<%BANNER%>

Effects of chronic methylmercury exposure on reproductive success, behavior and steroid hormones of the White Ibis (Eudo...

Permanent Link: http://ufdc.ufl.edu/UFE0041539/00001

Material Information

Title: Effects of chronic methylmercury exposure on reproductive success, behavior and steroid hormones of the White Ibis (Eudocimus albus).
Physical Description: 1 online resource (195 p.)
Language: english
Creator: Jayasena, Nilmini
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: corticosterone, courtship, endocrine, estradiol, eudocimus, homosexual, male, methylmercury, testosterone, white
Wildlife Ecology and Conservation -- Dissertations, Academic -- UF
Genre: Wildlife Ecology and Conservation thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Effects of chronic methylmercury exposure on reproductive success, behavior, and steroid hormones of the white ibis (Eudocimus albus) Methylmercury is a widespread environmental contaminant, with bioaccumulative properties. Effects of chronic ecologically relevant levels of mercury on wildlife are poorly known though some of the most sensitive end points of mercury toxicity are manifested as reproductive and behavioural impairment due to chronic exposure. I looked for experimental evidence of physiological, reproductive and behavioral impairment at environmentally relevant doses in the White Ibis (Eudocimus albus), a species known to be chronically exposed in the Florida Everglades. Wild-caught birds, kept in a free-flight aviary, were exposed to methylmercury from 90 days to 3 years of age. Treatment groups received 0.0001 (control), 0.01, 0.03 and 0.3 ppm (wet weight) methylmercury in their diet. Reproductive success of ibises was monitored in 2006, 2007 and 2008. Reproductive behavior and fecal concentrations of steroid hormone metabolites were monitored during 2007 and 2008 breeding seasons. I found dose-related changes in male pairing behavior, with up to 55% of dosed males nesting as male-male pairs, resulting in up to a 30% decrease in reproduction in comparison to control birds. Dosed males showed significantly reduced rates of display during courtship with the greatest reduction being in homosexual males. Females approached high dose group and homosexual males to a significantly lesser degree. Males nesting with males initiated nests earlier in the breeding season compared to heterosexual pairs. The probability of switching partners between breeding attempts was significantly lower in male-male pairs, with some pairs continuing to nest together for all three years. Overall fledging success in low and high dose group females were 33-35% lower than in control females. 20-30% of dosed females nested late or did not nest at all as males were unavailable due to male-male pairing. I report effects of methylmercury on fecal concentrations of estradiol, testosterone and corticosterone metabolites in both male and female ibises. The most prominent changes in estradiol were increased concentrations in dosed males during the courtship stage, seen to a greater extent in homosexual dosed males. Changes in estradiol concentrations of females were inconsistent and non-linear, but dosed birds generally had lower concentrations compared to controls. Testosterone concentrations in both sexes had significant interactions with stage of reproduction, and showed non-linear treatment effects. High dose group females had significantly elevated testosterone during laying, incubation and chick-rearing associated with reduced reproductive success in 2008. Corticosterone concentrations showed a general pattern of depression in dosed birds. Dosed birds had a higher corticosterone response than controls to an experimentally induced stressor, in the form of reduced food availability. Chronic, sub-lethal methylmercury exposure resulted in altered sexual preference of males, changes in courtship behavior, and altered concentrations of steroid hormones. I estimated a 15-50% reduction in reproductive success in dosed groups when compared to the controls as a result of male-male pairing and reduced female fledging success. Chronic, low-level mercury exposure could affect fitness and population parameters of wild populations, and has conservation implications as methylmercury pollution continues to be a problem in many regions.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Nilmini Jayasena.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Frederick, Peter C.
Local: Co-adviser: Oli, Madan K.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0041539:00001

Permanent Link: http://ufdc.ufl.edu/UFE0041539/00001

Material Information

Title: Effects of chronic methylmercury exposure on reproductive success, behavior and steroid hormones of the White Ibis (Eudocimus albus).
Physical Description: 1 online resource (195 p.)
Language: english
Creator: Jayasena, Nilmini
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: corticosterone, courtship, endocrine, estradiol, eudocimus, homosexual, male, methylmercury, testosterone, white
Wildlife Ecology and Conservation -- Dissertations, Academic -- UF
Genre: Wildlife Ecology and Conservation thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Effects of chronic methylmercury exposure on reproductive success, behavior, and steroid hormones of the white ibis (Eudocimus albus) Methylmercury is a widespread environmental contaminant, with bioaccumulative properties. Effects of chronic ecologically relevant levels of mercury on wildlife are poorly known though some of the most sensitive end points of mercury toxicity are manifested as reproductive and behavioural impairment due to chronic exposure. I looked for experimental evidence of physiological, reproductive and behavioral impairment at environmentally relevant doses in the White Ibis (Eudocimus albus), a species known to be chronically exposed in the Florida Everglades. Wild-caught birds, kept in a free-flight aviary, were exposed to methylmercury from 90 days to 3 years of age. Treatment groups received 0.0001 (control), 0.01, 0.03 and 0.3 ppm (wet weight) methylmercury in their diet. Reproductive success of ibises was monitored in 2006, 2007 and 2008. Reproductive behavior and fecal concentrations of steroid hormone metabolites were monitored during 2007 and 2008 breeding seasons. I found dose-related changes in male pairing behavior, with up to 55% of dosed males nesting as male-male pairs, resulting in up to a 30% decrease in reproduction in comparison to control birds. Dosed males showed significantly reduced rates of display during courtship with the greatest reduction being in homosexual males. Females approached high dose group and homosexual males to a significantly lesser degree. Males nesting with males initiated nests earlier in the breeding season compared to heterosexual pairs. The probability of switching partners between breeding attempts was significantly lower in male-male pairs, with some pairs continuing to nest together for all three years. Overall fledging success in low and high dose group females were 33-35% lower than in control females. 20-30% of dosed females nested late or did not nest at all as males were unavailable due to male-male pairing. I report effects of methylmercury on fecal concentrations of estradiol, testosterone and corticosterone metabolites in both male and female ibises. The most prominent changes in estradiol were increased concentrations in dosed males during the courtship stage, seen to a greater extent in homosexual dosed males. Changes in estradiol concentrations of females were inconsistent and non-linear, but dosed birds generally had lower concentrations compared to controls. Testosterone concentrations in both sexes had significant interactions with stage of reproduction, and showed non-linear treatment effects. High dose group females had significantly elevated testosterone during laying, incubation and chick-rearing associated with reduced reproductive success in 2008. Corticosterone concentrations showed a general pattern of depression in dosed birds. Dosed birds had a higher corticosterone response than controls to an experimentally induced stressor, in the form of reduced food availability. Chronic, sub-lethal methylmercury exposure resulted in altered sexual preference of males, changes in courtship behavior, and altered concentrations of steroid hormones. I estimated a 15-50% reduction in reproductive success in dosed groups when compared to the controls as a result of male-male pairing and reduced female fledging success. Chronic, low-level mercury exposure could affect fitness and population parameters of wild populations, and has conservation implications as methylmercury pollution continues to be a problem in many regions.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Nilmini Jayasena.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Frederick, Peter C.
Local: Co-adviser: Oli, Madan K.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0041539:00001


This item has the following downloads:


Full Text

PAGE 1

1 EFFEC TS OF CHRONIC METHYLMERCURY EXPOSURE ON REPRODUCTIVE SUCCESS, BEHAVIOR, AND STEROID HORMONES OF THE WHITE IBIS ( Eudocimus albus ) By NILMINI JAYASENA 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 2010

PAGE 2

2 2010 Nilmini Jayasena

PAGE 3

3 To my parents

PAGE 4

4 ACKNOWLEDGMENTS First and foremost, I wish to extend my sincere gratitude to my ma jor adviser, Dr. Peter Frederick for his support and guidance throughout this project I n addition to writing the initial proposal and getting funding, he was always there for advice, encouragement and help in herding ibises. He was truly a mentor in eve ry sense of the word. I would also like to acknowledge the guidance and input of my supervisory committee, Drs. Iske Larkin, Madan Oli, Katie Sieving and Mel Sunquist. I received tremendous support from Dr. Michael Avery, Kandy Keacher and other personne l at the USDA Wildlife Research Center, Gainesville where the aviary was located. Dr. Marilyn Spalding gave valuable advice and training in various aspects of ibis health care. I wish to thank Dr. Lou Guillette and his lab for graciously allowing me use of his lab for radioimmunoassays. I would like to thank Dr. Ben Bolker and James Colee for statistical consultations. Many volunteers and assistants helped in collecting the data, I am grateful for all their help. I extend my special thanks to Nancy Mon tes, Teresa Bryan and Pilar Jaramillo for being friends as well as enthusiastic assistants. Leslie Straub and Bobbi Jo Sampson helped in general upkeep of the aviary and in keeping the birds happy. I would like to thank all the staff in the department, C aprice McRae, Monica Lindberg, Jennifer Vann, Claire Williams, Delores Tillman, Laura Hayes, Elaine Culpepper and Sam Jones for patiently and cheerfully helping me in numerous ways and clearing up many logistical problems A large number of friends helped me in numerous ways a big thank goes out to all of them, especially to John Simon, Evan Adams, Kate Williams, Louise Venne, Ross Tsai, Fangyuan Hua, and Advait Edgaonkar. Many other friends helped me with words of encouragement, I would like to acknowle dge Gamika Prathapasinghe, Gehan

PAGE 5

5 Jayasuriya, Rasika Jinadasa, Harsha Ariyarathna, Naveen Wijesena, Bhagya Janananda and Utta ra Munasinghe for their friendship and support throughout. I would be remiss if I did not mention all my teachers, especially my sc ience teachers in primary and secondary school, as well as my lecturers at the University of Peradeniya, Sri Lanka, who kept me interested in science. Most of all, I wish to acknowledge my parents and sister for all their love, encouragement, guidance and support given throughout my life.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURES ................................ ................................ ................................ ........ 13 ABSTRACT ................................ ................................ ................................ ................... 16 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 19 2 E FFECTS OF CHRONIC METHYLMERCURY EXPOSURE ON PAIRING BEHAVIOR AND REPRODUCTIVE SUCCESS OF THE WHITE IBIS ................... 24 Introduction ................................ ................................ ................................ ............. 24 Metho ds ................................ ................................ ................................ .................. 29 Experimental Setup ................................ ................................ .......................... 29 Exposure of Captive Birds to Dietary Methylmercury ................................ ....... 29 Behavioral Sampling during Breeding ................................ .............................. 31 Statistical Analyses ................................ ................................ .......................... 32 Results ................................ ................................ ................................ .................... 35 Sex Ratio and Mercury Levels ................................ ................................ .......... 35 Pair Formation and Nesting Success ................................ ............................... 35 Clutch Size, Hatchability and Fledging Success ................................ ............... 38 Discussion ................................ ................................ ................................ .............. 40 Homosexual Pairing and Reproductive Success ................................ .............. 40 Clutch Size, Hatchability and Fledging Success ................................ ............... 43 Conclusions ................................ ................................ ................................ ............ 45 3 EFFECTS OF CHRONIC METHYLMERCURY EXPOSURE ON COURTSHIP AND PARENTAL BEHAVIOR OF WHITE IBISES ................................ .................. 58 Introduction ................................ ................................ ................................ ............. 58 Methods ................................ ................................ ................................ .................. 60 Experimental Setup and Dietary Methylmercury Exposure of White Ibises ...... 60 Behavioral Sampling during the Breeding Season ................................ ........... 61 Provision of breeding material and nest checks ................................ ......... 61 Courtship behavior ................................ ................................ ..................... 62 Incubation behavior ................................ ................................ .................... 63 Statistical Analyses ................................ ................................ .......................... 64 Results ................................ ................................ ................................ .................... 66 Nest Initiations and Courtship Behavior ................................ ........................... 66 Parental Behavior ................................ ................................ ............................. 67

PAGE 7

7 Discussion ................................ ................................ ................................ .............. 68 Nest Initiations and Courtship Behavio r ................................ ........................... 68 Parental Behavior ................................ ................................ ............................. 71 Conclusions ................................ ................................ ................................ ............ 72 4 EFFECTS OF CHRONIC ME THYLMERCURY EXPOSURE ON SEX STEROIDS OF WHITE IBISES DURING THE BREEDING SEASON .................... 83 Introduction ................................ ................................ ................................ ............. 83 Methods ................................ ................................ ................................ .................. 88 Dietary Methylmercury Exposure of Captive White Ibises ................................ 88 Behavioral Sampling during the Breeding Season ................................ ........... 88 Collection of Fecal Samples ................................ ................................ ............. 89 Radioimmunoassays for Steroid Hormone Metabolites in Fecal Extracts ........ 91 Statistical Analyses of Hormone Data ................................ .............................. 92 Results ................................ ................................ ................................ .................... 94 Reproductive Behavior ................................ ................................ ..................... 94 Effects of Methylmercury on Estradiol in Adult Females ................................ .. 94 Effects of Methylmercury on Estradiol in Adult Males ................................ ....... 95 Ef fects of Methylmercury on Testosterone in Adult Females ............................ 97 Effects of Methylmercury on Testosterone in Adult Males ................................ 99 Discussion ................................ ................................ ................................ ............ 100 Effects of Methylmercury on Sex Steroids ................................ ...................... 100 Homosexual Pairing and Endocrine Disruption of Sex Steroids in Males ....... 103 Reproductive Success and Endocrine Disruption in Females ........................ 106 Dose response Relationships between Methylmercury and Endocrine Disrupti on ................................ ................................ ................................ .... 107 Conclusions ................................ ................................ ................................ .......... 108 5 EFFECTS OF CHRONIC METHYLMERCURY EXPOSURE ON CORTICOSTERONE RESPONSES OF WHITE IBISES DURING REPRODUCTI ON ................................ ................................ ................................ 127 Introduction ................................ ................................ ................................ ........... 127 Methods ................................ ................................ ................................ ................ 131 Experimental Setup and Expos ure of Captive White Ibises to Methylmercury ................................ ................................ ............................. 131 Monitoring Breeding Behavior ................................ ................................ ........ 131 Experimental Reduction of Food Availability du ring Breeding ........................ 132 Collection of Fecal Samples ................................ ................................ ........... 132 Radioimmunoassays for Corticosterone Metabolites in Fecal Extracts .......... 133 Statistical Analyses ................................ ................................ ........................ 133 Results ................................ ................................ ................................ .................. 134 Effects of Methylmercury and Food Stress on Corticosterone in Adult Females ................................ ................................ ................................ ...... 134 Effects of Methylmercury and Food Stress on Corticosterone in Adult Males 136 Disc ussion ................................ ................................ ................................ ............ 138

PAGE 8

8 Effects of Methylmercury on Corticosterone during Reproduction .................. 138 Effects of Methylmercury and Food Stress on Cortic osterone ........................ 139 Effects of Methylmercury on Corticosterone in Homosexually paired Males .. 141 Conclusions ................................ ................................ ................................ .......... 142 6 CONCLUSION ................................ ................................ ................................ ...... 155 Introduction ................................ ................................ ................................ ........... 155 Synthesis ................................ ................................ ................................ .............. 155 Male male Pairing in Ibises Induced by Methylmercury Exposure ................. 155 Effects of Methylmercury on Reproduction in Females ................................ .. 159 Effects of Methylmercury on White Ibis Populations ................................ ....... 161 Conclusions ................................ ................................ ................................ .......... 162 APPENDIX A MEAN REPRODUCTIVE PARAMETE RS ................................ ............................ 168 B FECAL CONCENTRATIONS OF ESTRADIOL AND TESTOSTERONE METABOLITES IN WHITE IBISES ................................ ................................ ....... 170 C FECAL CONCENTRATIONS OF CORTICOSTE RONE METABOLITES IN WHITE IBISES ................................ ................................ ................................ ...... 174 LIST OF REFERENCES ................................ ................................ ............................. 177 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 195

PAGE 9

9 LIST OF TA BLES Table page 2 1 Results of chi 2 ) tests testing deviance of sex ratios from one, in each treatment group and year. ................................ ................................ .......... 47 2 2 Total mercury levels in feathers and blood for each treatment group. ................ 47 2 3 Results of linear models testing nest initiation dates of heterosexual and homosexual males in each treatment group in each year ................................ .. 47 2 4 Results of chi 2 ) tests for proportions of homosexual and heterosexual pair days. ................................ ................................ ...................... 48 2 5 Percentage of females potentially excluded from breeding in free living colonies due to male m ale pairing. ................................ ................................ ..... 48 2 6 Results of Kruskal Wallis tests for treatment effects on clutch sizes, hatchability and fledging rates. ................................ ................................ ........... 48 2 7 Means and standard errors (S.E.) of young fledged per female over three breeding attempts and the number of successful attempts over the same period.. ................................ ................................ ................................ ............... 49 3 1 Model results for number of hea d bobs per observation session by treatment group in 2007. ................................ ................................ ................................ ..... 74 3 2 Model results for average number of head bobs per male per observation session by treatment and pairing type in 2008. ................................ .................. 74 3 3 Model results for number of pair bows per observation session by treatment group in 2007. ................................ ................................ ................................ ..... 74 3 4 Model results for average number of pair bows per pair per observation session by treatment and pairing type in 2008. ................................ .................. 74 3 5 Model results for number of aggressive acts per observation session by treatment group in 2007. ................................ ................................ .................... 75 3 6 Model results for average number of aggressive acts per male per observation session by treatment and pairing type in 2008. ............................... 75 3 7 Model results for average number of approaches by females per displaying male per observation session by treatment and pairing type in 2008. ................ 75 3 8 Model results for average number of a pproaches by males per displaying male per observation session by pairing type in 2008. ................................ ....... 75

PAGE 10

10 3 9 Model summaries for proportion of time spent at nest per observation session during the incubatio n stage, by sex and treatment group in 2007 and 2008. ................................ ................................ ................................ .................. 76 4 1 Model summaries of the best models explaining estradiol concentrations in white ibis females. ................................ ................................ ............................ 110 4 2 Model summaries of the best models explaining estradiol concentrations in white ibis males. ................................ ................................ ............................... 111 4 3 Coefficients for each MeHg dosed group comparing estradio l concentrations in each breeding stage relative to control males. ................................ .............. 112 4 4 Model summaries of the best models explaining testosterone concentrations in white ibis females. ................................ ................................ ........................ 112 4 5 Model summaries of the best models explaining testosterone concentrations in white ibis males. ................................ ................................ ........................... 113 5 1 Model summaries of the best models ex plaining corticosterone concentrations in white ibis females. ................................ ................................ 144 5 2 Coefficients comparing each dosed group with control females during each breeding stage in 2008. ................................ ................................ .................... 144 5 3 Model summaries of the best models explaining corticosterone concentrations in white ibis males. ................................ ................................ ... 145 5 4 Coefficients comparing each dosed group with control males during each breeding stage in 2008 ................................ ................................ ..................... 146 5 5 Coefficients comparing homosexual males in each dosed group with heterosexual males within the same group, by breeding stage in 20 08. .......... 146 5 6 Coefficients comparing dosed males with control males during the period of food restriction in 2008. ................................ ................................ .................... 146 6 1 Estimat ed reduction in reproductive output of dosed birds due to male male pairing (unproductive nesting) and decreased fledgling production in females in comparison to the control group. ................................ ................................ .. 166 A 1 Mea n clutch sizes, hatchability rates and fledging rates for each treatment. .... 168 B 1 Mean and standard errors (S.E.) of fecal estradiol metabolites in female ibises in the first breeding attempt, 2007 for each treatment group and breeding stage. ................................ ................................ ................................ 170

PAGE 11

11 B 2 Mean and standard errors (S.E.) of fecal estradiol metabolites in female ibises in the second breeding attempt, 2007 for each treatme nt group and breeding stage. ................................ ................................ ................................ 170 B 3 Mean and standard errors (S.E.) of fecal estradiol metabolites in female ibises in the first breeding attempt, 2008 for each treatment group and breeding s tage. ................................ ................................ ................................ 170 B 4 Mean and standard errors (S.E.) of fecal estradiol metabolites in male ibises in the first breeding attempt, 2007 for each treatment group and breeding stage. ................................ ................................ ................................ ................ 171 B 5 Mean and standard errors (S.E.) of fecal estradiol metabolites in male ibises in the second breeding attempt, 2007 for each treatment group and breeding stage. ................................ ................................ ................................ ................ 171 B 6 Mean and standard errors (S.E.) of fecal estradiol metabolites in male ibises in the first breeding attempt, 2008 for each treatment group and breeding stage. ................................ ................................ ................................ ................ 171 B 7 Mean and standard errors (S.E.) of fecal testosterone metabolites in female ibises in the first breeding attempt, 2007 for each treatment group and breeding stage. ................................ ................................ ................................ 172 B 8 Mean and standar d errors (S.E.) of fecal testosterone metabolites in female ibises in the second breeding attempt, 2007 for each treatment group and breeding stage. ................................ ................................ ................................ 172 B 9 Mean and standard errors (S.E.) of fecal testosterone metabolites in female ibises in the first breeding attempt, 2008 for each treatment group and breeding stage. ................................ ................................ ................................ 172 B 10 Mean and standard errors (S.E.) of fecal testosterone metabolites in male ibises in the first breeding attempt, 2007 for each treatment group and breeding stage. ................................ ................................ ................................ 173 B 11 Mean and standard errors (S.E.) of fecal testosterone metabolites in male i bises in the second breeding attempt, 2007 for each treatment group and breeding stage. ................................ ................................ ................................ 173 B 12 Mean and standard errors (S.E.) of fecal testosterone metabolites in male ibises in the first b reeding attempt, 2008 for each treatment group and breeding stage. ................................ ................................ ................................ 173 C 1 Mean and standard errors (S.E.) of fecal corticosterone metabolites in female ibises in the first breeding attempt, 2 007 for each treatment group and breeding stage. ................................ ................................ ................................ 174

PAGE 12

12 C 2 Mean and standard errors (S.E.) of fecal corticosterone metabolites in female ibises in the second breeding attempt, 2007 for each trea tment group and breeding stage. ................................ ................................ ................................ 174 C 3 Mean and standard errors (S.E.) of fecal corticosterone metabolites in female ibises in the first breeding attempt, 2008 for each treatment group and br eeding stage. ................................ ................................ ................................ 174 C 4 Mean and standard errors (S.E.) of fecal corticosterone metabolites in female ibises during the period of food restriction, 2008 for each treatment group and breeding sta ge. ................................ ................................ .......................... 1 75 C 5 Mean and standard errors (S.E.) of fecal corticosterone metabolites in male ibises in the first breeding attempt, 2007 for each treatment group and breeding stage. ................................ ................................ ................................ 175 C 6 Mean and standard errors (S.E.) of fecal corticosterone metabolites in male ibises in the second breeding attempt, 2007 for each treatment group and breeding stage. ................................ ................................ ................................ 175 C 7 Mean and standard errors (S.E.) of fecal corticosterone metabolites in male ibises in the first breeding attempt, 2008 for each treatment group and breeding stage. ................................ ................................ ................................ 176 C 8 Mean and standard errors (S.E.) of fecal corticosterone metabolites in male ibises during the period of food restriction, 2008 for each treatment group and breeding stage. ................................ ................................ .......................... 176

PAGE 13

13 LIST OF FIGURES Figure page 2 1 The male: female ratio of birds in each treatment group in each year.. .............. 50 2 2 The mean and standard deviation of t otal feather mercury in each treatment group for each year. ................................ ................................ ........................... 51 2 3 Mean and standard deviation of blood mercury levels in 2008 for each treatment group ................................ ................................ ................................ 52 2 4 Proportions of nests that contained eggs for each treatment group in each breeding season.. ................................ ................................ ............................... 53 2 5 Proportions of unproductive nests due to male male pairing. ............................. 54 2 6 Proportions of males nesting homosexually for each treatment group in each year.. ................................ ................................ ................................ .................. 55 2 7 Proportions of pair days taken up by male male pair bonds (homosexual pair days) for each treatment in each year.. ................................ ....................... 56 2 8 Proportions of females successful in fledging at least one young per breeding season in 2007 and 2008.. ................................ ................................ ................. 57 3 1 Differences in head bobbing rates between treatment groups in 2007 and 2008.. ................................ ................................ ................................ ................. 77 3 2 Differences in pair bowing rates be tween treatment groups in 2007 and 2008 ................................ ................................ ................................ .................. 78 3 3 Differences in aggression rates between treatment groups in 2007 and 2008. .. 79 3 4 Mean ( standard error) number of approaches from females to a displaying male per observation session in each treatment group, and for homosexually paired males in 2008.. ................................ ................................ ........................ 80 3 5 Mean ( sta ndard error) number of approaches from males to a displaying male per observation session for heterosexual and homosexual males in 2008.. ................................ ................................ ................................ ................. 81 3 6 Nest attentiveness in male and female white ibi ses during incubation. .............. 82 4 1 Parameter estimates of estradiol concentrations of female white ibises, compared to the control group in 2007 (first and second breeding attempts) and 2008 (first breeding attempt). ................................ ................................ ..... 114

PAGE 14

14 4 2 Parameter estimates of estradiol concentrations of male white ibises during pre breeding, compared to the control group in 2007 (first and second breeding attempts) .. ................................ ................................ .......................... 115 4 3 Parameter estimates of estradiol concentrations of male white ibises during display, compared to the control group in 2007 (first and second breeding attempts) and 2008 (first breed ing attempt).. ................................ .................... 116 4 4 Parameter estimates of estradiol concentrations of male white ibises during nest building, compared to the control group in 2007 (first and second breeding attempts) and 2008 (first breeding attempt).. ................................ ..... 117 4 5 Parameter estimates of estradiol concentrations of male white ibises during egg laying, compared to the control group in 2007 (first and second breedin g attempts) and 2008 (first breeding attempt).. ................................ .................... 118 4 6 Parameter estimates of estradiol concentrations of male white ibises during incubation, compared to the control group in 2007 (first and second breeding attempts) and 2008 (first breeding attempt).. ................................ .................... 119 4 7 Parameter estimates of estradiol concentrations of male white ibises during chick rearing, compared to the control group in 2007 (first and second breeding attempts) and 2008 (first breeding attempt).. ................................ ..... 120 4 8 Parameter estimates of estradiol concentrations of homosexual male white ibises in the dosed groups, co mpared to heterosexual males within the same group for the respective breeding stage (2008 first breeding attempt). ............ 121 4 9 Parameter estimates of testosterone concentrations of female white i bises, compared to the control group in 2007 (first and second breeding attempts). .. 122 4 10 Parameter estimates of testosterone concentrations of female white ibises, compared to the control g roup during each breeding stage in 2008 (first breeding attempt).. ................................ ................................ ........................... 123 4 11 Parameter estimates of testosterone concentrations of male white ibises, compared to the control group in 2007 (firs t and second breeding attempts). .. 124 4 12 Parameter estimates of testosterone concentrations of male white ibises, compared to the control group during each breeding stage in 2008 (first breed ing attempt). ................................ ................................ ............................ 125 4 13 Parameter estimates of the testosterone concentrations of homosexual male white ibises in the dosed groups, compared to heterosexual males (2008 first breeding attempt). ................................ ................................ ............................ 126

PAGE 15

15 5 1 Parameter estimates of corticosterone concentrations of dosed females compared to control females in the first and second breeding attempts of 2007. ................................ ................................ ................................ ................ 147 5 2 Parameter estimates of corticosterone concentrations of dosed females compared to control females in each breeding stage during the first breeding attempt of 2008.. ................................ ................................ ............................... 148 5 3 Parameter estimates of corticosterone concentrations of females during the period of experimentally induced food stress in 2008, compared to the period of normal food availability. ................................ ................................ ................ 149 5 4 Parameter estimates of corticosterone concentrations of dosed and homosexually paired males compared to heterosexually paired control males during the first and second breeding attempts of 2007. ................................ .... 150 5 5 Parameter estimates of corticosterone concentrations of dosed males compared to control males in each breeding stage during the first breeding attempt of 2008.. ................................ ................................ ............................... 151 5 6 Parameter estimates of corticosterone concentrations of dosed homosexually paired males compared to heterosexually paired males in the respective treatment group during the first breeding attempt of 2008.. .............................. 152 5 7 Parameter estimates of corticosterone concentrations of dosed and homosexual males, compared to heterosexual control males, during the period of experimentally induced food stress in 2008. ................................ ...... 153 5 8 Parameter estimates of corticosterone concentrations of males during the period of experimentally induced food stress in 2008, compared to the period of normal food availability.. ................................ ................................ ............... 154 6 1 Possible pathways leading to altered sexual preference and male male pairing following methylmercury exposure. ................................ ....................... 167 A 1 Treatment assignments in the aviary for each b reeding season.. ..................... 169

PAGE 16

16 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 EFFEC TS OF CH RONIC METHYLMERCURY EXPOSURE ON REPRODUCTIVE SUCCESS, BEHAVIOR, AND STEROID HORMONES OF THE WHITE IBIS ( Eudocimus albus ) By Nilmini Jayasena May 2010 Chair: Peter C. Frederick Major: Wildlife Ecology and Conservation Methyl mercury is a widespread enviro nmental contaminant, with bioaccumulative properties. Effects of chronic ecologically relevant levels of mercury on wildlife are poorly known though some of the most sensitive end points of mercury toxicity are manifested as reproductive and behavioural i mpairment due to chronic exposure. I looked for experimental evidence of physiolo gical, reproductive and behavioral impairment at environmentally relevant doses in the White Ibis ( Eudocimus albus ), a species known to be chronically expo sed in the Florida Everglades. W ild caught birds, kept in a free flight aviary, were exposed to methylmercury from 90 days to 3 years of age. Treatment groups received 0.0001 (control), 0.01, 0.03 and 0.3 ppm (wet weight) methylmercury in their diet. R eproductive success of ibises was monitored in 2006, 2007 and 2008. Reproductive behavior and fecal concentrations of steroid hormone metabolites w ere monitored during 2007 and 2008 breeding seasons I found dose related changes in male pairing behavior, with up to 55% of do sed males nesting as male male pairs resulting in up to a 30% decrease in reproduction in comparison to control birds Dosed males showed significantly reduced rates of display

PAGE 17

17 during courtship with the greatest reduction being in homosexual males. Fema les approached high dose group and homosexual males to a significantly lesser degree. Males nesting with males initiated nests earlier in the breeding season compared to heterosexual pairs. The probability of switching partners between breeding attempts was significantly lower in male male pairs, with some pairs continuing to nest together for all three years. Overall fledging success in low and high dose group females were 33 35% lower than in control females. 20 30% of dosed females nested late or did not nest at all as males were unavailable due to male male pairing. I report effects of methylmercury on fecal concentrations of estradiol, testosterone and corticosterone metabolites in both male and female ibises. The most prominent changes in estradi ol were increased concentrations in dosed males during the courtship stage, seen to a greater extent in homosexual dosed males. Changes in estradiol concentrations of females were inconsistent and non linear, but dosed birds generally had lower concentrat ions compared to controls. Testosterone concentrations in both sexes had significant interactions with stage of reproduction, and showed non linear treatment effects. High dose group females had significantly elevated testosterone during laying, incubati on and chick rearing associated with reduced reproductive success in 2008. Corticosterone concentrations showed a general pattern of depression in dosed birds. Dosed birds had a higher corticosterone response than controls to a n experimentally induced st ressor, in the form of reduced food availability Chronic, sub lethal methylmercury exposure resulted in altered sexual preference of males, changes in courtship behavior, and altered concentrations of ste roid

PAGE 18

18 hormones I estimated a 15 50% reduction in r eproductive success in dosed groups when compared to the controls as a result of male male pairing and reduced female fledging success. Chronic, low level mercury exposure could affect fitness and population parameters of wild populations, and has conserv ation implications as methylmercury pollution continues to be a problem in many regions.

