sulphides and
hydrogen sulphide might have taken place. Fluctuations in
pH would also be expected to influence features such as iron
cycling with acid reducing conditions leading to the formation
of soluble Fe +2 from insoluble Fe 3+ . din eeper water layers.
Upon mixing with more oxygenated water at the surface or in
134
the early rainy season, oxidation of iron to the insoluble
form would result in precipitation or production of a
flocculent brown surface film and subsequent decrease of
total iron in the water column (Hutchinson 1957).
In contrast to this situation in the dry season, increased
discharges during floods and generally during the rainy season
resulted in dilution effects on most ions, total hardness and
specific conductance, and the establishment of moderate
dissolved oxygen levels. Gross seasonal changes similar to
those seen in the study sites have been recorded in other
tropical rivers (cf.Bishop 1973, Adebisi 1981 a, Leveque et aZ
1983) and in some local rivers (Bacon et aZ 1979).
Seasonal variation in water chemistry at Station 4 was
largely dictated by increased seawater influence during times
of low flow and decreased salinities in the rainy season.
Such fluctuations strongly resembled those seen in brackish
water outlets from the Nariva Swamp (Bacon et aZ1979) and
are consistent with what is known of chemical processes taking
place in estuaries (Burton 1976).
Phytoplankton associations in the freshwater study sites
135
may be considered according to Hutchinsons (1967)
classification of temperate lake plankton, as being a
euglenophyte plankton assemblage characteristically composed
of dense blooms of various Euglena species (and in some
instances Trachelomonas species) and found in very small
and organically polluted bodies of water rich in non-humic
organic matter (cf Pennak 1978). In addition, occurrence of
the colonial diatom Melosira is indicative of eutrophic
waters while there is a strong correlation between soft
waters deficient in calcium and magnesium and the occurrence
of large numbers of desmids (Hutchinson 1967). Hutchinson
(1967) further poi nts out that species of Euglena and Phacue ,
in particular can utilise 'ammonia as an inorganic nitrogen
source as well as organic nitrogen sources, a factor possibly
contributing to their abundance in the waters in the study
site which were of relatively high organic content. Such
euglenophyte-dominated communities have also been described
for turbid swamps and pools and small creeks in Suriname
(Leentvaar 1975, van der Heide 1976). Station 1 showed a
relatively impoverished phytoplanktonic assemblage with
desmids predominating in the dry season unlike other stations.
This was a rather oligotropic community despite generally
high inorganic nutrient levels but consistent with lower
organi c loads.
136
Poorly developed zooplankton communities such as those
seen at the freshwater study sites, have been recorded for
other Neotropical areas, for example in streams and swamps
of Guyana (Carter 1934) and in Suriname (Leentvaar 1975).
In contrast, very high zooplankton densities have been
recorded for floodplain rivers especially in quiet backwaters
or during periods of reduced flow, for example in the Blue
Nile peaks of zooplankton up to 100,000 individuals.m-3
have been reported (Talling & Rzoska 1967). Such factors as
current, high turbidity and low dissolved oxygen may act to
reduce the numbers of zooplankton (Welcomme 1979), in addition
to other environmental factors such as light, nutrients and
temperature which limit primary productivity and hence
zooplankton production (Winner 1975).
Hydrographic stability is also important in determining
.the development of plankton communities (Winner 1975). The
development of plankton communities in rivers as compared to
lakes, largely depends on the physical characteristics of
turbulent flow and the turbidity due to suspended particulate
loads such that only certain small, fast-growing phytoplankton
species may be suitably adapted to such conditions (Reynolds
1984). Furthermore, flow rates in tropical river~ such as
those of the study site are not seasonally consistent because
137
of the alternation of wet and dry seasons and this influences
the seasonal succession of plankton (Hutchinson 1967, Welcomme .
1979). The reduction in flow rates at certain times of the
year determines the 'agel or retention time of the water mass
such that flow loses impetus~ transparency increases and the
biological sequence of production of phyto- and zooplankton
occurs which in turn influences other stream biota (Winner
1975, Rzoska 1978).
The seasonal succession of phyto- and zooplankton at the
study sites was similar to that which has been described for
many large tropical rivers (summary in Welcomme 1979), that
is high plankton densities at times of low water and
decreases with the floods. During the floods, nutrient
dilution and turbidity may act to limit photosynthetic
activity and even in the dry season, phytoplankton growth
may occur to the point where nutrients become limiting (Prowse
& Talling 1958). Such seasonal succession may involve rapid
fluctuations in abundance of species (Reynolds 1984), for
example seasonal pulses of,rotifers~ copepods or cladocerans
(Welcomme 1979) or total populations of phytoplankton may
decrease to nil within a month or two (Holden & Green 1960).
L~v@que et aZ(1983) described the succession of phytoplankton
communities in the intermittent Bandama River in the Ivory
138
Coast as progressing from a diatom-dominated community
(principally Melosir2J TabellariaJ Synedra and Navicula)
during the main floods to euglenophytes and pyrrhophytes
(primarily Cryptomonas and PeridiniL@) just after the floods
and predominantly chlorophytes throughout the dry period with
a peak of blue-greens in the middle of the dry season. Holden
& Green (1960) described similarly well defined succession of
zooplankton communities in the Sokoto River, Nigeria with
rotifers and copepods increasing in densities in the dry
period and rhizopods and cladocerans becoming dominant with
the onset of the floods. The effects of floods may also
influence plankton composition and normally bottom-dwelling
species may occur in the plankton at these times (Holden &
Green 1960), for example high populations of ostracods
occurred in the plankton at Station 3 in August 1980. Survival
of planktonic organisms during flood periods may be accomplished
by a variety of strategies, for example reservoirs of popula-
tions may survive in small backwaters or upstream (Welcomme
1979). Also many species may have diapausing eggs or other
stages which survive in the bottom sediments (Moghraby 1977),
for example rhizopods and rotifers such as Conochilus
(Hutchinson 1967, Holden & Green 1960).
Brackish water plankton communities may be classified
139
into three major categories: autochthonous (permanent
residents); temporary authochthonous (introduced from outside
areas by water movements and capable of limited proliferation
but dependent on reinforcement from parent populations); and
allochthonous (recently introduced from freshwater or open
sea, have limited survival potential and are unable to
propagate) (Perkins 1974). Many planktonic taxa at Station
4 seem to be allochthonous or temporary autochthonous from
upstream during the rainy season, for example Synedr>a
3
Closterium and Gyrosigma, or marine, for example Coscinodiscus.
Dominant zooplankton taxa su~h as cladocerans and copepods
at the study site were characteristic of oligohaline estuaries
such as those of the Amazon and Maracaibo Lake (Rodriguez 1974).
The occurrence of diatom blooms at the beginning of the rainy
season at Station 4 is consistent with what is known about
increased diatom abundance in conditions of reduced salinity
accompanied by high inorganic and organic nutrient concentra-
tions (Perkins 1974). Increased populations of cladocerans
and cyclopoid cope pods at these times may be attributable to
increased food availability, that is diatoms and organic
debris respectively (Pennak 1978). Bacon (1968) also found
a similar trend of increased populations of both phyto- and
zooplankton during the rainy season in the Caroni Swamp.
140
The macrofaunal composition of the freshwater study
sites was generally comparable to that of lentic rather than
lotic freshwater communities described by Odum (1970) and
Maitland (1978). In particular, many of the aquatic insect
species found are characteristic of either littoral lentic
or depositional lotic environments (cf.Merritt & Cummins
1984). Overall community composition at the study site was
)1
different from that recorded for other local areas studied,
for example Maracas River (Thornhill et al 1969, Caesar 1985)
and the Arima River (Hynes 1971), both fast-flowing clear
water streams with groups such as simuliids, hydropsychids and
psephenids present. Substrate was a major influential factor
onthebenthos at the study sites with the occurrence of such
groups as tubificids, chironomids, prosobranch gastropods and
sphaeriid bivalves which are commonly found in the soft
substrates of slow-flowing silt-laden floodplain rivers
(Welcomme 1979). The study area also bore distinct similarities
in faunal composition to the periodically inundated forests
along the white and mixed white/black water rivers in the
Amazon (Irmler 1975, 1981). The range (0.05 - 4.2g.m-2) and
mean value (1.7g.m-2) of standing crop of benthic invertebrates
at the study sites coincided with the ranges recorded for
floodplain lagoons in Africa and Latin America, that is
from 0 to about 6 g.m-2, and mean values about 2 g.m-2
141
(Welcomme 1979). Irmler (1975) reported mean biomasses of
benthic organisms in inundation forests to be between 0.2
and 1.8 g.m -2 for white water forests and between 1.1 and
-2
7.4 g.m for mixed white/black water forests.
Seasonal variation in benthic macroinvertebrate
populations at the freshwater study sites was the result of
two major factors: washouts due to flood events in the rainy
season and stagnation and even total drying out during the
dry season. Floods are major catastrophic events in streams
in general (Hynes 1970, Uhlmann 1979, Fisher 1983) and their
effects on population sizes, community structure and life
history patterns of fauna have been documented in several
tropical streams (for example Petr 1970, Bishop 1973, J.D. Hynes
1975, Stout 1981, 1982, Leveque et al 1983). Consequences of
floods may be quite severe in some streams, for example
r
Siegfried & Knight (1977) reported that washouts in a Ip
California creek reduced benthic invertebrate populations by
more than 85% and benthic standing crop by more than 95%
resulting in decreased species diversity of the community.
Such factors as substrate type, topography, faunal composition,
degree of development of adaptations, and size of organisms
will determine the severity of scouring and washouts on
benthic fauna (Stehr & Branson 1938, Hynes 1970, Siegfried &
Knight 1977). In particular, lacustrine species are unable
142
to survive well in rivers which periodically flow swiftly
(Hynes 1970).
At the study site, densities of benthic macroinvertebrates
increased to high levels if low to moderate flow rates were
maintained and early in the dry season. However, long periods
of stagnation resulted in minimal densities of benthic
organisms. Recovery was rapid when favourable conditions
returned. Stagnation and drought are common occurrences of
floodplain rivers (Welcomme 1979) and intermittent streams
and rivers (Hynes 1970, Williams & Hynes 1977). Stagnation
may produce conditions unsuitable for rheophilous species
(Harrison 1966, Iversen et aZ 1978, Canton et aZ 1984);
induce stratification and produce unfavourable benthic
conditions (Irmler 1975, Williams & Hynes 1976); decrease
habitat size (Canton et aZ 1984); and concentrate prey and
predators in very small aquatic habitats (Lowe-McConnell
1975). Community composition may change, favouring certain
species capable of withstanding such conditions and belonging
to functional groups such as shredders and predators (Canton
et: al: 1984). However such conditions may be suitable for
the more lacustrine species for a number of reasons such as
increased food availability in the form of algal production
or accumulation of organic debris, or increased densities of
143
prey for predators, lack of spates and increased stability of
the substrate and water column, and higher water temperatures
encouraging faster rates of growth and reproduction (Abell
1959, Iversen et al. 1978, Bishop 1973, Extence 1981, Leveque
et al 1983, Canton et aZ 1984). In addition, migration of
individuals from drying areas of stream bed into refuge pools
or oviposition by adults into these last remaining aquatic
habitats may actually increase population densities in such
pools (Extence 1981). The above factors may have contributed
to the lack of clear seasonal trends in densities of benthic
organisms.at some stations.
With continued drought, standing water may disappear
completely and most aquatic species may die but water may
persist within the substrate allowing refuge for small
species capable of inhabiting these interstitial spaces
(Clifford 1966). Moist substrate and leaf litter also provide
suitable habitats for certain terrestrial species such as
oligochaetes, dipteran larvae, scavenging beetles and ants,
slugs and snails (Williams & Hynes 1976). This situation was
comparable to that seen at the study site during the dry
season.
With the onset of flow, recovery of the stream fauna
144
may be rapid, for example Harrison (1966) found that after a
period of drought in a South African stream, normal faunal
composition in pools was attained within a month after
resumption of flow and within two months in faster flowing
regions. Recolonisation may be by a number of routes, for
example downstream drift and upstream migration from refuge
pools, vertical migration from within the substrate and
aerial recolonisation via ovipositing adults (Williams 1977).
The relative importance of each route will vary depending on
the situation present. Williams (1977) found that vertical
migration from the substrate accounted for 95% of recolonisa-
tion in a Canadian site studied, while J.D. Hynes (1975)
suggested that most recolonisation in the Ghanian stream
studied was via aerial means.
From the above discussion it may be seen that the effects
of stagnation and drought on benthic macroinvertebrate
communities may range from being detrimental to advantageous.
The nature and magnitude of the effect of drought depend on
the degree and duration of drought, the season of drought
(for temperate situations drought coincident with adult
emergence and oviposition times or major larval growth periods
can be more detrimental than at other times), and the nature
of the species present (Iversen et: al: 1978). Clifford (1966)
145
summarised a classification of intermittent streams based on
their temporal flow characteristics, that is short flow or
long flow, and the effect of such hydrological regimes on
faunal composition. In addition, other factors such as the
nature of the stream bed, water table characteristics and
shading effects of vegetation influence such a classification.
For example, long flow streams which are reduced to isolated
pools during a regularly occurring dry period may support a
permanent and varied fauna and even a substantial fish
population. On the other hand, a stream subject to extensive
drought such that no standing water remains, in conjunction
with an impervious streambed subject to high temperatures and
~
exposure may not possess a permanent stream fauna but one
limited only to adventitious and migratory species, a
situation similar in many respects to a temporary stream. 11
I
The extent of drought and the nature of the streambed wi11 I
presumably also influence the possible routes for recolonisa- Iii
" f
tion.
'I
The composition of the fauna of the stream will also
determine the ability of the stream to recover, for example
Harrison (1966) noted that recolonisation in his study area
was rapid because the fauna was composed of species adapted
to conditions of regular drought unlike streams which are
146
subject to only occasional periods of drought such as the
We 1sh mountain stream studied by Hynes (1958) where many of
the organisms were specialised rhithronic species. Many
of the benthic and other macrofauna found at the study sites
belonged to groups reported to be able to withstand periodic
stagnation and even desiccation.
Adaptations of aquatic species to overcome or survive
periodic drought in temporary and intermittent streams have
been outl ined by Hynes (1970) and he classified species into
six main groups:
(i) Species which are able to tolerate extreme conditions
in pools such as high temperatures and low oxygen
concentrations, for example some fish and invertebrates
including flatworms, crustaceans, some Trichoptera
and snails. Some even take advantage of the periodical
lentic conditions in order to increase survival of
young.
(i i) Some species survive drying conditions by burrowing
down into the substratum, for example flatworms,
nematodes, oligochaetes, some amphipods and isopods,
crayfi sh , e 1mi nthi d beetl es, some Chironomi dae , some
caseless Trichoptera, snails and mites.
(i i t ) Organisms may produce eggs or nymphs capable of
147
surviving long periods of drought. Williams & Hynes
(1976) document the occurrence of drought-resistant
'stages in many organisms studied: cysts in tubificids,
copepods and triclads,and eggs of mayflies and
chironomids (cf Grant & Stewart 1980). Immature
forms of amphipods, ostracods and cyclopoid
copepods, larvae or pupae of stoneflies, some dipterans
and trichopterans, and adult enchytraeids, leeches,
hemipterans, coleopterans, gastropods and fish were
all found to be capable of surviving dry periods.
(iv) Species may reinvade from elsewhere, for example
fishes and some aquatic invertebrates may recolonise
the expanding habitat after flow resumes from
aquatic refuges or insects may fly in to oviposit as
soon as water reappears. The latter may be particularly
important in the tropics since habitat expansion and
oviposition by adult insects coincide with the
beginning of the rains.
(v) Some species may occupy pools or the damp river bed
only during the dry period, for example some
mosquitoes, Chironomidae, beetles and bugs.
(vi) A few species appear to be highly specialised
inhabitants of temporary waters such as a few snails
and some Trichoptera which are capable of aestivation.
148
Irmler (1975, 1981) documented the range of adaptations
seen in some species of the Amazonian inundation forests
such as migration in the more mobile species, for example
large crustacea; periods of inactivity or dormancy during
the dry period in ampullariid snails and sphaeriid bivalves;
increased parental care such as brooding in sphaeriid
bivalves and the crab TrichodactyZus (belonging to the same
family as DiZocarcinus dentatus (Smalley & Rodriguez 1972));
and high reproductive potential such as seen in the sphaeriid
bivalve Eupera simoni which matures rapidly and produces two
generations per year. Sphaeriid bivalves have also been
shown to have flexible life history traits to adapt to
temporary waters in temperate regions (Hornbachet al.1980,
Way et aZ1980). Irmler (1981) also pointed out that many
species found in inundation forests are preadapted to such
environments and are closely related to species which can
survive alternating inundation and drought. In particular,
many congeners of Pomacea glauca are found in such habitats
(Pain 1950, 1960). P. urceus has been studied extensively for
its ability to survive under floodplain savanna conditions
in Venezuela (Burky et at 1972, Burky 1973, 1974,Burky &
Burky 1977) and in Trinidad swamps (Lum Kong 1986). P. paludoea
is a common inhabitant of the Florida Everglades (Perry 1973,
Hurdle 1974) and P. Zineata is successful in the Amazonian
149
inundation forest (Irmler 1981). Although the ability to
aestivate and withstand droughts varies between species, other
adaptations such as rapid sexual maturation and high
fecunditi es compensate for lowered survival abilities (Innler
1981).
The other large benthic organism studied here,
Dilocarcinus dentatus, showed a life cycle consistent with
that recorded by Stonley (1971) who noted that this species
brooded eggs and juveniles from the beginning of the dry
season and released young in June and July. She suggested
an annual life cycle with adults dying off at the time of
release of juveniles as also indicated in the present study.
Brooding of eggs and juveniles is also seen in the only
other local freshwater crab Pseudothelphusa garmani (Stonley
1971) and is commonly found in freshwater crabs in order to
increase juvenile survival rates in such unstable environments
(Vernberg & Vernberg 1983).
In many small streams with heavily vegetated banks such
as those of the study site and the narrow upper reaches of
rivers, input of material from the surrounding terrestrial
ecosystem can playa major role in terms of providing food
material and nutrients to the aquatic ecosystem (Cummins
1974, H .B..N Hynes 1975). Subsequentl y, stream faunas play
150
a very important role in terms of importing, producing,
processing and storing such organic matter (Cummins et aZ
1983) and such a role may be essential for the retention of
nutrients within tropical forest ecosystems (Dudgeon 1983).
In the tropics leaf fall into aquatic habitats has been
estimated as being up to 6 t.ha-1.y-1 (Welcomme 1985) and
Geisler et aZ (1975) counted 56 individual pieces of material
fall ing onto a 0.25 m2 glue board over water during a 24-hour
period. This material included mainly Culicidae and
Chironomidae with lesser numbers of ants, Lepidoptera,
spiders and plant seeds.
A higher abundance of insects fall ing onto glue boards
at Station 2 in the rainy season as compared to the dry was
consistent with the general trends quoted widely in the
literature with regards to increased abundance of terrestrial
insects at this time in the tropics. Such an increase may
be due to emergence of aquatic insects (Bright 1982),
increased abundance of herbivores at this time of new leaf
emergence (Wolda 1978), or decreased numbers due to harsh
conditions in the dry season (Janzen 1973). Nevertheless,
work in some tropical areas has indicated that insect numbers
may in fact show a reverse trend (Janzen 1973, Wolda 1978),
an unclear periodicity (Wolda & Flowers 1985), or be dependent
151
on the extent of seasonality of the habitat (Wolda &
Broadhead 1985). In addition to the fact that there may be
greater productivity of terrestrial insects during the rainy
season, many insects move into moist refugia such as streams
in the dry season (Janzen 1973) thus increasing in local
abundance at these sites. Such factors may have contributed
to the unclear trends seen in the data presented.
In summary, the general trend of seasonal variation in
the streams studied showed the influence of a main dry season
from January to May and a main rainy season from June to
December with a short dry period falling between September
and November. Lentic conditions often developed during the dry ~
I
periods but did not appear to last longer than four months
continuously. Drought conditions experienced were never
severe enough to dry out the entire length or even the
majority of the watercourses studied. Therefore, these streams
may be considered to be long flow intermittent streams with
substantial refuge areas and thus capable of maintaining a
diverse and permanent faunal assemblage (Clifford 1966).
A general seasonal trend as outlined above was not
consistent however, being highly variable from year to year
,"nt erms 0f t ime of onset ' duration and intensity of any
152
season. Lentic conditions often developed more than once
each year in the same areas and not necessarily during the
main dry season. This is unlike clear seasonal trends in
other tropical intermittent streams where one definitive
lentic period existed in each year (Adebisi 1981 a, Leveque
et aZ 1983).
Seasonality of flow was a major influence on the nature
of the stream environment (both physical and chemical) and
consequently on the nature and seasonal variation of the
biota. However, variability in the duration and intensity
of each season made it diffi~ult to predict the nature of
biotic changes. Generally however, a dense eutrophic
planktonic community developed in some areas during prolonged
lentic conditions and was decreased drastically with the
onset of floods. Benthic standing crop and allochthonous
input to the stream appeared to vary equally widely in dry
and rainy seasons.
153
THE FISH
INTRODUCTION
The Neotropical freshwater fish fauna is the richest in
the world with probably over 2000 species inhabiting the
tropical regions. The greatest diversity appears to be in
the Amazon basin with diversity decreasing outward to the
Guianas and Venezuela in the north, the Parana system in the
south and westward to Colombia and Ecuador (Gery 1969, Lowe-
McConnell & Howes 1981). The large number of species belong
to relatively few basic groups, in particular the characoids
and siluroids are dominant owing to the extensive adaptive
,,
radiation within these orders (Lowe-McConnell 1975). Cichlids ,, '
are well represented although not as well as in Africa (Gery
1969) and the cyprinodonts are also quite important (UNESCO/
UNEP/FAO 1978).
In contrast to the mainland, there are no primary
freshwater fishes recorded for the West Indies north of
. Trinidad (Myers 1938) there being instead only secondary
freshwater fishes belonging primarily to four families
(Lepisosteidae, Cyprinodontidae, Poeciliidae and Cichlidae)
of which the Poeciliidae is the most important (Miller
154
1982). In the Lesser Antilles, it appears that only four
species of secondary freshwater fishes occur, that is
RivuZus hartii~ R. marmoratus~ PoeciZia vivipara and P.
reticuZata; the remainder of the fauna being comprised of
marine invaders to freshwater (Miller 1982).
The diversity and distribution of the freshwater fish
fauna of the island of Trinidad reflects its intermediate
position between the mainland to the south and the West
Indian islands to the north. Of a total of 76 freshwater
fish species recorded for Trinidad by Boeseman (1960, 1964)
only 37 are genuine freshwater species of which 32 also
occur on the South American mainland. The few species which
are restricted to Trinidad are relatively recent modifications
of closely related continental forms (Boeseman 1960). In
addition, no characoids, cichlids or catfishes have been
recorded for rivers draining the north slopes of the Northern
Range, instead mainly a euryhaline fish fauna similar to
that of the Antilles is found in this region (Price 1955).
South of the Northern Range, the island supports a fairly
diverse assemblage of primary freshwater fishes comprised
predominantly of characoids, a feature which characterises
the islandls continental origins (Kenny & Bacon 1981).
155
Studies on the freshwater fish of Trinidad have been
largely taxonomic in nature. Boeseman (1960, 1964) gave the
most recent listing of species and he also included a
comprehens ive hi stori cal rev iew of past studies. Some of
the earlier works included information on distribution,
habitats and general biology, for example Regan (1906),
Guppy (1934, 1936) and Price (1955). More recently, a
detailed biological study was conducted on a species of
commercial importance, Hoplosternum littorale (Singh 1978).
This is one of the three principal freshwater species
exploited commercially in the island (Kenny & Bacon 1981).
The quppy , PoeciZia ret.iculatcc, is the only other species
which has been studied locally but only selected aspects of
its biology have been looked at. For example anti predator
adaptations in behaviour (Seghers 1973), size (Liley &
Seghers 1975), colour (Endler 1978) and life history
evol~ttdn'- (Reznick 1980, Reznick & Endler 1982) have been
thoroughly investigated.
Despite the paucity of biological or ecological studies
conducted locally, a certain amount of information is availa-
ble on the biology of the same or closely related species on
the mainland. Reviews of this literature were given in Lowe-
McConnell (1975) and Welcomme (1979, 1985). The general
156
biology of many small species has been described in the
aquarists' literature, for example~Tropical Fish Hobbyist
series, Frey (1961) and Breder & Rosen (1966) among others.
Generally,ecological studies of local and Neotropical fresh-
water fishes are lacking. Even in the West Indies, the only
studies known are of a poeciliid, Limia vittata in Cuba
(Barus et al 1980) and a recent study of Sicydium pZumieri
in Puerto Rico (Erdman 1982). In 1978 a UNESCO/UNEP/FAO review
cited only four ecological studies in the Neotropical region
up to that time. This situation has been improving recently
(cf Zaret 1984) but much remains to be done on fish community
organisation, feeding, fish effects on their food resources
and seasonality effects on breeding (UNESCO/UNEP/FAO 1978).
The species selected for this study were the six most
common fishes in the study site. They were Gasteropelecus
sternicla (Linnaeus 1758), Corynopoma riisei Gill 1858,
Astyanax bimaculatus (Linnaeus 1758), Hemigrammus unilineatus
(Gill 1858), Corydoras aeneus (Gill 1858) and Poecilia
reticulata Peters 1859. With the exception of G. sternicla
these species have been reported to be common throughout most
of Trinidad south of the Northern Range (Regan 1906, Guppy
1934, 1936, Price 1955, Boeseman 1960, Nelson 1964) and
P. ret~.cu lata has also been found north of the Northern Range
157
(Price 1955). G. sternicla was reported by Price (1955) to
be restricted to the southwestern peninsula as it is
'believed to have invaded the island from Venezuela fairly
recently' (p.lO). Liley & Seghers (1975) also noted in their
collections in the region of the Northern Range that while
P. reticulata occurred widely from uppermost springs to
lowland rivers, c. riisei~ A. bimaculatus~ H. unilineatus
and c. aeneus were midstream or lowland river species. The
widespread distribution of most of the species and their
popularity in the aquarium trade makes them attractive for
local study.
Another reason for the choice of the six species was
the range of morphology and habits exhibited. They vary in
phylogeny, size, feeding requirements, microhabitats and
general reproductive biology; factors which should affect
their responses to environmental conditions in the study
site. Details of their biology in the wild are scattered
in the literature and are minimal for many species although
some information exists for their habits under aquarium
conditions.
The following objectives were therefore set for this
part of the study:
158
(1) To determine the diversity of fish species in the
study site.
(2) To monitor the population demography and dynamics of
the six species, in particular population structure,
growth rates, life spans and population fluctuations.
(3) To investigate the reproductive strategies of the
six species including timing and duration of breeding
activity, size at maturation, fecundity and spawning
patterns and general breeding behaviour.
(4) To determine the effects of environmental seasonality
on the population biology and reproductive strategies
of the six species.
REVIEW OF THE SPECIES
The six species selected for the study belonged to
three orders (Characiformes, Siluriformes and Cyprinodonti-
formes) and four families (Gasteropelecidae, Characidae,
Callichthyidae and Poeciliidae) (Table 17). They are
illustrated in Plate 6.
GasteropeZecus sternicla (Linnaeus 1758)
The common hatchetfish is a small uniquely shaped
0
V > V >
V > C ' > V > V >
- "
~
c : e
" 0 R > < : I » " ' I c :
c . . " t : l
R > ~ t T I »
t T
c r
~ .
< : . . . . R > e n
n 0 R >
0 -
. . . . . . . . .
e n
" ' I e n
- s
" ' I
R > 0 - 0
<
' <
' "
~ .
V I R >
" ' I
. . . . .
e n
" ' I c . .
/ I )
~
' <
" ' I 0
: : : J
( " " ) ( " " )
o o
- l
: : : r : : : r I I I
r o e n
. . . . .
. . .
< +
I » ' "
' " ' l " ' I I I I
r o r o
" ' I
1 » - - ' "
. . . .
n n
e n
n
o o
: : : r
< +
. . . . .
" t : l
r o
< +
0 -
: : : r
. ; :
r o
I I I
~
e n
~
e n
' "
( " " )
e n
~ . & ~
. . . . .
: : : r
I »
l ; ; ' . : : § ~
~
" ' I
\ I ) - - ~ - Q ,
I »
n
. . . o < :
n
~ . . . . . . . . . .
< +
' "
0 -
~ g
o
r o
~
' "
A ' "
< :
c ,
o
. . . .
~
< +
R >
: : : r
' "
R >
e n
x
- l
~
R >
e n
< +
: : : r
" ' I
e n
I Q
' "
" 0
o
R >
~
n
o
. . . .
" 0
R >
< +
I I I
R >
I I I
J .
< +
~
< :
0 -
R >
' "
r o
c . .
§ ~
. . . ~
N . . .
~ . - ~
. . . . .
~
\ 0
~ I
. . . . . .
. . . . . .
-
' ~"
c . .
: z
V >
~ & ( " " )
r o
. . . . .
e n
~ ~ ~ c :
o
~ - - ! } - - - - ; : ; -. - - - J .
~
~ ; ; : t i '
~
\ 0
C O
~ ~
0 - I I I
" " "
-
R >
' "
-
6 S 1
160
\
(~
I
(d)
1
(f)
/"111"1'1111'1111111"'11"11"111"1'11"'1""1""1""1""111111""1""1"1'111111'1'11"1'1"1'1""1'11'1""11"'1'11'1"'11'111111"11'111'1"111'11,
o 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
GERMANY .STAfDllH
PLATE 6: The six fish species studied: (a) G. et erni.cl.a,
(b) C. riisei (male above, female below),
(c) A. bimaculatusJ (d) H. unilineatusJ
(e) C. aeneus, and (f) P. reticulata (female
above, male below).
161
characoid attaining total lengths up to 7 cm (Paysan 1975).
It is predominantly surface-dwelling and has, along with
other members of the family, an abil ity to jump and skim
along the surface of the water. For thi s purpose it uses
its well developed keel for stability and attachment of
enlarged pectoral muscles and its elongated and curved
pectoral fins are rapidly vibrated to create the necessary
forces (Ridewood 1913, Frey 1961, Weitzman 1954,1958). This
is presumably an adaptation to escape underwater predators
(Weitzman 1958). In addition to extensive morphological
modifications for 'flight', the almost vertical orientation
of the jaws allows for grasping organisms at the surface of
the water (Weitzman 1954). While the Brazilian species,
G. Leirie, and the related CarnegieUa strigata have been
bred in captivity (Stoye 1935, Stanley 1956) it appears that
G.sternicla has not (Paysan 1975). Nevertheless this species
is the commonest Gasteropelecus species in the aquarium
trade although almost nothing is known of its biology (Gery
1977). Its geographical range includes the Guianas and
Trinidad; Mato Grosso, Brazil; and the Peruvian Amazon
(Gery 1977).
Corynopomariisei Gill 1858
The sword-tailed characin is a small slender fish
162
growi ng up to 6 cm in total 1ength. It belongs to the sub-
family Glandulocaudinae which is an artificial assemblage of
species based on the presence of a caudal gland in males
(Gery 1977). Sexual dimorphism, fairly complex courtship
activities and internal fertilisation occur in many species
of this subfamily (Nelson 1964). Dimorphism is extreme in
C. riisei to the extent that the species has been described
under three different genera on the basis of degree of
development of males (Regan 1906, Guppy 1936). Males possess
a pouch-like structure at the base of the caudal fin which is
postulated to secrete a pheromone to attract females. In
addition, a pair of paddle-shaped opercular extensions on
the male are also used during complex courtship activities
(Nelson 1964). The unpaired fins of mature mates are
significantly longer than in females and the anteriormost
anal fin rays bear minute hooks which are necessary for
sperm transfer to the female (Nelson 1964, Breder & Rosen
1966). The sperm transferred to the female are not encapsul-
ated in a spermatheca but are held together by a viscid
substance; they may remain viable for as long as 10 months
or the female's lifetime (Sterba 1962, Breder & Rosen 1966).
Newly laid eggs show between two and 64 blastomeres and
they are laid in the early morning onto any suitably secure
objects (Nelson 1964, Breder & Rosen 1966). Breeding habits
163
and courtship behaviour in captivity were described in detail
by Frey (1961), Nelson (1964, 1965) and Breder & Rosen (1966).
Development of juveniles and their behaviour was described by
Nelson (1964) who noted sexual differentiation at the age of
17 weeks in captivity. He also found that aggregation occurred
in juveniles and sporadically in adults, and dominance
hierarchies developed in groups of males.
C. riisei is generally a surface-feeding omnivore (Frey
1961, Nelson 1964). Its geographical range includes Trinidad,
Venezuela and Colombia (Boeseman 1960) but the Trinidad
form may be different from that of the mainland (Gery 1977).
Astyanax bimaculatus (Linnaeus 1758)
Locally called the pink-finned sardine (Boeseman 1960),
this is a fairly large characid growing up to 15 cm long
which is commonly found in large schools in the middle and
lower layers of the water column (Paysan 1975). It has a
generalised body shape which is moderately elongate with
dorsal and ventral profiles symmetrical (Gery 1977). This
species is widely distributed from La Plata to Panama and
consists of several subspecies (Gery 1977). Studies of
natural populations of this and related species have been
conducted in north-east and south-east Brazil (Azevedo &
164
Vieira 1938, Nomura 1975 a,b,c). These studies show that the
species is omnivorous including as much plant as animal
material in their diet (Azevedo & Vieira 1938, Nomura 1975 b).
Reproduction is seasonal coinciding with the onset of the
rains and females may spawn several times during the breeding
period ( Azevedo & Vieira 1938, Pel izaro et al: 1981). Sexual
dimorphism is exhibited: males are considerably smaller than
females and small hooks are borne on the anteriormost anal fin
rays of males (Azevedo & Vieira 1938, Breder & Rosen 1966).
It is unclear whether this species has been bred in captivity.
Guppy (1936) described the species as being hardy and prolific
and fry may be raised on prepared foods. Azevedo & Vieira
(1938) reported breeding 'p iabas ' (tetragonopterines including
A. bimaculatus) in captivity by inducing spawning with
injections of hypophysial extract but it was not stated which
species were bred. On the other hand, Sterba (1962) and
Paysan (1975) reported that this species had not yet been
bred in captivity. Larval development of 'piabas' was dealt
with by Azevedo & Vieira (1938).
Hemigrammus unilineatus (Gill 1858)
The feather-fin is a small characid attaining maximum
lengths up to 4 cm. It is similar to A. bimaculatus in
posSessln.g a general ised body shape and the genus is believed
165
to have evolved from some more conservative Astyanax- or
Moenkhausia-like tetra by the process of paedomorphosis (Gery
1977). H. uniZineatus is found distributed throughout
Trinidad, the Guianas and the Amazon (Gery 1977).
The species is well known by aquarists although it is
inconspicuous and therefore not as popular as its more
brightly coloured congeners. Azevedo & Vieira (1938) included
the species in their discussion of tetragonopterines in
north-east Brazil but specific comments on its biology were
not made. Otherwise almost nothing is known of the habits of
this species in the wild. In aquaria, individuals remain in
well defined areas of the tank and are omnivorous (Paysan
1975). Males are generally smaller and slimmer than females
and they have been bred productively in captivity (Frey
1961). Courtship is simple and the externally fertilised
eggs appear to be spawned at the water's surface (Stolzenhain
1927).
Corydoras aeneus (Gill 1858)
Locally called the pui-pui or goldfish, this small
armoured catfish is a popular aquarium species which is used
as a bottom scavenger. For the past three decades there has
been an expo rt trade of this species and Hypostomus robinii
166
to the temperate countries with C. aeneus accounting for up
to 56% of the tr~de in 1967 when 792,783 individuals were
exported (Fisheries Division 1976).
C. aeneus is relatively small, attaining maximum lengths
up to 7.5 cm (Paysan 1975). Unlike other catfish, it has
large eyes and short barbels and is diurnal in habits
(Alexander 1965). Like several other callichthyid catfishes,
C. aeneus is an air-breather, gulping air at the surface,
using the posterior intestine as a site of gas exchange and
releasing used air from the anus (Kramer & McClure 1980).
Under conditions with high oxygen levels, C. aeneusis not
a obligatory air-breather (Kramer & McClure 1980) presumably
because of the increased energetic costs and increased risk
of predation associated with air-breathing (Kramer & McClure
1981, Kramer 1984).
C. aeneus is commonly found in schools and they are
bottom-dwelling feeding on detritus and its component fauna
(Frey 1961, Paysan 1975). This species has been bred in
captivity but the actual mechanism of fertilisation of the
eggs is still unclear (Breder & Rosen 1966, Zukal 1982).
Balon (1984, 1985) classified this species as an external
egg-bearer and more specifically, a transfer brooder in which
167
the eggs are carried for some time before deposition but
after deposition it is similar to a non-guarding phytophil in
behaviour. Generally females are distinguished from males
by being larger with a wider body and by possessing more
rounded dorsal and pelvic fins (Schofield 1957). Unlike
its local relatives Hoplosternum littorale and Callichthys
callichthys~ C. aeneus does not build foam nests (Breder &
Rosen 1966). Instead the female lays up to five eggs at a
time into a pocket formed by the pelvic fins pressed
together and sticks these onto the previously cleaned surface
of the aquarium glass or plants (Frey 1961 , Breder & Rosen
1966, Zukal 1982). Several males are required to fertilise
all the egg~ spawned by a female (Frey 1961, Zukal 1982).
Development of the eggs and fry in captivity has been
described by Adams (1946), Hart (1947), Schofield (1957)
and Zukal (1982) with maturation occurring not before two
years (Schofield 1957).
The geographic range of the species is large extending
from La Plata to Venezuela and Trinidad and westwards to
Bolivia and Peru (Boeseman 1960, Frey 1961, Paysan 1975).
Poecilia ret~.cu lata Peters 1859
The guppy ,·s the most common freshwater fish in Trinidad
168
(Boeseman 1960, Nelson 1964). It is very well known in the
aquarium trade since it is easily bred in captivity and new
exotic varieties have been selected. It has also been
introduced into many tropical areas for mosquito control
although its original area of distribution appears to have
been rather small, that is Venezuela, the Guianas, Trinidad
and Barbados (Boeseman 1960). On account of its ease of
maintenance and breeding and short generation time, the
guppy has been used extensively in a variety of biological
studies especially laboratory-oriented ones. Despite this
however,comparatively few studies have been undertaken on its
biology under natural conditions in the tropics and almost
nothing is known of its population dynamics and reproductive
biology in nature (Kramer, pers. comm.).
The guppy is a small schooling fish attaining total lengths
up to 5 cm for large laboratory-reared females (Frey 1961).
Sexual dimorphism is clearly evident (Endler 1984). Males
are small, slender and brightly but highly variably coloured
and the anal fin rays (3-5) are modified to form an intro-
mittent organ, the gonopodium; females are larger, more
rounded in shape and are uniformly gray except for a dark
spot just above the anal fin. Male coloration and intense
courtship activities may play important roles in determining
169
mating Success (Farr 1980, Endler 1980, 1984, Kodric-Brown
1985) in addition to other factors such as the composition
of the mal e group (Farr 1976, 1977) and sperm competition
(Hilderman & Wagner 1954). The breeding system of guppies
can be defi ned as afema 1e-based polygyny where males
compete amongst themselves for access to females and their
sole parental contribution to offspring is in the form of
gametes (Kodric-Brown 1985). The courtship display has been
thoroughly described by Baerends et al: (I955) and Liley
(1966). Eggs are fertilised internally and develop within
the ovary (intrafollicular gestation) but maternal-embryo
exchange is limited to minerals and respiratory gases.
Nutrients are supplied entirely by the stored yolk of the
egg, that is they are strict lecithotrophes (Thibault &
Schultz 1978, Wourms 1981). Gestation 1asts between three
to five weeks to produce up to 80 actively swimmi ng young
each up to 6.5 mm in length depending on the size of the
female (Frey 1961, Thibault & Schultz 1978).
Wild-caught guppies from Trinidad have been shown to
be benthic feeders, taking algae, organic debris and benthic
invertebrates (Dussault & Kramer 1981) and the flattened
head and upturned mouth are not feeding adaptations but are
related to aquatic surface respiration in hypoxic habitats
170
(Lewis 1970, Kramer & Mehegan 1981).
METHODS
General sampling and morphometric studies
Once each month duri ng the study period May 1980 to
June 1981, each of Stations 1 to 4 was fished using a 3 mm-
mesh two-man push-seine for approximately the same distance
each time. Distances seined were 10 m at Station 1, 25 m
at Station 2, 15 m at Station 3 and 25 m at Station 4. During
the dry seasons when Stations 1 to 3 became isolated pools,
the entire pool was seined. Only one pull of the seine was
made each month in order to standardise fishing effort on
each occasion. Accepted catch~effort methods (Youngs &
Robson 1978) could not be utilised because of the relatively
small fish populations in the stream and the distinct,
possibility of overfishing during the dry season when fish
were concentrated into pools. All fish caught were collected
and immediately preserved in 4% formaldehyde solution.
During the period July 1981 to August 1982, Stations 1
and 2 were seined repeatedly to collect at least 20 mature-
171
sized individuals of each of the six species under study.
This was not always possible even after seineing upstream
and downstream of the stations or at other sites, for example
Station 3. All juveniles were preserved and retained when
caught.
In the laboratory, preserved fish from each monthly
sample were identified and counted. For each of the six
species under study, standard lengths (SL) were measured
to the nearest millimetre from the most anterior extremity
to the hidden base of~caudal fin rays (Lagler 1978), and fish
were weighed to the nearest milligram after blotting excess
moisture (total weight, TW). They were dissected to remove
the gonads which were weighed to the nearest milligram and
stored in 70% alcohol for further analysis. Specimens were
sexed on the basis of gonadal and secondary sexual
characteristics when mature or developing; small individuals
were classed as juveniles when they could not be definitively
assigned to one or either sex. Data from all freshwater
collections were pooled in most analyses because of the
relatively small numbers of fish caught at each station.
Certain characters were noted for G. stemicZa (greatest
body depth:SL ratio), H. uniZineatus (greatest body depth:SL
172
ratio) and c. aeneus (body width at pectoral fins:SL ratio
and pectoral fin spine length:SL ratio) in order to test
their reliability for use in differentiating the sexes of
adult specimens. Ratios for males and females were compared
using a d-test after arcsine transformation of data where
necessary (Sokal & Rohlf 1981).
Length-weight relationships were determined for each
species using data from all individuals caught during the
first year of the study. The following equation was used:
Studies were made to ascertain the effects of formalin
fixation on the 1engths and wei ghts of the fish as well as on
the length-weight relationship. Collections of each species
were made in June/July 1985 and fish were killed by chilling
in water for not more than 30 minutes for the characoids and
173
P. reticuZata and up to one hour for C. aeneU8. Observations
showed that these times were just sufficient to stop all
voluntary and respiratory movements. Fish were retained in
water until measurements were made (up to two or three hours
maximum). Standard lengths and total weights were measured
to the nearest 0.1 mm and 1 mg respectively. Fish were then
fixed and preserved individually in a buffered 4% formaldehyde/
freshwater solution and after 10 months they were again
measured and weighed. Changes in length and weight for each
individual and the average for each species were calculated
and expressed as preserved measurements as a percentage of
fresh measurements. Lengths and weights before and after
preservation were tested for significant differences using
a paired-sample t-test (Parker 1979). Length-weight
relationships were determined for each species for fresh and
preserved specimens separately and the regression coefficients
were compared after Bailey (1981, Appendix 6).
Population studies and reproductive seasonality
For each species length-frequency histograms for all
individuals collected over the study period were drawn noting
the number of individuals of each length at varying gonadal
development stages (Table 18 and below). From these, the
c . . .
; 0 o
c
r o
r o
<
\ 0
<
; 0 . . ,
c . . . C l l
r o
: : : J
r o o
V 1 o
V 1
' "
: : : J : : : J
< Q
\ 0
. . , V I - l G )
- l
- l
~ . g
o 3 r o r o : : : J 0
r o
Q I
c : : 0 V 1
V I V 1 0 . : : : J
Q I
: : : J 0 . . . . . 0
r t r t
C r o V I ~
0 . r t
r o
V 1
r o : : r V 1 V I V I . . . . V I
r o
o ,
0 .
o
V I <
: : : J <
Q I - . ,
c : : r o < 0 r o
U
0 0
-
c . : : 1
U V 1
~ ~
r o
V 1
Q I 3
: : r t i l
n C T
" "
. . ,
. . , ~ .
r o Q I 3
Q I
V I ( ; )
C T Q I
o
: : : J Q I ~
~ . - S
t u
n . . ,
r o ~
0 .
x 0
r o
V I
. . . . , V I
V 1
Q I
o
. . ,
~ .
r t
: : : J
N
r o
0 .
V I 3
r o
: : : J : : r 0
o
o
- S
. . . .
- S
r - t -
V I ' "
o
' " : : r
r t
r o 0
<
3 : n ~
- S
t u
V I
o
C T
: : : J r o < Q
c
V I
V I
C T
. . . .
n
r o V I
' "
: E :
- s
r - t -
" "
" "
~ .
r o
c : :
. . . .
t i l
: : : J
0 . n
z r
r t
C l l : : r
X
V 1
r o " "
r o
0 . - S
: : r
t i l
. . ,
V 1
n
" "
c
: E : c
r - t -
0 -
: : r . . ,
: : : J
r o
A " - . ,
V I
- S
" "
r o
~ .
r o
r t Q I " "
: : : J
r o n r t
0 . V I
< , r o r o
e - r -
. 0
' " - S
: : : J n
: : : J
V I
. . . . ,
o c : :
" ' "
- S V I
3 r o
" " 0 .
: : : J
I
. . . . , r t
~ . : : r
x
r o
r o
0 . n
- . ,0
3 0
r o
r o <
~ . ~ ~
< Q
" " <
: E : Q I
' < Q I
r t < t u < Q < Q " "
o V I
V I
: : l V I
: : : f ' r " t Q
r o . : : 1
o - ~
l O
r o
: : : J ' <
: : : J
0 . - . ,
l O V 1 " "
o
V I
o
V I ~ .
V 1 3
r t V I n
3 Q I
r t
3
n
Q I
: : r 3 n
: : : J
" " 0 . . ,
U ~
- - ' . ~ c
~
r t - a l O
<
. . ,
" " r t
" "
c : : Q I r o
n
r o •
: : : ~
- S . 0 . . ,
o
V I
o
V 1
~ .
r o r o c •
3
: : l
- S
: : : J
C l l
o
~ ~
~ g
0 - r t < Q
o 0
c : :
" "
o
. . . . Q I
r o c : :
o n
. . . . ,
- S r o n
C T
- S o
n n
3
c : :
- S r o
o
' < < 0 C
" "
f {
3
r t ' "
- s . . . . : : r - a
' "
o ' <
C l >
: : : J V I . . 0
C l l
C l l . . . . ' <
r +
" "
< Q
c : :
: : : J
c : :
. . . .
N
r o r t
V 1 ' < : : : J
< Q
. . ,
- S
o
r o
: : r
C l l l O
Q I
~
<
N
0 . " " r +
: : : J
r o : : : J ~ 0
C l l
r o
V 1 ' <
r o
o
: : : J
" "
V I
. . ,
o
. . . .
r o
: E : V I
V I
C
- S
: E : I
' "
r +
n
o : : : J
: : r
' <
r t
" "
C l >
r o r o
r o I < Q r o Q I
: : r
< 0
: : : J
~ . . . .
3
C l l l O
f 1 )
. . . .
0 -
- l : : r
l O
r t
: : : J
" "
V I
o
V 1
- s
. . ,
o
~ 0 -
: E :
" "
. . . .
o
. . ,
- . ,
o
0 .
< 0
: : : J
: : : J
" "
Q I
o
V 1
r o
V 1
V I
r o . . ,
: : : J
o
V 1 . . . .
I
l O
. . . . . . - S C T
: E : c : :
: : r
. . . .
r o
C T 0
n
< Q " " 0
: : r
. . . . . . .
: : : J 0 .
I
r o
. . ,
f 1 )
" "
. . . . ,
< . 0 ' < C l l . . . .
: : : J
r o
r o
: : r
r t
~
n
C T
0 -
C . . ,
o
' <
o " "
- S <
. . . . ~
r o
0 .
: : : J
V 1
' <
o
- S . . . .
: : : J
r o ' <
o C T
o
n
o
0 .
. . .0 .
C l >
: E :
. . ,
o
~
. . . . . . .
<
. . ,
' <
r t
' <
P L T
175
TABLE 19: Mature oocyte size ranges used in classifying
female reproductive state and minimum oocyte
sizes used in batch fecundity estimates for
the six fish species studied.
