Effects of seasonality and migratory prey on body

condition of Cichla species in a tropical oodplain

river
Introduction
Migratory shes are an important component of
aquatic food webs worldwide. Fish migration links
food webs across landscapes by transporting produc-
tion among seemingly disparate systems and providing
a resource subsidy to resident consumers (Polis et al.
1997, 2004). For example, marine derived nutrients
are transported by anadromous shes, such as salmon
and alewife, into food webs of riverine (Garman &
Macko 1998; MacAvoy et al. 2000; Chaloner et al.
2002) and terrestrial systems (Hilderbrand et al. 1999;
Szepanski et al. 1999; Darimont & Reimchen 2002)
through predation, gamete release, excretion or
decomposition (especially in the case of salmon). In
a similar manner, freshwater shes migrating from
nutrient-rich oodplains into oligotrophic rivers may
transport production among systems and represent a
substantial resource subsidy for resident piscivores
(Winemiller & Jepsen 1998, 2004).
In tropical river systems, massive reproductive
migrations crossing hundreds of kilometres are driven
by wet-dry seasonality, with reproduction of migratory
shes usually occurring during the high-water period
in productive oodplains or ooded forests (Goulding
1980; Lowe-McConnell 1987). Among the most
conspicuous of the Neotropical long-distance migrants
are the prochilodontid shes of the genera Prochilodus
and Semaprochilodus. These large characiform shes
feed on detritus, algae and associated microorganisms
(Bowen 1983; Bowen et al. 1984) and are the primary
species of many commercial and subsistence sheries
in South America (Welcomme 1990). Although spe-
cics of prochilodontid spawning migrations may vary
Ecology of Freshwater Fish 2006: 15: 398407
Printed in Singapore All rights reserved
2006 The Authors
Journal compilation 2006 Blackwell Munksgaard
ECOLOGY OF
FRESHWATER FISH
398 doi: 10.1111/j.1600-0633.2006.00152.x
Hoeinghaus DJ, Winemiller KO, Layman CA, Arrington DA, Jepsen DB.
Effects of seasonality and migratory prey on body condition of Cichla
species in a tropical oodplain river.
Ecology of Freshwater Fish 2006: 15: 398407. 2006 The Authors
Journal compilation 2006 Blackwell Munksgaard
Abstract In this study we examine trends in body condition of three
Cichla species native to the Cinaruco River, Venezuela. Specically, we
characterise the potential subsidy provided by migratory Semaprochilodus
kneri to Cichla by comparing condition among species, hydrologic seasons
and different size-classes (based on the ability to consume S. kneri
assuming gape-limitation). All three species had lowest body condition
during the rising-water season (following the reproduction period). Only
Cichla temensis had greater body condition during the falling-water season
than low-water season. During the falling-water season, C. temensis
>300 mm SL had signicantly greater condition than smaller conspecics.
Annual migration of S. kneri into the river during the falling-water season
provides a resource subsidy to large C. temensis. Transfer of whitewater
production into nutrient-poor rivers by migratory shes appears to be a
general phenomenon in the major Neotropical river basins and may
partially account for high standing stocks of large piscivores in these
unproductive ecosystems.
D.J. Hoeinghaus
1,2,
*,
K.O. Winemiller
1
, C.A. Layman
3
,
D.A. Arrington
4
, D.B. Jepsen
5
1
Section of Ecology, Evolutionary Biology and
Systematics, Department of Wildlife and Fisheries
Sciences, Texas A&M University, College Station,
Texas, USA,
2
Kansas State University, Division of
Biology, Manhattan, KS, USA,
3
Marine Biology
Program, Department of Biological Sciences,
Florida International University, North Miami, FL,
USA,
4
Loxahatchee River District, Jupiter, FL,
USA,
5
Oregon Department of Fish and Wildlife,
Corvallis Research Lab, Corvallis, OR, USA
Key words: gape limitation; ribazon; predation;
food web subsidy; Semaprochilodus kneri
*Present address: D. J. Hoeinghaus, Kansas
State University, Division of Biology, 104 Ackert
Hall, Manhattan, KS 66506, USA; e-mail:
[email protected]
Accepted for publication January 31, 2006
by region [e.g. Central Amazon Basin, Brazil
Ribeiro & Petrere (1990); Vazzoler & Amadio
(1990); Upper Parana Basin, Brazil Agostinho et al.
(1993, 2003; Orinoco Basin, Venezuela Lilyestrom
(1983); Barbarino-Duque et al. (1998)], in general,
reproduction and feeding occur in productive ood-
plains during the high-water period, and young-of-the-
year form dense schools with adults to disperse
throughout the river basin as water levels fall.
