Cohort-Dependent Sex Ratio Biases in the American Crocodiles (Crocodylus acutus) of the Tempisque Basin

Abstract

A male-biased sex ratio of 3:1 has been reported for a population of American crocodiles (Crocodylus acutus) in the Tempisque River Basin, Guanacaste, Costa Rica. If confirmed, this would constitute one of the largest male-biased sex ratios reported for any population of a member of the genus Crocodylus. Here, we examine the aforementioned population of C. acutus and report on sex ratios of hatchling, juvenile, and adult age classes within a sample of 474 crocodiles captured in the Tempisque Basin between May 2012 and June 2014. Hatchling sex ratio is exceptionally male biased (3.5:1), an imbalance that is maintained in juveniles but is reduced in adults (1.5:1). Mark–recapture data document that juvenile males disperse from the study site, potentially to avoid competition, a process that reduces male bias in the adult age class. An increased role of males in human–crocodile conflict may be a result of juvenile males dispersing to human-inhabited areas.

Full text

Cohort-Dependent Sex Ratio Biases in the American Crocodiles
(Crocodylus acutus) of the Tempisque Basin
Christopher M. Murray
1
, Michael Easter
2
, Sergio Padilla
3
, Davinia B. Garrigo´s
4
,
Julia Ann Stone
1
, Juan Bolan˜os-Montero
5
, Mahmood Sasa
6
, and Craig Guyer
1
A male-biased sex ratio of 3:1 has been reported for a population of American crocodiles (Crocodylus acutus) in the
Tempisque River Basin, Guanacaste, Costa Rica. If confirmed, this would constitute one of the largest male-biased sex
ratios reported for any population of a member of the genus Crocodylus. Here, we examine the aforementioned
population of C. acutus and report on sex ratios of hatchling, juvenile, and adult age classes within a sample of 474
crocodiles captured in the Tempisque Basin between May 2012 and June 2014. Hatchling sex ratio is exceptionally male
biased (3.5:1), an imbalance that is maintained in juveniles but is reduced in adults (1.5:1). Mark–recapture data
document that juvenile males disperse from the study site, potentially to avoid competition, a process that reduces
male bias in the adult age class. An increased role of males in human–crocodile conflict may be a result of juvenile males
dispersing to human-inhabited areas.
SEX allocation and sex ratio theory have been a topic of
experimental and theoretical discussion among ecol-
ogists and evolutionary biologists for over a century
(Darwin, 1871; Du¨sing, 1883; Fisher, 1930). Critical to the
viability of sexually reproducing organisms is an adequate
proportion of both sexes, most classically modeled by Fisher
(1930). This model relies on competition for mates based on
the number of individuals of each sex and associated mating
prospects. Parents that produce the minority sex will have
increased fitness based on number of grandchildren pro-
duced and the minority sex will be selected for, thus
maintaining sex ratio equilibrium (Fisher, 1930). Numerous
variables, such as sex-linked drive (Hamilton, 1967), alter-
native fitness influences (Charnov et al., 1981), dispersal
(Bulmer, 1986), longevity (Eberhardt, 2002), operational
sampling (Gibbons, 1990), and sibling interaction (Uller,
2006) have been added to the basic Fisherian model to
explain unusual sex ratios observed in nature.
For organisms exhibiting environmental sex determina-
tion (ESD), abiotic cues determine patterns of steroid
production, which canalizes development toward one gender
or another (Valenzuela and Lance, 2004). Temperature-
dependent sex determination (TSD), the most common
mechanism of ESD, has obvious relevance to sex allocation
theory. TSD is likely selected for when the environment
produces the sex with the highest fitness in a highly variable
environment (Charnov and Bull, 1977; Warner and Shine,
2008). Further, temperature is thought to lead to differential
growth rates between sexes and, thus, different fitness
consequences (Bull and Charnov, 1989). However, sex ratios
of organisms with TSD are less self-correcting than those with
genotypic sex determination so that organisms with TSD may
experience altered sex ratios that undermine demographic
viability, a feature that may make them susceptible to
anthropogenic change (Doody et al., 2006).