Ma ture oocyte Minimum oocyte
Species size range (mm) size (mm)
G. sternicla 0.7 - 0.9 0.559
C. riisei 0.6 0.9 0.486
A. bi.macul.at.ue 0.6 - 0.9 0.462
H. uni lineatus 0.6 - 0.95 0.486
C. aeneus 1.5 - 1.75 1.142
P. reticulata 1.2 - 2.0 1.20
176
minimum lengths at which sexual maturation took place were
determined. Minimum lengths at which developing and mature
gonads were observed (Minimum Developing SL and Minimum
Mature SL respectively) were noted for each species. The
length at which 50% of the individuals therein were mature
was also noted for each species (Median Length at first
maturity, Cambray & Bruton 1984). This had to be determined
graphically in some instances.
Population structure was determined monthly by the use
of length-frequency histograms based on data from all stations
combined. In species where data were sufficient, growth
curves were drawn using moving modes in the population
structure for particularly strong and temporally distinct
classes of juveniles (cohort analysis, Bagenal & Tesch 1978).
In order to relate growth with age, it was assumed that
reproductive maturity of these species occurred within the
first year of life, a realistic assumption considering the
changes in population structure over the period of study.
It was also assumed that average size at maturity could be
equated to the Median Length at first maturity. Once this
length was assumed to be attained at 12 months of age,
backward and forward projections using moving modes for
different cohorts were made at one month intervals. Minimum
170
age at maturation and lifespans were estimated from these
growth curves assuming equal growth rates for males and
females throughout. For species where data were sufficient,
the parameters of the von Bertalanffy growth equation were
determined using a Walford plot, plot of In(L - It) against
oo
t, and substitution methods (Bagenal & Tesch 1978).
In order to determine the reproductive state of
individuals of the six species under study, a number of
characteristics were utilised. Since multiple spawning
fishes may spawn a few eggs at a time for prolonged periods,
gonad weight and size or gonadosomatic indices (GSI) alone
are not reliable indicators of breeding activity. Females
may have a few mature eggs remaining in a small ovary yet
still be in breeding condition (Kramer 1978a). Therefore
other characteristics such as mature oocyte diameter,
secondary sexual characteristics and a limited amount of
histology were utilised in addition to GSI's and gross morpho-
logical features.
Gross morphological characteristics of gonads and mature
oocyte diameters used in assessing the reproductive state of
the fish are summarised in Tables 18 and 19. Mature oocyte
diameters were determined for each species on the basis of
178
this size range of oocytes being the maximum observed in
ovaries filling one-half to two~thirds of the body cavity.
Preliminary histological studies on the ovaries of all species
(except P. reticulata) were made to corroborate gross
characteristics used. Secondary sexual characters were noted
in c. riisei~ A. bimaculatus and P. reticulata to assist in
determining the degree of sexual maturation.
In reproductive analyses, only individuals of developing
or mature sizes were utilised in order to eliminate juveniles
and pre-reproductives from these data. Minimum Developing
SL was used in the analysis of monthly variation of gonadal
maturity stages for each species in order to include small
first-time developing individuals in the data. However for
other calculations (GSI,.monthly sex ratios and condition
factors) only individuals of Minimum Mature SL and above
were included. Where Minimum Mature SL differed for the
sexes, the smaller of the 'two was used. A minimum of 10 mm
SL was used for P. reticulata because of the difficulties
in sexing and weighing fish smaller than this length.
Gonadosomatic indices were calculated according to the
following:
GSI - weight of both gonads x 100%
- total weight
179
In order to assess the effects of environmental season-
ality and the reproductive cycle on the condition of the
fishes, condition factors were calculated according to the
following equations:
Total condition factor = total weight (g) x 100
(CFT) SL(cm)b
Somatic condition factor = total weight - gonad weight (g) x 100
(CFS) SL(cm)b
where b is the regression coefficient of the length-weight
relationship (Bagenal & Tesch 1978).
Mean values of GSIls and condition factors were
calculated separately for males and females of each species
for each month. On account of the small numbers of specimens
involved, data from all stations were pooled for these
calculations and in four species, data for equivalent months
from successive years were combined.
Fecundity and spawning patterns
The fecundity of multiple spawning fishes is difficult
to determine by standard methods of fecundity estimation
(Bagenal 1978, Bagenal & Braum 1978). Where spawning occurs
during a fairly well defined season it has been suggested
180
that a count of all developing eggs (indicated by the presence
of yolk or vacuoles) would give the fecundity for that season
(Bagenal 1978). However, where there is continuous recruit-
ment of large eggs from the small-sized eggs, fecundity will
be much more difficult to determine (Bagenal & Braum 1978).
In addition, as found by Macer (1974) ,a very high percentage
of atretic mature eggs late in the reproductive season
indicates that there may be a substantial difference between
potential and actual fecundity (Bagenal 1978). Anothermethod
of fecundity estimation involves estimating the size of each
,I
J
batch of eggs spawned and the number of batches spawned over
the total breeding season (Bagenal 1978, Gale & Deutsch 1985).
In situations such as the tropics where batches may follow
each other continuously over the year, fecundity must'only
include one batch (Bagenal 1978). In this study batch
fecundities were estimated for each species except P. reticulata
by the following methods.
Ovaries utilised for this purpose were all mature, filling
a substantial proportion of the abdominal cavity, and were
taken from fish caught during the study period. Ovaries were
teased apart to separate the oocytes. In the case of small
ovaries (from c. riisei~ H. unilineatus~ c. aeneus and some
G. sternicla) complete counts were made of all oocytes within
181
the largest size range of oocytes as determined from oocyte
size distribution analyses (Table 19). For larger ovaries
(from some G. sternicla and A. bimaculatus) a volumetric
subsampling method was used with at least four subsamples being
taken (Bagenal & Braum 1978). Data from Gilson's fixed
ovaries used in oocyte size distribution analyses were also
pooled with these data. For P. reticulata brood size was
determined by counting all ova of mature size and/or embryos
present in the ovaries of mature females caught during the
study. The relationship between batch fecundity or brood
size and standard length was determined according to Bagenal
& Braum (1978). Statistical analyses of the relationships
were conducted in the same way as for the length-weight rela-
tionships.
Investigations of spawning patterns and frequency as well
as fertility (the actual number of eggs shed (Bagenal 1978))
were attempted by laboratory breeding of the six species
under study. All specimens were collected from the study
area and kept in aerated 8S-litre aquaria which were fitted
with undergrave 1 fi1ters. The fish were fed on prepared
dried fish foods with occasional 1ive food. An artificial
spawning site was placed in each aquarium (except for
P. reticulata) and it consisted of a 15 cm diameter glass
182
petri dish covered with 3 mm-mesh nylon netting to which a
bundle of dark green nylon yarn was attached. Observations
of the fish were made every day initially and occasionally
thereafter within one or two hours after dawn in order to
witness courtship and spawning behaviour. Spawning sites
were checked every day (usually at mid-morning) to remove
any eggs that may have been spawned. This procedure was
I
carried out until the end of the experiments. Details of
specimens used in each instance and the duration of each
experiment are given with the results.
Oocyte size distribution analyses were performed on
ovaries over a range of maturity stages in order to investigate
spawning patterns and frequency, and mature oocyte size
ranges for all species except P. retiauZata. The method
utilised volumetric subsampling procedures similar to those
used by Macer (1974).
Collections of female fish were made during the 1985
rainy season. They were killed by chilling and standard
lengths and total weights measured. For each female both
ovaries were excised, weighed to the nearest milligram
and fixed in Gilson's fluid and were occasionally shaken
to separate the oocytes. After a period of at least a month
183
excess Gilson1s fluid was removed and distilled water added
to make up to a known volume: between 1 and 5 ml for small
ovaries, up to 20 ml for larger ones. Eggs were agitated
thoroughly and 1 ml subsamples taken using a Iml syringe.
Each subsample was inserted into a Sedgewick-Rafter Counting
Cell where all oocytes greater than or equal to 0.243 mm were
measured and counted (usually not less than 150 oocytes).
For the smaller oocytes four or five randomly chosen strips
of the Sedgewick-Rafter Cell were scanned and all oocytes less
than 0.243 mm were measured and counted (usually about 150
oocytes or more). For very small ovaries diluted to 1 or
2 ml, all oocytes greater than or equal to 0.243 mm were
measured and counted. For larger ovaries diluted to 5 ml or
more, up to four 1 ml subsamples were taken and each
subsample treated as above. All oocytes were measured using
an eyepiece graticule at a magnification of 40x, along their
horizontal axis which was parallel to the grid lines of the
Sedgewick-Rafter Cell. In two instances formalin-fixed
ovaries were utilised; they were treated in the same manner
as Gilsonls fluid-fixed ovaries.
In order to check the accuracy of the subsampling procedure,
coefficients of variation (COV) were calculated. For replicate
1 ml subsamples of A. bimacuZatus oocytes, COV for oocytes
184
greater than or equal to 0.243 mm was 5.22% (N = 6) while
for small oocytes less than 0.243 mm COV was 7.81% (N = 10).
Maximum egg sizes for the other characoids were similar to
that for A. bimacuZatus so it was assumed that sampling
errors would be similar. For C. aeneus which had larger
oocytes ,COV for oocytes greater than or equal to 0.243 mm
was 15.43% and for oocytes less than 0.243 mm COV was 16.27%
(N = 10 in both cases). From raw data on oocyte sizes and
frequency, the total number of oocytes in both ovaries were
estimated and the number and percentage frequency of oocytes
of each size class calculated. Using oocyte size frequency
distribution histograms, the size range of the largest oocytes
in mature ovaries was determined and used in batch fecundity
estimations.
RESULTS
Species composition and distribution
A complete taxonomic list of fish species caught in
the study area during the study period is given in Table 20.
Genera 11y the fish caught in fresh and brackish waters
belonged to nine orders, 21 families and 31 species of which
185
TABLE 20: Taxonomic list of teleost fish collected in the
Chatham streams during the study period
(classification after Gery (1977) and Nelson (1984)).
Stations
Taxon 1 2 3 4
Characiformes
Erythrinidae
HopZias maZabaricus (Bloch) x x x
Erythrinus erythrinus (Schneider) x
Gasteropelecidae
GasteropeZecus stemicZa (Linnaeus) x x
Characidae
. Brycon eiebenthalae Eigenmann x
Triportheus eZongatus Gunther x x
Corynopoma riisei Gill x x x
Astyanax bimaculatue (L innaeus) x x x
Moenkhausia bondi (Fowler)l
Hemigrarronus uni Zineatus (Gi11 ) x x x x
Si 1uriformes
Pimelodidae
Rhamdia sebae (Valenciennes) x x x
Callichthyidae
CaZZichthys caZZichthys (Linnaeus) x
Corydoras aeneus (Gill) x x x
Gymnotiformes
Gymnotidae
Gymnotus carapo Linnaeus x x
Cyprinodontiformes
Ap 1ocheil idae
RivuZus hartii (Boulenger) x x x
186
TABLE 20 (continued)
Stations
Taxon 1 2 3 4
Poeciliidae
Poecilia reticulata Peters x X X
P. picta Regan X
P. vivipara Bloch & Schnei der X
Atheriniformes
Atherinidae X
Syngnathiformes
Syngnathidae X
Synbranchiformes
Synbranchidae
Synbranchus marmoratus Bloch X X
Perciformes
Centropomidae
Centropomus parallelus Poey X
Serranidae
Epinephelus itajara (Lichtenstein) X
Gerreidae
Diapterus rhombeus (Cuvier) X
Haemulidae
Pomadasys sp X
Nandidae
polycentrus schomburgkii Muller & Troschel X
Cichlidae
Cichlasoma bimaculatum (Linnaeus) X X
Crenicichla alta Eigenmann X X
187
TABLE 20 (continued)
Stations
Taxon 1 2 3 4
Mugil idae
Mugil curema Valenciennes x
Gobiidae
Sicydium punctatum Perugia x
Pleuronectiformes
Bothidae
Citharichthys sp x
Soleidae
Trinectes sp x
1 Collected in the middle-lower reaches of the Carlisle River
only.
188
two species were new records for Trinidad. The most dominant
families in terms of species represented were Characidae (six
species), Poeciliidae (three species), Erythrinidae, Callich-
thyidae and Cichlidae (two species each).
Of the 19 freshwater fish species, the characids were
very common at the first three stations especially C. riisei~
A. bimaculatus and H. unilineatus. P. reticulata was also
abundant. G. sternicla and C. aeneus were commonly found
and especially towards the end of the study period the former
species became quite abundant. Hoplias malabaricus~ Rivulus
hartii and the two cichlid species, Cichlasoma bimaculatum
and Crenicichla alt~~ were caught occasionally as adults but
during their respective reproductive periods juveniles were
common. Other species such as Rhamdia sebae~ Gymnotus carapo
and Synbranchus marmoratus were only rarely caught possibly
due to their more nocturnal habits.
Certain freshwater species were restricted in their
distribution, for example Moenkhausia bondi was only found
during an extensive seineing effort along the middle and
lower reaches of the Carlisle River, an area not regularly
sampled. G. sterniCZawas collected only from the Carlisle
River. Erythrinus erythrinus and Callichthys callichthys
189
were found only at Station 3. This station was situated just
beside the Chatham South Road and it is possible that they
were introduced there. E. erythrinus was only found later in
the study period. Another restricted species was PoZycentrus
schomburgkii found only at Station 2 and whose distribution
was probably determined by habitat preference since this
pool was one of the deeper, more extensive and permanent ones
in the upper Carlisle River. Pool depth and permanence most
likely influenced distribution of fish at the freshwater
stations since Stations 2 and 3 each supported 15 species as
compared with only 10 at the shallower less permanent
Station 1.
Special mention must be made of the presence of two
species which were new records for Trinidad, namely Brycon
siebenthaZae and Triportheus eZongatus (Alkins & De Souza
1983/84, Sturm & De Souza 1984). Only two full grown
specimens of B. siebenthaZae were caught at Station 2 and
extensive fishing of both the Carlisle and Quarahoon Rivers
fai1ed to turn up others. It is presumed that this species
had recently colonised the Carlisle River but had not
established a population. T. eZongatus on the other hand,
appeared to have been caught in the process of colonisation
of these rivers. They were initially found at the mouth of
190
a small river at Los Blanquizales Lagoon and at the mouth of
the Quarahoon River in September 1981. At thi s time their
distribution extended to approximately 300 m upstream of
Station 4 (about 600 m from the river mouth). By January
1985 they were common in the middle reaches of the Carlisle
River and as far as Station 3 on the Quarahoon River (about
1.2 to 1.4 km from the river mouth). A restricted range of
sizes of individuals were caught, from 56 to 108 mm SL as
compared to average adult size on the mainland between 200
and 280 mm SL (Goulding 1980). This suggests that these are
pre-reproductives but considering the wide distribution of
the species and the need for revision of the genus (Weitzman,
pers. comm.), this cannot be certain without further
investigation. No juveniles have been noted however. It
remains to be determined whether this species has established
a breeding population in the area.
The collection of T. el.onqatue in brackish water
indicates that this characid has a well developed tolerance
to salinities up to 12%0 (Sturm & De Souza 1984) and a long
term tolerance of salinities up to 5%0 (Alkins & De Souza
1983/84). In April 1981 a specimen of H. uniZineatus
(17 mm SL) was found at Station 4 at a time when specific
conductivity was 800 ~mhos. The potential of these characids,
191
generally considered to be primary freshwater fishes (Myers
1938), to tolerate brackish and saline waters needs to be
pursued further in the light of these findings.
Of the 31 species found in the study area only 12 were
restricted to brackish water at Station 4. They included
Centiropomue paral.Lel.ue whi ch occurred inmost of the catches
and Poecilia vivipara~ Diapterus rhombeus~ Cithariehthys sp
and Trinectes sp which were only occasionally found. Other
species such as Poecilia piot:a, Epinephel.ue itajara~ Pomadasys
sp, Mugil curema, Sicydium punctatum, atherinids and
syngnathids were uncommon or rare. However, at the mouth of
the Quarahoon River schools of juvenile M. curema~ atherinids
and Anableps anableps were seen. Many of these species were
represented by juveniles only, for example C. parallelus~
D. rhombeus, Citharichthys and Trinectes, and indicated the
use of this estuarine area as a nursery ground for these
species.
Population studies and reproductive seasonality
General trends in population sizes of the six species
studied are given first with more detailed results on morpho-
metry, population demography and reproductive seasonality
192
after. The quality of the data varied for the different
species largely because of lack of availability of specimens
at certain times.
Population fluctuations:
Despite the fact that the sampling method used for the
first year of the study was only semi-quantitative, the data
showed distinctive trends in the fluctuations of population
sizes for the study species. During the first year of the
study period, total numbers of the six species collected
from all freshwater stations varied widely from one month to
the next (Fig. 28). Such variation was due mainly to two
recognisable factors: the effects of flow regime and the
input of juveniles into the population. The former factor
involved two extreme situations where increased discharge
resulted in washouts and drastic decreases in fi$h populations
(June/July, October 1980, April 1981) and on the other hand,
lentic conditions and contraction of the habitat concentrated
fish into pools making them more vulnerable to capture in
large numbers (May 1980 at the end of a three month lentic
period) . Total numbers of fi$h caught showed a significant
\
inverse correlation with mean discharge for Stations 1, 2 and
3 (Spearman Rank correlation coefficient = -0.567, N = 14,
p<0.05). Substantial increases in the fish population
also occurred as a result of inputs of juveniles into the
193
0-0 Total
........ lmmatures
s...:.. 300
CI
~
e
u
..c
-VI-0
~
Qj
.0
E
~
z
0.3
--I
VI
C""l 0.2
E
QI
C~I
st::l
u 0.1
VI
Cl
0
M J J A 5 0 N 0 J F M A M J
1980 1981
FIGURE 28: Monthly flvctuations of total fish populations in
relation to mean discharge for Stations 1 to 3.
194
populations (Fig. 28) and was particularly evident in August
1980 and to a lesser extent in December 1980/January 1981.
and May/June 1981. Migration of fish within the stream may
have accounted for population size variation but this was not
investigated.
The general trends exhibited during the first year of
the study were also seen during the later sampling periods.
Data collected at these times were not quantitative but
general observations showed that peaks in fish population
sizes occurred during the 1981 rainy season (June and
September 1981) and during the 1982 dry season (December 1981-
May 1982). The latter may be partially explained by inputs of
juveniles at these times. In particular. heavy rains in
November/December 1981 foll owing a dry petit careme appeared
to have triggered massive reproduction in H. uniZineatus.
Subsequently. moderate rainfall throughout the early part of
the 1982 dry season produced conditions favourable for
survival of these juyeniles in addition to P. petiauZata3
and high total populations of these species were seen at this
time. Addition of juveniles of other species in May 1982
also augmented total populations. However. nearly 400 mm
of rainfall in June 1982 caused washouts and decimated the
fish populations severely.
The effects of flow regime on the populations of the six
195
species could be seen even more clearly by looking at each
station separately (Fig. 29). Generally, numbers of fish
caught were greater at Stations 2 and 3 than at Station 1
and this was consistent with the larger and more permanent
habitat at the former two sites. Nevertheless, numbers of
fish caught fluctuated widely at these two stations from
month to month. At times of high discharge and floods,
severe decreases in fish catches were apparent. Increases
in population density occurred at times of more moderate
or low discharge. A statistically significant correlation
between numbers of fish caught and discharge rates was
found only at Station 2 (Spearman rank correlation
coefficient = -0.646, N = 13, p<0.05).
Analysis of the data by species allowed the determination
of general population cycles for each species over the first
year of the study (Figs. 30 and 31). Median catches were
calculated for each species as the median value of ~onthly
catches for all stations combined. Immatures were considered
to be those individuals less than Minimum Developing SL for
the species.
Numbers of G. sternicla caught were relatively constant
196
Stn
20 / -, .2
/ "- e
/ ·e /<, <, /
,/
/ e <, /_e- -eo- _ e
/
O 0
.....
s: Stn 2
CI --;:, I-
Ul
~200 ('I')
s: E
\I)
- CI.I/' .2 C\'0100 ~/ \ t's:
L.. / \ u
c» Ul
.0 ...e..... \
E 0
::::J 0
Z 0
Stn 3
200
M J AS 0 N D J F M AM J
_---- 198-0----- ----1981 ----
e>---O Number of fish e- -e Discharge
f - flooded
FIGURE 29: Monthly fluctuations of fish populations at each of
Stations 1 to 3 in relation to discharge.
197
(a)
20J
o~
(b)
40
s: 20
0'
::J
0 •-,
u
s:
VI
0
-0
L.. (c)
Q1
.&;)
E
::J 0--0 Total
z 60 • -. Immatures
40
20
O~-----=~-~-...----r--~--.=~-.----~-{J---S---r--~
M J AS 0 N D J F M A M J
----- 198-0------ -----1981----
FIGURE 30: Monthly fluctuations of populations of (a) .§.. sternicla,
(b) I. riisei, and (c) 12· bimaculatus at Stations 1 to 3.
(0)
120 198
80
40
. ..-
0
(b)
s:
Ol
::::J
0
u 40
x:
-til-0 0 •
L..
(1) (c)
..0
~ 160
z ~Totol
• - • Immatures
120
80
40
•-, /.,
-, '. ,.../0 •M J J A 5 N D J F M A M J
1980 1981
FIGURE 31: Monthly fluctuations of populations of (a) li. unilineatus,
(b) C. aeneus, and (c) f. reticulata at Stations 1 to 3.
199
over time but remained low throughout this period: median
catch was 3.5 individuals (Fig.30). Immatures appeared in
the population mainly during the 1980 rainy season and again
in much smaller numbers in the early 1981 dry season. During
the second year of the study G. eteimiol.a must have had
successful breeding periods and numbers of this species
increased significantly over time. By the end of the study
fish collections were dominated by this species.
C. riisei numbers fluctuated cycl ically over the study
period (Fig.30 ) with peaks in population size occurring
during the main rainy season (August 1980) and after the
second rains (December 1980/January 1981). At these times
the population consisted of immatures to a large extent,
thus suggesting reproduction occurring then. This pattern
was generally exhibited for the remainder of the study
although unusual rainfall patterns in 1981/1982 altered the
timing of such cycles. In general, the December/January
peak in population was slightly greater than that for July/
August in both years. Median catch for this species was
15 individuals for the three stations fished.
The majority of the individuals of A. bimaoulatue
caught were immature and these accounted for the population
200
changes seen (Fig. 30). Large increases in numbers of
immature individuals occurred during the first rainy season
(August 1980, May/June 1981) and a minor peak occurred in
January 1981 after the second rains. This pattern of major
and minor peaks in numbers of immatures continued fairly
regularly over the rest of the study. Adults of this species
were particularly low in numbers so that trends in adult
populations could not be observed. Median catch was 6.5
individuals for the three stations.
Numbers of H. uniZineatus caught were relatively large
compared with other species (median catch 30.5 individuals).
Total populations varied widely as a result of input of
immatures and washouts due to floods (Fig. 31). The
population consisted mainly of immatures during the main
rainy season (August 1980) while a smaller peak during the
second rains (December 1980) consisted largely of adults
with some immatures. A similar trend was seen throughout
the rest of the study although a large influx of immature
individuals occurred in January 1982 after particularly heavy
rains in November/December 1981. Survival of these small
individuals throughout the 1982 dry season was good when
low to moderate flows were maintained but very heavy rains in
June 1982 severely decreased the population of this and
201
other species.
Catches of C. aeneus were very variable and generally
low (median catch 3.5 individuals). Initially large numbers
were caught but numbers declined markedly thereafter (Fig. 31).
Presumably such large numbers were caught because of their
concentration in the pools during a long dry season. Generally
however, no clear seasonal cycle of population fluctuations
was discerned.
Populations of P. reticulata oscillated widely over the
study period with a median catch of 24 individuals for the
three stations sampled (Fig. 31). Populations consisted
almost entirely of reproductive individuals except in August
1980, January and June 1981 when immatures made up a minor
proportion of the population. Massive inputs of immatures
such as seen in some other species were not seen in this
species and fluctuations were due to variation in numbers
of mature individuals. Comparison of P. reticulata population
sizes and discharge rates indicated a significant inverse
correlation (Spearman rank correlation coeffi~ient =-0.552,
N = 14, p<O.05). Substantial decreases occurred during times
of heavy rains and flooding, but populations recovered afterwards
when moderate or low flows existed. Population recovery was
extremely rapid on most occasions, attaining very high
202
densities after only a month or two. In certain instances
population decline began before the advent of heavy rainfall.
In such cases population movements or some other type of
population control mechanism or even overfishing might have
been contributory.
Morphometric studies:
The six study species varied widely in terms of general
body size (Table 21). A. bimaculatus was by far the largest
species with P. reticulata the smallest. C. riisei and
H. uniZineatus were approximately equal in lengths but body
proportions differed. c. aeneus was generally larger than
G. stemicZa but there was a substantial overlap of sizes.
Males and females of all species differed in maximum sizes
attained with males being smaller than females in G. sternicZa~
H. uniZineatus~ C. aeneus and P. reticuZata. Maximum male
size was greater than that for females in c. riisei and
A. bimacuZatus but in general many males tended to be smaller
than females.
Secondary sexual characters made possible the determination
of sex using external features in mature or even developing
specimens of C. riisei~ A. bimaculatus and P. reticulata.
203
TABLE 21: Range of standard lengths (SL, mm) and total
weights (TW, g) recorded during the study
period for the six fish species studied.
Minimum Maximum
Species Female Male
SL TW SL TW SL TW
G. stemicZa 14 0.038 49 4.333 42 2.466
c. riisei 9 0.006 37 .0.828 39 0.824
A. bimacuZatus 8 0.010 76 15.757 82 17.801
H. uniZineatus 9 0.011 39 1.413 33 0.857
C. aeneus 24 0.671 60 10.414 53 6.120
P. reticuZata 6 0.002 25 0.394 16 0.076
204
However, certain differences in body proportions made it
possible to distinguish sexes in the other species after
some experi ence had been gained. Analys is of these body
proportions showed that they could be used as fairly reliable
external indicators of sex (Table 22). Mean body depth:SL
ratio in G. stemicZa and H. unilineatus showed a highly
significant difference between males and females (Table 22)
with females being deeper bodied than males especially
when ripe with eggs. C. aeneus males and females differed
significantly in terms of body width : SL ratio and
pectoral fin spine length: SL ratio (Table 22). Females
were wider bodied and possessed shorter pectoral spines
than males.
Length-weight relationships for the six species are
shown graphically in Figs. 32 to 34. Points were plotted
for data from June 1980 to January 1981 but the line of
best fit was derived from data from June 1980 to June 1981.
The parameters of the length-weight relationships derived
from a full year's data are expressed in Tabl e 23. In all
cases the correlition coeffi~ients, r, were highly significant
and the regression coefficients were highly significantly
different from zero (F-test, p<0.001). Regression coefficients
ranged f'rom 3.049 to 3.515 in C. aeneus and G. eteimiola
V l
c ; )
. . . . .
: : : : :
3 :
.
.
~ " ' C
C D
n
C O
~
. . . . .
\ 1 : )
~
s
3 :
\ 1 : )
C D
; : s
~ .
Q l
* =
t i l
. . . . . . \ 1 : )
~
*
~
~ .
~
C D
0 -
*
; : s
~ .
C O
I
Q
\ 1 : )
e T
- l
~
~
) : >
C D
~
~
t i l
t o
~
e T
r
C O
I T 1
" 0
N
' / \
N
"
0
.
" " ' 0
( J )
0
( " )
( J )
0
"
. . . . . e T
3
t o
t o
t o
t o
Q l 0 , . . . . . .
V >
t i l
0
. . . . . . 0
0
0
- S
( J )
" 0
t o
0 -
0 -
0 -
( J ) Q l
( J )
X
Q l
~
. . . . . .
«
- c
~
( " )
s : : :
t i l
. . . . . .
Q l
( J )
" ' C
0 -
: : ( 0 -
t i l . . . . . .
0 - 3
( J )
- S
. . . . . ( J )
" 0
( J )
0
" 0
. . . . . .
" 0
0 -
0 -
: : : 3
0
" ' C
e T
c - +
c - +
: : : 3
. . . . . .
t i l
: : : 3
0
~
~
( J )
~
3
- S
V >
0
0
c - +
. . . . . . - + >
I T 1
. . . . .
- S
0
- + >
( J )
" 0
- S
0
V >
t i l
t i l
: : : 3
~
Q
~ 3
r
r
r
1 . 0
. . . . . .
P I
e T
t i l
t i l
- - '
~
e T
. . . . . .
C O
3
Q l
~
: : : 3
. . . . . .
: : : 3
\ 1 : )
J
0 -
: : : 3
- + >
Q l
V >
. . . . . .
~
r
- S
~ . t i l
X
0 -
( J )
( J )
Q
. . . . . .
~
0 -
V l
I T 1
( J )
~
. , . ,
C D
" T l
3 :
3 :
: : : :
3 :
- S
. .
( " )
t i l
x
- S
" "
e T
" 0
0
( J )
( J )
: : : : :
- S
.
( " )
0 -
. . . . . .
e r
~
3
U ' l
U 1
e N e N
N
N
N N
; : s
( J )
0
. . . . .
0 3 :
U ' l
- . . . J
- . . . J
\ 0
~
~ .
. 0 -
: : : 3
- C D
.
' "
t i l ~
~ Q J «
. . . . .
C O ~
. . . . .
\ 0
. . . . . 0
O ' l
~ .
~
N
\ 0
. . . . .
. . . . . e N
~
m
~
; : s
" 0
1 +
: : z
\ 1 : )
- S
- s
1 + 1 +
1 +
1 +
1 +
1 +
1 +
1 +
V l Q J 0
~
" 0
~
f ' T I c - +
0
0
0
0
0
0
a 0
. . . . . . , . . . . .
0
~
: : z
c o
. . . . . - S
0
. . . . . N N
. . . . .
N
~
s : : : N
e T
W C O
. . . . . U ' l
C ' \
- . . . J
U 1
\ 0
3
. . . . . .
Q l
e r
0
: : : 3
( J )
: : : 3
0 -
- S
t i l
. . . . .
Z
~
C ' \
c o
- . . . J
c o
' - J C O
0
. . . . . . .
\ 0 U ' l
0
\ 0
U ' l
0
0
. . . . . . . C J ' l
' "
0
~ - S
t i l
\ 1 : )
" 0
; : s
3
( J )
\ 1 : )
Q l
~
c o
\ 0
( " )
' - J
.
e T
~
. . . . . .
c o s : : :
\ 0 0 -
N O ' l
\ . 0
3
. . . . .
- S
O ' l c o
( J )
~
' - J
( J )
- . . . J
c o
\ 0
~ N
l j o
t i l
*
*
*
*
*
*
*
*
*
*
*
S 0 2
(a) ( b) 206
10.,------------......---~ 1.0-r-------------r--.
1.0 0.1
-oo
~
0.1 0.01
TW = (4.375 x 10-6)(SL)3.515 TW = (4.519111Q-6)(SL)3.342
O.Q1+_-_ -,----;,---.---r-....-IT-rl O.OO1+-..,...,rr.,.,.----r----;r---r--4
10 100 10 50
Standard length (mm)
FIGURE 32: Length-weight relationships for (a) G. sternicla, and
(b) C. riisei.
(a) 207
20...------------~~
(b)
10 10.....--------------.
TW = (7.5861110-6)(SU3.323
--- 1.0
-0\.
10
..o...
o
f-
0.1 0.1
0.01 0.Q1-+-,...-,- ......---~-,.-----.---4i
10 100 10 50Standard length (rnrn)
FIGURE 33: Length-weight relationships for (a) A. bimaculatus,
and (b) l:i. unilineatus.
(a) (b)
208
l00I-r---------~~ 1.01-.-----------r-----.
10 0.1
0\
---
-o
o
I-
1.0 0.01
TW=(43.152x10-6)(SU3.049 0.001 TW= (7.745x10-6)(SU3.366
0.1
100 10 50
10
Standard length (mm)
FIGURE 34: Length-weight relationships for (a) C. aeneus, and
(b) P. reticulata.
209
TABLE 23: Parameters of the length-weight relationship for
preserved male and female specimens belonging to
the six study species (collected from June 1980
to June 1981).
Species N 10910a a(x10~6) b r t 1S
G. sternicla 86 -5.359 4.375 3.515*** 0.996*** 14.905***
C. riisei 233 -5.345 4.519 3.342*** 0.980*** 7.714***
A. bimaculatus 166 -5.242 5.728 3.427*** 0.992*** 12.934***
H. unilineatus 461 -5.120 7.586 3.323*** 0.989*** 13.696***
ns
C. aeneus 99 -4.365 43.152 3.049*** 0.984*** 0.879
P. reticulata 517 -5.111 7.745 3.366*** 0.968*** 9.542***
N : Number of specimens
10910 a: Intercept
b Regression coefficient
r Correlation coefficient
*** p<O.OOl
-
ns not significant
1 t-test for deviation of observed b from 3.0
210
respectively while the constant, a, of the length-weight equation
ranged from 43.152 x 10-6 in C. aeneus to 4.375 x 10-6 in
G. stemicZa. For all species except C. aeneus, length-weight
relationships were significantly different from cubic (Table
23). Comparison of regression coefficients between species
showed that b for C. aeneus differed significantly from that
for all other species (Appendix 13) and may be as a result
of denser body material owing to the presence of external
scale armour. Slopes of the length-weight equations for
C. riisei, A. bimacuZatus, H. uniZineatus and P. reticuZata
did not differ at a highly significant level. However their
intercepts did show a highly significant difference
(analysis of covariance, F = 512.31, df = 3/1372, p<O.OOl).
C. riisei had the lowest value of a consistent with its elongated
and slender body form as compared to the other deeper bodied
species.
Generally all species showed significant shrinkage in
length as a result of formalin fixation (Appendix 7).
Shrinkage in C. aeneus was significant (paired-sample t-test,
p<0.05) but not as highly significant as for the other species
(p<O.OOl). Average preserved lengths ranged from 96.98% of
fresh length in P. reticuZata to 99.20% for C. aeneus. Length
decreases were more variable between specimens as compared to
211
weight changes and this was reflected in standard errors of
the means.
All fish except C. aeneus decreased in total weight after
preservation. There was a highly significant increase in
total weight after preservation in C. aeneus (p<O.OOl) which
averaged 105.61% of fresh weight and which was greater than
110% for two specimens. Loss of weight due to preservation
was not significant for A. bimacuZatus despite average
preserved weight being 98.14% of fresh weight. Many
A. bimaculatus specimens gained weight slightly after
preservation (up to 102.38%). Weight loss was highly
significant for all other species (p<O.OOl) and means ranged
from 94.85% for H. unilineatus to 97.32% for C. riisei.
Overa 11, a trend towards greater and more significant
shrinkage and weight loss occurred in the smaller species
as compared to the larger ones where less highly significant
shrinkage or weight loss and even weight gain took place.
Within each species, a similar trend was observed where
smaller individuals tended to lose more weight than larger
ones but this was not always consistent. These general
trends agreed very closely with those seen in other experiments
of this kind (Parker 1963).
212
The effects of formalin preservation on the regression
coefficient of the length-weight relationship for each species
was not significant (Appendix 8, d-test, p>O.10 in all cases)
although d-values for A. bimacutatu3 and C. aeneus were
relatively high. Length-weight relationships of both fresh
and preserved specimens had highly significant regression
and correlation coefficients. Slight differences were
evident between parameters from these data and those summarised
in Table 23. This was probably due to the short time span
over which these experimental specimens were collected and
the possibility of seasonal variation in the length-weight
relationship (Bagenal & Tesch 1978).
Population demography and reproductive seasonality:
During the study period no direct evidence of reproduction
was obtained for the six species under study. No spawning or
courtship behaviour was actually observed in the field not
through lack of time spent making observations but because of
the low visibility of the turbid waters at the study sites.
As a result, a range of other types of evidence had to be
utilised in order to elucidate reproductive patterns for these
species. These included the timing of occurrence in the
population of juveniles and individuals with developing,
213
mature or spent gonads to indicate the potential for and the
existence of previous reproductive events. In addition,
analyses of gonad states of mature-sized individuals using
gonadosomatic indices were made.
Other species besides those being investigated were
collected occasionally and gonad states were noted. Mature
male and female Callichthys callichthys were collected in
July, September and December 1980, and January 1981. Hop l-i.ae
malabaricus were caught with mature gonads in May and June
1980 and juveniles were noted throughout the rainy season.
Cichlasoma bimaculatum juveniles were predominant during the
dry season of each year during the study whereas those of
Crenicichla alta were found during the rainy seasons.
Each of the six species studied in detail will be dealt
with separately to describe the results of this part of the
study.
Gasterope Lecue etexmiola
Sexual maturation and growth:
Females of this species possessed developing gonads at
a minimum length of 32 mm while males had developing gonads
ata smaller size, 30 mm (Table 24, Fig. 35). The smallest
- - l
( / )
; b .
c ; o
: z ) : : >
3 : ' l : : 1 \ . : l : : : r : :
; J
- 0
c o
"
r n
r
n
C ! l
o - ' - S
' - S ~ ~
r n
- - I .
, " .
, " .
C I : l c t -
C I : l
: z
~
3 :
r o
C I : l
r o P J c
c t -
~
N
" ' " .
' " .
. . . . . . V I
t i l
< ; - - l
C I : l ' - S
~
: 3
: 3
- P o
" " .
( \ )
Q
Q
~
r o
P J t : T
~
' " .
- - '
C D C ! l
~ ~
~
' " .
" " .
Q
< ; - - l
C I : l
r o
- S
" "
~ ~ R
~
0 c t - c t -
c t -
" "
- t ,
~ ~
~
r
3 :
- - - '
t i l
t i l
r o
r o
V I
: : : l
= '
- 0
= '
- - I .
t o
t o
C D
e - t - : 3
r l "
o
c
z r
~
/ J l : 3
: 3
. . , . ,
P J
C l
N N W
C D
W
- t .
r l " C l
r o
N
W
W
: : : l c o
- P o
r o
V I
<
= '
- t ,
<
r o
. . . . . .
. . . . . .
r o
: 3
. . . . . .
- s
: 3
0
/ J l
0
- 0
. . . . . . - 0
r l "
. . . . . .
: : : l
N N W 3 :
3 :
W - P o
- - I .
z s
: 3
t o
. . . . . . . 0
0 0 ' 1
~ 0
t o
P J
: : : l
r l "
c : P J
: 3
: : : l
C
- s
. . . . . .
Q .
: 3
r l "
. . , . ,
( / )
( . J 1
. . . . . . . N « 3 :
N W
W
( . J 1 P J
. . . . . . .
r
- . . . . J 0 ' 1 W
- P o
- t ,
r t
3 :
c :
P J 0
. ,
. ,
r t
r o
C
. ,
r l "
V I
~
r o
( . J 1
N N
W W r o r t
3 :
. . . . . . .
. . . . . . .
P J
~ 0 ' 1 0 ' 1
- P o
V I
: : : l
. . . . . .
Q .
P J
X
. ,
Q .
V l
r t
. . . . . . .
N N
W W
c :
. . , . ,
( X l ( X l
N Q .
~ 0 0
r o
3 : : : : l
«
- P o t T l
r o
t o
Q .
V l
r t
- - I .
- 0 ~
P J
r o
V l
( )
: : : l
- - I .
( / )
( / )
r o
. . . . . . .
W N W
W
r
V l
r
( . J 1
N
0 0 ' 1
- P o
3 :
( X l
t T l
P J
: : : l
) : : >
Q .
- - '
- - - '
3 :
r o
Q .
. . . . . . .
. . . . . . .
~ t T l N
. . . . . .
. . , . ,
. . . . . . . . . . . . . .
f - '
W - . . . J
~
P J
N
0 ' 1 0 ' 1 0 \ . 0
~
: : : l
: z
( . J 1
. . . . . . . . . . . . .
. . . . . . .
t T l N
0
- P o 0 ' 1 W
0 ' 1
3 :
( . J 1
W N
0 0 ' 1
" " " 0 ' 1
v I Z
f i s h
o f
N u m b e r
W
0
- -
I V
0
0
0
0
- -
o
z
-
0
z -
I l
0 -
I I
-
- -
- -
u : >
e n
- -
~
e n
t . . . .
O ~ : : : O
-
I t > 0 I I >
c
3
< . . . . 1 . 0
<
I t > C . . .
3
I t >
- . . . 1 1 >
: ; ,
0
o I I > V \
. . . . .
c " ' O V \
; ; -
. . . : ; " 5 '
1 1 > \ 0 1 0
S T Z
216
mature female was 33 mm while mature males were as small as
31 mm. Only a few individuals were found to be mature at
these sizes, the majority matured at larger sizes. Median
length at first maturity was 38 mm for females and 34 mm for
males. The latter estimates were subject to error because of
the small number of fish caught at these lengths. Approximately
78% of females over 45 mm and males over 34 mm were mature or
spent. Only three spent females were caught between 44 and 46
mm in length; no spent males were recorded.
The length-frequency histograms used to derive the above
data showed three distinct length classes for each sex (Fig.35 ).
Small males and juveniles constituted one class with lengths
lesst~an 27 mm. A second class was comprised of females of
lengths from 29 to 36 mm (modal length 33 mm), and males from
28 to 33 mm (modal length 32 mm). A third class was made up of
the largest individuals caught: females greater than 40 mm
(modal length 42 mm) and males greater than 37 mm (modal length
40 mm). Very few specimens of sizes intermediate to these
groups were caught. Considering that these data have been
derived from individuals caught over the study period, these
classes could not be indicative of differential recruitment but
must reflect differential capture rates either due to migration
patterns or to varying growth rates with age.
217
Most regular sampling was done in pool habitats and
G. sternicZa may have only entered these habitats in large
numbers when ready to spawn. This could account for catches
comprised of specific size classes and specimens with gonads
in the developing or mature stages. However, large numbers of
fish with inactive gonads were also caught within these size
classes, an unlikely prospect if these areas were spawning
grounds.
On the other hand, few individuals of certain lengths may
have been caught because of rapid growth over that range. In
this situation, the majority of individuals would have
belonged to certain size classes due to their slower growth
during that phase. The coincidence of the two larger size
classes with the appearance of developing and mature individuals
could indicate the distinct separation between growth and
reproductive phases in the lifetime of this species. If this
was true, then at least two separate periods of reproduction
during the lifetime of each fish would be indicated.
Some evidence for growth 'spurts' alternating with slower
growth rates was obtained from growth curves derived from
monthly population structure data (Fig.36). Changes in
population structure over time allowed the determination of
II
I
FIGURE 36: Monthly variation of population structure and the
occurrence of gonad maturity stages for ~. sternicla
(I-III: cohorts used in growth analysis).
219
growth for three juvenile cohorts: July 1980 (I), July 1981
(II) and February 1982 (III). It was assumed that sexual
maturation of both sexes occurred within 12 months at a size
intermediate between the Median Length at first maturity for
males and females (36mm). These data show a flat~ning of
the growth curve at 33 mm and 43 mm, that is at 8 to 11 months
and at 14 to 17 months of age respectively (Fig.37). This
was coincident with the first appearance of developing
individuals at about 7 to 9 months and the youngest mature
individuals at 8 to 11 months of age. Parameters of the von
Bertalanffy growth equation were estimated and the full
equation was:
~t = 44.18 ( 1 - e-O.227(t-2.175))
The curve did not fit the upper points on the graph because of
a lack of data from fish of these sizes. Maximum lifespans
estimated from this curve indicated that females live up to
28 months while males live for 17 months assuming equal growth
rates throughout life for both sexes.
Population structure and reproductive seasonality:
Changes in population structure of G. sternicla each
month indicated one major reproductive period at the beginning
of the rainy season each year. Juveniles as small as 14 mm
appeared in catches in July 1980, July 1981 and July 1982 (Fig.36)
l e n g t h ( r n m )
S t a n d a r d
. . . . . . N W . l ' ( J ' I
° r _ . . . _. . - . . . J _ ° _ . L . - _ _ O L _ - . L . - - o L - . . . . - . - . . o . L - - L - - - - . J °
O J
G ' l
r o
- s
- s 0
r t
~
O J
r t
: : : Y
O J
: : : : l n
c
- - t >
- - n
- s
<
' <
r o
r o
. D
- - n
c
0
O J
- s
r - t -
1 G ' l
0
: : : : l
- -
t i l
r t
°
r o
- s
: : : s
} >
. . . .
n
1 0
I I )
I I
O J
Z '
Z '
3 '
: 3
. . . . . .
0
O J
( X )
: : : : l
. . . .
r o
. . . . . .
z r -
t i l
I I I
/ l )
N
I
0
P
N
N
" ' - J
. . . .
-
0
I
N
. . . . . .
n
" ' - J
o
c . n
: 3
0
0 -
-
0
-
: : : : l
r o
o ,
w
n
°
c
- s
t > 0 0
<
•
r o
n
3 :
0
0
- - n
I I I
. . . .
z r
Q .
0
r t
. . ,
r t
0
3
r o
: : : : l
0
-
o ,
. . . .
c
: -: =
-
; 0
. . ,
r t
o
. . . . . : : : : l
- < 1 0
. . . .
<
o t r
: : : : l
o z z
221
suggesting an extremely regular annual reproductive cycle.
The presence of mature specimens over more extended periods,
for exampl e from May to December 1980, June to September 1981,
December 1981/January 1982 and March to August 1982, might
suggest a more prolonged breeding period than indicated by the
appearance of juveniles. However, most of these mature
individuals were males; females had a more synchronised
appearance. Nevertheless, the majority of reproductives were
caught during the rainy parts of the year especially the main
rainy season from May to August with a minor representation
during the second rains in the latter months of the year. An
exception to this trend was seen in 1982 when many reproductives
were caught as early as March. The rather wetter than average
conditions experienced early in the dry season might have
contributed to this trend.
If data for all the years of the study were combined on a
monthly basis, it appeared that mature or spent females were
caught only from March to September. Mature males occurred
in all months for which data were available except February
(Fig.38). In some months outside of the apparent breeding
period, mature-sized fish with inactive or immature gonads
were predominant (up to 100% for females and 47% for males).
The highest proportion of mature males and females was
recorded in June/July. Spent females were found only in
5 4 222
(a)
80-+-~-¥
60
o-e
40
20
9 5 19 5 49 3 14 19 10 7 0 11
1001"TTTT1'.---rtTlrrrrrT-TTT,r,.~~~"TT""T-M-,"""T""T""'1~r-r..:-.........-"':""- .......;.':."..".".'"
(b)
80-
60-
-oe
I2J Immature o Developing [IT] Mature ~ Regressing
FIGURE 38: Monthly variation of gonad maturity stages for
G. sternicla (a) females, and (b) males (numbers at
top indicate sample sizes).
223
July and September. These data are based on relatively small
sample sizes and are combined from two years sampling so they
should be treated with some caution (c~ Appendix 9 (a)).
Recognition of moving modes in the population structure
diagr~ms indicated that G. stemicZa matured within one year
and therefore juveniles spawned in Mayor June of one year
matured to spawn by June or July of the following year. At
least two periods of active reproduction in the lifetime of these
fishes may be normal and a second spawning may take place either
during the later rainy period of the year (November/December)
or early in the following main rainy season.
The above reproductive cycle was supported by variation
in GSIls (Fig. 39). Female GSI's were as high as 12.262%
while maximum male GSI was 2.955%. Highest GSI's were seen
from April to August, being maximal in June. A smaller peak
was evident for males in December. The protracted high GSI
values over a period of five months was largely a result of
pooling data from three successive years (Appendix 9 (b)).
In fact, in 1980 and 1981 GSIls of both sexes rose abruptly
with the beginning of the first heavy rains and then fell
quickly afterwards. In 1982 GSI's of both sexes rose gradually
over a protracted period from March to July. This latter
224
(a)
5 4 12 12 23 2 20 9 2 2 4
8
4
•-e 0 • • I • • •
Vl (b)
~
7 2 10 2 44 3 13 19 10 5 0 4
•
2 ,
Ii
~ t f j0 ~ ~
(c)
ns ns ns * * ns ns ns ns ns ns ns9 29 15 67 5 33 28 13 9 2 10
-el~i BFemales• Males
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
J F M A M J J A 5 0 N 0
FIGURE 39: Monthly variation of GSI's for (a) females, and (b)
males, and (c) sex ratios for §.. sternicla (bars
represent ranges, boxes represent ± 1 Standard Error; binom-
ial test: ns - not significant, p>O.05; * - p<0.05;
** - p<O.Ol; *** - p<O.OOl; numbers at top indicate
sample sizes).
225
extended period of GSI increase and the variation in time of
onset of first rains in the first two years, resulted in an
overall protracted period of high GSI's in Fig.39 and masks
what is evidently in reality a highly synchronised and regular
reproductive cycle in this ~pecies.
Sex ratios for all specimens collected over the study
period showed a significantly high proportion of males and
this was also true for mature-sized specimens; male to female
ratios were 60 : 40% and 59 : 41% respectively (Table 25).
