Citation
Developmental Mortality in American Alligators (Alligator mississippiensis) Exposed to Organochlorine Pesticides

Material Information

Title:
Developmental Mortality in American Alligators (Alligator mississippiensis) Exposed to Organochlorine Pesticides
Creator:
RAUSCHENBERGER, RICHARD HEATH
Copyright Date:
2008

Subjects

Subjects / Keywords:
Alligators ( jstor )
Clutches ( jstor )
Egg masses ( jstor )
Eggs ( jstor )
Embryos ( jstor )
Fecundity ( jstor )
Female animals ( jstor )
Mortality ( jstor )
Pesticides ( jstor )
Viability ( jstor )
Lake Apopka ( local )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Richard Heath Rauschenberger. Permission granted to University of Florida to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
12/18/2004
Resource Identifier:
71355636 ( OCLC )

Downloads

This item has the following downloads:

rauschenberger_r ( .pdf )

rauschenberger_r_Page_099.txt

rauschenberger_r_Page_137.txt

rauschenberger_r_Page_013.txt

rauschenberger_r_Page_131.txt

rauschenberger_r_Page_193.txt

rauschenberger_r_Page_155.txt

rauschenberger_r_Page_025.txt

rauschenberger_r_Page_023.txt

rauschenberger_r_Page_084.txt

rauschenberger_r_Page_076.txt

rauschenberger_r_Page_033.txt

rauschenberger_r_Page_113.txt

rauschenberger_r_Page_210.txt

rauschenberger_r_Page_126.txt

rauschenberger_r_Page_121.txt

rauschenberger_r_Page_158.txt

rauschenberger_r_Page_206.txt

rauschenberger_r_Page_204.txt

rauschenberger_r_Page_233.txt

rauschenberger_r_Page_095.txt

rauschenberger_r_Page_065.txt

rauschenberger_r_Page_173.txt

rauschenberger_r_Page_182.txt

rauschenberger_r_Page_037.txt

rauschenberger_r_Page_177.txt

rauschenberger_r_Page_102.txt

rauschenberger_r_Page_119.txt

rauschenberger_r_Page_194.txt

rauschenberger_r_Page_092.txt

rauschenberger_r_Page_003.txt

rauschenberger_r_Page_061.txt

rauschenberger_r_Page_007.txt

rauschenberger_r_Page_051.txt

rauschenberger_r_Page_198.txt

rauschenberger_r_Page_191.txt

rauschenberger_r_Page_189.txt

rauschenberger_r_Page_120.txt

rauschenberger_r_Page_229.txt

rauschenberger_r_Page_030.txt

rauschenberger_r_Page_020.txt

rauschenberger_r_Page_063.txt

rauschenberger_r_Page_002.txt

rauschenberger_r_Page_050.txt

rauschenberger_r_Page_100.txt

rauschenberger_r_Page_221.txt

rauschenberger_r_Page_145.txt

rauschenberger_r_Page_219.txt

rauschenberger_r_Page_163.txt

rauschenberger_r_Page_215.txt

rauschenberger_r_Page_039.txt

rauschenberger_r_Page_180.txt

rauschenberger_r_Page_167.txt

rauschenberger_r_Page_060.txt

rauschenberger_r_Page_227.txt

rauschenberger_r_Page_055.txt

rauschenberger_r_Page_021.txt

rauschenberger_r_Page_052.txt

rauschenberger_r_Page_207.txt

rauschenberger_r_Page_116.txt

rauschenberger_r_Page_212.txt

rauschenberger_r_Page_105.txt

rauschenberger_r_Page_006.txt

rauschenberger_r_Page_089.txt

rauschenberger_r_Page_083.txt

rauschenberger_r_Page_115.txt

rauschenberger_r_Page_195.txt

rauschenberger_r_Page_224.txt

rauschenberger_r_Page_008.txt

rauschenberger_r_Page_026.txt

rauschenberger_r_Page_217.txt

rauschenberger_r_Page_134.txt

rauschenberger_r_Page_074.txt

rauschenberger_r_Page_097.txt

rauschenberger_r_Page_152.txt

rauschenberger_r_Page_149.txt

rauschenberger_r_Page_098.txt

rauschenberger_r_Page_109.txt

rauschenberger_r_Page_111.txt

rauschenberger_r_Page_132.txt

rauschenberger_r_Page_058.txt

rauschenberger_r_Page_112.txt

rauschenberger_r_Page_016.txt

rauschenberger_r_Page_205.txt

rauschenberger_r_Page_213.txt

rauschenberger_r_Page_069.txt

rauschenberger_r_Page_057.txt

rauschenberger_r_Page_125.txt

rauschenberger_r_Page_029.txt

rauschenberger_r_Page_091.txt

rauschenberger_r_Page_159.txt

rauschenberger_r_Page_070.txt

rauschenberger_r_Page_135.txt

rauschenberger_r_Page_117.txt

rauschenberger_r_Page_190.txt

rauschenberger_r_Page_162.txt

rauschenberger_r_Page_019.txt

rauschenberger_r_Page_176.txt

rauschenberger_r_Page_144.txt

rauschenberger_r_Page_130.txt

rauschenberger_r_Page_202.txt

rauschenberger_r_Page_072.txt

rauschenberger_r_Page_073.txt

rauschenberger_r_Page_047.txt

rauschenberger_r_Page_077.txt

rauschenberger_r_Page_101.txt

rauschenberger_r_Page_054.txt

rauschenberger_r_Page_011.txt

rauschenberger_r_Page_150.txt

rauschenberger_r_Page_164.txt

rauschenberger_r_Page_080.txt

rauschenberger_r_Page_179.txt

rauschenberger_r_Page_032.txt

rauschenberger_r_Page_171.txt

rauschenberger_r_Page_031.txt

rauschenberger_r_Page_044.txt

rauschenberger_r_Page_168.txt

rauschenberger_r_Page_185.txt

rauschenberger_r_Page_174.txt

rauschenberger_r_Page_043.txt

rauschenberger_r_Page_088.txt

rauschenberger_r_Page_108.txt

rauschenberger_r_Page_184.txt

rauschenberger_r_Page_049.txt

rauschenberger_r_Page_228.txt

rauschenberger_r_Page_188.txt

rauschenberger_r_Page_106.txt

rauschenberger_r_Page_046.txt

rauschenberger_r_Page_012.txt

rauschenberger_r_Page_209.txt

rauschenberger_r_Page_087.txt

rauschenberger_r_Page_172.txt

rauschenberger_r_Page_139.txt

rauschenberger_r_Page_200.txt

rauschenberger_r_Page_181.txt

rauschenberger_r_Page_004.txt

rauschenberger_r_Page_133.txt

rauschenberger_r_Page_140.txt

rauschenberger_r_Page_123.txt

rauschenberger_r_Page_024.txt

rauschenberger_r_Page_220.txt

rauschenberger_r_Page_165.txt

rauschenberger_r_Page_085.txt

rauschenberger_r_Page_042.txt

rauschenberger_r_Page_009.txt

rauschenberger_r_Page_230.txt

rauschenberger_r_Page_216.txt

rauschenberger_r_Page_079.txt

rauschenberger_r_Page_211.txt

rauschenberger_r_Page_169.txt

rauschenberger_r_Page_166.txt

rauschenberger_r_Page_045.txt

rauschenberger_r_Page_199.txt

rauschenberger_r_Page_071.txt

rauschenberger_r_Page_005.txt

rauschenberger_r_Page_018.txt

rauschenberger_r_Page_231.txt

rauschenberger_r_Page_086.txt

rauschenberger_r_Page_059.txt

rauschenberger_r_Page_107.txt

rauschenberger_r_Page_053.txt

rauschenberger_r_Page_118.txt

rauschenberger_r_pdf.txt

rauschenberger_r_Page_124.txt

rauschenberger_r_Page_068.txt

rauschenberger_r_Page_225.txt

rauschenberger_r_Page_170.txt

rauschenberger_r_Page_094.txt

rauschenberger_r_Page_082.txt

rauschenberger_r_Page_183.txt

rauschenberger_r_Page_186.txt

rauschenberger_r_Page_067.txt

rauschenberger_r_Page_138.txt

rauschenberger_r_Page_156.txt

rauschenberger_r_Page_104.txt

rauschenberger_r_Page_110.txt

rauschenberger_r_Page_075.txt

rauschenberger_r_Page_036.txt

rauschenberger_r_Page_178.txt

rauschenberger_r_Page_034.txt

rauschenberger_r_Page_218.txt

rauschenberger_r_Page_161.txt

rauschenberger_r_Page_232.txt

rauschenberger_r_Page_223.txt

rauschenberger_r_Page_201.txt

rauschenberger_r_Page_226.txt

rauschenberger_r_Page_027.txt

rauschenberger_r_Page_129.txt

rauschenberger_r_Page_090.txt

rauschenberger_r_Page_154.txt

rauschenberger_r_Page_038.txt

rauschenberger_r_Page_014.txt

rauschenberger_r_Page_197.txt

rauschenberger_r_Page_187.txt

rauschenberger_r_Page_028.txt

rauschenberger_r_Page_066.txt

rauschenberger_r_Page_148.txt

rauschenberger_r_Page_103.txt

rauschenberger_r_Page_048.txt

rauschenberger_r_Page_017.txt

rauschenberger_r_Page_064.txt

rauschenberger_r_Page_062.txt

rauschenberger_r_Page_022.txt

rauschenberger_r_Page_081.txt

rauschenberger_r_Page_093.txt

rauschenberger_r_Page_015.txt

rauschenberger_r_Page_141.txt

rauschenberger_r_Page_035.txt

rauschenberger_r_Page_160.txt

rauschenberger_r_Page_175.txt

rauschenberger_r_Page_192.txt

rauschenberger_r_Page_128.txt

rauschenberger_r_Page_203.txt

rauschenberger_r_Page_056.txt

rauschenberger_r_Page_078.txt

rauschenberger_r_Page_222.txt

rauschenberger_r_Page_127.txt

rauschenberger_r_Page_122.txt

rauschenberger_r_Page_214.txt

rauschenberger_r_Page_010.txt

rauschenberger_r_Page_208.txt

rauschenberger_r_Page_143.txt

rauschenberger_r_Page_196.txt

rauschenberger_r_Page_151.txt

rauschenberger_r_Page_157.txt

rauschenberger_r_Page_041.txt

rauschenberger_r_Page_136.txt

rauschenberger_r_Page_153.txt

rauschenberger_r_Page_040.txt

rauschenberger_r_Page_096.txt

rauschenberger_r_Page_146.txt

rauschenberger_r_Page_147.txt

rauschenberger_r_Page_142.txt

rauschenberger_r_Page_001.txt

rauschenberger_r_Page_114.txt


Full Text










DEVELOPMENTAL MORTALITY IN AMERICAN ALLIGATORS (Alligator
mississippiensis) EXPOSED TO ORGANOCHLORINE PESTICIDES














By

RICHARD HEATH RAUSCHENBERGER


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2004

































Copyright 2004

by

Richard Heath Rauschenberger

































To Jesus, my personal Lord and Savior. John 14:6. "Jesus said to him, I am the
way, the truth, and the life: no man comes to the Father, except by me." Ephesians 2:8-9.
"For by grace you have been saved through faith, and that not of yourselves; it is the gift
of God, Not of works, lest anyone should boast."















ACKNOWLEDGMENTS

I thank my wonderful wife, Alison; and my two sons, Heath and Ben. Their

steadfast love, support, and sacrifices allowed me to successfully complete the arduous

task of earning a Ph.D. I thank my parents, Richard E. and Mary Elizabeth

Rauschenberger, for their ever-present love, faith, and encouragement. I thank my

mother-in-law, Sandra Pillow, for baby-sitting Heath and Ben while Alison and I were

away at work and for her support and encouragement. I thank my parents-in-law,

Tommy and Debbie Kirk, for their love and ever-vigilant prayers. I thank my brother-in-

law, Matt Kirk; and sister-in-law, Kristin Dessert; for their support and encouragement. I

thank my late grandfather, M. E. "Pappy" Walls, for showing me the outdoors; my high

school biology teacher, Joe David White, for making me a better student; and my high

school football coaches, Randy Tapley and Jim Massarelli, for strengthening my work

ethic and ability to deal with adversity. I am forever grateful to Tim Gross for taking me

in as a student. I thank him and his wife, Denise, for the kindness, generosity, and

encouragement they've shown to my family and me. I thank my committee members

(Marisol Sepulveda; Bill Castleman; Richard Miles, Jr.; Franklin Percival; and Steve

Roberts) for their support, friendship, and significant contributions to my development as

a research scientist. I also want to thank Kent Vliet for sharing his vast literature and

knowledge of alligator reproduction. I especially thank Jon Wiebe and Janet Buckland

for their friendship and hard work. I am privileged to have had the opportunity to work

with the staff and students of our laboratory. I thank Wendy Mathis, Travis Smith, Jesse









Grosso, Eileen Monck, James Basto, Shane Ruessler, Carla Wieser, Alfred Harvey,

Adriano Fazio, Nikki Kernaghen, Jennifer Muller, and Jessica Noggle for their help and

friendship. I thank Ken Portier, Gary Stevens, Ramon Littell, Ron Marks, Jon Maul, and

Linda Garzarella for providing statistical advice and assistance. I thank the National

Institutes of Environmental Health Sciences Superfund Basic Research Program (grant

number P42ES-07375) and the Lake County Water Authority for providing financial

support for my education and research project. My name is listed alone as the author of

this dissertation, but this work was the product of a team that I am honored to have been a

part of and will always remember.














TABLE OF CONTENTS
Page
A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES ..................................................... ix

LIST OF FIGURES ............................. ............ .................................... xii

CHAPTER

1 IN T R O D U C T IO N .................................................................. .. ... .... ............... 1

H habitat D egradation in the Ocklaw aha Basin.......................................... ...............1...
A lligators as Potential O CP R eceptors.................................................... ...............2...
Developmental Biology of the American Alligator................................................3...
Post-O vipositional D evelopm ent...................................................... ...............6...
Ferguson's Post-Ovipositional Staging Scheme .............................................8...
Organochlorine Pesticide Toxicity in Vertebrates ................................................ 15
Classification, Mode of Action, and Pathology............................................. 15
Exposure and Effects of OCPs in Crocodilians............................................. 17
Reproductive Problems in Florida Alligators ................................................ 18
S p ecific A im s.............................................................................................................. 2 0

2 EGG AND EMBRYO QUALITY OF ALLIGATORS FROM REFERENCE AND
ORGANOCHLORINE CONTAMINTED HABITATS ................. ..................... 23

M materials and M ethods ............................................ ........................... ................ 24
Egg C collections and Incubation...................................................... ................ 24
A analysis of O CPs in Y olk ........................................................ 26
G C/M S A analysis ... .. ................................. ......................................... 27
D ata A n aly sis....................................................................................................... 2 8
R e su lts.......................................................................................................... ........ 3 0
Inter-Site Comparisons of Clutch Characteristics ..........................................30
Organochlorine Pesticides Burdens and Clutch Characteristics ......................31
Clutch Survival and OCP Burdens in Egg Yolks...........................................32
Average Egg Mass, Clutch Size and OCP Burdens ......................................33
D isc u ssio n .. ............... ........ ... .... ........................................................... 3 4
Inter-Site Comparisons of Clutch Characteristics .........................................34
Clutch Survival Parameters and OCP Burdens ..............................................36
Egg and Clutch Size and OCP Burdens ......................................... ................ 38









3 MATERNAL TRANSFER OF ORGANOCHLORINE PESTICIDES.................. 54

M materials an d M eth o d s ...............................................................................................5 5
Site descriptions ..................................................................... ............... 55
A nim al C collections ....................................................... .... ......... ..... .. ........ ..... 56
Analysis of OCPs in Maternal Tissues and Yolk...........................................57
G C/M S A analysis ... .. ................................. ......................................... 59
D ata A n aly sis....................................................................................................... 6 0
R e su lts........................................... ... ............................................................. ........ 6 1
Female Morphological and Reproductive Characteristics ................................61
O C P concentrations in Y olk ........................................................... ................ 62
O CP concentrations in m aternal tissues ......................................... ................ 62
Relationships between Maternal Tissue and Yolk Burdens...............................63
Relationships between Maternal Mass and OCP concentrations in Eggs and
T issue e s......................................................................................................... . 6 4
D isc u ssio n ................................................................................................................... 6 5
Evaluation of Predictive M odels .................................................... ................ 67
Relationships between Maternal Mass and OCP concentrations in Eggs and
Tissues.................. ............ . .... .. ............... 69
Maternal body burdens: Toxicological Implications.....................................69

4 MATERNAL FACTORS ASSOCIATED WITH DEVELOPMENTAL
MORTALITY IN THE AMERICAN ALLIGATOR............................................80

M materials and M ethods .. ..................................................................... ................ 81
Site D escriptions......................................................................................... 82
A nim al C collections ....................................................... .... ......... ..... .. ........ ..... 82
Analysis of OCPs in Maternal Tissues and Yolk...........................................83
G C/M S A analysis ... .. ................................. ............................. ............ 86
D ata A n aly sis....................................................................................................... 8 7
R e su lts....................................................................................................... ....... .. 8 8
D isc u ssio n ............................................................................................................... ... 8 9

5 MORPHOLOGY AND HISTOPATHOLOGY OF AMERICAN ALLIGATOR
(Alligator mississippiensis) EMBRYOS FROM REFERENCE AND OCP-
CON TAM IN A TED H ABITA TS .......................................................... ................ 99

M materials and M ethods ................... .............................................................. 102
Site D escriptions... ................................................................... ........... 102
E gg C collections ........................................................................................... 103
Embryo Sam pling and M easurem ent ....... .......... ....................................... 103
H istopathology ..................................................................................... 105
A analysis of O CPs in Y olk ........................................................ 106
G C /M S A naly sis ... ................................................................................. 108
R esults.................................... .................................... ................... 109
Inter-Site C lutch C om prisons ..................................................... ................ 109
Intra-Site Live Embryo/Dead Embryo Morphological Comparisons .............110









Inter-Site Comparisons of Morphology of Live Embryos .............. ................111
Live Embryo Morphology and Embryo Survival Relationships.................... 112
Live Embryo Morphology and Egg Yolk OCP Burdens................................ 113
Embryo Morphological Age, Derived Morphometric Variables and Egg Yolk
O C P B u rd e n s .............. ..... ...... ...................................................................... 1 1 5
Histopathology of Live and Dead Embryos ......................... .................. 116
D isc u ssio n ............................................................................................................. .. 1 1 7

6 NUTRIENT AND CHLORINATED HYDROCARBON CONCENTRATIONS IN
AMERICAN ALLIGATOR EGGS AND ASSOCIATIONS WITH DECREASED
C L U T C H V IA B IL IT Y ............................................................................................. 143

M materials and M ethods ........................................ .......................... ............... 145
Egg Collections and Incubation...... ........ ......................145
F ield stu dies ..................................................................... ............ 14 6
Laboratory experiments............................................ 147
Analysis of Chlorinated Hydrocarbons in Yolk..................... .................. 149
G C /M S A analysis ... .................................................................. ............... 150
Nutrient Analysis ........................................................................ 151
D ata A n aly sis..................................................................................................... 152
R e su lts..................................................................................................... ......... 15 4
F ield S tu d y ..................................... .................................................................... 1 5 4
Case-control cohort study...... ........... ........ ..................... 154
E expanded fi eld study ....................................................... ............... 157
Laboratory Experim ents ........................................................ 160
D isc u ssio n ............................................................................................................... .. 1 6 2

7 REPRODUCTIVE EFFECTS OF ORGANOCHLORINE PESTICIDE EXPOSURE
IN A CAPTIVE POPULATION OF AMERICAN ALLIGATORS (Alligator
m ississip p ien sis) ....................................................................................................... 1 8 2

M materials and M ethods ................... .............................................................. 182
R e su lts..................................................................................................... ......... 18 5
D isc u ssio n ............................................................................................................. .. 1 8 6

8 CONCLUSIONS ..................................................... ...... .... ............... 196

In tro d u ctio n .............................................................................................................. 19 6
Sum m ary of Study's Findings ............................................................. 197
Future Considerations and Global Implications .............................................204

LIST OF REFERENCES .......................................................... ............ 208

BIOGRAPH ICAL SKETCH .................. .............................................................. 217















LIST OF TABLES


Table page

2-1. Reproductive, morphometric, and contaminant parameters measured on clutches of
alligator eggs collected during summer 2000, 2001, and 2002...............................41

2-2. Explanatory variables included in RDA with forward selection of four best
v a riab le s ............................................................................................................. .. 4 2

2-3. Summary of clutch parameters and site comparisons for clutches of American
alligator eggs collected during 2000-2002. ............... .................................... 43

2-4. Organochlorine pesticide burdens and clutch parameters and site comparisons for
clutches of American alligator eggs collected during 2000-2002.........................44

2-5. Results of RDA evaluating associations between clutch survival parameters and
O C P v ariab les........................................................................................................... 4 7

2-6. Results of RDA evaluating associations between egg and clutch size parameters and
O C P v ariab les........................................................................................................... 4 8

3-1. Morphological and reproductive characteristics of adult female alligators collected
during June 2001 and 2002 from Lakes Apopka, Griffin, and Lochloosa in central
F lo rid a .................................................................. ............................................... ... 7 3

3-2. Pesticide concentrations (ng/g wet wt.) in tissues and yolks of adult female alligators
collected during June 2001 and 2002 from Lakes Apopka, Griffin, and Lochloosa
in cen trial F lo rid a ..................................................................................................... 7 4

3-3. Regression equations for predicting organochlorine pesticide (OCP) concentrations
in m atern al tissue es .................................................................................................... 7 8

4-1. Reproductive, morphometric, and contaminant parameters measured on adult female
alligators collected during June 1999, 2000, 2001, and 2002..............................93

4-2. Explanatory variables included in RDA with forward selection of four best
v a riab le s ............................................................................................................. .. 9 4

4-3. Reproductive, morphometric, and contaminant summary statistics of adult female
alligators collected during June of 1999-2002. ................................... ................ 95









4-4. Results of redundancy analysis with automatic selection of four best maternal
factors associated with variation in reproductive efficiency...............................97

4-5. Results of redundancy analysis with automatic selection of four best maternal
factors associated with variation in clutch size characteristics ..............................97

5-1. Summary statistics for parameters measured on American alligator clutches
collected during June 2001 and 2002. ...... ... ........................................... 122

5-2. Comparisons of egg and embryo morphometrics of live and dead embryos collected
during June-August of 2001 and 2002. .......... ... ......................................... 124

5-3. Morphometric comparisons of live embryos collected during June-August 2001 and
2 0 0 2 ...................................................................................................... .......... 12 8

5-4. Explanatory variables included in partial redundancy analysis evaluating
relationship between organochlorine pesticide burdens in eggs and embryo
m orp h om etrics........................................................................................................ 13 1

5-6. Best five organochlorine pesticide (OCP) variables that account for embryo
morphological age and derived morphological parameters .............................133

6-1. Classification matrix for clutches collected during 2002 ................................165

6-2. Reproductive, morphometric, and contaminant parameters measured on clutches of
alligator eggs collected during summer 2000, 2001, and 2002........................... 165

6-3. Explanatory variables included in RDA with forward selection of four best variables
for case-control cohort and expanded field studies........................................166

6-4. Summary of clutch parameters on clutches collected during 2002 .....................168

6-5. Evaluation of the relationship between concentrations of nutrients, PAHs, and PCBs
in eggs and clutch survival parameters via RDA analysis. ...............................169

6-6. Evaluation of clutch size parameters and explanatory factors for clutches collected
d u rin g 2 0 0 2 ........................................................................................................... 16 9

6-7. Evaluation of the relationship between nutrient concentrations and explanatory
variables for clutches collected during 2002....... ... ..................................... 169

6-8. Summary and comparison of parameters measured on clutches collected during
2 0 0 0 -2 0 0 2 .............................................................................................................. 17 0

6-9. Evaluation of the relationships between clutch survival parameters and explanatory
variables via RDA using age as the covariate. ...... ... ................................... 171

6-10. Evaluation of the relationships between clutch size parameters and explanatory
variables via RDA using age as the covariate. ...... ... ................................... 171









6-11. Evaluation of the relationships between thiamine concentrations and explanatory
variables via RDA using age as the covariate. ...... ... ................ ................... 172

6-12. Site comparisons of parameters measured on clutches collected during 2003...... 173

7-1. Summary statistics and comparisons of clutch parameters among treated and control
groups for years 2002-2004................................... ...................... ............... 192

7-2. Organochlorine concentrations and blood chemistry values of captive adult female
alligators sacrificed during 2002. ...... ....... ........ ...................... 194

7-3. Explanatory parameters and clutch survival parameters with () indicating nature of
association and value equal to concordance percentage. .................................195















LIST OF FIGURES


Figure page

1-1. M ap of O cklaw aha B asin. ................. ............................................................. 22

2-1. Biplot of clutch survival parameters (solid lines) and organochlorine pesticide
variables (dashed lines) for clutches of alligator eggs collected from Lake
Lochloosa during sum m er 2001-2002................................................. ................ 49

2-2. Biplot of clutch survival parameters (solid lines) and organochlorine pesticide
variables (dashed lines) for clutches of alligator eggs collected from Lake Griffin
during sum m er 2000-2002 ........................................ ....................... ................ 50

2-3. Biplot of clutch survival parameters (solid lines) and organochlorine pesticide
variables (dashed lines) for clutches of alligator eggs collected from Lake Apopka
during sum m er 2000-2002 ........................................ ....................... ................ 5 1

2-4. Biplot of clutch survival parameters (solid lines) and organochlorine pesticide
variables (dashed lines) for clutches of alligator eggs collected from Emeralda
M arsh during sum m er 2000-2002 ...................................................... ................ 52

2-5. Biplot of egg and clutch size parameters (solid lines) and organochlorine pesticide
variables (dashed lines) for clutches of alligator eggs collected from Lake
Lochloosa during sum m er 2001 and 2002 ......................................... ................ 53

3-1. Linear regressions of total organochlorine pesticide (OCP) concentrations in
maternal tissues against total OCP concentrations in egg yolks. ..........................79

4-1. Biplot of maternal factors (dashed lines) and clutch survival parameters (solid lines)
of American alligators collected during June 1999-2002. ...............................98

5-1. Representative developmental stages of embryos that were collected from Lakes
Lochloosa (reference site), Apopka, and Griffin, and Emeralda Marsh during 2001-
2 0 0 2 .................................................................................................................... 1 3 4

5-2. Ordination biplot of embryo morphometric parameters (solid lines) and
organochlorine pesticide (OCP) variables (dashed lines) for embryos collected at
chronological age D ay 14..................................... ....................... ............... 135









5-3. Ordination biplot of embryo morphometric parameters (solid lines) and
organochlorine pesticide (OCP) variables (dashed lines) for embryos collected at
chronological age D ay 25 ........ ............................... ..................... 136

5-4. Ordination biplot of embryo morphometric parameters (solid lines) and
organochlorine pesticide (OCP) variables (dashed lines) for embryos collected at
chronological age D ay 33 .................. ......................................................... 137

5-5. Ordination biplot of embryo morphometric parameters (solid lines) and
organochlorine pesticide (OCP) variables (dashed lines) for embryos collected at
chronological age D ay 43 .................. ......................................................... 138

5-6. Ordination biplot of derived embryo morphometric parameters (solid lines) and
organochlorine pesticide (OCP) variables (dashed lines) for embryos collected at
chronological age D ay 14 .................................... ....................... ............... 139

5-7. Ordination biplot of derived embryo morphometric parameters (solid lines) and
organochlorine pesticide (OCP) variables (dashed lines) for embryos collected at
chronological age D ay 25 .................................... ....................... ............... 140

5-8. Ordination biplot of derived embryo morphometric parameters (solid lines) and
organochlorine pesticide (OCP) variables (dashed lines) for embryos collected at
chronological age D ay 33 .................. ......................................................... 141

5-9. Ordination biplot of derived embryo morphometric parameters (solid lines) and
organochlorine pesticide (OCP) variables (dashed lines) embryos collected at
chronological age D ay 43 .................. ......................................................... 142

6-1. Biplot of clutch survival parameters and explanatory factors for clutches collected
d u rin g 2 0 0 2 ........................................................................................................... 17 5

6-2. Biplot of clutch size parameters and explanatory variables for clutches collected
d u rin g 2 0 0 2 ........................................................................................................... 17 6

6-3. Biplot of nutrient concentrations in eggs (solid arrows) and explanatory variables
(d ash ed arrow s). ..................................................................................................... 17 7

6-4. Relationships between embryo age and thiamine phosphorylation in egg yolk for 29
clutches collected during 2002 from Lakes Lochloosa, Griffin, Apopka, and
Em eralda M arsh. .......... .. .... ............................. .............. ............... 178

6-5. Biplot of clutch survival parameters and explanatory variables for clutches collected
during 2000-2002.. ............... .......... .. ..... ............................... 179

6-6. Biplot of clutch size variables (solid lines) and explanatory variables (dashed lines)
for clutches collected during 2000-2002 ....... ... ....................................... 180









6-7. Biplot of thiamine egg yolk concentrations (solid lines) and explanatory variables
(dashed lines) measured on clutches collected during 2000-2003 .......................181















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

DEVELOPMENTAL MORTALITY IN AMERICAN ALLIGATORS (Alligator
mississippiensis) EXPOSED TO ORGANOCHLORINE PESTICIDES

By

Richard Heath Rauschenberger

December 2004

Chair: Timothy S. Gross
Major Department: Veterinary Medicine-Physiological Sciences

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 degradation has been the American alligator (Alligator mississippiensis).

Decreased clutch viability (proportion of eggs in a nest that yield a live hatchling) was

observed in the years after a chemical spill in which large amounts of sulfuric acid and

dicofol, an organochlorine pesticide (OCP), flowed into Lake Apopka. Lake Apopka and

other lakes in the Ocklawaha basin have also been contaminated by urban sewage and

agricultural chemicals, with agricultural chemicals entering the lakes via rainfall run-off

or back-pumping of water from agricultural lands). Decreased hatch rates are a problem

at Lake Apopka, as well as at other OCP-contaminated sites in Florida. The purpose of

my study was to determine the causes for decreased clutch viability, and to test the

hypothesis that maternal exposure to OCPs is associated with embryonic mortality in

alligators.









Field studies involved collecting and artificially incubating eggs from reference

sites (Lake Lochloosa) and from OCP-contaminated sites (Lakes Apopka, Griffin, and

Emeralda Marsh Restoration Area) to evaluate clutch viability as a function of egg and

maternal OCP concentrations. Nutrient content of eggs and histopathology and

morphometrics of embryos were also evaluated to identify potential factors associated

with embryo mortality. In addition, a novel laboratory experiment exposed a captive

population of adult alligators to an OCP mixture, and compared OCP burdens in eggs and

clutch viability with a captive control group.

Results of field studies suggested that OCP concentrations (ng total OCP/g egg

yolk, Mean SE) in reference site clutches (n = 19; 102 16) were significantly (a =

0.05) lower than those of Apopka (n = 23; 7,582 2,008), Griffin (n = 42; 1,169 423),

and Emeralda Marsh (n = 31; 15,480 + 2,265). Clutches from reference sites also had

significantly higher clutch viability (70 4%) than those of Apopka (51 6%), Griffin

(44 5%), and Emeralda Marsh (48 6%). Furthermore, decreased thiamine

concentrations in eggs may play a role in decreased clutch viability in wild clutches.

Results of the captive study suggested that treated females produced eggs containing

higher OCP concentrations (n = 7; 13,300 2,666) than controls (n = 9; 50 4). Eggs of

treated females also exhibited decreased viability (9 6%) as compared to controls (44 +

11%). These field and laboratory studies support the hypothesis that maternal exposure

to OCPs is associated with decreased clutch viability in American alligators, and that

thiamine deficiency may also be a contributing factor in reduced clutch viability.















CHAPTER 1
INTRODUCTION

Habitat Degradation in the Ocklawaha Basin

In central Florida, several lakes within the Ocklawaha River Basin (Fig. 1-1) have

experienced severe degradation of habitat quality since the early 1900s, as agricultural

and urban development progressed. Indeed, Lake Apopka (headwaters of the

Ocklawaha) was once renowned for its clear water and its excellent largemouth bass

fishing. More recently, Lake Apopka has gained world-wide notoriety as the "poster

child" for polluted lakes, because of highly publicized problems associated with

environmental contamination. Initial degradation of Lake Apopka and other lakes within

the Ocklawaha Basin occurred as the result of the loss of thousands of hectares of marsh

habitat through the agricultural practice known as muck farming (which involves

installing levees around an area of marsh, so the marsh can be drained; allowing the

fertile peat to be farmed). This farming practice began in the 1940s and continued into

the 1980s (Benton et al., 1991). In addition to sewer discharge from the city of Winter

Garden entering the Lake Apopka, organochlorine pesticides (OCPs) were heavily and

widely used to control crop-destroying insect pests.

Since the 1980s, use of most OCPs has been discontinued since they were

determined to be persistent environmental contaminants that resist metabolic degradation

and bioaccumulate in animal tissues, where they are potentially carcinogenic,

immunotoxic, endocrine disrupting, and developmentally toxic (Fairbrother et al., 1999;









Ecobichon, 2001). Altered function of the reproductive and endocrine systems of

wildlife and human populations have been suggested to occur after exposure to a variety

of OCPs and OCP metabolites such as dichlorodiphenyltrichloroethane (DDT),

dichlorodiphenyltrichloroethylene (DDE), methoxychlor, dicofol, chlordane, dieldrin,

and toxaphene (Colborn et al., 1993; Longnecker et al., 2002).

Further degradation and OCP contamination occurred in Lake Apopka in 1980. A

chemical spill occurred when a highly acidic wastewater pond at the Tower Chemical

Company's main facility overflowed into the Gourd Neck area of Lake Apopka (Fig. 1-

1). Because of the large volume and acidity (sulfuric acid), and the high levels of DDT,

dicofol, and related OCP contaminants that entered the relatively narrow area of the lake,

aquatic vegetation and animals were severely affected. In 1983, the area was placed on

the US Environmental Protection Agency's (EPA) National Priority Site List and became

a part of the Superfund program; which was created by the Comprehensive

Environmental Response, Compensation, and Liability Act (CERCLA), later amended by

the Superfund Amendments and Reauthorization Act (SARA). The CERCLA and SARA

provide authority for the government to respond to the release and/or threat of release of

hazardous wastes, and allow cleanup and enforcement actions. Lake Apopka is still

listed and groundwater toxicity testing is ongoing (EPA, 2004).

Alligators as Potential OCP Receptors

The American alligator is an important member of Florida wetlands and plays

important roles in the ecology, esthetics, and economy of Florida. Therefore, identifying

physiological and ecological characteristics related to potential susceptibility to effects of

contaminants, as well as potential exposure routes, is important in managing populations









for optimal human use. Especially important to consider, in regard to wildlife

populations, are potential effects of OCPs on reproduction.

One of the first qualities that may be related to an alligator's susceptibility to

reproductive effects of OCP contaminants is that alligators do not attain sexual maturity

until approximately 6-10 years of age, which allows exposure and bioaccumulation of

OCPs to occur before reproductive maturity. Potential implications are that, as females

begin to mobilize body stores during vitellogenesis, the lipophilic OCPs that have

accumulated in their fatty tissues during their lifespan would likely be deposited in what

will later be the embryos' sole source of nutrition (egg yolk). Secondly, adults exhibit a

long reproductive period (over 30 years), and a long life span (over 50 years) (Ferguson,

1985), and are higher order predators (which allows for increased OCP exposure and

bioaccumulation, possibly leading to altered endocrine and reproductive function).

Thirdly, alligators build nests that can be identified from considerable distances (which

aids in egg collections), lay a large number of eggs (approximately 40 eggs per clutch),

and have a long developmental period of 65-72 days (Ferguson, 1985), allowing extended

exposure at a potentially critical stage of development. Thus, the propensity for OCPs to

be bioaccumulated and biomagnified in biota (combined with the alligator's reproductive

biology, longevity, ecological trophic level, and relatively long in ovo developmental

period) suggests the potential for OCPs to alter reproductive function.

Developmental Biology of the American Alligator

Understanding normal embryonic development is an obvious necessity in

determining the occurrence of abnormal embryonic development and identifying critical

periods of development (e.g., organogenesis). Therefore, this brief review summarizes

pre-ovipositional and post-ovipositional development of the alligator embryo.









Pre-ovipositional Development

Overall, when compared to other vertebrate species such as the domestic chicken

and domestic pig, there is a paucity of data related to crocodilian development. Despite

the relatively low number of publications, the quality of papers covering early embryonic

development is fairly high, considering that much of the research took place

approximately a century ago. The most appropriate place to begin discussing embryonic

development would be the point when fertilization occurs. However, the precise timing

and location of fertilization within a female alligator's oviduct is unknown and

inadequately studied.

Pre-ovipositional development has been examined by sacrificing gravid females

and collecting their eggs and embryos. Sacrifice of gravid females was required since

alligator embryos are at a more advanced stage of development at the time of oviposition

(Clarke, 1891). The earliest developmental stage examined in these pre-ovipositional

studies were of Nile crocodile embryos (Crocodylus niloticus), in which all embryos

exhibited body folds, a neural medullary groove, an embryonic shield, area opaca, early

gut, and area pellucida (Voeltzkow, 1892).

After the appearance of the neural folds, the amniotic head fold is formed from an

anterior fold in the blastoderm. The head fold is crescent shaped, because it begins to

develop with its free ends pointing toward the posterior end of the embryo, and develops

craniocaudally. The amniotic primordium develops in continuity with the head, and is

derived from the somatopleure around the trunk.

Craniocaudal separation of the embryo from the blastoderm occurs partly as a

result of the development of the dorsal amniotic fold, but separation is not complete until









post-ovipositional stage 3 (Day 3). The neural groove and blastopore become clearly

demarcated as the ectoderm and endoderm of the blastoderm develop. The endoderm

may form extensions that penetrate the underlying yolk. The blastopore goes through the

entire blastoderm, with the primitive streak located posterior to the blastopore

(Voeltzkow, 1892).

As the body folds develop, the border between embryonic and extra-embryonic

tissues becomes visible. At this point, the beginning of the foregut is discernable, and the

notochord stretches from the midline of the head fold to the anterior border of the

blastopore. The primitive streak and primitive groove lie posterior to the blastopore, with

the primitive groove being continuous at its posterior end. The primitive streak extends

to a little less than halfway between the head fold and blastopore (Ferguson, 1985).

Neural folds have two origins. The first is a secondary fold located anteriorly in

the head region, and growing posteriorly along the median dorsal line to form a V-shaped

process, with the apex pointing toward the blastopore. The second is posterior folds that

arise as ectodermal ridges extending forward from the blastopore, circumventing the

neural groove. The apex of the V-shaped secondary head fold later disappears, and each

of the separate arms becomes continuous with the corresponding posterior neural fold.

Thus, the secondary head fold forms the anterior part of the neural folds. Closure of the

folds occurs first in the middle region of the embryo closer to the anterior end of the

neural groove in alligators (Ferguson, 1985) but closer to the posterior end in Nile

crocodiles (Voeltzkow, 1892).

After the closure of the neural canal, the blastoporal or neurenteric canal is no

longer visible. The neurenteric canal runs from its posterior cranioventral opening to









where it opens into the neural groove at its caudal limit. During this period, somites

develop along the median axis, with the first pair developing halfway between the

anterior and posterior ends. The peripheral somatic cells are compactly arranged, and

contain small myocoels within the center of the somites. The mesodermal layers cleave

and form the somatic and splanchnic components as the foregut develops.

The head fold of the embryo is positioned ventrally into the underlying yolk,

which is accentuated by the bending of the anterior neural folds, and by the cranial

flexure that occurs later. At this pre-ovipositional stage of development, the embryo has

not yet attached to the inner surface of the eggshell membrane.

Because embryos are at an advanced stage of development at the time of

oviposition (and because an entire clutch typically hatches within a 2-day period, with

most hatchlings being similar in size), it appears that fertilization occurs over a short time

period; and that embryos are actively developing during the next 2- to 3-week period in

which the ova receive albumin, eggshell membrane, and eggshell depositions (Ferguson,

1985). Presently, little information exists about gaseous exchange and embryonic

metabolism before oviposition, or about the processes that prevent the embryo from

attaching to the top of the egg before oviposition.

Post-Ovipositional Development

Post-ovipositional development is better understood than pre-ovipositional

development. Again, the amount of literature concerning crocodilian development is

miniscule compared to the amount of literature dealing with human and chicken

embryology.

One important area to address when discussing post-ovipositional development is

the staging scheme. Establishing a staging scheme or a normal table of development for









any species allows results of various studies to be compared (Billet et al., 1985). The

currently accepted staging scheme for crocodilian embryology was proposed by Ferguson

(1985). Before Ferguson's, the only other staging systems related to crocodilians came

from Voeltzkow (1892), Reese (1912), and Webb et al. (1983). These works were

impressive, considering the conditions these pioneers faced; but many stages were

missing, and incubation conditions were poorly controlled.

Ferguson (1985) improved on their work by monitoring and controlling the

temperature (300C) and the relative humidity (95-100%) at which the eggs were

incubated, allowing duplication of his experiment and standardization of the

characteristics one should see in an embryo, given its stage. This accepted staging

scheme is based on external morphological features, with limb and eye development

being important diagnostic elements. With respect to craniofacial development, a fair

amount of data exists, because of Ferguson's focus on the structure and development of

the palate in the alligator, and on how its development relates to stage (Ferguson, 1981).

Although the relationship between craniofacial development and developmental stage has

been studied, information relating stage and development in other organ systems is

somewhat lacking.

Alligator embryos are very sensitive to temperature. For example, 26-340C is the

optimum incubation temperature; anything above or below for an extended period will

result in increased mortality (Ferguson, 1985). Furthermore, 0.5-1 C changes can mean

the difference between an entire clutch of embryos being 100% females or 100% males,

since crocodilians exhibit temperature-dependent sex determination (Lang & Andrews,

1994).









Ferguson's Post-Ovipositional Staging Scheme

Because our study used Ferguson's staging scheme, a summary description of

Ferguson's (1985) staging scheme, it is summarized here. The normal table of

development for crocodilians was based on examination of 1500 Alligator

mississippiensis embryos, 300 Crocodylusporosus embryos, and 300 Crocodylus

johnsoni embryos. One bias is that all of the alligator embryos used in developing this

scheme originated from Rockefeller Wildlife Refuge, located in southern Louisiana.

Alligator embryos from other geographic areas may develop at different rates, given the

same incubation conditions. Alligators inhabiting Arkansas and North Carolina

experience a shorter summer compared to populations inhabiting southern Louisiana or

Florida. Shorter summers mean that optimal nest temperatures are maintained for a

shorter period of time. Thus, embryos from more northerly latitudes may develop at an

increased rate compared to embryos from southerly latitudes (given identical incubation

conditions), since the northern embryos must complete development within a shorter time

frame. This hypothesis is supported by evidence that crocodilian species (Crocodylus

porosus and C. johnsoni) living along the equator have longer and more variable

incubation periods and slower embryonic development than the (more northerly)

Louisiana alligator (Deeming & Ferguson, 1990).

Setting aside the potential bias described above, developmental "stages" are

determined by morphological characteristics alone, and are applicable to embryos

regardless of incubation temperature. However, the developmental day(s) associated

with each stage are only valid if the eggs are incubated at 300C with a relative humidity

of 95-100%. Temperatures lower than 300C slow the rate of development, and

temperatures above 300C have been shown to increase the rate of development. Low









humidity within the nest has been shown to dehydrate eggs, causing embryonic mortality

and alterations in growth patterns (Deeming & Ferguson, 1990).

Stage 1 covers the period from oviposition to the end of the first 24 hours, and is

characterized by the embryo and blastoderm being not attached to the top of the inner

eggshell membrane. The heart is a simple S-shaped tube. There are 16-18 pairs of

somites along the trunk, and 3 pairs of somitomeres anterior to the otic vesicle. Although

the brain has not yet regionalized, optic placodes and vesicles are present on the head.

Body torsion has not begun. The notocord is evident, the gut is incomplete caudally and

opens ventrally, and blood vessels are not present in the extraembryonic membranes.

Stage 2 (Day 2) embryos have 21-25 pairs of somites and a three-loop heart.

However, one of the most notable characteristics is that the embryo attaches to the top of

the egg, causing an opaque spot to form that is visible in an otherwise translucent egg,

when the egg is candled. Blood vessels are now visible, and the hindbrain is discernable

as a clear transparent region. The lens placode and optic cup are defined, and no body

torsion has occurred.

Stage 3 (Day 3) embryos have 26-30 somites, and are completely delineated from

blastoderm. Forebrain, midbrain, and hindbrain are now discernable, and the optic cup

has an elongated horseshoe shape, extending below the lens vesicle to the primitive

oronasal cavity. The head is positioned at a right angle to the body, but no body torsion

has occurred.

Stage 4 (Day 4) embryos have 31-35 pairs of somites with the tail being distinct,

straight, and unsegmented at the posterior end. Body torsion has started, with the cranial

half rotated so that the right surface is contacting the shell membrane, while the left is









parallel with underlying yolk. The caudal half of the embryo remains at a right angle to

the yolk. The heart is displaced from midline to the left side of the embryo. Three

cranial arches are present; and cranial nerves to the cranial arches are visible, using

oblique or transmitted illumination.

Stage 5 (Day 5) embryos have 36-40 pairs of somites, and the tail-tip bends

ventrally at a right angle to the body, with 3-5 somites visible at its base. Body torsion is

complete except for the tail. The otic pit is dorsal to the junction of the 2nd and 3rd

brachial arches, and its external opening is closed.

Stage 6 (Day 6) embryos have visible nasal placodes, and the hindlimbs are barely

discernable on each side; with the right hind limb slightly advanced over the left.

Forelimb buds are not yet present, and body torsion is complete. The olfactory bulbs,

forebrain, and midbrain are distinct. In the hindbrain, 4-6 neuromeres are discernible.

Foregut and hindgut are formed, but midgut is incomplete ventrally. Major vitellogenic

blood vessels emerge at the level of the 18th somite and smaller ones at the 6th and 11th

somites.

Embryos at Stage 7 (Day 7) have distinct hind limb buds. In addition, forelimb

buds are barely visible and extend over somites 12-15. The midbrain bulge is evident,

and the tail-tip is curled at 900 to the rest of the tail. Three brachial arches are present;

and at the level of the heart, the cranial end is bent at 900 to the rest of body.

Embryos at stage 8 (Day 8) have nasal pits external to the swellings of the olfactory

bulbs, and distinct forelimb and hind limb buds that extend over somites 11-16 and 27-

32, respectively. An apical ectodermal ridge is developing on the hind limb bud, and the

tail is coiled through 2 turns and has 12-18 somites.









Stage 9 (Day 9) embryos have four brachial arches, and a visible maxillary process

extending to the midpoint of the eye. The optic cup is large and round but unpigmented.

A distinct apical ectodermal ridge is present on the hind limb, and the hind limb bud

extends beyond the forelimb. The tail is curled through three 900 turns. The heart

exhibits distinct atria and ventricles, and lung primordia are visible through the

pericardial sac. Midgut and body walls are open ventrally from the caudal limit of the

pericardial sac to 2/3 of the way down the body, and the liver and mesonephros are barely

visible.

Stage 10 (Day 10-11) embryos have pigmented eyes (except for a central opaque

lens) with the right eye developing pigmentation earlier and darker than the left eye. Five

brachial arches are present, and medial and lateral processes are distinct elements on each

side of the nasal pits. Maxillary processes delimit a distinct groove beneath the eye. The

tail is coiled through four 900 turns, and the liver and mesonephros are clearly visible

through the body walls.

Stage 11 (Day 12) embryos have a visible nasal pit slit forming between the medial

and lateral processes. Forelimb and hind limb buds extend caudally from the body wall

and exhibit distinct apical, ectodermal ridges. The forelimb has a distinct constriction

that separates the distal and proximal elements, with constriction less obvious in the hind

limb. A loop of midgut is visible at the umbilicus, the eye exhibits a distinct black

pigment in the iris, and the chorioallantois extends 2/3 around the breadth of the shell

membrane.

Embryos at stage 12 (Day 13-14) have a distinct notch in the midline of the face

between the medial nasal processes. Forelimbs are starting to bend in the region of









constriction, so that they are positioned closer to the pleuron of the embryo. The

elongated hind limb shows little differentiation into proximal and distal elements and,

although there is a distinct apical ectodermal ridge, footplate formation is barely

discernable.

Stage 13 (Day 15) embryos have distinct nasal pit slits, and forelimbs are now bent

toward the pericardium. The distal portion of the hind limb is flattened and enlarged into

a footplate primordium. The chorioallantois now extends as a ring around the inner

circumference of the central eggshell membrane.

Embryos at Stage 14 (Day 16-17) have nasal pit slits that are closed due to the

merging of the medial nasal, lateral nasal, and maxillary processes. Foot and hand plates

are distinct, with the former more advanced than the latter. Lower jaw extends one-

quarter beneath the upper jaw, the upper earflap is overgrowing the external ear opening,

and the embryonic face rests on the large bulge of the thorax. A large loop of gut

herniates through the narrow umbilical stalk and touches the yolk, and the abdominal

viscera are visible through body walls. The tail is coiled and kinked at the tip, and

contralateral reflexes occur.

Stage 15 (Day 18-20) embryos have lower jaws that extend one-third to one-half

the length of the upper jaw. The anlage for the upper eyelid is an elevated rim of tissue

above each eye. Distinct and proximal and distal regions, as well as hand and foot plates

are present on both the fore and hind limb. There is a distinct hollow in the face beneath

the anterior one-third of the eye.

Stage 16 (Day 21) embryos exhibit faint digital condensations in the footplate but

not the hand plate. The lower jaw is now two-thirds the length of the upper jaw, with the









upper jaw being hook-shaped around the pericardial ridge. Caruncle development is

observed, with two tiny widely spaced thickenings that are just discernable on the tip of

the snout.

Embryos at stage 17 (Day 22-23) exhibit mesodermal condensations for the five

forelimb digits and four hind limb digits, the head is extended off of the pericardial sac

by neck elongation, and the external earflap is distinct.

Stage 18 (Day 24-26) embryos have discernable, distinct cartilaginous digital rays

on the hand and foot. The margins of upper eyelid anlage extend over the superior rim of

the iris, forming a distinct groove between the eyelids and the eye. Dorsal scalation is

now evident, and the pericardial sac is starting to submerge into the ventral thoracic wall.

Stage 19 (Day 27-28) embryos have upper and lower eyelids, and the lower jaw lies

behind the anterior margin of the upper jaw. Interdigital clefting has started, and slight

marginal notches can be seen, particularly in the footplates. White flecks representing

ossification are visible around the upper and lower jaws.

Stage 20 (Day 29-30) embryos have nail anlages starting to develop, first on the

most medial digit of the foot, then on adjacent digits; followed by the most medial digit

on the hand, and finally on the adjacent hand digits. Interdigital clefting now extends

one-quarter the length of the digits, and the lower jaw is in adult relationship with the

upper jaw. The pericardial sac is one-quarter withdrawn into the body, and ossification is

evident in the proximal and distal elements of limbs. Scale formation is evident dorsally,

and scutes (osteoderms) are beginning to appear in the neck region near the skull.

Stage 21 (Day 31-35) embryos have interdigital clefting now extending three-

quarters down the digits, and phalanges can be distinguished. Scales are now visible on









the ventral body wall; and dorsally on the snout, neck, body, and tail. Scutes on neck are

clearly defined. The pericardial sac is one-half withdrawn into the body, and a white ring

in the iris surrounds the outline of the lens of the eye. Both upper and lower eyelids

overlap the eye.

Stage 22 (Day 36-40) embryos have pigmented margins of the upper jaw, ventral

flank, and proximal and distal elements of the limbs. Interdigital clefting is at the adult

level, and the eyelids are typically closed from this point forward. The pericardial sac is

two-thirds withdrawn.

Stage 23 (Day 41-45) embryos have more extensive pigmentation, with the

embryos appearing light brown with dorsal stripes. Scales are present on distal and

proximal elements, and nails have a slight distal elevation. The sensory papillae are

present on lateral jaw margins, and scales are evident on gular skin. The midbrain is

visible as a white bulge at the back of the cranium, and the pericardial sac is three-

quarters withdrawn.

Stage 24 (Day 46-50) embryos are blacker. Nails on hands have elevations at

their tips, and the nails are starting to form curves. The midbrain is covered by

pigmented skin. The pericardial sac is fully withdrawn and the midline is closing. The

volume of yolk outside the body cavity is large, and scales and scutes are evident all over

embryo.

Stage 25 (Day 51-60) embryos look identical to hatchlings, except smaller. The

external yolk is beginning to be withdrawn, and few gross morphological changes are

evident at this and later stages. Growth relationships (head length: total length ratio) and

the amount of external yolk present are the major observable differences.









Stage 26 is not present in alligators. This stage was established using tooth

eruption sequences and is useful only for saltwater crocodiles (Crocodylusporosus) and

freshwater crocodiles (Crocodylusjohnsoni).

Stage 27 (Day 61-63) embryos have withdrawn the yolk sack into the body, ending

with skin forming across the umbilical scar. The last stage before hatching (Stage 28,

Day 64-70) ends with the umbilical scar being diminished in length and width.

Overall, the first 35 days are a period of rapid organogenesis, and the second 35

days are characterized by embryo growth. Since organogenesis has been shown to be a

sensitive period in regard to effects of developmental toxicants (Schmidt & Johnson,

1997), the first 35 days of incubation appear to be the most susceptible time points for

toxicant-induced mortality.

In summary, the established staging scheme provides a way to estimate the age of

the clutch at the time of collection, and allows one to later determine if a clutch is

undergoing normal development. One can determining if a clutch is undergoing normal

development by examining embryos at pre-selected time points and comparing their

morphological age to their calendar stage (i.e., does an embryo exhibit the normal

morphological characteristics that it should exhibit, given its calendar age?). In addition,

embryonic development may be compared among clutches and among populations, by

collecting embryos at pre-determined stages of development.

Organochlorine Pesticide Toxicity in Vertebrates

Classification, Mode of Action, and Pathology

Organochlorine pesticides (also known as chlorinated hydrocarbon insecticides)

may be separated into five classes of compounds. These classes are DDT and its analogs,

cyclodienes and similar compounds, toxaphene (composed of several congeners), mirex









and chlordecone (which have cage-like structures), and benzene hexachloride (BHC). In

rodent models, studies suggest that OCPs can adversely affect the function of neurons

and cause cellular damage to the liver and kidneys (Smith, 1991). Organochlorine

pesticides affect neural transmission by altering enzyme activity (Ca2+-ATPase,

phospokinase) and the electrophysical properties (K+, Na+ ion exchange) of nerve cell

membranes. Different analytes may elicit similar effects (neuronal hyperactivity), but by

different mechanisms. For example, studies suggest DDT and its analogs affect the nerve

axon by keeping Na+ channels open longer than normal. Cyclodienes, alternatively, may

affect neural transmission at presynaptic terminals and may affect the y-aminobutyric

acid (GABA)-regulated chloride channel. Although they can cause severe neural

dysfunction, little morphological changes are evident in neural tissue, even at lethal doses

(Smith, 1991).

Morphological changes are evident in the liver and include hepatocellular

hypertrophy and focal necrosis. Hypertrophy is due to enlargement of the smooth

endoplasmic reticulum (SER) and formation of a lipid droplet in the center of the SER

(caused by OCP-induced expression of microsomal enzymes within the SER).

Functional alterations may also occur in hepatocytes, with disruption of intercellular

communication (by hindering transfer of growth inhibitors) (Smith, 1991).

Morphological changes have also been found in the liver and kidney of fish

chronically exposed to organochlorine pesticides. For example, chronic exposure to

OCPs induce hepatic lesions, such as foci of vacuolated hepatocytes and spongiosis

hepatic (lesions of hepatic parenchyma). Renal lesions induced by chronic OCP









exposure include dilation of tubular lumina, and vacuolization (degeneration) and

necrosis of tubular epithelium (Metcalfe, 1998).

In addition to morphological changes, organochlorine pesticides may adversely

affect endocrine and reproductive function in laboratory models and wildlife populations.

Mechanisms include direct toxicity on endocrine glands (such as o,p'-DDD's ability to

permanently inactivate the adrenals), competitive binding of steroid hormone receptors,

increased expression of steroid-metabolizing hepatic microsomal enzymes, and inhibition

of hormone synthesis (such as DDE-induced inhibition of proglandin synthesis, leading

to eggshell thinning in raptors) (Gross et al., 2003).

Exposure and Effects of OCPs in Crocodilians

Current knowledge on the effects of environmental contaminants on crocodilian

reproductive physiology is important in understanding the likelihood of developmental

alterations occurring as a result of exposure; and understanding which mechanisms may

be involved.

Campbell (2003) reviewed the effects of organic and inorganic contaminants on

crocodilians. Campbell reported only 26 studies related to the bioaccumulation of

organic contaminants, with just 35% (8/23) of crocodilian species being represented. Of

the 26 studies, 38% involved American alligators (Alligator mississippiensis), 26%

involved Nile crocodiles (Crocodylus niloticus), 13% involved American crocodiles

(Crocodylus acutus), and 13% involved Morolet's crocodile (Crocodylus moreletii).

Slightly more studies were found that investigated effects of organic contaminants. With

respect to these 39 studies, only 13% (3/23) of crocodilian species were represented,

consisting of the American alligator (91% of studies), the Nile crocodile (5%), and the

African dwarf crocodile (Osteolaemus tetraspis, 4%). Of these studies, American









alligators are the only species in which an effort has been made to determine the

relationship between OCPs and depressed hatch rates, with most of this work involving

populations in central Florida.

Reproductive Problems in Florida Alligators

In the early to mid 1980s, studies showed that the population of juvenile alligators

inhabiting the aquatic ecosystem of Lake Apopka, Florida, declined by 90%. This

decline was preceded by a 1980 chemical spill and decades of OCP contamination via

anthropogenic activities described earlier. The loss of juveniles was attributed primarily

to a dramatic decrease in clutch viability (the proportion of eggs in a clutch that produce a

live hatchling) (Woodward et al., 1993).

Alterations in sexual differentiation, sex steroid hormone concentrations, and

metabolism were also documented among Lake Apopka alligators. For example,

testosterone was lower in male alligators from Lake Apopka as compared to those of

control sites. Ovaries of female juvenile alligators from Lake Apopka showed

abnormalities, suggesting that reproductive alterations were occurring in both sexes

(Gross et al., 1994; Guillette et al., 1994; Gross, 1997). In addition, high concentrations

of OCPs were measured in egg yolk, but concentrations were not clearly associated with

increased mortality (Heinz et al., 1991).

Later studies suggested that the cause for the population decline was potentially

more complex than previously suggested. First, poor egg viability for Lake Apopka

alligators was more closely associated with muck farm reclamation (wetland restoration)

sites than with tissue and egg concentrations of the predominant pesticide residue (DDE)

(Giroux, 1998). Second, altered endocrine function and decreased egg viability were

documented among alligators at another site, Lake Griffin, where tissue and egg









concentrations of residues such as DDE are modest or intermediate compared with those

of Lake Apopka. However, like Lake Apopka, Lake Griffin is highly eutrophic and has

adjacent muck farms and muck farm reclamation areas (Marburger et al., 1999). Third,

poor reproductive success among Lake Apopka alligators appeared to result from both

decreased proportions of fertile eggs that produce a live hatchling and decreased

proportions of hatchlings that survive through the first 20 days of life (which is the

toxicant-sensitive organogenesis period); and decreased proportions of unbanded eggs

(i.e., eggs that are nonviable on initial examination) (Masson, 1995; Wiebe et al., 2001).

Unbanded eggs show no evidence of embryo-eggshell attachment (as indicated by

the presence of an opaque spot or band that results from fusion of extraembryonic

membranes to the dorsal portion of the inner eggshell membrane). Unbanded eggs may

result from very early embryo mortality (fertilization has been confirmed in many cases

by the presence of paternal DNA, via DNA microsatellite analysis); or may result from

infertile eggs (Rotstein, 2000).

The last similarity between alterations in alligator populations of Lake Griffin and

Lake Apopka is increased mortality among adult Lake Griffin alligators (Schoeb et al.,

2002), which is similar to increased adult mortality on Lake Apopka in the early 1980s.

These data indicate that alligator populations are adversely affected at each of several life

stages. Although anatomic and endocrinologic effects of exposure to

endocrine-disrupting OCPs could account for many of these effects, additional

underlying mechanisms are almost certainly involved. Overall, these data point to a

complex process involving the introduction of OCPs into these aquatic ecosystems from









chemical spillage or from muck farming and reclamation activities; possibly leading to

developmental toxicity, in addition to endocrine disruption.

Specific Aims

The overall objective of our study was to determine the causes of decreased hatch

rates among alligators in contaminated sites, and to determine if causal links could be

established between specific adverse effects and exposure to individual OCPs or

combinations of OCPs. The project consisted of epidemiological field studies, which

evaluated embryonic development and mortality as a function of maternal and

environmental exposure to OCPs and egg nutrient composition; and controlled laboratory

experiments to test hypothesized links between decreased hatch rates, altered egg

composition, and exposure to selected OCPs.

Specific aim 1: Conduct field epidemiological studies to determine the relative

contributions of unbanded eggs, embryonic mortality in banded eggs, and decreased

perinatal mortality to the overall decreased reproductive success in alligators at OCP-

contaminated sites, to determine which OCPs or combinations of OCPs are most closely

associated with adverse effects at each life stage, and to examine the relationship between

OCP burdens in maternal tissues and eggs. For Specific Aim 1, the hypotheses were

H1a: Adverse effects at early life stages are associated with muck farm
environments, exposure to specific OCPs or OCP combinations, or both;

Hlb: Specific OCPs found in maternal tissues are highly correlated to those present
in eggs indicating maternal transfer of OCPs and that maternal size is correlated with
OCP burdens and hatch rates;

H1c: Eggs in which embryonic and perinatal mortality occur result from
developmental abnormalities, altered structure or composition of the egg, or both.

Specific aim 2: Conduct controlled in ovo and in vivo experiments with alligators

to confirm causal links between decreased hatch rates and affected life stages as a






21


function of exposure to selected OCPs or altered egg qualities, or both. For Specific Aim

2, the hypotheses were

H2a: Exposing a captive breeding population of adult alligators to an
environmentally relevant mixture of OCPs will elicit OCP concentrations in eggs and
developmental effects similar to those observed in wild eggs from OCP-contaminated
field sites;

H2b: Exogenous in ovo alteration of egg nutrients based on data from field studies
will alter embryonic development.



























































Figure 1-1. Map of Ocklawaha Basin.














CHAPTER 2
EGG AND EMBRYO QUALITY OF ALLIGATORS FROM REFERENCE AND
ORGANOCHLORINE CONTAMINATED HABITATS

In the southeastern US, aquatic ecosystems have experienced habitat degradation,

alterations in water quality, and in some cases, important declines in biodiversity due to

increases in land development and associated anthropogenic impacts. A case-in-point is

the Ocklawaha River Basin in central Florida. Within this basin, American alligators

(Alligator mississippiensis) from impacted lakes have exhibited poor clutch viability

(number eggs that yield a live hatchling / total number of eggs found in clutch) (Masson,

1995), abnormal reproductive hormone concentrations (Gross et al., 1994), and

unexplained adult mortality (Schoeb et al., 2002). During the mid 1980s, clutches from

alligators on Lake Apopka experienced severe declines in clutch viability (declined from

50% to 4%), and alligator clutches from other impacted lakes had only moderate

viabilities (range of 40 to 60%). These rates were below those observed in other less

impacted Florida lakes (reference sites), including Lake Woodruff National Wildlife

Refuge (79%), Orange Lake (82%), and the Everglades Water Conservation Areas (65-

75%) (Woodward et al., 1993; Masson, 1995; Rice, 1996).