All extant crocodilians exhibit TSD and members of the
genus Crocodylus produce viable offspring between 25 and
35uC, with males generated between 31 and 33.5uC and
females produced at all other temperatures in the viable
range (Lang and Andrews, 1994). With such a narrow male-
specific thermal window, it has been suggested that male
bias in clutches of crocodiles is difficult to produce in nature
(Thorbjarnarson, 1997; Lance et al., 2000). However, male-
biased sex ratios are known in crocodilian populations
(Thorbjarnarson, 1997). For example, American crocodile
(Platt and Thorbjarnarson, 2000) and Morelet’s crocodile
populations (Rainwater et al., 1998) in Belize as well as
a population of the American crocodile (Crocodylus acutus)
in the Tempisque Basin of Guanacaste, Costa Rica (3:1 male
bias; Bolan˜os-Montero, 2012) exhibit some of the strongest
male biases documented. Similarly, Charruau (2012) re-
ported a male bias in hatchling C. acutus in Banco
Chinchorro Biosphere Reserve, Mexico. However, the study
by Bolan˜os-Montero (2012) was based on limited sampling
in the Tempisque watershed and a biased sex ratio was not
found by Sa´nchez-Ramı´rez (2001) for the same area.
Additionally, these previous studies were based on raw
counts (observed ratios of captures animals) of individuals
detected rather than accounting for segments of the
population that were undetected. Therefore, the aim of this
study was to use mark–recapture techniques to estimate
cohort-specific population sizes of each sex independently,
compare such estimates, and assess whether a male-biased
sex ratio characterizes the population of C. acutus in the
Tempisque River Basin.
MATERIALS AND METHODS
Study area.—All sampling for this study was performed in
Palo Verde National Park and adjacent localities in the
Guanacaste Province, in northwestern Costa Rica. Seven
sampling localities in the Tempisque River basin were used;
the Humedal, the Drain, the Bebedero River, Varillal
Lagoon, La Bocana, Nicaragua Lagoon, and the Tower Ponds
(Fig. 1). All sampling localities were restricted to the areas
1
Department of Biological Sciences, Auburn University, 331 Funchess Hall, Auburn, Alabama 36849; E-mail: (CMM) cmm0054@auburn.edu;
and (CG) guyercr@auburn.edu. Send reprint requests to CMM.
2
Everglades Holiday Park, 21940 Griffin Rd., Fort Lauderdale, Florida 33332; E-mail: m.easter05@gmail.com.
3
Palo Verde Biological Station, Organization for Tropical Studies, Guanacaste, Costa Rica; E-mail: sergio.padilla@ots.ac.cr.
4
Biodiversity and Evolutionary Biology ‘‘Cavanilles’’ Institute, University of Valencia, Valencia, Spain; E-mail: davinia.beneyto@uv.es.
5
Grupo Aqua Corporacion Internacional SA, Can˜ as, Guanacaste, Costa Rica; E-mail: crocacutus@gmail.com.
6
Instituto Clodomiro Picado, Facultad de Microbiologı´a, Universidad de Costa Rica, San Jose´, Costa Rica; E-mail: msasamarin@gmail.com.
Submitted: 3 November 2014. Accepted: 11 April 2015. Associate Editor: D. S. Siegel.
F2015 by the American Society of Ichthyologists and Herpetologists DOI: 10.1643/CE-14-186 Published online: July 31, 2015
Copeia 103, No. 3, 2015, 541–545

within or immediately surrounding Palo Verde National
Park and are part of the same Tempisque River Basin
population. The Humedal, Varillal Lagoon, La Bocana,
Nicaragua Lagoon, and Tower Ponds are seasonal wetlands
characterized by open expanses of Typha,Thalia, and
Eichhornia in the wet season (May to December). These
areas are void of water during the dry season (December to
May) when they are characterized by cracked mud with
patches of vegetation. The Drain is a permanent canal that
drains irrigation water from rice fields north of the National
Park to the Tempisque River 10 km to the southwest. This
canal is influenced both by tide and agricultural discharge.
The Bebedero River is a major tributary to the Tempisque
River that extends to the northeast from the Tempisque
River to near the town of Can˜as, Guanacaste Province, Costa
Rica. This river is tidally influenced toward the mouth where
it meets the Tempisque River 5 km north of the Gulf of
Nicoya.
Crocodile sampling.—American crocodiles were captured in
the Tempisque Basin between May 2012 and June 2014.
