Chemosphere xxx (2017) xxx-xxx
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com
Detection of a synthetic sex steroid in the American crocodile (Crocodylus acutus):
Evidence for a novel environmental androgen
Christopher M. Murraya, ∗, Mark Merchantb, Michael Easterc, Sergio Padillad, Davinia B. Garrigóse,
Mahmood Sasa Marind, f, Craig Guyerg
aDepartment of Biology, Tennessee Technological University, PO Box 5063, Cookeville, TN 38505, USA
bDepartment of Chemistry, McNeese State University, Lake Charles, LA, USA
cScales and Tails of Ohio, Lakewood, OH 44107, USA
dPalo Verde Biological Station, Organization for Tropical Studies, Guanacaste, Costa Rica
eBiodiversity and Evolutionary Biology “Cavanilles” Institute, University of Valencia, Valencia, Spain
fInstituto Clodomiro Picado, Facultad de Microbiología, Universidad de Costa Rica, San José, Costa Rica
gDepartment of Biological Sciences, Auburn University, 331 Funchess Hall, Auburn, AL 36849, USA
Received 5 December 2016
Received in revised form 3 April 2017
Accepted 4 April 2017
Available online xxx
Handling Editor: Jim Lazorchak
Endocrine disrupting contaminants
Endocrine-disrupting contaminants (EDC's) are well known to alter sexual differentiation among vertebrates via estro-
genic effects during development, particularly in organisms characterized by temperature-dependent sex determination.
However, substances producing androgenic effects typically lack potency when tested in laboratory settings and are vir-
tually unstudied in field settings. Here, we assay levels of a synthetic androgen, 17α-methyltestosterone (MT), in a heav-
ily male-biased population of American crocodiles in the Tempisque River Basin of Costa Rica based on the recent hy-
pothesis that this chemical is an EDC in developing crocodilian embryos. The presence of MT was documented in all
field-collected samples of egg yolk and in plasma of all age classes in among population of crocodiles. Hatchlings ex-
hibited higher plasma MT concentrations (102.1 ± 82.8 ng/mL) than juveniles (33.8 ± 51.5) and adults (25.9 ± 20.8 ng/
mL). Among populations, crocodiles captured in the Tempisque River (62.9 ± 73.7 ng/mL) were higher in MT concen-
tration than those from Tarcoles (13.3 ± 11.4 ng/mL) and negative controls (0.001 ± 0.0002 ng/mL). A mechanism for
the bio-transport of MT and its subsequent effects is proposed.
© 2016 Published by Elsevier Ltd.
Pollutants, referred to as endocrine-disrupting contaminants (ED-
C's), have been shown to alter endocrine function in all vertebrate
classes (Hayes et al., 2002), affecting many organ systems and asso-
ciated processes within each system. The magnitude of this anthro-
pogenic epidemic has spurred pollutant-related operational terminol-
ogy, such as bio-indicators (Hyne et al., 2009) and sentinel species
(Milnes and Guillette, 2008). Commonly, EDC's mimic estrogens or
aromatizable androgens, or act as androgen antagonists, thus biasing
secondary sexual characteristics towards a female morphology and as-
sociated physiology (Guillette et al., 1995). Crocodilian exposure to
environmental estrogens has resulted in hermaphroditic males (fol-
licular growth within seminiferous tubules) or ‘super’ females [rapid
and poorly regulated follicular development, (Guillette et al., 1994)].
To date, the phenomenon of biasing sexual differentiation using an-
drogens is known from lab settings only around pivotal temperatures,
or the temperatures at which sex differentiations changes between
Email addresses: firstname.lastname@example.org (C.M. Murray); mmerchant@mcneese.
edu (M. Merchant); email@example.com (M. Easter); spa_sergio@hotmail.
com (S. Padilla); firstname.lastname@example.org (D.B. Garrigós); msasamarin@gmail.