PAGE 19

19 CHAPTER 1 INTRODUCTION Mercury (Hg) is a widespread global pollutant, with both natural and anthropogenic sources of emission, but the latter has increased up to thr ee fold over the past century (Bergan et al. 1999) Estimations of pre industrial natural emissions of Hg into the atmosphere were about 500 Mg/year; however, anthropogenic emissions at present are estimated to range fr om 2200 4000 Mg/year (Selin 2009) Combustion of fossil fuels is a major contributor to Hg emissions, and projected increase in use of coal powered electricity, especially in Asia, could raise emissions from this source b y 50% by the year 2050 (Streets et al. 2009) Hg being volatile, the atmosphere takes a major role in the biogeochemical cycling of this element and h igh levels of mercury have been reported in ecosystems far from sources of pollution (Schroeder and Munthe 1998, Bergan et al. 1999) Thus, Hg pollution is truly a global problem as regions far from sources may be contaminated. Hg contamination is prevalent in many aquatic ecosystems with direct atmospheric deposition, transpor t from watersheds and direct discharges from point sources all contributing to emissions (Zillioux et al. 1993, St. Louis et al. 1994, Ullrich et al. 2001, Merritt and Amirbahman 2009, Selin 2009) Aquatic environments including wetlands can be very efficient at converting elemental Hg t o the more toxic and bioaccumulative methylated form which is then accumulated in aquatic organisms (Gilmour et al. 1992, Zillioux et al. 1993, Loftus 2000, Ullrich et al. 2001) While effects of methylmercury (MeHg) are known to occur in a wide variety of species across various taxa, animals higher up the food chain are among the most vulnerable to high environmental exposure (Loftus 2000) This is due to the bioaccumulative property of MeHg, a lipid

PAGE 20

20 soluble compound (Wolfe et al. 1998) B ioconcent r ation factors of MeHg exceeding 10 6 water concentrations are the norm in fish in upper trophic levels in contaminat ed waters (Zillioux et al. 1993) Therefore, piscivorous wildlife are at high risk for environmental MeHg exposure (Scheuhammer et al. 2009) Due to higher absorption from the gut, lipophilicity, lower rates of metabolism and excretion resulting in longer half life, MeHg is more toxic than inorganic forms (Scheuhammer 1987) MeHg toxicity causes central nervous system damage as it is lipophilic and can cross the blood brain barrier in contrast to inorganic Hg which cannot penetrate this ba rrier so readily (Wolfe et al. 1998) MeHg is known to have adverse effects over a wide range of organ systems in wildlife (Thompson 1996, Wolfe et al. 1998, Scheuhammer et al. 2009) T hese effects include changes in motor coordination, neurotoxic effects, embryotoxic effects, immunotoxic effects, endocrine changes, and effects on growth, survival and reproduction (e.g. Heinz 1979, Spalding et al. 1994, Sepulveda et al. 1999b, Spal ding et al. 2000a, Heinz and Hoffman 2003a, Brasso and Cristol 2008, Evers et al. 2008, Hawley et al. 2009) Many aquatic birds that are upper trophic level species (e.g. piscivores ) are exposed to MeHg through their diet (Scheuhammer 1987, Sundlof et al. 1994, Frederick et al. 1 999, Sepulveda et al. 1999a, Frederick et al. 2002, Henny et al. 2007) MeHg effects on birds are species and dose dependent, but are capable of causing multiple adverse effects (reviewed in Wolfe et al. 1998, Scheuhammer et al. 2009) Acute MeHg toxicity in birds (usually at levels of > 15 ppm total Hg in liver in many species) produces progressive motor impairment following brain lesions, reduced appetite and weight loss and ultimately death (Scheuhammer 1987, Scheuhammer et al. 2009) Exposure to

PAGE 21

21 sublethal doses of MeHg has been associated with decrease d appetite, weight loss and emaciation (Scheuhammer et al. 1998, Spalding et al. 2000a) though causal relationships were not established in the former study. Sublethal doses of MeHg have less obvious but equally as devastating effects as it affects re productive p erformance. MeHg at sublethal doses (ranging from 0.25 to 2 ppm in eggs) has embryotoxic effects, decreases fertility of eggs, decreases hatchability due to early embryonic mortality (Heinz and Hoffman 2003a, b, Heinz et al. 2009, Scheuhammer et al. 2009) It has also been shown that Hg decreases survival of chicks and post fledging birds (Heinz and Locke 1976, Spalding et al. 2000b) Exposure to MeHg has been associated with altered levels of reproductive hormones in several avian species, including in Ciconiiforms (Heath and Frederick 2005, Tan et al. 2009) MeHg exposure is also associated with behavioral changes during reproduction such as decreased nesting effort abnor mal incubation and chick provisioning behavior (Barr 1986, Heath 2002, Evers et al. 2004, Heath and Frederick 2005) As is evident from the studies described above, MeHg has adverse effects on many aspects of reproduction. However, there have been few controlled studies studying long term low dose reproductive effects in birds, despite the fact that free living birds are chronically exposed in contaminated environments (but see Hei nz 1976b, a, 1979) Further, the links between chronic MeHg exposure, endocrine disruption, behavioral impairment and subsequent reduction in reproductive success have not been studied in a controlled environment in birds. For example, there is a positi ve correlation between increased MeHg exposure, alterations in sex steroids during reproduction and reduced nesting effort of white ibises ( Eudocimus albus ) in the Florida Everglades

PAGE 22

22 (Heath and Frederick 2 005) There is also a strong association between reduced reproductive success, impaired parenting behavior, increased stress hormones and Hg exposure in common loons ( Gavia immer Evers et al. 2004, Evers et al. 2008) Although the weight of evidence indicates that MeHg exposure was directly related to the above mentioned reproductive deficits and altere d hormone levels, the fact that they were field studies precluded finding causal links between the two. Confounding factors in the field such as reduced food availability, habitat degradation, predation, diseases, and presence of other contaminants preven t establishment of a causal link between MeHg exposure and the above reproductive effects. I hypothesized that exposure to chronic MeHg at sub lethal levels can cause physiological, reproductive and behavioral changes in upper trophic level aquatic birds, with endocrine disruption of reproductive hormones being a main mechanism for reproductive and behavioral impairment. I further hypothesized that these changes would be of a sufficient magnitude to cause population level effects. I specifically predicted the following effects in MeHg dosed birds in comparison to controls: reduced courtship behavior resulting in reduced nesting attempts reduced clutch size s and impaired h atchability impaired incubation behavior (less time spent on the nest) reduced fledg ing success increased nest abandonment with additional stressors changes in sex steroids associated with reduced courtship activity higher stress hormones (corticosterone) in dosed birds, and highest during a period of added stress leading to nest abandonm ent.

PAGE 23

23 I experimentally tested my predictions on an aquatic bird, the white ibis ( Eudocimus albus ), which is exposed to MeHg in its natural environment through its dietary habits of consuming macro invertebrates and small fishes (Kushlan 1979, Heath et al. 2009) Previous studies have found associations between altered endocrine function during reproduction of wild white ibises with dietary exposure to MeHg (Heath 2002, Heath and Frederick 2005) thus making it a suitable species to study effects of sublethal doses on the endocrine system and its effect on reproduction. To avoid confounding factors, I tested my hypotheses in a controlled experimental setting, by exposing captive white ibises to dietary MeHg at 0.05, 0.1 and 0.3 ppm wet weight (ran ge of MeHg in prey items, Loftus 2000) from 90 days of age to 3.5 years. I report the findings in the following chapters.

PAGE 24

24 CHAPTER 2 EFFECTS OF CHRONIC M ETHYLMERCURY EXPOSUR E ON PAIRING BEHAVIO R AND REPRODUCTIVE SUC CESS OF THE WHITE IB IS Introduction Me thylmercury (MeHg) is a globally distributed environmental contaminant, which has toxic effects on a variety of organisms. The rate of mercury (Hg) deposition has increased at least by 50%, and possibly up to three times since pre industrial times, mainly due to anthropogenic activities (Bergan et al. 1999) The majority of anthropogenic emissions are from combustion of fossil fuels, and although emissions are declining in Europe and North America, they are increasing i n Asia which accounts for 50% of total global emissions (reviewed in Selin 2009) There is uncertainty about future emissions, with projected changes by 2050 estimated to be 4% to 96% of the to tal emissions in 2006 (Streets et al. 2009) Of the two main forms of atmospheric Hg, elemental and divalent Hg, divalent Hg is the main form to be de surface (Mason et al. 2005, Selin et al. 2007) An issue of concern is that the amount of divalent Hg is projected to increase in the future which would lead to increased local deposi tion, and subsequent conversion to MeHg (Gilmour et al. 1992, S elin 2009, Streets et al. 2009) MeHg, an organic form of Hg, is more toxic than elemental Hg due to its higher absorption, and lower rates of metabolism resulting in a higher half life within organisms (Scheuhammer 1987, Wolfe et al. 1998) Due to its property of bio accumulation, organisms higher up the trophic chain have higher levels of MeHg in their tissues (Scheuhammer 1987, Zillioux et al. 1993) MeHg is capable of causing neurotoxicity, e mbryotoxicity, immunotoxicity, endocrine disruption, and various reproductive effects in a variety of taxa (Wolfe et al. 1998, Scheuhammer and Sandheinrich 2008, Scheuhammer et al. 2009, Tan et al. 2009)

PAGE 25

25 MeHg contamination has been recognised as a problem in ma ny aquatic systems (Zillioux et al. 1993, St. Louis et al. 1994, Ullrich et al. 2001, Canari o et al. 2007, Rumbold et al. 2008) Since MeHg is biomagnified through the food web, animals on top of the food web, such as piscivorous birds, are among the most highly exposed (Scheuhammer 1987, Sundlof et al. 1994, Wolfe et al. 1998, Frederick et al. 1999, Loftus 2000) Further, aquatic MeHg has been discovered to move into adjacent terrestrial food webs (Cristol et al. 2008) and there a re an increasing number of studies reporting effects on terrestrial feeding birds (Custer et al. 2007, Condon and Cristol 2009, Franceschini et al. 2009, Wada et al. 2009a) A wide range of effects of MeHg on aquatic birds have been documented, ranging fr om toxic or lethal effects to more subtle manifestations due to chronic sublethal exposure. While these effects vary depending on species, usually liver Hg concentrations of more than 15 ppm in birds have been associated with toxic consequences and death (Scheuhammer et al. 2009) Sublethal effects of MeHg in birds include reduced appet ite (Spalding et al. 2000a) immunosuppressive effects (Spalding et al. 2000b, Kenow et al. 2007, Hawley et al. 2009) impaired reproduction (Heinz 1979, Brasso and Cristol 2008, Burgess and Meyer 2008, Evers et al. 2008, Hill et al. 2008) behavioral effects (Nocera and Taylor 1998, Bouton et al. 1999, Evers et al. 2004) and endocrine disruption (Heath and Frederick 2005, Adams et al. 2009, Franceschini et al. 2009, Wada et al. 2009a) Some of the more sensitive endpoints of MeHg toxicity are manifested as reproductive effects (Scheuhammer 1987) Egg Hg concentrations of more than 1 ppm wet weight (ww) has been associated with embryotoxic effects and impaired hatchability in a variety of birds, with sensitive species having effects at less t han 0.25 ppm ww

PAGE 26

26 (Heinz and Hoffman 2003a, Hill et al. 2008, Heinz et al. 2009, Scheuhammer et al. 2009) It has also been shown to decrease reproductive output in free living birds at blood Hg levels of more than 3 ppm (Brasso and Cristol 2008, Burgess and Meyer 2008, Evers et al. 2008) For example, common loons ( Gavia immer ) had a 40% reduction in production of fledglings in birds with >3 ppm blood Hg, compared to birds with <1 ppm blood Hg (Evers et al. 2004, Evers et al. 2008) MeHg is known to be an endocrine disruptor, capable of affecting many systems including reproductive hormones (reviewed in Tan et al. 2009) and exposure has been associated with altered sex steroids in breeding birds with feather Hg concentrations at <20 ppm (Heath 2002, Heath and Frederick 2005) Reproductive behavior is another sensitive endpoint of MeHg effects. Extensive studies in common loons have shown that loons with blood Hg levels of >3 ppm spent less time sitting on nests, less time foraging for young, and increased time spent brooding and resting (Evers et al. 2004) In an experimental dosing study on mallards ( Anas platyrhynchos ), females exposed to dietary MeHg of 0.5 ppm laid more eggs outside nest boxes than control females (Heinz 1979) In young of Hg exposed birds, behavioral effects include less time spent brooding effects in loon chicks (Nocera and Taylor 1998) and lower response to maternal calls in mallard chicks (Heinz 1979) The Florida Everglades is an ecosystem that has been extensively contaminated by Hg and many upper trophic level fauna have been found to have high levels of MeHg (Frederick 1999, Loftus 2000, Rumbold et al. 2008) Great egret ( Ardea albus ) nestlings in this ecosystem were estimated to receive Hg ranging from 0.37 to 0.47 mg/kg in fish, a level high enough to cause detrimental effects on health and survival

PAGE 27

27 (Frederick et al. 1999) White ibises ( Eudocimus albus ) in the Everglades with feather Hg levels ranging from 0.33 20 ppm showed an association between altered reproductive hormone levels and increased Hg exposure (Heath and Frederick 2005) The same study showed correlative evidence of decreased numbers of white ibises nesting in years where a standardized Hg exposure index was high. The authors sugg ested that Hg exposure could be causing lower numbers due to decreased nesting effort or early abandonment of nests or a combination of both. While field evidence is correlative in nature, there is sufficient reason to believe that high levels of Hg expo sure could be related to decreased nesting attempts (i.e. fail to nest at all) and lower nest success in Ciconiiforms (nest failure of birds that proceed to nesting) in the Everglades. Laboratory studies have found Hg effects which could contribute to rep roductive effects in free living birds in several ways. For example, experimental MeHg exposure of 0.4 ppm ww in food resulted in suppressed immune responses in common loon chicks (Kenow et al. 2007) while 0.5 ppm ww MeHg in diet reduced appet ite in great egret nestlings (Spalding et al. 2000a) Both immunosuppression and nutritional deficiencies can result in lower survival of chicks and fledglings, and also reduce reproductive effort in adults (Martin 1987, Simons and Martin 1990, Gustafsson et al. 1994, Saino et al. 1997, Moller et al. 1998, Naef Daenzer et al. 2001) In addition, poor nutrition itself can hav e detrimental effects on immunity, causing positive feedback in the above cycle (Lochmiller et al. 1993, Alonso Alvar ez and Tella 2001) Therefore, in free living species, reduced food availability and diseases can interact with Hg effects on immunity and reproduction, and cause enhancement of Hg effects. Confounding factors in the field, such as in the example

PAGE 28

28 above, preclude finding a causal relationship between chronic MeHg exposure and impaired reproductive success. MeHg by itself can act via multiple interacting pathways to reduce reproductive success as outlined previously. MeHg can have direct embryotoxic effec ts in many avian species (Heinz et al. 2009, Scheuhammer et al. 2009) it c an reduce reproductive success via hormonal pathways (Drevnick and Sandheinrich 2003, Tan et al. 2009) and it can adversely affect reproductive behavior (Heinz 1979, Barr 1986, Nocera and Taylor 1998, Evers et al. 2004, Sandheinrich and Miller 2006) I hypothesized that chronic, sublethal exposure to MeHg will cause multiple physiological and behavioral effects which will result in reduced reproductive success in aquatic b irds. I specifically predicted that MeHg exposure will result in reduced nesting attempts, reduced clutch size, reduced hatchability and reduced fledging success, with all of these contributing to poorer reproductive success in exposed birds. I tested my predictions on an aquatic bird, the white ibis ( Eudocimus albus ); a species exposed to chronic MeHg exposure in the Florida Everglades. I chose this species since previous research shows an association between MeHg exposure and reduced nesting effort of white ibises in the Everglades during years of peak contamination as well as an association with altered levels of sex steroids during reproduction (Heath and Frederick 2005) In order to avoid confoundi ng factors present in the field (as outlined above), I conducted an experimental study on captive white ibises, by exposing them to MeHg (0.05 0.3 ppm ww, the range of MeHg found in their prey items in the Everglades, Loftus 2000) via their diet for more than three years. I monitored reproductive success for three breeding seasons, the results of which are reported in this chapter. Chapter 3

PAGE 29

29 d ocuments effects of MeHg on behavior, chapter 4 deals with effects on sex steroids, and chapter 5 on effects on corticosterone. Methods Experimental S etup White ibises were experimentally exposed to MeHg via diet over the course of 3.5 years. In spring 2005, 12 32 day old nestlings were collected from breeding colonies from 2 locations: the Alley North colony of the Everglades (Water Conservation Area 3, Broward Co. Florida, N 26 11.179, W 80 31.431) and near White Springs, Hamilton Co., Florida (N 3 0 19.900, W 82 45.367) Birds were genetically sexed (Avian Biotech International, Tallahassee, FL), individually tagged using leg bands and randomly assigned to four treatment groups control, low, medium and high. Each group comprised approximatel y 20 birds of each sex. They were housed in Gainesville, Florida in a 1200 m 2 free flight, circular aviary divided into quadrants by net walls Perches, nest platforms and a wading pool were placed in each cage in equal numbers and a similar orientation I switched cage assignments of the groups in the fall of 2006 and 2007 (after each breeding season) in order to reduce location dependent effects. Each of the new locations were determined by placing treatment groups in a novel quadrant, and one that a llowed new neighbouring groups each time. Thus, each breeding season was in a new location for each treatment group (Fig. A 1). Exposure of Captive Birds to Dietary M ethylmercury Exposure to dietary MeHg was started when the birds were 85 95 days old and continued for the duration of the experiment. Dosed groups were given methylmercury chloride (CH 3 HgCl, hereafter MeHg) via food at 0.05 (low), 0.1 (medium) and 0.3 (high)

PAGE 30

30 ppm wet weight. Solutions used were adjusted to reflect concentrations of CH 3 Hg, ra ther than CH 3 HgCl. These doses span the range of Hg that piscivorous wading birds were exposed to in the Everglades in the early 1990s (Frederick et al. 1999, Loftus 2000) Solutions of MeHg in a corn oil vehicle were sprayed onto food pellets (Flamingo and custom Ibis diets, Mazuri Company, Brentwood, MO, USA) while b eing rotated in cement mixers to ensure even distribution of the solution. The control group received pellets sprayed with the corn oil alone. Persons mixing food batches, observers and caretakers were blinded to the identity of treatment groups. The bi rds fed ad libitum as feed was present at all times in the cages. Thus, while each treatment group was exposed to a constant MeHg concentration in food, total exposure for individual birds depended on individual food intake, which simulates natural condit ions, and accounts for variability in Hg levels within a dose group. Concentrations in feed samples were measured as total Hg (THg) by cold vapor atomic absorption spectroscopy to ensure that doses were as calculated. THg concentrations in feather sampl es of the birds were also measured by the same method in spring 2006, 2007 and 2008 to cross check the relative validity of the dosing procedure. All feather and feed Hg analyses were conducted by the Florida Department of Environmental Protection Chemist ry Section (Tallahassee, Florida), using a modified version of USEPA method 245.6 (Standard Operating Procedure number HG 006 3.14). Blood samples were tested only in 2008, from a random sample of 10 birds per treatment group (five from each sex). THg le vels in blood were analyzed using the direct mercury analysis method by the South Florida Water Management District laboratory.

PAGE 31

31 Behavioral Sampling d uring B reeding I monitored breeding behavior in 2006, 2007 and 2008. While breeding in the first 12 mont hs of life has not been recorded in wild white ibises (Heath et al. 2009) the captive birds started breeding i n their first year (mid March to early August 2006), while still in partial juvenal plumage. Though not all pairs proceeded to the nest building stage in 2006, almost all displayed courtship behavior. Throughout the breeding periods of each year, I provi ded twigs and fresh cattail grass ( Typha sp ) as nesting material; these were replenished weekly to ensure an ad libitum supply. Nesting platforms were created by suspending pieces of plastic net like mesh underneath parallel bamboo poles in each cage bef ore the start of the first breeding season, with 48 nest platforms provided per cage (eight per perch). In each breeding season, there were un used nest platforms indicating these were supplied in excess. Identities of birds displaying, nest building, l aying, incubating or chick rearing were recorded daily at sunrise throughout each breeding season I observed both male female (heterosexual) and male male (homosexual) pairing in all groups. I inspected e ach nest platform daily for presence/absence of a nest and if present, the status of the nest (i.e. few stick s / partial nest/ full nest). The number of eggs and/or chicks and the pair associated with the nest were recorded. Eggs were recorded each day they were laid, and kept track of until hatching. Eggs that did not hatch during the normal incubation period (21 22 days) were removed from the nests at 25 30 days after the lay date. Chicks were individually marked by leg bands and monitored until fledging or death. Thus, the fate of each nest with regards to parents/eggs/chicks was individually known on a daily basis. Observers were kept blind to treatment groups until after the experiment was completed in 2008.

PAGE 32

32 Statistical Analyses R statistical software version 2.10.0 (R Development Core Team 2009 ) was used for statistical analyses. I used chi square tests to determine whether the sex ratio in each group was significantly different from one. contain eggs (u nproductive nests), and proportions of homosexually nesting males between the control and Hg dosed groups during each breeding season. I did a similar analysis combining the data from all three breeding seasons and comparing each dosed group to the contro l group. The day of first nest initiation in each year for each group was denoted as Julian day one (2006: 30 March; 2007: 20 February; 2008: 26 February). I used analyses of variance (ANOVA) to test for differences in nest initiation dates within each t reatment group, with the type of pair bond (male male or male female) as the explanatory variable. Only the first nesting attempt was considered in this latter analysis. I estimated the percentage of females who would potentially be prevented from breed ing in a free living colony with no re nesting attempts. For this analysis I made the assumption that those females whose first attempt were with re nesting males (whose previous attempts could be homosexual or heterosexual) would have nested earlier had males been available, and would have been excluded from breeding in wild colonies due to the shorter breeding season. I also assumed that females that were in breeding condition, yet did not nest, were excluded due to unavailability of males due to male m ale pairing. For the high dose group, the only group with a more females than males; I deducted the excess number of females (n=2) to avoid violating the previous assumption. I summed these two categories of females for each treatment group in

PAGE 33

33 2007 and 2 008 and expressed it as a percentage. I excluded 2006 from this analysis to prevent confounding effects of age. The proportion of the breeding season spent as male male or male female pairs was quantified by summing the number of days for each pairing typ e for all the individual pairs in each treatment group (heterosexual/homosexual pair days, including re nesting attempts). I used chi square tests to test for differences in total numbers of heterosexual and homosexual pair days by treatment group for eac h breeding season. Generalized linear models (GLMs; binomial distributions with logit links) were used to test whether pairing type of the previous breeding season affected whether birds switched partners in the subsequent breeding season. I used Wilcoxon rank sum tests to examine clutch size differences between the count of all eggs laid in a clutch irrespective of whether they survived to hatch age; date clutch date (21 days post laying). Some birds laid multiple clutches, and a clutch was defined as having at least 10 days interval between the last egg of the earlier clutch and the first egg of the subsequent clutch, and/or, if another courtship period intervened between the two clutches. Hatching and fledging rates were analyzed using non parametric tests, since the data were non normally distributed. Hatchability was defined as the proporti on of eggs were defined as young who survived at least up to 40 days, at wh ich time they gained

PAGE 34

34 proportion of eggs that survived till hatch date that day fledging Wallis tests and fledging rates. Where these were si gnificant (P <0.05) or marginally significant (P wise comparisons using Wilcoxon rank sum tests with Bonferroni adjusted p values were performed, since the main comparison of interest was whether each dosed group was different from the controls All clutch sizes, hatchability and fledging rates were analyzed separately by breeding attempt to avoid pseudoreplication (Hurlbert 1984) Only the first breeding attempt was included in 2006, while both first and second att empts for 2007 and 2008 were considered. I tested whether there were treatment (treatment = MeHg exposure group) differences in the number of birds that failed to produce nestlings in at least one of the first two breeding attempts of the 2007 and 2008 bre eding season. I limited this analysis to the first two breeding attempts since in the wild, nesting conditions would not usually remain suitable for more than a single re nesting attempt (Heath et al. 2009) I excluded the 2006 breeding season since free living birds do not usually reproduce during their first year (Heath et al. 2009) I tested this in each sex separately, using (males ), heterosexual pairing with no egg production, birds that with clutches where no eggs survived to hatch age, and birds that failed to nest at all. Nestlings were defined as a hatched chick, with post hatch survival not being taken into account. I also t ested

PAGE 35

35 whether there were differences in fledging success for individual females by treatment successful in producing at least one young per breeding season, in 2007 and 2008. The total number of fledglings produced per female was calculated on an individual basis for all three years. A GLM (using a quasi Poisson distribution) was used to determine whether there were treatment effects on total fledgling production. A similar anal ysis was performed for the total number of successful breeding attempts per individual female, summed over the three years. Results Sex Ratio and Mercury Levels The sex ratio was not significantly different from one in any group year combination (chi squar e tests, all p values >>0.05, Table 2 1; Fig. 2 1). Levels of feather Hg in the birds showed a clear dose dependent relationship each year (Table 2 2; Fig. 2 2), and feather Hg levels spanned the range found in wild birds both prior to and during the peri od of Hg contamination in the Everglades (Frederick et al. 2004) Blood Hg levels in 2008 sho wed a similar dose related pattern (Table 2 2; Fig. 2 3). Pair Formation and Nesting Success Even in the first year, some birds were successful in laying fertile eggs, and hatching young. In subsequent years all birds participated in courtship behavior an d almost the entire original cohort of birds was successfu l in pair formation and nest building. During each breeding season I documented nests without eggs (unproductive nests). In 2006, the medium and high dose groups showed a trend towards having more 0.13 respectively; Fig. 2 4) but there was no difference between low and control groups

PAGE 36

36 (P = 0.67). In 2008, low and medium groups had significantly more unproductive nest s a similar trend in the high dose group (P = 0.053). All three dosed groups showed similar non pared against the control; Low: P = 0.12; Medium: P = 0.12; High: P = 0.057). In the analysis combining data from all three breeding seasons, all dosed groups had significantly more 0.002; Medium: P = 0.001; High: P = 0.001). The loss of productivity in dosed groups ranged from 10 30%; 14 24% and 7 9% in 2006, 2007 and 2008 respectively. The majority of the unproductive nests were attended by male male pairs (2006: 75 85%, 2007: 82 100%, 2008: 50 100%; Fig. 2 5), although a few were attended by heterosexual pairs. For all three years combined, the loss of productivity compared to the control was 13.2%, 14.6% and 13.5% for low, medium and high groups respectively. Of this, th e proportions due to homosexual pairs were 73.7%, 91.3% and 77.8% for low, medium and high groups respectively. The high dose group had a significantly higher number of homosexual 6: P = 0.042; 2007: P = 0.027; 2008: P = 0.041; Fig. 2 6). The medium dose group had significantly more homosexual pairs in 2007 and 2008, with a non exact tests; 2006: P = 0.12; 2007: P = 0.019; 2008: P = 0.023) while the low dose group showed a tendency towards a higher number of homosexual males in 2007 and 2008 all three years, the control group had a maximum of 20% homosexual mal es (in 2006)

PAGE 37

37 which decreased to 0% by 2008. Respective maximum percentages for low, medium and high groups are 36.4% (in 2007), 43.5% (in 2006) and 55.6% (in 2006), which decreased respectively to 18.2%, 26.1% and 22.2% by 2008. Homosexually nesting males showed reproductive behaviors that were largely similar to that of heterosexually paired males (but see chapter 3 for important differences in courtship behavior). This included a courtship period prior to nest building, and maintenance of the pair bond after nest building, including taking turns sitting on the empty nests. Most male male pairs had well constructed nests, similar to that of male female pairs. Several of these males who subsequently formed heterosexual bonds, produced fertile eggs. In mo st treatment groups, in most years, males that nested homosexually initiated nest building significantly earlier than heterosexual males (Table 2 3). The only exceptions were the control group in 2006 and high dose group in 2007 where there were no signif icant relationships between nest initiation date and type of pair bond. Thus, males were pairing with males at a time of the season when many unpaired females were available. While some homosexually nesting males proceeded to pair with females within the same or subsequent breeding season, others continued with the same male partners. A significantly larger proportion of the total pair days in the breeding season were occupied by unproductive male male pairs in each of the dosed groups than in the contro l group in all three years (chi square tests, all p values <0.0001; Table 2 4; Fig. 2 7). Homosexually paired males were significantly less likely to switch their partners from one year to the next when compared with heterosexually

PAGE 38

38 nesting males, irrespec tive of treatment group (GLMs; 2006 to 2007: P = 0.0024; 2007 to 2008: P = 0.0033). The percentage of females who would potentially be prevented from breeding in a free living colony with no re nesting attempts (non nesting females + females nesting for th e first time with re nesting males) were 5.9%, 22.2%, 27.8% and 30% for control, low, medium and high dose groups in 2007 (Table 2 5). In 2008, the only group with females in this category was the high dose group, with the percentage being 20%. In each y ear, 50% of the late nesting females nesting had partners who nested homosexually in their previous attempts. Clutch Size, Hatchability and Fledging Success There were no significant differences in either uncorrected or hatch date clutch sizes between the control and any dosed group in any year (Table 2 6). The marginally significant trend (all differences show higher reproductive success in the respective metric for the control group unless explicitly stated to be not so) in uncorrected clutch size in 200 6 was due to a marginal difference between the control and low dose groups (pair wise comparison; P = 0.11). The significant difference in uncorrected hatchability in 2007, first attempt (Kruskal Wallis test; P = 0.04) was due to a difference between the medium and high dose groups (pair wise comparison: P = 0.047) and none of the pair wise comparisons between the control and dosed groups were significant. The similar effect in hatch date hatchability for the same year (Kruskal Wallis test; P = 0.04) was again not due to differences with the control, but was between the low and high dose group (pair wise comparison; P = 0.04). The marginal trend in uncorrected hatchability in 2008, first attempt (Kruskal Wallis test; P = 0. 07) was due to marginal differe nces between control and low dose groups (pair wise comparison; P = 0.17), while a similar

PAGE 39

39 trend in hatch date hatchability (Kruskal Wallis test; P = 0.05) was due to a difference between control and high dose groups (pair wise comparison; P = 0.07). Fledg ing success was extremely low in all groups in 2006, with no differences in comparison to the control in uncorrected, hatch date, or 40 day fledging rates (Kruskal Wallis tests; P>0.1; Table 2 6). In 2007 (second attempt), the Kruskal Wallis test for 40 d ay fledging rates was marginally significant (P = 0.09), but this was due to the medium dose group having a higher fledging rate than the low dose group (pair wise comparison; P = 0.08). In the first breeding attempt of 2008, the marginal trend in uncorre cted fledging rate (Kruskal Wallis test; P = 0.08) was due to differences between the medium and high dose groups (pair wise comparison; P = 0.19); while the similar trend in hatch date fledging rate (Kruskal Wallis test; P = 0.06) was due to difference be tween the control and medium groups (pair wise comparison; P = 0.15). The medium group had a marginally higher hatch date fledging rate (0.29 0.05 standard error) than the control group (0.11 0.06) in the first breeding attempt, while the positions we re switched for the second breeding attempt (control: 0.29 0.10; medium: 0.04 0.04). In the second breeding attempt of 2008, there were marginally significant difference in uncorrected fledging rates (Kruskal Wallis test; P = 0.07) and hatch date fle dging rates (Kruskal Wallis test; P = 0.08), but none of the pair wise comparisons showed any trends (P > 0.1). Means and standard errors of all clutch sizes, hatchability rates and fledging rates are presented in Appendix A. Nestling production of dose d males was significantly lower than in control males in Medium (M): P = 0.002; C vs. High (H): P = 0.04; 2008: C vs. L: P = 0.049; C vs. M: P =

PAGE 40

40 0.022; C vs. H: P = 0. 017). Nestling production of dosed females were significantly 0.003; C vs. M: P = 0.49; C vs. H: P = 0.049), but in 2008, only high dose females had significantly lower 0.009). There were no significant differences in females fledging at least one young (female fledging success) in either 2007 or 2008, though the low dose group in 2007 and the high dose group in 2008 showed a trend towards lower fledging success of 0.31; 2008: C vs. L: P = 0.49; C vs. M: P = 1; C vs. H: P = 0.050; Fig. 2 8). Table 2 7 shows means an d standard errors of the total number of young fledged per female and total number of successful attempts over the three year period I monitored breeding success. The high dose group fledged 34.8% less young per female than controls (GLM; P= 0.085) and ha d 35.4% fewer successful attempts (GLM; P = 0.07). Low dose females fledged 33.5% less young per female in comparison to control females (GLM, P = 0.10). None of the other treatments showed significant differences or trends in comparison with the control in these two metrics (GLMs; P > 0.1; Table 2 7). Discussion Homosexual Pairing and Reproductive Success The main reproductive effect attributable to MeHg dosing seen in my study was the occurrence of male male pairing in dosed birds, an effect which has, t o my knowledge, not been reported so far as an effect of MeHg exposure. However, in most experimental studies on reproductive effects of MeHg, mate choice was highly constrained (Heinz 1979, Heath 2002) Male male pairing behavior has been reported

PAGE 41

41 extensively in many species in both natural and laboratory settings (e.g. Bag emihl 1999, Bailey and Zuk 2009) In many of these cases, males were pairing with males in large part because females are unavailable. A large number of homosexual pairings in captivity, and in the wild, have been associated with strongly skewed sex rat ios (Bagemihl 1999) However, the sex ratio in all groups in my study were close to one (0.9 1.3), and none of them were statistically different from one. Further, the highest proportion of male male pairs occurred in the high dose group for two years, where the sex ratio was 0.9, with more females than males. This lack of relationship between male biased sex ratio and homosexual pairing makes lack of females an unlikely explanation. It also seems unlikely that location effects (= cage effect) caused male male pairing, since each breeding season was spent in a new location with new neighbouring groups. While there remains the possibility of cohort effects ( pairing patterns being due to t he chance and permanent social makeup of the groups) this hypothesis does not explain the relationship between dose and h omosexual pairing. Same sex pairing has not been reported in the white ibis in natural conditions. In a wild colony with minimal exp osure to MeHg, there were no male male pair bonds observed in 134 individually identified white ibis pairs, in more than 15,000 pair hours of observation over four breeding seasons (Frederick 1987a) Since male male pairs were observed at a low frequency in the control gr oup, it seems likely that some degree of homosexual behavior is fomented by some combination of captivity and first time breeding. As pairing behavior has not been studied in wild populations of white ibises heavily exposed to MeHg, it cannot be said for certain whether MeHg affects birds via

PAGE 42

42 male male pairing in the wild. Nonetheless, the evidence from this study indicates that MeHg is capable of impairing key processes in mate choice, and there is no obvious reason why wild populations would be immune f rom these same effects. Female female pairing has been reported in several other avian species and is thought to be a direct result of a strongly female biased sex ratios ( Larus occidentalis : Hunt and Hunt 1977, Larus delawarensis and Larus californicus : Conover et al. 1979, Larus delawarensis : Ryder and Somppi 1979, Larus ar gentatus : Shugart 1980, Sterna caspia : Conover 1983, Phoebastria immutabilis : Young et al. 2008) While organochloride exposure was likely to be the cause of a female biased sex ratio in some of these colonies (Fry and Toone 1981) the proximate cause of same sex p airing is believed to be the effect of biased sex ratios (Conover and Hunt 1984) and not a direct effect of the pollutant. In my study however, altered sexual preference seems to be directly mediated by exposure to MeHg rather than effects of the contaminant on sex ratio. The exact mechanism of this is not clear, but it could possibly be due to effects of MeHg on both behavior (chapter 3 ) and endocrine expression ( chapter 4). MeHg is known as an endocrine disrupting compound (Tan et al. 2009) and alterations in hormone expression in juvenile birds have been reported to cause altered sexual preference in adults (Adkins Regan and Leung 2006) Xenobiotic exposure causing aberrant sexual behavior has also been reported in other studies (Haegele a nd Hudson 1977, Fisher et al. 2001, Toft and Guillette 2005, Fernie et al. 2008, Lee et al. 2008) Few of these however, report same sex pairing to be among the effects of exposure (but see Lee et al. 2008)

PAGE 43

43 The reproductive loss resulting from homosexual pairing in my study ranged from 7 30% depending on treatment group and year, or an average 13 14 % in each treatment over all three years. Even by the third year, dosed groups had 18 26% more homosexual males than in the control group. While I suggest population modeling to determine the long term effects of male male pairing on white ibis populations, the above evidence indicates that pop ulation level effects would be considerable. Same sex pairing in breeding colonies with a sex ratio of one would not only remove the homosexual birds from productive breeding, but would also induce a shortage of partners for the opposite sex. This effect was not as strongly seen in the aviary where the longer breeding season made it possible for birds to have multiple re nesting attempts, thus making my estimates conservative. However, even in the aviary 22 30% of dosed females in 2007, and 20% of high d ose females in 2008, were late nesters or did not nest at all. Only 5.8% of control females were similarly excluded from early nesting in 2007 (none in 2008). At all times, there were more males involved in male male pairs than there were unattached fema les, therefore indicating these females would have formed pairs in the absence of male male pairs. Clutch Size, Hatchability and Fledging Success There was very low fledging success across treatment groups in 2006. This was due in part to lack of prior b reeding experience and to the cage setup. For example, even though feed was available ad libitum in the cages, chicks were not being fed sufficiently by parents. Once chicks gained locomotory ability, some would sometimes fall out of nests, and the struc ture of the perches on which the nests were built would prohibit them from climbing back. These chicks were also subjected to aggression, sometimes leading to mortality by cannibalism when on the ground. Use of

PAGE 44

44 supplemental food and better climbing struc tures in 2007 and 2008 resulted in higher chick survival. For these reasons, any treatment effects present in 2006 may have been masked by lack of paren tal experience and cage setup. Al though there were no large effects of MeHg treatment on clutch size, h atchability and fledging success overall high dose females had an average 35% reduction in the mean numbers of fledglings produced and successful attempts in comparison to the control females over the entire duration of the study, largely due to poor perf ormance in 2008. Although low dose females had a similar 33% reduction, this was mainly due to poor performance in 2007, when inexperience in parenting skills may have enhanced MeHg effects. Inexperience, however, cannot be cited as a factor contributing to failure in high dose females in 2008. Failure in the high dose group is comparable to the 40% loss in chick production seen in common loons ( Gavia immer ) at similar blood mercury levels (>3 ppm in loons, 3.95 0.68 ppm in my study; Evers et al. 2004, Evers et al. 2008) It is noteworthy that the high dose group rec eived doses that were within the range of MeHg exposure in the Everglades and elsewhere (Frederick et al. 1999, Brasso and Cristol 2008, Evers et al. 2008) making these findings relevan t to free living populations. The effects of male male pairing and reproductive deficits in females recorded in my study would be addi tive at the population level. In a worst case scenario, with 13% reduction of productive nesting attempts due to male male pairing, and 35% reduction of fledgling production in heterosexual pairs as in the high dose group, reductions of up to 48% might be possible, resulting in effects noticeable at the population level. The prese nt findings support my hypothesis that exposure to MeHg at ecologically relevant

PAGE 45

45 levels will lead to substantially reduced reproduction in the white ibis. The main pathway leading to reproductive deficits was, however, in large part through the unexpected mechanism of altered sexual preference. This finding adds to present knowledge about effects of chronic MeHg exposure, providing experimentally based insight into effects at the lower end of the MeHg exposure spectrum, as low as 0.05 ppm in the diet. Con clusions Hg contamination remains a problem to both aquatic and terrestrial avian wildlife in North America (Rimmer e t al. 2005, Brasso and Cristol 2008, Burgess and Meyer 2008, Cristol et al. 2008, Evers et al. 2008, Tsipoura et al. 2008) with several regions being strongly affected by contamination (Driscoll et al. 2007, Evers et al. 2007) Although Hg emissions in North America have generally decreased over the past decade, a decreasing trend in Hg deposition has not been seen in south east U.S.A. (Butler et al. 2008) Further, s everal biological hotspots of Hg have been documented, where exposure can impact ecological and human health (Rumbold and Fink 2006, Evers et al. 2007, Rumbold e t al. 2008, Liu et al. 2009) Criteria for identifying biological hotspots in the north east U.S.A. hav e included indicators such as > 3 ppm blood Hg in common loons, and whole fish concentrations (in yellow perch, Perca flavescens and brook trout, Salvel inus fontinalis ) of 0.16 ppm ww. Factors such as regional differences in Hg bioavailability and effects due to factors such as pH (Meyer et al. 1995) biomethylation (Gilmour and Henry 1991, Ullrich et al. 2001) and selenium availability (Cuvin Aralar and Furness 1991, Peakall and Burger 2003) can contribute to geographical differences in exposure. Further more there are species differences in susceptibility due to metabolic differences such as ability to demethylate MeHg

PAGE 46

46 (Spalding et al. 2000a, Henny et al. 2002, Eagles Smith et al. 2009) While these factors could contribute to variability in adverse effects due to Hg exposure across regions and species, information from my study indicates that f or species such as the white ibis, reproductive effects could occur at much lower levels (0.73 0.03 ppm in blood with dietary exposure of 0.05 ppm ww in the low dose group) than have been previously documented. I nformation about Hg effects on wildlife ha s increased over the past several years (see reviews in Wolfe et al. 1998, Scheuhammer et al. 2009, Tan et al. 2009, Weis 2009) however, there are few controlled studies of reproductive effects at lower exposures. Field studies are essenti al to determine integrated effects of multiple interacting stressors on populations, and they portray a more realistic picture of combined effects. However, experimental studies are better at defining lowest observed adverse effects levels (LOAELs) for co ntaminant exposure and for subsequently setting management guidelines for polluted environments (Spalding et al. 2000b) The present study is, to my knowledge, the first avian experimental study to look for effects of MeHg at chronic dietary exposure levels as low as 0.05 ppm (ww). Therefore, these findings will be relevant to establishing mercury LOAELs for piscivorous wildlife and monitoring wildlife population responses to environmental MeHg exposure.