Crocodiles were captured by hand, breakaway snare pole, or
top-jaw rope. Sex (by cloacal examination of the genitalia),
snout to vent length (SVL), and total length (TL) were
recorded followed by creation of a permanent mark via tail
scute removal. Individuals were categorized into a cohort
based on size: hatchling (,35 cm), juvenile (35–180 cm),
and adult (.180 cm; modified from Platt and Thorbjarnar-
son, 2000). A sub-adult size class was not included because
any size cut-off between the juvenile and sub-adult cohort
was deemed an artificial threshold.
Sampling of hatchlings included 13 clutches (204
hatchlings) within which all individuals were known to
have been captured because we monitored the nests from
which they emerged and the number of viable eggs was
matched to the number of hatchlings captured (n56) or
because we discovered aggregated individuals soon after
hatching from unmonitored nests. The risk of hatchling
predation and capture detection existed at unmonitored
nests, but was minimized by our timing. We captured two,
seven, and four clutches in 2012, 2013, and 2014,
respectively. Two clutches in 2013 were captured in the
Bebedero River while the rest were captured along a 10 km
stretch of drainage canal along the northwest border of the
national park. Because hatchlings remain aggregated at nest
sites immediately after hatching and because we accounted
for all individuals known to have hatched from monitored
nests, we assume that sex ratios based on raw counts of
hatchlings are not biased by undetected individuals emerging
from unmonitored nests. A hatchling crocodilian was de-
termined to be male if the clitero-penis possessed all of the
following character states: bi-lobed structure, extensive
vascularization, and length extending the length of the vent
(Fig. 2; Allsteadt and Lang, 1995).
For juveniles we used Bailey’s triple catch method (Bailey,
1951, 1952) to estimate gender-specific population size and,
thus, sex ratio. This method allowed us to estimate the
number of males and females by adjusting for individuals
not detected during our sampling. This algorithm assumes
an open population with variable gain and survival and
requires three sampling events with at least 20 captures per
event (Donnelly and Guyer, 1994). Each sampling event
represented a complete survey of the area, with one event
occurring in 2013 (January–May) and two events occurring
in 2014 (January–March and April–June). Separate estimates
for population size, loss rate, and gain rate were calculated
for male and female juveniles. Because the sample of adults
yielded few recaptures, unlike the juvenile cohort, we used
raw counts to characterize sex ratio of this cohort.
A G-test was used to assess differences in proportion of
males among cohorts and to test deviation from a 1:1 sex
ratio. A runs test for non-randomness was used to assess any
bias in the pattern of capture of males and females in the
field.
RESULTS
A total of 474 American crocodiles were captured during our
study. Analysis for raw count data provides G-test results
that indicate a significant difference in the sex ratio among
cohorts (G 550.828, df 52, P5,0.01). The hatchling
cohort exhibits a sex ratio of nearly 80%male, while
juvenile and adult cohorts both exhibit 60%male sex ratios.
Across all cohorts we recover a 2.2:1 male-biased sex ratio.
When data for juveniles were adjusted for undetected
individuals, we recovered a 3.4:1 male-biased sex ratio
overall, an imbalance that differs from the expected 1:1
sex ratio (G 5103.05, df 51, P5,0.001). The sex ratio
differed marginally among cohorts (G 55.447, df 52, P5
0.07; Fig. 3). The hatchling and juvenile cohorts consisted of
nearly 80%males, while the adult cohorts consisted of
approximately 60%males.
Our juvenile sex ratio was based on population estimates
associated with the second capture interval (N
1
; Table 1).
These values differ markedly from the female-biased pattern
indicated by juvenile raw counts. Vital demographic rates of
change for juvenile females indicated a positive growth rate
for this component of the population, with gains due to the
combined effects of growth of hatchlings and immigration
being sufficient to counteract losses due to the combined
effects of mortality and emigration. Vital rates of change for
juvenile males indicated a negative growth rate for this
component of the population, with gains due to the
combined effects of growth of hatchlings and immigration
being insufficient to counteract losses due to the combined
effects of mortality and emigration. Losses were more similar
for male and female juveniles than were gains, which were
Fig. 1. Map of the study area in Guanacaste, Costa Rica. Region shown
includes Palo Verde National Park and the Tempisque River Delta.
Seven sampling areas are noted: (Drain [A], Varialle Lagoon [B],
Humedal [C], La Bocana [D], Nicaragua Lagoon [E], Tower ponds [F], Rio
Bebedero [G]).