In the Cinaruco River of western Venezuela,
Semaprochilodus kneri (Pellegrin) migrate en mass
into the system from the Orinoco oodplain during
the falling-water period and provide a seasonally
abundant prey resource to large piscivores. The
consumption of migratory S. kneri may provide an
important resource subsidy that allows large pisci-
vores in this system to achieve greater condition,
fecundity and population densities than based on
in situ productivity alone (Winemiller & Jepsen
1998, 2004). In this study we analyse a database of
weight and standard length (SL) data collected over
4 years to examine trends in body condition of three
Cichla species, abundant piscivores native to the
Amazon, Essequibo and Orinoco basins. Specically,
we characterise the potential subsidy provided by
S. kneri to Cichla by comparing body condition
among species, during different hydrologic seasons,
and in relation to gape limitations of different Cichla
size classes and the ability to consume S. kneri.
Materials and methods
Regional description
The llanos is an extensive grassland savanna in
central and western Venezuela characterised by
strong annual wet and dry seasonality (see Mach-
ado-Allison 1990; Winemiller et al. 1996; Sarmiento
& Pinillos 2001). Whitewater rivers draining the
Andes, such as the Guanare and Portuguesa Rivers,
deposit nutrient-rich alluvial sediments in oodplains
of the Orinoco River during ood periods when
broad expanses of the lowland llanos are inundated
(Hoeinghaus et al. 2004). During the rainy season,
many sh species migrate into nutrient-rich lowland
rivers and ooded terrain where they spawn (Mach-
ado-Allison 1990; Winemiller 1996). During the
falling-water period, large schools dominated by
Prochilodus and Semaprochilodus, locally called
ribazo nes, ascend rivers of western Venezuela from
the Orinoco River oodplain and Apure River delta
(Barbarino-Duque et al. 1998; Layman & Winemiller
2005). School density gradually decreases as indivi-
duals remain in upstream habitat, are caught by
shermen (Rodriguez et al. in press), or are con-
sumed by predators.
Study area
The Cinaruco River is a low-gradient oodplain
tributary of the Orinoco River that originates and
drains the south-western llanos near the Colombian
border (632N, 6724W). Rivers such as the Cina-
ruco, that originate in the llanos rather than the Andes
Mountains, have low nutrient and suspended sediment
loads because of the dystrophic nature of llanos soils
(Montoya et al. 2006). Seasonal water level uctua-
tions exceed 5 m, with peak inundation occurring in
August. During the high-water period, thousands of
square kilometres of surrounding savanna and gallery
forest are ooded. Width of the main channel ranges
from 50 to 200 m during the lowest point in the
hydrologic cycle. Over 280 sh species have been
collected from the Cinaruco River, with taxa repre-
senting a wide range of ecological attributes and life
history strategies (Arrington & Winemiller 2003;
Hoeinghaus et al. 2003b; Arrington et al. 2005; Lay-
man & Winemiller 2005; Layman et al. 2005c;
Winemiller et al. 2006).
Three species of Cichla, locally called pavo n,
naturally occur in the Cinaruco River: C. temensis
(Humboldt), C. orinocensis (Humboldt), and C. inter-
media (Machado). Cichla intermedia is strongly
associated with structured habitats in the main channel
in or near fast currents and C. orinocensis is most
common in shallow areas of lagoons near submerged
branches or rocks (Jepsen et al. 1997). Cichla temensis
occurs in a broad range of habitats in both lagoons and
the main channel (Jepsen et al. 1997). All three
species construct nests and spawn during the late
low-water to early rising-water period (Winemiller
et al. 1997; Jepsen et al. 1999) and cease feeding
while brood guarding (Devick 1972; Zaret 1980;
Jepsen et al. 1997). Pavo n are voracious piscivores,
capable of causing community level shifts in prey
species abundances and distributions (Zaret & Paine
1973; Jepsen et al. 1997; Layman & Winemiller 2004;
Layman et al. 2005a).
Analyses
A database of weight and SL data for the three species
of Cichla (C. temensis N 650, C. orinocensis
N 675, C. intermedia N 73; Table 1) naturally
occurring in the Cinaruco River, Venezuela was
compiled from a mark-recapture study conducted
between January 1999 and May 2003 (Hoeinghaus
et al. 2003a). Fish were collected during all seasons
from all representative habitat types (e.g. sand banks,
lagoons, creeks and the main channel) in an effort to
collect representative samples of the populations. All
individuals were collected by hook and line, identied,
weighed (g), measured (mm SL) and released at the
Variation in Cichla body condition
399
location of capture. Although individuals were collec-
ted during all seasons, sample sizes are substantially
lower for all three species during the rising-water
period (May to August) because Cichla cease feeding
while brood-guarding and high water constraints limit
shing effort (Table 1). For this reason, statistical
analyses primarily compare condition during periods
of falling (September to December) and low water
(January to April). Comparisons are further limited for
C. intermedia because this species is less abundant
than its congeners.