com (M. Sasa Marin); email@example.com (C. Guyer)
male and female (Crews et al., 1994); however, environmental andro-
gens are virtually unknown from field settings, probably as a result
of the aromatizable nature of testosterone and it's structural analogs
that contribute to rapid conversion (Wibbels and Crews, 1995). In lim-
ited fashion, paper-mill effluent was identified as having androgenic
substances associated with the masculinization of female mosquitofish
(Parks et al., 2001), providing support for the only known environ-
17α-methyltestosterone (MT) is a synthetic androgen used widely
in the commercial farming of fish (Phelps and Popma, 2000). It is
commonly applied to fish fry to divert sexual differentiation away
from female development and towards male, the more profitable sex
in most cases. The suggested aquaculture concentrations for the treat-
ment of 300,000 fry is 60 mg/kg feed (Popma and Green, 1990), al-
though the strict application of this concentration among facilities
and countries is unknown. The half-life of MT in water and soil is
short, suggesting that MT is unlikely to be an EDC. However, the
persistence of this compound in a hydrophobic environment, such as
lipids, is poorly known (Phelps and Popma, 2000; Gupta and Acosta,
2004; Murray et al., 2016a) while potential for negative ecological ef-
fects have been suggested (Mlalila et al., 2015). Use of this synthetic
androgen is a regular practice in tilapia farming, perhaps the most
prominent New World fish industry, where it is applied liberally via
feed (Gupta and Acosta, 2004). Containment regulations of the com-
pound, either by water or animal exchange with the surrounding envi
0045-6535/© 2016 Published by Elsevier Ltd.
2 Chemosphere xxx (2017) xxx-xxx
ronment, vary widely by region and detailed studies of MT fate in the
environment are not available (Phelps and Popma, 2000).
A uniquely male-biased sex ratio of American crocodiles (Croco-
dylus acutus) in the Tempisque drainage of Costa Rica has been pre-
viously described (Bolaños-Montero, 2012; Murray et al., 2015). This
is one of multiple reports that describe recent male-biased crocodil-
ian populations throughout Central America (Charruau et al., 2005;
Escobedo-Galván, 2008). Specific to the Tempisque drainage, ther-
mal data from nests demonstrate that the male-bias is not a function
of temperature effects on hatchling sex determination (Murray et al.,
2016b). Additionally, exposure of crocodilian eggs to MT generates
male hatchlings when incubated at a temperature that should produce
only female offspring (Murray et al., 2016a). The goal of this study is
to determine if MT is present in eggs as well as free-ranging hatchling,
juvenile and adult crocodiles on the North Pacific versant of Costa
Rica. In addition, a mechanism for its physiological action in the Tem-
pisque system is proposed.
2.1. Sample collection
Between January 2013 and January 2016, we hand captured Amer-
ican crocodiles at Palo Verde National Park and surrounding areas.
Juveniles and adults were captured in four locations within the Park
and three outside of it. El Humedal (10.341128°, −85.343311°), La
Bocana (10.335950°, −85.278837°), Nicaragua Lagoon (10.328573°,
−85.273901°) and Varillal Lagoon (10.403402°, −85.358544°) are all
seasonally flooded wetlands within the Park. Varillal Lagoon is in
very close proximity to the Tempisque River while the other wetlands
are all relatively equidistant and farther from the permanent Temp-
isque and Bebedero waterways (Fig. 1). We also surveyed waterway
banks of a drainage canal along the northwestern border of Palo Verde
National Park (10.393632°, −85.396221°), as well as the Bebedero
River near its junction with the Tempisque River (10.327898°,
−85.206055°), in order to find fresh nests. These banks were then
surveyed frequently enough to assure that hatchlings were captured
within days of hatching. Other locations sampled included the Tar-
coles River south of Guanacaste (9.800419°, −84.606156°). Tarcoles
was hypothesized to be a negative control population, lacking MT ex-
posure, based on its distance from aquaculture facilities. For each indi-
vidual sampled we extracted 0.5–2.5 mL of whole blood (18–27 gauge
needles, depending on the size of the individual) from the spinal vein.
Needles and syringes were flushed with 0.5% heparin solution to pre-
vent coagulation. Whole blood was kept on ice for no more than 2 h
before being centrifuged so that plasma supernatant could be removed
and frozen (−20 °C).
During the 2015 nesting season, six eggs were collected from five
nests late in the first trimester or early in the second trimester of in-
cubation. Eggs were obtained from the waterway banks of a drainage
canal along the northwestern border of Palo Verde National Park.
Canal banks were monitored multiple times per week as part of an-
other study to ensure accuracy of deposition and hatching date to
within two days. Eggs were frozen at −20 °C. Upon capture, sex was
determined for individuals via inspection of the secondary sex organ.
