ArticlePDF Available


Advances in the understanding of ecological factors determining predator–prey interactions have provided a strong theoretical background on diet preferences of predators. We examined patterns of jaguar predation on caiman in southern Pantanal, Brazil. We investigated factors affecting predation rates and vulnerability of caiman to predation by jaguars. We recorded 114 caiman mortality incidents. Predation accounted for 62.3% (n = 71) of all caiman found dead, while other causes of mortality (nonpredation) accounted for 37.7% (n = 43). We found that jaguars prey on a broad size range of caiman body and caiman predation was influenced by distance to forests. During dry seasons, 70% (n = 49) of deaths were due to predation, while 30% (n = 21) were due to nonpredation causes. However, we found no significant relationship between annual and monthly killings of caiman and rainfall totals by year and month (r = 0.130, r = -0.316). The annual flooding regime may be a more important factor influencing prey selection by jaguars. Although neotropical crocodilians are relatively well studied, their interactions with jaguars have been mostly ignored and should be prioritized in future studies.
Predator–prey interactions: jaguar predation on caiman in a
floodplain forest
F. C. C. Azevedo1,2 & L. M. Verdade3
1 Departamento de Ciências Naturais, Universidade Federal de São João del Rei, São João del Rei, Brazil
2 Instituto Pró-Carnívoros, Atibaia, Brazil
3 Laboratório de Ecologia Isotópica/USP, Piracicaba, SP, Brazil
caiman; jaguar; Pantanal; predator–prey
interaction; temporal variation.
Fernando C. C. Azevedo, Departamento de
Ciências Naturais, Universidade Federal de
São João del Rei, Praça Dom Helvécio 74
Bairro Fábricas, São João del Rei, MG
36301-160, Brazil. Tel: +55 3233792483
Editor: Anne Braae
Received 7 February 2011; revised 9 August
2011; accepted 2 September 2011
Advances in the understanding of ecological factors determining predator–prey
interactions have provided a strong theoretical background on diet preferences of
predators. We examined patterns of jaguar predation on caiman in southern
Pantanal, Brazil. We investigated factors affecting predation rates and vulnerabil-
ity of caiman to predation by jaguars. We recorded 114 caiman mortality inci-
dents. Predation accounted for 62.3% (n=71) of all caiman found dead, while
other causes of mortality (nonpredation) accounted for 37.7% (n=43). We found
that jaguars prey on a broad size range of caiman body and caiman predation was
influenced by distance to forests. During dry seasons, 70% (n=49) of deaths were
due to predation, while 30% (n=21) were due to nonpredation causes. However,
we found no significant relationship between annual and monthly killings of
caiman and rainfall totals by year and month (r=0.130, r=-0.316). The annual
flooding regime may be a more important factor influencing prey selection by
jaguars. Although neotropical crocodilians are relatively well studied, their inter-
actions with jaguars have been mostly ignored and should be prioritized in future
Diet selection is a vital concept for understanding predator–
prey interactions and the relative impact of predators on dif-
ferent prey. In this context, the complex interaction between
predator and prey has long been addressed by the optimal
foraging theory (OFT) (MacArthur & Pianka, 1966). OFT
predicts that in a productive environment, predators should
be selective towards more profitable prey. However, prey
behavior can have important effects on active predator’s
choice (Abrams & Matsuda, 1993). For instance, an ambush
predator facing two alternative mobile prey species with dif-
ferences in behavior would prefer the type that is more active
if conditions of abundance, capturing and handling are similar
(Persson & Greenberg, 1990).
Diet selection by large terrestrial carnivores can be affected
by several ecological factors, such as habitat structure, prey
abundance, prey vulnerability, prey availability, rainfall/
weather conditions, seasonality and the presence of competi-
tors (Patterson et al., 2004; Owen-Smith & Mills, 2008). These
ecological factors may determine the level at which prey selec-
tion can occur. For instance, large terrestrial carnivores tend
to select larger prey species, particularly in very productive
environments (Karanth & Sunquist, 1995). In the case of
large felids, vulnerability and availability of appropriately
sized prey seem to be the most important factors determining
prey selectivity. This pattern has been observed for tigers,
Panthera tigris (Karanth & Sunquist, 1995; Bagchi, Goyal &
Sankar, 2003), lions, Panthera lion (Owen-Smith & Mills,
2008), jaguars, Panthera onca (Azevedo & Murray, 2007a),
leopards, Panthera pardus (Hayward et al., 2006), and
cougars, Puma concolor (Husseman et al., 2003) feeding pref-
erentially on the most vulnerable prey species. In addition,
seasonality (wet and dry seasons) may be important to deter-
mine selectivity by terrestrial carnivores towards certain prey
species, prey age and gender (Pole et al., 2004). For instance,
variation in rainfall levels has been suggested to affect the
vulnerability of certain prey to predation by lions due to
changes in vegetative cover and food resources (Owen-Smith
& Mills, 2008). Therefore, an evaluation of the conditions
affecting the relative vulnerability of prey species can be of
paramount importance for understanding the dynamics of
predator–prey interactions.
Despite advances in the understanding of ecological factors
determining predator–prey interactions and a strong theoreti-
cal background on diet selection by predators, few studies
have investigated prey vulnerability and selection patterns of
large felids in tropical regions. As one of the most secretive of
all tropical large cats, the jaguar has been little studied regard-
ing patterns of prey selection. Generally considered as oppor-
tunistic feeders, jaguars include a wide variety of prey in their
diet, but selectivity does occur towards large- and medium-
Journal of Zoology
Journal of Zoology. Print ISSN 0952-8369
Journal of Zoology •• (2011) ••–•• © 2011 The Authors. Journal of Zoology © 2011 The Zoological Society of London 1
sized prey species (Gonzalez & Miller, 2002). This pattern is
more pronounced in productive environments where prey
base is abundant (Polisar et al., 2003). However, large home
ranges, wide dietary breadth and dense forested habitats have
precluded precise quantification of diet constituents and accu-
rate sampling techniques for determining available prey base
in most of the jaguar’s geographical range.
In the Pantanal region of Brazil, predation of caiman,
Caiman yacare, by jaguars provides a unique opportunity to
test predictions regarding prey selection. First, an interesting
ecological interaction emerges as both jaguar and caiman are
large predators and predation of caiman by jaguar outside the
region has only been reported in Venezuela (Polisar et al.,
2003). Second, caiman are mainly active in water and may be
more vulnerable to jaguars because jaguars prefer forested
habitat with standing water and are often found on river and
lake banks (Quigley & Crawshaw, 1992). Third, seasonality
may be an important factor affecting predator–prey interaction
in the region as it is the driving force in the Pantanal ecosystem.
With the onset of the wet season in the Pantanal, most caiman
often leave lake and river areas and travel long distances
(Coutinho & Campos, 1996), which apparently makes indi-
viduals more vulnerable to predation (Campos et al., 2006).
Indeed, flooding and predation by medium-sized carnivores
are the main causes of mortality of crocodilian eggs (Campos,
1993). Finally, caiman are highly abundant in the Pantanal and
may represent an important prey item to jaguars. Population
estimates indicate that there may be as many as 35 000 000
non-hatchling caiman in the whole Pantanal (Mourão et al.,
2000), although density may vary locally.
We examined patterns of jaguar predation on caiman in the
southern Pantanal, Brazil. Despite the high local caiman
abundance in the area, a previous study reported jaguar pre-
dation to be less than expected with little or no selectivity
(Azevedo & Murray, 2007a). Our objectives were to investi-
gate factors that may affect predation rates and vulnerability
of caiman to predation by jaguars at three levels. On an
intraspecific scale, we analysed patterns of jaguar predation
among different caiman body sizes to determine whether selec-
tion was driven by body size. On a spatial scale, we analysed
jaguar kill sites to determine whether habitat factors affect
caiman vulnerability to predation by jaguars. On a temporal
scale, we analysed the influence of temporal variation to quan-
tify the effect of rainfall on jaguar predation rates on caiman.