Individuals were separated into two size-classes
based on the ability to consume S. kneri determined
from morphological data assuming gape limitation.
Maximum gape (mouth aperture in vertical dimension
with jaws fully protruded) and SL was measured (to
nearest mm) for a subset of Cichla specimens
(N 61) to generate gapeSL relationships for each
species. Semaprochilodus kneri SL and maximum
body depth (vertical distance from the anterior inser-
tion of the dorsal n to the ventrum) were measured
(to nearest mm) for young-of-the-year individuals and
post-spawn adults (age 1+) migrating into the Cina-
ruco River during the falling-water period of 2003.
Additional individuals were measured during the low-
water period of 2004. Semaprochilodus kneri were
collected by castnet and released following measure-
ment. To evaluate the accuracy of gape limitation
estimates based on morphological measures, stomach
contents of Cichla were examined using the gastric
lavage method (see Layman et al. 2005b) and SLs for
all S. kneri occurring in stomachs were recorded along
with the SL of the pavo n in which it was found.
It is important to note that other species in the river
undergo seasonal migrations similar to S. kneri, albeit
in much lower densities. Specically, we have
previously demonstrated that Leporinus spp. and
Myleus spp. are also part of the ribazo n on the
Cinaruco River (Layman & Winemiller 2005).
However, in more than 2400 individual Cichla
stomachs we have examined since 1994 (C. temensis
N 1,365; C. orinocensis N 755; C. intermedia
N 307), we have never found a member of these
two genera in the stomach contents. Likewise, other
sh species we believe may undergo seasonal
migrations (e.g. Piractus brachypumum, Mylossoma
aureus, Pseudoplatystoma fasciatum) have never
been identied in Cichla stomachs. Semaprochilodus
kneri, however, makes up approximately 50% of C.
temensis diet during the falling-water period (Jepsen
et al. 1997; Winemiller & Jepsen 1998), strongly
suggesting C. temensis preferentially utilises this prey
species and not others that undertake seasonal
migrations. As such, we base our analysis in this
paper on S. kneri only.
Cichla body condition was evaluated using the
relationship between weight and SL, W aL
b
, at
three levels: taxonomic (among species), seasonal
(among seasons within species) and size class
(between size-classes within species and seasons).
Weight and SL data were log-transformed to yield a
linear relationship [log
10
(W) a + b(log
10
L)] in
which b is the slope (allometric coefcient) and a
is the intercept. Analysis of covariance (ancova) was
used to test for differences among slopes and
intercepts (when slopes did not differ) of the log-
transformed data for the three levels of comparison.
For all tests, log
10
W was the dependent variable and
log
10
L the covariate. Depending on the level of
comparison, the independent variable was species,
season or size-class. For signicant main effects,
pairwise comparisons of intercepts were performed,
with the experiment-wide error rate maintained at
a 0.05 using Bonferroni correction. Although other
indices, such as Fultons K, are commonly used to
assess sh condition, the calculation of least-squares
regression parameters for a population or subgroup is
a more accurate method of comparing lengthweight
relationships (Cone 1989).
Results
Taxonomic and seasonal comparisons
Signicant differences in body condition were
observed at the taxonomic level. Intercepts of the
log-transformed weightSL relationships were signi-
cantly different among all species (Table 2). For
individuals of the same SL, C. orinocensis weighed
slightly more than its congeners and C. temensis
weighed the least (Fig. 1).
Table 1. Number of individuals available for analysis by species, season and
size-class.
Species N Season N Size-class N
C. temensis 650 Falling 172 <300 46
>300 126
Low 472 <300 116
>300 356
Rising 6 <300 2
>300 4
C. orinocensis 675 Falling 78 <300 48
>300 30
Low 590 <300 202
>300 388
Rising 7 <300 6
>300 1
C. intermedia 73 Falling 10 <300 2
>300 8
Low 61 <300 13
>300 48
Rising 2 <300 0
>300 2
Hoeinghaus et al.
400
Signicant seasonal differences in body condition
were observed for C. temensis and C. intermedia. Only
C. temensis had signicantly higher condition during
the falling water period (Table 2; Fig. 2b), whereas
body condition of both species was signicantly lower
during the rising water period (Table 2; Fig. 2b,d). No
difference was observed in condition of C. intermedia
between falling and low-water periods (comparisons
were not made among the rising-water period because
of limited sample size; Fig. 2f).