PAGE 47

47 Table 2 1. Results of chi 2 ) tests testing deviance of sex ratios from one, in each treatment group and year. Year Control Low Medium High 2 value P value 2 value P value 2 value P value 2 value P value 2006 0.012 0.912 0.050 0.823 0.107 0.744 0.000 0.995 2007 0.016 0.899 0.050 0.823 0.107 0.744 0.000 0.995 2008 0.000 1.000 0.166 0.684 0.166 0.684 0.000 1.000 Note: All chi 2 values) are for one degree of freedom Table 2 2. Total mercury levels in feathers and blood for each treatme nt group. Total mercury (mg/kg fw) Year Control Low Medium High Mean S.D. Mean S.D. Mean S.D. Mean S.D. Feathers 2006 0.74 0.25 7.15 2.60 15.24 8.65 23.86 8.77 2007 0.47 0.11 8.20 1.53 14.13 5.92 51.32 12.33 2008 0.62 0.21 4.31 1.28 17.96 9.15 35 .04 16.94 Blood 2008 0.07 0.01 0.73 0.09 1.60 0.32 3.95 0.68 Note: All concentrations are on fresh weight (fw) basis. S.D.: Standard deviation. Table 2 3. Results of linear models testing nest initiation dates of heterosexual and homosexual males in ea ch treatment group in each year Treatment group Year Intercept S.E. Coefficient S.E. R squared F value P value n Control 2006 21.50 5.14 6.50 9.62 0.04 0.46 0.51 14 2007 36.33 3.28 31.33 10.37 0.34 9.13 0.0007 20 2008 Low 2006 19.00 9.95 12.00 6.03 0.25 3.97 0.07 14 2007 24.00 3.09 14.75 5.13 0.29 8.27 0.009 22 2008 27.39 1.25 10.89 2.93 0.41 13.73 0.001 22 Medium 2006 31.22 4.64 15.82 6.39 0.26 6.13 0.02 19 2007 20.54 2.36 15.34 3.57 0.4 7 18.43 0.0003 23 2008 22.12 2.49 13.45 4.88 0.27 7.61 0.01 23 High 2006 31.29 3.57 19.69 4.65 0.54 17.92 0.0007 17 2007 22.80 3.62 6.80 5.44 0.09 1.57 0.23 18 2008 20.14 1.70 10.14 3.60 0.33 7.93 0.01 18 Note: S.E.: Standard error; n: sample size; Intercept term: Julian date for heterosexual nest initiation; Coefficient: Difference from intercept for homosexual nests. There were no homosexual nests in the control group for 2008.

PAGE 48

48 Table 2 4. Results of chi square ( 2 ) tests for proportions of homosexual and heterosexual pair days. Year Control vs. Low Control vs. Medium Control vs. High 2 value P value 2 value P value 2 value P value 2006 66.64 <0.0001 91.19 <0.0001 93.67 <0.0001 2007 563.02 <0.0001 354.72 < 0.0001 243.92 <0.0001 2008 311.74 <0.0001 448.55 <0.0001 206.68 <0.0001 Note: All chi 2 values) are for one degree of freedom Table 2 5. Percentage of females potentially excluded from breeding in free living colonies due to male male pa iring Treatment Total 2007 2008 females No. not nesting No. nesting with re nesting males Percentage potentially excluded No. not nesting No. nesting with re nesting males Percentage potentially excluded Control 17 0 1 5.8 0 0 0.0 Low 18 2 2 22.2 0 0 0.0 Medium 18 1 4 27.8 0 0 0.0 High 20 1 5 30.0 0 4 20.0 Table 2 6. Results of Kruskal Wallis tests for treatment effects on clutch sizes, hatchability and fledging rates. Parameter Statistics 2006 2007 2008 Attempt 1 Attempt 1 Attempt 2 Attemp t 1 Attempt 2 Uncorrected clutch size 2 value 6.83 1.03 0.98 2.72 1.05 P value 0.08 0.79 0.82 0.44 0.79 Hatch date clutch size 2 value 3.68 2.68 0.93 1.04 2.90 P value 0.30 0.44 0.82 0.79 0.41 Uncorrected hatchability 2 value 0.98 8.11 0.56 7.06 3.48 P value 0.81 0.04 0.90 0.07 0.32 Hatch date hatchability 2 value 0.85 8.17 1.55 7.95 4.48 P value 0.84 0.04 0.67 0.05 0.21 Uncorrected fledging rate 2 value 0.26 0.63 3.21 6.70 7.02 P value 0.97 0.89 0.36 0.08 0.07 Hatch date fledging rate 2 value 0.47 1.02 5.35 7.23 6.66 P value 0.93 0.80 0.15 0.06 0.08 40 day fledging rate 2 value 0.91 0.60 6.58 5.76 4.93 P value 0.82 0.90 0.09 0.12 0.18 Note: All chi 2 values) are for three degrees of freedom

PAGE 49

49 Table 2 7. Means and standard errors (S.E.) of young fledged per female over three breeding attempts and the number of successful attempts over the same period. The t value and P values from the generalized linear models are shown for each model. Each dosed group is compared to the control (intercept). Treatment Young per female Successful attempts Mean S.E. t value P value Mean S.E. t value P value Control (Intercept) 1.76 0.25 3.51 0.001 1.47 0.19 2.45 0.017 Low 1.17 0.27 1.64 0.105 1.11 0.25 1.19 0.239 Medium 1.83 0.23 0.17 0.865 1.61 0.18 0.43 0.672 High 1.15 0.22 1.75 0.085 0.95 0.17 1.83 0.072

PAGE 50

50 Figure 2 1. The male: female ratio of birds in each treatment group in each year. Shows only birds that were of breeding age (i.e. juveniles excluded). Only t he original cohort collected from the wild as nestlings were breeding during 2006 and 2007. In 2008, the few birds that were fledged during 2006 are included in the breeding cohorts. Differences in the ratio between 2006 and 2007 (only in the control gro up) were due to mortality.

PAGE 51

51 Figure 2 2. The mean and standard deviation of total feather mercury in each treatment group for each year.

PAGE 52

52 Figure 2 3. Mean and standard deviation of blood mercury levels in 2008 for each treatment group (n=10 for e ach treatment).

PAGE 53

53 Figure 2 4. Proportions of nests that contained eggs for each treatment group in each breeding season. All dosed groups were compared to the control group for m arginally significant d 0.1), two asterisks (**) indicat e a significant difference (P < 0.05) relative to the control group.

PAGE 54

54 Figure 2 5. Proportions of unproductive nests due to male male pairing. All nesting attempts within each breeding season are considered.

PAGE 55

55 Figure 2 6. Proportions of males nesting homosexually for each treatment group in each year. All males were from the original cohort, and had a similar duration of mercury exposure. All dosed groups were compared to the contro l group for marginally significant difference (P e a significant difference (P < 0.05) relative to the control group.

PAGE 56

56 Figure 2 7. Proportions of pair days taken up by male male pair bonds (homosexual pair days) for each treatment in each year. Chi square tests were used to compare each dosed group to the control group for the relevant year. All tests were highly significant (P<0.0001).

PAGE 57

57 Figure 2 8. Proportions of females successful in fledging at least one young per breeding season in 2007 and 2008. Fem ales in each dosed group compared (*) indicates a marginally significant difference (P 0.1).

PAGE 58

58 CHAPTER 3 EFFECTS OF CHRONIC M ETHYLMERCURY EXPOSUR E ON COURTSHIP AND PARENTAL BEHAVIOR OF WHITE IBISES Introduction Behavioral studies are being increasingly used to study effects of environmental contaminants (e.g. McCarty and Secord 1999a, Bell 2001, Weis et al. 2001, Grue et al. 2002, Gorissen et al. 2005) Behavioral endpoints are considered to be comprehensive and se nsitive indicators of exposure since they integrate both physiological and biochemical processes of contaminants (Peakall 1996, Zala and Penn 2004) Behavioral modification can result in many adverse effects that affect survival of individuals and even scale up to population, community and ecosystem level impacts (reviewed in Grue et al. 2002) For example, mummichogs ( Fundulus heteroclitus ) from a highly polluted estuary in New Jersey, U.S.A., had impaired prey capture and predator avoidance behavior, which apparently resulted in reduced longevity while prey species achieved higher densities p resumably due to reduced predation pressure (Weis et al. 1999) Effects of contaminants on behavior during reproduction, such as impaired courtship, nesting or parenting abilities, are especially capable of having population level impacts (Haegele and Hudson 1977, McCarty and Secord 1999a, b, Saaristo et al. 2009) It has been recogn ized for some time that higher trophic level organisms such as piscivorous birds are especially vulnerable to bioaccumulative contaminants (Grasman et al. 1998) The methylated form of mercury (CH3Hg, MeHg hereafter) is a widespread bioaccumulative contaminant, known to cause reproductive impairment in many bird species (Burgess and Meyer 2008, Evers et al. 2008, Hill et al. 2008) Behavioral effects of MeHg during avian reproduction include effects on both chicks and adults. In

PAGE 59

59 an experimental study, mallards ( Anas platyrhynchos ) exposed to MeHg showed abnormal e gg laying behavior as well as decreased responsiveness of chicks to maternal calls (Heinz 1979) Common loon ( Gavia immer ) chicks in mercury contaminated natural lakes showed decreased time spent brooding by back riding (Nocera and Taylor 1998) Pairs of zebra finches ( Taeniopygia guttata ) experimentally fed MeHg spent less time caring for eggs than control birds resulting in decreased hatching success (Heath 2002) Other than these effects during reproduction, behavioral effects include decreased activity, tendency to seek s hade and decreased motivation to hunt prey in great egret ( Ardea albus ) chicks experimentally exposed to dietary MeHg (Bouton et al. 1999) There is also correlative evidence suggesting MeHg exposure could affect pair formation and nest fidelity. There were lower numbers of territorial pairs of loons in mercury polluted lake systems in Canada, as well a s increased nest abandonments (Barr 1986) White ibises ( Eudocimus albus ), in the Florida Everglades showed an inverse relationship between nesting numbers and MeHg exposure, suggesting that exposure reduced the likelihood of nesting and/or increased nest abandonment (Heath and Frederick 2005) Nest abandonment during periods of adverse environmental conditions is a major reason for nest failure in this spe cies (Frederick and Collopy 1989) Following these studies, I hypothesized that chronic MeHg exposure can affect reproductive success by: a) impaired courtship behavior and subsequent pair formation leading to reduced numbers of nest initiations; and b) impaired incubation behavior leading to reduced hatchability. I predicted reduced rates of display behavior and

PAGE 60

60 increased rates of aggressive behavior would be associated with MeHg exposure and that b oth lead to would be reduced pair formation and nest initiations. I further predicted that MeHg exposed birds would have reduced nest attentiveness leading to poor nest success and poor hatching success. I also hypothesized that behavioral pathways would be associated with changes in hormone profiles, described in greater detail in the two subsequent chapters. I chose the white ibis ( Eudocimus albus ) as a representative species to test my hypotheses, since it is a upper trophic level wading bird with inc reased risk of exposure to dietary MeHg, and because previous studies showed that reproductive success may be affected by exposure (Heath and Frederick 2005) I tested these hypotheses by observing repro ductive behavior in captive ibises chronically exposed to dietary MeHg at levels found in their prey items in the Everglades (Loftus 2000) Methods Experimental Setup and Dietary Methylmercury Exposure of White Ibises White ibises were caught as nestlings from wild colonies in South Florida in spring 2005. They were kept in a 1200 m 2 circular, free flight aviary divided into four quadrants by net walls. Approximately 20 birds of each sex were randomly assigned to each of the four treatment gr oups. Birds were genetically sexed (Avian Biotech International, Tallahassee, FL), and individually tagged using leg bands prior to placement in treatment groups. MeHg exposure was started when the birds were 90 days old and continued throughout the durat ion of the experiment (2005 through 2008). The dietary MeHg exposure rates used in this experiment spanned the range found in prey of ibises in the Everglades during the mid 1990s (Frederick et al. 1999, Loftus 2000) Low, medium

PAGE 61

61 and high treatment groups corresponded respectively to 0.05, 0.1 and 0.3 ppm ww MeHg in di et. Doses were introduced into feed in a corn oil vehicle sprayed onto pelletized food (Flamingo and custom Ibis diets, Mazuri Company, Brentwood, MO, USA). The control group received food pellets sprayed with only the corn oil vehicle. Behavioral Sam p ling d uring the Breeding Season Provision of breeding material and nest checks Reproduction in white ibises is generally initiated during the second year (Heath et al. 2009) In the captive colony, birds started displaying courtship activity and nesting in 2006, as one year old juveniles. Captive birds were in breeding condition by mid February in 2007 and 2008, though in 2006 display behavior started mid March. Re nesting attempts were frequent and breeding continued through August. Each group was provided with 6 perch modules in a similar configuration with each perch having 8 platforms with plastic netting to provide a cup shaped base for nest building (48 per group). I provided all groups with an ad libitum supply of nesting material in the form of twigs and fresh cattail leaves ( Typha sp ), replenished weekly from mid February to end August. I performed da ily 10 15 minute observations of reproductive activity within each treat ment group from approximately 8 m outside the aviary edge using a spotting scope (16 48x 65mm). All observations were done within an hour of sunrise. During this time, I identified th rough leg bands all individuals that were showing courtship behavior, nest building behavior, incubation, brooding or other chick rearing behavior. Thereafter, I entered each cage and used a mirror pole to detect the presence of eggs and chicks. Eggs tha t did not hatch were removed from the nests at 25 30 days after the lay date. Chicks were individually marked by leg bands and kept track of until

PAGE 62

62 fledging or death. Thus, the status of each nest with regards to identity of reproductive adults, and numbe rs of eggs/chicks was individually known on a daily basis. Courtship behavior 2007 breeding season: Between February 26 th and 10 th May 2007, I observed each treatment group for one hour, within two hours of sunset, at approximately 3 day intervals. Two observers standing 8m from the outside of the cage at the junction of two treatment groups alternately scanned one of the two adjacent treatment groups every 2.5 minutes. Observers were pseudo randomly assigned to a treatment so that each person observed each treatment group twice in a random order within a four day observation period. Observations for any day were begun after a habituation period of five minutes after observers reached their stations. During each scan the number of birds performing the following behaviors were recorded: head bobbing (Rudegeair 1975) ; r aggressive acts performed while on perches. An aggressive act was defined as pecking/jabbing behavior performed to chase away other approaching birds. Head bobbing is a courtship display perfor med early in the display period, prior to pair formation (Rudegeair 1975) Females appr oach males displaying and if accepted by the male, the pair bow together. Males typically display to a number of females before establishing a pair bond and initiating a nest. In 2007, individual identities of birds were not recorded. I observed nesting behavior for a total of 5040 minutes over 21 days in 2007, and discontinued observations when display behavior had become very rare in early May.

PAGE 63

63 2008 breeding season: I modified the observation protocol in 2008 to focus on individual specific behavior. Each treatment group was observed for 40 minutes within two hours of sunset. I pseudo randomly switched the order in which each cage was observed (as well as observers) so that each treatment was observed at equal intensity for each time slot over a peri od of 4 days of observation. As in 2007, observers were placed outside the aviary and used spotting scopes (16 48x 65mm) for identification of leg bands of displaying birds. I recorded the number of head bobs performed per individual male displaying, iden tities of birds approaching each displaying male, and whether the bird was accepted or chased away (rejected) by the males. Rates of pair bowing for pairs, and aggressive acts of displaying males were also recorded. Behavioral observations were started o n 20 th February and continued until 6 th May 2008, by which time almost all birds had initiated nesting. Only data from the first nesting attempt of each bird were used in analyses. Incubation behavior In both 2007 and 2008 I recorded nest attendance durin g incubation. Each treatment group was observed for one hour per sampling session within two hours of local sunrise. Scans were performed every five minutes from outside the aviary with the aid of binoculars and a spotting scope. Nest attentiveness incl uded both incubation and other behaviors associated with caring for the eggs (e.g. shading eggs, defending from length of the nest, and was included as part of nest atte ntiveness. Nest attentiveness also included standing over the nest to shade it. Each nest site and its occupants were individually identified.

PAGE 64

64 Each treatment group was scanned at five minute intervals by an observer alternating attention between two ad jacent focal groups. Using two observers, scans were performed on alternate cages every 2.5 minutes and all groups were sampled at the same time within a session. I performed observations in 2007 from 29 th March to 21 st July (31 days, 7440 minutes) and i n 2008 from 18 th April to 22 nd July (28 days, 6720 minutes). Statistical Analyses R statistical software version 2.10.0 (R Development Core Team 2009) was used (treatment = MeHg exposure groups) differences in the numbers of males and females initiating nests in each breeding season. Generalized linear models (GLMs), with negative binomial errors to account for overdispersion, were used to model 2007 courtship behavior. Res ponse variables were numbers of head bobs, pair bows and aggressive acts while the explanatory variable was treatment group. I standardized 2008 courtship data based on nest initiation dates of individual males prior to statistical analysis. Nest initia tion was defined as the date a pair started building a nest, by placing sticks and other nesting material in a nesting platform. For each male, I averaged behavioral data (numbers of head bobs, pair bows, aggressions, approaches by females and approaches by males) recorded during the 14 days prior to nest th March 2008 (n = 6) were eliminated from the analysis since there were insufficient data from these birds. The data set used in the analyses was recorded over 31 days between 20 th February and 11 th April 2008 (4960 minutes of observation).

PAGE 65

65 The data from 2008 were modeled using GLMs, sometimes fitted with Poisson errors (those with no overdispersion) in addition to negative binom ial errors. Response variables in 2008 were averaged behaviors, and each behavior type was analyzed using a separate model. Explanatory variables were treatment group and/or type of pairing association (male male or male female). Quantile quantile plots and residual were used to determine whether the number of males that accepted advances by males differed according to pair type (male male or male female). For this anal ysis I used the total counts (not averages) of male approaches and the number of those that were accepted and rejected over the standardized 14 day courtship period. Parental attendance at nests were analyzed for each year separately, using linear mixed ef fects models (Pinheiro and Bates 2 000) The proportion of scans during which a bird was attending its nest (nest attentiveness = incubation + nest guarding) was calculated for each sex and sampling session. Proportions were analyzed separately by sex, with bird identity as a random effe ct to eliminate pseudoreplication, and treatment as the fixed effect. Quantile quantile plots of residuals of both random and fixed effects were examined for departures from normality. I tested whether the proportion of time any nest was left unattended (no parent in attendance) differed by treatment group using Kruskal Wallis tests since data did not conform to assumptions of normality. Whe re these were significant ( P < 0.05) or wise comparisons using Wilcoxon rank sum tests with Bonferroni adjusted p values were performed, since the main comparison of interest was whether each dosed group was different from the controls.

PAGE 66

66 Results Nest Initiations and Courtship Behavior There were no significant differences in the num bers of nest initiations in birds of 2008: P = 1; Females: 2006: P = 0.81; 2007: P = 0.5; 2008: P = 0.24). In both years, males in Hg groups had significantly lower hea d bobbing rates than in the control group (GLMs, P <0.05, Tables 3 1 & 3 2; Fig. 3 1). I could not discriminate between display rates of heterosexually and homosexually paired males in 2007 because identity of birds was not recorded in 2007, but in 2008 h omosexual males had significantly lower rates of head bobbing than heterosexual males. Pair bowing rates were significantly lower in both dosed and homosexual males than in control (heterosexual) males in 2008 (GLM, P <0.01) though not in 2007 (Tables 3 3 & 3 4; Fig. 3 2). Aggressive acts by males on perches (display territory) did not differ by treatment group in 2007 (GLM, P >0.1; Fig. 3 3 A; Table 3 5). In 2008 homosexual males performed significantly fewer aggressive acts compared to control males (G LM, P = 0.0096) and there was a non significant trend in the same direction in high dose males (GLM, P = 0.09; Fig. 3 3 B; Table 3 6). In 2008, both high dose males and homosexual males were approached during courtship significantly less often by females t han were control males (GLM, P <0.05; Fig.3 4; Table 3 7). Homosexual males were approached by other males to a greater degree than heterosexual males (GLM, P = 0.001; Fig. 3 5; Table 3 8). A significantly higher proportion of homosexual males accepted m ales who approached them than did

PAGE 67

67 Parental Behavior There were MeHg treatment effects on male nest attentiveness in 2007 (F 3, 63 = 2.8; P = 0.047; Table 3 9), but the only treatment coefficient that w as significantly different from controls was for the medium group males (t = 2.45, P = 0.017). Medium dose group males were present at the nest 42% of the time on average while control males were present 49% of the time (mean reduction 7%, 95% CI: 1 13 %; Fig. 3 6A). Coefficients for males in low and high dose groups did not show differences compared to control males (Low: t = 0.02, P = 0.98; High: t = 0.03; P = 0.97). Females in 2007 showed marginal effects of treatment on nest attentiveness (F 3, 68 = 1.85; P = 0.14; Table 3 9), but the treatment coefficient for medium group females was significantly different from controls (t = 2.08, P = 0.041). Medium group females were present at the nest 59% of the time, an average increase of 6% above control fem ales (95% CI: 0.2 12.5%; Fig. 3 6 A). Low and high group females showed a non significant trend of staying at the nests for a longer percentage of time than did control females (Low: t = 1.69, P = 0.09; High: t = 1.81, P = 0.07). The Kruskal Wallis tes t for proportion of time nests were unattended was marginally significant (P = 0.051) but none of the pair wise comparisons with the control group were significant (P = 1). In 2008, there were no significant differences between treatment groups in either m ale or female nest attendance (m ales: F 3, 69 = 1.04; P = 0.38; f emales: F 3, 70 = 0.63; P = 0.6; Table 3 9, Fig. 3 6 B). The Kruskal Wallis test for proportion time unattended was non significant (P = 0.11) as were the pair wise comparisons with the contro l group (P >0.1).

PAGE 68

68 Discussion Nest Initiations and Courtship Behavior I predicted that exposure to MeHg would cause impairment of courtship behavior with consequent decreases in nest initiations. While rates of key courtship behaviors (head bobbing and p air bowing) were reduced by MeHg exposure, I did not see a reduction in the numbers of nest initiations. However, MeHg exposure was also related to increased male male pairing, which was not a predicted effect ( c hapter 2 ), and which effectively reduced th e numbers of productive nests. A main behavioral effect noticed during courtship was consistently decreased display rates (both head bobbing and pair bowing) in dosed and homosexual males in 2008. In 2007, there was a similar pattern in head bobbing tho ugh not in pair bowing rates. This absence of difference in pair bowing rates in 2007 could be due to the sampling technique. Pair bows were a relatively rare event and it is possible that scan sampling techniques used in 2007 were inadequate to identify differences between groups. Homosexual males (and to a lesser degree, high dose heterosexual males) were also seen to have lower rates of aggression in 2008. MeHg exposure has been associated with decreased activity levels in both great egrets (Ardea albus ; Bouton et al. 1999) and common loons ( Gavia immer ; Evers et al. 2004) Both courtship displaying and aggre ssion are energetic activities, and therefore, the reduced rates I observed may be due to MeHg induced lethargy rather than impaired reproductive behavior per se It is also possible that endocrine changes associated with MeHg exposure (chapter 4) could ha ve affected courtship behavior. The association between sex steroids and reproductive behavior is well documented (Young et al. 1964, Adkins

PAGE 69

69 Regan 2005a, Nelson 2005, Fusani 2008) and I observed changes in sex steroids of dosed males, of which some changes were seen to be similar direction but of a greater magnitude in homosexual males. Another possibilit y is that behavioral changes were not due to changes in circulating levels of hormones, but to organizational effects of MeHg during development, mediated via hormones or through direct action of MeHg on the brain (Adkins Regan et al. 1997, Adkins Regan 2007) However, my study was not designed to reveal mechanistic relationships between MeHg exposure and behavioral changes Another behavioral change I recorded was the lower number of females approaching homosexual males (and to a les ser degree, high dose heterosexual males). One explanation could be that the reduced display rates made these males unattractive to females. Establishment of a display territory and performing head bobbing and similar actions advertise the white ibis mal (Rudegeair 1975) and it is possible that females were preferentia lly approaching males who had higher display rates. A second reason could be that MeHg dosed females were less likely to approach males due to reduced motivation of females to pair, independently of low display rates of males. However, other than in the high dose group, I did not see reduced approaches of females to heterosexual males; therefore, reduced female motivation does not seem to explain the lower number of approaches to homosexual males. It may be that a reduced motivation for display behavior in homosexual males could have been compounded by low female approach rates, further reducing motivation for display. This too is unlikely since unpaired males in my study often had higher rates of display, which indicates female approaches are not necess ary for

PAGE 70

70 motivation. A third reason for low female approach rates may be because these males were already displaying with another male and were therefore occupied. A fourth reason could be due to a lack of some unmeasured aspect of courtship behavior in h omosexual (and high dose group) males (e.g. greetin g calls; Rudegeair 1975) which made them less attractive to females. A recent field study found that oscine passerines in MeHg contaminated sites had a lower diversity of note types and shorter songs than in reference sites (Hallinger et al. 2010) indicating many aspects of courtship behavior may be impaired by MeHg. H omosexual males, in addition to having more males approach them, also accepted approaching males more often than heterosexual males. It seems likely that this was a key step in formation of male male pairs; however, my study is not able to provide insight s into the mechanism of action. Homosexual males could have accepted males because other males were more attractive than females; because they could not distinguish between males and females (impaired sensory perception or learning abilities); or because they had fewer females approaching them. If the last was true, it does not explain why homosexual males initiated nesting earlier than heterosexual males (chapter 2) while there were breeding females available. MeHg exposure at low levels can induce visu al and auditory sensory impairment (Rice and Gilbert 1982, 1992) ; thus it is possible that homosexual males were unable to distinguish between sexes. Male fruit flies ( Drosophila melanogaster ) with elevated dopamine (a neuromodulator) showed increased male m ale courtship and also had sensory deficits, suggesting impaired perception led to male male courtship (Liu et al. 2008) Another possibility is that a MeHg induced learning deficit contributing to inappropriate pairing. One

PAGE 71

71 explanation given for imp aired singing ability in oscine passerines, which require learning to acquire song ability, was that MeHg induced impaired sensory perception interfered with their ability to learn songs (Hallinger et al. 2010) In the same study, the only species that did not have altered song characteristics with MeHg exposure was a su boscine, which does not learn its song (Eastern Phoebe, Sayornis phoebe ), suggesting that altered singing in oscines occurred via learning deficits (Hallinger et al. 2010) Social environment too can contribute to sexual preferences, as shown in estradiol treated zebra finches ( Taeniopygia guttata ) that showed female fem ale pairing only when housed in all female groups (Adkins Regan 2005b) Since the ibises in my study were brought into captivity as nes tlings and reached sexual maturity in absence of other adults (as opposed to what would occur in a natural colony), it is possible that they lacked some aspect of forming sexual preference that are influenced by social experiences. This could explain the occurrence, albeit at a low frequency, of male male pairing in control birds, a behavior not recorded in free living ibises (Frederick 1987b) A MeHg induced learning deficit may possibly have accentuated this in dosed groups. An alternative pathway is MeHg induced demasculinization of behavior which resulted in m ales accepting other males. Although changes in male sexual behavioral patterns and demasculinization of behavior is a known effect of some environmental contaminants, notably those with estrogenic activity (Hayes et al. 2002, Zala and Penn 2004, Milnes et al. 2006) there are no prior records of MeHg having similar actions. Parental Behavior I saw some evidence for treatment effects on parental attentiveness, but did not find a general response in the direction I predicted, nor a clear dose dependent pattern. In at least one dose group (medium), males spent a lower proportion of tim e at the nest

PAGE 72

72 than control males, with a concomitant increase in the proportion of time females spent at the nest. Since there were no differences in the proportion of time nests were left unattended, this suggests that the females were compensating for t he absence of males. Other contaminant studies have observed similar compensatory behaviour when one of a pair is behaviorally impaired (reviewed in Grue et al. 2002) However, I did not see predicted dose related differences in parental attentiveness as there were no changes in the highest dose group. Conclusions Chronic exposure to MeHg at environmentally relevant levels (0.05 to 0.3 ppm ww) altered courtship behavior in the white ibis. Impaired courtship behavior was seen in all dosed groups, but within dosed groups, it was t he males that formed homosexual pair bonds that had the largest differences from the control group. This may indicate that male male pairing was facilitated by the MeHg induced behavioral impairment in dosed birds. However, the evidence from my study is not strong enough to indicate behavioral impairment as the sole or main cause for male male pairing, since formation of sexual preferences is controlled by multiple complex pathways involving phylogenetics, developmental pathways in the brain, hormones, be havior and social environment (Adkins Regan et al. 1997, Adkins Regan and Leung 2006, Tomaszycki et al. 2006, MacFarlane et al. 2007) There is recent information that mercury may have epigenetic actions (Pilsner et al. 2010) and epigenetic impr inting has been shown to alter mate preference (Crews et al. 2007) In the Crews et al. (2007) study, vinclozolin, an endocrine disruptor, was shown to alter mate preferences in female rats three generations removed from exposure, where they preferred non exposed male rats.

PAGE 73

73 While it is not possible to elucidate the mechanism of action of MeHg on behavior from the evidence presented here, it is likely that behaviorally impaired bir ds would have a lower chance of reproductive success in natural environments where they will be competing with unimpaired individuals. Another factor in the field is that other stressors (e.g. predation, competition, food availability, diseases) would int eract with MeHg exposure, possibly lowering thresholds for adverse effects to occur (Grue et al. 2002) It is possible that MeHg exposure may have been a factor in the lower numbers of white ibises nesting in the Florida Everglades during years of higher exposure (Heath and Frederick 2005) however I do not have conclusive evidence for this from my study. Future studies to determine mechanistic pathways may be of value in predicting behavioral responses of wildlife to chronic MeHg exposure.