Sex ratios for each month were not significantly different
from a 50 : 50 ratio except in April and May when ratios were
significantly biased towards a predominance of females in
April and males in May (Fig.39 ).
The impact of the reproductive cycle on the condition of
G. sternicla was not extreme and no clearly cyclical changes
could be discerned (Fig.40 ,Appendix 9{c)). There was a
tendency for female condition to be highest in the dry season,
decreasing gradually over the reproductive period when somatic
condition decreased drastically relative to total condition.
Variable fluctuations were apparent in the latter part of the
year. Male condition was lowest in the dry season and increased
immediately prior to breeding, during which time both somatic
V l
; b t ; l
: : : s
: z
' I : : J C 1
C J : : : t :
- 0
V I
*
*
*
r o
*
*
( )
C 1 l
* Q - '
~ ~ ' : i
' : i
t : " .
t : " . 0 -
~ ~
~
r o
t : " . t : " . ~
\ i -
~
V I
' : ;
t : " . N C 1 l
~
~
( J
( J
t : " . ~
~
~
t : " .
t : " .
C 1 l
~ ~
~
- 0 ( J
- 0 - 0
Z
: z ~
N N
/ \ / I
/ \
0 C N
~
~ l : : i
0
0
a
M -
3 ~
\ i - \ i - \ i -
.
0 -
l : : i ~ ~
0
0 V I
a
r o
C 1 l C 1 l
. . . . . . . . . . . .
0
U 1
- s
- j
. . . . . . .
l Q
) : : >
: : : s 0
I - ' I - '
c o
. . . . . .
z
N
. . . . . I - ' N
+ : > N + : >
r
- + ,
( J )
' - J
U 1
U 1 0
0 " \
r T 1
V I
N C J 1 0 " \
I - ' N
0
o - 0
N
P J
r o
U 1
( )
: : : s
I - '
M - . . . . . . .
I - ' I - '
0
0
; : 0
3
P J
r o
. . . . . . .
N U 1
U 1 I . D
O J
- 0 I - '
I - ' N M -
: : : s w
U 1 U 1
. . . . . .
) : : >
0 V l
V V I
. . . . . .
: : : s r o
0 0
. . . . . .
.
x
. . . . . . n
. . . . . . . I - '
0 I - ' I - '
I - '
. . . . . .
0
U 1 - s
- - '
P J
: : : s
- - '
0 - r t
. . . . . .
r o
W
U 1 0 ' > C J 1 0 " \
+ : >
0
n
U 1 w a
0 " \ I . D <
+ : >
. . . . . .
V I
r t
. . . . . .
0 -
~
0
C
- I >
. . j : : : >
: : : s 0
P J
0 " \ 3 :
+ : > U 1 w + : >
. . . . . .
)
V I
P J
U 1 I - ' ' - I
0 ' > a
+ : >
. . . . . .
V I
r o r t
3
P J
: T
I - '
0 . r o
. . . . . . .
I - ' C D
w
U 1 . . . . . . . I - ' V I
W 0 0 ' > ' I
r o 0
> <
N N
I . D 0 ' > 0 ' > C J 1
W x
3
<
0 " \ N C J 1 ' - I P J C D
0 " \
O J
. . . . . .
)
: : : s
: : : s : : : s
- I >
. . . . . .
V I
V I V I
r o
*
* *
V I
r t
* * *
V l : T : T
* *
*
r o C D
x V I
. . . . . .
V I
- 0
+ : > N N ; : 0
0 ' > N Z
r t r o
N I - '
N N ( )
+ : >
W P J
s : :
( J )
I - ' I - '
a
+ : > I . D
M -
0 .
. . . . . .
r o
' <
0
V I
- 0
: s :
a I - ' . . . . . . . I - '
0 . . . . . . .
P J
C D V I
; : 0
)
M -
r t
U 1 N . . . . . . . . . j : : : >
. . . . . .
w
O J
P J C
s : :
N . . j : : : >
W
' - J )
O J
O J
M -
0
0 .
. . . . . .
. . . . . .
r o 0 .
I
0
r o
V I
0 .
. . . . . . I - ' I - '
. . . . . . . . . . . . . . . . . . . .
I - '
N
0 -
r o
P J
0 -
V I
W U 1
C J 1
+ : > C J 1 U 1
r o
+ : > U 1
0 " \
+ : > O J I . D
C l . .
: : : s
~ 0 -
0 " \
+ : >
U 1
+ : >
+ : >
+ : >
<
0 " \ U 1 . . . . . . .
+ : > 0 ' > N
0 -
s : :
P J
I - '
. . . . . .
W
V l
0 " \ N
W
0 C J 1
' - J
> <
O J . . . . . . .
+ : > 0
' - I
' - J
N
0
0 " \ W
' - I
W
+ : >
: : : s
: : : s : : : s
V I
V l
V I
*
*
*
*
*
*
9 2 2
227
(a) 5 4 12 12 23 2 20 9 3 2 2 4
1.6
';;:
-!--!
!........ If\1.4 \ If/
\ f/
!
~
..0... 1.2
u
-0 (b) 9 5 17 3 44 3 13 19 10 7 0 6
c
0
"0 1.6
c
0
U
'\: -- o;::::~=------ ~1.4 ~ -r
1.2-L...-,----.---,---.--.------,,------.----r--.--..,.--.--.---
J F M A M J J A S o N o
0-0 Total condition .-. Somatic condition
FIGURE 40: Monthly variation of total and somatic condition for
§.. sternicla (a) females, and (b) males (upper and
lower bars represent 1 Standard Error for eFT and
CFS respectively; numbers at top indicate sample sizes).
228
and total condition declined.
In summary, G. sternicla had a very clearly defined
seasonal reproductive cycle where spawning took place over a
very short period (one or two months) usually at the beginning
of the main rainy season. This seemed to be the major breeding
period although males showed gonadal responses to rains later
in the year. Individuals matured within one year in time to
spawn with the first rains of the season. At least two periods
of active reproduction during the lifetime of each individual
may be norma 1.
Corynopoma i-i.ieei.
Sexual maturation and growth:
Females of this species possessed developing ovaries at a
minimum length of 23 mm as compared to males which first
started developing at 26 mm (Table 24, Fig. 41). Mature gonads
were first found in males and females at the same size, 26 mm.
Having attained a mature size, the majority of females (57%)
had mature or spent gonads. However, many males grew to quite
large sizes before becoming reproductively active. Several
large males (up to 34 mm) had immature gonads and incompletely
developed secondary sexual characters, thus suggesting delayed
first reproduction. Only 38% of all the males of mature size
6 Z Z
230
possessed mature testes. The presence of several large
females with undeveloped ovaries suggested either delayed
reproduction or more than one breeding period during the
lifetime of the individual. Only two females considered com-
pletely spent were caught during the entire study although many
females with ovaries which showed some loss of eggs were
recorded. The latter however, were considered to be still
capable of spawning and were therefore classed as mature. No
distinctly spent males Were found although a very few large
males with slightly more flaccid testes than normal were noted.
From population structure diagrams, three distinct cohorts
(Fig. 42, July 1980 (1), December 1980 (II) and June 1981 (III)
were utilised to derive combined growth curves for males and
females. It was assumed that sexual maturity was attained in
12 months at a length of 30.5 mm (average Median Length at
first maturity for males and females). On the basis of this
curve it was estimated that females started developing by six
months of age and males a month or two later (Fig.43 ).
Minimum age at maturity was estimated to be eight months for
both sexes. These estimates coincided with the occurrence of
developing and mature gonads at the appropriate times and lengths
in the cohorts used in the analysis (Fig.42). Extrapolation
of the growth curve indicated a maximum life span of two years
231
10 10
M N = 35 J N = 17
0 0
J N = 10
N = 5
A
J N = 13 5
N = 26
~O N = 12
N = 30
oA N = 23
co
0\
....5. N = 21 N
N = 8 0 N = 260
N N = 6 N = 42
J
N = 38
F N = 28
0
N = 35
J
('oj N = 29
N = 17 ~MF
N = 6 A N = 17.....M
co
0\
..A... N = 6 M
N = 25
J N = 0
M
J N = 26
J N = 4
32 40 0 8 16
24 32 40
0 8 16 24 Standard length Imrn)
FIGURE 42: Monthly variation of population structure and the
occurrence of gonad maturity stages for f· riisei
(key as for Figure 36; I-III: cohorts used in
growth analysis).
G )
c
; ; 0
r n
~
w
G )
S t a n d a r d l e n g t h ( r n r n )
- s
o
: E :
r t
: : : : ;
n
s : : :
- s
<
r o
~
o
- s
I n
· 0
U l
r o
- - -
0
} >
' 3
\ 0
O J
I I I
r o
. . . . . .
U l
3
0
O J
~
: : : l
~
0 -
: : T
\
I I I
~
r o
\
' 3
O J
\
- '
r o
\
0
U l
" ' "
\
n
o \
' 3
\
0 -
\
: : : l
r o
\
0 -
n
s : : :
[ J
0
t >
- s
•
<
( " ' )
3 : :
r o
0
0
, . .
I I I
z r
a .
~
0
- ,
o '
, . . . .
3
r t
; : ,
0
r t
, . .
r o
c :
0 -
( i )
. . . ,
-
-
_ . ; : ,
-
-
0 -
~ \ O
" < , . .
' <
: T
r o
' <
r o
Z t Z
233
for males and about 21 months for females. Data were insuffi-
cient to determine the parameters of the von Bertalanffy
equation.
Population structure and reproductive seasonality:
Small immature C. riisei individuals occurred in the
population twice each year and mainly in the rainy seasons
(Fig.42 ). Juveniles were predominant from July to October
1980, December 1980 to March 1981, June to September 1981 and
January to March 1982. These times coincided with periods of
high or continuous rainfall. Input of juveniles appeared to
be of equal magnitude whether during the main rainy season or
the rains later in the year and took place over prolonged
breeding seasons (up to four months) each time.
The presence of reproductively mature and spent individuals
in the population coincided with the occurrence of juveniles
but there was only one month during the entire study period
when a mature fish was not caught (Fig.42 , Appendix 10 (a)).
The proportion of mature females each month was irregular due
to small sample sizes but showed highest values (100%) in
January and April, and waS more than 50% in July, August and
October (Fig. 44 ). Proportions of mature males were highest
in May, June, August and October (more than 50% of mature-
234
(a)
o-e
9 11 6 8 20 12 18 6 10 8
(b)
80
60
-oe
40
20
J F M A N D
FIGURE 44: Monthly variation of gonad maturity stages for
c. riisei (a) females, and (b) males (key as for
Figure 38).
235
sized males) which coincided roughly with the timing of females.
Completely spent females occurred only in February and May. In
between the breeding periods, that is March, May and December,
the proportion of mature-sized individuals with inactive or
immature gonads was highest possibly suggesting new recruits
entering the reproductive population at these times (see also
Fig.42 ).
GSI values for females supported the above trends and
showed a clear bimodal pattern over the calendar year (Fig.45).
Maximum GSI attained was 14.13% in September but generally
high GSI I S were seen from May to October and again in December
to February. Periods of high GSI were prolonged, over two or
three months, and coincided closely with times of juvenile
input into the population. Maximum GSI for males was 2.398% in
September but no clear cyclical pattern was evident. Mean
monthly GSI's for males oscillated between 0.887 and 1:495% over
the study (Appendix 10(b)).
Sex ratios for the whole population over the study period
was 53% males to 47% females and showed no significant
deviation from a 50 : 50 ratio (Table 25). However, significantly
more mature-sized males than females were collected during the
study (ratio 58 : 42%, x2 test, p<O.05). Sex ratios varied
236
(a)
16 9 10 3 2 7 4 2 3 4 6 13 12
12
8
4 i
8
~ 0•
lJ) (b)
\!) 4 4 3 3 11 12 9 4 5 4 2 1
2
e ~ i i
~ ~
~ ! f ~ •~
0
(c)
ns ns ns ns ns ns ** ns ns ns ns ns
21 9 10 40 16 21 9 15 14 21 n
~:J~~i~~~~ii~~~
J F M A M J J A 5 0 ~ D
FIGURE 45: Monthly variation of GSI's for (a) females, and
(b) males, and (c) sex ratios for .f. riisei (key
as for Fi gure 39).
237
monthly (Fig. 45) and in many months males predominated,
especially in July when the deviation from a 50 : 50
ratio was significant (p<0.01). In January sex ratios were
50 : 50 and only in November and December did females
predominate but these ratios were not significantly different
from unity.
The reproductive cycle of C. riisei may have affected the
condition of the fish since cyclical changes in both total and
somatic condition which could be related to the breeding cycle
were evident (Fig. 46, Appendix 10 (c)). The difference
between total and somatic condition for females was large
except in March and November when GSI's were low. In addition,
both total and somatic condition were at their lowest values
at these times after the breeding seasons. Highest values of
both condition factors were at the onset of breeding and declined
steadily during the breeding period.
In summary, C. riisei had two peaks of reproductive
activity each year coincident with the rainy seasons although
mature individuals were found almost all year round. Some
juveniles which had been spawned in the earlier part of the
main rainy season (May/June) could have matured by the second
rainy period but it is likely that most would have matured
238
(a)
1.2 9 10 3 2 7 4 3 3 5 6 13 14
1.1
. 1.0
0.9
L-
.a...
u
-o
c a.8--'----r---.---,----,,------,-----;----,----.--.-----,,---,----r--
a
"U (b)
C
a
1.1 9 11 6 8 13 12 18 6 10 8 8 13u
1.0
0.9
M A M J J A 5 0 N DJ F
Monthly variation of total and somatic conditionFIGURE 46:
for C. riisei (a) females, and (b) males (key as
for Figure 40).
239
early in the main rainy season of the following year. Individuals
which had been spawned late in the main rainy season of any year
could have matured and reproduced in the following year's main
rains or later during the second rains. Prolonged breeding up
to three or four months during any season presumably allowed
continuous recruitment of new reproductives throughout a
prolonged breeding period in the following year. Estimates of
maximum lifespans suggested that both sexes may have had only
two periods of active reproduction during their lifetime.
Astyanax bimacuZatus
Sexual maturation and growth:
Insufficient data were obtained on adults of this species
to be able to make definite conclusions as to many aspects of
its reproductive biology. The minimum length of males with
developing gonads was 40 mm and with mature gonads, 56 mm
(Table 24). The smallest reproductively active female caught
was already mature at 57 mm indicating that the Minimum
Developing SL for this sex was less than this. Data were
inadequate to determine median size at first maturity. More
than 50% of the males of mature size were found with mature
testes but only one mature female was noted from 11 of mature
size (Fig. 47). No spent adults were caught.
f i s h
o f
N u m b e r
C 7 l
. r -
I V
0
0
z - . .
0
I V
0
0
o
0
- - - i
Z . . . . . .
O J
: : : Y
0
I ' D
Z . . . . . .
1 1
0 -
I
- -
I V
1 1
- t ,
- s ( J 1
I V
I ' D
r o
- . . . J
W
- -
3
0
C 7 l
O J O J
I
M -
r o
e s
V l
0
: : : l
O J
l / l
- -
- I -
: : : Y
: : l
( j ' \
0 . .
- 0
w .
e
0 -
I ' D
<
r o M -
: : l
: £
- ' .
I ' D
I ' D
r o : : : l
V l
L C l
0
O J : : : l
: : l O J
0 -
0 . .
3
O J
C T
M -
e
U l
- s
3
. . . .
- 0 .
O J
o
M '
: : l
r o
' <
a .
V l
V l
o
. . ,
M '
O J
7 '
a .
t o
r o
I ' D
' <
V l
t l l
O J
: : l
O J
V l
: : : l
1 0
- t ,
0 -
. . . . ~
0
: : r I
V l
- s
f f i
. . , . . ,
N
I ' D
L a
e ~ ~
- i 1
,
0 - 0
r o
- s
~
I
w
I : : P
U 1
f f i
0 -
- 0 .
~
3
O J
a
n
e - - o J
- - '
- -
O J
. ! . J
( J 1
M -
e
V l
~
~
O P Z
241
The predomi nance of j uvenil es and 1ack of adults in
catches could have been due either to difficulty in catching
these large active fish or the adults may have moved into the
sampling sites only infrequently and spent most of their time
elsewhere. It appeared that juveniles only spent their
initial months in the sampling areas since population structure
diagrams showed a gradual decrease in numbers of juveniles until
only a few were being caught during the dry season. Immature
fish over 30 to 40 mmwere not represented in the samples in
the numbers expected (Fi g . 48 ). As a result of thi s it was not
possible to identify moving modes of cohorts long enough to be
able to draw complete growth curves or to relate size to age
accurately.
Nevertheless, using three cohorts of juveniles (Fig.48 ,
July 1980 (I), May 1981 (II) and May 1982 (III)) it was possible
to draw a partial growth curve by assuming spawning with the
first heavy rains and coinciding modal lengths of 18 mmfor all
three cohorts and making backward and forward projections
from this point (Fig.49). This method is subject to high
levels of error because of the inherent variability in growth
rates due to environmental conditions. However, from a general
extrapolation of this curve it could be predicted that maturity
might be attained at approximately 16 months and maximum
242
1
N = 4 20
M
°1------==--=------II1II 10 N = 13
J N = 0 J
Ol-r----~ar::z::'-----------.
J N = 19 A N = 2
N = 4
o N = 36
N = 68
o
~A N N = 19
o N = 2
s N = 30 J N = 12
o N = 15 F N = 4
N = 2 M N = 3N
D N = 0 A N = 1
J _==.:..:.N. _= __9.; IU+-_...cz::::l<QIZ;~=__=_ _
N
~ N -- 52
F N = 0 ~
M N = 0
N = 0 M
J N = 0
N = 23
M J
N = 7
A
J N = 16 N = 15
o 10 20 30 1,0 50 60 70 80
Standard length (mm)
FIGURE 48: Monthly variation of population structure and the
occurrence of gonad maturity stages for ~. bimaculatus
(key as for Figure 36; I-III: cohorts used in
growth analyses).
E v Z
244
lifespans would be in the order of three years.
Population structure and reproductive seasonality:
Juveniles entered the population only in the rainy seasons
and predominantly in the main rainy season beginning May/June
each year. Recognisable cohorts of juveniles appeared in
catches in July and August 1980, May 1981 and May 1982 .(Fig.48 )
and indicate spawning occurring immediately upon the first
rains. A minor cohort of juveniles may be recognised in
January 1982 again coincident with heavy rains after a
pronounced petit careme in late 1981.
The occurrence of mature adults was noted in May 1980,
May 1981 and October/November 1981 (Fig. 48, Appendix l1(a))
and supports the evidence for early rainy season spawning.
Mean GSIls increased twice each year in Mayand November
(Table 26, Appendix l1(b)) when years were combined. However,
high GSI I S in November were only seen in 1981 for reasons
stated above. Maximum GSI for individuals collected monthly
was 4.49% for females and 1.447% for males both being recorded
in May. GSI values in the second rainy period of the year
were less than those for the main rainy season. Additional
females with well developed gonads were collected in June 1985
C J 3 :
3 :
3 :
: z
0 ~
0
P J
P J
( ) P J
C
0
- I
: : : s
: : : s
- s
u : : l
r T
' <
<
~
r T
c
n
( ! )
c
0
c : c
: : : T
P J
: J
I . f )
0 -
3
r -
e - C D
r T
" ,
( ! ) ~
" " ' S
- S
N
0 '
3 :
0
N
0 0
0
C D
.
.
P J
N
N C O
O ' l
1 . 0
. . . . . . . . .
3 :
: : : s
. . . . .
. . . . . . N
. . . . . . .
W
0
«
. . . . . . .
W
" " - J U 1
W
C D : : : s
" ' T l
P J
r T
C D
: : : r
" " ' S
3
~
I . f )
P J
~
' <
. . . . .
0
0 0
( )
C D
0
<
. . . . . .
. . . . .
. . . . .
. . . . .
P J
( j )
: 3
; ; 0
1 . . 0
. . . . . W
N
C , / ) C ' "
" " ' S
P J
( ) 1
W
" " - J 0
. . . . . .
. . . . . .
. . . . . .
: : : s
I
I I
I
P J
: : : s
u : : l
. j : : : o
. . . . .
0
W
. . . . . . . .
C D r T
. C D
. . . . . .
0 -
. j : : : o
~
. . . . . N
" " - J
. . . . . . . . . . . . . . -
0
. j : : : o
0 ' \
1 . 0
e x >
: : : s
. . . . . .
0 ' \ 0
0
0
- I )
( j )
. . . . . . : z
N
N 0
N 0 '
C , / )
. . . . . .
I I )
. . . . .
. . . . . .
0
0 0 3 :
- I )
.
C D
. j : : : o
P J 0
1 . 0 < . . T I c . . n
0
. j : : : o
. . . . . .
: : : s " " ' S
0 ' \ N C O
. . . . . .
0 ' \ C O
" " - J c . . n
~
.
3 :
P J
~
o -
~ .
C D
0 0 0
.
( j )
N
" ' . J e x > ~
. j : : : o
C , / )
; ; 0
N Q
C O
. . . . .
. . . . . . .
w 0 P J
~
I I I
: : : s ~
. . . . . . . .
. . . . .
u : : l
0 0
P
C D
~
~
. . . . . .
1 . 0 . . . . . . .
~
~
0 )
0 1 . 0
- 1 : : : 0
. j : : : o
~
N
. . . . . . . . . .
W
N 0
N
: z
5 1 7 2
246
and June 1986 and GSI's of these specimens ranged from 6.77 to
14.80% (fresh weights).
Numbers of mature-sized fish collected monthly were too
few to see any consistent trends or to make any definite
statements about sex ratios (Appendix l1(b)). There appeared
to be a predominance of females during times of breeding. For
the entire population (including immature individuals) caught
over the whole study period, the male : female sex ratio was
64 36% and showed a highly significant deviation from a
50 50 ratio (Table 25, x2 test, p<O.OOl). However, sex
ratios for mature-sized individuals showed no significant
difference from unity (p>0.05).
Trends in variation of condition factors v.erenot very
clear owing to the lack of data (Table 27, Appendix 11(c)).
However, it appeared that female total and somatic condition
increased in the latter half of the year after a minimum in
May and were maximal in August. Male condition was maximal
in January and showed minimal values during the breeding period.
In summary, A. bimacuZatus appeared to breed after a
relatively long period of growth up to two years. Breeding
took place at the very beginning of the main rainy season in
247
TABLE27: Monthly variation of mean total and somatic
condition factors for A. bimaculatus (years
combi ned) .
Females Males
Month CFT(%) CFS(%) N CFT(%) CFS(%) N
January 1.387 1.384 2 1.471 1.464 2
March 0 1.219 1. 212 1
May 1.378 1.337 2 1.123 1.106 1
August 1.575 1.572 1 0
October 1.444 1.435 6 1.379 1.369 4
November 1.449 1.436 2 1.347 1.332 2
248
May or June and may not have lasted longer than one or two
months judging from the synchronised appearance Of juveniles
and mature individuals in the population at these times. The
major reproductive period seemed to be during the main rainy
season and under certain conditions a second, minor breeding
period might have occurred later in the year coincident with
the second rains. Each individual might therefore have had at
least one, possibly two, active periods of reproduction each
year giving a total of at least two to three active breeding
periods during its lifetime.
Hemigrammus uniZineatus
Sexual maturation and growth:
The majority of specimens of this species which were
collected were immature individuals (Fig. 50). Minimum lengths
for developing fish were 20 mm for males and 23 mm for females
(Table 24). Females also became sexually mature at a larger
size than males: 25 mm as compared to 21 mm for males. Median
length at first maturity was 26 mm for males and 29 mm for
females. Of the females of mature size, 50% possessed mature
gon~ds while 17% had developing ovaries, whereas of the mature-
sized males, only 33% had mature gonads and 40% were developing.
By 26 mm in length all males had either developing or mature
testes but many large females (up to 35 mm) were still immature.
This suggested either delayed reproduction in females or they
I "
> - <
G l
C
; ; 0
f i s h
o f
N u m b e r
r r 1
( j \
J - -
0 ' \
0
J - -
0
U l t v
0
0
0
0 0
, . . . .
Z
0
U 1
Z
I I
C T
I I
- 1
O l
( J I
O J : : : : : r
N
( ] \
r o
0
N
- - t > - s
r o r o
: : 3
O J O J
- '
M
r o
0
V l
0
: : : : l
V l
O J
: : : : : r
: : : : l
C L
- 0
' - ' .
C
0 -
r o
<
r o
M
: : : : l
: E :
. . . . . .
r o
U i
- '
r o
r o : : : : l
V l
L C l
0
O J
: : : : l
O J
: : : : l
C L C L
3
I f )
O J
0 -
. -
M
~
e
c
: : : J
- S
: : 3
. . . . . .
a .
O J
M e
. . ,
r o
' <
a .
V l
V l
M
O J
A
I I >
L C l
r o
: : : J
r o
' <
t v
1 0
V l
U 1
O J
; ; T
V l O J
: : : : l
- t ,
C L
-
0 3
V l
- s
3
N
r o
"
L C l
c
- - t >
0
~
- s
r o
- s
w I I
U l
C
: : : : l
. . . . . .
- '
W
: : : : l
1 T I
I ~
6 i 7 Z
250
reproduce more than once during their lifetime. Ten spent
females were caught during the study period, all between the
lengths 29 to 35 mm, but many partially spent females were
noted and considered still capable of spawning.
Five cohorts of juveniles were followed in order to draw
a growth curve for males and females together (Fig.51 , July 1980
(I), December 1980 (II), May 1981 (III), September 1981 (IV)
and January 1982 (V)). It was assumed that sexual maturity
occurred within 12 months at 26.5 mm for both sexes. Parameters
of the von Bertalanffy growth model were estimated and the
following equation resulted:
lt = 37.23 ( 1 _ e-O.177(t-3.802))
From this curve it was estimated that minimum age at maturity
for males was nine months and females, 11 months (Fig.52 ).
Development of gonads would have begun at about eight months
in males and 10 months in females. Maximum lifespan was
estimated to be about 16 months in males, 20 months for females
assuming that both sexes had equal growth rates.
Population structure and reproductive seasonality:
Population structure diagrams for each month of the study
period indicated the presence of juveniles in both rainy seasons
of each year (Fig.51). Small immature individuals were caught
N=44 N=57
J M
o 8 16 24 32 l,() N= 4
N= 4
A N=10
o 8 16 24 32 40
Standard length (mm)
FIGURE 51: Monthly variation of population structure and the
occurrence of gonad maturity stages for !!. uni 1ineatus
(key as for Figure 36; I-V: cohorts used in growth
analyses) .
G l
<
. . ,
S t a n d a r d l e n g t h ( r n r n )
0
~ 0
: : : E
C J : l
M -
r o : : : : : l " "
- s
n
M -
O J e
- s
O J
<
r o
~
- - t ,
- - t , - - t ,
0
' <
. . ,
r o
- D
I I
e
O J
e
M -
: : l
. . . . . .
0
0
- - '
~
- "
~
0
r o
. . .
»
"
1 0
W
I ~ n >
: ' I
N
. . . . . .
w
: 3
3
O J
0
-
- "
- - '
: J
r o
. . . . .
( / )
: : : : r C D
I
V l
. . . . . . .
9
- "
' - o J
' - o J
N
0
. . . .
w
C D
( / )
C D
0
N
n
o
: 3
C T
~
r o
[ >
0
0 < ] 0
0 .
•
( " )
z
0
0
C D
n
. . . .
o ,
: : : T
e
. . ,
0
o ·
. . ,
3
<
. . . .
: J
0
r o
. . . . .
c
. . -
- - t ,
. . , . . -
r D -
< < :
-
: : l
. . . . .
M -
1 0
M - " <
. . . . .
r o
: : : T
0 .
M -
o
z s z
253
in July to September, and December 1980/January 1981, January,
March, May to July, and September 1981 and January to March and
July 1982. Relatively large numbers of juveniles were caught
in August 1980 and January 1982 indicating particularly
successful reproduction during these seasons. Considering the
growth rates of these fish and the sizes of juveniles caught, it
appeared that spawning took place immediately upon the first
rains of each season. Small individuals were caught over varying.
periods of time, from three consecutive months in 1980 to over
four months in 1981, indicating breeding seasons of these lengths.
Mature adults were found in every month of the study period
(Fig. 51) although in certain months no mature females were
present (November 1980, January, February and June 1981) and in
others no mature males (January and May 1982, Fig. 53). When all
years were combined (Fig. 54) the proportion of mature females
was seen to be highest in June/August and October/November (more
than 40%). Spent females were noted in February, March, May,
August and December indicating that spawning could have occurred
at any time throughout the year although not necessarily in any
one year. Mature males became progressively more dominant from
January to July when 54% of males of Minimum Developing SL were
reproductively active. Proportions of mature males gradually
decreased until September then increased in October and November
again. It was noted that proportions of males and females with
254
30 (0 )
10
s:
III
-
- 00
....
Cb (b)
.£:l
§ 30
z
20
10
o MJ JASONDJFMAMJJASONDJFMAMJJA
___ 1980 1981 1982 ---
FIGURE 53: Monthly variation of gonad maturity stages for
H. unilineatus (a) females, and (b) males over the
whole study period (key as for Figure 35).
255
(a)
100 14 38 23 50 43 22 20 22 24 11 28IT'"TTT"'TT"rrT"rrrT"TTTTTTT"l"'TIrrT"TTTT"TTTTT"TTTTT"l"'TIrTT1"'TTT~""TT""',.::n.T"1
(b)
80
60
40
20
o...JL.....L.......IL.-.L.....JIL-<:~-L-~....L-.JL.......L.....I'-L-...J<:--L.-..I<:.-.L......J<:..-L.-l...-_~L....J
J F M A M J J AS 0 N o
FIGURE 54: Monthly variation of gonad maturity stages for
------
li. unilineat~~ (a) females, and (b) males over the
calendar year (key as for Figure 38).
256
developing gonads increased immediately prior to the breeding
seasons (Fig. 54) and substantial numbers of fish greater than
Minimum Developing SL but with immature gonads, were present
during the reproductive season (Fig. 53). From population
structure diagrams it appeared that these immature fish might
then mature and spawn by the following breeding season. Thus
it seemed that juveniles spawned in one breeding season might
grow to MinimumDeveloping SL by the following season when some
larger individuals might spawn, but most others might not spawn
until early in the following season at approximately one year
of age.
GSI's supported the evidence for two breeding periods per
year with the May to August rainy season being the major period
of reproduction (Fig. 55). In addition, these data showed that
this pattern was very variable and subject to environmental
modification. In 1980 highest GSI's were recorded for
females from May to October and minor increases occurred in
December 1980. However, in 1981 high female GSI IS were
evident from as early as April (with the first rains of
the season) to December and then rose again from March to June
1982. Such a pattern could be explained largely by the somewhat
unusual rainfall regime for 1981/1982. The trends in variation
of male GSIls however, indicated a relatively regular cyclical
( ' I . )
G S I
' I .
. . . . .
. . . .
Q
n
C D
I " -
U ' l
0 -
0
8
0
I
I
. "
G } I - - - C £ }
C 5
- -
~ ~ ~ a
c
: ; 0
I ' T 1
U ' I
t - - - - - - '
•
[ I I
I ~
L . ~ " S C l ~ *
< J 1
< J 1
. .
0 \
~
~
~
L . ~ ~ a
U ' I
I : : I : 3 :
0 0
•
.
I - - - - C m - - - - l
0
. . . l . > ~ ~ ( J \ i
~
\ D
C r t
( X )
0 \
~ : : : r
z - .
. . . - r
~
. . . . . . . . . . . . •
O l / ) ~ ' \ ' \ [ J ~ a
. . . '. <. .
; : : :
. .
~ <
I - - C I J
U ' I
O ~ ~ i
( 1 ) ~
•
~ " " " )
r t . . . . . .
N
C ~
C I : J
Z ~ ~ ' I a
•
V l r t
. . . . . .
0
. .
W
I - - - - { I ) - - - - - l H I I
N
O ~
c ~ ~ ~ a
<
( 1 ) 0
. . . .
" " " ) - + >
0
N
. . . . . . . .
~
L . ~
I ~ a
r t G }
: : : r V ' l
0 \
( 1 ) - -
0 . . .
l o o a
' T I ~
-
~ V l
. . . .
: : : r
0
l J ' I
~
o - + >
I - + - - f
I ~ i j l
~ ~
. . . .0 . .
( 1 ) " " " )
W
1 - 0
V l . . . . . . . .
H I H
l > ~ ' I i j l
r t ~
c -
-
o .
~
J - - - - [ I J - - - - - - l
~
•
~ & \ ~ i D i j l
' < - + >
( 1 )
- 0 3
w
( 1 ) ~
l J ' I
t - - e : : I J - - i
. . . L. . ~ ~ a
" " " ) . . . . . .
•
\ D
. . . .( .1 .)
( X )
O V l
. . . .
. . . .
0 \ I c : : : : : J C : J
U ' l
0 . "
t - - - - t I : J - - - I
L . ~
I ~ a
. . . . . . . . ~
, , ~
. . . .
. . . .
W
I . . . . : : J
t - - - - f
( 1 ) 0 .
I C I J - I
0 ' 1 1 1
l > ~
•
< c
~ o -
I
0 \
z - .
H
t - - C I J - l
•
l / ) ~ I w i j l
V l -
- + > 3
. . . .
O ~
I - f ] ( ] - l w
Z '
" " " ) . . . . . .
O ~
•
I : : : j *
( 1 )
. . . .
" V l
. . . . . . . .
I . . . : . : J
l J ' I
C I J
I - - - C l I : J - - - - - i N
' 1 1 1 1
Z ~
l C
C ~
" " " ) ~
I . . . . : : J I D
0 \
( 1 ) 0 . ~
~
U ' l l l l
c ~
I
w . . . . . . . .
i
!
< . 0 ( " )
. . . . .
I W : : J
~
~
I - - - C I J - - - - - l
. . . . . . - - . -
L . ~
Z ' I I I
V I
( 1 )
Z '
N
C I J ~
' T I ~ a l i j l
X
- s
N
( X )
~
I - - - - - { ] [ J
I - - - C I I : r - - i
0 \
~ ~ ~ i j l
r t
. . . . . .
0
. . . .
N
t - - - { ] I ] - - - l
V I
. . . .
t - - - - C l D
. . . l . > ~ ~ i j l 0
1 . 0
- + >
( X )
0
~
I - - - - i W J
N ~ ~ l ~ i j l
- s ~
I d
N
C I J
I - ~ Z ' i j l N
•
0
H
! - - l
w
I W i j l
1 - 1
•
I
w
H
~
I I I
l o i j l
l > ~
•
L S Z
258
pattern with peaks in May and December 1981 and May 1982.
Fluctuation of female GSI's showed the regular occurrence
of an initially high GSI at the beginning of the breeding
season followed by a decrease and a further increase to previous
1eve 1s . The decrease of GSI in the middl e of the breedi ng
season was evident in July 1980, June and September 1981 and
July 1982 (Fig.55). This pattern was accompanied by a decrease
in the numbers of mature females as well as either in the same
or previous month, a loss of the larger females in the popula-
tion (Fig. 51). While these features may be expected at the
end of the breeding period, at this time it suggested the
presence of two groups of females reproducing successively
during anyone breeding season. It was likely that larger
females spawned at the very beginning of the breeding season
with the onset of the first rains, then probably die, followed
by a second group of smaller females which spawned later in
the season.
MaximumGSI's recorded for female and male H. unilineatus
were 13.101% in May 1981 and 2.083% in June 1980 for females
and males respectively. Female GSI's were generally high with
GSI I S greater than 10% being recorded for nine out of the 28
months of the study, again suggesting fairly prolonged
259
reproduction or continuous spawning-readiness for females.
Sex ratios for individuals caught over the study period
showed no significant variation from a 50 : 50 ratio whether
all individuals or only mature-sized fish were considered
(Table 25). However, mature-sized fish showed a slight predomi-
nance of females over males and x2 value was high (3.03,
p<O.1). Monthly sex ratios showed no significant variation
from unity except in June 1980 (male : female ratio 76 : 24%,
p<O.05) and October 1981 (male: female ratio 18 : 82%,
p<O.05). Slightly skewed ratios were found for some other
months (September, November 1980, November 1981, May, July
1982). In addition, monthly variation in sex ratios
appeared to follow a cyclical pattern of higher proportions
of males during breeding seasons (Fig. 55). These trends were
not entirely clear-cut or consistent however.
Both male and female condition factors showed a very
close correlation with the reproductive cycle (Fig. 56).
Female total condition increased relative to somatic condition
immediately prior to and during the breeding seasons as GSI's
increased. Both total and somatic condition decreased
drastically at the same time that GSI's decreased in the
middle of the breeding periods. Male condition showed a
f a c t o r
C o n d i t i o n
N
6 -
( ; x )
0
e x ,
s r
: r - . C 1 ' l
N
I
' "
,
I
I
,
. . , . ,
. . . . . . .
- -
- -
( J 1
G l
- -
3 :
c :
: ; 0
( J 1
r n
e n
l -
U 1
0 ' \
e n
I D
l -
. . . . .
~ >
( J l
- -
-
~
. . . . . 3. . : . .
. /
- -
- - »
e r o
- -
/ '
~
I D
- : : l
e n
c e
M -
f o e "
3 : : 5 " O l f )
l '
- -
0 ) - - - '
I
- - - ' ' <
U i
r o
H I
N
V l
0
<
0
\
0 )
0 - s
<
e n
0 )
r o
Z
M -
- s
. . . . . .
W
N
M - O
N
: : 5 " : : l C J
r o
0
0
~
: E : - I >
l -
: : 5 "
O M -
e n
- - - ' 0
N
r o M -
/
0 )
"
V l - - - '
0
H
M -
~
e 0 )
3 :
I
o . : : l
, < 0 .
W
~
l '
»
- 0 V l
I
r o o
- s 3
~
. . . . . . 0 )
~
N
- -
3 :
O M -
. /
0 . . . . . . .
. . . . . . . :
n
~
_ _ l -
I D
- -
" , n
I D
r o 0
" "
- "
( X )
' < : : l
C D
w
0 .
- -
- - l -
0 ) . . . . . .
V l M -
- "
( J l
- -
- 1 > 0
»
O : : l
- s
\ 0
- I > l '
I f )
. . , . , 0
. . . . . . . . . . ,
- "
( , 0
l '
w
e I : r :
0
. . . . , .
r o
- "
e
N
( J l
. j : : : o : : l
Z
0 . . . · . .
- - - '
\ 0
e n
: : l
C J
r o
0 )
"
M -
~
~
l -
"
: & : :
e
~
V l
~
\ 0
~
0 )
\
"
" ' =
N
. . . .
~
C D
~
- I >
3 :
r o
/
I .
3
N
0 )
- "
~ ,
- >
~
- - »
r o ~
. . . . . . . .
I D
V l
w
~
( X )
N
i l ' S
Y .
N 3 :
. . . . . . "
, / "
0 )
~
. /
. . . . e : :
: : : 5
. . . . e : :
. . /
N
0 .
t v
l -
, ,
H
" " -
, ,
. . . . . . .
. -
, ,
, ,
w
0
' ,
, ,
-
l -
, ,
. . . ~. . . .
~
w
~
»
0 9 2
261
slightly different pattern of fluctuations with increases
prior to breeding and decreases during the breeding period;
these fluctuations were very regular and cyclical over most
of the study period.
In summary, H. uniZineatus had two prolonged breeding
periods each year coincident with the rainy seasons. These
were quite regular in males but in females they were more
variable in response to environmental changes. It was possible
that some females which were spawned early in one season matured
quickly enough to spawn in the following breeding period but
most seemed to mature at the next after that (that is at one
year old). These latter individuals most likely spawned
immediately with the onset of the rains. It seemed that a
second group of smaller females might have then matured and
spawned later in the same season and possibly in the following
season as well. A female might therefore have one, possibly
two, periods of active reproduction in her lifetime and there
appeared to be two overlapping generations of females in the
population at anyone time. Males probably had only one active
breeding period in their lifetime judging by their small size,
short 1ifespan, early maturity and lack of large individuals
with inactive or immature gonads in the population.
262
Corudoi-ae aeneus
Sexual maturation and growth:
Unlike some of the other species studied~ most specimens
of this species were reproductively active with only a few inactive
or immature individuals collected (Fig.57). Minimum size of
specimens with developing gonads was 34 mm for females while
males started developing at a smaller istze , 31 mm (Table 24).
However, both males and females attained maturity at equal
minimum lengths~ 34 mm. Most males greater than this length
were mature (70%) and 72% of mature-sized females were also
mature or spent. Several large females with inactive or
developing gonads were caught and could have indicated repeated
spawning for females over their lifetime. Data for males
showed this trend more clearly with two peaks in the numbers of
males with developing testes: one at a modal length of 34 lM1
and another at 44mm. This could have suggested that males
undergo at least two reproductive cycles during their lifetime.
Since females grew to much larger sizes than males and
presumably live longer, it was not unreasonable to suggest that
they may undergo more than two reproductive cycles during their
lifetime. The majority of males were mature at a smaller size
than females and median length at first maturity was 35 mm fqr
males and 38 mm for females. Many partially spent females were
noted but classed as mature due to the number of ova still
f i s h
N u m b e r o f
I V
0
< 5
0
0
0
" " T l
l : 5
Z - ; ; :
Z o
G J
. . . . . .
C
I I
"
; : J : J
. . . . . .
- "
f T 1 l '
- "
( J 1
0 '
( J l
' J
O J
- 1
: : : l
: : : l " "
0 . . .
r o
. ,
c . . . . . •
c
r o
<
r o O J
: : : l
r t
- ' .
. . . . . . . .
0
r o
: : s
V 1
V l
: : : l " "
O J
- 0
: : : l
0 . . .
0 -
r o
r t
e r
: : £
r o
r o
: 3 : : s
O J
c o
r o
0
V 1
: : s
O J
0 . . .
A
r o
: 3
O J
" <
r t
O J
C
V 1
- s
- + ,
r t
0
' <
- s
V l
. ,
r t
O J
c . . o
c o
C
r o
. ,
V l
r o
O J
w
: : s
U " l
0 . . .
V l
N
r o
- + >
0
- s
I n
I ~
I ~
O J
- + >
r o
: 3
O J
r o
V l
E 9 Z
264
present; 14 fully spent females between 45 and 57 mm were noted.
Insufficient data were obtained to define cohorts or
moving modes in the monthly analyses of population structure
of this species (Fig.58 ). Therefore it was not possible to
make any definite conclusions as to growth rates or to correlate
size wi th age.
Population structure and reproductive seasonality:
No large cohorts of juveniles were recorded for this
species. The smallest individuals caught were 24 mm SL and
these occurred in catches, in September 1980 and March 1982
(Fig.58). It may be assumed that juveniles of this species
were either not being caught with the methods used or were not
available for capture. Even the adult population showed
inconsistent capture rates with either entire schools or few
if any fish being caught.
Mature individuals were caught in 19 out of 23 monthly
collections. Months when only developing or immature specimens
were caught were December 1980, February 1981, January and
June 1982 (Fig.58). Spent females appeared in May, September
and October 1980, January, July, October and November 1981,
indicating prolonged breeding seasons spanning both rainy

266
periods of the year. Proportions of females which were
mature or spent were somewhat irregular due to small sample
sizes and differences between years when data were combined
(Fig. 59, Appendix 12(a)). However, highest proportions of
mature or spent females (100%) were found in March, April and
September. Fairly high proportions were also found in
February, August and November. The highest percentage of
inact ive females was in December. The pattern for males was
clearer and if small samples in April and August/September
were not considered, proportions of mature males increased
from March to a maximum in July (100%) and decreased there-
after with a second peak in October.
If data for all years were combined, the pattern of GSI
variation for females was one of gradual increase in GSI
from January to a high level in April to July followed by a
further increase in September/October and a decrease at the
end of the year (Fig.60 ). However, this pattern was largely
due to the pooling of data for consecutive years and in fact a
fairly clear seasonal cycle emerged when the whole study
period was looked at. Unfortunately, because of small or
missing samples in many months the data. are incomplete
(Appendix 12(b)). Nevertheless, these data showed clear peaks
in female GSIls from June to September 1980, April to June and
267
100 7 8 8 19 E 14 5 7 14 4 5
L\~ (0)

 ~ ~~
~ 
 ~~
8
o-~
~ ~
-~ '"
60-[\
-e - t/jo
. 40J- ~
~[/;
20J- /;~
V; ~V~~[7;o j
4 12 4 42 42 12 2 3 14 4 2
100-,--.------rTTTTr-TTTTn-ITTTTTTI-r.,...::....,-n;.Trr~,-.::--r---=--.
( b)
80-
60-
-e
o
40-
FIGURE 59: Monthly variation of gonad maturity stages for
f. aeneus (a) females, and (b) males (key as for
Figure 38).
268
(a) 7 8 8 19 16 14 5 7 14 5
20
16
12 •
~ 8
o
•
0
(b) '- 12 3 0 32 40 12 2 3 14 '- 2
, t ... f J ! II , ~ j •
0
(c) ns ns ns ns * ** ns ns ns ns ns ns
11 20 11 1 55 56 26 7 10 28 5 7
100j
-e 50
0
~ ~ ~ ~ ~ ~ ~ ~ ~0 ~ ~ ~
J F M A M J
J A 5 0 N D
60: Monthly variation of GSI's for (a) females,
and
FIGURE
I
(b) males, and (c) sex ratios for c. aeneus-
(key as for Figure 39).
269
October to December 1981 and March to May 1982. A minor peak
occurred in January 1981. This indicated prolonged breeding
periods of up to four months associated with the two rainy seasons
each year. Increase in GSI was also consistently associated
with the presence of spent females. Pooled GSI data for all
years did not give a clear picture of breeding in this species
because of differences in timing of gonadal maturation in each
year as a response to variable onset and duration of the rainy
seasons as well as the confounding effect of prolonged breeding
periods. Generally there was one major breeding season each
year in the main rainy season and a second occurred depending on
environmental conditions. During the first year of the study,
the December/January breeding period was minor but in 1981 the
first breeding period started with the early rains in April
and the second from October to December, was of equal magnitude
to the first.
Male GSI variation was similar to that for females with
two reproductive periods possible each year and coinciding with
the rainy seasons. The major increase in GSI was in May/June
each year with the second peak of variable magnitude and
timing. Maximum male GSI recorded was 1.144% in May 1980 and
the female maximum was 21.169% in October 1981.
270
Sex ratios for specimens collected during the study were
slightly dominated by males (Table 25) but this was not
significant (x2 test, p>0.05). Monthly variation in sex ratios
indicated significantly higher proportions of males during
May and June when breeding was at a peak (p<O.05 (May), p<O.Ol
(June». For the other months sex ratios did not differ
significantly from unity (Fig. 60). Aquarium breeding showed
that more than one male may be necessary to fertilise all the
eggs of one female (Sterba 1962, Zukal 1982) and this could
have accounted for such male-biased ratios.
The variation in condition factors for males and females
was remarkably cyclical (Fig. 61, Appendix 12(c». Total
condition for females was much greater than somatic condition
in all months of the year consistent with continuously high
GSI's when data for all years were combined. Both total and
somatic condition of females increased along with increasing
GSI's at times of breeding and generally decreased after each
breeding period. Male condition tended to decrease at the time
of breeding and increased immediately afterwards.
In summary, C. aeneUs appeared to have two breeding periods
each year coincident with the rainy seasons. The first period
during the main rains was generally the major breeding time while
271
(0 )
5.i, 7 8 8 1 19 16 14 5 7 14 5
5.0
;1
4.6 !'-' J', 1--r--!--r--I' /<, '1.......!././,
<,.//
)/
L..
.0.... 4.2
u
..0..
c (b)
0
..... 12 3 0 36 40 12 2 3 14 4 24
-0
C
0
u 5.4
5.0
4.6
4.2-L.~_---.-_--.-----.-----r-----.-----r-----r-----r---r---r---r-
F M A M J J A 5 0
N o
J
Monthly variation of total and somatic condition
FIGURE 61:
for C. aeneus (a) females, and (b) males (key as
for Figure 40).
272
the second breeding period was of variable magnitude and timing.
Both were for prolonged periods up to four months and mature
individuals were found throughout the year. It was not clear
whether an individual could have one or two active breeding
periods each year. Population structure diagrams showed that
1arge females sp-awned in consecuti ve breeding seasons but
without more data it could not be ascertained whether these were
the same individuals or not. If they were the same individuals,
then a female might have a lifespan of about three years based
on sexual maturation in two years (Schofield 1957) and three
reproductive cycles per 1ifetime. If individuals spawned only
once each year then 1i fespans of more than four years might be
possible.
PoeciZia reticutata
Sexual maturation and growth:
Sexual development in this species began at a very small
size in both sexes, namely 9 mm(Table 24, Fig. 62). Maturation
occurred quickly particularly in males and minimummature size
was 9 mm for males and 11 mmfor females. The majority of
males matured at a smaller size than females and the median
length at first maturity was 10.8 mmfor males as compared to
12.4 mm for females. Some males of quite large size (14 mn)
were recorded as developing on the basis of gross characteristics
of the testes as well as incomplete development of the secondary
273
80
40 ;;1
~~
. 111111111111 ...1••••__- -
~ 200 (b)
'0 N: 503
lo...