Size-class comparisons
Relationships between maximum gape and SL did not
differ among Cichla species (ancova, F 0.835,
d.f. 2, P 0.439), so the combined relationship
(y 0.151x ) 5.999, R
2
0.924) was used to esti-
mate mouth gape for all individuals using measures of
SL attained at the time of capture. Maximum body
depth of S. kneri increased with SL according to the
relationship y 0.374x + 0.351, R
2
0.97. Based on
the above morphological relationships and the size of
young-of-the-year S. kneri captured migrating into the
river during the falling water period of 2003, Cichla
were separated into two size-classes: <300 and
>300 mm SL, with 300 mm representing the threshold
at which Cichla can consume S. kneri assuming gape
limitation.
The 300-mm threshold dened by morphological
measurements conforms to dietary data from stomach
contents analyses. The smallest Cichla from the
Cinaruco River with a S. kneri recorded in its stomach
was a C. temensis with 335 mm SL, captured
November 1999 (number of stomachs exam-
ined 1365). Stomach contents analyses illustrate
that C. temensis preferentially consume young-of-the-
year S. kneri even when their gape would allow them
to consume larger individuals (Fig. 3). No S. kneri
were identied from stomach contents of C. orino-
censis (number of stomachs 755) or C. intermedia
(number of stomachs 307), although both species
attain sizes capable of consuming S. kneri. For
Table 2. ancova results of body condition comparisons at the three levels
of comparison: among species, among seasons within species and between
size classes (<300 and >300 mm SL) within species and seasons.
Species Factor d.f. F-value P-value
All Species LogSL 2 0.445 0.641
Species 2 141.185 <0.001; O > I > T
C. temensis Season LogSL 2 0.601 0.549
Season 2 19.014 <0.001; F > L > R
Size-class
Falling 1 5.442 0.021
Low 1 1.781 0.183
C. orinocensis Season LogSL 2 0.187 0.830
Season 2 5.975 0.003; F L > R
Size-class
Falling 1 1.487 0.227
Low 1 0.819 0.366
C. intermedia Season LogSL 1 0.194 0.661
Season 1 0.123 0.727
Size-class
Low 1 0.674 0.415
For signicant main effects (bold values), pairwise comparisons of intercepts
are provided in the adjacent column with underlined pairs not signicantly
different (O, C. orinocensis; I, C. intermedia; T, C. temensis; F, falling; L, low;
R, rising). Not all possible size-class or seasonal comparisons were
performed because of insufcient sample sizes for some size-classes and
seasons (Table 1). For C. intermedia, seasonal comparisons are for the
falling and low-water periods only because of limited sample size.
(a) (b)
Fig. 1. Untransformed (a) and log-transformed (b) weight-standard length relationships for Cinaruco Cichla. Regression intercepts are
signicantly different, with C. orinocensis > C. intermedia > C. temensis (see Table 2).
Variation in Cichla body condition
401
Fig. 2. Untransformed and log-transformed weight-standard length relationships among seasons for each species. Seasons: falling,
n low, s rising. See Table 2 for ancova results.
Hoeinghaus et al.
402
example, this dataset contains 419 C. orinocensis and
58 C. intermedia that were >300 mm SL, more than
half of the individuals measured for each species
(Table 1).
A single signicant difference was observed in body
condition among size-classes for all intra-seasonal
species comparisons. Large C. temensis (>300 mm
SL) had signicantly higher condition than smaller
individuals during the falling-water period, as evi-
denced by a signicantly higher regression slope
(Table 2; Fig. 4).
Discussion
We observed signicant species, seasonal and size-
based differences in body condition of naturally
occurring Cichla species of the Cinaruco River,
Venezuela. Relationships between weight and SL
likely differed among species because of differences
in allocation of energy for growth. Based on growth
rate estimates, C. temensis are longer and heavier than
its congeners for individuals of the same age (Jepsen
et al. 1999). Cichla temensis therefore have lower
condition values because they increase body length at
a faster rate than the other species, i.e. more biomass is
allocated to increasing length rather than girth (Jepsen
et al. 1999). General trends in Cichla body condition
among seasons, on the other hand, are because of
seasonal variation in prey densities, as well as seasonal
reproduction by all three species.
Tropical shes reveal diverse morphological and
physiological adaptations to wetdry seasonality, most
notably reproductive seasons that coincide with rising
water levels that allow juveniles to feed in resource-
rich oodplains (Goulding 1980; Lowe-McConnell
1987). During oodwater recession, shes are forced
from the shallow refuges of the oodplain and prey
densities increase in shrinking aquatic habitats. Large-
bodied predators in tropical oodplain systems gener-
ally exhibit increased foraging activity during the
falling and low-water periods as prey are forced into
deeper creeks, lagoons and the main river channel
(Jepsen et al. 1997; Arrington et al. 2002; Hoeinghaus
et al. 2003b; Arrington et al. 2006).
