Title: Reproductive endocrinology of the Florida manatee (Trichechus manatus latirostris)
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00100693/00001
 Material Information
Title: Reproductive endocrinology of the Florida manatee (Trichechus manatus latirostris) estrous cycles, seasonal patterns and behavior
Physical Description: Book
Language: English
Creator: Larkin, Iskande Lieve Vandevelde, 1969-
Publisher: State University System of Florida
Place of Publication: Florida
Florida
Publication Date: 2000
Copyright Date: 2000
 Subjects
Subject: Veterinary Medicine thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Veterinary Medicine -- UF   ( lcsh )
Genre: government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )
 Notes
Summary: ABSTRACT: The study of Florida manatee reproduction is important to determine correctly their actual reproductive potential in establishing population models, making management decisions, identifying sensitive periods of the year, and recognizing differences between healthy and unhealthy states related to reproduction. The objectives of this study were to measure fecal steroid hormone concentrations (17b-estradiol, progesterone and testosterone) with a fecal radioimmunoassay to 1) determine the length and frequency of estrous cycles in female manatees, 2) identify seasonal hormone fluctuations, 3) correlate hormone concentrations with behavior, 4) correlate hormone concentrations with reproductive tissues, 5) determine if food availability and water temperature affect seasonal fluctuations. Elephants are used as a model from which to pose hypotheses. The results indicate that the mean gut transit time for the Florida manatee is 6-8 days. This suggests a time delay of approximately one-week between fecal hormone concentrations and plasma hormone concentrations. The current technique can not distinguish different reproductive groups (e.g. pregnant vs. non-pregnant, adult vs. calf), but it may be able to distinguish gender among adult manatees. The estrous cycle of the Florida manatee is approximately 28-42 days in length. Seasonal fluctuations of captive manatee hormone concentrations suggest slight peaks during the spring and fall. Data from wild manatees were non-significant; thus comparisons between wild and captive animals were inconclusive. The majority of behavioral data analyses were non-significant; however, a small number of behaviors were correlated with fluctuations in hormone concentrations. Vulva swelling and a behavior described as female mounting were correlated with increased estradiol and/or decreased progesterone concentrations. A model for manatee reproductive patterns and future research objective are presented.
Summary: KEYWORDS: Florida manatee, Trichechus manatus latirostris, fecal radioimmunoassay, estradiol, testosterone, progesterone
Thesis: Thesis (Ph. D.)--University of Florida, 2000.
Bibliography: Includes bibliographical references (p. 319-338).
System Details: System requirements: World Wide Web browser and PDF reader.
System Details: Mode of access: World Wide Web.
Statement of Responsibility: by Iskande Lieve Vandevelde Larkin.
General Note: Title from first page of PDF file.
General Note: Document formatted into pages; contains xv, 339 p.; also contains graphics.
General Note: Vita.
 Record Information
Bibliographic ID: UF00100693
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 81388889
alephbibnum - 002639435
notis - ANA6262

Downloads

This item has the following downloads:

Dissertation ( PDF )


Full Text











REPRODUCTIVE ENDOCRINOLOGY OF THE FLORIDA MANATEE
(TRICHECHUSMANATUSLA TIROSTRIS): ESTROUS CYCLES, SEASONAL
PATTERNS AND BEHAVIOR

















By

ISKANDE LIEVE VANDEVELDE LARKIN


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


2000





















To Roger Reep, for inspiring and encouraging me;
my mom, for reminding me I have what it takes;
and my husband, who has shown me how to appreciate what life offers.















ACKNOWLEDGMENTS

I thank my committee members for guiding me through this process and

providing unique perspectives that broadened my learning and introduced me to new

ideas and ways of thinking. Roger Reep was a very reassuring mentor that proved to

be the calm in the midst of many storms. Tim Gross provided the bulk of my

guidance in radioimmunoassays and technical support, without which this project

could not have been accomplished. I thank Tom Vickroy for being there to answer

questions at a moment's notice and remind me to look at things objectively. Dan

Odell and John Eisenberg were figures of inspiration to learn about new species,

scientific subjects and perspectives I otherwise would have neglected. Bob Bonde

was an ever-cheerful face and allowed me to join him in the field to meet wild

manatees for the first time. Bob always kept me up to date on new observations in

the field and allowed me to attend many small meetings and workshops on manatees.

Last but not least, Lynn Lefebvre was there when I needed her to step in and join the

committee cheerfully.

I would like to give a special thanks to all the people and the many

organizations that came together in helping me accomplish this study because they

saw the importance in learning more about manatee reproduction. I thank our lab

technician, Maggie Stoll, who spent a whole year getting up at the wee hours of the

morning to collect fecals for me. At Homosassa Springs State Wildlife Park, I thank

Tom Linley for allowing us to conduct our study there, Sally Lieb for trusting that not









all graduate students are made the same, and the staff for being so friendly. Randy

Runnels and staff at SeaWorld Florida provided as much help as possible despite their

busy schedules. Mark Barringer and staff at the Living Seas in Epcot went above and

beyond any of my hopes and expectations by collecting all of the samples from each

of their three males for the duration of the study time. I thank Dr. Dave Murphy and

Debbie Halin for collecting samples from their manatees at Lowry Park Zoo in

Tampa for as long as they were able. Dr. Greg Bossart, Dr. Maya Menchaca, Nerva

and staff at Miami Seaquarium were very helpful with their advice and calls to update

me on the status of their animals and collections of samples. I am sorry we did not

have enough funds to analyze the many samples they collected for this study. I would

like to extend the same apology to Jessica Koelsch at Mote Marine Laboratories who

provided boundless energy in collecting many samples from the wild manatees in

Sarasota Bay. Finally, I would like to thank the many veterinary students, teachers

and volunteers that that joined me in my behavioral observations, fecal collections

and lab work. Seeing their excitement in my work was most rewarding.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ........................................................................................ iii

LIST O F TA B LE S ............................................................................................viii

L IS T O F F IG U R E S .................................................................................................. ix

A B S T R A C T .................................................................................... x iv

1 INTRODUCTION AND BACKGROUND REVIEW ........................... ............... 1

In tro d u c tio n ........................................................................................................... 1
O objectives .......................................................................................................... 2
H ypotheses ........................................................................................................ 3
B background R eview .................................................................................. .. 5
C conservation ...................................................................................................... 5
Evolution .............................................................................. .. ..............6
D distribution ....................................................................................................... 9
M anatee R eproduction........................................................................ ............. 10
Other Mammalian Reproductive Patterns .......................................... ............. 14
Elephant R production ................................................................................... 21
Seasonality ....................................................................................... 29
General m am m alian patterns........................................................ .............. 29
Seasonality in Sirenians ............................................................................. 32
M easurem ent of H orm ones .............................................................. .............. 33
Captive Versus W ild Conditions.................................................................... 34

2 FECAL RADIOIMMUNOASSAY VALIDATION FOR THE FLORIDA
M AN A TEE .................................................................................................... 36

Introduction ......................................................................................................... 36
M materials and M ethods ..................................................................... ..............38
Subjects ......................................................................... .... . .. .............. 38
Fecal Collections ........................................................................................... 40
Experiment One: Gut Transit Time Collections................................ .............. 41
Experiment Two: Fecal Handling Collections.................................. ..............43
Experiment Three: Plasma and Necropsy Collections .................................... 44
R adioim m unoassay ........................................................................................ 45



v









17 3p-estradiol .................................................................... ..................... 47
Progesterone ........................................................................................ ... 48
Testosterone............................................. 49
A analysis of H orm onal Param eters................................................ .............. 51
R e su lts ......................................................................... . .................................. 5 3
Experiment One: Fecal Marker/Gut Transit Times..................................... 53
Experim ent Tw o: Fecal H handling ......................................... ..................... 54
Experiment Three: Fecal Comparisons with Plasma and Tissue Samples ......... 55
Plasm a and fecal ............................................................. ............. 55
F ecal and necrop sy tissu e ............................................................ .............. 58
Experiment Four: A Comparison of Gender, Location, Age and Individuals .... 59
G ender and location ................................................... .............. .......... 59
A g e .......................................................................................................... .. 6 1
Discussion ................ ......... ....... .. .. ......................... 62
Experiment One: Fecal Marker/Gut Transit Time ...................................... 62
Experim ent Tw o: Fecal H handling ......................................... ..................... 65
Experiment Three: Fecal Comparisons with Plasma and Tissue Samples ......... 67
Experiment Four: A Comparison of Gender, Location, Age and Individuals .... 70
C o n clu sio n s ......................................................................................................... 7 2

3 SEASONAL AND ESTROUS PATTERNS OF THE FLORIDA MANATEE .... 94

In tro d u c tio n ......................................................................................................... 9 4
M methods ........................ . ...... ......... ...... .......... .... ... .......... . 97
Experiment Five: Fecal Collections for Estrous Cycle Data ............................. 97
Experiment Six: Fecal Collections for Seasonality Data............................. 98
A n a ly sis .......................................................................................................... 9 8
Results .............................. .. ...... ..... ... ......... ....... .............. 99
Experiment Five: Individual Hormone Profiles and Estrous Patterns............. 99
Individual horm one profiles ....................... ......................................... 99
E strous patterns.................................................................................... 100
Experim ent Six: Seasonality ....... ......... ............ ................... 103
Seasonality at different locations...... ..... ..................... 103
Seasonality in individual manatees...................................... 105
D discussion .................... .. ......... .. ..... ........................ 107
Experiment Five: Individual Hormone Profiles and Estrous Patterns........... 107
Individual horm one profiles ....... ....... ......... ................... 107
E strous patterns.................................................................................... 109
Experim ent Six: Seasonality ....... ......... ............ ................... 110
Seasonality at different locations...... .... ..................... 110
C conclusions .............................................................. ......................... ............ 111

4 MANATEE REPRODUCTIVE BEHAVIOR......................................... 172

In tro d u c tio n ....................................................................................................... 1 7 2
M eth o d s .......................................................................................... .......... 17 4
Experiment Seven: Behavioral Data Collection..................... ................ 174


vi









Social interaction ..................................................................... ............. 177
O their interaction ...................................................................... ............. 178
O their ....................................................................... ..... . ............. 178
A analysis ......................................................................................................... 179
R e su lts .................. ...... ...... ....................... ................................................... 1 8 1
Experiment Seven: Behavioral Data........... ........ ................... 181
General Observations of Individual Manatees ........................................... 182
D discussion .......................................................................... ............. 184
Experiment Seven: Behavioral Data......................... .................... 184
General Observations of Individual Manatees....................... ................ 187
C o n c lu sio n s ...................................................................................................... 1 8 8

