Nutrient and Organochlorine Pesticide ... - Wiley Online Library

Journal of Aquatic Animal Health 21:249–261, 2009 Ó Copyright by the American Fisheries Society 2009 DOI: 10.1577/H07-051.1

[Article]

Nutrient and Organochlorine Pesticide Concentrations in American Alligator Eggs and Their Associations with Clutch Viability R. HEATH RAUSCHENBERGER* U.S. Fish and Wildlife Service, 7915 Baymeadows Way, Suite 200, Jacksonville, Florida 32256 , USA

MARIA S. SEPU´LVEDA Department of Forestry and Natural Resources and School of Civil Engineering, Purdue University, 195 Marsteller Street, West Lafayette, Indiana 47907, USA

JON J. WIEBE

AND

JANET E. WIEBE

Louisiana Department of Wildlife and Fisheries, 2415 Darnall Road, New Iberia, Louisiana 70560, USA

DALE C. HONEYFIELD U.S. Geological Survey, Northern Appalachian Research Laboratory, 176 Straight Run Road, Wellsboro, Pennsylvania 16901, USA

TIMOTHY S. GROSS Department of Physiological Sciences, College of Veterinary Medicine, University of Florida, Post Office Box 100144, Gainesville, Florida 32610, USA Abstract.—Since the early 1900s, the lakes of the Ocklawaha basin in central Florida have experienced ecological degradation due to anthropogenic development. One species affected by this degradation is the American alligator Alligator mississippiensis, which has suffered from poor clutch viability and embryo mortality. Although some studies indicate that organochlorine pesticides (OCPs) may be involved, OCPs do not account for all of the variation seen in hatch rates. Indeed, nutrition and non-OCP contaminants have been associated with developmental problems in fish and birds. Our study evaluated embryo mortality in alligators at reference and OCP-contaminated sites as a function of exposure to OCPs, polychlorinated biphenyls (PCBs), and polyaromatic hydrocarbons (PAHs), along with egg nutrients (Zn, Se, and vitamins A, E, and B1). The four-pronged study consisted of a case-control cohort study, an expanded field study, a topical egg treatment thiamine amelioration experiment, and a topical egg treatment thiamine antagonist experiment. The results from the two field studies suggested that the total thiamine levels in the eggs were positively associated with clutch viability and negatively associated with the lipid content and certain OCPs measured in egg yolks. In addition, PCBs, PAHs, Zn, Se, and vitamins A and E were not found to be associated with the observed clutch viability defects. The thiamine levels in the eggs explained 38% of the variation in clutch survival in the case-control cohort study and 27% in the expanded field study. The topical egg treatment experiments were successful in elevating the thiamine concentrations in the albumin but not the yolk. No significant differences were noted among treatment groups in either egg treatment experiment with respect to clutch survival. In summary, thiamine egg concentrations explain some of the variation in the clutch viability of freeranging alligators, but the cause–effect relationships are still unclear.

In central Florida, the populations of American alligators Alligator mississippiensis inhabiting lakes contaminated with organochlorine pesticides (OCPs) have poor reproductive success, primarily owing to embryo mortality (Woodward et al. 1989, 1993). During 2000–2002, clutch viability (the percentage of eggs that yield live hatchlings) was monitored for 168 * Corresponding author: [email protected] Received December 21, 2007; accepted May 15, 2009 Published online December 14, 2009

clutches from reference and OCP-contaminated sites. Clutches from a reference site, Lochloosa Lake, had higher clutch viability (mean ¼ 70%) than those from the OCP-contaminated sites, Lake Apopka (51%), Emeralda Marsh Conservation Area (EMCA; 48%), and Lake Griffin (44%) (Rauschenberger et al. 2007). Furthermore, 115 of these clutches were analyzed on a wet weight basis for OCPs, and the results indicated that the eggs of alligators inhabiting EMCA (total average egg yolk OCPs ¼ 15,480 ng/g), Lake Apopka (7,582 ng/g), and Lake Griffin (1,169 ng/g) had

249

250

RAUSCHENBERGER ET AL.

significantly higher OCP burdens than those of Lochloosa Lake (102 ng/g) (Rauschenberger et al. 2007). Although total embryo mortality was highest in eggs from sites with high OCPs, OCP egg burdens accounted for less than half the variation in mortality rates, suggesting that other factors also contribute to embryo mortality in alligators (Rauschenberger et al. 2007). Non-OCP factors that have been associated with increased embryo mortality in vertebrates include nutritional imbalances (both deficiencies and excesses; Wilson 1997; McEvoy et al. 2001) and exposure to other contaminants such as polychlorinated biphenyls (PCBs; Summer et al. 1996) and polyaromatic hydrocarbons (PAHs; Hoffman 1990). Indeed, recent reports suggest that vitamin B1 (thiamine) deficiency is involved in the increased incidence of embryo mortality in American alligators inhabiting the aforementioned OCP-contaminated lakes in central Florida. Thiamine concentrations in egg yolks were positively correlated with clutch viability and accounted for 40% of the variation in such viability at Lochloosa Lake, Lakes Griffin, and Apopka, and EMCA (Sepu´lveda et al. 2004). The reasons for continued work in this area are that the above-mentioned study only examined five clutches per site, sampling occurred during only one nesting season (2000), and the roles of other nutrients and contaminants were not evaluated. With respect to other nutrients, Lance et al. (1983) have suggested that vitamin E (tocopherol) has a role in the subnormal clutch viability of captive alligators from Louisiana. With respect to other contaminants, laboratory experiments suggest that contaminants such as PCBs reduce thiamine storage in animals (Yagi et al. 1979) and that high contaminant burdens affect thiamine’s role in the production of metabolic energy (de Roode et al. 2002a). These data support the need for further examination of contaminant burdens and the nutrient content of alligator eggs and their association with poor clutch viability in American alligators from OCPcontaminated sites in Florida. For these reasons, our study evaluated embryo mortality in alligators at reference and OCP-contaminated sites as a function of exposure to OCPs, PCBs, and PAHs and the presence of egg nutrients (Zn, Se, and vitamins A [retinol], E, and B1). The four-pronged study consisted of a case-control cohort study, an expanded field study, a topical egg treatment thiamine amelioration experiment, and a topical egg treatment thiamine antagonist experiment. Zinc and vitamins A, E, (Ashworth and Antipatis 2001), and B1 (de Roode et al. 2002b) were examined because they are important for embryo development