<11
.D
E 160
:;,
z
120
80
40
5 10 15 20 25
Standard length (mm)
The relationship between gonad maturity stages and
FIGURE 62:
size for P. reticulata (a) females and juveniles, and
(b) males (key as for Figure 35).
274
sexual characters. As many as 24% of the males of mature size
were classed as either immature or developing while 20% of
mature-sized females contained developing ovaries. A range of
females up to very large sizes (23 mm) were noted to have
. developing ovaries and this suggested repeat reproduction
throughout life in this species.
Monthly analyses of population structure did not provide
data for estimation of growth rates since no distinct cohorts
could be recognised or followed (Fig. 63). Laboratory-kept
individuals however showed the development of secondary sexual
characters in males by 4 weeks of age and females produced their
first broods when 11 to 14 weeks old. No estimates of maximum
lifespans were obtained.
Population structure and reproductive seasonality:
Population structure in this species showed changes which
correlated with fluctuations .of population size rather than the
reproductive cycle (Fig. 63). Most immature or developing
individuals were caught at times when population sizes were
increasing or at their highest levels. At times of low
population sizes, most of the individuals were mature and quite
large; very few if any, small immature individuals were present.
Population sizes were seen to fluctuate in response to stream
40 (a) (b)
N=163 40 (a) (b) 275
N=142
20
20
M
0 1L ~Scom~ 0
N=32
J
0
oj - ~
eo N
~0\ 30 N=lOl D
20
~
L N=140A 0
5
~
0 N= 5
J
= i
N
N=20 ~
D
~
F
J 1L N=154
N=150
N
co
~m
M
F
~ L
~M N=16 JIlL,
A N= 2
M N= 2 A
J A M ~
J N= 1 J N= 0
A N= 0 J N= 3
i i i i , i i I
4 8 12 16 20 24 8 12 16 4 8 12 16 20 24 8 12 16
Standard length (mm)
FIGURE 63: Monthly variation of population structure and the
occurrence of gonad maturity stages for f. reticulata
(a) females and juveniles, and (b) males (key as for
Figure 36).
276
flow regime and large numbers of immature and developing fish
were present as total numbers of individuals increased between
flood events. For exarnp l e , large population sizes with
significant proportions of immature fish were seen in May~
August 1980, February, September 1981 and January to Apri 1 1982
(Fig. 63). At these times females and males attained their
largest sizes. In other months when discharge was high or
immediately after floods had taken pl ace, small specimens were
absent and mainly adults were cauqht , for example November 1980,
April, May, July, October , November 1981 and July 1982. Small
fi sh were either washed away or had moved into more protected
sites away from the areas sampled. Generally however~ small
immature fi sh were found in most months of the year i ndi eating
that reproduction took place throughout the year.
This conclusion was supported by the presence of mature
males and females bearing ova or embryos in all months during
the study period when fish were collected (Fig. 63). If data
for all years were comb ined , the proportion of mature females
greater than Minimum Developing SL increased in May to July and
again in October/November. Males showed a similar trend
(Fig. 64). Conversely, the proportion of developing and
immature females increased (up to 40%) in March~ August and
December and for males in May, September and December (up to
277
100Trln13lTTnl~97mn96m-ri1T25mT1L.TnOlT~~61~211Tl~4r~rr88~7~~22~r!15~.j2~9~
(a)
80-
60-
-oe
40-
20-
o V / r//f//1t-....I-./.../..r-'""'''''''
72 67
100 63 S2 Sl 28 8 38 78 15 8 22
(b)
80-
601-
-oe
40]-
-
20
v /1/ / ~//V/ t//V/ ~o
J F M A M J J A SON D
FIGURE 64: Monthly variation of gonad maturity stages for
£. reti..f.UJ.~J:a (a) females, and (b) males over the
calendar year (key as for Figure 38).
278
50%). These trends were based on the above-mentioned
variations in population structure with flow regime and may not
indicate any real variation in reproductive activity (Fig. 65).
GSI's were calculated for females only since male gonads
could not be weighed accurately on account of their small size.
Female GSI variation during the study supported the hypothesis
that there were no clear seasonal trends in reproductive
activity in this species (Fig. 66). GSI's were maximal during
the drier periods of the year when populations were increasing
or at their highest (August 1980, February 1981, May 1982) but
GSI's were generally high all year round. MeanGSI was never
1ess than 6% and GSI' s greater than 10% were recorded for
every month that fish were available. Very low GSI values
were noted in months when population sizes were increasing and
there were large numbers of relatively small reproductive
females or large females with developing ovaries. Very high
or decreasing population sizes were not associated with low
GSI's indicating that population decrease was not due to a
fecundity-influenced population control mechanism. Maximum
GSI recorded for females was 25.714% and for 10 males in June/
July 1980, maximum GSI was 11.29%, mean GSI was 5.36% (standard
error 0.95).
279
(0 )
s:
III
- 0-0
'-
cv
.0 (b)
E
:::l
Z
1,0
O...J=Mo::IJ!!!!l!! JASONDJFMAMJJASONDJFMAMJJA
---1980 1981 1982--
FIGURE 65: Monthly variation of gonad maturity stages for
P. reticulata (a) females, and (b) males over the
whole study period (key as for Figure 35).

281
Sex ratios for this species over the study period were
significantly biased towards females whether all or only
mature-sized fish were considered (Table 25, x2 test, p<O.OOl).
However, a significant predominance of mature females was seen
only in some months (Fig. 66, May/June 1980, February 1981,
January to May 1982). These months were generally associated
with high population sizes. In other months males and females
occurred in more or less equal proportions.
Condition factors for both males and females showed no
clear seasonal trends in their variation (Fig.67). Female
total and somatic condition were widely divergent all year
round consistent with continuously high GSI's.
In summary, the data indicate that P. reticulata reproduced
continuously all through the year and its reproduction was
influenced by the impact of flooding on population sizes.
High flow rates decreased fish populations and mayhave.~-
resulted in increased levels of reproductive activity and a
population structure characteristic of growing populations as
numbers built up to previous levels. Major flood events
generally occurred once or twice each year resulting in the
above changes in population size and structure at this frequency.
f a c t o r
C o n d i t i o n
N
N
0
e n
N
0
N
i : r I
0
. . . "
0 -
o ~
. . . . . . .
a
G l
~
c
N
: ; 0
f o e N
r r 1
( J 1
L . /
m
/
' - . J
( X l
. . . . . . . .
r -
c e
L .
. . . . . . .
l ' -
- -
l ' -
w
/ - , .
- -
n
3 :
c o
. . . . . l >
0 - 0
0
. /
U )
: : : l : : : l
0 : > ~ . /
0 - r t
. . . . . U i
: : , -
0 -
O l f )
- . . . I
O J
r t
\
- s
' <
l ' -
V l
0
~
: : : l
<
0
/
O J
- s
r o
- s /
~
- 0
0 N
t - e
c e
O J
- s - s
Z
\
r o r t
. . . . .
V l
1 - 0
. . . . .
. . . .
r o
0
\ D
0
: : : l : : : l
I
- s
r t
N
r o
0
N
- t , C X l
1 + r - t -
L .
N
~ "
. . . . . . .
< ,
n
( J 1
c O J
N
~
C T \
< . n
" T ]
c e
O J
r t /
- t ,
O J
r - t -
O J r o
: : : l
- . . . . : : J
f - i
0 - ~
: 3 0
z
I
O J
0 /
O J
I
. . . ,
<
I
r o r o
0 - N
I
0
H f
l >
I I
- s
I
I
r t
r n
I I
0
r t
- s
I
I
a
: : Y
r t
- s
0
I
3 :
I
O J
r o
0
I
. . . ,
I
I
I
' - - - " W
: £
I
N
O l
N
: : Y O J
~ L .
f e ,
w
/ - O i
,
0 : : : l
U )
,
0 -
0 )
,
, ~
r o
,
0
V l L .
. - , I
,
V l
0
I I
,
r t
: 3 I I
,
O J
C
I
I
,
0
0
,
I
0 - r t
l >
,
I
,
,
' <
I
,
n
,
I
( J 1
- 0
I e
C 3
Q
I f )
( J 1
r o n
. . . ,
0
\
- - ' .
: : : l
0 -
0
~
0 ~
e o
, j
0 -
I
r t
I
I
~ /
W
0
7 ' 0
I
Z
r o : : : l
I
I
' <
I
. . . . .
. . . . .
"
; : .
O J
O J
0 - . . . . : : J
N
V l
: : : l
f
0 -
( J 1
( X l
~
L .
0
w
a
. . . ,
0 -
\
-
w
. . . "
- . . . I
" T ]
l ' -
\ D
: 3
O J •
t o
\
c
( J 1
. . . ,
( X l
I ' D
w
3 :
- . . . I
r o
, /
r t
*
0
~
/
( X l
r t
0
~
I e
~ l >
( J 1
O J
I . D
- '
\
0 )
. . . . .
W
N 3 :
~
~
N
N
\
\
\
\
\
\
L .
\
a
a
\
\
\
\
\
\
\
L .
0
w
. . . . . .
o - t
l > 0
a
Z 8 Z
283
Fecundity and spawning patterns
In order to determine fecundity and spawning patterns of
the six species studied, several different methods were
attempted to provide a complete picture of these aspects of their
reproductive biology. Conventional fecundity estimates based on
the numbers of mature eggs in both ovaries were obtained (batch
fecundity) as well as estimates based on laboratory spawning
experiments (fertility). The latter were expected to be more
realistic estimates of egg production for multiple spawning
fishes. Spawning patterns were also deduced from laboratory
spawning experiments as well as from oocyte size distribution
analyses of developing and mature ovaries. These data could
determine the pattern of ovarian cycling in a limited way.
Ovari an fecundity estimates:
Batch fecundity estimates were obtained for each species
utilising specimens with mature ovaries caught during regular
monthly sampling as well as from special collections made
afterwards for this purpose. The fecundity-length relationship
was determined for all species (except A. bimacul.atue for which
available data were inadequate for this purpose) according to
the equation:
284
logIQF = loglOa + b loglOSl
where F = batch fecundity (or brood size for P. reciculat.a't
and SL = standard length (Figs.68 and 69 ). The constants a
and b (± 95% confidence limits) and the correlation coefficient,
r , are given for each species in Table 28. The 95%confidence
limits of b gave an indication of the spread of the points
around aline of slope b and hence were used to reflect the
variability of fecundity about given lengths for each species.
Determination of batch fecundities for C. »i ieei ,
H. uniZineatus and C. aeneue were particularly difficult because
of the high proportion of partially spent fish in the samples.
In these cases, only ovaries which appeared full and for which
the GSI values were maximal for fish of that length, were used
in analyses. However, this severely limited the numbers of
ovaries used in analyses for C. riisei and C. aeneus in
particular. Such small samples especially over a wide size
range for C. aeneus may have resulted in the low statistical
significance of the parameters of the fecundity-length
relationship (Table 28). For C. aeneue , there was no significant
correlation between lo910F and loglOSL (p>O.05) and the slope
showed no significant variation from zero (F-test, p>O.05).
In addition, the slopes of the lines of best fit were much less
F e c u n d i t y N
~
8
§
l . T l
. . . .
0
0
0
8
N
0
0
" T l
I I
e n
R l
O J
N
U 1
0
o c t
I . D
0
x
0 0
0
O J
0
e n
( )
I
O J
0
W
r i -
. . . . . .
( )
: : : : r
I I
V l
J .
r
- t ,
. . . . . .
C
r o
w
( )
: : : : l
w
C
0
: : : : l
c e
. . . . .
0 -
: : : : l
. . . .
r o
r i -
O J
' <
8
J
r t
- '
C
U 1
r o
: : : : l
( . Q
e n
r i -
o
U 1
: : : : r
8
o
0
- S
l : : S
r o
O J
r i -
V ' l
O
. . . . .
" T l
: : : : l
0
U 1
I I
o
: : : : l
: : : : r
o 0
. . . . .
0 -
N
o
" 0
0
. . . .
. ,
U 1
c e
0 -
W
- t ,
x
0
. . . .
- S
0
e n
1 1 )
I
0
: : : : l
. . . ! f l
\ 0
. . . . .
-
I f )
: : r
r
. . . . . .
J "
c n
3
o o
N
3
8
8 l . T l 0 N
l : : S f : : : : : = = _ _ - - - 1 . . . L - _ - - l . . . . . _ ~ g L _ _ _ L . _ _ _ _ J ' _ _ _ . L . . _ ~ 8 _~_ _ _ ' _ _ J §
- s
. . . . .
o
U 1
0 0 o
I I
r o
o
o
. . . .
-
o
i j )
x
e s
I
N
I f )
r
w
o
w
c o
. . . . .
8 . . . . L - - - - - - - - - ' - - - - - - ~
S 8 Z
286
(0 ) (b)
2000.,----------~ 40r---------r-------.
0
0
1000
00
0
0
500 10
..>...-
"0
C
::l
u 0 5
QI
u, 0
0
0
o
100
F = 0.15(S9 1OO3L) 1.984 F = (1.291x 10-3 )(SU
501+- __ -.-----.---~__r___r----r___._l 1+---O<:>-OO-.-----.....--.----~
20 50 100 10 50
Standard length (mm)
FIGURE 69: (a) Batch fecundity-length relationship for I· aeneus,
and (b) brood fecundity-length relationship for
P. reticulata.
c ; ) V l
: : l Q
- s 0 ' "
: z
.
. . . . ~
~
~
- 0
I I I
* 0
*
*
( D
\ 0
* *
( ' )
C J ; J
. . . . . .
*
' " : :
~
~
§ - I .
( I : l ( I : l
~ . c t -
0
( D
( I : l - - I
l l J ~ .
~ .
~
c t -
l / )
( I : l »
C I . l
~ . ~
O J
( \ )
( ; )
~ .
~
~
r
- 0
n - " 0 ( " " )
" 0 ; : 0 C J ; J ~ .
: z ~ .
: z ~
~
I I ' T 1
( ; )
( I : l
1 \ 1 \ ( D
1 \
0
0 . . . . . . .
C
n - O a . . . . . .
0 t T
t o
- s ~
~
3
. .
( D
N
< ; i -
: : l
~
~ - s 0 ' "
c t -
" "
l / ) ( X )
0 ( D ( D
a ( D
a I I I
t T
~ ~
. . . . .
. . . . . . . . . . . .
l / )
o
n -
C J " I
m
c o
- s
. . . . .
l / )
\ 0
P I - s
. . . . . ,
- I .
- n
: : l
n - 0
. . . . . .
- I .
0
m
0
- f l
- s
0 " 0
- f l : : l
. . . . . .
- 0
0
: : l t T I I I
W
- t ,
0 - ( ' ) P I
n
N Z
- c N
N W
- 1 : = : 0
( D
( ' ) ( D U )
l l J - s
0
1 . 0 - . . . J
W \ . 0
( D
P I
: : l n -
0
<
o
- I . . . . . . ,
( D . . . . . .
~
t T
3
. . . . . ,
. . . . . ,
( l ) ( l )
P I
3
. . . . . ,
- I .
n - " 0
m n -
- I .
. . . . . .
( D
( I )
v n : : l
. . . . .
( ' )
l / )
0
a I - 0
I
I I I ~
.
. . . . . . ( l )
( D l / )
: : l
. j : : : >
. . . . . . . 0
N N
0
( D ( ' )
a t o
: : l
. . . . . .
0 C J " I . . . . .
( X )
n - . . . . . . 0
: : l
1 . 0 C T l
. . . . . ,
. . . . . ,
( l )
n - ( X )
0
\ . 0 1 . 0 C T l - 1 : = : 0
' "
l / )
. . . . . .
Q l
\ . 0 N
c o
0
n -
' "
l / )
O ' "
: : : r
l / )
~
( D
r t
( D
s : :
. . . . . ,
~
0 -
. . . . . .
( l )
. . . . . .
0 . . . . . . . N N
<
( D
.
( l ) ( ' )
0 -
. . . . . . . . . . . . .
N
0 N
0 - s : :
0 U ) ( ) ' I
. . . . . . .
< . T I
0 0
: : s
0 ' "
- . . . . . . . . . U )
\ . 0
1 . 0 W
0 -
. . . . . .
Q l
( D
- . . .
x
x
x X
n -
. 0
~
~
c
0
. . . . . . . . . . . . .
. . . . . . . . . . . .
I
l l J
. . . . . .
3
0
0 a 0
n -
I I
I I
( l )
- I .
W
( ) ' I
W
N
W
: : s
0
c . o
: : l
0
n -
: : : r
0
~
t o
. . . . .
. j : : : > ( ' )
W
W
W
( l )
. . . . . . .
. . . . . .
0
0
0
1 . 0 0
C T l
W
: : s
Q l
. . . . . ,
- n
0
c o w
0 0 0
n -
w ( X ) - I .
- 1 : = : 0 0 0 N
I I
: : l
0 - 0 ' "
0
I I I ( D
*
*
* *
: : s
* l / )
: : s
* * J +
0
( ' )
*
*
: : : r
t o
J + ( D
1 +
J +
1 +
1 +
1 . 0
. . . . . . .
( ) ' I
- 0
0
0 N . . . . . . .
N
~ ~
N
l l J
- I . . . . . . ,
. . . . .
W
U )
~ - . . . J
3
0
. . . . . .
+ . . . . . .
N . . . . . .
0
W
~
0
1 . 0 W
W
n -
O ' "
l / )
. . . . . ,
' "
- I .
<
0
( D
t o
. . . . . . .
0
0
0
0
0
0
V l
.
r
. . . . . .
w
U 1
C J " I
- 1 : = : 0
. . . . . .
. . . . . .
~ . . . . . .
' - J
~
. . . . .
\ . 0
U 1
C T l
0
: : l
l / )
*
*
*
*
*
*
*
*
*
0
0
. . . . . .
0
0
0
1 . 0
0
N
N
c - t -
0 ) -
. . . . .
U 1
. . . . . .
0 0
l / )
. . . . .
1 . 0
W
N
. . . .
: : l
: : l
: : l
: : l
: : s
' "
l / )
V I
V I
V I ( I )
L 8 Z
288
than three in the case of C.aeneus and larger than three for
C. riisei but these differences were not significant (Table 28).
Relatively low statistical significance was seen for the
parameters of the fecundity-length relationship for G. etemicl.a
(r, p<0.05; b , p<0.05). Small sample sizes as well as the use
of ovaries which may have been partially spent or early maturing
may have been contributory in this case.
The values of b for three species (G. eterniol.a, H. uniZineatus
and P. reiriculata) were very close to three indicating a near
cubic relationship between batch fecundity and standard length as
expected. There was no significant deviation of b from 3.0 for
these species (Table 28).
Variability of fecundity values about the line of best fit
was comparatively high in G. etermiol.a, C. riisei and C. aeneus
(95% confidence limits of b > 2.0). This was most likely due
to small sample sizes and the real variability inherent in
ovari es which may have been parti ally spent or early maturing.
Data for A. bimacuZatus were very 1imited and only three
fully mature females were analysed. Batch fecundities were 1948
(57 mm SL), 3075 (76 mmSL) and 6364 (70 mmSL) eggs. Maximum
batch fecundity for this species was much greater than that for
289
any of the other species investigated (Table 29). In contrast,
the ovoviviparous P. reticuZata had the lowest maximumfecundity
at 26 while the other species were intermediate between these
two extremes. Maximum batch fecundities were positively
correl ated with size for the six species (r = 0.827, p<O.05,
N = 6).
Batch fecundities corresponded very closely withGSI for
most species and in four cases, maximum fecundity estimates were
for females whose GSI values were maximal for their respective
species (Table 29).
Laboratory spawning experiments:
Successful spawning in the laboratory was achieved only for
C. x-i-ie ei., H. unilineatus and P. »et-ioul at:a. Other species were
maintained for prolonged periods and although courtship behaviour
was observed inA. irimaculatue and to a 1esser extent G. eterniala ,
no spawning took place. Detailed observations were made of these
species on many occasions at all times of the day and night but
at no time was spawni ng observed. Attempts were made on several
occasions to repl icate flood conditions by replacing tank water
with fresh water but this was unsuccessful. Presumably other
factors may be more relevant to inducing spawning in these
290
TABLE 29: Maximal batch/brood fecundities recorded for the
six fish species studied.
Speci es SL(mm) GSI(%)l Fecundity
G. etiernicl.a 45 12.262 1458
c. riisei 36 14.130 833
A. bimaculatue 70 14.800 6364
H. wz.i Zineatus 37 11.521 787
c. aeneus 56 21.169 1046
P. »etrioulata 19 18.033 26
1 GSI value refers to the same individual for which
fecundity was determined.
291
species in the wild. In the case of A. bimaculatue , one pair
showed intense but sporadic courtship activity over long
periods but spawnirg did not occur. In this case female
receptivity seemed to be low despite the persistent advances on
the part of the male.
Courtship seen in G. eteimiol.a, A. bimaculatue and
H. uniZineatus was simple, consisting of varying combinations of
certain recognisable actions performed by both sexes such as
chasing, circling and swimming parallel to each other in the
same direction for short distances. In A. bimaculatue , after
some period of chasing, the male would assume a stationary
head-down or head-up position in front of or parallel to the
female, meanwhile twitching or erecting the pectoral and all
unpaired fins. The female remained stationary or moved slowly
so that they circled each other. The male would then move away
from the female with exaggerated lateral undulations of the
body in a 'leading' motion. When performing circling movements,
male and female on occasion oriented themselves so as to touch
ventrally. During courtship in A. bimacul.aiue male coloration
intensified. Considerable aggression was seen in the behaviour
of this species and males often had to be removed or separated
from the larger females because of the damage inflicted to
them.
292
H.unilineatus showed courtship behaviour similar to
A. bi.maculatue with exaggerated undul atory 'leading I movements
performed by males. Head-up ori entat i on was performed by either
. sex or both at the same time and was followed by alternating
head-down, head-up orientation creating a vertical zig-zag
movement in the water. Parall el swimming, circl ing and chasing
were also seen in this species. Male coloration varied in
relation to the intensity of courtship activity.
Courtship in G. stemicZa also included circling movements
by the male around the female and when directly in front of the
female rapid vibration of the male's body occurred. Chasing and
leading also took place. Most of this behaviour was performed
at or a few centimetres below the water I s surface.
Intense and more complex courtship was seen in C. riisei
and P. ret iculata and it generally followed the patterns
described for these species in the literature. Only very
1imited courtship was seen in C. aeneue . Spawning in C. riisei
took place in the morning soon after dawn and up to mid-morning
for periods lasting up to an hour. Spawning was not observed in
H. unilineatus but appeared to take place either in the early
morning or at dusk.
293
Spawning experiments with C. riisei were successful and in
total, 11 females spawned under aquarium conditions (Table 30).
Female 11 (not included in analyses) spawned only once with a
clutch size of 36. Other females showed a clear 'small brood'
spawning pattern.
Total numbers of eggs spawned by each female (total fertility)
varied from 17 to 4987 over spawning periods ranging from 10 to
146 days. Numbers of clutches spawned per female ranged from 2
to 55. Generally, the size of the female was directly correlated
with number of clutches ( r = 0.816, p<O.OI, N = 9), total
fertility ( r = 0.730, p<0.05, N = 9) and total spawning period
( r = 0.827, p<O.OI, N = 9). Eggs were spawned in clutches
ranging from 1 to 266 eggs for different females. Larger females
spawned larger clutches than smaller females and mean clutch
sizes (ranging from 2.8 to 90.7) and maximumclutch sizes were
correlated with female size ( r = 0.589, p<0.05; r = 0.796,
p<O.Ol for mean and maximum clutch sizes respectively; N = 10 in
both cases).
Spawning interval s ranged from 1 to 65 days and showed no
recognisable periodicity in pattern varying from one female to
another. For example, female 2 spawned comparatively small
clutches every day for the first 18 days and subsequently spawned
- n
. . . . . .
V l . . , . ,
r o
. . , . ,
- n
. . , . ,
. . , . ,
. . , . ,
- n
, . ,
- n
- n
r r t
3
. . . . . .
0 . >
> - '
N
. p .
( j 1 W
' - l O ' l
~
0 < . 0 C O
. . . . . . .
r o
. . , . ,
V l
- l
r i " e n . - - . .
; : p
0 . >
3 V >
W
w
w
w
w
w
w
w C O
w w
~
: E :
3 r
C O
' - l
' - l
W
~
N
U l I
0 . > U 1 U l U l
0 .
I I I r r i
0 . >
- s
0 . > W
0 .
0
n
r r t n
n
- s
~ ~
- l
. . . . . ,
0 .
C C 0
r o
0
( J l M - 3 M -
W
N
. . . . . .
. . . . . .
N
- s ~
n o - O J
. . . . . .
e n
N 0 : >
0
. . . . . . C O
N C O
0 " >
r i "
V l
- - - '
: : s r o
0 . >
c
r o . . . . . ,
. . . . . .
: 3
I I I
. . . . . .
: 3
«
O J
- s
0 .
«
0 . >
I I I
3
0
- 0 - 0 - l
0 . >
- t )
r o O J 0
. . . . . .
t o
. - - . . ' " ' $ : : : : : M -
+ : : >
. . . . . . .
0 : >
. . . . . . e n
N
N ' - l
" - l
' - l
r o
a . . . . . ~.
- s
O J
e n
+ : : >
w
. . . . . .
0
w
' - l w
N
0 . + : : >
- . - - 0 . . . . . J .
r o
- - - '
V l
o . ~
0 . >
C
l O
~
. . . . . .
0 .
M -
I I I
0 .
. . . . . .
0
O ~
r o
- t )
- + , c
- l
. . . . . . .
. . . . . . .
+ : : >
0 .
3 0
U l c o
W
N W
N
N
r o e r M -
e n c o
W
N
0
. . . . . . N 0
+ : : >
U l
0 .
O J
t O r o O J
- . . . . J
. . . . . . . U l
0 " > e n
. . . . . .
0 e n
c " - l W
. . . . . .
c : : r
. . . . . , t o ' " ' $
0
1 I l
- - ;
~
O J
t o
M -
0
M -
- - ;
3 :
. . . . . .
N
~
N r o
. p . N «
. . . . . . N
\ 0
N
e n + : : >
w
r o
0 . >
I I I
~
- . . . . J
0 : >
W 0 W
\ 0
N ' - l 0 N
r o
. . . . . . .
- 0
X
O J
~
1 + 1 +
1 + 1 +
1 + 1 +
1 +
- 0 1 + 1 +
M -
: : : : :
r o
. . . . . , ( " )
r o ~
V l
. . . . . .
. . . . . . . . . . . ,
. . . . . .
. . . . . . .
r n
. . . . . . 0 0
0
N 0
0
c
~
<
3
( j 1
M - O J l O
N m w
0
W
" - l < . 0
r o
. . . . . .
( " )
~
~ I I I
r o
M -
x
r o
3 :
. . . . . .
. . . . . . . V l
. . . . . .
. . . . . . . . . . . . N 0 - 0
. . . . . . N c : : r
a . r o
r o
- - ;
( j 1
. - - . . M -
r o
U 1
0 . : : : : :
: 3
- - - r o
r o
r o
~
~
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . ; : > : l
. . . . . .
M -
I I
I I I I
I I
0 . > I I I
. . . . . . . . . . . .
. . . . . . . . . . . .
N
0 " > ~
( j 1
( J l
- . J - t )
< . 0 W W t o
U 1 + : : >
r o
0
- - ;
C J
. . . . . .
. . . . . . . . . . . . . . . . . .
N
W + : : > U l
< . 0
3 :
( j 1
. . . . . .
N
N W - . J
0 W 0
0
r o
' - s
0 . >
t " ' > .
U l '
. . . . . .
0 W
0 - . J U l ( )
0 0 0 0 - . . . . J
~ t " ' > .
. . . . . .
C J : l
1 +
1 + 1 + 1 + 1 +
1 + 1 + 1 + 1 +
1 + C
1 + < ' 0
t " ' > .
r i "
. . . . . .
. . . . . .
( " )
V l
N
0 0 W O ' l \ 0 \ 0 - . J ( J l
+ : : >
r n ~
. p .
. . . . . .
U l
\ 0 W O ' l N
0
U l
m
I I I
N
; : > : l
N
r o
. . . . . . ( j 1 . p .
. . . . . . . . . . . . I - - ' . . . . . .
< . 0
U 1
c o
0 . > V l
I
I
I I I
I I I
I
I
: : l
. . . . . . .
N
I - - '
N
t o
( j 1
N
+ : : >
+ : : > 0 N
U l
0
m
r o
. . . . . . ( j 1
N
" - l ' - l
W N
e n
- . . . . . J
m
1 7 6 Z
295
at more protracted intervals (Fig. 70). Female 1 however,
spawned at fairly regular intervals (mean interval 2.7 ± 0.3
days) for 146 days. No peaks in spawni ng were recogni sed in
the spawni ng patterns of either female. In other females peaks
of spawning activity could be recognised being separated by
prolonged periods of inactivity, for example females 3, 7 and 10
(Fig.70). Generally intervals between clutches were short with
modal values between 1 and 2.5 days.
Most of the females used in these experiments died within
two months of their last spawning episode. Female 1 survived
more than two months after spawning ceased but no attempts at
further spawning or courtship were seen.
Eggs were spawned onto the artificial weed provided and
these water-hardened eggs were on average 1.106 ± 0.014 (SE) mm
in diameter and weighed 0.305 mg each. A very variable proportion
of these eggs failed to develop but the proportion did not result
from sperm shortage. The male with female 1 died after 48 days
of the experiment and was not replaced. Ouring the rema i nder of
the experiment, fertilised eggs were produced and hatched
norma lly; a further period of 98 days over which 3841 eggs were
spawned. Sperm were found to be stored in spermatophores in the
oviduct of female C. riisei. Each spermatophore was slightly
296
100I-
o I II t ~U
50 100
F2 150 F3
100
QI
N 0
~1°OL 50 50-1 F5200
.L,
. 0 0
20 20 2011 100F7 F8
. - • .aJ. I ,Li I i 0
50 20
100 F9
Laboratory spawni ng patterns for C. ri i se i .
FIGURE 70:
297
smaller than the size of a mature ovum and consisted of large
numbers of sperm embedded in a viscous matrix without an outer
capsule. Single spermatophores were found in the oviducts of
wild-caught spent females suggesting that sperm shortage may
not be a problem for th i s speci es .
Hatching of C. riisei eggs occurred within 24 hours from
the time of spawn i ng to produce 1a rv ae whi ch on average were
3.1 mm in total length. The yolk sac was still present and
disappeared within two or three days. larvae remained more or
less inactive on the substrate or attached to the sides of the
container for the first two days. After this time they became
very active as their fins developed more fully and by the fourth
day were seen to be browsing on the substrate and weeds. Efforts
at rearing them further than this stage were unsuccessful.
Spawning experiments withH. uniZineatus were not as
successful as with C. riisei and only four females spawned on the
weed provided (Table 31). Differentiating between the sexes in
small specimens was often difficult and many combinations of
different individuals and trials were attempted before successful
spawni ng was attained. Despite the paucity of data for this
species however, certain differences could be seen when compared
with C. riisei.

299
Total fertil ity of H. uniZineatus ranged from 64 to 1014
eggs spawned over periods up to 37 days. Numbers of clutches
spawned varied from 2 to 13 and average clutch size ranged
from 16.2 to 145.7 eggs. Generally clutch sizes were comparable
to those of C. riisei but were spawned over shorter overall
periods with greater intervals between clutches (modal interval
3 to 4 days, mean intervals 2.8 to 4 days for H. uni.l.ineatue]:
Spawning patterns tended to indicate a gradually decreasing
clutch size over the breeding period, for example female 2
(Fig. 71). H. uniZineatus thus showed a small brood spawning
pattern similar to C. riisei but individuals did not spawn for
as prolonged periods as the latter species.
Eggs of H. unilineatus were approximately the same size as
those of C. riisei and hatched within 24 hours. Larval develop-
ment was similar to C.. riisei.
Only 1imited breeding of P. reticulata was attempted due to
lack of space and concentration on other species. In particular,
breedi ng was undertaken only to obta in data on age at maturation
and breeding periodicity. Periods between successive broods
were found to be approximately five weeks. Since, the exact
time of fertilisation could not be determined, the time for
brood development was not ascertained. Brood sizes varied from
300
200- F1
100-
F2
.-
VI
100
.c
..u...
::J
U
O ......... L..&...,
10
l00L:
o 20
l°OL
o 20 days
FIGURE 71: Laboratory spawning patterns for H. unilineatus.
301
4 to 11 with a predomi nance of ma1es produced. Young were born
at a size of 5 to 6 mm in total length.
In order to compare fecundity estimates by conventional
ovarian analyses and fertility as determined in laboratory
spawning experiments, batch fecundities (calculated from the
fecundity-length equation) and total fertilities for'different
sized females of C. riisei and H. uniZineatus were compared
(Table 32). For a C. riisei female of 38 mmcalculated batch
fecund ity was 544 eggs as compared to actual total ferti 1ity of
4987 eggs. The weight of these as water-hardened eggs amounted
to 22% and 198% respectively of total female body weight. Such
a high total fert il ity may be extreme for thi s speci esc ,Neverthe-
less it showed the vast underestimation made by ovarian estimates
of fecundity for a prol ific multiple spawner such as this. For
a smaller female 32 mm SL, calculated batch fecundity (243 eggs)
and total fertility (206 eggs) coincided closely and amounted
to between 14 and 17% of female body weight.
Data for H. uniZineatus showed a similar trend with ovarian
fecundity estimates being lower than fertility for a large
fema 1e but not with as great a ddscrepancy as seen inc. riisei.
In the examples used, total egg weight to female body weight
ratio (using the same unit egg weight as c. riisei) was low
( . / )
. . . . c o
N
: : : c :
- 0
I ' D
n
- t , ~
- t ,
r n
~
) : : >
=
=
=
0 .
; : 3
t o
0 0
<
- l
0 . I ' D
- ; ' C " .
t o
- S
r o
) : : >
V )
C 1 )
)
O J
( \ )
" " '
- t ,
' C " .
: : E :
C O
O J
r
0 .
; : 3
C D
- S
t o
r r I
. . . . .
( \ )
I ' D
r o
V )
t o
" : :
~
W
: : : y
' C " .
z r
< : " t -
- t ,
N
- - ' ' C " .
r t
~
r o
C ! l
C ! l
' <
- s
( \ )
( )
I
( " ' " t
Q J
' C " .
7 '
. . . . . . . . . . . .
. . . . . . . . .
- - '
o
( J
s : u
- - '
s : :
< : >
( " ' " t
V )
0
~
. . - - . .
I ' D
' <
0 -
( . / ) : 3
Q J
w : 3
W
0 - W
W W W
I ' D - 0
r
- - . . J
N : 3
< : > C O
r - t -
0 " 1 C O
- t ,
l O s : u
. . . . . . . . .
<
V )
U 1 C D
a
l O
- S
- 0
0 -
- s
. . . . . .
0 -
I ' D
: 3
V )
s : :
: : : :
Q J
l D
n
r - t -
I ' D
s : u 0
. . . . . .
V l
: : ; :
- '
~
: 3
a
u » l O
- 0
I ' D
: : E :
z s : 0
~ - S
- t , 0
~ C D
- t ,
- t ,
. . . . . c +
- t , e : 0
V ) I ' D
I ' D
r o
- t ,
0 -
t o
N U 1 : 3
U 1 0 " 1 + : >
: 3
: 3
0 - . . . . . .
0 -
. . - - . . Q J
I ' D
= s + : >
+ : > . J : : > + : > c o
O J
- t ,
. . . . . .
s : u
0 -
) : : > l O O J
I ' D n
N W 0 + : >
r t U 1
- '
Q J
I ' D
c +
l O c +
- S
- 0 C D
r o
V ) - t ,
V ) n
I ' D r t
: 3
- 0
0
V l
: : : y
- t ,
s : u
0 - n
I ' D
e r - h
- - '
C D
= s
~ 0
- ; - t ,
I ' D
0 -
0 -
: : E :
Q J I ' D
C D
' <
N
"
n
C D ~
0 -
r t
X
"
e :
0
r t
n
C D
: : E :
O J
- ; s : : 0 - ~
0 - r r 1
p : > C D
: : : : l
V l
0 -
O l D ~
' <
0 .
. . . . . .
0 -
O - l D
l D = s
. . . . .
: : : y
Q J . . - - . . , <
N
c +
: : : :
- ; C D I ' D
. . . . . . . . . . . . r t
r t t - - ' N
N
~ : : : :
' <
w
"
. . . . . .
U l
. . . . . . . . . : : : :
t - - ' N I ' D
0 - - - . . J - . . . . J - - . . J
' <
. . . . . .
i - ' I ' D l O s : u
n I i >
: : : y
Q J
. . . . .l D.
~ ~
l D = s 0 -
C D M -
: : : y c +
n 0 -
c + • •
s : :
- s c +
- '
0
C D s : u
s : u
t o M - c +
. . . . . .
t o
r t s : u
V l - - '
C D
0
0 -
s : : ; ; 0 - t , - t ,
0 ~
- t ,
V l - + , e :
C D I ' D
0
. . . . .
. . . . . .
t - - '
- s
+ : > n
- s
: 3 - s
0 ~ N . . j : : >
+ : > < : > r o
\ . 0 0 - 0
c +
. . . . . .
- - j
t o
W < : > U 1 l O I ' D C " " )
: 3 c o
- s
- - . . J t - - '
l O
. J : : > 0 " 1 ' - I 0 - - '
0 -
- s
V ) V )
. . . . . .
s : :
N C D
r t
Q J
I ' D
~
0 -
~
c +
. . . . . .
~
' <
" ' .
t o
r t
. , . ,
" ' .
c +
C i l
s : u
( \ )
= s r o
V )
C D
I t o
- ;
: : : ; : " ' .
t o
0 - r r 1
r t
~
. . . . . .
C D
O l D
s : u
e :
. . . . . .
: : E :
O - l D
: : : i
: 3
l D C D . . . . .
. . - - . . , <
0 -
0 -
= s . . . . . .
~
r t
: : E :
r o
t - - ' t - - ' . . . . . . . . . : : : ; :
t o N
r t
0 ' \
\ . 0
C D
' < t r : : : - s
. . . . . .
= s
+ : > - . . . . J
+ : > w
i - ' I ' D
C O
V )
- ;
r t
. . . ·. l . D
C D
l O = s
~
0
- - ' - ;
: : : y r t
; : 3
- + ,
Q J
C D
c + • •
( ) " ' .
r t
< ; - - - >
I ' D
. . . . .
0
l O
- ;
0
" ' .
; : : s
l O
0 -
~
( \ )
V )
U l
C D
~
= s
- + , . , . ,
0 -
< : - t -
s : u
I ' D
C D
~
~
- 0
- s n
- s
C i l
0 -
. . . . . .
0 . . . . . .
N
s : u
0 0 e :
c +
. . . . . .
M - ~
m
c o < : >
- - . J . 0 . . . - - '
c o N
- . I . - . I .
0
c + r t
' < ' <
Z O E
303
(not more than 27%) compared to C. riisei. Smaller H. unilineatus
females also showed close agreement between batch fecundity and
total fertility (Table 32).
From the above comparisons, it appeared that small individuals
may mature one batch of eggs and spawn all of these in several
clutches over relatively short periods of time. However, larger
females seemed to have a different strategy where the equivalent
-of several batches of eggs could be matured successively and
spawned in many clutches over prolonged periods. By comparing
total fertil ity and batch fecundities for C. riisei (Table 32)
it may be deduced that the equivalent of from 1 to 8 batches
may be matured successively and spawned almost continuously
over peri ods of 3 to 5 months dependi ng on the size of the
female. A large female H. uniZineatus showed less than the
equivalent of two batches of eggs maturing and being spawned
over a five week period.
Data on mean cl utch sizes for c. riisei suggested that one
batch of eggs coul d be spawned in 6 to 10 cl utches. For some
females which spawned in the laboratory, for example female 3
(Fig.70 ), numbers of ~lutches spawned closely together
coincided with this range.
304
Oocyte size distribution analyses:
Oocyte size distribution studies were conducted for all
species except P. reticulata. In most cases, ovaries which
were analysed were developing, developing/mature, mature and
spent. One immature ovary was analysed for H. unilineatus
for purposes of compari son.
Gas tieropelecue eternic La
Oocyte size frequency histograms for developing/mature
ovaries in this species showed a large proportion of reserve
previtellogenic oocytes (0.024 - 0.170 mm), an intermediate
group of small vitellogenic oocytes and a smaller group of
oocytes within the mature size range for this species (>0.559 mm)
(Fig. 72). As maturity was attained both modal and maximum
sizes of the ova within the largest size range increased.
Mature ovaries were all excised from ripe-running fish and
showed increasing proportions of mature-sized ova with increase
in GSI. There were also comparatively more small vitellogenic
oocytes within the 0.243 to 0.486 mmrange in larger ovaries
and in some ovaries they seemed to constitute an intermediate
peak of oocytes. In the mature ovaries, there was a fairly
distinct separation between mature-sized ova and these smaller
305
Developing/Mature
3 GSI 4.69%
2
Mature
3 GSI 4.84 v,
2
Mature
3 GSI 6.14%
2
e~o 1-J:.L-Jo-Ln-=-::\42~~~$!:!:-t--:-IJl-:?:--~------r------.----":"-===:!::;::=!-_--.-----=::::;:==--
x=r' Mature
~ 3 GSI 8.01%
LL
2
Mature
3 GSI9.52%
2
Regressing
3 GSI 0.84 v.
2
5 10 15 20 25 30 35
Oocyte size (micrometer units)
FIGURE 72: Oocyte size distribution analyses for ~. sternicla
(l micrometer unit = 0.0243 mm; numbers at top represent
percentage frequency).
306
vitellogenic oocytes indicating some synchrony of maturation
of ova. However, in some mature ovaries with the highest GSl1s,
this separation was not as clear and this might suggest the
possibility of continuous recruitment of smaller oocytes even
at this stage of maturity. These ovaries also showed loss of
some of the 1argest ova (0.851-0.875 mm) presumably as a result
of spawning or handling.
These oocyte size distributions indicated that ova mature
in a synchronised manner in this species but some maturation of
ova might take place while spawning is occurring. The spent
ovary showed a distinct decrease in the percentage of mature
ova and smaller vitellogenic oocytes and supported the idea of
some continuity of ovum maturation during spawning. Residual
ova would most 1 ikely be resorbed although no signs of atresia
were observed.
On the basis of other evidence indicating a short breeding
period in this species (rapid decrease in GSI and synchronised
appearance of juveniles in the population) it may be suggested
that although prolonged maturation and spawning might take
place, it may not occur for more than one or two months.
Prolonged spawning might account for the high variabil ity in the
fecundity-length relationship for this species.
307
Corynopoma riisei
Analysis of ovaries from this species showed that even in
mature ovaries oocyte distribution was almost continuous from
sma 11 previtellogeni c oocytes to mature-sized ova (Fig. 73).
A small peak of mature-sized ova could be seen in the largest
mature ovary. The partially spent ovary showed loss of many of
the large ova while the spent ovary was lacking the majority of
these as well as smaller vitellogenic oocytes; only very few
ova remained.
Oocyte distribution indicated asynchronous and possibly
continuous maturation of ova in time since there was no clear
distinction between mature ova and smaller vitellogenic oocytes.
Continuous recruitment of small oocytes into the mature size
range indicated the potential for multiple spawning over a
prolonged breeding period. This was confirmed with data
obta ined from 1aboratory spawning experiments. Similarly in
order to maintain the high level of egg production seen in
these experiments, continuous development of reserve oocytes
would also have to take place.
Astyanax bimaeulatus
The distribution of oocyte sizes in this speci~s showed
two main groups of oocytes: small reserve oocytes and small
308
Developing
3 GSI 6.51%
2
Developing/Mature
3 GSI 7.84 'I.
2
Mature
3 GSI 14.13'1.
o 00
Mature / Regressi ng
GS15.44'1.
3
2
Regressing
3
2
O·..L---~5-----:,r:-O=::"~~""""='=15:==~-::;';20;---~~25=---=="'---=3TO---~35
Oocyte size (micrometer units)
FIGURE 73: Oocyte size distribution analyses for C. riisei
(information as for Figure 72).
309
vitellogenic oocytes constituted one group, and the larger
group cons is ted of mature-sized ova (Fig. 74)'. These were
cl early separated especially in the most mature ovary analysed.
As maturation progressed, the percentage of mature ova increased
substantially. Maximum ovum size increased from 0.729 to 0.826mm
while modal ovum diameter increased from 0.608 to 0.632 mm.
The 1argest ovary ana lysed showed loss of some of the 1arger
ova and a distinct lack of oocytes between 0.437 and 0.462 mm
in diameter. The increase infrequency of oocyte sizes at
0.243 mmmight have been due to differences in methods of
estimation of oocytes smaller and larger than this size.
The distribution shown suggested synchronised maturation
of a large number of ova in this species with limited or no
recru itment of small oocytes into the mature size range once
spawning has begun. Owing to the lack of availability of a
spent ovary for analysis, no conclusions may be drawn on the
nature of spawning (whether multiple or total) in this species.
Hemigrammus ~nilineatus
Ovarian size distribution analysis was difficult for this
species because of lack of any developing ovaries for analysis
and the difficulties in differentiating between early mature
and partially spent ovaries. An immature ovary only contained
310
Developing
3 GSI 6.77 v.
2
Developing / Ma ture
3 GSI8.87%
2
.~
>-
u
C
'~"
cr Mature
'" GSI 10.76 v.~ 3
2
Mature
GSI 14.80'1.
3
2
o-L.l...-_--r- --,-__ ~;:==::!::JL-___L,---_.__-----,--=:,e....-.C:::l--.-
5 10 15 20 25 30 35
Oocyte size (micrometer units)
FIGURE 74: Oocyte size distribution analyses for A. bimaculatus
(information as for Figure 72).
311
oocytes which were without yolk and smaller than 0.243 mm in
diameter (Fig. 75). Mature ovaries showed a pattern of oocyte
size distribution similar to G. stemicZa with a fairly clear
distinction between mature-sized ova and smaller vitellogenic
oocytes and a comparatively large percentage of the latter.
However in the ovary with the highest GSI, an almost continuous
distribution of oocyte sizes from reserve oocytes to mature
ova was apparent. As with G. etierrriol-a this could indicate
conti nuity of oocyte maturation duri ng the breeding period.
Partially spent and spent ovaries contained only a small
proportion of mature ova clearly separated from the reserve
oocytes.
Data for this species suggested that ovarian maturation
may be somewhat synchronised but continuous recruitment of
small vitellogenic oocytes during breeding is not precluded.
Partially spent and spent ovaries indicated multiple spawning
to some degree since all the ova were not spawned at once.
Laboratory spawning experiments confirmed multiple spawning in
thi s species with a 1imited number of small clutches being
produced. Numbers of eggs spawned during theseexperiments
indicated continuous recruitment of smaller vitellogenic
oocytes must also take place during the breeding period in some
individuals.
Immature 312
3 GSI 1.19%
2
O-'J:-.:.:..~$?;-;::::l~~!;:.:.!.:..~:. .J:---....,.l----..-----.-----.-----,..------r-
Developing / Mature
3 GSI9.51%
2
Mature
3 GSI 8.24 v.
2
>-
u
C
<II
:J
CT Mature
~ 3 GSI 10.97 v.
1.L
2
Mature / Regressing
3 GSI 4.91 %
2
Regressing
3 GSI2.84%
2
0.L---r--~-=~~::::l-...=~-==---::;;;=-_~ __ -.-
5 10 15 20 25 30
Oocyte 35size (micrometer units)
FIGURE 75: Oocyte size distribution analyses for H. unilineatus
(information as for Figure 72).
313
Comjdorae aeneus
Analysis of ovaries of this species showed an unclear
distinction between mature and smaller vitellogenic oocytes
(Fig. 76). There was only a small peak in the distribution
of mature ova except perhaps in the ovary with the highest GSI.
Maximum ovum di ameter increased to 1.75 mmas maturity
progressed but determination of the lower size limit of
mature ova was difficult owing to lack of clear separation
between these ova and the small er oocytes. There was also an
almost continuous distribution between the small vitellogenic
oocytes and the non-yolked reserve oocytes.
Data indicated that ovarian maturation in this species
was not highly synchronised with maturation of ova probably
occurring continuously during the breeding season. The spent
ovary showed only a few ova greater than 0.608 mm. The
spawning pattern was most likely a multiple spawning one
over long periods. In some ovaries, many atretic ova were
found suggesting that a large proportion of residual ova
might be resorbed in this species. This was not quantified
however.