Possible causal factors for reduced hatch rates in alligator populations within the

impacted sites within the Ocklawaha River Basin include pesticides, algal toxins,

nutritional changes, density-related stress, and diseases. In one case, a chemical spill

from a chemical manufacturing plant in 1980 near Lake Apopka (EPA, 2004) was

temporally associated with the decline in reproductive success and consequent alligator









population decline on Lake Apopka during the early 1980s. However, decreases in

clutch viability for Lake Apopka appeared to be more related to proximity to muck farm

restoration areas as compared to yolk concentrations (Giroux, 1998), which is consistent

with decreases in clutch viability on Lake Griffin and Emeralda Marsh, Griffin's adjacent

muck farm restoration area (Sepulveda et al., 2001).

Poor reproductive success threatens the long-term conservation of alligators,

potentially altering the ecology of affected ecosystems, and substantially reducing the

aesthetic and economic values of alligators in affected areas. Understanding and

characterizing poor reproductive performance and determining associated factors is

needed so that efficacious mitigation strategies may be developed. Thus, the overall

objective of the present study was to determine the relative contributions of losses during

in ovo development in American alligators at impacted sites in central Florida, and to

evaluate whether organochlorine pesticides (OCPs) are associated with adverse

developmental effects and altered clutch characteristics.

Materials and Methods

Egg Collections and Incubation

Lakes Apopka (N 280 35', W 810 39'), Griffin (N 280 53', W 81 46'), Emeralda

Marsh Conservation Area ((N 280 55', W 810 47'), and Lochloosa (N 290 30', W 820

09') in Florida were selected as collection sites because prior studies indicate vastly

different levels of OCP exposure across these sites (Gross unpublished data, (Masson,

1995).

Alligator nests were located via aerial (helicopter) and ground surveys airboatt),

and clutches were subsequently collected by ground crews. The top of each egg was

marked before eggs were removed from the nest to ensure proper orientation; thus,









preventing embryo mortality due to inversion. Embryo mortality due to inversion occurs

if an embryo has attached to the top of the egg, inversion may either break embryonic

attachment or cause the yolk mass to settle on top of the embryo, crushing it.

After marking each egg and placing about 5 cm of nest substrate in a uniquely

numbered polypropylene pan (43 cm x 33 cm x 18 cm), all eggs found in each clutch

were placed in the pan 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 additional eggs were set.

The top layer of eggs was covered with nest substrate so that there was no space left

between the top of the pan and the top of the eggs (approximately 10 cm). Clutches were

transported to the US Geological Survey's Center for Aquatic Resources Studies,

Gainesville, Florida (CARS). Upon arrival, clutches were evaluated for embryonic

viability using a bright light candling procedure. Viable eggs (i.e. having a visible band)

were nested in pans containing moist sphagnum moss and incubated at 30.5C and -98%

humidity, in an incubation building (7.3 m x 3.7 m). This intermediate incubation

temperature will normally result in a 1:1 male/female sex ratio, since alligators have

temperature dependent sex (or gender) differentiation. One or two eggs were opened

from each clutch to identify the embryonic stage of development at the time of collection,

and to collect yolk samples for later measurement of OCP burdens. From each clutch,

information on the following parameters was collected: total number of eggs found per

nest (fecundity); number of unbanded eggs, number of damaged eggs, number of dead

banded eggs, number of live banded eggs, total clutch mass, and average egg mass of

clutch.









For years 2001 and 2002, some clutches were involved in an embryo development

study. For these clutches, each clutch was evenly divided between two pans, with half of

the clutch left relatively undisturbed (except for weekly monitoring of embryo mortality)

to determine clutch viability (the number of live hatchlings / fecundity), and the other

half of the clutch used to study embryo development and morphometry (Chapter 5).

Analysis of OCPs in Yolk

Analytical grade standards for the following compounds were purchased from the

sources indicated: aldrin, alpha-benzene hexachloride (a-BHC), P-BHC, lindane, 6-BHC,

p,p '-dichlorodiphenyldichloroethane (p,p '-DDD), p,p '-dichlorodiphenyldichloroethylene

(p,p '-DDE), dichlorodiphenyltrichloroethane (p,p '-DDT), dieldrin, endosulfan,

endosulfan II, endosulfan sulfate, endrin, endrin aldehyde, endrin ketone, heptachlor,

heptachlor epoxide, hexachlorobenzene, kepone, methoxychlor, mirex, cis-nonachlor,

and trans-nonachlor from Ultra Scientific (Kingstown, RI, USA); cis-chlordane, trans-

chlordane, and the 525, 525.1 polychlorinated biphenyl (PCB) Mix from Supelco

(Bellefonte, PA, USA); oxychlordane from Chem Service (West Chester, PA); o,p '-

DDD, o,p '-DDE, o,p '-DDT from Accustandard (New Haven, CT, USA); and toxaphene

from Restek (Bellefonte, PA, USA). All reagents were analytical grade unless otherwise

indicated. Water was doubly distilled and deionized.

Egg yolk samples were analyzed for OCP content using methods modified from

Holstege et al. (1994) and Schenck et al. (1994). For extraction, a 2 g tissue sample was

homogenized with -1 g of sodium sulfate and 8 mL of ethyl acetate. The supernatant

was decanted and filtered though a Btchner funnel lined with Whatman #4 filter paper

(Fisher Scientific, Hampton, NH, USA ) and filled to a depth of 1.25 cm with sodium

sulfate. The homogenate was extracted twice with the filtrates collected together. The









combined filtrate was concentrated to ~2 mL by rotary evaporation, and then further

concentrated until solvent-free under a stream of dry nitrogen. The residue was

reconstituted in 2 mL of acetonitrile. After vortexing (30 s), the supernatant was applied

to a C 18 solid phase extraction (SPE) cartridge (pre-conditioned with 3 mL of

acetonitrile; Agilent Technologies, Wilmington, DE, USA) and was allowed to pass

under gravity. This procedure was repeated twice with the combined eluent collected in a

culture tube. After the last addition, the cartridge was rinsed with 1 mL of acetonitrile

which was also collected. The eluent was then applied to a 0.5 g NH2 SPE cartridge

(Varian, Harbor City, CA, USA), was allowed to pass under gravity, and collected in a

graduated conical tube. The cartridge was rinsed with an additional 1 mL portion of

acetonitrile which was also collected. The combined eluents were concentrated under a

stream of dry nitrogen, to a volume of 300 [tL, and transferred to a gas chromatography

(GC) vial for analysis.

GC/MS Analysis

Analysis of all samples was performed using a Hewlett Packard HP-6890 gas

chromatograph (Wilmington, DE, USA) with a split/splitless inlet operated in splitless

mode. The analytes were introduced in a 1 [iL injection and separated across the HP-5MS

column (30 m x 0.25 mm; 0.25 [tm film thickness; J & W Scientific, Folsom, CA, USA)

under a temperature program that began at 600 C, increased at 10 C/min to 2700 C, was

held for 5 min, then increased at 250 C/min to 3000 C and was held for 5 min. Detection

utilized an HP 5973 mass spectrometer in electron impact mode. Identification for all

analytes and quantitation for toxaphene was conducted in full scan mode, where all ions

are monitored. To improve sensitivity, selected ion monitoring was used for the









quantitation for all other analytes, except kepone. The above program was used as a

screening tool for kepone which does not optimally extract with most organochlorines.

Samples found to contain kepone would be reextracted and analyzed specifically for this

compound.

For quantitation, a five-point standard curve was prepared for each analyte (r2 >

0.995). Fresh curves were analyzed with each set of twenty samples. Each standard and

sample was fortified to contain a deuterated internal standard, 5 [iL of US-108 (120

[g/mL; Ultra Scientific), added just prior to analysis. All samples also contained a

surrogate, 2 [g/mL of tetrachloroxylene (Ultra Scientific) added after homogenization.

Duplicate quality control samples were prepared and analyzed with every twenty samples

(typically at a level of 1.00 or 2.50 [g/mL of y-BHC, heptachlor, aldrin, dieldrin, endrin,

andp,p '-DDT) with an acceptable recovery ranging from 70 130%. Limit of detection

ranged from 0.1-1.5 ng/g for all OCP analytes, except toxaphene (120-236 ng/g), and

limit of quantitation was 1.5 ng/g for all analytes, except toxaphene (1500 ng/g).

Repeated analyses were conducted as allowed by matrix interference and sample

availability.

Data Analysis

Specific OCP analytes were removed from analysis if measurable concentrations

were found in < 5% of all clutches. Numerical data were log-transformed [ln(x)], while

proportional data were arcsine square root transformed to meet statistical assumptions.

ANOVA (PROC GLM; SAS Institute Inc., 2002) was used for inter-site

comparisons of adult female and clutch characteristics, and the Tukey test was used for

multiple comparisons among sites (a = 0.05). Because relationships between response

variables and explanatory variables (Table 2-1) in ecological studies are often complex









with interactions occurring, an indirect gradient multivariate analysis method, Detrended

Correspondence Analysis (DCA) (ter Braak, 1986) was used to initially evaluate

data structure. Two matrices were constructed for DCA, with the first representing the

response variables (clutch number x clutch parameters) and the second representing the

explanatory variables (clutch number x OCP burdens) (Table 2-2). DCA results

indicated that a direct gradient, multivariate linear analysis, redundancy analysis (RDA)

(Rao, 1964), was appropriate since the gradient lengths of the DCA ordination axes were

equal to or less than 2 standard deviations (ter Braak, 1995).

RDA is the canonical form of principal components analysis (PCA). In RDA, as in

PCA, a straight line is fitted to each the response variable (clutch survival parameters) in

an attempt to explain the data of all response variables. Similar to PCA, the lower the

residual sum of squares, the better the environmental variable is at explaining the

variation in response variables. RDA, unlike PCA, restricts the clutch scores (from the

response variables measured on each clutch) to a linear combination of the environmental

(explanatory variables). Because clutch scores are constrained to a linear combination of

environmental variables, RDA explains slightly less variance compared to PCA (ter

Braak & Tongeren, 1995; ter Braak, 1994). For RDA involving compositional data (i.e.,

clutch viability rates or percentages) and quantitative environmental variables,

compositional data is log-transformed (In (x + 1)) with correlation biplots being centered

by the response variables (i.e., unbanded egg percentage) and by the samples (i.e.,

clutches) (ter Braak, 1994). These correlation biplots provide a way to examine

relationships among a number of response variables and explanatory factors with

response variable arrows forming a biplot of correlations with each other, environmental









arrows forming a biplot among each other, and response variable arrows and

environmental arrows forming a biplot of correlations with each other (ter Braak, 1995).

For the RDA, separate matrices were constructed for response variables measured

as a percentage (i.e., clutch viability) and response variables measured as a number (i.e.,

clutch mass) because percentage data were ln(x+1) transformed and not standardized,

while continuous data were ln(x) transformed and standardized(ter Braak & Smilauer,

2002). Automatic forward selection of the best four explanatory variables was conducted

for both sets of RDA analyses and tested for significance by Monte Carlo permutation

tests. DCA and RDA were conducted using the program CANOCO (ter Braak &

Smilauer, 2002). Biplots of environmental variables and response variables were then

constructed to interpret relationship between clutch parameters (response variables) and

explanatory factors.

Results

Inter-Site Comparisons of Clutch Characteristics

From 2000-2002, 168 clutches were collected from Lakes Lochloosa (n = 44),

Apopka (n = 31), Griffin (n = 47), and Emeralda Marsh (n = 46). No significant

differences were determined among sites with respect to clutch mass (overall mean +

standard error: 3.7 0.08 kg), egg mass (83 + 1.4 g), or percentage of unbanded eggs (15

+ 1.7%) (Table 2-3).

In contrast, significant differences were determined among sites with respect to

fecundity, clutch viability, percentage of damaged eggs, percentage of early embryo

mortality, and percentage of late embryo mortality. Clutches from Lochloosa had lower

fecundity and late embryo mortality rates compared to all other sites. In addition,

Lochloosa clutches had greater clutch viability rates than all other sites and lower early









embryo mortality rates than all other sites, except for Apopka. Clutches from Emeralda

Marsh had greater incidence of damaged eggs than all other sites, except for those of

Lake Griffin (Table 2-3).

Organochlorine Pesticides Burdens and Clutch Characteristics

From 2000-2002, clutch characteristics and OCP burdens were measured on 115

clutches collected from Lakes Lochloosa (n = 19), Apopka (n = 23), Griffin (n = 42), and

Emeralda Marsh (n = 31). No significant differences were determined among sites with

respect to clutch mass (overall mean standard error: 3.8 0.09), clutch viability (50 +

3.1), percentage of damaged eggs (4 1), unbanded eggs (13 1.6), early embryo

mortality (21 2.3), and late embryo mortality (11 + 1.7) (Table 2-4). However,

significant differences were determined for fecundity and egg mass, with Lochloosa

clutches having lower fecundity than all other sites, and greater average egg mass

compared to those of all other sites, except for Lake Apopka. Furthermore, significant

differences were detected among sites with respect to individual OCP concentrations in

egg yolks, total OCP concentrations in egg yolks, and number of OCPs detected at

measurable levels. For total OCP concentrations and number of analytes detected at

measurable levels, egg yolks of Lake Lochloosa clutches had significantly lower total

concentrations and a lower number of analytes detected at measurable levels (Table 2-4).

Individual OCP analyte concentrations in egg yolks of Lochloosa clutches were

significantly less than those of the other sites, except for Lake Griffin with respect to

aldrin and trans-nonachlor. Aldrin and trans-nonachlor egg yolk concentrations of

Lochloosa clutches did not significantly differ from Lake Griffin, but egg burdens of

these analytes of both sites were significantly less than those of Lake Apopka and

Emeralda Marsh (Table 2-4).









Clutch Survival and OCP Burdens in Egg Yolks

Because a number of site specific factors may potentially affect clutch survival

parameters and since OCP burdens varied greatly among sites, relationships between

OCP egg yolk variables and clutch survival were evaluated on a site-by-site basis.

For Lake Lochloosa, redundancy analysis with Monte Carlo permutation tests for

significance indicated that none of the four extracted OCP variables (Table 2-5) were

found to be significantly correlated with the clutch survival variables. Number of OCPs

detected (NOC) approached significance (P = 0.07), was negatively associated with

clutch viability, positively correlated with percentage unbanded eggs and late embryo

mortality, and accounted for 11% of the variation in clutch survival parameters (Fig. 2-1).

For Lake Griffin, redundancy analysis with Monte Carlo permutation tests for

significance indicated that three of the four extracted OCP variables were found to be

significantly correlated with the clutch survival variables and together accounted for 21%

of the variance in clutch survival parameters. The extracted OCP variables were

concentration of p,p '-DDE, toxaphene, and p,p '-DDT, accounting for 8, 7, and 6%,

respectively, of variation in clutch survival variables (Table 2-5). Clutch viability was

positively associated with toxaphene and p,p '-DDE egg yolk concentrations, but had little

to no correlation with p,p '-DDT yolk burdens. Early embryo mortality rates were

negatively associated withp,p '-DDE and toxaphene. Late embryo mortality rates were

positively associated with toxaphene, and negatively associated withp,p '-DDT, andp,p '-

DDE. Unbanded egg rates were positively associated with p,p '-DDT and p,p '-DDE, but

negatively associated with toxaphene (Fig. 2-2).

For Lake Apopka, redundancy analysis with Monte Carlo permutation tests for

significance also indicated that three of the four extracted OCP variables were found to









be significantly correlated with the clutch survival variables. These OCP variables were

percentage dieldrin (lambda A = 17%), percentage trans-chlordane (12%), and percentage

aldrin (10%), and together accounted for 3% (Z lambda A's) of the variance in the clutch

survival parameters (Table 2-5). Clutch viability was positively associated with aldrin,

weakly associated with trans-chlordane, and negatively associated with dieldrin. Early

embryo mortality and unbanded egg rates were positively associated with dieldrin and

trans-chlordane, and negatively associated with aldrin. Late embryo mortality rates were

negatively with all three OCP variables (Fig. 2-3).

For Emeralda Marsh, redundancy analysis with Monte Carlo permutation tests for

significance also indicated that only percentage toxaphene was found to be significantly

correlated with the clutch survival variables, and it accounted for 9% of the variance in

the clutch survival parameters (Table 2-5). Percentage toxaphene was positively

associated with clutch viability, weakly associated with late embryo mortality, and

negatively associated with early embryo mortality and unbanded egg rates (Fig. 2-4).

Percentage of heptachlor epoxide showed a near significant association (P = 0.09) with

clutch parameters, being negatively correlated with clutch viability and positively

correlated with early and late embryo mortality rates.

Average Egg Mass, Clutch Size and OCP Burdens

For Lochloosa clutches, three of four OCP variables were determined (via RDA

analysis) to be significantly associated with egg and clutch size parameters and accounted

for 64% of the variation in egg and clutch size parameters. These OCP variables

included number of OCPs detected at measurable levels (NOC) (lambda A = 31%), p,p '-

DDT concentrations (20%), and trans-nonachlor concentrations (13%) (Table 2-6).









NOC and trans-nonachlor concentrations were negatively associated with average

egg mass but positively associated with fecundity and clutch mass. In contrast, p,p '-DDT

concentrations were positively associated with egg mass, negatively associated with

fecundity, and had little to no association with clutch mass (Fig. 2-5).

For Lake Griffin clutches, however, no significant associations were found between

OCP variables and egg and clutch size variables. In contrast, percentage o,p '-DDT in

Emeralda Marsh clutches was found to be positively associated with increasing egg and

clutch mass but negatively associated with fecundity. Lastly, Lake Apopka clutches were

somewhat similar to Emeralda clutches in that one OCP variable (p,p '-DDD

concentration) was found to be positively associated with egg and clutch mass and

negatively associated with fecundity (Table 2-6).

Discussion

Inter-Site Comparisons of Clutch Characteristics

The results of the present study suggested that the relative contributions of losses

during in ovo development in alligators at impacted sites in Florida are lower clutch

viability, higher rates of damaged eggs, higher rates of early embryo mortality, and

higher rates of late embryo mortality. Although not significantly different among sites,

infertility and/or embryo mortality before embryo attachment (unbanded eggs) also

appears to be a major constituent of reduced clutch viability among all sites. In order of

importance, major constituents of reduced clutch viability for all sites include early

embryo mortality, unbanded eggs, late embryo mortality, and damaged eggs. In addition,

clutches from OCP-contaminated sites had an average of 10 more eggs per clutch as

compared to the reference site, but average clutch mass was not significantly different,









making average egg mass of reference site clutches greater than that of clutches of OCP-

contaminated sites.

The reduced clutch viability, increased rates of unbanded eggs and embryo

mortality, and concurrent increase in fecundity without proportional increase in clutch

mass observed in clutches from OCP-contaminated sites, as compared to the reference

site (Lochloosa), suggest that females and their embryos from contaminated sites may be

responding to one or more environmental factors common to the three OCP-contaminated

sites. Although measurement of all environmental factors is impractical, the large

differences in OCP concentrations in alligator eggs between reference and OCP-

contaminated sites were found. Specifically, total OCP egg yolk burdens and number of

OCPs detected at measurable levels in Lake Lochloosa were significantly less than those

of Lake Griffin clutches, and OCP burdens in Lake Griffin clutches were, in turn,

significantly less than those of Lake Apopka and Emeralda Marsh.

Although Lake Apopka and Emeralda Marsh were not determined to be

significantly different with respect to total OCP concentrations in egg yolks, significant

differences were determined between these two high OCP exposure sites in regard to

certain analytes, as well as the total number of OCPs detected at measurable levels.

Clutches from Emeralda Marsh had a greater number of OCP analytes in their egg yolks

and contained higher concentrations of cis-chlordane, p,p'-DDD, o,p-DDD, trans-

chlordane, and toxaphene compared to those from Lake Apopka. Conversely, clutches

from Lake Apopka had higher concentrations of aldrin, dieldrin, heptachlor epoxide, and

oxychlordane compared to those of Emeralda Marsh.









The differences in OCP exposure profiles among sites likely reflect the differences

in historic land-use and OCP applications, as opposed to differences in xenobiotic

biotransformation among the different alligator populations inhabiting the respective

sites. Importantly, although Emeralda Marsh is separated from Lake Griffin by only a

levee easily traversed by alligators, large differences in OCP egg burdens were noted

between the two sites. Such differences in exposure suggest that the highly exposed adult

females which oviposite within Emeralda Marsh likely have established territories and

reside year round within Emeralda Marsh (a former agricultural property). Furthermore,

the relatively high egg burdens in clutches of Emeralda Marsh likely occurred over a

relatively short period since this 2,630 ha area was not flooded until the early 1990s

(Marburger et al., 1999).

In summary, the differences in OCP egg burdens between the reference site and the

contaminated sites support the hypothesis that OCP contaminants may be associated with

reduced clutch viability, given that OCPs have been causally linked to reduced

reproductive success in other oviparous species (Donaldson & Fites, 1970; Fry, 1995).

Clutch Survival Parameters and OCP Burdens

Results of redundancy analyses more directly addressed the question of whether

OCPs are associated with reduced clutch viability by relationships on a site-by-site basis

to control for potential site-associated confounding factors. For Lake Lochloosa, no

significant correlations were determined although significance might have been detected

given a greater sample size. The positive but insignificant correlations between increases

in unbanded egg and late embryo mortality percentage and number of OCPs may suggest

that increased OCP burdens in eggs play a role in clutch viability or it may simply

indicate that older females have increased levels of OCPs due to increased exposure time









and that decreased clutch viability is due to decreased egg quality associated with

senescence.

For Emeralda Marsh, the weak associations between OCP variables and clutch

survival variables suggests that other factors may be involved in reduced embryo survival

and increased rates of unbanded eggs. The weak associations for Emeralda Marsh are

surprising given that relatively stronger associations were determined for the other high

exposure site (Lake Apopka; Table 2-5), as well as the intermediate exposure site (Lake

Griffin, Table 2-5), with Emeralda Marsh being separated from Lake Griffin by only a

non-fenced levee.

Stronger associations were noted for Lake Apopka in contrast to the weak,

associations noted for Emeralda Marsh. The positive association between early embryo

mortality and unbanded egg rates and extracted OCP variables for Lake Apopka clutches

suggests that the percentages of dieldrin and trans-chlordane in eggs may play an

important role in altered egg fertility and/or early embryo survival Interestingly, the

percentage of aldrin, (dieldrin's parent compound) had a negative association with late

embryo mortality, a positive association with clutch viability, and near-zero correlations

with percentage unbanded eggs and early embryo mortality. However, dieldrin (a

metabolite formed from aldrin) had strong, positive correlations with percentage

unbanded eggs and early embryo mortality, and a negative correlation with clutch

viability, suggesting this metabolite has greater efficacy than its parent compound in

affecting embryo survival. The potential consequence exists that increasing a female

alligator's ability to biotransform aldrin to dieldrin may increase the risk of early embryo

mortality. Another important note is that the level of dieldrin in Apopka clutches was









two-fold greater than those of Emeralda Marsh, suggesting that OCP mixture

composition may be more important than sum OCP concentrations.

For Lake Griffin, the negative to near-zero association between early embryo

mortality rates and extracted OCP variables suggests that OCP burdens in eggs may not

play an important role in early embryo mortality. However, the positive association

between toxaphene burdens and late embryo mortality suggests that as toxaphene burdens

increase, so does the risk for increased embryo death during the last 35 days of

development. Furthermore, the positive association between p,p '-DDT concentrations

and unbanded egg rates suggests that these analytes may be involved in altered egg

fertility and/or embryo survival (prior to eggshell membrane attachment) (Fig. 2-2).

Egg and Clutch Size and OCP Burdens

For Lochloosa, the strong associations between OCP burdens and egg and clutch

size parameters suggest that, although a low OCP exposure site, certain patterns of OCP

exposure are strongly associated with egg and clutch size characteristics. The positive

associations p,p '-DDT concentrations have with clutch weight and average egg weight

and p,p '-DDT's negative association with fecundity may be potentially related to

senescent females, since older females have been reported to lay smaller clutches of

larger eggs (Ferguson, 1985) and would likely have higher OCP burdens due to extended

exposure period. In contrast, the positive associations that NOC and trans-nonachlor

have with fecundity and clutch mass, and the negative associations these OCP variables

have with egg mass, suggests that increased OCP exposure may have altered clutch and

egg size characteristics, as opposed to female age. Although these speculations are

interesting from a low exposure effect standpoint, they are irrelevant at the population-

effect level since clutch viability rates were unrelated.









Since the low exposure site had stronger associations between OCP variables and

egg and clutch size variables than intermediate and high exposure sites, one might

initially conclude that other factors are more important than OCP burdens in influencing

egg and clutch size characteristics. While this may be the case, the fact that the

intermediate and high exposure sites have significantly greater fecundity (averaging 10

more eggs per clutch compared to the low exposure site), significantly less average egg

mass, and similar clutch mass suggest that females attaining their maximum

physiological response in regard to number of eggs ovulated. These intermediate and

highly exposed females appear to be producing more ova but are unable to sequester

additional egg components (i.e., lighter eggs), in effect decreasing the amount of energy

and structural supplies available to each embryo and resulting in lighter eggs and higher

embryo mortality rates.

In summary, our results suggest that, over all sampled clutches, clutch survival

parameters and egg and clutch size parameters vary between the low OCP exposure site

(Lochloosa) and the intermediate-high OCP exposure sites. Furthermore, OCP burdens

do not appear to be related to clutch survival for the low exposure site but are associated

with clutch survival for the intermediate-high OCP contaminated sites. In contrast, egg

and clutch size parameters appear to be a sensitive endpoint in OCP response in alligators

due to the strong associations noted between OCP and clutch size variables for the low

exposure site and the lack thereof for the intermediate-high exposure sites, suggesting

attainment of maximum response. In order to better determine the role of OCPs in the

reduced reproductive efficiency of OCP-exposed alligator populations, suggested future

studies should examine the relationship between maternal OCP burdens and respective






40


egg burdens, presence of other environmental contaminants, maternal factors associated

with clutch survival and OCP burdens, and how egg composition relates to clutch

survival and OCP burdens.









Table 2-1. Reproductive, morphometric, and contaminant parameters measured on
clutches of alligator eggs collected during summer 2000, 2001, and 2002.


Clutch Parameter
Response variables
Fecundity
Clutch mass
Ave. Egg Weight
Unbanded eggs% a
Early embryo mortality%

Late embryo mortality%

Clutch Viability

Explanatory variables
[OCP analyte] in egg yolk
% OCP analyte


Definition

Total No. of eggs in one clutch
Total mass of eggs in one clutch
Clutch mass / Fecundity
No. of unbanded eggs / fecundity x 100
No. of deaths < dev. Day 35 / fecundity
x 100
No. of deaths > dev. Day 35
/ fecundity x 100
No. eggs yielding live hatchling
/ fecundity x 100

ng OCP analyte / g egg yolk wet weight
[OCP analyte] / 1 [OCP] x 100


aAn egg with no evidence of embryonic attachment


Measured
as

n
kg
g
Percentage
Percentage

Percentage

Percentage


ppb
Percentage









Table 2-2. Explanatory variables included in RDA with forward selection of four best
variables (a = 0.05).
Variable Code
Age Age
No. OCPs at measurable levels NOC
Z [OCP] TOC
% Aldrin ALD%
[Aldrin] [ALD]
% cis-Chlordane CC%
[cis-Chlordane] [CC]
% cis-Nonachlor CN%
[cis-Nonachlor] [CN]
% Dieldrin DL%
[Dieldrin] [DL]
% Heptochlor epoxide HE%
[Heptachlor epoxide] [HE]
%Lipid content LPC%
% Mirex MX%
[Mirex] [MX]
% o,p-DDT ODDT%
[o,p-DDT] [ODDT]
% o,p-DDD ODDD%
[o,p-DDD] [ODDD]
% Oxychlordane OX%
[Oxychlordane] [OX]
% p,p'-DDE PDDE%
[p,p'-DDE] [PDDE]
% p,p'-DDD PDDD%
[p,p'-DDD] [PDDD]
% p,p'-DDT PDDT%
[p,p'-DDT] [PDDT]
% trans-Chlordane TC%
trans-Chlordane [TC]
% trans-Nonachlor TN%
[trans-Nonachlor] [TN]
% Toxaphene TX%
[Toxaphene] [TX]












Table 2-3. Summary of clutch parameters and site comparisons for clutches of American alligator eggs collected during 2000-2002.


Parameter'a
N. Clutches
Fecundity (n)

Clutch mass (kg)

Egg mass (g)

Clutch viability (%)

Damaged eggs (%)

Unbanded eggs (%)

Early Emb. Mort. (%)

Late Emb. Mort. (%)


Lochloosa
44
36 1.2 B
(22-56)
3.4 0.15
(1.6-4.8)
87 + 2.2
(61-139)
70 + 3.9 A
(0-100)
2+ 1.4 B
(0-60)
11 2.2
(0-84)
12 2.7 B
(0-69)
6+ 1.7B
(0-34)


Apopka
31
46 1.3 A
(28-56)
4+ 0.13
(2.4-5.1)
86 +2
(62-120)
51 + 5.8 B
(0-98)
2+0.6B
(0-16)
21 4.9
(0-100)
15 + 4.2 AB
(0-94)
12 + 3.5 A
(0-77)


Emeralda Griffin
46 47
46 1.1A 45+ 1.2A
(27-64) (19-58)
3.8 0.21 3.6 0.13
(1.9-9.2) (1.5-5.2)
83 4 80 1.6
(58-180) (46-113)
48 5.5B 44 4.9B
(0-97) (0-92)
5+1.3A 4+ 1.8AB
(0-33) (0-63)
14 3.7 17 3.2
(0-100) (0-100)
23 3.9 A 22 3.9 A
(0-95) (0-100)
10 2.4A 13 3.1A
(0-82) (0-89)


Summary
168
43 0.7
(19-64)
3.7 + 0.08
(1.5-9.2)
83 1.4
(46-180)
53 + 2.6
(0-100)
3 0.7
(0-63)
15 1.7
(0-100)
19 2
(0-100)
11 + 1.4
(0-89)


aValues indicate mean standard error of mean with ranges in parentheses. Values with different letters (A-B) indicate significant
differences (a = 0.05); same letters indicate significant differences were not detected. Clutch viability = No. of eggs yielding a live
hatchling / Fecundity x 100, Damaged eggs = No. damaged eggs / fecundity x 100, Unbanded eggs = No. of unbanded eggs /
fecundity x 100, Early Emb. Mort. = No. of embryonic deaths on or before developmental Day 35 / fecundity x 100, and Late Emb.
Mort. = No. of embryonic deaths post dev. Day 35 / fecundity x 100).












Table 2-4. Organochlorine pesticide burdens and clutch parameters and site comparisons for clutches of American alligator eggs
collected during 2000-2002.
Parameters Loch. Apopka Emeralda Griffin Summary


N. Clutches
Fecundity (n)

Clutch mass (kg)

Egg mass (g)

Clutch viability (%)

Damaged eggs (%)

Unbanded eggs (%)

Early Emb. Mort. (%)

Late Emb. Mort. (%)

Aldrin (ng/g)

Methoxychlor (ng/g)


19
40 1.7B
(26-56)
3.6 + 0.17
(2.2-4.8)
90 + 2.9 A
(78-139)
65 + 5.5
(0-95)
4 3.1
(0-60)
11+2
(0-33)
13 3
(0-36)
8 2.5
(0-34)
0 +OC
(0-0)
0 0C
(0-0)


23
47 1.4 A
(31-56)
4 0.16
(2.5-5.1)
86 2.5 AB
(62-120)
52 + 6.4
(0-98)
2 0.8
(0-16)
17 4.2
(0-81)
15 4.2
(0-90)
14 4.6
(0-77)
4 0.3 A
(2.9-5.2)
8+ 1B
(5.7-16.4)


31
46 1.3 A
(27-64)
3.8 + 0.25
(2.1-9.2)
82 4.8 B
(58-180)
50 + 6.9
(0-97)
6+ 1.6
(0-32)
10 2.3
(0-58)
26 + 5.1
(0-95)
10+ 2.5
(0-61)
2+ 0.3 B
(1.5-4.3)
9+ 1B
(5.8-18.4)


42
46 1.2 A
(24-58)
3.7 + 0.13
(1.5-5.2)
79 1.5B
(46-105)
43 5.1
(0-92)
5+2
(0-63)
15 3.2
(0-100)
24 4.3
(0-100)
13 + 3.3
(0-89)
0+0C
(0-0)
17 + 0.3 A
(16.9-17.5)


115
45 + 0.7 A
(24-64)
3.8 + 0.09
(1.5-9.2)
83 1.6
(46-180)
50 3.1
(0-98)
4+ 1
(0-63)
13 1.6
(0-100)
21 + 2.3
(0-100)
11 + 1.7
(0-89)
3 0.3
(1.5-5.2)
9 0.8
(5.7-18.4)












Table. 2-4. Continued.


Parameter'a
Mirex (ng/g)

Dieldrin (ng/g)

Hep. Epoxide (ng/g)

cis-Chlordane (ng/g)

cis-Nonachlor (ng/g)

Oxychlordane (ng/g)

p,p'-DDE (ng/g)

p,p'-DDD (ng/g)

p,p'-DDT (ng/g)

o,p'-DDD (ng/g)

o,p'-DDT (ng/g)


Orange/Loch
2+ 0.4 B
(1.2-2.7)
4+ 0.5 D
(1.3-8.2)
3 0.8 C
(1.2-9.7)
2+0.2D
(1.2-4.1)
5 0.6 C
(2.4-12.5)
4+ I D
(1.2-17.8)
74 11.7 C
(28-231)
2+0.2 D
(1.2-3)
1 0C
(1.2-1.3)
0+0C
(0-0)
1 0C
(1.2-1.4)


Apopka
6+ 1.1 A
(1.1-17.2)
344 + 80.9 A
(12.5-1783.2)
17 5.6A
(1.2-135.5)
43 + 7.6 B
(6.6-179.2)
88 + 27.3 A
(10.5-656.2)
51 + 14.6 A
(3.9-353.8)
5794 1794.7 A
(18.3-42653.4)
42 + 8.5 B
(10.6-192.8)
9+ 2.1 AB
(1.2-45.6)
5+0.7B
(3.1-9.2)
11 + 1.9A
(1.2-38.5)


Emeralda
3 0.5 AB
(0.1-10.3)
142 20.4 B
(8.7-386.7)
7+ 1.4 B
(0.1-32.1)
90 13 A
(8.9-281)
66 9.7 A
(11.6-232.2)
23 3.8 B
(3.2-109.5)
8069 1402 A
(36.2-33554.8)
1289 196.1 A
(10.3-2962.8)
12 1.2 A
(5.8-25.5)
37 5.1 A
(0.1-104)
170 161.6 A
(4.2-4372.8)


Griffin
3 + 0.2 AB
(1.1-4.5)
23 3.8 C
(2.9-124)
7+ B
(1.1-29.6)
11 0.9C
(4.3-31.8)
18 1.6B
(6.5-54.2)
10 1.3 C
(1.1-41.9)
271 + 31.3 B
(62.9-979.1)
7+0.9 C
(2.7-28.9)
5 0.8 B
(1.1-7.2)
1 0B
(1.3-1.3)
4+ 0.3 B
(1.1-7.4)


Summary
4 0.4
(0.1-17.2)
118+ 20.9
(1.3-1783.2)
8 1.4
(0.1-135.5)
37 5
(1.2-281)
43 6.7
(2.4-656.2)
21 3.4
(1.1-353.8)
3445 + 610.6
(18.3-42653.4)
382 + 78.7
(1.2-2962.8)
10 1
(1.1-45.6)
29 4.5
(0.1-104)
48 + 42
(1.1-4372.8)












Table. 2-4. Continued.


Parameter'a
trans-Chlordane (ng/g)

Toxaphene (ng/g)

trans-Nonachlor (ng/g)

-OCPs (ng/g)


Orange/Loch
3 0.7 C
(1.2-3.7)
0 0OC
(0-0)
8 1.6 C
(2.5-24.6)
102 15.5 C
(42.7-289.4)


N. OCPs 9 0.3 D
(7-11)


Apopka
8 1.5 B
(1.3-27.4)
2738 224.5 B
(1896.1-3809.1)
212 + 66.9 A
(10.5-1569.2)
7582 2008.2 A
(472.5-47333.8)
13 + 0.3 B
(10-16)


Emeralda
25 + 3.3 A
(2.9-58.2)
6865 + 552.4 A
(2300.6-12975.4)
191 + 30.5 A
(14.2-718.6)
15480 2265.4 A
(269.6-53559.7)
14 + 0.2 A
(13-17)


Griffin
2+ 0.2 C
(1.1-8.7)
3043 + 425.9 B
(1927.9-4533.2)
36 4.7 B
(8.6-155.2)
1169 422.8 B
(101.5-16795.4)
11 + 0.1 C
(9-13)


Summary
11 + 1.5
(1.1-58.2)
5456 + 483
(1896.1-12975.4)
108 17.5
(2.5-1569.2)
6133 + 940.8
(42.7-53559.7)
12 + 0.2
(7-17)









Table 2-5. Results of RDA evaluating associations between clutch survival parameters


and OCP variables.
Site Variablea
Lochloosa NOC
[DL]
PDDT%
PDDE%

Apopka DL%
TC%
ALD%
LPC%

Emeralda Marsh TX%
HE%
ME%
[HE]

Griffin [PDDE]
[TX]
[PDDT]
[ODDD]


LambdaA
0.11
0.09
0.08
0.11

0.17
0.12
0.10
0.06

0.09
0.06
0.06
0.06

0.08
0.07
0.06
0.04


P
0.074
0.194
0.166
0.104

0.004
0.024
0.042
0.16

0.044
0.09
0.15
0.15

0.024
0.016
0.04
0.09


F
2.25
1.59
1.72
2.45

4.25
3.32
3.16
1.85

2.99
2.27
1.85
1.89

3.67
3.16
2.71
1.96


aSee Table 2-2 for definition of variable codes.









Table 2-6. Results of RDA evaluating associations between egg and clutch size
parameters and OCP variables.
Site Variable LambdaA P F
Lochloosa NOC 0.31 0.004 10.15
[PDDT] 0.20 0.042 4.29
[TN] 0.13 0.006 6.77
OX% 0.08 0.088 2.8

Griffin PDDD% 0.05 0.134 2.32
[ODDT] 0.03 0.406 0.91
[PDDT] 0.02 0.236 0.95
[CC] 0.01 0.54 0.33

Emeralda [ODDT] 0.22 0.01 8.07
CC% 0.05 0.146 2
ODDT% 0.05 0.182 1.82
LPC% 0.04 0.21 1.7

Apopka [PDDD] 0.24 0.01 6.51
[ME] 0.08 0.112 2.54
[PDDT] 0.05 0.218 1.5
PDDE% 0.05 0.294 1.29









(C
C6 Early Emb. Mort.



[DL]
PDDE% '


Clutch viability '

------------- -- 4 ------- -------- - ^ -- -- -- -- -- -- -- -- -
PDDTP




Unbanded egg% 0
B


N A Late Emb. Mort.
CO A


-0.8 0.6


Figure 2-1. Biplot of clutch survival parameters (solid lines) and organochlorine
pesticide variables (dashed lines) for clutches of alligator eggs collected from
Lake Lochloosa during summer 2001-2002. Arrows pointing in the same
direction indicate a positive correlation (e.g., clutch viability and PDDE%),
arrows that are approximately perpendicular indicate near-zero correlation
(e.g., late emb. mort. and [DL]), and arrows pointing in opposite directions
indicate negative correlations (e.g., clutch viability and [DL]. Arrow lengths
indicate rank order of correlations. For example, late emb. mort. has higher
positive correlation with NOC (A) compared to unbanded egg% (B). Cosine
of angle formed between individual clutch variables and individual OCP
variables (see Table 2-2 for code definitions) equals correlation coefficient (r)
(ter Braak, 1995). For example, arrows pointing in exactly opposite directions
have an angle of 1800, and since cos(180) = -1.0, the arrows are perfectly,
negatively correlated (r) (ter Braak, 1995).














CO
Late Emb. Mort.
Early Emb. Mort.


[TX]
\,-- [ODDD]









Clutch Viability /


[PDDT] Unhanded egg%
[PDDT]
00/
C) [PDDE]

-1.0 1.0



Figure 2-2. Biplot of clutch survival parameters (solid lines) and organochlorine
pesticide variables (dashed lines) for clutches of alligator eggs collected from
Lake Griffin during summer 2000-2002.










00 A4 AD%
6 Clutch Viability
\ TC%



















Late Emb. Mort LPC%
6

-1.0 1.0



Figure 2-3. Biplot of clutch survival parameters (solid lines) and organochlorine
pesticide variables (dashed lines) for clutches of alligator eggs collected from
Lake Apopka during summer 2000-2002.












CME%
/
/1




// Early Emb Mort


Late Emb Mot / ,-- HE%
TX% -/ / ------ -

Clutch Viability--- ------ ---










N1 Unbanded egg%
[HE]
(0



-0.8 0.6



Figure 2-4. Biplot of clutch survival parameters (solid lines) and organochlorine
pesticide variables (dashed lines) for clutches of alligator eggs collected from
Emeralda Marsh during summer 2000-2002.



















Fecundity



OX%/o Egg Mass



NOC y\




(D0 [PDDT]
0 Clutch Mass

-1.0 1.0



Figure 2-5. Biplot of egg and clutch size parameters (solid lines) and organochlorine
pesticide variables (dashed lines) for clutches of alligator eggs collected from
Lake Lochloosa during summer 2001 and 2002.














CHAPTER 3
MATERNAL TRANSFER OF ORGANOCHLORINE PESTICIDES

Studies have documented organochlorine pesticide (OCP residues) in eggs and/or

somatic tissues of several crocodilian species including the American alligator, Alligator

mississippiensis (Heinz et al., 1991), Morelet's crocodile, C. moreletti (Wu et al., 2000a),

the American crocodile, Crocodylus acutus (Hall et al., 1979; Wu et al., 2000b), and the

Nile crocodile, C. niloticus (Skaare et al., 1991). Indeed, alligator populations inhabiting

Lake Apopka, where an OCP spill occurred in the 1980s, and other central Florida lakes

contaminated with OCPs (through historic OCP use) produce eggs that contain

concentrations of total OCPs that are over 100 times higher than concentrations found in

eggs from reference lakes (Gross, unpublished data). In addition, the alligator

populations inhabiting the OCP-contaminated lakes experience increased (and highly

variable) rates of embryonic mortality, leading to reduced clutch success, and juvenile

alligators appeared to have abnormal sex hormone concentrations as compared to those of

reference sites (Masson, 1995; Rice, 1996; Woodward et al., 1993). However, a clear

dose-response relationship has not been established with respect to individual or total

OCP concentrations in egg yolks and reduced clutch success (Heinz et al., 1991). The

lack of a clear dose-response suggests other factors (e.g., diet, population dynamics, and

specific OCP mixtures) might be involved and/or that developmental effects result from

altered maternal physiology resulting from OCP exposure, as opposed to direct

embryotoxicity.









With respect to altered maternal physiology, alterations in steroid hormone levels

have also been shown in alligators inhabiting OCP-contaminated sites (Guillette et al.,

1994). Furthermore, maternal exposure suggests that OCPs may be maternally

transferred from the adult female alligator to her offspring, as has been reported in other

oviparous vertebrates (Russell et al., 1999). Assuming OCPs are maternally transferred,

the possibility exists that yolks could be used as predictors of maternal exposure. A

noninvasive method such as this would aid ecological risk assessments in understanding

exposure levels for rare/endangered crocodilian species without having to capture and/or

remove adults from the breeding population. Therefore, the objectives of the present

study were to examine maternal transfer as a potential route for embryonic OCP

exposure, and to evaluate the use of yolk burdens for predicting OCP burdens in maternal

tissues in alligators. Our hypothesis was that OCP burdens in maternal tissues and yolks

would be strongly correlated, which would allow yolk burdens to be used to predict

maternal body burdens and suggest maternal transfer of OCPs as the major route for

embryonic OCP exposure.

Materials and Methods

Site descriptions

Lakes Apopka (N 280 35', W 810 39'), Griffin (N 280 53', W 81 49'), and

Lochloosa (N 290 30', W 820 09') in Florida were selected as collection sites because

prior studies by our laboratory indicate vastly different levels of OCP exposure across

these sites. All three lakes are part of the Ocklawaha Basin. Lake Lochloosa (which is

connected to Orange Lake) was selected as a low exposure (reference) site. Four years

(1999-2002) of data indicate mean total OCP concentrations in egg yolks from the

reference sites (Lakes Orange and Lochloosa) were 231 30 ppb (mean standard









deviation [SD], n = 56 clutches) with a concurrent mean clutch viability rate (number of

live hatchlings/total number of eggs in a nest) of 71 21% (Gross, unpublished data).

Lake Griffin was selected as an intermediate exposure site since yolk concentrations

averaged 4,414 + 617 ppb (n = 47 clutches) and Lake Apopka was selected as a high

exposure site since yolk concentrations averaged 15,911 + 1,786 ppb (n = 42) for the

same time period (Gross, unpublished data). Furthermore, mean clutch viability rates

during this time period for Lakes Apopka (51 31%, n = 42) and Griffin (44 33%, n =

47) have been below rates observed for the reference site.

Animal Collections

Adult female alligators and their corresponding clutches of eggs were collected

from Lakes Apopka (n = 4), Griffin (n = 8), and Lochloosa (n = 3) over the course of two

nesting seasons (June 2001 and June 2002). Nests were located by aerial survey

(helicopter) and/or from the ground airboatt). Once nests were located, all eggs were

collected, and the nest cavity was covered. A snare-trap was set perpendicular to the tail-

drag in order to capture the female as she crossed over the nest. After the traps were set,

one member of the trapping crew subsequently transported the eggs to the Florida Fish

and Wildlife Conservation Commission's Wildlife Research Unit (FWC; Gainesville, FL,

USA) and placed the eggs in a temperature-controlled incubator. Snare-traps were

checked later in the evening and early the next morning.

Trapped females were secured and transported from each lake to the United States

Geological Survey's Florida Integrated Science Center (USGS; Gainesville, FL, USA).

Upon arrival, the animals were weighed, measured, and blood samples were collected

from the post-occipital sinus. Adult alligators were then euthanized by cervical

dislocation followed by double pithing. A full necropsy was performed on each female.









Bile, liver, adipose (composite of abdominal fat and the abdominal fat pad), and tail

muscle samples were collected for later determination of OCP burdens. Liver, adipose

tissue, and muscle were wrapped in aluminum foil, while bile and blood were placed in

scintillation vials. All samples were grouped according to nest identification number

(ID), placed in plastic bags labeled with the appropriate ID, and stored in a -80 C

freezer. Each female's corresponding clutch of eggs was then transferred from FWC to

USGS where yolk samples were collected (two eggs/clutch) and stored with the

corresponding maternal tissues. The remaining eggs were set for incubation in a

temperature/humidity-controlled incubator (31-33 C, 88-92% relative humidity) located

at USGS.

Analysis of OCPs in Maternal Tissues and Yolk

Analytical grade standards for the following compounds were purchased from the

sources indicated: aldrin, alpha-benzene hexachloride (a-BHC), P-BHC, lindane, 6-BHC,

p,p '-dichlorodiphenyldichloroethane (p,p '-DDD), p,p '-dichlorodiphenyldichloroethylene

(p,p '-DDE), dichlorodiphenyltrichloroethane (p,p '-DDT), dieldrin, endosulfan,

endosulfan II, endosulfan sulfate, endrin, endrin aldehyde, endrin ketone, heptachlor,

heptachlor epoxide, hexachlorobenzene, kepone, methoxychlor, mirex, cis-nonachlor,

and trans-nonachlor from Ultra Scientific (Kingstown, RI, USA); cis-chlordane, trans-

chlordane, and the 525, 525.1 polychlorinated biphenyl (PCB) Mix from Supelco

(Bellefonte, PA, USA); oxychlordane from Chem Service (West Chester, PA); o,p '-

DDD, o,p '-DDE, o,p '-DDT from Accustandard (New Haven, CT, USA); and toxaphene

from Restek (Bellefonte, PA, USA). All reagents were analytical grade unless otherwise

indicated. Water was doubly distilled and deionized.









Adipose, liver, bile, and yolk samples were analyzed for OCP content using

methods modified from Holstege et al. (1994 and Schenck et al. (1994). For extraction, a

2 g tissue sample was homogenized with ~1 g of sodium sulfate and 8 mL of ethyl

acetate. The supernatant was decanted and filtered though a Btichner funnel lined with

Whatman #4 filter paper (Fisher Scientific, Hampton, NH, USA) and filled to a depth of

1.25 cm with sodium sulfate. The homogenate was extracted twice with the filtrates

collected together. The combined filtrate was concentrated to ~2 mL by rotary

evaporation, and then further concentrated until solvent-free under a stream of dry

nitrogen. The residue was reconstituted in 2 mL of acetonitrile. After vortexing (30 s),

the supernatant was applied to a C 18 solid phase extraction (SPE) cartridge (pre-

conditioned with 3 mL of acetonitrile; Agilent Technologies, Wilmington, DE, USA) and

was allowed to pass under gravity. This procedure was repeated twice with the combined

eluent collected in a culture tube. After the last addition, the cartridge was rinsed with 1

mL of acetonitrile which was also collected. The eluent was then applied to a 0.5 g NH2

SPE cartridge (Varian, Harbor City, CA, USA), was allowed to pass under gravity, and

collected in a graduated conical tube. The cartridge was rinsed with an additional 1 mL

portion of acetonitrile which was also collected. The combined eluents were

concentrated under a stream of dry nitrogen, to a volume of 300 [tL, and transferred to a

gas chromotography (GC) vial for analysis.

Whole blood was analyzed for OCP content using methods modified from

Guillette et al. (1999). A 10 mL aliquot was transferred from the homogenized bulk

sample and extracted in 15 mL of acetone by vortex mixer. The mixture was centrifuged

for 5 min at 3000 rpm, after which the supernatant was transferred to a clean culture tube.









This process was repeated with the supernatants collected and concentrated under a

stream of dry nitrogen until solvent-free. The residue was re-extracted in 11.5 mL of 1:1

methylene chloride-petroleum ether. After mixing, the sample was allowed to settle and

the upper layer was transferred to a clean culture tube. This extraction was performed

twice with the extracts collected together. The combined extracts were then applied to a

prepared florisil cartridge (5 mL Fisher PrepSep, Fisher Scientific, Hampton, NH, USA).

The cartridge had been prepared by filling the reservoir to a depth of 1.25 cm with

anhydrous sodium sulfate and by prewashing the modified cartridge with 10 mL of 2:1:1

acetone: methylene chloride: petroleum ether. After the sample passed under gravity

with the eluent collected in a 15-mL graduated conical tube, the cartridge was eluted with

4 mL of the 2:1:1 solvent mixture which was also collected. The combined eluents were

concentrated under a stream of dry nitrogen, to a volume of 300 [tL, and transferred to a

GC vial for analysis.

GC/MS Analysis

Analysis of all samples was performed using a Hewlett Packard HP-6890 gas

chromatograph (Wilmington, DE, USA) with a split/splitless inlet operated in splitless

mode. The analytes were introduced in a 1 [iL injection and separated across the HP-5MS

column (30 m x 0.25 mm; 0.25 [tm film thickness; J & W Scientific, Folsom, CA, USA)

under a temperature program that began at 600 C, increased at 10 C/min to 2700 C, was

held for 5 min, then increased at 250 C/min to 3000 C and was held for 5 min. Detection

utilized an HP 5973 mass spectrometer in electron impact mode. Identification for all

analytes and quantitation for toxaphene was conducted in full scan mode, where all ions

are monitored. To improve sensitivity, selected ion monitoring was used for the









quantitation for all other analytes, except kepone. The above program was used as a

screening tool for kepone which does not optimally extract with most organochlorines.

Samples found to contain kepone would be reextracted and analyzed specifically for this

compound.

For quantitation, a five-point standard curve was prepared for each analyte (r2 >

0.995). Fresh curves were analyzed with each set of twenty samples. Each standard and

sample was fortified to contain a deuterated internal standard, 5 [iL of US-108 (120

[g/mL; Ultra Scientific), added just prior to analysis. All samples also contained a

surrogate, 2 [g/mL of tetrachloroxylene (Ultra Scientific) added after homogenization.

Duplicate quality control samples were prepared and analyzed with every twenty samples

(typically at a level of 1.00 or 2.50 [g/mL ofy-BHC, heptachlor, aldrin, dieldrin, endrin,

andp,p '-DDT) with an acceptable recovery ranging from 70 130%. Limit of detection

ranged from 0.1-1.5 ng/g for all OCP analytes, except toxaphene (120-236 ng/g), and

limit of quantitation was 1.5 ng/g for all analytes, except toxaphene (1500 ng/g).

Repeated analyses were conducted as allowed by matrix interference and sample

availability.

Data Analysis

OCP concentrations in maternal tissues and egg yolks were lipid-adjusted (wet

weight concentration / proportion of lipid in tissue), and lipid-adjusted tissue-to-egg yolk

ratios (maternal tissue OCP concentrations /egg OCP concentrations) were examined.

Predictive models were determined by linear regression analysis of OCP concentrations

in yolk against those of maternal tissues (log-transformed wet weight concentrations).

Each model's ability to fit the data was evaluated by examining the p-value (a = 0.05),

the r2 value, and the residual plots (SAS Institute Inc., 2002). ANOVA was used for









inter-site comparisons of adult female and clutch characteristics, and the Tukey test was

used for multiple comparisons among sites. The relationship between maternal mass (kg)

and concentrations of OCPs in eggs and maternal tissues (log-transformed wet weight

concentrations) were evaluated using linear regression to assess whether increasing mass

was associated with increasing concentrations of OCPs in eggs and maternal tissues,

which may suggest adult females continue to bioaccumulate OCPs as they grow

throughout their life. Adult females were grouped by site since the extreme differences

in OCP exposure among sites would likely confound results. Unless otherwise noted,

values are reported as mean standard deviation.

Results

Female Morphological and Reproductive Characteristics

For all females, mass and snout-vent length (SVL) averaged 74 20 kg (range:

44-114) and 135 11 cm (119-156), respectively. Clutch mass (mass of all eggs from a

single nest) and fecundity (number of eggs collected from a single nest) of these

individuals were 3.65 0.86 kg (1.84-4.82) and 43 10 eggs/nest (19-56), respectively.

No significant differences were detected across sites with respect to female mass (p =

0.14), total length (p = 0.90), SVL (p = 0.25), tail girth (p = 0.98), head length (p = 0.55),

clutch mass (p = 0.23), or fecundity (p = 0.40, Table 1).

With respect to lipid concentrations in egg yolk and muscle, no significant

differences were detected across sites (p > 0.05). However, lipid concentration in liver of

Lochloosa females was significantly higher (p < 0.05) than that of Apopka and Griffin

females (which were not significantly different from one another). Furthermore, lipid

concentration in abdominal adipose tissue of Apopka females was significantly less (p <

0.05) than that of Lochloosa and Griffin females (Table 1).










OCP concentrations in Yolk

Egg yolks from Lake Apopka females contained the highest total OCP

concentration (15,108 13,704) and greatest number of individual OCPs detected above

the limit of quantitation (n = 18) with p,p'-DDE (66%) and toxaphene (32%) being main

constituents. Lake Griffin females produced eggs with the next highest total OCP

burdens (393 300 ng/g; n = 13) being mainly composed of p,p'-DDE (69%), trans-

nonachlor (10%), and dieldrin (7%). Lake Lochloosa females produced egg yolks with

the smallest total OCP burden (124 53 ng/g, n = 9), with main constituents being p,p'-

DDE (73%), trans-nonachlor (10%), and cis-nonachlor (4%; Table 3-2). The OCP

analytes with the highest average egg yolk concentrations were toxaphene (4,862 + 4,177

ng/g), which was detected above the limit of quantitation in 3 of 15 clutches, followed by

p, p'-DDE (2,828 5,968 ng/g), dieldrin (191 474 ng/g), and trans-nonachlor (126 +

209 ng/g), which were above quantitation limit in all 15 clutches.

OCP concentrations in maternal tissues

Adipose tissue (a composite of abdominal fat and fat pad) contained the highest

concentration of total OCPs (12,805 31,678 ng/g wet weight) of all tissues. p,p'-DDE

(67%) composed the majority of the total burden, followed by dieldrin (5%), and trans-

nonachlor (3%). Although toxaphene was only detected in 3 individuals from Lake

Apopka, its average burden in adipose tissue was 13,463 1,267 ng/g (Table 3-2). In

liver, OCP analytes were detected above the quantitation limit in 9 of 15 individuals, and

total OCP concentrations averaged 1,008 1,245 ng/g. Liver burdens were primarily

composed of p,p'-DDE (76%) and dieldrin (6%). Total OCP concentrations in muscle

averaged 716 1,053 ng/g and were above quantitation limits in 10 of 15 individuals









with most of the burden being composed of p,p'-DDE (83%), dieldrin (6%) and trans-

nonachlor (6%). Total OCP burdens in bile (412 483 ng/g) were above quantitation

limits in five individuals with p,p'-DDE (86%) and dieldrin (6%) comprising the majority

of the burden. Total OCP concentrations in blood (43 + 21 ng/g) were above quantitation

limits in 4 individuals with p,p'-DDE (64%) and dieldrin (14%) comprising most of the

burden. Overall, Lake Apopka alligators exhibited the highest OCP concentrations in

maternal tissues and egg yolks, followed by Lakes Griffin and Lochloosa, respectively

(Table 3-2).

Relationships between Maternal Tissue and Yolk Burdens

Examination of lipid-adjusted maternal tissue-to-egg yolk burdens showed

differences among tissues. With respect to total OCPs, the adipose burden-yolk burden

ratio was close to 1 (95% confidence interval (CI), 0.76 K< K < 1.11). In contrast, the

liver-yolk ratio was significantly greater than 1 (95% CI, 1.49 < [ < 9.19), and muscle

ratios showed considerable variation (95% CI, -1.17 < K < 37.35). As would be expected,

most individual OCPs followed the above trend. However, cis-chlordane was an

exception as liver ratios (95% CI, 2.85 < [ < 6.75) and muscle ratios (95% CI, 1.78 < [ <

15.1) were greater than 1, while adipose ratios (95% CI, 0.59 < [ < 0.84) were less than

1. With respect to total OCP concentrations, significant linear relationships (predictive

models) were found for adipose, liver, muscle, and bile (p < 0.05, Fig. 1). With respect to

individual OCP analytes, predictive models were derived for 12 of 14 (78%) of the OCPs

co-detected in adipose tissue and egg yolk, followed by liver (9/12, 75%), bile (8/11,

73%), and muscle (2/12, 17%; Table 3-3). Although nine OCP analytes were

concurrently detected in blood of the females and their respective egg yolks, no

significant linear correlations were detected (p > 0.05).









As for individual OCP analytes, p,p'-DDE concentrations in yolk was significantly

correlated with those of liver, muscle, bile, and adipose tissue. Blood p,p'-DDE

concentrations did not exhibit a significant linear relationship (p > 0.05) with yolk p,p'-

DDE concentrations. Heptachlor epoxide, trans-chlordane, cis-chlordane, trans-

nonachlor, cis-nonachlor, mirex, and dieldrin concentrations in yolk were significantly

correlated to their respective concentrations in adipose, liver, and bile. With respect to

oxychlordane, significant correlations were only derived for liver and adipose tissue, and

significant correlations for p,p'-DDD concentrations were found only for adipose and

bile. Toxaphene and o,p'-DDT concentrations in adipose tissue were significantly

correlated with respective egg yolk concentrations (Table 3-3).

Relationships between Maternal Mass and OCP concentrations in Eggs and Tissues

For females collected from Lakes Apopka (n = 4) and Lochloosa (n = 3), no

significant correlations (p > 0.05) were found when maternal mass (kg) was compared

against either individual or total OCP concentrations (log-transformed wet weight) in

maternal tissues and eggs. However, significant correlations might have been difficult to

detect because of the small sample size. In contrast, a larger number of Lake Griffin

females (n = 8) were collected, and analyses indicated significant correlations between

maternal mass and OCP concentrations in tissues and eggs indicating that larger females

have higher concentrations of OCPs in their tissues and eggs, which may suggest females

continue to bioaccumulate OCPs as they grow (increase in mass). For Lake Griffin

females, OCP burdens in eggs had the greatest number of significant correlations (p <

0.05) with body mass (kg), which consisted of cis-nonachlor (r2 = 0.87), cis-chlordane (r2

= 0.75), trans-nonachlor (r2 = 0.73), dieldrin (r2 = 0.69), p,p'-DDE (r2 = 0.66), o,p'-DDT

(r2 = 0.61), heptachlor epoxide (r2 = 0.59), oxychlordane (r2 = 0.58), trans-chlordane (r2 =









0.57), and total OCPs (r2 = 0.71). Following egg concentrations, abdominal fat OCP

burdens-to-body mass correlations consisted of cis-nonachlor (r2 = 0.67), cis-chlordane

(r2 = 0.81), trans-nonachlor (r2 = 0.63), dieldrin (r2 = 0.62), p,p'-DDE (r2 = 0.58),,

heptachlor epoxide (r2 = 0.53), oxychlordane (r2 = 0.51), and total OCPs (r2= 0.64).

Although egg burdens of o,p'-DDT and trans-chlordane were correlated with body mass,

abdominal fat burdens were not. Lastly, liver OCP burdens-to-body mass correlations

included only trans-nonachlor (r2= 0.99) and p,p'-DDT (r2= 0.99). No significant

correlations were found for cis-chlordane, trans-chlordane, oxychlordane, dieldrin,

heptachlor epoxide, o,p'-DDT, and cis-nonachlor.

Discussion

The presence of OCPs in the eggs and tissues of alligators is not novel; however,

the value of our study was that OCP concentrations in maternal tissues and yolks

appeared to be strongly correlated with one another, allowing yolk burdens to be used as

predictors of OCP burdens in tissues of adult reproductive alligators, which may be a

useful noninvasive technique that would aid risk assessments involving endangered

crocodilians. Furthermore, our results are consistent with other studies that suggest OCPs

are maternally transferred in wild alligators (Rauschenberger et al., 2004).

Several OCP analytes were detected in both maternal tissues and yolk (Table 3-3)

suggesting that mixture composition may be an important consideration in risk

assessment. One reason for this is that different xenobiotic compounds may induce or

inhibit certain biotransformation enzymes. Specifically, alligators from Louisiana

express several different xenobiotic biotransformation enzymes (e.g., liver cytochrome P-

450 enzymes [CYP] such as CYP1A, CYP2B) in response to xenobiotic exposure (Ertl et

al., 1999). Furthermore, genetic partitioning has been reported in spatially separated









alligator populations (Ryberg et al., 2002). Therefore, the possibility exists that certain

individuals or populations may lack the genetic or epigenetic ability to produce a

particular biotransformation enzyme, which may lead to increased risk of xenobiotic-

induced toxicity. For example, certain populations of black-banded rainbowfish

(Melanotaenia nigrans) were able to tolerate copper exposures (96-hr EC50) that were 8.3

fold greater than the tolerance limits of other, spatially-separated populations of the same

species. Genetic analyses suggested that allozyme frequencies of tolerant and susceptible

populations were significantly different at AAT-1 and GPI-1 loci, suggesting differences

in allozymes of exposed fish may have assisted in the increased copper tolerance

(Woosley, 1996).