and survival and their activity or level may be affected by chlorinated hydrocarbons. This strategy aided in forming hypotheses pertaining to the associations between non-OCP factors and embryo mortality and between non-OCP factors and OCP exposure. For example, if a high level of a non-OCP factor showed a strong positive association with embryo viability regardless of OCP burden, and the level did not differ among OCP exposure groups, it could be hypothesized that the effects were related to the non-OCP factor and unrelated to OCP exposure. Conversely, if a high level of the non-OCP factor showed a strong positive association with embryo viability, but only in lowOCP exposure groups, it could be hypothesized that the effects were due to a combination of OCP exposure and the non-OCP factor. Based on the case-control cohort study, hypotheses were derived that focused on the major non-OCP factors associated with embryo mortality as well as OCP exposure. The results of the expanded field study were then used to design an egg treatment experiment to examine some of the hypotheses in a more controlled setting. Methods Egg collection and incubation.—American alligator eggs were collected during the 2001, 2002, and 2003 nesting seasons (June–July) from three OCP-contaminated sites—Lake Apopka (28835 0 N, 81839 0 W), Lake Griffin (28853 0 N, 81846 0 W), and EMCA (28855 0 N, 81847 0 W)—and two uncontaminated reference sites— Lochloosa Lake (29830 0 N, 82809 0 W) and Lake Dexter (29898 0 N, 81847 0 W)—in central Florida. Alligator nests were located by means of aerial (helicopter) and ground (airboat) surveys, and clutches were subsequently collected by ground crews. The top of each egg was marked before the eggs were removed from the nests to ensure proper orientation, thereby preventing embryo mortality owing to inversion. Such mortality occurs because inverting the egg’s orientation may either break the embryonic attachment or cause the yolk mass to settle on top of the embryo, crushing it. After the eggs were marked, all of those from a given clutch were placed in a uniquely numbered polypropylene pan (43 3 33 3 18 cm) containing about 5 cm of nest substrate. The eggs were arranged in five rows with six eggs per row; if a clutch contained more than 30 eggs, a second layer of nest substrate was added and the remaining eggs were placed on top of the others. The top layer of eggs was covered with nest substrate so that there was no space left between the eggs and the top of the pan (approximately 10 cm). The clutches were then transported to the U.S. Geological

NUTRIENTS, PESTICIDES, AND ALLIGATOR EGG VIABILITY

251

TABLE 1.—Reproductive, morphometric, and contaminant variables measured in American alligator eggs collected during summer 2000, 2001, and 2002. Abbreviations are as follows: OCP ¼ organochlorine pesticide, PCBs ¼ polychlorinated biphenyls, and PAHs ¼ polyaromatic hydrocarbons. Variable

Definition

Units

Response variables Fecundity Clutch mass Mean egg mass Unbanded eggsa Early embryo mortality Late embryo mortality Clutch viability

Number of eggs in one clutch Mass of eggs in one clutch Clutch mass/fecundity (Number of unbanded eggs/fecundity) 3 100 (Number of deaths before development day 35/fecundity) 3 100 (Number of deaths on or after development day 35/fecundity) 3 100 (Number of eggs yielding live hatchlings/fecundity) 3 100

Number kg g Percent Percent Percent Percent

Explanatory variables OCP analyte in egg yolk Percent OCP analyte PCBs in egg yolk PAHs in egg yolk Thiamine in egg yolk Zn, Se, vitamin A, and vitamin E in egg yolkb

OCP analyte/egg yolk wet weight (OCP analyte/R[OCPs]) 3 100 PCBs/egg yolk wet weight PAHs/egg yolk wet weight Thiamine/egg yolk wet weight Analyte/egg yolk wet weight

ng/g Percent ng/g ng/g pmol/g ng/g

a b

Eggs with no evidence of embryonic attachment. Clutches collected in 2002 only.

Survey’s Center for Aquatic Resources Studies in Gainesville, Florida. Upon arrival, the clutches were evaluated for embryonic viability using a bright-light candling procedure. One or two eggs were opened from each clutch to document the embryonic stage of development for each clutch at the time of collection. Yolk samples were collected and stored at 808C for later determination of the concentrations of OCPs, PAHs, and PCBs (ng/g of egg yolk wet weight) as well as those of Se, Zn, and vitamins A, E, and B1 (pmol/g of egg yolk wet weight). For each clutch the following parameters were recorded: the number of eggs (fecundity); the numbers of unbanded, damaged, dead-banded, and live-banded eggs; total clutch mass (kg); and average egg mass (g; see Table 1). Unbanded eggs are eggs in which there is no evidence of embryo attachment to the top of the inner eggshell membrane; these may be nonfertilized eggs or eggs in which embryo mortality occurred before attachment (Rotstein et al. 2002). The viable eggs, which had a visible band and a reddish-pink hue when candled, were nested in pans containing moist sphagnum moss and incubated at 30.58C and approximately 98% humidity in an incubation building (7.3 3 3.7 m). The incubation temperature used results in a 1:1 male : female sex ratio, as alligators have temperature-dependent sexual differentiation (Ferguson 1985). On a daily basis, the temperature and humidity were monitored at several locations throughout the incubator, the clutches were rotated within the incubator, and air was circulated to mitigate any thermal gradients. The eggs were monitored for viability via illumination every 10 d during incubation. The clutches collected in 2002 were

used for the case-control cohort field study, which was expanded by adding data from clutches collected and analyzed during the 2000 and 2001 nesting seasons. The clutches collected in 2003 were used for the laboratory egg treatment experiments. Field studies.—The case-control cohort study conducted in 2002 involved the selection of clutches based on their clutch viability (good, intermediate, or poor) and OCP egg burden (high, intermediate, or low). Each clutch was thus assigned to a category in a 3 3 3 matrix (Table 2). The purpose of the case-control cohort study was to determine whether the concentrations of PAHs, PCBs, Zn, Se, and vitamins A, E, and B1 differed or showed trends among the clutch viability–OCP categories. The purpose of the expanded field study was to further examine the relationships between the thiamine and OCP concentrations in eggs and clutch viability and clutch size. Thiamine, OCP, and clutch survival data for clutches collected from 2000 (Sepu´lveda et al. 2004), 2001, and 2002 were combined to increase the ecological relevance and validity of our conclusions as well as the power of the tests of the association between thiamine level and clutch viability. Standardized methods were used for the OCP and thiamine analyses as well as for the egg collections and incubation for all of the clutches (n ¼ 72) included in this analysis. Laboratory experiments.—In 2003, two complementary laboratory experiments were conducted using clutches collected from Lake Dexter, Lake Griffin, and EMCA. The first experiment tested the hypothesis that increasing the thiamine levels of the eggs would decrease embryo mortality. This experiment consisted

252


TABLE 2.—Morphometric, nutrient, and contaminant measurements (means 6 SEs) in American alligator egg clutches collected for the 2002 case-control cohort study from Lochloosa Lake, Lake Apopka, Lake Griffin, and EMCA. Values are presented for six categories of clutch viability and organochlorine pesticide (OCP) level, as follows: good–high (.70% viability and .3,700 ng of OCPs/g of yolk wet weight), good–intermediate (.70% viability and 350–3,700 ng/g), good–low (.70% viability and ,350 ng/g), intermediate–low (48–70% viability and ,350 ng/g), poor–high (,48% viability and .3,700 ng/g), and poor–low (,48% viability and ,350 ng/g). Values with letters in common are not significantly different (P . 0.05). See Table 1 and the text for further information on the variables studied. Variable Clutches Fecundity Clutch mass Egg mass Clutch viability Damaged eggs Unbanded eggs Early embryo mortality Late embryo mortality Dieldrin Heptachlor epoxide cis-chlordane cis-nonachlor Oxychlordane Toxaphene p,p 0 -DDD p,p 0 -DDE p,p 0 -DDT trans-chlordane trans-nonachlor OCPs NOCa PAHs PCBs Se Zn Thiamine monophosphate Thiamine pyrophosphate Free thiamine Total thiamine Vitamin Eb a b

Good–high

51 4 86 92 0 6 2 0 248 3 161 82 23 10,289 2,614 19,136 24 51 251 32,959 14 21 39 1,000 15,900 23 14 463 500 16,287