We hypothesized that (1) jaguar would prey indiscriminantly
over a broad range of caiman body sizes; (2) jaguar kills would
not be disproportionate in sites where caiman and jaguar both
forage, indicating that other factors influence vulnerability to
predation; (3) jaguar depredation patterns would be uniform
throughout the year.
Study area
The study was conducted in the southern part of the Pantanal
region in the state of Mato Grosso do Sul in Brazil (20°05’S
and 56°36’W) during February 2003–December 2008. The
Pantanal is a floodplain covering approximately 140 000 km2
of land on the borders of Bolivia and Paraguay. The Pantanal
is characterized by an annual regime of flooding and lowland
relief, with the landscape being composed of wetlands, season-
ally inundated grasslands and woodlands, and non-flooding
forests. The study site is a working 150 km2cattle ranch/
wildlife reserve, located in the southern region of the state.
Three major habitat types occur on the ranch: (1) open areas,
with grasslands composed mainly of grass species interspersed
with palm trees such as the carandá, Copernicia australis, and
acuri palms, Attalea phalerata, and commercial rice fields; (2)
closed woodlands constituting a one-story forest with trees
such as the ipê (Tabebuia sp.) lacking clear vertical stratifica-
tion and a grass understory that is flooded during wet season;
(3) semi-deciduous forests having vertical differentiation,
including ipê and acuri palms (Dubs, 1994).
Sampling method
The density of all sizes/age classes of caiman in the study area
based on systematic night counts in permanent rice irrigation
canals and swamps was reported to be approximately 16.4
2.1 individuals km-2(mean se), yielding a biomass of
327.3 kg km-2(Azevedo & Murray, 2007a). Rainfall data were
taken from the weather station at San Francisco Ranch.
Average annual precipitation at the study site between 2003
and 2008 was 1393.2 mm (197.6). We considered wet season
commenced when a threshold of monthly rain reached
100 mm for more than 2 consecutive months. Wet seasons
usually comprised the period between October and April and
dry seasons from May to September.
Caiman captures
We captured caiman in permanent rice irrigation canals,
swamps and along a tributary of the Miranda River from June
to September 2008. We attempted to capture caiman with
different body sizes in order to represent a wide spectrum of
body sizes and both gender. Capture techniques consisted of
approaching the animals on foot or by boat at night with a
spotlight. Animals were then snared and physically restrained
during data collection. No chemical immobilization was used.
Animal handling procedures were approved by IBAMA, the
Brazilian Environment Institute.
Body- and head-length measurements
Three body-length and 11 head-length variables were taken
from captured animals, as described by Hall & Portier (1994).
Body measurements such as the snout–vent length (SVL; cm)
were taken with a tape measure (1 mm precision). Head meas-
urements were taken with a caliper (.02 mm precision). Body
mass (BM) was taken with Pesola hanging scales (PESOLA
AG, Baar, Switzerland). The gender of caiman was deter-
mined through manual probing of the cloaca and/or visual
examination of genital morphology (Allstead & Lang, 1995)
with a speculum of appropriate size.
Jaguar predation on caiman F. C. C. Azevedo and L. M. Verdade
2Journal of Zoology •• (2011) ••–•• © 2011 The Authors. Journal of Zoology © 2011 The Zoological Society of London
We used measurements from living animals captured to
predict the BM (kg) in from body and head measurements by
general linear regression. We used quadratic regression
because some proportions on caiman skulls accelerate or
decelerate on the inflexion during ontogenetic process
(Verdade, 2000). We then used these equations to estimate the
size distribution of a collection of predated caiman from skulls
collected in the study area between 2003 and 2008. We took
three cranial measurements from the skulls: dorsal cranium
length (DCL), snout length (SL) and cranium width (CW).
Measurements of captured animals were also used to deter-
mine whether there was a significant morphometric variation
between male and female caiman using multivariate statistical
analysis. We compared BM of captured caiman with esti-
mated BM of caiman predated by jaguars using t-tests in order
to analyse whether selection was driven by size.
Predator consumption of caiman
Predator consumption of caiman was determined from analy-
ses of kills found in the study area. Carcasses of caiman were
found opportunistically while checking transects or by the
presence of vultures. Transects have been systematically used in
the study area to estimate the density of potential jaguar prey
species (Azevedo & Murray, 2007a). Predation of caiman,
specifically by jaguars, was determined if 2 of the following
were observed: signs of bites on carcasses, presence of scats
near kill sites, presence of fresh tracks near the site, cached
remains or signs of a trail where the prey had been carried or
dragged. The presence of such signs indicated death due to
predation by jaguar, herein called predation, as opposed to
other mortality causes, for instance, hunting, diseases, drought
and undetermined causes, herein called nonpredation.
The estimated area where carcasses could be reasonably
found due to habitat accessibility represented 60% of the total
area of the ranch (see Azevedo & Murray, 2007a). Whenever
possible, we performed necropsies on caiman found dead from
2003 to 2008 to evaluate the probable cause of death. We
collected skulls of predated caiman and recorded information
on the precise location using a Global Positioning System
(GPS) unit. We later used GPS locations of predation and
nonpredation mortalities to estimate habitat type and prox-
imity to nearest forest and water.
Vulnerability to predation
We used logistic regression to investigate the importance of
habitat factors such as forest and standing water on caiman
predation risk. We used cause of mortality (predation vs. non-
predation causes) as the dependent variable. The distance
(meters) of dead caiman from the nearest forest and water were
treated as independent variables. To assess the extent that
proximity to suitable habitat affected caiman vulnerability to
predation, we performed simple correlation analyses using
distance from nearest forest or water versus the number of
carcasses of caiman killed by jaguars. For comparison, we also
examined distance from nearest forest or water versus the
number of nonpredation carcasses within the study site. We
tested data for normality using the Kolmogorov–Smirnov test
for goodness-of-fit. When variances were comparable, we used
t-tests; otherwise, we used nonparametric Mann–Whitney tests
(U). The fit of regression models was compared using Akaike’s
information criterion, with the Akaike weights (wi) for each
model serving as evidence that the ith model is actually the best
model of the models being considered given the data (Burnham
& Anderson, 1998). We also measured the association between
death due to predation by jaguars and due to nonpredation
causes in relation to distance to forests and water sources using
the odds ratio (OR) produced by the logistic regression, i.e. the
relative amount by which the odds of predation by jaguars
increase (OR greater than 1.0) or decrease (OR less than 1.0)
when the distance to forests and water sources is increased by
1.0 unit (meters). Probability levels were two-tailed in all analy-
ses of determinants of prey mortality.
Temporal variation
The impact of jaguar predation on caiman was annually cal-
culated based on a kill proportion (K) as follows: K=S/A,
where Sis the number of events of caiman predation by
jaguars recorded in a year and Ais the total number of
deaths (predation and nonpredation causes) recorded in that
year. To investigate temporal patterns of predation of
caiman by jaguars, we compared seasonal distribution of
mortality factors using Pearson chi-square tests. To assess
the effect of seasonality on predation, we investigated the
correlations between annual, seasonal and monthly killings
of caiman and available local records of rainfall totals by
year, season (wet/dry) and month for 2003–2008 using
bivariate correlation analysis. We also attempted to predict
caiman annual and monthly kill proportion from available
local records of rainfall totals by year and month using
simple regression analysis. We performed analysis of the
temporal autocorrelation of the kills using monthly intervals
with time lags of 12 months using the autocorrelation func-
tion of SPSS (Statistical Package for the Social Sciences
Inc., Chicago, IL, USA, 2008.). Assumptions of normality
were met by our temporal data; therefore, we used paramet-
ric tests in our correlation analyses.