PAGE 74

74 Table 3 1. Model results for number of head bobs per observation session by treatment group in 2007. Coefficient Estimate Standard error P value Intercept (Control) Low Medium High 2.0851 1.0986 0.7941 1.0018 0.3002 0.4376 0.4325 0.4358 <0.0001 0.0121 0.0663 0.0215 Results of a generalized linear model (GLM) with a negative binomial distribution (theta = 0.54). Sample size (N) = 88; control males are the reference (inter cept) to which other treatment coefficients are compared. Table 3 2. Model results for average number of head bobs per male per observation session by treatment and pairing type in 2008. Coefficient Estimate Standard error P value Intercept (Control) L ow Medium High Homosexual 3.2142 0.7751 0.6905 1.5632 1.4534 0.2166 0.2993 0.3111 0.3339 0.3493 <0.0001 0.0096 0.0264 <0.0001 <0.0001 Results of a generalized linear model (GLM) with a negative binomial distribution (theta = 1.32). N = 75; hete rosexual control males are the reference (intercept) to which other treatment coefficients are compared. Table 3 3. Model results for number of pair bows per observation session by treatment group in 2007. Coefficient Estimate Standard error P value In tercept (Control) Low Medium High 2.068 0.1618 0.2032 0.2436 0.2240 0.3184 0.3189 0.3149 <0.0001 0.611 0.524 0.439 Results of a generalized linear model (GLM) with a negative binomial distribution (theta = 1.02). N = 88; control males are the reference (intercept) to which other treatment coefficients are compared. Table 3 4. Model results for average number of pair bows per pair per observation session by treatment and pairing type in 2008. Coefficient Estimate Standard error P value Inte rcept (Control) Low Medium High Homosexual 2.4551 0.6585 0.8532 0.9346 0.864 0.1530 0.2185 0.2328 0.2440 0.2745 <0.0001 0.0026 0.0002 0.0001 0.0016 Results of a generalized linear model (GLM) with a negative binomial distribution (theta = 3.2) N = 75; heterosexual control males are the reference (intercept) to which other treatment coefficients are compared.

PAGE 75

75 Table 3 5. Model results for number of aggressive acts per observation session by treatment group in 2007. Coefficient Estimate Standa rd error P value Intercept (Control) Low Medium High 2.0389 0.0850 0.0741 0.1634 0.1193 0.1673 0.1675 0.1660 <0.0001 0.6110 0.6580 0.3250 Results of a generalized linear model (GLM) with a negative binomial distribution (theta = 5.47). N = 88; con trol males are the reference (intercept) to which other treatment coefficients are compared. Table 3 6. Model results for average number of aggressive acts per male per observation session by treatment and pairing type in 2008. Coefficient Estimate Stan dard error P value Intercept (Control) Low Medium High Homosexual 1.4572 0.0392 0.0064 0.3167 0.5349 0.1170 0.1618 0.1673 0.1885 0.2067 <0.0001 0.8087 0.9697 0.0929 0.0096 Results of a generalized linear model (GLM) with a Poisson distribut ion. N = 75; heterosexual control males are the reference (intercept) to which other treatment coefficients are compared. Table 3 7. Model results for average number of approaches by females per displaying male per observation session by treatment and pa iring type in 2008. Coefficient Estimate Standard error P value Intercept (Control) Low Medium High Homosexual 0.8557 0.0780 0.2565 0.6162 2.3547 0.1581 0.2241 0.2470 0.2902 0.7157 <0.0001 0.7277 0.2991 0.0337 0.0010 Results of a generali zed linear model (GLM) with a Poisson distribution. N = 75; heterosexual control males are the reference (intercept) to which other treatment coefficients are compared. Table 3 8. Model results for average number of approaches by males per displaying mal e per observation session by pairing type in 2008 Coefficient Estimate Standard error P value Intercept (Heterosexual) Homosexual 3.45 3.05 0.707 0.791 <0.0001 0.0001 Results of a generalized linear model (GLM) with a Poisson distribution. N = 75 ; The intercept is for heterosexual males of all treatment groups.

PAGE 76

76 Table 3 9. Model summaries for proportion of time spent at nest per observation session during the incubation stage, by sex and treatment group in 2007 and 2008. Year Sex Degrees of freed om Explanatory variable (df) F test Total Denominator F value P value 2007 Male 1121 63 Treatment (3) 2.8117 0.0465 2007 Female 1115 68 Treatment (3) 1.8544 0.1356 2008 Male 1352 69 Treatment (3) 1.0415 0.3798 2008 Female 1351 70 Treatment (3) 0.6264 0.6003 df: Degrees of freedom.

PAGE 77

77 Figure 3 1. Differences in head bobbing rates between treatment groups in 2007 and 2008. Asterisks above the error bars signify whether each coefficient was stat istically significant, compared to the control gro up in the statistical models (* .1; **: P <0.05; ***: P < 0.01). A) Mean ( standard error) number of head bobs per observation session (scan sampling) in each treatment group in 2007. B) Mean ( stan dard error) number of head bobs per male per observation session (focal sampling) in each treatment group, and for homosexually paired males in 2008.

PAGE 78

78 Figure 3 2. Differences in pair bowing rates between treatment groups in 2007 and 2008 Asterisks a bove the error bars signify whether each coefficient was statistically significant, compared to the control group in the statistical models (*** : P <0.01). A) Mean ( standard error) number of pair s bows per observation session (scan sampling) in each tre atment group in 2007. B) Mean ( standard error) number of pair bows per pair per observation session (focal sampling) in each treatment group, and for homosexually paired males in 2008.

PAGE 79

79 Figure 3 3. Differences in aggression rates between treatment g roups in 2007 and 2008. Asterisks above the error bars signify whether each coefficient was statistically significant, compared to the control gro up in the statistical models (*: P 0. 1; **: P <0.05 ). A) Mean ( standard error) number of aggressive acts per observation session (scan sampling) in each treatment group in 2007. B) Mean ( standard error) number of aggressive acts per male per observation session (focal sampling) in ea ch treatment group, and for homosexually paired males in 2008.

PAGE 80

80 Figure 3 4. Mean ( standard error) number of approaches from females to a displaying male per observation session in each treatment group, and for homosexually paired males in 2008. Aster isks above the error bars signify whether each coefficient was statistically significant, compared to the control group in the statistical mod el (**: P <0.05).

PAGE 81

81 Figure 3 5. Mean ( standard error) number of approaches from males to a displaying male per observation session for heterosexual and homosexual males in 2008. Asterisks above the error bars signify that the coefficient for homosexual males was statistically significant, compared to heterosexual males in the statistical mod el (***: P <0.01).

PAGE 82

82 Figure 3 6. Nest attentiveness in male and female white ibises during incubation. Figures show proportion of time spent at nest by each sex for 2007 and 2008. Asterisks above the error bars signify whether each coefficient was statistically significan t, compared to the control group for the relevant sex in the statistical mod error) of time spent at nest by each sex in 2007. B) Mean proportion ( standard error) of time spent at nest by each sex in 2008.

PAGE 83

83 CHAPTER 4 EFFECTS OF CHRONIC M ETHYLMERCURY EXPOSUR E ON SEX STEROIDS OF WHITE IBISES DURING THE BREEDING SEASON Introduction Mercury (Hg) a globally distributed pollutant, is known to have adverse effects on the health of many wildlife sp ecies (Clarkson 1993, Thompson 1996, Wolfe et al. 1998, Crump and Trudeau 2009, Scheuhammer et al. 2009, Tan et al. 2009) Methylmercury (MeHg), the more potent and bioaccumulative form of Hg is formed by sulphate methylation in aqu atic ecosystems (Gilmour et al. 1992, Cleckner et al. 1999) Wetlands are especially good environments for methylating Hg and biomagnificat ion factors exceeding 10 6 water concentrations are the norm in fish in upper trophic levels in contaminated waters (Z illioux et al. 1993) Thus, higher trophic level organisms such as piscivorous fish and birds are at higher risk for MeHg exposure in these ecosystems (Scheuhammer et al. 2009) While being dose and species dependent, effects of MeHg include neurotoxicity, embryotoxicity, immunosuppression, depressed appetite, decreased survival, and v arious detrimental effects on reproduction across taxonomic groups (Heinz 1979, Barr 1986, Spalding et al. 1994, Thompson 1996, Nocera and Taylor 1998, Wolfe et al. 1998, Bouton et al. 1999, Sepulveda et al. 1999b, S palding et al. 2000a, Spalding et al. 2000b, Heinz and Hoffman 2003a, Sandheinrich and Miller 2006, Brasso and Cristol 2008, Burgess and Meyer 2008, Evers et al. 2008, Crump and Trudeau 2009, Heinz et al. 2009, Scheuhammer et al. 2009, Weis 2009) Though less well documented than its other effects, Hg is also known to have effects on hormones (Drevnick and Sandheinrich 2003, Franceschini et al. 2009, Tan et al. 2009, Wada et al. 2009b) Endocrine disrupting compounds (EDCs) can act via multiple mechanisms, such as interfering in synthesis, transport, metabolism and

PAGE 84

84 binding of hormones to their receptors (Guillette and Gunderson 2001) Three main hormone systems: the hypothalamic pituitary adrenal axis, the hypothalamic pituitary gonadal (HPG) axis an d the hypothalamic pituitary thyroid axis can all be affected by Hg exposure (Tan et al. 2009) Since the HPG axis is overwhelmingly important in regulating reproductive behavior and physiology (Nelson 2005) it foll ows that disruption of the HPG axis by Hg could adversely affect reproduction (Ottinger et al. 2009) Several studies have documented adverse effects on various aspects of reproduction due to Hg exposure though not all of thes e have been linked to hormone disruption. In fathead minnows ( Pimephales promales ), MeHg at concentrations of 0.87 and 3.93 ppm dry weight (dw) in the diet affected both sex hormones and reproductive success (Drevnick and Sandheinrich 2003) In this study male s had lower testosterone and females had lower estradiol in comparison to controls. In female minnows decreased estradiol concentrations were associated with ovarian cell apoptosis suggesting this as a mechanism of suppressing sex hormones (Drevnick et al. 2006) In rats, MeHg injection (10 mg/kg subcutaneously for eight days) resulted in decreased spermatogenesis and apoptosis of germ cells (Homma Takeda et al. 2001) In the latter study, exposure levels were high enough to result in acut e MeHg induced neuropathy, but it also seems likely that cellular death in gonadal tissue is a general mechanism by which reproductive hormones and success can be affected across taxa. MeHg may also affect steroidogenic enzymes (Tan et al. 2009) In rats on various protein and lipid diets, MeHg at 3 mg/kg body weight per day, reduced activity of two steroidogenic enzymes, namely C17,20 lyase and 17 hydroxylase (McVey et al. 2008) However, that study noted that some diets seemed to have a protective effect in maintaining activity of

PAGE 85

85 steroidogenic enzymes. Reduced chole sterol (the precursor of steroid hormones) was observed in ovarian tissue of catfish ( Clarias batrachus ) after exposure to both organic and inorganic mercurials (Kirubagaran and Joy 1995) Vitellogenin gene expression in livers of female fathead minnows was significantly suppressed by dietary MeHg at 0.87 and 3.93 ppm dw (Klaper et al. 2006) Vitellogenin is a yolk precursor in all oviparous animals and a marker for endocrine disrupting chemicals since it can be induced by estrogen mimics in males (Klaper et al. 2006) White ibises ( Eudocimus albus ) in the Florida Everglades showed an association with increased MeHg levels and altered sex steroids (Heath and Fre derick 2005) Pre breeding f emale ibises showed a significant negative correlation between mercury exposure and estradiol. Incubating m ale ibises showed a significant positive correlation between mercury exposure and testosterone levels. The same study showed a negative correlation between inter annual differences in Hg exposure and numbers of white ibises nesting. The authors suggested that MeHg exposure caused decreased nesting and/or increased nest abandonment in these birds with endocrine disruption being a possible mechanism. Several studies showed effects of MeHg on aspects of reproduction known to be mediated via endocrine pathways, though endocrine disruption was not specifically examined as a mechanism. In a study of wild common loons ( Gavia immer ), females with highest risk of MeHg exposure (>3 ppm in blood) had 40% reductio n in production of fledglings, and incubation behavior was also affected, with birds at higher exposure levels leaving nests unattended for longer periods of time (Evers et al. 2008) Increased abandonment of nesting territories and poor incubation success associated with higher mercury exposure was also noted in another study of common loons (Nocera and

PAGE 86

86 Taylor 1998) In a study of mallard ducks ( Anas platyrhynchos ) fed MeHg at 0.5 ppm dw (equivalent to 0.1 ppm w et w eight ) over three generations, it was observed that a greater percentage of dosed females laid eggs outside the nest (Heinz 1979) In a fie ld study on common loons, decreased numbers of nest initiations and increased nest abandonment was reported in areas with higher Hg exposure (Barr 1986) however, the positive correlation between high Hg lake s and low lake productivity was a confounding factor in this study. Nesting, territoriality and incubation are all behaviors mediated by reproductive hormones (Nelson 2005) therefore it seems very possible that the above effects were associated wit h changes in hormone profiles. It is possible that many endocrine effects of MeHg on wildlife are more prevalent than reported, as many of the reported effects may be mediated by the endocrine system, and few studies have examined endocrine responses. The combined inf ormation from laboratory studies showing MeHg induced endocrine impairment and from field based behavioral studies showing MeHg associated reproductive impairment led me to form the hypothesis that chronic sublethal MeHg exposure can cause reproductive imp airment via endocrine disruption in Ciconiiforms (wading birds). Due to the known associations between sex steroids and reproduction (Young et al. 1964, Nelson 2005, Adkins Regan 2007, Fusani 2008, O'Neal et al. 2008) and the ability of MeHg to act on the HPG axis (Tan et al. 2009) I predicted changes in estradiol and testosterone with chronic sublethal MeHg exposure. With the information from the Heath and Frederick (2005) study, I specifically predicted that estradiol levels in females would be lowered while testosterone levels would be elevated in males, but I did not have s ufficient information to predict other associations

PAGE 87

87 (e.g. estradiol in males, testosterone in females, interactions with stage of breeding). In all cases I predicted dose dependent effects on hormone levels, with more changes with higher exposure. Furthe rmore, I predicted behavioral changes associated with alterations in hormone profiles: lower estradiol levels to be associated with decreased motivation for courtship and nest initiation; and increased testosterone to be associated with higher rates of agg ression and poor parental behavior. Field studies of MeHg effects are constrained to some extent in that they provide only correlative evidence, and cannot establish lowest observed adverse effect levels (LOAELs). On the other hand, many captive studies h ave been performed in laboratory species at doses higher than what wildlife are exposed to in the field (e.g. Homma Takeda et al. 2001, McVey et al. 2008) Thus, a combin ation of field based ecological understanding, and controlled laboratory studies at environmentally relevant exposure levels are needed to test whether a causal relationship between low dose chr onic MeHg toxicity and reproductive hormones exists and to mea sure the magnitude of the e ffect of MeHg on reproductive success. Controlled studies on MeHg associated hormone disruption and associated effects on reproductive success and behavior in birds are especially lacking. I experimentally tested my hypotheses, using the white ibis ( Eudocimus albus ) as a representative wading bird species, since it is chronically exposed to sublethal MeHg in t he Florida Everglades (Heath and Frederick 2005) Further, there is evidence suggesting that decreased reproductive success and endocrine disruption in the white ibis is linked to Hg exposure in the Everglades (Heath and Frederick 2005) In order to determine causal relationships between Hg exposure at Everglades levels and

PAGE 88

88 reproductive hormone profile s in white ibises, I exposed captive ibises to dietary MeHg ranging from 0.05 0.3 ppm ww (range found in prey of white ibis in the Everglades; Loftus 2000) for 3.5 years a nd measured fecal metabolite concentrations of estradiol and testosterone metabolites and reproductive success over two consecutive breeding seasons. Methods Dietary Methylmercury Exposure of Captive White Ibises W hite ibis nestlings were collected from co lonies in the Florida Everglades and raised to adulthood in captivity. They received dietary MeHg that spanned the range of concentrations found in their food items in the Everglades during peak contamination in the 1990s (Loftus 2000) Nest lings were randomly assigned to one of four groups of 20 males and 20 females in each, receiving 0.00 (control), 0.05(low), 0.1 (medium) or 0.3 (high) ppm MeHg wet weight (ww) in pelletiz ed food. Birds were continuously exposed to these diets from 90 days of age through out the duration of the experiment (2005 to 2008) The birds were housed together in a 1200 m 2 circular free flight aviary divided into four quadrants by net walls, with perches, a wading pond and feeding area located in identical configur ations in each quadrant. Cage locations were switched each year prior to the breeding season, so that each breeding season was spent in a different location thus controlling for location effects. All birds were individually identifiable by leg bands, and were genetically sexed (Avian Biotech International, Tallahassee, Florida) Behavioral Sampling d uring the Breeding Season Reproductive behavior and su ccess was observed in February to August of 2006, 2007 and 2008 In depth behavioral sampling results a re reported in Chapter 3.

PAGE 89

89 Nesting was encouraged by providing nesting structures (cup shaped platforms of plastic netting) in the perches (total of 48 platforms, eight per perch). I provided ad libitum nesting material in the form of twigs and fresh catt ail ( Typha sp. ), replenished weekly throughout the breeding season (February to August). Identities of birds displaying, nest building, laying, incubating or chick rearing were recorded daily at sunrise throughout the breeding season. Each nest platform was inspected for presence/absence of a nest and if present, the status of the nest (i.e. few sticks/ partial nest/ full nest). The number of eggs and/or chicks and the pair associated with the nest were recorded. Eggs were recorded each day they were la id, and kept track of until hatching. Chicks were individually marked and kept track of until fledging or death. Thus, the fate of each nest with regards to parents/eggs/chicks was individually known on a daily basis. Observers were kept blind to treatm ent groups in order to keep out biases. Collection of Fecal Samples I measured fecal concentrations of estradiol and testosterone metabolites from individually identified fecal samples. While plasma hormone levels have been traditionally used and validat ed in animals (e.g. Deviche et al. 1980, Runfeldt and Wingfield 1980, Ludders et al. 1998), t his option was not viable in my experiment. The repeated sampling of individual birds at sho rt time intervals needed for my purposes would have entailed too much disturbance to the flock (in terms of catching and handling birds) and potentially affect ed breeding behavior Measurement of steroid hormone metabolites in fecal samples is a non invasive and accepted method (e.g. Wasser et al. 1997, Was ser et al. 2000, Palme 2005).

PAGE 90

90 Fecal samples needed to be individually identified together with breeding stage of the bird. To identify feces from individual birds, I fed birds with baits (small fishes typically used as feed supplements), that were stuffed with glass beads of different colors (sizes 18/0 or 15/0). A pilot study demonstrated that the beads did pass along the gut with other food material in 2 3 hours time, and the feces from known individuals contained the color combinati ons that had been fed to them. I fe d baits containing known color combinations of beads to individually identifiable birds, usually by throwing the baits to target birds using binoculars to c onfirm that baits were eaten. I also plac ed fish in the nests of tightly incubating birds and obse rved the birds to be sure they ate the baits. B aits were fed between 09:00 to 11:00 hrs, and samples collected 3 4 hr after feeding to control for diurnal variability of hormone levels. Since fecal hormone metabolites represent circulating levels as int egrated over the gut passage time, they are not as sensitive to short term hormone fluctuation as plasma samples (Palme 2005) During the chick rearing stage, nestlings and fledglings sometimes ingested beads via food fed to them by parents. To avoid the resultin g fecal identification error, I collected samples during chick rearing by approaching individual adult birds on the nest or on a perch, and directly observed where the ir feces fell as they flew off. I collected fecal material only from areas of the floor that had been recently washed, or from 0.5 2 m 2 framed panel s of clean plastic sheeting placed beneath perches Fecal splays that appeared to have more than one contributor were discarded and only fresh, un dried fecal samples were collected. Samples were collected using clean spatulas into 2ml polypropylene cryo tubes (Fisher Scientific), placed on ice immediately after collection,

PAGE 91

91 stored temporarily for up to eight hours in a 4 Celsius freezer, and transferred to a 20 Celsius freezer for longer term storage. Fecal samples were collected several times a week thr oughout the breeding season in 2007 and 2008 The breeding season was divided into six stages: pre breeding (only in 2007), display, nest building, egg laying, incubation and chick rearing. Since male male pairs ( chapter 2) had no eggs to incubate or young to raise, the period following completion of nest building was defined as the incubation stage. During this time, males were usually observed sitting on the nests, and seemed behaviorally similar to incubating heterosexual birds. Radioimmunoassays for St eroid Hormone Metabolites in Fecal Extracts Fecal samples were freeze dried, homogenized, and 0.05g of each sample used for extraction (by 80% ethanol) of steroid hormone metabolites (Adams et al. 2009) Fecal extracts were stored in a 20 Celsius freezer until analysis for steroid hormones by radioimmunoassay. Each sample extract was assayed for est radiol and testosterone metabolites. I followed radi oimmunoassay methods previously validated for these hormones in w hite i bis fecal extracts (Adams et al. 2009) Estradiol 125 I Coat a Count RIA kits (Diagnostic Products, Los Angeles, CA, USA) were used for measuring estradiol metabolites. Testosterone 125 I double antibody RIA kits (MP Biomedicals Solon, OH, USA) were used to measure testosterone metabolites. Samples were analyzed in duplicate, and those with a coefficient of variation more than 15% were reanalyzed (if there was sufficient sample for reanalysis) or discarded. All intra assay coe fficients of variation were <15% for both hormones. The inter assay coefficient of variation for estradiol was

PAGE 92

92 16.9% and 8.9% for 2007 and 2008 samples respectively, while for testosterone the respective numbers were 11.7% and 12.9%. Statistical Analyses of Hormone Data Hormone concentrations ( testosterone as nan ograms per gram feces dw and estradiol as picograms per gram feces dw) were transformed to their natural logarithms in order to norm alize the data. I used package nlme (Pinheiro et al. 2009) in R version 2.10.0 (R Development Core Team 2009) f or statistical analyses. Each hormone was analyzed separately by sex and by year since the structure of the data was too complex ost sets of data had a heterogeneous of mating behavior ( for males: heterosexual/ homosexual) which was accounted for by adding a varia nce function to the models (Pinheiro and Bates 2000) I accounte d for correlation between repeated samples of individuals within a breeding season by using extended linear models with a correlation structure (Pinheiro and Bates 2000) Models were estimated using generalized least squares methods with maximum likelihood (Pinheiro et al. 2009) I used an information theoretic approach and variance function (Burnham and Anderson 20 02) While an auto regressive moving average or compound symmetry correlation structure was sufficient for most models, a more complex general correlation structure was needed for a few models. All predictor variables determined a priori (breeding stage, treatment group, type of mating behavior for males) and all possible combinations of interactions were used in the full models to test my hypotheses. For computational reasons, I tested two models for males in each year; one included data from all breedi ng stages and treatments (full

PAGE 93

93 model), while the other comprised data only from dosed males in display, nest building and incubation stages (pairing type model; breeding stages relevant to both pairing types). There were no homosexual males in the control group in 2008 and all breeding stages were not common between the two mating types (no stages of egg laying and chic k rearing for homosexual males). T he modeling software did not support interaction terms of treatment*mating type and breeding stage*matin g type interactions (and three way terms with these terms) where there were missing data in some factor combinations. Therefore, I restricted data to treatments and breeding stages which had data for both homosexual and heterosexual males in the pairing t ype model to test the a bove interactions. In all models n on significant interaction term s (P<0.05; c onditional F test) were removed I report results of conditional F tests (F tests) which test the significance of explanatory variables that could include several coefficients (e.g. the explanatory variable for treatment effects has coefficients for each dose group) and of conditional t tests (t tests) which test the marginal significance of each separate coefficient when all other coefficients are present in the model (Pinheiro and Bates 2 000) Residual plots and quantile quantile plots of the models were examined to see whether data conformed to the assumption of normality. I tested one a posteriori model for testosterone concentrations in female ibises contrasting only high dose group females with the control females. The a priori model comprising the full data set had effects mainly in the high dose group and the a posteriori model was tested to get a clearer picture of the effects.

PAGE 94

94 Results Reproductive Behavior The most prominent ef fect of MeHg exposure on breeding was on pairing and nesting behavior of male ibises. In all three years that breeding was monitored, male ibises courted, made pair bonds and nested as male male pairs to a greater degree in all dosed groups than in the un dosed control (chapter 2). Due to the strong probability that pairing behavior was influenced by endocrine expression, the hormone analyses took into account the type of pairing behavior (male male or male female) exhibited by male ibises. Other behavio ral changes during courtship included lowered rates of head bobbing and pair bowing in dosed males; lower number of aggressive acts in high dose and all homosexual males; lower number of female approaches to high dose and all homosexual males; and higher n umber of male approaches to homosexual males (chapter 3). Notable changes in reproductive success (chapter 2) were significantly reduced nestling production in both sexes in 2007, in all dosed males in 2008 and high dose females in 2008; marginally reduced fledging success of high dose females in 2008; marginally reduced total fledged and total number of successful attempts per high dose female over all breeding seasons (35% reduction in comparison with control females for the two last metrics). Effects of Methylmercury on Estradiol in Adult Females Breeding stage was strongly significant as a predictor of estradiol concentrations in both nesting attempts of 2007 (first: F 5, 272 = 9.2, P <0.0001; second: F 5, 171 = 23.9, P <0.0001; Table 4 1). Treatment (MeH g) effects on estradiol concentrations of females were only seen in the first attempt of 2007 (F 3 272 = 4.07, P = 0.0075). During the first

PAGE 95

95 attempt the medium group had lower estradiol than control females (t = 2.8, P = 0.006; Fig. 4 1), but there were no significant differences in the low (t = 0.7, P = 0.47) or high (t = 1.5, P = 0.25) dose groups. Though the F test for treatment effects in the second attempt was non significant (F 3, 171 = 1.3, P = 0.27), medium (M) and high (H) dose females showed a trend towards lower estradiol concentrations (M: t = 1.9, P = 0.06; H: t = 1.6, P = 0.12) with no significant differences in low (L) dose females (t = 1.6; P = 0.29). Actual concentrations of fecal estradiol metabolites are reported in Appendix B. Whil e breeding stage was again a significant predictor of estradiol in the first attempt of 2008 (F 4, 293 = 41.0, P <0.0001; Table 4 1); overall treatment effects were non significant (F 3, 293 = 1.66, P = 0.18; Fig. 4 1). However, the low dose group had reduc ed estradiol compared to the controls (t = 2.06, P = 0.04) and the medium group showed a similar trend (t = 1.62, P = 0.11). The high dose group did not show any significant difference from control (t = 0.73, P = 0.47). Effects of Methylmercury on Estr adiol in Adult Males Males showed a more complex pattern of estradiol response to dosing than females did. In the first nesting attempt of 2007, there were significant effects of breeding stage on estradiol (F 5, 245 = 6.75, P <0.0001) as well as a signifi cant interactive effect of stage and treatment on estradiol (F 1 5, 245 = 1.76, P = 0.040; Table 4 2). There were no significant main effects of either treatment or pairing type (F 3, 245 = 0.68, P = 0.57 and F 1, 245 = 1.13, P = 0.29 respectively). In the second nesting attempt of 2007, breeding stage had only a marginal effect on male estradiol concentrations (F 5, 183 = 2.24, P = 0.053; Table 4 2), while there was a significant interaction between treatment and breeding stage (F 1 5, 183 = 1.76, P = 0.044).

PAGE 96

96 Pairing type did not have a significant effect on estradiol (F 1, 183 = 0.04, P = 0.84) and main effects of treatment were only marginally significant (F 3, 183 = 1.88, P = 0.13). In the first attempt of 2008 the full model showed significant effects of bre eding stage on male estradiol concentrations (F 4, 317 = 73.3, P <0.0001; Table 4 2) and a significant stage*treatment interaction (F 12, 317 = 4.2, P <0.0001). There were no significant main effects of treatment and pair type (F 3, 317 = 0.7, P = 0.54 and F 1, 317 = 0.1, P = 0.79 respectively). Some of the main changes in individual coefficients comparing each dosed group with the control males are summarized herewith (Table 4 3). The only time pre breeding estradiol concentrations differed from controls was for low dose males in the first attempt of 2007, when they had significantly lower concentrations (Fig. 4 2). During display, low and high dose groups showed negative trends in estradiol in the second attempt of 2007, but during 2008 (first attempt) both these groups had significantly elevated estradiol in comparison with controls (Fig. 4 3). Though not always consistent across treatment groups, in all three nesting attempts across years, changes during nest building showed lower estradiol concentrations compared to control males (Fig. 4 4). During the second breeding attempt of 2007, all groups had significantly decreased estradiol during nest building. Changes during laying and incubation were mainly shown in high dose males. High dose males had decr eased estradiol compared to control males during egg laying in the second attempt of 2007 (Fig. 4 5) and also during incubation in both 2007 (second attempt) and 2008 (first attempt; Fig. 4 6). During chick rearing, low dose males had significantly elevat ed concentrations in the first

PAGE 97

97 breeding attempt of 2007, and high dose males showed a similar non significant trend in the first attempt of 2008 (Fig. 4 7). The pairing type model for 2007 showed no main effects of either treatment or pair type (F 2, 114 = 0.32, P = 0.73 and F 1, 114 = 0.61, P = 0.44 respectively; Table 4 2) and there was only a marginally significant effect of breeding stage (F 2, 114 = 3.02, P = 0.053). The pairing type model for 2008 showed significant effects of breeding stage (F 2, 162 = 51.99, P < 0.0001; Table 4 2), pair type (F 1, 162 = 9.3, P = 0.0027), stage*treatment (F 4, 162 = 3.15, P = 0.016), treatment*pair type (F 2, 162 = 9.91, P = 0.0001), and stage*pair type (F 2, 162 = 15.57, P < 0.0001) but not a main effect of treatment (F 2, 162 = 0.13, P = 0.88). The comparisons of interest in this model were between homosexual and heterosexual males within groups. During display, medium and high dose group homosexual males had significantly elevated estradiol in comparison to heterosexual mal es of each group (M: t = 2.13, P = 0.035; H: t = 3.05, P = 0.0026; Fig. 4 8) while low dose homosexual males showed marginally lower concentrations (t = 1.57, P = 0.12). During nest building, homosexual males had lower concentrations than heterosexual ma les within a group (L: t = 5.16, P < 0.0001; M: t = 2.56, P = 0.012; H: t = 1.85, P = 0.066). During incubation, only low dose homosexual males showed significant differences from their heterosexual counterparts during incubation, with depressed estradio l concentrat ions (L: t = 6.76, P < 0.0001; M: t = 0.58, P = 0.56; H: t = 0.44, P = 0.66). Effects of Methylmercury on Testosterone in Adult Females In the first breeding attempt of 2007, I found significant effects of breeding stage (F 5, 273 = 6, P <0.000 1) and marginally significant effects of treatment (F 3, 273 = 2.15, P = 0.094) on female testosterone (Table 4 4). Medium and high dose group females had

PAGE 98

98 suppressed concentrations of testosterone compared to controls in all breeding stages but there were no effects in the low dose group (L: t = 1.09, P = 0.28; M: t = 2.06, P = 0.041; H: t = 2.33, P = 0.021; Fig. 4 9). In the second breeding attempt, there were significant effects of breeding stage (F 5, 172 = 2.5, P = 0.03; Table 4 4), but none of treat ment (F 3, 172 = 1.31, P = 0.27). However, treatment coefficients for the low and medium dose groups showed a negative trend in testosterone concentrations compared to controls (L: t = 1.61, P = 0.11; M: t = 1.7, P = 0.091; Fig. 4 9), but showed no effec ts in the high dose group (t = 0.81, P = 0.42). Actual concentrations of fecal testosterone metabolites are reported in Appendix B. In 2008 (first attempt), there were significant effects of breeding stage (F 4, 266 = 25.08, P <0.0001; Table 4 4), a margi nally significant interaction between breeding stage and treatment (F 12, 266 = 1.43, P = 0.15), and a non significant main effect of treatment (F 3, 266 = 1.45, P = 0.23). During the display stage, medium dose females showed a significantly lower concentra tion of testosterone, with a similar trend in the high dose group, but no effect in the low dose group (L: t = 0.69, P = 0.49; M: t = 2.32, P = 0.021; H: t = 1.79, P = 0.075; Fig. 4 10). The low and medium dose groups did not show significant difference s from control females in any other stage (t tests, P >0.1). High dose group females, however, showed positive trends in testosterone during laying (t = 1.51, P = 0.13), incubation (t = 1.44, P = 0.15) and chick rearing (t = 1.86, P = 0.051). The only st age where high dose females did not show effects was during nest building (t = 0.15, P = 0.88). The model contrasting just the high dose group with the control group showed significant effects of breeding stage (F 4, 132 = 4.04, P = 0.004; Table 4 4), tre atment (F 1, 132 = 4.89, P = 0.029), and stage*treatment (F 4, 132 = 2.46, P =

PAGE 99

99 0.049). The coefficients for testosterone in high dose females showed a negative trend in during display (t = 1.92, P = 0.058), and significantly higher concentrations during eg g laying (t = 2.68, P = 0.0082), incubation (t = 2.13, P = 0.035), and chick rearing (t = 2.81, P = 0.0058), but no effects during nest building (t = 1.32, P = 19). Effects of Methylmercury on Testosterone in Adult Males I found effects of breeding stage o n male testosterone concentrations only in the first nesting attempt of 2007 (F 5, 260 = 3.19, P = 0.0082; Table 4 5), with no effect of either treatment (F 3, 260 = 0.81, P = 0.49) or pairing type (F 1, 260 = 0.58, P = 0.45). In the second attempt, there we re significant effects of breeding stage (F 5, 197 = 5.8, P = 0.0001; Table 4 5), and treatment (F 3, 197 = 4.8, P = 0.003), but no effects of pairing type (F 1, 197 = 1.17, P = 0.28). There were significant decreases of testosterone concentrations in medium and high dose group males, with a similar trend in low dose males, in comparison with the control group in the second attempt (L: t = 1.43, P = 0.15; M: t = 2.84, P = 0.005; H: t = 3.59, P = 0.0004; Fig. 4 11). In the first attempt of 2008, there were significant effects of breeding stage (F 4, 302 = 34.41, P <0.0001; Table 4 5), pairing type (F 1, 302 = 17.34, P <0.0001), and stage*treatment interactions (F 12, 302 = 2.06, P = 0.019); and a marginally significant effect of treatment (F 3, 302 = 2.09, P = 0 .10). During display, high dose males showed a negative trend in testosterone concentrations compared to the controls, however, there were no effects in other dosed groups (L: t = 0.39, P = 0.69; M: t = 0.52, P = 0.60; H: t = 1.49, P = 0.14; Fig. 4 12). Only the medium group showed changes during nest building with significantly depressed testosterone concentrations (L: t = 0.14, P = 0.89; M: t = 1.99, P = 0.047; H: t = 1.03, P = 0.30). In the laying stage, low dose males had elevated testosterone, a similar trend in the medium dose group, and no effects in high