2 65 6 0.5 6 0.8 6 3.4 z 60 6 5.7 6 2.3 60 6 20.2 6 0.8 6 14.9 6 11.2 6 4.1 6 313.6 6 348.7 6 3,277.6 60 6 1.5 6 46 6 3,891.1 z 60 6 1.3 yx 60y 6 100 6 300 6 5.7 zy 6 3.6 z 6 92.6 6 83.3 6 2389.7

Good–intermediate

49 3 69 80 1 6 10 3 72 11 15 24 25 0 10 1,167 0 1 55 1,391 11 33 168 1,233 15,433 13 0 719 733 26,397

3 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6

1.9 0.7 12.5 3.8 z 1.3 1.5 5.1 1.4 47.8 4.2 1.1 4.1 13.4 0 4.9 830.3 0 0 13.9 923 y 0 2.8 zy 28.9 z 176.4 809 1.6 zy 0.3 x 427.9 427.7 2,978.6

Good–low

41 4 88 79 0 10 12 0 6 4 4 7 8 0 3 139 0 1 16 188 10 34 83 1,067 24,900 41 21 868 931 19,118

3 62 6 0.2 66 6 4.3 z 60 6 4.9 6 8.1 60 6 1.1 6 2.3 6 1.3 6 1.3 65 60 6 0.6 6 30.9 60 60 6 4.7 6 41.9 x 6 0.7 6 7.6 zy 6 14.6 z 6 66.7 6 10,570.9 6 8.4 z 6 3.1 z 6 231.1 6 220.0 6 7,352.9

Intermediate–low

49 4 86 60 0 15 7 18 9 2 7 9 4 0 4 117 0 1 15 168 10 31 111 1,133 15,300 24 16 238 278 12,255

3 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6

5.2 0.4 1.5 5z 0 4.2 3.3 4 5 0.5 3.6 4.2 1.4 0 0.7 45.6 0 0 5.8 67.9 x 0.9 2.9 zy 47 z 166.7 750.6 1.7 zy 2.1 z 40.6 38.9 122.5

Poor–high

55 4 76 11 2 25 54 8 157 6 145 89 31 4,670 897 9,149 10 34 273 15,508 14 18 40 1,000 15,267 2 18 891 912 21,054

3 6 5.8 6 0.4 6 6.4 6 9.4 y 6 1.1 6 10.5 6 21.6 6 4.3 6 66 6 2.6 6 69.7 6 38.4 6 6.8 6 1,268.9 6 578.1 6 3,668.7 6 2.1 6 15.3 6 141.4 6 5,426.2 z 6 0.3 6 2.5 x 6 2.5 z 6 57.7 6 437.2 6 1.8 x 6 8.1 zy 6 261.2 6 267.4 6 1,889.2

Number of OCP analytes above the limit of quantitation. a-tocopheral.

of topically treating eggs from Lake Griffin and EMCA with thiamine to increase their thiamine levels. The eggs from these two sites are known to have low thiamine concentrations, moderate to high yolk OCP concentrations, and high embryo mortality (Sepu´lveda et al. 2004). A high (60 mg/mL of dimethyl sulfoxide [DMSO]) or low (12 mg/mL) concentration (application volume ¼ 50 lL) of thiamine-HCl (ICN Biomedicals, Montreal, Quebec) was applied to the surface of each egg with a micropipette. Controls received only vehicle treatment (50 lL of DMSO). These doses were calculated to achieve yolk thiamine concentrations similar to those measured (3,123 6 477 pmol/g) in clutches from the reference sites (Sepu´lveda et al. 2004). The eggs from each site were labeled and randomly distributed across treatment groups, so that all clutches were equally represented in all groups. There were two replicates per treatment, with a minimum of 26 eggs (maximum, 31) per replicate.

After being dosed, the eggs were placed in the incubator and monitored weekly to determine embryo and hatchling mortality. Three eggs from each replicate were sampled 7 d after treatment to determine the amount of thiamine present in the albumin and yolk. The embryo mortality rate (i.e., the percentage of treated eggs that failed to hatch) was recorded for each treatment group. The second experiment tested the hypothesis that, in the absence of high OCP exposure, a reduction in thiamine bioactivity (functional deficiency) would increase embryo mortality. This experiment involved topically applying a thiamine antagonist to eggs from Lake Dexter, which are known to have relatively high thiamine concentrations, low OCP burdens, and low embryo mortality (A. R. Woodward, Florida Fish and Wildlife Conservation Commission, personal communication). To decrease thiamine bioactivity, oxythiamineHCl (ICN Biomedicals), a thiamine antagonist (Aker-


TABLE 2.—Extended.

Variable Clutches Fecundity Clutch mass Egg mass Clutch viability Damaged eggs Unbanded eggs Early embryo mortality Late embryo mortality Dieldrin Heptachlor epoxide cis-chlordane cis-nonachlor Oxychlordane Toxaphene p,p 0 -DDD p,p 0 -DDE p,p 0 -DDT trans-chlordane trans-nonachlor OCPs NOCa PAHs PCBs Se Zn Thiamine monophosphate Thiamine pyrophosphate Free thiamine Total thiamine Vitamin Eb

Poor–intermediate

46 4 78 15 1 5 36 42 264 4 25 21 14 1,928 18 1,019 0 1 41 2,057 11 27 55 933 12,367 12 4 553 569 20,025

3 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6

4.5 0.6 6.2 10.9 y 1.1 2.7 26.9 21.8 134.8 1.8 9.2 6.3 5.2 0 4.8 795 0 0 15.8 770.2 y 0.7 3.5 zyx 10.7 z 120.2 800.7 6.7 yx 3.8 yx 482.3 486.5 5,831

Poor–low

47 4 79 25 0 21 30 24 16 5 13 11 7 0 4 153 0 2 22 234 11 43 92 833 12,967 13 7 326 345 15,221

3 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6

0.9 0.1 0.3 13 y 0 6.8 3.9 16.9 5.4 1.7 0.8 0.8 2.8 0 0.8 28.6 0 0.5 2.8 42.5 x 0 7.3 z 16 z 145.3 788.1 6.4 zy 5.3 zy 115.4 104.3 1,784.5

man et al. 1998), was topically applied at concentrations of 12 or 60 mg/mL (application volume, 50 lL) using DMSO as the carrier. Controls received 50 lL of DMSO. The eggs were labeled and randomly distributed among treatment groups, so that all clutches were equally represented in each group. There were two replicates per treatment, with a minimum of 20 eggs (maximum, 21) per replicate. After being dosed, the eggs were placed in the incubator and monitored weekly to determine embryo and hatchling mortality. For each replicate, hatch rates were determined as the number of live hatchlings per the number of eggs in the replicate. Analysis of chlorinated hydrocarbons in yolk.—Egg yolk samples were analyzed for chlorinated hydrocarbon content using the methods described in Rauschenberger et al. (2004b) by the University of Florida’s Analytical Toxicology Core Laboratory in Gainesville. Briefly, analytical grade standards for the following compounds were purchased from the sources indicated: aldrin, alpha-benzene hexachloride (a-BHC), b-BHC, lindane, d-BHC, p,p 0 -dichlorodiphenyldichloroethane