Caiman biometry
We captured 33 caiman. Their mean SVL was 79.1 24.2 cm
(sd) (range: 39.0–129.0) and the mean BM was 15.7
14.0 kg (range: 1.2–55.0 kg). Male caiman comprised 60.6% of
our sample. Male caiman (SVL =86.5 25.6 mm; mean sd)
were longer than female caiman (67.8 17.4 mm) [t=2.30,
degrees of freedom (d.f.) =31, P=0.028]. In addition, male
caimans (BM =19.7 15.9 kg, n=20) were marginally
heavier than female caimans (9.6 7.4 kg, n=13) (U=81.5,
P=0.074). However, no sexual dimorphism in head and body
form was observed (multivariate analysis of variance – Pillai’s
trace criterion: F=1.630; d.f. =14, P=0.163). DCL was the
F. C. C. Azevedo and L. M. Verdade Jaguar predation on caiman
Journal of Zoology •• (2011) ••–•• © 2011 The Authors. Journal of Zoology © 2011 The Zoological Society of London 3
most accurate predictor of BM estimation for caiman (DCL
Radj2=0.983, n=21; Table 1). Using DCL equation, it was
possible to estimate BM of living caiman from skulls
(Table 2).
Caiman consumption by jaguars
For the entire study period, we recorded 114 caiman mortali-
ties. Predation accounted for 62.3% (n=71), while nonpreda-
tion accounted for 37.7% (n=43) of all caiman found dead.
We were able to retrieve a random sample of 22 skulls of
caiman predated by jaguars from 2003 to 2008. Their esti-
mated BM was 27.87 14.4 kg (mean sd). Jaguars choose
prey across a wide range of caiman body mass, from subadults
to adults (estimated BM range: 6.8–49.2 kg). We found that
caiman predated by jaguars were heavier than captured
caiman (t=-2.91, d.f. =53, P=0.005).
Vulnerability to predation
Our analysis on the importance of habitat factors in caiman
mortality for our study area suggested that distance to forest
was the most important variable explaining cause of mortality
(Table 3). The habitat-related OR for the best model [0.994;
95% confidence interval (CI) =0.991–0.998] indicated that
probability of death from jaguar predation decreases as dis-
tance to forest increases. Indeed, based on carcass locations,
the number of caiman killed by jaguars decreased with
increased distances from both forest (r=-0.639, n=71) and
water (r =-0.734, n=71). Those carcasses determined not to
be killed by jaguars did not have a significant correlation with
either distance to forest (r=-0.422, n=43) or distance to
water (r=-0.256, n=43). However, when comparing the
average distances to forest, locations of caiman killed by
jaguars (160.79 274.33 m, n=68) were closer to forest (U=
1135, P=0.048) than those not killed by jaguars (305.63
462.25 m, n=43). In contrast, distances to water were similar
(U=1420, P=0.799) between caiman killed by jaguars (38.18
95.05 m, n=68) and caiman not killed by jaguars (91.58
233.03 m, n=43).
Temporal patterns of killings
Our results yielded 11.8 5.8 caiman mortalities annually by
jaguars and a monthly average of 1.0 0.6 (sd) caiman
mortalities by jaguars. The autocorrelation function value of
monthly jaguar predation on caiman was not significant (all
P>0.950) at any time-lag intervals. We therefore considered
jaguar killing of caiman to be independent of monthly time
Table 1 Regression equations between body- and head-size measurements and the body mass in Pantanal caiman, Caiman yacare (y=a+bx +cx2).
Xa(SE)b(SE)c(SE)Radj2SE P
DCL 9.49 (3.63) -1.90 (0.38) 0.11 (0.01) 0.98 1.85 <0.001
WSR 6.50 (4.09) -2.56 (0.76) 0.30 (0.03) 0.97 2.19 <0.001
CW 4.76 (4.37) -1.95 (0.73) 0.22 (0.03) 0.97 2.45 <0.001
SW 6.86 (5.34) -3.60 (1.21) 0.49 (0.06) 0.97 2.59 <0.001
SL 4.24 (4.17) -1.85 (0.77) 0.25 (0.03) 0.96 2.74 <0.001
WN 13.83 (5.50) -20.57 (5.14) 9.20 (1.13) 0.96 2.74 <0.001
LCR 21.38 (10.74) -18.01 (5.22) 3.92 (0.62) 0.95 3.14 <0.001
IOW 2.34 (6.49) -7.87 (7.75) 8.77 (2.16) 0.90 4.43 <0.001
OL -13.97 (15.55) 1.17 (7.72) 1.70 (0.93) 0.86 5.25 <0.001
OW 5.19 (32.25) -20.18 (24.71) 9.41 (4.64) 0.83 5.78 <0.001
DCL, dorsal cranial length; WSR, surangular width; CW, cranial width; SW, basal snout width; SL, snout length; WN, maximal width of external nares;
LCR, length of the postorbital cranial roof; IOW, minimal orbital length; OL, maximal orbital length; OW, maximal inter-orbital width; SE, standard error
of the constants and the regression coefficient.
Table 2 Estimated body mass (BM) of 22 predated caiman, Caiman
yacare. Body mass in kilograms was predicted by linear regression from
dorsal cranial length (DCL) measured from skulls collected from 2003 to
2008. (BM =9.5 -1.9x+0.1x2).
27.8 41.33
27.7 40.92
18.3 11.43
20.5 16.60
28.8 45.63
28.7 45.19
28.3 43.46
19.0 12.96
15.8 6.85
27.4 39.68
29.6 49.22
25.4 31.92
26.2 34.92
26.9 37.66
18.0 10.81
21.5 19.30
25.0 30.47
18.2 11.22
20.1 15.58
19.8 14.84
23.5 25.35
aBody mass predicted from cranium width measured from a skull par-
tially broken.
Jaguar predation on caiman F. C. C. Azevedo and L. M. Verdade
4Journal of Zoology •• (2011) ••–•• © 2011 The Authors. Journal of Zoology © 2011 The Zoological Society of London
intervals. The highest number of mortalities occurred in Sep-
tember (Fig. 1a). Temporal variation on caiman predation
suggested there was an association between season and
cause of mortality. Proportion of caiman mortalities in wet
versus dry seasons differed significantly (c21=4.60, P<
0.001). During dry seasons, 72.7% (n=40) of deaths were
due to predation, while 27.3% (n=15) were due to other
causes (Fig. 1b). During wet seasons, 52.5% (n=31) of
mortalities were due to predation, while 47.5% (n=28) were
due to other causes (Fig. 1). During dry seasons, caiman
mortalities were 2.2 times more likely to occur due to preda-
tion than due to other causes (OR =2.2), and more predation
mortalities occurred in 2007 (n=19; 26.8% of all killings)
(Fig. 1c).
Correlation between caiman mortality and local records of
rainfall indicated no significant relationship between annual
and monthly caiman mortality and rainfall totals by year and
month (r=0.130, P=0.807, n=6; r=-0.316, P=0.317, n=12,
respectively). A similar pattern was observed between sea-
sonal mortalities and rainfall by season. Caiman killings were
not correlated with local rainfall during dry (r=-0.583, P=
0.224, n=6) or wet seasons (r=-0.092, P=0.863, n=6). Our
results suggested that no significant relationship between
caiman’s annual and monthly mortality proportions could be
predicted by rainfall totals either yearly or monthly (Radj2=
-0.234, P=0.831, n=6; Radj2=0.164, P=0.106, n=12,
Our results indicate that predation by jaguars was more
important than other forms of mortality in caiman popula-
tion. However, the proportions of mortalities due to jaguar
predation that we report may be underestimates. Although we
did not test for differences in probability of detection of car-
casses, our sample may have been biased in favor of carcasses
of caiman not killed by jaguars. Jaguar’s behavior with their
kills may make caiman killed by jaguars harder to find than
caiman carcasses not killed by jaguars.