Examination of maternal tissue-to-egg concentration ratios (lipid-adjusted) showed

differences among tissues. The adipose-to-yolk concentration ratio was close to 1,

suggesting that OCPs reach equilibrium within abdominal adipose tissue, and that lipids

and OCPs are mobilized and subsequently incorporated into the developing yolks. In

contrast, liver-to-yolk concentration ratios were significantly greater than 1, and muscle-

to-yolk concentration ratios showed considerable variation. One suggested explanation

for the high liver-to-yolk ratios relates to one major function of the liver cells

hepatocytess), which is to accumulate and convert hydrophobic xenobiotics into

hydrophilic metabolites to facilitate detoxication, excretion, and elimination. In addition,

the low lipid content of liver (relative to the lipid content of adipose tissue and yolk,

Table 3-1) may have contributed to the marked differences. With respect to the muscle-

to-yolk ratios, the reasons for the large degree of variability are not as clear. One

possible explanation is that muscle lipids are not mobilized during yolk formation and, as









a result, OCP burdens may continually accumulate in muscle lipids. Another potential

explanation relates to the low lipid content of muscle when compared to yolk (Table 3-1).

Lastly, cis-chlordane's exceptional liver, muscle, and adipose ratios underscore the fact

that different OCP analytes may not always exhibit identical pharmacokinetics.

When compared to other vertebrates, adipose tissue-to-egg ratios in alligators are

similar to those reported in the freshwater catfish, Clarias batrachus, in that adipose-to-

egg ratios are approximately equal to 1. Furthermore, C. batrachus mobilizes lipids from

its abdominal adipose tissue during vitellogenesis (Lal & Singh, 1987), similar to what

this study suggests occurs in the American alligator. In contrast to adipose tissue OCP

concentrations, muscle-to-egg OCP ratios in alligators appear to be quite different from

fish. Alligator muscle-to-egg ratios were highly variable and, for the most part, greater

than 1, while fish ratios appear to be consistently close to 1. With respect to more closely

related species, muscle-to-egg OCP ratios are similar to those reported for the common

snapping turtle (Chelydra serpentina) and several bird species with ratios exhibiting a

great deal of variability and being greater than 1 (Russell et al., 1999). These differences

suggest that fish differ from terrestrial vertebrates in regards to lipid content of muscle

and/or lipid mobilization strategy (during vitellogenesis), which could lead to differences

in embryonal exposure given equivalent maternal exposure.

Evaluation of Predictive Models

Although significant linear models were found for most tissues with respect to total

OCP concentrations, caution should be used in the application of these "total OCP"

models since it is probable that the concentrations and ratios of individual OCP analytes

may vary across different locations. The greatest number of predictive linear models was

derived for adipose tissue. This was not surprising considering that (for most analytes)









adipose-yolk lipid normalized ratios were close to 1. Next, with respect to the number of

significant linear models, were liver and bile. The similarities between liver and bile

should be expected since the liver produces bile, which transports OCP analytes to the

intestinal lumen, leading to their eventual elimination from the body. However, OCP

analytes may be reabsorbed from the intestine and redirected back to the liver via the

portal vein through a process known as enterohepatic circulation, which may delay

elimination of lipophilic xenobiotics, increase hepatic exposure and bioaccumulation

(Stenner et al., 1997). For OCP concentrations in muscle, regression analysis indicated

that only two out of 12 mutually detected analytes could be predicted using OCP

concentrations in eggs. Lastly, nine OCP analytes were concurrently detected in blood

and egg yolk with none exhibiting significant relationships. Possible explanations for the

few significant linear relationships include the low lipid content of these tissues and thus

the relatively low concentrations of OCP analytes in these tissues, as well as the

possibility that each of these tissue burdens exhibit a nonlinear relationship with yolk

burdens. In addition, blood samples were collected after the female had oviposited.

Since blood was collected after eggs were excreted from the body, it is likely that the

overall maternal body burden decreased, which would in turn lower the steady-state OCP

concentrations in blood.

As for individual OCP analytes, predictive models for p,p'-DDE were derived for

four of the five maternal tissues. One likely reason for this is that p,p'-DDE was detected

in considerable concentrations in all eggs and in almost all tissues for all 15 females.

Similarly, predictive models were derived for commonly detected analytes such as

heptachlor epoxide, trans-chlordane, cis-chlordane, trans-nonachlor, cis-nonachlor,









mirex, and dieldrin for most tissues. Somewhat surprisingly, oxychlordane (a metabolite

of cis- and trans-chlordane) and p,p'-DDD (an intermediate metabolite of p,p'-DDT)

showed significant linear models only with respect to liver and adipose tissue. The fact

that linear models for toxaphene and o,p'-DDT were derived only for adipose tissue was

likely related to their low concentrations and infrequent detections in other tissues (Table

3-3).

Relationships between Maternal Mass and OCP concentrations in Eggs and Tissues

Although a portion of a female alligator's OCP body burden may be eliminated

through egg deposition, adult female alligators from Lake Griffin had increased OCP

concentrations in their tissues and eggs as they increased in mass, similar to size-related

OCP bioaccumulation in smallmouth bass inhabiting contaminated sites in Michigan

(Henry et al., 1998). Corresponding increases in OCP burdens and mass indicate that

larger and possibly older females accumulate OCPs faster than they can excrete them. In

addition, the relationship between OCP burdens in eggs and body mass was very similar

to the relationship between abdominal fat burdens and body mass.

The correlation between OCP burdens in liver and body mass was significant for

trans-nonachlor and p,p'-DDT; however the major metabolites of these compounds

(oxychlordane and p,p'-DDE, respectively) were not significantly correlated with body

mass. These results contrast those of egg and abdominal fat burdens and suggest that that

alligator liver may not sequester OCP metabolites to the same extent as abdominal fat or

egg.

Maternal body burdens: Toxicological Implications

Although our study's objective was to evaluate maternal transfer and prediction of

the maternal OCP body burdens carried by the American alligator, we would be remiss if









we did not discuss whether these reported body burdens were capable of eliciting harmful

effects. Although several studies report body and egg burdens in crocodilians, relatively

few studies directly relate body and egg burdens to acute toxicological effects (Campbell,

2003), so we will briefly discuss how p,p'-DDE burdens in maternal alligator liver

compare to reported p,p'-DDE burdens in liver of birds (birds were not from the present

study areas) that have been associated with mortality (Blus, 1996).

In previous studies, mean DDE liver residues in birds which died due to DDT

exposure ranged from 19,000-55,000 ng/g. When birds were exposed to DDE alone,

liver residues of dead birds averaged 3,883,000 ng/g (range 460,000-11,725,000 ng/g)

(Blus, 1996). When compared to the liver residues of the most contaminated alligators

(Lake Apopka, upper 95% CI < 7,000 ng/g), it appears that death due to DDT/DDE

exposure might be unlikely assuming bird and alligator susceptibilities are similar.

However, since p,p'-DDE liver concentrations in alligators are almost half of lethal liver

concentrations in birds, there is reason for some concern. In addition, the assumption that

bird and alligator susceptibilities are similar might be argued as unfounded considering

the variability in toxic responses between individuals of the same species, different

species, and different vertebrate classes (James et al., 2000). To account for these

uncertainties the risk assessment process identifies the different sources of uncertainty

and incorporates the uncertainty in attempting to determine a "safe" tissue concentration

based on levels associated with no adverse effects (NOAEL) or lowest observed adverse

effect levels (LOAEL). Typically, interspecies extrapolation is assigned an uncertainty

factor of 10, as are inter-individual uncertainty, uncertainty related to comparing different

study designs (e.g., acute doses related to experimental bird studies, in contrast to chronic









exposure studies in wild alligators), and uncertainty related to database quality since

DDE (p,p'-DDE + o,p-DDE) liver residues were reported, instead of p,p'-DDE. These

four uncertainty factors constitute an overall uncertainty factor of 10,000, which is an

order of magnitude greater than commonly used uncertainty factors (range: 300-1000)

(James et al,. 2000). Considering the high degree of uncertainty, we suggest that more

information is required before a "safe" level of p,p'-DDE exposure is determined for the

American alligator based upon actual or predicted liver concentrations.

Sublethal effects are another possible consequence of OCP exposure. For example,

exposure of the freshwater catfish, Clarias batrachus, to an OCP analyte (y-BHC) at

sublethal levels (2,000-8,000 ng/g) during vitellogenesis significantly decreased the

biosynthesis and mobilization of phospholipids from liver to the developing follicles (Lal

& Singh, 1987). Interestingly, alterations in fatty acid profiles of alligator eggs have

been associated with reduced clutch success. Specifically, fatty acid profiles from wild,

alligator eggs (normal hatch rates) showed considerable differences when compared to

those of eggs from captive alligators (reduced hatch rates). One suggested explanation

for this association between altered fatty acid profiles and reduced clutch success in

captive alligators was that certain fatty acids are critical for reproductive success and that

captive diets were deficient in essential fatty acids (Noble et al., 1993). Thus, the

possibility exists that exposure to OCPs may alter the liver's ability to synthesize

necessary fatty acids, leading to altered egg quality and decreased clutch success in wild

alligators that inhabit OCP-contaminated sites. Chronic exposure to low doses of OCPs

prior to and during vitellogenesis has been suggested as a cause for significant increases

in OCP concentrations in egg yolk, as well as significantly decreased hatch rates in









captive adult female alligators. Importantly, the doses did not appear to induce acute

toxicity in the adult females (Rauschenberger et al., 2004). Presently, we are using a

captive breeding population of adult alligators, as well as data from field studies, to

further evaluate the relationships between OCP exposure, altered fatty acid biosynthesis,

nutritional content of eggs, and embryonic mortality.

In summary, the significant levels of OCP analytes observed across such a wide

range of crocodilian species and geography suggests the need for a greater understanding

of xenobiotic metabolism and toxicological responses in crocodilians. Such

understanding would aid in the conservation of this ancient group by determining what

risks are posed by contaminants with respect to species survival and how contaminant-

related risks compare to other risks, such as habitat destruction. The results of the present

study provide some evidence suggesting that maternal transfer of OCP analytes is the

major route for embryonic exposure. In addition, it provides several models for the

prediction of OCP concentrations in maternal tissues of American alligators, which may

be extrapolated to other crocodilians. Hopefully, the present study will encourage new

investigations into the pharmacokinetics and pharmacodynamics of contaminants in other

crocodilian species.









Table 3-1. Morphological and reproductive characteristics of adult female alligators
collected during June 2001 and 2002 from Lakes Apopka, Griffin, and
Lochloosa in central Florida.
Parameter a,b Apopka Griffin Lochloosa
Number of females collected 4 8 3
Total Length (cm) 252 38 258 + 17 258 7
Snout-Vent Length (cm) 142 + 15 134 + 9 129 5
Mass (kg) 94 30 70 17 63 4
Clutch Mass (kg) 3.78 + 0.98 3.33 + 0.82 4.31 + 0.45
Fecundity (# eggs/clutch) 43 10 40 10 49 6
Lipid % Adipose 47.0 32.5 B 78.1 + 8.0 A 81.4 + 4.0 A
Lipid % Liver 1.3 1.0 A 0.8 0.2 A 5.0 2.3 B
Lipid % Muscle 0.8 + 0.9 1.3 + 0.9 0.2 0.02
Lipid % Yolk 19.9 1.1 18.1 + 1.7 18.2 1.6
a Values represent mean standard deviation. b Different letters indicate significant
differences (p < 0.05).












Table 3-2. Pesticide concentrations (ng/g wet wt.) in tissues and yolks of adult female alligators collected during June 2001 and 2002
from Lakes Apopka, Griffin, and Lochloosa in central Florida.


Lake a
Apopka
(4)


Chemical b,c
Aldrin
a-BHC
(3-BHC
cis-Nonachlor
cis-Chlordane
6-BHC
Dieldrin
Endosulfan I
Endosulfan II
Endosulfan Sulfate
Endrin
Endrin Aldehyde
Endrin Ketone
y-BHC
Heptachlor
Heptachlor Epoxide
Hexachlorobenzene
Kepone
Methoxychlor
Mirex
o,p '-DDD
o,p '-DDE
o,p '-DDT
Oxychlordane
p,p '-DDD
p,p '-DDE
Toxaphene
trans-Nonachlor
Total OCP


Bile
X
X
X
10 + 3.2
4 1.9
X
38 + 10.2
X
X
X
X
3 0.4
X
X
X
3 2.1
1 0
X
X
2 2.6
X
X
2 0.2
7 1.3
2 0.2
806 341
X
21 + 7.8
900 369.7


Blood
X
X
X
2 0.4
1 0.4
X
5 0.4
X
X
X
X
X
X
X
X
0.3 0
1+ 0
X
X
X
X
X

1 0.2
1 0.2
42 5.7
X
3 0.4
55 7


Adipose
X
X
7.5 6.7
521 + 602.7
190 + 241.2
X
2,376 + 3,770.9
X
X
X
X
X
X
X
X
67 + 81.5
10
X
X
19 + 13.1
3 3.4
52 + 55.8
27 26.4
247 + 336.4
43 + 67.5
29,840 34,366
13,436 + 12,670.2
1,153 + 1,378.7
44,650 + 53,230


Liver
X
X
X
31 6.7
11 9.1
X
105 80.2
X
21
X
X
X
X
X
X
6 2.2
1 0.0
X
5
7+ 9.0
X
X
4 1.8
17 9.9
11 10.3
1,846 918.1
X
65 22.7
2,140 1,024


Muscle
X
X
X
23 18.3
14.1 11.7
X
68 48.3
X
X
X
X
X
X
X
8 11.7
4 3.6
1
X
X
1 0.4
X
X
4 2.2
12 10.7
17 8.0
1,392 1,0782
X
68 57.0
1,610 1,226


Yolk
1
X
2 1.4
123 81.9
62 59.2
X
663 803.0
X
X
X
X
X
X
X
1 0.04
26 15.0
1 0.0
X
X
7 7.1
X
45 17.7
17 7.8
75 68.2
52 61.4
9,994 8,529
4,862 4,177
387 277.7
15,108 13,704













Table 3-2. (Continued)


Lake
Griffin
(8)


Chemical
Aldrin
a-BHC
(3-BHC
cis-Nonachlor
cis-Chlordane
6-BHC
Dieldrin
Endosulfan I
Endosulfan II
Endosulfan Sulfate
Endrin
Endrin Aldehyde
Endrin Ketone
y-BHC
Heptachlor
Heptachlor Epoxide
Hexachlorobenzene
Kepone
Methoxychlor
Mirex
o,p '-DDD
o,p '-DDE
o,p '-DDT
Oxychlordane
p,p '-DDE
p,p'-DDT
Toxaphene
trans-Chlordane
trans-Nonachlor
Total OCP


Bile
X
X
X
4 2.4
2 0.5
X
13 4.9
X
X
X
X
X
X
X
X
3 3.3
10
X
X
0.3 0
X
X
1 0.2
7 4.9
54 25.7
1
X
1 0.3
9 5.9
87 46.4


Blood
X
X
X
1 0.4
10
X
7
X
X
X
5
X
2
X
X
1
1
X
X
1
X
X
1 0.0
X
13 9.1
13
X
10
1 0.09
31 28.4


Adipose
X
X
2.1 + 1.1
75 + 74.9
30 + 10.8
X
109 + 133.4
X
X
X
X
X
X
2
X
34 45.7
1+0
X
X
5+ 3.6
X
X
106.9
56 + 84
1,030 931.3
3 1.5
X
3 1.7
171 213.5
1,533 1,439


Liver
X
X
X
8 4.4
2 0.7
X
17 + 8.0
X
X
X
X
X
X
X
X
5 3.6
1 0
X
X
1
X
X
X
8 6.2
75 46.6
29 0.9
X
1 0.3
18 13.2
153 78.1


Muscle
X
X
X
9 10.6
3 3.4
X
22 20.8
X
X
X
X
X
X
X
2 1.4
10 12.0
1
X
X
1 0.5
X
X
2 0.1
16 17.8
131 132.4
X
X
1
28 36.0
208 227


Yolk
X
X
X
14 7.6
11 3.7
X
26 25.7
X
X
X
X
X
X
X
X
8 8.8
1 0.0
X
2
1 0.2
X
3
3 1.8
12 14.9
273 204.0
3
X
2 1.0
40 38.3
393 299












Table 3-2. Continued.
Lake Chemical Bile Blood Adipose Liver Muscle Yolk
Lochloosa Aldrin X X X X X X
(3) ca-BHC NA X X X X X
(3-BHC NA X 1 X X X
cis-Nonachlor NA X 17 + 1.8 1 X 5 1.9
cis-Chlordane NA X 8 + 1.4 X X 3 + 0.1
6-BHC NA X X X X X
Dieldrin NA X 14 4.7 2.6 1.4 4 2.8
Endosulfan I NA X X X 15.6 X
Endosulfan II NA X X X X X
Endosulfan Sulfate NA X X X X X
Endrin NA X X X X X
Endrin Aldehyde NA X X X X X
Endrin Ketone NA X X X X X
y-BHC NA X X X X X
Heptachlor NA X X X 18 +9.1 X
Heptachlor Epoxide NA X 11 9.6 X X 3 2.9
Hexachlorobenzene NA X X X X X
Kepone NA X X X X X
Methoxychlor NA X X X X X
Mirex NA X 2.6 X X X
o,p'-DDD NA X X X X X
o,p'-DDE NA X X X X X
o,p'-DDT NA X 3 0.2 X 7.1 1 0.0
Oxychlordane NA X 17 11.0 1 X 5 4.3
p,p'-DDD NA X 1 0.1 X 1 20.9
p,p'-DDE NA X 297 +90.1 20 +20.9 11 6.7 91 32.5
p,p'-DDT NA X 1.4 0.1 X 1.4 X
Toxaphene NA X X X X X
trans-Chlordane NA X 1 + 0.1 X 1.4 X
trans-Nonachlor NA X 38 + 24.6 2.6 1.4 12 + 8.8
Total OCP NA X 407 143.6 28 +32.6 33 33.9 124 53.3
a Number of females and clutches collected noted in parentheses beneath name of lake. b Values represent mean + standard deviation










[SD], values without SD indicate a single measurement. X indicates values which were below limit of detection (LOD) or below limit
of quantitation (LOQ) and NA indicates not analyzed. LOD ranged from 0.1-1.5 ng/g for most OCP analytes (toxaphene LOD ranged
from 120-236 ng/g), and LOQ ranged was 1.5 ng/g for all analytes except for toxaphene (1500 ng/g). Percent recovery ranged from
70-130%. The following chemicals were neither detected in females nor their eggs: a-BHC, 6-BHC, endosulfan sulfate, and kepone.
BHC = Benzene hexachloride; DDD = Dichlorodiphenyldichloroethane; DDE = Dichlorodiphenyldichloroethylene; DDT =
Dichlorodiphenyltrichloroethane; Total OCP = organochlorine pesticide concentrations for all analytes.











Table 3-3. Regression equations for predicting organochlorine pesticide (OCP)
concentrations in maternal tissues, where LOG [Tissue-OCP] = bo + b LOG
[Yolk-OCP].
Tissue Chemicala bo bi n r2 p
Adipose Dieldrin 0.6624 0.8785 15 0.87 < 0.0001
cis-Nonachlor 0.6737 0.9136 15 0.75 < 0.0001
cis-Chlordane 0.4037 0.9633 15 0.69 0.0001
Heptachlor Epoxide 0.6294 0.8134 14 0.62 0.0008
Mirex 0.8217 0.6030 6 0.89 0.0028
o,p -DDT 0.5840 0.6040 14 0.41 0.0141
Oxychlordane 0.6694 0.8544 15 0.80 <.0001
p,p'-DDD 0.2375 0.7 597 14 0.50 0.0046
p,p'-DDE 0.6968 0.9216 15 0.93 <.0001
Toxaphene 0.0880 1.0928 3 0.99 0.0486
trans-Chlordane 0.1733 0.9397 12 0.58 0.0041
trans-Nonachlor 0.6430 0.8960 15 0.84 < 0.0001
Bile Dieldrin -0.6196 0.9559 4 0.90 0.0494
cis-Nonachlor -0.3863 0.7646 5 0.97 0.0017
cis-Chlordane -0.4308 0.6314 5 0.83 0.0301
Heptachlor Epoxide -0.3207 0.6959 5 0.79 0.0435
p,p'-DDD -1.1407 1.0748 4 0.95 0.0246
p,p'-DDE -0.6385 0.9472 5 0.94 0.0057
trans-Nonachlor -0.2919 0.6867 5 0.96 0.0039
trans-Chlordane -0.2245 -0.4531 5 0.87 0.0220
Blood NSb
Liver Dieldrin 0.0248 0.7162 7 0.98 <0.0001
cis-Nonachlor -0.2471 0.8448 8 0.92 0.0002
cis-Chlordane -0.5557 0.8876 7 0.97 <0.0001
Heptachlor Epoxide -0.3878 0.8323 6 0.85 0.0084
Mirex -0.0547 0.9557 5 0.89 0.0155
Oxychlordane -0.2855 0.8123 7 0.92 0.0005
p,p'-DDE -0.7696 1.0156 10 0.93 <.0001
trans-Chlordane -0.0722 0.3300 7 0.94 0.0003
trans-Nonachlor -0.2854 0.8263 8 0.98 <.0001
Muscle p,p'-DDE -0.3733 0.8153 10 0.54 0.0160
Mirex 0.1816 -0.2797 0.96 0.0040
a BHC = Benzene hexachloride; DDD = Dichlorodiphenyldichloroethane; DDE =
Dichlorodiphenyldichloroethylene; DDT = Dichlorodiphenyltrichloroethane. b NS = no
significant linear regressions were determined for the 9 chemicals which were detected
both in blood and in yolk.











106
10 -

) 105

0 104
.2
"o
S 103 -
C)
0 102

0) 101 -
o
_J


C. y = -0.4817 + 0.8342x
(r2 = 0.89,p <0.05)













100 101 102 103 104
Log Total OCPs in Yolk (ng/g)


D. y = -0.0865 + 0.6688x
(r2 = 0.55,p <0.05)





O
O *

O
*0 0


100 101 102 103 104
Log Total OCPs in Yolk (ng/g)


Figure 3-1. Linear regressions of total organochlorine pesticide (OCP) concentrations in
maternal tissues against total OCP concentrations in egg yolks. A. Adipose
tissue. B. Liver. C. Bile. D. Muscle.


A. y= -1338.60 + 3.318x
(r2 = 0.95,p <0.05)


100 101 102 103 104 105


100 101 102 103 104 105













CHAPTER 4
MATERNAL FACTORS ASSOCIATED WITH DEVELOPMENTAL MORTALITY
IN THE AMERICAN ALLIGATOR

Recent data suggested maternal organochlorine pesticide (OCP) body burdens and

OCP egg yolk concentrations are significantly correlated, and that significant

relationships between maternal size and maternal body burdens exist. Maternal age and

size has also been shown to have a strong relationship with clutch viability (number of

live hatchings / total number of eggs) and clutch size characteristics (i.e., fecundity,

clutch mass). Specifically, females between 15 and 30 years old (~ 2.3-2.8 m in total

length) produce larger clutches (35-40 eggs / clutch) with increased clutch viability

compared to younger females, which themselves produce smaller clutches (15-25 eggs)

with smaller eggs and have decreased clutch viability. Females older than 30 years tend

to produce clutches similar to 15-30 year old females, with the only exception being

smaller clutches (15-25 eggs) (Ferguson, 1985). Therefore, female size or age may be a

confounding factor when examining the relationship between OCP burdens in yolk and

reproductive performance. In addition, age (or size) and maternal OCP exposure could

cause interactive effects. For example, females of optimum reproductive age may be

more resistant to effects of OCPs; while, younger (or older) females may show increased

susceptibility. Therefore, the objective of the present study was to test the hypotheses

that reproductive efficiency, clutch viability, and mortality rates are significantly

correlated with maternal OCP body burdens, maternal size, or both; and (2) that clutch

size characteristics are significantly correlated with maternal OCP body burdens,

maternal size, or both.









Materials and Methods

The greatest difficulty in examining the relationship between maternal age and

OCP exposure and effects is that determining the age of an alligator requires either long

term monitoring or counting the rings that form in the femur as a result of annual calcium

deposition (Ferguson, 1985). However, this technique is not valid for reproductive

females since femoral bone resorption provides calcium necessary for eggshell formation

and egg yolk nutrition, and subsequently causes the removal of "bone rings" and

underestimation of age (Elsey & Wink, 1985; Wink & Elsey, 1986). In addition,

removing an alligator's limb simply to age it is ethically unacceptable. Given these

difficulties with assigning a chronological age, female size will be used lieu of age. One

potential limitation in using female size as an indicator of age class is that female growth

rates between lakes may differ since dietary composition has been suggested to differ

among OCP-contaminated sites and reference sites (Rice, 2004). Therefore, the

possibility exists that a female from a reference site may be smaller than one from a

contaminated site, even though both are of the same age. This is important since age, in

addition to size, has been shown to be an important determinant of sexual maturity in

alligators. Indeed, alligator ranchers are able to accelerate growth so that a female may

reach six feet in length in 3-4 years, however, these females do not seem to be able to

reproduce until they reach 8-10 years of age (Ferguson, 1985). To control for potential

confounding due to differential growth rates, relationships between female size and OCP

burdens and clutch viability will be evaluated using site and year as covariates. If the

effects of covariates are determined statistically negligible, female data will be grouped

together.









Site Descriptions

Lakes Apopka (N 280 35', W 810 39'), Griffin (N 280 53', W 81 49'), and

Lochloosa (N 290 30', W 820 09') in Florida were selected as collection sites because

prior studies by our laboratory indicate vastly different levels of OCP exposure across

these sites. All three lakes are part of the Ocklawaha Basin. Lake Lochloosa (which is

connected to Orange Lake) was selected as a low exposure (reference) site. Three years

(2000-2002) of data indicate mean total OCP concentrations in egg yolks from the

reference sites (Lakes Orange and Lochloosa) were 102 16 ppb (mean standard

deviation [SD], n = 19 clutches) with a concurrent mean clutch viability rate (number of

live hatchlings/total number of eggs in a nest) of 70 4% (Gross, unpublished data).

Lake Griffin was selected as an intermediate exposure site since yolk concentrations

averaged 1169 423 ppb (n = 42 clutches) and Lake Apopka was selected as a high

exposure site since yolk concentrations averaged 7,582 2,008 ppb (n = 23) for the same

time period (Chapter 2). Furthermore, mean clutch viability rates during this time period

for Lakes Apopka (52 6%, n = 23) and Griffin (43 5%, n = 42) have been below rates

observed for the reference site.

Animal Collections

Adult female alligators and their corresponding clutches of eggs were collected

from Lakes Apopka (n = 19), Griffin (n = 18), and Lochloosa (n = 3) over the course of

four nesting seasons (June 1999 to June 2002). Nests were located by aerial survey

(helicopter) and/or from the ground airboatt). Once nests were located, all eggs were

collected, and the nest cavity was covered. A snare-trap was set perpendicular to the tail-

drag in order to capture the female as she crossed over the nest. After the traps were set,

one member of the trapping crew subsequently transported the eggs to the Florida Fish









and Wildlife Conservation Commission's Wildlife Research Unit (FWC; Gainesville, FL,

USA) and placed the eggs in a temperature-controlled incubator. Snare-traps were

checked later in the evening and early the next morning.

In 1999 and 2000, trapped females were secured and measurements (total length,

snout-vent length, head length and tail girth) were collected along with a blood sample

and a scute for OCP analysis. These females were then immediately released. In 2001

and 2002, females were captured and transported from each lake to the United States

Geological Survey's Florida Integrated Science Center (USGS; Gainesville, FL, USA).

Upon arrival, the animals were weighed, measured, and blood samples were collected

from the post-occipital sinus. Adult alligators were then euthanized by cervical

dislocation followed by double pithing. A full necropsy was performed on each female.

Bile, liver, adipose (composite of abdominal fat and the abdominal fat pad), and tail

muscle samples were collected for later determination of OCP burdens. Liver, adipose

tissue, and muscle were wrapped in aluminum foil, while bile and blood were placed in

scintillation vials. All samples were grouped according to nest identification number

(ID), placed in plastic bags labeled with the appropriate ID, and stored in a -80 C

freezer. Each female's corresponding clutch of eggs was then transferred from FWC to

USGS where yolk samples were collected (two eggs/clutch) and stored with the

corresponding maternal tissues. The remaining eggs were set for incubation in a

temperature/humidity-controlled incubator (31-33 C, 88-92% relative humidity) located

at USGS.

Analysis of OCPs in Maternal Tissues and Yolk

Analytical grade standards for the following compounds were purchased from the

sources indicated: aldrin, alpha-benzene hexachloride (a-BHC), P-BHC, lindane, 6-BHC,









p,p '-dichlorodiphenyldichloroethane (p,p '-DDD), p,p '-dichlorodiphenyldichloroethylene

(p,p '-DDE), dichlorodiphenyltrichloroethane (p,p '-DDT), dieldrin, endosulfan,

endosulfan II, endosulfan sulfate, endrin, endrin aldehyde, endrin ketone, heptachlor,

heptachlor epoxide, hexachlorobenzene, kepone, methoxychlor, mirex, cis-nonachlor,

and trans-nonachlor from Ultra Scientific (Kingstown, RI, USA); cis-chlordane, trans-

chlordane, and the 525, 525.1 polychlorinated biphenyl (PCB) Mix from Supelco

(Bellefonte, PA, USA); oxychlordane from Chem Service (West Chester, PA); o,p '-

DDD, o,p '-DDE, o,p '-DDT from Accustandard (New Haven, CT, USA); and toxaphene

from Restek (Bellefonte, PA, USA). All reagents were analytical grade unless otherwise

indicated. Water was doubly distilled and deionized.

Adipose, liver, bile, and yolk samples were analyzed for OCP content using

methods modified from Holstege et al. (1994) and Schenck et al. (1994). For extraction,

a 2 g tissue sample was homogenized with ~1 g of sodium sulfate and 8 mL of ethyl

acetate. The supernatant was decanted and filtered though a Btichner funnel lined with

Whatman #4 filter paper (Fisher Scientific, Hampton, NH, USA) and filled to a depth of

1.25 cm with sodium sulfate. The homogenate was extracted twice with the filtrates

collected together. The combined filtrate was concentrated to ~2 mL by rotary

evaporation, and then further concentrated until solvent-free under a stream of dry

nitrogen. The residue was reconstituted in 2 mL of acetonitrile. After vortexing (30 s),

the supernatant was applied to a C 18 solid phase extraction (SPE) cartridge (pre-

conditioned with 3 mL of acetonitrile; Agilent Technologies, Wilmington, DE, USA) and

was allowed to pass under gravity. This procedure was repeated twice with the combined

eluent collected in a culture tube. After the last addition, the cartridge was rinsed with 1




Full Text

PAGE 1

DEVELOPMENTAL MORTALITY IN AMERICAN ALLIGATORS ( Alligator mississippiensis ) EXPOSED TO ORGANOCHLORINE PESTICIDES By RICHARD HEATH RAUSCHENBERGER A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2004

PAGE 2

Copyright 2004 by Richard Heath Rauschenberger

PAGE 3

To Jesus, my personal Lord and Savior. John 14:6. “Jesus said to him, I am the way, the truth, and the life: no man comes to th e Father, except by me.” Ephesians 2:8-9. “For by grace you have been sa ved through faith, and that not of yourselves; it is the gift of God, Not of works, le st anyone should boast.”

PAGE 4

iv ACKNOWLEDGMENTS I thank my wonderful wife, Alison; a nd my two sons, Heath and Ben. Their steadfast love, support, and s acrifices allowed me to succes sfully complete the arduous task of earning a Ph.D. I thank my pa rents, Richard E. and Mary Elizabeth Rauschenberger, for their ever-present love faith, and encouragement. I thank my mother-in-law, Sandra Pillow, for baby-sitting Heath and Ben while Alison and I were away at work and for her support and enc ouragement. I thank my parents-in-law, Tommy and Debbie Kirk, for their love and eve r-vigilant prayers. I thank my brother-inlaw, Matt Kirk; and sister-in-law, Kristin De ssert; for their support and encouragement. I thank my late grandfather, M. E. “Pappy” Walls, for showing me the outdoors; my high school biology teacher, Joe David White, for making me a better student; and my high school football coaches, Randy Tapley and Ji m Massarelli, for strengthening my work ethic and ability to deal with adversity. I am forever grateful to Tim Gross for taking me in as a student. I thank him and his wife Denise, for the kindness, generosity, and encouragement they’ve shown to my family and me. I thank my committee members (Marisol Seplveda; Bill Castleman; Richard Miles, Jr.; Franklin Percival; and Steve Roberts) for their support, friendship, and signi ficant contributions to my development as a research scientist. I also want to thank Kent Vliet for sh aring his vast literature and knowledge of alligator reproduction. I espe cially thank Jon Wiebe and Janet Buckland for their friendship and hard work. I am pr ivileged to have had the opportunity to work with the staff and students of our laborator y. I thank Wendy Mathis, Travis Smith, Jesse

PAGE 5

v Grosso, Eileen Monck, James Basto, Shane Ruessler, Carla Wieser, Alfred Harvey, Adriano Fazio, Nikki Kernaghen, Jennifer Mu ller, and Jessica Noggle for their help and friendship. I thank Ken Portier, Gary Steven s, Ramon Littell, Ron Marks, Jon Maul, and Linda Garzarella for providing statistical advice and assistance. I thank the National Institutes of Environmental Health Scienc es Superfund Basic Research Program (grant number P42ES-07375) and the Lake County Wa ter Authority for providing financial support for my education and research project. My name is listed al one as the author of this dissertation, but this work was the product of a team that I am honored to have been a part of and will always remember.

PAGE 6

vi TABLE OF CONTENTS Page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES.............................................................................................................ix LIST OF FIGURES..........................................................................................................xii CHAPTER 1 INTRODUCTION........................................................................................................1 Habitat Degradation in the Ocklawaha Basin...............................................................1 Alligators as Potential OCP Receptors.........................................................................2 Developmental Biology of the American Alligator......................................................3 Post-Ovipositional Development...........................................................................6 FergusonÂ’s Post-Ovipositional Staging Scheme...................................................8 Organochlorine Pesticide T oxicity in Vertebrates......................................................15 Classification, Mode of Action, and Pathology...................................................15 Exposure and Effects of OCPs in Crocodilians...................................................17 Reproductive Problems in Florida Alligators......................................................18 Specific Aims..............................................................................................................20 2 EGG AND EMBRYO QUALITY OF ALLI GATORS FROM REFERENCE AND ORGANOCHLORINE CONTAMINTED HABITATS............................................23 Materials and Methods...............................................................................................24 Egg Collections and Incubation...........................................................................24 Analysis of OCPs in Yolk...................................................................................26 GC/MS Analysis..................................................................................................27 Data Analysis.......................................................................................................28 Results........................................................................................................................ .30 Inter-Site Comparisons of Clutch Characteristics...............................................30 Organochlorine Pesticides Burden s and Clutch Characteristics.........................31 Clutch Survival and OCP Burdens in Egg Yolks................................................32 Average Egg Mass, Clutch Size and OCP Burdens............................................33 Discussion...................................................................................................................34 Inter-Site Comparisons of Clutch Characteristics...............................................34 Clutch Survival Parameters and OCP Burdens...................................................36 Egg and Clutch Size and OCP Burdens..............................................................38

PAGE 7

vii 3 MATERNAL TRANSFER OF OR GANOCHLORINE PESTICIDES.....................54 Materials and Methods...............................................................................................55 Site descriptions...................................................................................................55 Animal Collections..............................................................................................56 Analysis of OCPs in Maternal Tissues and Yolk................................................57 GC/MS Analysis..................................................................................................59 Data Analysis.......................................................................................................60 Results........................................................................................................................ .61 Female Morphological and Reproductive Characteristics..................................61 OCP concentrations in Yolk................................................................................62 OCP concentrations in maternal tissues..............................................................62 Relationships between Maternal Tissue and Yolk Burdens................................63 Relationships between Maternal Mass and OCP concentrations in Eggs and Tissues..............................................................................................................64 Discussion...................................................................................................................65 Evaluation of Predictive Models.........................................................................67 Relationships between Maternal Mass and OCP concentrations in Eggs and Tissues..............................................................................................................69 Maternal body burdens: Toxi cological Implications...........................................69 4 MATERNAL FACTORS ASSOCIA TED WITH DEVELOPMENTAL MORTALITY IN THE AMERICAN ALLIGATOR.................................................80 Materials and Methods...............................................................................................81 Site Descriptions..................................................................................................82 Animal Collections..............................................................................................82 Analysis of OCPs in Maternal Tissues and Yolk................................................83 GC/MS Analysis..................................................................................................86 Data Analysis.......................................................................................................87 Results........................................................................................................................ .88 Discussion...................................................................................................................89 5 MORPHOLOGY AND HISTOPATHOLOGY OF AMERICAN ALLIGATOR ( Alligator mississippiensis ) EMBRYOS FROM REFERENCE AND OCPCONTAMINATED HABITATS...............................................................................99 Materials and Methods.............................................................................................102 Site Descriptions................................................................................................102 Egg Collections.................................................................................................103 Embryo Sampling and Measurement................................................................103 Histopathology..................................................................................................105 Analysis of OCPs in Yolk.................................................................................106 GC/MS Analysis................................................................................................108 Results.......................................................................................................................1 09 Inter-Site Clutch Comparisons..........................................................................109 Intra-Site Live Embryo/Dead Embryo Morphological Comparisons...............110

PAGE 8

viii Inter-Site Comparisons of Morphology of Live Embryos................................111 Live Embryo Morphology and Embryo Survival Relationships.......................112 Live Embryo Morphology and Egg Yolk OCP Burdens...................................113 Embryo Morphological Age, Derived Mo rphometric Variables and Egg Yolk OCP Burdens.................................................................................................115 Histopathology of Live and Dead Embryos......................................................116 Discussion.................................................................................................................117 6 NUTRIENT AND CHLORINATED HYDRO CARBON CONCENTRATIONS IN AMERICAN ALLIGATOR EGGS AND ASSOCIATIONS WITH DECREASED CLUTCH VIABILITY.............................................................................................143 Materials and Methods.............................................................................................145 Egg Collections and Incubation.........................................................................145 Field studies................................................................................................146 Laboratory experiments..............................................................................147 Analysis of Chlorinated Hydrocarbons in Yolk................................................149 GC/MS Analysis................................................................................................150 Nutrient Analysis...............................................................................................151 Data Analysis.....................................................................................................152 Results.......................................................................................................................1 54 Field Study.........................................................................................................154 Case-control cohort study...........................................................................154 Expanded field study..................................................................................157 Laboratory Experiments....................................................................................160 Discussion.................................................................................................................162 7 REPRODUCTIVE EFFECTS OF ORGANOCHLORINE PESTICIDE EXPOSURE IN A CAPTIVE POPULATION OF AMERICAN ALLIGATORS ( Alligator mississippiensis ).......................................................................................................182 Materials and Methods.............................................................................................182 Results.......................................................................................................................1 85 Discussion.................................................................................................................186 8 CONCLUSIONS......................................................................................................196 Introduction...............................................................................................................196 Summary of StudyÂ’s Findings..................................................................................197 Future Considerations and Global Implications.......................................................204 LIST OF REFERENCESÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….208 BIOGRAPHICAL SKETCH...........................................................................................217

PAGE 9

ix LIST OF TABLES Table page 2-1. Reproductive, morphometric, and contam inant parameters measured on clutches of alligator eggs collected during summer 2000, 2001, and 2002................................41 2-2. Explanatory variables included in R DA with forward selection of four best variables...................................................................................................................42 2-3. Summary of clutch parameters and si te comparisons for clutches of American alligator eggs collected during 2000-2002...............................................................43 2-4. Organochlorine pesticide burdens and cl utch parameters and site comparisons for clutches of American alligator eggs collected during 2000-2002............................44 2-5. Results of RDA evaluating associations between clutch survival parameters and OCP variables...........................................................................................................47 2-6. Results of RDA evaluating associations between egg and clutch size parameters and OCP variables...........................................................................................................48 3-1. Morphological and reproductive characteris tics of adult female alligators collected during June 2001 and 2002 from Lakes Apopka Griffin, and Lochloosa in central Florida......................................................................................................................73 3-2. Pesticide concentrations (ng/g wet wt.) in tissues and yolks of adult female alligators collected during June 2001 and 2002 from Lakes Apopka, Griffin, and Lochloosa in central Florida......................................................................................................74 3-3. Regression equations for predicting orga nochlorine pesticide (O CP) concentrations in maternal tissues....................................................................................................78 4-1. Reproductive, morphometric, and contaminant parameters measured on adult female alligators collected during June 1999, 2000, 2001, and 2002..................................93 4-2. Explanatory variables included in R DA with forward selection of four best variables...................................................................................................................94 4-3. Reproductive, morphometric, and contaminant summary statisticsa of adult female alligators collected during June of 1999-2002.........................................................95

PAGE 10

x 4-4. Results of redundancy analysis with au tomatic selection of four best maternal factors associated with varia tion in reproductive efficiency....................................97 4-5. Results of redundancy analysis with au tomatic selection of four best maternal factors associated with variation in clutch size characteristics................................97 5-1. Summary statistics for parameters measured on American alligator clutches collected during June 2001 and 2002.....................................................................122 5-2. Comparisons of egg and embryo morpho metrics of live and dead embryos collected during June-August of 2001 and 2002...................................................................124 5-3. Morphometric comparisons of live embryos collected during June-August 2001 and 2002........................................................................................................................128 5-4. Explanatory variables included in partial redundancy an alysis evaluating relationship between organochlorine pest icide burdens in eggs and embryo morphometrics........................................................................................................131 5-6. Best five organochlorine pesticid e (OCP) variables that account for embryo morphological age and derived morphological parameters ..................................133 6-1. Classification matrix for clutches collected during 2002........................................165 6-2. Reproductive, morphometric, and contam inant parameters measured on clutches of alligator eggs collected during summer 2000, 2001, and 2002..............................165 6-3. Explanatory variables incl uded in RDA with forward selec tion of four best variables for case-control cohort and expanded field studies................................................166 6-4. Summary of clutch parameters on clutches collected during 2002 .......................168 6-5. Evaluation of the relationship between co ncentrations of nutrients, PAHs, and PCBs in eggs and clutch survival parameters via RDA analysis.....................................169 6-6. Evaluation of clutch size parameters a nd explanatory factors for clutches collected during 2002............................................................................................................169 6-7. Evaluation of the relationship between nutrient concentrati ons and explanatory variables for clutches collected during 2002..........................................................169 6-8. Summary and comparison of parameters measured on clutches collected during 2000-2002...............................................................................................................170 6-9. Evaluation of the relationships between clutch survival parameters and explanatory variables via RDA using age as the covariate........................................................171 6-10. Evaluation of the relationships between clutch size parameters and explanatory variables via RDA using age as the covariate........................................................171

PAGE 11

xi 6-11. Evaluation of the relationships between thiamine concentrations and explanatory variables via RDA using age as the covariate........................................................172 6-12. Site comparisons of parameters m easured on clutches collected during 2003......173 7-1. Summary statistics and comparisons of clutch parameters among treated and control groups for years 2002-2004....................................................................................192 7-2. Organochlorine concentrations and bl ood chemistry values of captive adult female alligators sacrificed during 2002............................................................................194 7-3. Explanatory parameters a nd clutch survival parameters with () indicating nature of association and value equal to concordance percentage........................................195

PAGE 12

xii LIST OF FIGURES Figure page 1-1. Map of Ocklawaha Basin..........................................................................................22 2-1. Biplot of clutch survival parameters (solid lines) and or ganochlorine pesticide variables (dashed lines) for clutches of alligator eggs collected from Lake Lochloosa during summer 2001-2002......................................................................49 2-2. Biplot of clutch survival parameters (solid lines) and or ganochlorine pesticide variables (dashed lines) for clutches of a lligator eggs collected from Lake Griffin during summer 2000-2002.......................................................................................50 2-3. Biplot of clutch survival parameters (solid lines) and or ganochlorine pesticide variables (dashed lines) for clutches of alligator eggs collected from Lake Apopka during summer 2000-2002.......................................................................................51 2-4. Biplot of clutch survival parameters (solid lines) and or ganochlorine pesticide variables (dashed lines) for clutches of alligator eggs collected from Emeralda Marsh during summer 2000-2002............................................................................52 2-5. Biplot of egg and clutch size parameters (solid lines) and organochlorine pesticide variables (dashed lines) for clutches of alligator eggs collected from Lake Lochloosa during summer 2001 and 2002...............................................................53 3-1. Linear regressi ons of total organochlorine pest icide (OCP) concentrations in maternal tissues against total OCP concentrations in egg yolks. ...........................79 4-1. Biplot of maternal factors (dashed lines) and clutch su rvival parameters (solid lines) of American alligators collected during June 1999-2002. .....................................98 5-1. Representative developmental stages of embryos that were collected from Lakes Lochloosa (reference site), Apopka, and Griffin, and Emeralda Marsh during 20012002. .....................................................................................................................134 5-2. Ordination biplot of embryo morphometric parameters (solid lines) and organochlorine pesticide (OCP) variables (dashed lines) for embryos collected at chronological age Day 14.......................................................................................135

PAGE 13

xiii 5-3. Ordination biplot of embryo morphometric parameters (solid lines) and organochlorine pesticide (OCP) variables (dashed lines) for embryos collected at chronological age Day 25 ......................................................................................136 5-4. Ordination biplot of embryo morphometric parameters (solid lines) and organochlorine pesticide (OCP) variables (dashed lines) for embryos collected at chronological age Day 33.......................................................................................137 5-5. Ordination biplot of embryo morphometric parameters (solid lines) and organochlorine pesticide (OCP) variables (dashed lines) for embryos collected at chronological age Day 43.......................................................................................138 5-6. Ordination biplot of derived embryo morphometric pa rameters (solid lines) and organochlorine pesticide (OCP) variables (dashed lines) for embryos collected at chronological age Day 14.......................................................................................139 5-7. Ordination biplot of derived embryo morphometric pa rameters (solid lines) and organochlorine pesticide (OCP) variables (dashed lines) for embryos collected at chronological age Day 25.......................................................................................140 5-8. Ordination biplot of derived embryo morphometric pa rameters (solid lines) and organochlorine pesticide (OCP) variables (dashed lines) for embryos collected at chronological age Day 33.......................................................................................141 5-9. Ordination biplot of derived embryo morphometric pa rameters (solid lines) and organochlorine pesticide (OCP) variable s (dashed lines) embryos collected at chronological age Day 43.......................................................................................142 6-1. Biplot of clutch surviv al parameters and explanatory factors for clutches collected during 2002............................................................................................................175 6-2. Biplot of clutch size parameters and explanatory vari ables for clutches collected during 2002............................................................................................................176 6-3. Biplot of nutrient concentrations in e ggs (solid arrows) and explanatory variables (dashed arrows)......................................................................................................177 6-4. Relationships between embryo age and thiamine phosphorylation in egg yolk for 29 clutches collected duri ng 2002 from Lakes Lochloosa, Griffin, Apopka, and Emeralda Marsh.....................................................................................................178 6-5. Biplot of clutch survival parameters and explanatory va riables for clutches collected during 2000-2002...................................................................................................179 6-6. Biplot of clutch size va riables (solid lines) and expl anatory variables (dashed lines) for clutches collected during 2000-2002................................................................180

PAGE 14

xiv 6-7. Biplot of thiamine egg yolk concentra tions (solid lines) and explanatory variables (dashed lines) measured on clut ches collected during 2000-2003.........................181

PAGE 15

xv Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DEVELOPMENTAL MORTALITY IN AMERICAN ALLIGATORS ( Alligator mississippiensis ) EXPOSED TO ORGANOCHLORINE PESTICIDES By Richard Heath Rauschenberger December 2004 Chair: Timothy S. Gross Major Department: Veterinary Medicine—Physiological Sciences Since the early 1900s, the lake s of the Ocklawaha Basin in central Florida have experienced ecological degradation due to anthropogenic development. One species affected by degradation has been the American alligator ( Alligator mississippiensis ). Decreased clutch viability (pr oportion of eggs in a nest that yield a live hatchling) was observed in the years after a chemical spill in which large amounts of sulfuric acid and dicofol, an organochlorine pesticide (OCP), flowed into Lake Apopka. Lake Apopka and other lakes in the Ocklawaha basin have al so been contaminated by urban sewage and agricultural chemicals, with agricultural chem icals entering the lakes via rainfall run-off or back-pumping of water from agricultural la nds). Decreased hatch rates are a problem at Lake Apopka, as well as at other OCP-cont aminated sites in Florida. The purpose of my study was to determine the causes for decr eased clutch viabilit y, and to test the hypothesis that maternal exposure to OCPs is associated with embryonic mortality in alligators.

PAGE 16

xvi Field studies involved collec ting and artificially incuba ting eggs from reference sites (Lake Lochloosa) and from OCP-cont aminated sites (Lakes Apopka, Griffin, and Emeralda Marsh Restoration Area) to evaluate clutch viability as a function of egg and maternal OCP concentrations. Nutrient content of eggs and histopathology and morphometrics of embryos were also evaluate d to identify potential factors associated with embryo mortality. In addition, a novel laboratory experiment exposed a captive population of adult alligators to an OCP mixtur e, and compared OCP burdens in eggs and clutch viability with a captive control group. Results of field studies suggested that OCP concentrations (ng total OCP/g egg yolk, Mean SE) in reference site clutches (n = 19; 102 16) were significantly ( = 0.05) lower than those of Apopka (n = 23; 7,582 2,008), Griffin (n = 42; 1,169 423), and Emeralda Marsh (n = 31; 15,480 2,265). Clutches from reference sites also had significantly higher clutch viability (70 4 %) than those of Apopka (51 6%), Griffin (44 5%), and Emeralda Marsh (48 6 %). Furthermore, decreased thiamine concentrations in eggs may play a role in de creased clutch viability in wild clutches. Results of the captive study suggested that treated females produced eggs containing higher OCP concentrations (n = 7; 13,300 2,666) than controls (n = 9; 50 4). Eggs of treated females also exhibited decreased viability (9 6%) as compared to controls (44 11%). These field and laboratory studies s upport the hypothesis that maternal exposure to OCPs is associated with decreased clutch viability in American alligators, and that thiamine deficiency may also be a contribu ting factor in reduced clutch viability.

PAGE 17

1 CHAPTER 1 INTRODUCTION Habitat Degradation in the Ocklawaha Basin In central Florida, several lakes within the Ocklawaha River Basin (Fig. 1-1) have experienced severe degradation of habitat quality since the earl y 1900s, as agricultural and urban development progressed. Ind eed, Lake Apopka (headwaters of the Ocklawaha) was once renowned for its clear water and its excellent largemouth bass fishing. More recently, Lake Apopka has gain ed world-wide notorie ty as the “poster child” for polluted lakes, because of hi ghly publicized problem s associated with environmental contamination. Initial degradation of Lake Apopka and other lakes within the Ocklawaha Basin occurred as the result of the loss of thousands of hectares of marsh habitat through the agricu ltural practice known as muck farming (which involves installing levees around an area of marsh, so the marsh can be drained; allowing the fertile peat to be farmed). This farming practice began in the 1940s and continued into the 1980s (Benton et al., 1991). In addition to sewer discharge from the city of Winter Garden entering the Lake Apopka, organochlor ine pesticides (OCPs) were heavily and widely used to control crop-dest roying insect pests. Since the 1980s, use of most OCPs ha s been discontinued since they were determined to be persistent environmental c ontaminants that resist metabolic degradation and bioaccumulate in animal tissues, wh ere they are potenti ally carcinogenic, immunotoxic, endocrine disrupti ng, and developmentally toxic (Fairbrother et al., 1999;

PAGE 18

2 Ecobichon, 2001). Altered func tion of the reproductive a nd endocrine systems of wildlife and human populations have been sugge sted to occur after exposure to a variety of OCPs and OCP metabolites such as dichlorodiphenyltrichloroethane (DDT), dichlorodiphenyltrichloroethyl ene (DDE), methoxychlor, dico fol, chlordane, dieldrin, and toxaphene (Colborn et al., 1993 ; Longnecker et al., 2002). Further degradation and OCP contamina tion occurred in Lake Apopka in 1980. A chemical spill occurred when a highly acidi c wastewater pond at the Tower Chemical CompanyÂ’s main facility overflowed into th e Gourd Neck area of Lake Apopka (Fig. 11). Because of the large volume and acidity (sulfuric acid), and the high levels of DDT, dicofol, and related OCP contaminants that en tered the relatively narrow area of the lake, aquatic vegetation and animals were severely affected. In 1983, the area was placed on the US Environmental Protection AgencyÂ’s (EPA ) National Priority Site List and became a part of the Superfund program; whic h was created by the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA), later amended by the Superfund Amendments and Reauthorizat ion Act (SARA). The CERCLA and SARA provide authority for the governme nt to respond to the release a nd/or threat of release of hazardous wastes, and allow cleanup and enfo rcement actions. Lake Apopka is still listed and groundwater toxicity testing is ongoing (EPA, 2004). Alligators as Potential OCP Receptors The American alligator is an important member of Florida wetlands and plays important roles in the ecology, es thetics, and economy of Flor ida. Therefore, identifying physiological and ecological charac teristics related to potential sus ceptibility to effects of contaminants, as well as potential exposure routes, is important in managing populations

PAGE 19

3 for optimal human use. Especially importa nt to consider, in regard to wildlife populations, are potential effect s of OCPs on reproduction. One of the first qualities that may be rela ted to an alligatorÂ’s susceptibility to reproductive effects of OCP contaminants is that alligators do not at tain sexual maturity until approximately 6-10 years of age, whic h allows exposure and bioaccumulation of OCPs to occur before reproductive maturity. Potential implications are that, as females begin to mobilize body stores during vitell ogenesis, the lipophilic OCPs that have accumulated in their fatty tissues during their li fespan would likely be deposited in what will later be the embryosÂ’ sole source of nutrition (egg yolk) Secondly, adults exhibit a long reproductive period (over 30 years), and a long life span (over 50 years) (Ferguson, 1985), and are higher order predators (which allows for increased OCP exposure and bioaccumulation, possibly leading to altere d endocrine and reproductive function). Thirdly, alligators build nests th at can be identified from considerable distances (which aids in egg collections), lay a large number of eggs (approximately 40 eggs per clutch), and have a long developmental period of 65-72 days (Ferguson, 1985), allowing extended exposure at a potentially critic al stage of development. Thus, the propensity for OCPs to be bioaccumulated and biomagnified in biot a (combined with the alligatorÂ’s reproductive biology, longevity, ecological troph ic level, and relatively long in ovo developmental period) suggests the poten tial for OCPs to alter reproductive function. Developmental Biology of the American Alligator Understanding normal embryonic devel opment is an obvious necessity in determining the occurrence of abnormal embryonic development and identifying critical periods of development (e.g., organogenesis). Therefore, this brie f review summarizes pre-ovipositional and post-ovi positional development of the alligator embryo.

PAGE 20

4 Pre-ovipositional Development Overall, when compared to other verteb rate species such as the domestic chicken and domestic pig, there is a paucity of data re lated to crocodilian development. Despite the relatively low number of publications, th e quality of papers c overing early embryonic development is fairly high, considering that much of the research took place approximately a century ago. The most a ppropriate place to begin discussing embryonic development would be the point when fertiliz ation occurs. However, the precise timing and location of fertilization within a fe male alligatorÂ’s oviduct is unknown and inadequately studied. Pre-ovipositional development has been examined by sacrificing gravid females and collecting their eggs and embryos. Sacr ifice of gravid females was required since alligator embryos are at a more advanced stag e of development at the time of oviposition (Clarke, 1891). The earliest developmental stage examined in these pre-ovipositional studies were of Nile crocodile embryos ( Crocodylus niloticus ), in which all embryos exhibited body folds, a neural medullary gr oove, an embryonic shield, area opaca, early gut, and area pellucida (Voeltzkow, 1892). After the appearance of the neural folds, the amniotic head fold is formed from an anterior fold in the blastoderm. The head fo ld is crescent shaped, because it begins to develop with its free ends pointing toward th e posterior end of the embryo, and develops craniocaudally. The amniotic primordium deve lops in continuity with the head, and is derived from the somatopleure around the trunk. Craniocaudal separation of the embryo fr om the blastoderm occurs partly as a result of the development of the dorsal amniotic fold, but separation is not complete until

PAGE 21

5 post-ovipositional stage 3 (Day 3). The neural groove and blastopore become clearly demarcated as the ectoderm and endoderm of the blastoderm develop. The endoderm may form extensions that pe netrate the underlying yolk. Th e blastopore goes through the entire blastoderm, with the primitive str eak located posterior to the blastopore (Voeltzkow, 1892). As the body folds develop, the border between embryonic and extra-embryonic tissues becomes visible. At th is point, the beginning of the foregut is discernable, and the notochord stretches from the midline of the head fold to the anterior border of the blastopore. The primitive streak and primitive groove lie posterior to the blastopore, with the primitive groove being continuous at its posterior end. The primitive streak extends to a little less than halfwa y between the head fold and blastopore (Ferguson, 1985). Neural folds have two origins. The first is a secondary fold located anteriorly in the head region, and growing posteriorly along the median dorsal line to form a V-shaped process, with the apex pointing toward the bl astopore. The second is posterior folds that arise as ectodermal ridges extending forwar d from the blastopore, circumventing the neural groove. The apex of the V-shaped s econdary head fold later disappears, and each of the separate arms becomes continuous with the corresponding posterior neural fold. Thus, the secondary head fold forms the anterior part of the neural fo lds. Closure of the folds occurs first in the middle region of th e embryo closer to the anterior end of the neural groove in alligators (Ferguson, 1985) but closer to the posterior end in Nile crocodiles (Voeltzkow, 1892). After the closure of the neural canal, th e blastoporal or neurenteric canal is no longer visible. The neurenteric canal runs from its posterior cranioventral opening to

PAGE 22

6 where it opens into the neural groove at it s caudal limit. During this period, somites develop along the median axis, with the fi rst pair developing halfway between the anterior and posterior ends. The periphera l somatic cells are compactly arranged, and contain small myocoels within the center of the somites. The mesodermal layers cleave and form the somatic and splanchnic components as the foregut develops. The head fold of the embryo is positi oned ventrally into the underlying yolk, which is accentuated by the bending of the an terior neural folds, and by the cranial flexure that occurs later. At this pre-ovi positional stage of development, the embryo has not yet attached to the inner su rface of the eggshell membrane. Because embryos are at an advanced st age of development at the time of oviposition (and because an entire clutch ty pically hatches within a 2-day period, with most hatchlings being similar in size), it app ears that fertilization occurs over a short time period; and that embryos are actively developi ng during the next 2to 3-week period in which the ova receive albumin, eggshell me mbrane, and eggshell depositions (Ferguson, 1985). Presently, little information exists about gaseous exch ange and embryonic metabolism before oviposition, or about the processes that prev ent the embryo from attaching to the top of th e egg before oviposition. Post-Ovipositional Development Post-ovipositional development is better understood than pre-ovipositional development. Again, the amount of literat ure concerning crocodilian development is miniscule compared to the amount of literature deali ng with human and chicken embryology. One important area to address when disc ussing post-ovipositi onal development is the staging scheme. Establishing a staging scheme or a normal table of development for

PAGE 23

7 any species allows results of various studies to be compared (B illet et al., 1985). The currently accepted staging scheme for cr ocodilian embryology was proposed by Ferguson (1985). Before FergusonÂ’s, the only other staging systems related to crocodilians came from Voeltzkow (1892), Reese (1912), and We bb et al. (1983). These works were impressive, considering the conditions th ese pioneers faced; but many stages were missing, and incubation conditions were poorly controlled. Ferguson (1985) improved on their work by monitoring and controlling the temperature (30C) and the relative humi dity (95-100%) at which the eggs were incubated, allowing duplication of his e xperiment and standa rdization of the characteristics one should see in an embryo, given its stage. This accepted staging scheme is based on external morphological features, with limb and eye development being important diagnostic elements. With re spect to craniofacial development, a fair amount of data exists, because of FergusonÂ’ s focus on the structure and development of the palate in the alligator, and on how its de velopment relate s to stage (Ferguson, 1981). Although the relationship between craniofacial development and developmental stage has been studied, information relating stage a nd development in other organ systems is somewhat lacking. Alligator embryos are very sensitive to temperature. For example, 26-34C is the optimum incubation temperature; anything a bove or below for an extended period will result in increased mortality (Ferguson, 1985) Furthermore, 0.5-1 C changes can mean the difference between an entire clutch of embryos being 100% females or 100% males, since crocodilians exhibit temperature-depe ndent sex determinati on (Lang & Andrews, 1994).

PAGE 24

8 Ferguson’s Post-Oviposi tional Staging Scheme Because our study used Ferguson’s stagi ng scheme, a summary description of Ferguson’s (1985) staging scheme, it is su mmarized here. The normal table of development for crocodilians was based on examination of 1500 Alligator mississippiensis embryos, 300 Crocodylus porosus embryos, and 300 Crocodylus johnsoni embryos. One bias is that all of the alligator embryos used in developing this scheme originated from Rockefeller Wildlife Refuge, located in southern Louisiana. Alligator embryos from other geographic areas may develop at different rates, given the same incubation conditions. Alligators inhabiting Arkansas and North Carolina experience a shorter summer compared to popul ations inhabiting southern Louisiana or Florida. Shorter summers mean that optimal nest temperatures are maintained for a shorter period of time. Thus, embryos from more northerly latitudes may develop at an increased rate compared to embryos from sout herly latitudes (given identical incubation conditions), since the northern embryos must complete development within a shorter time frame. This hypothesis is supported by evidence that crocodilian species ( Crocodylus porosus and C johnsoni ) living along the equator have longer and more variable incubation periods and slower embryonic development than the (more northerly) Louisiana alligator (Deeming & Ferguson, 1990). Setting aside the potential bias descri bed above, developmental “stages” are determined by morphological characteristics alone, and are applicable to embryos regardless of incubation temperature. Howe ver, the developmental day(s) associated with each stage are only valid if the eggs ar e incubated at 30C with a relative humidity of 95-100%. Temperatures lower than 30 C slow the rate of development, and temperatures above 30C have been shown to increase the rate of development. Low

PAGE 25

9 humidity within the nest has been shown to dehydrate eggs, causing embryonic mortality and alterations in growth pa tterns (Deeming & Ferguson, 1990). Stage 1 covers the period from oviposition to the end of the first 24 hours, and is characterized by the embryo and blastoderm be ing not attached to the top of the inner eggshell membrane. The heart is a simple S-shaped tube. There are 16-18 pairs of somites along the trunk, and 3 pair s of somitomeres anterior to the otic vesicle. Although the brain has not yet regionalized, optic placode s and vesicles are present on the head. Body torsion has not begun. The notocord is ev ident, the gut is incomplete caudally and opens ventrally, and blood ve ssels are not present in th e extraembryonic membranes. Stage 2 (Day 2) embryos have 21-25 pair s of somites and a three-loop heart. However, one of the most notable characteristics is that the embryo attaches to the top of the egg, causing an opaque spot to form that is visible in an otherw ise translucent egg, when the egg is candled. Blood vessels are now visible, and the hindb rain is discernable as a clear transparent region. The lens placode and optic cup are defined, and no body torsion has occurred. Stage 3 (Day 3) embryos have 26-30 som ites, and are completely delineated from blastoderm. Forebrain, midbrain, and hindbrai n are now discernable, and the optic cup has an elongated horseshoe shape, extending below the lens vesicle to the primitive oronasal cavity. The head is positioned at a right angle to th e body, but no body torsion has occurred. Stage 4 (Day 4) embryos have 31-35 pairs of somites with the tail being distinct, straight, and unsegmented at the posterior e nd. Body torsion has started, with the cranial half rotated so that the right surface is contacting the shell membrane, while the left is

PAGE 26

10 parallel with underlying yolk. The caudal half of the embryo remains at a right angle to the yolk. The heart is displaced from midline to the left side of the embryo. Three cranial arches are present; and cranial nerves to the cranial arches are visible, using oblique or transmitted illumination. Stage 5 (Day 5) embryos have 36-40 pair s of somites, and the tail-tip bends ventrally at a right angle to th e body, with 3-5 somites visible at its base. Body torsion is complete except for the tail. The otic pit is dorsal to the junction of the 2nd and 3rd brachial arches, and its ex ternal opening is closed. Stage 6 (Day 6) embryos have visible nasa l placodes, and the hi ndlimbs are barely discernable on each side; with the right hind limb slightly advanced over the left. Forelimb buds are not yet present, and body tors ion is complete. The olfactory bulbs, forebrain, and midbrain are distinct. In the hindbrain, 4-6 neuromeres are discernible. Foregut and hindgut are formed, but midgut is incomplete ventrally. Major vitellogenic blood vessels emerge at the level of the 18th somite and smaller ones at the 6th and 11th somites. Embryos at Stage 7 (Day 7) have distin ct hind limb buds. In addition, forelimb buds are barely visible and extend over som ites 12-15. The midbrain bulge is evident, and the tail-tip is curled at 90 to the rest of the tail. Three brachial arches are present; and at the level of the heart, the cranial end is bent at 90 to the rest of body. Embryos at stage 8 (Day 8) have nasal pits external to the swellings of the olfactory bulbs, and distinct forelimb and hind limb buds that extend over somites 11-16 and 2732, respectively. An apical ectodermal ridge is developing on the hind limb bud, and the tail is coiled through 2 turns and has 12-18 somites.