314
Developing / Ma t ure
3 GSI 3.02 %
2
Mature
3 GSI 11.0 v.
2
O...L--__ --.-- __ --,- __ --r __ ~~-L...L...,_--.___-===?
Regressing
GSI 1.78 v,
3
2
0.L.._--,~_-:::~L-~.-.e:=~r:=~~Q..-r:::1----r-
10 20 30 1,0 50 60 70
Oocyte size (micrometer units)
FIGURE 76: Oocyte size distribution analyses for C. aeneus
(information as for Figure 72).
315
DISCUSSION
The species composition of the freshwater fish communities
of the Quarahoon and Carlisle Rivers was dominated by primary
freshwater fishes such as the Characiformes which made up almost
50% of the species recorded. Other primary freshwater families
recorded were Pimelodidae, Callichthyidae, Gymnotidae and
Nandidae accounting for a further 26% of the freshwater fish
species while the remainder of the species belonged to
secondary freshwater fish families (Darlington 1957). This
composition was consistent with the general pattern of
distribution of freshwater fishes in Trinidad (Price 1955,
Boeseman 1960, 1964).
In spite of the small drainage area of the Chatham streams
and the exacting environmental conditions, the study area
possessed a notably high species richness, supporting almost
half of the true freshwater species recorded for Trinidad
(Boeseman 1960) in addition to two new species records for the
island. In contrast, Williams & Coad (1979) reported low
diversity of fishes in Canadian temporary streams because of the
need for physiological tolerance or migratory patterns to
withstand the harsh conditions of this habitat. A study of
316
the Bandama River, Ivory Coast (Leveque et: al. 1983) noted that
more than 80 species were recorded for this drainage within a
catchment area of 97,500 km2; approximately that predicted by
species number/catchment area relationships (Welcomme 1979).
However such relationships are not reliable predictors of
diversity for basins of extremely small area such as the study
site.
Factors whi ch have genera 11y been proposed to affect the
species diversity of communities include time, environmental
stability, spatial heterogeneity, productivity, competition
and predation (Lowe-McConnell 1975, Krebs 1985). In addition,
the intermediate disturbance hypothesis proposes that extreme
levels of disturbance of communities would result in low
species diversity whereas intermediate levels would favour
increased divers i ty (Connell 1978, Krebs 1985). Such
disturbance levels may be dictated by both frequency or intensity
of disturbance (Krebs 1985). It was proposed by Sanders
(quoted in Lowe-McConnell 1975) that predictable oscillations
of long-term sequence could produce high levels of diversity,
a situation exemplified by tropical rivers with regular cycles
of flooding (Lowe-McConnell 1975, Beumer 1980). Cyclical
fluctuations in conditions did occur in the streams studied
although predictability was comparatively low. Nevertheless
317
extensive refuge pools and long flow conditions may have
mediated the detrimental effects associated with unpredictability
or particularly harsh flood/drought conditions to promote high
species richness (Clifford 1966).
A further factor which may have been contri butory to high
species ri chness at the study site was the dynamic state of
local faunas in the southwest peninsula as a result of coloni-
sation of species from the nearby mainland. Of the nine
characoid fishes recorded, three were recently established
co 1oni sers and two were new records duri ng the study peri od ,
G. etemiela and M. bondi are reported only in a restricted
number of streams in southwestern Tri nidad (Pri ce 1955, Boeseman
1960) and evidence points to their being recent colonists
(Price 1955). E. erythrinus is not recorded by Boeseman
(1960, 1964) having been caught only within recent times in a
small north-flowing watercourse in a drainage immediately
adjacent to the Chatham basin (Kenny, pers. comm.). While
not all colonist species establish populations, at least these
three relatively recent arrival s have done so and a fourth
T. elonqatue , may be in the process ofestabl ishment. Faunal
diversity trends in some North American river systems have
been shown to be explained in part by the proximity of basins
to rich source areas (Horowitz 1978).
318
With the exception of the above species which are of
restricted distribution, all of the other freshwater species
are widely distributed and common throughout Trinidad south of
the Northern Range (Guppy 1934,1936, Price 1955, Boeseman
1960). Some species are commonly found in ponds, ditches and
slow-flowing watercourses, for example G. carapo and
C.caUichthys whil e others are found in middl e to lower course
rivers, for example R. sebae~ C. aeneus and s. marmoratus~ and
the characids A. bimacuZatus~ H. unilineatus and C. riisei.
In particular, P. ve t-iculat a, C. r-i iee i and A. bimacul.atue
have been reported to be the three most common species in
Trinidad (in order of decreasing abundance) based on analysis
of the numbers of collections made by Price (1955) containing
these species (Nel son 1964). In addition, Nelson showed that
C. riisei and A. bimaculatiuewere significantly associated with
each other and with three other species: H. unilineatus~
C. aeneus and the cichlid, C. bimaculatum. He attributed this
association to common exclusion from particular habitats,
specifically those with high gradients or those tending to
become brackish in the dry season.
t~any of the species collected from the Chatham streams
have been shown to have some abi 1 ity to withstand fl uctuating
environmental conditions, particularly stagnation and associated
319
hypoxia, increased predation and crowding (Carter 1935, Lowe-
McConnell 1964). Air-breathing has been recorded for
E. eruthi-inue , Cal.l ichthue , C. aeneue and S. marmoratiue
(Kramer 1978b,Kramer & McClure 1980) and suggested for
. C. bimacuZatum (Lowe-McConnell 1964). An air-breathing
ability and a capacity for overland movement (for example in
R. hart-i i , C. caUichthys and S. marmoratue'i permits the
colonisation of new habitats which may be more favourable.
Use of the oxygen-ri ch surface water for respiration has been
reported for many Neotropical fish species (Carter 1935, Lowe-
McConnell 1964, Lewis 1970, Kramer & McClure 1982) including
P. reticulata (Kramer & Mehegan 1981), and species of the genera
Hopl.i.ae, Astyanax~ Rhamdia; Rivulus and CichZasoma in Panama
(Kramer 1983). Myers (1947) and Kramer et al. (1978) recorded
the ability of S. marmoratus to survive in an active state
without free water in burrow systems in the Amazon. Despite
the capacity for many species to withstand harsh environmental
conditions, species richness was greater for the deeper, more
permanent pools in the streams. In addition, the larger species
were found in the bigger, deeper pool s comparable with other
studies elsewhere (Holden 1963).
The species of the brackish water Station 4 included
representatives of a euryhal ine freshwater family, the
320
Poeciliidae, while all the other species belonged to euryhaline
marine families (Miller 1982). c. paral.leiue and M. curema
have been commonly found in brackish water habitats in Trinidad
(Boeseman 1960). The use of the lower estuari ne reaches of
rivers as a nursery ground is relatively well documented
(Beumer 1980).
Total populations of the six species studied showed wide
variation over time and correlated with stream discharge rates
and inputs of j uven i 1es duri ng the reproducti ve seasons.
Migratory movements may have also played a role in determining
population densities of certain species. Prolonged rainfall
into the early part of the 1982 dry season indicated that
moderate flow regimes were ideal for both reproduction and
survival leading to high population sizes of most species at
this time. The role of flood events in decreasing fish
popul ation sizes has been well documented in many rivers and
streams which are subject to highly seasonal flows (Hynes 1970,
Brown 1971, Deacon & Minckley 1974, Beumer 1980, Collins et
al: 1981). In part i cul ar, streams with narrow courses and steep
banks are more 1 ikely to show the destructive effects of
flood i ng than those with low banks or wi de bra i ded courses
(Deacon & Minckley 1974). Presence of small tributaries or
floodplain pools have also been shown to be important as
321
refuges duri ng fl oods in desert ri vers (Deacon & Minck1 ey 1974).
Severe floods may affect adul ts and juveni 1es as well as
recently spawned eggs (Pa1oumpis 1958, Seegrist & Gard 1972)
although John (1964) maintained that young of the year were
most affected due to their reduced swimming abil ity.
In the Bandama River, Leveque et al. (1983) suggested that
decreases in fish catches with floods were not due to reduction
in population sizes but the inefficiency of sampling methods
under these conditions. Increased habitat size resulting in
decreased dens it i es of fi sh is also an important cons i derati on
especially in floodplain rivers where mortality is assumed to
be low at the time of floods (Welcorrme 1979). At the study
sites, washouts due to the narrow channel were likely to have
been the major cause of low catches during the rainy seasons
with habitat expansion into small drainage channels playing
a minor role. It was presumed that tributaries and deeper
poo 1s where current flow was reduced acted as refuges duri ng
floods. Recovery of populations after flooding was rapid as
documented in other studies (for example Deacon & Minckley
1974, Beumer 1980).
In contrast, during drier periods, population densities
increased primarily due to habitat contraction. Although
322
Shirvell (1981) postulated that stream flow was not always
inversely related to fish density because behaviour such as
territoriality at times of habitat contraction could act to
maintain individual spacing and promote migration to other
pools. Mortality did not appear to be particularly high at
these times except when complete drying out of smaller pools
took place. The ability of many species to withstand stagnant
conditions and crowding and access to an extensive series of
refuge pools most 1 ikely played major roles in enhancing fish
survival unl ike other situations where drought induced high
mortalities in North American intermittent streams (Paloumpis
1958, John 1964) and on tropical floodplains (Lowe-McConnell
1975, Welcomme 1979).
Population sizes also fluctuated as a result of large
numbers of juveniles, particularly of A. bimaaulatus , H.
uniZineatus and C. riisei3 entering the population during the
rainy seasons once or twice each year. Highly synchronised
seasonal spawning patterns of high fecundity species have
been shown to lead to great fluctuations in fish populations
of tropical rivers (Lowe-McConnell 1969). Leveque et al: (1983)
also noted that failure of floods in certain years could lead
to relatively smaller increases in population at these times
due to alack of a reproductive stimul us.
323
Populationfl uctuations in P. retiiculata in particular
reflected a pattern of mortality very much dictated by periodic
flooding although some possibil ity of a density dependent
population control mechanism might have existed. Reproductive
capacity of female guppies in 1aboratory-kept popul at ions is
known to be affected by a variety of factors including photo-
period (Turner 1937), artificial illumination (Scrimshaw 1944),
water temperature (Dildine 1936), food (Silliman 1948, Hester
1964), copulation frequency (Breder & C,oates 1932), living
space (Rose 1959, Rose & Rose 1965) and population density
(Warren 1973 a, b , c, Dahlgren 1979). However, this study
showed no obvious density-rel ated changes in fecundity. In
other studies, Rosenthal (1952) and Yamagishi (1976) found
cyclical reproductive patterns where numbers of young increased
to a maximumthen decreased over time. Yamagishi (1976)
attributed these to peaks in reproductive activity of
successive generations of females occurring at 24 to 36 week
intervals; no such trends were observed in the present study.
Relative abundances of the six species changed over the
study period with G. etiernicl.a becoming comparatively abundant
towards the end of the study. This could be attributed to
several consecutive successful reproductive seasons for this
species. Evidence for such shifts in relative abundances of
324
fish species has been widely documented for North American
streams (Starrett 1951~ Seegrist & Gard 1972~ Deacon &
Minckley 1974, Grossman et aZ 1982, Waters 1983, Schlosser
1982, 1985, Minshall et aZ 1985). These studies correlated
such changes in community composition with variability in
reproductive success and juvenile survival for different
species due to interannual differences in hydrological regimes.
Grossman et aZ (1982) after a 12-year study of an Indiana
stream fish assemblage concluded that stochastic forces play
a decisive role in the determination of structural and
functional properties of many stream fish assemblages. ,Changes
in relative abundance of fishes have also been noted in
tropical systems, for example Holden (1963) and Lowe-McConnell
(1964).
Life history parameters and reproductive characteristics
of the six species studied are summarised in Table 33. Generally,
individuals matured early and had relatively short lifespans.
The largest species A. bimacuZatus showed a greater estimated
age at first maturation and lifespan than the other smaller
species. In some species, the reproductive strategy seemed to
be to maximise production of eggs or young at any favourable
time of the year but usually breeding was associated with one
or both of the twice yearly rainfall maxima. Reproductive timing
ranged from close coincidence with the onset of the rains for
3 : - 0
V l C J j
. . " 3 :
3 : 3 : 3 : 3 : 0
o 0
I I I
" " 0 ~ Q . l P J Q . l Q . l - + .
x
a . " " 0
X X x r t . . . . . . . - s
' "
' " I I I
I I I c
c V l . . . .
: : : : r t >
C J j V l
~ 3
: : l a . 3 3 3
3 3 3 ~ - o
~ n . . . . J 0 P J c : c : c < : :
c : c : c r o ,
o : : T . . . - j . . . . . . . r - t - 3
: : l : : l
3 3 3
3 3 3
o 0
< . C < . C
V l
o : : >
a . . . . . .
V l o V l
- s
. . . . . . C T ' G ' ) < ( , Q 3 3 :
t i l
t i l V l
< , t i l r
: : l
r o Q , I V ' l C I I I
n r T
' "
' " Q >
r t >
I 1 l I I I
V l r t > " " 0 . . . . . r t - . . . .
~ . . . . 3
" " 0
t i l r t > c : < : :
. . . . n o - s
- s
' "
I I I ' "
. . , ~
t i l V l
r t > a . ~ . : : T
o
t i l N I I I
: : >
' " ' "
' " t i l
~
~ o 0 V l 0 a . N t i l r t > I I I
I I I
: : I : : >
< : : r t > " " 0
o : : l N
. . . . .
t i l V l
. . . . r t > V l
: : > r t >
n Q > ' "
. . . . . . . . . . . .
r t < : ; , 1 . O
r
' < n
Q > I 1 l
c
I I I
' <
I I I : : l
o ~
: : l
< . C
a . : : l
3 3
. . . , ' "
I 1 l ~
r - t - o
3
3
3 r t -
: : T o
. . . .
r t >
' <
3
I 1 l
3
o
. . . .
' <
V l
c
3
w . . . . .
. - . ) . 3 3
N
I I I 0
Q >
< : :
) ( <
o
. . ,
N
" " ' l . . . . 3
: : >
r t > N w ' <
. - . C T \ " " ' - J Z < l > " " 0
o
' " . . . . . . . . . .
" " ' l I w
Z . p ; . N I 0 ~ Q > a . C O
: : >
o
/ 1 )
I I I - 0 N C J ( J ' I 0 0 : >
: : l ~
. . . .
. . . . .
I
c
C O
N
' " 0
. . . . .
~ C
N \ D
' < I I I 3
z N . - . )
. . . . .
t i l
• - . ) N
\ D
C J . . . . . .
. . . . .
( J ' I
w
\ D
r t >
. . . . . .
. . . . . ,
0 1
: : l C O
: : T
~ .
o
t i l
. . . .
. . . .
o
a .
r t >
~
. j : >
. . . . V l o
o
r t > ~ .
E !
o
: : > <
" 0
~
~ N C O . . . . .
c o
: : > . . . .
' "
Q >
. . . . . , C O W
. . . . . . . . . N N
3 r t > < l > " " 0 : : T N
O " l o
. . ,
~ .
I N I ~ Q > . . . . . . 0 \ c o
I W
. . . . . . . W 0
: : >
' "
Q >
: : l Q > / 1 )
N - O . j : > o V l W
: : > ~ u : : l 0 : >
o a
. - . ) ~
/ 1 ) 3
I
r t > . . . .
' " 0 : : T
r o
~ < : :
a . Q > \ D o
I I I N \ D
. . . .
t i l V >
r o
c o
o
W
~
. N N W
: : l \ D
t i l
. j : > < D
0 ' 1
0 : > '- l "' >
Q >
: : l
a .
~
r o
. . . . . .
. . . . .
: : l
" " 0
. j : >
U 1
o ~
~
. . . . . . . .
3
o
. . . . . .
: : T
N
0 : >
3
a .
t . . . o . U 1
< : : < l > 0 <
o 0
' "
c
) ( < r t > . . . . . .
o o
n
M - . . . . . . w . . . . .
M - ~
" " ' l : : >
. . . .
V I 0 \ 0 \
: z ' w" o < l > - 0 ' <
. " , . " ,
. . . . . .
N
- 0 N
: : >
~ o ' " " " ' l ' "
. j : > . . . . . .
<
I I I : : > " " ' l : : T
r o
/ 1 )
. . , ' " 0
~ < : : u : : l
c o
3
~ .
n
U 1 o
V l : : T c o
: t -
: : T
. . . . . . N
r t > N 0 ' 1
: : >
Q >
r t >
o
~
" " ' l
<
Q >
r t >
n
- n o .
. . . .
o
I 1 l
. . ,
l > "
t i l
. . <. . .
. . . .
. . . . . . . . . .
r t >
. . . . . .
I w
3 r t >
n
N
o
Q > 3
. - . )
t i l
< : : . . . . . . . . . . 0 / 1 ) 0 <
~
~
W
. j : > o 0 ) ( < r o
r t > Q >
< D o
N - . . . . " " ' - J ~
M -
£ " . ; : ; .~
. . . . .
I V l C O
N
. j : > I I I . . . . . ,
- 0
~ ' "
. . . .
: : > ~ : : T o
: : T
t i l
I I I \ D
' "
' " 0
~ < : : u : : l
r t >
c o
o 0 1
o
t i l : : T U 1 W
. . . . .
: : l 0 : >
. - . )
t i l
. . . . . . W
W
x
3
Q >
. . . . .
~
r t >
t i l
: : T
: t -
t i l
N
w
o .
. - . )
" 0
Q > < l > 0
' . . . . . . . . . . .
x < r t >
3 N
3
r + ~ . 3
< : : 0 . . . . . n
: : >
t i l
r o ~ ( D W w
o . . . . .O .' l ( ] l 0 ' 1
o
c - t - . . . . . . . . . . . . . . . . 0 l . . O I - s : P J 0 . c : : n + : : a . o r t >
: : >
: : : 3 " " " ' S I N
N I " < ~ . . . . . t i l
. . . . . I I I
Q J O C + : : a . + : o
' " C . , J : : : : : a . ( D 0 ' \
\ D
- - - - c 3 ( X ) ~
. j : > t i l
' " . . . . . ,
V l . - . )
I I I ~ 0 1 r t
O " l
. . . . . .
C
N
' " . . . . .
. - . )
. j : >
U 1
a .
. j : >
o
W
' "
r t >
a .
N o
: r :
0 1
<
. . ,
N
c : n o <
' "
. . . . . ,
I
< < l >
o
" 0
r t 3
. . . . .
W . . . . . .
: : > ~ N
c
: : l
. . . . . .
Z N ~ N U 1 o U 1
. . . . < ' <
r t : z
o 0 ' \ I I . . . . . . .
: : >
o : t -
" " 0 : : T ( J ' I
N m I I I
: : l
Q > N
o
C
~ u : : l . . . . .
~ 0
o
I
o : : T
o
c
. . . . . .
. . . . . .
N
< : :
V l N
V l
m 0 ' 1
\ D
o
s z £
326
relatively short periods (one or two months in G. etiemicl-a
and A. bimaculatus) to more prolonged breeding periods during
the rainy season (up to four months in c. riisei, H. unilineatus
and c. aeneus) to continuous breeding (P. retdoul-ata ). Female
GSI's were highest for P. ret.icul-at:aand c. aeneus and were
roughly equivalent for the other species. Generally batch
fecundity was correlated with individual size and maximum batch
fecundity for each species was directly dependent on species
size. Batch fecundities were highest for A. bimacul.atue and
lowest in P. »et-ioulatas other species had intermediate batch
sizes. However, maximum fertilities for C. riisei and
H. unilineatus were high and comparable with batch fecundities
of the larger species as a resul t of the multipl e spawning
habit.
Data available for some of these species either locally
or in other regions agreed in many respects with that from the
present study. In particular, data were available for c. riisei,
A. bimaculatus and P. reticulata.
Nelson (1964) reported that spawning of C. riisei
occurred predominantly at the beginning of the rainy season
in Trinidad streams although they spawned at all times of the
327
year in the 1aboratory. Under aquarium conditions sexual
differentiation was seen in males at 17 weeks (Nelson 1964).
In comparison full gonadal maturation of males and females was
observed between eight and nine months of age in the present
study,. but the development of secondary sexual characters
began earlier in males. In addition male size may be an
important factor in female choice for this internally
fertilising species where female receptivity is constantly low
(Nelson 1964). Since males are normally larger than females,
delayed maturity to a larger size would be advantageous and
enhance the male's attractiveness to females.
A. bimacul.atue is also reported to be a ra i ny season
spawner in Brazil (Azevedo & Vieira 1938, de Godoy 1975, Nomura
1975 c, Pelizaro et at 1981). Azevedo & Vieira (1938)
described 'p iabas ' (including A. bimaculaiues spawning
immediately after the first rainfall and spawning more than
once during the season. Multiple spawning over a prolonged
season was confirmed by Pelizaro et al (1981) although the
source of their specimens was denoted as 'tanks in the
Department of Zoology, UNESP, Sao Paulo' (p. 416). Data of
Nomura (1975 c) indicated a breeding period when GSI's were
high up to three months long and he concluded that this species
was a total spawner.
328
Upstream movement for spawning was described for 'piabas'
by Azevedo & Vieira (1938) and for two Brazilian subspecies of
A. bimaculabue by de Godoy (l975) with spawning actually
taking place in small tributaries or shallows. Such migratory
behaviour could account for the lack of adults in catches at
the study si te and the movement of 1arger juveni 1es out of the
areas sampled and probably into other feeding areas or larger
pools downstream. Azevedo & Vieira (1938) noted female-biased
sex ratios but attributed them to the inefficiency in capturing
the smaller males.
Data on age and growth for A. bimaculatue in the present
study were somewhat unrel iable and seemed to contrast with
Nomura's (1975 c) estimates of median length at maturity at
92.5 mmtotal 1ength (TL) at two years for males and 104.0 mm
TL at three years for females. However, Nomura found that the
minimum size class at maturity was 65 mmTL for males and
70 mmTL for females; almost identical sizes to those recorded
in the present study (56 to 57 mm SL). Further, it would not
be unexpected that under more exacting environmental conditions
in a smaller habitat as that of the study site, maturation
might occur at a smaller size or younger age. Nomura (1975 c)
noted that fecundity of A. bimaculatus was significantly
correlated with total lengths of fish and batch fecundity esti-
329
mates from the Chatham population fitted the lower portion of
his fecundity-length curve. Maximum fecundity recorded by
Nomura was 31,778 eggs for a 137 mm TL female. In contrast,
GSI's from the present study were higher than those recorded
by Nomura, that is maximum female GSI 10.90%, maximum male
GSI 0.98%.
The only other species for which life history data from
natural populations are available is P. reticulata. Liley &
Seghers (1975) found that sizes of lowland stream guppies in
Trinidad were smaller on average than those from upper or middle
course habitats. For three lowland sites, mean SL of females
ranged from 16 to 17 mm whereas male SL ranged from 12 to 14
mm. Such differences were attributed to size selective predation
by large predators such as Crenicichla alta. Reznick (1980)
and Reznick & Endler (1982) analysed this phenomenon in more
detail and found that for localities where C. alta was the
dominant predator, guppy sizes were smaller than at other sites
where either RivuZus hartii or Aequidens pulcher were dominant.
At Crenioichl.a sites mean male size at maturity for guppies
(mean of 10 randomly chosen males) was 14.88 mm SL while mini-
mum size of gravid female guppies (SL at which two-thirds of
females therein contained developing embryos) was 14. 6 mm SL.
At the Chatham sites guppy sizes were found to be even smaller
330
Male guppy sizes are generally smaller than females as
with most poeciliids due to cessation of growth of males after
maturity and shunting of energy into courtship and other
related activities (Endler 1984). Persistent growth of
females after maturity takes advantage of increased fecundities
being possible with increased body size (Miller 1979, Endler
1984). Endler (1984) also suggested that in areas of high
predation or low food availability, average female body size
may decrease because of the greater efficiency of small females
in bearing young. At the same time however, males may also be
sma11er but the extent of body size reduction woul d not be as
great as for females and under these conditions sexual
dimorphism in size tends to be reduced, that is male: female
size ratio approaches unity. In the present study mean male
mean female size ratio was 0.79 while mean male: minimum
gravid female size ratio was 0.93. The latter ratio was lower
than those quoted for areas with high predation (1.01) and areas
with weak predation (0.94) by Endler (1984, from data of Reznick
& Endler 1982) thus suggesting low levels of predation on
331
guppies at the study site. From observations however, the
study area supported a high diversity of potential guppy
predators including Crenicichla alta3 Astyanax bimacuZatus3
CichZasoma bimaculatum3 Rivulus hartiiJ PoZycentrus schomburgkii~
Hopl.iae malabaxricue, Gymnotus carapo and Synbranchus marmoratus
in addition to arthropod predators such as the nepid, Ranatra
which was commonly found. Such a range of predators as well
as the seasonal concentration of guppies into small pools
during the dry season with these predators woul d in fact
suggest very high levels of predation on Chatham guppy
populations. Si ze reduction may be due more to envi ronmenta1
seasonality and unpredictable mortal ity associated with floods
and drought.
Reznick & Endler (1982) also determined that guppies from
Crenicichla-dominant sites had shorter interbrood intervals
than at other sites. Their estimate of mean interbrood
interval for CrenicichZa sites was 24.1 days compared to five
weeks for Chatham guppies. However, the latter estimate was
based on very few samples and shoul d be treated with some
caution as a generalisation for this population.
Crenicichla-dominant sites were also found to support
guppies with high brood fecundities (Reznick & Endler 1982).
332
For example they found the highest predicted fecundity for a
20 to 21 mmSL female was eight offspring while in the present
study, guppies of the same size had a predicted brood size of
12. High fecundity could be an adaptation to compensate for
high mortal ity from a range of potential predators as well as
being a trait favoured by the seasonal conditions at the
study site. Dahlgren (1979) suggested that if guppy populations
frequently pass through periods of low density, genotypes for
comparatively high fecundity or fertility would increase in
frequency under these conditions and remain common in the gene
pool at other times.
Continuous reproduction throughout the year would also
increase the reproductive capacity of the species. It is
generally quoted that guppies reproduce continuously (Scott 1979,
Lam 1983) but all studies have been laboratory-based unlike the
present one which now confirms continuous breeding in natural
populations of this species. Other factors which could
potentially affect reproductive capacity of guppies appear not
to be important at the study sites, that is population density
and food availability (Dahlgren 1979,1980 a, b). Neither
density-related reduction in fecundity nor seasonal variation
in brood sizes were found; at all times brood sizes and GSI's
were high. Dahlgren (1980 a .b) stressed the very adaptable
333
nature of this species with respect to utilising food of
variable quality.
Other characteristics of the Chatham guppies compared
favourably with those obtained from laboratory-bred guppies.
Female age at first maturation has been recorded as 48 to 63
days (Thibault & Schultz 1978) and 25 weeks (Yamagishi 1976)
compared to 11 to 14 weeks in the present study. Brood
i nterva 1s reported were 21 days (Scrimshaw 1944, Thi baul t &
Schultz 1978) while Yamagishi (1976) found most intervals
between 26 to 140 days (mean interval 38.2 days) with intermit-
tent spawninqs at 59 - 66 and 90 days; intervals at the study
site were about 35 days. Average female GSI's at low population
densities were 9.671% (Dahlgren 1979) while mean GSI for the
present study was 9.749% ( N = 575). Other data not obtained
in the present study but available in laboratory studies were
gestation period - 25 days (Dahlgren 1979), longest period of
reproduction for a female - 400 days (Thibault & Schultz 1978),
maximumnumber of broods recorded for one female was 17 over
344 days (Bowden 1970 quoted in Miller 1979).
Population structure and many 1ife history characteristics
of the species studied were indicative of r -selected species.
Fish populations were subjected to drastic density independent
334
morta 1ity factors such as flood i ng and drought with the former
being particularly important. Populations of species such as
c. riisei~ A. bimaeulatus~ H. unilineatus and P. reticulata
showed extreme oscillations of numbers due to alternating
washouts and recolonisation and/or reproduction. Such
oscillations could be environmentally influenced or inherent
in the biology of the species (cf. Le Cren 1965, Lowe-McConnell
1969). Further evidence for the variable and density independent
mortality these populations were subjected to was seen in changes
in community composition due to stochastic influences.
Popul ati ons of speci es such as c. »i ieei , A. b-imaculatue
and H. unilineatus were largely comprised of recruits rather
than repeat spawners as a resul t of their qui ck maturation and
short life spans. This characteristic of fish populations is
an r-selected trait as compared with populations of longer lived
species which are comprised mainly of repeat spawners
(K-selected) (Nikolsky 1963). This feature was not seen in
P. »ebiculata where reproductive individuals dominated the
population and may be rel ated to the 1ive-bearing nature of these
species and the small number of young produced. With the exception of
335
P. ret iculata, all other species produced comparatively large
numbers of young each breeding season and they decreased
rapidly over time suggesting high levels of early mortality
(type III survivorship curve, Krebs 1985), another r-selected
trait (Le Cren 1962, Pianka 1970).
Other r-selected traits are small body size, rapid develop-
ment and early sexual maturity. These characteri sti cs were
seen in the species for which data were available. All the
species studied were less than 100 mmSL and coul d be
considered to be small according to Miller's (l979 b) definition.
Even the largest species, A. bimacul.atus, only attained a
1ength of 82 mm SL as compared to sizes attained in other
drainages (Guppy 1936, and personal observations). Early
development was rapid with eggs of C. riisei and H. unil.ineatue
hatching in less than 24 hours and larvae becoming free-swirrming
and feeding in four days. The smallest species, P. »etioulata,
matured withi n about two months compared to the 1argest species
A. bimaculatue which appeared to mature within its second year.
Monthly changes in population structure and growth curves showed
maturation ages of less than one year for all the other species
except C. aeneus for which data were inadequate to draw any
definite conclusions. Rapid maturation rates such as these are
typical of small species of seasonal tropical habitats (Lowe-
336
McConnell 1975, Welcomme 1979) as well as seasonal temperate
aquatic systems (Deacon & Minckley 1974). Reduction in age
at maturity is recognised as a potent mechanism for increasing
growth rates of popul ations (Wootton 1984 a).
Overall sex ratios for popul ations of four species
deviated from unity but no investigations were made to
determine reasons for this. Likely factors include migratory
behavtour of juveniles and/or reproductives (possibly in
G. eterniol.a and A. bimaculatues , mating systems of the
internally fertilising species depending on female choice,
that is female-based polygyny (C. riisei and P. reticulata),
and differential predation of more active colourful males in
P. retriculata (Endler 1980). Biased sex ratios have also been
recorded for other poeciliids. Female-dominated ratios
occurred at times of high population densities in Limia
vittata (Barus et al. 1980) similar to situations in the present
study. Krumholz (1948) attributed female dominance in mature
Gambusia affinis to differential death rates of males while
Milton & Atherington (1983) reported habitat preference as a
factor influencing sex ratios in the same species. Experimental
studies by Russian workers using guppies have suggested that
brood sex ratio adjustment by females allowed production of the
rarer sex thus returning population sex ratios to unity
337
following environmental disturbances (Geodakyan et al 1967,
Geodakyan and Kosobutski i 1969, quoted in Brown 1982). However,
replication of these experiments do not confirm this phenomenon
(Brown 1982). Farr (1981) also found female-biased sex ratios
in laboratory strains of guppies and attributed this to long
term inbreeding.
A variety of evidence including presence of mature
reproductives, high GSIls and presence of juveniles pointed to
a seasonal reproductive cycle for all species except
P. rebiculata, Mature individuals of c. »i.ieei., H. uniZineatus
and c. aeneue were found throughout the year but peaks of
reproducti ve act i v i ty coul d be inferred from high GSI IS and the
presence of juveniles in the population. Peaks in reproduction
in these species and spawning of G. etemicla and A. bimaeulatue
took place during the rainy seasons. c. i-i.ieei , H. uniZineatus
and c. aeneus could have had two active breeding periods each
year but often only one was evident for G. eterni.cl.a and
A. bimaaulatue , Kapetsky et al(l977, quoted in Welcomme 1979)
distinguished three groups of fish in the Magdalena River,
Colombia as those which bred once per year, those with two or
more breedi ng peri ods per year and those whi ch bred almost
continuously. Breeding twice each year has been
recorded by Lowe-McConnell (1975, 1979) and Welcomme
(1979) in tropical systems where two flood peaks or
338
twice yearly rainfall maxima occur. In the present study major
breedings seasons twi ce each year were seen cons i stently only
in the smaller species such as C. riisei and H. uniZineatus
which, because of their small size, fast growth rates and
early maturation, had two overl appi ng generations each year
and it was possible that some individuals, especially of
H. unil.ineatus .. bred twice in one year. Larger species such
as G. eteimiol.a and A. bimaculatue showed only minor peaks of
breeding in the second rainy season while c. aeneus seemed
more responsive to very high rainfall at this time. It was
unclear whether the same individuals of these larger species
were able to breed twice in one year. Under conditions where
two breeding periods per year might occur, strategies to
utilise these seasons might differ. Breeding cycles reported
for Barbue species in Lake Victoria, Africa (cf. Payne 1975)
show that two spawning periods may result from different
sections of the population spawning at different times with
each individual having an annual cycle, and for Barbue
ap leuroqranma .. fast maturation allows some young spawned in
one breeding season to mature and spawn by the next season
(Welcomme 1969).
Close coincidence between reproductive activity and the
onset and duration of the rains each year despite the
339
variability of the latter, indicated that the species studied
must have been cued to the stimul i of high rainfall and floods
to initiate reproduction. In 1980, rainfall maxima were well
defined in May and October while in 1981, rains started earlier
in April and again in November after a very dry petit careme.
However, rainfall continued intermittently through the
beginning of the dry season to February. When they occurred,
peaks of reproduction of the five species coincided almost
exactly with the timing of these rainfall maxima. In addition,
early reproduction in April 1981 seemed to allow more time for
first time maturation of young of the year of the short-lived
species such as C. riisei and H. uniZineatus~ as well as
development of gonads again in large individuals of these
species and the larger species in preparation for reproduction
during the second rains through to February. This could
explain the particularly high juvenile abundance at this
time for most species.
The most marked environmental changes seen at the study
site associated with the beginning of the rainy seasons were:
habitat expansion into small drainage canals as well as increases
in depth and width of the main channel and reconnection of
isolated pools; increased discharge and resumption of now after
stagnant conditions; water temperatures decreased with flood
340
events but were generally sl ightly higher and more stable
during the rainy season; increased turbidity and decreased
conductivity. The stream biota showed a reduction in
densities of both benthic and planktonic organisms at the
begi nni ng of the rai ny season when floods occurred but benthi c
organisms in particular showed rapid recovery to high densities
once moderate flow regimes were establ ished.
Many previous workers have found peaks iii spawning
activity of freshwater fish associated with seasonal rainfall
or floods but the exact nature of cues stimulating gonadal
development and spawning are sti 11 unclear (Lam 1983). Factors
which may affect gametogenesis in tropical fish include
photoperfod, temperature, water quality and nutrition (Lam
1983). Studies of the neon tetra, Paracheivodon innesi3 by
Tay (1983, quoted in Lam1983) demonstrated that temperature,
water quality (pH and conductivity) and light intensity were
important in influencing gonadal development. High conductivi-
ties inhibited gonadal development in P. innesi as well as in
the gymnotid, Eigenmannia irireecene (Kirschbaum 1975, 1979).
Th is factor is of some relevance to the present study
considering the high conductivities attained in contracting
poo 1s at the study area duri ng the 1ate dry season when
gametogenesis in most species should be occurring. The ionic
341
Photoperiod has been noted to affect gametogenesis in the
guppy, P. vet-icul-atia, but its effect is unclear (review in
Lam 1983). Gametogenesis has been enhanced in this species by
increased light intensities and may be relevant in the present
study to high levels of reproductive activity during the dry
season when cloud cover was marginally less and leaf-fall
a 11owed more 1i ght penetration to the stream surface. In
addition, guppy populations attained higher densities at most
times at Station 3 which was the most exposed site, although
the additional factor of high algal and planktonic productivity
at this site should not be ignored.
In contrast to the above exogenous infl uences on gonadal
development, endogenous rhythms of reproductive activity have
been proven to occur in temperate fish species and may be
involved in sexual cycling in subtropical and tropical species
but rigorous testi ng of this phenomenon is necessary (Scott
1979, Lam1983). Nevertheless, Schwassmann (l971, 1978) and Bye
(1984) suggested that an endogenous rhythm in tropical fish may
342
take gonadal development to the final stage of maturity which
is then maintained until spawning is triggered by sudden
environmental fluctuations. This phenomenon would produce a
relatively long-lasting phase of spawning-readiness (Schwassmann
1978) and could account for the presence of mature individuals
of some species throughout the year.
Proximal factors influencing the timing of spawning events
in tropical fishes appear to be associated with annual flooding
but the exact stimulus is unclear (Lowe-McConnell 1975,
Welcomme 1979, Scott 1979, Lam 1983). Factors involved as
cues include changes in water chemistry and water temperature,
di ssolved substances from newly wetted soil, flow rates, food
supply and availability of spawning sites among others. Lam
(1983) summarised that no single factor has been identified
a nd perhaps a consortium of factors is i nvo1ved. More speci fi c
factors which induce spawning in some species may be the
presence of a spawning substrate and social cues such as the
visual stimulus of a mate or pheromones (Chen & Martinich
1979, Bye 1984).
In the present study the most 1i kely stimul i for initiating
spawning activity in the species studied were increased flow
rates, water quality changes, particularly dilution effects, and
343
reduced temperatures of flood waters. These factors were also
influential during the early part of the 1982 dry season when
reproduction of most species was apparent. In particular, the
small brood spawners such as c. riisei and H. unilineatus as
well as c. aeneue continued spawning as long as environmental
conditions were favourable. Continual spawning has been noted
in Eigenmannia virescens by Kirschbaum (1979) once water levels
were constant and conductivities were low and despite the
absence of stimul i which were necessary for gonadal recrudescence
(that is rain simulation, rising water level and decreasing
conductivity). Longer than normal spawning periods in the
cardinal tetra, Paracheirodon axeZrodi3 were also related to
hydrological conditions in the Rio Negro basin (Geisler & Annibal
1986). Both E. virescens and P. axel.rodi are multiple spawners
(Schwassmann 1980, Geisler & Annibal 1986). Therefore in some
species, especially multiple spawners, environmental factors
appear to act as initiators of spawning behaviour and as
conditions remain favourable and constant, spawning and repeated
gonadal maturation could occur.
Breeding activity was not noted in the species studied
(except P. reticuZata) during the 1981 dry season which was
characteristically low in rainfall resulting in Stations 2 and
3 bei ng stagnant for up to three months. Stagnant conditions
344
most likely allowed a but l d-up of metabolites which could
repress spawning. Subsequent dilution of these metabolic
wastes at the beginning of the rainy season would then result
in spawning (Johnson 1963, Chen & Martinich 1979). These
metabol ites are thought to be ammonia and are not species
specific (Chen & Martinich 1979). In contrast to this situation
breeding in the dry season or all year round has been recorded
by Kramer (1978a), Lowe-McConnell (1979) and de Silva et al:
(1985) for streams with relatively minor seasonal variation and
continuous flow all year.
Reproductive activity in P. retriculata was continuous
throughout the study period. A comparable lack of seasonality
inbreeding was noted in the poecil i id, Brachyraphis epieoopi.,
in Panama where no monthly variation in population structure,
brood size, oocyte development or numbers of gravid females
were seen (Turner 1938). Other poeciliids, Gambueia affinis
and Kiphophorue spp. were found to have prolonged but seasonal
reproductive cycles curtailed by temperature and photoperiod
in Queensland, Australia where they have been introduced
(Milton & Atherington 1983). Eight cyprinodont species in a
small relatively aseasonal stream in Gabon, West Africa showed
a variety of breeding strategies, some being very seasonal,
others aseasonal (Brosset 1982), a situation similar to that
found by Kramer (1978a) working on characins in Panama.
345
Condition factors of all the species studied (except
P. reticulata), showed a fairly close correlation with seasonal
influences of the reproductive cycle. Generally total condition
increased relative to somatic condition with the onset of
breeding due to increased gonad weights. In some species
(c. eterniol-a , H. unil ineatue and C. aeneus) and especially in
females, decreases in somatic condition also occurred at this
time indicating the mobil isation of somatic reserves into
gonadal tissue (Le Cren 1951, Payne 1975). Both male and
female condition tended to show minimal values after the
reproductive period (for example in C. riisei and H. unilineatus)
and in H. uniZineatus females minimal condition factors were
also noted in the middle of the breeding period when GSI's
decreased. In small species such as C. riisei and H. unilineatus,
very few full y spent females were found at the end of the breedi ng
season and it is 1 i kely that death occurred soon after cessation
of spawning coincident with the low condition of these individuals.
Death of specimens of C. riisei bred in the laboratory within
one or two months after spawning ceased supported this possibility.
During the 1engthy reproductive season however, c. riisei and
c. aeneus in particular, maintained or even showed increases in
condition. Cambray & Bruton (l984) reported that in species
with protracted breeding seasons, the first batches of eggs may
depend on fat reserves accumulated previously while later
346
batches are derived from feeding during the breeding season.
In the small species which bred twice each year, seasonal
variation in condition factors was largely determined by the
reproductive cycle, a situation similar to that reported by
Payne (1975) for Barbus l.ibei-iene ie in Sierra Leone. It was
not possible to investigate condition cycles of immature
individuals because they matured within a year. As a result,
no evidence was obtained as to whether there were any food- or
metabolism-induced changes in condition over the year
comparable with those of temperate fishes. Larger species
whi ch had one major peak of breedi ng each year showed recovery
of condition after breeding (for example A. bimacul.atuei and
inC. aeneue , decreases in condition late in the year might
have been associated with limited food availability during the
second rains or possibly a second breeding period.
Variation in condition for P. reticuZata showed no
seasonal trends consistent with lack of seasonality in breeding
and indicating no obvious seasonal trends in food availability
or metabolism. Dahlgren (1980 a, b) suggested that P. iebioulata
possessed considerable adaptability to food sources of varying
quality. Further, in a study of six cyprinodonts in Texas
which all had protracted breeding periods, de Vlaming et al:
347
(1978) found no correlation between monthly variation of GSI's ,
fat reserves and standi ng crops of phyto- and zooplankton.
The six species studied exhibited varying lengths of their
breeding seasons: short, protracted and continuous. Generally,
an inverse relationship existed between maximumbatch fecundity
and 1ength of the breed i ng season. The speci es with the
highest batch fecundity (A. bimaculatuei spawned for only one
or two months each year with one major peak of'spawning at the
Qeginning of the main rainy season and rarely, a second minor
spawning 1ater in the year. G. etieimiola had a simi 1ar
reproductive pattern and its maximum batch fecundity was the
highest of the other species. c. riisei3 H. uni.l ineatue and
C. aeneus had more prolonged breeding periods up to four
months each season and usually twi ce each year and all had
roughly equivalent ranges of batch fecundity. P. retriculata
produced young continuously through the year tand had the smallest
fecundity of all the species. Batch fecundity was pos itively
correlated with individual size in most species and maximum
batch fecundity was dependent on average species size.
Analyses of spawning patterns indicated that in species
with short breeding seasons such as G. eternicl.a and
A. bimaculatue , ovum maturation was relatively synchronised.
348
however it caul d not be determi ned whether all the eggs were
spawned at one time or in a number of sessions. Ovaries of
H. uniZineatus also indicated some synchrony of development of
ova but laboratory observations showed that these ova were
shed in sma11 broods over 1imited peri ods , the 1e ngth of whi ch
was dependent on the size of the female. C. riisei was also
a small brood spawner and there was evidence of continuous
maturation of ova in individuals which spawned over prolonged
periods. Total egg production was dependent on female size
and for large females greatly exceeded that predicted by
batch fecundities. C. aeneue al so showed evidence of
continuous maturation of ova which indicated the potential for
continual egg production over time although the high levels
of atresia noted might suggest that not all ova produced are
shed.
Data for determining spawning patterns in these species
were obtained from a variety of sources such as GSI fluctuations,
ovarian analyses and laboratory spawning experiments. Certain
shortcomings are inherent in the use of anyone of these
techniques but in general they tended to corroborate each
other. GSI's fell rapidly after a maximumlevel in species
which had 1imited breeding periods such as G. sternicla but in
species with prolonged high GSI levels it could not be certain
349
whether this resulted from protracted breeding by individuals
or successive waves of different individuals. GSI value alone
cannot indicate whether an individual is partially spent or
developing in species where the eggs are not shed all at once
(Delahunty & de Vlaming 1980).
Shortcomings in the use of ovarian analyses have been noted
by several workers especially in situations where the reproductive
cycle of the species is not known adequately (Cambray & Bruton
1984). Macer (1974) found that histologically recognisable
groups of oocytes were not equivalent to groups defined on the
basis of size. In addition, only a 1imited number of batches
of oocytes may be discernable using this method of analysis so
that underestimates of potential spawnings may be made (Miller
1979 a).
Laboratory spawning experiments have one major drawback
in that they can only be marginally indicative of reality since
they are conducted under controlled environmental conditions
and in the absence of competition and predation. Other factors
such as female choice of males depending on the previous
mating history of males and sperm 1imitation (Nakatsuru & Kramer
1982) are important in the design of experimental set ups. The
effects of constant manipulation such as daily removal of
350
spawning substrates and eggs also need to be considered although
Heins & Rabito (1986) found no evidence that removal of ovi-
posited eggs affected spawning frequency in Notropis leedei .
Despite these shortcomings however, it could be concluded
that for all the species studied multiple spawning (or multiple
brood producti on in P. ret-iculata ) was either observed in the
study or recorded in the literature for the same species. The
nature of the multiple spawning strategy exhibited was not
exactly equivalent for all species however. Ovarian maturation
was highly synchronised in A. bimaculatue but more or less
continuous in C. riisei and C. aeneue: G. eteimiol.a and H.
uniZineatus ovaries included one batch of mature ova which was
separated, to varyi ng degrees, from what coul d have been a
second batch. The presence of small eggs in ovarian analyses
need not indicate fractional or multiple spawrrinq and these
oocytes may eventually be resorbed (Nikolsky 1963). This might
be the case in G. e teimiola but laboratory spawning evidence
showed that up to two batches of eggs could be spawned by
H. unil.ineatue during a breeding period of five weeks.
The sizes of clutches spawned and intervals between clutches
for each of these species also differed. A. bimaoul.atue are
reported to spawn 100 to 500 eggs at a time (Azevedo & Vieira
351
1938) and they may have several spawning acts in one season
(Pel izaro et: al 1981) but further details were not available.
Cl utch size of c. aeneue is reported to be between 200 to 380
eggs over periods up to five hours (Hart 1947, Breder & Rosen
1966). Sterba (1962) recorded spawnings occurring at four to
seven day interval s whil e Zuka 1 (1982) estimated two to three
weeks between spawni ng events. Clutch sizes for c. riisei and
H. unilineatus were comparable but C. riisei spawned more
frequently and for longer breeding periods than H. unilineatus.
Hails & Abdullah (1982) distinguished between three
spawning types based on ovarian analyses: single batch total
spawning, multiple batch spawning with two or more distinct
batches of ripening eggs, and multiple batches without clear
distinctions. A. bimaculatus seemed to fall into the single
batch total spawning category although it is still unclear
whether the entire batch is spawned within a short period or
in several spawn i ng acts duri ng the season. G. sternicla
and H. unilineatus seemed to fit the second category while
c. riisei and c. aeneus belong to the third category but
spawni ng patterns in G. sternicla and C. aeneue need to be
confirmed. The repeated brood production of P. reticulata
could be considered a form of multiple batch spawning with
two or more distinct batches complicated by one of these
352
I batches' be i ng develop i ng embryos. Superfoetat i on was not
observed in the present study consistent with Thibault & Schultz'
(1978) observations.
The data ava i 1able on breedi ng periodicity, batch
fecundities and spawning patterns for the six species studied
supported the general conclusions of Lowe-McConnell (1975)
where she described total spawners as having clearly defined
spawning seasons, being more fecund, produci ng numerous small
eggs and making long migrations to do so. In contrast, multiple
spawners have 1ess cl early defi ned breedi ng seasons, produce
fewer eggs at a time, often large ones and may make only local
movements to spawning areas. She however associated multiple
spawning, especially small brood spawning, with some form of
parenta 1 care, a feature not seen in any of the speci es studi ed.
Furthermore, examp 1es of small brood spawners do not include
any small species similar to those in the present study,
presumably because of lack of available data at the time.
It is becoming more widely recognised that some degree of
multiple spawning is of crucial importance to small fish for
a variety of reasons (Nikolsky 1963, Miller 1979 b, Cambray &
Bruton 1984). Most importantly it allows for considerable
increases in fecundity which would not otherwise be possible
353
due to size and nutritional constraints (Nikolsky 1961, 1963,
Loiselle & Welcomme 1971, Miller 1979 b, Wootton 1984 a and
others) .