Reproductive behaviour of Cichla species (nesting
and brood-guarding) and low densities of optimally
sized prey (e.g. depleted populations of S. kneri)
combine to create a physiological winter during the
end of the dry season and initial stages of the rising-
water period (when remaining prey become widely
dispersed), as evidenced by annual otolith bands
(Lowe-McConnell 1964; Jepsen et al. 1999). We
observed signicantly lower condition of both
C. temensis and C. orinocensis during the rising-water
period, but do not have sufcient data to fully
investigate seasonal trends in condition for C. inter-
media. Jepsen et al. (1999) observed otolith bands for
all three Cichla species of the Cinaruco River,
revealing that growth of all three species is lowest
following the reproductive and brood-guarding period.
Cichla temensis was the only species with signi-
cantly greater body condition during the falling-water
period than the low-water and rising-water periods.
Comparison of C. temensis body condition among
seasons, while excluding individuals >300 mm SL
during the falling water period, yields the same trend
observed for C. orinocensis and suggested for C. inter-
media: condition did not differ signicantly among the
falling and low-water periods, but was signicantly
lower during the rising-water period. Cichla temensis
large enough to consume S. kneri had signicantly
higher condition than smaller individuals during the
falling-water period. Although all three Cichla species
in the Cinaruco exhibit seasonal variations in diet
Fig. 3. Relationship between (a) Semaprochilodus kneri body
depth and Cichla temensis maximum gape based on stomach
contents analysis and (b) standard length and maximum body depth
for S. kneri across the range of size classes that occur in the
Cinaruco River throughout the hydrologic cycle. The solid line
denotes mouth gapebody depth ratio 1 and dashed lines bound
the size of S. kneri found in stomach contents.
Variation in Cichla body condition
403
(Jepsen et al. 1997), large C. temensis appear to gain a
growth advantage over its congeners and smaller
conspecics by exploiting the annual arrival of large
schools of S. kneri.
Results from dietary and stable isotope analyses
support the conclusion that greater body condition of
large C. temensis during the falling-water period is
because of the consumption of young-of-the-year
S. kneri migrating into the system. During the falling
water period, young-of-the-year S. kneri comprises
approximately 50% of the prey volume in stomach
contents of large (>40 cm SL) C. temensis (Jepsen
et al. 1997; Winemiller & Jepsen 1998). Stable carbon
isotope results directly tie the observed growth
advantage of large C. temensis during the falling
water period to assimilation of carbon ultimately
derived from the Orinoco oodplain, i.e. through
consumption of S. kneri (Winemiller & Jepsen 2004).
While other species may move into the Cinaruco from
the Orinoco oodplain, diet data suggests that S. kneri
are by far the most important prey species (Layman
et al. 2005c).
Lipid concentration of dorsal muscle tissue in
C. temensis is greatest during the falling-water
period (Arrington et al. 2006). Energy stored in the
form of lipids can be readily used for reproduction
in the subsequent low water period. The previous
studies mentioned above, combined with the body
condition results presented here, indicate that young-
of-the-year S. kneri provide large C. temensis with a
substantial resource subsidy prior to the spawning
season. This additional nutrition should permit them
to attain higher fecundities and population densities
than would be possible based on in situ production
alone.
Movement patterns and behavioural observations
also suggest that young S. kneri are especially
important for large C. temensis, whereas habitat
associations or behaviours of C. orinocensis and
C. intermedia may prevent or limit consumption of
S. kneri. Hoeinghaus et al. (2003a) observed short-
distance movement (500 m) of C. orinocensis,
C. intermedia and small C. temensis in the Cinaruco
River throughout the hydrologic cycle, however large
C. temensis moved into the main river channel and
exhibited longer downstream movements (up to
21 km) associated with the falling-water period.
Observations of aggregations of large C. temensis
attacking migrating schools of young-of-the-year
S. kneri within lower reaches of the Cinaruco River
during the early falling water period (October to
December) suggest that some C. temensis move
downstream prior to the upstream migration of
S. kneri in order to maximise their opportunity to
consume this seasonally abundant food resource. In
contrast, C. orinocensis is widely and evenly distri-
buted within ooded gallery forest during the falling-
water period and C. intermedia is strongly associated
with structured habitats with high current velocity.
Semaprochilodus kneri are most abundant in relat-
ively open low-velocity areas of the main river
channel and connected lagoons, especially sandy
beaches where they graze algae and detritus. Thus,
large C. orinocensis and C. intermedia may not
consume S. kneri simply because they do not occur
in the same habitats.
(a) (b)
Fig. 4. Untransformed (a) and log-transformed (b) weightstandard length relationships among size classes (<300 and >300 mm SL) for
Cichla temensis during the falling water period. Slopes are signicantly different (P 0.021; Table 2).
Hoeinghaus et al.