253

(p,p 0 -DDD), p,p 0 -dichlorodiphenyldichloroethylene (p,p 0 -DDE), dichlorodiphenyltrichloroethane (p,p 0 DDT), dieldrin, endosulfan, endosulfan II, endosulfan sulfate, endrin, endrin aldehyde, endrin ketone, heptachlor, heptachlor epoxide, hexachlorobenzene, kepone, methoxychlor, mirex, cis-nonachlor, and transnonachlor from Ultra Scientific (Kingstown, Rhode Island); cis-chlordane, trans-chlordane, and the 525, 525.1 polychlorinated biphenyl mix from Supelco (Bellefonte, Pennsylvania); oxychlordane from Chem Service (West Chester, Pennsylvania); o,p 0 -DDD, o,p 0 DDE, and o,p 0 -DDT from Accustandard (New Haven, Connecticut); and toxaphene from Restek (Bellefonte, Pennsylvania). All reagents were analytical grade unless otherwise indicated. Water was doubly distilled and deionized. Duplicate quality control samples were prepared and analyzed with every 20 samples (typically at a level of 1.00 or 2.50 lg/mL of lindane, heptachlor, aldrin, dieldrin, endrin, and p,p 0 -DDT) with an acceptable recovery ranging from 70% to 130%. The limit of detection ranged from 0.1 to 1.5 ng/g for all OCP analytes except toxaphene (120–236 ng/g); the limit of quantitation was 1.5 ng/g for all analytes except toxaphene (1,500 ng/g). Repeated analyses were conducted as allowed by matrix interferences and sample availability. Nutrient analysis.—Thiamine concentrations were measured in clutches collected during 2001, 2002, and 2003. At the time of collection, each sample was placed in a 5-mL cryovial, immediately frozen in a methanol–dry ice bath, shielded from UV light, and stored in a 808C freezer. The samples were later shipped overnight on dry ice (solid CO2) to the U.S. Geological Survey’s Appalachian Research Laboratory in Wellsboro, Pennsylvania, for analysis. Thiamine concentrations were determined as described in Brown et al. (1998). Briefly, a known amount of the frozen yolk sample was first placed in a 2% homogenization solution of trichloroacetic acid (Sigma, St. Louis, Missouri). The extract was then washed with ethyl acetate : hexane (3:2 by volume; Sigma) to remove excess acid. An aliquot of the washed solution was reacted with potassium ferricyanide (Sigma) to produce thiochrome derivatives. The resulting derivatives were separated on a Hamilton PRP-1 column (Alltech, Deerfield, Illinois) and detected with a spectrofluorometer set at 375-nm excitation and 433-nm emission wavelengths (Shimadzu, Columbia, Maryland). Authentic standards for thiamine pyrophosphate (TPP), thiamine monophosphate (TP), and thiamine-HCl (Sigma-Aldrich, St. Louis, Missouri) were used to quantify the amount of thiamine in each sample. In addition to being subjected to thiamine analysis,

254


samples from selected clutches collected during 2002 were analyzed for vitamin A (carotene, retinol, and activity; Association of Official Analytical Chemists [AOAC] 960.45 and 941.15), vitamin E (AOAC 948.26), Zn, and Se (AOAC 990.8) using AOAC methods (Horwitz 2000) by ABC Research Corporation in Gainesville, Florida. Because of the cost of screening for numerous analytes and because oxythiamine has membrane transfer properties similar to those of thiamine (oxythiamine-HCl and thiamine-HCl have molecular weights of 338.26 g/mol and 337.27 g/ mol, respectively, and topological polar surface areas of 75.9 and 65.6 square angstroms; U.S. Department of Health and Human Services 2005a, 2005b), the oxythiamine concentrations were not measured. Therefore, the transfer rates across the eggshell and eggshell membranes were assumed to be similar for these compounds. Data analysis.—In all phases of our study, analysis of variance (ANOVA; PROC GLM; SAS Institute 2002) was used for intersite and intergroup comparisons of the summary clutch characteristics, followed by Tukey’s test for multiple comparisons among sites and groups (a ¼ 0.05). Specific OCP analytes were removed from the analysis if measurable concentrations were found in less than 5% of all clutches. Because the relationships between the response and explanatory variables in ecological studies are often complex and involve interactions, an indirect gradient multivariate analysis method, detrended correspondence analysis (DCA; ter Braak 1986), was used to initially evaluate data structure for the case-control cohort and expanded field studies. Two matrices were constructed for the DCA, the first representing the response variables (clutch identification number 3 clutch variable) and the second representing the explanatory variables (clutch identification number 3 contaminant and nutrient variable). The DCA results indicated that redundancy analysis (RDA; Rao 1964) was appropriate for the case-control cohort and expanded field studies since the gradient lengths of the DCA ordination axes were no more than 2 standard deviations (ter Braak 1995). For the RDA, response variables measured as a percentage (i.e., clutch viability) or a number (i.e., clutch mass) were divided into separate matrices. The data were evaluated for normality and homogeneity of variance and transformed to meet the assumptions for parametric tests. Percentage data were loge(x þ 1) transformed and not standardized, while continuous data were loge(x) transformed and standardized (ter Braak and Smilauer 2002). For the case-control cohort study, no OCP variables were included in the RDA involving clutch survival or size since the clutches

were initially selected and categorized according to total OCP burden. Automatic forward selection of the four best explanatory variables was used for all RDA analyses. Monte Carlo permutation tests were used to determine significance (a ¼ 0.05). The DCA and RDA were conducted with the program CANOCO (ter Braak and Smilauer 2002), and CANODRAW (ter Braak and Smilauer 2002) was used to construct biplots of the environmental and response variables so as to highlight their relationships. Results Field Studies Case-control cohort study.—In 2002, 32 clutches were collected from EMCA, Lake Apopka, Lake Griffin, and Lochloosa Lake, for which clutch viability and OCP concentrations were determined. Of these clutches, 20 were selected for nutrient and non-OCP analyses. Twelve clutches were excluded because including them would have caused there to be uneven numbers of clutches across the 3 3 3 case-control cohort matrix. When more than three clutches met the requirements for a particular cell in the matrix, each clutch was assigned a number and three were randomly selected according to number. Seven of the nine categories were filled, six categories being represented by three clutches (Table 2; the good viability–high OCP burden category was represented by two clutches). Although the number of clutches in each category was not large, the clutches assigned to the good and intermediate viability categories had significantly greater viability rates than those assigned to the poor viability category, supporting the assignment of these clutches to their respective categories. Similarly, the clutches assigned to the high, intermediate, and low OCP categories were significantly different from one another with respect to total OCP burden (Table 2). In addition to differences in clutch viability rates and total OCP burdens among categories, differences were found with respect to the total PCB and total PAH burdens and the TP and TPP contents in the eggs (Table 2). Vitamin A was not detected in any of the eggs, probably because of the high detection limits (0.3 mg/kg) of the assay. Because vitamin A was not detected in any eggs, no conclusions were drawn as to its potential role in embryo mortality in alligators. No significant differences were found for Se, Zn, vitamin E, free thiamine (FT), and total thiamine (TT) (Table 2). Redundancy analysis with forward selection of the four best explanatory variables provided a way to evaluate the relationships between the non-OCP and