Enhancement of fecundity in small fish could be achieved
by a reduction in egg size but a lower limit to this process
is set by the increasing disadvantage suffered by smaller and
less provisioned young which ultimately hatch. Miller (1979 b)
estimated that the minimumteleost egg size was 0.25 to 0.40 mm
in several species of gobies. In freshwater habitats, especially
fluviatile ones, and where parental care is non-existent, it
could be expected that minimum egg size would be much larger.
Compared to north European freshwater fish species, egg sizes
of A. bimaoul.atiue, C. riisei and H. unilineatus were below the
modal size of 1.25 - 1.75 mmdiameter, while egg size of
C. aeneus and P. netriculata fell within this modal class
(Wootton 1979). Large egg, size in c. aeneus was similar to
that of the other local callichthyid, Hoploeternum littorale
whose eggs were 1.0 - 1.2 mmin diameter (Singh 1978). Egg
size of P. »et-ioul.atia was dictated by endowment of large
quantities of yolk for embryoni c development since maternal
contributions duri ng gestation are mi nimal (Wourms 1981).
In species where brood size or batch fecundity cannot be
354
increased further, increase in fecundity can be effected by
repeat spawning (Miller 1979 b). This phenomenon has been
reported for a variety of small fishes for example gobies
(Miller 1979 b , 1984), sticklebacks (Wootton 1984 a), darters
(Gale & Deutsch 1985) and minnows (Gale & Gale 1977, Gale &
Buynak 1978, 1982, Heins & Rabito 1986) as well as the
tropical zebra fish (Hisaoka & Firl it 1962) and poecil iids
(Miller 1979 b).
Comparison of reproductive output for the above multiple
spawning species is difficult owing to the different methods
used in expressing the data. Miller (l979 b) calculated
reproductive effort/3D days (total dry weight of offspring/
somatic dry weight of female/3D days) for several species
including a specimen of Brachydanio »ex-io which produced 5,530
eggs over a five month period (data from Hisaoka & Firl it 1962)
and which had a reproductive effort of 0.7725. In comparison,
reproductive effort for P. »etiioul-atia ranged from 0.1061 to 1.504
(Miller 1979 b, Table II).
In summarising their previous studies, Gale & Deutsch
(l985) proposed the use of the term 'proportional fecundity',
that is the vol umetri c ratio of water hardened eggs spawned
to the female that spawned them. In their studies, proportional
355
P. retrioul.atia could be considered a continuous multiple
spawner but when compared to the other species studiedt
reproductive output in the form of numbers of offspri ng
produced, was low. However, the ovoviviparous mode of
reproduction eliminates mortality from extrinsic factors during
the early course of development and permits a saving in the
number of offspri ng that need to be produced for representation
in the next generation (Miller 1979 b). In addition, although
superfoetation does not occur, maturation and fertil isation of
ova takes place over a period of days to reduce the nutritional
dra in on the female than if all the ova were to be endowed
simultaneously (Th i bault & Schultz 1978).
Gale & Deutsch (I985) and Heins & Rabito (I986) discussed
the shortcomings of conventional fecundity estimates for
356
multiple spawning fishes in the light of their results. The
former authors suggested that conventional fecundity estimates
be termed 'ovar ian egg counts I in order to avoid the implication
that batch fecundities are true reflections of potential egg
production in multiple spawning fishes. They suggested that
fecundity shoul d be measured as mean number of eggs spawned per
cl utch times the number of cl utches per year, and such data
should be obtained from captive laboratory or natural environment
spawning experiments.
In addition to allowing significant increases in fecundity
for small fi sh , a multi p1e spawni ng strategy has other advantages.
It is widely recognised that several batches of eggs spawned
over time reduces the chances of loss of all eggs by unfavourable
conditions (Nikolsky 1963, Lowe-McConnell 1975, 1979 and others).
In situations such as the study site where the beginning of the
rains are unpredictable and often followed by dry spells, this
is especially advantageous. Multiple spawning also allows use
of a spawni ng site more than once each season (Cambray & Bruton
1984), reduces intra- and interspecific competition between
fry (Nikolsky 1961, Cambray & Bruton 1984), may produce year
classes of uniform strength (Gale & Gale 1977), and allows
for control of egg numbers duri ng the current season by
continuous ovum maturation or resorption depending on ensuing
357
conditions (Macer 1974).
The observed correlations between size and the reproductive
characteristics of the six species studied raises some
interesting but speculative ideas on the interaction between
size and phylogeny in shaping reproductive and life history
patterns of these fishes. Of the six species studied, the
four characoids constituted one closely related group within
which a spectrum of reproductive strategies was noted, from
what appears to be relatively synchronised total spawning in
A. bimaculatue and to a lesser extent in G. eternio La, to
protracted small brood spawning in C. riisei and H. unilineatus.
The large size of A. bimaculatue could provide the potential for
the synchronised maturation of a 1arge number of eggs for any
one season. As a result of smaller size and a somewhat
restricted abdominal cavity due to its unusual shape, that
potential would be reduced in G. et.ex-niola, Small brood
spawning in H. uniZineatus an~ C. riisei could be related
directly to small size acting as a constraint to increased
batch fecunditi es. The fact that H. unilineatus may be closely
rel ated to A. bimacuZatus by the process of paedomorphosis
(Gery 1977) might explain the relative synchrony of ovum
maturation and spawni ng bei ng 1imited to fairly short periods
in the former species. In fact, Goul d (1977) suggested that
358
r-selection and early maturation could be related to the process
of paedomorphosis. However, C. riisei being a glandulocaudine,
is capable of continuous ovum maturation and spawning because
internal fertil isation has allowed ovulation and spawning to
be dissociated, both temporally and spatially, from male presence
and mati ng events (Nelson 1964). The importance of size as a
factor infl uencing the pattern of covariation of 1 ife history
traits has been noted by Stearns (1983) who found that it
accounted for a significant portion of ordering species onto a
"slow-rast ' continuum although he did also note the effects of
phylogenetic constraints on the evolution of life history traits.
In summary, the fish community at the study site could be
considered to be a diverse assemblage which might be maintained
by two major features of the area. The seasonally fluctuating
nature of the stream in conjunction with a 19n9 flow regime and
extensive refuge pools may have created an environment with
intermediate levels of disturbance which maintained non-equilibrium
communities of high species richness. Secondly, species richness
could be related to the proximity of the study area to a source
of high diversity, that is the mainl and, from which continual
colonisation has taken place.
The fi sh popul ations studied seemed to be strongly infl uenced by
359
the seasonal flow regime at the study sites. Possible mechanisms
may have been mortality, dispersal and concentration of
individuals or the stimulation or inhibition of reproduction.
Life history characteristics of the six species studied
reflected the impact of seasonal conditions on populations and
were characteristically r-selected traits such as small size,
short 1ife spans, rapid development, early maturation and
high early mortal ity. Biotic interactions such as predation
may also have been influential, especially on the smaller species
such as P. r-et-ioulata, Populations also showed r-selected
characteristics such as high variability in size and dominance
of age structure by recruit spawners.
Reproductive seasonality in most species studied was
determined primarily by the flood-drought cycle which may have
acted to both stimulate and inhibit breeding activity. Features
associated with the floods such as increased flow and water
quality changes may have been influential in stimulating
reproduction. Peaks of reproductive activity coincided closely
with rainfall maxima each year for all seasonally breeding
species. Numbers of breedi ng seasons per year and their 1engths
varied for different species. Some species had one major
breeding period and rarely, a second minor period each lasting
360
not more than one or two months (c. etemicl.a and A. bimaculat.ue t ,
Others consistently bred twice each year for periods up to
four months (c. r-i.ieei , H. uni.l-i.neatue and c. aeneus).
P. »et icul.ata showed continuous reproductive activity through-
out the study with no obvious seasonal variation in intensity.
Variability in the seasonal regime as a result of altered
timing of ra i nfa 11 and stream flow were responded to by
variability in reproductive timing in all seasonally breeding
species indicating a high degree of adaptability in these
species. The condition of fish was affected primarily by their
reproductive cycles and there was no direct evidence of other
factors such as food availability being highly influential.
Spawning patterns of the six species studied were variable
but consistent with those described for other seasonal tropical
rivers. They ranged from what appeared to be a seasonal and
highly synchronised total spawning pattern in A. bimacuiatue
to continuous small brood production in P. ret-icul.atia, Ovarian
analyses of A. bimaculatue suggested a total spawning pattern
but no laboratory confirmation was obtained and the 1 iterature
is ambiguous. Multiple spawni ng was confirmed inC. riisei and
H. uniZineatus by laboratory experiments and ovarian analyses.
Some degree of multiple spawning was suggested for c. eteimiola
361
and C. aeneue by ovarian analyses but this was unconfirmed by
1aboratory experi ments.
Maximum batch fecundities were directly proportional to
the size of the species but in the smaller species high fecundity
was attained by means of a multiple spawning strategy which
-
overcame the constraints imposed by small size. Multiple
spawning in turn increased the adaptability of the species by
allowing variabil ity in reproductive output and timing in
addit i on to other advantages.
Generally, the reproductive strategies of the six fish
species studied were strongly i nfl uenced by the seasonal
nature of the environment. However, the strategies used were
diverse and highly adaptable to the inherent stochasticity of
'""' .......~.
the system.
362
CONCLUSIONS
At the beginning of this study several predictions were
made as to the nature of the habitat and the responses of the
bi ota to seasonal changes. An assessment of the data shows
that most of these predictions made were true.
The streams studied were highly seasonal in nature with
variable physical and chemical parameters which in turn influenced
the biota including the fishes studied. Predictability of
conditions was low with duration, onset and intensity of any
season being highly variable. Hydrological patterns of the
streams were equivalent to those recorded for long-flow inter-
mittent streams with extens i ve refuge pools (Cl i fford 1966) and
differed somewhat from more highly intermittent tropical streams
where regular and extensive drying out was noted (Adebisi 1981 a).
On the other hand, the Chatham streams were much more seasonal
and were harsher environments than many small forest creeks or
larger streams studied elsewhere in the tropics (for example
Bishop 1973, Schwassman 1978, Kramer 1978a and others ).
Biological standing crops at the study sites were variable
showing no striking or consistent seasonal trends and high
levels were attained in both wet and dry seasons unlike more
363
obviously seasonal habitats such as floodplain rivers (Lowe-
McConne11 1975, We1conme 1979). Standi ng crops approximated
those attained in some aquatic habitats on the mainland.
Life history parameters of the fishes studied showed
r-selected traits as predicted including small size, short
life spans, rapid development, early maturation and high
early mortality. It was possible that high levels of predation
may have interacted with environmental factors to emphasise
these tra its in some speci es •
Despite the range of habits, sizes and phylogeny of the
fishes studied, their reproductive strategies strongly
reflected the seasonal ity of the environment and with the
exception of P. »etiiculata which bred continuously, all other
species had peaks of reproductive activity in the wetter parts
of the year. However, although onset of reproduction was
usually cued to the rainfall maxima, duration and intensity of
breeding during any season varied depending on the ensuing
environmental conditions. Breeding was recorded for some
species as long as conditions of flow were maintained.
Since breeding normally coincides with the most favourable
time for survival of young, it could be assumed that conditions
364
during the rainy season were favourable for survi va 1 of young
in the species studied.' Such factors as decreased crowding
and predation pressures could be proposed as ultimate factors
influenci ng reproductive timi ng in these species. Food
availabil ity could not be assessed definitively as a factor
because of lack of data on adult and larval feeding habits.
However, food limitation for adults did not appear to be
important because of the lack of clear seasonal variation in
food supply and the recorded feedi ng habits for many species
indicate a high degree of adaptability.
Breeding was generally synchronised in most species to
peak during the rainy season and massive production of young
in- species such as A. bimaculatue , H. uniZineatus and C. riisei
at approximately the same time may imply that swamping of
predators might be an important consideration. Kramer's (l978a)
proposition that phylogenetic constraints may be influential
also apply to this study. Many of the species studied are also
found in seasonally flooded environments on the mainland where
it is advantageous to breed at times of high water. Physiological
responses to initiate reproduction upon certain external cues
have evolved (for example Kirschbaum 1975) and it is unlikely
that such responses would change except under very strong and
contrary selection pressures. In addition, inhibition of
365
reproductive activity in stagnant pools by accumul ated metabol ites
or high conductivity and subsequent release of inhibition by
increased discharge could result in coordinated rainy season
spawning.
It was predicted that multiple spawning should occur in
a high frequency of fishes due to its advantageous nature in
unpredictable habitats. Multiple spawning (or brood production)
was definitely observed in three out of six species studied
and can be tentatively suggested for the other three species
on the basis of circumstantial evidence.
In addition to multiple spawning which itself allows for
variable reproductive output, reproductive strategies varied
depending on the number of breeding periods each year and the
length of the breeding season. Obvious reproductive Dutput
in the form of amount of gonadal tissue and duration or
intensity of breeding was considered but other aspects of
output should also be taken into account, for example courtship
and other behaviour, coloration and sexual dimorphism (Williams
1966) and these also varied between species. In general terms,
the predicted overall reproductive strategy in this fluctuating
habitat would be to expend high reproductive effort. However,
the tactics (cf Wootton 1984 b) by which this high output was
366
achieved differed due to phylogenetic and size constraints.
In conclusion, the six fish species studied were the
most common in the Chatham streams and as such coul d be
considered to be successful in this type of habitat. Their
reproductive strategies were diverse in nature differing in
timing, duration and intensity of breeding between species and
in different years. In general, they were highly adaptable
to the inherent stochasticity of the environment and this
undoubtedly contributed to their success. In this situation,
the relationship between strategy and the environment is best
surrrnarised by Dobzhansky (1950, p. 216): 'Chanqeab l e environ-
ments put the hi ghest premium on versatil ity rather than on
perfecti on in adaptation. I
367
BIBLIOGR-WHY
Abell, D.A. 1959. Observations on mosquito popul at ions of an
intermittent foothills stream in California. Ecology 40:
186-193.
Adams, B.E. 1946 . Spawni ng Corudonae aeneue . Aquarium,
Philad. 15(2):21-23.
Adebisi, A.A. 1981 a. The phys tco-chemical hydrology of a
tropical seasonal river - upper Ogun River. Hydrobiologia
79: 157-165.
Adebisi, A.A. 1981 b. Analyses of the stomach contents of
the piscivorous fishes of the upper Ogun River in Nigeria.
Hydrobiologia 79:167-177.
Alexander, R. MeN. 1965. Structure and function in the catfish.
J. Zoo l , Lond. 148:88-152.
Alkins,M.& G. De Souza. 1983/84. Two new freshwater fish
records for Trinidad and some comments on the zoogeography
of the southwestern peni nsul a. Living World 1983/1984:
8-12.
American Publ ic Health Association (APHA). 1971. Standard
Methods for the Examination of Water and Wastewater.
13th edition. American Public Health Association,
Washington, 874p.
Awachie, J.B.E. 1981. Running water ecology in Africa. Pages
339-366 in M.A. Locke & D.O. Williams (eds ) , Perspectives
in Running Water Ecology. Pl enum Press, N. Y•
Azevedo, P. de & B.B. Vieira. 1938. Contr tbu icdo para 0
catalogo biologico dos peixes fluviais do nordeste do
Brasil. Boletim da Inspetoria Federal de Obr-ae contra as
SecasJ Rio de Janeiro 1938(3):1-13. (In Portuguese, pp.1-7
seen).
Bacon, P.R. 1968. The ecology of Caroni Swamp, Trinidad, with
special reference to the plankton and cirripedes. PhD thesis,
University of the West Indies, Trinidad, 252p.
368
Bacon, P.R., J.S. Kenny, M.E. Alkins, S.N. Mootoosingh,
E.K. Ramcharan & G.S.B. Seeberan. 1979. Studies on the
biological resources of Nariva Swamp, Trinidad. Occasional
Papers No.4, Zoology Department, U.W.I., Trinidad, 455p.
Baerends, G.P., R. Brouwer & H.T.J. Waterbolk. 1955. Ethological
studies on Lebistes ret icul-atiue (Peters). I. An analysis
of the male courtship pattern. Behaviour 8:249-334.
(Quoted in Liley 1966).
Bagenal, LB. 1978. Aspects of fish fecundity. Pages 75-101
in S.D. Gerking (ed). Ecology of Freshwater Fish Production.
John Wil ey, New York.
Bagenal, LB. & E. Braum. 1978. Eggs and early 1 ife history.
Pages 165- 20 1 in T. Bagena 1 (ed). Methods for Assessment
of Fish Production in Fresh Waters. 3rd ed, Bl ackwell
Scientific Publications, Oxford.
Bagena 1, T. B. & F. W. Tesch. 1978. Age and growth. Pages 101-
136 in T. Bagena 1 (ed). Methods for Assessment of Fish
Production in Fresh Waters. 3rd ed, Blackwell Scientific
Publications, Oxford.
Bailey, N.T.J. 1981. Statistical Methods in Biology. 2nd ed,
John Wiley, New York, 216p.
Balek, J. 1983. Hydrology and Water Resources in Tropical
Regions. Elsevier, N.Y., 271p.
Balon, E.K. 1984. Patterns in the evolution of reproductive
styles in fishes. Pages 35-53 in G.W. Potts & R.J. Wootton
(eds). Fish Reproduction. Academic Press, London.
Balon, E.K. 1985. Additions and amendments to the classification
of reproductive styles in fishes. Pages 59-72 in E.K. Balon
(ed). Early life histories of fishes. W. Junk, Dordrecht.
Barr, K.W. & J.B. Saunders. 1965. An outline of the geology
of Trinidad. Fourth Caribbean Geological Conference,
Tri ni dad, lOp.
Barus, V., J. l.tbos varsky & J. de la Cruz. 1980. Observations
on Limia vittata (Poeciliidae) from Kuba. Folia Zooloqica
29(3):267-287. .
369
Bayley, P.B. 1980. The limits of limnological theory and
approaches as applied to river-floodplain systems and
their fish production. Pages 739-746 in J.I. Furtado
(ed). Tropical Ecology and Development. Proc. 5th
International Symposium of Tropical Ecology, Kuala Lumpur.
Beadle, L.C. 1974. The Inland Waters of Tropical Africa.
Longman, London, 365p.
Beard, J.S. 1946. The Natural Vegetation of Trinidad. Oxford
Univers ity Press, Oxford, 152p.
Beaumont, P. 1975. Hydrology. Pages 1-38 in B.A. Whitton (ed).
River Ecology. University of California Press, Berkeley,
725p.
Berridge, C.E. 1981. Climate. Pages 1-12 in St. G.C. Cooper &
P. R. Baco n ( eds ) . The Natura L Resou:t>cesof Trinidad and
Tobago. Edward Arnold, London, 223p.
Bertness, M.D. 1981. Seasonal ity in tropical hermit crab
reproduction in the Bay of Panama. Biotropica 13(4):
2.92-300 .',
Beumer, J.P. 1979. Reproductive cycles of two Australian
freshwater fi shes: the spangl ed perch, Therapon unicolor
Gunther, 1859 and the East Queensland rainbowfish,
Nematocentris splendida Peters, 1866. J. Fish Biol. 15:
111-134.
Beumer, J.P. 1980. Hydrology and fish diversity of a North
Queensland tropical stream. Aust. J. Ecol-, 5(2): 159-186.
Billard, R. & B. Breton. 1978. Rhythms of reproduction in
teleost fish. Pages 31-53 in J.E. Thorpe (ed). Rhythmic
Activity in Fishes. Academi c Press, N. Y •
Bishop, J.E. 1973. Limnology of a small Malayan River Sungai
Gombak. W. Junk, The Hague, 493p.
Boeseman, M. 1960. The freshwater fishes of the island of
Trinidad. stud. Fauna Curacao 10(48) :73-153.
Boeseman, M. 1964. The freshwater fishes of the island of
Trinidad. Addenda, Errata et Corrigenda. stud. Fauna
Curacao 20 (82) :~2-57 .
370
Bowden, B.S. 1970. The relationship of light and temperature
to reproduct i on in the guppy, Poeci.l ia reticulata Peters.
PhD thesis, University of Connecticut. (Quoted in Miller
1979 b).
Breder, C.M. 1927. The fishes of the Rio Chucunaque drainage,
eastern Panama. Bull. Am. Mus. nat. Hiet . 57(3) :91-176.
Breder, C.M. & C.W. Coates. 1932. A preliminary study of
population stability and sex ratio of Lebi.e tee . Copeia
1932 (3): 147-155. (Quoted in Dahlgren 1979).
Breder, C.M. & D.E. Rosen. 1966. Modes of Reproduction in
Fishes. Natural History Press, New York, 941p.
Bright, G.H. 1982. Secondary benthic production in a tropical
island stream. Limnol . Oceanoqr , 27:472-480.
Brosset, A. 1982. Le peuplement de cyprinodontes du Bassin de
L'lvindo, Gabon. Rep. Eool., (Terre Vie) 36:230-292.
(In French).
Brown, J.H. 1971. The desert pupfish. Pages 65-71 in
Ecol.oqu , Evolution and Population Biology. Readings from
Scientific American, W.H. Freeman, San Francisco.
Brown, L.P. 1982. Can guppies adjust the sex ratio? Am.
Nat. 120(5):694-698.
Bullough, W.S. 1961. Vertebrate Reproduct-ive Cycles. Methuen,
London.
Burky, A.J. 1973. Organic content of eggs and juveniles of an
amphibious snail, Pomacea urceue (Muller) from the
Venezuelan savannah and their ecological significance.
Malac. Rev. 6: 59.
Burky, A.J. 1974. Growth and biomass production of an
amphibious snail, Pomacea urceue (Muller), .f rom the
Venezuelan savannah. Proc , ma'l.ac; Soc. Lond. 41:127-143.
Burky, A.J., J. Pacheco & E. Pereyra. 1972. Temperature,
water, and respiratory regimes of an amphibious snail,
Pomacea urceus (Muller), from the Venezuelan savannah.
Biol. Bull. 143:304-316.
Burky, K.A. & A.J. Burky. 1977. Buoyancy changes as related
to respiratory behaviour in an amphibious snail, Pomacea
urceue (Muller), from Venezuela, Nautilus 91(3):97-104.
371
Burton, J.D. 1976. Basic properties and processes in estuarine
chemistry. Pages 1-36 in J.D. Burton & P.S. l.'i s s (eds).
Estuarine Chemistry. Academi c Press, London, 229p.
Bye, V.J. 1984. The role of environmental factors in the
timing of reproductive cycles. Pages 187-205 in G.W. Potts
& R.J. Wootton (eds). Fish Reproduction. Academi c Press,
London.
Caesar. K. 1985. The effects of quarrying activities on the
aquatic fauna of the Acono River. Final year undergraduate
report, Zoology Department, U.W.I., Trinidad, 47p.
Cambray, J.A. & M.N. Bruton. 1984. The reproductive strategy
of a barb, Barbue anopl.ue (Pisces: Cyprinidae), colonizing
a man-made lake in South Africa. J. Zool., Land. 2C,.1:
143-168.
Canton, S.P., L.D. Cline, R.A. Short & J.J. Ward. 1984. The
macroi nvertebrates and fi sh of a Colorado stream dur ing a
period of fluctuating discharge. Ereehia, Biol. 14:311-316.
Carrington, L. 1980. A study of the physical characteristics
of a tropical stream. Final year undergraduate report,
Zoology Department, U.W.I., Trinidad, 56p.
Carter, G.S. 1934. Results of the Cambridge Expedition to
British Guiana, 1933. The freshwaters of the rainforest
areas of British Guiana. J. Linn. Soc. Land. (Zool.)
39(264): 147-193.
Carter, G.S. 1935. Reports of the Cambridge Expedition to
British Guiana. Respiratory adaptations of the fishes
of the forest waters, with descriptions of the accessory
respiratory organs of Electrophorus electx-icue (Linn.) and
Plecoe tomue p Lecoetiomue (Linn.). J. Linn. Soc. Lond.
(Zool., ) 39(265) :219-233.
Chen, L. & R.L. Martinich. 1979. Pheromonal stimulation and
metabolite inhibition of ovulation in the zebrafish,
Brachydanio nei-io, Pages 394-398 in M.S. Love & G.M.
Caillet (eds). Readings in Ichthyology. Goodyear
Publications, California.
Cl arke, G.M. 1980. Statistics and Experimental Design. 2nd
ed , Edward Arnold, London, 188p.
372
Clifford, H.F. 1966. The ecology of invertebrates in an
intermittent stream. Invest. Indiana Lakes Streams 7
(2):57-98.
Cole, G.A. 1979. Textbook on Limnology. C.V. Mosby, St Louis,
426p.
Collins, J.P., C. Young, J. Howell & W.L. Minckley. 1981.
Impact of flooding in a Sonoran Desert stream, including
elimination of an endangered fish population (Poecil.iopei.e
o. occodent al.ie, Poeciliidae). SWest. Nat. 26:415-423.
Connell, J.H. 1978. Diversity in tropical rain forests and
coral reefs. Science 199: 1302-1310.
Cummins, K.W. 1974. Structure and function of stream eco-
systems. BioScience 24:631-641.
Cummins, K.W., J.R. Sedell, F.J. Swanson, G.W. Minshall,
S.G. Fisher, C.E. Cushing, R.C. Petersen & R.L. Vannote.
1983. Organi c matter budgets for stream ecosystems:
problems in their evaluation. Pages 299-353 in J.R. Barnes
& G.W. Minshall (eds). Stream Ecology. Plenum Press ;
N.Y., 399p.
Dahlgren, B. T. 1979. The effects of population density on
fecundity and fertil ity in the guppy, Poecilia reticulata
(Peters). J. Fish BioZ. 15:71-91.
Dahlgren, B.T. 1980 a. The influence of three dietary non-
protei n 1eve 1s on fecundity and fertil ity in the guppy
PoeciZia re t-ioulata, Oikos 34: 337-346.
Dahlgren, B.T. 1980 b. The effects of three different dietary
protei n 1eve 1 s on the fecund i ty in the guppy, Poeci l.ia
reticulata (Peters). J. Fish BioZ. 16(1) :83-98.
Darlington, P.J. 1957. Zoogeography. John Wiley, New York,
675p.
Davies, B.R. 1979. Stream regulation in Africa: a review.
Pages 113-142 in J.V. Ward & J.A. Stanford (eds). The
Ecology of Regulated Simeame , Plenum Press, N.Y.
Davies, B.R., T. Davies, J. Frazer & F.M. Chutter. 1982. A
Bibliography of African Inland Water Invertebrates (to 1980).
Council for Scientific and Industrial Research, Pretoria,
418p.
373
Davies, B.R. & R.C. Hart, 1981. Invertebrates. Pages 51-68
in J.J. Symoens, M. Burgis & J.J. Gaudet (eds). The
Ecology and Utilization of African Inland Waters. UNEP.
Deacon, J.E. & W.L. Minckley. 1974. Desert fishes. Pages
385-488 in G. W. Brown Jr (ed). Desert Bio logy. Academi c
Press, N. Y.
de Godoy, M.P. 1975. Peixes do Brasil. Franciscana,
Piracicaba SP, 846p. (In Portuguese).
Delahunty, G. & De Vlaming, V.L. 1980. Seasonal relationships
of ovary wei ght, 1 iver wei ght and fat stores wi th body
wei ght in the go1dfi sh, Caraee iue auraiue (L.). J. Fish
BioZ. 1 6 : 5 -13 .
De Silva, S.S., J. Schut & K. Kortmulder. 1985. Reproductive
biology of six Bacbue species indigenous to Sri Lanka.
Environ. Biol. Fishes 12(3) :201-218.
De Vlaming, V.L. 1974. Environmental and endocrine control of
teleost reproduction. Pages 13-83 in C.B. Schreck (ed).
Control of Sex in Fishes. Virginia Polytechnic Institute
and State University, Blackburg.
De Vlaming, V.L., A. Kuris & F.R. Parker. 1978 .. Seasonal
variation of reproduction and 1 ipid reserve in some
subtropi cal cyprinodontids. Trans. Am. Fish. Soc. 107(3):
464-472 .
Dildine, G.C. 1936. The effect of light and temperature on
the gonads of Lebistes. Anat. Rec, 67:61-66. (Quoted
in Dahlgren 1979).
Dobzhanskyv T. 1950. Evolution in the tropics. Am. Sci. 38:
209-221.
Dudgeon, D. 1983. The importance of streams in tropical rain-
forest systems. Pages 71-82 in S.L. Sutton, TvC. Whitmore
& A.C. Chadwick (eds). Tropical rain forest. Blackwell
Scientific, Oxford, 498p.
Dussault, G.V. & D.L. Kramer. 1981. Food and feeding behaviour
of the guppy, Poecilia ret-ioul.at:a (Pisces: Poecil iidae).
Can. J. Zoo l., 59(4) :684-701.
Endler, J.A. 1978. A predator t s view of animal colour patterns.
Evol. Biol. 11:319-364.
374
Endler, J.A. 1980. Natural selection on colour patterns in
PoeciZia reticulata. Evolution 34(1):76-91.
Endler, J.A. 1984. Natural and sexual selection on colour
patterns in poeciliid fishes. Pages 95-111 in LM. Zaret
(ed). Evolutionary ecology of neoteopi.oal. freshwater
fishes. W. Junk, The Hague.
Erdman,D.S. 1982. Sicydium pl.umi.er-i (Bloch, 1786), Family
Gobiidae, reproductive cycle of a fish in the mountain
streams of Puerto Rico. Poster paper presented at the
Fisheries Society of the British Isles International
Symposium, Plymouth, Devon, July 1982.
Extence, C.A. 1981. The effect of drought on benthic invertebrate
communities in a lowland river. Hydrobiologia 83(2)
217-224.
Farnworth, E.G. & F.B. Golley (eds). 1974. Fragile Ecosystems.
Springer-Verlag, N.Y., 258p.
Farr, J.A. 1976. Social facilitation of male sexual behaviour,
i ntrasexua 1 compet it i on, and sexua 1 select i on in the guppy,
Poeci.l.ia reticuZata (Pisces: Poeciliidae). Evolution 30:
707-717.
Farr, J.A. 1977. Male rarity or novelty, female choice
behaviour, and sexual selection in the guppy, PoeciZia
reticulate (Pisces, Poeciliidae). Evolution 31:162-168.
Farr, J.A. 1980. Social behaviour patterns as determinants
of reproducti ve success in the guppy, Poecilia reticulata
Peters (Pisces: Poeciliidae). Behaviour 74:38-91.
Farr, J.A. 1981. Biased sex ratios in laboratory strains of
guppies, Poecilia recicul ata, Heredity 47(2):237-248.
Fisher, S.G. 1983. Succession in streams. Pages 7-27 in
J.R. Barnes & G.W. Minshall (eds). Stream Ecology.
Plenum Press, N.Y.
Fisheries Division. 1976. Annual Report. Ministry of
Agriculture, Lands and Fisheries, P.O.S., Trinidad, 24p.
Fittkau, E.J. 1964. Remarks on limnology of Central-Amazon
rain-forest streams. Yerh, into Verein. Limnol , 15:
1092-1096.
375
Foldats, E. 1962. La concentracion de oxigeno disuelta en las
aguas negras. Acta biol.. Venez. 3(10):149-159 (In
Spani sh ) .
Frey, H. 1961. Illustrated Dictionary of Tropical Fishes.
T.F.H. Publications, New Jersey.
Furch, K. & W.J. Junk. 1980. Water chemistry and macrophytes
of creeks and rivers in southern Amazonia and the Central
Brasilien shield. Pages 771-796 in J.I. Furtado (ed).
Tropical Ecology and Development. Kuala Lumpur.
Gale, W.F. & G.L. Buynak. 1978. Spawning frequency and fecundity
of satinfin shiner (Notropis analoetanue i - a fractional,
crevice spawner. Trans. Am. Fish. Soc. 107:460-463.
Gale, W.F. & G.L. Buynak. 1982. Fecundity and spawning
frequency of the fathead minnow - a fractional spawner.
Trans. Am. Fish. Soc. 111 (1) :35-40.
Gale, W.F. & W.G. Deutsch. 1985. Fecundity and spawning
frequency of captive tessellated darters - fractional
spawners. Trans. Am. Fish. Soc. 114:220-229.
Gale, W.F. & C.A. Gale. 1977. Spawning habits of spotfin
shiners (Notropis epi.lop terue ) - a fractional, crevice
spawner. Trans. Am. Fish. Soc. 106:170-177.
Geisler, R. & S.R. Annibal. 1986. Ecology of the cardinal-
tetra Paracheirodon axel.rodi. (Pisces, Characoidea) in
the river basin of the Rio Negro/Brazil as well as
breeding-related factors. Anim. Res. Dev. 23:7-39.
Geisler, R., H.A. Knoppel & H. Sial i. 1975. The ecology of
freshwater fi shes in Amazonia: present status and future
tasks for research. Anim. Res. Dev. 1:102-119.
Geodakyan,V.A. & V.I. Kosobutskii. 1969. Nature of the
feedback mechanism of sex regul ation. Genetika 5(6):
119-126. (Quoted in Brown 1982).
Geodakyan, V.A., V.l. Kosobutskii & D.S. Bileva. 1967.
Regulation of sex ratio by negative feedback. Genetika
3(9):152-163 .(Quoted in Brown 1982).
Gery, J. 1969. The freshwater fi shes of South Ameri ca. Pages
828-848 in LJ. Fittkau et al. (eds). Biogeography and
Ecology in South America~ Vol. 1. W. Junk, The Hague.
376
Gery, J. 1977. Characoids of the World. T.F.H. Publications,
New Jersey, 672p.
Gi bon, F.-M. & B. Statzner. 1985. Longitudi na 1 zonation of
lotic insects in the Bandama River system (Ivory Coast).
HydrobioZ.ogia 122:61-64.
Golterman, H.L. 1975. Chemistry. Pages 39-80 in B.A. Whitton
(ed). River Ecology. University of California Press,
Berkel ey.
Gould, S.J. 1977. Ontogeny and Phylogeny. Harvard Univers ity
Press, Cambridge, Mass. 501p.
Goulding, M. 1980. The Fishes and the Forest. University of
California Press, Los Angeles, 280p.
Grant, P.M. & K.W. Stewart. 1980. The life history of
Isonychia ei cca (Ephemeroptera: 01 igoneuridae) in an
intermittent stream in north central Texas. Ann. ent.
Soc. Am. 73(6):747-755.
Griffiths, J.F. 1959. Bioclimatology and the meteorological
services. Pages 282-300 in Proceedings of the Symposium
on Tropical Meteorology in Africa, Nairobi, Kenya.
Grossman, G.D., P.B. Moyle & J.O. Whitaker. 1982. Stochasticity
in structural and functional characteristics of an Indiana
stream fish assemblage: a test of community theory. Am. Nat.
120(4) :423-454.
Guppy, P.L. 1934. Observations on Trinidad larvicidal fishes.
Trop . Agric.~Trin. 11(5):117-122.
Guppy, P.L. 1936. A descriptive catalogue of the fishes of
Trinidad & Tobago. Memoirs of the Department of Agriculture,
Trinidad & Tobago, 165p.
Hails, A.J. & Z. Abdullah. 1982. Reproductive biology of the
tropical fish Trichogaster pectoralis (Regan). J. Fish
Biol. 21(2):157-170.
Harrison, A.D. 1966. Recolonisation of a Rhodesian stream
after drought. Arch. HydrobioZ.. 62:405-421.
Harrison, A.D. & J.J. Rankin. 1975. Forest litter and stream
fauna on a tropical island, St Vincent, West Indies.
Verh. into Verein. Limnol. 19: 1736-1745.
377
Harrison, A.D. & J.J. Rankin. 1976 a. Hydrobiological studies
of Eastern Lesser Antillean islands. I. St Vincent: Fresh-
water habitats and water chemistry. Arch. Hydrobiol.
Monogr. Beitr. Suppl. 50:96-144.
Harrison, A.D. & J.J. Rankin. 1976 b. Hydrobiological studies
of Eastern Lesser Antillean islands. II. St Vincent:
Freshwater fauna - its distribution, tropical river
zonation and biogeography. Arch. Hydrobiol. Monogr. Beitr.
Suppl. 50:275-311.
Harrison, A.D. & J.J. Rankin. 1978. Hydrobiological studies
of Eastern Lesser Antillean islands. III. St Vincent:
Freshwater Moll usca - thei r di stribution, popul ation
dynamics and biology. Arch. Hydrobiol. Monogr. Beitr.
Suppl., 54(2):123-188.
Hart, C.W. & D.G. Hart. 1969. A contribution to the limnology
of Dominica, West Indies. Proc • Acad, nat. Sci. Phi lad,
121: 109-126.
Hart, F.B. 1947. Corydoras aeneus and their spawning. Aquar.
J.18(3):1O-13,29.
Head, P.C. 1976. Organic processes in estuaries. Pages 53-91
in J.D. Burton & P.S. Liss (eds). Estuarine Chemistry.
Academic Press, London.
Heins, D.C. & F.G. Rabito. 1986. Spawning performance in
North American minnows: direct evidence of the occurrence
of multiple clutches in the genus Notropis. J. Fish Biol.
28:343-357.
Hester, F.J. 1964. Effects of food supply on fecundity in the
female guppy Lebistes reticulatus (Peters). J. Fish. Res.
Bd Can. 21: 757-764.
Hildemann, W.H. & E.D. Wagner. 1954. Intraspecifi c sperm
compet it i on in Lebistes. Am. Nat. 88:87-91.
Hisaoka, K.K. & C.F. Firlit. 1962. Ovarian cycle and egg
production in the zebrafish, Brachydanio rerio. Copeia
1962:788-792.
Holden, M.J. 1963. The populations of fish in dry season
pool s of the River Sokoto. Fishery Publ.e colon. Off.
19:1-58.
378
Holden, M.J. & J. Green. 1960. The hydrology and plankton of
the River Sokoto. J. Anim. Ecol. 29:65-84.
Hornbach, D.J., C.M. Way& A.J. Burky. 1980. Reproducti ve
strategies in the freshwater sphaeriid clam, Musculiwn
partumei.um (Say), from a permanent and temporary pond.
Oecol.oqi.a (Berl.) 44: 164-170.
Horowitz, R.J. 1978. Temporal variability patterns and the
distributional patterns of stream fishes. Ecol. Monogr.
48:307-321.
Hunte, W. 1976. Biological studies of freshwater shrimps
(Atyidae and Palaemonidae) in Jamaica. PhD thesis,
University of the West Indies, Jamaica, 379p.
Hurdle, M.T. 1974. Life history studies and habitat require-
ments of the apple snail at Lake Woodruff National
Wildl ife Refuge. Proc. Southeast Association of Game
& Fish Commissioners 27:215-224.
Hurlbert, S.H., G. Rodriguez & N.D. dos Santos (eds). 1981.
Aquatic Biota of Tropical South America. 2 vol s ,
San Diego State University, San Diego, 323+298p.
Hurlbert, S.H. & A. Villalobos-Figueroa (eds). 1982. Aquatic
Biota of Mexico, Central America and the West Indies.
San Diego State Univers ity, San Diego, 529p.
Hutchinson, G.E. 1957. A Treatise on Limnology. Vol. 1.
Geography, physics and chemistry, Part 2: Chemistry of
lakes. John Wiley, N.Y.
Hutchinson, G.E. 1967. A Treatise on Limnology. Vol. 2.
Introduction to lake biology and the limnoplankton.
John Wiley, N.Y., 1115p.
Hynes, H.B.N. 1958. The effect of drought on the fauna of a
small mountain stream in Wales. Verh. into Verein. Idmnol.,
13:826-833.
Hynes, H.B.N. 1960. The Bio logy of Pol.luted Waters. Liverpoo 1
Univers i ty Press, Liverpoo 1, 202p.
Hynes, H.B.N. 1970. The Ecology of Running Waters. Liverpool
Univers ity Press, Liverpool, 555p.
379
Hynes, H.B.N. 1971. Zonation of the invertebrate fauna in a
West Indian stream. Hydrobiologia 38(1) :1-8.
Hynes, H.B.N. 1975. The stream and its valley. Verh. into
Verein. Limnol. 19:1-15.
Hynes, J.D. 1975. Annual cycles of macro-invertebrates of a
river in southern Ghana. Preehia.iBiol., 5:71-83.
Irmler, U. 1975. Ecological studies of the aquatic soil
invertebrates in three inundation forests of Central
Amazonia. Amazoniana 5:337-409.
Irmler, U. 1981. Survival strategies of animals in the
seasonally flooded Amazonian inundation forest. Zool.
Anz.~ Jena 206(l/2):26-38.(In German).
Iversen, T.M., P. Wiberg-Larsen, S.B. Hansen & F.S. Hansen.
1978. The effect of part i a 1 and total drought on the
macroinvertebrate communities of three small Danish
streams. Hydrobiologia 60:235-242.
Jackson, LJ. 1977. Climate ~ Water and Agriculture in the
Tropics. Longman, London, 248p.
Janzen, D.H. 1973. Sweep samples of tropical foliage insects:
effects of seasons, vegetation types, elevation, time
of day, and insularity. Ecology 54:687-708.
John, K.R. 1964. Survival of fish in intermittent streams of
the Chiricachua Mountains, Arizona. Ecology 45:112-119.
Johnson, R.P. 1963. Studies on the life history and ecology
of the bigmouth buffalo Ictiobus cupx-inel.l.ue (Valenciennes).
J. Fish. Res. Ed Can. 20:1397-1429.
Junk, W.J. 1983. Aquatic habitats in Amazonia. The Environ-
mentalist :3 (euppl.. 5) :24-34.
Kapetsky, J.M. et al., 1977. Some ecological aspects of the
shallow lakes of the Magdalena floodplains, Colombia.
Paper presented to the Intertropical Symposium on Stability
and Diversity in Tropical Communities, Panama, March 1977.
(Quoted in Welcomme 1979).
Karr, J.R. 1976. Seasonality, resource availability, and
community diversity in tropical bird communities. Am.
Nat. 110:973-994.
380
Kenny, J.S. 1977. The Amphibia of Trinidad: an addendum.
Stud. Fauna Curacao 51(169) :92-95.
Kenny, J.S. & P.R. Bacon. 1981. Aquatic resources. Pages
112-144 in St G.C. Cooper & P.R. Bacon (eds). The
Natural Resources of Trinidad and Tobago. Edward Arnol d,
London.
Khanna, D.V. 1958. Observations on the spawning of the major
carps at a fi sh farm in the Punjab. Indian J. Fish. 5:
282-290.
Kirschbaum, F. 1975. Environmental factors control the
periodical reproduction of tropical electric fish.
Experientia 31:1159-1160.
Kirschbaum, F. 1979. Reproduction of the weakly electric fish
Eigenmannia ui.reecene (Rhamphichtyidae, Teleostei) in
captivity. Behav. Ecol., Sociobiol. 4: 331-355. (Quoted
in Lam 1983).
Kleerekoper, H. 1955. Limnological observations in north-
eastern Rio Grande do Sul, Brazil. Arch. Hydrobiol. 50:
553-567.
Klinge, H. & W. Ohle. 1964. Chemical properties of rivers in
the Amazonian area in relation to soil conditions. Verh.
into Verein. Irimnol., 15: 1067-1076.
Kodric-Brown, A. 1985. Female preference and sexual selection
for male coloration in the guppy (Poecilia reticulata).
Behav. Ecol. Sociobiol., 17(3): 199-206.
Kramer, D.L. 1978 a. Reproductive seasonality in the fishes
of a tropical stream. Ecology 59(5):976-985.
Kramer, D.L. 1978 b. Ventilation of the respiratory gas
bladder in Hoplerythrinus unitaeniatus (Pisces, Characoidei,
Erythrinidae). Can. J. Zoo l-, 56(4):931-938.
Kramer, D.L. 1983. Aquatic surface respiration in the fishes
of Panama: distribution in relation to risk of hypoxia.
Environ. Riol. Fishes 8(1):49-54.
Kramer, D.L. 1984. The evolutionary ecology of respiratory
mode in fishes: an analysis based on the costs of breathing.
Pages 67-80 in LM. Zaret (ed). Evolutionary ecology of
neotropical freshwater fishes. W. Junk, The Hague.
381
Kramer, D.L., C.C. Lindsey, G.E.E. Moodie & E.D. Stevens. 1978.
The fishes and the aquatic environment of the central
Amazon basin, with particular reference to respiratory
patterns. Can J. Zool. 56(4):717-729.
Kramer, D.L. & M. McClure. 1980. Aerial respiration in the
catfish, Oamjdorae aeneus (Call ichthyidae). Can. J. Zool.,
58(11) :1984-1991.
Kramer, D.L. & M. McClure. 1981. The transit cost of aerial
respiration in the catfish Comjdoeae aeneus (Callichthyidae).
Physiol. Zoo L: 54(2):189-194.
Kramer, D.L. & M. McClure. 1982. Aquatic surface respiration,
a widespread adaptation to hypoxia in tropical freshwater
fishes. Environ. Biol., Fishes 7(1):47-55.
Kramer, D.L. & J.P. Mehegan. 1981. Aquatic surface respiration,
an adaptive response to hypoxia in the guppy, Poecilia
retricul.atia (Pisces, Poeciliidae). Environ. Bi.o l., Fishes
6:299-313.
Krebs, C.J. 1985. Ecology. 3rd ed, Harper & Row, New York,
800p.
Krumholz, L.A. 1948. Reproduction in the western mosquitofish
Gambusia affinis and its use in mosquito control. Ecol;
Monogr. 18:1-43. (Quoted in Milton & Atherington 1983).
Lagler, K.F. 1978. Capture, sampling and examination of
fishes. Pages 7-47 in T. Bagenal (ed). Methods for
Assessment of Fish Production in Fresh Waters. 3rd ed,
Blackwell Scientific Publications, Oxford.
Lam, T.J. 1983. Environmental influences on gonadal activity
in fish. Pages 65-116 in W.S. Hoar, D.J. Randall & E.M.
Donaldson (eds). Fish Physiology vol. 9: Reproduction,
part B: Behaviour and Fertility Control. Academic Press,
N.Y.
Larimore, R.W., W.F. Childers & C. Heckrotte. 1959. Destruction
and re-estab 1 i shment of stream fi sh and invertebrates
affected by drought. Trans. Am. Fish. Soc. 88: 261-285.
Le Cren, LD. 1951. The length-weight relationship and seasonal
cycle in gonad weight and condition in the perch. J. Anim.
Ecol., 20:201-219.
382
Le Cren, E. D. 1962. The effi ciency of reproduction and
recruitment in freshwater fish. Pages 283-296 in E.D. Le
Cren & M.W. Holdgate (eds). The Exploitation of Natural
Animal Populations. Blackwell Scientific Publications,
OXford.
Le Cren, E.D. 1965. Some factors regulating the size of
populations of freshwater fish. Mitt. into Vaein. theor.
angew. Limnol. 13:88-105.
Leentvaar, P. 1975. Hydrobiological observations in Surinam
with special reference to the man-made Brokopondo Lake.
Stud. Fauna Suriname 15 (56) : 1-173 .
Levgque, C., C. Dejoux & A. Iltis. 1983. Limnologie du
fl euve Bandama, Cote d I Ivoire. Hydrobio Loqia 100:
113-141. (In French).
Levins, R. 1968. Evolution in changing environments. Prince-
ton University Press, Princeton, N.J., 120p.
Lewis, W.M. Jr. 1970. Morphological adaptations of cyprinodontoids
for i nhabiti ng oxygen defi cient waters. Copeia 1970: 319-326.
Lewis, W.M. Jr & F.H. Weibezahn. 1976. Chemistry, energy flow
and community structure in some Venezuel an fresh waters.
Arch. Hudrobio'l , Supp I , 50(2/3): 145-207.
Lev-lis, W.M. Jr & F. Weibezahn. 1981. The chemistry and phyto-
plankton of the Ori noco and Caroni Rivers, Venezuela. Arch.
Hydrobiol. 91 (4) :521-528.
Liley, N.R. 1966. Ethological isolating mechanisms in four
sympatri c species of poeci 1i id fi shes. Behaviour (Buppl , )
13:1-197.
Liley, N.R. & B.H. Seghers. 1975. Factors affecting the
morphology and behaviour of guppies in Trinidad. Pages
92-118 in G. Baerends, C. Beer & A. Manning (eds).
Function and Evolution in Behaviour. Cl arendon Press,
Oxford.
Liss, P.S. 1976. Conservative and non-conservative behaviour
of dissolved constituents during estuarine mixing. Pages
93-130 in J.D. Burton & P.S. Liss (eds). Estuarine
Chemistry. Academic Press, London.
383
Livingstone,D.A. 1963. Chemical composition of rivers and
lakes. Pl'of. Pap. U.S. Geol. Sum», 440G~ 64p.
Loiselle, P.V. & R.L. Welcomme. 1971. Two new species of
Barbue (Pisces: Cyprinidae) from southeastern Dahomey.
Revue Zoo l , Bot. afro 83:1-15.
Lowe-McConnell, R.H. 1964. The fishes of the Rupununi
Savannah district of British Guiana, South America. Part I.