PAGE 27

11 Stage 9 (Day 9) embryos have four brachia l arches, and a visible maxillary process extending to the midpoint of the eye. The optic cup is large and round but unpigmented. A distinct apical ectodermal ridge is present on the hind limb, and the hind limb bud extends beyond the forelimb. The tail is cu rled through three 90 turns. The heart exhibits distinct atria an d ventricles, and lung primord ia are visible through the pericardial sac. Midgut and body walls are open ventrally from the caudal limit of the pericardial sac to 2/3 of the way down the body, and the liver and mesonephros are barely visible. Stage 10 (Day 10-11) embryos have pigmen ted eyes (except for a central opaque lens) with the right eye developi ng pigmentation earlier and darker than the left eye. Five brachial arches are present, and medial and la teral processes are distinct elements on each side of the nasal pits. Maxillary processes delimit a distinct groove beneath the eye. The tail is coiled through four 90 turns, and th e liver and mesonephros are clearly visible through the body walls. Stage 11 (Day 12) embryos have a visible nasal pit slit forming between the medial and lateral processes. Forelimb and hi nd limb buds extend caudally from the body wall and exhibit distinct ap ical, ectodermal ridges. The fo relimb has a distinct constriction that separates the distal and proximal elemen ts, with constriction less obvious in the hind limb. A loop of midgut is visible at the um bilicus, the eye exhibits a distinct black pigment in the iris, and the chorioallantois extends 2/3 around the breadth of the shell membrane. Embryos at stage 12 (Day 13-14) have a di stinct notch in the midline of the face between the medial nasal processes. Fore limbs are starting to bend in the region of

PAGE 28

12 constriction, so that they are positioned cl oser to the pleuron of the embryo. The elongated hind limb shows little differentiati on into proximal and distal elements and, although there is a distinct ap ical ectodermal ridge, footplate formation is barely discernable. Stage 13 (Day 15) embryos have distinct na sal pit slits, and forelimbs are now bent toward the pericardium. The distal portion of the hind limb is flattened and enlarged into a footplate primordium. The chorioallant ois now extends as a ring around the inner circumference of the centr al eggshell membrane. Embryos at Stage 14 (Day 16-17) have nasa l pit slits that are closed due to the merging of the medial nasal, lateral nasal, a nd maxillary processes. Foot and hand plates are distinct, with the former more advanced than the latter. Lower jaw extends onequarter beneath the upper jaw, the upper ea rflap is overgrowing the external ear opening, and the embryonic face rests on the large bulge of the thorax. A large loop of gut herniates through the narrow umbilical stalk and touches the yolk, and the abdominal viscera are visible through body walls. The tail is coile d and kinked at the tip, and contralateral reflexes occur. Stage 15 (Day 18-20) embryos have lower jaws that extend one-third to one-half the length of the upper jaw. Th e anlage for the upper eyelid is an elevated rim of tissue above each eye. Distinct and proximal and dist al regions, as well as hand and foot plates are present on both the fore and hind limb. Th ere is a distinct hollow in the face beneath the anterior one-third of the eye. Stage 16 (Day 21) embryos exhibit faint di gital condensations in the footplate but not the hand plate. The lower jaw is now tw o-thirds the length of the upper jaw, with the

PAGE 29

13 upper jaw being hook-shaped around the perica rdial ridge. Caruncle development is observed, with two tiny widely spaced thickeni ngs that are just disc ernable on the tip of the snout. Embryos at stage 17 (Day 22-23) exhibit mesodermal condensations for the five forelimb digits and four hind limb digits, the head is extended off of the pericardial sac by neck elongation, and the exte rnal earflap is distinct. Stage 18 (Day 24-26) embryos have discernabl e, distinct cartilag inous digital rays on the hand and foot. The margins of upper eye lid anlage extend over the superior rim of the iris, forming a distinct groove between the eyelids and the eye. Dorsal scalation is now evident, and the pericardial sac is starting to submerge into the ventral thoracic wall. Stage 19 (Day 27-28) embryos have upper and lower eyelids, and the lower jaw lies behind the anterior margin of the upper jaw. Interdigital clefting has started, and slight marginal notches can be seen, particularly in the footplates. White flecks representing ossification are visible around the upper and lower jaws. Stage 20 (Day 29-30) embryos have nail an lages starting to develop, first on the most medial digit of the foot, then on adj acent digits; followed by the most medial digit on the hand, and finally on the adjacent hand digits. Interdigital clefting now extends one-quarter the length of the digits, and the lower jaw is in adult relationship with the upper jaw. The pericardial sac is one-quarter withdrawn into the body, and ossification is evident in the proximal and distal elements of limbs. Scale formation is evident dorsally, and scutes (osteoderms) are beginning to a ppear in the neck region near the skull. Stage 21 (Day 31-35) embryos have inte rdigital clefting no w extending threequarters down the digits, and phalanges can be distinguished. Scal es are now visible on

PAGE 30

14 the ventral body wall; and dorsally on the s nout, neck, body, and tail. Scutes on neck are clearly defined. The pericardial sac is one -half withdrawn into the body, and a white ring in the iris surrounds the outline of the lens of the eye. Both upper and lower eyelids overlap the eye. Stage 22 (Day 36-40) embryos have pigmented margins of the upper jaw, ventral flank, and proximal and distal elements of the limbs. Interdigital clef ting is at the adult level, and the eyelids are typica lly closed from this point fo rward. The pericardial sac is two-thirds withdrawn. Stage 23 (Day 41-45) embryos have more extensive pigmentation, with the embryos appearing light brown with dorsal st ripes. Scales are present on distal and proximal elements, and nails have a slight distal elevation. Th e sensory papillae are present on lateral jaw margins, and scales are evident on gular skin. The midbrain is visible as a white bulge at th e back of the cranium, and the pericardial sac is threequarters withdrawn. Stage 24 (Day 46-50) embryos are blacker Nails on hands have elevations at their tips, and the nails are starting to form curves. The midbrain is covered by pigmented skin. The pericardial sac is fully withdrawn and the mid line is closing. The volume of yolk outside the body cavity is large, and scales and scutes are evident all over embryo. Stage 25 (Day 51-60) embryos look identical to hatchlings, except smaller. The external yolk is beginning to be withdr awn, and few gross morphological changes are evident at this and later stages. Growth rela tionships (head length: total length ratio) and the amount of external yolk present are the major observable differences.

PAGE 31

15 Stage 26 is not present in alligators. This stage was esta blished using tooth eruption sequences and is useful only for saltwater crocodiles ( Crocodylus porosus ) and freshwater crocodiles ( Crocodylus johnsoni ). Stage 27 (Day 61-63) embryos have withdr awn the yolk sack into the body, ending with skin forming across the umbilical scar. The last stage before hatching (Stage 28, Day 64-70) ends with the umbilical scar being diminished in length and width. Overall, the first 35 days are a period of rapid orga nogenesis, and the second 35 days are characterized by embryo growth. Since organogenesis has been shown to be a sensitive period in regard to effects of developmental toxicants (Schmidt & Johnson, 1997), the first 35 days of incubation appear to be the most susceptible time points for toxicant-induced mortality. In summary, the establishe d staging scheme provides a way to estimate the age of the clutch at the time of co llection, and allows on e to later determine if a clutch is undergoing normal development. One can dete rmining if a clutch is undergoing normal development by examining embryos at preselected time points and comparing their morphological age to their calendar stage (i.e., does an embryo exhibit the normal morphological characteristics that it should exhibit, given its calendar age?). In addition, embryonic development may be compared among clutches and among populations, by collecting embryos at pre-determin ed stages of development. Organochlorine Pesticide T oxicity in Vertebrates Classification, Mode of Action, and Pathology Organochlorine pesticides (also known as chlorinated hydrocarbon insecticides) may be separated into five classes of com pounds. These classes are DDT and its analogs, cyclodienes and similar compounds, toxaphene (composed of several congeners), mirex

PAGE 32

16 and chlordecone (which have cage-like structur es), and benzene hexachloride (BHC). In rodent models, studies suggest that OCPs can adversely affect the function of neurons and cause cellular damage to the liver and kidneys (Smith, 1991). Organochlorine pesticides affect neural transmi ssion by altering enzyme activity (Ca2+-ATPase, phospokinase) and the electrophysical properties (K+, Na+ ion exchange) of nerve cell membranes. Different analytes may elicit si milar effects (neurona l hyperactivity), but by different mechanisms. For example, studies suggest DDT and its analogs affect the nerve axon by keeping Na+ channels open longer than normal. Cyclodienes, alternatively, may affect neural transmission at pres ynaptic terminals and may affect the -aminobutyric acid (GABA)-regulated chloride channel. Although they can cause severe neural dysfunction, little morphological ch anges are evident in neural ti ssue, even at lethal doses (Smith, 1991). Morphological changes are evident in the liver and include hepatocellular hypertrophy and focal necrosis. Hypertr ophy is due to enlargement of the smooth endoplasmic reticulum (SER) and formation of a lipid droplet in the center of the SER (caused by OCP-induced expression of mi crosomal enzymes within the SER). Functional alterations may also occur in hepatocytes, with disruption of intercellular communication (by hindering transfer of growth inhibitors) (Smith, 1991). Morphological changes have also been found in the liver and kidney of fish chronically exposed to organochlorine pestic ides. For example, chronic exposure to OCPs induce hepatic lesions, such as foci of vacuolated hepatocytes and spongiosis hepatic (lesions of hepatic parenchyma). Renal lesions induced by chronic OCP

PAGE 33

17 exposure include dilation of tubular lumina, and vacuol ization (degeneration) and necrosis of tubular epithelium (Metcalfe, 1998). In addition to morphological changes, or ganochlorine pesticides may adversely affect endocrine and reproductive function in la boratory models and wildlife populations. Mechanisms include direct toxicity on endoc rine glands (such as o,pÂ’-DDDÂ’s ability to permanently inactivate the adrenals), comp etitive binding of steroid hormone receptors, increased expression of steroid-metabolizing hepatic microsomal enzymes, and inhibition of hormone synthesis (such as DDE-induced in hibition of proglandin synthesis, leading to eggshell thinning in raptors) (Gross et al., 2003). Exposure and Effects of OCPs in Crocodilians Current knowledge on the effects of envi ronmental contaminants on crocodilian reproductive physiology is important in unders tanding the likelihood of developmental alterations occurring as a result of expos ure; and understanding which mechanisms may be involved. Campbell (2003) reviewed the effects of organic and inorganic contaminants on crocodilians. Campbell reported only 26 studi es related to the bioaccumulation of organic contaminants, with just 35% (8/23) of crocodilian sp ecies being represented. Of the 26 studies, 38% involve d American alligators ( Alligator mississippiensis ), 26% involved Nile crocodiles ( Crocodylus niloticus ), 13% involved American crocodiles ( Crocodylus acutus ), and 13% involved MoroletÂ’s crocodile ( Crocodylus moreletii ). Slightly more studies were found that investig ated effects of organi c contaminants. With respect to these 39 studies, only 13% (3/23) of crocodilian species were represented, consisting of the American alligator (91% of studies), the Nile cr ocodile (5%), and the African dwarf crocodile ( Osteolaemus tetraspis 4%). Of these studies, American

PAGE 34

18 alligators are the only species in which an effort has been made to determine the relationship between OCPs and depressed hatch rates, with most of this work involving populations in central Florida. Reproductive Problems in Florida Alligators In the early to mid 1980s, studies showed that the population of juvenile alligators inhabiting the aquatic ecosys tem of Lake Apopka, Florida, declined by 90%. This decline was preceded by a 1980 chemical spil l and decades of OCP contamination via anthropogenic activities described earlier. Th e loss of juveniles was attributed primarily to a dramatic decrease in clutch viability (the proportion of eggs in a clutch that produce a live hatchling) (Woodward et al., 1993). Alterations in sexual diff erentiation, sex steroid ho rmone concentrations, and metabolism were also documented among La ke Apopka alligators. For example, testosterone was lower in male alligators from Lake Apopka as compared to those of control sites. Ovaries of female juven ile alligators from Lake Apopka showed abnormalities, suggesting that reproductive al terations were occurring in both sexes (Gross et al., 1994; Guillette et al., 1994; Gr oss, 1997). In addition, high concentrations of OCPs were measured in egg yolk, but concen trations were not clea rly associated with increased mortality (Heinz et al., 1991). Later studies suggested that the cause for the population decline was potentially more complex than previously suggested. First, poor egg viability for Lake Apopka alligators was more closely associated with muck farm reclamation (wetland restoration) sites than with tissue and egg concentrations of the predominant pe sticide residue (DDE) (Giroux, 1998). Second, altered endocrine f unction and decreased egg viability were documented among alligators at another si te, Lake Griffin, where tissue and egg

PAGE 35

19 concentrations of residues such as DDE are mo dest or intermediate compared with those of Lake Apopka. However, like Lake Apopka Lake Griffin is hi ghly eutrophic and has adjacent muck farms and muck farm reclama tion areas (Marburger et al., 1999). Third, poor reproductive success among Lake Apopka a lligators appeared to result from both decreased proportions of fertile eggs th at produce a live hatchling and decreased proportions of hatchlings that survive thr ough the first 20 days of life (which is the toxicant-sensitive organogenesis period); and decreased proportions of unbanded eggs (i.e., eggs that are nonviable on initial examination) (Masson, 1995; Wiebe et al., 2001). Unbanded eggs show no evidence of embr yo-eggshell attachment (as indicated by the presence of an opaque spot or band th at results from fusion of extraembryonic membranes to the dorsal portion of the inne r eggshell membrane). Unbanded eggs may result from very early embr yo mortality (fertilization has been confirmed in many cases by the presence of paternal DNA, via DNA micros atellite analysis); or may result from infertile eggs (Rotstein, 2000). The last similarity between alterations in alligator populations of Lake Griffin and Lake Apopka is increased mort ality among adult Lake Griffin alligators (Schoeb et al., 2002), which is similar to increased adult mo rtality on Lake Apopka in the early 1980s. These data indicate that alligato r populations are adversely aff ected at each of several life stages. Although anatomic and endoc rinologic effects of exposure to endocrine-disrupting OCPs could account for many of these effects, additional underlying mechanisms are almost certainly i nvolved. Overall, th ese data point to a complex process involving the introduction of OCPs into th ese aquatic ecosystems from

PAGE 36

20 chemical spillage or from muck farming a nd reclamation activities; possibly leading to developmental toxicity, in addi tion to endocrine disruption. Specific Aims The overall objective of our study was to determine the causes of decreased hatch rates among alligators in contaminated sites, and to determine if causal links could be established between specific adverse eff ects and exposure to individual OCPs or combinations of OCPs. The project consiste d of epidemiological field studies, which evaluated embryonic development and mort ality as a function of maternal and environmental exposure to OCPs and egg nutri ent composition; and controlled laboratory experiments to test hypothesized links be tween decreased hatch rates, altered egg composition, and exposure to selected OCPs. Specific aim 1 : Conduct field epidemiological studi es to determine the relative contributions of unbanded eggs, embryonic mo rtality in banded eggs, and decreased perinatal mortality to the overall decreased reproductive success in alligators at OCPcontaminated sites, to determine which OCPs or combinations of OCPs are most closely associated with adverse effect s at each life stage, and to ex amine the relationship between OCP burdens in maternal tissues and eggs For Specific Aim 1, the hypotheses were H1a : Adverse effects at early life stages are associ ated with muck farm environments, exposure to specific OCPs or OCP combinations, or both; H1b : Specific OCPs found in maternal tissues are highly correlate d to those present in eggs indicating maternal transfer of OCPs and that maternal size is correlated with OCP burdens and hatch rates; H1c : Eggs in which embryonic and perina tal mortality occur result from developmental abnormalities, altered stru cture or composition of the egg, or both. Specific aim 2 : Conduct controlled in ovo and in vivo experiments with alligators to confirm causal links between decreased ha tch rates and affected life stages as a

PAGE 37

21 function of exposure to selected OCPs or al tered egg qualities, or both. For Specific Aim 2, the hypotheses were H2a : Exposing a captive breeding populat ion of adult alligators to an environmentally relevant mixture of OCPs w ill elicit OCP concentrations in eggs and developmental effects similar to those obser ved in wild eggs from OCP-contaminated field sites; H2b : Exogenous in ovo alteration of egg nutrien ts based on data from field studies will alter embryonic development.

PAGE 38

22 Figure 1-1. Map of Ocklawaha Basin. Ocklawaha River Gourd Neck Area Lake Apopka Lake Griffin LakeLochloosa Orange Lake Lake George

PAGE 39

23 CHAPTER 2 EGG AND EMBRYO QUALITY OF ALLI GATORS FROM REFERENCE AND ORGANOCHLORINE CONTAMINTED HABITATS In the southeastern US, aquatic ecosystem s have experienced habitat degradation, alterations in water quality, and in some cases important declines in biodiversity due to increases in land development and associated anthropogenic im pacts. A case-in-point is the Ocklawaha River Basin in central Florida. Within this basin, American alligators ( Alligator mississippiensis ) from impacted lakes have e xhibited poor clutch viability (number eggs that yield a live hatchling / to tal number of eggs found in clutch) (Masson, 1995), abnormal reproductive hormone concen trations (Gross et al., 1994), and unexplained adult mortality (Schoeb et al., 2002). During the mid 1980s, clutches from alligators on Lake Apopka experienced severe declines in clutch viability (declined from 50% to 4%), and alligator clutches from other impacted lakes had only moderate viabilities (range of 40 to 60 %). These rates were below those observed in other less impacted Florida lakes (reference sites), including Lake Woodruff National Wildlife Refuge (79%), Orange Lake (82%), and th e Everglades Water Conservation Areas (6575%) (Woodward et al., 1993; Masson, 1995; Rice, 1996). Possible causal factors for reduced hatch ra tes in alligator popul ations within the impacted sites within the Ocklawaha Rive r Basin include pesticides, algal toxins, nutritional changes, density-related stress, and diseases. In one case, a chemical spill from a chemical manufacturing plant in 1980 near Lake Apopka (EPA, 2004) was temporally associated with the decline in reproductive success and consequent alligator

PAGE 40

24 population decline on Lake Apopka during th e early 1980s. However, decreases in clutch viability for Lake Apopka appeared to be more related to proximity to muck farm restoration areas as compared to yolk con centrations (Giroux, 1998), which is consistent with decreases in clutch viability on Lake Griffin and Emeralda Marsh, GriffinÂ’s adjacent muck farm restoration area (S eplveda et al., 2001). Poor reproductive success threatens the long-term conservation of alligators, potentially altering the ecology of affected ecosystems, a nd substantially reducing the aesthetic and economic values of alligators in affected areas. Understanding and characterizing poor reproductive performance a nd determining associated factors is needed so that efficacious mitigation strategies may be developed. Thus, the overall objective of the present study was to determin e the relative contribu tions of losses during in ovo development in American alligators at imp acted sites in central Florida, and to evaluate whether organochlorine pesticid es (OCPs) are associated with adverse developmental effects and altered clutch characteristics. Materials and Methods Egg Collections and Incubation Lakes Apopka (N 28 35Â’, W 81 39Â’), Griffin (N 28 53Â’, W 81 46Â’), Emeralda Marsh Conservation Area ((N 28 55Â’, W 81 47Â’), and Lochloosa (N 29 30Â’, W 82 09Â’) in Florida were selected as collection sites because prior studies indicate vastly different levels of OCP exposure across th ese sites (Gross unpublis hed data, (Masson, 1995). Alligator nests were located via aerial (helicopter) and ground surveys (airboat), and clutches were subsequently collected by ground crews. The top of each egg was marked before eggs were removed from the nest to ensure prope r orientation; thus,

PAGE 41

25 preventing embryo mortality due to inversion. Embryo mortalit y due to inversion occurs if an embryo has attached to the top of the egg, inversion may either break embryonic attachment or cause the yolk mass to sett le on top of the embryo, crushing it. After marking each egg and placing about 5 cm of nest substrate in a uniquely numbered polypropylene pan (43 cm x 33 cm x 18 cm), all eggs found in each clutch were placed in the pan 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 additi onal eggs were set. The top layer of eggs was covered with nest substrate so that there was no space left between the top of the pan and the top of the eggs (approximat ely 10 cm). Clutches were transported to the US Geological SurveyÂ’s Center for Aquatic Resources Studies, Gainesville, Florida (CARS). Upon arrival, clutches were evaluated for embryonic viability using a bright light candling procedure. Viable eggs (i.e. having a visible band) were nested in pans containing moist spha gnum moss and incubated at 30.5C and ~98% humidity, in an incubation building (7.3 m x 3.7 m). This intermediate incubation temperature will normally result in a 1:1 ma le/female sex ratio, since alligators have temperature dependent sex (or gender) diffe rentiation. One or two eggs were opened from each clutch to identify the embryonic st age of development at the time of collection, and to collect yolk samples for later measur ement of OCP burdens. From each clutch, information on the following parameters was collected: total number of eggs found per nest (fecundity); number of unbanded eggs, num ber of damaged eggs, number of dead banded eggs, number of live banded eggs, to tal clutch mass, and average egg mass of clutch.

PAGE 42

26 For years 2001 and 2002, some clutches were involved in an embryo development study. For these clutches, each clutch was even ly divided between two pans, with half of the clutch left relatively undi sturbed (except for weekly mo nitoring of embryo mortality) to determine clutch viability (the number of live hatchlings / fecundity), and the other half of the clutch used to study embryo de velopment and morphometry (Chapter 5). Analysis of OCPs in Yolk Analytical grade standards for the following compounds were purchased from the sources indicated: aldrin, al pha-benzene hexachloride ( -BHC), -BHC, lindane, -BHC, p,pÂ’ -dichlorodiphenyldichloroethane ( p,pÂ’ -DDD), p,pÂ’ -dichlorodiphenyldichloroethylene ( p,pÂ’ -DDE), dichlorodiphe nyltrichloroethane ( p,pÂ’ -DDT), dieldrin, endosulfan, endosulfan II, endosulfan sulfate, endrin, e ndrin aldehyde, endrin ketone, heptachlor, heptachlor epoxide, hexachlorobenz ene, kepone, methoxychlor, mirex, cis -nonachlor, and trans -nonachlor from Ultra Scientific (Kingstown, RI, USA); cis -chlordane, trans chlordane, and the 525, 525.1 polychlorinated biphenyl (PCB) Mix from Supelco (Bellefonte, PA, USA); oxychlordane from Chem Service (West Chester, PA); o,pÂ’DDD, o,pÂ’DDE, o,pÂ’DDT from Accustandard (New Haven, CT, USA); and toxaphene from Restek (Bellefonte, PA, USA). All reag ents were analytical grade unless otherwise indicated. Water was doubly distilled and deionized. Egg yolk samples were analyzed for OC P content using methods modified from Holstege et al. (1994) and Sc henck et al. (1994). For extraction, a 2 g tissue sample was homogenized with ~1 g of sodium sulfate a nd 8 mL of ethyl acetate. The supernatant was decanted and filtered t hough a Bchner funnel lined with Whatman #4 filter paper (Fisher Scientific, Hampton, NH, USA ) and filled to a depth of 1.25 cm with sodium sulfate. The homogenate was extracted twice with the filtrates collected together. The

PAGE 43

27 combined filtrate was concentrated to ~2 mL by rotary evaporation, and then further concentrated until solvent-free under a stre am of dry nitrogen. The residue was reconstituted in 2 mL of acetonitrile. Afte r vortexing (30 s), the supernatant was applied to a C18 solid phase extraction (SPE) car tridge (pre-conditio ned with 3 mL of acetonitrile; Agilent Technologies, Wilmingt on, DE, USA) and was allowed to pass under gravity. This procedure was repeated twice with the comb ined eluent collected in a culture tube. After the last addition, the car tridge was rinsed with 1 mL of acetonitrile which was also collected. The eluent was then applied to a 0.5 g NH2 SPE cartridge (Varian, Harbor City, CA, USA), was allowe d to pass under gravity, and collected in a graduated conical tube. The cartridge was rinsed with an additional 1 mL portion of acetonitrile which was also collected. The combined eluents were concentrated under a stream of dry nitrogen, to a volume of 300 L, and transferred to a gas chromatography (GC) vial for analysis. GC/MS Analysis Analysis of all samples was performed using a Hewlett Packard HP-6890 gas chromatograph (Wilmington, DE, USA) with a split/splitless inlet ope rated in splitless mode. The analytes were introduced in a 1 L injection and separa ted across the HP-5MS column (30 m x 0.25 mm; 0.25 m film thickne ss; J & W Scientific, Folsom, CA, USA) under a temperature program that began at 60 C, increased at 10 C/min to 270 C, was held for 5 min, then increased at 25 C/min to 300 C and was held for 5 min. Detection utilized an HP 5973 mass spectro meter in electron impact m ode. Identification for all analytes and quantitation for toxaphene was c onducted in full scan mode, where all ions are monitored. To improve sensitivity, se lected ion monitoring was used for the

PAGE 44

28 quantitation for all other analytes, except kepone. The above program was used as a screening tool for kepone which does not optim ally extract with mo st organochlorines. Samples found to contain kepone would be reex tracted and analyzed specifically for this compound. For quantitation, a five-point standard curve was prepared for each analyte ( r2 0.995). Fresh curves were analyzed with each se t of twenty samples. Each standard and sample was fortified to contain a deuterat ed internal standard, 5 L of US-108 (120 g/mL; Ultra Scientific), added just prior to analysis. All samples also contained a surrogate, 2 g/mL of tetrach loroxylene (Ultra Scientific) added after homogenization. Duplicate quality control samples were prepar ed and analyzed with every twenty samples (typically at a level of 1.00 or 2.50 g/mL of -BHC, heptachlor, aldr in, dieldrin, endrin, and p,p’ -DDT) with an acceptable recovery rangi ng from 70 – 130%. Limit of detection ranged from 0.1-1.5 ng/g for all OCP analyt es, except toxaphene (120-236 ng/g), and limit of quantitation was 1.5 ng/g for all anal ytes, except toxaphene (1500 ng/g). Repeated analyses were conducted as allo wed by matrix interferences and sample availability. Data Analysis Specific OCP analytes were removed from analysis if measurable concentrations were found in < 5% of all clutches. Numerical data were log-transf ormed [ln(x)], while proportional data were arcsine s quare root transformed to meet statistical assumptions. ANOVA (PROC GLM; SAS Institute Inc., 2002) was used for inter-site comparisons of adult female and clutch char acteristics, and the Tukey test was used for multiple comparisons among sites ( = 0.05). Because relationships between response variables and explanatory vari ables (Table 2-1) in ecologi cal studies are often complex

PAGE 45

29 with interactions occurring, an indirect gr adient multivariate analysis method, Detrended Correspondence Analysis (DCA) (ter Braak, 1986) was used to initially evaluate data structure. Two matrices were construc ted for DCA, with the first representing the response variables (clutch number x clutch parameters) and the second representing the explanatory variables (clutch number x OC P burdens) (Table 2-2). DCA results indicated that a direct gradient, multivariat e linear analysis, re dundancy analysis (RDA) (Rao, 1964), was appropriate since the gradient lengths of the DCA ordination axes were equal to or less than 2 standard deviations (ter Braak, 1995). RDA is the canonical form of principal compon ents analysis (PCA). In RDA, as in PCA, a straight line is fitted to each the re sponse variable (clutch su rvival parameters) in an attempt to explain the data of all response variables. Similar to PCA, the lower the residual sum of squares, the better the e nvironmental variable is at explaining the variation in response variables. RDA, unlike PCA, restricts the clutch scores (from the response variables measured on each clutch) to a linear combination of the environmental (explanatory variables). Because clutch scores are constrained to a linear combination of environmental variables, RDA explains slig htly less variance compared to PCA (ter Braak & Tongeren, 1995; ter Braak, 1994). Fo r RDA involving compositional data (i.e., clutch viability rates or percentages) and quantitative environmental variables, compositional data is log-transformed (ln (x + 1)) with correlation biplots being centered by the response variables (i.e., unbanded e gg percentage) and by the samples (i.e., clutches) (ter Braak, 1994). These corre lation biplots provide a way to examine relationships among a number of response variables and explanatory factors with response variable arrows forming a biplot of correlations with each other, environmental

PAGE 46

30 arrows forming a biplot among each othe r, and response variable arrows and environmental arrows forming a biplot of correlations with each other (ter Braak, 1995). For the RDA, separate matrices were cons tructed for response variables measured as a percentage (i.e., clutch viability) and re sponse variables measured as a number (i.e., clutch mass) because percentage data were ln(x+1) transformed and not standardized, while continuous data were ln (x) transformed and standardized(ter Braak & Smilauer, 2002). Automatic forward selection of the be st four explanatory variables was conducted for both sets of RDA analyses and tested for significance by Monte Carlo permutation tests. DCA and RDA were conducted us ing the program CA NOCO (ter Braak & Smilauer, 2002). Biplots of environmental variables and response variables were then constructed to interpret relationship between clutch parameters (response variables) and explanatory factors. Results Inter-Site Comparisons of Clutch Characteristics From 2000-2002, 168 clutches were collected from Lakes Lochloosa ( n = 44), Apopka ( n = 31), Griffin ( n = 47), and Emeralda Marsh ( n = 46). No significant differences were determined among sites with respect to clutch mass (overall mean standard error: 3.7 0.08 kg), egg mass (83 1.4 g), or percentage of unbanded eggs (15 1.7%) (Table 2-3). In contrast, significant diffe rences were determined am ong sites with respect to fecundity, clutch viability, pe rcentage of damaged eggs, percentage of early embryo mortality, and percentage of la te embryo mortality. Clutches from Lochloosa had lower fecundity and late embryo mortality rates co mpared to all other sites. In addition, Lochloosa clutches had greater clutch viability rates than all other sites and lower early

PAGE 47

31 embryo mortality rates than all other sites, except for Apopka. Clutches from Emeralda Marsh had greater incidence of damaged eggs than all other sites, except for those of Lake Griffin (Table 2-3). Organochlorine Pesticides Burdens and Clutch Characteristics From 2000-2002, clutch characteristics a nd OCP burdens were measured on 115 clutches collected from Lakes Lochloosa ( n = 19), Apopka ( n = 23), Griffin ( n = 42), and Emeralda Marsh ( n = 31). No significant differences were determined among sites with respect to clutch mass (overall mean standa rd error: 3.8 0.09), clutch viability (50 3.1), percentage of damaged eggs (4 1) unbanded eggs (13 1.6), early embryo mortality (21 2.3), and late embryo mort ality (11 1.7) (Table 2-4). However, significant differences were determined fo r fecundity and egg mass, with Lochloosa clutches having lower fecundity than all other sites, and greater average egg mass compared to those of all other sites, excep t for Lake Apopka. Furthermore, significant differences were detected among sites with re spect to individual OCP concentrations in egg yolks, total OCP concentrations in e gg yolks, and number of OCPs detected at measurable levels. For total OCP concentra tions and number of analytes detected at measurable levels, egg yolks of Lake Loch loosa clutches had si gnificantly lower total concentrations and a lower number of analytes de tected at measurable levels (Table 2-4). Individual OCP analyte concentrations in egg yolks of Lochloosa clutches were significantly less than those of the other sites, except for Lake Griffin with respect to aldrin and trans -nonachlor. Aldrin and trans -nonachlor egg yolk concentrations of Lochloosa clutches did not significantly diffe r from Lake Griffin, but egg burdens of these analytes of both sites were signifi cantly less than those of Lake Apopka and Emeralda Marsh (Table 2-4).

PAGE 48

32 Clutch Survival and OCP Burdens in Egg Yolks Because a number of site specific factor s may potentially affect clutch survival parameters and since OCP burdens varied greatly among sites, relationships between OCP egg yolk variables and clutch survival we re evaluated on a site-by-site basis. For Lake Lochloosa, redundancy analysis with Monte Carlo permutation tests for significance indicated that none of the four extracted OCP variables (Table 2-5) were found to be significantly correlate d with the clutch survival va riables. Number of OCPs detected (NOC) approached significance ( P = 0.07), was negatively associated with clutch viability, positively correlated with percentage unbanded eggs and late embryo mortality, and accounted for 11% of the variation in clutch surviv al parameters (Fig. 2-1). For Lake Griffin, redundancy analysis w ith Monte Carlo permutation tests for significance indicated th at three of the four extracted OCP variables were found to be significantly correlated w ith the clutch survival variable s and together accounted for 21% of the variance in clutch survival parame ters. The extracted OCP variables were concentration of p,pÂ’ -DDE, toxaphene, and p,pÂ’ -DDT, accounting for 8, 7, and 6%, respectively, of variation in clutch survival variables (Table 2-5). Clutch viability was positively associated with toxaphene and p,pÂ’ -DDE egg yolk concentrations, but had little to no correlation with p,pÂ’ -DDT yolk burdens. Early embryo mortality rates were negatively associated with p,pÂ’ -DDE and toxaphene. Late embryo mortality rates were positively associated with toxaphene, and negatively associated with p,pÂ’ -DDT, and p,pÂ’ DDE. Unbanded egg rates were positively associated with p,pÂ’ -DDT and p,pÂ’ -DDE, but negatively associated with toxaphene (Fig. 2-2). For Lake Apopka, redundancy analysis w ith Monte Carlo permutation tests for significance also indicated that three of the four extracte d OCP variables were found to

PAGE 49

33 be significantly correlated with the clutch survival variables. These OCP variables were percentage dieldrin (lambda A = 17%), percen tage trans-chlordane (1 2%), and percentage aldrin (10%), and together accounted for 3% ( lambda AÂ’s) of the variance in the clutch survival parameters (Table 2-5). Clutch vi ability was positively associated with aldrin, weakly associated with trans -chlordane, and negatively associ ated with dieldrin. Early embryo mortality and unbanded egg rates were positively associated with dieldrin and trans -chlordane and negatively associated with aldrin. Late embryo mortality rates were negatively with all three OC P variables (Fig. 2-3). For Emeralda Marsh, redundancy analysis with Monte Carlo permutation tests for significance also indicated that only percentage toxaphene was found to be significantly correlated with the clutch survival variables, and it accounted for 9% of the variance in the clutch survival parameters (Table 2-5). Percentage toxaphene was positively associated with clutch viability, weakly associated with late embryo mortality, and negatively associated with early embryo mo rtality and unbanded egg rates (Fig. 2-4). Percentage of heptachlor epoxide show ed a near signifi cant association ( P = 0.09) with clutch parameters, being negatively correla ted with clutch viability and positively correlated with early and late embryo mortality rates. Average Egg Mass, Clutch Size and OCP Burdens For Lochloosa clutches, three of four OC P variables were determined (via RDA analysis) to be significantly a ssociated with egg and clutch size parameters and accounted for 64% of the variation in egg and clut ch size parameters. These OCP variables included number of OCPs detected at m easurable levels (NOC) (lambda A = 31%), p,pÂ’ DDT concentrations (20%), and trans -nonachlor concentrations (13%) (Table 2-6).

PAGE 50

34 NOC and trans -nonachlor concentrations were negatively associated with average egg mass but positively associated with f ecundity and clutch mass. In contrast, p,pÂ’ -DDT concentrations were positively associated with egg mass, negativ ely associated with fecundity, and had little to no associ ation with clutch mass (Fig. 2-5). For Lake Griffin clutches, however, no signi ficant associations were found between OCP variables and egg and clutch size variables. In contrast, percentage o,pÂ’ -DDT in Emeralda Marsh clutches was found to be pos itively associated with increasing egg and clutch mass but negatively associated with f ecundity. Lastly, Lake Apopka clutches were somewhat similar to Emeralda clutches in that one OCP variable ( p,pÂ’ -DDD concentration) was found to be positively associated with egg and clutch mass and negatively associated with fecundity (Table 2-6). Discussion Inter-Site Comparisons of Clutch Characteristics The results of the present study suggested that the relative cont ributions of losses during in ovo development in alligators at impacted sites in Florida are lower clutch viability, higher rates of damaged eggs, higher rates of early embryo mortality, and higher rates of late embryo mortality. A lthough not significantly different among sites, infertility and/or embryo mo rtality before embryo attachment (unbanded eggs) also appears to be a major constituent of reduced cl utch viability among all sites. In order of importance, major constituents of reduced cl utch viability for all sites include early embryo mortality, unbanded eggs, late embryo mortality, and damaged eggs. In addition, clutches from OCP-contaminated sites had an average of 10 more eggs per clutch as compared to the reference site, but averag e clutch mass was not significantly different,

PAGE 51

35 making average egg mass of reference site clut ches greater than that of clutches of OCPcontaminated sites. The reduced clutch viability, increased rates of unbanded eggs and embryo mortality, and concurrent increa se in fecundity without propo rtional increase in clutch mass observed in clutches from OCP-contaminated sites, as compared to the reference site (Lochloosa), suggest that females and th eir embryos from contaminated sites may be responding to one or more environmental fact ors common to the three OCP-contaminated sites. Although measurement of all envir onmental factors is imp ractical, the large differences in OCP concentrations in al ligator eggs between reference and OCPcontaminated sites were found. Specifically, total OCP egg yolk burdens and number of OCPs detected at measurable levels in Lake Lochloosa were significantly less than those of Lake Griffin clutches, and OCP burdens in Lake Griffin clutches were, in turn, significantly less than those of Lake Apopka and Emeralda Marsh. Although Lake Apopka and Emeralda Ma rsh were not determined to be significantly different with respect to total OCP concentrations in egg yolks, significant differences were determined between these two high OCP exposure sites in regard to certain analytes, as well as the total number of OCPs detected at measurable levels. Clutches from Emeralda Marsh had a greater number of OCP analytes in their egg yolks and contained higher concentrations of cis-chlordane, p,pÂ’-DDD, o,p-DDD, transchlordane, and toxaphene compared to thos e from Lake Apopka. Conversely, clutches from Lake Apopka had higher concentrations of aldrin, dieldrin, heptachlor epoxide, and oxychlordane compared to those of Emeralda Marsh.

PAGE 52

36 The differences in OCP exposure profiles am ong sites likely refl ect the differences in historic land-use and OCP applications, as opposed to differences in xenobiotic biotransformation among the different allig ator populations inha biting the respective sites. Importantly, although Emeralda Marsh is separate d from Lake Griffin by only a levee easily traversed by alligators, large differences in OCP egg burdens were noted between the two sites. Such differences in exposure suggest that the highly exposed adult females which oviposite within Emeralda Mars h likely have established territories and reside year round within Emeralda Marsh (a fo rmer agricultural property). Furthermore, the relatively high egg burdens in clutches of Emeralda Marsh likely occurred over a relatively short period since this 2,630 ha area was not flooded until the early 1990s (Marburger et al., 1999). In summary, the differences in OCP egg bur dens between the reference site and the contaminated sites support th e hypothesis that OCP contaminants may be associated with reduced clutch viability, given that OCPs have been causally linked to reduced reproductive success in other oviparous species (Donaldson & Fites, 1970; Fry, 1995). Clutch Survival Parameters and OCP Burdens Results of redundancy analyses more dire ctly addressed the question of whether OCPs are associated with reduced clutch viab ility by relationships on a site-by-site basis to control for potential site -associated confounding factors. For Lake Lochloosa, no significant correlations were determined alt hough significance might have been detected given a greater sample size. The positive bu t insignificant correlations between increases in unbanded egg and late embryo mortality perc entage and number of OCPs may suggest that increased OCP burdens in eggs play a role in clutch viability or it may simply indicate that older females ha ve increased levels of OCPs due to increased exposure time

PAGE 53

37 and that decreased clutch viability is due to decreased egg quality associated with senescence. For Emeralda Marsh, the weak associati ons between OCP variables and clutch survival variables suggests that other factor s may be involved in reduced embryo survival and increased rates of unbanded eggs. The weak associations for Emeralda Marsh are surprising given that relatively stronger associations were de termined for the other high exposure site (Lake Apopka; Table 2-5), as we ll as the intermediate exposure site (Lake Griffin, Table 2-5), with Emeralda Marsh be ing separated from La ke Griffin by only a non-fenced levee. Stronger associations were noted for La ke Apopka in contrast to the weak, associations noted for Emeralda Marsh. Th e positive association between early embryo mortality and unbanded egg rates and extracte d OCP variables for Lake Apopka clutches suggests that the percentages of dieldrin and trans -chlordane in eggs may play an important role in altered egg fertility and/or early embryo survival Interestingly, the percentage of aldrin, (dieldrinÂ’s parent co mpound) had a negative association with late embryo mortality, a positive asso ciation with clutch viabilit y, and near-zero correlations with percentage unbanded eggs and early em bryo mortality. However, dieldrin (a metabolite formed from aldrin) had strong, positive correlations with percentage unbanded eggs and early embryo mortality, a nd a negative correla tion with clutch viability, suggesting this metabolite has gr eater efficacy than its parent compound in affecting embryo survival. Th e potential consequence exists that increasing a female alligatorÂ’s ability to biotransform aldrin to dieldrin may increase th e risk of early embryo mortality. Another important not e is that the level of diel drin in Apopka clutches was

PAGE 54

38 two-fold greater than those of Emeral da Marsh, suggesting that OCP mixture composition may be more important than sum OCP concentrations. For Lake Griffin, the negative to nea r-zero association between early embryo mortality rates and extracted OCP variables s uggests that OCP burdens in eggs may not play an important role in early embryo mort ality. However, the positive association between toxaphene burdens and late embryo mo rtality suggests that as toxaphene burdens increase, so does the risk for increased embryo death during the last 35 days of development. Furthermore, th e positive association between p,pÂ’ -DDT concentrations and unbanded egg rates suggests that these an alytes may be involved in altered egg fertility and/or embryo survival (prior to eggshell membrane attachment) (Fig. 2-2). Egg and Clutch Size and OCP Burdens For Lochloosa, the strong associations between OCP burdens and egg and clutch size parameters suggest that, although a low OCP exposure site, certain patterns of OCP exposure are strongly associated with egg a nd clutch size characteristics. The positive associations p,pÂ’ -DDT concentrations have with clut ch weight and average egg weight and p,pÂ’ -DDTÂ’s negative association with fec undity may be potentially related to senescent females, since older females have been reported to lay smaller clutches of larger eggs (Ferguson, 1985) and would likely have higher OCP burdens due to extended exposure period. In contrast, the positive associations that NOC and trans-nonachlor have with fecundity and clutch mass, and th e negative associations these OCP variables have with egg mass, suggests that increased OCP exposure may have altered clutch and egg size characteristics, as opposed to fema le age. Although these speculations are interesting from a low exposure effect sta ndpoint, they are irrelevant at the populationeffect level since clutch viability rates were unrelated.

PAGE 55

39 Since the low exposure site had stronger associations between OCP variables and egg and clutch size variables than intermed iate and high exposure sites, one might initially conclude that other factors are more important than OCP burdens in influencing egg and clutch size characteristics. While this may be the case, the fact that the intermediate and high exposure sites have significantly grea ter fecundity (averaging 10 more eggs per clutch compared to the low exposure site), signifi cantly less average egg mass, and similar clutch mass suggest that females attaining their maximum physiological response in regard to number of eggs ovulate d. These intermediate and highly exposed females appear to be produci ng more ova but are unable to sequester additional egg components (i.e., lighter eggs), in effect decreasi ng the amount of energy and structural supplies available to each embr yo and resulting in light er eggs and higher embryo mortality rates. In summary, our results suggest that, ove r all sampled clutches, clutch survival parameters and egg and clutch size paramete rs vary between the low OCP exposure site (Lochloosa) and the intermediate-high OCP e xposure sites. Furthermore, OCP burdens do not appear to be related to clutch survival for the low exposure site but are associated with clutch survival for the intermediate-hi gh OCP contaminated site s. In contrast, egg and clutch size parameters appear to be a se nsitive endpoint in OCP response in alligators due to the strong associations noted between OCP and clutch size variables for the low exposure site and the lack thereof for the intermediate-high expos ure sites, suggesting attainment of maximum response. In order to better determ ine the role of OCPs in the reduced reproductive efficiency of OCP-exposed alligator populations, suggested future studies should examine the relationship betw een maternal OCP burdens and respective

PAGE 56

40 egg burdens, presence of other environmental contaminants, maternal factors associated with clutch survival and OCP burdens, a nd how egg composition relates to clutch survival and OCP burdens.

PAGE 57

41 Table 2-1. Reproductive, morphometric, a nd contaminant parameters measured on clutches of alligator eggs collected during summer 2000, 2001, and 2002. Clutch Parameter Definition Measured as Response variables Fecundity Total No. of eggs in one clutch n Clutch mass Total mass of eggs in one clutch kg Ave. Egg Weight Clutch mass / Fecundity g Unbanded eggs% a No. of unbanded eggs / fecundity x 100 Percentage Early embryo mortality% No. of deaths < dev. Day 35 / fecundity x 100 Percentage Late embryo mortality% No. of deaths dev. Day 35 / fecundity x 100 Percentage Clutch Viability No. eggs yielding live hatchling / fecundity x 100 Percentage Explanatory variables [OCP analyte] in egg yolk ng OCP analyte / g egg yolk wet weight ppb % OCP analyte [OCP analyte] / [OCP] x 100 Percentage aAn egg with no evidence of embryonic attachment

PAGE 58

42 Table 2-2. Explanatory variab les included in RDA with forw ard selection of four best variables ( = 0.05). Variable Code Age Age No. OCPs at measurable levels NOC [OCP] TOC % Aldrin ALD% [Aldrin] [ALD] % cis -Chlordane CC% [ cis -Chlordane] [CC] % cis -Nonachlor CN% [ cis -Nonachlor] [CN] % Dieldrin DL% [Dieldrin] [DL] % Heptochlor epoxide HE% [Heptachlor epoxide] [HE] %Lipid content LPC% % Mirex MX% [Mirex] [MX] % o,p -DDT ODDT% [ o,p -DDT] [ODDT] % o,p -DDD ODDD% [ o,p -DDD] [ODDD] % Oxychlordane OX% [Oxychlordane] [OX] % p,p '-DDE PDDE% [ p,p '-DDE] [PDDE] % p,p '-DDD PDDD% [ p,p '-DDD] [PDDD] % p,p '-DDT PDDT% [ p,p '-DDT] [PDDT] % trans -Chlordane TC% trans -Chlordane [TC] % trans -Nonachlor TN% [ trans -Nonachlor] [TN] % Toxaphene TX% [Toxaphene] [TX]

PAGE 59

43Table 2-3. Summary of clutch parameters and site comparisons for clutches of American alligator eggs collected during 2000-200 2. Parametera Lochloosa Apopka Emeralda Griffin Summary No. Clutches 44 31 46 47 168 Fecundity ( n ) 36 1.2 B 46 1.3 A 46 1.1 A 45 1.2 A 43 0.7 (22–56) (28–56) (27–64) (19–58) (19–64) Clutch mass (kg) 3.4 0.15 4 0.13 3.8 0.21 3.6 0.13 3.7 0.08 (1.6–4.8) (2.4–5.1) (1.9–9.2) (1.5–5.2) (1.5–9.2) Egg mass (g) 87 2.2 86 2 83 4 80 1.6 83 1.4 (61–139) (62–120) (58–180) (46–113) (46–180) Clutch viability (%) 70 3.9 A 51 5.8 B 48 5.5 B 44 4.9 B 53 2.6 (0–100) (0–98) (0–97) (0–92) (0–100) Damaged eggs (%) 2 1.4 B 2 0.6 B 5 1.3 A 4 1.8 AB 3 0.7 (0–60) (0–16) (0–33) (0–63) (0–63) Unbanded eggs (%) 11 2.2 21 4.9 14 3.7 17 3.2 15 1.7 (0–84) (0–100) (0–100) (0–100) (0–100) Early Emb. Mort. (%) 12 2.7 B 15 4.2 AB 23 3.9 A 22 3.9 A 19 2 (0–69) (0–94) (0–95) (0–100) (0–100) Late Emb. Mort. (%) 6 1.7 B 12 3.5 A 10 2.4 A 13 3.1 A 11 1.4 (0–34) (0–77) (0–82) (0–89) (0–89) aValues indicate mean standard error of m ean with ranges in parentheses. Values w ith different letters (A-B) indicate signifi cant differences ( = 0.05); same letters indicate significant differences were not detected. Clutch viability = No. of eggs yielding a live hatchling / Fecundity x 100, Damaged eggs = No. damaged eggs / fecundity x 100, Unbanded eggs = No. of unbanded eggs / fecundity x 100, Early Emb. Mort. = No. of em bryonic deaths on or before developmenta l Day 35 / fecundity x 100, and Late Emb. Mort. = No. of embryonic deaths post dev. Day 35 / fecundity x 100).

PAGE 60

44Table 2-4. Organochlorine pesticide burdens and clutch parameters and site comparisons for clutches of American alligator eggs collected during 2000-2002. Parametera Loch. Apopka Emeralda Griffin Summary No. Clutches 19 23 31 42 115 Fecundity ( n ) 40 1.7 B 47 1.4 A 46 1.3 A 46 1.2 A 45 0.7 A (26–56) (31–56) (27–64) (24–58) (24–64) Clutch mass (kg) 3.6 0.17 4 0.16 3.8 0.25 3.7 0.13 3.8 0.09 (2.2–4.8) (2.5–5.1) (2.1–9.2) (1.5–5.2) (1.5–9.2) Egg mass (g) 90 2.9 A 86 2.5 AB 82 4.8 B 79 1.5 B 83 1.6 (78–139) (62–120) (58–180) (46–105) (46–180) Clutch viability (%) 65 5.5 52 6.4 50 6.9 43 5.1 50 3.1 (0–95) (0–98) (0–97) (0–92) (0–98) Damaged eggs (%) 4 3.1 2 0.8 6 1.6 5 2 4 1 (0–60) (0–16) (0–32) (0–63) (0–63) Unbanded eggs (%) 11 2 17 4.2 10 2.3 15 3.2 13 1.6 (0–33) (0–81) (0–58) (0–100) (0–100) Early Emb. Mort. (%) 13 3 15 4.2 26 5.1 24 4.3 21 2.3 (0–36) (0–90) (0–95) (0–100) (0–100) Late Emb. Mort. (%) 8 2.5 14 4.6 10 2.5 13 3.3 11 1.7 (0–34) (0–77) (0–61) (0–89) (0–89) Aldrin (ng/g) 0 0 C 4 0.3 A 2 0.3 B 0 0 C 3 0.3 (0–0) (2.9–5.2) (1.5–4.3) (0–0) (1.5–5.2) Methoxychlor (ng/g) 0 0 C 8 1 B 9 1 B 17 0.3 A 9 0.8 (0–0) (5.7–16.4) (5.8–18.4) (16.9–17.5) (5.7–18.4)

PAGE 61

45Table. 2-4. Continued. Parametera Orange/Loch Apopka Emer alda Griffin Summary Mirex (ng/g) 2 0.4 B 6 1.1 A 3 0.5 AB 3 0.2 AB 4 0.4 (1.2–2.7) (1.1–17.2) (0.1–10.3) (1.1–4.5) (0.1–17.2) Dieldrin (ng/g) 4 0.5 D 344 80.9 A 142 20.4 B 23 3.8 C 118 20.9 (1.3–8.2) (12.5–1783.2) (8.7–386.7) (2.9–124) (1.3–1783.2) Hep. Epoxide (ng/g) 3 0.8 C 17 5.6 A 7 1.4 B 7 1 B 8 1.4 (1.2–9.7) (1.2–135.5) (0.1–32.1) (1.1–29.6) (0.1–135.5) cis-Chlordane (ng/g) 2 0.2 D 43 7.6 B 90 13 A 11 0.9 C 37 5 (1.2–4.1) (6.6–179.2) (8.9–281) (4.3–31.8) (1.2–281) cis-Nonachlor (ng/g) 5 0.6 C 88 27.3 A 66 9.7 A 18 1.6 B 43 6.7 (2.4–12.5) (10.5–656.2) (11.6–232.2) (6.5–54.2) (2.4–656.2) Oxychlordane (ng/g) 4 1 D 51 14.6 A 23 3.8 B 10 1.3 C 21 3.4 (1.2–17.8) (3.9–353.8) (3.2–109.5) (1.1–41.9) (1.1–353.8) p,p' -DDE (ng/g) 74 11.7 C 5794 1794.7 A 8069 1402 A 271 31.3 B 3445 610.6 (28–231) (18.3–42653.4) (36.2–33554.8) (62.9–979.1) (18.3–42653.4) p,p'-DDD (ng/g) 2 0.2 D 42 8.5 B 1289 196.1 A 7 0.9 C 382 78.7 (1.2–3) (10.6–192.8) (10.3–2962.8) (2.7–28.9) (1.2–2962.8) p,p'-DDT (ng/g) 1 0 C 9 2.1 AB 12 1.2 A 5 0.8 B 10 1 (1.2–1.3) (1.2–45.6) (5.8–25.5) (1.1–7.2) (1.1–45.6) o,p'-DDD (ng/g) 0 0 C 5 0.7 B 37 5.1 A 1 0 B 29 4.5 (0–0) (3.1–9.2) (0.1–104) (1.3–1.3) (0.1–104) o,p'-DDT (ng/g) 1 0 C 11 1.9 A 170 161.6 A 4 0.3 B 48 42 (1.2–1.4) (1.2–38.5) (4.2–4372.8) (1.1–7.4) (1.1–4372.8)

PAGE 62

46Table. 2-4. Continued. Parametera Orange/Loch Apopka Emer alda Griffin Summary trans-Chlordane (ng/g) 3 0.7 C 8 1.5 B 25 3.3 A 2 0.2 C 11 1.5 (1.2–3.7) (1.3–27.4) (2.9–58.2) (1.1–8.7) (1.1–58.2) Toxaphene (ng/g) 0 0 C 2738 224.5 B 6865 552.4 A 3043 425.9 B 5456 483 (0–0) (1896.1–3809.1) (2300.6–12975.4) (1927.9–4533.2) (1896.1–12975.4) trans-Nonachlor (ng/g) 8 1.6 C 212 66.9 A 191 30.5 A 36 4.7 B 108 17.5 (2.5–24.6) (10.5–1569.2) (14.2–718.6) (8.6–155.2) (2.5–1569.2) OCPs (ng/g) 102 15.5 C 7582 2008.2 A 15480 2265.4 A 1169 422.8 B 6133 940.8 (42.7–289.4) (472.5–47333.8) (269.6–53559.7) (101.5–16795.4) (42.7–53559.7) No. OCPs 9 0.3 D 13 0.3 B 14 0.2 A 11 0.1 C 12 0.2 (7–11) (10–16) (13–17) (9–13) (7–17)

PAGE 63

47 Table 2-5. Results of RDA evaluating associat ions between clutch survival parameters and OCP variables. Site Variablea LambdaA P F Lochloosa NOC 0.11 0.074 2.25 [DL] 0.09 0.194 1.59 PDDT% 0.08 0.166 1.72 PDDE% 0.11 0.104 2.45 Apopka DL% 0.17 0.004 4.25 TC% 0.12 0.024 3.32 ALD% 0.10 0.042 3.16 LPC% 0.06 0.16 1.85 Emeralda Marsh TX% 0.09 0.044 2.99 HE% 0.06 0.09 2.27 ME% 0.06 0.15 1.85 [HE] 0.06 0.15 1.89 Griffin [PDDE] 0.08 0.024 3.67 [TX] 0.07 0.016 3.16 [PDDT] 0.06 0.04 2.71 [ODDD] 0.04 0.09 1.96 aSee Table 2-2 for definition of variable codes.

PAGE 64

48 Table 2-6. Results of RDA evaluating asso ciations between egg and clutch size parameters and OCP variables. Site Variable LambdaA P F Lochloosa NOC 0.31 0.004 10.15 [PDDT] 0.20 0.042 4.29 [TN] 0.13 0.006 6.77 OX% 0.08 0.088 2.8 Griffin PDDD% 0.05 0.134 2.32 [ODDT] 0.03 0.406 0.91 [PDDT] 0.02 0.236 0.95 [CC] 0.01 0.54 0.33 Emeralda [ODDT] 0.22 0.01 8.07 CC% 0.05 0.146 2 ODDT% 0.05 0.182 1.82 LPC% 0.04 0.21 1.7 Apopka [PDDD] 0.24 0.01 6.51 [ME] 0.08 0.112 2.54 [PDDT] 0.05 0.218 1.5 PDDE% 0.05 0.294 1.29

PAGE 65

49 -0.80.6-0.60.6 Clutch viability Unbanded egg% Early Emb. Mort. Late Emb. Mort. NOC [DL] PDDE% PDDT% A B Figure 2-1. Biplot of clutch survival parameters (solid lines) and organochlorine pesticide variables (dashed lines) for clutches of alligator eggs collected from Lake Lochloosa during summer 2001-2002. Arrows pointing in the same direction indicate a positiv e correlation (e.g., clutch viability and PDDE%), arrows that are approximately perpendicular indicate near-zero correlation (e.g., late emb. mort. and [DL]), and a rrows pointing in opposite directions indicate negative correlations (e.g., clut ch viability and [DL]. Arrow lengths indicate rank order of correlations. For example, late emb. mort. has higher positive correlation with NOC (A) compared to unbanded egg% (B). Cosine of angle formed between individual clutch variables and individual OCP variables (see Table 2-2 for code defin itions) equals correlation coefficient (r) (ter Braak, 1995). For example, arrows pointing in exactly opposite directions have an angle of 180, and since cos( 180) = -1.0, the arro ws are perfectly, negatively correlated (r ) (ter Braak, 1995).

PAGE 66

50 -1.01.0-0.80.6 Clutch Viability Unbanded egg% Early Emb. Mort. Late Emb. Mort. [ODDD] [PDDE] [PDDT] [TX] Figure 2-2. Biplot of clutch survival parameters (solid lines) and organochlorine pesticide variables (dashed lines) for clutches of alligator eggs collected from Lake Griffin during summer 2000-2002.

PAGE 67

51 -1.01.0-0.80.8 Clutch Viability Unbanded egg% Early Emb. Mort. Late Emb. Mort. LPC% ALD% DL% TC% Figure 2-3. Biplot of clutch survival parameters (solid lines) and organochlorine pesticide variables (dashed lines) for clutches of alligator eggs collected from Lake Apopka during summer 2000-2002.

PAGE 68

52 -0.80.6-0.60.6 Clutch Viability Unbanded egg% Early Emb. Mort. Late Emb. Mort. [HE] HE% ME% TX% Figure 2-4. Biplot of clutch survival parameters (solid lines) and organochlorine pesticide variables (dashed lines) for clutches of alligator eggs collected from Emeralda Marsh during summer 2000-2002.

PAGE 69

53 -1.01.0-0.60.4 Clutch Mass Egg Mass Fecundity NOC [PDDT] [TN] OX% Figure 2-5. Biplot of egg a nd clutch size parameters (sol id lines) and organochlorine pesticide variables (dashed lines) for clutches of alligator eggs collected from Lake Lochloosa during summer 2001 and 2002.

PAGE 70

54 CHAPTER 3 MATERNAL TRANSFER OF ORGANOCHLORINE PESTICIDES Studies have documented organochlorine pe sticide (OCP residues) in eggs and/or somatic tissues of several crocodilian species including the American alligator, Alligator mississippiensis (Heinz et al., 1991), MoreletÂ’s crocodile, C moreletti (Wu et al., 2000a), the American crocodile, Crocodylus acutus (Hall et al., 1979; Wu et al., 2000b), and the Nile crocodile, C niloticus (Skaare et al., 1991). Indeed, alligator populations inhabiting Lake Apopka, where an OCP spill occurred in the 1980s, and other central Florida lakes contaminated with OCPs (through historic OCP use) produce eggs that contain concentrations of total OCPs that are over 100 times higher than concentrations found in eggs from reference lakes (Gross, unpublis hed data). In addition, the alligator populations inhabiting the OCP-contaminated lakes experience increased (and highly variable) rates of embryonic mortality, leading to reduced clutch success, and juvenile alligators appeared to have abnormal sex hormo ne concentrations as compared to those of reference sites (Masson, 1995; Rice, 1996; W oodward et al., 1993). However, a clear dose-response relationship has not been estab lished with respect to individual or total OCP concentrations in egg yolks and reduced clutch success (Heinz et al., 1991). The lack of a clear dose-response suggests othe r factors (e.g., diet, population dynamics, and specific OCP mixtures) might be involved and/or that developmental effects result from altered maternal physiology resulting from OCP exposure, as opposed to direct embryotoxicity.

PAGE 71

55 With respect to altered maternal physiol ogy, alterations in steroid hormone levels have also been shown in alligators inhabiti ng OCP-contaminated sites (Guillette et al., 1994). Furthermore, maternal exposure s uggests that OCPs may be maternally transferred from the adult female alligator to her offspring, as has been reported in other oviparous vertebrates (Russell et al., 1999). Assuming OCPs are maternally transferred, the possibility exists that yolks could be used as predictors of maternal exposure. A noninvasive method such as this would aid eco logical risk assessmen ts in understanding exposure levels for rare/endangered crocodilian species without having to capture and/or remove adults from the breeding population. Therefore, the objectives of the present study were to examine maternal transfer as a potential route for embryonic OCP exposure, and to evaluate the use of yolk bur dens for predicting OCP burdens in maternal tissues in alligators. Our hypothesis was that OCP burdens in maternal tissues and yolks would be strongly correlated, which would a llow yolk burdens to be used to predict maternal body burdens and suggest maternal tr ansfer of OCPs as the major route for embryonic OCP exposure. Materials and Methods Site descriptions Lakes Apopka (N 28 35Â’, W 81 39Â’), Griffin (N 28 53Â’, W 81 49Â’), and Lochloosa (N 29 30Â’, W 82 09Â’) in Florida were selected as co llection sites because prior studies by our laboratory indicate vastly different le vels of OCP exposure across these sites. All three lakes are part of th e Ocklawaha Basin. Lake Lochloosa (which is connected to Orange Lake) was selected as a low exposure (reference) site. Four years (1999-2002) of data indicate mean total OCP concentrations in egg yolks from the reference sites (Lakes Orange and Loch loosa) were 231 30 ppb (mean standard

PAGE 72

56 deviation [SD], n = 56 clutches) with a concurrent mean clutch viability rate (number of live hatchlings/total number of eggs in a ne st) of 71 21% (Gross, unpublished data). Lake Griffin was selected as an intermedia te exposure site since yolk concentrations averaged 4,414 617 ppb ( n = 47 clutches) and Lake A popka was selected as a high exposure site since yolk concen trations averaged 15,911 1,786 ppb ( n = 42) for the same time period (Gross, unpublished data). Furthermore, mean clutch viability rates during this time period for Lakes Apopka (51 31%, n = 42) and Griffin (44 33%, n = 47) have been below rates observed for the reference site. Animal Collections Adult female alligators and their corres ponding clutches of eggs were collected from Lakes Apopka ( n = 4), Griffin ( n = 8), and Lochloosa ( n = 3) over the course of two nesting seasons (June 2001 and June 2002). Nests were located by aerial survey (helicopter) and/or from the ground (airboat). Once nests were located, all eggs were collected, and the nest cavity was covered. A snare-trap was set perpendicular to the taildrag in order to capture the female as she cr ossed over the nest. After the traps were set, one member of the trapping crew subsequently transported the eggs to the Florida Fish and Wildlife Conservation CommissionÂ’s Wild life Research Unit (FWC; Gainesville, FL, USA) and placed the eggs in a temperaturecontrolled incubator. Snare-traps were checked later in the evening a nd early the next morning. Trapped females were secured and transported from each lake to the United States Geological SurveyÂ’s Florida In tegrated Science Center (USGS; Gainesville, FL, USA). Upon arrival, the animals were weighed, m easured, and blood samples were collected from the post-occipital sinus. Adult alligators were then euthanized by cervical dislocation followed by double pithing. A full necropsy was performed on each female.