542 Copeia 103, No. 3, 2015

reduced in males relative to females (Table 1). We found no
significant runs of either sex in our capture records (Z 5
–0.447, P50.327; Table 1); thus, non-random sampling of
sexes is rejected.
DISCUSSION
Our data document a male-biased sex ratio in the Palo Verde
National Park (Tempisque Basin) population of Crocodylus
acutus. The extent of this bias is extreme at hatching,
becoming reduced for adults. Our data for juveniles provide
evidence for how such an extreme male bias is adjusted
over time. Of particular interest is our documentation that
juvenile males exhibit a negative population growth rate,
likely caused by dispersal of this stage from the Tempisque
sampling localities, and no immigration of juvenile males to
the population. Juvenile females, apparently, have balanced
rates of immigration and emigration that maintain a positive
population growth rate for this segment of the population.
Based on raw counts of individuals captured, juvenile
females were more numerous than juvenile males. This
suggests either that migration of juvenile males is extensive
and immediate, that males are more adept at avoiding
capture, or that both processes are operating. Our data for
adults are limited because recaptures are too low to account
for undetected animals and the important parameters of an
open population. However, if our raw counts accurately
reflect the real adult sex ratio, then increased dispersal or
mortality of males beyond the juvenile stage does not
appear to be occurring. If anything, immigration of
juveniles to the study site to retain a male bias in detected
adults may be indicated by our data.
The extent of male bias in our hatchling cohort is
unprecedented in crocodilian populations. Charruau
(2012) reported male-biased sex ratios at birth for American
crocodiles at a site in Mexico, but there only 66%of
hatchlings were male, a percentage closer to that of adults at
the Tempisque population. Because the sex of our hatch-
lings was determined by visual inspection of the relative size
of developing genitalia, our extreme male bias might result
from misidentification of sex. However, we argue that this is
unlikely, based on a sample of ten hatchlings that were
Fig. 2. Clitero-penis (CTP) examples depicting conservative sex determination of hatchlings based on the number of lobes, vascularization, and
relative length. Female exhibiting small vascularized nub (A); female exhibiting small non-vascularized nub (B); female exhibiting single-lobed
vascularized projection (C); male exhibiting bi-lobed vascularized projection that extends the length of the cloacal opening (D, E).
Murray et al.—Crocodile sex ratio biases 543

identified as to sex at birth and then raised for a year at the
Palo Verde Biological Station. When re-examined a year
later, none of these individuals changed in status. The
extreme male bias at hatching might also result from
a biased sample of nests. However, hatchlings were captured
within a short time following emergence from their nests
during a time when nest mates remain aggregated. For
monitored nests the number of hatchlings captured corre-
sponded to the number of viable eggs known to be in the
nests. We assume this to also be true of unmonitored nests.
Further, the temporal and spatial disparity of the clutches
analyzed accounts for climatic bias within years as well as
spatial thermal variation between habitats. Therefore, the
only way that our count of hatchling males and females is
biased is if the pool of undetected nests produced a sex ratio
differing from our detected nests. Additionally, the window
of nest temperatures yielding male American crocodiles is
quite narrow (Thorbjarnarson, 1997; Lance et al., 2000),
making it unlikely that we selected nests within this narrow
window and missed others outside this range.
The demographic patterns that we document for juvenile
males and females are consistent with interpretations
inferred for other crocodilians. The territoriality of adult
male crocodilians has been used to suggest that juvenile
males disperse from the natal population to avoid compe-
tition (Thorbjarnarson, 1989; Platt and Thorbjarnarson,
2000), features consistent with our estimates of gain and
loss rates for juveniles (Fig. 4).
Across the entire population of Tempisque crocodiles, we
recovered a male bias that is as skewed as that reported by
Bolan˜os-Montero (2012) and far more skewed than that
reported by Sa´nchez-Ramı´rez (2001). The sampling of
Bolan˜os-Montero (2012) and Sa´ nchez-Ramı´rez (2001) may
have been inadvertently susceptible to location-based biases
or biases resulting from undetected individuals. Here, we
present increased sampling and mark–recapture analysis to
estimate animals not detected. The difference in observed
versus estimated sex ratios of juveniles is extreme and
renders strict observation-based counts unreliable. Dispersal
of juvenile males from the core sampling areas would have
likely resulted in an underestimation of males in the
population based on strict count sampling.