Our data suggest that caiman predation by jaguars varied
between years, heavier in 2003, 2007 and 2008 and lower in
2005 and 2006 (Fig. 1c). Similar to other studies where large
felids’ predation upon wildlife increases during dry seasons
(Patterson et al., 2004), in our study, caiman mortality was
2.2 times more likely to occur due to predation than due to
other causes during dry seasons. Indeed, predation risk of
caimans by jaguars was particularly heavy at the peak of dry
seasons in September. During wet seasons, caiman mortality
Table 3 Logistic regression models of variables found to be significant in predicting causes of caiman, Caiman yacare, mortality (predated vs. other
causes) on San Francisco Ranch, Brazil during January–December 2003–2008 as ranked by Akaike information criterion (AIC), where the other cause
is the reference cell
Forest 0.994 – 0.991–0.998 – 2 0.00 0.62
Forest +water 0.994 1.001 0.991–0.998 0.999–1.003 3 1.01 0.38
Water 1.001 – 0.999–1.003 – 2 22.41 0.00
Constant 1.581 – 1 22.40 0.00
For each modeli,Kiis the number of parameters in the model, Diis the change in AIC between the model and the model with the lowest AIC (best
model) and wiis the Akaike weight. Odds ratio (OR) and the corresponding 95% confidence interval (CI) are provided. Model significance was
determined by the likelihood ratio test. Forest and water models represent distance to forests and distance to water, respectively. OR indicates the
relative amount by which the odds of predation by jaguars increase (OR greater than 1.0) or decrease (OR less than 1.0) when the distance to forests
and water sources is increased by 1.0 unit (meters).
Jan Feb Mar Apr MayJun Jul Aug Sep Oct Nov Dec
Months from 2003 to 2008
Number of caiman found dead
Number of caiman deaths
Number of caiman dead
Predation dry
Predation wet
Nonpredation dry
Nonpredation wet
2003 2004 2005 2006 2007 2008
2003 2004 2005 2006 2007 2008
(a) (b) (c)
Figure 1 Monthly (a), seasonal (b) and annual (c) patterns of predation and nonpredation mortality of caiman, Caiman yacare (n=114) recorded at
San Francisco Ranch, Pantanal region, Brazil, from 2003 to 2008.
F. C. C. Azevedo and L. M. Verdade Jaguar predation on caiman
Journal of Zoology •• (2011) ••–•• © 2011 The Authors. Journal of Zoology © 2011 The Zoological Society of London 5
associated with jaguar predation was similar to other mor-
tality causes. The concentration of predation cases during
dry seasons seems to be related to the highest density of
caiman congregated in a few sites containing water of a suit-
able depth (Campos, Coutinho & Magnusson, 2003) and to
the longer caiman movement distances during dry seasons,
up to 10 km (Schaller & Crawshaw, 1980). Therefore, we
might tentatively conclude that jaguar predation on caiman
seems to be seasonally influenced. However, we could not
test for the possibility of variability in our carcass detection
by season or habitat. As any variability in this probability
could affect our results, our ability to speculate further is
Our results indicated that caiman predation by jaguars
was influenced by distance to forests. While the distribution
of carcasses of predated caiman was negatively associated
with distance to closest forest, those carcasses of nonpreda-
tion causes were not. In general, forests provide cover for
predators while stalking prey (Husseman et al., 2003). The
importance of forests with regard to vulnerability of live-
stock to predation by jaguars has already been reported in
the Pantanal region, where livestock predation risk increased
with proximity to forest cover declined (Azevedo & Murray,
2007b). A similar pattern was found in Southern Amazon
for cattle depredation by pumas and jaguars (Palmeira et al.,
2008). Adult caiman in the Pantanal are known to move
long distances in short periods of time (Campos et al., 2006),
which could increase their vulnerability to predation with
increasing distances from permanent water sources. Indeed,
our results indicate different correlation patterns of caiman
mortalities due to predation or nonpredation causes in rela-
tion to closest permanent water sources, suggesting that dis-
tance to water did influence caiman predation by jaguars.
While the distributions of caiman carcasses killed by jaguar
decreased with increased distance to water, those caiman
carcasses not killed by jaguar did not. In the Pantanal, the
availability of ephemeral water bodies that persist through-
out the first months of the dry season likely reflected the lack
of influence of permanent water sources on the vulnerability
of cattle to predation by jaguars. Cattle can find water to
drink during dry seasons even far from permanent water
sources (Azevedo & Murray, 2007b). Contrary to what has
been reported for jaguar depredation on livestock in the
Pantanal, caiman seemed to be more dependent on perma-
nent water holes not only for food resources but also to
increase their chance to escape from jaguar attacks. Caiman
probably find refuge from predation by jaguars in deep
As reported for other large predators that spend most of
their time near available water sources where they can find and
kill prey during dry seasons (Patterson et al., 2004), we specu-
late that jaguars concentrate their hunting near those perma-
nent water sites to increase their hunting success. Such
vulnerability seems to decrease during wet seasons when
flooding yields a very rich and productive habitat in which
caimans disperse (Junk et al., 2006). However, even though
more predation occurred during dry seasons, we found no
significant relationship between local rainfall levels and
caiman predation by jaguars. Our results suggest that neither
annual nor monthly rainfall totals at the study site influenced
predation of caiman by jaguars. Although rainfall has been
reported as an explanatory variable for predation patterns in
ecosystems where an array of large predators and prey interact
(Owen-Smith & Mills, 2008), local precipitation was not rel-
evant in the present study. A possible explanation for this
pattern is that the water level in Pantanal is determined by
precipitation upstream from a huge drainage, including vast
areas of Cerrado and the Amazon forest. In fact, the annual
flooding regime as a consequence of the rainfall in the whole
drainage seems to be more decisive for jaguar prey selection
(Crawshaw & Quigley, 1991).
In general, there is a bias in predation towards younger
individuals (Temple, 1987). Although body masses of the
carcasses of caiman killed by jaguars were estimates from the
regression equations and have an additional source of error
associated with them, our findings regarding the range of
estimated body masses for caiman killed by jaguars support
the hypothesis that jaguars prey upon a wide range of
caiman body sizes, including adult male and female caimans.
In addition, despite the fact that captures and carcass recov-
eries may have different biases towards smaller or larger
caimans, our results suggest that jaguars selected heavier
caiman compared with captured caiman. As a result, in areas
with jaguars, such as the Pantanal, we can expect a shorter
reproductive period and, consequently, a faster turnover of
reproductive male and female jaguars. Under these circum-
stances, the local extinction of the jaguar, which has already
occurred on a significant part of its range, could potentially
affect behavioral ecology and reproductive success of
caiman, with regard to mating systems, recruitment, disper-
sal patterns, use of areas near or in forests and population
Our results indicate that the caiman population structure
in our study area was male biased, and male caimans were
larger than female caimans, which is similar to the pattern
found in other central areas of the Pantanal region (Campos,
Coutinho & Magnusson, 2003). However, we did not find any
evidence of cranial sexual dimorphism in Pantanal caiman as
described for other species such as the American alligator
(Viosca, Jr, 1939), the Philippine’s crocodile (Hall & Portier,
1994) and the broad-snouted caiman (Piña et al., 2007). This
unexpected pattern should be prioritized in future studies and
could possibly be related to the species social behavior (Vliet,
The jaguar occurs throughout the geographic range of
caimans. However, jaguars have undergone local extinction in
many areas due to human impacts. Caiman, which are an
important prey species for jaguar, may also be affected as we
describe here. Further research on the influence of jaguar local
extinctions on caiman populations, behavior and genetics
should be prioritized in future studies.
We thank FAPESP for providing funds for the study
(FAPESP Biota Program, Proc. Nos. 2006/60954-4 and 2007/
Jaguar predation on caiman F. C. C. Azevedo and L. M. Verdade
6Journal of Zoology •• (2011) ••–•• © 2011 The Authors. Journal of Zoology © 2011 The Zoological Society of London
00976-7). Marina Lacorte, Nuno Pedroso, Luis Rosalino,
Emilie Henst, Cynthia Widmer, Phil Wilkinson, Christy
Wilkinson and Allan Woodward participated on data collec-
tion from living caiman. Instituto Pró-Carnívoros and San
Francisco Ranch provided logistics. We thank Henrique
Concone and Ricardo Costa for the help with mortality
surveys. We thank Jocelyn Aycrigg for constructive comments
on this paper.