PAGE 73

57 Bile, liver, adipose (composite of abdomina l fat and the abdominal fat pad), and tail muscle samples were collected for later dete rmination of OCP burdens. Liver, adipose tissue, and muscle were wrapped in aluminum foil, while bile and blood were placed in scintillation vials. All sa mples were grouped according to nest identification number (ID), placed in plastic bags labeled with the appropri ate ID, and stored in a –80 oC freezer. Each female’s corresponding clutch of eggs was then transferred from FWC to USGS where yolk samples were collected (t wo eggs/clutch) and stored with the corresponding maternal tissues. The remain ing eggs were set for incubation in a temperature/humidity-contr olled incubator (31-33 oC, 88-92% relative humidity) located at USGS. Analysis of OCPs in Ma ternal Tissues and Yolk Analytical grade standards for the following compounds were purchased from the sources indicated: aldrin, al pha-benzene hexachloride ( -BHC), -BHC, lindane, -BHC, p,p’ -dichlorodiphenyldichloroethane ( p,p’ -DDD), p,p’ -dichlorodiphenyldichloroethylene ( p,p’ -DDE), dichlorodiphe nyltrichloroethane ( p,p’ -DDT), dieldrin, endosulfan, endosulfan II, endosulfan sulfate, endrin, e ndrin aldehyde, endrin ketone, heptachlor, heptachlor epoxide, hexachlorobenz ene, kepone, methoxychlor, mirex, cis -nonachlor, and trans -nonachlor from Ultra Scientific (Kingstown, RI, USA); cis -chlordane, trans chlordane, and the 525, 525.1 polychlorinated biphenyl (PCB) Mix from Supelco (Bellefonte, PA, USA); oxychlordane from Chem Service (West Chester, PA); o,p’DDD, o,p’DDE, o,p’DDT from Accustandard (New Haven, CT, USA); and toxaphene from Restek (Bellefonte, PA, USA). All reag ents were analytical grade unless otherwise indicated. Water was doubly distilled and deionized.

PAGE 74

58 Adipose, liver, bile, and yolk samples were analyzed for OCP content using methods modified from Holste ge et al. (1994 and Schenck et al. (1994). For extraction, a 2 g tissue sample was homogenized with ~1 g of sodium sulfate and 8 mL of ethyl acetate. The supernatant was decanted and filtered though a Bchner funnel lined with Whatman #4 filter paper (Fisher Scientific, Ha mpton, NH, USA ) and filled to a depth of 1.25 cm with sodium sulfate. The homogenate was extracted twice with the filtrates collected together. The combined filtrat e was concentrated to ~2 mL by rotary evaporation, and then further concentrated until solvent-free unde r a stream of dry nitrogen. The residue was reconstituted in 2 mL of acetonitrile. After vortexing (30 s), the supernatant was applied to a C18 solid phase extrac tion (SPE) cartridge (preconditioned with 3 mL of acetonitrile; Agile nt Technologies, Wilmi ngton, DE, USA) and was allowed to pass under gravity. This proced ure was repeated twice with the combined eluent collected in a culture t ube. After the last addition, th e cartridge was rinsed with 1 mL of acetonitrile which was also collected. The eluent was then applied to a 0.5 g NH2 SPE cartridge (Varian, Harbor City, CA, US A), was allowed to pass under gravity, and collected in a graduated conical tube. The car tridge was rinsed with an additional 1 mL portion of acetonitrile whic h was also collected. The combined eluents were concentrated under a stream of dry nitrogen, to a volume of 300 L, and transferred to a gas chromotography (GC) vial for analysis. Whole blood was analyzed for OCP c ontent using methods modified from Guillette et al. (1999). A 10 mL aliquot was transferred from the homogenized bulk sample and extracted in 15 mL of acetone by vortex mixer. The mixture was centrifuged for 5 min at 3000 rpm, after which the supernatan t was transferred to a clean culture tube.

PAGE 75

59 This process was repeated with the supern atants collected and concentrated under a stream of dry nitrogen until solvent-free. Th e residue was re-extract ed in 11.5 mL of 1:1 methylene chloride-petroleum ether. After mixing, the sample was allowed to settle and the upper layer was tran sferred to a clean culture tube. This extraction was performed twice with the extracts collected together. Th e combined extracts we re then applied to a prepared florisil cartridge (5 mL Fisher Pr epSep, Fisher Scientific, Hampton, NH, USA). The cartridge had been prepared by filling the reservoir to a depth of 1.25 cm with anhydrous sodium sulfate and by prewashing the modified cartridge with 10 mL of 2:1:1 acetone: methylene chloride: petroleum ether. After the sample passed under gravity with the eluent collected in a 15-mL graduate d conical tube, the cart ridge was eluted with 4 mL of the 2:1:1 solv ent mixture which was also collected. The combined eluents were concentrated under a stream of dry nitrogen, to a volume of 300 L, and transferred to a GC vial for analysis. GC/MS Analysis Analysis of all samples was performed using a Hewlett Packard HP-6890 gas chromatograph (Wilmington, DE, USA) with a split/splitless inlet ope rated in splitless mode. The analytes were introduced in a 1 L injection and separa ted across the HP-5MS column (30 m x 0.25 mm; 0.25 m film thickne ss; J & W Scientific, Folsom, CA, USA) under a temperature program that began at 60 C, increased at 10 C/min to 270 C, was held for 5 min, then increased at 25 C/min to 300 C and was held for 5 min. Detection utilized an HP 5973 mass spectro meter in electron impact m ode. Identification for all analytes and quantitation for toxaphene was c onducted in full scan mode, where all ions are monitored. To improve sensitivity, se lected ion monitoring was used for the

PAGE 76

60 quantitation for all other analytes, except kepone. The above program was used as a screening tool for kepone which does not optim ally extract with mo st organochlorines. Samples found to contain kepone would be reex tracted and analyzed specifically for this compound. For quantitation, a five-point standard curve was prepared for each analyte ( r2 0.995). Fresh curves were analyzed with each se t of twenty samples. Each standard and sample was fortified to contain a deuterat ed internal standard, 5 L of US-108 (120 g/mL; Ultra Scientific), added just prior to analysis. All samples also contained a surrogate, 2 g/mL of tetrach loroxylene (Ultra Scientific) added after homogenization. Duplicate quality control samples were prepar ed and analyzed with every twenty samples (typically at a level of 1.00 or 2.50 g/mL of -BHC, heptachlor, aldr in, dieldrin, endrin, and p,p’ -DDT) with an acceptable recovery rangi ng from 70 – 130%. Limit of detection ranged from 0.1-1.5 ng/g for all OCP analyt es, except toxaphene (120-236 ng/g), and limit of quantitation was 1.5 ng/g for all anal ytes, except toxaphene (1500 ng/g). Repeated analyses were conducted as allo wed by matrix interferences and sample availability. Data Analysis OCP concentrations in maternal tissues and egg yolks were lipid-adjusted (wet weight concentration / proportion of lipid in tissue), and lipidadjusted tissue-to-egg yolk ratios (maternal tissue OCP concentrations /egg OCP concentrations) were examined. Predictive models were determined by linear regression analysis of OCP concentrations in yolk against those of maternal tissues (log -transformed wet weight concentrations). Each model’s ability to fit the data was evaluated by exam ining the p-value ( = 0.05), the r2 value and the residual plots (SAS Institute Inc., 2002). ANOVA was used for

PAGE 77

61 inter-site comparisons of adult female and clut ch characteristics, and the Tukey test was used for multiple comparisons among sites. The relationship between maternal mass (kg) and concentrations of OCPs in eggs and ma ternal tissues (log-transformed wet weight concentrations) were evaluated using linear regression to assess whether increasing mass was associated with increasing concentrations of OCPs in eggs and maternal tissues, which may suggest adult females continue to bioaccumulate OCPs as they grow throughout their life. Adult females were gr ouped by site since the extreme differences in OCP exposure among sites would likely confound results. Unless otherwise noted, values are reported as mean standard deviation. Results Female Morphological and Re productive Characteristics For all females, mass and snout-vent length (SVL) averaged 74 20 kg (range: 44-114) and 135 11 cm (119-156), respectively. Clutch mass (mass of all eggs from a single nest) and fecundity (number of eggs collected from a single nest) of these individuals were 3.65 0.86 kg ( 1.84-4.82) and 43 10 eggs/nest (19-56), respectively. No significant differences were detected across sites with respect to female mass (p = 0.14), total length (p = 0.90), SVL (p = 0.25), tail girth (p = 0.98), head length (p = 0.55), clutch mass (p = 0.23), or fecundity (p = 0.40, Table 1). With respect to lipid concentrations in egg yolk and muscle, no significant differences were detected across sites (p > 0.05) However, lipid concentration in liver of Lochloosa females was significantly higher (p < 0.05) than that of Apopka and Griffin females (which were not significantly different from one another). Furthermore, lipid concentration in abdominal adipose tissue of Apopka females was si gnificantly less (p < 0.05) than that of Lochloosa and Griffin females (Table 1).

PAGE 78

62 OCP concentrations in Yolk Egg yolks from Lake Apopka female s contained the highest total OCP concentration (15,108 13,704) and greatest number of individual OCPs detected above the limit of quantitation (n = 18) with p,p Â’-DDE (66%) and toxaphene (32%) being main constituents. Lake Griffin females produced eggs with the next highest total OCP burdens (393 300 ng/g; n = 13) being main ly composed of p,pÂ’-DDE (69%), transnonachlor (10%), and dieldrin (7%). Lake Lochloosa females produ ced egg yolks with the smallest total OCP burden (124 53 ng/g, n = 9), with main constituents being p,pÂ’DDE (73%), trans-nonachlor (10%), and ci s-nonachlor (4%; Table 3-2). The OCP analytes with the highest average egg yolk concentrations were toxaphene (4,862 4,177 ng/g), which was detected above the limit of quantitation in 3 of 15 clutches, followed by p, pÂ’-DDE (2,828 5,968 ng/g), dieldrin ( 191 474 ng/g), and trans-nonachlor (126 209 ng/g), which were above quantita tion limit in all 15 clutches. OCP concentrations in maternal tissues Adipose tissue (a composite of abdominal fat and fat pad) contained the highest concentration of total OCPs (12,805 31,678 ng/ g wet weight) of all tissues. p,pÂ’-DDE (67%) composed the majority of the total burden, followed by dieldrin (5%), and transnonachlor (3%). Although toxaphene was only detected in 3 individuals from Lake Apopka, its average burden in adipose tis sue was 13,463 1,267 ng/g (Table 3-2). In liver, OCP analytes were detected above the quantitation limit in 9 of 15 individuals, and total OCP concentrations averaged 1,008 1,245 ng/g. Liver burdens were primarily composed of p,pÂ’-DDE (76%) and dieldrin (6 %). Total OCP concentrations in muscle averaged 716 1,053 ng/g and were above qua ntitation limits in 10 of 15 individuals

PAGE 79

63 with most of the burden being composed of p,pÂ’-DDE (83%), dieldrin (6%) and transnonachlor (6%). Total OCP burdens in bile (412 483 ng/g) were above quantitation limits in five individuals with p,pÂ’-DDE (86 %) and dieldrin (6%) comprising the majority of the burden. Total OCP concentrations in blood (43 21 ng/g) we re above quantitation limits in 4 individuals with p,pÂ’-DDE (64%) and dieldrin (14%) comprising most of the burden. Overall, Lake Apopka alligators ex hibited the highest OCP concentrations in maternal tissues and egg yolks, followed by Lakes Griffin and Lochloosa, respectively (Table 3-2). Relationships between Maternal Tissue and Yolk Burdens Examination of lipid-adjusted matern al tissue-to-egg yolk burdens showed differences among tissues. With respect to total OCPs, the adipose burden-yolk burden ratio was close to 1 (95% confidence interval (CI), 0.76 1.11). In contrast, the liver-yolk ratio was si gnificantly greater than 1 (95% CI, 1.49 9.19), and muscle ratios showed considerable variation (95% CI, -1.17 37.35). As would be expected, most individual OCPs follo wed the above trend. Howeve r, cis-chlordane was an exception as liver ratios (95% CI, 2.85 6.75) and muscle ratios (95% CI, 1.78 15.1) were greater than 1, while adipose ratios (95% CI, 0.59 0.84) were less than 1. With respect to total OCP concentrations significant linear relationships (predictive models) were found for adipose, liver, muscle, and bile (p 0.05, Fig. 1). With respect to individual OCP analytes, predictive models we re derived for 12 of 14 (78%) of the OCPs co-detected in adipose tissue and egg yolk, followed by liver (9/12, 75%), bile (8/11, 73%), and muscle (2/12, 17%; Table 3-3) Although nine OCP analytes were concurrently detected in blood of the females and their respective egg yolks, no significant linear correlations we re detected (p > 0.05).

PAGE 80

64 As for individual OCP analytes, p,pÂ’-DDE c oncentrations in yolk was significantly correlated with those of liver, muscle, bile, and adipose tissue. Blood p,pÂ’-DDE concentrations did not exhib it a significant linear relations hip (p > 0.05) with yolk p,pÂ’DDE concentrations. Heptachlor epoxide, trans-chlordane, cis-chlordane, transnonachlor, cis-nonachlor, mirex, and dieldrin concentrations in yol k were significantly correlated to their respective concentrations in adipose, liver, and bile. With respect to oxychlordane, significant correl ations were only derived for liver and adipose tissue, and significant correlations for p,pÂ’-DDD concentr ations were found only for adipose and bile. Toxaphene and o,pÂ’-DDT concentrati ons in adipose tissue were significantly correlated with respective egg yol k concentrations (Table 3-3). Relationships between Maternal Mass and OC P concentrations in Eggs and Tissues For females collected from Lakes Apopka (n = 4) and Lochloosa (n = 3), no significant correlations (p > 0.05) were found when maternal mass (kg) was compared against either individual or total OCP con centrations (log-transformed wet weight) in maternal tissues and eggs. However, significan t correlations might have been difficult to detect because of the small sample size. In contrast, a larger nu mber of Lake Griffin females (n = 8) were collected, and analys es indicated significan t correlations between maternal mass and OCP concentrations in tissu es and eggs indicating that larger females have higher concentrations of OCPs in their tissues and eggs, which may suggest females continue to bioaccumulate OCPs as they gr ow (increase in mass). For Lake Griffin females, OCP burdens in eggs had the grea test number of significant correlations (p 0.05) with body mass (kg), which consisted of cis-nonachlor (r2 = 0.87), cis-chlordane (r2 = 0.75), trans-nonachlor (r2 = 0.73), dieldrin (r2 = 0.69), p,pÂ’-DDE (r2 = 0.66), o,pÂ’-DDT (r2 = 0.61), heptachlor epoxide (r2 = 0.59), oxychlordane (r2 = 0.58), trans-chlordane (r2 =

PAGE 81

65 0.57), and total OCPs (r2 = 0.71). Following egg concen trations, abdominal fat OCP burdens-to-body mass correlations consisted of cis-nonachlor (r2 = 0.67), cis-chlordane (r2 = 0.81), trans-nonachlor (r2 = 0.63), dieldrin (r2 = 0.62), p,pÂ’-DDE (r2 = 0.58), heptachlor epoxide (r2 = 0.53), oxychlordane (r2 = 0.51), and total OCPs (r2= 0.64). Although egg burdens of o,pÂ’-DDT and transchlordane were correlated with body mass, abdominal fat burdens were not. Lastl y, liver OCP burdens-to-body mass correlations included only trans-nonachlor (r2= 0.99) and p,pÂ’-DDT (r2= 0.99). No significant correlations were found for ci s-chlordane, trans-chlordan e, oxychlordane, dieldrin, heptachlor epoxide, o,pÂ’-DDT, and cis-nonachlor. Discussion The presence of OCPs in the eggs and tissu es of alligators is not novel; however, the value of our study was that OCP concen trations in maternal tissues and yolks appeared to be strongly correla ted with one another, allowing yolk burdens to be used as predictors of OCP burdens in tissues of adult reproductive alligators, which may be a useful noninvasive technique that would aid risk assessments involving endangered crocodilians. Furthermore, our results are cons istent with other stud ies that suggest OCPs are maternally transferred in wild alligators (Rausche nberger et al., 2004). Several OCP analytes were detected in both maternal tissues and yolk (Table 3-3) suggesting that mixture composition may be an important consideration in risk assessment. One reason for this is that different xenobiotic compounds may induce or inhibit certain biotransformation enzymes. Specifically, alligators from Louisiana express several different xenobi otic biotransformation enzymes (e.g., liver cytochrome P450 enzymes [CYP] such as CYP1A, CYP2B) in response to xenobiotic exposure (Ertl et al., 1999). Furthermore, genetic partitioning has been reported in spatially separated

PAGE 82

66 alligator populations (Ryberg et al., 2002). Therefore, the po ssibility exists that certain individuals or populations may lack the ge netic or epigenetic ability to produce a particular biotransformation enzyme, which may lead to increased risk of xenobioticinduced toxicity. For example, certai n populations of black -banded rainbowfish ( Melanotaenia nigrans ) were able to tolerate copper exposures (96-hr EC50) that were 8.3 fold greater than the toleran ce limits of other, spatially-sep arated populations of the same species. Genetic analyses sugge sted that allozyme frequencies of tolerant and susceptible populations were significantly di fferent at AAT-1 and GPI-1 loci, suggesting differences in allozymes of exposed fish may have a ssisted in the increas ed copper tolerance (Woosley, 1996). Examination of maternal tissue-to-egg con centration ratios (lipid-adjusted) showed differences among tissues. The adipose-toyolk concentration ratio was close to 1, suggesting that OCPs reach equilibrium within abdominal adipose tissue, and that lipids and OCPs are mobilized and subsequently in corporated into the developing yolks. In contrast, liver-to-yolk concentr ation ratios were significantly greater than 1, and muscleto-yolk concentration ratios showed consider able variation. One suggested explanation for the high liver-to-yolk ratios relates to one major function of the liver cells (hepatocytes), which is to accumulate and convert hydrophobic xenobiotics into hydrophilic metabolites to facilitate detoxicatio n, excretion, and elimination. In addition, the low lipid content of liver (relative to the lipid content of adipose tissue and yolk, Table 3-1) may have contribute d to the marked differences. With respect to the muscleto-yolk ratios, the reasons fo r the large degree of variabil ity are not as clear. One possible explanation is that muscle lipids are not mobilized during yolk formation and, as

PAGE 83

67 a result, OCP burdens may continually accumulate in muscle lipids. Another potential explanation relates to the low lipid content of muscle when compared to yolk (Table 3-1). Lastly, cis-chlordane’s exceptional liver, musc le, and adipose ratios underscore the fact that different OCP analytes may not al ways exhibit identi cal pharmacokinetics. When compared to other vertebrates, ad ipose tissue-to-egg ratios in alligators are similar to those reported in the freshwater catfish, Clarias batrachus in that adipose-toegg ratios are approximately equal to 1. Furthermore, C batrachus mobilizes lipids from its abdominal adipose tissue during vitelloge nesis (Lal & Singh, 1987) similar to what this study suggests occurs in the American alli gator. In contrast to adipose tissue OCP concentrations, muscle-to-egg OCP ratios in al ligators appear to be quite different from fish. Alligator muscle-to-egg ratios were highly variable and, for the most part, greater than 1, while fish ratios appear to be consiste ntly close to 1. With respect to more closely related species, muscle-to-egg OCP ratios ar e similar to those reported for the common snapping turtle ( Chelydra serpentina ) and several bird specie s with ratios exhibiting a great deal of variability and being greater than 1 (Russell et al., 1999). These differences suggest that fish differ from te rrestrial vertebrates in regard s to lipid content of muscle and/or lipid mobilization strategy (during vitell ogenesis), which could lead to differences in embryonal exposure given equivalent maternal exposure. Evaluation of Predictive Models Although significant linear models were found fo r most tissues with respect to total OCP concentrations, caution should be used in the application of these “total OCP” models since it is probable that the concentrations and ratios of individual OCP analytes may vary across different locations. The great est number of predictive linear models was derived for adipose tissue. This was not surp rising considering that (for most analytes)

PAGE 84

68 adipose-yolk lipid normalized ratios were close to 1. Next, with respect to the number of significant linear models, were liver and bile. The similarities between liver and bile should be expected since the liver produces bi le, which transports OCP analytes to the intestinal lumen, leading to their eventual elimination from the body. However, OCP analytes may be reabsorbed from the intestin e and redirected back to the liver via the portal vein through a process known as ente rohepatic circulati on, which may delay elimination of lipophilic xenobiotics, increa se hepatic exposure and bioaccumulation (Stenner et al., 1997). For OCP concentrations in muscle, regressi on analysis indicated that only two out of 12 mutually detected analytes could be predicted using OCP concentrations in eggs. Las tly, nine OCP analytes were c oncurrently detected in blood and egg yolk with none exhibiting significant re lationships. Possible explanations for the few significant linear re lationships include the low lipid co ntent of these tissues and thus the relatively low concentrations of OCP an alytes in these tissues, as well as the possibility that each of these tissue burden s exhibit a nonlinear relationship with yolk burdens. In addition, blood samples were co llected after the female had oviposited. Since blood was collected after eggs were ex creted from the body, it is likely that the overall maternal body burden decreased, which would in turn lower the steady-state OCP concentrations in blood. As for individual OCP analytes, predictiv e models for p,pÂ’-DDE were derived for four of the five maternal tissues. One likely reason for this is that p,pÂ’-DDE was detected in considerable concentrations in all eggs and in almost a ll tissues for all 15 females. Similarly, predictive models were derived for commonly detected analytes such as heptachlor epoxide, trans-chlo rdane, cis-chlordane, tran s-nonachlor, cis-nonachlor,

PAGE 85

69 mirex, and dieldrin for most tissues. Some what surprisingly, oxychlordane (a metabolite of cisand trans-chlordane) and p,pÂ’-DDD (an intermediate metabolite of p,pÂ’-DDT) showed significant linear models only with respect to liver an d adipose tissue. The fact that linear models for toxaphene and o,pÂ’-DDT were derived only fo r adipose tissue was likely related to their low concentrations and in frequent detections in other tissues (Table 3-3). Relationships between Maternal Mass and OC P concentrations in Eggs and Tissues Although a portion of a female alligato rÂ’s OCP body burden may be eliminated through egg deposition, adult female alligato rs from Lake Griffin had increased OCP concentrations in their tissues and eggs as they increased in mass, similar to size-related OCP bioaccumulation in smallmouth bass inha biting contaminated sites in Michigan (Henry et al., 1998). Corresponding increases in OCP burdens and mass indicate that larger and possibly older females accumulate OC Ps faster than they can excrete them. In addition, the relationship between OCP burdens in eggs and body mass was very similar to the relationship between abdomin al fat burdens and body mass. The correlation between OCP burdens in liver and body mass was significant for trans-nonachlor and p,pÂ’-DDT; however the major metabolites of these compounds (oxychlordane and p,pÂ’-DDE, respectively) we re not significantly correlated with body mass. These results contrast those of egg a nd abdominal fat burdens and suggest that that alligator liver may not sequester OCP metabolites to the same extent as abdominal fat or egg. Maternal body burdens: Toxicological Implications Although our studyÂ’s objective was to evaluate maternal transfer and prediction of the maternal OCP body burdens carried by the Am erican alligator, we would be remiss if

PAGE 86

70 we did not discuss whether these reported body burdens were capable of eliciting harmful effects. Although several stud ies report body and egg burdens in crocodilians, relatively few studies directly relate body and egg burdens to acute t oxicological effects (Campbell, 2003), so we will briefly discuss how p,p’-DDE burdens in maternal alligator liver compare to reported p,p’-DDE burdens in liver of birds (birds were not from the present study areas) that have been associated with mortality (Blus, 1996). In previous studies, mean DDE liver re sidues in birds which died due to DDT exposure ranged from 19,000–55,000 ng/g. When birds were exposed to DDE alone, liver residues of dead birds av eraged 3,883,000 ng/g (ra nge 460,000–11,725,000 ng/g) (Blus, 1996). When compared to the liver re sidues of the most contaminated alligators (Lake Apopka, upper 95% CI < 7,000 ng/g), it appears that death due to DDT/DDE exposure might be unlikely assuming bird and alligator susceptibilities are similar. However, since p,p’-DDE liver conc entrations in alligators are almost half of lethal liver concentrations in birds, there is reason for some concern. In addition, the assumption that bird and alligator susceptibil ities are similar might be ar gued as unfounded considering the variability in toxic responses between i ndividuals of the same species, different species, and different verteb rate classes (James et al., 2000). To account for these uncertainties the risk assessment process identifies the different sources of uncertainty and incorporates the uncertainty in attempti ng to determine a “safe” tissue concentration based on levels associated with no adverse effects (NOAEL) or lowest observed adverse effect levels (LOAEL). Typical ly, interspecies extrapolati on is assigned an uncertainty factor of 10, as are inter-indi vidual uncertainty, uncertainty related to comparing different study designs (e.g., acute doses rela ted to experimental bird studi es, in contrast to chronic

PAGE 87

71 exposure studies in wild alligators), and un certainty related to da tabase quality since DDE (p,p’-DDE + o,p-DDE) liver residues were reported, instead of p,p’-DDE. These four uncertainty factors cons titute an overall uncertainty factor of 10,000, which is an order of magnitude greater than commonly used uncertainty factors (range: 300-1000) (James et al,. 2000). Considering the high de gree of uncertainty, we suggest that more information is required before a “safe” leve l of p,p’-DDE exposure is determined for the American alligator based upon actual or predicted liver c oncentrations. Sublethal effects are anothe r possible consequence of OC P exposure. For example, exposure of the freshwater catfish, Clarias batrachus to an OCP analyte ( -BHC) at sublethal levels (2,000–8,000 ng/g) during vite llogenesis significantly decreased the biosynthesis and mobilization of phospholipids from liver to the developing follicles (Lal & Singh, 1987). Interestingly, alterations in fa tty acid profiles of alligator eggs have been associated with reduced clutch success. Specifically, fatty acid profiles from wild, alligator eggs (normal hatch rates) showed considerable differences when compared to those of eggs from captive alligators (reduc ed hatch rates). One suggested explanation for this association between altered fatty acid profiles and reduced clutch success in captive alligators was that certain fatty acids are critical for reproduc tive success and that captive diets were deficient in essential fatty acids (Noble et al., 1993). Thus, the possibility exists that exposure to OCPs may alter the liver’s ability to synthesize necessary fatty acids, leading to altered e gg quality and decreased clutch success in wild alligators that inhabit OCP-contaminated site s. Chronic exposure to low doses of OCPs prior to and during vitellogenesis has been s uggested as a cause for significant increases in OCP concentrations in egg yolk, as well as significantly decreased hatch rates in

PAGE 88

72 captive adult female alligators Importantly, the doses did not appear to induce acute toxicity in the adult female s (Rauschenberger et al., 2004). Presently, we are using a captive breeding population of adu lt alligators, as well as da ta from field studies, to further evaluate the relationships between OC P exposure, altered fatty acid biosynthesis, nutritional content of eggs, and embryonic mortality. In summary, the significant levels of OCP analytes observed across such a wide range of crocodilian species and geography s uggests the need for a greater understanding of xenobiotic metabolism and toxicologica l responses in crocodilians. Such understanding would aid in the conservation of this ancient group by determining what risks are posed by contaminants with respec t to species survival and how contaminantrelated risks compare to other risks, such as habitat destruction. The results of the present study provide some evidence suggesting that ma ternal transfer of OCP analytes is the major route for embryonic exposure. In a ddition, it provides several models for the prediction of OCP concentrations in maternal tissues of American alligators, which may be extrapolated to other crocodilians. H opefully, the present study will encourage new investigations into the phar macokinetics and pharmacodynamics of contaminants in other crocodilian species.

PAGE 89

73 Table 3-1. Morphological and reproductive characteristics of adult female alligators collected during June 2001 and 2002 from Lakes Apopka, Griffin, and Lochloosa in central Florida. Parameter a,b ApopkaGriffinLochloosa Number of females collected 483 Total Length (cm) 252 38258 17258 7 Snout-Vent Length (cm) 142 15134 9129 5 Mass (kg) 94 3070 1763 4 Clutch Mass (kg) 3.78 0.983.33 0.824.31 0.45 Fecundity (# eggs/clutch) 43 1040 1049 6 Lipid % Adipose 47.0 32.5 B78.1 8.0 A81.4 4.0 A Lipid % Liver 1.3 1.0 A0.8 0.2 A5.0 2.3 B Lipid % Muscle 0.8 0.91.3 0.90.2 0.02 Lipid % Yolk 19.9 1.118.1 1.718.2 1.6 a Values represent mean standard deviation. b Different letters i ndicate significant differences ( p < 0.05).

PAGE 90

74Table 3-2. Pesticide concentratio ns (ng/g wet wt.) in tissues and yolks of adult female alligators collected during June 2001 and 2002 from Lakes Apopka, Griffin, and Loch loosa in central Florida. Lake a Chemical b, c Bile Blood Adipose Liver Muscle Yolk Apopka Aldrin X X X X X 1 ( 4 ) -BHC X X X X X X -BHC X X 7.5 6.7 X X 2 1.4 cis -Nonachlor 10 3.2 2 0.4 521 602.7 31 6.7 23 18.3 123 81.9 cis -Chlordane 4 1.9 1 0.4 190 241.2 11 9.1 14.1 11.7 62 59.2 -BHC X X X X X X Dieldrin 38 10.2 5 0.4 2,376 3,770.9 105 80.2 68 48.3 663 803.0 Endosulfan I X X X X X X Endosulfan II X X X 21 X X Endosulfan Sulfate X X X X X X Endrin X X X X X X Endrin Aldehyde 3 0.4 X X X X X Endrin Ketone X X X X X X -BHC X X X X X X Heptachlor X X X X 8 11.7 1 0.04 Heptachlor Epoxide 3 2.1 0.3 0 67 81.5 6 2.2 4 3.6 26 15.0 Hexachlorobenzene 1 0 1 0 1 0 1 0.0 1 1 0.0 Kepone X X X X X X Methoxychlor X X X 5 X X Mirex 2 2.6 X 19 13.1 7 9.0 1 0.4 7 7.1 o,pÂ’ -DDD X X 3 3.4 X X X o,pÂ’ -DDE X X 52 55.8 X X 45 17.7 o,pÂ’ -DDT 2 0.2 27 26.4 4 1.8 4 2.2 17 7.8 Oxychlordane 7 1.3 1 0.2 247 336.4 17 9.9 12 10.7 75 68.2 p,pÂ’ -DDD 2 0.2 1 0.2 43 67.5 11 10.3 17 8.0 52 61.4 a p,pÂ’ -DDE 806 341 42 5.7 29,840 34,366 1,846 918.1 1,392 1,0782 9,994 8,529 Toxaphene X X 13,436 12,670.2 X X 4,862 4,177 trans -Nonachlor 21 7.8 3 0.4 1,153 1,378.7 65 22.7 68 57.0 387 277.7 Total OCP 900 369.7 55 7 44,650 53,230 2,140 1,024 1,610 1,226 15,108 13,704

PAGE 91

75Table 3-2. (Continued) Lake Chemical Bile Blood Adipose Liver Muscle Yolk Griffin Aldrin X X X X X X ( 8 ) -BHC X X X X X X -BHC X X 2.1 1.1 X X X cis -Nonachlor 4 2.4 1 0.4 75 74.9 8 4.4 9 10.6 14 7.6 cis -Chlordane 2 0.5 1 0 30 10.8 2 0.7 3 3.4 11 3.7 -BHC X X X X X X Dieldrin 13 4.9 7 109 133.4 17 8.0 22 20.8 26 25.7 Endosulfan I X X X X X X Endosulfan II X X X X X X Endosulfan Sulfate X X X X X X Endrin X 5 X X X X Endrin Aldehyde X X X X X X Endrin Ketone X 2 X X X X -BHC X X 2 X X X Heptachlor X X X X 2 1.4 X Heptachlor Epoxide 3 3.3 1 34 45.7 5 3.6 10 12.0 8 8.8 Hexachlorobenzene 1 0 1 1 0 1 0 1 1 0.0 Kepone X X X X X X Methoxychlor X X X X X 2 Mirex 0.3 0 1 5 3.6 1 1 0.5 1 0.2 o,pÂ’ -DDD X X X X X X o,pÂ’ -DDE X X X X X 3 o,pÂ’ -DDT 1 0.2 1 0.0 10 6.9 X 2 0.1 3 1.8 Oxychlordane 7 4.9 X 56 84 8 6.2 16 17.8 12 14.9 p,pÂ’ -DDE 54 25.7 13 9.1 1,030 931.3 75 46.6 131 132.4 273 204.0 p,pÂ’ -DDT 1 13 3 1.5 29 0.9 X 3 Toxaphene X X X X X X trans -Chlordane 1 0.3 1 0 3 1.7 1 0.3 1 2 1.0 trans -Nonachlor 9 5.9 1 0.09 171 213.5 18 13.2 28 36.0 40 38.3 Total OCP 87 46.4 31 28.4 1,533 1,439 153 78.1 208 227 393 299

PAGE 92

76Table 3-2. Continued. Lake Chemical Bile Blood Adipose Liver Muscle Yolk Lochloosa Aldrin X X X X X X ( 3 ) -BHC NA X X X X X -BHC NA X 1 X X X cis -Nonachlor NA X 17 1.8 1 X 5 1.9 cis -Chlordane NA X 8 1.4 X X 3 0.1 -BHC NA X X X X X Dieldrin NA X 14 4.7 2.6 1.4 4 2.8 Endosulfan I NA X X X 15.6 X Endosulfan II NA X X X X X Endosulfan Sulfate NA X X X X X Endrin NA X X X X X Endrin Aldehyde NA X X X X X Endrin Ketone NA X X X X X -BHC NA X X X X X Heptachlor NA X X X 18 9.1 X Heptachlor Epoxide NA X 11 9.6 X X 3 2.9 Hexachlorobenzene NA X X X X X Kepone NA X X X X X Methoxychlor NA X X X X X Mirex NA X 2.6 X X X o,pÂ’ -DDD NA X X X X X o,pÂ’ -DDE NA X X X X X o,pÂ’ -DDT NA X 3 0.2 X 7.1 1 0.0 Oxychlordane NA X 17 11.0 1 X 5 4.3 p,pÂ’ -DDD NA X 1 0.1 X 1 2 0.9 p,pÂ’ -DDE NA X 297 90.1 20 20.9 11 6.7 91 32.5 p,pÂ’ -DDT NA X 1.4 0.1 X 1.4 X Toxaphene NA X X X X X trans -Chlordane NA X 1 0.1 X 1.4 X trans -Nonachlor NA X 38 24.6 2.6 1.4 12 8.8 Total OCP NA X 407 143.6 28 32.6 33 33.9 124 53.3 a Number of females and clutches collected not ed in parentheses beneath name of lake. b Values represent mean standard deviation

PAGE 93

77[SD], values without SD indicate a single measurement. X indicat es values which were below lim it of detection (LOD) or below l imit of quantitation (LOQ) and NA indicates not analyzed. LOD range d from 0.1-1.5 ng/g for most OCP analytes (toxaphene LOD ranged from 120-236 ng/g), and LOQ ranged was 1.5 ng/g fo r all analytes except for toxaphene ( 1500 ng/g). Percent recovery ranged fro m 70-130%. The following chemicals were neithe r detected in females nor their eggs: -BHC, -BHC, endosulfan sulfate, and kepone. c BHC = Benzene hexachloride; DDD = Dichlorodiphenyldichloro ethane; DDE = Dichlorodiphenyldichloroethylene; DDT = Dichlorodiphenyltrichloroethane; Total OCP = organochlorine pesticide concentrations for all analytes.

PAGE 94

78 Table 3-3. Regression equations for pr edicting organochlorine pesticide (OCP) concentrations in maternal tissues, where LOG [Tissue-OCP] = bo + b1 LOG [Yolk-OCP]. Tissue Chemicala bo b1 n r2 p Adipose Dieldrin 0.6624 0.8785 150.87 < 0.0001 cis -Nonachlor 0.6737 0.9136 150.75 < 0.0001 cis -Chlordane 0.4037 0.9633 150.69 0.0001 Heptachlor Epoxide 0.6294 0.8134 140.62 0.0008 Mirex 0.8217 0.6030 6 0.89 0.0028 o,p -DDT 0.5840 0.6040 140.41 0.0141 Oxychlordane 0.6694 0.8544 150.80 <.0001 p,pÂ’ -DDD 0.2375 0.7 597 140.50 0.0046 p,pÂ’ -DDE 0.6968 0.9216 150.93 <.0001 Toxaphene 0.0880 1.0928 3 0.99 0.0486 trans -Chlordane 0.1733 0.9397 120.58 0.0041 trans -Nonachlor 0.6430 0.8960 150.84 < 0.0001 Bile Dieldrin -0.6196 0.9559 4 0.90 0.0494 cis -Nonachlor -0.3863 0.7646 5 0.97 0.0017 cis -Chlordane -0.4308 0.6314 5 0.83 0.0301 Heptachlor Epoxide -0.3207 0.6959 5 0.79 0.0435 p,pÂ’ -DDD -1.1407 1.0748 4 0.95 0.0246 p,pÂ’ -DDE -0.6385 0.9472 5 0.94 0.0057 trans -Nonachlor -0.2919 0.6867 5 0.96 0.0039 trans -Chlordane -0.2245 -0.4531 5 0.87 0.0220 Blood NSb Liver Dieldrin 0.0248 0.7162 7 0.98 <0.0001 cis -Nonachlor -0.2471 0.8448 8 0.92 0.0002 cis -Chlordane -0.5557 0.8876 7 0.97 <0.0001 Heptachlor Epoxide -0.3878 0.8323 6 0.85 0.0084 Mirex -0.0547 0.9557 5 0.89 0.0155 Oxychlordane -0.2855 0.8123 7 0.92 0.0005 p,pÂ’ -DDE -0.7696 1.0156 100.93 <.0001 trans -Chlordane -0.0722 0.3300 7 0.94 0.0003 trans -Nonachlor -0.2854 0.8263 8 0.98 <.0001 Muscle p,pÂ’ -DDE -0.3733 0.8153 100.54 0.0160 Mirex 0.1816 -0.2797 0.96 0.0040 a BHC = Benzene hexachloride; DDD = Dich lorodiphenyldichloroethane; DDE = Dichlorodiphenyldichloroethylene; DDT = Dichlorodiphenyltrichloroethane. b NS = no significant linear regressions were determined for the 9 chemicals which were detected both in blood and in yolk.

PAGE 95

79 Log Total OCPs in Yolk (ng/g) 100101102103104105 Log Total OCPs in Tail Muscle(ng/g) 100101102103104105106 100101102103104105 Log Total OCPs in Liver (ng/g) 100101102103104105106 100101102103104105 Log Total OCPs in Adipose (ng/g) 100101102103104105106 Apopka Griffin Lochloosa Log Total OCPs in Yolk (ng/g) 100101102103104105 Log Total OCPs in Bile (ng/g) 100101102103104105106 Figure 3-1. Linear regressions of total organochlorine pesticide (OCP) concentrations in maternal tissues against total OCP concentrations in egg yolks. A. Adipose tissue. B. Liver. C. Bile. D. Muscle. D. y = -0.0865 + 0.6688 x ( r2 = 0.55, p <0.05) A. y = -1338.60 + 3.318 x ( r2 = 0.95, p <0.05) B. y = -0.6330 + 0.9626 x ( r2 = 0.88, p <0.05) C. y = -0.4817 + 0.8342 x ( r2 = 0.89, p <0.05)

PAGE 96

80 CHAPTER 4 MATERNAL FACTORS ASSOCIATED WI TH DEVELOPMENTAL MORTALITY IN THE AMERICAN ALLIGATOR Recent data suggested maternal organoc hlorine pesticide (OCP) body burdens and OCP egg yolk concentrations are signif icantly correlated, an d that significant relationships between maternal size and maternal body burdens exist. Maternal age and size has also been shown to have a strong re lationship with clutch viability (number of live hatchings / total number of eggs) and cl utch size characteristics (i.e., fecundity, clutch mass). Specifically, females between 15 and 30 years old (~ 2.3–2.8 m in total length) produce larger clutches (35-40 eggs / clutch) with increased clutch viability compared to younger females, which themselves produce smaller clutches (15–25 eggs) with smaller eggs and have decreased clutch viability. Females older than 30 years tend to produce clutches similar to 15-30 year old females, with the only exception being smaller clutches (15–25 eggs) (Ferguson, 1985). Therefore, female size or age may be a confounding factor when examining the rela tionship between OCP burdens in yolk and reproductive performance. In addition, age (or size) and maternal OCP exposure could cause interactive effects. For example, females of optimum reproductive age may be more resistant to effects of OCPs; while, younger (or older) females may show increased susceptibility. Therefore, th e objective of the present study was to test the hypotheses that reproductive efficiency, clutch viabi lity, and mortality rate s are significantly correlated with maternal OCP body burdens, mate rnal size, or both; and (2) that clutch size characteristics are significantly correlated with maternal OCP body burdens, maternal size, or both.

PAGE 97

81 Materials and Methods The greatest difficulty in examining the relationship between maternal age and OCP exposure and effects is that determining th e age of an alligator requires either long term monitoring or counting the rings that form in the femur as a result of annual calcium deposition (Ferguson, 1985). However, this technique is not valid for reproductive females since femoral bone resorption provides calcium necessary for eggshell formation and egg yolk nutrition, and subsequently ca uses the removal of “bone rings” and underestimation of age (Elsey & Wink, 1985; Wink & Elsey, 1986). In addition, removing an alligator’s limb simply to age it is ethically unacceptable. Given these difficulties with assigning a chr onological age, female size will be used lieu of age. One potential limitation in using female size as an i ndicator of age class is that female growth rates between lakes may differ since dietar y composition has been suggested to differ among OCP-contaminated sites and referen ce sites (Rice, 2004). Therefore, the possibility exists that a fema le from a reference site may be smaller than one from a contaminated site, even though both are of the sa me age. This is important since age, in addition to size, has been shown to be an important determinant of sexual maturity in alligators. Indeed, alligator ranchers are able to accelerate growth so that a female may reach six feet in length in 3-4 years, howeve r, these females do not seem to be able to reproduce until they reach 8-10 years of ag e (Ferguson, 1985). To control for potential confounding due to differential growth rates, relationships between female size and OCP burdens and clutch viability will be evaluated using site and year as covariates. If the effects of covariates are determined statistically negligible, female data will be grouped together.

PAGE 98

82 Site Descriptions Lakes Apopka (N 28 35Â’, W 81 39Â’), Griffin (N 28 53Â’, W 81 49Â’), and Lochloosa (N 29 30Â’, W 82 09Â’) in Florida were selected as co llection sites because prior studies by our laboratory indicate vastly different le vels of OCP exposure across these sites. All three lakes are part of th e Ocklawaha Basin. Lake Lochloosa (which is connected to Orange Lake) was selected as a low exposure (reference) site. Three years (2000-2002) of data indicate mean total OCP concentrations in egg yolks from the reference sites (Lakes Orange and Loch loosa) were 102 16 ppb (mean standard deviation [SD], n = 19 clutches) with a concurre nt mean clutch viability rate (number of live hatchlings/total number of eggs in a ne st) of 70 4% (Gross, unpublished data). Lake Griffin was selected as an intermedia te exposure site since yolk concentrations averaged 1169 423 ppb (n = 42 clutches) a nd Lake Apopka was selected as a high exposure site since yolk concentrations av eraged 7,582 2,008 ppb (n = 23) for the same time period (Chapter 2). Furthermore, mean clutch viability rates during this time period for Lakes Apopka (52 6%, n = 23) and Griffin (43 5%, n = 42) have been below rates observed for the reference site. Animal Collections Adult female alligators and their corres ponding clutches of eggs were collected from Lakes Apopka ( n = 19), Griffin ( n = 18), and Lochloosa ( n = 3) over the course of four nesting seasons (June 1999 to June 2002) Nests were located by aerial survey (helicopter) and/or from the ground (airboat). Once nests were located, all eggs were collected, and the nest cavity was covered. A snare-trap was set perpendicular to the taildrag in order to capture the female as she cr ossed over the nest. After the traps were set, one member of the trapping crew subsequently transported the eggs to the Florida Fish

PAGE 99

83 and Wildlife Conservation Commission’s Wild life Research Unit (FWC; Gainesville, FL, USA) and placed the eggs in a temperaturecontrolled incubator. Snare-traps were checked later in the evening a nd early the next morning. In 1999 and 2000, trapped females were secu red and measurements (total length, snout-vent length, head length and tail girth) were collec ted along with a blood sample and a scute for OCP analysis. These females were then immediately released. In 2001 and 2002, females were captured and transporte d from each lake to the United States Geological Survey’s Florida In tegrated Science Center (USGS; Gainesville, FL, USA). Upon arrival, the animals were weighed, m easured, and blood samples were collected from the post-occipital sinus. Adult alligators were then euthanized by cervical dislocation followed by double pithing. A full necropsy was performed on each female. Bile, liver, adipose (composite of abdomina l fat and the abdominal fat pad), and tail muscle samples were collected for later dete rmination of OCP burdens. Liver, adipose tissue, and muscle were wrapped in aluminum foil, while bile and blood were placed in scintillation vials. All sa mples were grouped according to nest identification number (ID), placed in plastic bags labeled with the appropri ate ID, and stored in a –80 oC freezer. Each female’s corresponding clutch of eggs was then transferred from FWC to USGS where yolk samples were collected (t wo eggs/clutch) and stored with the corresponding maternal tissues. The remain ing eggs were set for incubation in a temperature/humidity-contr olled incubator (31-33 oC, 88-92% relative humidity) located at USGS. Analysis of OCPs in Ma ternal Tissues and Yolk Analytical grade standards for the following compounds were purchased from the sources indicated: aldrin, al pha-benzene hexachloride ( -BHC), -BHC, lindane, -BHC,

PAGE 100

84 p,pÂ’ -dichlorodiphenyldichloroethane ( p,pÂ’ -DDD), p,pÂ’ -dichlorodiphenyldichloroethylene ( p,pÂ’ -DDE), dichlorodiphe nyltrichloroethane ( p,pÂ’ -DDT), dieldrin, endosulfan, endosulfan II, endosulfan sulfate, endrin, e ndrin aldehyde, endrin ketone, heptachlor, heptachlor epoxide, hexachlorobenz ene, kepone, methoxychlor, mirex, cis -nonachlor, and trans -nonachlor from Ultra Scientific (Kingstown, RI, USA); cis -chlordane, trans chlordane, and the 525, 525.1 polychlorinated biphenyl (PCB) Mix from Supelco (Bellefonte, PA, USA); oxychlordane from Chem Service (West Chester, PA); o,pÂ’DDD, o,pÂ’DDE, o,pÂ’DDT from Accustandard (New Haven, CT, USA); and toxaphene from Restek (Bellefonte, PA, USA). All reag ents were analytical grade unless otherwise indicated. Water was doubly distilled and deionized. Adipose, liver, bile, and yolk samples were analyzed for OCP content using methods modified from Holste ge et al. (1994) an d Schenck et al. (1994). For extraction, a 2 g tissue sample was homogenized with ~1 g of sodium sulfate and 8 mL of ethyl acetate. The supernatant was decanted and filtered though a Bchner funnel lined with Whatman #4 filter paper (Fisher Scientific, Ha mpton, NH, USA) and filled to a depth of 1.25 cm with sodium sulfate. The homogenate was extracted twice with the filtrates collected together. The combined filtrat e was concentrated to ~2 mL by rotary evaporation, and then further concentrated until solvent-free unde r a stream of dry nitrogen. The residue was reconstituted in 2 mL of acetonitrile. After vortexing (30 s), the supernatant was applied to a C18 solid phase extrac tion (SPE) cartridge (preconditioned with 3 mL of acetonitrile; Agile nt Technologies, Wilmi ngton, DE, USA) and was allowed to pass under gravity. This proced ure was repeated twice with the combined eluent collected in a culture t ube. After the last addition, th e cartridge was rinsed with 1

PAGE 101

85 mL of acetonitrile which was also collected. The eluent was then applied to a 0.5 g NH2 SPE cartridge (Varian, Harbor City, CA, US A), was allowed to pass under gravity, and collected in a graduated conical tube. The car tridge was rinsed with an additional 1 mL portion of acetonitrile whic h was also collected. The combined eluents were concentrated under a stream of dry nitrogen, to a volume of 300 L, and transferred to a gas chromotography (GC) vial for analysis. Whole blood was analyzed for OCP c ontent using methods modified from Guillette et al. (1999). A 10 mL aliquot was transferred from the homogenized bulk sample and extracted in 15 mL of acetone by vortex mixer. The mixture was centrifuged for 5 min at 3000 rpm, after which the supernatan t was transferred to a clean culture tube. This process was repeated with the supern atants collected and concentrated under a stream of dry nitrogen until solvent-free. Th e residue was re-extract ed in 11.5 mL of 1:1 methylene chloride-petroleum ether. After mixing, the sample was allowed to settle and the upper layer was tran sferred to a clean culture tube. This extraction was performed twice with the extracts collected together. Th e combined extracts we re then applied to a prepared florisil cartridge (5 mL Fisher Pr epSep, Fisher Scientific, Hampton, NH, USA). The cartridge had been prepared by filling the reservoir to a depth of 1.25 cm with anhydrous sodium sulfate and by prewashing the modified cartridge with 10 mL of 2:1:1 acetone: methylene chloride: petroleum ether. After the sample passed under gravity with the eluent collected in a 15-mL graduate d conical tube, the cart ridge was eluted with 4 mL of the 2:1:1 solv ent mixture which was also collected. The combined eluents were concentrated under a stream of dry nitrogen, to a volume of 300 L, and transferred to a GC vial for analysis.

PAGE 102

86 GC/MS Analysis Analysis of all samples was performed using a Hewlett Packard HP-6890 gas chromatograph (Wilmington, DE, USA) with a split/splitless inlet ope rated in splitless mode. The analytes were introduced in a 1 L injection and separa ted across the HP-5MS column (30 m x 0.25 mm; 0.25 m film thickne ss; J & W Scientific, Folsom, CA, USA) under a temperature program that began at 60 C, increased at 10 C/min to 270 C, was held for 5 min, then increased at 25 C/min to 300 C and was held for 5 min. Detection utilized an HP 5973 mass spectro meter in electron impact m ode. Identification for all analytes and quantitation for toxaphene was c onducted in full scan mode, where all ions are monitored. To improve sensitivity, se lected ion monitoring was used for the quantitation for all other analytes, except kepone. The above program was used as a screening tool for kepone which does not optim ally extract with mo st organochlorines. Samples found to contain kepone would be reex tracted and analyzed specifically for this compound. For quantitation, a five-point standard curve was prepared for each analyte ( r2 0.995). Fresh curves were analyzed with each se t of twenty samples. Each standard and sample was fortified to contain a deuterat ed internal standard, 5 L of US-108 (120 g/mL; Ultra Scientific), added just prior to analysis. All samples also contained a surrogate, 2 g/mL of tetrach loroxylene (Ultra Scientific) added after homogenization. Duplicate quality control samples were prepar ed and analyzed with every twenty samples (typically at a level of 1.00 or 2.50 g/mL of -BHC, heptachlor, aldr in, dieldrin, endrin, and p,p’ -DDT) with an acceptable recovery rangi ng from 70 – 130%. Limit of detection (LOD) ranged from 0.1-1.5 ng/g for all OCP an alytes, except toxa phene (120-236 ng/g), and limit of quantitation (LOQ) was 1.5 ng/g for all analytes, except toxaphene (1500

PAGE 103

87 ng/g). Repeated analyses were conducted as allowed by matrix interferences and sample availability. Data Analysis Specific OCP analytes were removed from analysis if measurable concentrations were not found in at least 5% of the clutch es. Numerical data were log-transformed [ln(x)], while proportional da ta were arcsine square root transformed to conform to statistical assumptions. Maternal OCP bur dens for females collected during 1999 and 2000 were estimated using the femalesÂ’ resp ective yolk burdens and predictive models described in Chapter 3. ANOVA (PROC GLM; SAS Institute Inc., 2002) was used for inter-site comparisons of adult female and clutch characteristics, and the Tukey test was used for multiple comparisons among sites ( = 0.05 since no interactions were tested). Because relationships between respons e variables and explanatory variables (Table 4-1) in ecological studies are often complex with inte ractions occurring, an indirect gradient multivariate analysis method, Detrended Co rrespondence Analysis (DCA) (ter Braak, 1986) was used to initially evaluate data stru cture. Two matrices were constructed for DCA, with the first representing the respons e variables (female-clutch pair number x clutch parameters) and the second representi ng the explanatory vari ables (female-clutch pair number x maternal size and OCP burdens) (Table 2). DCA results indicated that a direct gradient, multivariate linear anal ysis, Redundancy Analysis (RDA) (Rao, 1964), was appropriate since the lengths of the DCA ordination axes were equal to or less than 2 standard deviations (t er Braak, 1995). For the RDA, sim ilar matrices were constructed with the exception that response variables meas ured as a percentage (i.e., clutch viability) and response variables measured as a number (i .e., clutch mass) were split into separate

PAGE 104

88 matrices because percentage data were ln(x +1) transformed and not standardized, while continuous data were ln(x) transformed and st andardized (ter Braak & Smilauer, 2002). Automatic forward selection of the best f our explanatory variables was conducted for both sets of RDA analyses and tested for significance by Monte Carlo permutation test DCA and RDA were conducted using the pr ogram CANOCO (ter Braak & Smilauer, 2002). Biplots of environmental variables a nd response variables we re then constructed to facilitate interpretation. Results A total of 40 female alligators and their respective clutches (female-clutch pairs) were collected during the summers of 1999-2002 from Lake Apopka ( n = 19), Lake Griffin ( n = 18), and Lake Lochloosa ( n = 3). No significant di fferences between lakes were determined with respect to fecundity, av erage egg mass, clutch viability, percentage of unbanded eggs, percentage of early embryoni c mortality, percentage of late embryonic mortality, female head length, female snout-v ent length, female tail girth, female total length, and female body condition index (Table 4-3). Significant differences were detected with respect to total OCP concentration in female adipose tissue with Lake Apopka female burdens (22,737 5,767.6 ng/g) being greater than those of Lake Griffin (1,821 702.7 ng/g), and Lake Lochloosa female burdens (375 63.1 ng/g). No significant differences were determined betw een adipose burdens of Lake Griffin and Lake Lochloosa females (Table 4-3). Results of the forward selection RDA ev aluating reproductive efficiency (clutch viability, percentage unbanded eggs, early em bryo mortality and la te embryo mortality) indicated that the four explanat ory variables that best accounted for the variance of clutch survival parameters were interaction variable: % trans -chlordane (TC%) (lambda A =

PAGE 105

89 12%), percentage p pÂ’ -DDE (DDE%) (lambda A = 6%), heptachlor epoxide ([HE]) (lambda A = 4%), and percentage oxychlordane ( A = 4%). These factors accounted for 28% of the total variation of reproductive efficiency. Two of the four variables, TC% and PDDE% were determined significant a nd together accounted for 18% of total variation (Table 4-4). Biplots of the extracted fact ors and reproductive variables (Fig. 4-1) suggested that clutch viability had strong ne gative correlations with TC% ( r = -0.5451) and %OX-[OX] ( r = -0.2885), but was weakly correlated with PDDE% ( r = 0.0548). Unbanded egg percentages, however, were positiv ely strongly correlated with DDE% ( r = 0.4012), weakly correlated with %OX-[OX] ( r = -0.1298) and TC%-[TC] ( r = 0.0228). Early embryonic mortality percentages were weakly correlated with TC%-[TC] ( r = 0.1860), %OX-[OX] ( r = 0.1556), and PDDE% ( r = 0.133). Late embryo mortality percentages were positively correlated with TC%-[TC] ( r = 0.3361) and %OX-[OX] ( r = 0.3144), but showed negative correl ations with DDE% ( r = -0.3230). For clutch size characteristics (fecundity, clutch mass, and average egg mass), RDA results indicated that the four explanatory variables that best explained clutch size variance were concentration of cis-chlordane ([CC]) (L ambdaA = 6%), (LambdaA = 6%), percentage dieldrin (DL%) (LambdaA = 6%), concentration of p,p Â’-DDD ([PDDD]) (LambdaA = 4%), and concentration of toxa phene ([TOX]) (LambdaA = 4%). None of these variables were determined to be significan tly associated with clutch size parameters (Table 4-5). Discussion With respect to the first hypothesis, results of the present study suggest that certain OCPs in maternal adipose tissue were signifi cantly associated with decreased clutch

PAGE 106

90 survival parameters (clutch viability, per centage, unbanded eggs, ear ly and late embryo mortality), but that clutch survival parameters were neither significantly correlated with maternal morphometrics. Although maternal bur dens of certain OCPs were correlated with clutch survival parameters, extracted va riables only explained 18% of the variation. However, it is important to note that compos itional percentage of an OCP analyte appears to be an important factor with respect to clutch survival parame ters. Indeed both OCP variables found to be significantly associated with clutch survival parameters were compositional variables (TC% and pDE%) (T able 4-4). This suggests that the composition of the OCP mixture may be more im portant in altering clutch viability than the total OCP burden or total number of OCPs de tected in maternal tissues. Furthermore, biplots (Fig. 4-1) suggest that the rates of unbanded eggs, early embryo mortality, and late embryo mortality, which all contribute to reproductive efficien cy, have different relationships with the each of the extract ed OCP variables, suggesting that certain mixtures differentially affect certain aspects of reproducti ve function. Two reasons for the importance of mixture com position are that the effects and the toxicity (potency) of OCPs vary considerably from analyte to anal yte. For example, Japanese quail orally dosed with technical grade DDE (300 ppm) had decreased rates of fe rtility and increased rates of mortality; however, similar exposures to technical grade DDT (300 ppm) did not cause adverse effects (Robson et al., 1976). D ecreased fertility in quail exposed to DDE is consistent with the results of the presen t study in that as DDE% increased in female alligator adipose tissue, the percentage of unbanded eggs also increased. Although total OCP burdens in maternal ad ipose tissue were significantly higher in females from Lake Apopka, no significant diffe rences were detected between lakes with

PAGE 107

91 respect to clutch viability, percentages of unbanded eggs, early embryo mortality, or late embryo mortality,. The significant differences in total OCP burdens among sites and lack thereof for clutch parameters, again sugge sts that total OCP burdens may be less important than mixture composition. With respect to the second hypothesis, OCP burdens and maternal morphometrics were not found to be associated with clutch size characteristics. This may be due to similar sized females and clutches being collect ed with in and among sites (Table 4-3). The correlations between certain OCPs a nd clutch survival parameters suggest decreased reproductive efficiency may be re lated to increased maternal OCP burdens, however, correlations alone do not establis h causal relationships. Indeed, several viewpoints must be considered before cau sality is concluded, however, the only viewpoint, or “criterion”, that can rule out a cause-effect relationship is temporality (i.e., exposure must precede effect ) (Hill, 1965). The major cr iteria used in the current practice of causal inference is temporality, biological plau sibility, consistency of the association, strength of th e association, and biological gradient, (Weed et al., 2002; Gadbury & Schreuder, 2003). Since females were exposed to OCPs prior to vitellogenesis and oviposition, temporality is satisfied. The second criteri on, biological plausibilit y, is also satisfied since studies in other oviparous vertebrate s have shown that OC P exposure can cause adverse reproductive effects through a variety of mechanisms (Fry, 1995) The third criterion is consistency of asso ciation, which means that similar results have been found among other studies examini ng the same problem. Few studies have examined reproductive effects of maternal OC P exposure in alligators with one of these

PAGE 108

92 studies reporting no significan t correlations between DDE concentrations in maternal tissues and clutch anomalies for Lake A popka alligators (Giroux, 1998). The earlier studyÂ’s focus was on a single analyte while this study looked a several analytes concurrently, so the two studies differ some what. Given relatively small number of studies and the ambiguous interpretations, no clear conclusion can be reached as to whether the consistency criter ion has been satisfied. The fourth and fifth criteria, strength of association and biological gradient, are somewhat similar in nature. Strength of association refers to how strongly correlated the causal factor is to the respons e variable, and the biological gr adient refers to whether the response variable increases as the dose increases. With resp ect to strength of association, the present studyÂ’s results indicate weak-mode rate associations between maternal OCP burdens and clutch survival parameters (18% of variance explained). With respect to dose-response, biplots indicat ed that biological gradient s existed between certain maternal factors and reproductive responses in that as percentage p,pÂ’-DDE and percentage trans-chlordane in creased, incidence of unbanded eggs increased and clutch viability decreased, respectively (Fig. 4-1). In summary, rarely does a single obser vational study estab lish clear causal relationships, and the presen t study is no exception. Howe ver, the present study does satisfy some of the criteria used for establ ishing causality. Importa ntly, results suggest that a moderate part of th e variation associated with reproductive function in the American alligator can be attributed to maternal OCP body burdens. Hopefully, the results of the present study will stimulat e future efforts aimed at increasing our understanding of the effects of enviro nmental contaminants of crocodilians.

PAGE 109

93 Table 4-1. Reproductive, mor phometric, and contaminant pa rameters measured on adult female alligators collected during June 1999, 2000, 2001, and 2002. Female Parameter Definition Measured as Response variables Fecundity Total No. of eggs in one clutch n Clutch mass Total mass of eggs in one clutch kg Ave. Egg Weight Clutch mass / Fecundity g % Unbanded eggsa No. of unbanded eggs / fecundity x 100 Percentage % Early embryo mortality No. of deaths < dev. Day 35 / fecundity x 100 Percentage % Late embryo mortality No. of deaths dev. Day 35 / fecundity x 100 Percentage Clutch Viability No. eggs yielding live hatchling / fecundity x 100 Percentage Explanatory variables Head Length Tip of snout to posterior base of skull (dorsal) cm Snout-Vent Length Tip of snout to posterior base of vent (dorsal) cm Tail Girth (cm) Circumferen ce of tail at vent cm Total Length (cm) Tip of snout to tip of tail (dorsal) cm Body condition index Snout-vent lengt h / Tail girth x 100 Percentage [OCP analyte] in adipose tissueb ng OCP analyte / g adipose tissue wet weight ppb % OCP analyte [OCP analyte] / [OCP] x 100 Percentage aAn egg with no evidence of embryonic attachment bSee text for list of measured OCP analytes. For 1999 and 2000 females, adipose OCP concentrations were estimated using pr edictive equations (see Chapter 3).