5 D ISC U S SIO N ........................................................ 190

In tro d u c tio n ........................................................................................ ............. 1 9 0
Elephants as a Model for Manatee Reproduction ........................ ................ 191
R eview of D ata............................................................................................. 195
C chapter T w o .......................................................................................... 195
C chapter T hree ......................................................................................... 198
Chapter Four ................................................................................ ............ 199
Results of Elephant M odel...... ................................. ................... 200
M anatee M odel ............................................................................................. 206
Future R research ............................................................................................ 208

APPENDIX A ASSAY TRIALS & TRIBULATIONS...........................213

APPENDIX B HORMONE CONCENTRATIONS...............................219

APPENDIX C BEHAVIOR DATA ................................. ................ 250

LIST OF REFERENCES ..................................................... ............321

BIOGRAPHICAL SKETCH............................................................ 339















LIST OF TABLES



Table Page

2-1. Values of normality characteristics, before and after log transformation of fecal
hormone concentrations. ........... .... ............... ............... 74

2-2. Gut transit and retention time of colored corn fed to adult manatees.................75

2-3. Lengths and weights of gut transit time manatees .................................... 76

2-4. Handling study normalized hormone concentrations. ................................ 77

2-5. Necropsied female manatee measurements of reproductive tracts, body length and
hormone concentrations. ........... .... ............... ............... 83

3-1. Background information of captive manatees ............................ ............ 113















LIST OF FIGURES


Figure Page

1-1. Serum hormonal changes in the rat during the estrous cycle, metestrus (Met),
diestrus (Di), proestrus (Pro) and estrus (Est) ...................... ........... 17

1-2. Serum hormonal changes in the rhesus monkey during the estrous cycle. ........... 19

1-3. Serum hormonal changes in the unmated cat during polyestrus cyclic activity ...... 19

1-4. Serum hormonal changes in the African elephant ......................... .............22

1-5. Seasonal plasma progesterone concentrations in the mare. ...........................31

2-1. Matching fecal and plasma concentrations of progesterone, estradiol, and
testosterone for individual m anatees ..................... ......... ...............78

2-2. Mean plasma concentrations of progesterone, estradiol and testosterone that have
matching fecal samples, grouped by reproductive state. ....................... 79

2-3. Mean fecal concentrations of progesterone, estradiol and testosterone that have
matching plasma samples, grouped by reproductive state. .....................80

2-4. A comparison between male and female plasma concentrations of progesterone,
estradiol, and testosterone. ........................................ ............ 81

2-5. Comparison between male and female fecal concentrations of progesterone, estradiol
and testosterone. .......... ....... .................... ............. 82

2-6. Fecal hormone concentration from necropsied animals. ...............................86

2-7. Male and female fecal concentrations of testosterone, progesterone and estradiol,
grouped by each location and gender ..................................... 87

2-8. Female and male fecal hormone comparisons of Progesterone (P), Estradiol (E) and
Testosterone (T), plus ratios of P/T, E/T and E/P. ........... ................. 88

2-9. Wild male and female hormone concentrations grouped by age and reproductive
status. ............ ...... ........................... ............. 90










2-10. Homosassa mean concentrations of estradiol and progesterone per individual
fem ale anim als. ........................................................ ............ 9 1

2-11. SeaWorld mean concentrations of estradiol and progesterone per individual
anim als. ................................................................. ........ . 92

2-12. Epcot and Lowry mean concentrations of testosterone per individual animals. .... 93

3-1. Amanda weekly mean estradiol and progesterone hormone concentrations. ....... 114

3-2. Ariel estradiol and progesterone hormone concentrations. ..........................115

3-3. Betsy estradiol and progesterone hormone concentrations. ....................... 116

3-4. Lorelei estradiol and progesterone hormone concentrations .........................117

3-5. Rachel estradiol and progesterone hormone concentrations ........................ 118

3-6. Rosie estradiol and progesterone hormone concentrations. ....................... 119

3-7. Star estradiol and progesterone hormone concentrations. .......................... 120

3-8. Charlotte estradiol and progesterone hormone concentrations. .................... 121

3-9. Georgia estradiol and progesterone hormone concentrations. ...................... 122

3-10. Rita estradiol and progesterone hormone concentrations. ......................... 123

3-11. Sara estradiol and progesterone hormone concentrations. ........................ 124

3-12. Stubbie estradiol and progesterone hormone concentrations. .................... 125

3-13. Amanda daily hormone concentrations of estradiol and progesterone. ........... 126

3-14. Rosie daily hormone concentrations of estradiol and progesterone. ............... 127

3-15. Charlotte daily hormone concentrations of estradiol and progesterone. ........... 128

3-16. Rita daily hormone concentrations of estradiol and progesterone. ................. 129

3-17. Amanda normalized estradiol and progesterone concentrations, indicating estradiol
peaks above 1STD ........ .............................. ......... ..... 130

3-18. Ariel normalized estradiol and progesterone concentrations, indicating estradiol
peaks above 1STD .......... ....... ............................. 131










3-19. Betsy normalized estradiol and progesterone concentrations, indicating estradiol
peaks above 1STD ........ .............................. ......... ..... 132

3-20. Rachel normalized estradiol and progesterone concentrations, indicating estradiol
peaks above 1STD ........ .............................. ......... ..... 133

3-21. Rosie normalized estradiol and progesterone concentrations, indicating estradiol
peaks above 1STD ........ .............................. ......... ..... 134

3-22. Star normalized estradiol and progesterone concentrations, indicating estradiol
peaks above 1STD ........ .............................. ......... ..... 135

3-23. Charlotte normalized estradiol and progesterone concentrations, indicating estradiol
peaks above 1STD ........ .............................. ......... ..... 136

3-24. Georgia normalized estradiol and progesterone concentrations, indicating estradiol
peaks above 1STD. .................. .......... .............. 137

3-25. Rita normalized estradiol and progesterone concentrations, indicating estradiol
peaks above 1STD. .................. .......... .............. 138

3-26. Sara normalized estradiol and progesterone concentrations, indicating estradiol
peaks above 1STD. .................. .......... .............. 139

3-27. Stubbie normalized estradiol and progesterone concentrations, indicating estradiol
peaks above 1STD. .................. .......... .............. 140

3-28. Amanda mean estrous cycle pattern. ................... .......................... 141

3-29. Ariel mean estrous cycle pattern. ......................... ......................... 142

3-30. Betsy mean estrous cycle pattern. ................ ... ............ ... .. .............. 143

3-31. Charlotte mean estrous cycle pattern ............. ... ........... ....... ........... 144

3-32. Georgia mean estrous cycle pattern. ............ ....... ........ ..... .......... 145

3-33. Rita mean estrous cycle pattern. ......... .. ......... ...... .... ............. 146

3-34. Sara mean estrous cycle pattern. ............... .... ................. ............... 147

3-35. Stubbie mean estrous cycle pattern. ............. ..... ...... 148

3-36. SeaWorld and Homosassa separate and combined mean estrous cycle patterns +
SEM .......... ........ ............................ ........... 149










3-37. Non-normalized SeaWorld and Homosassa separate and combined mean estrous
cycle patterns SE M .............................................. ............ 150

3-38. The number of days between peaks above the mean for each female at Homosassa
and SeaWorld, with the frequency of the time periods presented. ............151

3-39. The number of days between peaks above 1 STD for each female at Homosassa and
SeaWorld, with the frequency of the time periods presented. ................ 152

3-40. The number of days between peaks above 2 STD for each female at Homosassa and
SeaWorld, with the frequency of the time periods presented. .................153

3-41. Wild male testosterone and female estradiol and progesterone mean monthly
concentrations SE M .......................................................... 154

3-42. Homosassa female normalized estradiol and progesterone monthly LS means. ...155

3-43. SeaWorld female normalized estradiol and progesterone monthly, LS means. ....156

3-44. Epcot, Lowry and SeaWorld male normalized testosterone monthly LS means. .157

3-45. Homosassa normalized weekly mean SEM estradiol and progesterone
concentrations. ........................................................ ........... 158

3-46. SeaWorld normalized weekly mean SEM estradiol and progesterone
concentrations. ........... ...... ..................... ............ 159

3-47. Homosassa estradiol and progesterone frequency of peaks above the mean, 1 STD
and 2 STD per month. .......... ........................ ............ 160

3-48. SeaWorld estradiol and progesterone frequency of peaks above the mean, 1 STD
and 2 STD per month. .......... ........................ ............ 161

3-49. Individual Homosassa female estradiol concentrations by month, Amanda, Ariel,
Betsy, and Lorelei. ................................................... ........... 162

3-50. Individual Homosassa female estradiol concentrations by month, Rachel, Rosie,
and Star. .......... ........ .......................... ........... 163

3-51. Individual Homosassa female progesterone concentrations by month, Amanda,
Ariel, Betsy, and Lorelei ..................................................... .. 164

3-52. Individual Homosassa female progesterone concentrations by month, Rachel,
Rosie, and Star. ........... ...... ..................... ........... 165









3-53. Individual SeaWorld female estradiol concentrations by month, Charlotte, Georgia,
and R ita. .................. ........................... ............ 166

3-54. Individual SeaWorld female estradiol concentrations by month, Sara and Stubbie.
.................................................................................. . . 1 6 7

3-55. Individual SeaWorld female progesterone concentrations by month, Charlotte,
Georgia, and Rita. .......... ....... ................... ............ 168

3-56. Individual SeaWorld female progesterone concentrations by month, Sara and
Stubbie. .......... ........ .......................... ............ 169

3-57. Individual Epcot male testosterone concentrations by month, Chester, Hurricane,
and G ene. .......... ........ ......................... ........... 170

3-58. Individual Lowry male testosterone concentrations by month, Hugh and New Bob.
.................................................................................. . . 1 7 1

4-1. Female mounting at Homosassa Springs State Wildlife Park. ......................185

4-2. Female mounting behavior at Homosassa Springs involving two manatees, with a
third on her back next to them, during summer of 1996 .....................186















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

REPRODUCTIVE ENDOCRINOLOGY OF THE FLORIDA MANATEE
(TRICHECHUSMANATUSLATIROSTRIS): ESTROUS CYCLES, SEASONAL
PATTERNS AND BEHAVIOR

By

Iskande Lieve Vandevelde Larkin

August 2000


Chairman: Roger L. Reep
Major Department: Veterinary Medicine

The study of Florida manatee reproduction is important to determine correctly

their actual reproductive potential in establishing population models, making

management decisions, identifying sensitive periods of the year, and recognizing

differences between healthy and unhealthy states related to reproduction. The

objectives of this study were to measure fecal steroid hormone concentrations (17p3-

estradiol, progesterone and testosterone) with a fecal radioimmunoassay to 1)

determine the length and frequency of estrous cycles in female manatees, 2) identify

seasonal hormone fluctuations, 3) correlate hormone concentrations with behavior, 4)

correlate hormone concentrations with reproductive tissues, 5) determine if food

availability and water temperature affect seasonal fluctuations. Elephants are used as

a model from which to pose hypotheses. The results indicate that the mean gut transit

time for the Florida manatee is 6-8 days. This suggests a time delay of approximately









one-week between fecal hormone concentrations and plasma hormone concentrations.