Ecological groupings of fish species and effects of the
seasonal cycle on the fish. J. Linn. Soc. Lond. LZool., )
45:103-144.
Lowe-Ncconne l l , R.H. 1969. Speciation in tropical freshwater
fishes. Biol. J. Linn. Soc. Lond. 1:51-75.
Lowe-McConne 11, R. H. 1975. Fish Communities in Tropical
Px-eehioat.ere , Longman, London, 337p.
Lowe-McConnell, R.H. 1979. Ecological aspects of seasonality
in fishes of tropical waters. Syrrrp.zool , Soc. Lond. 44:
219..:241.
Lowe-McConnell, R.H. & G.J. Howes. 1981. Pisces. Pages 218-
229 in S.H. Hurlbert, G. Rodriguez & N.D. Dos Santos (eds).
Aquatic Biota of Tropical South America, Part 2-, San Diego
State University, San Diego, California.
Lum Kong, A. 1986. Aspects of the biology of the river conch,
Pomacea ux-ceue (Muller 1774). MPhil thesis, University
of the West Indies, Trinidad.
Macer, C.T. 1974. The reproductive biology of the horse
mackerel Tr-achurue tvachurue (L.) in the North Sea and
English Channel. J. Fish Biol. 6:415-438.
Maitland, P.S. 1978. Biology of Preehuatere , Blackie,
Glasgow, 240p.
McKillop, W.B. & A.D. Harrison. 1980. Hydrobiological studies
of Eastern Lesser Antillean islands. V. St Lucia:
Freshwater habitats, water chemistry and distribution of
freshwater molluscs. Arch. Hydrobiol. Supp l , 57(3):
251-290.
McKillop, W.B. & A.D. Harrison. 1982. Hydrobiological studies
of Eastern Lesser Antillean islands. VII. St Lucia:
Behavioural drift and other movements of freshwater marsh
molluscs. Arch. Hydrobiol. 94 (l): 53-69.
384
McKillop, W.B., A.D. Harrison & J.J. Rankin. 1981. Hydrobiological
studies of Eastern Lesser Antillean islands. VI. St Lucia:
Freshwater molluscs and the marsh environment. Arch.
Hydrobiol. Suppl., 58(4) :357-419.
Merritt, R.W. & K.W. Cummins (eds). 1984. An Introduction to
the Aquatic Insects of North America. 2nd ed, Kendall/Hunt,
Dubuque, Iowa, 722p.
Miller, P.J. 1979 a. A concept of fish phenology. Syrrp. zool.,
Soc. Lond. 44: 1-28.
Miller, P.J. 1979 b. Adaptiveness and implications of small size
in teleosts. Syrrp. zoo'l., Soc. Lond. 44:263-306.
Miller, P.J. 1984. The tokology of gobioid fishes. Pages
119-153 in G.W. Potts & R.J. Wootton (eds ) . Fish
Reproduction. Academic Press, London.
Miller, R.R. 1982. Pisces. Pages 486-501 in S.H. Hurlbert &
A. Villalobos-Figueroa (eds). Aquatic Biota of Mexico~
Central America and the West Indies. San Diego State
University, San Diego, California.
Milton, D.A. & A.H. Arthington. 1983. Reproductive biology of
Gambusia affinis halbvooki. Ba i rd and Girard, Xi.phophorue
hel.ler-i. (Gunther) and X. macul.atue (Heckel) (Pisces;
Poeciliidae) in Queensland, Australia. J. Fish Biol: 23:
23-41.
Minshall, G.W., R.C. Petersen & C.F. Nimz. 1985. Species
richness in streams of different size from the same
drainage basin. Am. Nat. 125(1):16-38.
Moghraby, A.I. el. 1977. A study on diapause of zooplankton
in a tropical river - the Blue Nile. Preehia, Bi.ol .
7(3) :207-212.
Moon, H.P. 1956. Observations on a small portion of a drying
cha 1k stream. Proc; zoo l , Soc. Lond. 126:327-333.
Moyle, P.B. & J.J. Cech Jr. 1982. Fishes: an introduction to
ichthyology. Prentice-Hall, N.J., 593p.
Myers, G.S. 1938. Freshwater fishes and West Indian zoogeo-
graphy. Rep. Smithson. Instn 1937:339-364.
385
Myers, G.S. 1947. The Amazon and its fishes. Part 2. The
fishes. Aquar . J. 18(4):13-20.
Nakatsuru, K. & D.L. Kramer. 1982. Is sperm cheap? Limited
male fertility and female choice in the lemon tetra (Pisces,
Characidae). Science 216:753-755.
Nelson, J.S. 1984. Fishes of the World. 2nd ed, John Wiley,
New York, 523p.
Nelson, K. 1964. Behaviour and morphology in the glandulocaudine
fishes (Ostariophysi, Characidae). Univ. Calif. Puble Zool.,
75(2):59-152. .
Nelson, K. 1965. The temporal patterning of courtship behaviour
in the glandulocaudine fishes (Ostariophysi, Characidae).
Behaviour 24: 90-146.
Nikolskii, G.V. 1961. Special Ichthyology. 2nd ed, National
Science Foundation & Smithsonian Institution, Washington,
D.C., 538p.
Nikolsky, G.V. 1963. The Ecology of Fishes. Academic Press,
London, 352p.
Nomura, H. 1975 a. Compar-acao da idade e do crescimento de
tr@s e spec ies de peixes do genero Astyanax Baird & Girard,
1854 (Osteichthyes, Characidae) do Rio Mogi Guacu, SP.
Revta bras. Bi.ol.. 35(3):531-547. (In Portuguese).
Nomura, H. 1975 b. Alimenta~ao de tres especies de peixes do
genera Astyanax (Osteichthyes, Characidae) do Rio Mogi
Gua<;;u, SP. Revta bras. Biol. 35:595-614. (In Portuguese).
Nomura, H. 1975 c. Fecundidade, maturacfo sexual e fndice
gonadosomatico de lambaris do genero Astyanax (Osteichthyes,
Characidae), relacionados com fatores ambientais. Revta
bras. Biol. 35:775-798. (In Portuguese).
Odum, LP. 1971. Fundamentals of Ecology. 3rd ed, W.B. Saunders,
Philadelphia, 574p.
Orr, T.M. & N.E. Milward. 1984. Reproduction and development of
Neoei.l.urue at.er (Perugi a) and Neoei/lurue hyrtlii Stei ndachner
(Teleostei: Plotosidae) in a tropical Queensland stream.
Aust. J. mar. Presho. Res. 35:187-195.
386
Padqet t , D.E. 1976. Leaf decomposition by fungi in a tropical
ra in-f'or-e s t stream. Biotropica 8(3) :166-178.
Pain, T. 1950. Pomacea (Ampullariidae) of British Guiana.
Proc; mal.ao , Soc. Lond. 28 :63-74.
Pain, T. 1960. Pomacea (Ampullariidae) of the Amazon River
system. J. Conch. Lond, 24 (12) :421-432.
Paloumpis, A.A. 1958. Responses of someminnows to flood and
drought conditions in an intermittent stream. Iowa St. coit:
J. Sci. 32: 547-561.
Parker, R.E. 1979. Introductory statistics for Biology. 2nd
ed, Edward Arnold, London, 122p.
Parker, R.R. 1963. Effects of formalin on length and weight
of fishes. J. Fish. Res. Bd Can. 20:1441-1445.
Payne, A.!. 1975. The reproductive cycle, condition and
feeding in Barbue I-iber-i.ene-i e , a tropical stream-dwelling
cyprinid. J. Zool., Lond, 176:247-269.
Payne, A. I. 1986. The Ecology of Tropical Lakes and Rivers.
John Wiley & Sons, Chichester, 301p.
Paysan, K. 1975. The Hamlyn Guide to Aquarium Fishes. Hamlyn,
London, 239p.
Pelizaro, M.G., H.S. Lieme dos Santos, R.A. Lopes & N. Castagnoli.
1981. Rhythm of development in the oocyte of Tambiu
Astyanax b-imaculatue (Reinhardt, 1874) (Pisces:Characidae),
a morphologic, morphometric and histochemical study. Arch.
Biol . (BruxeUes) 92(4) :415-431. .
Pennak, R.W. 1978. Freshwater Invertebrates of the United
States. 2nd ed, John Wiley, N.Y., 803p.
Perkins, E.J. 1974. The Biology of Estuaries and Coastal Waters.
Academic Press, London, 678p.
Perry, M.C. 1973. Ecological studies of the apple snail at
. Woodruff National Wildlife Refuge. Florida Sci. 36(1):
22-30.
Petr, T. 1970. The bottom fauna of the rapi ds of the Black
Volta River in Ghana. Hydrobiologia 35:399-418.
Pianka, E.R. 1970. On r- and K-selection. Am. Nat. 104:
592-597.
387
Price, J.L. 1955. A survey on the freshwater fishes of the
island of Trinidad. J. aqr-io • Soc. Trin. (Society Paper
863): 28p.
Prowse, G.A. & J.F. Talling. 1958. The seasonal growth and
succession of plankton algae in the White Nile. Irimnol ,
Oceanogr. 3:222-238.
Rankin,J.J. & A.D. Harrison. 1979. Hydrobiological studies
of Eastern Lesser Antillean islands. IV. St Vincent:
Compari son of fiel d and 1aboratory popul ations of Physa
maxmopaba Guilding (Gastropoda: Pulmonata). Arch.
Hydrobiol. Suppl., 57(l) :89-116.
Regan, C.T. 1906. On the freshwater fishes of the island of
Trini dad based on coll ection, notes, and sketches made
by Mr Lechmere Guppy, Junr. Proe, zool., Soc. Land. 1906:
378-393. .
Reyno1ds , C. S . 1984. The Eco logy of Freshwater Phytop Lankion
Cambridge University Press, Cambridge, 384p.
Reznick, D.N. 1980. Life history evolution in the guppy
(Poecilia »et-iculata) , PhD thesis, University of.
Pennsylvania, 224p.
Reznick, D.N. & J.A. Endler. 1982. The impact of predation
on 1ife history evolution in Trinidadian guppies (Poeeil-ia
»etricul-ata ) , Evolution 36(1):160-177.
Ricklefs, R.E. 1966. The temporal component of diversity
among species of birds. Evolution 20:235-242.
Ridewood, W.G. 1913. Notes on South American freshwater
flying-fish, Gaeteropel eeue and the commonflying-fish,
Exoooetiue , A. Mag. nat. Hist. 12(72):544-548.
Robinson, M.H. 1978. Is tropical biology real? Trop. EcoZ. 19:
30~50.
Rodriguez, G. 1974. Some aspects of the ecology of tropical
estuaries. Pages 313-333 in F.B. Golley & E. Medina (eds).
Tropical Ecological Systems. Springer-Verlag, NewYork.
Rose, S.M. 1959. Population control in guppies. Am. Midl. Nat.
. 62:474-481. (Quoted in Dahl gren 1979).
Rose, S.M. & F .C. Rose. 1965. The control of growth and
reproducti on in Freshwater organisms by spec ific products.
Mitt. into Verein. theon, anqeia, Irimnol., 13:21-35.
(Quoted in Dahlgren 1979).
388
Rosenthal, H.L. 1952. Observations on reproduction of the
poecil i id Lebistes ret icul.at.ue (Peters). Biol. Bull. 102:
30-38. (Quoted in Hester 1964).
Rzoska , J. 1961. Observations on tropical rain pools and
general remarks on temporary waters. Hydrobiologia 17:
265-286.
Rzcska, J. 1978. On the Nature of Rivers. W. Junk. The Hague,
67p.
Sanchez,L.,E. Vasquez & L. Blanco. 1985. Limnological
investigation of the rivers Uracoa, Yabo, Morichal Largo
and Claro in the eastern plains of Venezuela. Verh. into
Yerei.n, Limnol. 22: 2153-2160.
Schlosser, I.J. 1982. Fish community structure and function
along two habitat gradi ents ina headwater stream. Ecol.
Monoqr • 52 ( 4) :395-414.
Schlosser, I.J. 1985. Flow regime, juvenile abundance, and
the assemblage structure of stream fishes. Ecology 66(5):
1484-1490.
Schmidt, G.W. 1972. Chemical properties of some waters in the
tropical rain-forest region of the Central-Amazonia along
the new road Manaus-Caracarai. Amazoni.ana .3(2):199-207.
Schneller, M.V. 1955. Oxygen depletion in Salt Creek, Indiana.
Invest. Indiana Lakes Streams 4: 163-175.
Schofield, D. 1957. The cooperative cat Comrdoxrae aeneue .
Aquar-ium, Philad. 26(3) :84-86.
Schwassmann, H.O. 1971. Biological rhythms. Pages 371-428
in W.S. Hoar & D.J. Randall (eds). Fish Physiology, Vol.
6. Academi c Press, New York.
Schwassmann, H.O. 1978. Times of annual spawning and reproductive
strategies in Amazonian fishes. Pages 187-200 in J.E. Thorpe
(ed). Rhythmic Activity in Fishes. Academic Press, N.Y.
Schwassmann, H.O. 1980. Biological rhythms: their adaptive
significance. Pages 613-630 in M.A. Ali (ed). Environ-
mental Physiology of Fishes. Plenum Press, N.Y.
389
Scott,D.B.C. 1979. Environmental timing and the control of
reproduction in teleost fish. Syrrrp. zool.. Soc. Lond. 44:
105-132.
Scrimshaw, N.S. 1944. Superfoetation in poeciliid fishes.
Copeia 1944 (3) : 180-183.
Seegrist, D.W. & R. Gard. 1972. Effects of floods on trout
in Sagehen Creek, California. Trans. Am. Fish. Soc. 101:
478-482.
Seghers, B.H. 1973. An analysis of geographic variation in
the antipredator adaptations of the guppy, Poecilia
reticulata. PhD thesis, University of British Colombia,
272p.
Shirvell, C.S. 1981. Computer analysis of fish populations
in fluctuating streams. Fish. Mgmt 12(1):33-36.
Siegfried, C.A. & A.W. Knight. 1977. The effects of washout
in a Sierra foothill stream. Am. Midl. Nat. 98:200-207.
Silliman, R.P. 1948. Factors affecting population levels in
Lebistes reticuZatus. Copeia 1948 (l) :40-47 . (Quoted in
Dahl gren 1979).
Singh, 1.B. 197i£3. The biology of the cascadura Hop Loet.ernum
littorale Hancock 1834 with reference to its reproductive
biology and population dynamics. PhD thesis, University
of the West Indies, Trinidad, 298p.
Sioli, H. ·1964. General features of the limnology of Amazonia.
XI. Amazoni ana. Yerh, into Verein. Irimnol., 15: 1053-1058.
Sioli, H. 1975 a. Tropical rivers as expressions of their
terrestrial environments. Pages 275-288 in F.B. Golley &
E. Medina (eds). Tropical EcoZogical Systems. Springer-
Verlag, N. Y.
Sioli, H. 1975 b. Tropical River: the Amazon. Pages 461-488
in B.A. Whitton (ed). River EcoZogy. University of
California Press, Berkeley.
Sial i, H. 1984. The Amazon: LimnoZogy and Landscape. W. Junk,
Dordrecht, 763p.
390
Smalley, A.E. & G. Rodriguez. 1972. Trichodactylidae from
Venezuela, Colombia, and Ecuador (Crustacea: Brachyura).
Tulane Stud. Zool., Bot. 17(3/4):41-55.
Sakal, R.R. & F .J. Rohlf. 1981. Biometry. 2nd ed, W.H. Freeman,
San Francisco, 859p.
Stanley, A.E. 1956. Breeding the silver hatchet fish
(Gaetieropel.ecue loeoie ) , Aquarist Pondkp» 21 (6): 129.
Starrett, W.C. 1951. Some factors affecting the abundance of
minnows in the Des Moines River, Iowa. Ecology 32:13-27.
Statzner, B. 1982. Population dynamics of Hydropsychidae
(Insecta;Trichoptera) in the WZi River (Ivory Coast),
a temporary stream partly treated with the insecticide
Chlorphoxim. Rev. Hydrobio"l. trap. 15(2): 157-176.
Stearns, S.C. 1976. Life history tactics: a review of the
ideas. Quart. Rev. etot, 51(1):3-47.
Stearns, S.C. 1983. The influence of size and phylogeny on
patterns of covariation among life-history traits in
mamm1as. Oikos 41: 173-187 •
Stehr, W.C. & J.W. Branson. 1938. An ecological study of an
intermittent stream. Ecology 19:294-310.
Sterba, G. 1962. Freshwater Fishes of the WOT'ld. Studio
Vista Books, London, 877p.
Stolzenhain, F. 1927. Hemigrammus unilineatus Gill. Aquat ,
Life 11(6) : 105-107, 118.
Stonley, J.M. 1971. A monograph of the crabs of Trinidad.
MSc thesis,University of the West Indies, Trinidad, 402p.
Stout, R.J. 1981. Howabiotic factors affect the distribution
of two species of tropical predaceous aquatic bugs (Family:
Naucoridae). Ecology 62:1170-1178.
Stout, R.J. 1982. Effects of harsh environment on the 1ife
history patterns of two species of tropical aquatic.
Hemiptera (Family: Naucoridae). Ecology 63: (1) :75-83.
Stoye, F.H. 1935. CamegieUa etii-iqata (Guenther). Aquarium,
Philad. 4 (1):6-8 .
391
Sturm, M.G. de L. & G. De Souza. 1984. Triportheus e lonqatue :
a new record of a characid fish from Trinidad. Cope ia
1984(1) :262-263.
Talling, J.F. & J. Rz6ska, 1967. The development of plankton
in relation to hydrological regime in the Blue Nile. J.
Ecol. 55: 637 -662 .
. Tan, E.S.P. 1980. Some aspects of the biology of Malaysian
riverine cyprinids. Aquaculture 20(3):281-289.
Tay, L.L. 1983. PhD thesis ( in preparation). (Quoted in
Lam 1983).
Templeton, W.L., J.M. Dean, D.G. Watson & L.A. Rancitelli. 1969.
Freshwater ecological studies in Panama and Colombia.
BioScience 19:804-808.
Thibault, R. E. & R.J. Schultz. 1978. Reproductive adaptations
among viviparous fishes (Cyprinodontiformes: Poeciliidae).
EvaZution 32: 320-333.
Thornhill, E., L. Ramcharitar, D. Herrera, R. Aaron, R. Rudder,
C. Thomas, K. James & R. Moore. 1967. Ecological study
of a part of the Maracas River. BioZ. J. U.W.I.~ Trinidad
2(2) :29-54.
Thornton, J.A. 1980. Preliminary results from the ~1uchakata
Dam preimpoundment study. Trans. Zimbabwe scient. Ass.
60(5) :24-31.
Trinidad & Tobago, Government. 1979. Rainfall Report, 1953-1977.
Ministry of Works, Transport and Communications, Trinidad.
Trinidad & Tobago, Water & Sewerage Authority. 1976 a. Caroni
River Basin Study. Preliminary Report, Inventory and
Ana1ys is: Hydrology Report, 65p.
Trinidad & Tobago, Water & Sewerage Authority. 1976 b. Caroni
River Basin Study. Preliminary Report, Inventory and
Analysis: Report by Public Health Engineering, 82p.
Turner, C.L. 1937. Reproductive cycles and superfoetation in
poeciliid fishes. Biol. BuZZ. 72: 145-164. (Quoted in
Dahlgren 1979).
392
Turner, C.L. 1938. The reproductive cycle of Brachyraphis
episcopi .. an ovoviviparous poeciliid fish, in the natural
tropi ca 1 habi tat. Biol. eau: 75: 56-65.
Uh1mann, D. 1979 . Hydrobiology: a text for engineers and
scientists. John Wi1ey, London, 313p.
UNESCOjUNEPjFAO. 1978. Tropical Forest Ecosystems: a state
of knowledge report. UNESCO.. Pari s , 683p.
van der Heide, J. 1976. Brokopondo Research Report, Suriname.
Part II: Hydrology of the man-made Brokopondo Lake. Uitg.
natuuma, Stud-Kring Suriname 90:1-95.
Vernberg, W.B. & F.J. Vernberg. 1983. Freshwater adaptations.
Pages 335-363 in F.J. Vernberg & W.B. Vernberg (eds).
The Biology of Crustacea.. Vol. 8: Environmental Adaptations.
Academic Press, N.Y.
Ward, R.C. 1975. Principles of Hydrology. 2nd ed, McGraw-Hill,
London, 367p.
Warren, E.W. 1973 a. The establ ishment of a I normal I population
and its behavioural maintenance in the guppy - Poeci.l ia
ret.iculat-a (Peters). J. Fish Biol.. 5:285-304. (Quoted in
Dahl gren 1979).
Warren, E.W. 1973 b. Modifications of the response to high
density conditions in the guppy Poecilia reticulata
(Peters). J. Fish Biol. 5:737-752.
Warren, E.W. 1973 c. The effect of relative density upon some
aspects of the behavi our of the guppy - Poeci lia ret.iculata
(Peters). J. Fish Biol. 5:753-765.
Waters, T. F. 1983. Repl acement of brook trout by brown trout
over 15 years in a Minnesota stream: production and
abundance. Trans. Am. Fish. Soc. 112:137-146.
Way, C.M., D.J. Hornbach & A.J. Burky. 1980. Comparative
1ife hi story tacti cs of the sphaeri id cl am, Musculium
partumeium (Say), from a permanent and a temporary pond.
Aril. Midl. Nat. 104(2) :319-327.
Wehekind, L. 1955. Trinidad Rainfall, 1933-1952. Government
Printing Office, Trinidad, 36p.
393
Weitzman, S.H. 1954. The osteology and the relationships of
the South Ameri can characi d fi shes of the subfami ly
Gasteropelecinae. Stanford ichthyoZ. Bull. 4(4):213-263.
Weitzman, S.H. 1958. Oie'Fliegenden salmler'. Aquar . Terrar .
z. 11(2):42-45. (In German).
Welcornme, R.L. 1969. The biology and ecology of the fishes of
a small tropical stream. J. Zool. Lond. 158:485-529.
We1comme, R. L . 1979. Fisheries Eco logy of Pl.oodp lain Rivers.
Longman, London, 317p.
Welcomme, R.L. 1985. River Fisheries. FAGFish. Tech. Pap.
(262), 330p.
Wetze 1, R. G. & G. E. Likens. 1979. Lirrmological Analyses.
W.B. Saunders, Philadelphia, 357p.
Williams, D.O. 1977. Movements of benthos during the recoloni-
sati on of temporary streams. Oi.koe 29: 306-312.
Williams, D.O. & B.W. Coad. 1979. The ecology of temporary
streams. I I I: Temporary stream fishes in Southern Ontario,
Canada. Int. Revue ges. HydrobioZ. 64(4):501-516.
Wil,liams, D.O. & H.B.N. Hynes. 1976. The ecology of temporary
streams. I. The faunas of two Canadian streams. Int.
Revue ges. Hydrobiol. 61:761-787.
Williams, D.O. & H.B.N. Hynes. 1977. The ecology of temporary
streams. II. General remarks on temporary streams. Int.
Revue ges. Hydrobiol. 62(1):53-61.
Williams, G.C. 1966. Adaptation and Natural Selection.
Princeton University Press, New Jersey, 307p.
Williams, W.O. 1981. Running water ecology in Australia.
Pages 367-392 in M.A. Locke & D.O. Williams (eds).
Perspectives in Running Water Ecology. Plenum Press, N.Y.
Wi nner, J .M. 1975. Zoopl ankton. Pages 155-169 in B.A. Whitton
(ed) . River Ecology. University of California Press,
Berkel ey.
Walda, H. 1978. Seasonal fluctuations in rainfall, food and
abundance of tropical insects. J. Anim. EcoZ. 47:369-381.
394
Wolda, H. & E. Broadhead. 1985. Seasonality of Psocoptera in
two tropical forests in Panama. J. Anim. Ecol. 54(2):
519-530.
Wolda, H. & R.W. Flowers. 1985. Seasonality and diversity of
mayfly adults (Ephemeroptera) in a 'nonseasonal' tropical
environment. Biotropica 17(4) :330-335.
Wootton, R.J. 1979. Energy costs of egg production and
environmental determinants of fecundity in teleost fishes.
Symp. zool., Soc. Lond, 44: 133-159.
Wootton, R.J. 1984 a. A Functional Biology of Sticklebacks.
Croom Helm, London, 265p.
Wootton, R.J. 1984 b. Introduction: strategies and tactics
in fish reproduction. Pages 1-12 in G.W. Potts & R.J.
Wootton (eds). Fish Reproduction. Academi c Press, London.
Wourms, J.P. 1981. Viviparity: the maternal-fetal relation-
ship in fi shes. Am. Zool , 21: 473-515.
Yamagishi, H. 1976. Experimental study on population dynamics
in the guppy, Poeci.l ia reticulata (Peters). Effect of
shelters on the increase of population density. J. Fish
Bio Z. 9: 51- 6 5 .
Youngs, W.O. & D.S. Robson. 1978. Estimation of population
number and mortality rates. Pages 137-164 in T. Bagenal
(ed). Methods for Assessment of Fish Production in Fresh
Waters. 3rd ed, Blackwell Scientific Publications, Oxford.
Zaret, T.M. (ed) 1984. Evolutionary ecology of neotropical
freshwater fishes. W. Junk, The Hague, 173p.
Ziesler, R. & G.D. Ardizzone. 1979. The inland waters of
Latin America. FAa, Rome, 171p.
Zu ka 1, R. 1982. Breed i ng Corydoras aeneus. Trop. Fish Hobby.
May 1982: 9-11 .
395
APPENDIX 1
Procedure for pentade analysis of rainfall data.
The use of pentades (5-day periods) in analysing rainfall
patterns has been preferred by some workers because of the
inadequacy of data based on longer periods (months or weeks)
in reflecting short term rainfall variability (Jackson 1977).
The onset or end of various seasons, either on average or for
individual years, does not coincide with the calendar months
which are often used to make general isations about rainfall
regimes. Jackson (1977) reviews the use of pentade analysis
in the definition of rainy seasons, analysis of length of dry
seasons and in determining the occurrence of wet and dry spells
during the rainy season in various tropical regions.
The method used by Griffiths (1959, quoted in Jackson
1977) involves initially the determination of total rainfall
for consecutive 5-day periods from daily rainfall data.
When the totals for three consecutive pentades are
compared, the middle pentade is considered 'rainy' if the
following conditions are met:
396
APPENDIX1 (continued).
( t) The total ra i nfall for the three pentades together
amount to at least 76 mm,
(f t ) The rainfall amounts to 7.6 mm or more in each of at
1east two of the three pentades.
Thi s procedure is repeated for success ive pentades of
the period under analysis.
397
APPENDIX2
Tests of significance between means of two samples.
The Student's t-test was used to test for significance of
differences between the means of two small samples (t .e. N
1
and N2 < 30) under the condition that population variances
were not assumed to be equal (Parker 1979).
The fo 11owi ng formul ae were used:
= t wi th f degrees of freedom
f is given by:
Iff = u2/(N1 - 1) + (l - u)2/(N2 - 1)
where u =
Xl X means of samples 1 and 2
2
s s standard deviations of samples 1 and 2
1 2
N N number of measurements or replicates in samples 1 and 2.
1 2
398
APPENDIX2 (continued).
For comparison of means of two large samples (N > 30), a
d-t.est was used (Parker 1979) where:
= d
399
APPENDIX 3
Taxonomic 1 ist of aquatic macrofauna collected at Stations 1
to 4 during the study period.
Stations
Taxon 1 2 3 4
Nematoda x
Nemertea x
Anne1 ida
01 i gochaeta
Tubifi ci dae x x x
Naididae x x x
Enchytrae i dae x x
Other x x x
Hirudinea
Gloss i phon i dae
Glossiphonia x x x
Pl.acobde Ll.a x
Polychaeta
Nereidae x
Capitell idae x
Crustacea
C1adocera x x
Ostracoda x
Copepoda x x
Isopoda x
Amphipoda x
Decapoda
Penaei dae
Penaeue notialis Perez-Farfante x
Pa1aemon i dae
Palaemon pandalifoY'mis (Stimpson) x
Macrobrachium jelskii (Mi ers) x
M. he tex-och irue (Wiegmann) x
MacY'obrachium sp. x
Alphaeidae
Alphaeue sp. x
Portunidae
Callinectes sapidus Rathbun x
400
APPENDIX 3 (conti nued)
401
APPENDIX 3 (continued)
Stations
Taxon. 1 2 3 4
Gerridae
Brachymetra albinervis (Amyot & Serville) x x x
Limnogonus aduncue Ora ke & Harri s x x
TeZmatometra fusca Kenaga x x
Veliidae
Rhagovelia ? insularis Champion x
Coleoptera
Dyt isc idae "
Thermonectus sp.nov.? x
Laccophilus proximus Say x
Gyr inidae "
Gyretes ? distinguendus Reg. x
Hydroph i 1 i dae 4
Hel.ochai-ee sp.nov.? x
Carabidae
?Omophron x
Diptera
Chironomidae x x x
Heleidae x x x
Culicoides Sp.5 x
Moll usca
Gastropoda
Pl anorb i dae x x x
Ampul1ari i dae
Pomaceaglauca (Linnaeus) x x x
Hydrobiidae x
Ancyl i dae (2 s pp . ) x x
Bivalvia
Sphaeriidae (2 spp. ) x x x
Myte 11 i dae 6
My ti lops is dominigensis Recluz x
Vertebrata
Teleostei (see separate list)
Amphibia
Bufonidae
Bufa marinus (Linnaeus) x
B. granulosus beebei Gall ardo x
402
APPENDIX 3 (continued)
Stations
Taxon 1 2 3 4
Hylidae
HyZa qeoqnaphi.ca geographica Spix x
Leptodactyl i dae
Leptodacty Lue podicipinus pe terei.
(Steindachner) x
Reptilia
Chelidae
Phmmope (Heeocl.emmje ) gibbus
( Schwe i gger) x x
Emydidae
RhinocZemmys punctulca-ia punctulax-ia
(Daudin) x
Ki nosterni dae
Kinosternon scorpioides scorpioides
(Linnaeus) x x
Crocodyl i dae
Caiman erocodil.ue (L i nnaeus) x x x
Specimens determi ned by:
1 Dr. W.L. Peters, Florida A & M University, Tallahassee,
FL, U.S.A.
2 Prof. M.J. Westfall Jr., University of Florida, Gainesville,
FL, U.S.A.
3 Dr. N. Nieser, Tiel, The Netherlands.
4 Dr. P.J. Spangler, NMNH, Washington DC, U.S.A.
5 Mr. R. Martinez, CAREC, Trinidad.
6 Dr. P.R. Bacon, U.W.I., Mona, Jamaica.
403
APPENDIX 4
Taxonomic list of phyto- and zooplankton collected at Stations
1 to 4 between July 1980 and June 1981.
Stations
Taxon 1 2 3 4
Cyanophyta
Myxophyceae
Spirulina~ Osci l.latot-ia, other
fil amentous spp. x x x
Ch 1orophyta
Chlorophyceae
Desmidiaceae
Cloe t.er-i.um , Pleurotaeniwn x x x x
Zygnemataceae
Spirogyra x x
Oedogoni aceae
Oedogoniwn x' x
Ch rysophyta
Sacill ariophyceae
Cosci nodi scaceae
Coscinodiscus x
. Melosira X x x
Fragi 1ari aceae
Synedra x
Navi cul aceae
Navicula~ Pinnularia~ Gyrosigma X X X x
Protozoa
Phytomastigophorea
Euglenidae
Euglena~ Phacue , I'rache Lomonae x x x x
Zoomastigophorea x x x
Rhizopoda
Arce 11 i dae
Arcella x x x
Diffl ug i i dae
Difflugia x x x x
404
APPENDIX 4 (continued)
Stations
Taxon 1 2 3 4
Actinopoda
C1athru1 inidae
Clat/hrul.i.na x x
Cil iata x x x
Cae 1enterata
Hydrozoa
Petasi dae
Cnaepedacueta x
Ratifera
F1oscu1ariaceae
Conochi.Lue x x
Arthropoda
Crustacea
C1adocera x x
Ostracoda x x
Copepoda (Calanoida, Cyc1opaida) x x x x
Amphipada x x
Arachnoi dea (Hydracari na) x x x
Q l
: l >
: s :
- - l
t i l
; ; 0 ; ; 0


<
- f
\ J
C D
: : : s
: T
$ l I O
: : r
0
C D
C D
( ' " 1 -
Q l
\ J
C D
( ' " 1 -
C D ~ c
V l c . o
I T 1
- - ' : : r
. . . . . . ~
. . . . . .
Q l
- s
- S
: z
0
. . . . . .
$ l i n
C D - - '
0 -
C D
V l 0 - 0
( ' " 1 - C D
0
c V I
. . . . . .
1 - 1
~
. . . . . . V l
$ l I
V I
~
C X
V l
. . . . . .
. . . . . .
0
V l
0
M
: : : s 0
0
U 1
C D
0
X
: E
- t >
~
. . . . . .
0 -
- t >
' <
~
( ' " 1 -
c . o
<
Q l
0
I I I
- s
c - +
. . . . . .
( ' " 1 -
P I
M
Q l
C D
t : T
' <
~
V l
: : : s
M
N V l t i l
( ' " 1 -
x n
C C D
. 0
C D
M
- < M M - c 3
( / ) Q l
c - t
V l
X
: E
X
' <
I \ )
: : r
I \ )
P I
( / ) ~
' <
Q l
C D
V I
- C D 0
I \ )
-
V l - t >
- t >
( ' " 1 -
V l
V I
M -
M
. . . . . .
C D
X C D
x
I \ )
M
- s
c - +
' <
x
: : : s
. . . . . .
-N - 0
s : :
Q l
: z
" " C - t >
. . . . . .
- s
-M
o
7 '
: E
. . . . . .
Q l
- < C D
0
- t > C D c - + - s : : : s
z
z ~ t Q
: T
n
. . . . . . .
C D - s C D
. . . . . .
C D C D c - + 0 . 0
: T
O - C D - . . . J
0
. . . . . .
N
O V l
C D
0 . 0
- t >
3
( ' " 1 -
0
- s
C D : : r
- t >
~
C D
C D
- s
<
P I
C D
~
; ; 0
c - +
- s
; ; 0
C D
C D
( Q
C D
0 . . V l
V l
V l
- s
s : u
V l
: z
. . . . . .
C D - . 0
: s : c - +
0 -
V I
: S : C C D
P I
0
( / ) Q l
C Q l
. . . . . .
V I
: : : s
. . . . . .
$ l I
: : : s
- ~
. . . . . .
N
0
C D I
n
~ V l
0
t i l
C D
t i l
U l
- t >
U l
- t >
. . . . . .
n
. . . . . .
C D
: : : s
; ; 0
c - t
; ; 0
r o
( Q
C D
C
V l
- s
V l
. . . . . .
C D
0 - V I
: : : s
c
V I
$ l I
. . . . . .
0
~
"
: s :
. . . . . .
t i l
3 :
: : : s
U l
Q l
: : : s
S O P
) : : >
: E :
P >
\ J
" T 1
O J
1 . 0
: : : : : l " "
: : s
\ J
U l V l
r o
0 . .
: : J "
r r t
~
- s
- t ,
O J
: : z
r o
> <
V l : z
- <
0 C l
. . . . . .
. . . . .
- I
. . . . . .
I I
I I
. . . . .
z r
: 3
> <
. . . . .
r o
0
M
M
- I .
< ,
0 . .
: J
r - t
< . . n
: : E :
x
«
: : s
c r o
V l
V l ~
N
0 . .
- 0
: 3
< . . n
' "
r o
: z r o
0 -
I I
~
- 0
C D z s
r o
0 . .
n - s
0 -
n
I I I
I I I
r o z s
0
0
0 . .
N 0 : : s
: : s
1 + : : s
- t , r o
M -
- t >
M -
: : s
- ' .
r - t
M
0 - M
V l M -
<
0 . .
: : s
C D > <
- <
P >
O J
r o
: z :
C
L Q
,
N
<
z s
: 3
r o
' "
- s
P >
u
n
0 . .
,
C D
. . . . . .
P >
r o
C D
- I .
C D c r
N
V l
- - - '
P >
V l
. .
- ' . c r
r o
0
0
, . . . . . .
- - - '
- ; ,
3
M
M
- I .
r o
0
x
: z :
: : z
- <
- ; ,
M -
U l
. . . . . . ,
t n
- s
N
C D
' "
C D
0
0 -
- t >
; ; > : J 0
C D M -
: 3
U l : : : r
r o
M
x 0 -
,
C
r o
' " O J
. . . . . .
l . C l
,
r o
3 :
t n
t i l
t n
- ' .
0
: : : l
n
0
r o
- n
- t >
- ' .
n
- ' .
r o
: : : l
M -
: E :
r o
,
r o
0 . .
r o
M -
r o
,
3
: : : l
r o
0 . .
g o t ?
407
APPENDIX 6
Method used to compare two regression coefficients (after
Bailey 1981).
The res i dua 1 variance, S2, was calculated for each set of
data as follows:
1 (L:xy) 2
S2 = (z::y2_
N - 2 EX2
The d-statistic was then computed according to the
following formul a for sample sizes >30:
bl - b2
d JC:l: s 2+ 2
LX 2
1 2 J
where b = regression coefficient
other symbols are as in Appendix 5 with subscripts
denot i ng data sets 1 and 2.
To compare the observed regression coefficient with a predicted
value (b '}, the following procedure was used (after Sokal & Rohlf 1981):
standard error of b (sb) =jz::~~
= for n - 2 degrees of freedom.
- {
- f l : l : : >
; t o . V >
t ; J
: : J ; j
: : : : l
: z C J
V >
' c J C " : l
: : : r
- - 0
- 0
V l
f T l
*
*
I I I r o - - 0
r o
*
: : : r r r t
( ' )
C O
" $
C '
~
" S ~
*
r o : : z :
. . . . . .
( \ )
; : : s
~ . c t -
< ' : >
I I I
" . ( \ ) C l
- + >
; : : s r o
~ .
I i -
. . . . . . .
" .
( \ ) I I I
- + >
" $
C O
~ . N ~
r o
r o
; : : s
( \ ) > <
( ; ) Q
~
( ' )
( ' )
" .
; : : s ~ .
~ .
C O
~
: > : '
. . . . . .
. . . . . . . .
r l "
Q
( \ )
- 0
- 0 : z < : - - I N
: : z :
V >
I I I
r o
/ \
/ \ c :
0 R ~ ~
r l "
I I I
~
0
0 s : u ~ . . - t -
M -
3
" "
0
.
.
: : : : l
B " ~ ~
0 -
I I I
- + >
C O
0 V l C O
0 0 - r o
. . . . . r l "
0
0 1 s : u
- S
c :
. . . . . .
- + >
l O
- S
0
0 -
0
: : : : l 0 -
. . . . .
- S
- f l
r o
3
. . . . . . .
f T l
s : u
( J l 0 -
I I I U 1 : z
0 1 N
- S W W
- - '
. . . . . .
. . . . . .
s : u
n W
- S w O " l O " l
. . . . . .
O J
0
3
: : : : : l
: : : : l - 0
- S
. . . . . .
M -
r o
I I I - S
r o
" " 0
~ \ D \ D \ . 0
\ D \ D
I I I
V
0 1 - . . . . J
\ D . . . . . . . . C O
C O
r o
a
3 :
.
( J l
. . . . . . .
~ N m - s
w
c o
( »
a . . . . . . . .
W s : u
0 N
- l = = >
<
U ' 1
s : u
V > : : : : l
r -
r l "
1 +
1 + 1 + 1 + 1 + 1 +
. . . . . .
" " 0
0 0 0
0 0 0 0 - s
: : : : : l
~ m
. . . . . . . . . . . . .
N W
N N V l
' 0 1
. j : : o
\ D
. . . . . . . . N W 0
. . . . . . . m
: : : : : l
- s - s
m
<
V l
r o
: : : : s - 0 -
r o
: : : : : l
. . . . . .
1 + <
~
0 \ D
\ D \ D \ D
O J
r l "
( j )
U l
- I ' : >
- l = = > - . . . . J V > - - '
c o
: : : r
f T l c :
. . . . . . .
~ O " l
c o W N
m s : u
. . . . . .
N
U l N
- l = = > U 1
- l : : : : : l
: ; : :
s : u
0 -
1 +
1 +
1 +
1 + V l
1 + 1 +
: : E :
0
0
0
0
0 0
r o
. . . . . .
0 1 O " l
. . . . . . . .
O " l
O " l
- + > 0
( »
- l = = >
\ D . . . . . .
O " l
W
: : : r
r l "
3
r o
s : u
. . . . .
. . . . . . .
. . . . . .
. . . . . .
I I I
N
N . . . . . . .
. . . . . .
c o
W
- - 0
c :
s : u
- s
- I ' : > N
N
\ D
w
U 1
r i -
r o
. . . . . .
+ : >
. . . . . . . .
U ' 1
W
0 V l
- s
3
- I ' : > \ D
. . . . . .
c o ( j )
0
r -
m
r o
0 -
: : : : : l
*
*
*
*
* I
*
r l "
*
*
*
*
* V l
I I I
*
*
*
*
*
O J
: 3
0
" " 0
- + >
m
r l "
' J . . . . . . . .
O " l
0 w
\ D
: : : r
r i -
r o
- I ' : >
0
\ D ( J l
U l
' - J
r i -
I
0
W
W
. . . . . .
0
- {
0 0
M -
I I I
O l . . . . . .
- l = = >
\ D
U 1
U 1
: : E :
m
: : : : l
V l
X
*
*
V l
*
*
*
M -
*
*
*
*
*
*
*
*
*
*
8 0 P
) : : >
- i
U 1
Q
( J
. . . . :. . z
: : : J , : t > .
( J
0 - : : t : :
- 0
' t J
: : r
0
- 0
"
V l
0
- 0
r o
- s
r o
< . Q
r r t
C )
C J l
~
. . . . . . o -
$ : : l
~
~
: z
r o
c T
< : - , . 0 -
( \ l ; : s
( \ \ ~ .
a
0
: : r
( \ \
r o
< : - , .
< : - ' .
; : 3
P I 3 l
0 -
. . . . . .
r o
V l
C J l
( \ l "
$ : : l
\ ' - . I
< : - ' .
r o
> <
( \ l
~
Q
Q < : - ' . "
~
e )
U l
< : - , .
< : - , .
~
C 1 l ~
~
c o
c T
Q
( \ \
< ; - - >
z n ; : 0
> - < : z \ ' - . I
x
\ ' - . I
$ : : l
0 0 c o ~
: : s e
~
0
~
r l ' , < . Q
c t -
0 -
0 -
r t : 3
~
r o 0 - ~
~
, c o C 1 l
V l r o r o C J l "
"
"
U l
- 1 . - - ' V l
e )
- s
: : r
c o P I V l
c o
: : : J c T . . . . . • - 0 0
U l
. . · . . . · . 0 .
, - - . . .
r t "
. - . . .
- n o : : s
. , . , . ,
" T 1
\ J
. . . J . : : : l
- 0 " T 1
\ J
V l \ J
\ J - 0 " T 1
" T 1
' - ' "
( " )
n
- 0
P J e )
0 c o
( " )
: : : J O c o
. . . . . .
r l ' r o - t ,
,
~ , , - t ,
: 3
U l r o
: : z
t T l
. p o t T l
U 1
N r - o U 1
r o W
t T l t T l W W
" . . . . . .
U l
c T
. . . . .
- c . . . . . ( " )
. . . . . . . 0 0
. . . . .
~ W
: : s N 0 ' \ 0 ' \
W 0 ' \
o P >
e r o
e ) . . . . . .
V l
,
0 -
v . . . . · c o
<
r o : : s
P I
r o
O : : : J r t
0 - c T
I
I
I
I I
I I I I I I
I
c T
. p o . p o
t T l 0
t T l
. . . . . U 1
t T l U 1
t T l t T l o P > o P > - I ' : >
0
. c . . C l
0
. . . . . . : : : l
W 0
N N W
W N " ' - J 0 ~
W 0 0
" ' - J 0
. . . . .
N . . . . . . W 0
1 . 0 \ . C 0
\ . C \ . C - I ' : >
~
. p o
0
P J
. . . . . \ . C
N U 1
O " l " ' - J U 1
a 0 W ~
: : s
0
- t ,
0
< . Q
c T
: : r
. . . . . . . . . . .
I - ' . . . . . P J
r t o P >
O " l
. p o
. p o . p o
. . . . . . . r o
: : r
0 t T l N
0 a t T l 0 0 O " l 0 0
X
c o
- I
. p o
. . . . . .
\ . C
. . . . . . . \ . C N \ . C
0 0 c o 0 0 0 m
: : E :
. . . . . .
\ . C P I
c o 0 0 t T l " ' - J N W 0
W 0 0 " " W 0
- I ' : >
,
. p o
. . . . . . . .
I
. 0
I I 0 0 " ' - J N \ . C N \ . C W
c o o P >
~
e
Q J
~
P I
: 3
0
r t
r o
< . Q
c T
. . . . .
0
r o
,
: : s w W N
0 w W
W W W W
W W W
P I
U l
. p o . p o
. . . . . . p o
. . . . . . .
a 1 . 0 0 N
N
W N
W W
\ . 0 a w \ . C
" ' - J ~ N N
0
" ' - J C T
0
+
. p o
0 0 N O " l
o P > U 1
0 0 e x > o P > U 1 . . . . . . . t T l
0 -
"
* *
* *
* * *
* * r t
* *
*
* *
* * *
* * : : r
* * *
* *
0
* * *
* *
* *
* *
* * r o
*
( Q
. . . . .
0
r o
( / )
0 0
a a
0 0
0 0
0 0
0
0
: : : l
r
\ . 0 \ . 0
1 . 0 1 . 0
\ . C \ . C
\ . 0 ~
1 . . 0
1 . 0
\ . C \ . C
\ . 0
0 0 c o c o 0 0
0 0 \ . C
~ . . . . . . .
1 . . 0
t T l t T l
- s
. p o
. p o
N 1 . 0
0 0
O " l
" ' - J
' - J
a
0 0
" ' - J
0
* * * *
*
* *
*
*
*
*
*
- 0 . , . ,
* * * *
* * *
*
*
*
*
*
* *
* * *
*
* *
*
*
*
* *
*
: : r
*
c T
. . . . . .
0
. . . . . .
0
0
0
,
. , . , - 0 . p o
- 0
t T l . . . . . .
0
m
" ' - J
, ,
C D
. . . . .
1 \
U 1
W
. p o
e x >
0 ' \ . . . . . .
C L
. . . . . .
c o 0
c o
0 0
" ' - J
. p o
0
Q J
V l •
U l V l
: : : J
: : s
: : : J
: : s
: : : l
" "
: : s
c T
: : r 0
- 0 V l V l
c o
V l
V l
U l
V l
,
C D 0
. . . . . . 0
V l
e ) <
: : : l
. . . ·. C O - 0
U l
3 0 . C O
( " )
: : r
C D
: : : l
U l
: 3
c o
: : s
t i l
6 0 1 7
410
APPENDIX 9(a)
Monthly vari at i on of numbers of G. steY'nicla at each maturity
stage over the whole study peri ad.
Femal es Males
Month I 0 M R I D M R
1980/05 1 3 4
06 2
07 1 1
08 1 3
09 1
10 2 4
11 2
12 2 1 2 4 1
1981/01 3 2
02 4 1 2
03 1 1
04 1
05
06 2 1
07 9 2 2 1 6
08 5 1 8
09 2 5 5
10 2 1
11
12 1 4
1982/01 1 1 3 4
02 2
03 7 2 1 9 3
6
04 6 1 3 2
2 1
13 27
05 8 5 5 5
06
07 2 3 1
6
08 1 2
2 5
) : : >
3 :
3 :
t - '
- 0
0
. . . . . . .
0
\ . 0
- 0
: : l
\ . 0
: : = I
C O
r n
r l "
C O c - r
0
: z
: : l "
. . . . . . .
: : : r
< ,
. . . . . . . .
C l
0
. . . . . . .
0
0
a
. . . . . . . a
. . . . . . . t - '
a
a
0 0 0 a
U 1
m
. . . . . .
0 0 > <
- ~
. . . . . . . t - '
. j : : : > a
N
m U 1 N
W
<
\ . 0
O J
- s
0 -
O J
. : z
r l "
c
_ _ 0 -
3 : 3
. j : : : >
N
. . . . . . .
w
a 0
0 " ' 1
N N 0
. . . . . . . . . . W
0
0
< ,
< ,
< , < ,
: : l
- n r o
" -
. j : : : >
. . . . 0
. . . . . .
. . . . . .
N a
w
. . . . N w
. + : : > - ;
N 0
-
- -
-
-
-
- -
-
0
V l
- t >
V I
c o
X
. . . .
w
\ . 0
a
. . . . 0
a 0
- . . . J a a
- s
.
.