PAGE 100

100 dose males (L: t = 3.15, P = 0.0017; M: t = 1.83, P = 0.068; H: t = 0.59, P = 0.56). There was a positive trend in testosterone concentrations in low dose males during incubation, with no ef fects in the other groups (L: t = 1.60, P = 0.11; M: t = 0.69; P = 0.48; H: t = 1.03, P = 0.30). During chick rearing, only the high dose group showed a positive trend (L: t = 0.40, P = 0.69; M: t = 0.27, P = 0.79; H: t = 1.53, P = 0.13). The pairing ty pe model for 2007 only showed effects of breeding stage (F 2, 114 = 3.4, P = 0.037; Table 4 5), and no effects of treatment (F 2, 114 = 1.09, P = 0.34), or pairing type (F 2, 114 = 0.19, P = 0.67). The pairing type model for 2008 showed significant effects o f breeding stage (F 2, 158 = 25.26, P <0.0001; Table 4 5), pair type (F 1, 158 = 16.68, P = 0.0001), and a significant interaction between breeding stage and pair type (F 2, 158 = 11.16, P <0.0001). Main effects of treatment were non significant (F 2, 158 = 0 .63, P = 0.54); thus all homosexual males showed a similar direction of change from heterosexual males, irrespective of treatment group. During display, there were no changes in testosterone concentrations of homosexual males compared to heterosexual male s (t = 1.03, P = 0.30; Fig. 4 13); during nest building, homosexual males had significantly lower testosterone (t = 2.45, P = 0.015), and during incubation, they had significantly elevated concentrations in comparison to heterosexual males (t = 5.73, P <0 .0001). Discussion Effects of Methylmercury on Sex Steroids MeHg exposure altered fecal concentrations of estradiol and testosterone in both sexes during breeding; the first experimental study detecting evidence of endocrine disruption at such low levels o f exposure (0.05 to 0.3 ppm ww) of MeHg in birds. This finding is corroborated by field evidence where white ibises in the Everglades had

PAGE 101

101 changes in these sex steroids correlated with MeHg exposure at similar ranges (Heath and Frederick 2005) In that study, feather Hg levels ranged from 0.33 to 20 ppm, while in my study, mean feather Hg levels over all three years wer e 0.61, 6.53, 15.79 and 36.92 ppm in control, low, medium and high groups respectively. Since I compared endocrine response to MeHg in the same individuals over all breeding stages and multiple breeding seasons; I was able to control for, and find interac tions with these factors. The changes I documented were not always consistent across sexes, nesting attempts (in 2007), years, breeding stages or treatment groups. However, there were some broadly discernible patterns as follows. All changes of estradio l concentrations in dosed females, though not related linearly with dose group, were negative in comparison to control females as predicted, and in agreement with the results of the Heath and Frederick (2005) field study on MeHg effects in white ibises. I did not, however, see any decreased numbers of nest initiations associated with depressed estradiol as predicted. Males showed interactions between estradiol concentrations and breeding stage. A prominent change was increased estradiol in dosed males during display, with this change also reflected at a higher magnitude in homosexual dosed males. Dosed males also showed a general pattern of decreased estradiol during nest building in comparison to controls. Of the dosed groups, high dose males showed more changes in estradiol concentrations, and this was accentuated in 2008 when there were changes duri ng display, laying, incubation and chick rearing in the high dose group when compared with controls. Changes in estradiol concentrations of homosexual males during nest building and incubation too, were in the same direction as for

PAGE 102

102 heterosexual males when compared within dose groups, with changes in homosexual males being of greater magnitude. As with estradiol, female ibises generally showed decreased testosterone concentrations compared to control females though again with non linear treatment effects an d some inconsistencies among years. One prominent deviation from this pattern was in high dose females in 2008, when they showed increased testosterone during laying, incubation and chick rearing, which was also associated with lower reproductive success in this year. In males, testosterone concentrations were decreased in relation to controls during all breeding stages in 2007 and in pre laying stages in 2008. In 2008, there were elevated concentrations in dosed birds during laying as predicted and agai n corresponding to MeHg associated changes seen in ibises in the Everglades (Heath and Frederick 2005) High dose ma les showed elevated testosterone during incubation and chick rearing stages in 2008, associated with lower reproductive success. Homosexual males showed decreased testosterone in comparison to heterosexual males, irrespective of treatment group, during ne st building and increased concentrations during incubation. These changes were similar in direction and larger in magnitude to changes of dosed males to control males in the same stages. The sex related differences that I found in the patterns of MeHg ind uced endocrine disruption were not unexpected, since susceptibility to MeHg as well as pathways for metabolism and excretion differ between sexes in birds (Lewis et al. 1993, Monteiro and Furness 2001, Tan et al. 2009) Further, the white ibis has been found to be sexually

PAGE 103

103 dimorphic with regard to sex steroid levels in the breeding season (Heath et al. 2003) which could lead to differ ential pathways being affected by MeHg. Homosexual Pairing and Endocrine Disruption of Sex Steroids in Males In chapter 2, I report that MeHg exposure altered sexual partner preferences of male ibises. In the present chapter I also document differences be tween sex hormone profiles of homosexually and heterosexually paired males although I do not know the relationship between the observed endocrine disruption and sexual preference. This is because I do not know a) whether changes in circulating sex hormone s preceded homosexuality or vice versa; b) whether MeHg acted on neuro hormonal or other hormone independent pathways of sexual preference before sexual maturity in these birds (i.e. induced developmental changes); and c) whether both MeHg induced changes in circulating levels of sex hormones and developmental processes interacted to change sexual preferences. While my study cannot provide causal evidence for altering sexual preferences through the mechanism of endocrine disruption, it is well known that ho rmones play a major role in sexual partner preferences, and the development of preference in animals including birds (Adkins Regan 1988, Adkins Regan et al. 1997, Adkins Regan 1998, Adkins Regan and Wade 2001, Adkins Regan and Leung 2006, Adkins Regan 2007, 2009) Hormonal effects of sex steroids can be either organizational or activational (Adkins Regan 2007) Organizational effects are those that are established during the developmental period of an animal wherein they determine the behavioral phenotype expressed permanently in later life. Activational effects a re those that come into play at sexual maturity and adulthood and regulate behavior such as seasonally occurring breeding cycles of birds or estrous cycles of mammals (Adkins Regan 2007) In order

PAGE 104

104 for activational effects to occur, it is necessary for organizational effects to have been established beforehand. The establishment of sexual partner preferences appears to be an organizational effect in the zebra finch ( Taeniopygia guttata ; Mansukhani et al. 1996, Adkins Regan et al. 1997) Female zeb ra finches that were injected with estradiol in the first two weeks post hatch were subsequently housed in uni sex or mixed sex cages. Birds that were hormonally manipulated and housed in uni sex cages exhibited a pairing preference for females later in l ife indicating both a hormonal and social component in forming sexual partner preferences (Mansukhani et al. 1996) Zebra finches are socially monogamous, colonially breedin g birds similar to white ibises (Frederick 1987b, Zann 1994) ; therefore, it is likely that at least some aspects of organizational effects in est ablishing sexual partner preferences are common in the two species. In my study, MeHg exposure started at 90 days of age, at a time when the birds were still developing into juveniles, but considerably later in life than the period of hormone manipulation in zebra finches. Since sexual maturity occurred during the period of MeHg exposure it is possible that MeHg influenced organizational effects of sexual partner preference. This could have either occurred via endocrine pathways altered by MeHg exposure or by direct action of MeHg on structures that determine sexual partner preference. It has been postulated that these sites exist in the brain (Adkins Regan et al. 1997) and MeHg is a compound capable of crossing the blood brain barrier, and indeed shows an affinity for brain tissue (Wolfe et al. 1998, Tan et al. 2009) However, the ibises in my study did not show strong MeHg effects on fecal hormone concentrations as juveniles (Adams et al. 2009)

PAGE 105

105 which suggests MeHg induced changes, at least on hormone expression, are more prominent in adults. I noted that for the most part, endocrine changes shown by homosexual mal es in comparison to dosed heterosexuals were com parable in direction to the changes all dosed males showed in comparison to control males. This suggests that the same process was at work, though exaggerated in the homosexual birds and that altered pairing behavior was associated with the most extreme end of a continuum of MeHg induced hormone disruption While this reasoning is plausible, it should be emphasized that I did not test any predictions related to this hypothesis and cannot defend any part of i t. Further, similar hormonal correlates are lacking in other avian studies where same sex pairing was observed. In the graylag goose ( Anser anser ), where male male pairing is a common strategy when female availability is low, testosterone levels did not differ between heterosexual and male paired birds (Hirschenhauser et al. 2000) Female female pairing in Western gulls ( Larus occidentalis ) was not associated with changes in testosterone between heterosexu al and homosexual birds (Wingfield et al. 198 0) Although findings from the above studies indicate that same sex pairing can occur in absence of changes in circulating hormone profiles, it does not rule out hormonal pathways in my study as both these occurrences of same sex pairing resulted from bi ased sex ratios (Conover and Hunt 1984, Huber and Martys 1993) and were not analogous to my study. In the fruit fly ( Drosophila melanogaster ), males with increased levels of the neuromodulator dopamine, show ed enhanced male male courtship, but did not show changes in their courtship towards females (Liu et al. 2008) The authors suggest that

PAGE 106

106 altered sensory perception of other males increased male male courtship behavior in this species. I do not know w hether MeHg affected sensory pathways or dopamine levels in my study, or whether these pathways influence sexual preference in the white ibis. However, chronic MeHg exposure at levels comparable or lower than in my study led to visual and auditory impairm ent in monkeys ( Macaca fascicularis ; 50 g/kg/day leading to 0.6 0.9 ppm in blood; in my study average blood Hg levels were 0.73, 1.6 and 3.95 ppm in low, medium and high groups respectively ) indicating that MeHg is capable of affecting sensory pathways (Rice and Gilbert 1982, 1992) I also observed depressed rates of courtship display (head bobbing, pair bowing) and aggressive acts in dosed (and homosexual) males, as reported in chapter 3. As discussed therein, while I did not test for mechanistic pathways, hormonal control is one possibility. During courtship, testosterone concentrations of dosed males were generally depressed compared to control males, though not in all treatment/year combinations. Although hormonal control of sexual and agnostic behavio r is far from clear, androgens are generally believed to be a contributing factor (Crews and Moore 1986, Owens and Short 1995, Wingfield et al. 2001, Fusani 2008) Therefore, it is plausible that differences in testosterone levels resulting from MeHg exposure played a part in altering courtship behavior. It is also possible that altered courtship behavior made homosexual males unattractive to females (chapter 3), and male male pairing was therefore a result of female mate choice. Reproductive Success and Endocrine D isruption in Females In females, reduced reproductive success was most marked in the high dose group in 2008 (chapter 2). In 2008, 35% of high dose females failed to produce any nestlings, and over all three breeding seasons, they had 35% fewer average nu mbers of

PAGE 107

107 fledglings and numbers of successful breeding attempts (chapter 2). These reproductive failures were associated with significantly higher testosterone concentrations in high dose females during laying, incubation and chick rearing in 2008. Exper imentally elevated testosterone levels have been associated with reduced hatching success, fledging success, and nest survival i n passerines (O'Neal et al. 2008, Lopez Rull and Gil 2009) Generally, birds have higher testosterone levels during pre laying stages (Ketterson et al. 2005) and this is so in the white ibis as well (Heath et al. 2003) Therefore, elevated testosterone in post laying females may have altered behavior during these stages, leading to low reproductive success. Dosed male ibises also showed elevated testosterone during egg laying, incubation and chick rearing stages in 2008. The study by Heath and Frederick (2005) too showed a positive correlation between MeHg exposure and male testosterone levels dur ing the laying stage. Direct comparison cannot be done between hormone levels of the above study and mine since the former dealt with plasma levels while I measured concentrations of hormone metabolites in feces. However, the agreement in direction of ef fect between the two, coupled with poor nesting success in the high dose group adds strength to the argument that MeHg exposure contributed to decreased nesting effort of white ibises in the Everglades through endocrine disruption (Heath and Frederick 2005) Dose response Relationships between Methylmercury and Endocrine Disruption The effects of MeHg on sex steroids in ibises did not show a linear dose response relationship except in a few instances (e.g. estradiol in males during incubation 2008; testosterone in males during the second breeding attempt 2007). I sometimes observed endocrine effects at the lowest MeHg dose level (0.05 ppm), with no effects at higher dose levels, contrary to my predictions. However it is known that the dose response

PAGE 108

108 relationship of EDCs can be non monotonic even with monotonically increasing dose resulting in U shaped or inverted U shaped response curves (Welshons et al. 2003, Clotfelter et al. 2004) Another factor which might have affected responses was the long term exposure, since MeHg can accumulate in the hypothalamic pituitary axis, the main regulator for sex steroid pathways (Tan et al. 2009) Indeed, I found more effects in 2008 when birds had been exposed the longest. Though time effects complicates inter pretation and could affect directionality of effects making them more unpredictable, the most important implication was that even very low chronic MeHg exposure can result in endocrine effects, which is directly of importance for wildlife in contaminated e nvironments. Conclusions I observed endocrine effects even at the lowest exposure level of MeHg (0.05 ppm ww), at dose levels not tested in previous experimental studies. Furthermore, these endocrine effects were coupled with behavioral changes that resul ted in non optimal mating strategies leading to loss of fitness even at the lowest levels of MeHg exposure in this study. Endocrine disruption of sex steroids is a plausible mechanism for the altered sexual preference, courtship behavior, aggression and r educed reproductive success observed in dosed birds, though it does not exclude other pathways of action, and/or many interacting mechanisms of action (e.g. organizational effects of hormones, differences in brain architecture, MeHg effects on neural pathw ays of behavior). The effects observed in my study, i.e., altered behavior and sexual preference, endocrine disruption and reduced reproductive success, are all of large enough magnitude to have the ability to affect population level processes. The weig ht of evidence indicates that MeHg exposure was very likely a contributing factor for reduced

PAGE 109

109 nesting effort in ibises during years of peak mercury contamination in the Everglades (Heath and Frederick 2005) Thus, my study underlines the importance of understanding detrimental effects of ecologically re levant low dose MeHg exposure in wild wad ing bird populations

PAGE 110

110 Table 4 1. Model summaries of the best models explaining estradiol concentrations in white ibis females. Year Breeding attempt Degrees of freedom Explanatory variables (df) F test Total Denominator F value P value 2007 Attempt 1 281 272 Breeding stage (5) 9.20 <0.0001 Treatment (3) 4.07 0.0075 2007 Attempt 2 180 171 Breeding stage (5) 23.86 <0.0001 Treatment (3) 1.31 0.2716 2008 Attempt 1 301 293 Breeding stage (4) 41.02 <0.0001 Tre atment (3) 1.66 0.1754

PAGE 111

111 Table 4 2. Model summaries of the best models explaining estradiol concentrations in white ibis males. Year Breeding attempt Degrees of freedom Explanatory variables (df) F test Total Denominator F value P value 20 07 Attempt 1 270 245 Breeding stage (5) 6.75 <0.0001 Treatment (3) 0.68 0.5668 Pairing type (1) 1.13 0.2887 Breeding stage*Treatment (15) 1.76 0.0413 2007 Attempt 2 208 183 Breeding stage (5) 2.24 0.0525 Treatment (3) 1.88 0.13 41 Pairing type (1) 0.04 0.8357 Breeding stage*Treatment (15) 1.76 0.0439 2008 Attempt 1 338 317 Breeding stage (4) 73.30 <0.0001 Treatment (3) 0.70 0.5375 Pairing type (1) 0.10 0.7900 Breeding stage*Treatment (12) 4.20 <0 .0001 2007 Attempt 1 120 114 Breeding stage (2) 3.02 0.0525 Pairing type model Treatment (2) 0.32 0.7284 Pairing type (1) 0.61 0.4382 2008 Attempt 1 176 162 Breeding stage (2) 51.99 <0.0001 Pairing type model Treatment (2) 0.13 0.8776 Pairing type (1) 9.30 0.0027 Breeding stage*Treatment (4) 3.15 0.0159 Treatment* Pairing type (2) 9.91 0.0001 Breeding stage* Pairing type (2) 15.57 <0.0001

PAGE 112

112 Table 4 3. Coefficients for each MeHg dosed group comparing estradiol concentrations in each breeding stage relative to control males. Breeding Year Low Medium High stage attempt t value P value t value P value t value P value Pre breeding 2007 2.35 0.0193 0.58 0.5639 0.43 0.6698 Display attempt 1 0 .38 0.7061 0.44 0.6638 1.27 0.2048 Nest building 2.37 0.0185 1.16 0.2479 0.06 0.9539 Egg laying 1.98 0.0489 1.28 0.2009 0.42 0.6771 Incubation 0.39 0.6955 0.21 0.8342 0.49 0.6243 Chick rearing 2.10 0.0363 0.22 0.8252 1.05 0.2969 Pre breeding 2007 0.87 0.3871 0.36 0.7169 0.39 0.6938 Display attempt 2 1.65 0.1007 0.55 0.5864 1.94 0.0541 Nest building 2.59 0.0105 2.61 0.0097 2.32 0.0213 Egg laying 0.34 0.7374 1.38 0.1679 1.30 0.1958 Incubation 0.88 0.3827 1.28 0.2028 2.61 0.0097 Chick rearing 0.53 0.5976 0.53 0.5981 0.02 0.9827 Display 2008 2.24 0.0260 1.19 0.2367 3.56 0.0004 Nest building attempt 1 0.05 0.9567 1.56 0.1186 0.90 0.3673 Egg laying 0.64 0.5205 1.01 0.3117 3.32 0.0010 Incubation 1. 22 0.2216 1.88 0.0608 2.87 0.0043 Chick rearing 0.47 0.6410 0.29 0.7691 1.87 0.0630 Table 4 4. Model summaries of the best models explaining testosterone concentrations in white ibis females. Year Breeding attempt Degrees of freedom Explanator y variables (df) F test Total Denominator F value P value 2007 Attempt 1 282 273 Breeding stage (5) 6.00 <0.0001 Treatment (3) 2.15 0.0940 2007 Attempt 2 181 172 Breeding stage (5) 2.50 0.0323 Treatment (3) 1.31 0.27 20 2008 Attempt 1 286 266 Breeding stage (4) 25.08 <0.0001 Treatment (3) 1.45 0.2283 Breeding stage*Treatment (12) 1.43 0.1502 2008 Attempt 1 142 132 Breeding stage (4) 4.04 0.0040 Control vs. High Treatment (1) 4.89 0.0288 Breeding stage*Treatment (4) 2.46 0.0487

PAGE 113

113 Table 4 5. Model summaries of the best models explaining testosterone concentrations in white ibis males. Year Breeding attempt Degrees of freedom Explanatory variables (df) F test Tot al Denominator F value P value 2007 Attempt 1 270 260 Breeding stage (5) 3.187 0.0082 Treatment (3) 0.811 0.4888 Type (1) 0.584 0.4455 2007 Attempt 2 207 197 Breeding stage (5) 5.801 0.0001 Treatment (3) 4.8 0.003 Typ e (1) 1.166 0.2815 2008 Attempt 1 323 302 Breeding stage (4) 34.41 <0.0001 Treatment (3) 2.09 0.1019 Type (1) 17.34 <0.0001 Breeding stage*Treatment (12) 2.06 0.0191 2007 Attempt 1 120 114 Breeding stage (2) 3.4 0.0368 Pairing type model Treatment (2) 1.09 0.3396 Type (1) 0.186 0.6675 2008 Attempt 1 166 158 Breeding stage (2) 25.26 <0.0001 Pairing type model Treatment (2) 0.63 0.5362 Type (1) 16.68 0.0001 Breeding stage*Type (2) 11.16 <0 .0001

PAGE 114

114 Figure 4 1. Parameter estimates of estradiol concentrations of female white ibises, compared to the control group in 2007 (first and second breeding attempts) and 2008 (first breeding attempt). Asterisks denote whether each coefficient was sta tistically significant, compared to the control group in the statistical model s (t 0.05).

PAGE 115

115 Figure 4 2. Parameter estimates of estradiol concentrations of male white ibises during p re breeding compared to the control group in 20 07 (first and second breeding attempts). Asterisks denote whether each coefficient was statistically significant, compared to the control group in the statistical model s (t 0.05).

PAGE 116

116 Figure 4 3. Parameter estimates of estradiol co ncentrations of male white ibises during display compared to the control group in 2007 (first and second breeding attempts) and 2008 (first breeding attempt). Asterisks denote whether each coefficient was statistically significant, compared to the contro l group in the statistical model s (t 0.05).

PAGE 117

117 Figure 4 4. Parameter estimates of estradiol concentrations of male white ibises during nest building compared to the control group in 2007 (first and second breeding attempts) and 2008 (first breeding attempt). A sterisks denote whether each coefficient was statistically significant, compared to the control group in the statistical model s (t 0.05).

PAGE 118

118 Figure 4 5. Parameter estimates of estradiol concentrations of male white ibises during e gg laying compared to the control group in 2007 (first and second breeding attempts) and 2008 (first breeding attempt). Asterisks denote whether each coefficient was statistically significant, compared to the control group in the statistical models (t te 0.1; **: P < 0.05).

PAGE 119

119 Figure 4 6. Parameter estimates of estradiol concentrations of male white ibises during incubation compared to the control group in 2007 (first and second breeding attempts) and 2008 (first breeding attempt). Asterisks d enote whether each coefficient was statistically significant, compared to the control group in the statistical model s (t 0.05).

PAGE 120

120 Figure 4 7. Parameter estimates of estradiol concentrations of male white ibises during chick rearin g compared to the control group in 2007 (first and second breeding attempts) and 2008 (first breeding attempt). Asterisks denote whether each coefficient was statistically significant, compared to the control group in the statistical model s (t tests; *: 0.05).

PAGE 121

121 Figure 4 8. Parameter estimates of estradiol concentrations of homosexual male white ibises in the dosed groups, compared to heterosexual males within the same group for the respective breeding stage (2008 first breeding attempt). Asterisks denote whether each coefficient was statistically significant, compared within group to heterosexual males in the statistical model s (t tests; 0.05).

PAGE 122

122 Figure 4 9. Parameter estimates of testosterone concentrations of female white ibises, compared to the control group in 2007 (first and second breeding attempts). Asterisks denote whether each coefficient was statistically significant, compared to the control group in the stat istical models (t < 0.05).

PAGE 123

123 Figure 4 10. Parameter estimates of testosterone concentrations of female white ibises, compared to the control group during each breeding stage in 2008 (first breeding attempt). Asterisks denote whether each coefficient was statistically significant, compared to the control group in the statistical model s (t tests; *: P 0.05).

PAGE 124

124 Figure 4 11. Parameter estimates of testosterone concentrations of male white ibises, compared to the control group in 2007 (first and second breeding attempts). Asterisks denote whether each coefficient was statis tically significant, compared to the control group in the statistical models (t 0.1; **: P < 0.05).

PAGE 125

125 Figure 4 12. Parameter estimates of testosterone concentrations of male white ibises, compared to the control group during each breeding sta ge in 2008 (first breeding attempt). Asterisks denote whether each coefficient was statistically significant, compared to the control group in the statistical model s (t tests; *: P 0.05)

PAGE 126

126 Figure 4 13. Parameter estimates of the testostero ne concentrations of homosexual male white ibises in the dosed groups, compared to heterosexual males (2008 first breeding attempt). There were no treatment effects, thus, the deviation of testosterone concentrations of homosexual males from heterosexual males in each breeding stage was similar across all dosed groups. Asterisks denote whether each coefficient was statistically significant, compared within group to heterosexual males in the statistical models (t tests; 0.05).

PAGE 127

127 CHAPTER 5 EFFECTS OF CHRONIC M ETHYLMERCURY EXPOSUR E ON CORTICOSTERONE RESPONSES OF W HITE IBISES DURING R EPRODUCTION Introduction Physiological and hormonal pathways of the vertebrate stress response are highly conserved across ta xa (Romero 2004) Glucocorticoid hormones released by the adrenal gland are a major end player in the stress response (Axelrod and Reisine 1984, Sapolsky et al. 2000) and have been widely studied in the context of wildlife responses to stressors (e.g. Wingfield and Silverin 1986, Wingfield et al. 1995, Romero 2004, Wingfield 2005, Busch and Hayward 2009, Schoech et al. 2009) Glucocorticoid hormones ar e regulated by the hypothalamo pituitary adrenal axis (HPA axis) depending on the duration and magnitude of the stressor; thus, acute and chronic stressors can result in different physiological responses (Dallman 1993, Wingfield et al. 1998, Romero 2004) Exposure to chronic stressors may result in acclimation where continued exposure results in a depressed corticosteroid response (reviewed in Romero 2004) Following acclimation, another novel stressor can induce a higher response by the HPA axis than in a non acclimated animal, a process known as facilitation (Bhatnagar and Vining 2003, Romero 2004) Chronic stressors can also increase baseline glucocorticoids, direct resources away from reproduction and reduce fitness in animals (Wingfield and Sapolsky 2003, Romero 2004, Bonier et al. 2009) and there are well documented examples of elevated glucocorticoids suppressing reproductive activity and hormones (e.g. Silverin 1998, Schoech et al. 2009, Sheriff et al. 2009) One of the best examples is that of elevated glucocorticoids due to higher predation risk causing large declines in rep roduction in the snowshoe hare ( Lepus americanus ; Boonstra and Singleton 1993,

PAGE 128

128 Krebs et al. 2001, Sheriff et al. 2009) There is also evidence that elevated cort icosterone levels can mediate nest abandonment in birds. Male pied flycatchers ( Ficedula hypoleuca ) with experimentally elevated corticosterone levels abandoned nests at the nestling stage (Silverin 1998) and free living European starlings ( Sturnus vulgaris ) that abandoned nests had higher levels of corticosterone than birds that did no t abandon nests ( Love et al. 2004) Anthropogenic environmental contaminants are a comparatively novel source of stressors to free living animals (Norris 2000, Boonstra 2004, Busch and Hayward 2009) Depending on contaminant levels, and duration of exposure, these may co nstitute acute or chronic stressors. While some contaminants have elicited elevated glucocorticoid levels (e.g. Wikelski et al. 2001, Wikelski et al. 2002) sometimes exposure results in depressed hormone levels (Norris et al. 1999, Oskam et al. 2004) Responses may also depend on the mechanism of action of the pollutant as well as phys iology and life history of the organism (Busch and Hayward 2009) Some contaminants may not be perceived as a stressor and thus not elicit a response. There could be potentially co nfounding effects as some contaminants can directly modulate the HPA axis even in absence of being perceived as a stressor per se (Busch and Hayward 2009) Methylmercury (MeHg), a globally distributed environmental pollutant, is capable of affecting endocrine pathways including the HPA axis (Tan et al. 2009) Acute experimental MeHg exposure resulted in elevated cortisol levels in rainb ow trout ( Onchorynchus mykiss ); however, exposure levels were higher than typically found in the field (Bleau et al. 1996) In a river system polluted by mercury (Hg), polycyclic aromatic hydrocarbons and polychlorinated biphenyls, fish showed inability to re spond

PAGE 129

129 to acute stress with increased cortisol (Hontela et al. 1992) Both adult and nestling free living tree swallows ( Tachycineta bicolor ) showed a negative relationship between blood Hg and baseline cort icosterone (the main glucocorticoid in birds; Franceschini et al. 2009) Another study showed elevated baseline corticosterone in tree swallow nestlings exposed to a Hg contaminated site in comparison with nestlings from a reference site (Wada et al. 2009b) However, the birds in the contaminated site showed a depressed response to capture stress, compared with reference birds. W hile several studies show depressed reproduction following chronic MeHg exposure (e.g. Brasso and Cristol 2008, Evers et al. 2008, Crump and Trudeau 2009) there i s paucity of information about MeHg effects on corticosterone during reproduction. An exception is in the common loon ( Gavia immer ) where a strong correlation was found between plasma corticosterone, Hg exposure categories based on blood Hg levels, and re productive success (Evers et al. 2004, Evers et al. 2008) An increase of 1 ppm of blood Hg resulted in a 14.6% increase in plasma c orticosterone in loons (Evers et al. 2004) and breeding loons in the hig h r isk category (> 3 ppm blood mercury) had 40% less fledged young than birds with < 1 ppm blood mercury (Evers et al. 2008) Though not conclusive, given the association between high glucocorticoids and decreased reproduction (e.g. Sheriff et al. 2009) it is possible that high corticosterone was a factor in reduced loon production. White ibises ( Eudocimus albus ) have been chronically exposed to sub lethal levels of dietary MeHg in the Florida Everglades (Frederick 1999, Frederick et al. 2004, Heath and Frederick 2005) MeHg exposure has been correlated with altered sex steroid levels during reproduction in free living ibises in the Everglades ( Heath and Frederick

PAGE 130

130 2005) In the same study, there was a negative relationship between a MeHg exposure index and numbers of ibises nesting, although it was not clear whether decreased nesting effort or increased nest abandonment, or both, were responsib le for the lower numbers. A controlled dosing experiment using dietary MeHg exposure ranging from 0.05 0.3 ppm wet weight (ww) found increased occurrence of male male pairs (chapter 2) altered rates of courtship behavior (chapter 3) and altered concent rations of estradiol and testosterone (chapter 4) in dosed whit e ibises The same experiment found highly non linear effects of MeHg on corticosterone during the juvenile period, but high dose birds had consistently elevated concentrations in comparison t o control birds (Adams et al. 2009) To summarize the evidence from the above studies, chronic exposure to MeHg can modulate corticosterone responses in birds, both MeHg and corticosterone can depress reproduction, and corticosterone is associated with nest abandonment in birds. The white ibis population in the Everglades has been chronically expo sed to MeHg and has shown a positive correlation between exposure and reduced reproduction. With this evidence, I hypothesized that chronic exposure to sub lethal levels of MeHg increase basal corticosterone levels in the white ibis during reproduction. I further hypothesized that an additional acute stressor during this period would increase corticosterone levels to a greater extent in MeHg exposed birds via facilitation of the corticosterone response. Furthermore, I predicted that MeHg exposed birds wou ld abandon nests to a greater degree than unexposed birds due to higher corticosterone levels. I tested these hypotheses in captive white ibises, experimentally exposed to dietary MeHg (at 0.05, 0.1 and 0.3 ppm ww) for 3.5 years by monitoring fecal

PAGE 131

131 cortic osterone metabolites during two breeding seasons. I also experimentally induced a period of reduced food availability during the incubation stage to test the hypothesis that elevated corticosterone induced by this additional stressor would cause increased nest abandonment in MeHg exposed birds. Methods Experimental Setup and Exposure of Captive White Ibises to Methylmercury W hite ibis nestlings were collected from colonies in the Florida Everglades in spring 2005 Nestlings were randomly assigned to one o f four groups of 20 males and 20 females in each, receiving 0.00 (control), 0.05(low), 0.1 (medium) or 0.3 (high) ppm MeHg wet weight (ww) in pelletiz ed food. These MeHg concentrations spanned the range found in their food items in the Everglades during p eak mercury contamination in the 1990s (Frederick et al. 1999, Loftus 2000, Frederick et al. 2004) Birds were continuously exposed to these diets from 90 days of age onwards through out the durati on of the experiment (2005 to 2008). The birds were housed together in a 1200m 2 circular free flight aviary with the different treatment groups being separated only by net walls. Locations of groups were switched each year prior to the breeding season, so that each breeding season was spent in a different location thus controlling for location effects. All birds were individually identifiable by leg bands, and were genetically sexed (Avian Biotech International, Tallahassee, Florida) Monitoring Breedi ng Behavior All treatment groups were provided with 48 platforms on perches to provide nesting structures and an ad libitum supply of twigs and nesting materials throughout the nesting season (generally February August). Birds bred for the first time a s sub adults in 2006 and annually in 2006, 2007 and 2008. Identities of birds displaying, nest

PAGE 132

132 building, laying, incubating or chick rearing and nest contents were recorded daily at sunrise throughout each breeding season Observers were kept blind to t reatment groups until the end of the study Experimental Reduction of Food Availability d uring Breeding During the 2008 breeding season I restricted food in all treatments to test the prediction that food stress would increase nest abandonment in mercury d osed birds. The ad libitum food supply was reduced for seven days (24 30 March, 168 hours) to 80% of the food consumed in each treatment group in the three weeks prior to the experiment. Food reduction occurred when at least ten pairs incubating in each treatment group. At this time there were 11, 12, 11 and 14 pairs incubating in the control, low, medium and high dose groups respectively. There was one nest with 1 5 day old hatchlings in the medium group. Collection of Fecal Samples This is described in detail in the methods section of chapter 4. I collected individually identified fresh fecal samples from birds several times weekly during the breeding seasons of 2007 and 2008. Samples were collected using clean wooden spatulas into individual 2 ml p olypropylene cryotubes (Fisher Scientific) and immediately placed on ice. They were stored temporarily for up to eight hours in a 4 Celsius freezer, and transferred to a 20 Celsius freezer for longer term storage. The breeding season for each bird was d ivided into six stages: pre breeding (only in 2007), display, nest building, egg laying, incubation and chick rearing. Since male male pairs (chapter 2) had no eggs to incubate or young to raise, the period following completion of nest building was define d as the incubation stage. Fecal samples were

PAGE 133

133 collected in the first two breeding attempts of 2007, and the first breeding attempt of 2008. Radioimmunoassays for Corticosterone Metabolites in Fecal Extracts Fecal samples were freeze dried, homogenized, an d 0.05g of each sample used for extraction (by 80% ethanol) of steroid hormone metabolites (Adams et al. 2009) Fecal extracts were stored in a 20 Celsius freezer until analysis for corticosterone hormones by radioimmunoassay. I followed radioimmunoassay methods previously validated for corticosterone in white ibis fecal extracts (Adams et al. 2009) Corticosterone 125 I double antibody RIA kits (MP Biomedicals, Solon, OH, USA) were used to measure fecal corticosterone metabolites. Samples were analyzed in duplicate, and those with a coefficient of variation more than 15% were re analyzed or discarded if there were in s ufficient samples for reanalysis Intra assay coefficients of variation we re <15%, while the inter assay coefficient of variation was 11.7% and 12.9% for 2007 and 2008 samples respectively. Statistical Analyses I used natural logarithms of hormone concentrations (nanograms per gram feces, dry weight) to normalize data. R statis tical software, version 2. 10 0 (R Development Core Team, 2009) was used for statistical analyses Extended linear models with a correlation structure and variance function were used for data analysis (Pinheiro and Bates 2000, Pinheiro et al. 200 9; see chapter 4 for details) Response variables were the log transformed corticosterone concentrations To test whether corticosterone concentrations were affected by MeHg exposure group (= treatment) during the period of normal food allowance, I con ducted separate analyses by gender, breeding attempt (first and second in 2007) and year (2007 and