The current technique can not distinguish different reproductive groups (e.g. pregnant

vs. non-pregnant, adult vs. calf), but it may be able to distinguish gender among adult

manatees. The estrous cycle of the Florida manatee is approximately 28-42 days in

length. Seasonal fluctuations of captive manatee hormone concentrations suggest

slight peaks during the spring and fall. Data from wild manatees were non-

significant; thus comparisons between wild and captive animals were inconclusive.

The majority of behavioral data analyses were non-significant; however, a small

number of behaviors were correlated with fluctuations in hormone concentrations.

Vulva swelling and a behavior described as female mounting were correlated with

increased estradiol and/or decreased progesterone concentrations. A model for

manatee reproductive patterns and future research objective are presented.














CHAPTER 1
INTRODUCTION AND BACKGROUND REVIEW


Introduction

Sirenians represent a unique mammalian order that has a number of

specialized adaptations for a life of aquatic herbivory. Sirenians are the only true

aquatic herbivores among living mammals. Sirenians have been known to exist for

about 60 million years. Manatees have evolved to inhabit a tropical distribution in

both fresh and salt waters (Domning 1982; Owen 1855; Savage et al. 1994). Unlike

some well-studied marine mammals, manatees do not form stable social groups.

Manatees are considered semi-social, meaning that they do interact with other

individuals, but not on a long term or permanent basis (Hartman 1979; Reynolds

1981). The only long-term interaction found to occur in Florida manatees is between

a mother and her calf, which lasts approximately 1-2 years. A reproductively mature

female (-5 years in age), has a calf about every 3 years. Calves are born throughout

the year; however, peaks in calving and reproductive activity have been documented

in spring months (Hernandez et al. 1995; Rathbun et al. 1995; Reid et al. 1995).

Currently, the survival of the Florida manatee has been jeopardized both

directly and indirectly by human expansion. To ensure and maintain this endangered

species' reproductive viability, more knowledge of their reproductive physiology

needs to be obtained. A better understanding of the manatee estrous cycle will

increase our knowledge of the actual reproductive potential of female manatees. The









identification of seasonal hormonal fluctuations will indicate reproductively sensitive

periods of the year. In addition, a clear understanding of the basic reproductive

physiology of manatees will allow us to identify abnormal or unhealthy states. All of

these data will contribute to more accurate modeling of population dynamics of the

species. These population models are key tools used by management agencies to

make decisions on how best to protect a species and curtail human impact. Human-

related collisions with watercraft represents the largest cause of known, identifiable

manatee mortality, which further emphasizes the importance of management for

manatee survival (Ackerman et al. 1995).



Objectives

The Florida manatee is protected by legislation, including the Endangered

Species Act (1973), Marine Mammal Protection Act (1972) and Florida Manatee

Sanctuary Act (1978). In addition, there is the Manatee Recovery Plan (1995), a draft

of non-regulatory objectives and goals to aid in the management and recovery of

Florida manatees. The current studies are directly related to the above mentioned

objectives of the Florida Manatee Recovery Plan (U.S. Fish & Wildlife Service

1995). The goals of this study are as follows:



1) To document manatee reproductive steroid hormone concentrations

(17p-estradiol, progesterone and testosterone) for both males and females in wild and

captive populations:









a) To define the length and frequency of estrous cycles in female manatees.

Known patterns in elephants will be used as a model for comparison with manatee

reproductive physiology.

b) To determine if seasonal fluctuations of these hormone concentrations

occur in male and female manatees.

2) To correlate measured hormone concentrations with the behavior of female

manatees observed in captivity.

3) To correlate measured hormone concentrations from necropsied animals

with the status of their reproductive tissues.

4) To determine whether light cycle dynamics and water temperature

influence any seasonal changes in captive manatee hormone concentrations.



Hypotheses

These objectives will allow testing of the following hypotheses:



1) Female manatees have a reproductive cycle similar to that of elephants, with low

hormonal concentrations and very long luteal phases (10 week periods of

increased progesterone concentrations) between follicular phases and estrus.

2) Male manatees show seasonal fluctuations, having higher testosterone

concentrations that correspond to active spermatogenesis in warmer months.

Female manatees also show seasonal fluctuations, with increased reproductive

hormone concentrations from spring to fall.









3) There are differences between captive and wild manatees with regard to seasonal

reproductive activity. Cold water temperatures and food availability in the

winter are important factors.

Elephants (order Proboscidea) have been chosen as a model for comparison to

Florida manatee reproductive patterns because of key reproductive similarities that

are unique among mammals. Female manatees and elephants have multiple corpora

lutea on their ovaries (Hanks & Short 1972; Marmontel 1988; Perry 1953; Short

1966). These corpora lutea may be formed either by ovulation or luteinization.

Horses are one of the few other known mammals to have multiple corpora lutea;

however, they are strongly seasonal breeders (Allen et al. 1987; Urwin & Allen

1982). Elephants may have calves at any time of the year but are influenced by

proximate factors such as rain, in locations where water and nutritious food

availability are limited during dry periods (Eisenberg et al. 1971; Jainudeen et al.

1972a; Jainudeen et al. 1972b; Katugaha 1993; Poole 1987; Poole & Moss 1981).

Proboscidea are also evolutionarily related to sirenia, so other reproductive

similarities may still be discovered despite their separation through time and

differences in ecology and habitat. Information gathered from what is already known

about manatees, what this study contributes, and what we know of other mammalian

breeding patterns, especially elephants, will be utilized to create a model for Florida

manatee reproductive biology.









Background Review

Conservation

The Florida manatee is protected legislatively by the Endangered Species Act

(1973), Marine Mammal Protection Act (1972) and Florida Manatee Sanctuary Act

(1978). In addition, there is the Manatee Recovery Plan (1995), a draft of non-

regulatory objectives and goals to aid in the management and recovery of Florida

manatees. The Endangered Species Act (ESA) is designed to stimulate development

of a harmonious balance between the welfare of species and bio-diversity of the

environment with the health and well-being of mankind. The ESA is a declaration to

protect those species that are "of aesthetic, ecological, educational, historical,

recreational, and scientific value to the Nation and its people". The Marine Mammal

Protection Act and Florida Manatee Sanctuary Act further specify legislation to

protect the health and stability of marine ecosystems, obtain stable populations of

marine mammals and decrease the harassment and/or taking of marine mammals. To

take things one step further, the Florida Manatee Recovery Plan (FMRP) and part of

the ESA have set forth goals, objectives and tasks to restore the Florida manatee

population, and ultimately de-list the species under the ESA. Assessment of

population trends is imperative to accomplish these goals. Data and models are

utilized to determine population trends, based on estimates of survival and birth rates.

There are four main objectives outlined by the FMRP: 1) identify and minimize

causes of injury and mortality, 2) protect important habitat, 3) monitor the population

and important habitat, 4) establish recovery activities, monitor progress and revise the

Recovery Plan. In particular the third objective further outlines the need for studying

the physiology, ecology and life history of the Florida manatee, including analysis of









calf production and expanding long term studies of individual animals. The long-

term studies should provide information on reproductive traits, behavior, age specific

birth rates and success in calf rearing.

The present study will provide information on manatee endocrinology related

to their estrous cycle and possible seasonal fluctuations in hormone concentrations.

These data can be utilized by population models to refine information on calving

intervals and the fertility of manatees. Reproductively sensitive periods of the year

may be identified and subsequently management would be able to curtail negative

human impact. In addition, this study provides a non-invasive tool that can have

important future uses on captive and wild animals: to monitor hormone

concentrations that provide information on reproductive status, characterize hormone

concentrations during breeding events, and identify clinical signs of unhealthy states

related to steroid hormone concentrations, such as neonatal mortality. This study is

only the first step in accessing the physiological information that new non-invasive,

scientific techniques can provide from living animals.



Evolution

The known fossil record of the Florida manatee can be traced back to the

origin of sirenians in the early Eocene and has introduced us to a unique herbivorous

marine mammal unlike other extant aquatic mammals. Manatees are derived from

the same lineage as other unique species including the elephant, hyrax (order

Hyracoidea) and aardvark (order Tubulidentata). This section reviews the evolution

of manatees in relation to the development of other sirenia and other orders. By

understanding how the Florida manatee has evolved and adapted until the present,









and how other species similarly developed, then perhaps important ecological factors

related to reproduction may be identified. The early studies based lineage on bone

morphology, but more recent studies corroborate these associations based on DNA,

amino acid sequencing and immunological comparisons of proteins from bone

extracts and sera.

Within the order Sirenia there are two extant families, Trichechidae and

Dugongidae. Presently the Dugongidae is represented by only one species, the

dugong (Dugong dugon, Miuller, 1776). The most recently extinct species of the

Dugongidae is Steller's sea cow (Hydrodamalis gigas, Zimmermann, 1780). There

are three species of manatees (Trichechidae) the West Indian manatee (Trichechus

manatus Linnaeus, 1758), which is further subdivided into two subspecies, the

Florida manatee (T. manatus latirostris Harlan, 1824) and the Antillean manatee (T

manatus manatus Linnaeus). There are also the Amazonian manatee (T inunguis

Natterer, 1883) and the West African manatee (T senegalensis Link, 1795) (Bertram

& Bertram 1973; Domning 1978; Domning & Hayek 1986).

In comparison to the number of extant species, there has been a much richer

history of sirenians throughout geological time. Excluding the five extant sirenians,

there have been 31 species and subspecies discovered from the Eocene to the

Pliocene epochs (Domning 1994). The earliest known sirenian, Prorastomus

sirenoides (Owen 1855), is dated from the early Eocene age and is the best ancestral

representative for the order Sirenia (Savage et al. 1994). Domning (1982) speculated

that protosiren-like sirenians had an Old World Tethyan origin and dispersed to South

America. The earliest known trichechid Potamosiren dates from the Middle









Miocene. As the Andes Mountains formed minerals and nutrients eroded into lower

waterways and true aquatic grasses developed. Manatees of this Mio-Pliocene age

(Ribodon) were able to take full advantage of this rough herbage because they

possessed supernumerary molars that were continually replaced throughout their

lifetime. With further evolution manatees of the Pliocene-Recent (genus Trichechus)

gained more teeth, which were smaller in size to aid in resistance of wear. As access

to the Pacific was closed off, forming the Amazonian lakes, some manatees, similar

to modem T. inunguis, specialized even further by developing smaller more complex

wear-resistant teeth as they continued to eat aquatic grasses. Although T manatus

did not specialize to the extent of T inunguis they both were able to adapt, unlike the

Dugong which, outside of the Indopacific region, did not develop root hypsodonty in

response to the abrasiveness of seagrasses. It is thought that the West African T

senegalensis resulted from a waif dispersal of primitive Trichechus during the

Pliocene or Pleistocene (Domning 1982).

