Final Report: Exposure and Response of Morelet's Crocodile (Crocodylus moreletii) Populations to Endocrine Disrupting Compounds in Belize, Central America

EPA Grant Number: R826310
Title: Exposure and Response of Morelet's Crocodile (Crocodylus moreletii) Populations to Endocrine Disrupting Compounds in Belize, Central America
Investigators: McMurry, Scott T. , Anderson, Todd A.
Institution: Towson University
EPA Project Officer: Carleton, James N
Project Period: December 1, 1997 through November 30, 2000 (Extended to December 31, 2002)
Project Amount: $159,788
RFA: Endocrine Disruptors (1997) RFA Text |  Recipients Lists
Research Category: Economics and Decision Sciences , Health , Safer Chemicals , Endocrine Disruptors


The objectives of this research project were to determine: (1) organochlorine (OC) concentrations in soil, sediment, and nesting material of Morelet’s crocodile from contaminated and noncontaminated sites; (2) OC concentrations in nonviable eggs, caudal scute samples, and fortuitous carcasses of Morelet’s crocodile; (3) biomarker response of crocodiles to OC exposure at the individual level by monitoring plasma vitellogenin induction, plasma hormone (estradiol-17ß, testosterone) concentrations, serum chemistry, and penis size in crocodiles from contaminated and non-contaminated sites; and (4) the response of crocodile populations to OC exposure by monitoring reproductive activity (nesting success), population size structure, sex ratios, and juvenile survival rates on contaminated and noncontaminated sites.

Our research was based on the following hypotheses: (1) Morelet's crocodiles and their eggs from contaminated lagoons in Belize contain higher OC burdens than crocodiles from non-contaminated areas; (2) biochemical (hormones, vitellogenin) and morphological (penis size) differences exist between Morelet's crocodiles from contaminated and noncontaminated areas; and (3) reproductive success, population size structure, sex ratios, and juvenile survival rates are altered in crocodiles inhabiting noncontaminated sites compared to contaminated areas.

Summary/Accomplishments (Outputs/Outcomes):

Over the last 20 years, evidence of population declines and reproductive impairment in American alligators (Alligator mississippiensis) in Florida has increased concerns over the effects of endocrine-disrupting contaminants (EDCs) on wildlife and stressed the importance and utility of reptiles, particularly crocodilians, as a focal species in the field of ecotoxicology. The focal species for this study was Morelet's crocodile (Crocodylus moreletii), an endangered, freshwater crocodile found in Mexico, Guatemala, and Belize. During a pilot study in 1995, multiple OC pesticides considered to be EDCs were found in the eggs of Morelet’s crocodiles from three localities in northern Belize. Based on these findings and previous data from Florida showing egg contamination, population declines, and reproductive abnormalities in alligators exposed to many of the same chemicals, this project was initiated to examine various endpoints of contaminant exposure and response in Morelet’s crocodiles living on contaminated and reference sites in northern Belize.

Contaminant profiles of each study site were examined by analyzing OC residues in soil, sediment, and crocodile nest material. Multiple OCs, some of which are considered to be EDCs, were found in all three matrices at both sites at similar concentrations. Actual OC exposure in crocodiles was examined through residue analysis of nonviable and unhatched eggs and caudal scute samples. As with the environmental samples, multiple OCs were found in all matrices examined. p,p´-Dichlorodiphenylethane (DDE) was found in every egg analyzed, while OC occurrence and concentrations in scutes were more variable. Upon finding that both study sites and the crocodiles inhabiting them were contaminated with OCs, and to approximately the same degree, considerable effort was made to locate a noncontaminated crocodile habitat to use as a reference site for comparisons of individual- and population-level endpoints of endocrine disruption. Two additional sites in northern Belize and four additional sites in southern Belize were examined, but all crocodile eggs from each locality were shown to contain environmental contaminants, suggesting contamination of each site and the crocodiles inhabiting them. Similar contaminant concentrations also were found in American crocodile (C. acutus) eggs from four sites in the coastal zone of Belize, further illustrating the ubiquitous nature of environmental contamination in the country. After failing to find a noncontaminated reference site, we again focused on the New River Watershed and the Gold Button Lagoon, and continued our original course of action. After documenting contamination of both study sites with multiple OCs, we next wished to examine exposure of crocodiles to these contaminants, both at the individual and population levels.