O J
3 :
c o
\ . 0
. j : : > ~
N
N W
\ . 0
. + : : > . + : : >
0 0
r l "
r o
U 1
0
. j : : > - . J
W
. . . . .
. . . . U 1
~
0 . + : : > 0 0
P J
\ . 0
. . . . . . . . . . . . .
. . . . U 1 - P >
. . . . 0
- . J W . + : : >
0
: : = I
V I
1 +
1 +
1 + 1 + 1 +
1 + 1 +
1 +
O J
0
. j : : >
0
a 0 0
a
: : l
. "
V >
0 -
r o
r r l
U 1
. j : : > . . . .
0 0
a a
3
. j : : : >
. . . .
N N
U 1 ~
0 0
G " >
P I
. . . . . .
t i l
. . . . . .
r o
G ) V I
V ' >
. . . .
0 0 0
w a a
. . . . . . .
- t >
0
0
. j : : >
N N
W W
W
- s
~
" - J
. j : : : >
0 " \ U l
m ~ 0
; ; 0
U 1
N
. . . . \ . 0 U 1
w U l
P J
~
: : = I
u : l
C 1 )
r o
, . . . : . .
c t -
w
· N a 0 0 0
0
( \ \
.
' " : :
. p o
. j : : : >
U 1 w
W
N ~
; s
. . . . . . . " - J
w 0 0 - . . . J 0 ' >
m
~ .
. . . .
. . . . . . N U 1
0 0
N 0 " \
( ; )
< : - >
~
0
<
. . . . . .
N . . . .
N a 0
0
c o
- s
. j : : : > 3 :
a \ . 0
0 " \ a \ . 0
c o
r o
N W
- . J W N
0 0 U 1
P J
r l "
. . . . . .
N
0 " \ N 0 " ' 1
0 0 W
: : l "
: : = I
c o
1 + 1 +
1 + 1 +
1 +
: : : : :
a
a
0 0
: : l "
V >
r r l
0
. j : : : > 3 :
0
~
0
- '
P I
0 0
0 0 0 0 ' >
- - '
c o
r o
V I
G )
r l "
C
V ' >
. . . . . . .
0 0 -
a N
0
.
. + : : >
~
. + : : >
\ . 0 0 ' >
. + : : >
- . J
. . . . . .
U 1
. . . . . . ; ; 0
a . . . . c o
0
P J
- s
. . . . .
: : = I
u : l
0
0
N r o
0 -
N N
e o
N
\ . 0
" - J
W
. + : : >
U 1
U 1
. + : : > . . . . . .
U 1
N
1 1 1 7
I - - '
. . . . . .
0
. . . . . . 0 0
. . . . . . . . . . . . 0
. . . . . . 0
< = >
3 :
C > C D
+ : > + : > \ . 0
U 1
> - ' 0 ' 1 + : >
+ : >
0 ' - J c o
I - - ' O J
. . . . . . U 1
- . . . . . J W
e n
N W ~
~
0 \ D
( X l W ~
' - J U 1 0 : : : s
N e n
0
W
\ D + : : >
1 + 1 + 1 + 1 +
1 +
1 + 1 + 1 + 1 +
1 +
1 + 1 +
0 C > U l
0 0 0
0 > - '
0
< = > < = >
< = >
r r i
. . . . . . . . . . . . . . . . . . . . . . . . I - - '
0 > - '
0
. . . . . . N
3 :
< = >
. . . . . . N
- . . . . . J O J
. . . . . . W O " l + : : > e n < . T 1
O " l
< = >
- l
C D
G J
( / l
. . . . . .
. . . . . .
0 0 . . . . . .
0 0 0 C >
0
< = > < = >
. . . . . .
N > - ' 0 + : : > w w
W N
> - ' O ' l
~
' - J W
' - J W 0 - . . . . . J ; : 0
O " l O " l
0 < . . v
< = >
0 e n
w w 0 : >
0
+ : : > - . . . . . J O J
c o C >
< = >
: : : s
( . Q
C D
. . . . . .
. . . . . .
. . . . . . N N > - '
N 0
I - - '
0 > - '
0 ' )
> - ' + : : > ' - J
0 ' 1 U 1
+ : : >
\ . 0 U 1
< = >
O ' l
0 ' )
N
U 1 U 1 0 : >
\ D 0 ' 1
0 : >
N I - - '
U 1
W
' - J U 1 N 0 W
+ : : > N
+ : : > - . . . . . J
- . . . J
2 1 1 7
: : t : >
3 :
> - ' 3 :
. . . . . .
- 0
0
3 :
0
s o
- 0
< . 0
: : : l
C D
: : : : l
C O
r n
C O
O l r - t -
M -
0
. . . . . .
: z :
~
: : : l
< ,
: : r
< ;
- - '
C J
V I
0
0
0
0
. . . . . . . . . . . . 0
. . . . . .
. . . . . . .
0
0 0
0 0 0
U l
e n
- - J
< . . 0 C O
0
. . . . . . . N > - '
N
e n U 1 W
+ > > <
1 +
<
( / )
0 1 < . . 0
- S
r n
n
0 1
r - t -
> - '
. . . . . . . . . . . . .
. . . . . . .
. . . . . . . . . . . . . . . . . . .
. . . . . . . > - '
. . . . . .
> - '
0
W
. j : O >
N
W
. j : O >
W W
U 1
U 1
+ >
+ >
: : : l
n
~
C O
U 1
- . . . J
N 0
N
e n 0
C O C O
- n
. j : O > ' - l
. . . . . . .
e n
N
. . . . . . . 0
N 0 " \
0 N
0
- l
- l J
1 +
1 + 1 + 1 +
1 + 1 +
1 +
o - . Q .
o
0
0 0
0 0
0 0
0
.
: : : l
0
. . . . . . .
. . . . . .
0 0
0 0
a .
N
. . . . . . .
. . . . . . e n e n
- - J
c o
- n
r - t -
r o
3
0
O J
- - ' : : : l
r o
- l J
> - '
. . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . .
. . . . . . .
> - '
0 1
( )
W
N
. j : O > W
N W
W
W U 1
- I ' : >
+ >
r - t -
' - l n
. . . . . . . W
. . . . . . - . . . J
- - J 1 . 0 N c o
- P o U l
0
. j : O > - n
> - '
N
N W
N U l U 1
w - I ' : > C O
- S
c . . n
V I
1 +
1 + 1 + 1 +
1 + 1 +
1 +
o - . Q .
- l J
0
0 0
0 0 0 0
0
- S
0
. . . . . .
0 0
0 0 0
0 )
. . . . . . . w
W - . . . . . J e n
e x : >
c ; ' : )
C J l
I " i -
\ \ l
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
> - ' " : :
; : s
. j : O >
. j : O > . " .
0 ' \ W
W - P o U 1 - I ' : > - P o W
. . . . . . .
N - . . . . . J ' - l n C O
U 1 C O W U l - P o
~
. . . . . . , . ,
. j : O >
. . . . . . . N
W - . . . J 0 ' - l W
- . . . . . J
C O
i )
- l
1 + 1 +
1 + 1 + 1 + 1 + 1 +
~ 0
0 0 0
0 0 0 0
<
.
C D
0
0 0 0
0 0 0
- S
. . . . . . ,
0 )
W
0 U 1 W
0
3 : c - t
O J
~
C D
r o
: : E :
~
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . .
. . . . . . . . . . . .
. . . . . .
n
0
. ,
. j : O >
U 1
w - P o - P o W . j : O >
- I ' : > 0 ' \ W
c . . n ( 1 )
. . . . . . .
N U 1 - . . . . . J
U l . j : O >
C O
N
0
U 1
. . . . . . ,
. . . . . . .
N
- . . . . . J 0 N
N
N N
W
~ U l
c - t
1 +
1 +
1 +
1 +
1 +
1 + 1 +
C
a .
0 0
0 0
0
0
0
0
0 0
0
0
0
0
W - - . . J
0 U 1 0 )
- + : : >
0
( 1 )
- s
- ' .
0
a .
) : : >
. . . . . . . 3 :
. . . . . . .
- 0
0
~
~
- 0
: : : s
C O
C O
r r t
. . . . . . .
M '
N
: z
: : r
0
< : )
0
. . . . . . . 0
. . . . . . .
. . . . . . .
0
0 -
0
0 0 0 . . - .
0 0
~
-
\ . D C O
. . . . . . .
0
N
. . . . . . .
( J " I N
W
~ 0 " \ ~
0 0
> <
~
( " )
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . .
( " )
. . j : : : > 0
U ' 1
( J " I
W
W
. . j : : : > ( J " I ( J " I
0 " \
U 1
~
C O )
: : s
< : )
U ' 1
. p - . . . . . . .
~
. . . . . . . . . . . .
. . . . .
N
' - J e n
r l "
N
. . . . . . .
1 . 0
. p -
0
O " >
. . j : : : >
N W
W
' - J
. . . . . .
- I
"
: : s
1 +
1 +
1 +
1 +
1 +
1 + 1 + 1 +
1 + 1 +
s : :
~
C D
< : )
0
0
0
0
0 0
0 0
0
.
0 -
.
.
< : )
0
0
0
0
0 0 0 0
0
< . . . V
. . . . . . . U ' 1
W
~
N N N W W
C O
"
3
O J
. . . . . .
C O
. . . . . . .
. . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . .
. . j : : : >
( J " I
U ' 1
. j : : > . . j : : : > W
0 " \ W
U 1
- P > ~
< : )
. p - C J
< . . . V
. . . . . . . O ~
0
~ 0
0 0 0 0
. . . . . . . O ' l e n
( J " I . . . . . . .
- P >
\ . D 0 " \
e n w
~
V I
"
1 +
1 +
1 +
1 +
1 + 1 + 1 + 1 + 1 + 1 +
~
< : )
0
0 0
0 0 0 0
0 0
.
.
.
< : )
0
0
0
0 0 0
0 0 0
W
W U ' 1
N N 0
W W W
~
. . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. j : : > . . j : : : > . p -
( J " I . p - . . j : : : >
( J " I . . j : : : >
. . . . . . .
~ O " l - P >
. . j : : : >
' - J . . j : : : >
. . . . . . .
O " l N . . . . . . .
0 N - . . . . J W
C J
C O
. . . . .
C O
0 0 0 0 ~ - . . . . J
C O \ . D \ . D 0
C O
- l
"
- 1 +
1 + 1 +
1 +
1 +
1 + 1 + 1 + 1 +
1 +
1 +
~
0 0 < : )
0 0
0 0
0
0 0
.
.
0
.
0 0 0 < : )
0 0 0 0
0 0
W N
N N . . . . . . .
O " l
U ' 1 w -
0 N
N
U ' 1
3 :
O J
. . . . .
C O
. . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . .
. . . . . . . . . . . . . .
. . . . . . .
. . . . . . . . . . . . . .
. . . . . . .
. . j : : : >
( J " I
~ . . j : : : > . p -
. . . . . . .
- P > ~
. . j : : : >
O " l . . j : : : >
~
N
U 1
' - J
~ N
N
~ . . . . . . .
U ' 1 < : )
W C O )
N
' l N
. . . . . . .
N
0 0 \ . D
. . . . . . .
C O U ' 1 . . j : : : >
W
U ' 1
V I
"
1 +
1 +
1 +
1 +
1 + 1 +
1 +
1 +
1 +
1 +
1 +
~
0
0
0
0 0
0 0
0 < : )
0
0
.
0 0
0 0
0
0
0
0
0
0
0
W N
N . . . . . . .
' - J
W
U ' 1
U 1
W
N
N
415
APPENDIX lO(a)
Monthly variation of numbers of c. riisei at each maturity
stage over the whole study period.
Females Males
Month I D M R I D M R
1980/05 13 3 1 3-
06 1 1 1 4
07 1 5 1 2
08 1 3 2
09 3 1 2 1 2
10 1 1 2
11 1 2
12 17 1 6
1981/01 5 2 2
02 2 2 4 3 2 1
03 1 1
04 1 2
05 1 6 1 3
06 2 1 2 5 3
07 1 1 2 2 4
08 2 2
09 1 1 3 2 2
10 5 3 2
11 1 7 5 5 1 2
12 3 6 6 1
1982/01 4 3 1 1
02 3 2 1 3 2
03 3 1 2 1
2
04 1 3 2
1
05 2 1 2 2
6
06
1 1
07
08
) : : : >
3 :
. . . . . . 3 :
- 0
0
~
0
~
- 0
~
~
: : l
0 0
f T 1
r t
C O
r l "
0
: z :
z r
~
< ,
= >
- '
< , 0
0
0
0
0 . . . . . . .
0
~
. . . . . .
t - - '
0
0
0 0 0 0
( J )
U 1
- . . . . J
. . , .
C O
1 . . 0
. . . . . .
0
( J ) > <
~ N
U 1 W N
<
~
P o >
0
- s
. . . . . .
0 -
P o >
: : z :
r t
s : : : :
. p o
3 : 3
- . . . . J
N C O
( J ) W
. . j : : > N W
~
W N 0 ' 1 0
0 0 _ 0 -
< ,
< ,
< ,
. . . . . . . .
< ,
< , . . . . . . . .
< ; . . . . . . . .
< , < , < ;
< , . . . . . . . .
~
- n r o
. . . . . .
t - - '
~
w
( J 1 ~ w
. . . . . .
( J )
\ . 0
~
. . . . . . 0 ' 1
W
- s
w
0
V l
. . . . . ,
V l
( 1 )
X
t - - '
U 1
t - - '
~
~
- . J
t - - '
( J ) W
w
0 C O
0 0 w
- s
3 :
P o >
C D
. . , .
. . , . W
0
- P o
1 . . 0
N
. . . . . . ' - J
0 " 1 0 0
C O C O
r t
O J
. . , . U 1
0
. p o N
. . j : : > C >
1 . . 0
~ W
' - J W
N
: : l
t - - '
\ . 0
0 ' 1
~
( J )
~ s o 0 0 - P o
C O
W
0 0 0 0
0
V l
1 +
1 +
1 +
1 +
1 + 1 +
1 + 1 +
P o >
C / )
~
W
t - - '
~ 0
N
0
~
- n
r T l
0 -
C D
. . j : : > O ' l
N O ' l
0 " 1
- . . . . J 0 0
. . , . 3
( J ) ~
O ' l
" " - J 0
W
G l
O J
- - '
V l
I - - <
C D
V l
G >
( / )
( J ) ~ " " - J
. . . . . .
" " - J
0
0
. . . . . ,
. . . . . . . .
0
. p .
W
~ N
N - P o
W
. . , . ; ; 0
~
~
~ - . J
N
0 0
U 1
. . , .
O J
( J ) ~ *
w
. . . . . .
0 " 1
1 . . 0
C J
: : l
u : : : l
C D
' " S
~ ~
~
t - - ' ~
t - - '
" , .
. p .
N
t - - ' \ . 0
0 0
0
" , .
C 1 l
( J 1
~
" " - J W
0
U 1 ~
( \ )
( J )
~ \ . 0 W C O
- . . . . J N
" , .
( J 1
. . j : : >
t - - ' - . . . . J C >
~ 0
0
<
( 1 )
~
. . . . . . ~ ~ t - - '
~
t - - ' 0 0
0
.
r t
3 :
. p .
. . . . . .
N C D : : s
0 C O \ . 0 0 C O C O
( 1 )
. . j : : > ( J 1
~
O J
w ~ 1 . . 0 - . J C O
C O
( J )
~
- . J N : : l
W " " - J " " - J " " - J - . . . . J
: : : E :
: : r
1 + 1 + 1 + 1 + 1 + 1 +
1 + 1 + 1 +
0
C / )
- '
0 0 0 0
0 0 0 0
( 1 )
r T l
. . , .
. . . . . .
t - - '
N W ~
U " 1
W
3 :
. . , .
. . , .
t - - '
1 . . 0 - P o - . . . . J
C O O J V l
C O
r t
C D C
0 -
G )
U ' )
0 0
0
0 C > t - - ' . . . . . . . .
0
0
.
.
( 1 )
. . , .
w ( J )
U 1
C O U " 1
C O N
~
( J )
' - J
C T I
W . . j : : >
N t - - ' . . . . . .
- . J
; ; 0
*
( J 1
0
0 0 U 1 . . j : : >
0 N
V o l
O J
0
: : l
0 -
u : : : l
C D
. . . . . .
. . . . . .
t - - '
~
N ~
t - - '
N
U 1 ~ N
O ' l
W
N
U 1
V o l
. . , .
W C T I
. . j : : >
0 1 . . 0
C O
0 0
( J )
N
\ . 0
- P o
C O
0 t - - '
V o l
9 t V
: D
. . . . . . . 3
. . . . . . .
" " " ' 0
0
1 . O
1 . O
" " " ' 0
~
C O
[ " T l
C O
r +
. . . . . . .
I ' V
: z
z r
0
C )
0
. . . . . . . 0
. . . . . . .
. . . . . . .
- - - . . . . . . .
0
0
0
0
0 0 0 0
- . . J
- - -
i l l 0 : >
. . , .
. . . . . . . 0
r - o
. . . . . . .
( j ) N
W
U 1
C O > <
"
. . . . . . .
0
0 -
Z
C
. . . . . . .
3 3
C O
- P >
( j )
e n
- . J
U " 1
C )
C ) U 1 U 1
0 " >
N
0
_ C 1 " n
< ,
< ,
< , "
< ,
0
- n c o
C )
. . , . I ' V
N
. . . . . . .
U 1
U 1
C )
C ) . . . . . . . . . . . . . . N - P >
0
- - -
- -
~
-
- S
- - -
- - -
-
- - N
- - -
( / )
r +
~
c
C O
0 -
. . , . 1 . O
N
N C O
I - - ' U 1
N O " ' l
- P >
3
. . , .
C O
c : : >
C ) N
- P o
. . . . . . . I - - ' N
0 " > + : >
. . , .
O J
I ' V
W
. . . . . . .
0
0 " > + : >
c c U 1 i l l
~
1 . O
i l l
i l l
I ' V . . . . . . . C O
c o 0
0 U 1
1 +
1 +
1 +
1 + 1 +
1 + 1 +
C . / l
C ) . . . . . . .
0
I ' V
0
0
I T 1
. . , .
C D
( j ) . . . . . . .
. . . . . . . "
- P o
U 1
3
0 " >
. . . . . . .
- . J W
C O i l l
O J
- '
C D
G )
C . / l
I - - ' 0 U " l
< . . v w
0
. . . . . .
I - - '
i l l U 1 i l l
< . . v
; ; > ; J
~
I - - '
~
+ : >
"
O J
I - - '
0 " > 1 . O < . . v U 1
- P o
" "
"
~
c o
C O
. . . . . . .
. . . . . . .
( j )
N I ' V N
W O " ' l
. . . . . . . . . . . . . .
N - . J . . . . . . .
< . . v
. . . . . . .
U " l
U " l - . . J < . . v
C O N
C O C O U 1
C O
"
C )
. . . . . . . . . . . . . . C )
I - - ' . . . . . . .
. . . . . . .
. . . . . . . I - - ' . . . . . . .
0
0
3
. . , .
. . . . . . .
I - - '
~
r v . . . . . . .
- P o . . . . . . .
C O
~ 0 I . D
C O
( j )
~ ( j )
i l l 0 " >
C O U " 1
0 0 - P o . . . . . . .
0 " >
W O J
. . , .
W
- P o U " l
C O 0 U 1
C O - . J c : : >
~
"
"
1 +
1 + 1 +
1 +
1 +
1 + 1 +
1 +
1 +
1 +
0
0
0 0 C )
0 0
0
c : : >
C . / l
r n
. . . . . . .
W N ( j )
W
0 . . . . . . .
W
0
3 :
. . , .
. . , .
U 1 w
0 " >
. . . . . . .
W
< . . v
O J
- '
"
C D
G )
t n
0 0
0
0
. . . . . .
0
0
0
. . . . . .
c : : >
.
N
C O 0 " >
N
. . . . . . .
I . D
C O
~
~
i l l
N
" C O
W r v
U " 1 ; ; > ; J
"
. . , .
l O
W
0
W
N
"
"
W
- . J
< . . v
O J
~
u : : l
C O
. . . . . . . I - - '
I - - '
. . . . . . .
. . . . . . .
. . . . . . .
C )
I - - '
. . . . . . .
i l l . . , .
U 1
C O ( j )
N
1 . O
+ : >
0
C O
. . . . . .
" . . . . . . .
0
. . , . - P >
+ : >
l O
0 N
"
"
W . . . . . . . ( j )
N
1 . O
L T 1 7
. . . . . .
. . . . . . . .
1 0
< . D
c o
C O
o
. . . . . . . .
o 0
o
-
o 0
0 0 0
o
O ' l C . J l
- - -
C O
N . . . . . . . .
C . J l + : > W
O ' l
. . . . . .
o
( )
. . . . . . . . . . . . . . . . . . 0.
. . . . . .
. . . . . . . . . . . . . . . . . . 0. . . . 0
. . . . . . . . . . . . . . . . . . . . - 0 . . . . . . . .
o
O O O C O
0 0 0 \ . 0 < . 0
: : ' l
0 . . . . . . . . . . " . ' . 0
n
- . . J . p : . c o N
O ' l N O O O U 1
o . p : . N ' " C O
" T 1
W - . . J N ' - J
- . . J . . . . . . . . C O O . . . . . . . .
N C . J l C . J l ( J " 1 0
o
- I
- t ,
1 + 1 +
1 + 1 + 1 +
1 + 1 + 1 +
( )
o 0
o 0 0
o 0 0
o
: : ' l
o 0
o 0 0
o 0 0
0 .
C O N
O ' l N < . 0
N . . . . . . U 1 . . . . . .
r l -
o
: : ' l
- t ,
o
o o
o
o
o
Q I
o
o c o
< . 0
< . D
C O
< . 0
o
< . 0
0 0 ' "
r l -
n
N N
W
- . . J
< . 0
o N 8
W O " l " - l
< . 0
o
. p : .
" T 1
' - J
N
. . . . . . . . - . . J o
o
+ : > " - l " - l - . . J
W
- s
V l
V I
1 +
1 +
1 +
1 + 1 + 1 +
1 +
1 +
o
o o
o
o o
o
o
o
o o
o
o o
o o
N
U 1 O ' l
U 1
. . . . . . . . N N
W
. . . . . . 0 o o
o
o
1 0
< . 0 < . D o o
o o C O
o 1 0 0 0
o ' "
. p : .
. . . . . .
. . . . . . . . n
' - J
. . . . . . . . W W o o C . J l
O " l W 1 0 o
. p : . . p : .
. . . . . .
" T 1
U 1 ' - J 0 0 W W
N W 0 0 N C . J l < . 0
- I
1 + 1 + 1 + 1 +
1 + 1 + 1 + 1 + 1 + 1 +
1 + 1 + 1 +
o o o o
o 0 o o o 0 o o
o
. . . . . . 0 o o o o 0 o
o o o o
o
. p : . . p : .
. . . . . . N N
W W + : > N W
W N N
r o
C D
. . . . . . 0 . . . . . . . . . . . . . . . . . . 0 . . . . . . . . . . . . . 6. . 0 . . . . . . . . . . . . . 0. 0
V I
0 1 0 0 0 0
< . 0 0 0 1 0 C O
o
o 1 0 c o
r l -
C O . p : . W W < . 0
O C . J l W \ . O O ( J
o
o C . J l C . J l
C
. . . . . .
. . . . . . . . W O O N W 0 C . J l ' - J 0 0
U 1 C . J l W
" T 1
0 .
V l
1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 +
1 +
1 + 1 + 1 +
0 0 0 0 o 0 0 0 0
o
0 0 0
C D
~
o 0 0
o 0 0 0 0
o
. . . . . .
0 0 0
. p : . N . p : .
W W + : > W N
W
W N N
o
0 .
8 1 t 7
) : : >
. . . . . . . .
3 :
. . . . . . . .
- 0
0
1 . . 0
1 . . 0
- 0
~
0 0
0 0
r r i
r i -
. . . . . . . .
N
: : z
z r
0
0
0
I - ' 0
I - '
. . . . . . . .
0
0
0 0
0 - - - . . . . . . .
a
0 0
- . J
- - -
~ c o
0
I - '
N
. . . . . . . .
N
W
- . . . . . J U 1 ~
0 ' 1
O J
> <
. . . . . . . .
0
( )
. . . . . . . .
I - - '
I - '
. . . . . . . .
. . . . . . . . 0
0
I - ' I - ' a
.
( )
I - - ' 0
0
c o
0
< - 0 0
s o
N 0
( )
0
1 . . 0
I - - '
0
- . . . . . J
W
W - . J
N W ~
~
. . . . . . . .
- n
I - - '
( J 1 w
m
- . J
I - - '
0 0
- . . . J
0
M -
- l
- ' .
1 +
1 +
1 +
1 + 1 +
1 +
1 +
1 +
~
~
c :
0
a
0 0
0
0
a 0
.
C D
0 -
0
a
0
0
0 0
0
0
I - - '
+ : >
W W
I - - ' N w
+ : >
- n
C D
: 3
O J
- - '
C D
. . . . . . . .
I - - '
0
. . . . . . . . 0
I - ' 0
a 0
0
0
\ . 0 0
\ . 0 \ . 0
I - ' < - 0 c o 0 c o
c »
+ : >
N
I - ' U 1 ~
' - J 0
m < - 0 c o
0 - n
' - J
0
- - . J + : >
\ . 0 + : >
~ ~ + : >
V >
1 +
1 + 1 + 1 + 1 +
1 + 1 + 1 +
~
a 0
0 0
a 0 0 0
.
a 0
0 0 0 0
a 0
I - - ' . . . . . . . . N
W W a
+ : > N
6 1 1 7
420
APPENDIX ll(a)
Month 1y varia t ion of numbers of A. bimacu Latus at each maturity
stage over the whole study peri od.
Females Males
Month I 0 M R I D M R
1980/05 1 1 1 1
06
07
08 2
09 1
10
11 1
12
1981/01 2 1
02
03
04
05 1
06
07
08
09 1
7 1 4 4 210
11 3 1 3
12
1982/01 2
02 1
03 1 1
04 1
05 2
06
07 1
08 2
: t : : >
3 :
I - ' 3
I - '
- 0
I - ' 0
0
. o
< . . D
- 0
: : : 5
l O
~
O J
C O
r n
r - t -
O J
r t
0
I - '
: z
N z r
< , : : r
< ,
- '
0
I - ' 0
0
0
. . . . . . . .
I - ' > - '
0
0
0
( J 1
-
I - '
I - '
< . n
I - ' 0
I - '
> <
W
C O
<
> - '
0 1
. . . . . .
- s
. . . . . .
0 1 C T
: z :
c - t -
C
3 : : 3
I - '
0
. . p o
0
0
N
N
I - '
0
0
_ C T
< ,
- . . . . .
< , - . . . . .
: : : 5
- n r o
I - '
I - '
N
. . . . . .
I - ' 0 " >
0
I - ' 0
-
- - - S
-
-
l J )
0
- + ,
V l
r o
X
I - '
. . p o
0
0
I - '
0
0
- s
3
0 1
r o
. . p o I - '
I - '
N
- . J 0 " >
N
r t
O J
W
N
N
0 . . 0
- . . J
~
I - '
. . . . . .
( J 1
~
( . V - . J
- . J
0
O ' l
U 1
0
V l
1 +
1 +
1 +
0 1
U ' 1
0
0
: : : 5
" " T l
r n
r o 0 -
( J l
0
: 3
w
0
G )
O J
- - '
V l
. . . . . .
r o
V l
G : l
V l
0
0
- + ,
. . . . . . . .
0
I - '
. . . . . . . .
; 0
- s
C l - l ?
< . . D
. . . . . . . .
O J
W
0
~
~
~
r o
( . V I : Y
0
" ' .
. . . . . . . . N
~
O ' l
C O
Q
I - '
0
~
I N
P
I i -
~
( J )
I - ' I - '
0 0 0
0
3
( J 1
( J 1
0 . . 0
0
- P > r o
<
I - '
O ' l
N O J
C O - P >
r o
( J 1
. . . . . . . .
C O O ' l ' - . l
~
- s
1 + 1 +
1 +
1 +
r t
= s -
0 0
0
U ' 1
r o
r T 1
N
W
0
3 :
: E
W
~
< . n
O J
= s -
O
- '
r o
r n
G : l
l / )
V l
0
a
a
. . . . . . . .
r t
.
C
N
- . J
c o
C l - l ? 0 -
. . p o
N
O J
; 0
' - . l W
a
O J
~
~
r o
. . . . . ,
r o
I - '
0
a
0
. . p o
' - . l
0 . . 0
0 -
- P > 0
0 . . 0
. . p o
N
< . . 0
t 2 1 7
) : : : >
: : s :
. . . . .
: : s :
. . . . . . .
. . . . .
- - 0
0
: : s :
0
\ . 0
\ . 0
\ . 0
: : : : : l - - 0
r o
: : : : : l
e n
C O
e n
r n
O J
M -
M -
. . . . . . . 0
N
- . : : J
: : : : : l : z :
< ,
: : : : : l " "
< ,
. . . . . . .
V l
0
. . . . . .
0
. . . . .
0
. . . . . 0
0
0 0
> - <
' <
C J 1
. . . . . . . . . . . . .
. . . . . . . . . .
U " l
0
W
0 0
1 + > <
<
. . . . . . .
O J
V l
) . . . . . . .
f T 1
- ' .
( )
O J
M -
. . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . .
0
. p .
. j : : : >
N
W c . . n
W
U " l
: : : : : l
. . . . . . n
e n
- . . . J - . . . J C O
e n - P >
. p .
. "
C J 1
- . . . J 0
0
U " l
e n
0
- I
- h
1 +
1 +
~
( )
0
0
.
0
: : : : : l
0 0
. p .
0 .
0
. ,
r o M -
: 3
O J
0
: : : : : l
r o
. . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
- h
O J
. p .
. j : : : > ( )
U " l
N
W W c . . n
( J )
- . . . J
W 0
0 0 0 ' - J n
M -
. ,
. j : : : >
N . j : : : >
c . . n
- P >
0
e n 0
V l
- s
V l
1 +
1 +
~
0
0
- h
.
0
0
0
- s
. p .
0
; b
0 '
t " .
. . . . . .
. . . . .
. . . . . . . . . . .
. . . . . . .
~
c - .
N
- P > W
W
. . . . . . .
~
. . . . . .
' - l
- P > ' - J
N
n
' N
. ,
. . . . .
\ . 0
' - l
\ . 0
W
R
- I < ; t
1 +
1 +
1 +
~
~ .
\ J )
0
0
0
.
0
. . . . .
0
0
<
. p .
0
C J 1
r o
)
3 :
O J
- - '
M -
e o
- . : : J
r o
. . . . . .
. . . . .
. . . . . .
. . . . . .
. . . . . . .
: < :
- . : : J
N
- P > W
W
. . . . . . .
. . . . . .
0
0 " >
W
0 ' \
0
n . . . . .
. p .
N
N \ . 0
0 " >
r o
V l
"
1 +
1 +
1 +
V l
~
M -
0
0
0
e
. . . . .
0 .
0
0
. p .
. . . . . . .
- P >
' <
r o
)
. . . . . .
0
0 .
423
APPENDIX 12 (a)
Monthly variation of numbers of C. aeneue at each maturity
stage over the whole study period.
Females Males
Month 0 M R I D M R
1980/05 2 5 3 5 10 29
06 3 10 7 33
07 3
08 1
09 5 2 2 1
10 1 1 1
11
12 3 2
1981/01 1 2 3 2
02
03 2 1
04 1
05 2 2
06 1 2 2
07 1 4 5 1 12
08
09
10 1 4 6 1 4 9
11 1 2 1 4
12 2
1982/01 1
2
12
02 1 7
2
03 6 1
1
04
2 105
06
07
08 1 3
2
3 :
. . . . . . > -
3 :


0 - - 0
0
1 . O
~
- - 0
~
~
e x >
O J
r t r r t
r l -
0


: : : r
: z
< ,
: : : r
- - - '
0
0
0
0
0
t - - '
. . . . . . 0
. . . . . .
. . . . . . .
0
0
0 0 0 0
( } 1 ' <
0 ' \
- . . . . . J
( ) : J
I . . D
-
. . . . . . 0

 N
( } 1
I ' : >
L V
0 ' \ ~
> <
<
t - - '
O J
I ' : >
- s
O J 0 -
: : z :
s : : : r t
W
W
_ _ 0 -
- J .
3 : 3
~
0 C O
0
t - - '
u . . >
N 0
t - - ' N
0
N N 0
0
< ,
< ,
< , < ,
~
- n C D
. . . . . . . . . . .
- . , J W
t - - '
I ' : >
0
W
. . . . . . 0 O ' l
W N I ' : >
-
- ;
- - -
( } 1
- -
-
W
-
-
- - - -
-
0
V l
- t ,
U J
r o
> <
t - - '
. . . . . .
. . . . . .
N
' - I
( ) : J ~
c . . n
I ' : >
. . . . . . N 0
' - I N c . . n
" " " \
. 3 :
( 0
O J
( ) : J C O
t - - ' O ' l C O
t - - '
+ : > w
0 e x > w 0
r t
P J
N
' - I
( ) : J t - - '
O J
t - - '
- . , J
' - I
0 ' \ 0
W
- J .
~
- . , J
0
N
I ' : >
t o - ' 0 " >
N N O ' l
0 ' \ C O
"
0
"
U J
1 +
1 +
1 +
1 +
1 + 1 +
1 +
1 + 1 + 1 + 1 +
O J
. . . . .
V J
0
N
t o - ' I ' : >
t o - '
W 0 0 0
~
- n
r r t
( J 1 C D 0 -
. . . . .
N
. . . . . . c . . n
' - I I ' : > 0 0
O ' l
3
N
~ + : >
t o - ' N u . . > + : >
' - I - . . . . . J
e x >
P I
G l
- - - '
V l
. . . . . .
C D
G )
U J
. . . . . .
V l
. . . . . . N 0
0 L V I ' : > t T l
0 0 0
. . . . . . .
.
- t ,
( ) : J
- . , J 
 0
0 c . . n w + : >
+ : > w
. . p .
( } 1 ( } 1
; ; 0
( ) : J
' - I
w 0 " > I ' : > - . . . . . J " " " \
O J
"
. . p .
( J 1
- . , J
P J
0 c . . n W 0 ' \ O ' l + : >
+ : >
*
~
C J
< . . 0
( 0
. . . . .
. . . . . .
t o - ' . . . . .
. . . . .
~
( \ )
( J 1
N
0
0 " > N O ' l
0 L V 0 " >
+ : >
~
( } 1 I ' . ) . . p .
( \ )
0 ' \
0
0 " > O ' l O ' l
u . . > W
. . . . . .
- . , J
W
O ' l
0 u . . > ' - I - . . . . . J
L V
O ' l ~
- . , J
. . . . . .
- . , J
~ . . . . . . ( ) : J
~
O ' l W - . . . . . J O l
0
<
r o
" " " \
0 0
0
0
0 0
0 0
0
3 : r t
( } 1
' - I I ' . )
u . . > . . . . . .
+ : > ( 0
L V
W O ' l
: : : r
N 1 . O
c . . n
- . , J
W
1 . 0 ( } 1
+ : > N
P J
r o
- . , J
W 1 . O
N - . , J
( ) : J
0 ' \
( } 1


~
: : :
1 + 1 +
1 + 1 +
1 +
1 + 1 +
1 + : : : r
0
0 0
0
0
0
0 0 - - - '
V J
.
.
f T 1
r o
. . . . . .
0
0
0
0
0 0
3 : :
. . . . . .
. . p .
0 ' \
N
c . . n
N
t T l
P I
U J
r t
r o
c
0 -
G )
V l
0 0
0
0
0
0
0 . . . . . . .
r o
( } 1 ( } 1
. . . . . .
w
I ' . )
. . . . . .
N
( J 1
" " " \
e x >
1 . 0
+ : >
1 . O ; 0
+ : > - J •
*
C O
. . . . .
0 ' \
0 " >
"
W
W
P I
0
"
: : : : : : s
0 -
l D
C D
0 0
0
0
0
. . . . . .
0
.
. . p .
0 ' \
e x >
N
+ : >
. . . . . .
. . . . . .
e x >
. . . . . .
' - I
I ' . )
0
+ : >
. . . . . .
"
N
0 ' \
I . . D
N
+ : >
"
: >
3 :
- - 0
0
w
-
< o
- - 0
~
0 0
-
0 0
r n
r l "
N
z
z r
-
0
0
0
. . . . . .
0
. . . . . .
0
-
0
0 . . . . . .
0
0
0
0
0 ' - J
- C O
1 . 0
0
. . . . . .
N
N
, . \ : : >
W
U 1 -
0 "
' - J > <
0 : >
-
N
-
: z 0 -
C
. . . . . .
3 : 3
N
0
0
W
~
0
N
N
N -
0
0
N 0 0
< , 0 -
-
< ,
< ,
< ,
( ' )
- n 1 D
0
. . . . . .
0
. . . . . .
N
O " l 0 0
0 -
. j : : : > N
-
0
0
- 0
- - s
-
- -
- N
-
-
-
-
-
V I : : : 5
-
r l -
- ' .
: : : 5
s : :
r o
. . . . . .
0 .
W
( ) 1
- . . . . . J
. . p .
W
0
- . . . J
. . . . . .
N
3 :
. . p .
I D
- . . . . . J
~
' - J 0
0
W
W
i o
O J
. . . . . .
( J 1
. . . . . .
( J " l
N
. . . . . .
O " l
~
0 0
~
W
N
N
W
- . . . J O " l
-
O " l
-
1 +
1 +
1 +
1 +
1 +
1 +
1 +
1 +
( / )
0
N
. . . . . .
0
0
0
I T l - n
.
-
I D
' - J
0
W
0 "
W
- . . . J C O
' - J : 3
. . p . . . . . . .
C O
O " l
0
W
O J
. . . . .
I D
G ' l
. . . . . .
V >
- . . . . , J 0
0
I - '
N
0
0
. . . . .
U " 1
I - ' O " l
W W
U 1
C O
; 0
, . \ : : > ' - J
~
. . . . . . W
e n
0 0
l O
N O J
W
C O
1 . 0 s o U " 1
- . . . J
~
c o
I D
. . . . . .
. . . . . . I - ' N
. . . . . . . . . . . . 0 0
, . \ : : >
N 0 " 0
. . . . . .
. . . . . .
U " 1
1 . 0
0 0 0 0
. . . . . . . . . . . . I - ' N
O " l
W 0 "
. . p .
( J 1
N W l O
U 1 W
0 0 0
0 0 0 0
3 :
( J " l
W W N
~ W
W I D
. . p .
. . . . . . ( ) 1
C O W
0 0 " O J
( J 1
~
- . . . J 0
C O ~ O " l
~
1 +
1 +
1 + 1 +
1 +
1 + 1 +
1 +
0
( / )
0 0
0
0 0
0
.
. .
f T 1
0
0
0 0
0
0
0
: : s :
0 )
, . \ : : > . . p .
. j : : : >
W
~
w O J
. . . . .
r o
G ' l
V >
0
0
0
0
0
0 . . . . .
0
W
I - '
W
N
N
W
N
~
( ) 1
( J 1
~ O " ' l ( ) 1
0
: ; : 0
0
C O
W N
0 )
C O . . p .
l O
O J
~
l . Q
I D
0
0
0
0
0
0
0
.
~
~
U " 1
W
~ - . . . J . . p .
- . . . . , J
W
U " 1 0 )
W
- . . . J
C O
0
. . . . . .
- . J
' - J
. j : : : >
N
W
S Z v
~
3 :
. . . . . .
3 :
u
0
f - ' 3 :
0
c o
U
: : 5
C D
c o
: : 5
C O
r r t
( X ) ~ r t
M "
0
z r : z
: : 5
f - '
< , : : 5 " " "
- - '
0
t i l
0
0
0
a
t o - ' 0 . . . . . . . .
. . . . . .
. . . . . .
0
0
0
0 ( J 1
0 0 Q )
( X )
- . I
c o
-
. . . . . . 0
N
f - ' > <
I ' V
( J 1 1 +
w
+ : >
0 " \
<
f - '
~
V l
I ' V
r n
- s
~ .
~ n
r t
+ : >
( J l
U 1 - P o
( J l
. . . . . .
+ : >
( J 1 - P o
+ : : >
( J 1
+ : >
+ : : > .
0
+ : >
c o
. . . . . . . I ' V
W
- . . J
s o
- . I
( J 1
- . . J
N : : 5
c o ( )
. . . . . .
< o
I ' V
I ' V e n
( J 1
' 0
- . I
' 0
W
W
0 " \
. "
~
N
. . . . . . . - . I
0 0
f - ' - . . . . J
0 0
0
c o
+ : : >
- - t
' "
- - t >
' " 1 +
1 +
1 +
1 +
1 +
1 +
1 +
1 +
1 +
1 +
n
0
0
0
*
0
0
0
0
0
0 0 0
.
: : 5
·
· . . . . . .
0
0
0
0
0
0
0
. . . . . ,
f - ' N
0 .
N
- . I
- . I
N
- . I c o
' 0 . . . . . .
0 0
. "
r t
C D
3
0
O J
- - ' : : 5
r o
- - t >
+ : : > .
+ : >
. j : : > - P o
+ : : >
+ : >
+ : : > - P o
+ : > + : >
+ : >
+ : : >
~
o
( J 1
N
Q " )
- . I
U 1
( J 1
( J 1
c o
0 " \
0 " \ c o
( ) r t
. . . . . .
' 0
f - '
Q " )
( J l + : : >
( J 1 c o
w
' 0
0 c o
' "
0
. "
W
N
+ : : >
. . . . . . . - . I
( X ) - . . J
c o
W
- . I N 0
l . / )
- s
t i l
1 +
1 +
1 +
1 +
1 +
1 +
1 +
1 +
1 + 1 +
- - t >
0
0
0
0 *
0
0
0 0
0
0
0
.
.
·
~
t o - '
f - '
. . . . . . . 0
0
0
0
0
N
0
N
o .
c o
' 0 W
l . D
c o
W
N
0 " \
C J
~
( \ )
~
( \ )
( J 1 . j : : >
( J 1 - P o + : >
+ : >
+ : > + : : >
+ : : >
~
·
( J 1 C ! l
w
\ D - . I 0 e x >
N
0 " \
( " " ' )
. . . . . .
Q " )
\ D
' "
W 0 + : >
N
U 1 0
( J 1
0
( X ) . "
( J 1 ~
N
N W
c o
l . D
- - t
<
C D
1 + 1 +
1 + 1 + 1 +
1 + 1 +
~
0
0 0
0 0
0 0
*
r t
·
= s -
f - ' 0
. . . . . . I ' V 0
+ : >
N
Q )
( J 1
f - '
I ' V - . . J C D
l . D ' 0
3 :
O J
: E :
- - '
= s -
r o
0
- - '
C D
( J 1
. j : : >
- P o - P o + : : > + : > - P o
- P o + : >
t i l
·
( J 1
. . . . . . \ D
0 ' \ - . I ~
e x > N
r t
. j : : >
f - ' ( " " ' )
N - . . . . J 0 ~
W
c o ' 0 C
' "
Q )
( J 1 Q " )
U 1 - . I N
+ : >
N . . . . . .
. "
0 .
l . / )
1 + 1 +
1 + 1 + 1 +
1 + 1 +
~
0 0 0
0
0
0 0
C D
.
~
. . . . . .
f - '
W
- + : : > I ' V
0
0
( J 1
f - '
0 Q )
c o t - - "
- . . J
0
0 .
9 Z 1 7
: t >
. - .
: : s :
. - .
\ J
0
\ . 0
\ . 0
\ J
: : l
C P
C P
r n
M -
. - .
I ' )
z
z r
0
0
0
. . . . . .
0
I - '
0
0
0 - . . . . . .
0
0
0
0
- . . . . . J
0
C P
\ . 0
0
-
N
. - .
. j : : >
( J 1 N
W
e n
" " ' - J
- > <
0 0
-
I - '
N
( ' )
~
. j : : >
. j : : >
. j : : > U 1
. p .
~
- I : >
- I : >
( ' )
- . . . . . J
c o
" " ' - J
( J ) 0
O ' l ~
1 . 0
0 0
0
n
W
. j : : >
1 . 0
I - '
. p .
0
W
- I : >
W
: : l
- n
. - .
0
N
. - . U 1
C O
0 0
N
0
r l -
~
1 +
1 +
1 +
1 +
1 +
1 +
1 +
: : l
~
C
0
0
0
0
0
0
0
. C ' D
.
.
.
0 -
0
0
0
0
0
- . . . . . J
. p .
0
I - '
0 0
\ . 0
-
" ' T 1
-
c t >
-
3
P I
. . . . . .
c t >
. p .
. p . . j : : >
. j : : >
. p .
. j : : >
~
- I : >
- I : >
.
.
- U 1
. p . . j : : >
O ' l
W U 1
W
W
0 ' 1
n
e n
( J 1
" " ' - J 0
e n
- . . . . . J
0
C O
0 0
- n
\ . 0
1 . 0
N
0
- I : >
W
0 ' 1
V I
-
-
1 +
1 +
1 +
1 + 1 +
1 +
1 +
< l ' . ! 1 .
0
0
0
0
0
0
0
.
.
.
. . . . . .
C &
I - '
0
0
0
. . . . . .
- . . . . . J
W
0
W 0 0
0 0
-
L Z 1 7
) : : 0
n
Q
; J o .
: : r : : " " 0
0
~ .
~
#
< :
- I
: u
~
" " 0
3
s : »
s : »
I I I
I T l
*
*
*
C i l
- '
~
C ' " o -
: z
~ l 1 J
* ~
*
< : +
~ .
§
- - ' C
~ .
~
0
~ - S
\ \ )
*
( l )
~ .
( l )
~
~ .
. . . . . .
- ' .
~
- +
I I I C i l
~
~
~ X
~ . I I I
\ \ )
~
Q
N
~ .
~ 0
0
- ' . ~ .
~ .
w
C / ) ~
~ . . . . . .
~
~ Q
~
< : - - >
~
W
~
< ; - - >
~
~
0
~
~
c - ! "
- 0 < : +
" ' 0
" ' C
- 0 - + - t ,
; : ; - +
V
~
A
A
~
A
~
( I )
C / ) C i l
0
0
0
0
- S
.
.
.
r o
c - ! "
0
0
0
0
l 1 J
( ] " I
( J ' I
. . . . . .
0
Q
- S
x r
. . . . . .
.
r o
- '
I I I
( I )
W
. p -
. . . . . .
. . . . . . . . .
C / )
N I I I
.
- ' .
- S
- +
0
C O
< . T I
I - '
C O
0
( I ) ~
O " l
. . . . . . . . . W
< . T I
C O ~
- 0
0
W
W
0
< . T I
~
- S
~
n
( I )
~ .
I I I
*
*
0
*
* 0
I I I
*
*
* r o
( I )
* ~
*
- t ,
*
~ ~
- t ,
c - ! "
- ' .
n
c - ! "
- ' .
; : ;
~
I
( I )
r o
. . . . . .
. p -
O
0
~
' " : ; c - ! "
0 -
( J ' l
W
I - '
W
I I I
I
~ .
. p - N
C O
1 . 0
I I I
~ .
. . . . . . . .
I - '
. . . . . . . . . O " l
C / )
0
c - ! "
~
~
: : : s
l 1 J
- t >
~
I I I
I I I
e n
*
c - ! " ~ .
- ' . *
c - ! "
; : ;
*
I I I
( I )
c - ! "
; J : : .
- '
n
( I )
n
~
l ; J - '
N
I - '
< . T I
l 1 J
~ .
.
- '
( ] " I
C O
N
n
~
( ] " I
. . . . . . . . .
I - '
C
O
. p -
C O
0
- '
~
: : : s
l 1 J
< : - - I
e n
*
*
c - ! "
R
( I )
*
- +
*
0 -
; : ;
~
C / )
c - ! "
l 1 J
I I I
- S
r o
- 0
- '
( I )
~
l 1 J
- S
c - ! "
I
- ' .
. p -
0
: J : : o
§
0
- 0
~ . ~
< . T I
1 . 0
- 0
I I I
< : - - I
. . . . . . . . .
~
r o
; : ;
~ .
I - '
W
~
- ' .
~
: : : s
0 -
~
- ' . e n
*
I I I
R
X
*
- +
*
- t ,
~
O ' l
0
C O
- S
c
c - ! "
I I I
~
C ' : l
- ' . I
r o
~
~
c . c
I I I
R
. . . . . . . . .
- ' .
~
I - '
0 -
; : s
X
l 1 J
< . T I
~
c - ! "
I I I
~
l 1 J
*
C O
*
r o
I I I
*
n
C
- ' .
' < : J
3
r o
3
I I I
l 1 J
~
- s
~ I I I
- ' .
c - ! "
I I I
- +
C
~ .
r o
\ . )
0 -
0 -
- ' .
~
- ' . < : - - I r o
R 0 -
~
- +
R
8 Z v