PAGE 134

134 2008). Predictor variables were breeding stage, treatment, and for males, type of pairing behavior (heterosexual or homosexual pairing). For males, two se ts of models were tested in each year for the period of normal food availability. One included data from all breeding stages and treatments (full model). The second comprised data from dosed males in display, nest building and incubation stages of the fi rst attempt (stages relevant to both heterosexually and homosexually nesting males). This model (pairing type model) was to test interactive effects of mating type in males and treatment (see chapter 4 for details). To test whether food restriction affect ed corticosterone concentrations, data was truncated to include only nest building, egg laying and incubation stages, as other stages did not have sufficient individuals during food restriction. An additional categorical predictor of food stress was added to this set of models. For all models, interactions were tested and removed if non significant (P<0.05; F test s ) Residual plots and quantile quantile plots of the models were examined to see whether data conformed to the assumption of normality. I repo rt results of conditional F tests (F tests) which test the significance of explanatory variables that could include several coefficients (e.g. the explanatory variable for treatment effects has coefficients for each dose group) and of conditional t tests ( t tests) which test the marginal significance of each separate coefficient when all other coefficients are present in the model (Pinheiro and Bates 2000) Results Effects of Methylmercury and Food Stress on Corticosterone in Adult Females There were significant effects of breeding stage o n fecal corticosterone metabolites of females during both breeding attempts of 2007 (first: F 5, 274 = 18.2, P

PAGE 135

135 <0.0001; second: F 5, 172 = 4.5, P = 0.0008; Table 5 1). I did not find any significant effects of MeHg sterone in the first attempt (F 3, 274 = 0.8, P = 0.49, Fig. 5 1). In the second attempt, the coefficients for the low (L) and medium (M) dose groups showed a non significant trend towards decreased corticosterone by comparison with controls, though there were no changes in the high (H) dose group (L: t = 1.9, P = 0.067; M: t = 1.8, P = 0. 07; H: t = 1.15, P = 0.25; Fig. 5 1) and the overall F test for treatment was non significant (F 3, 172 = 1.5, P = 0.2). There were no significant interactive effect s of breeding stage*treatment group in either year. Actual fecal concentrations of corticosterone metabolites are reported in Appendix C. In 2008, both treatment (F 3, 279 = 4.7, P = 0.003), and breeding stage*treatment interaction (F 12, 279 = 2, P = 0.021 Table 5 1) were significant predictors of corticosterone level. However, not all coefficients for treatments across stages showed significant effects in comparison with the control group (Table 5 2; Fig. 5 2). The most obvious differences included incr eased corticosterone concentrations during incubation in the low dose females (t = 2.6, P = 0.09) with a similar trend in the high dose females (t = 1.7, P = 0.097); and depressed corticosterone concentrations during chick rearing in the medium dose group (t = 2.7, P = 0.008) with similar trends in the other groups (low: t = 1.6, P = 0.11; high: t = 1.5, P = 0.13). There were no nest abandonments during the period of artificially induced food stress in 2008 in any group, but corticosterone concentrations were elevated, irrespective of treatment group, during nest building, incubation and egg laying stages of food stressed females compared to the non food stressed period (F 1, 190 = 300, P

PAGE 136

136 <0.0001, Table 5 1, Figure 5 3). The largest increase was during ne st building and the smallest during incubation. E ffects of Methylmercury and Food Stress on Corticosterone in Adult Males I found marginally significant effects of treatment on male corticosterone during the first breeding attempt of 2007 (F 3, 259 = 2.4, P = 0.07, Table 5 3) but no effects of pairing type (F 1, 259 = 1.7, P = 0.19). The pairing type model for 2007 also did not show a significant effect of MeHg exposure on corticosterone concentrations of homosexual males (F 1, 114 = 0.5, P = 0.48; Table 5 3) In the full model, only the high dose group had significant changes compared to the control, which were negative in direction (L: t = 0.93; P = 0.35; M: t = 0.65; P = 0.52; H: t = 2.12, P = 0.035; Fig. 5 4). In the second attempt of 2007, there were significant treatment effects (F 3, 198 = 8.9, P <0.0001; Table 5 3) as well as effects of pairing type (F 1, 198 = 4.4, P = 0.037) in the full model. All significant differences in the second attempt showed decreased parameter estimates when compared to th e control group (L: t = 1.8, P = 0.07; M: t = 4.18, P < 0.0001; H: t = 4.49; P < 0.0001; Fig. 5 4). While breeding stage was an important predictor of corticosterone concentrations in both attempts (first: F 5, 259 = 22.9, P < 0.0001; second: F 5, 198 = 13.3, P <0.0001), there were no interactions between breeding stage and treatment group in either attempt. In 2008, I found significant effects of breeding stage*treatment interaction (F 12, 318 = 1.9, P = 0.035; Table 5 3) and pairing type (F 1, 318 = 7.3 P = 0.007) on corticosterone, although not all treatment stage coefficients were significant (Table 5 4; Fig. 5 5). Males in the high dose group had significantly decreased concentrations of corticosterone during nest building (t = 2.04, P = 0.04). Du ring chick rearing both medium and high dose groups showed negative trends (M: t = 1.9, P = 0.06; H: t = 1.5,

PAGE 137

137 P = 0.13). The low dose group had non significant trends towards decreased corticosterone concentrations during display (t = 1.7, P = 0.09), a nd elevation during incubation (t = 1.6, P = 0.099) but none of the other dose groups showed changes in comparison with controls during these breeding stages. In the pairing type model of 2008, I found a significant interaction of breeding stage*pairing ty pe (F 2, 167 = 49.9, P <0.001; Table 5 3) and a marginally significant interaction of treatment*pairing type (F 2, 167 = 2.6, P = 0.078). Main effects of both treatment and pairing type were highly significant in this model (F 2, 167 = 9.4, P <0.0001 and F 1, 167 = 30, P <0.0001 respectively). Homosexual males generally showed depressed corticosterone compared with heterosexuals within treatment groups (Fig.5 6; Table 5 5), but high dose males had elevated concentrations during the display stage (t = 2.03, P = 0.04) and the medium dose group showed a similar trend (t = 1.8, P = 0.078) compared to heterosexual males within respective groups. During incubation homosexual males in all dosed groups had significantly decreased corticosterone concentrations in comp arison with heterosexual males (t tests, P <0.0001). In the food stress model, I found significant effects of food stress (F 1, 206 = 332.9, P <0.0001; Table 5 3) and pairing type (F 1, 206 = 7.5, P = 0.007), significant interactions of food stress*treatment (F 3, 206 = 3.9, P = 0.009) and food stress*pairing type (F 1, 206 = 13.7, P = 0.0003) and a marginally significant interaction in breeding stage*food stress (F 2, 206 = 2.5, P = 0.09). The main differences in corticosterone were during incubation (Table 5 6; Fig. 5 7), with food stress related to significantly elevated corticosterone in high dose males (t = 2.3, P = 0.023) and a similar trend in the medium dose group (t = 1.7, P = 0.085). Homosexual males had depressed corticosterone during incubation (t

PAGE 138

138 = 2.03, P = 0.043). There was a positive trend in corticosterone of high dose heterosexual males during laying (t = 1.5, P = 0.12) but no changes in any group during nest building. Compared within each treatment group and breeding stage though, mean corti costerone concentrations were always elevated during food stress relative to the period of normal food availability (t tests, P <0.0001; Figure 5 8). Discussion Effects of Methylmercury on Corticosterone during Reproduction Exposure to MeHg, in the absence of food stress, had an overall depressing effect on basal corticosterone concentrations during reproduction in heterosexual male and female ibises. This was contrary to my prediction that chronic exposure to MeHg would elevate basal corticosterone. Howe ver, negative relationships of corticosteroids with chronic exposure to stress has been reported in other studies (Rich and Romero 2005, Cyr and Romero 2007) Exposure to MeHg has also been associated with decreased baseline corticosterone levels in free living tree swallows (Franceschini et al. 2009) It is possible that chronic exposure to MeHg depressed the physiological corticosterone response in ibises via acclimation (Romero 200 4) There is also a potentially confounding factor of MeHg being a endocrine disruptor of steroid hormones (reviewed in Tan et al. 2009, also see chapter 4, this study) thereby possibly affecting corticosterone metabolism pathways other than by modulating the stress response. Glucocorticoid hormones play a role during reproduction by mediating metabolic pathways that provide for the higher energetic requirements of reproductive activity and are typically high during chick rearing (Love et al. 2004) In my study too, control ibises of both sexes had the highest corticosterone concentrations during chick rearing but dosed groups had lower corticoster one during chick rearing when compared to controls.

PAGE 139

139 Corticosterone promotes gluconeogenesis and foraging behavior in birds (Astheimer et al. 1992, Wingfield et al. 1 998) Thus, MeHg induced depression of corticosterone could be detrimental in free living birds particularly during times of high energy demand such as chick rearing, and when food availability is more variable. The one exception to depressed corticoste rone concentrations in dosed heterosexual birds in the non food stressed period was the incubation stage. Both low and high dose group females had elevated corticosterone during incubation relative to controls in 2008. Effects of Methylmercury and Food St ress on Corticosterone During food stress, incubating males in the medium and high dose groups had higher corticosterone concentrations compared to control males. Though there were no nest abandonments, corticosterone did show facilitation in response to an acute stressor in mercury dosed birds, as predicted (Bhatnagar and Vining 2003, Romero 2004) The duration and magnitude of the food stress I induced was much lower than would be typical for free living birds where food availability could be nil for some periods and might fluctuate more than the stable 20% reduction in this study. Energy expenditure is higher in free living birds and would be even higher under conditions of low food availability, thus i ntensifying the stressor. The aviary food supply was also very predictable even during food restriction, and predictable events are known to be much less stressful than unpredictable ones (Wingfield 2005) My best prediction is therefore, that a much higher level of food stress in the field would probably r esult in a major glucocorticoid response that might well result in nest abandonment. This is especially true if there are other stressors acting in concert with food availability. It is also possible that in incubating birds normal reproductive signals we re stronger than those altered by MeHg exposure and application of the stressor.

PAGE 140

140 Evidence that responses to environmental variability has endocrinologically mediated inbuilt constraints in birds (Jacobs and Wingfield 2000) is provided by arctic breeding birds that are able to down regulate their stress r esponse to continue breeding under adverse environmental conditions (Wingfield and Hunt 2002) Though arctic breeding is not analogous to my study, physiological and behavioral pathways in incubating ibises may be more resistant to change by low intensi ty external modulators such as that imposed by food restriction, since birds have invested considerable resources in reproduction by this time. Not surprisingly, I saw elevated corticosterone in all birds during the period of food stress, irrespective of s ex, treatment, breeding stage or sexual preference. There is ample evidence that restricted food availability elevates glucocorticoids in birds and is known to promote foraging behavior (Astheimer et al. 1992, Rogers et al. 1993, Wingfield et al. 1998, Kitaysky et al. 1999, Kita ysky et al. 2007) It is also not surprising that the smallest rise in corticosterone concentrations was during incubation, in both sexes and across treatments. As noted before, higher levels of corticosterone are associated with nest desertion in birds (Silverin 1998, Love et al. 2004) Relative to nest building and laying stages, parental investment is higher by the incubation stage. Therefore, it is adaptive to suppress stress responses when parental investments are higher and indeed, birds are able to suppress their stress response when it is reproductively advantageous to do so (Wingfield et al. 1995, Wingfield and Kitaysky 2002, Wingfield and Sapolsky 2003, B okony et al. 2009) It is interesting that in spite of MeHg exposure facilitating the stress response to food restriction (in males) their stress

PAGE 141

141 responses were relatively attenuated during incubation. It is also true that energetic needs during incuba tion are probably the lowest of any stage in reproduction. Effects of Methylmercury on Corticosterone in Homosexually paired Males On almost all occasions I found corticosterone responses in homosexual males were depressed relative to heterosexual males. The single exception was during the display stage (in 2008), where homosexual males of medium and high dose groups had higher corticosterone concentrations compared to heterosexual males within their same groups. During the courtship period I also observ ed differences in behavior of these males (see chapter 3), where they showed lower rates of display activity and aggression. High levels of corticosterone are known to inhibit reproductive behavior in birds (Wingfield and Sapolsky 2003) For example, experimental elevation o f corticosterone in male song sparrows ( Melospiza melodia crowned sparrow ( Zonotrichia leucophrys gambelii ) resulted in suppression of territorial behavior during breeding (Wingfield and Silverin 1986, Meddle et al. 2002) Therefore, it is possible that high corticos terone levels were involved in this apparent reduction of courtship behavior in homosexual white ibises. During the courtship period, I also observed lower numbers of females approaching males that formed male male pairs later on (chapter 3) and it is pos sible that females were avoiding males with higher corticosterone. In mate choice trials, female zebra finches ( Taeniopygia guttata ) preferred males who were selected to produce lower levels of circulating corticosteroids indicating a role for glucocortic oids in mate choice (Roberts et al. 2007, Husak a nd Moore 2008) It is possible that altered sexual preference and behavioral modifications in male ibises in this study were mediated by MeHg induced changes in circulating steroid

PAGE 142

142 levels, leading to organizational effects during development (Adkins Regan et al. 1997, Adkins Regan and Leung 2006, see also chapter 4) In the Florida scrub jay ( Aphelocoma coerulescens ), levels of corticosterone during the nestling stage were strongly correlated wi th fearful behavior later, leading the authors to speculate that early exposure to glucocorticoids can program later behavioral patterns (Schoech et al. 2009) The birds in my study were exposed to MeHg since they were 90 days of age, and non linear effects of treatment on corticosterone and sex steroid metabolites were found before sexual maturity (Adams et al. 2009) possibly having organizational effects. Conclusions Long term exposure to MeHg appeared to acclimat e birds to this stressor, resulting in basal corticosterone concentrations being generally depressed compared to control birds. However, experimentally induced food stress evoked a strong increase in corticosterone in all birds. Facilitation of the corti costerone response to food stress by MeHg treatment was most pronounced in the high dose group. The food stress I induced was not sufficient to cause birds to abandon nests, but the direction and nature of the corticosterone response suggests that MeHg ex posure predisposes birds to abandon in response to the more acute food fluctuations in the field, or to multiple stressors. I have high internal and external validity in my study system since the experimental MeHg exposure was similar in chronicity and l evel to that of wild ibises (Frederick et al. 1999, Heath and Frederick 2005) and effects seen in my experiment are therefore likely to be manifested in free living birds at similar exposure levels. Though mercury levels in fish and wading birds have generally decreased by

PAGE 143

143 compari son with the peak contamination in the 1990s, there are still areas of mercury hotspots in the Everglades where exposure levels are deemed high (Rumbold et al. 2008) and MeHg contamination remains a concern in this ecosystem (Perry 2008) Therefore these findings are relevant to wildlife management in the Everglades.

PAGE 144

144 Table 5 1. Model summaries of the best models explaining corticosterone concentrations in white ibis females. Year Breeding attempt Degrees of freedom Explanatory variables (df) F test Total Denominator F value P value 2007 Attempt 1 283 274 Breeding stage (5) 18.2 < 0.0001 Treatment (3) 0.8 0.4969 2007 Attempt 2 1 81 172 Breeding stage (5) 4.5 0.0008 Treatment (3) 1.5 0.2194 2008 Attempt 1 299 279 Breeding stage (4) 42.8 <0.0001 Treatment (3) 4.7 0.0031 Breeding stage*Treatment (12) 2.0 0.0206 2008 Attempt 1 199 190 Breeding stage (2) 7.8 0.0006 Food stress Treatment (3) 1.9 0.1237 Food stress (1) 300.6 <0.0001 Breeding stage*Food stress (2) 12.0 <0.0001 Table 5 2. Coefficients comparing each dosed group with control females during each breeding stage i n 2008. Breeding stage Low Medium High t value P value t value P value t value P value Display 3.22 0.002 1.11 0.268 0.30 0.768 Nest building 0.99 0.325 0.81 0.420 0.17 0.868 Egg laying 0.25 0.803 0.62 0.533 0.39 0.698 Incubation 2.63 0.009 0.45 0.651 1.67 0.097 Chick rearing 1.58 0.115 2.66 0.008 1.50 0.135

PAGE 145

145 Table 5 3. Model summaries of the best models explaining corticosterone concentrations in white ibis males. Year Breeding attempt Degrees of freedom Explanatory variables (df ) F test Total Denominator F value P value 2007 Attempt 1 269 258 Breeding stage (5) 22.9 <0.0001 Full model Treatment (3) 2.4 0.0708 Pairing type (1) 1.7 0.1878 2007 Attempt 2 208 198 Breeding stage (5) 13.3 <0.0001 Ful l model Treatment (3) 8.9 <0.0001 Pairing type (1) 4.4 0.0371 2008 Attempt 1 339 318 Breeding stage (4) 89.0 <0.0001 Full model Treatment (3) 2.6 0.0549 Pairing type (1) 7.3 0.0074 Breeding stage*Pairing type (12) 1.9 0.0353 2007 Attempt 1 120 114 Breeding stage (2) 12.1 <0.0001 Pairing type model Treatment (2) 3.2 0.0445 Pairing type (1) 0.5 0.4839 2008 Attempt 1 177 167 Breeding stage (2) 172.2 <0.0001 Pairing type model Treatment (2) 9. 4 0.0001 Pairing type (1) 30.0 <0.0001 Treatment*Pairing type (2) 2.6 0.0785 Breeding stage*Pairing type (2) 49.9 <0.0001 2008 Attempt 1 226 206 Breeding stage (2) 1.1 0.3471 Food stress Treatment (3) 0.5 0.6592 Pairing type (1) 7.5 0.0067 Food stress (1) 332.9 <0.0001 Breeding stage*Treatment (6) 1.5 0.1901 Breeding stage*Food stress (2) 2.5 0.0887 Treatment*Food stress (3) 3.9 0.0094 Pairing type*Food stress (1) 13.7 0.0003

PAGE 146

146 Tab le 5 4. Coefficients comparing each dosed group with control males during each breeding stage in 2008 Breeding stage Low Medium High t value P value t value P value t value P value Display 1.71 0.089 1.07 0.284 0.66 0.508 Nest building 0.71 0.48 1 1.35 0.178 2.04 0.042 Egg laying 0.56 0.575 0.13 0.900 1.44 0.150 Incubation 1.65 0.100 0.88 0.379 0.36 0.719 Chick rearing 1.31 0.191 1.86 0.064 1.52 0.129 Table 5 5. Coefficients comparing homosexual males in each dosed group with heterosexu al males within the same group, by breeding stage in 2008. Breeding stage Low Medium High t value P value t value P value t value P value Display 2.55 0.0118 1.77 0.0781 2.03 0.0439 Nest building 2.68 0.0080 0.38 0.7064 0.20 0.8411 Incubation 1 1.85 <0.0001 5.76 <0.0001 5.42 <0.0001 Table 5 6. Coefficients comparing dosed males with control males during the period of food restriction in 2008. Breeding stage Low Medium High t value P value t value P value t value P value Nest building 0. 70 0.485 0.83 0.410 1.21 0.229 Egg laying 0.96 0.338 1.01 0.315 1.55 0.123 Incubation 0.42 0.673 1.73 0.085 2.30 0.023

PAGE 147

147 Figure 5 1. Parameter estimates of corticosterone concentrations of dosed females compared to control females in the first and second breeding attempts of 2007. Asterisks denote whether each coefficient was statistically significant, compared to the control group in the statistical models (t tests; < 0.05).

PAGE 148

148 Figure 5 2. Parameter estimates of corticosterone concentrations of dosed females compared to control females in each breeding stage during the first breeding attempt of 2008. Asterisks denote whether each coefficient was stati stically significant, compared to the control group in the statistical model s (t tests; *: P 0.05).

PAGE 149

149 Figure 5 3. Parameter estimates of corticosterone concentrations of females during the period of experimentally induced food stress in 200 8, compared to the period of normal food availability. Effects of food stress on corticosterone differed due to breeding stage as indicated in the figure, but not due to treatment group. Asterisks denote whether each coefficient was statistically signifi cant, compared within stage, to the period of normal food availability in the statistical models (t tests; : P 0.05).

PAGE 150

150 Figure 5 4. Parameter estimates of corticosterone concentrations of dosed and homosexually paired males compared to heterosexually paired control males during the first and second breeding attempts of 2007. Note that homosexual mal es could be in any treatment group. Asterisks denote whether each coefficient was statistically significant, compared to the control group in the statistical models (t

PAGE 151

151 Figure 5 5. Parameter estimates of corticosterone c oncentrations of dosed males compared to control males in each breeding stage during the first breeding attempt of 2008. Asterisks denote whether each coefficient was statistically significant, compared to the control group in the statistical models (t te sts; *: P 0.05).

PAGE 152

152 Figure 5 6. Parameter estimates of corticosterone concentrations of dosed homosexually paired males compared to heterosexually paired males in the respective treatment group during the first breeding attempt of 2008. Asterisks de note whether each coefficient was statistically significant, compared to heterosexual males in the stat istical models (t 0.1; **: P <0.05).

PAGE 153

153 Figure 5 7. Parameter estimates of corticosterone concentrations of dosed and homosexual males, com pared to heterosexual control males, during the period of experimentally induced food stress in 2008. Effects of food stress on corticosterone differed due to breeding stage and pairing type as indicated in the figure. Asterisks denote whether each coeff icient was statistically significant, compared to the control group in the statistical models (t tests; : P 0.05).

PAGE 154

154 Figure 5 8. Parameter estimates of corticosterone concentrations of males during the period of experimentally induced food stress in 2008, compared to the period of normal food availability. Although effects of food stress on co rticosterone differed due to breeding stage and treatment group, all groups had significant ly elevated concentrations when compared to itself during normal food availability; denoted by asterisks (**; t tests; P < 0.05).

PAGE 155

155 CHAPTER 6 CONCLUSION Introduction In the previous chapters I reported the results of experiments examining the effects of chronic methylmercury ( MeHg ) exposure (0.05 0.3 ppm ww in diet) on reproductive success, behavior and steroid hormones in white ibises ( Eudocimus albus ). This sectio n will be a synthesis of the results, and suggestions for future directions. I will address the following questions: By what mechanisms can male male pairing occur as a result of MeHg exposure? How were females affected by MeHg exposure? What effects migh t be felt at the population level in the wild? Synthesis Male male Pairing in Ibises Induced by Methylmercury Exposure As reported in the previous chapters, I documented dose related patterns in male male pairing. There was an average 13 14% loss in egg p roduction in dosed groups over three years (range 7 30% across years and treatment groups) of which the majority (74 91%) was due to homosexual pairing behavior (range 50 100% across years and treatment groups). The maximum proportions of homosexual males were 20% in the control group, and 36%, 43% and 56% in low, medium and high dose groups which dropped to 0%, 18%, 26%, and 22% respectively by the third year. Homosexual males showed lower rates of key courtship behaviors (head bobbing and pair bowing), reduced rates of aggression, had fewer females approach them, and had other males approaching them during courtship. Homosexual males also initiated nest building

PAGE 156

156 earlier in the breeding season than heterosexual pairs, and showed a lower probability of sw itching partners between breeding attempts. While I found evidence of MeHg induced endocrine disruption in estradiol and testosterone levels of dosed males, the changes in sex steroid levels in heterosexual dosed males were, on the whole, mirrored in dire ction and exaggerated in homosexual males. The most prominent change in homosexual males was elevated estradiol levels during the courtship period. The magnitude of male male pairing was considerable, since even in 2008, when male male pairing was at its minimum, there were 18 26% more homosexual males in dosed groups than in the control group. This translates to a similar percentage of MeHg exposed males not being involved in productive nesting at a given time with an equal percentage of females not havi ng males available in a breeding colony with a sex ratio of one. As discussed in previous chapters, there are several mechanistic pathways which could possibly result in homosexual pairing. Male male pairing may have been due to behavioral alteration duri ng courtship in males, due to female mate choice (avoidance of these males), alteration of sexual preference due to changes in circulating levels of steroid hormones, or organizational effects during development (Fig. 6 1). Since behavior, sexual preferen ce and hormones can all interact (Adkins Regan et al. 1997, Adkins Regan and Leung 2006, Ottinger et al. 2008) it is possible that two or more of these postulated mechanisms acted in concert to produce changes in sexual preference of MeHg exposed males. Changes in sex steroids of homosexual males, for the most pa rt, showed a similar direction to changes seen in heterosexual dosed males, but were exaggerated in

PAGE 157

157 magnitude. This indicates two patterns. First, male male pairing was associated with similar hormonal changes as seen in dosed heterosexual males, which i s indicative of hormonal changes in male male pairing being related to MeHg exposure. Second, the exaggerated hormonal response in homosexual males indicates that male male pairing was associated with the most extreme end of a continuum of MeHg induced ho rmone disruption However, it seems likely that MeHg exposure before development of sexual maturity was a key factor in altering sexual preferences, since available evidence indicates this occurs early in life (e.g. Adkins Regan et al. 1997, Adkins Regan and Leung 2006) Changes during development may occur from MeHg induced changes in sex steroid expression which influence organizational pathways and/or by MeHg induced changes in organization independently of changes in sex steroids (e.g. direct MeHg action on neuronal pathways, MeHg induced changes in neuromodulators). I hypothesize that MeHg induced organizational change which occurred before sexual maturity influenced sexual preference and altered hormone expression of ibises in later life. This hypothesis leads to several specific questions: 1) can MeHg exposure induce organizational changes; 2) can organizational changes induced by MeHg exposure lead to changes in expression of sex steroids during adult hood; 3) is there a relationship between MeHg induced organizational changes and altered sexual preferences; 4) do MeHg induced organizational changes influence sexual behavior in adulthood; 5) is there a relationship between sex steroid expression and sex ual preference; 6) what is the relationship between sex steroid expression and sexual behavior; and 7) is there a relationship between sexual behavior and sexual preference. These questions could be

PAGE 158

158 tested by comparisons of endocrine profiles and mate cho ice of birds exposed to MeHg before and after sexual maturation. This would address the question of whether sexual preferences and hormonal changes were organizational, activational or both. The window of time within which sexual preferences are formed c ould be quite narrow (Adkins Regan et al. 1994) Therefore, using different time periods for MeHg exposure (egg, nestling, juvenile, and adult) and switching them at different periods from dosed to control groups would be a potential method of determining the sensitive period during which sexual preferences can be altered. If birds exposed in ovo, or through stages prior to sexual maturity, but subsequently brought up in a mercury free environment show altered sexual preference (resulting in male male pairing) and changes in sex ste roids similar to birds continuously exposed to MeHg, this would indicate organizational effects of MeHg on both sexual preferences and expression of sex steroids. However, this would not prove a link between sex steroids and sexual preferences. While oth er studies do show such a link (Adkins Regan et al. 1997, Adkins Regan and Leung 2006) ; testing this would require experimentally inducing changes sex steroids during the sensitive period independently of MeHg. Another crucial que stion is whether changes in sexual behavior are mediated via organizational effects or circulating levels of sex steroids, and how these affect sexual preference. These relationships are not well understood at present. Although sex steroids are generally believed to be associated with sexual behavior, it is acknowledged that hormones are one of many possible components which can affect behavior (Crews and Moore 1986, Owens and Short 1995, Wingfield et al. 2001, Fusani

PAGE 159

159 2008) Organizational changes associated with a ltered sex steroid expression in adult hood could both be responsible for changes in sexual preference. There is evidence that altering circulating sex steroids can influence some aspects of sexual behavior (Hutchison 1970, Day et al. 2006) while others remain unaffected (Moore and Kranz 1983, Fusani and Hutchison 2003, Fusani et al. 2007) While t he relationship between sexual behavior and circulating sex steroids are somewhat equivocal, available evidence indicates that changing levels of sex steroids after the critical development period does not change sexual preferences (reviewed in Adkins Regan 2007) Another tool to find evidence of mechanistic relationsh ips between endocrine changes and sexual preferences is using genetic techniques. Gene array and quantitative genetic techniques have been used to determine whether MeHg exposure results in differential gene expression in fish (e.g. Gonzalez et al. 2005, Klaper et al. 2008) It is possible to develop similar techn iques to determine whether changes occur in gene expression of specific genes involved in hormone expression of MeHg exposed animals at various stages (e.g. during development, after sexual maturity). Effects of Methylmercury on Reproduction in Females Fem ale ibises in low and high dose groups had on average 33% and 35% fewer numbers of young fledged over three years of breeding in comparison with control females. Changes in steroid hormone levels of dosed females during reproduction showed, for the most p art, decreased levels in comparison to control females. However, the greatest loss of reproductive success occurred in the high dose group in 2008, and was associated with increased levels of testosterone. Again, several interacting mechanisms could be re sponsible for the decreased reproductive success in dosed females. As with the males, MeHg may directly alter

PAGE 160

160 behavioral pathways, act via organizational changes, or act via changes in circulating levels of hormones. Given the changes in endocrine profil es seen in dosed birds, it seems likely that endocrine disruption was involved. I hypothesize that endocrine disruption in sex steroids, (especially elevated testosterone) were responsible for poor reproductive success in females. Further, I suggest that the altered testosterone was probably caused by MeHg induced activational changes rather than organizational changes. The increase in reproductive failure and endocrine changes with longer and higher exposure lead me to believe, that unlike in males, cha nges in circulating hormone levels are caused more by current MeHg exposure than they are by exposure during development. I predict longer MeHg exposure would increase reproductive failure and produce more endocrine changes in females. More specifically I predict that longer MeHg exposure would result in females having behavioral alterations during nesting such as laying outside nests, reduced incubation of eggs, reduced parental attention to nestlings, and increased aggression to nestlings, as has been f ound in other species. Based on trends in my studies, I predict that these changes will be associated with increased levels of testosterone in post laying stages of females. I further predict that breeding success would increase after a period of depurat ion of MeHg in females. A long term dosing study would be necessary to test these predictions. Starting MeHg exposure at various stages of development (in ovo, nestling, juvenile, adult) can test whether MeHg effects are organizational or activational. S witching adult birds between dosed and control groups would test the prediction that mercury (Hg) depuration can increase breeding success and normalize endocrine profiles in formerly

PAGE 161

161 dosed females and vice versa. A direct way of testing whether increased testosterone is responsible for behavioral effects is by experimentally increasing testosterone levels in un dosed females. Other studies have found evidence that experimentally elevated testosterone can increase aggression (Adkins Regan 1999) alter parental behavior and reduce fledging success (O'Neal et al. 2008) in females as well as inhibit incubation in male spotted sandpipers ( Actitis mecularia ), a sex role reversed bird (Oring et al. 1989) Effects of Methylmercury on White Ibis Populations I recorded dose related effects of MeHg exposure resulting in substantially decreased reproductive success in both ma le and female ibises as outlined above. Some of these effects would be exacerbated in a wild colony for the following reasons. In the aviary, as a result of constant unlimited food availability, the captive birds had an unnaturally long breeding season, typically having at least two breeding cycles per season, and not infrequently three or more. Though free living ibises do re nest following nest failure, it is not known in what proportion this occurs, and is believed to be small (Heath et al. 2009) Therefore, the 13 14% loss of production seen in all dosed groups (averaged across all three years) due mainly to male male nesting is a fairly conservative estimate since it takes multiple attempts within a season into consideration. Absence of multiple breeding cycles in wild colonies would also mean that male male pairing would result in more females left unpaire d than in the aviary. My attempt to quantify this showed that 22%, 28% and 30% of low, medium and high dose females were late nesters or did not nest at all in 2007 versus just 6% of control females. In 2008, 20% of high dose females were late nesters (n one of the other groups were simi larly constrained in 2008).

PAGE 162

162 I believe that loss of productivity due to male male pairing and decreased reproductive success in MeHg exposed females would be additive in a free living population. This seems reasonable since excluding females in reproductive condition from breeding (due to male unavailability) would directly reduce their output to zero, whilst MeHg induced reproductive impairment in exposed females who are breeding would also have a reduced output comparative to unexposed females. A conservative estimate would therefore mean 13% 14% loss of production due to male male pairing (including multiple nesting attempts) added to the loss due to reduced success in females would both contribute to reduced reproductive output (Table 6 1). This combined reduction shows that dosed groups could have a 15 50% lower reproductive output comparative to controls, which could mean a considerable impact at the population level. I propose to quantify these impact s more rigorous ly by using a population model which can take into account both male and female effects on population level parameters. Different models taking into account the variability seen in my study would be able to model different scenarios and would be useful in predicting population responses at different levels of exposure. This would also be useful for long term wildlife conservation and management plans in regions affected by Hg contamination. Conclusions I believe that my study provides valuable information with regard to conservation and management implications of Hg affected wildlife populations. I have found experimental evidence of MeHg induced reproductive impairment even at very low levels of exposure (0.05 0.3 ppm ww in diet). To my knowledge, thi s is the first avian experimental study finding evidence of impairment at such low exposure levels. In

PAGE 163

163 addition, this is the first evidence of Hg effects on sexual preference. It has been reported that the detectability of thresholds for sublethal endpoi nts are higher in laboratory experiments than in field studies, and that thresholds for life threatening endpoints seem to be lower in wild birds due to combined effects from multiple stressors (Spalding et al. 2000b) While the endpoints measured in my study were not of survival, they are integrated responses of several physiol ogical and behavioral pathways, all capable of being affected by stressors other than contaminant exposure. For example, reduced food availability is a severe limitation of reproduction in free living wading birds (Frederick and Ogden 2001, Gawlik 2002) which could be exacerbated by MeHg induced inhibition of reproduction and enhanced corticosterone responses. Ther efore, effects seen in my experiment are likely to be manifested in free living species at similar exposure levels, or at even lower levels due to interacting effects of other stressors in the field. My study can help establish lowest observed adverse ef fect levels (LOAELs) for white ibises at the very least, but can also provide information towards establishing similar points for other water birds. As mentioned above, this is the first experimental study to record adverse effects on reproduction at leve ls as low as 0.05 ppm ww in the where 0.5 ppm of dietary MeHg was the lowest dose level used (Rumbold et al. 2008) Anas platyrynchos ), a herbivorous bird also recognized to be relatively resistant to MeHg effects at the embryo stage (Heinz et al. 2009) There are speci es differences in MeHg metabolism for example in embryonic sensitivity (Heinz et al. 2009) and the ability of some species to

PAGE 164

164 demethylate MeHg to the less toxic elemental form (Spalding et al. 2000a, Henny et al. 2002) which will contribute to the uncertainty of establishing LOAELs for species other than the white ibis from the data reported here. Therefore, many more studies in different species are needed to establish general mechanisms and susceptibility of birds to MeHg exposure. In the specific example of the Everglades ecosy stem, the fact that the white ibis is the major contributor to the wading bird breeding population in the Everglades, and the fact that it has been recommended as an indicator species for monitoring restoration of the Everglades ecosystem (Frederick et al. 2009) underline the importance of knowing drivers of population processes for this species. Therefore, the information from my study can help in predicting population responses of changing MeHg availability in the white ibis in the Everglades. Though Hg levels are much lower now in Everglades wildlife, including in wading birds, than during levels of peak cont amination in the 1990s, there are still areas of Hg hotspots where exposure is estimated to be above currently acceptable levels (Rumbold et al. 2008) therefore predicting population responses to MeHg contamination remains important in this ecosystem. In addition, Hg availability is known to respond dynamically to hydrological changes such as hydro period and inter drought intervals, suggesting that Hg availability should be considered strongly in hydrological restoration plans. MeHg contamination remains a problem in other parts of North America as well, with both aquatic and terrestrial avian species being chronically exposed through their diet (Driscoll et al. 2007, Evers et al. 2007, Cristol et al. 2008) Increasing Hg emissions in Asia, which contrib ute to both local and global deposition, are a threat to ecosystems

PAGE 165

165 worldwide (Wu et al. 2006, Zhang and Won g 2007, Jaffe and Strode 2008) Therefore findings from my study are relevant to conservation and management implications of Hg affected wildlife beyond the context of white ibises in the Florida Everglades.