Negative perceptions of crocodiles have reached a critical
level in Costa Rica as a result of increased media coverage of
crocodile/human conflicts and/or an actual statistical in-
crease in attacks by crocodiles on humans (Valdelomar et al.,
2012). A unique male bias, to the extent described by
Bolan˜os-Montero (2012), is likely to contribute to this
negative perception. The demographics discussed here
indicate an expanding population, from a spatial perspec-
tive, that may lead to increased overlap with human-
inhabited areas and potential for human-induced conflict
such as feeding (Rainwater et al., 2011). Furthermore, if this
problem persists for generations then selection for more
aggressive males based on heightened reproductive compe-
tition may occur. Of utmost importance is an organized
collaborative effort to assess the demographics of Crocodylus
acutus along the Pacific versant of Costa Rica prior to
management action.
ACKNOWLEDGMENTS
We thank M. Mendonc¸a, M. Merchant, and T. Wibbels for
guidance and review, T. Connors, A. Blanco, V. Salvatico, F.
Bonilla, S. Bermudez, and J. Serrano for logistic support, and
MINAET, Vicerrectorı´a de Investigacio´ n Uni versidad de
Costa Rica, and the Organization for Tropical Studies for
permitting assistance and funding.
Table 1. Populations estimates, birth/immigration rates, death/
emigration rates, and population growth rates of males and females
in the juvenile cohort based on Bailey’s Triple Catch algorithm (top).
Bailey’s Triple Catch Males Females
Population size at time 1 (N
1
) 3106208.2 89671.8
Raw count 98 117
Birth/immigration rate (B
12
) 0.322 1.241
Death/emigration rate (D
01
) 0.565 0.752
Instantaneous birth/immigration
rate (b
12
)
–1.13 0.216
Instantaneous death/emigration
rate (d
01
)
0.833 1.39
Population growth rate (r) –0.586 0.073
Fig. 3. Cohort-specific sex ratios with juvenile cohort based on triple
catch algorithm estimates. Sample sizes do not include recaptures. G-
test results show independence in male frequency among cohorts (G 5
50.828, df 52, P5,0.01).
Fig. 4. Schematic depicting proposed demographics among cohorts of
the two sexes, showing emigration of juvenile males from
the population.
544 Copeia 103, No. 3, 2015

LITERATURE CITED
Allsteadt, J., and J. W. Lang. 1995. Sexual dimorphism in
the genital morphology of young American alligators,
Alligator mississippiensis. Herpetologica 51:314–325.
Bailey, N. T. 1951. On estimating the size of mobile
populations from recapture data. Biometrika 38:293–306.
Bailey, N. T. 1952. Improvements in the interpretation of
recapture data. The Journal of Animal Ecology 21:120–127.
Bull, J. J., and E. L. Charnov. 1989. Enigmatic reptilian sex
ratios. Evolution 43:1561–1566.
Bulmer, M. G. 1986. Sex ratio theory in geographically
structured populations. Heredity 56:69–73.
Bolan˜ os-Montero, J. R. 2012. American crocodile (Crocody-
lus acutus) (Crocodylia: Crocodylidae) (Cuvier 1807)
population status in the Great Tempisque Wetland,
p. 167–178. In: Proceedings of the 21
st
Working Meeting
of the IUCN SSC Crocodile Specialist Group. IUCN, Gland,
Switzerland.
Charnov, E. L., and J. Bull. 1977. When is sex environ-
mentally determined? Nature 266:828–830.
Charnov, E., R. L. Los-den Hartogh, W. T. Jones, and J.
Van den Assem. 1981. Sex ratio evolution in a variable
environment. Nature 289:27–33.
Charruau, P. 2012. Microclimate of American crocodile
nests in Banco Chinchorro biosphere reserve, Mexico:
effect on incubation length, embryos survival and hatch-
lings sex. Journal of Thermal Biology 37:6–14.
Darwin, C. 1871. The Descent of Man and Selection in
Relation to Sex. John Murray, London, England.
Donnelly, M. A., and C. Guyer. 1994. Estimating popula-
tion size, p. 183–205. In: Measuring and Monitoring
Biological Diversity: Standard Methods for Amphibians.
W. R. Heyer, M. A. Donnelly, R. W. McDiarmid, L. C.
Hayek, and M. S. Foster (eds.). Smithsonian Institution
Press, Washington, D.C.