Individuals were diagnosed as male is the clitero-penis exhibited all
of the following characteristics: bi-lobed structure, extensive vascu-
larization, and length extending the length of the vent (Murray et al.,
Fig. 1. A. Map of general sampling localities in Costa Rica including the Palo Verde
National Park region (Area 1) and Tarcoles River [Area 2, (9.800419°, −84.606156°).
B. Map of Area 1 including egg sampling locality [A (10.393632°, −85.396221°)],
hatchling sampling localities [A, B (10.327898°, −85.206055°)], Varillal Lagoon [C,
(10.403402°, −85.358544°) Humedal Palo Verde [D, (10.341128°, −85.343311°)], La
Bocana [E, (10.335950°, −85.278837°)] and Nicaragua Lagoon [F, (10.328573°,
−85.273901°)]. Specific tilapia farm localities are not included to protect privacy of the
2.2. Sample preparation and analysis
Steroid hormones were extracted from egg yolk and plasma us-
ing a 3:2 solution (volume: volume) of ethyl acetate and hexane, re-
spectively. Samples were dried under vacuum at 25 °C and the re-
sulting residue was dissolved in 100 μL assay buffer supplemented
with 10 μL of dimethyl sulfoxide (DMSO) to encourage dissolution.
We quantified 17α-methyltestosterone using commercially available
sandwich ELISA kits (MaxSignal®methyltestosterone kit, Bioo Sci-
entific, Austin, TX). This competitive enzyme immunoassay reliably
quantifies methyltestosterone concentrations in feed, fish, shrimp and
meat from skeletal muscle, urine and serum samples with a high re-
covery rate (>80%), high sensitivity (0.1 ng/g) and low detection limit
(0.3 ng/mg). Cross-reactivity with testosterone was listed as <0.1%
and samples for 1 kit were not analyzed in duplicate. However, fif-
teen samples from kit 1 were also run in kit 2 to test the variabil-
ity within samples among kits. The average difference of sample
Chemosphere xxx (2017) xxx-xxx 3
MT concentrations between kits was 27% (3.52 ng/mL). All samples
in kit 2 were run in duplicate. Optical density was determined using
a Benchmark Plus microtiter plate spectrophotometer (Bio-Rad, Her-
cules, CA) at 450 nm. Six negative control samples were added to the
analysis to detect false positive readings. These included six captive
American alligators from McNeese State University, Lake Charles,
2.3. Statistical analysis
MT concentration data did not meet the assumptions for the
Shapiro-Wilk test of normality. A Wilcoxon signed rank test was per-
formed to assess difference in means between yolk MT concentrations
of alligator eggs given doses of MT known to cause sex reversal from
female to male (Murray et al., 2016a) and wild yolk concentrations
collected for crocodiles sampled during this study. Concentrations of
MT in blood plasma and eggs were compared among crocodile co-
horts using Kruskal-Wallis H Test. Dunn's post-hoc test was used to
elucidate differences among cohorts. Among localities, all Palo Verde
field sites were pooled and plasma MT concentrations were compared
to Tarcoles samples and negative controls via a Kruskal-Wallis H
Test. Dunn's post-hoc test was used to elucidate differences among
locations. This test determined whether exposure is spatially acute to
the Tempisque or widespread. Within cohorts, Wilcoxon signed rank
tests were used to assess statistical differences in MT concentrations
between morphological sexes. Standard deviation was used as a vari-
ability estimator and is reported in association with treatment means.
All statistical analyses were performed in program R (R Development
Core Team, 2011).
Efficiency for the extraction method above was determined by
spiking 300 μL egg yolk samples to 1, 15, and 50 ng/mL with MT.
Each sample was spiked and extracted in triplicate, and the results
compared to un-spiked samples. The samples were vortexed vigor-
ously for 1 min to ensure the absorption of the MT spike into the
yolk prior to extraction. Recovery concentrations were measured to be
0.82 ± 0.04, 14.6 ± 0.9, 56.2 ± 8.1, respectively, for 1, 15, and 50 ng/
mL of MT. Likewise, fresh plasma samples from American alligators
were spiked 1, 15, and 50 ng/mL MT. Recovery of spiked samples
were found to be 0.9 ± 0.1, 16.0 ± 2.0, 52.4 ± 4.2, respectively, in the
plasma samples. For all yolk and plasma samples that were not spiked,
the concentrations of MT were determined to be below the detection
limit of 0.1 ng/mL.