PAGE 166

166 Table 6 1. Estimated reduction in reproductive output of dosed birds due to male male pairing (unproductive nesting) and decreased fledgling production in females in comparison to the control group. Treatment Reduction due to unproductive nesting Reduction due to female impairment Total reduction Lo w 13.2 33.5 46.7 Medium 14.6 NA 14.6 High 13.5 35.4 48.9 Note that the medium dose group did not show differences in fledgling production of females in comparison to control females.

PAGE 167

167 Figure 6 1. Possible pathways leading to altered sexual preferenc e and male male pairing following methylmercury exposure.

PAGE 168

168 APPENDIX A MEAN REPRODUCTIVE PA RAMETERS Table A 1. Mean clutch sizes, hatchability rates and fledging rates for each treatment. Metric Year Control Low Medium High breeding attempt Mean S. E. Mean S.E. Mean S.E. Mean S.E. Uncorrected clutch size 2006 attempt 1 2.30 0.21 1.29 0.29 2.11 0.31 2.13 0.35 2007 attempt 1 2.82 0.33 2.56 0.32 2.88 0.15 2.87 0.24 2007 attempt 2 2.64 0.32 2.58 0.31 2.43 0.23 2.73 0.24 2008 attempt 1 3.88 0.46 3.78 0.39 3.50 0.22 4.21 0.42 2008 attempt 2 3.06 0.37 3.00 0.40 3.35 0.34 2.69 0.30 Hatch date clutch size 2006 attempt 1 1.80 0.36 0.86 0.34 1.78 0.40 1.63 0.42 2007 attempt 1 2.00 0.30 2.13 0.34 1.76 0.28 2.40 0.32 2007 a ttempt 2 2.07 0.27 1.83 0.37 1.79 0.26 1.64 0.36 2008 attempt 1 2.65 0.19 2.39 0.30 2.67 0.28 2.21 0.39 2008 attempt 2 1.59 0.32 1.06 0.31 1.29 0.33 1.81 0.34 Uncorrected 2006 attempt 1 0.48 0.14 0.33 0.21 0.54 0.12 0.55 0.16 hatcha bility 2007 attempt 1 0.58 0.10 0.43 0.10 0.42 0.07 0.76 0.09 2007 attempt 2 0.56 0.11 0.59 0.13 0.51 0.09 0.47 0.14 2008 attempt 1 0.76 0.07 0.53 0.07 0.69 0.08 0.50 0.09 2008 attempt 2 0.42 0.09 0.20 0.07 0.37 0.10 0.44 0.11 Hat ch date hatchability 2006 attempt 1 0.67 0.15 0.50 0.29 0.69 0.12 0.78 0.16 2007 attempt 1 0.77 0.08 0.55 0.10 0.70 0.09 0.90 0.05 2007 attempt 2 0.69 0.10 0.83 0.12 0.78 0.09 0.73 0.14 2008 attempt 1 0.94 0.06 0.81 0.06 0.91 0.04 0.77 0.08 2008 attempt 2 0.91 0.05 0.61 0.14 0.93 0.05 0.68 0.12 Uncorrected fledging 2006 attempt 1 0.03 0.03 0.08 0.08 0.04 0.04 0.10 0.10 rate 2007 attempt 1 0.08 0.04 0.04 0.03 0.09 0.04 0.07 0.04 2007 attempt 2 0.23 0.09 0.08 0.04 0.28 0.08 0.23 0.10 2008 attempt 1 0.09 0.04 0.09 0.03 0.21 0.05 0.08 0.04 2008 attempt 2 0.10 0.04 0.01 0.01 0.01 0.01 0.11 0.07 Hatch date fledging 2006 attempt 1 0.04 0.04 0.13 0.13 0.05 0.05 0.17 0.17 rate 2007 attempt 1 0.13 0.06 0.06 0.04 0.14 0.06 0.10 0.06 2007 attempt 2 0.27 0.10 0.10 0.05 0.42 0.11 0.38 0.15 2008 attempt 1 0.11 0.05 0.13 0.04 0.29 0.06 0.12 0.05 2008 attempt 2 0.29 0.10 0.06 0.06 0.04 0.04 0.14 0.08 40 day fledging rate 2006 attempt 1 0.08 0.08 0.25 0.25 0.14 0.14 0.20 0.20 2007 attempt 1 0.22 0.11 0.13 0.10 0.19 0.09 0.14 0.08 2007 attempt 2 0.41 0.13 0.11 0.06 0.56 0.13 0.52 0.18 2008 attempt 1 0.11 0.05 0.16 0.05 0.32 0.07 0.19 0.09 2008 attempt 2 0.32 0.10 0.14 0.14 0.06 0.06 0.18 0.11 Note: S.E.: Standard error

PAGE 169

169 Figure A 1. Treatment assignments in the aviary for each breeding season. C: Control; L: Low; M: Medium; H: High.

PAGE 170

170 APPENDIX B FECAL CONCENTRATIONS OF ESTRADIOL AND TES TOSTERONE METABOLITE S IN WHITE IB ISES Table B 1. Mean and standard errors (S.E.) of fecal estradiol metabolites in female ibises in the first breeding attempt, 2007 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E Pre breeding 9072.7 770.4 8514.3 1044.5 6919.4 592.3 9742.0 1639.7 Display 8996.2 1301.6 10500.6 1361.0 8117.3 739.9 6776.1 581.4 Nest building 16590.6 2740.4 19204.8 7039.9 17433.1 3164.6 16681.4 3576.8 Egg laying 19867.7 6447.4 11129.5 1031.2 10783 .3 1682.0 14452.0 3306.9 Incubation 9241.2 760.2 12632.4 1798.8 7038.3 783.8 9729.1 842.5 Chick rearing 13956.8 1810.5 11575.8 1269.1 10663.5 941.3 11776.6 1451.5 Note: Concentrations are in picograms per gram of feces, dry weight. Table B 2. Mean an d standard errors (S.E.) of fecal estradiol metabolites in female ibises in the second breeding attempt, 2007 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Pre breeding 21100. 3 3915.8 17544.5 3626.5 14100.6 2868.1 12660.7 2843.8 Display 23713.7 4296.4 25164.5 7750.7 19820.2 3268.5 13890.7 2238.3 Nest building 18116.6 1061.6 20217.8 3459.2 22100.5 2217.5 21855.6 NA Egg laying 20413.3 3906.2 12627.2 2297.9 13115.0 1916.2 16880 .7 5186.6 Incubation 11048.7 1068.3 9398.8 1004.9 8163.0 454.3 9547.7 1250.4 Chick rearing 9343.2 1287.2 10583.2 1259.1 9085.6 641.0 10111.4 2376.0 Note: Concentrations are in picograms per gram of feces, dry weight Table B 3. Mean and standard errors (S.E.) of fecal estradiol metabolites in female ibises in the first breeding attempt, 2008 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Display 15110.8 2149.6 10463.8 848.5 12596.4 1274.5 13549.3 1488.5 Nest building 39538.1 7780.6 38747.4 6434.9 42962.3 10504.4 45858.9 8514.3 Egg laying 36416.6 5880.7 28754.4 5135.8 43419.9 10907.0 71412.4 19778.9 Incubation 22315.4 2435.1 18377.7 979.6 17320.7 2037.9 17622.0 1295.1 Chic k rearing 20574.9 1175.7 22775.8 4671.1 18701.0 2560.5 20570.2 1585.8 Note: Concentrations are in picograms per gram of feces, dry weight

PAGE 171

171 Table B 4. Mean and standard errors (S.E.) of fecal estradiol metabolites in male ibises in the first breeding attem pt, 2007 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Pre breeding 6396.6 589.3 4697.2 573.3 7030.7 801.4 6814.6 891.2 Display 5694.5 659.9 5090.2 644.9 4910.5 510.0 4179.3 428.7 Nest building 7947.2 1657.5 4381.0 393.0 5560.5 499.0 8147.1 1550.4 Egg laying 5777.2 545.6 10963.6 2552.7 7320.7 1212.5 5425.5 784.8 Incubation 5763.7 441.0 6340.1 748.6 5797.4 590.0 6705.0 812.4 Chick rearing 7262.3 634.6 13651.1 3519.3 8330.0 2062.4 9608.4 2135.7 Note: Concentrations are in picograms per gram of feces, dry weight Table B 5. Mean and standard errors (S.E.) of fecal estradiol metabolites in male ibises in the second breeding attempt, 2007 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Pre breeding 6444.0 549.5 7497.3 1274.7 6852.4 649.6 5958.1 731.7 Display 10790.5 2420.7 6646.3 851.7 9583.6 1094.8 6749.7 1633.7 Nest building 31173.5 11798.5 8245. 8 1275.0 10151.9 2341.1 8094.4 1509.9 Egg laying 7905.8 533.8 8719.8 1260.4 6645.3 564.4 10845.3 1955.5 Incubation 8990.9 885.9 9765.6 809.2 7571.6 433.3 6335.1 538.2 Chick rearing 8267.4 717.3 8556.5 1128.6 8839.0 822.7 8376.4 1011.2 Note: Concentrat ions are in picograms per gram of feces, dry weight Table B 6. Mean and standard errors (S.E.) of fecal estradiol metabolites in male ibises in the first breeding attempt, 2008 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Display 6051.1 277.5 7421.7 501.8 7331.9 692.0 9387.0 1106.3 Nest building 29769.4 9645.0 23784.0 5632.1 12232.2 1363.4 15603.7 2552.0 Egg laying 12965.2 1057.6 16827.8 2350.2 11813.8 903.5 8964.1 488.9 I ncubation 14379.7 1415.6 12270.6 667.3 11241.8 731.4 10239.8 482.7 Chick rearing 13193.8 596.4 12851.6 529.6 12963.9 440.8 15372.0 1151.6 Note: Concentrations are in picograms per gram of feces, dry weight

PAGE 172

172 Table B 7. Mean and standard errors (S.E.) of fecal testosterone metabolites in female ibises in the first breeding attempt, 2007 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Pre breeding 51.0 12.4 39.6 3.2 41.2 3.2 41. 4 3.5 Display 49.1 4.2 49.6 3.6 39.1 2.1 37.9 2.1 Nest building 45.6 2.9 55.1 17.5 56.8 8.6 46.2 3.1 Egg laying 49.9 9.9 35.9 1.4 36.8 2.5 39.5 3.0 Incubation 40.9 2.7 43.7 3.8 36.8 2.4 37.3 3.6 Chick rearing 39.0 4.5 34.4 2.0 35.6 4.9 34.1 3.0 Note: Concentrations are in nanograms per gram of feces, dry weight Table B 8. Mean and standard errors (S.E.) of fecal testosterone metabolites in female ibises in the second breeding attempt, 2007 for each treatment group and breeding stage. Breeding Contr ol Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Pre breeding 49.9 5.0 36.4 6.5 36.0 4.2 45.7 3.8 Display 53.3 8.6 47.0 6.0 49.6 9.8 46.3 3.9 Nest building 52.2 7.8 54.2 9.3 47.6 5.0 63.2 NA Egg laying 49.1 3.2 43.1 6.0 43.2 3.4 40.7 2 .6 Incubation 44.7 5.0 46.2 5.8 49.3 5.6 41.2 1.9 Chick rearing 59.5 5.2 44.7 5.1 43.1 2.8 56.5 6.8 Note: Concentrations are in nanograms per gram of feces, dry weight Table B 9. Mean and standard errors (S.E.) of fecal testosterone metabolites in fem ale ibises in the first breeding attempt, 2008 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Display 40.1 4.4 35.9 4.1 30.4 1.6 31.9 2.2 Nest building 43.9 4.2 45.2 3.4 50.9 12.6 41.6 3.8 Egg laying 36.5 2.2 34.8 3.3 36.7 3.2 49.3 7.9 Incubation 32.9 2.5 34.1 1.3 36.3 2.1 37.9 2.3 Chick rearing 27.9 1.7 25.4 2.2 28.8 0.7 32.9 1.8 Note: Concentrations are in nanograms per gram of feces, dry weight

PAGE 173

173 Table B 10. Mean and standard errors (S.E.) of fecal testosterone metabolites in male ibises in the first breeding attempt, 2007 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Pre breeding 46.6 4.1 39.3 3.6 40.8 3.5 50.5 12.2 Display 51.1 5.3 40.1 2.6 48.2 4.3 40.0 2.5 Nest building 56.8 9.1 49.4 4.5 49.1 5.6 49.3 3.5 Egg laying 43.6 4.7 48.7 5.1 52.2 9.7 35.6 2.5 Incubation 38.3 2.6 44.8 4.6 46.6 3.9 40.2 2.8 Chick rearing 35.1 4.4 51.4 10.0 3 1.7 1.8 39.2 6.3 Note: Concentrations are in nanograms per gram of feces, dry weight Table B 11. Mean and standard errors (S.E.) of fecal testosterone metabolites in male ibises in the second breeding attempt, 2007 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Pre breeding 46.3 8.6 51.2 11.3 42.9 12.9 20.0 4.7 Display 70.4 8.9 62.0 12.4 67.3 8.8 82.5 18.4 Nest building 92.2 28.5 61.1 5.9 46.8 4.3 47.4 8.0 Egg laying 48.8 3.2 42.7 4.0 40.6 3.8 32.9 3.2 Incubation 54.5 4.9 51.8 4.2 57.7 11.5 38.1 1.7 Chick rearing 55.6 5.3 43.8 3.6 42.8 5.0 48.6 6.9 Note: Concentrations are in nanograms per gram of feces, dry weight Table B 12. Mean and standard errors (S.E.) of fecal t estosterone metabolites in male ibises in the first breeding attempt, 2008 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Display 36.6 3.0 39.1 2.3 41.1 3.2 33.2 1.6 Nest buil ding 93.5 19.6 86.2 16.1 54.5 4.4 74.9 15.5 Egg laying 36.0 2.2 49.6 4.5 46.7 6.3 38.4 2.8 Incubation 32.8 1.4 37.1 1.5 37.7 2.3 37.3 1.9 Chick rearing 28.3 1.8 27.2 2.3 28.0 1.0 30.6 1.7 Note: Concentrations are in nanograms per gram of feces, dry we ight

PAGE 174

174 APPENDIX C FECAL CONCENTRATIONS OF CORTICOSTERONE ME TABOLITES IN WHITE IBISES Table C 1. Mean and standard errors (S.E.) of fecal corticosterone metabolites in female ibises in the first breeding attempt, 2007 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Pre breeding 10876.5 454.0 9988.8 355.1 11341.5 1205.0 10438.9 356.3 Display 8899.1 344.4 9496.7 381.4 8907.5 309.6 8515.4 248.3 Nest building 8966.3 546.8 7334.8 1 179.7 9455.6 527.4 8735.9 347.2 Egg laying 8236.0 296.7 7570.4 211.3 8356.3 315.1 8653.6 457.4 Incubation 8224.5 187.6 8347.2 475.9 8097.7 317.5 8060.4 217.7 Chick rearing 10590.4 984.4 9578.5 642.9 10232.1 793.6 9894.1 519.5 Note: Concentrations are in nanograms per gram of feces, dry weight Table C 2. Mean and standard errors (S.E.) of fecal corticosterone metabolites in female ibises in the second breeding attempt, 2007 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Pre breeding 10171.2 701.5 9269.3 676.4 9054.9 457.7 9264.9 608.8 Display 10598.5 764.9 9868.7 443.1 8776.5 521.0 9385.2 694.1 Nest building 8494.4 44.9 10043.4 953.0 9433.4 738.7 13174.0 NA Egg laying 970 3.1 349.5 8950.5 457.5 9201.7 427.3 9932.9 1306.8 Incubation 9746.5 434.1 8897.1 368.0 9655.0 493.9 9403.4 344.9 Chick rearing 12454.2 1148.7 11901.2 501.6 11299.3 535.1 9520.8 623.3 Note: Concentrations are in nanograms per gram of feces, dry weight T able C 3. Mean and standard errors (S.E.) of fecal corticosterone metabolites in female ibises in the first breeding attempt, 2008 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E Display 13134.3 361.7 11821.1 129.7 12442.7 648.5 12881.4 448.9 Nest building 7191.7 607.5 8233.3 848.4 8273.4 789.4 7246.4 720.6 Egg laying 9374.2 864.0 8784.2 853.6 7938.3 535.7 8522.3 785.9 Incubation 9517.1 643.5 11205.0 392.7 8992.8 629.1 10575. 1 412.9 Chick rearing 12496.2 233.4 12130.4 242.4 11579.8 316.5 11781.6 527.9 Note: Concentrations are in nanograms per gram of feces, dry weight

PAGE 175

175 Table C 4. Mean and standard errors (S.E.) of fecal corticosterone metabolites in female ibises during the period of food restriction, 2008 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Nest building 12417.7 357.0 12950.3 299.1 13393.9 834.7 14006.1 1534.6 Egg laying 12686.6 723.0 13645.1 302.5 13758.0 NA 13238.8 560.3 Incubation 14173.0 657.9 12803.7 625.8 13485.4 302.8 13810.7 366.2 Note: Concentrations are in nanograms per gram of feces, dry weight Table C 5. Mean and standard errors (S.E.) of fecal corticosterone metabolite s in male ibises in the first breeding attempt, 2007 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Pre breeding 10788.6 265.7 10301.8 322.5 10606.4 344.7 10366.5 626.2 Displa y 9602.2 548.2 9129.6 415.1 10205.7 310.9 10046.7 348.9 Nest building 9614.9 829.0 9224.2 357.5 9726.9 441.3 10103.0 762.7 Egg laying 8656.5 368.5 8766.2 408.1 8469.0 234.9 8149.8 320.1 Incubation 8568.1 227.1 8558.9 355.5 9268.9 521.5 8157.0 212.7 Chi ck rearing 10223.7 630.3 10745.7 1351.9 9501.5 345.4 9254.3 501.2 Note: Concentrations are in nanograms per gram of feces, dry weight Table C 6. Mean and standard errors (S.E.) of fecal corticosterone metabolites in male ibises in the second breeding at tempt, 2007 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Pre breeding 8789.1 576.5 8439.3 607.0 9099.0 788.2 8233.3 339.1 Display 9911.9 538.6 8963.8 417.5 9627.3 425.5 9090 .4 638.3 Nest building 13612.8 1919.8 9639.2 306.0 8827.7 343.8 9144.4 662.0 Egg laying 10089.4 484.1 9342.5 218.5 9115.4 499.3 7776.1 215.7 Incubation 10641.1 371.0 10115.0 533.5 8884.7 374.1 8871.6 391.6 Chick rearing 12747.5 540.0 12726.4 1410.3 119 96.0 829.3 11102.0 1051.1 Note: Concentrations are in nanograms per gram of feces, dry weight

PAGE 176

176 Table C 7. Mean and standard errors (S.E.) of fecal corticosterone metabolites in male ibises in the first breeding attempt, 2008 for each treatment group a nd breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Display 12503.3 241.0 11933.8 162.4 13167.9 397.7 13143.4 495.1 Nest building 10472.1 1468.6 9195.6 808.0 8624.9 589.2 7159.4 393.0 Egg laying 8513.2 506 .6 9082.7 641.2 8743.8 607.4 7366.6 808.9 Incubation 9987.4 570.4 10229.8 491.3 9901.4 528.9 9712.7 656.0 Chick rearing 12820.6 301.2 12267.0 343.0 11927.0 412.3 11988.9 550.9 Note: Concentrations are in nanograms per gram of feces, dry weight Tab le C 8. Mean and standard errors (S.E.) of fecal corticosterone metabolites in male ibises during the period of food restriction, 2008 for each treatment group and breeding stage. Breeding Control Low Medium High stage Mean S.E. Mean S.E. Mean S.E. Mean S.E. Nest building 14171.6 621.9 13546.0 427.9 13542.2 401.1 13267.5 936.1 Egg laying 12703.2 364.9 13263.2 199.9 13732.7 1172.2 14122.7 728.3 Incubation 13337.8 237.9 13247.3 358.2 13853.6 360.9 14223.8 639.5 Note: Concentrations are in nanograms per gra m of feces, dry weight

PAGE 177

177 LIST OF REFERENCES Adams, E. M., P. C. Frederick, I. L. V. Larkin, and L. J. Guillette. 2009. Sublethal Effects of Methylmercury on Fecal Metabolites of Testosterone, Estradiol, and Corticosterone in Captive Ju venile White Ibises ( Eudocimus Albus ). Environmental Toxicology and Chemistry 28 :982 989. Adkins Regan, E. 1988. Sex hormones and sexual orientation in animals. Psychobiology 16 :335 347. Adkins Regan, E. 1998. Hormonal mechanisms of mate choice. American Z oologist 38 :166 178. Adkins Regan, E. 1999. Testosterone increases singing and aggression but not male typical sexual partner preference in early estrogen treated female zebra finches. Hormones and Behavior 35 :63 70. Adkins Regan, E. 2005a. Hormones and An imal Social Behavior. Princeton University Press, Princeton. Adkins Regan, E. 2005b. Tactile contact is required for early estrogen treatment to alter the sexual partner preference of female zebra finches. Hormones and Behavior 48 :180 186. Adkins Regan, E. 2007. Hormones and the development of sex differences in behavior. Journal of Ornithology 148 :S17 S26. Adkins Regan, E. 2009. Hormones and Sexual Differentiation of Avian Social Behavior. Developmental Neuroscience 31 :342 350. Adkins Regan, E., and C. H. Leung. 2006. Sex steroids modulate changes in social and sexual preference during juvenile development in zebra finches. Hormones and Behavior 50 :772 778. Adkins Regan, E., V. Mansukhani, C. Seiwert, and R. Thompson. 1994. Sexual Differentiation of Brain a nd Behavior in the Zebra Finch Critical Periods for Effects of Early Estrogen Treatment. Journal of Neurobiology 25 :865 877. Adkins Regan, E., V. Mansukhani, R. Thompson, and S. Yang. 1997. Organizational actions of sex hormones on sexual partner prefere nce. Brain Research Bulletin 44 :497 502. Adkins Regan, E., and J. Wade. 2001. Masculinized Sexual Partner Preference in Female Zebra Finches with Sex Reversed Gonads. Hormones and Behavior 39 :22 28.

PAGE 178

178 Alonso Alvarez, C., and J. L. Tella. 2001. Effects of exp erimental food restriction and body mass changes on the avian T cell mediated immune response. Canadian Journal of Zoology Revue Canadienne De Zoologie 79 :101 105. Astheimer, L. B., W. A. Buttemer, and J. C. Wingfield. 1992. Interactions of Corticosterone with Feeding, Activity and Metabolism in Passerine Birds. Ornis Scandinavica 23 :355 365. Axelrod, J., and T. D. Reisine. 1984. Stress Hormones Their Interaction and Regulation. Science 224 :452 459. Bagemihl, B. 1999. Biological exuberance: animal homosex uality and natural diversity. Pages i xv, 1 751 in Biological exuberance: animal homosexuality and natural diversity. Profile Books Ltd. Bailey, N. W., and M. Zuk. 2009. Same sex sexual behavior and evolution. Trends in Ecology & Evolution 24 :439 446. Barr J. F. 1986. Population dynamics of the common loon ( Gavia immer ) associated with mercury contaminated waters in northwestern Ontario. Canadian Wildlife Service Occasional Paper:1 25. Bell, A. M. 2001. Effects of an endocrine disrupter on courtship and ag gressive behaviour of male three spined stickleback, Gasterosteus aculeatus Animal Behaviour 62 :775 780. Bergan, T., L. Gallardo, and H. Rodhe. 1999. Mercury in the global troposphere: a three dimensional model study. Atmospheric Environment 33 :1575 1585. Bhatnagar, S., and C. Vining. 2003. Facilitation of hypothalamic pituitary adrenal responses to novel stress following repeated social stress using the resident/intruder paradigm. Hormones and Behavior 43 :158 165. Bleau, H., C. Daniel, G. Chevalier, H. va nTra, and A. Hontela. 1996. Effects of acute exposure to mercury chloride and methylmercury on plasma cortisol, T3, T4, glucose and liver glycogen in rainbow trout ( Oncorhynchus mykiss ). Aquatic Toxicology 34 :221 235. Bokony, V., A. Z. Lendvai, A. Liker, F Angelier, J. C. Wingfield, and O. Chastel. 2009. Stress Response and the Value of Reproduction: Are Birds Prudent Parents? American Naturalist 173 :589 598. Bonier, F., P. R. Martin, I. T. Moore, and J. C. Wingfield. 2009. Do baseline glucocorticoids pred ict fitness? Trends in Ecology & Evolution 24 :634 642. Boonstra, R. 2004. Coping with changing northern environments: The role of the stress axis in birds and mammals. Integrative and Comparative Biology 44 :95 108.

PAGE 179

179 Boonstra, R., and G. R. Singleton. 1993. Population Declines in the Snowshoe Hare and the Role of Stress. General and Comparative Endocrinology 91 :126 143. Bouton, S. N., P. C. Frederick, M. G. Spalding, and H. McGill. 1999. Effects of chronic, low concentrations of dietary methylmercury on the b ehavior of juvenile great egrets. Environmental Toxicology and Chemistry 18 :1934 1939. Brasso, R. L., and D. A. Cristol. 2008. Effects of mercury exposure on the reproductive success of tree swallows ( Tachycineta bicolor ). Ecotoxicology 17 :133 141. Burgess N. M., and M. W. Meyer. 2008. Methylmercury exposure associated with reduced productivity in common loons. Ecotoxicology 17 :83 91. Burnham, K. P., and D. Anderson. 2002. Model Selection and Multi Model Inference: A Practical Information Theoretic Approac h, 2 edition. Springer Science, New York. Busch, D. S., and L. S. Hayward. 2009. Stress in a conservation context: A discussion of glucocorticoid actions and how levels change with conservation relevant variables. Biological Conservation 142 :2844 2853. But ler, T. J., M. D. Cohen, F. M. Vermeylen, G. E. Likens, D. Schmeltz, and R. S. Artz. 2008. Regional precipitation mercury trends in the eastern USA, 1998 2005: Declines in the Northeast and Midwest, no trend in the Southeast. Atmospheric Environment 42 :158 2 1592. Canario, J., M. Caetano, C. Vale, and R. Cesario. 2007. Evidence for elevated production of methylmercury in salt marshes. Environmental Science & Technology 41 :7376 7382. Clarkson, T. W. 1993. Mercury Major Issues in Environmental Health. Enviro nmental Health Perspectives 100 :31 38. Cleckner, L. B., C. C. Gilmour, J. P. Hurley, and D. P. Krabbenhoft. 1999. Mercury methylation in periphyton of the Florida Everglades. Limnology and Oceanography 44 :1815 1825. Clotfelter, E. D., A. M. Bell, and K. R. Levering. 2004. The role of animal behaviour in the study of endocrine disrupting chemicals. Animal Behaviour 68 :665 676. Condon, A. M., and D. A. Cristol. 2009. Feather Growth Influences Blood Mercury Level of Young Songbirds. Environmental Toxicology an d Chemistry 28 :395 401. Conover, M. R. 1983. Female Female Pairings in Caspian Terns. Condor 85 :346 349. Conover, M. R., and G. L. Hunt. 1984. Experimental Evidence That Female Female Pairs in Gulls Result from a Shortage of Breeding Males. Condor 86 :472 4 76.

PAGE 180

180 Conover, M. R., D. E. Miller, and G. L. Hunt. 1979. Female Female Pairs and Other Unusual Reproductive Associations in Ring Billed and California Gulls. Auk 96 :6 9. Crews, D., A. C. Gore, T. S. Hsu, N. L. Dangleben, M. Spinetta, T. Schallert, M. D. Anw ay, and M. K. Skinner. 2007. Transgenerational epigenetic imprints on mate preference. Proceedings of the National Academy of Sciences 104 :5942 5946. Crews, D., and M. C. Moore. 1986. Evolution of Mechanisms Controlling Mating Behavior. Science 231 :121 125 Cristol, D. A., R. L. Brasso, A. M. Condon, R. E. Fovargue, S. L. Friedman, K. K. Hallinger, A. P. Monroe, and A. E. White. 2008. The movement of aquatic Mercury through terrestrial food webs. Science 320 :335 335. Crump, K. L., and V. L. Trudeau. 2009. M ercury Induced Reproductive Impairment in Fish. Environmental Toxicology and Chemistry 28 :895 907. Custer, C. M., T. W. Custer, and E. F. Hill. 2007. Mercury exposure and effects on cavity nesting birds from the Carson River, Nevada. Archives of Environmen tal Contamination and Toxicology 52 :129 136. Cuvin Aralar, M. L. A., and R. W. Furness. 1991. Mercury and selenium interaction: A review. Ecotoxicology and Environmental Safety 21 :348 364. Cyr, N. E., and L. M. Romero. 2007. Chronic stress in free living E uropean starlings reduces corticosterone concentrations and reproductive success. General and Comparative Endocrinology 151 :82 89. Dallman, M. F. 1993. Stress Update Adaptation of the Hypothalamic Pituitary Adrenal Axis to Chronic Stress. Trends in Endoc rinology and Metabolism 4 :62 69. Day, L. B., J. T. McBroom, and B. A. Schlinger. 2006. Testosterone increases display behaviors but does not stimulate growth of adult plumage in male golden collared manakins ( Manacus vitellinus ). Hormones and Behavior 49 :2 23 232. Drevnick, P. E., and M. B. Sandheinrich. 2003. Effects of dietary methylmercury on reproductive endocrinology of fathead minnows. Environmental Science & Technology 37 :4390 4396. Drevnick, P. E., M. B. Sandheinrich, and J. T. Oris. 2006. Increased ovarian follicular apoptosis in fathead minnows ( Pimephales promelas ) exposed to dietary methylmercury. Aquatic Toxicology 79 :49 54. Driscoll, C. T., Y. J. Han, C. Y. Chen, D. C. Evers, K. F. Lambert, T. M. Holsen, N. C. Kamman, and R. K. Munson. 2007. Mer cury contamination in forest and freshwater ecosystems in the Northeastern United States. Bioscience 57 :17 28.

PAGE 181

1 81 Eagles Smith, C. A., J. T. Ackerman, J. Yee, and T. L. Adelsbach. 2009. Mercury Demethylation in Waterbird Livers: Dose Response Thresholds and D ifferences among Species. Environmental Toxicology and Chemistry 28 :568 577. Evers, D. C., Y. J. Han, C. T. Driscoll, N. C. Kamman, M. W. Goodale, K. F. Lambert, T. M. Holsen, C. Y. Chen, T. A. Clair, and T. Butler. 2007. Biological mercury hotspots in the northeastern United States and southeastern Canada. Bioscience 57 :29 43. Evers, D. C., O. P. Lane, L. Savoy, and W. Goodale. 2004. Assessing the impacts of methylmercury on piscivorous wildlife using a wildlife criterion value based on the Common Loon, 19 98 2003. Report to Maine Department of Environmental Protection no. BRI 2004 05. BioDiversity Research Institute, Gorham, Maine. Evers, D. C., L. J. Savoy, C. R. DeSorbo, D. E. Yates, W. Hanson, K. M. Taylor, L. S. Siegel, J. H. Cooley, M. S. Bank, A. Maj or, K. Munney, B. F. Mower, H. S. Vogel, N. Schoch, M. Pokras, M. W. Goodale, and J. Fair. 2008. Adverse effects from environmental mercury loads on breeding common loons. Ecotoxicology 17 :69 81. Fernie, K. J., J. L. Shutt, R. J. Letcher, J. I. Ritchie, K. Sullivan, and D. M. Bird. 2008. Changes in reproductive courtship behaviors of adult American kestrels ( Falco sparverius ) exposed to environmentally relevant levels of the polybrominated diphenyl ether mixture, DE 71. Toxicological Sciences 102 :171 178. F isher, S. A., G. R. Bortolotti, K. J. Fernie, J. E. Smits, T. A. Marchant, K. G. Drouillard, and D. M. Bird. 2001. Courtship behavior of captive American kestrels ( Falco sparverius ) exposed to polychlorinated biphenyls. Archives of Environmental Contaminat ion and Toxicology 41 :215 220. Franceschini, M. D., O. P. Lane, D. C. Evers, J. M. Reed, B. Hoskins, and L. M. Romero. 2009. The corticosterone stress response and mercury contamination in free living tree swallows, Tachycineta bicolor Ecotoxicology 18 :51 4 521. Frederick, P., D. E. Gawlik, J. C. Ogden, M. I. Cook, and M. Lusk. 2009. The White Ibis and Wood Stork as indicators for restoration of the everglades ecosystem. Ecological Indicators 9 :S83 S95. Frederick, P. C. 1987a. Extrapair Copulations in the M ating System of White Ibis ( Eudocimus a lbus ). Behaviour 100 :170 201. Frederick, P. C. 1987b. Mating strategies of White Ibis. University of North Carolina. Frederick, P. C. 1999. Mercury contamination and its effects in the Everglades ecosystem. Reviews in Toxicology 3 :213 255. Frederick, P. C., and M. W. Collopy. 1989. Nesting Success of 5 Ciconiiform Species in Relation to Water Conditions in the Florida Everglades. Auk 106 :625 634.