Abrams, P. & Matsuda, P. (1993). Effects of adaptive preda-
tory anti-predator behavior in a two-prey-one-predator
system. Evol. Ecol. 7, 312–326.
Allstead, J. & Lang, J.W. (1995). Sexual dimorphism in the
genital morphology of young American alligators, Alligator
mississippiensis.Herpetologica 51, 314–325.
Azevedo, F.C.C. & Murray, D.L. (2007a). Spatial organiza-
tion and food habits of jaguars (Panthera onca) in a flood-
plain forest. Biol. Conserv. 137, 391–402.
Azevedo, F.C.C. & Murray, D.L. (2007b). Evaluation of
potential factors predisposing livestock to predation by
jaguars. J. Wildl. Manage. 71, 2379–2386.
Bagchi, S., Goyal, S.P. & Sankar, K. (2003). Prey abundance
and prey selection by tigers (Panthera tigris) in a semi-arid,
dry deciduous forest in western India. J. Zool. (Lond.) 260,
Burnham, K.P. & Anderson, D.R. (1998). Model selection and
inference: a practical information-theoretic approach. New
York: Springer-Verlag.
Campos, Z. (1993). Effect of habitat on survival of eggs and
sex ratio of hatchlings of Caiman crocodilus yacare in the
Pantanal, Brazil. J. Herpetol. 27, 127–132.
Campos, Z., Coutinho, M. & Magnusson, W.E. (2003). Ter-
restrial activity of caiman in the Pantanal, Brazil. Copeia 3,
Campos, Z., Coutinho, M., Mourão, G.M., Bayliss, P. &
Magnusson, W.E. (2006). Long distance movements by
Caiman crocodiles yacare: implications for management of
the species in the Brazilian Pantanal. Herpetol. J. 16, 123–
Coutinho, M. & Campos, Z. (1996). Effect of habitat and sea-
sonality on the densities of caiman in southern Pantanal,
Brazil. J. Trop. Ecol. 12, 741–747.
Crawshaw, P.G. & Quigley H. (1991). Jaguar spacing, activity
and habitat use in a seasonally flooded environment in
Brazil. J. Zool. 223, 357–370.
Dubs, B. (1994). Differentiation of woodland and wet savanna
habitats in the Pantanal of Mato Grosso, Brazil. Disserta-
tion, University of Zürich, Edinburgh, UK.
Gonzalez, C.A.L. & Miller, B.J. (2002). Do jaguars Panthera
onca depend on large prey. West. North Am. Nat. 62, 218–
Hall, P.M. & Portier, K.M. (1994). Cranial morphometry of
New Guinea crocodiles (Crocodylus novaeguineae): ontoge-
netic variation in relative growth of the skull and an assess-
ment of its utility as a predictor of the sex and size of
individuals. Herpetol. Monogr. 8, 203–225.
Hayward, M.W., Henschell, P., O’Brien, J., Hofmeyr, M,
Balme, G. & Kerley, G.H.I. (2006). Prey preferences of
the leopard (Panthera pardus). J. Zool. (Lond.) 270, 298–
Husseman, J.S., D.L. Murray, G. Power, C. Mack, C.R.
Wenger & H. Quigley (2003). Assessing differential prey
selection patterns between two sympatric large carnivores.
Oikos 101, 591–601.
Junk, W.J., Cunha, C.N., Wantzen, K.N., Petermann, P.,
Strüssmann, C., Marques, M.I. & Adis, J. (2006). Biodiver-
sity and its conservation in the Pantanal of Mato Grosso,
Brazil. Aquat. Sci. 68, 278–309.
Karanth, K.U. & Sunquist, M.E. (1995). Prey selection by
tiger, leopard and dhole in tropical forests. J. Anim. Ecol.
64, 439–450.
MacArthur, R.H. & Pianka, E.R. (1966). On the optimal use
of a patchy environment. Am. Nat. 100, 603–609.
Mourão, G. M., Coutinho, M., Mauro, R., Campos, Z.,
Tomás, W. & Magnusson, W. E. (2000). Aerial surveys of
caiman, marsh deer and pampas deer in the Pantanal
wetland of Brazil. Biol. Conserv. 92, 175–183.
Owen-Smith, N. & Mills, M.G.L. (2008). Shifting prey
selection generates contrasting herbivore dynamics within
a large-mammal predator-prey web. Ecology 89,
Palmeira, F.B.L., Crawshaw, P.G., Jr, Haddad, C. M.,
Ferraz, K.M.P.M. & Verdade, L.M. (2008). Cattle depreda-
tion by puma (Puma concolor) and jaguar (Panthera onca)
in Northern Goiás, Central-Western Brazil. Biol. Conserv.
141, 118–125.
Patterson, B.D., Kasiki, S.M., Selempo, E. & Kays. R.W.
(2004). Livestock predation by lions (Panthera leo) and
other carnivores on ranches neighboring Tsavo National
Parks, Kenya. Biol. Conserv. 119, 507–516.
Persson, L. & Greenberg, L.A. (1990). Optimal foraging and
habitat shifts in perch (Perca fluviatilus) in a resource gradi-
ent. Ecology 71, 1699–1713.
Piña, C.I., Larriera, A., Siroski, P. & Verdade. L.M. (2007).
Cranial sexual discrimination in hatchling broad-snouted
caiman (Caiman latirostris). Iheringia Ser. Zool. 97, 17–20.
Pole, A., Gordon, I.J., Gorman, M.L. & MacAskill, M.
(2004). Prey selection by African wild dogs in southern
Zimbabwe. J. Zool. (Lond.) 262, 207–215.
Polisar, J., Maxit, I., Scognamillo, D., Farrel, L., Sunquist,
M.E. & Eisenberg, J.E. (2003). Jaguars, pumas, their prey
base, and cattle ranching: ecological interpretations of a
management problem. Biol. Conserv. 109, 297–310.
Quigley, H.B. & Crawshaw, P.G., Jr (1992). A conservation
plan for the jaguar Panthera onca in the Pantanal region of
Brazil. Biol. Conserv. 61, 149–157.
Schaller, G.B. & Crawshaw P.G. Jr. (1980). Movement pat-
terns of jaguar. Biotropica 12, 161–168.
F. C. C. Azevedo and L. M. Verdade Jaguar predation on caiman
Journal of Zoology •• (2011) ••–•• © 2011 The Authors. Journal of Zoology © 2011 The Zoological Society of London 7
Temple, S.A. (1987). Do predators always captured substand-
ard individuals disproportionately from prey populations?
Ecology 68, 116–125.
Verdade, L.M. (2000). Regression equations between body
and head measurements in the broad-snouted caiman
(Caiman latirostris). Rev. Bras. Biol. 60, 469–482.
Viosca, P., Jr (1939). External sexual differences in the
alligator, Alligator mississippiensis.Herpetologica 1,
Vliet, K.A. (1989). Social displays of the American alligator.
Am. Zool. 29, 1019–1031.
Jaguar predation on caiman F. C. C. Azevedo and L. M. Verdade
8Journal of Zoology •• (2011) ••–•• © 2011 The Authors. Journal of Zoology © 2011 The Zoological Society of London
... Although several studies have investigated jaguar ecology in the Brazilian Pantanal (e.g. Crawshaw and Quigley, 1991;Silveira, 2004;Soisalo and Cavalcanti, 2006;Azevedo and Murray, 2007;Cavalcanti and Gese, 2009;Azevedo and Verdade, 2012), so far only four genetic analyses have been published (Eizirik et al., 2008;Roques et al., 2014Roques et al., , 2016Valdez et al., 2015) and all have focused on the southern portion of the biome. The initial studies indicated that southern Pantanal jaguars maintain considerable levels of genetic variability and connectivity, a pattern that must be further tested with expanded sampling into the northern portion of the biome. ...