In relation to other Orders, evolutionary development of the Sirenia can be

traced back to the Paleocene where ancestral Ungulata differentiated into at least five

lines: Eparctocyon, Cete, Meridungulata, Phenacodonta, and Tethytheria. Simpson

(1945) originally defined the superorder Paenungulata with African ungulates and

included the orders Proboscidea (elephants), Sirenia, and Hyracoidea (hyrax), plus

Desmostyliformes, an extinct herbivore inhabiting inter-tidal coastal areas (Inuzuka et

al. 1995) as a suborder. Later, McKenna (1975) classified Proboscidea, Sirenia and

Desmostylia as orders within the mirorder Tethytheria, separating the Hyracoidea









order under the mirorder Phenacodonta along with Perissodactyla (e.g. horses, tapirs,

and rhinoceroses).

Several molecular studies have provided data suggesting phylogenetic

relationships among the manatee, elephant, hyrax and aardvark. The eye lens protein

ca-crystallin A amino acid sequence of the aardvark was compared with several

species and found to have unique similarities with Sirenia, Hydracoidea and

Proboscidea (de Jong et al. 1981). Similarly, cladistics utilizing a- and

P-hemoglobin sequences grouped the elephant, hyrax and manatee into a

monophyletic clade supporting Simpson's (1945) superorder Paenungulata

(Kleinschmidt et al. 1986). Additional DNA sequences of mitochondrial cytochrome

b gene segments (Ozawa et al. 1997), and 16S rRNA plus, 12S rRNA and tRNA

(Springer et al. 1997) group the elephant, hyrax and manatee together. These findings

suggest a considerable radiation from a common ancestor for these diverse species.

Studying these related, extant species may provide a point from which to explore

reproductive similarities, and identify a model species similar to Florida manatees.



Distribution

Environmental factors, such as ambient temperature, food availability and the

severity of fluctuations and changes in the environment, directly affect the ability of a

species to reproduce. The Sirenia represent the only marine mammals that are

obligate herbivores. Manatees and dugongs, unlike other marine mammals, have a

relatively low metabolic rate and are not adapted to cold water temperatures (Irvine

1983; Scholander & Irving 1941). Their distribution throughout the world is









primarily tropical (Bertram & Bertram 1973). This is in contrast to the North Pacific

range of the now extinct Steller's sea cow (Anderson 1995; Domning 1977; Steller

1751). The West Indian manatee is found in fresh, brackish and salt waters along the

coasts of the Gulf of Mexico (including both the U.S. and Mexico), Central America,

the north and northeast of South America, and throughout the Caribbean and

southeastern U.S. along the Atlantic (Husar 1977; Lefebvre et al. 1989; Odell 1982).

Within the United States Florida manatees are found in relatively calm, shallow

waters that have grass beds, such as estuaries that are protected by barrier islands,

bays, and large rivers. During summer months, Florida manatees spread north to

Georgia and the Carolinas, and west to Louisiana and Mississippi. This range

contracts drastically during colder months of the year, with the majority of animals

wintering in Florida and often utilizing warm water refugia (Powell & Rathbun 1984;

Rathbun et al. 1982). Even within Florida there is a seasonal north/south migration

depending upon the individual manatee's range (Deutsch et al. 1998; Reid et al.

1991). Water temperatures and food availability are important proximate factors

influencing these migrations and in turn affect reproductive patterns.



Manatee Reproduction

This section will review information on manatee reproductive anatomy,

behavior, and natural history, ending with some reproductive parameters that are still

unknown.

Reproductive anatomy of the female manatee has been characterized

extensively. Some of the earliest studies examined gross anatomical characteristics

and a few histological aspects of female reproductive tracts (Freund 1930; Garrod









1877; Grasse 1948; Murie 1872; Quiring & Harlan 1953; Wislocki 1935).

Marmontel (1988) wrote a brief review of these early studies. Ovaries of female

manatees are located just postero-lateral to the kidneys, suspended by mesovarium

from the dorsal abdominal wall. The ovaries are oval in shape and are unusual in that

the active cortex is a two-dimensional layer on one side of the ovary, as is the case in

dugong (Bonde et al. 1983; Marsh et al. 1984a; Mossman & Duke 1973; Quiring &

Harlan 1953). Follicles and corporal lutea developing on the ovarian surface protrude

outward from the two-dimensional surface. Corpora lutea (CL) may range in size

from 4.8 to 7.6 mm in diameter; however, large Graafian follicles may be twice the

size of CL, ranging from 5 mm to a maximum of 12 mm in diameter (Marmontel

1988; Odell 1982). The uterus is bicornuate with the ovaries ovulating with equal

frequency to either uterine horn (Marmontel 1988). However, not all corpora lutea

are representative of follicles that have ovulated, as mentioned above.

In comparison to the female data, relatively few studies have focused on male

manatee anatomy. Externally, females can be distinguished from males by the closer

association of the genital aperture with the anus, where in males the genital aperture

is in closer proximity to the umbilicus in the center of the abdomen (Bonde et al.

1983; Husar 1977). Male gross anatomical descriptions and figures are provided by

Vrolik (1852). One early study (Chapman 1875) briefly describes the

ischiocavernosus, bulbourethrae, and retractor penis muscles as well developed and a

penis length of 12 inches. The mean testicle diameter is 3 inches and the testis

appears to be divided into two or three lobes. Male manatees have internal testes, like

most other marine mammals including cetaceans (Odell 1982; Schroeder 1990). The









testes are located postero-lateral to the kidneys and measure up to 15 x 10 cm in an

adult male (Bonde et al. 1983). A study of testicular weights relative to body length

suggests that sexual maturity occurs in males having a body length of 275 cm or

greater (Odell et al. 1981). However, a more recent study suggests that smaller

animals, as young as two years of age, may produce sperm capable of impregnating

females (Hernandez et al. 1995). The seminal vesicles are relatively large and the

prostate gland, differing from other species, is characterized by erectile muscular

tissue rather than glandular tissue (Harrison & King 1965).

Florida manatees are considered a semi-social species, interacting with each

other without forming long term bonds, except in the case of a mother and her calf

(Hartman 1979; Reynolds 1981). A calf may remain with its mother for 1-2 years

and the calving interval is approximately 2.5-3 years (Hartman 1979; Rathbun et al.

1995; Reid et al. 1995). Manatees reach sexual maturity at approximately 2-5 years

of age (Marmontel 1995; Odell et al. 1995; Rathbun et al. 1995). Once a female

becomes sexually mature and comes into estrus, it is generally thought that manatees

form a mating herd or herd of consorting males. However, copulation has rarely been

directly observed in the wild and it has not yet been verified that the female in any

such aggregate is actually in estrus (Rathbun et al. 1995). The mating herd is

certainly the most visible potential breeding behavior manatees exhibit, but a more

subtle, breeding scenario noted from field observations, may include a single male

diligently shadowing a female until she becomes receptive (personal communication,

R. K. Bonde, U.S.G.S., 2000). The relative frequency of these contrasting behaviors

is unknown. The mating herd consists of one focal female being pursued by several









males. The individual male members participating in the herd are transitory, try

relentlessly to hold on to her, and roll over in attempts to gain access to her ventrum.

A major difficulty in consistently identifying the exact composition of these herds

derives from the hazardous nature of swimming among manatees to identify the sex

of individuals on their ventral side, from otherwise indistinguishable individuals. The

female may be pursued for 2-4 weeks (Hartman 1979; Rathbun et al. 1995).

However, physiological estrus may not necessarily be indicative of this entire period

of pursuit, but instead last only a brief period during the whole mating herd scenario.

The majority of 2-4 weeks of male pursuit may characterize behavioral estrus on the

part of the female, the establishment of dominance among males, or relate to a

strategy of sperm competition, with males breeding as frequently as possible while

she is receptive (Gomendio et al. 1998). Female manatees are thought to exhibit

promiscuous breeding behavior, mating with several males in the herd (sensu Wilson

1975; Wittenberger 1978). This mating system may more specifically be described as

"scramble competition polygyny" (Alcock 1983). This reproductive strategy has also

been described for the humpback whale (Megaptera novaeangliae) (Tyack &

Whitehead 1982). In manatees, there appears to be a peak of this reproductive

behavior occurring in April-May (Rathbun et al. 1995). Gestation length is estimated

at 12-14 months (Odell et al. 1995; Rathbun et al. 1995; Reid et al. 1995). It has been

suggested that Florida manatees may have a suppression of reproductive activity

during the colder months of the year, indicating a diffusely seasonal reproductive

pattern (Hernandez et al. 1995). The fact that fewer mating herds are seen during the

winter months when individuals are in closer proximity to each other as they









congregate in warm water refuges further supports the hypothesis that reproductive

behavior is suppressed during the winter.

Among many unknowns are the following: What are the hormonal parameters

associated with estrus in female manatees and how frequently does this occur? Can

the presence of males or other females influence cycling, by accelerating or inhibiting

estrus? What hormonal parameters and profiles are associated with manatee

reproductive cyclicity and seasonality, the latter potentially in both males and

females? Can females conceive while lactating? By answering these questions a

better understanding of the reproductive biology of manatees, and the relationship

between their physiology and behavior will be gained. Ultimately these answers will

provide information on the growth potential of the Florida manatee population and

will improve ongoing population modeling efforts. However, the current study can

only address some of these questions.



Other Mammalian Reproductive Patterns

This section is presented to give some examples of hormonal patterns possible

within the range of mammalian physiology, as well as ecological factors that

influence reproduction. Information on the physiology and natural history of other

mammalian species aids in outlining what may or may not be a possible pattern of

reproduction given what is currently known of Florida manatees.