clutch variables and allowed each clutch’s site to be included in the analysis. Including site in the analysis aided in identifying whether site differences (as opposed to other factors) were related to clutch survival and related variables. For the 20 clutches included in the RDA, TP (lambda A ¼ 26%) and TPP (12%) were significantly correlated with clutch survival, accounting for 38% of the variation in the clutch survival variables. Indeed, TP had a strong positive association with clutch viability and a strong negative association with early embryo mortality, while TPP had a strong negative association with late embryo mortality (Figure 1). In contrast, the PAH and PCB burdens did not appear to be significantly associated with embryo survival. In contrast to clutch survival, clutch size (e.g., fecundity) appeared to be associated with site, as three of the four extracted explanatory variables were the nominal site variables. Of these extracted variables, only Lochloosa Lake was determined to be significantly associated with clutch size, accounting for 27% of the variation. Furthermore, the Lochloosa Lake clutches appeared to have higher average egg masses and lower fecundity than those from other sites (Figure 1B). Lastly, the relationship between nutrients and chlorinated hydrocarbons were examined via RDA. Interestingly, all four extracted explanatory variables— the heptachlor epoxide concentration (lambda A ¼ 18%), dieldrin (17%; 100 [dieldrin concentration/total OCP concentration]), the trans-chlordane concentration (15%), and collection from Lochloosa Lake (9%)—were found to be significantly associated with the nutrient levels in eggs, accounting for 59% of the variation in nutrient content. The heptachlor epoxide concentration had a strong negative correlation with TPP but weak positive correlations with the other thiamine forms and nutrients. Dieldrin had strong negative associations with FT and TT but strong positive relationships with vitamin E, Zn, and Se. The trans-chlordane concentration had strong negative correlations with vitamin E, Zn, and Se and little to no correlation with the thiamine concentrations. The Lochloosa Lake clutches appeared to be associated with high thiamine concentrations and low Zn, Se, and vitamin E concentrations (Figure 1C). In summary, the results of the case-control cohort study suggested that thiamine was positively associated with clutch viability and negatively associated with embryo mortality (Figure 1A) and that the thiamine concentration was negatively associated with certain OCPs (Figure 1C). Expanded field study.—Clutches collected from Lochloosa Lake (n ¼ 18), Lakes Griffin (21) and

255

Apopka (14), and EMCA (19) were compared in the expanded field study. No significant differences were noted among sites with respect to clutch survival, clutch size, or the four forms of thiamine. Significant differences were found among sites with respect to the OCP burden in eggs. Indeed, the mean total OCP burden and the number of OCP analytes detected at quantifiable levels differed significantly among sites (Table 3). The initial RDA showed that embryo age at the time of collection was an important factor, but not a specific factor of interest. Therefore, another RDA was conducted using age as a covariate. The four best explanatory variables determined by the second RDA were significant and accounted for 30% of the variation in clutch survival. These variables were the concentrations of TT (lambda A ¼ 16%), TPP (7%), and TP (4%), and the relative concentration of methoxychlor (3%; 100 [methoxychlor concentration/total OCP concentration]). Total thiamine and thiamine monophosphate were strongly and positively correlated with clutch viability and negatively correlated with the percentage of unbanded eggs and early embryo mortality but showed almost no correlation with late embryo mortality (Figure 1D). Thiamine pyrophosphate was strongly and negatively correlated with late embryo mortality but had weak to near-zero correlations with the remaining clutch survival variables. The concentration of methoxychlor had positive correlations with the percentage of unbanded eggs and early embryo mortality but almost no correlation with other clutch survival parameters (Figure 1D). Redundancy analysis was also used to examine the relationships between clutch size and its explanatory variables. The results of this analysis indicated that two of the four extracted variables—FT (lambda A ¼ 9%) and TPP (6%)—were significant, explaining 15% of the variation in clutch size. All of the thiamine forms were positively associated with mean egg weight and negatively associated with fecundity. Thiamine pyrophosphate was negatively correlated with collection from Lake Griffin, indicating that depressed TPP levels were associated with clutches from that lake. Total and free thiamine had strong, negative correlations with clutch mass (Figure 1E). Lastly, RDA was used to examine the relationship between the various thiamine forms (as response variables) and the explanatory variables to see whether thiamine deficiency was associated with the OCP variables and other clutch variables. The results indicated that four extracted variables were significantly correlated with the thiamine concentrations, accounting for 31% of the variation in thiamine levels. Interestingly, the lipid content of the eggs accounted

256


FIGURE 1.—Ordination biplots of American alligator clutch viability and various explanatory variables obtained from redundancy analysis (RDA). Panel (A) shows the biplot of clutch survival and explanatory variables for clutches collected at Lakes Apopka and Griffin, Lochloosa Lake, and the Emeralda Marsh Conservation Area during 2002 (the case-control cohort study), panel (B) the corresponding biplot of clutch size; panel (C) shows the biplot of nutrient concentrations in eggs (solid arrows) and explanatory variables (dashed arrows) in the same study. Panel (D) shows the biplot of clutch survival and explanatory variables for clutches collected during 2000–2002 (the expanded field study), panel (E) the corresponding biplot of clutch size (solid lines) and explanatory variables (dashed lines); panel (F) shows the biplot of thiamine egg yolk concentration (solid lines) and explanatory variables (dashed lines) for clutches collected during 2000–2002. The first two RDA axes are shown. Arrows pointing in the same direction indicate a positive correlation, arrows that are perpendicular indicate a near-zero correlation, and arrows pointing in opposite directions indicate negative correlations. The cosine of the angle formed at the origin between individual variables is the correlation coefficient (r). For example, if arrows pointing in exactly opposite directions have an angle of 1808 and cos(180) ¼ 1.0, then the variables would be perfectly negatively correlated (ter Braak 1995). The explanatory variables can be ordered according to their degree of association with the response variables by perpendicularly projecting the explanatory variable points onto the response variable arrow. For example, panel (A) shows that a perpendicular projection (z) of thiamine monophosphate (TP) onto clutch viability indicates that TP has a stronger association with clutch viability that total thiamine (TT; y). Other abbreviations are as follows: FT ¼ free thiamine, TPP ¼ thiamine pyrophosphate, PCBs ¼ polychlorinated biphenyls, DL% ¼ percent dieldrin, HE ¼ heptachlor epoxide, TC ¼ trans-chlordane, ME% ¼ percent methoxychlor, MI ¼ mirex, and OX ¼ oxychlordane.