404
By consuming migratory S. kneri, C. temensis,
which only occur in clear or blackwater rivers,
assimilate carbon originating from the vast and
productive oodplains of the whitewater Orinoco
River (Winemiller & Jepsen 2004). Consumption of
S. kneri also increases ecological efciency as large
piscivores, such as Cichla, are able to exploit prey of
nearly optimal size (Winemiller & Taylor 1987) that
also occupy low positions in the food web (Winemiller
& Jepsen 1998; Layman et al. 2005b). The importa-
tion of production from whitewater oodplains,
together with the increased ecological efciency
associated with short food chains, could account for
the impressive secondary productivity of relatively
nutrient-poor tropical rivers such as the Cinaruco.
The transport of primary production from nutrient-
rich whitewater rivers into oligotrophic clearwater and
blackwater ecosystems by migratory shes appears to
be a general phenomenon within the major Neotropi-
cal river basins. In the Amazon Basin for example,
migratory species of the genera Mylossoma, Colosso-
ma, Anodus, Triportheus, Semaprochilodus, Prochil-
dous and Brycon (which all feed at relatively low
positions in the food web) occur in multiple river types
as adults, but only reproduce in whitewater systems
during the high-water period (Carvalho de Lima &
Araujo-Lima 2004). Whitewater oodplains function
as source habitats, and upstream migrations that
distribute these species within the basin also serve to
transport whitewater production into oligotrophic
systems. Disruption of seasonal migrations by river
impoundment or overshing, already widespread in
the Orinoco Basin (Winemiller et al. 1996; Barbarino-
Duque et al. 1998), is likely to have profound effects
on ecosystem processes of the Cinaruco and other
oligotrophic tropical rivers.
Acknowledgements
Edgar Pelaez, Jim Marzoula, Carol Marzoula and Glenn Webb
provided logistical support through the Cinaruco River Fishing
Club and Tour Apure. Don Taphorn at UNELLEZ provided
logistical support and helped obtain sampling permits. Fish
collections were made with shing permits 0439 and 0192
issued by the Servicio Autonomo de los Recursos Pesqueros y
Acuicolas of Venezuela. We acknowledge reviewer sugges-
tions, which improved the manuscript. This study was funded
in part by a Texas Water Resources Institute Mills Scholarship
and Research Scholarships from the Society of Wetland
Scientists and TAMU Ofce of Graduate Studies (DJH), a US
Environmental Protection Agency STAR Graduate Fellowship
and Texas A & M Regents Fellowship (CAL), Tom Slick and
William Bill Baab Conservation Fellowships (CAL and
DAA), The L.T. Jordan Institute (DAA), The International
Sportsh Fund (KOW, CAL and DAA), and National Science
Foundation grants DEB-0411978 (KOW and CAL) and DEB-
0089834 (KOW). Facilities were made available to DJH at the
Research Nucleus in Limnology, Ichthyology and Aquaculture
(Nupelia) at the State University of Maringa (UEM), Brazil,
during manuscript preparation.
References
Agostinho, A.A., Vazzoler, A.E.A.d.M., Gomes, L.C. & Okada,
E.K. 1993. Estraticacio n espacial y comportamiento de
Prochilodus scrofa en distintas fases del ciclo de vida, en las
planicie de inundacio n del alto r o Parana y embalse de
Itaipu, Parana, Brasil. Revista de Hydrobiologia Tropical 26:
7990.
Agostinho, A.A., Gomes, L.C., Suzuki, H.I. & Ju lio, H.F. Jr
2003. Migratory shes of the upper Parana River basin,
Brazil. In: Carolsfeld, J., Harvey, J., Ross, C. & Baer, A., eds.
Migratory shes of South America: biology, sheries and
conservation status. Victoria, Canada: International Develop-
ment Research Centre and The World Bank, pp. 1998.
Arrington, D.A. & Winemiller, K.O. 2003. Diel changeover in
sandbank sh assemblages in a neotropical oodplain river.
Journal of Fish Biology 63: 118.
Arrington, D.A., Winemiller, K.O., Loftus, W.F. & Akin, S.
2002. How often do shes run on empty? Ecology 83:
21452151.
Arrington, D.A., Winemiller, K.O. & Layman, C.A. 2005.
Community assembly at the patch scale in a species rich
tropical river. Oecologia 144: 157167.
Arrington, D.A., Davidson, B.K., Winemiller, K.O. & Layman,
C.A. 2006. Inuence of life history and seasonal hydrology
on lipid storage in three Neotropical sh species. Journal of
Fish Biology 68: 13471361.
Barbarino-Duque, A., Taphorn, D.C. & Winemiller, K.O. 1998.
Ecology of the coporo, Prochilodus mariae (Characiformes,
Prochilodontidae), and status of annual migrations in western
Venezuela. Environmental Biology of Fishes 53: 3346.
Bowen, S.H. 1983. Detritivory in neotropical sh communities.
Environmental Biology of Fishes 9: 137144.