In general, mammalian females are born with an excessive but fixed number

of primordial follicles in their ovaries. Only some of these follicles will begin

development towards ovulation, and a smaller number of those will actually reach the

potential to ovulate (Bronson 1989). Theca cells associated with the follicle are









influenced by luteinizing hormone (LH) and produce androgens that diffuse into the

follicular fluid surrounding the oocyte. Follicle stimulating hormone (FSH)

stimulates the granulosa cells of the follicle to produce enzymes that convert the

androgens into estrogens (Gore-Langton & Armstrong 1988). Pre-ovulatory follicles

secrete relatively high concentrations of estrogens, which characterize the period of

estrus (Clark & Markaverich 1988). The actual release of the oocyte during ovulation

is induced by a LH surge (Lipner 1988). The corpora lutea (CL) then develops from

the luteinization of the remaining granulosa and theca cells, which shift to the

secretion of progesterone, beginning the luteal phase of the estrous cycle (Niswender

& Nett 1988).

Hormonal variations of the female reproductive cycle varies greatly among

different mammalian species (Feder 1981; Short 1984). Some of the various factors

that affect this cycle are: the life span of the CL; length of the estrus cycle;

spontaneous or induced ovulation; the degree to which behavioral or pheromonal

cueing is involved; whether hormones originate from the ovary or adrenals; the types

of hormones secreted (i.e. estrone, estradiol or other metabolites); and the number of

cycles per season (i.e. polyestrous, monestrous and use of postpartum estrus).

Information on life history strategies among different species, in addition to

reproductive physiology, will also provide clues to the type of breeding system a

species may exhibit. One general life history concept is referred to as r and K

selection (Boyce 1984; MacArthur & Wilson 1967; Pianka 1970). In brief, the

concept stems from types of habitat and resource dispersion that select for r- or K-

species. K- and r- are parameters in the logistic equation. An r- selected species is









usually characterized by a small body size, a large allocation of energy to

reproduction, with a greater number of small offspring, and early sexual maturity and

activity. There is typically little investment in survivorship of the offspring, and

survivorship is highly dependent on environmental conditions. These species live in

unpredictable or ephemeral habitats that directly affect the mortality of both adults

and juveniles with periods of rapid population growth and large-scale mortality. In

contrast, K- selected species live in more resource stable environments with either

constant or predictable seasonal fluctuations. K- selected species are often

characterized by a large body size, a proportionally smaller allocation of energy to

reproduction, and a small number of larger offspring, with a greater amount of

investment into the survivorship of those offspring. Sexual activity is often delayed

beyond sexual maturity due to competition with conspecifics (Begon et al. 1990).

Within this r/K concept manatees best fit the K- selected characteristics given their

large size, long lives, single calves born, calves weaned at 1-2 years, and sexual

maturity between 2-5 years. The r/K theory would also suggest that there is some

type of competition between breeding individuals so that sexual activity is delayed

beyond sexual maturity, as might be the case with male competition in the mating

herd.

An example of an r-selected species would be the rat (Rattus norvegicus)

(Baker 1979; Freeman 1988; Robinson 1979). The adult female rat lives a maximum

of 3 years, has her maximum fertility between 100-300 days of age, is polyovular,

bearing several young per litter and has a short estrous cycle of 4 to 5 days (see

Figure 1-1). The rat estrous cycle is entrained to light/dark changes and ovulation











FSH
Rat

Progesterone
Estrogens

Relative ...***
Hormone
Cone. .....




Met Di Pro Est

Days of Four Day Estrous Cycle


Figure 1-1. Serum hormonal changes in the rat during the estrous
cycle, metestrus (Met), diestrus (Di), proestrus (Pro) and estrus (Est).
The dark boxes indicate time at night. The arrows indicate period
receptive to breeding. Diagram developed from information provided
in Bronson (1989) and Freeman (1988).




can be timed to within a few hours in the early morning of the day of vaginal estrus

(i.e. detection of cellular cornification) (Feder 1981). The reproductive physiology is

characterized by FSH at low concentrations except for a surge that occurs around the

time of ovulation. This surge overcomes normal inhibitory influences of inhibin and

gonadal steroids, thus stimulating follicular development that will mature after the

current ovulation occurs. LH is secreted in a pulsatile manner and a surge of LH is

caused by an increased frequency of these pulses, coordinated with high levels of

estradiol which cease inhibiting LH secretion, thus promoting ovulation. High

concentrations of estradiol are necessary to promote female searching activity and









receptivity. Without estradiol, female rats will not exhibit receptivity and lordosis to

males. Sexual receptivity in the rat may be enhanced by progesterone concentrations

(part of which is adrenal in origin), and progesterone may also be important in

terminating receptivity (Beach 1976; Fahrbach & Pfaff 1982; Feder & Marrone 1977;

Pfaff 1980; Pfaff 1982; Pfaff & Schwartz-Giblin 1988).

By comparison, the rhesus monkey (Macaca mulatta) is a more K-selected

species with a longer life span and a longer estrous or menstrual cycle of about 28

days, which is characterized by bleeding as the uterine endometrium is sloughed off

(see Figure 1-2). Generally, the female is monovulatory, having one or occasionally

two offspring. The CL persists much longer, relative to the rat. With a longer

menstrual cycle, both the follicular and luteal phases are protracted. The follicular

phase is characterized by high concentrations of estradiol, and the luteal phase by

high concentrations of progesterone. Unlike the rat, hormones play a more

diminished role in female receptivity to mating, since rhesus monkeys copulate

throughout the female's cycle. There is, however, a greater frequency of copulation

near ovulation. The rhesus monkey also lacks the circadian affects that dominate the

female rat's cycle (Bronson 1989; Pohl & Hotchkiss 1983). A model utilizing

characteristics similar to the rhesus monkey would be a step closer to something

representative of the Florida manatee, in comparison to the rat.

For a further comparison, the rhesus monkey and rat are spontaneous

ovulators, meaning that preovulatory maturation of follicles and ovulation can occur

in the absence of males. Other mammals, such as cats (Felis) (see Figure 1-3), rabbits

(Oryctolagus) and voles (subfamily Arvicolinae) are induced ovulators, defined as












Rhesus monkey


FSH

-LH


-- Progesterone
........ Estrogens


I I I I I 12
8 4 0 4 8 12


Days of Estrus
Figure 1-2. Serum hormonal changes in the rhesus monkey during the
estrous cycle. Day zero marks the beginning of estrus. Diagram
developed from Feder (1977).


Domestic Cat

Unmated Polyestrous


Mated


Pregnant or Psp.


* 0 0 N


41
U..


............... Estrogens

Progesterone


Pseudopregnancy (Psp.)


Pregnancy


0 20 40 60 -40 -20 0 20 40 60
Days
Figure 1-3. Serum hormonal changes in the unmated cat during polyestrus
cyclic activity. The dark boxes indicate periods of estrus, which last 2-10 days.
The average interval between estrous periods is 17 days. The arrow denotes
when mating occurred. Day zero marks the beginning of estrus that leads to
ovulation. Diagram developed from Feder (1977).


Relative
Hormone
Conc.


Relative
Hormone
Conc.









species whose preovulatory maturation of follicles and ovulation generally fail in the

absence of males but are stimulated by copulation to induce ovulation (Bearden &

Fuquay 1992; Concannon & Lein 1983; Feder 1981). The ovaries of induced

ovulators constantly have waves of follicular development and atresia (degeneration)

occurring when the animals are in season. Thus, estradiol concentrations and

receptivity are at a relatively high level throughout the breeding season. With

stimulation of copulation a neural impulse elicits the ovulatory surge of LH (Bronson

1989; Feder 1981). The presence of multiple CL on manatee ovaries throughout

various stages of the estrous cycle is a good indicator that manatees are spontaneous

ovulators. Fluctuations of progesterone without mating would further suggest that

Florida manatees are spontaneous ovulators.

Essentially, all of these species mentioned have variations of hormonal

patterns that revolve around the physiological conditions stated above. There are two

other groups of species that may be potential models for manatee reproduction, those

species that live in a similar aquatic environment (i.e. other marine mammals such as

cetaceans and pinnipeds) and species that are evolutionarily related. Additionally, a

model species should be well studied so that questions and hypotheses may be posed

from known data. Of the two groups, some pinnipeds and the bottlenose dolphin

(Tursiops truncatus) have been extensively studied among marine mammals.

Elephants, both the Asian (Elephas maximus) and African (Loxodonta africana), are

well-researched distant relatives of the manatee. Reasons for choosing elephants as a

model over other marine mammals include the following: extensive field work has

been done on male and female cues and detection of reproductive state between









elephants, both elephants and manatees have similar energetic factors related to life as

a herbivore, and because a unique characteristic such as multiple CL was conserved

through time, perhaps other reproductive features such as hormone metabolites might

be conserved. This does not mean to indicate that another species might not also be a

good model for manatee reproduction, but only that elephants are a good starting

point from which to propose hypotheses. A graph of current data related to elephant

hormone fluctuations during the estrous cycle is presented in Figure 1-4.



Elephant Reproduction

Some of the decisions made in approaching the present study were based on

elephants as a model due to reasons mentioned above, such as important reproductive

similarities between manatees and elephants (Marmontel 1988). This section will

review elephant reproduction and discuss some of the similarities and differences

between elephants and manatees.

Elephants, unlike manatees, have a permanent social structure that is

organized around a matriarch, her direct offspring and their young. The matriarch is

usually the mother, aunt, older sister, or grandmother of the other elephants in the

family unit. Unlike young females that remain with their natal unit, once young

males in the group become sexually mature, they disperse to travel independently or

with bull herds that consist of males ranging in age from puberty and upwards (Buss

1990; Moss 1983; Sikes 1971). Sexual maturity occurs in both sexes near 9-10 years

of age (Perry 1953), although this may be extended in certain environmental

conditions (Laws 1969). Young males, despite being sexually mature, generally are











African Elephant


- Estrogen


Relative
Hormone
Conc.


............... FSH
LH

Progesterone


I I I I I I I 120
-90 -60 -30 0 30 60 90 120


Figure 1-4. Serum hormonal changes in the African elephant. The LH peaks
indicated by the arrows are considered ovulatory LH surges. The duration
between consecutive LH surges is 23-30days. Diagram developed from Hodges
(1998).



not able to compete successfully against older bull males to gain access to females

until they are about 20 or 30 years old.

Sexually mature male elephants periodically go through a physiological state

known as musth, similar to the rutting behavior in ungulates (Eisenberg et al. 1971;

Poole 1987). This was first described in detail for Asian elephants (Jainudeen et al.