PAGE 182

182 Frederick, P. C., B. Hylton, J. A. Heath, and M. G. Spalding. 2004. A his torical record of mercury contamination in southern Florida (USA) as inferred from avian feather tissue. Environmental Toxicology and Chemistry 23 :1474 1478. Frederick, P. C., and J. C. Ogden. 2001. Pulsed breeding of long legged wading birds and the impor tance of infrequent severe drought conditions in the Florida Everglades. Wetlands 21 :484 491. Frederick, P. C., M. G. Spalding, and R. Dusek. 2002. Wading birds as bioindicators of mercury contamination in Florida, USA: Annual and geographic variation. Env ironmental Toxicology and Chemistry 21 :163 167. Frederick, P. C., M. G. Spalding, M. S. Sepulveda, G. E. Williams, L. Nico, and R. Robins. 1999. Exposure of great egret ( Ardea albus ) nestlings to mercury through diet in the Everglades ecosystem. Environmen tal Toxicology and Chemistry 18 :1940 1947. Fry, D. M., and C. K. Toone. 1981. DDT Induced Feminization of Gull Embryos. Science 213 :922 924. Fusani, L. 2008. Testosterone control of male courtship in birds. Hormones and Behavior 54 :227 233. Fusani, L., L. B. Day, V. Canoine, D. Reinemann, E. Hernandez, and B. A. Schlinger. 2007. Androgen and the elaborate courtship behavior of a tropical lekking bird. Hormones and Behavior 51 :62 68. Fusani, L., and J. B. Hutchison. 2003. Lack of changes in the courtship beh aviour of male ring doves after testosterone treatment. Ethology Ecology & Evolution 15 :143 157. Gawlik, D. E. 2002. The effects of prey availability on the numerical response of wading birds. Ecological Monographs 72 :329 346. Gilmour, C. C., and E. A. Hen ry. 1991. Mercury Methylation in Aquatic Systems Affected by Acid Deposition. Environmental Pollution 71 :131 169. Gilmour, C. C., E. A. Henry, and R. Mitchell. 1992. Sulfate Stimulation of Mercury Methylation in Fresh Water Sediments. Environmental Science & Technology 26 :2281 2287. Gonzalez, P., Y. Dominique, J. C. Massabuau, A. Boudou, and J. P. Bourdineaud. 2005. Comparative effects of dietary methylmercury on gene expression in liver, skeletal muscle, and brain of the zebrafish ( Danio rerio ). Environmen tal Science & Technology 39 :3972 3980. Gorissen, L., T. Snoeijs, E. Van Duyse, and M. Eens. 2005. Heavy metal pollution affects dawn singing behaviour in a small passerine bird. Oecologia 145 :504 509.

PAGE 183

183 Grasman, K. A., P. F. Scanlon, and G. A. Fox. 1998. Rep roductive and physiological effects of environmental contaminants in fish eating birds of the Great Lakes: A review of historical trends. Environmental Monitoring and Assessment 53 :117 145. Grue, C. E., S. C. Garner, and P. L. Gilbert. 2002. On the signifi cance of pollutant induced alterations in the behaviour of fish and wildlife. Pages 1 90 in G. U.K. Guillette, L. J., and M. P. Gunderson. 2001. Alterations in development of reproductive and endocrine systems of wildlife populations exposed to endocrine disrupting contaminants. Reproduction 122 :857 864. Gustafsson, L., D. Nordling, M. S. Andersson, B. C. Sheldon, and A. Qvarnstrom. 1994. Infectious Diseases, Reproductive Ef fort and the Cost of Reproduction in Birds. Philosophical Transactions of the Royal Society of London Series B Biological Sciences 346 :323 331. Haegele, M. A., and R. H. Hudson. 1977. Reduction of Courtship Behavior Induced by DDE in Male Ringed Turtle Dov es. Wilson Bulletin 89 :593 601. Hallinger, K. K., D. J. Zabransky, K. A. Kazmer, and D. A. Cristol. 2010. Birdsong Differs between Mercury Polluted and Reference Sites. Auk 127 :156 161. Hawley, D. M., K. K. Hallinger, and D. A. Cristol. 2009. Compromised i mmune competence in free living tree swallows exposed to mercury. Ecotoxicology 18 :499 503. Hayes, T. B., A. Collins, M. Lee, M. Mendoza, N. Noriega, A. A. Stuart, and A. Vonk. 2002. Hermaphroditic, demasculinized frogs after exposure to the herbicide atra zine at low ecologically relevant doses. Proceedings of the National Academy of Sciences of the United States of America 99 :5476 5480. Heath, J. A. 2002. White Ibis ( Eudocimus albus ) Reproductive Physiology. Ph.D. University of Florida, Gainesville, FL. He ath, J. A., and P. C. Frederick. 2005. Relationships among mercury concentrations, hormones, and nesting effort of White Ibises ( Eudocimus albus ) in the Florida Everglades. Auk 122 :255 267. Heath, J. A., P. C. Frederick, T. M. Edwards, and L. J. Guillette. 2003. Reproductive physiology of free living White Ibises ( Eudocimus albus ) in the Florida Everglades. General and Comparative Endocrinology 133 :118 131. Heath, J. A., P. C. Frederick, J. A. Kushlan, and K. L. Bildstein. 2009. White Ibis ( Eudocimus albus ) The Birds of North America Online (A. Poole, Ed.). Ithaca: Cornell Lab of Ornithology; Retrieved from the Birds of North America Online: http://bna.birds.cornell.edu/bna/species/009

PAGE 184

184 Heinz, G. H. 1976a. Methylmercury 2nd Generation Reproductive and Behavioral Effects on Mallard Ducks. Journal of Wildlife Management 40 :710 715. Heinz, G. H. 1976b. Methylmercury 2nd Year Feeding Effects on Mallard Reproduction and Duckling Behavior. Journal o f Wildlife Management 40 :82 90. Heinz, G. H. 1979. Methylmercury Reproductive and Behavioral Effects on 3 Generations of Mallard Ducks. Journal of Wildlife Management 43 :394 401. Heinz, G. H., and D. J. Hoffman. 2003a. Embryotoxic thresholds of mercury: Estimates from individual mallard eggs. Archives of Environmental Contamination and Toxicology 44 :257 264. Heinz, G. H., and D. J. Hoffman. 2003b. Predicting mercury in mallard ducklings from mercury in chorioallantoic membranes. Bulletin of Environmental Contamination and Toxicology 70 :1242 1246. Heinz, G. H., D. J. Hoffman, J. D. Klimstra, K. R. Stebbins, S. L. Kondrad, and C. A. Erwin. 2009. Species Differences in the Sensitivity of Avian Embryos to Methylmercury. Archives of Environmental Contamination and Toxicology 56 :129 138. Heinz, G. H., and L. N. Locke. 1976. Brain Lesions in Mallard Ducklings from Parents Fed Methylmercury. Avian Diseases 20 :9 17. Henny, C. J., E. F. Hill, R. A. Grove, and J. L. Kaiser. 2007. Mercury and drought along the lower Ca rson River, Nevada: I. Snowy egret and black crowned night heron annual exposure to mercury, 1997 2006. Archives of Environmental Contamination and Toxicology 53 :269 280. Henny, C. J., E. F. Hill, D. J. Hoffman, M. G. Spalding, and R. A. Grove. 2002. Ninet eenth century mercury: Hazard to wading birds and cormorants of the Carson River, Nevada. Ecotoxicology 11 :213 231. Hill, E. F., C. J. Henny, and R. A. Grove. 2008. Mercury and drought along the lower Carson River, Nevada: II. Snowy egret and black crowned night heron reproduction on Lahontan Reservoir, 1997 2006. Ecotoxicology 17 :117 131. Hirschenhauser, K., E. Mostl, B. Wallner, J. Dittami, and K. Kotrschal. 2000. Endocrine and behavioural responses of male greylag geese ( Anser anser ) to pairbond challeng es during the reproductive season. Ethology 106 :63 77. Homma Takeda, S., Y. Kugenuma, T. Iwamuro, Y. Kumagai, and N. Shimojo. 2001. Impairment of spermatogenesis in rats by methylmercury: involvement of stage and cell specific germ cell apoptosis. Toxicol ogy 169 :25 35.

PAGE 185

185 Hontela, A., J. B. Rasmussen, C. Audet, and G. Chevalier. 1992. Impaired Cortisol Stress Response in Fish f rom Environments Polluted by PAHs, PCB s, and Mercury. Archives of Environmental Contamination and Toxicology 22 :278 283. Huber, R., an d M. Martys. 1993. Male Male Pairs in Greylag Geese ( Anser a nser ). Journal Fur Ornithologie 134 :155 164. Hunt, G. L., and M. W. Hunt. 1977. Female Female Pairing in Western Gulls ( Larus o ccidentalis ) in Southern California. Science 196 :1466 1467. Hurlbert, S. H. 1984. Pseudoreplication and the Design of Ecological Field Experiments. Ecological Monographs 54 :187 211. Husak, J. F., and I. T. Moore. 2008. Stress hormones and mate choice. Trends in Ecology & Evolution 23 :532 534. Hutchison, J. B. 1970. Differen tial Effects of Testosterone and Oestradiol on Male Courtship in Barbary Doves ( Streptopelia r isoria ). Animal Behaviour 18 :41 &. Jacobs, J. D., and J. C. Wingfield. 2000. Endocrine control of life cycle stages: A constraint on response to the environment? Condor 102 :35 51. Jaffe, D., and S. Strode. 2008. Sources, fate and transport of atmospheric mercury from Asia. Environmental Chemistry 5 :121 126. Kenow, K. P., K. A. Grasman, R. K. Hines, M. W. Meyer, A. Gendron Fitzpatrick, M. G. Spalding, and B. R. Gray 2007. Effects of methylmercury exposure on the immune function of juvenile common loons ( Gavia immer ). Environmental Toxicology and Chemistry 26 :1460 1469. Ketterson, E. D., V. Nolan, and M. Sandell. 2005. Testosterone in females: Mediator of adaptive tr aits, constraint on sexual dimorphism, or both? American Naturalist 166 :S85 S98. Kirubagaran, R., and K. P. Joy. 1995. Changes in lipid profiles and 32P uptake into phosphoprotein (Vitellogenin) content of the ovary and liver in the female catfish, Clarias batrachus exposed to mercury. Biomedical and Environmental Sciences 8 :35 44. Kitaysky, A. S., J. F. Piatt, and J. C. Wingfield. 2007. Stress hormones link food availability and population processes in seabirds. Marine Ecology Progress Series 352 :245 258. Kitaysky, A. S., J. C. Wingfield, and J. F. Piatt. 1999. Dynamics of food availability, body condition and physiological stress response in breeding Black legged Kittiwakes. Functional Ecology 13 :577 584.

PAGE 186

186 Klaper, R., B. J. Carter, C. A. Richter, P. E. Dre vnick, M. B. Sandheinrich, and D. E. Tillitt. 2008. Use of a 15 k gene microarray to determine gene expression changes in response to acute and chronic methylmercury exposure in the fathead minnow Pimephales promelas Rafinesque. Journal of Fish Biology 72 : 2207 2280. Klaper, R., C. B. Rees, P. Drevnick, D. Weber, M. Sandheinrich, and M. J. Carvan. 2006. Gene expression changes related to endocrine function and decline in reproduction in fathead minnow ( Pimephales promelas ) after dietary methylmercury exposur e. Environmental Health Perspectives 114 :1337 1343. Krebs, C. J., R. Boonstra, S. Boutin, and A. R. E. Sinclair. 2001. What drives the 10 year cycle of snowshoe haves? Bioscience 51 :25 35. Kushlan, J. A. 1979. Feeding Ecology and Prey Selection in the Whit e Ibis. Condor 81 :376 389. Lee, H. G., Y. C. Kim, J. S. Dunning, and K. A. Han. 2008. Recurring Ethanol Exposure Induces Disinhibited Courtship in Drosophila. Plos One 3 Lewis, S. A., P. H. Becker, and R. W. Furness. 1993. Mercury Levels in Eggs, Tissues, and Feathers of Herring Gulls Larus Argentatus from the German Wadden Sea Coast. Environmental Pollution 80 :293 299. Liu, G. L., Y. Cai, Y. X. Mao, D. Scheidt, P. Kalla, J. Richards, L. J. Scinto, G. Tachiev, D. Roelant, and C. Appleby. 2009. Spatial Vari ability in Mercury Cycling and Relevant Biogeochemical Controls in the Florida Everglades. Environmental Science & Technology 43 :4361 4366. Liu, T., L. Dartevelle, C. Y. Yuan, H. P. Wei, Y. Wang, J. F. Ferveur, and A. Guo. 2008. Increased dopamine level en hances male male courtship in Drosophila Journal of Neuroscience 28 :5539 5546. Lochmiller, R. L., M. R. Vestey, and J. C. Boren. 1993. Relationship between Protein Nutritional Status and Immunocompetence in Northern Bobwhite Chicks. Auk 110 :503 510. Loftu s, W. F. 2000. Accumulation and fate of mercury in an Everglades aquatic food web. Ph.D. Florida International University, Miami, FL. Lopez Rull, I., and D. Gil. 2009. Elevated testosterone levels affect female breeding success and yolk androgen deposition in a passerine bird. Behavioural Processes 82 :312 318. Love, O. P., C. W. Breuner, F. Vezina, and T. D. Williams. 2004. Mediation of a corticosterone induced reproductive conflict. Hormones and Behavior 46 :59 65.

PAGE 187

187 MacFarlane, G. R., S. P. Blomberg, G. Kapl an, and L. J. Rogers. 2007. Same sex sexual behavior in birds: expression is related to social mating system and state of development at hatching. Behavioral Ecology 18 :21 33. Mansukhani, V., E. AdkinsRegan, and S. Yang. 1996. Sexual partner preference in female zebra finches: The role of early hormones and social environment. Hormones and Behavior 30 :506 513. Martin, T. E. 1987. Food as a Limit on Breeding Birds a Life History Perspective. Annual Review of Ecology and Systematics 18 :453 487. Mason, R. R. M. L. Abbot, R. A. Bodaly, O. R. Bullock, C. T. Driscoll, D. Evers, S. E. Lindberg, M. Murray, and E. B. Swain. 2005. Monitoring the response to changing mercury deposition. Environmental Science & Technology 39 :14A 22A. McCarty, J. P., and A. L. Secord. 1999a. Nest building behavior in PCB contaminated tree swallows. Auk 116 :55 63. McCarty, J. P., and A. L. Secord. 1999b. Reproductive ecology of tree swallows ( Tachycineta bicolor ) with high levels of polychlorinated biphenyl contamination. Environmental Toxicology and Chemistry 18 :1433 1439. McVey, M. J., G. M. Cooke, I. H. A. Curran, H. M. Chan, S. Kubow, E. Lok, and R. Mehta. 2008. An investigation of the effects of methylmercury in rats fed different dietary fats and proteins: Testicular steroidogenic enzymes and serum testosterone levels. Food and Chemical Toxicology 46 :270 279. Meddle, S. L., L. M. Romero, L. B. Astheimer, W. A. Buttemer, I. T. Moore, and J. C. Wingfield. 2002. Steroid hormone interrelationships with territorial aggression in an arcti c breeding songbird, Gambel's white crowned sparrow, Zonotrichia leucophrys gambelii Hormones and Behavior 42 :212 221. Merritt, K. A., and A. Amirbahman. 2009. Mercury methylation dynamics in estuarine and coastal marine environments A critical review. Earth Science Reviews 96 :54 66. Meyer, M. W., D. C. Evers, T. Daulton, and W. E. Braselton. 1995. Common loons ( Gavia immer ) nesting on low ph lakes in northern Wisconsin have elevated blood mercury content. Water, Air, & Soil Pollution 80 :871 880. Milnes, M. R., D. S. Bermudez, T. A. Bryan, T. M. Edwards, M. P. Gunderson, I. L. V. Larkin, B. C. Moore, and L. J. Guillette. 2006. Contaminant induced feminization and demasculinization of nonmammalian vertebrate males in aquatic environments. Environmental Res earch 100 :3 17. Moller, A. P., P. Christe, J. Erritzoe, and J. Mavarez. 1998. Condition, disease and immune defence. Oikos 83 :301 306.

PAGE 188

188 Monteiro, L. R., and R. W. Furness. 2001. Kinetics, dose response, and excretion of methylmercury in free living adult Co ry's shearwaters. Environmental Science & Technology 35 :739 746. Moore, M. C., and R. Kranz. 1983. Evidence for Androgen Independence of Male Mounting Behavior in White Crowned Sparrows ( Zo notrichia l eucophrys g ambelii ). Hormones and Behavior 17 :414 423. N aef Daenzer, B., F. Widmer, and M. Nuber. 2001. Differential post fledging survival of great and coal tits in relation to their condition and fledging date. Journal of Animal Ecology 70 :730 738. Nelson, R. J. 2005. An Introduction to Behavioral Endocrinolo gy, Third edition. Sinauer Associates, Sunderland, MA Nocera, J. J., and P. D. Taylor. 1998. In situ behavioral response of common loons associated with elevated mercury (Hg) exposure. Conservation Ecology 2 Norris, D. O. 2000. Endocrine disrupters of the stress axis in natural populations: How can we tell? American Zoologist 40 :393 401. Norris, D. O., S. Donahue, R. M. Dores, J. K. Lee, T. A. Maldonado, T. Ruth, and J. D. Woodling. 1999. Impaired adrenocortical response to stress by brown trout, Salmo tru tta living in metal contaminated waters of the Eagle River, Colorado. General and Comparative Endocrinology 113 :1 8. O'Neal, D. M., D. G. Reichard, K. Pavilis, and E. D. Ketterson. 2008. Experimentally elevated testosterone, female parental care, and repr oductive success in a songbird, the Dark eyed Junco ( Junco hyemalis ). Hormones and Behavior 54 :571 578. Oring, L. W., A. J. Fivizzani, and M. E. Elhalawani. 1989. Testosterone Induced Inhibition of Incubation in the Spotted Sandpiper ( Actitis m ecularia ). H ormones and Behavior 23 :412 423. Oskam, I. C., E. Ropstad, E. Lie, A. E. Derocher, O. Wiig, E. Dahl, S. Larsen, and J. U. Skaare. 2004. Organochlorines affect the steroid hormone cortisol in free ranging polar bears ( Ursus maritimus ) at Svalbard, Norway. J ournal of Toxicology and Environmental Health Part a Current Issues 67 :959 977. Ottinger, M. A., E. Lauoie, N. Thompson, A. Barton, K. Whitehouse, M. Barton, M. Abdelnabi, M. Quinn, G. Panzica, and C. Viglietti Panzica. 2008. Neuroendocrine and behavioral effects of embryonic exposure to endocrine disrupting chemicals in birds. Brain Research Reviews 57 :376 385. Ottinger, M. A., E. T. Lavoie, N. Thompson, M. Bohannon, K. Dean, and M. J. Quinn. 2009. Is the gonadotropin releasing hormone system vulnerable to endocrine disruption in birds? General and Comparative Endocrinology 163 :104 108.

PAGE 189

189 Owens, I. P. F., and R. V. Short. 1995. Hormonal Basis of Sexual Dimorphism in Birds Implications for New Theories of Sexual Selection. Trends in Ecology & Evolution 10 :44 47. Palme, R. 2005. Measuring fecal steroids Guidelines for practical application. Bird Hormones and Bird Migrations: Analyzing Hormones in Droppings and Egg Yolks and Assessing Adaptations in Long Distance Migration 1046 :75 80. Peakall, D., and J. Burg er. 2003. Methodologies for assessing exposure to metals: speciation, bioavailability of metals, and ecological host factors. Ecotoxicology and Environmental Safety 56 :110 121. Peakall, D. B. 1996. Disrupted patterns of behavior in natural populations as a n index of ecotoxicity. Environmental Health Perspectives 104 :331 335. Perry, W. B. 2008. Everglades restoration and water quality challenges in south Florida. Ecotoxicology 17 :569 578. Pilsner, J. R., A. L. Lazarus, D. H. Nam, R. J. Letcher, C. Sonne, R. Dietz, and N. Basu. 2010. Mercury associated DNA hypomethylation in polar bear brains via the LUminometric Methylation Assay: a sensitive method to study epigenetics in wildlife. Molecular Ecology 19 :307 314. Pinheiro, J., D. Bates, S. DebRoy, D. Sarkar, a nd the R. C ore T eam. 2009. nlme: Linear and Nonlinear Mixed Effects Mo dels. R package version 3.1 96 Pinheiro, J. C., and D. M. Bates. 2000. Mixed Effects Models in S and S Plus Springer Verlag New York, Inc. R Development Core Team. 2009. R: A Language a nd Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3 900051 07 0, URL http://www.R project.org Rice, D. C., and S. G. Gilbert. 1982. Early Chronic Low Level Me thylmercury Poisoning in Monkeys Impairs Spatial Vision. Science 216 :759 761. Rice, D. C., and S. G. Gilbert. 1992. Exposure to Methyl Mercury from Birth to Adulthood Impairs High Frequency Hearing in Monkeys. Toxicology and Applied Pharmacology 115 :6 10. Rich, E. L., and L. M. Romero. 2005. Exposure to chronic stress downregulates corticosterone responses to acute stressors. American Journal of Physiology Regulatory Integrative and Comparative Physiology 288 :R1628 R1636. Rimmer, C. C., K. P. McFarland, D. C. Evers, E. K. Miller, Y. Aubry, D. Busby, and R. J. Taylor. 2005. Mercury concentrations in Bicknell's thrush and other insectivorous passerines in Montane forests of northeastern North America. Ecotoxicology 14 :223 240.

PAGE 190

190 Roberts, M. L., K. L. Buchanan, A T. D. Bennett, and M. R. Evans. 2007. Mate choice in zebra finches: does corticosterone play a role? Animal Behaviour 74 :921 929. Rogers, C. M., M. Ramenofsky, E. D. Ketterson, V. Nolan, and J. C. Wingfield. 1993. Plasma Corticosterone, Adrenal Mass, Win ter Weather, and Season in Nonbreeding Populati ons of Dark Eyed Juncos ( Junco hyemalis h yemalis ). Auk 110 :279 285. Romero, L. M. 2004. Physiological stress in ecology: lessons from biomedical research. Trends in Ecology & Evolution 19 :249 255. Rudegeair, T 1975. The reproductive behavior and ecology of the White Ibis ( Eudocimus albus ). Ph.D. University of Florida, Gainesville, FL. Rumbold, D. G., and L. E. Fink. 2006. Extreme spatial variability and unprecedented methylmercury concentrations within a const ructed wetland. Environmental Monitoring and Assessment 112 :115 135. Rumbold, D. G., T. R. Lange, D. M. Axelrad, and T. D. Atkeson. 2008. Ecological risk of methylmercury in Everglades National Park, Florida, USA. Ecotoxicology 17 :632 641. Ryder, J. P., an d P. L. Somppi. 1979. Female Female Pairing in Ring Billed Gulls. Auk 96 :1 5. Saaristo, M., J. A. Craft, K. K. Lehtonen, and K. Lindstrom. 2009. Sand goby ( Pomatoschistus minutus ) males exposed to an endocrine disrupting chemical fail in nest and mate comp etition. Hormones and Behavior 56 :315 321. Saino, N., S. Calza, and A. P. Moller. 1997. Immunocompetence of nestling barn swallows in relation to brood size and parental effort. Journal of Animal Ecology 66 :827 836. Sandheinrich, M. B., and K. M. Miller. 2 006. Effects of dietary methylmercury on reproductive behavior of fathead minnows ( Pimephales promelas ). Environmental Toxicology and Chemistry 25 :3053 3057. Sapolsky, R. M., L. M. Romero, and A. U. Munck. 2000. How do glucocorticoids influence stress resp onses? Integrating permissive, suppressive, stimulatory, and preparative actions. Endocrine Reviews 21 :55 89. Scheuhammer, A. M. 1987. The Chronic Toxicity of Aluminum, Cadmium, Mercury, and Lead in Birds a Review. Environmental Pollution 46 :263 295. Sch euhammer, A. M., M. W. Meyer, M. B. Sandheinrich, and M. W. Murray. 2009. Effects of Environmental Methylmercury on the Health of Wild Birds, Mammals, and Fish. AMBIO: A Journal of the Human Environment 36 :12 19.

PAGE 191

191 Scheuhammer, A. M., and M. B. Sandheinrich. 2008. Recent advances in the toxicology of methylmercury in wildlife. Ecotoxicology 17 :67 68. Scheuhammer, A. M., A. H. K. Wong, and D. Bond. 1998. Mercury and selenium accumulation in common loons ( Gavia immer ) and common mergansers ( Mergus merganser ) fr om eastern Canada. Environmental Toxicology and Chemistry 17 :197 201. Schoech, S. J., M. A. Rensel, E. S. Bridge, R. K. Boughton, and T. E. Wilcoxen. 2009. Environment, glucocorticoids, and the timing of reproduction. General and Comparative Endocrinology 163 :201 207. Schroeder, W. H., and J. Munthe. 1998. Atmospheric mercury An overview. Atmospheric Environment 32 :809 822. Selin, N. E. 2009. Global Biogeochemical Cycling of Mercury: A Review. Annual Review of Environment and Resources 34 :43 63. Selin, N. E., D. J. Jacob, R. J. Park, R. M. Yantosca, S. Strode, L. Jaegle, and D. Jaffe. 2007. Chemical cycling and deposition of atmospheric mercury: Global constraints from observations. Journal of Geophysical Research Atmospheres 112 Sepulveda, M. S., P. C. F rederick, M. G. Spalding, and G. E. Williams. 1999a. Mercury contamination in free ranging great egret nestlings ( Ardea albus ) from southern Florida, USA. Environmental Toxicology and Chemistry 18 :985 992. Sepulveda, M. S., G. E. Williams, P. C. Frederick, and M. G. Spalding. 1999b. Effects of mercury on health and first year survival of free ranging great egrets ( Ardea albus ) from southern Florida. Archives of Environmental Contamination and Toxicology 37 :369 376. Sheriff, M. J., C. J. Krebs, and R. Boonst ra. 2009. The sensitive hare: sublethal effects of predator stress on reproduction in snowshoe hares. Journal of Animal Ecology 78 :1249 1258. Shugart, G. W. 1980. Frequency and Distribution of Polygyny in Great Lakes Herring Gulls in 1978. Condor 82 :426 42 9. Silverin, B. 1998. Behavioural and hormonal responses of the pied flycatcher to environmental stressors. Animal Behaviour 55 :1411 1420. Simons, L. S., and T. E. Martin. 1990. Food Limitation of Avian Reproduction an Experiment with the Cactus Wren. Ec ology 71 :869 876. Spalding, M. G., R. D. Bjork, G. V. N. Powell, and S. F. Sundlof. 1994. Mercury and Cause of Death in Great White Herons. Journal of Wildlife Management 58 :735 739.

PAGE 192

192 Spalding, M. G., P. C. Frederick, H. C. McGill, S. N. Bouton, and L. R. M cDowell. 2000a. Methylmercury accumulation in tissues and its effects on growth and appetite in captive great egrets. Journal of Wildlife Diseases 36 :411 422. Spalding, M. G., P. C. Frederick, H. C. McGill, S. N. Bouton, L. J. Richey, I. M. Schumacher, C. G. M. Blackmore, and J. Harrison. 2000b. Histologic, neurologic, and immunologic effects of methylmercury in captive great egrets. Journal of Wildlife Diseases 36 :423 435. St. Louis, V. L., J. W. M. Rudd, C. A. Kelly, K. G. Beaty, N. S. Bloom, and R. J. Fl ett. 1994. Importance of Wetlands as Sources of Methyl Mercury to Boreal Forest Ecosystems. Canadian Journal of Fisheries and Aquatic Sciences 51 :1065 1076. Streets, D. G., Q. Zhang, and Y. Wu. 2009. Projections of Global Mercury Emissions in 2050. Environ mental Science & Technology 43 :2983 2988. Sundlof, S. F., M. G. Spalding, J. D. Wentworth, and C. K. Steible. 1994. Mercury in Livers of Wading Birds (Ciconiiformes) in Southern Florida. Archives of Environmental Contamination and Toxicology 27 :299 305. Ta n, S. W., J. C. Meiller, and K. R. Mahaffey. 2009. The endocrine effects of mercury in humans and wildlife. Critical Reviews in Toxicology 39 :228 269. Thompson, D. R. 1996. Mercury in birds and terrestrial mammals. Pages 341 356 in Environmental contaminan ts in wildlife: Interpreting tissue concentrations. CRC Press, Inc.; CRC Press. Toft, G., and J. L. J. Guillette. 2005. Decreased sperm count and sexual behavior in mosquitofish exposed to water from a pesticide contaminated lake. Ecotoxicology and Environ mental Safety 60 :15 20. Tomaszycki, M. L., S. B. Banerjee, and E. Adkins Regan. 2006. The role of sex steroids in courtship, pairing and pairing behaviors in the socially monogamous zebra finch. Hormones and Behavior 50 :141 147. Tsipoura, N., J. Burger, R. Feltes, J. Yacabucci, D. Mizrahi, C. Jeitner, and M. Gochfeld. 2008. Metal concentrations in three species of passerine birds breeding in the Hackensack Meadowlands of New Jersey. Environmental Research 107 :218 228. Ullrich, S. M., T. W. Tanton, and S. A. Abdrashitova. 2001. Mercury in the aquatic environment: A review of factors affecting methylation. Critical Reviews in Environmental Science and Technology 31 :241 293. Wada, H., D. A. Cristol, F. M. A. McNabb, and W. A. Hopkins. 2009a. Suppressed Adrenoco rtical Responses and Thyroid Hormone Levels in Birds near a Mercury Contaminated River. Environmental Science & Technology 43 :6031 6038.

PAGE 193

193 Wada, H., D. A. Cristol, F. M. A. McNabb, and W. A. Hopkins. 2009b. Suppressed Adrenocortical Responses and Thyroid Hor mone Levels in Birds near a Mercury Contaminated River. Environmental Science & Technology. Weis, J. S. 2009. Reproductive, Developmental, and Neurobehavioral Effects of Methylmercury in Fishes. Journal of Environmental Science and Health Part C Environmen tal Carcinogenesis & Ecotoxicology Reviews 27 :212 225. Weis, J. S., G. Smith, T. Zhou, C. Santiago Bass, and P. Weis. 2001. Effects of contaminants on behavior: Biochemical mechanisms and ecological consequences. Bioscience 51 :209 217. Weis, J. S., G. M. S mith, and T. Zhou. 1999. Altered predator prey behavior in polluted environments: implications for fish conservation. Environmental Biology of Fishes 55 :43 51. Welshons, W. V., K. A. Thayer, B. M. Judy, J. A. Taylor, E. M. Curran, and F. S. vom Saal. 2003. Large effects from small exposures. I. Mechanisms for endocrine disrupting chemicals with estrogenic activity. Environmental Health Perspectives 111 :994 1006. Wikelski, M., L. M. Romero, and H. L. Snell. 2001. Marine iguanas oiled in the Galapagos. Scienc e 292 :437 438. Wikelski, M., V. Wong, B. Chevalier, N. Rattenborg, and H. L. Snell. 2002. Marine iguanas die from trace oil pollution. Nature 417 :607 608. Wingfield, J. C. 2005. The concept of allostasis: Coping with a capricious environment. Journal of Ma mmalogy 86 :248 254. Wingfield, J. C., and K. E. Hunt. 2002. Arctic spring: hormone behavior interactions in a severe environment. Comparative Biochemistry and Physiology B Biochemistry & Molecular Biology 132 :275 286. Wingfield, J. C., and A. S. Kitaysky. 2002. Endocrine responses to unpredictable environmental events: Stress or anti stress hormones? Integrative and Comparative Biology 42 :600 609. Wingfield, J. C., S. E. Lynn, and K. K. Soma. 2001. Avoiding the 'costs' of testosterone: Ecological bases of h ormone behavior interactions. Brain Behavior and Evolution 57 :239 251. Wingfield, J. C., D. L. Maney, C. W. Breuner, J. D. Jacobs, S. Lynn, M. Ramenofsky, and R. D. Richardson. 1998. Ecological bases of hormone behavior interactions: The "emergency life hi story stage". American Zoologist 38 :191 206.

PAGE 194

194 Wingfield, J. C., A. Newman, G. L. Hunt, and D. S. Farner. 1980. Androgen in High Concentrations in the Blood of Female Western Gulls, Larus o ccidentalis Naturwissenschaften 67 :514 515. Wingfield, J. C., K. M. Oreilly, and L. B. Astheimer. 1995. Modulation of the Adrenocortical Responses to Acute Stress in Arctic Birds a Possible Ecological Basis. American Zoologist 35 :285 294. Wingfield, J. C., and R. M. Sapolsky. 2003. Reproduction and resistance to stress: When and how. Journal of Neuroendocrinology 15 :711 724. Wingfield, J. C., and B. Silverin. 1986. Effects of Corticosterone on Territorial Behavior of Free Living Male Song Sparrows Melospiza Melodia. Hormones and Behavior 20 :405 417. Wolfe, M. F., S. Schwa rzbach, and R. A. Sulaiman. 1998. Effects of mercury on wildlife: A comprehensive review. Environmental Toxicology and Chemistry 17 :146 160. Wu, Y., S. X. Wang, D. G. Streets, J. M. Hao, M. Chan, and J. K. Jiang. 2006. Trends in anthropogenic mercury emiss ions in China from 1995 to 2003. Environmental Science & Technology 40 :5312 5318. Young, L. C., B. J. Zaun, and E. A. VanderWerf. 2008. Successful same sex pairing in Laysan albatross. Biology Letters 4 :323 325. Young, W. C., R. W. Goy, and C. H. Phoenix. 1964. Hormones and Sexual Behavior Science 143 :212 218. Zala, S. M., and D. J. Penn. 2004. Abnormal behaviours induced by chemical pollution: a review of the evidence and new challenges. Animal Behaviour 68 :649 664. Zann, R. A. 1994. Reproduction in a Zebr a Finch Colony in South Eastern Australia the Significance of Monogamy, Precocial Breeding and Multiple Broods in a Highly Mobile Species. Emu 94 :285 299. Zhang, L., and M. H. Wong. 2007. Environmental mercury contamination in China: Sources and impacts. Environment International 33 :108 121. Zillioux, E. J., D. B. Porcella, and J. M. Benoit. 1993. Mercury Cycling and Effects in Fresh Water Wetland Ecosystems. Environmental Toxicology and Chemistry 12 :2245 2264.

PAGE 195

195 BIOGRAPHICAL SKETCH Nilmini Jayasena was born in 1978 in Kandy, Sri Lanka. She received her Bachelor of Veterinary Science degree from the University of Peradeniya, Sri Lanka prior to obtaining her Ph.D. at the University of Florida She has been interested in wildlife and natural history from an early age, and hopes to continue doing research in the field of wildlife ecology. Nilmini joined the Faculty of Veterinary Medicine and Animal Science at the University of Peradeniya Sri Lanka upon completion of her Ph.D