Bowen, S.H., Bonetto, A.A. & Ahlgren, M.O. 1984. Micro-
organisms and detritus in the diet of a typical neotropical
riverine detritivore, Prochilodus platensis (Pisces: Prochilo-
dontidae). Limnology and Oceanography 29: 11201122.
Carvalho de Lima, A

. & Araujo-Lima, C.A.R.M. 2004. The

distributions of larval and juvenile shes in Amazonian
rivers of different nutrient status. Freshwater Biology 49:
787800.
Chaloner, D.T., Martin, K.M., Wipi, M.S., Ostrom, P.H. &
Lamberti, G.A. 2002. Marine carbon and nitrogen in
southeastern Alaska stream food webs: evidence from
articial and natural streams. Canadian Journal of Fisheries
and Aquatic Sciences 59: 12571265.
Cone, R.S. 1989. The need to reconsider the use of condition
indices in shery science. Transactions of the American
Fisheries Society 118: 510514.
Darimont, C.T. & Reimchen, T.E. 2002. Intra-hair stable
isotope analysis implies seasonal shift to salmon in gray wolf
diet. Canadian Journal of Zoology 80: 16381642.
Devick, W.S. 1972. Life history study of the tucunare, Cichla
ocellaris. Federal Aid in Sportsh Restoration Project F-9-1,
Job Completion Report. Honolulu: Hawaii Department of
Land and Natural Resources.
Variation in Cichla body condition
405
Garman, G.C. & Macko, S.A. 1998. Contribution of marine-
derived organic matter to an Atlantic coast, freshwater, tidal
stream by anadromous clupeid shes. Journal of the North
American Benthological Society 17: 277285.
Goulding, M. 1980. The shes and the forest. Berkeley, CA:
University of California Press.
Hilderbrand, G.V., Hanley, T.A., Robbins, C.T. & Schwartz,
C.C. 1999. Role of brown bears (Ursus arctos) in the ow of
marine nitrogen into a terrestrial ecosystem. Oecologia 121:
546550.
Hoeinghaus, D.J., Layman, C.A., Arrington, D.A. & Wine-
miller, K.O. 2003a. Movement of Cichla species (Cichlidae)
in a Venezuelan oodplain river. Neotropical Ichthyology 1:
121126.
Hoeinghaus, D.J., Layman, C.A., Arrington, D.A. & Wine-
miller, K.O. 2003b. Spatiotemporal variation in sh assem-
blage structure in tropical oodplain creeks. Environmental
Biology of Fishes 67: 379387.
Hoeinghaus, D.J., Winemiller, K.O. & Taphorn, D.C. 2004.
Compositional change in sh assemblages along the Andean
piedmont Llanos oodplain gradient of the r o Portuguesa,
Venezuela. Neotropical Ichthyology 2: 8592.
Jepsen, D.B., Winemiller, K.O. & Taphorn, D.C. 1997.
Temporal patterns of resource partitioning among Cichla
species in a Venezuelan blackwater river. Journal of Fish
Biology 51: 10851108.
Jepsen, D.B., Winemiller, K.O., Taphorn, D.C. & Rodriguez-
Olarte, D. 1999. Age structure and growth of peacock
cichlids from rivers and reservoirs of Venezuela. Journal of
Fish Biology 55: 433450.
Layman, C.A. & Winemiller, K.O. 2004. Size-based prey
response to piscivore exclusion in a Neotropical river.
Ecology 85: 13111320.
Layman, C.A. & Winemiller, K.O. 2005. Patterns of habitat
segregation among large shes in a Venezuelan oodplain
river. Neotropical Ichthyology 3: 103109.
Layman, C.A., Langerhans, R.B. & Winemiller, K.O. 2005a.
Body size, not other morphological traits, characterizes
cascading effects in sh assemblage composition following
commercial netting. Canadian Journal of Fisheries and
Aquatic Sciences 62: 28022810.
Layman, C.A., Winemiller, K.O., Arrington, D.A. & Jepsen,
D.B. 2005b. Body size and trophic position in a diverse
tropical food web. Ecology 86: 25302535.
Layman, C.A., Winemiller, K.O. & Arrington, D.A. 2005c.
Describing the structure and function of a Neotropical river
food web using stable isotopes, stomach contents, and
functional experiments. In: de Ruiter, P.C., Wolters, V. &
Moore, J.C., eds. Dynamic food webs: multispecies assem-
blages, ecosystem development, and environmental change.
Amsterdam: Elsevier, pp. 395406.
Lilyestrom, C.G. 1983. Aspectos de la biolog a del coporo
(Prochilodus mariae). Revista UNELLEZ de Ciencia y
Tecnolog a 1: 511.
Lowe-McConnell, R.H. 1964. The shes of the Rupununi
savanna district of British Guiana, Pt. 1. Groupings of sh
species and effects of the seasonal cycles on the sh.