1972a; Jainudeen et al. 1972b) and subsequently identified in African elephants

(Poole & Moss 1981). Musth is characterized as a period of aggressive behavior,

temporal gland enlargement and secretion, a recurrent discharge of urine, increased

testosterone concentrations, and tends to last 2-3 months. It is suggested that,









although males are continually capable of mating with females, musth aids male

elephants in securing dominance over other non-musth male elephants, as well as

overcoming the aggressive and protective behaviors of matriarchal females. A

positive correlation has been found between musth and males searching for and

associating with female herds (Eisenberg et al. 1971; Poole 1987; Poole & Moss

1981). Younger, sexually mature males sporadically experience musth, but older

males (20 or 30+ yr.) have an annual musth period, thus securing their access to

breeding females on a regular basis. Annual periods of musth, however, are not

synchronized as a population and it is suggested that older musth males can inhibit

musth in lower ranking males (Poole 1982). This de-synchronization of musth

between males consequently decreases the frequency of fights between males to

establish dominance. However, there may be peaks of musth activity during

environmentally beneficial times of the year, such as the rainy season, when

vegetation is more abundant (Eisenberg et al. 1971; Jainudeen et al. 1972a; Jainudeen

et al. 1972b; Katugaha 1993; Poole 1987; Poole & Moss 1981).

Anatomically, female manatees and elephants have a similar deciduate

placentation and bicornuate uterine structure (Cooper et al. 1964; Perry 1974;

Shoshani & Eisenberg 1982; Wislocki 1935). Both species have ovaries that lie near

the kidneys as well as mammary glands that are axillary in their location (Laursen &

Bekoff 1978; Marmontel 1988; Perry 1953; Sikes 1971). A unique characteristic of

elephant reproductive anatomy is the particularly long urogenital canal (-90 cm -

Balke et al. 1988a; Balke et al. 1988b) which travels ventrally to an anterior position

relative to the hind legs. Unlike manatees, female elephants have a glans clitoris that









is relatively large and well developed, with a prepuce. The glans clitoris is long

enough to extend beyond the vulval orifice and may have resulted in incorrect sexing

of some individuals (Perry 1953; Shoshani & Eisenberg 1982; Sikes 1971).

In contrast to other land mammals, but similar to marine mammals including

manatees, male elephants have intra-abdominal testes that weigh between 2-3 kg,

lying medial and slightly posterior to the kidneys (Perry 1953; Sikes 1971). The

development of mature sperm and the cellular components of the seminiferous

tubules have been described by 9 phases of development and suggest cyclicity in the

spermatogenic process (Johnson & Buss 1967). Another feature of the elephant testis

is the absence of a distinct epididymis (Eales 1929; Short et al. 1967). The elephant

penis has large paired elevator penis muscles and a very well developed corpus

cavernosum penis, without an os penis or cartilage (Sikes 1971).

When a female elephant comes into estrus, some behavioral signs may

indicate her receptivity to males. A male will often 'test' a female by placing his

trunk on the female's vulva or her urine on the ground and then put his trunk in his

mouth (Eisenberg et al. 1971; Jainudeen et al. 1971). This is reflective of a flehmen

response, an important means of chemical communication between males and

females, indicating a female's reproductive state (Rasmussen et al. 1993; Rasmussen

et al. 1996). A study of wild African elephants has suggested 5 types of behavior that

indicate a female is receptive to breeding: wariness, the estrus walk, the chase,

mounting and consort behavior. In brief, wariness is a behavioral indication that the

female is nervous in the presence of a male. A female exhibiting the estrus walk will

move away from her group to avoid a male who is following her. The chase is an









increase in the intensity of an estrus walk. The female may run a considerable

distance (1 km or more) from her family group with a male in pursuit, and this may

last several hours. Mounting often occurs after a chase. The female will stand while

the male places his trunk, then head on her back. The male will then rear up and put

his forelegs on her back while he gains intromission, which lasts less than a minute.

Consort behavior is more subtle, involving participation of both the male and female

to remain in close proximity to each other, with the male following the female and the

female staying close to the male to avoid advances of other males (Moss 1983).

Females exhibit some mate choice, with older, large males successfully stopping a

female after a chase and mounting more frequently, in comparison to smaller,

younger males. In addition, large males in musth more frequently display consort

behavior with females, allowing the female to avoid harassment from other, younger

males (Moss 1983). Both elephant species are reported to be polyestrus and breed

throughout the entire year; however, they may have a peak of births during more

beneficial times of year (Katugaha et al. 1999; Laursen & Bekoff 1978; Shoshani &

Eisenberg 1982), similar to manatees. Behaviorally, female elephants are

polyandrous while they are receptive to males, as are female manatees, and the

intensity of sexual activity is periodic (Buss 1990; Buss & Smith 1966; Hartman

1979; Short 1966). However, it has been suggested that female elephants mating with

several males may be the result of unnatural conditions due to over crowding or

captivity (Sikes 1971).

Initial studies of elephant behavior, hormonal estrogens and estrogen driven

characteristics such as vaginal cytology suggested an estrous cycle length of









approximately 3-4 weeks (Eisenberg et al. 1971; Jainudeen et al. 1971; Ramsay et al.

1981; Watson & D'Souza 1975). Estrus was defined as the period a female would

stand to be mounted by a male and tended to last 4 days. Male behavioral cues such

as urine testing or a flehmen response could help to detect the onset of estrus.

Ramsay et al. (1981) found a correlation between behavior and concentrations of

estrone and estradiol, and confirmed previous indications of a 3-week cyclical

pattern. However, vaginal cytology and mucous smears did not correlate well with

behavior and the apparent onset of estrus.

In contradiction to the information provided by estrogen related cyclicity,

progesterone concentrations suggest a much longer estrous cycle of 14-16 weeks,

with a luteal phase of -10 weeks and an interluteal phase of -4-5 weeks (Brannian et

al. 1988; Hess et al. 1983; Plotka et al. 1988). This longer estrous cycle was

subsequently supported by measurements of testosterone (Taya et al. 1991), FSH and

inhibin (Brown 1991). To reconcile the differences between the estrogen and

progesterone data sets, Plotka et al. (1988) proposed a model of follicular waves

concurrent with the 3-week estrogen cycle which eventually would culminate in an

ovulatory event that produced a functional CL and the 10 week luteal phase. This 3-

week estrogen cycle during the interluteal phase was subsequently supported by

measurements of LH peaks 3 weeks apart, with the second peak classified as an

ovulatory peak which initiated the 10 week luteal phase (Kapustin et al. 1996).

However, the data from Brown et al. (1991) would not support a continuation of the

3-week estrogen cycle to produce waves of follicles during the 10-week luteal phase.









A recent model incorporating this hormonal data has been proposed by Hodges

(1998) and is represented graphically in figure 1-4.

African and Indian female elephants, like adult female manatees, exhibit a

large number of CL on their ovaries. Both elephants and manatees have CL that

develop from ovulated and unovulated follicles. This is a rare characteristic of

mammalian reproduction. The function of these numerous CL is not well understood.

Several previous studies have tried to correlate the function of the CL in relation to

progesterone concentrations and the number, size or mass of CL on the ovaries in

different reproductive states from wild elephants shot for population control (Hanks

& Short 1972; Laws 1969; Ogle et al. 1973; Perry 1953; Short & Buss 1965; Smith et

al. 1969; Smith & Buss 1975). However, extremely low progesterone concentrations

in elephants, compared to other mammals, and a wide range in the number of CL (2-

50) have proven difficult to interpret. An early hypothesis suggested for elephants,

and subsequently for manatees, is that a certain number or mass of CL had to develop

during several non-fertile cycles or silent heats before enough progesterone could be

secreted by the CL to result in an ovulation, lead to a subsequent luteal phase, and

support pregnancy (Hanks & Short 1972; Marmontel 1988; Perry 1953; Short 1966).

Recent studies of elephants have indicated that the unusually low progesterone

concentrations were actually indicative of another progestin metabolite being

biosynthesized and secreted by the elephant CL. It was discovered that two 5c-

reduced metabolites, 5c-pregnane-3,20-dione (5ca-DHP) and 5u-pregnane-3-ol-20-

one (5a-P-3-OH), are the major circulating progestins (Heistermann et al. 1997;

Hodges et al. 1997; Hodges et al. 1994). A correlation between progesterone and 5c-









reduced progestins indicated that the pattern and length of the estrous cycle was

similar to that previously outline by progesterone, however, concentrations of the 5cc-

reduced progestins were several fold higher.

The formation of multiple CL, both ovulated and unovulated (or luteinized),

begs the question for both elephants and manatees of whether they are monovular or

polyovular. The fact that both species generally have only one calf would suggest

that they are monovular. However, the numerous CL with stigmata, which are scars

that indicate the rupturing event of ovulation, would suggest that both species are

polyovular. To date the question has not been satisfactorily answered (Hodges 1998).

However, in horses, another species that possesses accessory CL, pregnancy may

continue to parturition without the accessory CL and the accessory CL may actually

be the secondary result of excess chorionic gonadotrophic activity (Allen et al., 1987;

Urwin and Allen, 1982).

When an elephant is pregnant, gestation will last approximately 22 months

(Flower 1943; Krishne Gowda 1969; Lang 1967; Moss 1983). Generally a single calf

is born. Twinning has been reported, but is rare (Hundley 1927; Laws 1969). An

elephant calf is generally weaned at 2-4 years of age at the birth of the subsequent

calf, with the youngest age to survive without milk at -24 months. However, calves

may suckle for an extended period, as long as 8 years, depending upon the mother's

tolerance of nursing (Lee & Moss 1986). A calving interval ranging from 3-13 years

has been recorded, and variation in environmental factors are an important influence

(Laws 1969; Shoshani & Eisenberg 1982). The elephant sex ratio of male to female

calves born is generally close to 1:1. This changes as the animals mature, due to









increased male calf metabolic needs, dispersion of males at puberty, and the greater

selective pressures on males due to sexual dimorphism, as well as artificial influences

such as poaching of large males with tusks (Katugaha et al. 1999; Lee & Moss 1986).



Seasonality

As mentioned above, the identification of seasonal hormonal fluctuations in

the Florida manatee may determine reproductively sensitive periods of the year.

Periods when reproduction is suppressed may highlight factors that represent costs to

manatee energetic such as limits in food availability or extreme temperatures.

Intervals of increased reproductive activity would allow managers to focus fieldwork

identifying the reproductive rate or growth potential of the population. This section

reviews factors that can influence seasonality in a number of species and the degree to

which various species are affected by seasonal cues.



General mammalian patterns

Many species, including whales, dolphins, dugongs, pinnipeds and horses

have reproductively active and inactive seasons within a year (Boyd et al. 1999; Daels

& Hughes 1993; Heinsohn 1972; Kasuya et al. 1974; Mackintosh 1965; Marsh 1988;

Marsh 1995; Marsh et al. 1984c; McBride & Kritzler 1964; Yoshioka et al. 1986).