... We genotyped for this study samples collected between 2008 and 2015 during field ecology and behavioral projects (Silveira, 2004;Azevedo and Murray, 2007;Cavalcanti and Gese, 2009;Azevedo and Verdade, 2012). We obtained 58 blood samples from free-ranging, captured animals inhabiting five different study sites ( Table S1). ...
Habitat loss and fragmentation are important threats to carnivores worldwide and are especially intense for large predators. Jaguars have been extirpated from over half of their original distribution, and few regions still maintain large populations. The Pantanal is among the best examples of such regions and can be used to better understand several aspects of jaguar biology that are relevant for conservation planning throughout the species' range. Thus, in this study we used microsatellite markers and field data to: (i) assess the genetic structure and gene flow of jaguars (n = 110) from the northern and southern Pantanal; (ii) verify if females are more phil-opatric than males; (iii) produce a timeline pedigree to allow the identification of distances from offspring to their mothers; and (iv) estimate the generation time for jaguars. Our results are consistent with the hypothesis that Pantanal jaguars represent a panmictic population, although exhibiting some degree of local differentiation. The Paraguay River seems to be an important factor promoting gene flow among the studied populations, highlighting its relevance for regional conservation efforts. Our data provide the first genetic evidence of female philopatry and male-biased dispersal in jaguars. In addition, we report the first timeline pedigree for a wild jaguar population and the first direct estimate of the species' generation time. Our results contribute to the construction of more realistic assessments of jaguar population dynamics and long-term genetic viability, thus contributing to the design of improved conservation strategies on behalf of this species.
... This low occupancy estimate may suggest that human-altered modified landscapes may have a negative influence on jaguar occupancy. However, contrary to our expectations, increasing distances from cities, Eucalyptus plantations, and pastures did not water habitats has long been documented (Azevedo & Verdade, 2012;Figel et al., 2019;Ramalho et al., 2021) and jaguars in RDSP seemed to follow this pattern. ...
Full-text available
The jaguar (Panthera onca) plays an important role in maintaining biodiversity and ecological processes. We evaluated the status of a jaguar population in one of the last stronghold habitats for its conservation in the Atlantic Forest, the Rio Doce State Park (RDSP). We used a random survey design from 2016/17 to estimate jaguar abundance and density as well as its occupancy and detection probabilities in the entire Park's area. To monitor for temporal fluctuations in density and abundance, we used a systematic survey design in the southern portion of the Park where jaguars were more recorded when using the random approach. We then conducted two surveys in 2017/18 and 2020. Our 2016/17 random survey revealed that jaguar density (0.11 ± SE 0.28 individuals/100 km²) was the lowest obtained for the species across the Atlantic Forest. We noticed that jaguar density increased three times from 2017/18 (0.55 ± SE 0.45 individuals/100 km²) to 2020 (1.61 ± SE 0.6 individuals/100 km²). Jaguar occupancy and detection probability were 0.40 and 0.08, respectively. The low jaguar occupancy probability was positively associated with smaller distances from lakes and records of potential prey. The detection probability was positively associated with prey detection, the rainy season, and smaller distances from lakes. Our work contributes to a growing awareness of the potential conservation value of a protected area in a human-dominated landscape as one of the last strongholds for jaguars across the Atlantic Forest.
... The dry season concentrates wildlife in and around remaining water bodies, while the wet season can force terrestrial species into remaining and restricted dry areas. Thus, both seasons can cause greater abundance of prey in a relatively accessible area, possibly favoring predators (Smith 1981;Anderson et al. 2009;Azevedo and Verdade 2012). In addition to seasonal flooding that raises water levels and reduces unflooded areas, ESEC-MJ experiences daily tides of 10-m amplitude, which flood extensive areas (Santos et al. 2016;Ferreira et al. 2017). ...
Full-text available
Limited space on islands usually cannot sustain stable populations of large predators. However, jaguars, the largest cat species in the Americas, unexpectedly occur in the Estação Ecológica Maracá-Jipioca reserve, a system of continental Amazonian islands in the Atlantic Ocean. We investigated jaguar population structure, density, and activity patterns. We placed 25 camera-trap stations across 149.19 km 2 and used spatially explicit capture-recapture to estimate density, and Rayleigh's test to assess activity patterns. We identified 21 individuals (12 females, six males, and three cubs) and estimated an adult density of 6.7 individuals per 100 km 2 , which equals a population of approximately 43 jaguars. The population is composed mostly of females (66%) in relation to males (33%). Male and female activity patterns overlapped and showed more activity during daytime. The high jaguar density and the presence of females with cubs indicate that these islands are likely natural refugia for jaguars, reinforcing the importance of this protected area for jaguar conservation and possibly challenging the paradigm that large-mammal populations are not feasible in restricted islands.
... They are seldom sighted in forested habitats, which may explain the relative low number of marsh deer killed by jaguars in our study (n = 47), mostly in flat grasslands or water. As for caiman, predation by jaguars in the Pantanal is influenced by the fact that they move longer distances on land during the dry season, which exposes them to higher predation risk (Azevedo and Verdade 2012;Campos et al. 2003). Likewise, forested cover habitats represented the second most important type of habitat used by jaguars to kill caiman in both seasons (n = 48, 37.2% of all caiman killed). ...
Ecologists increasingly recognize the importance of habitat selection as a multi-level, hierarchical process, but individual-level variation in selection patterns have yet to be fully evaluated within populations of large carnivores. We assessed jaguar (Panthera onca) selection of forest across seasonally variable forest availability in the Pantanal region of South America. Using resource selection functions (RSF), we evaluated the importance of forest cover for jaguars in a heterogeneous, dynamic landscape, and how seasonal variability in habitat availability influences habitat selection and predation patterns. A multi-level hierarchical analytical framework revealed that jaguars increasingly selected forest cover as availability of other habitats declined due to seasonal flooding. At the population range level, jaguars selected water, forest, and bushy grassland comparably during the dry season, and bushy grassland, forest, and flat grassland during the wet season. At the home range level, bushy grassland and forest were selected during both seasons. Jaguars killed prey animals with comparable frequency across forest, water, and bushy grasslands during the dry season, with more preference detected for forest and avoidance of grasslands. During the wet season, jaguars killed prey disproportionately to their availability in forest and bushy grasslands. These results clearly indicate that jaguars undergo marked dynamic shifts in habitat and kill site selection in response to seasonal flooding, with forest patches becoming more important as the availability of other habitats decreases. Land conservation efforts need to consider variability in jaguar-landscape dynamics associated with seasonal flooding, given increasing pressure to modify native landscapes to more open livestock grazing areas.
... In general, mammalian carnivores occur in low density, since they need large areas to survive (Karanth & Chellam, 2009;Loveridge et al., 2010). Their populations decline may lead to several ecosystem imbalances, considering that these species are efficient prey regulators (Azevedo & Verdade, 2012;Terborgh et al., 2001). In current reduced and fragmented natural areas, some species of more resilient mammalian carnivores use areas modified by humans or in a regeneration stage to prey searching (Karanth & Chellam, 2009). ...