MT concentrations in yolk sampled from Crocodylus acutus eggs
in natural nests (n = 6) did not differ from concentrations known from
yolk of alligator eggs dosed with MT concentrations known to cause
masculinization in the lab (n = 7) (Fig. 2, [Murray et al., 2016a]). MT
was detected in every egg yolk sample analyzed.
Hatchlings, juveniles and adults differed significantly in plasma
MT concentrations (χ2= 29.52, DF = 3, p < 0.0001; Fig. 3). Hatch-
lings exhibited higher plasma MT concentrations (102.1 ± 82.8 ng/
mL, n = 35) than juveniles and adults, who did not differ from each
other 33.8 ± 51.5 (n = 32) and 25.9 ± 20.8 ng/mL (n = 10) respec-
tively). All cohorts were significantly higher than negative control
concentrations (0.001 ± 0.0002 ng/mL, n = 6). Individuals were iden-
tified to sex via inspection of the secondary sex organ and MT con-
centrations did not differ between males and females within any co
Fig. 2. Bar graph comparing yolk MT concentrations between experimental eggs from
Murray et al. (2016a) and wild eggs collected for this study. MT concentrations did not
differ indicating the potential for biasing sexual differentiation towards males in wild
eggs in Palo Verde. Standard deviation is used as the variability estimator.
Fig. 3. Bar graph showing variation in plasma MT concentrations among crocodile
age cohorts with associated standard deviation (χ = 29.52, df = 3, p < 0.0001). Coding
above bars indicate post-hoc Dunn's statistical difference and sample size, respectively.
hort. All plasma samples contained a measureable level of MT higher
than the highest negative control concentration and MT concentrations
of hatchlings were much greater than those in yolk of eggs producing
those hatchlings (mean = 0.04 ± 0.19 ng/g).
Concentrations of plasma MT differed significantly among collec-
tion locations (χ2= 14.3, df = 3, p = 0.002) (Fig. 4). Tempisque sam-
ples (62.9 ± 73.7 ng/mL, n = 73) were higher in MT concentration
than negative controls (0.001 ± 0.0002 ng/mL, n = 6) however not sta-
tistically different than Tarcoles (13.3 ± 11.4 ng/mL, n = 3).
Analysis of crocodile egg yolk and blood plasma from samples col-
lected in the Tempisque Basin and Tarcoles River indicated that ani-
mals experience heavy MT exposure and/or retain MT as they grow.
The presence of MT in blood plasma of juveniles and adults, how-
ever, suggests that exposure to this synthetic hormone is chronic. If
exposure to MT were acute within individuals, a steroid of this nature
would be filtered by the liver and stored (Guillette et al., 1995), mak-
ing acute past exposure of MT undetectable in plasma. Further, degra-
dation of MT in the environment is rapid (Gupta and Acosta, 2004)
potentially lending support for a bio-transporter.
4 Chemosphere xxx (2017) xxx-xxx
Fig. 4. Bar graph showing variation in plasma MT concentrations among crocodile sam-
pling localities with associated standard deviation (χ = 14.3, df = 3, p = 0.002). Coding
above bars indicate post-hoc Dunn's statistical difference and sample size, respectively.
Our observation that yolk MT concentrations do not differ between
wild eggs and those experimentally dosed with MT (Murray et al.,
2016a) suggests that Palo Verde crocodile eggs contain MT at levels
known to produce male offspring despite incubation at female-produc-
ing temperatures (Murray et al., 2016b). In fact, of 357 eggs monitored
at Palo Verde, 225 were estimated to produce females and 132 were
estimated to produce males, based on temperature alone (Murray et al.,
2016b). Data collected by Murray et al., 2016a suggests that 60% of
the females would convert to males in the presence of MT. Thus, 75%
of eggs incubated at field temperatures known from Palo Verde would
produce males in the presence of known levels of MT in the yolk. This
value approaches the 80% male bias for hatchling crocodiles recorded
at Palo Verde (Murray et al., 2015).