The active season, which includes periods of breeding and/or parturition, is often

expressed during the most energetically beneficial times of year (e.g., increased food

availability) (Bronson 1989; MacDonald 1984; Vaughan 1986). Ultimate factors,

which are important in the long-term or evolutionary sense, and proximate factors,

that have an immediate impact on the initiation or cessation of reproductive activity









both play important roles in the types of breeding patterns exhibited by various

species. Examples of ultimate and proximate factors include: food availability,

rainfall, temperature, competition and predation (Bronson 1989). The context in

which these factors occur will determine the classification of a factor. An example of

seasonal hormonal fluctuations is well characterized by the horse (Equus caballus)

(see Figure 1-5). The ultimate factor affecting seasonality in the horse is photoperiod;

however, proximate factors such as nutrition and climatic temperatures also influence

reproductive patterns (Daels & Hughes 1993).

Studies in controlled settings have found that some daily and seasonal

rhythms may persist without any changes in environmental cues, and therefore may

be defined as circadian (daily) or circannual (annual) rhythms. Photoperiod is a

strong influence on seasonality in many species and studies of circadian and

circannual rhythms have worked towards identifying the physiological mechanisms

that are involved in entraining animal's reproductive patterns to photoperiod or other

environmental cues. Artificial manipulations of daylight cycles change circadian and

circannual rhythms in some mammals, such as sheep (Ovis aries) and syrian hamsters

(Mesocricetus auratus) (Bronson 1989; Martin et al. 1990). However, circannual

rhythms are thought to be independent of circadian rhythms to some degree (Gwinner

1980). For example, while kept under constant environmental conditions (e.g. food,

temperature, light:dark, 12:12), golden-mantled ground squirrels (Spermophilus

lateralis) still will continue to hibernate once a year, and starlings (Sturnus vulgaris)

still will molt and have seasonal gonadal changes (Gwinner 1980; Gwinner 1986;

Pengelley & Fisher 1957; Pengelley & Fisher 1963).











Domestic Horse


Progesterone
Winter
Relative Anestmns Transitional
Prog. Period
Conc.







Jan Feb Mar Apr May Jun

Figure 1-5. Seasonal plasma progesterone concentrations in the mare. The
dark bars represent estrus behavior. The arrows indicate the time of
ovulation. Diagram developed from Daels (1993).




This suggests that the degree to which environmental factors impact seasonal

behavior can vary greatly and depends upon the species being studied.

An interesting example of seasonal reproductive flexibility in relation to

latitude and energetic involves deer of the genus Odocoileus, which is distributed

from Brazil to Canada. In northern climates above 300 latitude the breeding season is

relatively short, during the fall and early winter. In contrast, breeding in the Florida

Everglades may occur sporadically throughout the year, with a peak of reproductive

activity in September (Smith et al. 1996). In Venezuela, females may breed more

than once a year and individuals within the population are asynchronous in their

reproductive condition (Brokx 1972; Bronson 1989; Goss 1983; Lee 1970; Richter &

Labisky 1985). With this information, it is important to take into consideration all of









the factors, whether proximate or ultimate, that provide the most optimal timing of

reproduction. Some species are hard wired to the extent of exhibiting seasonal

changes without any alterations in the environment, whereas others are flexible

enough to adapt reproductive patterns to complement environmental factors.



Seasonality in Sirenians

Factors such as photoperiod, water temperature, and food availability will

probably be the most crucial for determining Florida manatee reproductive patterns.

It is already known that the Amazonian manatee has a strong, seasonal breeding cycle

due to the influence of rainfall on the growth of aquatic vegetation (Best 1982;

Marmontel et al. 1992). The breeding patterns of the dugong are not so clearly

defined. Mature male and female dugongs are not in breeding condition continuously

throughout the year. The proportion of the population having active, rather than

regressed or resting reproductive organs, appears to increase during the spring and

beginning of summer. This suggests a diffusely seasonal breeding pattern. Dugong

calving also suggests a diffusely seasonal pattern with an increase during the same

time period as reproductive activity (Marsh et al. 1984a; Marsh et al. 1984b; Marsh et

al. 1984c). Some evidence indicates that food availability may have been a crucial

factor causing a drop in the number of female dugongs found pregnant from October

1976 July 1977. This period was correlated with a die-back of seagrasses (Marsh

1995).

Winter months (December-February) are energetically stressful for Florida

manatees, accounting for 10% of all deaths in northeastern Florida during 1986-92

and only 2-4% in other regions of Florida. This may result from long- or short-term









exposure to low water temperatures (Ackerman et al. 1995). Manatees actively seek

out warm water areas, such as natural warm springs, industrial warm water effluents

or by migrating south (Hartman 1974; Hartman 1979; Powell & Waldron 1981; Rose

& McCutcheon 1980; Shane 1983). In a cold environment, manatees have to choose

between the risk of hypothermia or remaining in warm water effluents where food

availability is often eliminated due to over-grazing. During these situations it may be

beneficial for manatee survival if energy expenditures, such as those associated with

reproductive activity, are decreased. Once temperatures become warmer in the spring

and early summer, reproductive activity becomes more energetically feasible. In fact,

peaks in breeding behavior and number of calves born during the spring and lulls in

the winter have been suggested recently by several studies (Hernandez et al. 1995;

Marmontel 1995; O'Shea & Hartley 1995; Rathbun et al. 1995; Reid et al. 1995).



Measurement of Hormones

Hormonal concentrations in animals are typically determined from plasma or

serum samples. In the present study blood plasma was collected opportunistically.

However, frequently repeated blood sampling is not practical with manatees for

several reasons. There is a complex vasculature rather than a single vein accessible

for sampling (Walsh & Bossart 1999) and it takes several individuals to restrain a

manatee to draw blood. At the time of this study, captive manatees had not been

trained to present a flipper for blood sampling as has been done for some cetaceans,

therefore, blood sampling represented a stressful event for manatees. When working

with an endangered species it would be particularly advantageous to develop a non-

invasive procedure for monitoring endocrine function and thereby decrease any









associated handling stress. Techniques utilizing feces are a rather recent development

that allow us to study an animal's reproductive status with minimal disruptive

interaction with the animal (Schaftenaar et al. 1992). Measurements of fecal steroid

concentrations have previously been used successfully in studying several species

(Lasley & Kirkpatrick 1991) including carnivores (Gross 1992; Jurke et al. 1997;

Monfort et al. 1997), primates (Hodges et al. 1992; Wasser et al. 1988), and Asian

elephants (Hoppen et al. 1992). However, one of the difficulties in utilizing fecals is

the inability to measure protein hormones such as follicle stimulating hormone (FSH)

and lutenizing hormone (LH), due to their breakdown in the gastrointestinal tract.



Captive Versus Wild Conditions

Caution must be applied in any study comparing animals housed in captivity

with those living in their natural habitat. This is especially the case if the findings are

gathered on a small number of captive individuals and will be generalized to the

population as a whole, which lives predominantly in the wild. Captivity can greatly

influence behavioral changes in various species (Carlstead 1996; Price 1984). For

example, animals in captivity no longer have to forage for food and water, and most

harsh weather conditions are controlled to some degree. There are no predators

present to evade, thus less energy is expended on vigilantly watching for danger.

However, movement is limited to the size of the enclosure and there is no escape

from possible stressors such as humans and social interactions with conspecifics,

other individuals of the same species. Animals are often housed in high densities

which may or may not be representative of wild conditions. Close association with

conspecifics may facilitate the spread of disease among the group, yet veterinary









treatment may prolong an individual's life well beyond its natural span in the wild.

Reproduction in captivity is limited by the individuals placed in the same enclosure,

as well as the physical, pheromonal and social cues that may be necessary to

stimulate a species to breed. All of these factors may play a part in changing the

behavior exhibited in captivity versus the wild and the effects will differ depending

upon the species. The important role of the researcher is to take into consideration all

of the information available on behavior in the wild and make careful comparisons to

behaviors observed in captivity. In the case of Florida manatees it is important to

note that they are not normally found in a social group or herd, unless it is a mating

herd, and can be considered semi-social (Reynolds 1981). Currently, all adult captive

manatees are housed in single sex groups to prevent breeding and the production of

additional calves that would be raised in captivity. In the wild manatees spend a large

portion of their day swimming as they forage for a variety of aquatic plants (Hartman

1979), whereas captivity limits both space and food by amount, quality and variety.

Environmental factors such as water temperature and daylight fluctuations may be

held close to constant depending upon the type of enclosure. These are some of the

elements that affect manatee behavior in captivity and must be taken into

consideration when evaluating the behavioral data collected.














CHAPTER 2
FECAL RADIOIMMUNOASSAY VALIDATION FOR THE FLORIDA
MANATEE


Introduction

The present study of manatee reproductive endocrinology was conducted with

the intention of providing detailed information on steroid hormone concentrations in

living Florida manatees, whereas previous studies of reproduction concentrated on

behavior and anatomy. The data generated from the technique described in this

chapter provide information on basic reproductive patterns of potential estrous cycles

and seasonality, as discussed in chapter three. This reproductive information may be

used to identify sensitive time periods of breeding, and to refine population models

that utilize information on the frequency of conception.

The majority of data collected for this study rely on a fecal radioimmunoassay

(RIA) utilized to measure steroid hormone concentrations (1713-estradiol,

progesterone and testosterone) from manatee fecal samples. These steroids were

chosen because of their conserved expression across taxonomic groups and

predominant roles as indicators of reproductive status. The most important benefit

this technique provided was a non-invasive means of gathering physiological data

from live animals, whether in captivity or the wild (Hodges 1986). Frequently

repeated blood sampling was not practical with manatees for several reasons. There

is a complex vasculature rather than a single vein accessible for sampling (Walsh &

Bossart 1999) and it takes several individuals to restrain a manatee to draw blood,









which is stressful to the animal. In contrast to plasma or urine samples, the fecal RIA

provided the ability to perform collections of samples without significant restraint and

without behavioral training for collection. At the time the study was conducted no

training of manatees for collection of urine or blood samples had been undertaken.

Techniques utilizing feces are a rather recent development that allow the study of an

animal's reproductive status with minimal disruptive interaction with the animal

(Schaftenaar et al. 1992). Due to species variation in hormonal metabolites, assays

must be validated specifically for each species to be studied. Measurements of fecal

steroid concentrations have been used successfully in studying the reproductive

parameters of several species (Bamberg et al. 1991; Lasley & Kirkpatrick 1991)

including carnivores (Gross 1992; Jurke et al. 1997; Monfort et al. 1997), herbivores

(Desaulniers et al. 1989; Kirkpatrick et al. 1991; Kirkpatrick et al. 1990), primates

(Heistermann et al. 1993; Hodges et al. 1992; Wasser et al. 1994; Wasser et al. 1991;

Wasser et al. 1988; Wasser et al. 1993), and elephants (Fieb et al. 1999; Hoppen et al.