257


TABLE 3.—Measurements (means 6 SEs) for American alligator egg clutches collected during the 2000–2002 expanded field study. Values with letters in common are not significantly different (P . 0.05). See Tables 1–2 and text for further information. Variable Clutches Fecundity Clutch mass Egg mass Clutch viability Damaged eggs Unbanded eggs Early embryo mortality Late embryo mortality Thiamine monophosphate Thiamine pyrophosphate Free thiamine Total thiamine Aldrin o,p 0 -DDD o,p 0 -DDT Methoxychlor Mirex Dieldrin Heptachlor epoxide cis-chlordane cis-nonachlor Oxychlordane Toxaphene p,p 0 -DDD p,p 0 -DDE p,p 0 -DDT trans-chlordane trans-nonachlor OCPs NOC

Lochloosa

40 3.6 90 63 4 11 13 8 20 12 747 780 0 0 1 0 2 4 3 2 5 4 0 2 76 1 3 8 104 9

18 6 1.8 6 0.18 6 3.1 6 5.6 6 3.3 62 63 6 2.6 6 3.7 6 2.6 6 102.8 6 102.8 60 60x 60y 60 6 0.4 6 0.5 x 6 0.8 x 6 0.2 w 6 0.6 x 6 1.1 x 60x 6 0.2 w 6 12.3 w 60x 6 0.7 yx 6 1.7 x 6 16.2 w 6 0.3 w

Griffin

45 3.5 78 40 5 12 26 18 19 7 536 562 0 1 3 17 2 20 6 11 18 11 2,678 7 283 2 2 37 783 11

for 16% of the thiamine variation, followed by the mirex (5%), trans-chlordane (6%), and oxychlordane concentrations (4%). Lipid content had strong negative correlations with FT and TT and positive correlations with TP and TPP. The trans-chlordane concentration was positively correlated with TPP and negatively correlated with the remaining thiamine forms. The oxychlordane and mirex concentrations were positively correlated with TP, had near-zero correlations with TT and FT, and had weak negative correlations with TPP (Figure 1F). Laboratory Experiments Ten clutches were used in the thiamine amelioration experiment, all 10 from sites known to have relatively high embryo mortality rates and low thiamine levels in eggs. Five of the clutches, however, were from a site known to produce clutches with intermediate OCP egg yolk burdens (Lake Griffin), and the other five were from a site (EMCA) known to produce clutches with high OCP egg yolk burdens. Conversely, clutches (Lake Dexter; n ¼ 4) with relatively low embryo mortality, high thiamine levels in eggs, and low OCPs were used in the thiamine antagonist topical exposure study. Clutch characteristics differed significantly

21 6 1.8 6 0.19 6 2.2 6 6.7 6 2.5 6 2.3 6 6.3 6 4.9 6 3.8 6 2.7 6 93.1 6 94.2 60 60y 6 0.4 y 6 0.3 6 0.4 6 3.4 y 6 1.4 y 6 0.7 x 6 2.4 y 6 2.1 y 6 376.5 y 61x 6 47.4 x 6 0.7 yx 6 0.2 x 6 7.2 y 6 264.9 x 6 0.2 x

Apopka

47 4 86 49 2 13 17 19 20 8 573 601 3 5 10 8 4 323 12 46 61 38 2,738 49 4,576 9 7 157 6,855 13

14 62 6 0.2 6 3.5 6 8.6 6 0.9 6 3.6 6 6.4 6 6.8 6 3.9 6 3.6 6 110.8 6 111.1 6 0.1 6 2.3 y 6 2.1 z 6 2.6 6 1.4 6 66.8 z 6 2.2 z 6 12.2 y 6 12.4 z 6 6.1 z 6 224.5 y 6 13.3 y 6 948.3 y 6 3.8 y 6 2.2 y 6 36.8 z 6 1,267.2 y 6 0.4 y

EMCA

46 4 87 46 5 12 25 11 19 11 657 688 3 47 301 10 3 186 6 109 71 24 7,558 1,711 11,304 15 31 208 20,417 14

19 6 1.8 6 0.38 6 7.4 6 9.1 6 1.8 6 3.5 6 6.7 6 3.8 6 5.2 6 3.8 6 125.9 6 128.5 6 0.3 6 5.7 z 6 290.9 z 6 1.9 6 0.9 6 25.4 z 6 1.6 y 6 15.6 z 6 8.9 z 6 3.4 z 6 703.6 z 6 225.1 z 6 1,872.1 z 6 1.7 z 6 4.1 z 6 30.8 z 6 2,969.9 z 6 0.2 z

among sites with respect to fecundity, clutch mass, egg mass, most OCP analytes, total OCP burdens, and the number of OCPs detected at quantifiable levels (Table 4). Seven days after topical treatment, three eggs from three different clutches (the same clutches for all replicates) were analyzed to determine the amount of thiamine that was transferred to the egg. Because of funding limitations and because the TT egg yolk concentrations were of primary interest, the TT concentrations in albumin were only measured in EMCA clutches to validate that topical applications would cross the eggshell and eggshell membranes. For the EMCA clutches, the TT in the egg albumin of the high- and low-thiamine treatment groups was significantly greater than in the controls (Table 5). Indeed, these concentrations were more than 40- and 30-fold greater, confirming that there was a significant increase in the thiamine levels of the albumin. In contrast, the TT concentrations in the egg yolks of the EMCA clutches from the high- and low-thiamine treatment groups were not significantly higher than those of the controls (Table 5). Similarly, no significant differences were found between treatment groups for the Lake Griffin clutches (Table 5). The thiamine concentrations

258


TABLE 4.—Measurements (means 6 SEs) for American alligator egg clutches collected from Lake Dexter, Lake Griffin, and EMCA for the 2003 topical egg treatment studies. Values with letters in common are not significantly different (P . 0.05). See Tables 1–2 and text for further information. Variable Clutches Fecundity Clutch mass Egg mass Unbanded eggs Damaged eggs Dieldrin Heptachlor epoxide cis-chlordane cis-nonachlor Oxychlordane Toxaphene o,p 0 -DDD o,p 0 -DDT p,p 0 -DDD p,p 0 -DDE p,p 0 -DDT trans-chlordane Endrin ketone Mirex trans-nonachlor OCPs NOC

Dexter

37 3 82 3 0 5 2 1 6 4 0 1 1 1 117 1 1 0 4 10 171 12

Griffin

4 6 3.8 y 6 0.5 y 6 4.1 y 62 60 6 1.4 6 0.4 60y 6 1.8 y 61y 60y 60y 60 60 6 28.1 y 60y 60 60y 6 1.3 6 3.7 y 6 38 y 6 0.5 y

46 4 93 6 1 29 8 1 22 14 0 0 3 3 399 2 1 0 2 54 556 12

in the egg yolk of the control groups for both the EMCA and Lake Griffin clutches were higher than the means reported in the expanded field study but still within their respective ranges. Although the EMCA clutches from both thiamine treatment groups had embryo mortality rates averaging 10% less than those of the controls, no significant differences were found among the treatment groups (Table 5). In the thiamine antagonist study, embryo survival

6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6

EMCA

5 1.2 zy 0.1 z 2.7 z 3 0.5 9.3 3.1 0y 7 zy 5 zy 0y 0y 0.8 0.4 114.7 y 0.4 y 0 0y 0.3 19.9 zy 159.2 y 0.4 zy

46 4 80 5 4 188 4 8 60 26 6,765 13 3 981 13,166 16 3 3 2 168 21,410 14

5 6 1.9 z 6 0.1 zy 6 2.6 y 6 3.4 6 3.1 6 78.2 6 1.5 6 2.6 z 6 19.8 z 6 10.7 z 6 2,240.4 z 60z 6 0.4 6 407.8 6 5,918.5 z 65z 6 0.6 60z 6 0.5 6 60.3 z 6 8,499.4 z 6 0.5 z

of the controls (81 6 9% [mean 6 SE]) was slightly less than that of the low- (98 6 2%) and highexposure groups (88 6 8%), but no significant differences were found between the treatment groups and the controls. Discussion The present study examined the associations among egg nutrients, OCP egg burdens, and clutch survival and