PAGE 110

94 Table 4-2. Explanatory variab les included in RDA with forw ard selection of four best variables ( = 0.05). Variable Code Head Length HL Snout-vent length SVL Total Length TL Tail Girth TG Body Index BI Age Age No. OCPs quantitated NOC [OCP] TOC % cis -Chlordane CC% [ cis -Chlordane] [CC] % cis -Nonachlor CN% [ cis -Nonachlor] [CN] % Dieldrin DL% [Dieldrin] [DL] % Heptochlor epoxide HE% [Heptachlor epoxide] [HE] % Mirex MX% [Mirex] [MX] % o,p -DDT ODDT% [ o,p -DDT] [ODDT] % Oxychlordane OX% [Oxychlordane] [OX] % p,p '-DDE PDDE% [ p,p '-DDE] PDDE] % p,p '-DDD PDDD% [ p,p '-DDD] [PDDD] % p,p '-DDT PDDT% [ p,p '-DDT] [PDDT] % trans -Chlordane TC% trans -Chlordane [TC] % trans -Nonachlor TN% [ trans -Nonachlor] [TN] % Toxaphene TX% [Toxaphene] [TX]

PAGE 111

95Table 4-3. Reproductive, morphometric and contaminant summary statisticsa of adult female alligators collected during June of 1999-2002. Parameter Apopka Griffin Lochloosa Summary Female-clutch pairs (n) 19 18 3 40 Fecundity (n) 45 2 44 1.9 49 3.5 44 1.3 (22–54) (19–56) (45–56) (19–56) Clutch mass (kg) 4.2 0.14 3.6 0.18 4.3 0.26 3.9 0.11 (2.4–5.1) (1.8–5.2) (4–4.8) (1.8–5.2) Egg Mass (g) 89 1.2 84 2.8 88 1.1 87 1.4 (77.6–100) (70.8–112.6) (86.1–90) (70.8–112.6) Clutch 'viability (%)b 61 6.8 42 8.5 60 8.3 52 5.2 (0–98) (0–92) (48–76) (0–98) Unbanded eggs (%)c 19 6.9 21 5.3 9 5.5 19 4 (0–100) (0–70) (3–20) (0–100) Early embryo mortality (%)d 12 3 14 3.4 25 3.2 14 2.1 (0–45) (0–52) (20–31) (0–52) Late embryo mortality (%)e 7 2 22 6.8 6 3 14 3.4 (0–25) (0–89) (0–10) (0–89) Head Length (cm) 37 1.3 36 0.7 35 0.4 36 0.7 (22–52) (28–41) (35–36) (22–52) Snout-Vent Length (cm) 140 3.6 135 1.9 129 2.8 137 2 (83–156) (120–148) (125–134) (83–156) Tail Girth (cm) 68 2.8 66 1.7 62 2 67 1.5 (36–92) (52–77) (59–66) (36–92) Total Length (cm) 263 7.9 260 4.5 258 4.2 262 4.2 (161–304) (220–298) (253–266) (161–304) Body indexf 2.09 0.059 2.05 0.04 2.07 0.023 2.07 0.033 (1.46–2.7) (1.79–2.38) (2.03–2.11) (1.46–2.7) TotalOCPs (ng/g)g 22,734 5767.6 A 1,821 702.7 B 375 63.1 B 11,648 3,200.9 (5,224.7–123,081.5) (355.6–12,938.7) (289.8–498.1) (289.8–123,081.5)

PAGE 112

96 96aValues represent mean standard error with ra nge in parentheses. Significant differences ( = 0.05) between sites indicated by letters (A-B) beside mean. Same letters = not significant b % of eggs in a clutch th at yield a live hatchling. c % of eggs with no eviden ce of embryonic attachment d % of embryos in a clutch that perish duri ng first half (35 days) of development. e % of embryos in a clutch that perish dur ing last half (35 days) of development. f snout-vent length / tail girth g [ ng OCP analyte / g adipos e tissue (wet weight)]

PAGE 113

97 Table 4-4. Results of redundancy analysis with automatic selection of four best maternal factors associated with varia tion in reproductive efficiency. Variable LambdaAa P F TC%* 0.12 0.004 5.18 pDE%* 0.06 0.04 2.78 [HE] 0.04 0.202 1.65 Lochloosa 0.04 0.146 1.80 aProportion of total variance explained by each variable (total variance = 1.0). P < 0.05 Table 4-5. Results of redundancy analysis with automatic selection of four best maternal factors associated with variation in clutch size characteristics. Variable LambdaA P F [CC] 0.06 0.174 2.21 [pDD] 0.04 0.190 1.94 [TOX] 0.04 0.162 1.49 [DL] 0.06 0.134 2.54

PAGE 114

98 -1.01.0-1.01.0 CLutch viability Unbanded egg% Early Emb. Mortality Late Emb. Mortality DDE% trans-Chlordane% [HE] Lochloosa Figure 4-1. Biplot of maternal factors (das hed lines) and clutch survival parameters (solid lines) of American alligators collected during June 1999-2002. Arrows pointing in the same direction indi cate a positive correlation (e.g., unbanded egg% and DDE%), arrows that are approxi mately perpendicular indicate nearzero correlation (e.g., clutch viability and DDE%), and arrows pointing in opposite directions indicate negative co rrelations (clutch viability and transchlordane%]. Arrow lengths indicate ra nk order of correlatio ns. For example, extending a perpendicular line from the early emb. mort. axis to tip of transchlordane% arrow indicates that early emb. mort. and trans-chlordane have a stronger positive correlation compared to early emb. mort. correlation and [HE]. The cosine of the angle formed at the origin between individual clutch variables and individual OC P variables is the correlation coefficient (r). For example, arrows pointing in exactly oppos ite directions have an angle of 180, and since cos(180) = -1.0, the arrows are perfectly negatively correlated (r) (ter Braak, 1995).

PAGE 115

CHAPTER 5 MORPHOLOGY AND HISTOPATHOLOGY OF AMERICAN ALLIGATOR ( ALLIGATOR MISSISSIPPIENSIS ) EMBRYOS FROM REFERENCE AND OCPCONTAMINATED HABITATS In central Florida, American alligators living in habitats contaminated with organochlorine pesticides (OCPs) have poor reproductive success in comparison to populations inhabiting reference sites (Wood ward et al., 1993) (Wiebe et al., 2001). Decreased reproductive efficiency has been largely attributed to increased rates of early embryo mortality (mortality occurring first 35 days of development) and, to a lesser extent, late embryo mortality (mortality afte r day 35), as well as increased incidence of unbanded eggs, with unbanded eggs likely being a product of in fertility, or preovipositional embryo mortality, or a combin ation of both (Masson, 1995; Wiebe et al., 2001; Rotstein et al., 2002). A clear dose-response relationship betw een embryo mortality and total OCP burdens in eggs has not been established (H einz et al., 1991), and recent studies, on Lake Apopka, suggested poor egg viab ility was more closely associated with muck farm reclamation (wetland restorati on) sites than with tissue a nd egg concentrations of the predominant pesticide residue (DDE) (Giroux, 19 98). Muck farming typically refers to a farming practice where a dike is built around a marshy area adjacent to a lake, then the water is pumped out of the marsh, and the fert ile peat (i.e., “muck”) is then used for crop production. In addition, altered endocrine f unction and decreased egg viability were documented among alligators at another si te, Lake Griffin, where tissue and egg concentrations of OCP residue s such as DDE are intermediate in comparison with Lake

PAGE 116

100 Apopka, but Lake Griffin is also highly eu trophic and has adjacent muck farms and muck farm reclamation areas. Although a clear dose-response relationship ha s not been established with respect to embryo mortality and total organochlorine pesticide burdens, great differences exist, nonetheless, between sites with respect to OCP egg burdens and OCP constituent composition, suggesting mixture composition may pl ay a role or even be more important than simple cumulative OCP burdens. OCPs are of concern because they are prevalent and persistent environmental contaminants that are lipid soluble, resistan t to metabolic degradation, bioaccumulate in animal tissues, and may cause altered function of the immune system, as well as neural toxicity (Blus, 1996). Furthermore, in vitro and in vivo experiments using laboratory organisms, as well as epidemiology studies involving OCP-exposed human and wildlife populations, suggest a variety of OC Ps and OCP metabolites, such as dichlorodiphenyltrichloroethane (DDT), dichlorodiphenyltrichloroethylene (DDE), methoxychlor, dicofol, chlordane, dieldri n, and toxaphene, may be associated with disrupted endocrine function and altered embryo development (Colborn et al., 1993; Gray et al.; 1997; Fairbrother et al., 1999; Longnecker et al., 2002; Rattner & Heath, 2003). Although increased early embryonic mortality and late embryo mortality have been documented, few histopathology studies have ex amined live, moribund, or dead alligator embryos to determine whether alterations in morphology and/or spec ific pathogenicities are associated with increased mortality rates or specific OCPs and OCP burdens in eggs. Such histopathology studies are arduous due to rapidity of tissue autolysis, confounded by the difficulty in determining whether an embryo is alive or dead. Determining

PAGE 117

101 whether an embryo is alive or dead, when it is still inside the egg, is difficult because bright light candling is currently the onl y practical method for determining embryo viability in large studies involving thousands of eggs. Using bright light candling, live embryos are differentiated from dead embryos based on the color of the illuminated egg, with bright red indicating a live embryo (or very recently dead) and orange-pink indicating a dead embryo (color changes may be related to breakdown of red bloods cells and general autolysis of egg and embryo membranes). Although difficult, examining morphologica l development and histopathology of live and dead embryos would aid in understand ing the causes and mechanisms associated with the embryo mortality. For example, histopathology may indicate occurrence of acute chemical toxicity since it is known that many types of pesticides, including OCP compounds, induce toxicopathic lesions in vita l organs, with liver being the predominant target organ (Metcalfe, 1998). Furthermor e, evaluating how changes in morphology and histopathology relate to clutch mortality ra tes and OCP egg burdens may provide insight as to whether OCPs play a role in the in creased incidence of em bryo mortality observed in OCP-contaminated lakes. Therefore, the objective of the presen t study was to evaluate embryo morphology as a function of embryo conditi on (live/dead), lake of origination, clutch quality, and OCP egg burden, and to evalua te the histopathology of em bryos from clutches with diverse OCP egg burdens and mortality rates. To accomplish this objective the following hypotheses will be tested. Firs t, morphological development of live alligator embryos is different from dead embryos of the same chronological age. Second, morphological development of live embryos of the same ch ronological age is different among reference

PAGE 118

102 and OCP-contaminated sites. Third, morphol ogy of live embryos from clutches with low mortality rates is different from those of cl utches with high mortality rates, and fourth, variation in morphological deve lopment of live embryos is associated with composition and/or concentration of OCPs in eggs. Lastly, histopathology of live embryos from clutches with low mortality rates and low OCP egg burdens is different from those of clutches with high mortality rates and high OCP egg burdens. Materials and Methods Site Descriptions Lakes Apopka (N 28 35Â’, W 81 39Â’), Griffin (N 28 53Â’, W 81 46Â’), Emeralda Marsh Conservation Area ((N 28 55Â’, W 81 47Â’), and Lochloosa (N 29 30Â’, W 82 09Â’) in Florida were selected as collection sites because prior studies by our laboratory indicate vastly different levels of OCP expos ure across these sites. All three lakes are part of the Ocklawaha Basin. Lake Lochloos a (which is connected to Orange Lake) was selected as a low exposure (reference) site. Four years (2000-2002) of data indicate mean total OCP concentrations in egg yolks from the reference sites (Lake Lochloosa) were 102 15 ppb (mean standard error [SE], n = 19 clutches) with a concurrent mean clutch viability rate (number of live hatchli ngs/total number of eggs in a nest) of 70 4% Lake Griffin was selected as an intermed iate exposure site sin ce yolk concentrations averaged 1,169 423 ppb ( n = 42 clutches) and Lake A popka was selected as a high exposure site since yolk concen trations averaged 7,582 2,008 ppb ( n = 23) for the same time period (Gross, unpublished data). Furthe rmore, mean clutch viability rates during this time period for Lakes Apopka (51 6%, n = 23) and Griffin (44 5%, n = 42) have been below rates observed for the reference site.

PAGE 119

103 Egg Collections In the field, clutches were located vi a aerial surveys (helicopter) and ground surveys (airboat). Each clutch was provide d with a unique identif ication number, and immediately transported in plastic pans (43 cm x 33 cm x 18 cm) c ontaining the original nest substrate material to the US Geologi cal SurveyÂ’s Center for Aquatic Resources Studies, Gainesville, Florida (CARS). Upon a rrival, complete clutches were evaluated for embryonic viability using a bright light ca ndling procedure. Viable eggs (i.e. having a visible band) were nested in pans cont aining moist sphagnum moss and incubated at 30.5C and ~98% humidity, in an incubation building (7.3 m x 3.7 m). This intermediate incubation temperature will normally result in a 1:1 male/female sex ratio. One or two eggs were sacrificed from each clutch to identify the embryonic stage of development at the time of collection, and to evaluate the concentrations of OCPs in yolk. From each clutch, information on the following paramete rs was collected: total number of eggs found per nest (fecundity); number of unbande d eggs, number of damaged eggs, number of dead banded eggs, number of live bande d eggs, total clutch mass and average egg mass of clutch. Then, each clutch was evenly divided between two pans, with half of the clutch left relatively undisturbed (except fo r weekly monitoring of embryo viability) to determine clutch viability (the number of live hatchlings / fe cundity), and the other half of the clutch used to study embr yo development and morphometry. Embryo Sampling and Measurement After initial determination of morphological ages (MA) for all clutches (Ferguson, 1985), 2-4 live embryos were collected from each clutch at each of four selected chronological ages. Morphol ogical age (MA) refers to the age of the embryo as determined by level of morphological developm ent and chronological ag e (CA) refers to

PAGE 120

104 the calendar age when an embryo was sampled. For example, two clutches are initially examined and it is determined that embryos of clutch A are of MA Day 12 and those of clutch B are of MA Day 10. W ith respect to the initial ag e determination, it is assumed that MA = CA. Furthermore, if embryos of both clutches are to be sampled at CA age14, then clutch A would be sampled two days after initial examination and Clutch B would be sampled 4 days after initial examina tion. MA can then be determined (based on morphological features), and can be compar ed to CA to see if actual morphological development (MA) differs from what would be expected, given the particular CA. The four chronological ages sampled were Day 14, Day 25, Day 33, Day 43. These ages were selected because each are clearly distingui shable from other ages, and provide a good representation of progression of organogenesi s and growth (Ferguson, 1985). These ages also correspond to periods of increased em bryo mortality, as determined by previous studies (Masson, 1995). The following parameters were measured on fresh embryos (live and dead): egg mass; embryo condition (live or dead), em bryo mass, embryo morphological age, eye length, head length, and total le ngth of embryo. Dead embryo s were differentiated from live embryos based on the lack of a visible ca rdiac contractions and signs of autolysis, such as atypical coloration and loss of tissue integrity. Other parame ters were derived in an attempt to determine more subtle differences in development. These derived parameters included the following ratios: eye leng th / head length (Eye L.: Head L.); head length / total length (Head L.: Total L.); a nd total length / embryo mass (Total L.: Emb. M.).

PAGE 121

105 Morphological age was determined using Ferguson’s (1985) crocodilian developmental staging scheme. Fresh em bryos were photographed with either a Olympus model DPII digital cam era mounted to a Zeiss model Stemi SV 6 dissecting scope (for embryos of age Day 9 or less) or with a Canon EOS D30 digital camera mounted to a photographer’s stand (for embryos of age Day 10 or greater) (Fig. 5-1). Embryos were measured from digitized photographs using an image analysis software program, SigmaScan Pro (Systat Inc., 1999). After being photographed, embryos were fixed in formalin and stored in labeled containers for histopathology. Histopathology Subsamples of live embryos from “best case” clutches (clutch viability > 71%, which is equal to overall mean clutch viabi lity + 1 standard deviation) and low total OCP egg burdens ( 350 ng/g) and live embryos from “wor st case” clutches (< 47%) and high total OCP egg burdens (i.e., 3,700) were selected and pr ocessed for histopathology. Comparing best case to worst case provided the best opportunity for determining if differences existed with respect to frequency of lesions and identifyi ng target organs and tissues. If large differences were found be tween embryos of best case and worst case clutches, subsequent examin ations could be conducted on embryos of intermediate quality clutches. Conversely, if no differences were found, it would be unlikely to detect differences in intermediate quality clutches ; therefore, subsequent examinations would not be warranted. Embryos were cross-sectioned, and then f our equidistance step-sections were taken from each of the following regions: the hea d, the thorax, and the abdomen. For each of the 7 m sections, distances between st ep-sections ranged from ~42-300 m, depending on the age of the embryo, with inter-sectiona l distances increasing with embryo size.

PAGE 122

106 Sections were mounted to slides and staine d with hematoxylin and eosin. Slides were screened for lesions, the type of lesion present, and the organ or tissue involved. Expected morphological changes due to chronic OCP exposure include hepatocellular hypertrophy and focal necrosis. Hypertrophy is due to enlargement of the smooth endoplasmic reticulum (SER) and formati on of a lipid droplet in the center of the SER (caused by OCP-induced expression of microsomal enzymes within the SER) (Smith, 1991). Other morphological changes found in the liver include foci of vacuolated hepatocytes and spongiosis hepatic (lesions of hepatic parenchyma). Renal lesions induced by chronic OCP exposure include dilati on of tubular lumina, and vacuolization (degeneration) and necrosis of tubul ar epithelium (Metcalfe, 1998). Other than hepatic and renal toxicopath ic lesions, OCPs may cause death by disrupting neural transmission to the point of cardiovascular failure. Neural morphology is rarely changed by OCP exposure, which causes difficulty in determining whether cardiovascular failure was caused by OCP exposure or some other factor. Analysis of OCPs in Yolk Analytical grade standards for the following compounds were purchased from the sources indicated: aldrin, al pha-benzene hexachloride ( -BHC), -BHC, lindane, -BHC, p,pÂ’ -dichlorodiphenyldichloroethane ( p,pÂ’ -DDD), p,pÂ’ -dichlorodiphenyldichloroethylene ( p,pÂ’ -DDE), dichlorodiphe nyltrichloroethane ( p,pÂ’ -DDT), dieldrin, endosulfan, endosulfan II, endosulfan sulfate, endrin, e ndrin aldehyde, endrin ketone, heptachlor, heptachlor epoxide, hexachlorobenz ene, kepone, methoxychlor, mirex, cis -nonachlor, and trans -nonachlor from Ultra Scientific (Kingstown, RI, USA); cis -chlordane, trans chlordane, and the 525, 525.1 polychlorinated biphenyl (PCB) Mix from Supelco (Bellefonte, PA, USA); oxychlordane from Chem Service (West Chester, PA); o,pÂ’-

PAGE 123

107 DDD, o,pÂ’DDE, o,pÂ’DDT from Accustandard (New Haven, CT, USA); and toxaphene from Restek (Bellefonte, PA, USA). All reag ents were analytical grade unless otherwise indicated. Water was doubly distilled and deionized. Egg yolk samples were analyzed for OC P content using methods modified from Holstege et al. (1994) and Sc henck et al. (1994). For extraction, a 2 g tissue sample was homogenized with ~1 g of sodium sulfate a nd 8 mL of ethyl acetate. The supernatant was decanted and filtered t hough a Bchner funnel lined with Whatman #4 filter paper (Fisher Scientific, Hampton, NH, USA ) and filled to a depth of 1.25 cm with sodium sulfate. The homogenate was extracted twice with the filtrates collected together. The combined filtrate was concentrated to ~2 mL by rotary evaporation, and then further concentrated until solvent-free under a stre am of dry nitrogen. The residue was reconstituted in 2 mL of acetonitrile. Afte r vortexing (30 s), the supernatant was applied to a C18 solid phase extraction (SPE) car tridge (pre-conditio ned with 3 mL of acetonitrile; Agilent Technologies, Wilmingt on, DE, USA) and was allowed to pass under gravity. This procedure was repeated twice with the comb ined eluent collected in a culture tube. After the last addition, the car tridge was rinsed with 1 mL of acetonitrile which was also collected. The eluent was then applied to a 0.5 g NH2 SPE cartridge (Varian, Harbor City, CA, USA), was allowe d to pass under gravity, and collected in a graduated conical tube. The cartridge was rinsed with an additional 1 mL portion of acetonitrile which was also collected. The combined eluents were concentrated under a stream of dry nitrogen, to a volume of 300 L, and transferred to a gas chromatography (GC) vial for analysis.

PAGE 124

108 GC/MS Analysis Analysis of all samples was performed using a Hewlett Packard HP-6890 gas chromatograph (Wilmington, DE, USA) with a split/splitless inlet ope rated in splitless mode. The analytes were introduced in a 1 L injection and separa ted across the HP-5MS column (30 m x 0.25 mm; 0.25 m film thickne ss; J & W Scientific, Folsom, CA, USA) under a temperature program that began at 60 C, increased at 10 C/min to 270 C, was held for 5 min, then increased at 25 C/min to 300 C and was held for 5 min. Detection utilized an HP 5973 mass spectro meter in electron impact m ode. Identification for all analytes and quantitation for toxaphene was c onducted in full scan mode, where all ions are monitored. To improve sensitivity, se lected ion monitoring was used for the quantitation for all other analytes, except kepone. The above program was used as a screening tool for kepone which does not optim ally extract with mo st organochlorines. Samples found to contain kepone would be reex tracted and analyzed specifically for this compound. For quantitation, a five-point standard curve was prepared for each analyte ( r2 0.995). Fresh curves were analyzed with each se t of twenty samples. Each standard and sample was fortified to contain a deuterat ed internal standard, 5 L of US-108 (120 g/mL; Ultra Scientific), added just prior to analysis. All samples also contained a surrogate, 2 g/mL of tetrach loroxylene (Ultra Scientific) added after homogenization. Duplicate quality control samples were prepar ed and analyzed with every twenty samples (typically at a level of 1.00 or 2.50 g/mL of -BHC, heptachlor, aldr in, dieldrin, endrin, and p,p’ -DDT) with an acceptable recovery rangi ng from 70 – 130%. Limit of detection ranged from 0.1-1.5 ng/g for all OCP analyt es, except toxaphene (120-236 ng/g), and limit of quantitation was 1.5 ng/g for all anal ytes, except toxaphene (1500 ng/g).

PAGE 125

109 Repeated analyses were conducted as allo wed by matrix interferences and sample availability. Results Inter-Site Clutch Comparisons A total of 58 clutches were coll ected during June 2001 and 2002 from Lakes Apopka, Griffin, Lochloosa and Emeralda Mars h Conservation Area. No differences ( = 0.05) were determined between sites with re spect to the following clutch parameters: fecundity (overall mean standard error: 43 1), clutch viability (56 3.9%), damaged eggs (4 1.7%), unbanded eggs (12 1.9%), early embryo mortality (i.e., mortality prior to Day 36; 14 2.7%), and late embryo mort ality (i.e., on or afte r Day 36; 14 2.5%). However, the average egg mass of clutches fr om Emeralda Marsh was greater than that of Lake Griffin (Table 5-1). Significant differences were noted between sites in recent studies (Chapter 2) which a had a larger total sample size (n = 168). The lack of significant differences was likely due to the large variance noted in Emeralda clutches (Table 5-3). Many differences were detected between sites with respect to egg yolk OCP concentrations. Alligator eggs from Emer alda Marsh and Lake Apopka were found to have a greater number of OCP analytes ( n = 12 and 11, respectively) as compared to Lake Griffin ( n = 10), which was greater than Lochloosa ( n = 8.5) (Table 1). Eggs from Emeralda Marsh yielded the highest tota l OCP concentrations (29,838 4,844 ng/g), which were over three-fold greater than thos e of Lake Apopka, 32-fold greater than those of Lake Griffin, and 290-fold greater than those of Lake Lochloosa. Furthermore, 45% of individual OCP analytes were at greater concentrations in Emeralda eggs as compared to those of Lake Apopka, with major differe nces in total OCP egg yolk concentrations

PAGE 126

110 related to amounts of toxaphene (thr ee-fold greater in Emeralda) and p,pÂ’ -DDE (two-fold greater in Emeralda). Other OCP analyte c oncentrations were similar between Emeralda and Lake Apopka, except for a few DDT and ch lordane analytes. When individual OCP egg yolk concentrations from Apopka and Emeralda were compared against Lakes Griffin and Lochloosa, 82% of the individual OCP analyte con centrations were greater in eggs from Emeralda Marsh and Lake Apopka as compared to the other lakes (Table 5-1). Intra-Site Live Embryo/Dead Em bryo Morphological Comparisons Comparisons between live and dead embr yos sampled at chronological age (CA) Day 14 yielded the following results. For Lakes Lochloosa, Apopka, and Griffin clutches, live embryos sampled at CA Day 14 had an overall morphological age (MA) of 15 0.3 (mean standard error), which was greater than the MA (11 1) of dead embryos, and live embryos were of greater mass compared to dead embryos, suggesting that dead embryos may have been devel opmentally retarded. One other notable difference between live and dead embryos samp led at CA Day 14 was that eggs of dead embryos were greater in mass compared to live cohorts for Lakes Apopka and Griffin; however, for Lake Lochloosa, eggs of live embryos were greater in mass compared to dead cohorts (Table 5-2). Comparisons betw een eye length, head le ngth, and total length were not made because dead embryos coul d not be uniformLy positioned for photographs due to their size and fragility of tissues resulting from the early stage of development and limited autolysis. For Lake Lochloosa, live and dead embryos of CA Day 25 differed with respect to egg mass, embryo mass, and MA (for all endpoints: live > dead), and Total L.: Emb. M., with live Total L.: Emb. M. ratios being less than those of dead embryos. For Lake Apopka, live and dead embryos differed with respect to embryo mass, Eye L.: Head L.,

PAGE 127

111 morphological day (for all endpoin ts: live > dead), and Head L. : Total L. (live < dead). No significant differences we re detected between live and dead embryos from Emeralda clutches for the measured endpoints. For La ke Griffin, differences between live and dead embryos were determined for embryo mass, eye length, head lengt h, and MA (for all endpoints: live > dead) (Table 5-2). For Lake Lochloosa, live and dead embryos of CA Day 33 differed with respect to embryo mass and MA (for both endpoints: liv e > dead). For Lake Apopka, live embryos had greater mass, eye length, and MA than d ead embryos, but dead embryos had greater Total L.: Emb. M.. Live embryos from Emeralda Marsh clutches were of greater mass and Total L.: Emb. M. than dead cohorts. Li ve embryos from Lake Griffin clutches were also of greater mass and were of greater MA in comparison to dead cohorts (Table 5-2). For Lake Lochloosa, live and dead embryos sampled at CA Day 43 differed with respect to embryo mass, eye length, head length, Head L.:Total L., and MA (for all endpoints: live > dead). For Lake Apopka, live embryos were of greater mass, head length, total length, and morphological age compared to dead c ohorts. Live embryos of Emeralda Marsh were of greater morphological age than dead cohorts, however, power of detection was low since only one dead embryo was sampled. Lastly, live embryos of Lake Griffin were of greater mass and morphological age th an their dead cohorts (Table 5-2). Inter-Site Comparisons of Morphology of Live Embryos Egg mass of live embryos of CA Day 14 di ffered between sites, with respect to egg mass and MA. Eggs of CA Day 14 embryos from Lake Lochloosa clutches were of greater mass compared to those of Emeral da and Griffin, and MA of Lake Griffin embryos was greater than that of Emeralda Ma rsh (Table 3). For embryos sampled at CA

PAGE 128

112 Day 25, egg mass, embryo mass, eye lengt h, head length, total length, and TL: EM differed among sites. Eggs from Lake Loch loosa were of greater mass than all other sites, but embryo mass for Lochloosa clutches was less than that of Griffin. Embryos from Emeralda and Lake Griffin had the gr eater eye lengths in comparison to Lake Apopka with no significant differences dete cted for Lake Lochloosa embryos. With respect to head lengths and to tal lengths, embryos from La ke Apopka clutches were less than those of all other sites and embryos from Lake Griffin were greater than those of all other sites except for Emeralda. Lastly, Lochloosa embryos had higher Total L.: Emb. M. than all other sites except fo r Lake Apopka (Table 5-3). For CA Day 33 embryos, only egg mass and MA differed between sites. Similar to earlier sampling periods, Lake Lochloosa eggs were of greater mass than all other sites except for Apopka. Lake Griffin embryos were of greater MA than all other sites, and Lake Apopka embryos were of lesser mass than all other sites except for those of Emeralda. For CA Day 43 embryos, Lochloosa eggs were of greater mass than all other sites, and Lochloosa embryos were of lesser mass than all other sites except for Lake Apopka. Embryos of Lake Griffin were of greater ma ss than all other sites. Embryos of Lake Apopka and Lake Griffin also had greater eye lengths than those of Emeralda. For TL: EM ratios, Lake Griffin embryos had smaller ra tios compared to all other sites except for Emeralda. In addition, the MA of embryos of Lochloosa was less than all other sites except Emeralda (Table 5-3). Live Embryo Morphology and Em bryo Survival Relationships Redundancy analysis with forward select ion (Monte Carlo permutation tests for significance) was used to examine whethe r embryo morphometric parameters (eye

PAGE 129

113 length, head length, total lengt h, and embryo mass) were strongly associated with embryo survival parameters (clutch viability, early embryo mortality, and late embryo mortality percentages) for each of the chronological ages (CA) sampled (Day 14, Day 25, Day 33, Day 43). For CA Day 14 embryos, results of the RDA indicat ed only late embryo mortality percentage was significantly asso ciated with the observed variation embryo morphology, but accounted for only 7% of the morphological variation. For CA Day 25 and Day 33 embryos, no significant associa tions were found between morphological and embryo survival parameters. For CA Day 43, cl utch viability was determined significant but accounted for only 5% of the variation in morphology. Live Embryo Morphology and Eg g Yolk OCP Burdens In contrast to embryo survival paramete rs, OCP concentrations in egg yolks were significantly associated with variation in embryo morphology. For CA Day 14 embryos, partial-redundancy analysis usi ng site (i.e., Emeralda, Apopka Griffin, Lochloosa) as the covariate (since OCP burdens differed among site s), with forward selection of the best five OCP variables (Table 4), indicated that four of five sele cted variables were determined to be significant via Monte Carlo permutation tests. The four OCP variables were oxychlordane concentra tion ([OX]), heptachlor epoxid e percentage of total OCP burden (HE%), toxaphene percentage of total OCP burden (TX%), and trans -nonachlor percentage of total OCP burden (TN%). Th ese OCP variables accounted for 47% of the observed variation in embryo morphometric pa rameters. Individually, [OX] explained 20% of the variation in the morphometric parameters, followed by HE% (11%), TX% (10%), and TN% (6%) (Table 5). Embryo h ead length was negatively correlated with [OX] and HE%. Embryo eye length was ne gatively correlated with TN% and HE%.

PAGE 130

114 Embryo mass was positively correlated with [OX] and HE%, and embryo total length was negatively correlated with [OX] but pos itively correlated with TN%.(Fig. 5-2). For CA Day 25 embryos, four of five sele cted variables were determined to be significant, with HE%, cis-Nonachlor percen tage of total OCP burden (CN%), dieldrin percentage of total OCP burden (DL%), and dieldrin concentrati on ([DL]) accounting for 24% of variation of embryo morphological parameters. HE% accounted for 11% of embryo morphological variation, followed by CN% (6%), DL% (5%) and [DL] (2%) (Table 5). Embryo head length was positivel y correlated with DL%, CN%, HE%, and [DL]. Total embryo length was also positively correlated with HE%, DL%, and [DL], but showed little correlation with CN%. In contrast to head length and total length, embryo mass and eye length were negatively correlated with HE%, DL%, and [DL], but showed little correlation with CN% (Fig. 5-3). For CA Day 33 embryos, three of five sele cted variables were determined to be significant and consisted of CN%, DL%, cischlordane percentage of total OCP burden (CC%) and accounted for 24% of morphological variation (Table 5). Embryo head length and total length were positively correl ated with DL% and negatively correlated with CC% and CN%. In contrast, embryo mass and eye length were positively correlated with CC% and CN%, but showed little correlation with DL% (Fig. 5-4). For CA Day 43 embryos, three of five sele cted variables were determined to be significant and together accounted for 20% of the morphological variation. These variables consisted of p,pÂ’ -DDT concentration ([pDDT]), cis -chlordane concentration ([CC]), and total number of individual OCP anal ytes detected in yolk (NOC) (Table 5). Embryo mass was positively correlated w ith NOC and [pDDT], but showed little

PAGE 131

115 correlation with [CC]. Embryo eye lengt h was positively correlated with NOC, negatively correlated with [CC], and showed no correlation with [pDDT]. Embryo head length was positively correlated with NOC and negatively correlated with [CC] and [pDDT]. Total embryo length was negatively correlated with NOC, not correlated with [pDDT], and positively correlated with [CC] (Fig. 5-4). Embryo Morphological Age, Derived Morphometric Variables and Egg Yolk OCP Burdens For embryos sampled at CA Day 14, four of five RDA-selected OCP variables were determined to be signifi cant and accounted for 44% of th e variation associated with morphological age (MA) and deri ved morphometric variables (DMV), which consisted of Eye L.: Head L., Head L.: Total L., and To tal L.: Emb. M.. The four extracted OCP variables were [OX], CN%, OX%, and HE%, and each respectively accounted for 21%, 12%, 7%, and 4% of variance associated w ith MA and DMV (Table 5-6). With the exception of Total L.: Emb. M., all DMV and MA were positively correlated with were [OX], OX%, and HE%. Total L.: Emb. M. was positively correlated with CN%, and CN% was negatively correla ted with MA and the other DMV (Fig. 5-6). For CA Day 25 embryos, three of five selected OCP variables were determined significant and accounted for 22% of MA and DMV variance. The three OCP variables consisted of HE%, NOC, and [pDD] and each respectively accounted for 12%, 7%, and 3% of the variation in MA and DMV (Table 6). HE% was positively correlated with Total L.: Emb. M. and negatively correlated with MA, Head L.: Total L., and Eye L.: Head L. NOC was negatively correlated with Total L.: Emb. M. and positively correlated with MA, Head L.: Total L., and Eye L. [p DD] was positively correlated with Head L.:

PAGE 132

116 Total L. and Eye L.: Head L., but showed little correlation with MA and Total L.: Emb. M (Fig. 5-7). For Day 33 embryos, three OCP variables (DL%, PDE%, and CN%) evenly accounted for 15% of variation in MA a nd DMV (Table 6). DL% was negatively correlated with Head L.: Total L. and Eye L. : Head L., but showed little correlation with MA and was positively correlated with Total L.: Emb. Mass. PDE%, and CN% were positively correlated with MA, Head L.: Tota l L. and Eye L.: Head L., but showed a negative correlation with Tota l L.: Emb. M. (Fig. 5-8). Lastly, for CA Day 43 embryos, four of five OCP variables selected via partial RDA were determined to be significant and accounted for 24% of variation in MA and DMV. These four OCP variables consisted of, with respect to amount of variation accounted for, PDT% (8%), [CC] (5%), NOC (5%), and [PDT] (4%) (Table 6). PDT%, NOC, and [PDT] were positively correlated with Head L.: Total L., Eye L.: Head L., and MA, but were negatively correlated with Total L.: Emb. M. [CC] was negatively correlated with Head L.: Total L., Eye L.: H ead L., and MA, but was positively correlated with Total L.: Emb. M. (Fig. 5-9). Histopathology of Live and Dead Embryos Results of histopathology of live embryos ( n = 34) from five reference clutches (clutch viability > 71% and OCP yolk burdens < 350 ng/g) and live embryos ( n = 26) from four OCP-contaminated clutches (clu tch viability < 47% and OCP yolk burdens > 3,700 ng/g) indicated that 16% of all embryos exhibited at least one type of hepatic lesion, followed by lesions of the skeletal mu scle (5%), and kidney (3%). Hepatic lesions included necrosis (characterized by pyknotic nuclei and vacuolated hepatocytes) and cholestasis. Lesions detect ed in skeletal muscle incl uded necrosis characterized by

PAGE 133

117 pyknotic nuclei and segmented sarcoplasm. Ki dney lesions included necrosis of tubules characterized by vacuolizati on and pyknotic nuclei. No significant differences were determined between reference embryos and OCP-contaminated embryos with respect to incidence of hepatic lesions ( 2 = 0.87, p = 0.49), renal lesions ( 2 = 1.58, p = 0.50), or muscular lesions ( 2 = 2.41, p = 0.25). Histopathology results of dead embryos ( n = 20) from OCP-contaminated sites indicated that generalized autolysis was the predominant finding, with fungi hyphae present in 2 embryos, and a si ngle case of menigoencephalitis that would be consistent with a bacterial infection. Advance autolysis, in some cases, likely impeded detection of cytoxic lesions. Discussion With respect to the first hypothesis, re sults suggest that certain morphological parameters of live alligator embryos differ from those of dead embryos of the same chronological age. Intra-site comparisons s uggested that among all sites and all sampled ages (CA) embryo MA and mass were greate r for live embryos as compared to dead embryos. Importantly, the concurrent decr eases in MA and mass of dead embryos suggests that embryos may have been deve loping normally up to a point at which development stalled and the embryo eventu ally perished, or embryos could have developed at a much slower overall rate until the point at which they perished. Either way it appears that the mass of dead embryos was appropriate for their MA. For example, live embryos of Lake Griffin samp led at CA Day 14 had an average MA of ~ Day 16 and an average mass of 0.41 g, which wa s similar to the MA (~ Day 15) and mass (0.41 g) of dead embryos sampled at CA Day 25 (Table 2). Other measured parameters and derived parameters showed variation in patterns among sites and ages, but one

PAGE 134

118 consistency was that when differences were detected, measured parameters were almost always greater in live embryos as compared to dead embryos. With the exception of Lake Apopka, few significant differences were found between live and dead embryos with respect to derived morphometric parameters (i.e., Eye L.: Head L.; Head L.: Total L.; and To tal L.: Emb. M.). Differences were found between derived morphometric parameters of live and dead embryos sampled at older ages (CA) from Lake Apopka, and suggest that morphology of dead embryos of Lake Apopka is disproportionate compared to liv e cohorts. The differences between the patterns of morphometric relationships of live and dead embryos from Lake Apopka as compared to other sites, may indicate the cause s or mechanisms associated with mortality of Lake Apopka embryos differ from other site s, since it has been s hown that the type of teratogenic effect may depend on the specifi c teratogenic agent or cause (Schmidt & Johnson, 1997). With respect to the second hypothesis, results suggested that morphology of live embryos was not consistently different among sites, except for live embryos of CA Day 25. For Day 25 live embryos, embryos of Lake Griffin and Emeralda Marsh were consistently larger, with respect to measurem ent parameters, than those of Lakes Apopka and Lochloosa. The only differences found, w ith respect to derived parameters, was for Total L.: Emb. M., with Lochloosa embryos app earing to be leaner embryos compared to those of Emeralda and Griffi n. Since Day 25 is during the middle of organogenesis, this stage of development may be more sensit ive to OCP exposure or variation in yolk nutrient content since it has been shown in other species the peri od of organogenesis is

PAGE 135

119 most susceptible to altera tions caused by teratogen expos ure or nutrient excess or deficiency (Schmidt & Johnson, 1997). With respect to the third hypothesis, re dundancy analysis results indicated that variation in morphometry of live embryos is not significantly related to variation in clutch mortality rates, suggesting that live embryos from clutches with high mortality rates develop similarly to those of low mort ality clutches. This finding may suggest a threshold-type response in which embryos expos ed to stressors below a certain threshold have the ability to overcome stressors through various cellu lar homeostatic mechanisms, but above a certain threshold, developmental retardation a nd lethality occur. Such threshold dose-response patterns have been acc epted as a major dose-response pattern in mammalian developmental toxico logy (Rogers & Kavlock, 2001). With respect to the fourth hypothesis, va riation in morphologi cal development of live embryos was significantly associated with variation in the composition and concentration of OCPs in eggs. However, th e strength of the relationships appeared to decrease with the age sampled (CA), with youngest embryos sampled (CA Day 14) showing the strongest rela tionships between OCP egg burden and morphometric parameters, followed by each subsequent CA, respectively (Table 5-5). Interestingly, the percentage of the total OCP burden (concentr ation) composed by an OCP analyte (i.e., HE%), appeared to be more important than OCP analyte concentrations alone. With respect to all sampled ages, except the elde st (CA Day 43), OCP percentage variables accounted from a minimum of 47% to a ma ximum of 100% of the total variation attributed to all OCP variables found to be significantly associated with variation in measured morphometric parameters (Table 55). For derived morphometric parameters

PAGE 136

120 and morphological age (MA), similar patterns were observed in that embryos sampled at younger CA showed stronger rela tionships with OCP burdens than older cohorts (Table 5-6). With respect to individual OCP analytes, the cyclodienes appear to be more important than the dichlorodi phenylethanes, in that cyclodienes accounted for an average of 70% of the morphometric variation that co uld be attributed to OCP variables for all sampled ages. This is surprising cons idering that dichlorodiphenylethanes ( p,pÂ’ -DDT + p,pÂ’ -DDD + p,pÂ’ -DDE) make up an average of 66% of the total OCP burden among all sites (Table 5-1). Another important observation was that di fferent cyclodienes appeared to be associated with morphological variation of embryos of different ages (CA). Most important were the components of technical grade chlordane and its metabolites, which include cis and trans -chlordane, cis and trans -nonachlor, oxychlordane, and heptachlor epoxide. One or more of these components were found to be significantly associated with variation in embryo morphology for each CA sampled. These data suggest that the chlordane group may merit furthe r study in relation to developm ental effects in reptiles, especially considering other st udies have suggested that sexu al differentiation in turtles may be altered by low dose in ovo exposures of these compounds (Willingham, 2004). With respect to the final hypothesis, no si gnificant differences were found between the histopathology of live and dead embryos fr om best-case clutches (low mortality rates and low OCP egg burdens) compared to those of worst-case clutches (high mortality rates and high OCP egg burdens). Few signs of b acterial or fungal infections were found. These results may suggest that lesions were not a causal factor in death, and may not be

PAGE 137

121 associated with variation in OCP exposure or increased mortality rates. However, the death and autolysis of delicate embryonic tissu es may have obscured lesions associated with death and/or OCP exposure. In addi tion, OCPs may cause dysfunction in neural transmission, leading to cardiovascular fa ilure and mortality. Future studies may consider in ovo monitoring of neural transmissi on and cardiovascular function to determine if increased OCP exposure is associ ated with altered neural transmission and cardiovascular failure in alligator embryos. In conclusion, the present study found that embryo mortality occurring in alligator populations inhabiting refere nce and OCP-contaminated sites was characterized by developmental retardation without gross deformities or overt presence of lesions to vital organs. However, variation in embryo mo rphology appeared to be associated with variation in OCP burdens of eggs and the percentage composition composed by an OCP analyte was equally as important as concen tration, suggesting the importance of mixture composition. Younger embryos appeared more susceptible to OCP influence but OCP influence may not necessarily be the result of direct embryo effects. Similar types of embryo mortality has been documented in quail, with embryo mortality determined to be maternally mediated, where maternal liver function was altered, resulting in nutrient deficiencies in eggs that were severe enough to induce em bryo mortality (Donaldson & Fites, 1970). In summary, subsequent studi es should evaluate embryo mortality in alligators as a function of OCP exposure and egg nutrient content.

PAGE 138

122Table 5-1. Summary statistics for paramete rs measured on American alligator clut ches collected during June 2001 and 2002. Parameter Apopka Emeralda Griffin Lochloosa Summary No. Clutches (n) 15 7 18 18 58 Fecundity ( n ) 46 2 50 3 43 2 40 2 43 1 (28–56) (42–64) (19–58) (26–56) (19–64) Egg mass (g) 86 3.3 AB 107 18.1 A 79 3.2 B 89 3.2 AB 87 2.8 (67–120) (65–180) (46–113) (71–139) (46–180) Clutch viability (%) 53 8.6 63 13.8 46 6.7 64 5.7 56 3.9 (0–92) (0–96) (0–87) (0–95) (0–96) Damaged eggs (%) 1 0.4 1 0.7 7 4.2 4 3.3 4 1.7 (0–4) (0–4) (0–63) (0–60) (0–63) Unbanded eggs (%) 13 3.4 10 3.9 15 4.8 11 2.1 12 1.9 (0–40) (0–30) (0–65) (0–33) (0–65) Early Emb. Mort. (%) 14 6.2 24 12.7 13 4.4 13 3.1 14 2.7 (0–90) (0–95) (0–73) (0–36) (0–95) Late Emb. Mort. (%) 19 6.3 3 2.1 19 4.8 9 2.6 14 2.5 (0–77) (0–15) (0–58) (0–34) (0–77) Dieldrin (ng/g) 405.7 121.32 A 227.3 40.09 A 24.8 5.18 B 3.6 0.5 C 146.1 39.53 (23.5–1,859) (88–386.7) (4.4–76.9) (1.3–8.2) (1.3–1,859) Hep. Epoxide (ng/g) 14.4 3.01 A 5.3 1.38 AB 7.1 2.05 B 2.8 0.76 B 7.8 1.24 (1.2–46.8) (1.4–11.6) (1.1–29.6) (1.2–9.7) (1.1–46.8) cis -Chlordane (ng/g) 48.3 13.74 B 150.4 26.21 A 10.7 1.01 C 1.9 0.21 D 34.3 7.71 (6.6–179.2) (62.4–281) (4.3–16.9) (1.2–4.1) (1.2–281) cis -Nonachlor (ng/g) 70.4 16.41 A 89.2 15.7 A 17.1 2.7 B 4.6 0.63 C 35.7 6.29 (10.5–238.4) (55–166) (4.4–54.2) (2.4–12.5) (2.4–238.4) Oxychlordane (ng/g) 45.7 10.94 A 29.3 4.1 A 12.6 3.17 B 3.8 1.06 C 20.4 3.71 (3.9–176) (17.9–46.1) (1.1–45.9) (1.2–17.8) (1.1–176)

PAGE 139

123Table 5-1. Continued. Parameter Apopka Emeralda Griffin Lochloosa Summary Toxaphene (ng/g) 3,308 658.5 B 8,269 1,077.2 A 2,677 376.5 B nd C 4872 722.8 (1,896.1–9,678.3) (4,512.8–11,485.4) (1,927.9–3,110.8) (0–0) (1,896–11,485.4) p, p' -DDD (ng/g) 45.8 13.49 B 1,986 333.9 A 5.7 1.1 C 1.9 0.21 D 277.6 101.28 (10.6–192.8) (617.3–2962.8) (1.5–18.5) (1.2–2.9) (1.2–2962.8) p, p' -DDE (ng/g) 5,792 1,490.4 B 18,056.3 3,113.7 A 337 55.2 C 74.8 12.38 D 3,805 924 (18.3–22,421.9) (6,811.7–33,554.8) (94.6–979.1) (28–231) (18.3–33,554.8) p, p' -DDT (ng/g) 9.8 3.81 A 17.7 3.24 A 2.7 0.25 AB 1.3 0.02 B 9.4 2.17 (1.2–45.6) (5.8–25.3) (2.5–3) (1.2–1.3) (1.2–45.6) trans -Chlor. (ng/g) 7.4 2.38 B 44 4.61 A 1.6 0.19 C 2.6 0.73 BC 11.3 2.81 (1.3–27.4) (23.2–58.2) (0.8–3.4) (1.2–3.7) (0.8–58.2) trans -Nonachl. (ng/g) 202.5 54.77 A 278.2 56.93 A 42.4 9.31 B 8.4 1.67 C 101.7 20.46 (10.5–787.6) (148–554.9) (8.9–155.2) (2.5–24.6) (2.5–787.6) [OCP] (ng/g) 9,177 2,391.2 B 29,838 4,844.3 A 911 302.4 C 103 16.3 D 6,238.6 1,508.35 (555.2–35,587.8) (13,183.5–53,559.7) (128.7–4,487.7) (42.7–289.4) (42.7–53,559.7) No. OCPs ( n ) 11 0.28 A 12 0.22 A 10 0.18 B 8.6 0.28 C 10.1 0.2 (9–12) (11–13) (9–12) (6–10) (6–13) aValues represent mean standard error with ranges in parenthe ses. Letters beside values (A -D) indicate differences between si tes ( = 0.05). Clutch viability % = number of liv e hatchlings / fecundity x 100, damaged eggs % = number of damaged egg / fecundity x 100, unbanded eggs % = number of unbanded eggs / fecundity x 100, early embryo mortality % = number of embryos that died at age s Day 35 / fecundity x 100, late embryo mortality % = number of embr yos that died at ages >Day 35 / fecundity x 100, Hep. Epoxide = heptachlor epoxide, trans -Chlor. = trans -chlordane, trans -Nonachl. = trans -nonachlor, and No. OCPs = nu mber of OCPs detected at measurable levels.

PAGE 140

124Table 5-2. Comparisons of egg and embr yo morphometrics of live and dead embryos collected during June-August of 2001 and 2002. Agea Parameterb Apopka Emeralda Griffin Live Dead Live Dead Live Dead 14 n 7 9 12 23 8 Egg mass (g) 86.3 1.23 92.5 1.35* 78.6 1.35 81.6 1.92 86.7 2.6 Embryo mass (g) 0.38 0.12* 0.14 0.089 0.35 0.019 0.41 0.032 0.2 0 Eye L. (mm) 2.9 0 0 0 0 0 3.5 0.06 0 0 Head L. (mm) 8 0 0 0 0 0 9.1 0.28 0 0 Total L. (mm) 51.4 0 0 0 0 0 57.5 0.57 0 0 Eye L.: Head L. 0.36 0 0 0 0 0 0.39 0.012 0 0 Head L.: Total L. 0.16 0 0 0 0 0 0.16 0.006 0 0 Total L.: Emb. M. 171.33 0 0 0 0 0 111.5 15.353 0 0 Morph. Day 16 1.195 12.6 1.661 14 0 15.96 0.4* 12 3 25 n 20 13 24 6 30 37 Egg mass (g) 81.1 2.01 87.1 2.1 80 0.98 75.7 2.55 80.7 1.49 79.7 1.05 Embryo mass (g) 1.15 0.11* 0.56 0.131 1.45 0.116 1.57 0.27 1.51 0.088* 0.41 0.188 Eye L. (mm) 5 0.18 4.2 1.17 5.5 0.12 5.9 0 5.9 0.17* 2.5 0.98 Head L. (mm) 11.5 0.55 11.2 2.66 13.5 0.41 16.7 0 14.9 0.65* 7.7 3.54 Total L. (mm) 73.7 3.38 65.6 14.12 87.6 2.44 98.7 0 89.1 2.32 72.3 0 Eye L.: Head L. 0.44 0.01* 0.36 0.024 0.41 0.007 0.35 0 0.4 0.011 0.33 0.029 Head L.: Total L. 0.16 0.012 0.17 0.015 0.15 0.002 0.17 0 0.16 0.003 0.2 0* Total L.: Emb. M. 68.25 4.72 146.51 41.27* 58.89 2.47 47.02 0 52.24 3.526 0 0 Morph. Day 24.2 0.99* 17.63 1.475 26.42 0.58 28 0 26.13 0.619* 14.63 1.75 33 n 15 13 29 3 29 17 Egg mass (g) 82.4 2.08 81.5 3.4 80.8 1.15 75 3.09 77.7 1.42 81 1.4

PAGE 141

125Table 5-2. Continued. Agea Parameterb Apopka Emeralda Griffin Live Dead Live Dead Live Dead 33 Embryo mass (g) 3.14 0.09* 1.2 0.423 3.8 0.196* 1.2 0 4.04 0.203* 0.75 0.552 Eye L. (mm) 6.2 0.1* 3.3 0.52 6.4 0.12 6 0 6.5 0.1 6.6 0.27 Head L. (mm) 19.1 0.31* 10.7 3.21 20.1 0.49 19.1 0 20.6 0.35 19.9 0.34 Total L. (mm) 108.2 1.1* 67.8 13.25 114.4 2.54 108.8 0 117.8 2.19 111.9 5.02 Eye L.: Head L. 0.32 0.008 0.32 0.049 0.32 0.007 0.31 0 0.32 0.006 0.33 0.008 Head L.: Total L. 0.18 0.00* 0.15 0.017 0.18 0.002 0.18 0 0.18 0.001 0.18 0.005 Total L.: Emb. M. 34.77 0.81 291.4 254.57* 31.36 1.56 90.69 0* 31.41 1.467 30.53 0 Morph. Day 33.3 0.33* 19.83 3.323 35.69 0.57 33 0 39.28 0.854* 19 3.167 43 n 44 10 24 1 41 24 Egg mass (g) 81.2 1.45 85.4 2.56 79.8 1.19 73.5 0 78 1.32 80.6 1.34 Embryo mass (g) 9.91 0.27* 3.11 1.463 10.69 0.30 0 0 13.01 0.514* 6.77 2.706 Eye L. (mm) 7.3 0.14 7 0.4 6.5 0.11 0 0 7.5 0.2 6.8 0.75 Head L. (mm) 27.5 0.53* 21.6 5.57 26.1 0.41 0 0 27.4 0.9 24.6 4.08 Total L. (mm) 172.1 5.2* 130.1 36.33 170.1 2.14 0 0 184.9 7.05 167.8 29.4 Eye L.: Head L. 0.27 0.004 0.34 0.069* 0.25 0.006 0 0 0.35 0.08 0.29 0.041 Head L.: Total L. 0.16 0.003 0.17 0.004 0.15 0.002 0 0 0.19 0.042 0.15 0.003 Total L.: Emb. M. 15.89 1.00 238 188.667* 16.59 0.35 0 0 14.26 0.741 21.2 7.317 Morph. Day 47.8 0.68* 30.5 4.119 47.6 0.42* 38 0 48.5 0.482* 36 5.04

PAGE 142

126 Table 5-2. Continued. Agea Parameterb Lochloosa Live Dead 14 n 13 8 Egg mass (g) 90.4 1.64* 79.5 0.98 Embryo mass (g) 0.32 0.032 0 0 Eye L. (mm) 4.3 1.22 0 0 Head L. (mm) 12.2 3.18 0 0 Total L. (mm) 56.6 12.21 0 0 Eye L.: Head L. 0.35 0.025 0 0 Head L.: Total L. 0.18 0.03 0 0 Total L.: Emb. M. 151.48 52.437 0 0 Morph. Day 14.46 0.666* 8.33 1.202 25 n 32 22 Egg mass (g) 87.5 1.25* 81.7 1.36 Embryo mass (g) 1.03 0.051* 0.52 0.175 Eye L. (mm) 5.5 0.16 5 0.37 Head L. (mm) 13.4 0.51 13 1.89 Total L. (mm) 78.8 2.18 80.9 9.26 Eye L.: Head L. 0.42 0.007 0.39 0.025 Head L.: Total L. 0.16 0.003 0.16 0.005 Total L.: Emb. M. 74.64 12.461 186.08 78.758* Morph. Day 24.28 0.49* 14.5 2.045 33 n 27 11 Egg mass (g) 86.5 1.28 82.9 1.83 Embryo mass (g) 3.41 0.226* 3.13 1.788 Eye L. (mm) 6.4 0.18 7.4 0 Head L. (mm) 19.8 0.37 25.8 0* Total L. (mm) 115.1 2.63 144.8 0* Eye L.: Head L. 0.33 0.008 0.29 0 Head L.: Total L. 0.17 0.001 0.18 0 Total L.: Egg M. 31.93 1.262 0 0 Morph. Day 36.48 0.676* 16.75 4.304 43 n 40 14 Egg mass (g) 85.7 1 84.4 1.11 Embryo mass (g) 9.25 0.346* 5.74 3.25 Eye L. (mm) 7 0.14* 3 0 Head L. (mm) 26.2 0.63* 18.3 9.75 Total L. (mm) 165 4.15 212 0 Eye L.: Head L. 0.27 0.006 0.35 0 Head L.: Total L. 0.16 0.002* 0.13 0

PAGE 143

Table 5-2. Continued. Agea Parameterb Lochloosa Live Dead 43 Total L.: Emb. M. 14.29 1.212 139.47 139.47* Morph. Day 45.69 0.496* 20.74 4.456 aAge = chronological (calendar) age of embryo (days). bValues = mean standard error. L. = le ngth, Eye L.: Head length = eye length / head length, Head L.: Total L. = head length / to tal length, Total L.: Emb. M. = total length / embryo mass, and Morph. Day = age as dete rmined by morphological characteristics. *indicate significant differences ( = 0.05).

PAGE 144

128Table 5-3. Morphometric comparisons of live em bryos collected during June-August 2001 and 2002. Agea Parameterb Apopka Emeralda Griffin Lochloosa Summary 14 n 7 12 23 13 55 Egg mass (g) 86.3 1.23 AB 78.6 1.35 B 81.6 1.92 B 90.4 1.64 A 83.6 1.11 (82.8–90.9) (70.8–85.4) (61.8–100.1) (77–99.7) (61.8–100.1) Embryo mass (g) 0.38 0.115 A 0.35 0.02 AB 0.41 0.032 AB 0.32 0.03 B 2 0.37 0.022 (0–0.8) (0.2–0.4) (0.1–0.8) (0.2–0.5) (0–0.8) Eye length (mm) 2.9 0 0 0 3.5 0.06 4.3 1.22 3.8 0.53 (2.9–2.9) (0–0) (3.4–3.7) (2–6.6) (2–6.6) Head length (mm) 8 0 0 0 9.1 0.28 12.2 3.18 10.4 1.43 (8–8) (0–0) (8.7–9.9) (5.8–19.1) (5.8–19.1) Total length (mm) 51.4 0 0 0 57.5 0.57 56.6 12.21 56.4 4.07 (51.4–51.4) (0–0) (56.3–58.9) (44–81) (44–81) Eye L.: Head L. 0.36 0 0 0 0.39 0.012 0.35 0.025 0.37 0.013 (0.36–0.36) (0–0) (0.36–0.42) (0.29–0.41) (0.29–0.42) Head L.: Total L. 0.16 0 0 0 0.16 0.006 0.18 0.03 0.17 0.011 (0.16–0.16) (0–0) (0.15–0.18) (0.13–0.24) (0.13–0.24) Total L.: Emb. M. 171.33 0 0 0 111.49 15.353 151.48 52.437 135.91 23.698 (171.33–171.33) (0–0) (72.25–147.3) (0–223.86) (0–223.86) Morph. Day 16 1.195 AB 14 0 B 15.96 0.4 A 14.46 0.67 AB 15.18 0.291 (10–19) (14–14) (12–21) (9–18) (9–21) 25 n 20 24 30 32 106 Egg mass (g) 81.1 2.01 B 80 0.98 B 80.7 1.49 B 87.5 1.25 A 82.7 0.77 (62.4–96.4) (71.6–90.5) (59.4–100.2) (73.3–99) (59.4–100.2) Embryo mass (g) 1.15 0.111 AB 1.45 0.12 AB 1.51 0.088 A 1.03 0.051 B 1.29 0.049 (0.2–1.9) (0.2–2.7) (0.7–2.4) (0.3–1.6) (0.2–2.7) Eye length (mm) 5 0.18 B 5.5 0.12 A 5.9 0.17 A 5.5 0.16 AB 5.5 0.08 (4.5–7.2) (4.6–6.5) (4.9–8.6) (4.5–7.6) (4.5–8.6) Head length (mm) 11.5 0.55 B 13.5 0.41 A 14.9 0.65 A 13.4 0.51 A 13.5 0.29 (9.1–17.4) (10.4–17.7) (10.8–24.3) (10–18.5) (9.1–24.3)

PAGE 145

129Table 5-3. Continued. Agea Parameterb Apopka Emeralda Griffin Lochloosa Summary 25 Total length (mm) 73.7 3.38 C 87.6 2.44 AB 89.1 2.32 A 78.8 2.18 BC 82.9 1.43 (38.4–102.9) (71–105.6) (72.8–118.5) (63.4–102.5) (38.4–118.5) Eye L.: Head L. 0.44 0.01 0.41 0.007 0.4 0.011 0.42 0.007 0.42 0.005 (0.38–0.51) (0.36–0.48) (0.24–0.48) (0.34–0.46) (0.24–0.51) Head L.: Total L. 0.16 0.012 0.15 0.002 0.16 0.003 0.16 0.003 0.16 0.003 (0.13–0.32) (0.13–0.18) (0.14–0.21) (0.14–0.18) (0.13–0.32) Total L.: Emb. M. 68.25 4.717 AB 58.89 2.473 B52.24 3.526 B 74.64 12.461 A63.36 4.046 (46.28–102.39) (39.12–75.47) (0–79.12) (0–317) (0–317) Morph. Day 24.17 0.988 26.42 0.583 26.13 0.619 24.28 0.49 25.29 0.33 (12–33) (15–28) (17–30) (17–28) (12–33) 33 n 15 29 29 27 100 Egg mass (g) 82.4 2.08 AB 80.8 1.15 B 77.7 1.42 B 86.5 1.28 A 81.7 0.77 (65.9–92.6) (69.5–91.3) (62.4–90.7) (69.5–97.3) (62.4–97.3) Embryo mass (g) 3.14 0.09 3.81 0.196 4.04 0.203 3.41 0.226 3.67 0.107 (2.4–3.7) (1.7–6.1) (1.4–5.8) (1.1–8) (1.1–8) Eye length (mm) 6.2 0.1 6.4 0.12 6.5 0.1 6.4 0.18 6.4 0.06 (5.6–7) (4.9–7.7) (5.5–7.5) (4.3–7.5) (4.3–7.7) Head length (mm) 19.1 0.31 20.1 0.49 20.6 0.35 19.8 0.37 20 0.21 (16.7–21.4) (11.8–23.4) (14.5–23.2) (15.7–23.4) (11.8–23.4) Total length (mm) 108.2 1.11 114.4 2.54 117.8 2.19 115.1 2.63 114.6 1.22 (101–114.7) (76.6–135.9) (79.1–132.9) (91.2–153.6) (76.6–153.6) Eye L.: Head L. 0.32 0.008 0.32 0.007 0.32 0.006 0.33 0.008 0.32 0.004 (0.28–0.41) (0.26–0.42) (0.24–0.38) (0.26–0.38) (0.24–0.42) Head L.: Total L. 0.18 0.002 0.18 0.002 0.18 0.001 0.17 0.001 0.17 0.001 (0.16–0.19) (0.15–0.2) (0.17–0.19) (0.15–0.18) (0.15–0.2) Total L.: Emb. M. 34.77 0.812 31.36 1.562 31.41 1.467 31.93 1.262 32.07 0.726 (28.6–42.06) (22.28–52.18) (22.58–56.5) (19.2–49.51) (19.2–56.5) Morph. Day 33.33 0.333 C 35.7 0.57 BC 39.28 0.854 A 36.48 0.676 B 36.62 0.406 (33–38) (28–38) (28–48) (25–43) (25–48)

PAGE 146

130Table 5-3. Continued. Agea Parameterb Apopka Emeralda Griffin Lochloosa Summary 43 n 44 24 41 40 Egg mass (g) 81.2 1.45 B 79.8 1.19 B 78 1.32 B 85.7 1 A 81.4 0.68 (63.7–96) (67.5–91.5) (60.3–91.3) (69.6–97.9) (60.3–97.9) Embryo mass (g) 9.91 0.274 BC 10.69 0.297 B13.01 0.514 A 9.25 0.346 C 10.69 0.229 (7.3–14.41) (7.5–13.4) (7.71–23.53) (4.5–17.14) (4.5–23.53) Eye length (mm) 7.3 0.14 A 6.5 0.11 B 7.5 0.2 A 7 0.14 AB 7.2 0.08 (5.8–9.5) (5.5–7.1) (5.9–10.9) (4.3–8.6) (4.3–10.9) Head length (mm) 27.5 0.53 26.1 0.41 27.4 0.9 26.2 0.63 26.9 0.35 (15.8–36.9) (21.6–28.4) (3.2–35.8) (9.6–33.8) (3.2–36.9) Total length (mm) 172.1 5.19 170.1 2.14 184.9 7.05 165 4.15 173.6 2.87 (92.6–250.4) (147.5–181.3) (19.9–278.8) (122.5–251.3) (19.9–278.8) Eye L.: Head L. 0.27 0.004 0.25 0.006 0.35 0.08 0.27 0.006 0.29 0.02 (0.22–0.37) (0.21–0.3) (0.21–2.66) (0.23–0.45) (0.21–2.66) Head L.: Total L. 0.16 0.003 0.15 0.002 0.19 0.042 0.16 0.002 0.17 0.012 (0.13–0.22) (0.13–0.16) (0.01–1.48) (0.13–0.18) (0.01–1.48) Total L.: Emb.M. 15.89 1.004 A 16.6 0.35 AB 14.26 0.741 B 14.29 1.212 A 15.06 0.517 (0–23.01) (14.39–19.67) (0–24.48) (0–27.22) (0–27.22) Morph. Day 47.76 0.676 A 47.6 0.42 AB 48.54 0.482 A 45.69 0.496 B 47.35 0.289 (41.45–55) (38–48) (43–55) (37–48) (37–55) aAge = chronological (calendar) age of embryo (days). bValues represent mean standard error with ranges in parentheses. Letters beside values (A-D) indicate differences between si tes ( = 0.05). L. = length, Eye L.: Head length = eye length / head le ngth, Head L.: Total L. = head length / total length, Total L. : Emb. M. = total length / embryo mass, and Morph. Day = age as determined by morphological characteristics.