The first individual-level endpoint of response was plasma vitellogenin induction. Vitellogenin is an egg-yolk precursor protein expressed in all oviparous and ovoviviparous vertebrates. Males and juveniles normally have no detectable concentration of vitellogenin in their blood but can produce it following stimulation by an exogenous estrogen, such as an EDC. Thus, the presence of vitellogenin in the blood of male and immature crocodiles can serve as an indicator of exposure to estrogen-mimicking chemicals. Of 358 males and juvenile females sampled in this study, no vitellogenin induction was observed, suggesting these animals likely were not exposed to estrogenic xenobiotics. Many of the animals sampled, however, later were found to contain OC pesticides in their caudal scutes, confirming they in fact had been exposed to OCs (and EDCs). Previous researchers have stressed that vitellogenin induction is a measure of a biological effect, not merely the presence of a contaminant in the body of an animal. Our results support this notion, and suggest plasma vitellogenin induction still may serve as a reliable biomarker of estrogen exposure in crocodilians. The lack of a vitellogenic response, however, should not be interpreted necessarily as an indication that no exposure or other contaminant-induced biological response has occurred.

The second, individual-level endpoint of response was plasma steroid hormone concentrations. The selection of this endpoint was based on numerous studies reporting altered concentrations of estradiol-17ß (E2) and testosterone (T) in alligators from Lake Apopka and other contaminated lakes in Florida. In the present study, few intersite differences in plasma hormone concentrations were noted. No significant differences in plasma E2 concentrations were detected between sites. However, large juvenile males and females from the contaminated site exhibited significantly reduced plasma T concentrations compared to large juvenile males and females from the reference site, respectively. This finding was consistent with results from previous studies on alligators in Florida. No other intersite differences in hormone concentrations were observed.

The third individual-level endpoint of response was male phallus size. Concurrent with reductions in plasma T concentrations, male alligators from Lake Apopka and other contaminated lakes in Florida have exhibited smaller phallus size compared to animals from a reference lake. Researchers speculate that abnormal hormone concentrations during critical early life stages may affect anatomical structures dependent on these hormones for proper growth and development (e.g., genitalia). p,p´-DDE is one of the primary contaminants of concern at Lake Apopka and has been shown to be anti-androgenic in laboratory mammals, inhibiting normal androgen function in vivo. This persistent OC has been detected in alligator eggs and serum from Lake Apopka, suggesting its potential role in the reproductive anomalies observed in juvenile males. p,p´-DDE also has been detected in Morelet's crocodile eggs and scutes in Belize, confirming EDC exposure in maternal females, neonates, juveniles, and other adults. Thus, in the present study, male crocodile phallus size and plasma T concentrations were examined as endpoints of response to p,p´-DDE exposure, as well as exposure to other contaminants.

No differences in mean phallus size were observed between sites, whereas mean plasma T concentrations in juveniles from the Gold Button Lagoon again were significantly reduced compared with those from the New River Watershed. Juvenile males from both sites exhibited positive relationships between body size and phallus size. Whereas juvenile males from the New River Watershed also exhibited positive relationships between plasma T and body size and plasma T and phallus size, no such relationships were observed for juveniles from the Gold Button Lagoon. For adults, no significant intersite differences were observed in phallus size, plasma T concentrations, or relationships between plasma T and body size or phallus size. Because of the similarity in contaminant profiles between the New River Watershed and the Gold Button Lagoon, it is unclear whether phallus size and plasma T concentrations observed in crocodiles from these two sites are normal or altered by some stressor (e.g., EDCs). Thus, the biological significance of the few site differences observed in this study is difficult to interpret.

The examination of population-level endpoints also was confounded by the fact that both sites exhibited similar contaminant profiles. Data on site-specific size class distribution and sex ratios, however, do not indicate feminization of either population. More data on the natural pattern of temperature-dependent sex determination in Morelet's crocodile, population size class distribution, and sex ratios at both sites, and the influence of embryonic exposure to OCs during the temperature-sensitive period are needed to adequately examine the impacts of these chemicals on developing crocodiles and subsequent repercussions at the population level. Intersite differences in population densities (as a function of crocodile encounter rates), population size structure, and sex ratios were observed. We attribute these differences, however, to physical differences between the sites (water level fluctuations, amount of aquatic vegetation, open versus closed systems) and age- and sex-specific wariness and microhabitat use, all of which may have influenced animal catchability and thus interpretation of data on these endpoints. Overall, no population-level effects of EDC exposure on Morelet's crocodiles inhabiting the two study sites were observed.