Concentrations of plasma MT for adults at Palo Verde do not dif-
fer from levels measured at Tarcoles samples. Initially, we thought
that crocodiles at Palo Verde were affected by tilapia farms located
just outside the Park boundary. Therefore, the Tarcoles population was
sampled to serve as an uncontaminated control site. However, our re-
sults instead demonstrate either that MT is more widespread in Costa
Rica or that Palo Verde crocodiles are dispersing south. The Tarcoles
River drains San Jose, the major industrial and residential center of
Costa Rica. It is known to be a highly polluted river, accumulating and
concentrating waste from many sources (Rainwater et al., 2011) and
may very well be accumulating MT input from other pollution sources
or smaller unknown fish farming operations. Further, restaurants and
tourists surround the site, and fish waste deposited in the river may be
frequent despite not being directly downstream from a tilapia farm.
Among cohorts, eggs collected after one-third of incubation ex-
hibit concentrations of MT that are barely detectable, while hatch-
lings exhibit extremely high concentrations of MT in their blood
plasma without sufficient time for exposure after hatching. Juveniles
and adults exhibit lower MT concentrations than hatchlings. This may
be indicative of less frequent exposure and/or more rapid utiliza-
tion or storage of this exogenous androgen in growing or reproduc-
tively mature individuals. Based on these results we propose a likely
mechanism for chronic crocodile exposure to MT in the Tempisque
Basin. Initially, tilapia bio-accumulate MT in adipose tissue, a com-
mon storehouse for cholesterol-based exogenous steroids (Guillette
et al., 1995), from the supplied feed. Retention of MT in exposed
tilapia has been documented at 67 ng/g fish for ≥21 days (Goudie
et al., 1986), a potent concentration for endocrine-disrupting effects
(Murray et al., 2016a). Since the escape of tilapia from farms to sur-
rounding natural waterways is observable and tilapia are now com-
mon the Tempisque Basin, and crocodiles readily move in and out of
tilapia farm ponds, tilapia has become an available food source for
crocodiles in the system. Crocodiles, like the tilapia, sequester the MT
acquired via the fish consumption, and must also use adipose tissue
for storage after plasma circulation. Some processes during folliculo-
genesis likely mobilize MT stored in adipose tissue during the prepa-
ration and deposition of yolk, allowing deposit of MT in eggs (Paitz
and Bowden, 2010). Such processes have been suggested for andro-
gens (Guillette et al., 1995) and the detectability of MT in hatchlings
and documented masculinizing effects (Murray et al., 2016a) imply
that MT is not aromatizable in Alligator mississippiensis or Crocody-
lus acutus, or is aromatized at a low rate. Supplied doses are known
to bias sexual differentiation in crocodilians (Murray et al., 2016a)
and act as an environmental androgen within the eggs in the Temp-
isque basin, resulting in the male-biased sex ratio previously recov-
ered (Murray et al., 2015).
Of interest is the low concentration of MT detected in egg yolk and
high concentration detected in hatchling plasma before the hatchlings
were large enough to consume tilapia. The persistence of MT in water
and soil is on the order of hours or days (Phelps and Popma, 2000),
so environmental exposure without a bio-transporter is unlikely. Be-
cause the concentrations of MT recovered in wild eggs in this study is
comparable to the concentrations of experimentally-dosed crocodilian
eggs from (Murray et al., 2016a), it is likely that much MT persists
within developing eggs but is unrecognizable in our analysis. Paitz
and Bowden (2011) and Paitz et al. (2012) posit that maternally-de-
rived steroid hormones (in their case, estradiol) are conjugated to a
water-soluble sulphates in painted turtles, a reptilian species that also
exhibits TSD. Such conjugates are transported from yolk to embryo
and utilized later in development, while the intended embryonic re-
sponse to the modified steroid is the same as if it lacked conjugation.
When experimentally treated, conjugation of sex steroid hormones can
occur within hours (Paitz and Bowden, 2015). Such a process would
make MT undetectable to our analyses but result in high hatchling
concentrations when (and if) the steroid is biologically reactivated.
Alternatively, it may be a result of low extraction efficiency of MT
from the hydrophobic environment of the yolk. However, this seems
very unlikely given that the extraction of androgens from avian eggs
has been shown to be efficient with the use of hexane as a solvent
(Von Engelhardt and Groothuis, 2005).