TABLE 5.—Comparison of the thiamine contents of American alligator eggs collected from EMCA and Lake Griffin and treated with different amounts of thiamine-HCl. The low-treatment group received 50 lL of 12 mg of thiamine-HCl/mL of dimethyl sulfoxide, the high-treatment group 60 mg/mL; the control group received only 50 lL of dimethyl sulfoxide. Values with letters in common are not significantly different (P . 0.05). Treatment group Site

Component

EMCA

Albumin

Yolk

Griffin

a

Yolk

Thiamine form and embryo mortalitya FT TP TPP TT FT TP TPP TT Embryo mortality FT TP TPP TT Embryo

Control 68 2 4 76 1,047 41 21 1,125 25 1,041 6 2 1,052 36

6 6 6 6 6 6 6 6 6 6 6 6 6 6

9.2 y 0.3 y 2.5 5.6 y 54.5 5.5 0.2 60.5 1.5 115.9 0.6 0.1 116.7 2.3

Low 2,414 7 9 2,436 1,122 31 18 1,185 12 1,215 9 2 1,229 43

6 6 6 6 6 6 6 6 6 6 6 6 6 6

464.3 z 1.2 z 2.5 459.2 z 80.7 4.1 2.8 71.8 4.3 29.9 0.4 0.6 28.6 10.1

High 3,120 6 7 3,138 1,176 36 24 1,254 18 1,233 11 4 1,252 41

6 6 6 6 6 6 6 6 6 6 6 6 6 6

56.8 z 1.0 z 1.2 53.7 z 4.3 3.1 0.3 8.4 9.5 18.7 4.0 0.9 24.7 3.4

The thiamine forms are as follows: FT ¼ free thiamine, TP ¼ thiamine monophosphate, TPP ¼ thiamine pyrophosphate, and TT ¼ total thiamine. Embryo mortality is defined as (the number of eggs in which the embryos died/the number of eggs in the treatment group) 3 100.


size characteristics via field studies and laboratory experiments. The first phase, a case-control cohort study, suggested that PAHs, PCBs, and nonthiamine nutrients were unlikely causes of subnormal clutch viability, as their levels did not show large differences across sites nor were they significantly associated with differences among clutch survival. Furthermore, the total PCB and PAH burdens were below those known to have adverse effects on avian development (Summer et al. 1996). In contrast, TP and TPP in eggs accounted for 38% of the variation in clutch survival, lower levels of TP and TPP being associated with lower clutch viability, which is consistent with the results of similar studies involving fish (Fitzsimons et al. 1999). Deficiencies in TP and TPP are also known to cause intrauterine growth retardation in laboratory models (Roecklein et al. 1985). Lastly, the case-cohort control study indicated that thiamine levels were negatively associated with the ratio of dieldrin to the total OCP burden. This association suggests that OCPs affect thiamine levels, as OCP exposure depressed thiamine levels in laboratory models (Yagi et al. 1979). The results of the expanded field study differed from those of the case-control cohort study in the amount of variation in clutch survival accounted for by egg thiamine concentrations. For example, the expanded field study found that TT, TPP, and TP accounted for only 27% of the variation in clutch survival, whereas the case-control cohort study found that TP and TPP accounted for 38%. Another finding of the expanded field study was that the thiamine concentrations in eggs were weakly associated with the OCP burdens and lipid content of egg yolk. Specifically, depressed thiamine levels were associated with elevated lipid content in egg yolks. Some of the maternal factors that may affect egg yolk composition include maternal liver function, OCP exposure, body condition, and dietary factors. Indeed, laboratory studies of walking catfish Clarias batrachus have shown that exposure to similar pesticides altered liver function, leading to altered yolk composition (Lal and Singh 1987). Another factor to consider is that diet and body condition differed among the alligators from two of the lakes (Lakes Apopka and Griffin) in the present study (Rice 2004). In contrast to the two field studies, the laboratory egg treatment experiments did not support the hypothesis that thiamine is related to embryo viability in alligators, as neither thiamine amelioration nor antagonism altered embryo mortality rates. Possible reasons for this are that the thiamine levels were already sufficient for embryo survival, thiamine is not a significant factor in alligator egg viability, and the

259

amounts of thiamine or oxythiamine transferred to the yolk were below the effect threshold. In regard to the transfer of thiamine to the egg yolk, we were successful in increasing the level in the albumin, but our data indicate that thiamine did not move into the yolk compartment and might not have been available to the embryo. Indeed, other studies have also noted problems in achieving target egg yolk concentrations in reptilian eggs via topical treatments (Muller et al. 2007a, 2007b). Although our thiamine and oxythiamine treatment experiments did not indicate any effects of treatment, similar studies involving treatment of fish eggs have demonstrated the effects of induced thiamine deficiency and thiamine amelioration on embryo and fry survival (Fitzsimons et al. 2001). The disparity between the findings of the present alligator egg treatment studies and fish studies may be due to species-specific differences in thiamine requirements or differences in the ability to deliver an effective dose to the embryo. In conclusion, low thiamine levels in American alligator eggs may be associated with subnormal clutch viability, and the lipid content and OCP burden of the eggs may account for some of the variation in thiamine concentration. However, thiamine levels only explained 38% of the variation in clutch survival in the case-control cohort study and 27% in the expanded field study, suggesting that other factors are also involved. Because of the lack of effects in the experimental egg treatment studies, no clear cause– effect relationships can yet be established. Given the difficulty of achieving target thiamine concentrations in yolk via topical egg treatments and the fact that alligator embryos are relatively advanced in development at the time of oviposition (Clarke 1891), future efforts should consider maternally mediated exposures and mechanisms. Similar maternally mediated exposures to OCPs have been conducted with captive alligators (Rauschenberger et al. 2004a), and these showed a cause–effect relationship between OCPs and clutch viability (Rauschenberger et al. 2007). Additional studies with thiamine may elucidate its role in alligator development and embryo mortality. Acknowledgments We recognize Jesse Grosso, Eileen Monck, Shane Ruessler, Carla Wieser, Alfred Harvey, Nancy Szabo, Carolyn Dı´az, Nicola Kernaghen, Jennifer Muller, Jessica Noggle, Beverly Arnold, Ken Portier, Raymond Littell, Jon Maul, Allan Woodward, and the late Travis Smith for their important contributions. Funding was provided by Lake County Water Authority through a grant awarded to M.S.S. and by National Institutes of Environmental Health Services Superfund basic re-