The primary confounding factor in this study is the lack of a reference site. Similarity in contamination profiles between the New River Watershed and the Gold Button Lagoon precludes the comparison of endpoint measurements to legitimate reference values. Thus, it is difficult to determine the toxicological significance of any intersite differences observed in this study. In addition, in instances where no intersite difference was observed for a particular endpoint (e.g., plasma E2 concentration), it is difficult to discern whether the lack of a difference indicates that that particular endpoint measurement is normal (unaltered) at both sites or if the endpoint is altered in some way, but to the same degree, at both sites. In turn, uncertainty as to whether endpoint measures in this study are altered or unaltered may render comparisons of these data to ecotoxicological data on other crocodilians (such as Florida alligators) less meaningful.

The toxicological significance of OC concentrations in sediments at the Gold Button Lagoon and the New River Watershed is unknown. Because of the paucity of toxicity data pertaining to OC effects on crocodilians and other reptiles, it is unknown what concentrations pose a risk to these animals, and extrapolations based on data from other organisms may be inappropriate. Mean concentrations of p,p´-DDE in crocodile eggs from Belize are among the lowest reported for any crocodilian species. Despite low levels of OC contamination at the New River Watershed and the Gold Button Lagoon compared to other areas of the world, potential chemical-induced effects on Morelet’s crocodiles should not be ignored. Currently, it is unknown at what OC concentrations endocrine disruption may occur in crocodilians, but recent research suggests low concentrations (e.g., 100 ppb p,p´-DDE) similar to those detected at the Belize study sites may cause reproductive impairment in alligators.

Another confounding factor in the present study is the lack of previous research on Morelet's crocodiles focusing on the response endpoints examined in this study. In the absence of a legitimate reference site, basic information on vitellogenin induction, plasma steroid hormone concentrations, serum chemistry, male phallus size, nesting success, population density, size class structure, and sex ratios in other Morelet's crocodile populations would be particularly useful for comparative purposes. Although such data may be limited in their applicability because of multiple sources of interstudy variation, information on seasonal and age- and sex-specific patterns in these endpoints in other populations might aid in interpreting the results observed in this study. Apart from this study, however, data on the endocrinology of Morelet's crocodile are nonexistent.

A third uncertainty associated with this study is the relationship between contaminant exposure and the magnitude of response, if any, in crocodiles at the two study sites. Because of the current endangered status of Morelet’s crocodile, collection of internal tissues for contaminant residue analysis is not feasible. Caudal scutes, collected from crocodiles as a byproduct of the marking procedure, have been analyzed for OC pesticides and have confirmed OC exposure in animals at both sites. Although these samples provide valuable qualitative data on crocodile exposure, their utility as indicators of OC concentrations in internal tissues is unknown. Caudal scutes from alligators have been shown to be relatively poor predictors of mercury in internal tissues, but may provide a rough estimation of contamination in populations without sacrificing animals. Our observations in this study indicate an allometric relationship between crocodile size and fat content in caudal scutes. No fat is visibly present in the scutes of hatchlings and small juveniles, but fat content increases with size such that substantial fat cores are present in the scutes of large adults. Because of the size-specific variation in crocodile scute fat content and the inability to investigate the relationships between OC concentrations in scutes and internal tissues, the relationship between OC exposure and biomarker response in these animals remains unclear.

Natural variability between sites and site-specific influences of other stressors (e.g., disease, injury, and malnutrition) also may have influenced the individual-level endpoints measured in this study. In addition, capture stress and intersite variability in the timing of blood collection may have affected the steroid hormone concentrations in crocodiles sampled at each locality. Physical differences between sites, differences in age- and sex-specific catchability of crocodiles, and the inability to maintain proper incubation temperatures for clutches in captivity likely influenced population-level endpoints. Lastly, slight differences in size class designations, possible differences in species sensitivity to the endpoints measured, and temporal differences in sample collection (e.g., different months) may reduce the validity of comparisons between the present study and similar studies on Florida alligators.