Results presented here, in combination with previous experimental
studies, suggest that 17α-methyltestosterone (MT) acts as a non-arom-
atizable environmental androgen on crocodiles in the Tempisque and,
likely, surrounding basins in Costa Rica and is responsible for a
male-biased sex ratio. Interestingly, because the short environmen-
tal half-life of MT limits exposure to bio-transport and accumulation,
and MT concentrations are highly detectable in juveniles and adults
as well as newborn hatchlings, MT likely follows a rare pathway for
maternally supplied exogenous sex steroids or their mimics in devel-
oping embryos. Further hypothesis testing regarding the physiologi-
cal mechanism of maternal supply, and bio-reactivation during late
development, is needed. Specifically, analysis of MT concentration
among yolking follicles within female crocodiles is critical to our pro-
posed mechanism. Additionally, study of the travel and conjugation
of MT within eggs among yolk, albumin, and embryo is needed. En-
vironmentally, it is necessary to quantify MT in tilapia and other po-
tential crocodile food sources to isolate the mechanism of bioaccu
Chemosphere xxx (2017) xxx-xxx 5
mulation. Lastly, the sex ratios and endocrine profiles of other verte-
brates in the area require quantification to fully understand the breadth
and mechanism of effects of this environmental androgen. If the pro-
posed mechanism is isolated to diet, then predatory fish and piscivorus
birds may likely be affected, either in sex ratio and/or endocrine pro-
files. However, if MT is widely present in the food chain then one may
expect a male-bias in the TSD mud turtle population or other aquatic
reptiles and amphibians. Testing of such hypotheses is critical in falsi-
fying the proposed model.
We thank the Palo Verde Biological Station staff and MINAET
for logistic support and permitting as well as J. Bolaños for guidance
and logistic support. We also acknowledge J. Goessling for proof-
reading and A. Cooper for logistic support and thank Vicerrectoría
de Investigación Universidad de Costa Rica VI 741-B5-270 and the
Organization for Tropical Studies for permitting assistance and fund-
ing (OTS Fund 507). The authors declare they have no actual or poten-
tial competing financial interests. Institutional guidelines for the care
and use of animals were followed under approved IACUC proposal
Bolaños-Montero, J.R., 2012. American crocodile (Crocodylus acutus) (Crocodylia:
Crocodylidae) (Cuvier1807) population status in the Great Tempisque wetland. In:
Crocodiles. Proceedings of the 21st Working Meeting of the IUCN-SSC Crocodile
Specialist Group. IUCN, Gland, Switzerland, pp. 167–178.
Charruau, P., Rogelio Cedeno-Vasquez, J., Calme, S., 2005. Status and conservation of
the American crocodile (Crocodylus acutus) in Banco Chinchorro biosphere re-
serve, Quintana Roo, Mexico. Herpetol. Rev. 36, 390–395.
Crews, D., Bergeron, J.M., Bull, J.J., Flores, D., Tousignant, A., Skipper, J.K.,
Wibbels, T., 1994. Temperature-dependent sex determination in reptiles: proxi-
mate mechanisms, ultimate outcomes, and practical applications. Dev.
Genet. 15 (3), 297–312.
Escobedo-Galván, A.H., 2008. Estructura poblacional y proporción de sexos en
Caiman crocodilus en Caño Negro, Costa Rica. Iheringia Sér. Zool. 98 (4),
Goudie, C.A., Shelton, W.L., Parker, N.C., 1986. Tissue distribution and elimination of
radiolabelled methyltestosterone fed to adult blue tilapia. Aquacul-
ture 58, 227–240.
Guillette Jr., L.J., Crain, D.A., Rooney, A.A., Pickford, D.B., 1995. Organization ver-
sus activation: the role of endocrine-disrupting contaminants (EDCs) during em-
bryonic development in wildlife. Environ. Health Perspect. 103 (Suppl. 7),
Guillette Jr., L.J., Gross, T.S., Masson, G.R., Matter, J.M., Percival, H.F., Woodward,
A.R., 1994. Developmental abnormalities of the gonad and abnormal sex hormone
concentrations in juvenile alligators from contaminated and control lakes in
Florida. Environ. Health Perspect. 102 (8), 680–688.
Gupta, M.V., Acosta, B.O., 2004. A review of global tilapia farming practices. Aquac.
Asia 9, 7–16.