260


search program grant P42ES-07375 to T.S.G. Alligator eggs were collected under a Florida Fish and Wildlife Conservation Commission permit, and all animal work was conducted according to guidelines set forth by the U.S. Geological Survey and University of Florida Institutional Animal Care and Use Committees. The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the U.S. Fish and Wildlife Service. References Akerman, G., U. Tjarnlund, E. Noaksson, and L. Balk. 1998. Studies with oxythiamine to mimic reproduction disorders among fish early life stages. Marine Environmental Research 46:493–497. Ashworth, C. J., and C. Antipatis. 2001. Micronutrient programming of development throughout gestation. Reproduction 122:527–535. Brown, S., D. C. Honeyfield, and D. L. Vandenbyllaardt. 1998. Thiamine analysis in fish tissues. Pages 73–81 in G. McDonald, J. D. Fitzsimons, and D. C. Honeyfield, editors. Early life stage mortality syndrome in fishes of the Great Lakes and Baltic Sea. American Fisheries Society, Symposium 21, Bethesda, Maryland. Clarke, S. F. 1891. The habits and embryology of the American alligator. Journal of Morphology 5:181–205. de Roode, D. F., M. B. Gustavsson, A. L. Rantalainen, A. V. Klomp, J. H. Koeman, and A. T. Bosveld. 2002a. Embryotoxic potential of persistent organic pollutants extracted from tissues of guillemots (Uria aalge) from the Baltic Sea and the Atlantic Ocean. Environmental Toxicology and Chemistry 21:2401–2411. de Roode, D. F., P. J. Vuorinen, and A. T. Bosveld. 2002b. Effects of furazolidone, PCB77, PCB126, Aroclor 1248, paraquat and p,p 0 -DDE on transketolase activity in embryonal chicken brain. Toxicology 173:203–210. Ferguson, M. W. J. 1985. The reproductive biology and embryology of crocodilians. Pages 329–491 in C. Gans, editor. Biology of the Reptilia, volume 14. Development. Wiley, New York. Fitzsimons, J. D., S. B. Brown, D. C. Honeyfield, and J. G. Hnath. 1999. A review of early mortality syndrome (EMS) in Great Lakes salmonids: relationship with thiamine deficiency. Ambio 28:9–15. Fitzsimons, J. D., L. Vandenbyllaardt, and S. B. Brown. 2001. The use of thiamine and thiamine antagonists to investigate the etiology of early mortality syndrome in lake trout (Salvelinus namaycush). Aquatic Toxicology 52:229–239. Hoffman, D. J. 1990. Embryotoxicity and teratogenicity of environmental contaminants to bird eggs. Reviews of Environmental Contamination and Toxicology 115:40–89. Horwitz, W. 2000. Official methods of analysis of AOAC International. AOAC International, Gaithersburg, Maryland. Lal, B., and T. P. Singh. 1987. Impact of pesticides on lipid metabolism in the freshwater catfish, Clarias batrachus, during the vitellogenic phase of its annual reproductive cycle. Ecotoxicology and Environmental Safety 13:13– 23.

Lance, V., T. Joanene, and L. McNease. 1983. Zinc, vitamin E, and trace elements in the plasma of wild and farmreared alligators during the reproductive cycle. Canadian Journal of Zoology 61:1744–1751. McEvoy, T. G., C. J. Robinson, J. A. Ashworth, J. A. Rooke, and K. D. Sinclair. 2001. Feed and forage toxicants affecting embryo survival and fetal development. Theriogenology 55:113–129. Muller, J. K., T. S. Gross, and C. J. Borgert. 2007a. Topical dose delivery in the reptilian egg treatment model. Environmental Toxicology and Chemistry 26:914–919. Muller, J. K., J. E. Scarborough, M. S. Sepu´lveda, G. Casella, T. S. Gross, and C. J. Borgert. 2007b. Dose verification after topical treatment of alligator (Alligator mississippiensis) eggs. Environmental Toxicology and Chemistry 26:908–913. Rao, C. R. 1964. The use and interpretation of principal component analysis in applied research. Sankhya A 26:329–358. Rauschenberger, R. H., J. J. Wiebe, J. E. Buckland, J. T. Smith, M. S. Sepu´lveda, and T. S. Gross. 2004a. Achieving environmentally relevant organochlorine pesticide concentrations in eggs through maternal exposure in Alligator mississippiensis. Marine Environmental Research 58:851–856. Rauschenberger, R. H., J. J. Wiebe, M. S. Sepu´lveda, J. E. Buckland, and T. S. Gross. 2004b. Predicting maternal body burdens of organochlorine pesticides from eggs and evidence of maternal transfer in Alligator mississippiensis. Environmental Toxicology and Chemistry 23:2906– 2915. Rauschenberger, R. H., J. J. Wiebe, M. S. Sepu´lveda, J. E. Scarborough, and T. S. Gross. 2007. Parental exposure to pesticides and reduced clutch viability in American alligators. Environmental Science and Technology 41:5559–5563. Rice, A. K. 2004. Diet and condition of alligators (Alligator mississippiensis) in three central Florida lakes. Master’s thesis. University of Florida, Gainesville. Roecklein, B., S. W. Levin, M. ComLy, and A. B. Mukherjee. 1985. Intrauterine growth retardation induced by thiamine deficiency and pyrithiamine during pregnancy in the rat. American Journal of Obstetrics and Gynecology 151:455–460. Rotstein, D. S., T. R. Schoeb, L. M. Davis, T. C. Glenn, B. S. Arnold, and T. S. Gross. 2002. Detection by microsatellite analysis of early embryonic mortality in an alligator population in Florida. Journal of Wildlife Diseases 38:160–165. SAS Institute. 2002. SAS version 9.0. SAS Institute, Cary, North Carolina. Sepu´lveda, M. S., R. H. Rauschenberger, J. J. Wiebe, D. Honeyfield, J. Hinterkopf, W. E. Johnson, and T. S. Gross. 2004. Relationship between thiamine concentrations and organochlorine pesticides in eggs of largemouth bass and the American alligator and early life stage mortalities. Journal of Wildlife Diseases 40:782– 786. Summer, C. L., J. P. Giesy, S. J. Bursian, J. A. Render, T. J. Kubiak, P. D. Jones, D. A. Verbrugge, and R. J. Aulerich. 1996. Effects induced by feeding organochlorinecontaminated carp from Saginaw Bay, Lake Huron, to


laying white leghorn hens, II. Embryotoxic and teratogenic effects. Journal of Toxicology and Environmental Health 49:409–438. ter Braak, C. J. F. 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology 67:1167–1179. ter Braak, C. J. F. 1995. Ordination. Pages 91–173 in R. H. G. Jongman, C. J. F. ter Braak, and O. F. R. Van Tongeren, editors. Data analysis in community and landscape ecology. Cambridge University Press, Cambridge, UK. ter Braak, C. J. F., and P. Smilauer. 2002. CANOCO reference manual and Canodraw for Windows users’ guide: software for canonical community ordination (version 4.52). Microcomputer Power, Ithaca, New York. U.S. Department of Health. 2005a. Oxythiamine HCl: substance summary. Available: http://pubchem.ncbi. nlm.nih.gov/. (April 2009.)

261

U.S. Department of Health. 2005b. Thiamine HCl: substance summary. Available: http://pubchem.ncbi.nlm.nih.gov/. (April 2009.) Wilson, H. R. 1997. Effects of maternal nutrition on hatchability. Poultry Science 76:134–143. Woodward, A. R., M. L. Jennings, and H. F. Percival. 1989. Egg collecting and hatch rates of American alligator eggs in Florida. Wildlife Society Bulletin 17:124–130. Woodward, A. R., H. F. Percival, M. L. Jennings, and C. T. Moore. 1993. Low clutch viability of American alligators on Lake Apopka. Florida Scientist 56:52–63. Yagi, M., K. Kamohara, and Y. Itowaka. 1979. Thiamine deficiency induced by polychlorinated biphenyls (PCB) and dichlorodiphenyltrichloroethane (DDT) administration to rats. Journal of Environmental Pathology and Toxicology 2:1119–1125.