Full-text available
Ocelots (Leopardus pardalis) are widely distributed throughout the Americas, being dependent on forested areas to survive. Although ocelot ecology is broadly studied throughout the species range distribution, studies concerning factors that may affect ocelot occupancy in the Atlantic Forest are still scarce. We used camera traps to evaluate factors influencing the probabilities of detection and occupancy of ocelots in a protected area of the Atlantic Forest, the Rio Doce State Park (RDSP), southeastern Brazil. To assess ocelot occupancy and detection probabilities, we measured the distances between sampling stations and rivers, lakes, cities, pasture, and Eucalyptus plantations. In addition, we recorded the mean rainfall levels for each sampling occasion, and native grassland areas within a 500 m-buffer around each sampling station. We found a strong and positive association between ocelot detection and the dry season, which might be due to a higher number of individuals moving through the Park during this season. Moreover, we found a strong and positive association of ocelot detection with native grassland areas around lakes, which may be related to the ocelot behavior of searching for prey in these areas. Conversely, the ocelot occupancy probability was intermediate ( Ψ ^ = 0.53, 95% CI = 0.36-0.69) and was not strongly associated with the evaluated covariates, which may be explained by the high-quality of forest habitats and water resources that are homogeneously distributed within the Park. Our study indicates that the RDSP still provides a structurally suitable forest habitat for ocelots, but because of the current worrying scenario of over fragmentation, reduction of forest cover, and weakness of the protective legislation of this biome, the long-term persistence of the species in RDSP is uncertain.
... Large felids (>30 kg) have exclusively carnivorous diets and depend upon medium and large terrestrial prey to fuel high metabolic demands (Sunquist andSunquist 2002, Carbone et al. 2007). Although some of them commonly use trees for resting, hunting, avoiding predators or competitors (e.g., leopards Panthera pardus [Le Roux and Skinner 1989] and pumas Puma concolor [Santos et al. 2014]), and others may prey on arboreal, aquatic, and semiaquatic species (e.g., jaguars Panthera onca [Azevedo and Verdade 2012] and tigers Panthera tigris [Mukherjee and Sen Sarkar 2013]), there are no documented cases of large felids living a primarily arboreal existence for extended periods. Here, we report the evolution of a unique lifestyle for a large terrestrial top predator, in which jaguars live an arboreal and semiaquatic existence for 3-4 months of the year in a completely flooded environment during the annual highwater season of the Amazon River Basin (Fig. 1). ...
Large felids (>30 kg) have exclusively carnivorous diets and depend upon medium and large terrestrial prey to fuel high metabolic demands (Sunquist and Sunquist 2002, Carbone et al. 2007). Although some of them commonly use trees for resting, hunting, avoiding predators or competitors (e.g., leopards Panthera pardus ‐ Le Roux & Skinner, 1989 and pumas Puma concolor ‐ Santos et al. 2014), and others may prey on arboreal, aquatic and semi‐aquatic species (e.g., jaguars Panthera onca ‐ Azevedo & Verdade, 2012; and tigers Panthera tigris ‐ Mukherjee & Sen Sarkar, 2013), there are no documented cases of large felids living a primarily arboreal existence for extended periods. Here, we report the evolution of a unique lifestyle for a large terrestrial top predator, in which jaguars live an arboreal and semi‐aquatic existence for 3‐4 months of the year in a completely flooded environment during the annual high‐water season of the Amazon River Basin (Fig. 1).
... All crocodilians are highly susceptible to predation in the first months of life by a large variety of other animals, including large fish, snakes, monitor lizards, raptors, wading birds, small mammalian carnivores, and also conspecifics (Somaweera et al. 2013). However, members of certain species (e.g., saltwater crocodile Corocodylus porosus, American alligator) become apex predators in their respective habitats when they reach maturity (Grigg and Kirshner 2015), while others (e.g., Yacare caiman Caiman yacare) remain susceptible to predation into adulthood (Azevedo and Verdade 2012). Because of the similar risks of predation in early life, one might predict that hatchlings of any species would show similar responses to novel stimuli, such as little exploration behavior in a novel environment and overall lower levels of activity. ...
Full-text available
Behavioral predispositions are innate tendencies of animals to behave in a given way without the input of learning. They increase survival chances and, due to environmental and ecological challenges, may vary substantially even between closely related taxa. These differences are likely to be especially pronounced in long-lived species like crocodilians. This order is particularly relevant for comparative cognition due to its phylogenetic proximity to birds. Here we compared early life behavioral predispositions in two Alligatoridae species. We exposed American alligator and spectacled caiman hatchlings to three different novel situations: a novel object, a novel environment that was open and a novel environment with a shelter. This was then repeated a week later. During exposure to the novel environments, alligators moved around more and explored a larger range of the arena than the caimans. When exposed to the novel object, the alligators reduced the mean distance to the novel object in the second phase, while the caimans further increased it, indicating diametrically opposite ontogenetic development in behavioral predispositions. Although all crocodilian hatchlings face comparable challenges, e.g., high predation pressure, the effectiveness of parental protection might explain the observed pattern. American alligators are apex predators capable of protecting their offspring against most dangers, whereas adult spectacled caimans are frequently predated themselves. Their distancing behavior might be related to increased predator avoidance and also explain the success of invasive spectacled caimans in the natural habitats of other crocodilians.
... During prey selection, besides nutritional value and vulnerability, predators also assess the risk of injuries associated with the prey's defensive weapons and size-related strength. Still, numerous other factors, such as physical threats and barriers in the prey's habitat or the potential to hunt in packs, may alter a predator's prey choice (Lendrem, 1986;Hayward and Kerley, 2005;Azevedo and Verdade, 2011;Mukherjee and Heithaus, 2013). ...
Gregarious behaviour of large bodied herbivorous dinosaurs, such as ceratopsians, hadrosaurs and sauropods, has received much attention due to their iconic mass death assemblages (MDAs). Yet, social lifestyle of ankylosaurs, a highly specialized group of armoured herbivores that flourished predominantly during the Cretaceous Period, remains largely ambiguous. Whereas most ankylosaurs are found as isolated individuals, which may suggest a dominantly solitary lifestyle, the few examples of ankylosaur MDAs indicate that some members of this clade could have been gregarious. In this review, we assess taphonomic history, ontogenetic composition of the MDAs, defence system and other comparative anatomical attributes, and inferred habitat characteristics of ankylosaurs; aspects that may indicate and/ or influence group formation in extant herbivores and can also be studied in fossils. We show that the ankylosaurian gross anatomy, such as their heavy armour, barrel-shaped body and usually stocky limbs, combined with the rarity of their MDAs and multiple parallel trackways, all suggest a solitary adult life with efficient anti-predator defence system, limited agility, and confined foraging range. However, characteristics of the known MDAs of Pinacosaurus, Gastonia, and the Iharkút nodosaurids evaluated in this study imply that at least some ankylosaurs formed groups. Nevertheless, we found no common and consistent set of features to explain why these particular ankylosaurs were gregarious. While inefficient anti-predator defence along with likely higher agility of juvenile Pinacosaurus living in open habitats could account for their gregarious behaviour, such ontogenetic, anatomical and habitat features are not combined either in Gastonia or in the Iharkút nodosaurid MDAs. Instead, members of each MDA likely had their own specific conditions driving them to form relatively small herds, indicating a more complex social structuring in ankylosaurs than previously acknowledged. Studying morphological and functional disparity within Ankylosauria may help explain the repertoire of their social behaviour. Our holistic approach shows that combining palaeontological and biological information is essential and can provide new insights into the behavioural ecology of long extinct vertebrates.
Full-text available
Most large felids are classified as solitary species, with only lions ( Panthera leo ) and cheetahs ( Acinonyx jubatus ) exhibiting social, collaborative behaviours. Herein, we present evidence of the formation of male coalitions by jaguars ( Panthera onca ), based on data from five studies conducted with camera trapping, GPS telemetry, and direct observations in the Venezuelan Llanos and Brazilian Pantanal. Out of 7062 male records obtained with camera traps or visual observations, we detected 105 cases of male-male interactions, of which we classified 18 as aggression, nine as tolerance, 70 as cooperation/coalition, and eight as unidentified. In two studies, two male jaguars formed stable coalitions lasting over 7 years each. In the Llanos, each coalition male paired and mated with several females. For male jaguar coalitions, we documented similar behaviours as recorded earlier in lions or cheetahs, which included patrolling and marking territory together, invading territories of other males, collaborative chasing and killing other jaguars, and sharing prey. However, different from lions or cheetahs, associated male jaguars spent less time together, did not cooperate with females, and did not hunt cooperatively together. Our analysis of literature suggested that male jaguar coalitions were more likely to form when females had small home range size, a proxy of females’ concentration, while in lions, the male group size was directly correlated with the female group size. Similarly, locally concentrated access to females may drive formation of male coalitions in cheetahs. We conclude that high biomass and aggregation of prey are likely drivers of sociality in felids. Significance statement The division into social and solitary species in large felids has so far seemed unambiguous, with only lions and cheetahs classified as social species, in which male coalitions also occurred. Our data show that, under certain conditions, male coalitions may also form in jaguar populations. Factors that drive formation of male coalitions in lions and cheetahs, but not in other species of large cats, have not been clear until now. Our analyses indicate that in jaguars, lions, and cheetahs, the concentration of females likely plays the most important role. In jaguars, the probability of male coalition occurrence is highest in populations with the smallest mean female home range size (and thus likely high local density of females), while in lions, male group size is most strongly correlated with female group size.