Hayes, T.B., Collins, A., Lee, M., Mendoza, M., Noriega, N., Stuart, A.A., Vonk, A.,
2002. Hermaphroditic, demasculinized frogs after exposure to the herbicide
atrazine at low ecologically relevant doses. Proc. Natl. Acad. Sci. U. S. A. 99 (8),
Hyne, R.V., Wilson, S., Byrne, M., 2009. Frogs as bioindicators of chemical usage and
farm practices in an irrigated agricultural area. Final Rep. Land Water Aust.
Milnes, M.R., Guillette, L.J., 2008. Alligator tales: new lessons about environmental
contaminants from a sentinel species. Bioscience 58 (11), 1027–1036.
Mlalila, N., Mahika, C., Kalombo, L., Swai, H., Hilonga, A., 2015. Human food safety
and environmental hazards associated with the use of methyltestosterone and other
steroids in production of all-male tilapia. Environ. Sci. Pollut. Res. 22 (7),
Murray, C.M., Easter, M., Merchant, M., Rheubert, J.L., Wilson, K.A., Cooper, A.,
Mendonça, M., Sasa Marin, M., Guyer, C., 2016a. Methyltestosterone alters sex
determination in the American alligator (Alligator mississippiensis). Gen. Comp.
Endocrinol. 236, 63–69.
Murray, C.M., Easter, M., Padilla, S., Sasa Marin, M., Guyer, C., 2016b. Regional
warming and the thermal regimes of American crocodile nests in the Tempisque
basin, Costa Rica. J. Therm. Biol. 60, 49–59.
Murray, C.M., Easter, M., Padilla, S., Garrigós, D.B., Stone, J.A., Bolaños-Montero, J.,
Sasa, M., Guyer, C., 2015. Cohort-dependent sex ratio biases in the American
crocodiles (Crocodylus acutus) of the Tempisque basin. Copeia 103 (3), 541–545.
Paitz, R.T., Bowden, R.M., 2015. The in ovo conversion of oestrone to oestrone sulfate
is rapid and subject to inhibition by Bisphenol A.. Biol. Lett. U. K. 11, 20140946.
Paitz, R.T., Sawa, A.R., Bowden, R.M., 2012. Characterizing the metabolism and
movement of yolk estradiol during embryonic development in the red-eared slider
(Trachemys scripta). Gen. Comp. Endocrinol. 176 (3), 507–512.
Paitz, R.T., Bowden, R.M., 2011. Biological activity of oestradiol sulphate in an
oviparous amniote: implications for maternal steroid effects. Proc. R. Soc.
B 278, 2005–2010.
Paitz, R.T., Bowden, R.M., 2010. Progesterone metabolites, ‘xenobiotic-sensing’ nu-
clear receptors, and the metabolism of maternal steroids. Gen. Comp. En-
docrinol. 166, 217–221.
Parks, L.G., Lambright, C.S., Orlando, E.F., Guillette Jr., L.J., Ankley, G.T., Gray,
L.E., 2001. Masculinization of female mosquitofish in kraft mill effluent-contami-
nated Fenholloway River water is associated with androgen receptor agonist activ-
ity. Toxicol. Sci. 62 (2), 257–267.
Phelps, R.P., Popma, T.J., 2000. Sex reversal of tilapia. In: Costa-Pierce, B.-A.,
Rakocy, J.-E. (Eds.), Tilapia Aquaculture in the Americas. The World Aquaculture
Society Baton Rouge, Louisiana, United States, p. 2. 34–59.
Popma, T.J., Green, B.W., 1990. Sex Reversal of Tilapia in Earthen Ponds. Interna-
tional Center for Aquaculture Research and Development Series No. 35.
R Development Core Team, 2011. R: a Language and Environment for Statistical
Computing. R Foundation for Statistical Computing, Vienna, Austria.
Rainwater, T.R., Millichamp, N.J., Barrantes, L.D., Barr, B.R., Bolaños-Montero, J.R.,
Platt, S.G., Abel, M.T., Cobb, G.P., Anderson, T.A., 2011. Ocular disease in
American crocodiles (Crocodylus acutus) in Costa Rica. J. Wildl. Dis. 47 (2),
Von Engelhardt, N., Groothuis, T., 2005. Measuring steroid hormones in avian eggs.
Ann. N. Y. Acad. Sci. 1046 (1), 181–192.
Wibbels, T., Crews, D., 1995. Steroid-induced sex determination at incubation temper-
atures producing mixed sex ratios in a turtle with TSD. Gen. Comp. En-
docr. 100 (1), 53–60.