In general, the results of this study indicate few or no effects of EDC exposure on Morelet’s crocodiles inhabiting contaminated wetlands in northern Belize. The only major difference observed between crocodiles from the New River Watershed and the Gold Button Lagoon in this study was that juveniles from the Gold Button Lagoon exhibited lower plasma T concentrations than juveniles at the New River Watershed. This same difference has been consistently observed in juvenile male alligators from Lake Apopka compared to Lake Woodruff. However, in the Florida studies, reduced plasma T also was concurrent with reduced phallus size, suggesting a potential link between reduced T and alterations in anatomical structures (e.g., male phallus) dependent on androgens for proper growth and development. In the present study, no intersite differences in phallus size were observed. This suggests the lower T concentrations in Gold Button Lagoon juveniles, whether contaminant-induced or not, may not be biologically significant. In addition, no intersite differences in plasma E2 concentrations were observed in this study, and vitellogenin induction was not observed in any of the 358 male or juvenile female crocodiles examined. These results suggest minimal or no alteration of E2 concentrations in crocodiles from the two sites, despite the fact that many of these animals have been exposed to environmental contaminants considered to be xenobiotic estrogens. It is possible that the concentrations of EDCs to which Morelet's crocodiles in northern Belize are exposed are insufficient to influence these endpoints, while EDC concentrations at Lake Apopka are sufficiently high to induce an effect. Indeed, mean p,p´-DDE concentrations in alligator eggs from Lake Apopka are 27- to 45-fold higher than those observed in eggs from the Gold Button Lagoon or the New River Watershed. No population-level effects of EDC exposure on Morelet's crocodiles inhabiting the two study sites were observed.

Currently, Morelet's crocodile populations in northern Belize appear to have recovered from past over-harvesting, and threats related to habitat loss and human exploitation appear minimal. Researchers recently have speculated that, although Morelet's crocodiles in northern Belize seemingly face no immediate threats, exposure to environmental contaminants may present a subtle yet significant longterm threat to populations in certain areas. Results of the present study provide little evidence of contaminant-induced effects on crocodiles from two polluted habitats in northern Belize. Multiple confounding factors and uncertainties encountered in this study, however, make intersite and interstudy (crocodile to alligator) comparisons difficult and some results equivocal. Thus, the potential effects of EDCs and other contaminants on crocodiles inhabiting these sites should not be assumed to be negligible. Longterm studies are essential to adequately assess the effects of EDCs on crocodilian populations, as many of the contaminant-induced effects are organizational in nature, occurring during embryonic development, but not appearing until later in life.

Journal Articles on this Report : 2 Displayed | Download in RIS Format

Other project views: All 2 publications 2 publications in selected types All 2 journal articles
Type Citation Project Document Sources
Journal Article Rainwater TR, Wu TH, Finger AG, Canas JE, Yu L, Reynolds KD, Coimbatore G, Barr B, Platt SG, Cobb GP, Anderson TA, McMurry ST. Metals and organochlorine pesticides in caudal scutes of crocodiles from Belize and Costa Rica. Science of the Total Environment 2007;373(1):146-156 R826310 (Final)
not available
Journal Article Wu TH, Canas JE, Rainwater TR, Platt SG, McMurry ST, Anderson TA. Organochlorine contaminants in complete clutches of Morelet's crocodile (Crocodylus moreletii) eggs from Belize. Environmental Pollution 2006;144(1):151-157 R826310 (Final)
not available

Supplemental Keywords:

Belize, Central America, Latin America, tropical ecosystems, tropics, aquatic ecosystems, crocodiles, crocodilians, reptiles, sentinel species, endangered species, endocrine disruptors, dichlorodiphenylethane, DDE, dichlorodiphenyltrichloroethane, DDT, organochlorines, mercury, heavy metals, health effects, morphometrics, reproductive effects, residue concentrations, population effects, individual health effects, ecological effects, organics, aquatic, biology, Florida, FL, agriculture, ecosystem protection, environmental exposure, risk, health, toxicants, biochemistry, biology, chemical mixtures, ecological effects, ecological indicators, ecosystem assessment, indicators., RFA, Health, Scientific Discipline, Toxics, Geographic Area, Waste, Ecosystem Protection/Environmental Exposure & Risk, Bioavailability, Environmental Chemistry, Health Risk Assessment, Ecosystem/Assessment/Indicators, exploratory research environmental biology, State, Endocrine Disruptors - Environmental Exposure & Risk, pesticides, Ecological Effects - Environmental Exposure & Risk, endocrine disruptors, Biochemistry, Children's Health, Molecular Biology/Genetics, Biology, International, Ecological Indicators, ecological exposure, adverse outcomes, aquatic ecosystem, Lake Apopka, Central America, endocrine disrupting chemical, endocrine disrupting chemicals, exposure studies, Morelet's Crocodile, Belize, conservation, animal models, reproductive processes, Florida, biological effects, Lake Apopka Alligators, FL, DDE, exposure assessment

Progress and Final Reports:

Original Abstract
  • 1998
  • 1999
  • 2000
  • 2001
  • 2002