Full-text available
Agricultural development was the major contributor to South America's designation as the continent with the highest rates of forest loss from 2000-2012. As the apex predator in the Neotropics, jaguars (Panthera onca) are dependent on forest cover but the species' response to habitat fragmentation in heterogeneous agricultural landscapes has not been a subject of extensive research. We used occupancy as a measure of jaguar habitat use in Colombia's middle Magdalena River valley which, as part of the intercontinental Tumbes-Chocó-Magdalena biodiversity hotspot, is exceedingly fragmented by expanding cattle pastures and oil palm plantations. We used single-season occupancy models to analyze 9 months of data (2015-2016) from 70 camera trap sites. Given the middle Magdalena's status as a "jaguar corridor" and our possible violation of the occupancy models' demographic closure assumption, we interpreted our results as "probability of habitat use (Ψ)" by jaguars. We measured the associations between jaguar presence and coverage of forest, oil palm, and wetlands in radii buffers of 1, 3, and 5 km around each camera trap. Our camera traps recorded 77 jaguar detections at 25 of the camera trap sites (36%) during 15,305 trap nights. The probability of detecting jaguars, given their presence at a site, was 0.28 (0.03 SE). In the top-ranked model, jaguar habitat use was positively influenced by wetland coverage (β = 7.16, 3.20 SE) and negatively influenced by cattle pastures (β = -1.40, 0.63 SE), both in the 3 km buffers. We conclude that wetlands may serve as keystone habitats for jaguars in landscapes fragmented by cattle ranches and oil palm plantations. Greater focus on wetland preservation could facilitate jaguar persistence in one of the most important yet vulnerable areas of their distribution.
Full-text available
Jaguar on two small ranches in southwestern Brazil had a density of about one animal per 25 sq km. Females ranged over at least 25-38 sq km and males over twice that much terrain. The ranges of females overlapped, and the range of a resident male included the ranges of several females. Jaguar and puma ranges also overlapped, but each species favored parts not much used by the other. Day-and-night radio tracking revealed precise travel and activity patterns of two female jaguar. The social system of jaguar is in most respects similar to that of other large solitary cats such as puma, leopard, and tiger.
Full-text available
Flooding of nests and predation were identified as the major mortality agents of eggs of Caiman crocodilus yacare in Brazil. Nests on floating grass mats and in forests were flooded in 1989, but only floating grass mats nests were flooded in 1990. Predation on eggs was high in forest nests, but my presence may have increased predation. No nest on floating grass mats was attacked by predators. The period of sex determination of C. c. yacare can extend to 40 d of incubation. I predicted nest temperature with a multiple regression model incorporating weather variables during incubation. The predicted nest temperatures for nests on floating grass mats and nests in forest were correlated with the sex ratio of the clutch. High nest temperatures (> 31.5 C) produced mostly males, and low nest temperatures (<30.5 C) produced only females. The estimated sex ratio varied between years in nests on floating grass mats, but not in forest nests. Estimated nest temperatures significantly affected the snout-vent lengths of hatchlings but not their masses. Fifty percent of nesting occurred on floating grass mats, so the destruction of this habitat by introduced animals such as the water buffalo (Bubalus bubalis) will adversely affect the productivity and hence long-term density of caimans in the Pantanal.
The jaguar (Panthera onca) has been classified as an opportunistic hunter that takes as many as 85 prey species, according to availability. In this study we analyzed jaguar food habits throughout its range to quantify the importance of small, medium, and large prey in the diet. Because peccaries (Tayassu) are present in most studies, we also tested their importance in relation to other prey items. We conclude that jaguars are equally using medium- and large-size prey, with a trend toward use of larger prey as distance increases from the equator. There was no significant difference between the importance of peccaries and other large prey.
1. Ecological factors influencing prey selection by tiger Panthera tigris, leopard Panthera pardus and dhole Cuon alpinus were investigated in an intact assemblage of large mammals in the tropical forests of Nagarahole, southern India, between 1986 and 1990. 2. Densities of large herbivores were estimated using line transects, and population structures from area counts. Carnivore diets were determined from analyses of scats (faeces) and kills. Selectivity for prey species was inferred from likelihood ratio tests comparing observed counts of scats to hypothesized scat frequencies generated from prey density estimates using parametric bootstrap simulations. Predator selectivity for size, age, sex and physical condition of prey was estimated using selection indices. 3. Ungulate and primate prey attained a density of 91 animals km-2 and comprised 89-98% of the biomass killed. Predators showed significant (P
The cranial morphometry of New Guinea crocodiles (Crocodylus novaeguineae) was investigated to quantify ontogenetic variation in relative growth of the skull to provide possible demographic insights into commercially harvested populations. Growth attributes were, in general, linear, and univariate statistics provided estimates of size from untransformed values. Curvilinear size (total length) to age transformations were attempted through applications of three growth models: a reparameterized Richards curve, the von Bertalannfy model, and a Brody curve. Satisfactory performance was attained only through the Brody model, with reliable age estimates restricted to smaller (≤2 m) individuals. The utility of relative growth ratios to correctly predict the sex of known gender animals was assessed by both parametric (multivariate discriminant analysis--DISCRIM) and nonparametric (binary tree classification analysis--CART) methods. Crossvalidation data sets were created to reduce model biases and to evaluate the correctness of classification rates. CART models consistently showed greater congruence of apparent error rates with true error rates than did DISCRIM models, while the 'best fit' models of each method closely paralleled one another in providing conservative overall ability to correctly discriminate gender from measured attributes. Skull growth occurs in three distinctive stages that mimic a saltatory ontogenetic response and are hypothesized to be related to functional foraging responses.
Predators are commonly thought to capture substandard individuals (those in poorer condition than the average individual) in higher than expected proportions, but evidence for this paradigm is scant, biased, and inconsistent. I describe the outcomes of 447 attacks by a trained Red-tailed Hawk (Buteo jamaicensis) on three species of typical mammalian prey (eastern chipmunks, cottontail rabbits, and gray squirrels). A strong correlation existed between the difficulty the hawk had capturing individuals of each species and the difference between the proportions of substandard individuals in the hawk's diet and the prey populations. The degree to which substandard individuals of a particular prey species are taken disproportionately by a predator seems to be a direct function of how difficult it normally is for the predator to capture and kill individuals of that species. A review of other studies shows a similar relationship between vertebrate predators and their prey.
Coordinated terrestrial movement has not previously been reported in crocodilians. However, between 1989 and 1999, 94% of 525 Caiman crocodilus yacare found on land in the Pantanal were in coordinated groups (n = 73) walking head to tail and forming nearly straight lines. Caimans left pools and initiated terrestrial movements spontaneously and in response to disturbance by researchers and hunters. The sex ratio of the groups was biased toward males (0.8 +/- 0.24) and was similar to that found in aquatic habitats in the study area. However, two groups consisted only of females. When caimans left pools subjected to disturbances, such as hunting and capture for research, they walked head to tail in lines. Caimans that left pools in response to disturbance buried in mud near pools or in leaf litter in forest.