PAGE 147

131 Table 5-4. Explanatory vari ables included in partial re dundancy analysis evaluating relationship between organochlorine pest icide burdens in eggs and embryo morphometrics. Variablea Code Measured as cis -Chlordane [CC] ng/g yolk wet weight cis -Nonachlor [CN] ng/g yolk wet weight Dieldrin [DL] ng/g yolk wet weight Hep. Epoxide [HE] ng/g yolk wet weight o,p -DDD [ODD] ng/g yolk wet weight Oxychlordane [OX] ng/g yolk wet weight p,pÂ’ -DDD [PDD] ng/g yolk wet weight p,pÂ’ -DDE [PDE] ng/g yolk wet weight p,pÂ’ -DDT [PDT] ng/g yolk wet weight Toxaphene [TX] ng/g yolk wet weight trans -Chlordane [TC] ng/g yolk wet weight trans -Nonachlor [TN] ng/g yolk wet weight All OCP burdens [TOC] ng/g yolk wet weight No. OCPs measured NOC n cis -Chlordane% CC% [CC] / [TOC] x 100 cis -Nonachlor% CN% [CN] / [TOC] x 100 Dieldrin% DL% [DL] / [TOC] x 100 Hep. Epoxide% HE% [HE] / [TOC] x 100 o,p -DDD% ODD% [ODD] / [TOC] x 100 o,p -DDT% ODT% [ODT] / [TOC] x 100 Oxychlordane% OX% [OX] / [TOC] x 100 p,pÂ’ -DDD% PDD% [PDD] / [TOC] x 100 p,pÂ’ -DDE% PDE% [PDE] / [TOC] x 100 p,pÂ’ -DDT% PDT% [PDT] / [TOC] x 100 Toxaphene% TX% [TX] / [TOC] x 100 trans -Chlordane% TC% [TC] / [TOC] x 100 trans -Nonachlor% TN% [TN] / [TOC] x 100

PAGE 148

132 Table 5-5. Best five organochl orine pesticide (OCP ) variables accounting for variation in embryo morphology, selected using redundancy analysis with forward selection and Monte Carlo permutation tests for significance. Agea OCP variable LambdaAb P F 14 [OX] 0.20 0.012 12.31 HE% 0.11 0.02 7.03 TX% 0.1 0.028 8.48 ODT% 0.05 0.178 3.87 TN% 0.06 0.018 6.25 25 HE% 0.11 0.002 12.09 CN% 0.06 0.01 7.79 DL% 0.05 0.016 5.80 [DL] 0.02 0.04 3.49 PDD% 0.02 0.076 2.80 33 CC% 0.06 0.042 6.68 DL% 0.08 0.012 9.85 PDE% 0.03 0.082 4.00 CN% 0.1 0.006 14.51 PDD% 0.03 0.052 6.01 43 [CC] 0.07 0.002 11.11 [PDT] 0.08 0.002 12.58 NOC 0.05 0.002 9.60 [DL] 0.02 0.108 2.14 [HE] 0.01 0.064 2.84 aAge = chronological (calendar) age of embryo (days). bLambdaA = amount of morphometric varian ce accounted for by explanatory variable.

PAGE 149

133 Table 5-6. Best five organoc hlorine pesticide (OCP) variab les that account for embryo morphological age and derived morphologi cal parameters as determined by redundancy analysis with forward sel ection and Monte Carlo permutation tests for significance. Age Code LambdaA P F 14 [OX] 0.21 0.004 14.66 ODT% 0.08 0.064 5.41 CN% 0.12 0.006 11.56 OX% 0.07 0.026 6.67 [HE] 0.04 0.012 4.58 25 HE% 0.12 0.002 14.46 NOC 0.07 0.01 8.14 [PDD] 0.03 0.044 3.63 [CN] 0.01 0.154 2 [OX] 0.03 0.056 4.14 33 CC% 0.04 0.054 4.94 DL% 0.06 0.016 7 PDD% 0.03 0.082 4.07 PDE% 0.05 0.006 6.36 CN% 0.06 0.004 8.39 43 PDT% 0.08 0.008 11.2 [CC] 0.05 0.006 8.42 [PDT] 0.04 0.018 6.7 NOC 0.05 0.008 8.82 [DL] 0.02 0.068 3.02

PAGE 150

134 Figure 5-1. Representative developmental stag es of embryos that were collected from Lakes Lochloosa (reference site), Ap opka, and Griffin, and Emeralda Marsh during 2001-2002. A) Live embryo at Day 14 with red line indicating eye length. B) Saggital section of Day 14 embryo. C) Day 25 live embryo. D) Saggital section of Day 25 embryo. E) Day 33 live embryo with red line indicating head length. F) Saggital section of Day 33 embryo. G) Day 43 embryo with red line indicating total le ngth. H) Saggital section of Day 43 embryo (organogenesis nearly complete). 5 mm 5 mm 5 mm 5 mm E. F. G. H. 5mm B. 5 mm A. 5 mm 5 mm C. D.

PAGE 151

-0.80.8-0.60.8 Eye Length Head Length Total Length Embryo mass Heptachlor epoxide% o,p-DDT% Toxaphene% trans-Nonachlor% [Oxychlordane] Figure 5-2. Ordination biplot of embryo morphometric parameters (solid lines) and organochlorine pesticide (OCP) variable s (dashed lines) for embryos collected at chronological age Day 14. Arrows poin ting in the same direction indicate a positive correlation (e.g., embryo mass and [oxychlordane), arrows that are approximately perpendicular indicate near-zero correlation, and arrows pointing in opposite direc tions indicate negative corre lations (head length and [oxychlordane]. Arrow lengths indi cate rank order of correlations. For example, extending a perpendicular line (A) from the embryo mass axis to tip of [oxychlordane] arrow indicates that [oxychlordane] has a stronger positive correlation with embryo mass than hept achlor epoxide% (B). The cosine of the angle formed at the origin betw een individual clutch variables and individual OCP variables is the correla tion coefficient (r). For example, if arrows pointing in exactly opposite dire ctions have an angle of 180, and cos(180) = -1.0, then the arrows would be perfectly, negatively correlated (r) (ter Braak, 1995). A B

PAGE 152

136 -0.60.6-0.10.6 Eye Length Head Length Total Length Embryo mass cis-Nonachlor Dieldrin% Heptachlor expoxide p,p'-DDD% [Dieldrin] Figure 5-3. Ordination biplot of embryo morphometric parameters (solid lines) and organochlorine pesticide (OCP) variable s (dashed lines) for embryos collected at chronological age Day 25

PAGE 153

137 -0.60.6-0.30.4 Eye Length Head Length Total Length Embryo mass cis-Chlordane% cis-Nonachlor% Dieldrin% p,p'-DDD% p,p'-DDE% Figure 5-4. Ordination biplot of embryo morphometric parameters (solid lines) and organochlorine pesticide (OCP) variable s (dashed lines) for embryos collected at chronological age Day 33.

PAGE 154

138 -0.60.6-0.30.4 Eye Length Head Length Total Length Embryo mass NOC [cis-Chlordane] [Dieldrin] [Heptachlor epoxide [p,p'-DDT] Figure 5-5. Ordination biplot of embryo morphometric parameters (solid lines) and organochlorine pesticide (OCP) variable s (dashed lines) for embryos collected at chronological age Day 43.

PAGE 155

139 -0.80.8-0.40.2 Morphological day Eye L.: Head L. Head L.: Total L. Total L.: Emb. M. cis-Nonachlor% o,p-DDT% Oxychlordane% [Heptachlor epoxide] [Oxychlordane] Figure 5-6. Ordination bi plot of derived embryo morphometr ic parameters (solid lines) and organochlorine pesticide (OCP) va riables (dashed lines) for embryos collected at chronol ogical age Day 14.

PAGE 156

140 -0.600.60-0.150.15 Morphological day Eye L.: Head L. Head L.: Total L. Total L.: Emb. M. NOC Heptachlor epoxide% [cis-Nonachlor] [Oxychlordane] [p,p'-DDD] Figure 5-7. Ordination bi plot of derived embryo morphometr ic parameters (solid lines) and organochlorine pesticide (OCP) va riables (dashed lines) for embryos collected at chronol ogical age Day 25.

PAGE 157

141 -0.60.6-0.40.3 Morphological day Eye L.: Head L. Head L.: Total L. Total L.: Emb. M. cis-Chlordane% cis-Nonachlor% Dieldrin% p,p'-DDD% p,p'-DDE% Figure 5-8. Ordination bi plot of derived embryo morphometr ic parameters (solid lines) and organochlorine pesticide (OCP) va riables (dashed lines) for embryos collected at chronol ogical age Day 33.

PAGE 158

142 -0.60.6-0.20.8 Morphological day Eye L.: Head L. Head L.: Total L. Total L.: Emb. M. NOC p,p'-DDT% [cis-Chlordane] [Dieldrin] [p,p'-DDT] Figure 5-9. Ordination bi plot of derived embryo morphometr ic parameters (solid lines) and organochlorine pesticide (OCP) variables (dashed lines) embryos collected at chronol ogical age Day 43.

PAGE 159

143 CHAPTER 6 NUTRIENT AND CHLORINATED HYDRO CARBON CONCENTRATIONS IN AMERICAN ALLIGATOR EGGS AND ASSOCIATIONS WITH DECREASED CLUTCH VIABILITY In central Florida, American alligator ( Alligator mississippiensis ) populations inhabiting lakes contaminated with orga nochlorine pesticides (OCPs) have poor reproductive success, primarily due to increa sed embryo mortality ( Woodward et al., 1993; Woodward et al., 1989). During 2000-2002, cl utch viability (percentage of eggs that yield a live hatchling) was monitore d on 168 clutches from reference and OCPcontaminated sites, indicating that clutch es from a reference site, Lake LochloosaOrange, had higher clutch viability (mean clutch viability = 70%) as compared than the OCP-contaminated sites, Lake Apopka (51%), Emeralda Marsh Restoration Area (48%), and Lake Griffin (44%). Furthermore, 115 of these clutches were analyzed for OCPs, and results indicated that allig ators inhabiting Emeralda Rest oration Marsh (total average egg OCPs = 15,480 ng/g), Lake Apopka (7,582 ng/g), and Lake Griffin (1,169 ng/g) contained significantly higher OCP burdens in eggs compared to those of Lake Lochloosa-Orange (102 ng/g) (Chapter 2). Although total embryo mortality was highest in eggs from sites with high OCPs, the am ount of variation in embryo mortality rates explained by OCP egg burdens differs among OCP-contaminated sites (Chapter 2), suggesting the presence of additional factor(s). With respect to vertebrates, examples of non-OCP factors that ha ve been associated with increased embryo mortality include nut ritional deficiencies and excesses (Wilson, 1997; McEvoy et al., 2001), exposure to polyc hlorinated biphenyls (PCBs) (Summer et

PAGE 160

144 al., 1996), and exposure to polyaromatic hyd rocarbons (PAHs) (Hoffman, 1990). For example, early-life stage (embryo) morta lity has been associated with: thiamine deficiency in trout and salmon (Fitzsimons et al., 1999); PCB exposure in chickens (Summer et al., 1996); and PAH exposure in mallard eggs (Hoffman & Gay, 1981). More recent data suggested that thiami ne deficiency may be involved in the increased incidence of embryo mortality in American alligato rs inhabiting the aforementioned OCP-contaminated lakes in central Florida. Indeed, thiamine concentrations in egg yolks were positively co rrelated with clutch viability and accounted for 40% of variation in cl utch viability among Lakes Lochloosa, Griffin, Apopka, and Emeralda Marsh (Seplveda et al., 2004). Howe ver, further investigation into thiamineÂ’s potential role in embryo mort ality is warranted before any conclusions are drawn. Reasons for further study are that only five clutches were sampled per site, sampling occurred during a single nes ting season (2000), and the potentia l role of other nutrients (i.e., vitamin E) and contaminants (i.e., PCBs) were not evaluated. With respect to other vitamins, vitamin E (tocopherol) has been sugge sted as having a potential role in the reduced clutch viability of captive alligators from Louisiana (Lance et al., 1983). Lastly, besides the embryotoxic eff ects of PCBs and PAHs, st udies indicate that these contaminants may reduce thiamine storage in laboratory animals (Yagi et al., 1979), and that the presence of high contaminant burde ns may affect thiamineÂ’s role in the production of metabolic energy (d e Roode et al., 2002a). Toge ther these data suggest the need for a detailed examination of contaminan t burdens and nutrient content of eggs, and their association with clutch viability and embryo mortality in American alligators from OCP-contaminated sites in Florida.

PAGE 161

145 The present studyÂ’s spec ific aims were to conduct a case-control, cohort study to examine the relationship between multiple nutri ents and contaminants, to further examine hypotheses derived from the case-cohort study vi a an expanded field study, and to test hypotheses derived from the expanded field st udy using laboratory experiments. Materials and Methods Egg Collections and Incubation Alligator eggs were collected during 2001, 2002, and 2003 nesting seasons (JuneJuly) from the following OCP-contaminated s ites: Lakes Apopka (N 28 35Â’, W 81 39Â’), Griffin (N 28 53Â’, W 81 46Â’), and Emeral da Marsh Conservation Area ((N 28 55Â’, W 81 47Â’), and from a reference site, Lake Lochloosa (N 29 30Â’, W 82 09Â’) in central Florida. Alligator nests were located via ae rial (helicopter) and ground surveys (airboat), and clutches were subsequen tly collected by ground crews. The top of each egg was marked before eggs were removed from the nest to ensure prope r orientation; thus, preventing embryo mortality due to inversion. Embryo mortalit y due to inversion occurs because, once an embryo has attached to the top of the egg, inverting the eggÂ’s orientation may either break embryonic attachme nt or cause the yolk mass to settle on top of the embryo, crushing it. After marking each egg and placing about 5 cm of nest substrate in a uniquely numbered plastic pan (43 cm x 33 cm x 18 cm), all eggs found in each clutch were placed in the pan 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 collected. The top layer of eggs was covered with nest substrat e so that there was no space left between the top of the pan and the top of the eggs (approx imately 10 cm). Clutches were transported to the US Geological SurveyÂ’s Center for Aqua tic Resources Studies in Gainesville, FL.

PAGE 162

146 Upon arrival, clutches were evaluated fo r embryonic viability using a bright light candling procedure. One or two eggs were opened from each clutch to identify the embryonic stage of development at the time of collection, and to collect yolk samples for later measurement of OCP, PAH, and PCB burde ns and selected nutrie nt content with all yolk and albumin samples being stored at -80 C. From each clutch, the following parameters were measured: total number of eggs per nest (fecundity); number of unbanded eggs, number of damaged eggs, numbe r of dead, banded eggs, number of live banded eggs, total clutch mass, and average e gg mass of clutch. Viable eggs (i.e. having a visible band) were nested in pans cont aining moist sphagnum moss and incubated at 30.5C and ~98% humidity in an incubation bu ilding (7.3 m x 3.7 m). This intermediate incubation temperature will normally result in a 1:1 male/female sex ratio, as alligators have temperature dependent sexual differen tiation (Ferguson, 1985). On a daily basis, temperature and humidity were monitored at several locations thr oughout the incubator, clutches were rotated within the incubator, and air was circulated to mitigate any thermal gradients. Eggs were monitored for viability via bright-light candling every 10 days during incubation. Clutches collected duri ng 2001 and 2002 were used for the field study and those collected during 2003 were used for th e laboratory experiment. Experimental Design Field studies A case-control cohort study was conducted th at involved the selection of clutches based upon their viability and their OCP egg burd ens. Clutches were assigned to one of nine possible categories based on clutch viabil ity and OCP egg burdens (Table 6-1). The purpose of the case-control cohort study was to determine if PAH, PCB, zinc, selenium, vitamins A, E, and B1 concentrations differed or show ed trends among clutch viability-

PAGE 163

147 OCP categories. Selenium (Spallholz & Hoffman, 2002), zinc, vitamins A, E, (Ashworth & Antipatis, 2001) and B1 (de Roode et al., 2002b)were examined because they are important for embryo development and survival, and their activity and/or levels may be affected by chlorinated hydrocarbons. This strategy aided in forming hypotheses related to the association between non-OCP factors and embryo mort ality and the relationship between non-OCP factors and OCP exposure. Fo r example, if increasing levels of a nonOCP factor showed a strong pos itive association with embr yo viability, regardless of OCP burden, and levels did not differ betw een OCP exposure groups, then it could be hypothesized that the potential effects were likely related to the non-OCP factor(s) and unrelated to OCP exposure(s). Conversely, if increasing leve ls of the non-OCP factor(s) showed a strong positive associ ation with embryo viability, bu t only with respect to low OCP exposure groups, then it could be hypothesize d that the potential effects were likely due to a combination of OCP exposure and the level of the non-OCP factor(s). Based on the case-control cohort study, hypot heses were derived that focused on the major non-OCP factors associated with em bryo mortality and OCP exposure. Lastly, results of this expanded field study were used to design an egg treatment experiment to examine the hypotheses in a more controlled setting. Laboratory experiments In 2003, laboratory experiments were c onducted using clutches collected from Lakes Dexter and Griffin, and from Emeral da Marsh. Based upon the case-control cohort study and the expanded field study (see results), the purpose of this experiment was to test the hypothesis that increas ing thiamine levels in eggs would result in decreased embryo mortality. This experiment consisted of increasing thiamine concentrations in eggs that were known to have low thiamine concentrations, moderate to high yolk OCP

PAGE 164

148 concentrations (Lakes Griffin and Emeral da Marsh), and high embryo mortality. Thiamine HCL was applied at high (60 mg thiamine/mL dimethyl sulfoxide, DMSO) and low concentrations (12 mg/mL DMSO) over th e surface of each egg (application volume of 50 l) using a micropipette. Controls receiv ed only vehicle treatment (DMSO). These doses were calculated to achie ve yolk thiamine concentrations similar to those measured from the reference site (Lakes Orange-Lochl oosa complex). Eggs from each site were labeled and randomLy distributed among each treat ment group, so that all clutches were equally represented in the study. There were two repli cates per treatment with a minimum of 26 eggs (maximum of 31) per re plicate. After being dosed, eggs were placed in the incubator, and candled weekly to determine effects on embryo and hatchling survival. For Emeralda Marsh clutches, thre e eggs from each replicate were sampled 7 days after treatment to determine the amount of thiamine present in albumin and yolk. Embryo mortality rates were recorded for each treatment group as the percentage of eggs failing to hatch over the number of eggs treated. The second experimentÂ’s purpose was to test the hypothesis that, in the absence of high OCP exposure, decreased thiamine bioa ctivity (functional deficiency) would result in increased embryo mortality rates. This experiment involved inducing decreases in thiamine bioactivity in eggs known to have relatively high thiamine concentrations, low OCP burdens, and low embryo mo rtality. Since clutches fr om Lake Lochloosa-Orange were assigned to another study during 2003, eggs were collected from another reference site (Lake Dexter, N 29 98Â’, W 81 47Â’). To decrease thiamine bioactivity, oxythiamineHCL, a thiamine antagonist (Akerman et al., 1998), was topically applied at concentrations of 12 or 60 mg/mL using DMSO as the carrier. Cont rols received only

PAGE 165

149 DMSO. Eggs were labeled and randomLy distributed among each treatment group, so that all clutches were equally represented. There were two replicates per treatment with a minimum of 20 eggs (maximum of 21) per re plicate. After being dosed, eggs were placed in the incubator, and candled weekly to determine effects on embryo and hatchling survival. Hatch rates for each replicate were de termined as the percentage of eggs that produced a live hatchling. Analysis of Chlorinated Hydrocarbons in Yolk Analytical grade standards for the following compounds were purchased from the sources indicated: aldrin, al pha-benzene hexachloride ( -BHC), -BHC, lindane, -BHC, p,pÂ’ -dichlorodiphenyldichloroethane ( p,pÂ’ -DDD), p,pÂ’ -dichlorodiphenyldichloroethylene ( p,pÂ’ -DDE), dichlorodiphe nyltrichloroethane ( p,pÂ’ -DDT), dieldrin, endosulfan, endosulfan II, endosulfan sulfate, endrin, e ndrin aldehyde, endrin ketone, heptachlor, heptachlor epoxide, hexachlorobenz ene, kepone, methoxychlor, mirex, cis -nonachlor, and trans -nonachlor from Ultra Scientific (Kingstown, RI, USA); cis -chlordane, trans chlordane, and the 525, 525.1 polychlorinated biphenyl (PCB) Mix from Supelco (Bellefonte, PA, USA); oxychlordane from Chem Service (West Chester, PA); o,pÂ’DDD, o,pÂ’DDE, o,pÂ’DDT from Accustandard (New Haven, CT, USA); and toxaphene from Restek (Bellefonte, PA, USA). All reag ents were analytical grade unless otherwise indicated. Water was doubly distilled and deionized. Egg yolk samples were analyzed for chlorinated hydrocarbon content using methods modified from Holste ge et al. (1994) an d Schenck et al. (1994). For extraction, a 2 g tissue sample was homogenized with ~1 g of sodium sulfate and 8 mL of ethyl acetate. The supernatant was decanted and filtered though a Bchner funnel lined with Whatman #4 filter paper (Fisher Scientific, Ha mpton, NH, USA ) and filled to a depth of

PAGE 166

150 1.25 cm with sodium sulfate. The homogenate was extracted twice with the filtrates collected together. The combined filtrat e was concentrated to ~2 mL by rotary evaporation, and then further concentrated until solvent-free unde r a stream of dry nitrogen. The residue was reconstituted in 2 mL of acetonitrile. After vortexing (30 s), the supernatant was applied to a C18 solid phase extrac tion (SPE) cartridge (preconditioned with 3 mL of acetonitrile; Agile nt Technologies, Wilmi ngton, DE, USA) and was allowed to pass under gravity. This proced ure was repeated twice with the combined eluent collected in a culture t ube. After the last addition, th e cartridge was rinsed with 1 mL of acetonitrile which was also collected. The eluent was then applied to a 0.5 g NH2 SPE cartridge (Varian, Harbor City, CA, US A), was allowed to pass under gravity, and collected in a graduated conical tube. The car tridge was rinsed with an additional 1 mL portion of acetonitrile whic h was also collected. The combined eluents were concentrated under a stream of dry nitrogen, to a volume of 300 L, and transferred to a gas chromatography (GC) vial for analysis. GC/MS Analysis Analysis of all samples was performed using a Hewlett Packard HP-6890 gas chromatograph (Wilmington, DE, USA) with a split/splitless inlet ope rated in splitless mode. The analytes were introduced in a 1 L injection and separa ted across the HP-5MS column (30 m x 0.25 mm; 0.25 m film thickne ss; J & W Scientific, Folsom, CA, USA) under a temperature program that began at 60 C, increased at 10 C/min to 270 C, was held for 5 min, then increased at 25 C/min to 300 C and was held for 5 min. Detection utilized an HP 5973 mass spectro meter in electron impact m ode. Identification for all analytes and quantitation for toxaphene was c onducted in full scan mode, where all ions

PAGE 167

151 are monitored. To improve sensitivity, se lected ion monitoring was used for the quantitation for all other analytes, except kepone. The above program was used as a screening tool for kepone which does not optim ally extract with mo st organochlorines. Samples found to contain kepone would be reex tracted and analyzed specifically for this compound. For quantitation, a five-point standard curve was prepared for each analyte ( r2 0.995). Fresh curves were analyzed with each se t of twenty samples. Each standard and sample was fortified to contain a deuterat ed internal standard, 5 L of US-108 (120 g/mL; Ultra Scientific), added just prior to analysis. All samples also contained a surrogate, 2 g/mL of tetrach loroxylene (Ultra Scientific) added after homogenization. Duplicate quality control samples were prepar ed and analyzed with every twenty samples (typically at a level of 1.00 or 2.50 g/mL of -BHC, heptachlor, aldr in, dieldrin, endrin, and p,p’ -DDT) with an acceptable recovery rangi ng from 70 – 130%. Limit of detection ranged from 0.1-1.5 ng/g for all OCP analyt es, except toxaphene (120-236 ng/g), and limit of quantitation was 1.5 ng/g for all anal ytes, except toxaphene (1500 ng/g). Repeated analyses were conducted as allo wed by matrix interferences and sample availability. Nutrient Analysis Thiamine concentrations were measured in clutches coll ected during years 2001, 2002, and 2003. For analysis, samples were shipped overnight on dry ice (solid CO2) to the USGS Leetown Science Center, Appalach ian Research Laborator y in Wellsboro, PA. Thiamine concentrations were determined as described in (Brown et al., 1998). Briefly, a known amount of the frozen yolk sample was first placed in 2% trichloroacetic acid (TCA, Sigma, St. Louis, Missouri, USA) ho mogenization solution. The extract was then

PAGE 168

152 washed with ethyl acetate:he xane (3/2, vol/vol, Sigma) to remove excess TCA. An aliquot of the washed solution was reacted with potassium ferricyanide (Sigma) to produce thiochrome derivatives. The resulti ng derivatives were separated on a Hamilton PRP-1 column (Alltech, Deerfield, Illinois, USA) and detected with a spectrofluorometer set at 375 nm excitation wavelength a nd 433 nm emission (Shimadzu, Columbia, Maryland, USA). Authentic standards of thiamine pyrophosphate, thiamine monophosphate and thiamine-HCL (ICN Biomed icals, Montreal, Quebec, Canada) were used to quantify the amount of thiamine in each sample. In addition to thiamine analyses, samples from selected clutches collected during 2002 were sent to ABC Research Corp. in Ga inesville, FL, and analyzed for vitamin A (carotene, retinol, and activity) ( AOAC 960.45 and 941.15), vitamin E (tocopherol) (AOAC 948.26), zinc and sele nium (AOAC 990.8) using AOCAC methods (Horwitz, 2000). Data Analysis For the case-control cohort study, expande d field study, and laboratory experiments ANOVA (PROC GLM; SAS Institute Inc., 2002) was used for inter-site and inter-group comparisons of summary clutch characteristics, with the Tukey test for multiple comparisons among sites and groups ( = 0.05). Because relationships between response variables and explanatory vari ables (Table 6-2) in ecologi cal studies are often complex with interactions occurring, an indirect gr adient multivariate analysis method, Detrended Correspondence Analysis (DCA) (ter Braak, 1986) was used to initially evaluate data structure for the case-control cohort study, as well as the expanded field study. Two matrices were constructed for DCA, with th e first representing th e response variables clutch ID number x clutch parameters) a nd the second represen ting the explanatory

PAGE 169

153 variables (clutch ID number x OCP burdens) (Table 6-3). DCA results indicated that a direct gradient, multivariate linear analys is, redundancy analysis (RDA) (Rao, 1964), was appropriate for the case-control cohort st udy and the expanded field study since the gradient lengths of the DCA ordination axes were never more than (approximately 2 standard deviations (ter Braak, 1995). For the RDA, similar matrices were cons tructed with the exception that response variables measured as a percentage (i.e ., clutch viability) and response variables measured as a number (i.e., clutch mass) were divided into separate matrices because percentage data were ln(x+1) transformed and not standardized, while continuous data were ln(x) transformed and standard ized(ter Braak & Smilauer, 2002). Automatic forward selection of the best f our explanatory variables was for all RDA analyses and Monte Carlo permutation test s were used to determine significance ( = 0.05). DCA and RDA were conducted usi ng the program CANOC O (ter Braak & Smilauer, 2002), and CANODRAW (ter Braak & Smilauer, 2002) was used to construct biplots of environmental vari ables and response variables to interpret relationships between clutch parameters (response va riables) and explanatory factors. Specific OCP analytes were removed from analysis if measurable concentrations were found in less than 5% of all clutches. Numerical data, such as fecundity, were logtransformed [ln(x)], while proportional data (c lutch viability) were arcsine square root transformed to meet statistical assumptions and [ln (x+1)] transformed for RDA analysis

PAGE 170

154 Results Field Study Case-control cohort study In 2002, 32 clutches were collected from Emeralda Marsh, and Lakes Apopka, Griffin, and Lochloosa. Of the 32 clutches 20 were selected and each of the 20 was assigned to one of nine categ ories based on clutch viabilit y and total OCP burdens in eggs (Table 6-1). Only seven categories were filled with six cate gories being represented by three clutches. The remaining category, “good viability-high OCP burden”, was represented by two. Although the number of cl utches within each ca tegory was not large, clutches assigned to good and intermediate vi ability categories had significantly greater viability rates compared to poor category cl utches, which supports the assignment of these clutches to their resp ective categories. Similarly, clutches assigned to high, intermediate, and low OCP categories were sign ificantly different from one another with respect to total OCP burdens, further supporting the statistical and bi ological validity of assigned categories (Table 6-4) Differences among OCP analytes were not determined. In addition to the somewhat expected diffe rences in clutch viability rates and OCP burdens among categories, significant differenc es were found with respect to total PCB burdens, total PAH burdens, thiamine m onophosphate (TP), and thiamine pyrophosphate (TPP) in eggs (Table 6-4). Although total PCB and PAH burdens differed among categories, levels were below those known to elicit adverse effect s on avian development (Summer et al., 1996). Vitamin A was not detect ed in any of the eggs with the lack of detection likely due to the relatively higher limit of detec tion (0.3 ppm) compared to the other nutrients (e.g., 0.1 ppm for vitamin E) Since vitamin A was not detected or

PAGE 171

155 quantified in any eggs, no conclusions can be reached regarding its potential role in embryo mortality in alligators. No other significant differences were not ed for non-OCP variables likely due in part to the relatively small sample sizes and considerable variation in values of clutch parameters. Since the purpose of the study was to develop hypothe sis; some important non-significant differences shoul d be pointed out. For exam ple, mean values of total thiamine and free thiamine concentrations of good viability-low OCP clutches and poor viability-high OCP clutches were nearly fou r-fold those of intermediate viability-low OCP clutches (Table 6-4). This four-fold di fference may suggest that reduced viability in clutches with low OCP burdens may be associat ed with reduced thiamine levels, and that poor viability in clutches with high OCP bur dens may not be associated with reduced thiamine levels. Redundancy analysis (RDA) with forward selection of best four explanatory variables (Table 6-3) provided a way to ev aluate the relationships between the non-OCP variables and clutch variables, and allowed each clutchÂ’s site to be included in the analysis. Including site in th e analysis aided in identifying whether site differences, as opposed to other factors, were related to clutch survival a nd related parameters. For the 20 clutches included in the RDA, thiami ne monophosphate, TP, (lambda A = 26%) and thiamine pyrophosphate, TPP, (12%) were signifi cantly correlated with clutch survival parameters, accounting for 38% of th e variation in clutch survival parameters (Table 6-5). Indeed, TP had a strong positive associat ion with clutch viability and a strong negative association with early embryo mort ality, while TPP had a strong negative association with late embr yo mortality (Fig. 6-1).

PAGE 172

156 These results are biologically plausible because TP and TPP are the bioactive forms of thiamine needed for the production of me tabolic energy and deficiencies have been associated with intrauterine growth retarda tion in laboratory models (Roecklein et al., 1985). In contrast, the positive relationshi p between unbanded egg% and TPP (Fig. 6-1) has little biological implications because an embryo must be present for TPP to be produced. Important to note is that PAH and PCB burdens did not appear to be significantly associated with em bryo survival parameters. In contrast to clutch su rvival parameters, clutch si ze parameters (e.g., fecundity) appeared to be associated with the site, as three of the four extracted explanatory variables were the nominal site variables. Of these extracted variables, only Lochloosa was determined to be significantly associated with clutch size pa rameters (Table 6-6), accounting for 27% of the variation. Furtherm ore, Lochloosa clutches appear to have higher average egg masses and lower fecundity compared to other sites (Fig. 6-2). Lastly, the relationship between nutrien ts and chlorinated hydrocarbons were examined via RDA. Interestingly, all four extracted explanatory variables, heptachlor epoxide concentration (lambda A = 18%), dieldrin% (17%), trans-chlordane concentration (15%), Lochloosa (site eff ect, 9%) were found to be significantly associated with nutrient levels in eggs, accounting for 59% of the variation in egg nutrient content (Table 6-7). Heptachlor e poxide concentrations had a strong negative correlation with thiamine pyrophosphate, but weak positive correlations with the other thiamine forms and nutrients. Dieldrin% had strong negative asso ciations with free thiamine and total thiamine, but strong positive relationships with vitamin E, zinc, and selenium. Trans-chlordane concentrations ha d strong negative correlations with vitamin

PAGE 173

157 E, zinc, and selenium and little to near-zero correl ation with thiamine concentrations. Lochloosa clutches appeared to be associat ed with increasing thiamine concentrations and decreasing zinc, selenium, and vitamin E concentrations (Fig. 6-3). In summary, results of the case-contro l cohort study suggest the main factors associated with reduced clutch viability a nd increased embryo mortality are decreasing thiamine concentrations (Table 6-5, Fig. 6-1), and that reductions in thiamine concentrations may be associated with orga nochlorine pesticides (T able 6-7, Fig. 6-3). Therefore, the expanded field study was de signed in order to examine how clutch survival parameters vary as a function of OCP burdens and thiamine concentrations in eggs. Expanded field study The purpose of the expanded field study was to examine the relationships between thiamine and OCP concentrations in eggs and clutch viability and cl utch size parameters. Since consistent methods were used for OCP and thiamine analysis, as well as for egg collections and incubation, data from year 2000 (Seplveda et al., 2004), was combined with data from years 2001 and 2002. Using a larger number of clutches (n = 72) over multiple nesting seasons increased ecological validity of conclusions, as well as power in testing the hypothesis that th iamine deficiency and OCP ex posure are associated with altered clutch survival parameters and altered clutch size. Clutches from the Lochloosa-Orange co mplex (n = 18), Emeralda Marsh (n = 19), and Lakes Apopka (n = 14) and Griffin (n=21) No significant di fferences were noted among sites with respect to clutch survival pa rameters, clutch size pa rameters, or the four thiamine parameters. However, biologi cal significance should be noted in that Lochloosa-Orange complex clutches had m ean clutch viability rates that were

PAGE 174

158 consistently greater than all other sites by an average of 18%. Furthermore, LochloosaOrange clutches had lower embryo mortality rates that were less than all other sites, by an average of 8%. The paucity of statistically significant differences is likely due to the high variance of clutch survival in OCP-contamin ated sites. Significant differences were found among sites with respect to many OCP anal yte burdens in eggs. Indeed, mean total OCP burdens and number of OCP analytes dete cted at quantifiable levels significantly differed among all sites (Table 6-8). RDA was used to evaluate the relationships between the many OCP and thiamine parameters (explanatory variables) and th e clutch survival parameters (response variables) (Table 6-3). Initial RDA showed that embryo age at the time of collection was an important factor, but not a specific factor of interest. Further examination of age effects indicated that for all sites, phosphorylation of free thiamine increased with age (Fig. 6-4). Therefore, another RDA was conducted usi ng age as a covariate. The best four explanatory variables determined via this RDA accounted for 30% of the variation in clutch survival parameters and consisted of total thiamine concentration (lambda A = 16%), thiamine pyrophosphate (7%), thiamine monophosphate (4%), and methoxychlor% (3%), with all explanatory variables dete rmined to be significant (Table 6-9). Total thiamine (TT) and thiamine monophosphate (TP) were strongly and positively correlated with clutch viability, and negatively correlated with unbanded egg% and early embryo mortality but showed near-zero correlation with late embryo mortality. Thiamine pyrophosphate (TPP) was strongly and negatively correlated with late embryo mortality and had weak to near-zero corre lations with remaining clutch survival

PAGE 175

159 parameters. Methoxychlor% (ME%) had posit ive correlations with unbanded egg% and early embryo mortality and near -zero correlations with other clutch survival parameters (Fig. 6-5). In addition to clutch survival parameters redundancy analysis was used to examine relationships between clutch size variables a nd explanatory variable s. Results of the RDA indicated that two of four extracted variables were found to be significant and explained 15% of the variation in clutch size parameters. Extracted variables found to be significantly associated with clutch size va riables included free thiamine (lambda A = 9%) and thiamine pyrophosphate (6 %). Site effect may be im portant regarding variation in clutch size parameters, as the nominal variable “GR” (Lake Griffin) approached significance (Table 6-10). Interestingly, all thiamine forms were pos itively associated with egg weight and negatively associated with fecundity. Thia min pyrophosphate was negatively correlated with Griffin (meaning clutches from Lake Gr iffin had reduced levels of TPP), and had near-zero correlations with clutch mass. Total thiamine and free thiamine had strong, negative correlations with clutch mass and n ear-zero correlations with GR (Fig. 6-6). Lastly, redundancy analysis was used to examine the relationship between the various thiamine forms (response variables) and explanatory va riables to see if thiamine deficiency was associated with OCP variables or other clutch variables. Results indicated that four extracted variable were significan tly correlated with thiamine concentrations and accounted for 31% of the variation in thiami ne levels. Interestin gly, lipid content of eggs (%) accounted for 16% of thiamine variation, followed by mirex concentrations (5%), trans-chlordane concentr ations (6%), and oxychlordane concentrations (4%) (Table

PAGE 176

160 6-11). Lipid content had strong negative co rrelations with free thiamine and total thiamine and positive correlations with thiamine monoand pyrophosphate. Transchlordane concentrations were positively correlated with thiamine pyrophosphate and negatively correlated with the remaining th iamine forms. Oxychlordane and mirex concentrations in eggs were positively correlated with thiamine monophosphate, had near-zero correlations with total and free th iamine, and weak negative correlations with thiamine pyrophosphate (Fig. 6-7). In summary, results of the expande d field study suggested that thiamine concentrations and certain OCP variables accounted for a signif icant amount of the variation in clutch survival and size char acteristics, supporting the hypothesis that thiamine deficiency and OCP exposure contri butes to decreased clutch viability and altered clutch size characteristics (Figs. 66; 6-5). Furthermore, decreasing thiamine levels were associated with increasing lip id content in egg yolks, suggesting that alterations in yolk composition are occurring and may be indicative of altered maternal liver function, possibly due to a number of reasons including OCP exposure and female age. In addition, alterations may be related to dietary fact ors. Indeed, altered liver function, leading to altered yolk composition ha s been documented in laboratory studies in catfish exposed to similar pesticides (Lal & Singh, 1987), and diets and body condition of alligators have been suggested to differ am ong two of the lakes included in the present study (Lakes Apopka and Griffin) in central Florida (Rice, 2004). Laboratory Experiments Because the case-control cohort and expanded field studies supported the hypothesis that thiamine deficiency and OCPs ar e associated with altered clutch survival and clutch size parameters, two laboratory ex periments were conducted to more directly

PAGE 177

161 test this hypothesis. The purpose of the firs t experiment was to te st the hypothesis that increasing in ovo thiamine concentrations would in crease embryo survival (thiamine topical exposure experiment ) in clutches with high OCP burdens, and the second experiment tested the hypothesis that decrea sing thiamine concentrations (via thiamine activity inhibitor) would decr ease embryo survival in clut ches with low OCP burdens. A total of 14 clutches were used in the two experime nts, with clutches having relatively high embryo mortality, low thiamine levels in eggs, and intermediate (Lake Griffin, n = 5) to high OCP burdens in eggs (E meralda Marsh, n = 5) used in the thiamine topical exposure study. Convers ely, clutches (Lake Dexter, n = 4) having relatively low embryo mortality, high thiamine levels in eggs, and low OCPs were used in the oxythiamine (thiamine-antagonist) topical expos ure study. Clutch characteristics differed significantly among sites with respect to f ecundity, clutch mass, egg mass, many OCP analytes, total OCP burdens, and number of OC Ps detected at quantifiable levels (Table 6-12). Seven days after topical treatment, three eggs from three different clutches (same clutches sampled for all replicates) were anal yzed to determine the amount of thiamine that was transferred into the egg. For Emeralda clutches, results indicated that total thiamine concentrations in egg albumin of the high and low thiamine treatment groups were significantly greater than controls. Ind eed, total thiamine concentrations in albumin of the high thiamine and low thiamine treatment groups were over 40-fold and over 30fold greater, respectively, than those of c ontrols, confirming a significant increase in thiamine levels in these eggs. Thiamine con centrations in egg yolk of Emeralda clutches were also greater in high and low thiamine treatment groups in a dose-dependent manner,

PAGE 178

162 but the difference was not significant, with thiamine concentrations in high treatment groups being only 1.1-fold greater than controls (Table 6-13). Similar results were noted for Lake Griffin clutches, with thiamine tr eatments showing a dose-dependent, but nonsignificant increase in thiamine concentrati ons. High treatment clutches had thiamine concentrations that were 1.2-fold those of controls (Table 6-13). Thiamine concentrations in egg yolk of control groups for both Emeralda and Gr iffin clutches were high than means reported in the expanded fiel d study but still within respective ranges. Changes in embryo mortality rates were the primary interest and analysis indicated no significant differe nces were noted between th iamine treatment groups and controls. However, Emeralda clutches fr om both thiamine treatment groups had embryo mortality rates which averaged 10% less than those of controls. However, for Lake Griffin clutches, thiamine treatments were associated with a 5-7% increase in embryo mortality (Table 6-13). For the oxythiamine (thiamine antagonist) study using Lake De xter clutches, no significant differences were noted between oxythiamine treatment groups and controls. Surprisingly, embryo survival of controls (mean standard error: 81 9%) was slightly less than those of the low exposure groups (98 2%), and high exposure groups (88 8%). Oxythiamine concentrations were not measured because oxythiamine has physicochemical properties very similar to thiamine ; therefore, transf er rates across the eggshell were assumed to be similar. Discussion The present study examined associations between egg nutrients, OCP egg burdens, and clutch survival and size characteristics us ing a three-tiered approach that identified potentially important associations, and th en more rigorously examined hypothesized

PAGE 179

163 associations using large field studies and laboratory experiments. The first tier of the present study, case-control cohor t study, suggested that PAH and PCB concentrations as well as non-thiamine nutrients, were not likely to be the cause of decreased clutch viability as their levels did not show larg e differences across sites, nor were they significantly associated with alter clutch su rvival parameters. In addition, the casecontrol study indicated that th iamine concentrations were significantly associated with clutch survival parameters and that the associ ation suggested decreased thiamine levels in eggs were associated with decreased clutch viability, which is consistent with similar studies involving fish ((Fitzsim ons et al., 1999). Lastly, as dieldrin% increased (i.e., the proportion of total OCP burden composed by dieldrin) thiamine levels decreased, suggesting that OCPs may be indirectly i nvolved in decreased clutch viability via thiamine reduction, as OCP exposure has been suggested to decrease thiamine concentrations in laboratory models (Yagi et al., 1979). Results of the expanded field study provi ded more support for the hypothesis that thiamine deficiency may be involved in decr eased clutch viability and that OCP burdens and lipid content were significantly associated with variation of thia mine concentrations. However, the laboratory experiments, ove rall, did not support the hypothesis that thiamine is related to embryo viability in al ligators as thiamine amelioration or inhibition did not altered embryo mortality rates. One poten tial reason for the lack of effects is that thiamine levels were already sufficient fo r adequate embryo survival and therefore increasing concentrations were biologically irrelevant. With the thiamine antagonist experiment, two potential reasons for the la ck of effects are that oxythiamine may not have transferred into the yolk compartment and/or the concentra tion was not high enough

PAGE 180

164 to inhibit thiamine activity to the point that effects were elicited. Although thiamine and oxythiamine treatment experiments were ineff ective in the present study, similar studies involving in ovo treatments of fish eggs have been effective in demonstrating the effects of induced thiamine deficiency and thiami ne amelioration on embryo and fry survival (Fitzsimons et al., 2001). In conclusion, decreasing thiamine levels in eggs may be associated with decreased clutch success and lipid content, and OCP burde ns may be associated with variation in thiamine concentrations. However, it should be noted that thiamine levels in eggs only explained 38% of the variation in clutch su rvival parameters in the case-control cohort study and 27% of the clutch survival varia tion in the expanded fi eld study, which suggest that other factors are likely invo lved as well. Because of th e lack of effects observed in the experimental studies, future studies should try to induce th iamine deficiency in eggs through maternal dietary restriction, especially since embryos are at a relatively advanced stage of development by the time oviposition oc curs (Clarke, 1891). A concurrent study involving a captive adult alligator breeding popula tion will be able to control for diet and examine relationships between maternal OC P exposure and thiamine levels in eggs.

PAGE 181

165 Table 6-1. Classification matrix for clutches coll ected during 2002. Total OCP Burdena Clutch Viability >3700 3700 x > 350 350 (10071%) Good viab./High OCP Good via b./ Inter. OCP Good viab./Low OCP (70-48%) Inter. viab./High OCP Inter. viab./ Inter OCP Inter viab./Low OCP (47-0%) Poor viab./High OCP Poor via b./ Inter. OCP Poor viab./Low OCP ang/g yolk wet weight. Good and High = grea ter than mean + 1 standard deviation, Intermediate = mean 1 standard deviati on, Low and Poor = less than mean–1 standard deviation. Table 6-2. Reproductive, morphometric, a nd contaminant parameters measured on clutches of alligator eggs collected during summer 2000, 2001, and 2002. Parameter Definition Measured as Response variables Fecundity Total No. of eggs in one clutch n Clutch mass Total mass of eggs in one clutch kg Ave. Egg Weight Clutch mass / Fecundity g Unbanded eggs% a No. of unbanded eggs / fecundity x 100 Percentage Early embryo mort.% No. of deaths < dev. Day 35 / fecundity x 100 Percentage Late embryo mort.% No. of deaths dev. Day 35 / fecundity x 100 Percentage Clutch Viability No. eggs yielding live hatchling / fecundity x 100 Percentage Explanatory variables [OCP analyte] in ng OCP analyte / g egg yolk wet weight ppb OCP analyte% [OCP analyte] / [OCP] x 100 Percentage [PCBs] in egg yolk ng PCBs / g egg yolk wet weight ppb [PAHs] in egg yolk ng PAHs analyt e / g egg yolk wet weight ppb Thiamine in egg yolkb Pmoles / g egg yolk wet weight pmol/g Zn, Se, Vit. A, Ec ng analyte / g egg yolk wet weight ppb aAn egg with no evidence of embryonic attachment. bThiamine was measured in pmoles because various bioactive forms of were measured. cThese analytes were only measured in clutches collected during year 2002.

PAGE 182

166 Table 6-3. Explanatory variab les included in RDA with forw ard selection of four best variables for case-control cohor t and expanded field studies. Variablea Code Embryo age at time of collection Age Lake Griffin GR Lake Apopka AP Lake Lochloosa-Orange LO Emeralda Marsh EM No. OCP analytes at measurable levels NOC [OCP] TOC % Aldrin ALD% [Aldrin] [ALD] % cis -Chlordane CC% [ cis -Chlordane] [CC] % cis -Nonachlor CN% [ cis -Nonachlor] [CN] % Dieldrin DL% [Dieldrin] [DL] % Heptochlor epoxide HE% [Heptachlor epoxide] [HE] %Lipid content LPC% % Mirex MX% [Mirex] [MX] % o,p -DDT ODDT% [ o,p -DDT] [ODDT] [Methoxychlor] [ME] % Methoxychlor ME% % o,p -DDD ODDD% [ o,p -DDD] [ODDD] % Oxychlordane OX% [Oxychlordane] [OX] % p,p '-DDE PDDE% [ p,p '-DDE] [PDDE] % p,p '-DDD PDDD% [ p,p '-DDD] [PDDD] % p,p '-DDT PDDT% [ p,p '-DDT] [PDDT] % trans -Chlordane TC% trans -Chlordane [TC] % trans -Nonachlor TN% [ trans -Nonachlor] [TN] % Toxaphene TX% [Toxaphene] [TX] PCBs [PCB] PAHs [PAHs] Free Thiamine FT

PAGE 183

167 Table 6-3. Continued. Variablea Code Thiamine monophosphate TP Thiamine pyrophosphate TPP Vitamin E Vit.E Zinc Zn Selenium Se aFor the case-control cohort study, no OCP va riables were included in RDA involving clutch survival or size parameters, since clutches were selected a priori based on total OCP egg burdens. OCP variables were incl uded in the RDA evaluating the relationship between egg nutrients and chlorinated hydro carbons. For the expanded field study, only thiamine variables and OCP variables were in cluded after it was determined they were the more important explanatory factors (see results).

PAGE 184

168Table 6-4. Summary of cl utch parameters on clut ches collected during 2002. Parametera Good-High Good-Int. Good-Low Int.-Low Poor-High Poor-Int. Poor-Low No. Clutches 2 3 3 3 3 3 3 Fecundity ( n ) 51 5 49 1.9 41 2 49 5.2 55 5.8 46 4.5 47 0.9 Clutch mass 4 0.5 3 0.7 4 0.2 4 0.4 4 0.4 4 0.6 4 0.1 Egg mass (g) 86 0.8 69 12.5 88 6 86 1.5 76 6.4 78 6.2 79 0.3 Clutch viability 92 3.4 A 80 3.8 A 79 4.3 A 60 5 A 11 9.4 B 15 10.9 B 25 13 B Damaged eggs 0 0 1 1.3 0 0 0 0 2 1.1 1 1.1 0 0 Unbanded eggs 6 5.7 6 1.5 10 4.9 15 4.2 25 10.5 5 2.7 21 6.8 Early emb. mort. 2 2.3 10 5.1 12 8.1 7 3.3 54 21.6 36 26.9 30 3.9 Late emb. mort. 0 0 3 1.4 0 0 18 4 8 4.3 42 21.8 24 16.9 Dieldrin 248 20.2 72 47.8 6 1.1 9 5 157 66 264 134.8 16 5.4 Hep.Epoxide 3 0.8 11 4.2 4 2.3 2 0.5 6 2.6 4 1.8 5 1.7 cis-Chlordane 161 14.9 15 1.1 4 1.3 7 3.6 145 69.7 25 9.2 13 0.8 cis-Nonachlor 82 11.2 24 4.1 7 1.3 9 4.2 89 38.4 21 6.3 11 0.8 Oxychlordane 23 4.1 25 13.4 8 5 4 1.4 31 6.8 14 5.2 7 2.8 Toxaphene 10289 313.6 0 0 0 0 0 0 4670 1268.9 1928 0 0 0 p,p'DDD 2614 348.7 10 4.9 3 0.6 4 0.7 897 578.1 18 4.8 4 0.8 p,p'-DDE 19136 3277.6 1167 830.3 139 30.9 117 45.6 9149 3668.7 1019 795 153 28.6 p,p'-DDT 24 0 0 0 0 0 0 0 10 2.1 0 0 0 0 trans-Chlordane 51 1.5 1 0 1 0 1 0 34 15.3 1 0 2 0.5 trans-Nonachlor 251 46 55 13.9 16 4.7 15 5.8 273 141.4 41 15.8 22 2.8 [OCPs] 32959 3891.1 A 1391 923 B 188 41.9 C 168 67.9 C 15508 5426.2 A 2057 770.2 B 234 42.5 C NOC 14 0 11 0 10 0.7 10 0.9 14 0.3 11 0.7 11 0 [PAHs] 21 1.3 BC 33 2.8 AB 34 7.6 AB 31 2.9 AB 18 2.5 C 27 3.5 ABC 43 7.3 A [PCBs] 39 0 B 168 28.9 A 83 14.6 A 111 47 A 40 2.5 A 55 10.7 A 92 16 A Selenium 1000 100 1233 176.4 1067 66.7 1133 166.7 1000 57.7 933 120.2 833 145.3 TP 23 5.7 AB 13 1.6 AB 41 8.4 A 24 1.7 AB 2 1.8 C 12 6.7 BC 13 6.4 AB TPP 14 3.6 A 0 0.3 C 21 3.1 A 16 2.1 A 18 8.1 AB 4 3.8 BC 7 5.3 AB FT 463 92.6 719 427.9 868 231.1 238 40.6 891 261.2 553 482.3 326 115.4 Thiamine 500 83.3 733 427.7 931 220.0 278 38.9 912 267.4 569 486.5 345 104.3 Vit. E 16287 2389.7 26397 2978.6 19118 7352.9 12255 122.5 21054 1889.2 20025 5831 15221 1784.5 Zinc 15900 300 15433 809 24900 10570.9 15300 750.6 15267 437.2 12367 800.7 12967 788.1 aCodes for parameters are listed in Table 6-3.

PAGE 185

169 Table 6-5. Evaluation of the relationship betw een concentrations of nutrients, PAHs, and PCBs in eggs and clutch survival parameters via RDA analysis ( = 0.05). Variable Lambda A P F Thiamine monophosphate 0.26 0.004 6.28 Thiamine pyrophosphate 0.12 0.014 3.95 Free thiamine 0.08 0.102 2.1 Thiamine forms 0.08 0.142 2.08 Table 6-6. Evaluation of clut ch size parameters and explan atory factors for clutches collected during 2002. Variable Lambda A P F Lochloosa 0.27 0.018 6.77 Apopka 0.06 0.198 1.54 PCB concentrations 0.05 0.272 1.25 Emeralda Marsh 0.07 0.164 1.81 Table 6-7. Evaluation of the relationshi p between nutrient co ncentrations and explanatory variables for cl utches collected during 2002. Variable Lambda A P F Heptachlor epoxide conc. 0.18 0.006 4.07 trans-Chlordane conc. 0.15 0.022 3.67 Dieldrin% 0.17 0.002 5.43 Lochloosa 0.09 0.048 3.13

PAGE 186

170 Table 6-8. Summary and comparison of para meters measured on clutches collected during 2000-2002. Parametera Loch.-Orange Griffin Apopka Emeralda No. clutches 18 21 14 19 Fecundity 40 1.8 45 1.8 47 2 46 1.8 (26–56) (24–58) (31–56) (34–64) Clutch mass 3.6 0.18 3.5 0.19 4 0.2 4 0.38 (2.2–4.8) (1.8–4.8) (2.6–4.9) (2.3–9.2) Egg mass 90 3.1 78 2.2 86 3.5 87 7.4 (78–139) (46–89) (67–120) (60–180) Clutch viability 63 5.6 40 6.7 49 8.6 46 9.1 (0–95) (0–87) (0–80) (0–97) Damaged% 4 3.3 5 2.5 2 0.9 5 1.8 (0–60) (0–46) (0–12) (0–27) Unbanded% 11 2 12 2.3 13 3.6 12 3.5 (0–33) (0–32) (0–40) (0–58) Early Emb. Mort. 13 3 26 6.3 17 6.4 25 6.7 (0–36) (0–93) (0–90) (0–95) Late Emb. Mort. 8 2.6 18 4.9 19 6.8 11 3.8 (0–34) (0–58) (0–77) (0–61) TP 20 3.7 19 3.8 20 3.9 19 5.2 (0–52) (0–72) (0–60) (0–67) TPP 12 2.6 7 2.7 8 3.6 11 3.8 (0–31) (0–53) (0–46) (0–54) FT 747 102.8 536 93.1 573 110.8 657 125.9 (77–1324) (109–1570) (50–1412) (57–2171) TT 780 102.8 562 94.2 601 111.1 688 128.5 (77–1364) (152–1583) (62–1431) (57–2212) ALD 0 0 0 0 3 0.1 3 0.3 (0–0) (0–0) (3–3) (3–4) oDD 0 0 C 1 0 B 5 2.3 B 47 5.7 A (0–0) (1–1) (1–9) (8–104) oDT 1 0 B 3 0.4 B 10 2.1 A 301 290.9 A (1–1) (1–6) (1–29) (4–4373) ME 0 0 17 0.3 8 2.6 10 1.9 (0–0) (17–17) (6–16) (6–18) MI 2 0.4 2 0.4 4 1.4 3 0.9 (1–3) (1–4) (1–17) (0–10) DL 4 0.5 D 20 3.4 C 323 66.8 A 186 25.4 A (1–8) (6–70) (24–957) (30–387) HE 3 0.8 C 6 1.4 B 12 2.2 A 6 1.6 B

PAGE 187

171 Table 6-8. Continued. Parametera Loch.-Orange Griffin Apopka Emeralda (1–10) (1–30) (1–30) (0–29) CC 2 0.2 D 11 0.7 C 46 12.2 B 109 15.6 A (1–4) (6–17) (7–179) (15–281) CN 5 0.6 C 18 2.4 B 61 12.4 A 71 8.9 A (2–13) (8–54) (10–171) (17–166) OX 4 1.1 C 11 2.1 B 38 6.1 A 24 3.4 A (1–18) (1–42) (4–72) (3–57) TX 0 0 C 2678 376.5 B 2738 224.5 B 7558 703.6 A (0–0) (1928–3111) (1896–3809) (3216–12975) pDD 2 0.2 D 7 1 C 49 13.3 B 1711 225.1 A (1–3) (3–18) (11–193) (10–2963) pDE 76 12.3 D 283 47.4 C 4576 948.3 B 11304 1872.1 A (28–231) (70–979) (18–13294) (36–33555) pDT 1 0 C 2 0.7 BC 9 3.8 B 15 1.7 A (1–1) (1–2) (1–46) (6–25) TC 3 0.7 BC 2 0.2 C 7 2.2 B 31 4.1 A (1–4) (1–3) (1–27) (3–58) TN 8 1.7 C 37 7.2 B 157 36.8 A 208 30.8 A (3–25) (10–155) (10–532) (14–555) TOC 104 16.2 D 783 264.9 C 6855 1267.2 B 20417 2969.9 A (43–289) (127–4488) (555–18471) (672–53560) NOC 9 0.3 D 11 0.2 C 13 0.4 B 14 0.2 A (7–11) (10–13) (10–15) (13–16) aSee Table 6-3 for parameter codes. Values = mean standard error with range in parentheses. Table 6-9. Evaluation of the relationships between clutch survival parameters and explanatory variables via RDA using age as the covariate. Explanatory Variable LambdaA P F Total Thiamine 0.16 0.002 13.96 Thiamine Pyrophosphate 0.07 0.002 6.14 Thiamine Monophosphate 0.04 0.01 3.56 Methoxychlor% 0.03 0.036 2.71 Table 6-10. Evaluation of the relationships between clutch size parameters and explanatory variables via RDA using age as the covariate. Explanatory Variable LambdaA P F Free Thiamine 0.09 0.012 6.52 Thiamine Pyrophosphate 0.04 0.01 4.82 GR 0.03 0.054 3.36 Total Thiamine 0.04 0.06 3.9

PAGE 188

172 Table 6-11. Evaluation of the relationships between thiamine concentrations and explanatory variables via RDA using age as the covariate. Variable LambdaAP F Lipid content % 0.16 0.002 13.4 Trans-chlordane concentrations 0.06 0.002 6.5 Mirex concentrations 0.05 0.024 4.23 Oxychlordane concentrations 0.04 0.048 3.62

PAGE 189

173 Table 6-12. Site comparisons of parameters measured on clutches collected during 2003. Parameter Dexter Griffin Emeralda No. Clutches 4 5 5 Fecundity 37 3.8 B 46 1.2 AB 46 1.9 A Clutch mass 3 0.5 B 4 0.1 A 4 0.1 AB Egg mass 82 4.1 B 93 2.7 A 80 2.6 B Unbanded eggs 3 2 6 3 5 3.4 Damaged eggs 0 0 1 0.5 4 3.1 Dieldrin 5 1.4 29 9.3 188 78.2 Hep. Epoxide 2 0.4 8 3.1 4 1.5 cis-Chlordane 1 0 B 1 0 B 8 2.6 A cis-Nonachlor 6 1.8 B 22 7 AB 60 19.8 A Oxychlordane 4 1 B 14 5 AB 26 10.7 A Toxaphene 0 0 B 0 0 B 6765 2240.4 A o,p'-DDD 1 0 B 0 0 B 13 0 A o,p'-DDT 1 0 3 0.8 3 0.4 p,p'-DDD 1 0 3 0.4 981 407.8 p,p'-DDE 117 28.1 B 399 114.7 B 13166 5918.5 A p,p'-DDT 1 0 B 2 0.4 B 16 5 A trans-chlordane 1 0 1 0 3 0.6 Endrin ketone 0 0 B 0 0 B 3 0 A Mirex 4 1.3 2 0.3 2 0.5 trans-Nonachlor 10 3.7 B 54 19.9 AB 168 60.3 A OCP burdens 171 38 B 556 159.2 B 21410 8499.4 A No. OCP analytes 12 0.5 B 12 0.4 AB 14 0.5 A

PAGE 190

174 Table 6-13. Comparisons of parameters meas ured on the three thiamine treatment groups during 2003. Treatment Group Site Component Parametera control low high Emeralda Albumin FT (g/ng) 68 9.2 B 2414 464.3 A 3120 56.8 A (58–77) (1950–2878) (3063–3176) TMP 2 0.3 B 7 1.2 A 6 1 A (2–3) (6–8) (5–7) TPP 4 2.5 9 2.5 7 1.2 (1–6) (6–11) (6–8) TT 76 5.6 B 2436 459.2 A 3138 53.7 A (70–82) (1976–2895) (3084–3191) Yolk FT 1047 54.5 1122 80.7 1176 4.3 (992–1101) (1041–1203) (1172–1181) TMP 41 5.5 31 4.1 36 3.1 (35–46) (27–35) (33–39) TPP 21 0.2 18 2.8 24 0.3 (20–21) (15–21) (24–24) TT 1125 60.5 1185 71.8 1254 8.4 (1064–1185) (1113–1257) (1246–1263) Embryo Mort. 25 1.5 12 4.3 18 9.5 (23–26) (8–17) (8–27) Griffin Yolk FT 1041 115.9 1215 29.9 1233 18.7 (925–1157) (1185–1245) (1214–1252) TMP 6 0.6 9 0.4 11 4 (6–7) (9–9) (7–15) TPP 2 0.1 2 0.6 4 0.9 (2–2) (2–3) (3–5) TT 1052 116.7 1229 28.6 1252 24.7 (935–1168) (1200–1257) (1227–1277) Embryo Mort. 36 2.3 43 10.1 41 3.4 (33–38) (33–54) (38–45)

PAGE 191

175 -1.01.0-0.60.6 Clutch viability Unbanded eggs% Early Emb. Mort.% Late Emb. Mort.% TPP TP FT TT Figure 6-1. Biplot of clutch survival parameters and expl anatory factors for clutches collected during 2002.