Zoological Journal of the Linnaean Society 45: 103144.
Lowe-McConnell, R.H. 1987. Ecological studies in tropical
sh communities. Cambridge, UK: Cambridge University
Press.
MacAvoy, S.E., Macko, S.A., McIninch, S.P. & Garman, G.C.
2000. Marine nutrient contributions to freshwater apex
predators. Oecologia 122: 568573.
Machado-Allison, A. 1990. Ecology of shes of the oodplain
areas of the Venezuelan Llanos. Interciencia 15: 411423.
Montoya, J.V., Roelke, D.L., Winemiller, K.O., Cotner, J.B. &
Snider, J.A. 2006. Effects of hydrological seasonality and
lateral connectivity on environmental variables and algal
biomass of a tropical oodplain river. Journal of the North
American Benthological Society 25: 157170.
Polis, G.A., Anderson, W.B. & Holt, R.D. 1997. Toward an
integration of landscape and food web ecology: the dynamics
of spatially subsidized food webs. Annual Review of Ecology
and Systematics 28: 289316.
Polis, G.A., Power, M.E. & Huxel, G.R., editors. 2004. Food
webs at the landscape level. Chicago: University of Chicago
Press.
Ribeiro, M.C.L.d.B. & Petrere, M. Jr 1990. Fisheries ecology
and management of the jaraqui (Semaprochilodus taeniurus,
S. insignis) in Central Amazonia. Regulated Rivers Research
and Management 5: 195215.
Rodriguez, M.A., Winemiller, K.O., Lewis, W.M. & Taphorn,
D.C. in press. The freshwater habitats, shes, and sheries of
the Orinoco River Basin. In: Minns, C.K. & Munawar, M.,
eds. Freshwater shes, their biodiversity, sheries and
habitats: health and prospects. Ecovision World Monograph
Series.
Sarmiento, G. & Pinillos, M. 2001. Patterns and processes in a
seasonally ooded tropical plain: the Apure Llanos, Vene-
zuela. Journal of Biogeography 28: 985996.
Szepanski, M.M., Ben-David, M. & Van Ballenberghe, V. 1999.
Assessment of anadromous salmon resources in the diet of
the Alexander Archipelago wolf using stable isotope analysis.
Oecologia 120: 327335.
Vazzoler, A.E.A.d.M. & Amadio, S.A. 1990. Aspectos
biolo gicos de peixes Amazo nicos. XIII. Estrutura e
comportamento de cardumes multiespecif cos de Semapro-
chilodus (Characiformes, Prochilodontidae) no baixo rio
Negro, Amazonas, Brasil. Revista Brasiliera Biologica 50:
537546.
Welcomme, R.L. 1990. Status of sheries in South American
rivers. Interciencia 15: 337345.
Winemiller, K.O. 1996. Dynamic diversity in sh assemblages
of tropical rivers. In: Cody, M.L. & Smallwood, J.A., eds.
Long-term studies of vertebrate communities. Orlando:
Academic Press, pp. 99132.
Winemiller, K.O. & Jepsen, D.B. 1998. Effects of seasonality
and sh movement on tropical river food webs. Journal of
Fish Biology 53: 267296.
Winemiller, K.O. & Jepsen, D.B. 2004. Migratory neotropical
sh subsidize food webs of oligotrophic blackwater rivers. In:
Polis, G.A., Power, M.E. & Huxel, G.R., eds. Food webs at
the landscape level. Chicago: University of Chicago Press,
pp. 115132.
Winemiller, K.O. & Taylor, D.H. 1987. Predatory behavior and
competition between smallmouth and largemouth bass.
American Midland Naturalist 117: 148166.
Winemiller, K.O., Marrero, C. & Taphorn, D.C. 1996.
Perturbaciones causadas por el hombre a las poblaciones de
peces de los llanos y del piedemonte Andino de Venezuela.
Biollania 12: 1348.
Hoeinghaus et al.
406
Winemiller, K.O., Taphorn, D.C. & Barbarino-Duque, A. 1997.
Ecology of Cichla (Cichlidae) in two blackwater rivers of
southern Venezuela. Copeia 1997: 690696.
Winemiller, K.O., Montoya, J.V., Layman, C.A., Roelke, D.L.
& Cotner, J.B. 2006. Experimental demonstration of seasonal
sh effects on benthic ecology of a Neotropical oodplain
river. Journal of the North American Benthological Society
25: 250262.
Zaret, T.M. 1980. Life-history and growth relationships of
Cichla ocellaris, a predatory South American cichlid.
Biotropica 12: 144157.
Zaret, T.M. & Paine, R.T. 1973. Species introduction in a
tropical lake. Science 182: 449455.
Variation in Cichla body condition
407