Doody, J. S., E. Guarino, A. Georges, B. Corey, G. Murray,
and M. Ewert. 2006. Nest site choice compensates for
climate effects on sex ratios in a lizard with environmental
sex determination. Evolutionary Ecology 20:307–330.
Du¨sing, K. 1883. Die Factoren, welche die Sexualita¨t
entscheiden. Jenaische Zeitschrift fu¨r Naturwissenschaf-
ten 16:428–464.
Eberhardt, L. L. 2002. A paradigm for population analysis of
long-lived vertebrates. Ecology 83:2841–2854.
Fisher, R. A. 1930. The Genetical Theory of Natural
Selection. Clarendon Press, Oxford, England.
Gibbons, J. W. 1990. Sex ratios and their significance
among turtle populations, p. 171–182. In: Life History and
Ecology of the Slider Turtle. Smithsonian Institution Press,
Washington, D.C.
Hamilton, W. D. 1967. Extraordinary sex ratios. Science
156:477–488.
Lance, V. A., R. M. Elsey, and J. W. Lang. 2000. Sex ratios
of American alligators (Crocodylidae): male or female
biased? Journal of Zoology 252:71–78.
Lang, J. W., and H. V. Andrews. 1994. Temperature-
dependent sex determination in crocodilians. The Journal
of Experimental Zoology 270:28–44.
Platt, S. G., and J. B. Thorbjarnarson. 2000. Status and
conservation of the American crocodile, Crocodylus acutus,
in Belize. Biological Conservation 96:13–20.
Rainwater, T. R., N. J. Millichamp, L. D. B. Barrantes, B. R.
Barr, J. R. B. Montero, S. G. Platt, M. T. Abel, G. P. Cobb,
and T. A. Anderson. 2011. Ocular disease in American
crocodiles (Crocodylus acutus) in Costa Rica. Journal of
Wildlife Diseases 47:415–426.
Rainwater, T. R., S. G. Platt, and S. T. McMurry. 1998.
A population study of Morelet’s crocodile (Crocodylus
moreletii) in the New River watershed of northern Belize,
p. 206–220. In: Crocodiles. Proceedings of the 14
th
Working Meeting of the Crocodile Specialist Group,
IUCN, The World Conservation Union, Gland, Switzer-
land and Cambridge, U.K.
Sa´ nchez-Ramı´rez, J. 2001. Estado de la poblacio´n de
cocodrilos (Crocodylus acutus)enelrı´o Tempisque,
Guanacaste, Costa Rica. INBIO, Heredia, Costa Rica.
Thorbjarnarson, J. 1989. Ecology of the American crocodile
(Crocodylus acutus), p. 228–258. In: Crocodiles: Their
Ecology, Management, and Conservation. P. M. Hall
(ed.). IUCN, The World Conservation Union Publications,
Gland, Switzerland.
Thorbjarnarson, J. 1997. Are crocodilian sex ratios female
biased? The data are equivocal. Copeia 1997:451–455.
Uller, T. 2006. Sex-specific sibling interactions and
offspring fitness in vertebrates: patterns and implica-
tions for maternal sex ratios. Biological Reviews
81:207–217.
Valdelomar, V., M. A. Ramı´rez-Vargas, S. G. Quesada-
Acun˜ a, C. Arrieta, I. Carranza, G. Ruiz-Morales, S.
Espinoza-Bolan˜ os, J. M. Mena-Villalobos, C. Brizuela,
L. Miranda-Fonseca, M. Matarrita-Herrera, J. Gonza´lez-
Venegas, E. Calderon-Sancho, J. F. Araya, A. Sauma-
Rossi, I. Sandoval-Herna´ ndez, and A. Go´ mez-Le´piz.
2012. Percepcio´ n y conocimiento popular sobre el coco-
drilo Crocodylus acutus (Reptilia: Crocodylidae) en zonas
aledan˜as al rı´o Tempisque, Guanacaste, Costa Rica. Re-
search Journal of the Costa Rican Distance Education
University 4:191–202.
Valenzuela, N., and V. Lance (Eds.). 2004. Temperature-
Dependent Sex Determination in Vertebrates. Smithso-
nian Institution, Washington, D.C.
Warner, D. A., and R. Shine. 2008. The adaptive signifi-
cance of temperature-dependent sex determination in
a reptile. Nature 451:566–568.
Murray et al.—Crocodile sex ratio biases 545