PAGE 192

176 -1.01.0-0.80.8 Clutch mass Egg mass Fecundity [PCBs] Apopka Emeralda Marsh Griffin Lochloosa Figure 6-2. Biplot of clutch size parameters and explanatory variables for clutches collected during 2002.

PAGE 193

177 -1.01.0-1.01.0 Se Vit. E Zn TPP TP FT TT [HE] [TC] DL% LO Figure 6-3. Biplot of nutrien t concentrations in eggs (s olid arrows) and explanatory variables (dashed arrows).

PAGE 194

178 0% 20% 40% 60% 80% 100% 14253343 Figure 6-4. Relationships between embryo ag e and thiamine phosphorylation in egg yolk for 29 clutches collected during 2002 from Lakes Lochloosa (n = 6), Griffin (n = 10, Apopka (6), and Emeralda Marsh (n = 7). Embr y o a g e ( da y s ) FT TP TPP

PAGE 195

179 -1.00.6-0.61.0 Clutch viability Unbanded egg% Early Emb. Mort. Late Emb. Mort. TPP TP TT ME% Figure 6-5. Biplot of clutch survival parameters and explan atory variables for clutches collected during 2000-2002. See text and Table 6-3 for definition of explanatory variable codes.

PAGE 196

180 -1.01.0-1.00.4 Clutch mass Egg mass Fecundity TPP FT TT GR Figure 6-6. Biplot of clutch size variables (solid lines) a nd explanatory va riables (dashed lines) for clutches co llected during 2000-2002.

PAGE 197

181 -1.00.6-0.40.8 TPP TP FT TT LPC [MI] [OX] [TC] Figure 6-7. Biplot of thiamine egg yolk c oncentrations (solid lines) and explanatory variables (dashed lines) measured on clutches collected during 2000-2003.

PAGE 198

182 CHAPTER 7 REPRODUCTIVE EFFECTS OF ORGANOCHLORINE PESTICIDE EXPOSURE IN A CAPTIVE POPULATION OF AM ERICAN ALLIGATORS (ALLIGATOR MISSISSIPPIENSIS) In central Florida, American alligator ( Alligator mississippiensis ) eggs collected from organochlorine pesticide (OCP) contam inated sites (Lakes Apopka, Griffin, and Emeralda Marsh) contain total concentrati ons of OCPs that range from 4,000-30,000 ng/g yolk wet weight. This is several orders of magnitude greater than the reference sites (Lakes Orange and Lochloosa) (Chapter 2). In addition, alligato r populations inhabiting OCP-contaminated sites have experienced increased embryonic mortality resulting in reduced clutch success (Masson, 1995; Rotste in et al., 2002; Chapter 2). One possible explanation for these increased rates of embryonic mortality is embryonic exposure to OCPs, as similar effects have been reported in birds (Summer et al., 1996). The present study utilized a population of cap tive adult alligators to test the hypotheses that maternal exposure to OCPs would increase OCP burde ns in egg yolks, leading to increased embryonic mortality and decreased hatch rates. Materials and Methods Alligators were obtained from JungleL and Zoo (Kissimmee, FL) and Gatorland Zoo (Orlando, FL). Thirteen male and 14 female adult alligators were randomLy assigned to one of 13 pens (a pproximately 30 m x 30 m) at a ratio of 1 male: 1 female, except for one large pen which housed two females and one male. Seven pens were designated as treatment pens and six as cont rol. Prior to random group assignment (i.e.,

PAGE 199

183 control vs. OCP group), head length, total leng th, tail girth, and estimated mass [calculated using total lengt h and tail girth (Woodward et al., 1992)] were determined, with by-sex comparisons (T-test) indicati ng no significant differ ences among endpoints ( P > 0.3617 for all comparisons). Male and fe male head lengths averaged 35 3 cm (mean standard deviation) and 30 2 cm, respectively. Total lengths for males and females were 2.53 0.15 m and 2.23 0.18 m, respectively, and estimated mass was 69.5 13.1 kg for males and 46 13.8 kg for females. In addition to the breeding pairs, two extra treated females were housed separa tely to monitor bioaccumulation of OCPs and health status via monthly blood assessments of hematocrit, glucose, and total protein (Mader, 1996). Selection of specific OCP analytes and dose calculations were based on OCP concentrations in alligator yol ks collected from contaminated sites in Florida and avian maternal transfer rates (Fairbrother et al ., 1999). The dosing regime was designed to coincide with oocyte development and yol k formation (vitellogenesis), which for alligators begins in early fall and continues through late spri ng (Lance, 1986; Guillette, et al., 1997). Dosing began on 16 and 17 October, 2001. Animals were randomized with OCP-treated individuals receiving one intram uscular (IM) and one intraperitoneal (IP) injection consisting of a mixture of p,p Â’-DDE (36.5 mg/kg), t oxaphene (2.6 mg/kg), chlordane (2.5 mg/kg), and diel drin (8.4 mg/kg) solubilized in reagent grade olive oil (cumulative injection volume of 40 mL). C ontrol animals received the same volume of olive oil. Animals did not receive oral doses of OCPs until they resumed feeding the following spring. On 16 April, 2002 oral dos ing began and continued to 20 October, 2002 when animals went began winter fast. Th is pattern of animals receiving oral doses

PAGE 200

184 from April to October continued through 2003 and 2004, so animals were exposed for a period of three years. Treat ed animals received oral dos es of p,p’-DDE (0.18 mg/kg), toxaphene (0.13 mg/kg), chlordane (0.014 mg /kg), and dieldrin (0.018 mg/kg). The chemicals were mixed with reagent grade oliv e oil (total mixture volume per weekly dose = 8 mL). Control animals received the sa me feed ration minus the OCP mixture. When females in breeding groups began nesting (24 June–10 July 2002), the two extra OCP-treated females housed in separate enclosures for monthly health status monitoring, and two females which did not produce clutches were sacrificed (via decapitation/cervical dislocation with double p ithing) to determine bioaccumulation rates of OCPs. Tissue samples (adipose, liver, a nd blood) were collected for analytical chemistry, along with one or two egg yolks from each of the females that oviposited, with eggs being collected and incubated using methods described in Chapter 2, except no helicopter or airboat was necessary. Ti ssues and yolks were screened for 30 OCP analytes by GC-MS according to procedures de scribed in Chapter 3. Lipid content (%) was determined gravimetrically for liver, and adipose tissue, wh ile GC-MS techniques were used for blood (Chapter 3). For 2002 and 2003 clutches, a subset of yolks from five control clutches and four treated clutches were analyzed for thiamine content to determine if thiamine levels were related to OCP exposure and/or clutch viability. Thiamine analysis was conducted using methods described in Chapter 6. For treated versus control comparisons, T-tests (PROC TTEST; SAS Institute Inc., 2002) and Wilcoxon two-sample tests (PROC NPAR1WAY WILCOXON) were used for parametric and nonparametric clutch paramete rs, respectively. Nume rical data were logtransformed [ln(x)], while propor tional data were arcsine squa re root transformed to aid

PAGE 201

185 in meeting statistical assumptions for parame tric tests. Logistic regression (PROC LOGISTIC; SAS Institute Inc., 2002) was used to evaluate associations between clutch survival parameters and potential explanat ory factors (i.e., tota l OCP concentrations, thiamine concentrations, and clutch size parameters). Results Nine clutches were collected from the control group and seven from the treated group over a period of three years. Clutch pa rameters that signif icantly differed among the groups included clutch vi ability, incidence of unbanded eggs, lipid content, egg concentrations of seven of eight OCP analyt es, and total OCP concentrations in eggs (Table 7-1). Specifically, clutch viability of the control group wa s 30% higher than the treated group, and the inciden ce of unbanded eggs was 40% lo wer in the control group as compared to the treated group. In addition, eggs of the treated group had significantly higher lipid content and total OCP concentra tions over those of controls. Importantly, OCP burdens in yolks from the control group (5 0 3.6 ng/g) were less than those of the reference site (102 15.5 ng/g), and the treated group yielded yolk burdens (13,300 2,666 ng/g) that fell within the range of the mean OCP concentrations (1,169-15,480 ng/g) observed in contaminated sites (Chapt er 2). No significant differences were detected with respect to number of clut ches produced by each group, fecundity, clutch mass, egg mass, oxychlordane c oncentrations, or thiamine concentrations, with thiamine being analyzed on five control and four treated clutches during year 2002-2003. Monthly health status assessments on two “ex tra” females, which were housed apart from breeding females, indicated that blood chemis try values appeared to be within normal limits (Table 7-2). After 10 months of dos ing and concurrent with nesting of other captive females (June 2002), the two extra females and two non-reproductive females

PAGE 202

186 were sacrificed and tissues were analyzed for OCP content. Individual chemicals exhibited differing concentrations among ti ssues, with the differing levels possibly related to varying lipid content of the tissue and the level of the administered dose. Results of logistic regression indicated that OCP variables an d clutch-egg size variables appeared to be associated with clutch survival parameters. Based on the differences between treated and control groups it was not surprising to find that total OCP concentrations (TOC) in egg yolk was ne gatively associated w ith clutch viability and positively associated with incidence of unbanded eggs. Fecundity, egg mass, and clutch mass, which were not correlated with (TOC), but were positively correlated with clutch viability, early embryo mortality, and late embryo mortality, and negatively correlated with unbanded egg incidence. Lipi d content in eggs wa s positively correlated with early and late embryo mortality, and as shown in group comparisons, lipid content was positively correlated with TOC (i.e., si gnificantly higher in the treated group). Thiamine concentrations in eggs were determin ed to be significantly correlated with one another. In addition, thiamine monophosphate (TP) was found to be positively associated with clutch viability, clutch mass, and f ecundity, and negatively associated with unbanded egg incidence. Thiamine pyrophosphate (TPP) was not associated with any of the clutch survival parameters. In contra st, free thiamine and total thiamine were negatively associated with unbanded egg incide nce and positively associated with early and late embryo mortality (Table 7-3). Discussion The results of this study support th e hypothesis that OCPs are maternally transferred to the developing e gg, and that maternal exposure is associated with reduced clutch success and increased embryonic mortalit y. In addition, this is the first study to

PAGE 203

187 develop a method for exposing alligator embryos to endogenous concentrations of OCPs, in contrast to prior studies that have exogenously applied OCPs to eggs to elicit embryonic exposure (Matter, et al., 1998). Im portantly, the dosing regime did not induce adult mortality, or alterations in monthl y blood chemistry assessments, food intake, weight gain, and behavior (e.g., females fi ercely defended their nests). However, subclinical, cytotoxic effects on the liver, gona ds, or kidneys may have been undetected. In addition, OCPs may cause f unctional defects in neural tr ansmission, leading to subtle increases in stress due to subl ethal neuronal hyperactivity. The decreased clutch viability noted in the OCP treated group was due to increased incidence of unbanded eggs. Sin ce unbanded eggs may be the result of embryo mortality occurring prior to embryo attachme nt (Rotstein et al., 2002) or possibly the result of infertility (or both), and since bot h parents received sim ilar doses of OCPs and since effects of OCPs vary from species to species and analyte to analyte (Rattner & Heath, 2003), the potential mechanisms by which OCP exposure might possibly induce increased incidence of unbanded eggs are many and may include altered egg quality, direct embryo toxicity, or decreased reproductive function in males. Other factors besides OCP exposure and e gg concentrations that were associated with variation in clutch survival parameters (and that werenÂ’t concurrently associated with OCP exposure) included: fecundity, clut ch mass (fecundity and clutch mass were collinear), egg mass, lipid c ontent, thiamine monophosphate (TP), free thiamine (FT), and total thiamine (TT) concentrations (all thiamine forms were collinear with one another).

PAGE 204

188 In order to accurately interpret associa tions, consideration should be given to biological relevance and plausibility, as well as the experimental design and results. Given that the major contribution to decrea sed clutch viability for both control and treated groups was the increased incidence of unbanded eggs, fact ors associated with unbanded eggs are due careful consideration. In this respect, concentrations of OCPs in eggs were positively associated with unbande d egg incidence and negatively associated with clutch viability, as e xpected given the experimental design and results of group comparisons. In contrast to OCPs, clutch fecundity, clutch mass, egg mass, and TP had negative associations with unbanded egg incide nce and positive associations with clutch viability. However, none of these explan atory factors were correlated with OCPs, suggesting that they may independently ac count for a portion of the incidence of unbanded eggs in OCP exposed group and, mo re importantly, the control group. The biological implications are that in a control situ ation the incidence of unbanded eggs are related to smaller clutches and lighter e ggs, and that the rate of unbanded eggs is basically doubled when captive animals are exposed to OCPs. Factors that were not correlated with clutch viability were FT, TT, and lipid content. FT and TT were negatively co rrelated with unbanded egg incidence and positively correlated with early and late embryo mortality. These associations may be a result of the significant correlation FT and TT ha ve with TP, and might not be a result of a direct link with unbanded egg incidence. Positive associations between FT, TT, and early and late embryo mortality, suggest pos sible thiamine hypervitaminosis. Although unlikely, thiamine concentrations of experime ntal clutches (both gr oups) were three-fold to five-fold greater than thos e of wild clutches, and thiamine hypervitaminosis has been

PAGE 205

189 shown to induce neurotoxicosis in laborator y models (Snodgrass, 1992), which suggests that thiamine toxicity canÂ’t be completely discounted. Also correlated with increased early and late embryo mortality were fecundity, clutch mass, and egg mass. Although a biologically relevant reason for this association may be pres ent, the association between embryo mortality and fecundity, clutch and egg ma ss may be attributed to the fact that as unbanded egg incidence decreases, the more embryos are pres ent and allows the potential for embryo mortality to occur. In contrast to thiamine concentrations, lipid content was found to be significantly associated only with early and late embryo mortality, and OCP concentrations, which is to be expected as treated groups had signifi cantly higher levels comp ared to controls. One might conclude that the reasons for the positive association between OCPs and lipid content is simply because OCPs are hydr ophobic and lipophilic; however, because lipid content was different between treatment gr oups, it may be that OCP exposure altered liver and/or follicle function in producing and sequestering yolk components. These results may suggest maternally-mediated alterations in egg quality and resulting decreased clutch viability. Ma ternally-mediated reductions in clutch viability is a likely scenario as liver is a known target organ for OCP-induced toxicity (Metcalfe, 1998), and exposure to similar organochlorine compounds has been shown to alter liver function (phospholipid production and transfer), vitello genesis, egg component profiles of other oviparous vertebrates ((Lal & Singh, 1987), a nd up-regulates biotransformation enzymes that are involved in xenobiotic biotransforma tion and lipid metabolism (Ertl et al., 1998). With respect to lipid contents associ ation with embryo mortality, an important finding was that OCP concentrations were not associated with early or late embryo

PAGE 206

190 mortality, suggesting lipid associ ation is not simply due to multicollinearity with OCPs. These results may indicate that maternal exposur e to OCPs alters lipid content, leading to increased embryo mortality. This association may be biologically re levant and plausible since differences in egg yolk fatty acid prof iles are suggested to be related to reduced clutch viability in captive alligators (Noble et al., 1993; Millstein, 1995). In summary, results of the present stu dy supports the hypothesis that parental OCP exposure may decrease clutch viability by increasing the incidence of unbanded eggs. These results differ from observations in wild clutches from OCP-contaminated sites in which reduced clutch viability is primarily due to increases in early and late embryo mortality (Chapter 2). However, unbanded eggs may be products of very early embryo mortality (Rotstein et al., 2002), with very early embr yo mortality in the captive population being likely related to OCP effects that have been exacerbated by the stress of captivity. Also important to consider is that alligators (less than 50 years old) from OCPcontaminated sites have likely been exposed to OCPs since con ception, and therefore may have been reproductively altered duri ng development {Gross et al., 1994} and may respond differently to OCP exposure as compared to previously unexposed adults. This study confirms, as somewhat expected, th at OCPs are maternally transferred in the alligator and that this is likely the major ro ute for embryonic exposure. This study is also the first induce, via maternal OCP exposur e, endogenous OCP exposure in developing alligator embryos. Importantl y, this ecological relevant e xperiment demonstrates that parental exposure to OCPs results in decreases in clutch viability sim ilar to what has been observed in wild alligator populations inhabiti ng OCP-contaminated sites. Lastly, this study provides experimental evidence linking pa rental OCP exposure to decreased clutch

PAGE 207

191 viability in the American alligator, and suggests a maternally-mediated mechanism may be involved.

PAGE 208

192 Table 7-1. Summary statistics and comparisons of clutch parameters among treated and control groups for years 2002-2004. Parameter Control Treated Summary No. Clutches 9 7 16 Fecundity ( n ) 32 2.4 30 2.5 31 1.7 (19–40) (20–37) (19–40) Clutch mass (kg) 2.31 0.227 2.22 0.203 2.27 0.151 (1.33–3.3) (1.51–2.9) (1.33–3.3) Egg mass (g) 73 3 73 1.2 73 1.7 (53–83) (70–78) (53–83) Clutch viability (%) 44 11* 9 6 29 7.9 (0–95) (0–35) (0–95) Unbanded eggs (%) 39 12.4* 81 12.3 58 10.1 (3–100) (22–100) (3–100) Damaged eggs (%) 3 3 0 0 2 1.7 (0–27) (0–0) (0–27) Early emb. mort. (%) 8 4 4 2.8 6 2.5 (0–36) (0–16) (0–36) Late emb. mort. (%) 6 2.7 5 5 6 2.6 (0–20) (0–35) (0–35) Lipid content (%) 19 0.7* 22 0.7 20 0.7 (15–21) (20–25) (15–25) TP (pmoles/g) 24 6.1 16 8 20 4.8 (4–39) (3–39) (3–39) TPP (pmoles/g) 21 6 15 3.7 18 3.6 (11–44) (6–22) (6–44) Thiamine (pmoles/g) 3088 182.8 3035 343.4 3065 170.4 (2623–3694) (2576–4045) (2576–4045) Thiamine (pmoles/g) 3133 182.6 3066 352.8 3103 173.6 (2640–3731) (2585–4105) (2585–4105) CC (ng/g) 1 0* 24 9.4 14 6.1 (1–1) (3–64) (1–64) CN (ng/g) 1 0* 10 2.6 7 2 (1–2) (1–18) (1–18) Dield. (ng/g) 6 1.2* 773 122.1 335 116.4 (3–11) (475–1143) (3–1143) Oxychl. (ng/g) 1 0.1 2 0.4 2 0.2 (1–2) (1–3) (1–3) p,p'-DDE (ng/g) 19 2.6* 11729 2200.4 5038 1838.8 (6–30) (5801–18448) (6–18448)

PAGE 209

193 Table 7-1. Continued. Parameter Control Treated Summary Oxychl. (ng/g) 1 0.1 2 0.4 2 0.2 (1–2) (1–3) (1–3) p,p'-DDE (ng/g) 19 2.6* 11729 2200.4 5038 1838.8 (6–30) (5801–18448) (6–18448) TC (ng/g) 1 0* 25 9.9 17 7.5 (1–1) (3–66) (1–66) TN (ng/g) 2 0.4* 36 8 17 5.6 (1–4) (11–56) (1–56) Toxa. (ng/g) 0 0* 2035 720 2035 720 (0–0) (1315–2755) (1315–2755) [OCPs] (ng/g) 50 3.6* 13300 2666.1 5728 2116.4 (29–60) (6393–21991) (29–21991) No. OCPs ( n ) 5 0.7 7 1.2 6 0.7 (0–7) (0–9) (0–9)

PAGE 210

194 Table 7-2. Organochlorine concentrations and blood chemistry values of captive adult female alligators sacrificed during 2 002 (Rauschenberger et al., 2004). Mean standard deviation (sample size). Parameter Control Treated Adipose Tissue (ng/ga) p,pÂ’-DDE No datab 68,315 35,275 (4) Toxaphene No datab 8,385 1486 (4) Chlordane No datab 708 200 (4) Dieldrin No datab 4,372 1,237 (4) Lipid Content (%) No datab 82 6 (4) Liver Tissue (ng/ga) p,pÂ’-DDE No datab 8,168 3,750 (4) Toxaphene No datab Not detectedc Chlordane No datab 23 13 (4) Dieldrin No datab 143 92 (4) Lipid Content (%) No datab 4 2 (4) Whole Blood (ng/ga) p,pÂ’-DDE No datab 179 184 (4) Toxaphene No datab Not detectedc Chlordane No datab Not detectedd Dieldrin No datab 15 5 (4) Lipid Content (%) No datab 0.10 0.02 (4) Hematocrit ( %) e 20-30 20 4 (2) Glucose (mg/dl) e 74 63 17 (2) Tot.Plasma Protein mg/dl) e5.1 6 1 (2) ang chemical/g yolk wet weight (not lipid normalized). bNo control females were sacrificed. cLimit of detection for toxaphene = 230 ng/g. dLimit of detection for chlordane = 0.2 ng/g. eFor controls, blood chemistry valu es reported by Mader (1996). Values for treated group reflect mean of 10 samples collected evenly over 10 months.

PAGE 211

195 Table 7-3. Explanatory parame ters and clutch survival parameters with () indicating nature of association and value equal to concordance percentage. Parametera Unbanded egg% Clutch Viability Early Emb. Mort. Late Emb. Mort. TOCb (+) 64 (–) 65 ns ns Fecundityc (–) 71 (+) 65 (+) 54 (+) 65 clutch massc (–) 77 (+) 72 (+) 67 (+) 76 egg mass (–) 63 (+) 58 (+) 55 (+) 66 lipid %b ns ns (+) 58 (+) 68 TPcd (–) 59 (+) 59 ns ns TPPd ns ns ns ns FTd (–) 68 ns (+) 74 (+) 72 TTd (–) 69 ns (+) 74 (+) 72 aTOC = total OCP concentrations in egg yolk; TP = thiamine monophosphate; TPP = thiamine pyrophosphate; FT = free thiamine; TT = total ( ) thiamine concentrations in yolk. b-d Parameters sharing same superscript letters are significantly correlated with each other, those not sharing letters are not correlated. NS = not significant ( P > 0.05)

PAGE 212

196 CHAPTER 8 CONCLUSIONS Introduction The American alligator has a significant role in the ecology, esthetics, and economy of Florida. Thus, maintaining viable popul ations of alligators is desirable for many reasons. As reproduction is critical for ma intaining a species p opulation, any relatively sudden or sustained decrease in repro duction may be cause for concern. Over the last quarter century, alligat or populations in organochlorine (OCP) contaminated lakes in central Florida have garnered intense study and much attention due to their decreased reproductive performan ce (Woodward et al., 1989; Woodward et al., 1993; Wiebe et al., 2001). Decreased reproduct ive performance has been attributed to decreased clutch viability due to increased embryo mortality, with mortality typically occurring during the first 20 days of devel opment (Masson, 1995). Furthermore, prior study suggests OCP exposure may be a potential contributing factor to increased embryo mortality since increased mortality had been reported only in sites heavily contaminated with OCPs and since alligator eggs from thes e sites contained increased levels of OCPs, but a clear relationship was not ev ident (Heinz et al., 1991). Understanding biological and environmental characteristics related to embryo development, egg quality, and clutch viability in alligators is necessary in evaluating whether OCPs and/or some other factor(s) may be causally linked to decreased clutch viability. Identifying an d understanding factors associated with decreased clutch viability may benefit management of alligator populations and ensure sustainable human use. On

PAGE 213

197 a larger scale, understandi ng the relationship between OCP exposure and developmental mortality in alligators ma y provide some insight into the potential impact of organochlorine pesticides on the ecol ogical health of FloridaÂ’s wetlands. Evaluating the relationships between cl utch viability, OCP exposure, embryo development, and other potentially important biological and environmental factors has been the theme of this disse rtation. These evaluations have been accomplished through field studies designed to identify importan t factors and associations and to test hypothesized associations with laboratory experiments. Summary of StudyÂ’s Findings The first study (Chapter 2) examined clutch viability on OCP-contaminated and reference sites from 2000-2002. Results indicate d that clutch viability is significantly lower in contaminated sites, with these site s having higher rates of early and late embryo mortality, and that unbanded eggs also appear to be an important constituent of reduced clutch viability for reference and contaminat ed sites. In order of importance, major constituents of reduced clutch viability for all sites include early embryo mortality, unbanded eggs, late embryo mortality, and dama ged eggs. In addition, clutches from OCP-contaminated sites had an average of 10 more eggs per clutch as compared to the reference site, but average clutch mass wa s not significantly different, making average egg mass of reference site clutches greater than that of clutches of OCP-contaminated sites. In addition to differences in clutch viability and size, la rge differences in OCP concentrations in alligator eggs between reference and OCP-contaminated sites were found. Although not surprising given the histor y of the sites, these results support the continued problems with alligator repr oduction in OCP-contaminated sites.

PAGE 214

198 Results of redundancy analyses indicate that for Lake Lochloosa, no significant correlations were determined although significance might have been detected given a greater sample size. For Emeralda Marsh, th e weak associations between OCP variables and clutch survival variables suggests that other factors may be involved in reduced embryo survival and increased rates of unbanded eggs. The weak associations for Emeralda Marsh are surprising given that relatively stronger associations were determined for the other high exposure site (Lake Apopka; Table 2-5), as well as the intermediate exposure site (Lake Griffin, Table 2-5), with Emeralda Marsh being separated from Lake Griffin by easily traver sable, non-fenced levee. The positive association between early em bryo mortality and unbanded eg g rates and extracted OCP variables for Lake Apopka clutches suggest s that the percentage s of dieldrin and trans chlordane in eggs may play an important role in altered egg fertility and/or early embryo survival. For Lake Griffin, the negative to near-zero association between early embryo mortality rates and extracted OCP variables s uggests that OCP burdens in eggs may not play an important role in early embryo mort ality. However, the positive association between toxaphene burdens and late embryo mo rtality suggests that as toxaphene burdens increase, so does the risk for increased embryo death during the last 35 days of development. Furthermore, th e positive association between p,pÂ’ -DDT concentrations and unbanded egg rates suggests that these an alytes may be involved in altered egg fertility and/or embryo survival (prior to eggshell membrane attachment) (Fig. 2-2). In summary, the first study suggested that, over all sampled clutches, clutch survival parameters and egg and clutch size parameters vary between the low OCP exposure site (Lochloosa) and the intermedia te-high OCP exposure sites. Furthermore,

PAGE 215

199 OCP burdens do not appear to be related to cl utch survival for the low exposure site but are associated with clutch survival for the intermediately OCP contaminated site and one of the highly OCP contaminated sites. The next study (Chapter 3) evaluated the relationship between OCP burdens in eggs and in maternal tissues in order to exam ine the extent of maternal transfer of OCPs in the alligator and to determine if eggs coul d be used as predicto rs of maternal tissue burdens. Major finding of this study were that adipose tissue and yolk burdens were similar when adjusted for lipid content a nd that yolk was an ex cellent predictor of adipose tissue burdens. Convers ely, blood and yolk burdens were not linearly related. Importantly, liver had higher burdens than yolk after adjustment for lipid content suggesting liver may sequester OCPs, supporti ng the possibility that liver function may be altered due to chronic OCP exposure. Altered liver function due to OCP exposure may affect lipid metabolism, vitellogene sis, and egg quality (Lal & Singh, 1987), potentially resulting in maternally-med iated embryo mortality. However, non-OCP related maternal factor(s), such as size and ag e, could also affect clutch viability rates. The following study (Chapter 4) addressed the potential influence of maternal factors on clutch viability. Re sults indicated maternal body si ze was not associated with variation in clutch survival parameters, but moderate associations existed between maternal OCP burdens and clutch survival parameters (18% of variance explained, P < 0.05). Specifically, as p,pÂ’-DDE proportions increased in relation to total OCP egg burdens, the incidence of unbanded eggs incr eased, and as trans-chlordane proportions increased in relation to total OCP burdens in eggs, clutch viability decreased and early embryo mortality increased.

PAGE 216

200 Since increased rates of embryo mortality were the reason for decreased clutch viability, the fourth study (Cha pter 5) sought to evaluate the histopathology, growth, and development of embryos, and their associati ons with OCP exposure (egg burdens). Intrasite comparisons suggested that among all sites and all sampled ages (calendar age = CA), that morphological age (MA) of embr yos and embryo mass were greater for live embryos as compared to dead embryos, whic h suggested that embryos may have been developing normally up to a point at wh ich development stalled and the embryo eventually died, or embryos could have devel oped at a much slower overall rate until the point at which they perished. Either way it appears that the mass of dead embryos was appropriate for their MA. Morphometry of live embryos did not appear to be significantly related to variati on in clutch mortality rates, suggesting that live embryos from clutches with high mortality rates de velop similarly to those of low mortality clutches. This finding may suggest a threshol d-type response in which embryos exposed to stressors below a certain threshold have the ability to overcome stressors through various cellular homeostatic mechanisms, but above a certain threshold, developmental retardation and lethality occu r. Such threshold dose-response patterns have been accepted as a major dose-response pattern in mammalian developmental toxicology (Rogers & Kavlock, 2001). Furthermore, variation in morphologi cal development of live embryos was significantly associated with variation in th e composition and concentration of OCPs in eggs. The strength of the relationships appear ed to decrease with the age sampled (CA), with youngest embryos sampled (CA Day 14) showing the strongest relationships between OCP egg burden and morphometric parameters, followed by each subsequent

PAGE 217

201 CA, respectively (Table 5-5). Interestingly, the percenta ge of the total OCP burden (concentration) composed by an OCP analyte (i .e., HE%), appeared to be more important than OCP analyte concentrations alone. W ith respect to all sampled ages, except the eldest (CA Day 43), relative proportions of th e OCP analyte(s) appeared to be more important than concentrations alone (Table 5-5). For derived mor phometric parameters and morphological age (MA), similar patterns were observed in that embryos sampled at younger CA showed stronger rela tionships with OCP burdens than older cohorts (Table 5-6). Another important observation was that di fferent cyclodienes appeared to be associated with morphological variation of embryos of different ages (CA). Most important were the components of technical grade chlordane and its metabolites, which include cis and trans -chlordane, cis and trans -nonachlor, oxychlordane, and heptachlor epoxide. One or more of these components were found to be significantly associated with variation in embryo morphology for each CA sampled. These data suggest that the chlordane group may merit furthe r study in relation to developm ental effects in reptiles, especially considering other st udies have suggested that sexu al differentiation in turtles may be altered by low dose in ovo exposures of these compounds (Willingham, 2004). In conclusion, the embryo morphology a nd histopathology study found that embryo mortality occurring in alligator populations inhabiting reference and OCP-contaminated sites was characterized by developmental reta rdation without gross deformities, or overt presence of lesions to vital organs. Howe ver, variation in embryo morphology appeared to be associated with variation in OCP bur dens of eggs and the percentage composition composed by an OCP analyte was equally as important as concen tration, suggesting the

PAGE 218

202 importance of mixture composition. Younger embryos appeared more susceptible to OCP influence but OCP influence may not nece ssarily be the result of direct embryo effects. Similar types of embryo mortality, characterized by developmental retardation) has been documented in quail, with embryo mortality determined to be maternally mediated, where maternal liver function was al tered, resulting in nutrient deficiencies in eggs that were severe enough to induce em bryo mortality (Donaldson & Fites, 1970). Because nutrition and non-OCP contaminants have been associated with developmental retardation in salmonids (Fitzsimons et al ., 1999) and birds (Wilson, 1997; Gilbertson et al., 1991)., the next study (Chapter 6) ev aluated embryo mortality in alligators of reference and OCP-contaminated sites as a function of exposure to OCPs, polychlorinated biphenyls (PCBs), and pol yaromatic hydrocarbons (PAHs), as well as egg nutrient content. Results of this study suggested that decreasing thiamine levels in eggs may be associated with decreased clutch success and lipid content, and OCP burdens may be associated with variation in thiamine concentrations In addition, PCBs, PAHs, and non-thiamine nutrients were not found to be significan tly associated with clutch viability. Thiamine levels in eggs explained 38% of the variation in clutch survival parameters in the case-control c ohort study and 27% of the clutch survival variation in the expanded fiel d study, which suggest that factors in addition to thiamine are likely involved. A lack of effects on clutch viability observed in experiments involving thiamine amelioration and inhibition via topical egg treatments may suggest a number of potential conclusions including th at either thiamine has no effect on the embryo or that the embryo was not exposed to enough thiamine or th iamine inhibitor to elicit effects.

PAGE 219

203 The last experiment of this dissertati on (Chapter 7) attempted to examine the relationship between OCP exposur e and decreased clutch viabili ty in alligators in a more direct way which involved orally dosing a cap tive breeding population of adult alligators with OCPs to test the hypothesi s that adult alligators exposed to OCPs yield clutches with decreased clutch viability as compared to controls. The results of this study found that clutch viability was decreased in the OCP tr eated group, but that the decrease was due to increased incidence of unbanded eggs and not mortality in banded eggs (i.e., after embryo attachment). However, very early embryo (conceptus) mortality has been documented in unbanded eggs via determination of paternal DNA {Rotstein et al., 2002), so it is likely that a portion of the unbanded eggs were products of early embryo mortality, especially considering captives had not been raised in a contaminated habitat and may have responded more severely to OCP exposure in comparison to their wild cohorts. In addition to OCP factors, the captive study s uggested other factors, not concurrently associated with OCP exposure, were associat ed with clutch viability. These factors included fecundity, clutch mass (fecundity a nd clutch mass were correlated with each other), egg mass, lipid content, thiamine monophosphate (TP), free thiamine (FT), and total thiamine (TT) concentrations (all thiamine forms were correlated with one another). In summary, results of the captive parental dosing study support the hypothesis that OCP exposure may decrease clutch viab ility by increasing the incidence of unbanded eggs and/or early embryo mortality. The bi ological implications ar e that in a control situation incidence of unbanded eggs are related to smaller cl utches and lighter eggs, and that the rate of unbanded eggs is basically doubled when captive animals are exposed to

PAGE 220

204 OCPs. The study confirms that OCPs are matern ally transferred in th e alligator and that maternal OCP exposure alters lipid content of eggs and reduces clutch viability. Overall, our studies (Chapt ers 2-7) suggest that OCPs may indeed be contributing to the decreased clutch viab ility in alligator populations inhabiting OCP-contaminated sites. In addition, thiamine leve ls and clutch size parameters appear to be associated with embryo mortality and decreased clutch viab ility. These data combined with dead embryos being developmentally retarded s uggest that alterations in growth and metabolism are the probable mechanism by whic h mortality results, as opposed to acute toxicity to organs or specific deformities since these were not gr eatly observed. Lastly, the captive exposure study provided experiment al evidence that parental OCP exposure can reduce clutch viability. Continuing studi es beyond this dissertation are investigating the relationship between fatty acid prof iles, OCPs, and clutch viability. Future Considerations and Global Implications Although difficult and expensive, conducti ng an expanded captive dosing study is likely the only way to separate which OCPs are actually causing the decreased clutch viability from those that are just collinear. A large number of alligators would have to be involved so that hypothesized OCP analytes could be given individually to determine what component of the OCP mixture was resp onsible or if decreased clutch viability resulted from some type of mixture effect. A challenge in trying to re late the present findings to other OCP exposure studies involving birds, mammals, and fish is that the basic metabolic function of an adult alligator is vastly different from most models. For example, blood flow of a 70 kg alligator (0.26 L/min) is less than 8% of th at of a 70 kg human, and 0.3% of that of a 70 kg shrew (Coulson & Hernandez, 1983). Thes e differences mean that xenobiotics and

PAGE 221

205 endogenous compounds circulate throughout the al ligator at a decreased rate which can affect excretion and elimination, as well as the amount of time target organs are exposed (i.e., decreased blood flow through liver may mean increased OCP e xposure to liver). Another factor that may affect OCP toxicity is the temperature of the alligator, as low temperatures have been associated with incr eased DDT toxicity in exposed fish (Rattner & Heath, 2003). Although it is often believed that being cold-blooded causes low blood flow in alligators, low blood flow is actually related to their relati vely small hearts and low blood hemoglobin, and not temperature (Coulson & Hernandez, 1983). Speculatively, low blood flow, seasonally lo wer body temperatures, and seasonal fasting (possibly resulting in mobilization of lipids and hydrophobic contaminants) may contribute to this species su sceptibility to reproductive modulation via OCP exposure. Another aspect concerning the biochemistry of the alligator is that they are true predators in that they cannot di gest complex sugars or starches or plant proteins. Indeed, their sources of glucose are mainly glucone ogenesis, in which the liver uses carbon skeletons of catabolized amino acids to synt hesize glucose, and u tilization of glucose stores in the carcasses of pr ey. The implications are the importance of normal liver function in producing energy. Furthermore, liver functions in amino acid storage, in that it has been shown that alligator s can store excess amino acids in liver that can be later mobilized for protein production. The ecologi cal trophic level of th e alligator obviously contributes to increased orga nochlorine pesticide exposure vi a biomagnification, but their apparent susceptibility to maternally mediated development mortality may be exacerbated by the physiological requirements of being a predator. This speculation is supported by studies indicating p,pÂ’-DDE causes eggshell thinning in predatory birds but

PAGE 222

206 not domestic fowl, suggesting predatory sp ecies may be more susceptible to the reproductive effects of OCP exposure (Fairbroth er et al., 1999). Lastly, alligators are a poikilothermic species that fast for up to si x months during a time when vitellogenesis is underway. The concurrent fasting and vite llogenesis means that OCPs are mobilized along with lipid stores to meet the meta bolic demands of homeostasis and follicle development, likely increasing risk of OCP-a ssociated alterations in liver function. Identifying other species that may be susceptible to similar organochlorineassociated embryo mortality is important for both helping to maintain biodiversity and for better understanding of the mechanisms of organochlorine-associated developmental mortality. Key ecological and physiological ch aracteristics to look for in a potential model are that the species be a season ally fasting, oviparous, highly fecund, poikilothermic predator. Given these attribut es, species which may be potential models for examining OCP-induced reproductive toxicity include predatory turtles, such as the common snapper ( Chelydra serpentina ), the softshell ( Apalone muticus ), the alligator snapper ( Macrochelys temminckii) water snakes ( Nerodia spp .), and predatory fish, such as largemouth bass ( Micropterus salmoides ), and bowfin ( Amia calva ). Indeed, alterations in endocrine function and increase d developmental mortality have been noted in largemouth bass inhabiting Emeralda Mars h (Seplveda et al ., 2004). In addition, turtle eggs from Lake Apopka were found to have abnormalities and poor hatch rates around the time reproductive problems with alligat ors began to be investigated (Franklin Percival, pers. comm..). The global implications of this dissertati onÂ’s results and postula tions suggest that predatory reptiles and fish inhabiting areas of the world that receive(d) high inputs of

PAGE 223

207 organochlorine pesticides may be at risk of increased rates of embryo mortality or decreased reproductive performance. Many of these areas are in tropical, third-world countries that continue to buy DDT from U. S. manufacturers because it is an economical way to control malarial mosquitoes and cropdestroying pests (Brema n et al., 2004). The combination of concern for human health and ecological inte grity underscore the exigency for better understanding of the eff ects associated with OCPs and similar persistent organic compounds, so that best management practices may be developed in order to protect human health and ecological integrity.

PAGE 224

208 LIST OF REFERENCES Akerman, G., Tjarnlund, U., Noaksson, E., & Balk, L. (1998). Studies with oxythiamine to mimic reproduction disorders among fish early life stages. Marine Environ Res 46: 493-497. Ashworth, C. J. & Antipatis, C. (2001) Micronutrient programming of development throughout gestation. Reproduction 122: 527-35. Benton, J., Douglas, D., & Prevatt, L. (1991) Completion report as required by federal aid in fish restoration: Wallop-Breaux project F-30-18, Ocklawaha Basin Fisheries Investigations Study XII. State of Flor ida Game and Freshwater Fish Commission. Tallahassee, FL. Billet, F., Gans, C., & Maderson, P. F. A. (1985). Why study reptilian development? In Biology of the Repti lia. 14. Development A : 3-25. Gans, C., Billet, F. & Maderson, P. F. A. (Eds.). New York, NY: John Wiley and Sons. Blus, L. J. (1996). DDT, DDD, and DDE in Birds. In Environmental contaminants in wildlife : Beyer, W. (Ed.). Boca Raton: Lewis Publishers. Breman, J. G., Alilio, M. S., & Mills, A. ( 2004). Conquering the intolerable burden of malaria: whatÂ’s new, whatÂ’s needed: a summary. Am J Trop Med Hyg 71: 1-15. Brown, S., Honeyfield, D. C., & Vandenbyllaardt, D. L. (1998). Thiamine analysis in fish tissues. American Fisheries Society Sym posium 21. 21, 73-81. American Fisheries Society. Campbell, K. R. (2003). Ecotoxicology of crocodilians. Applied Herpetology 1: 45-163. Clarke, S. F. (1891). The habits and embryology of the American alligator. J Morph 5: 181-205. Colborn, T., Von Saal, F. S., & Soto, A. M. (1993). Developmental effects of endocrinedisrupting chemicals in wildlife and humans. Environ Health Perspect 101: 378384. Coulson, R. A. and Hernandez, T. (1983). Alligator metabolism: studies on chemical reactions in vivo. Oxford, UK, Pergamon.

PAGE 225

209 Deeming, D. C. & Ferguson, M. W. J. ( 1990). Morphometric analysis of embryonic development in Alligator mississippiensis Crocodylus johnstoni and Crocodylus porosus J Zool 221: 419-439. de Roode, D. F., Gustavsson, M. B., Rantalaine n, A. L., Klomp, A. V., Koeman, J. H., & Bosveldt, A. T. (2002a). Embryotoxic potenti al of persistent organic pollutants extracted from tissues of guillemots (Uria aalge) from the Baltic Sea and the Atlantic Ocean. Environ Toxicol Chem 21: 2401-11. de Roode, D. F., Vuorinen, P. J., & Bosveld, A. T. (2002b). Effects of furazolidone, PCB77, PCB126, Aroclor 1248, paraquat and p, p'-DDE on transketolase activity in embryonal chicken brain. Toxicology 173: 203-10. Donaldson, W. E. & Fites, B. L. (1970) Embryo mortality in quail induced by cyclopropene fatty acids: Re duction by maternal diets high in unsaturated fatty acids. J Nutr 100: 605-10. Ecobichon, D. J. (2001). Toxic effects of pesticides. In Caasarett and Doull's toxicology: the basic science of poisons : 763-810. Klaasen, C. D. (Ed.). New York, NY: McGraw-Hill. Elsey, R. M. & Wink, C. S. (1985). Femoral bone as a possible source of calcium for eggshell deposition in Alligator mississippiensis Anat Rec 211: A57. EPA. (2004). http://www.epa.gov/region4/wa ste/npl/nplfls/towercfl.htm, accessed October 8. US Environmental Pr otection Agency, Washington DC. Ertl, R. P., Bandiera, S. M., Buhler, D. R ., Stegeman, J. J., & Winston, G. W. (1999). Immunochemical analysis of liver microsomal cytochromes P450 of the American alligator, Alligator mississippiensis Toxicol Appl Pharmacol 157: 157-165. Ertl, R. P., Stegeman, J. J., & Winston, G. W. (1998). Induction time course of cytochromes P450 by phenobarbital and 3-met hylcholanthrene pretreatment in liver microsomes of Alligator mississippiensis Biochem Pharmacol 55: 1513-1521. Fairbrother, A., Ankley, G. T ., Birnbaum, L. S., Bradbury, S. P., Francis, B., Gray, E., Hinton, D., Johnson, L. L., Peterson, R. E., & Van Der Kraak, G. (1999). Reproductive and developmental toxicology of contaminants in oviparous animals. In Reproductive and developmental effect s of contaminants in oviparous vertebrates : 283-361. Di Giulio, R. T. & Tillit, D. E. (Ed.). Pensacola, FL : SETAC press. Ferguson, M. W. J. (1981). The valu e of the American alligator ( Alligator mississippiensis ) as a model for research in craniofacial development. J Craniofac Genet Dev Biol 1: 123-144.

PAGE 226

210 Ferguson, M. W. J. (1985). The reproductiv e biology and embryology of crocodilians. In Biology of the reptilia: Vol. 14, Development A : 329-491. Gans, C. (Ed.). New York, NY: John Wiley and Sons. Fitzsimons, J. D., Brown, S. B., Honeyfield, D. C., & Hnath, J. G. (1999). A Review of Early Mortality Syndrome (Ems) in Great Lakes Salmonids: Relationship With Thiamine Deficiency. Ambio 28: 9-15. Fitzsimons, J. D., Vandenbyllaardt, L., & Brow n, S. B. (2001). The use of thiamine and thiamine antagonists to inves tigate the etiology of early mortality syndrome in lake trout (Salvelinus namaycush). Aquat Toxicol 52: 229-39. Fry, D. M. (1995). Reproductive effects in bi rds exposed to pestic ides and industrial chemicals. Environ Health Perspect 103 Suppl 7: 165-71. Gadbury, G. L. & Schreuder, H. T. (2003). Ca use-effect relationshi ps in analytical surveys: an illustration of statistical issues. Environ Monit Assess 83: 205-227. Gilbertson, M., Kubiak, T., Ludwig, J., & F ox, G. (1991). Great La kes embryo mortality, edema, and deformities syndrome (GLEMEDS) in colonial fish-eating birds: similarity to chick-edema disease. J Toxicol Environ Health 33: 455-520. Giroux, D. (1998). Lake Apopka revisited: a co rrelational analysis of nesting anomalies and DDT contaminants. Master of Science Thesis. Gainesville, Florida. University of Florida. Gray, L. E. Jr., Kelce, W. R., Wiese, T., Tyl, R., Gaido, K., Cook, J., Klinefelder, G., Desaulniers, D., Wilson, E., Zacharewski, T ., Waller, C., Foster, P., Lasky, J., Reel, J. G. J., Laws, S., McLachlan, J., Bres lin, W., Cooper, R., DiGiulio, R., Johnson, R., Purdy, R., Mihaich, E., Safe, S., S onnenschein, C., Weshons, W., Miller, R., McMaster, S., & Colborn. T. (1997). Endocrine Screening Methods. Workshop report: detection of estrogenic and an drogenic hormonal and antihormonal activity for chemicals that act via receptor or steroidogenic enzyme mechanisms. Reprod Toxicol 11: 719-750. Gross, D. A. (1997). Thymus, spleen and bone marrow hypoplasia and decreased antibody responses in hatchli ng lake Apopka alligators. Ma ster of Science Thesis. Gainesville, Florida. University of Florida. Gross, T. S., Arnold, B. S., Sepulveda, M. S., & McDonald, K. (2003). Endocrine disrupting chemicals and e ndocrine active agents. In Handbook of Ecotoxicology : 1033–1098. Hoffman, D. J., Rattner, B. A., Burt on, G. A. Jr. & Cairns, J. Jr. (Ed.). Boca Raton, FL USA: Lewis Publishers. Gross, T. S., Guillette, L. J., Percival, H. F., Masson, G. R., Matter, J. M., & Woodward, A. R. (1994). Contaminant-induced reproducti ve anomalies in Florida alligators. J Comp Pathol 26: 1-8.

PAGE 227

211 Guillette, L. J., Brock, J. W., Rooney, A. A., & Woodward, A. R. (1999). Serum Concentrations of Various Environmental Contaminants and Their Relationship to Sex Steroid Concentrations and Phallus Size in Juvenile American Alligators. Arch Environ Contam Toxicol 36: 447-455. Guillette, L. J. Jr, Gross, T. S., Masson, G. R., Matter, J. M., Percival, H. F., & Woodward, A. R. (1994). Developmental a bnormalities of the gonad and abnormal sex hormone concentrations in juvenile al ligators from contaminated and control lakes in Florida. Environ Health Perspect 102: 680-8. Guillette, L. J. Jr, Woodward, A. R., Crain, D. A., Masson, G. R., Palmer, B. D., Cox, M. C., You-Xiang, Q., & Orlando, E. F. (1997) The reproductive cycle of the female American alligator (Alligator mississippiensis). Gen Comp Endocrinol 108: 87101. Hall, R. J., Kaiser, T. E., Robertson, W. B. Jr, & Patty, P. C. (1979). Organochlorine residues in eggs in the endangered Amer ican crocodile (Crocodylus acutus). Bull Environ Contam Toxicol 23: 87-90. Heinz, G. H., Percival, H. F., & Jennings, M. L. (1991). Contaminants in American Alligator Eggs From Lake Apopka, Lake Griffin, and Lake Okeechobee, Florida. Environ Monit Assess 16: 277-285. Henry, K. S., Kannan, K., Nagy, B. W., Kever n, N. R., Zabik, M. J., & Giesy, J. P. (1998). Concentrations and hazard assessmen t of organochlorine contaminants and mercury in smallmouth bass from a remote lake in the upper peninsula of Michigan. Arch Environ Contam Toxicol 34: 81-86. Hill, A. B. (1965). The environment a nd disease: association or causation? Proc R Soc Med 58: 295-300. Hoffman, D. J. (1990). Embryotoxicity and tera togenicity of environmental contaminants to bird eggs. Rev Environ Contam Toxicol 115: 40-89. Hoffman, D. J. & Gay, M. L. (1981). Embryot oxic effects of benzo[ a]pyrene, chrysene, and 7,12-dimethylbenz[a]anthracene in petr oleum hydrocarbon mixtures in mallard ducks. J Toxicol Environ Health 7: 775-787. Holstege, D. M., Scharberg, D. L., Tor, E. R ., Hart, L. C., & Galey, F. D. (1994). A rapid multiresidue screen for organophosphorus, organochlorine, and N-methyl carbamate insecticides in plant and animal tissues. J AOAC Int 77: 1263-74. Horwitz, W. (2000). Official methods of analys is of AOAC Interna tional. Gaithersburg, MD, AOAC International. James, R. C. Warren, D. A. Halmes, N. C., & Roberts, S. M. (2000). Risk assessment. In Principles of Toxicology : 437-478. Williams, P. L. James, R. C. & Roberts, S. M. (Ed.). New York: John Wiley & Sons.

PAGE 228

212 Lal, B. & Singh, T. P. (1987). Impact of pesticides on lipid metabolism in the freshwater catfish, Clarias batrachus during the vitellogenic phase of its annual reproductive cycle. Ecotoxicol Environ Saf 13: 13-23. Lance, V. (1986). Hormonal control of reproduction in crocodilians. In Wildlife Management: Crocodiles and Alligators : 409-415. Webb, G. J. W., Manolis, S. C. & Whitehead, P. J. (Ed.). Chipping No rton, NSW: Surrey Beatty and Sons. Lance, V., Joanene, T., & McNease, L. (1983) Selenium, vitamin E, and trace elements in the plasma of wild and farm reared alligators during the reproductive cycle. Can J Zool 61: 1744-1751. Lang, J. W. & Andrews, H. V. (1994). Temp erature-dependent sex determination in crocodilians. J Exp Zool 270: 28-44. Longnecker, M. P., Klebanoff, M. A., Brock, J. W., Zhou, H., Gray, K. A., Needham, L. L., & Wilcox, A. J. (2002). Maternal serum level of 1,1-dichloro-2,2-bis(pchlorophenyl)ethylene and ri sk of cryptorchidism, hypos padias, and polythelia among male offspring. Am J Epidemiol 155: 313-22. Mader, D. R. (1996). Reptile Medicine and Surgery Philadelphia, PA: W. B. Saunders. Marburger, J. E., Johnson, W. E., Douglas, D. R., & Gross, T. S. (1999). Pesticide contamination of fish and sediments in the Emeralda Marsh Conservation Area: relevance to fisheries establishmen t in flooded muck farms. Technical Memorandum No. 31 to the St. JohnÂ’s Wate r Management District. Palatka, FL. Masson, G. R. (1995). Environmental influe nces on reproductive potential, clutch viability and embryonic mortality of the American alligator in Florida. PhD dissertation. Gainesville, FL, University of Florida. Matter, J. M., McMurry, C. S., Anthony, A. B., & Dickerson, R. L. (1998). Development and implementation of endocrine biomarkers of exposure and effects in American alligators (Alligator mississippiensis). Chemosphere 37: 1905-1914. McEvoy, T. G., Robinson, C. J., Ashworth, J. A., Rooke, J. A., & Sinclair, K. D. (2001). Feed and forage toxicants affecting em bryo survival and fetal development. Theriogenology 55: 113-129. Metcalfe, C. D. (1998). Toxicopathic responses to organic compounds. In Fish diseases and disorders : 133-161. Leatherland, J. F. & Woo, P. T. K. (Ed.). Cambridge, MA: CABI Publishing. Millstein, S. R. (1995). Nutrient therapy in the American allig ator and its association with egg yolk long-chain fatty acids and egg fertility. Master of Science Thesis Gainesville, FL. University of Florida.

PAGE 229

213 Noble, R. C., Mccartney, R., & Ferguson, M. W. J. (1993). Lipid and fatty-acid compositional differences between eggs of wild and captive-breeding alligators ( Alligatormississippiensis )–an association with reduced hatchability. J Zool 230: 639-649. Rao, C. R. (1964). The use and interpretation of principal component analysis in applied research. Sankhya A 26: 329-358. Rattner, B. A. & Heath, A. G. (2003). Envi ronmental factors affecting contaminant toxicity in aquatic and te rrestrial vertebrates. In Handbook of Ecotoxicology : 679699. Hoffman, D. J., Rattner, B. A., Burton, G. A. Jr. & Cairns, J. Jr. (Ed.). Boca Raton, FL USA: Lewis Publishers. Rauschenberger, R. H., Wiebe, J. J., Buckland, J. E., Smith, J. T., Seplveda, M. S., & Gross, T. S. (2004). Achieving environmen tally relevant organochlorine pesticide concentrations in eggs through maternal exposure in Alligator mississippiensis Marine Environ Res 58: 851-856. Reese, A. M. (1912). The embryol ogy of the Florida alligator ( A. mississippiensis ) Proc 7th Int Zool Congr 1907: 535-537. Rice, A. K. (2004). Diet a nd condition of alligators ( Alligator mississippiensis ) in three central Florida lakes. Master of Science Thesis. Gainesville, FL. University of Florida. Rice, K. G. (1996). Dynamics of exploitati on on the American alligator: environmental contaminants and harvest. PhD Dissertation. Ga inesville, FL. University of Florida. Robson, W. A., Arscott, G. H., & Tinsley, I. J. (1976). Effect of DDE, DDT and calcium on the performance of adult Japanese quail ( Coturnix coturnix japonica ). Poult Sci 55: 2222-7. Roecklein, B., Levin, S. W., ComLy, M., & Mukh erjee, A. B. (1985). Intrauterine growth retardation induced by thiamine deficiency and pyrithiamine during pregnancy in the rat. Am J Obstet Gynecol 151: 455-60. Rogers, J. M. & Kavlock, R. J. (2001). Developmental toxicology. In Casarett and Doull's Toxicology: The Basic Science of Poisons : 351-386. Klaasen, C. D. (Ed.). New York: McGraw-Hill. Rotstein, D. S. (2000). An evaluation of deve lopmental mortality and fertilization failure in alligators using microsatelli te analyses. Master of Science Thesis. Gainesville, FL. University of Florida. Rotstein, D. S., Schoeb, T. R., Davis, L. M., Glenn, T. C., Arnold, B. S., & Gross, T. S. (2002). Detection by microsatellite analys is of early embryonic mortality in an alligator population in Florida. J Wildl Dis 38: 160-165.

PAGE 230

214 Russell R. W, Gobas, F. P. C., & Haffner, G. D. (1999). Maternal transfer and in ovo exposure of organochlorines in ovipa rous organisms: A model and field verification. Environ Sci Technol 33: 416-420. Ryberg, W. A., Fitzgerald, L. A., Honeycutt, R. L., & Cathey, J. C. (2002). Genetic relationships of American alligator p opulations distributed across different ecological and geographic scales. J Exp Zool 294: 325-333. SAS Institute Inc. (2002). SAS version 9.0. Cary, NC. Schenck, F. J., Wagner, R., Hennessy, M. K., & Okrasinski, J. L. Jr. (1994). Screening procedure for organochlorine and orga nophosphorus pesticide residues in eggs using a solid-phase extraction cleanup and gas chromatographic detection. J AOAC Int 77: 1036-40. Schmidt, R. R. & Johnson, E. M. (1997). Principles of teratology. In Handbook of Developmental Toxicology : 3-12. Hood, R. D. (Ed.). Boca Raton, Florida: CRC Press. Schoeb, T. R., Heaton-Jones, T. G., Clemm ons, R. M., Carbonnea u, D. A., Woodward, A. R., Shelton, D., & Poppenga, R. H. (2002). Clinical and necropsy findings associated with increased mortality am ong American alligators of Lake Griffin, Florida. J Wildl Dis 38: 320-337. Seplveda, M. S., Wiebe, J. J., Harvey, A ., Basto, J., Ruessler, D. S., Roldan, E., & Gross, T. S. (2001). An evaluation of environmental contaminants and developmental toxicity for the Ameri can alligator in central Florida. 21st Society of Toxicology Annual Meeting Abstract B ook. March 15-19. San Francisco, CA. pp. 151. Seplveda, M. S., Wiebe, J. J., Honeyfield, D. C., Rauschenberger, R. H., Hinterkopf, J. P., Johnson, W. E., & Gross, T. S. (2004). Organochlorine pesticides and thiamine in eggs of largemouth bass and American alligators and their relationship with early life-stage mortality. J Wildl Dis In press Skaare, J. U., Ingebrigsten, K., Aulie, A., & Kanui, T. I. (1991). Organochlorines in crocodile ( Crocodylus niloticus) Bull Environ Contam Toxicol 47: 126-130. Smith, A. G. (1991). Chlorinate d hydrocarbon insecticides. In Handbook of pesticide toxicology: Vol. 2: 731–915. Hayes, W. J., Jr. & Laws, E. R., Jr. (Ed). San Diego, CA: Academic Press. Snodgrass, S. R. (1992). Vitamin neurotoxicity. Mol Neurobiol 6: 41-73. Spallholz, J. E. & Hoffman, D. J. (2002). Sele nium toxicity: cause a nd effects in aquatic birds. Aquat Toxicol 57: 27-37.

PAGE 231

215 Stenner, R. D., Merdink, J. L., Stevens, D. K., Springer, D. L., & Bull, R. J. (1997). Enterohepatic recirculation of trichloroeth anol glucuronide as a significant source of trichloroacetic acid. Metabolit es of trichloroethylene. Drug Metab Dispos 25 : 529-535. Summer, C. L., Giesy, J. P., Bursian, S. J., Re nder, J. A., Kubiak, T. J., Jones, P. D., Verbrugge, D. A., & Aulerich, R. J. (1996). Effects induced by feeding organochlorine-contaminated carp from Saginaw Bay, Lake Huron, to laying White Leghorn hens. II. Embryotoxic and teratogenic effects. J Toxicol Environ Health 49: 409-38. Systat, Inc. (1999). SigmaScan Pro. Point Richmond, CA. ter Braak, C. J. F. (1986). Canonical co rrespondence analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology 67: 1167-1179. ter Braak, C. J. F. (1994). Canonical commun ity ordination. Part I: Basic theory and linear methods. Ecoscience 1: 127-140. ter Braak, C. J. F. (1995). Ordination. In Data analysis in community and landscape ecology : 91-173. Jongman, R. H. G., ter Braak, C. J. F. & Van Tongeren, O. F. R. (Ed.). Cambridge, UK: Cambridge University Press. ter Braak, C. J. F. & Smilauer, P. (2002) CANOCO reference manual and Canodraw for Windows Users' guide: software for canonical community ordination (version 4.52). Ithaca, NY, USA, Microcomputer Power. Voeltzkow, A. (1892). On the oviposition and embryonic development of the crocodile. Ann Mag Nat Hist (Ser 6) 9: 66-72. Webb, G. J. W., Buckworkth, R., Sack, G. C., & Manolis, S. C. (1983). An interim method for estimating the age of Crocodylus porosus embryos. Aust Wildl Res 10: 563-570. Weed, D. L. (2002). Environmental epidemio logy: basics and proof of cause-effect. Toxicology 181: 399-403. Wiebe, J. J., Rotstein, D., Percival, H. F., Woodward, A., Schoeb, T., & Gross, T. S. (2001). Evidence of developmen tal toxicity in th e American alligator from central Florida lakes. 20th Annual meeting of Society for Environmental Toxicology and Chemistry Abstract Book. November 15-19. San Francisco, CA. pp. 231. Wiebe, J. J., Sepulveda, M., Woodward, A., Pe rcival, H. F., Abecrombie, A., Wilkinson, P., Harvey, A., Basto, J., Ruessler, D. S ., Roldan, E., & Gross, T. S. (2001). An evaluation of environmental contaminants and decreased egg viability in the American Alligator ( Alligator mississippiensis ). 2001 Society of Toxicology Annual Meeting, Nov. 13-17. Baltimore, MD. pp. 160.

PAGE 232

216 Willingham, E. (2004). Endocrine-disrupti ng compounds and mixtures: unexpected doseresponse. Arch Environ Contam Toxicol 46: 265-9. Wilson, H. R. (1997). Effects of ma ternal nutrition on hatchability. Poult Sci 76: 134143. Wink, C. S. & Elsey, R. M. (1986). Changes in femoral morphology during egg-laying in Alligator mississippiensis J Morphol 189: 183-188. Woodward, A. R., Jennings, M. L., & Percival H. F. (1989). Egg collecting and hatch rates of American alligat or eggs in Florida. Wildl Soc Bull 17: 124-130. Woodward, A. R., Moore, C. T., & Delany, M. F. (1992). Experimental alligator harvest. Florida Fish and Wildlife Conservati on Commission. Gainesville, FL. pp. 170. Woodward, A. R., Percival, H. F., Jennings, M. L., & Moore, C. T. (1993). Low clutch viability of American alligators on Lake Apopka. Florida-Scientist 56 (1) 52-63. Woosley, R. L. (1996). Cardiac Actions of Antihistamines. Annu Rev PharmacolToxicol 36: 233-252. Wu, T. H., Rainwater, T. R., Platt, S. G ., McMurry, S. T., & Anderson, T. A. (2000a). Organochlorine contaminants in Morelet's crocodile (Crocodylus moreletii) eggs from Belize. Chemosphere 40: 671-8. Wu, T. H., Rainwater, T. R., Platt, S. G ., McMurry, S. T., & Anderson, T. A. (2000b). DDE in Eggs of Two Crocodile Species From Belize. J A g ric Food Chem 48: 6416-6420. Yagi, M., Kamohara, K., & Itowaka, Y. (1979). Thiamine deficiency induced by polychlorinated biphenyls (PCB) and di chlorodiphenyltrichloroethane (DDT) administration to rats. J Environ Pathol Toxicol 2: 1119-1125.

PAGE 233

217 BIOGRAPHICAL SKETCH Richard Heath Rauschenberger was born in North Little Rock, Arkansas in 1970, and is the son of Richard Edward and Mary Elizabeth Rauschenberger. Heath graduated from Greenbrier High School in Greenbrier, Arkansas in 1988; and received a BS in wildlife management in 1993, from Arkansas St ate University in Jonesboro, Arkansas. After gaining professional work experience as a pest contro l technician and later as a private lands wildlife biologist Heath returned to Arkansas State in 1999. He entered graduate school and received his MS in biology in 2001. After ear ning his MS degree, Heath immediately entered the University of Florida College of Ve terinary MedicineÂ’s doctoral program (under the mentorship of Dr. Timothy S. Gross) and majored in physiological sciences, with a concentration in interdisciplinary toxicology. Heath has held diverse positions such as lifeguard, vete rinary technician, grocery store clerk, pest control technician, wildlife bi ologist, and currently, research graduate assistant, which have aided in rounding out his professional experience. Heat h is married, has two sons, and enjoys spending time with them and the re st of his family. He also enjoys outdoor activities and is an active Christian a nd member of a local church.