1992; Wasser et al. 1996), just to list a few. However, one of the difficulties in

utilizing feces is the inability to measure protein hormones such as follicle stimulating

hormone (FSH) and lutenizing hormone (LH), due to their breakdown in the

gastrointestinal tract.

The purpose of the experiments included in this chapter was to validate the

use of the fecal RIA as a technique to measure 17p3-estradiol, progesterone and

testosterone in both male and female Florida manatees. Experiment One measured

the transit time of digesta, from ingestion to excretion, through the gastrointestinal

tract of the manatee. Experiment Two looked at the possible loss or metabolism of









steroids due to different means of handling the samples before being assayed. These

first two preliminary studies were each conducted to aid in the use of a fecal marker

(experiment one) and to identify variables that may affect apparent hormone

concentration (experiment two). Experiment Three compared hormone

concentrations from fecal samples to a set of manatee plasma samples and to

reproductive tissues collected from necropsied animals, addressing the third objective

of this dissertation. Some of the first objective is presented in Experiment Four,

which measured differences between gender, location, age and individual manatees.

A final section is devoted to Assay Trials and Tribulations encountered in developing

this technique, that is located in the Appendix.




Materials and Methods

Subjects

The captive animals utilized in these experiments for the collection of fecal

samples were held at SeaWorld Florida, Lowry Park Zoo in Tampa, Living Seas at

Epcot, and Homosassa Springs State Wildlife Park, all in the state of Florida. A list

of background information on these captive animals is presented in Table 3-1.

Visible scars and size can be used to distinguish individual manatees. The majority

of fecal samples collected from wild manatees were from individuals found at

Homosassa River and Crystal River. A small subset of these samples also included

known individuals that had been rehabilitated in captivity and were subsequently

released to the wild. Additional samples were collected from captive animals at









Miami Seaquarium and from wild animals in Sarasota Bay, Florida; however, we

were not able to analyze these samples due to a lack of funds.

Animals were divided into six categories: adult males, juvenile males/calves,

pregnant females, lactating females, cycling females and juvenile females/calves.

Determination of these categories was based on sex, estimated age of the animal and

visual assessment of lactation or pregnancy. Age estimates were based on total body

length or life history reports. Three size classes were used, as defined by the Sirenia

Project, Florida Caribbean Science Center, U.S. Geological Survey (U.S.G.S.), for

their manatee photo-identification system. These classes were developed using data

from Odell (1977), O'Shea and Reep (1990), and Marmontel (1993). Manatees

having a body length less than 245 cm were considered to be calves up to 2+ years of

age; those of 246-265 cm in body length were classified as subadults or juveniles,

independent from their mother and non-reproductive; and those greater than 266 cm

were assumed to be sexually mature adults. These parameters were reassessed if

conflicting behavioral data were gathered (e.g. a female shorter than 266 cm nursing

a calf). In addition, wild and captive populations were distinguished. All collections

from wild individuals were opportunistic, and as many collections as possible were

made.

Fecal and reproductive tissue samples were collected from necropsied animals

found throughout the coastal waters of Florida. All fecal and blood samples collected

in this study were covered under Endangered Species Permit PRT-791721, issued to

U.S.G.S.. Tissue samples were covered under the Dept. of Environmental Protection









(DEP) Endangered Species Permit PRT-773494. The University of Florida project

approval number to utilize animal tissue was A164.

Year-long fecal collections were made from 12 captive adult females (5 at

SeaWorld Florida and 7 at Homosassa Springs State Wildlife Park) and 14 captive

males (9 at SeaWorld Florida, 3 at Epcot, and 2 at Lowry Park Zoo). Sixty-one

winter and 41 non-winter fecal samples were collected from wild manatees. Seven

male and 16 female reproductive tracts with fecals from necropsied manatees were

collected. Six additional female reproductive tracts were collected without

corresponding fecal samples. Thirty-three plasma samples were received from health

related or re-tagging projects conducted by other individuals, 22 of which had

corresponding fecal samples collected.



Fecal Collections

Fecal samples of approximately 5 g or more were collected weekly from

particular individuals as consistently as possible for 1 year from Epcot (Aug 95-Nov

96) and Homosassa Springs (Jan 96-Dec 96), 17 months from SeaWorld Florida (Mar

96-Aug 97), and 6 months from Lowry Park Zoo (Aug 95 Feb 96). Samples were

collected 2 times per week from captive females and once per week from captive

males using a hand held pool net. A fecal sample could often be seen as an identified

manatee excreted it. The fecal could be attached to an animal for 5-10 minutes, but

often was dropped within five minutes of the first visual sighting. The samples were

stored in plastic bags and labeled with the name or identification number of the

animal, its sex, the time and date. Consistent sampling from the same individual was

important in order to detect trends. Colored corn was utilized when possible with









captive manatees to increase the reliability of fecal identification from individuals and

decrease collection time (See Experiment One for details). However, collections

from the males housed at SeaWorld proved to be very difficult. At SeaWorld Florida

the females are housed on exhibit to the public with viewing of the animals from

above the water, as well as underwater viewing. However, the male manatees at

SeaWorld are not on exhibit, but in an enclosure backstage with viewing available

only above water. The priority at SeaWorld was to collect samples from females

rather than males and glare from the sun, plus the design of the male enclosure made

it difficult to see. Therefore, the majority of male fecal samples were collected from

their enclosure as a group. Fecal collections from wild manatees did not provide

repeated samples from individuals. Samples were stored on ice in the field, then

frozen at -20 C in the lab, 2-7 hours later, until they were utilized for assay analysis,

approximately 2-3 years later.



Experiment One: Gut Transit Time Collections

The first preliminary gut transit study utilized one animal from Homosassa

Springs, an adult female, who was fed red corn in gel cubes with monkey biscuit

(Zupreem, http://www.zupreem.com/ or Purina/Mazuri, http://www.mazuri.com/)

powder for taste, prepared as follows: five hundred milliliters of water were heated in

a beaker until boiling. To this 25 g of gelatin (Sigma, http://www.sigma-aldrich.com)

were added slowly and mixed for 5 minutes. It was then removed from the heat and

70 g of monkey biscuit powder were then added for flavor and stirred with a spoon.

One hundred fifty grams of colored grits (Micro Tracers Inc., San Francisco, CA)









were added and stirred again. The mixture was spooned into ice cube trays and

refrigerated overnight. The gelatinized corn was fed, 1.4 kg per day, and fecal

samples collected each day until corn was found in the feces. This delay was defined

as transit time. Once corn was identified in the feces, feeding of the marker was

stopped to measure the time necessary for the total amount of marker fed to pass

through the gastrointestinal tract. This delay was defined as retention time.

The second preliminary gut transit study was conducted at Miami

Seaquarium. Fecal samples were collected daily from four manatees, 2 females and 2

males. These manatees were already receiving the colored corn to mark their fecal

samples for the reproduction study, so the color of the corn was changed to a new

color. Two aspects of gut transit time were determined, as done similarly for the

Homosassa female mentioned above. The first aspect was transit time from first

feeding for corn to be transported through the gut to its first identification in the fecal

samples. This measured the fastest rate of corn transfer through the gut. The second

aspect was retention time for all the corn fed to clear the digestive tract. Feeding of

the new color stopped, and the old color resumed, once the new color was identified

in the fecal sample, and the time it took for the new color to no longer be found in the

fecal samples was measured. This duration identified the slowest transfer of corn

through the digestive tract, see timeline in Table 2-2 (all figures for this chapter are

located at the end of the chapter). Measurement of the old color retention time when

the animal was switched to the new color, and measurement of the old color transit

time once the new color was terminated, allowed this study to be repeated for each

animal. The manatees were fed 550 g of colored corn per day each morning. It was









found that this smaller amount of 550 g was sufficient to identify the colored corn in

fecals compared to the initial amount of 1.4 kg utilized in the first study at Homosassa

Springs.



Experiment Two: Fecal Handling Collections

Experiment Two was conducted to identify and decrease possible variance in

the data due to the handling of fecal samples in the field. Large samples were

collected from 5 known captive female manatees at Homosassa Springs State

Wildlife Park and subdivided to assess the influence of different factors. These

factors included: normal, number of times subjected to freeze and thaw, length of

time in water before collection, degree of water content (saturated or dry), and length

of time at ambient temperature. The first category 'normal' is based on the standard

procedure by which samples were collected, wherein the fecal was collected with a

net directly from the animal, placed in a zip-lock bag, and put in a cooler on ice until

it could be frozen upon return to the laboratory, 2-7 hours later. The other conditions

represent variations from this standard means of collecting fecals. The dry condition

was included to identify the effects of eliminating water content as a variable. The

mean dry weight of 1 g wet weight fecals for each of the five females was 0.17 g +

.02 SD. A total of nine categories was assessed: normal, in water 30 min, in water 2

hrs, in air 30 min, in air 2 hrs, freeze immediately, freeze/thaw 3x, freeze/thaw 5x,

and dry.

Testosterone concentrations were measured with the original fecal RIA

protocol (see Appendix: Assay Trials & Tribulations) before use of the double









antibody kits from ICN. Testosterone was chosen as the hormone to measure for two

reasons. 1) At the time of the preliminary study, this was the only hormone assay that

appeared to be working properly, reflecting biological concentrations similar to other

mammals, with higher adult male concentrations than other groups, such as adult

females or calves. 2) Enzymes utilized to cleave sulfide or glucuronide bonds from

the basic hormone structure, thus aiding in antibody binding to the hormone, were not

used in the testosterone assay. Metabolic activity often affects the side chains and

their presence or absence can influence the antibody's ability to bind to the steroid in

the assay. The change in binding capacity of the antibody would therefore produce

apparent increases or decreases in hormone concentrations measured in the assay.

Data were normalized, setting the 'normal' condition for each female at

100%. Results from other handling conditions were represented as percent changes

from the 'normal' 100%. Paired t-tests were used to identify any breakdown of

hormone viability when subjected to these different categories.



Experiment Three: Plasma and Necropsy Collections

Plasma samples were collected by veterinarians during routine medical

examinations or by individuals participating in the capture, tag and release program

of wild manatees. A total of 33 plasma samples was analyzed for each of the three

hormones. Twenty-two fecal samples were collected from the same individuals for

which plasma samples were collected. Of these 22 fecal samples, 17 were collected

on the same day as the plasma sample. The remaining 5 fecal samples were collected

from a single female, Georgia, ranging from a week to 40 days after the plasma









sample collection date. All plasma samples and matching fecal samples from the

same individuals were assayed for each of the three hormones.

Tissue and fecal samples were collected from manatees that were brought into

the Manatee Salvage Program, Florida Department of Environmental Protection

(FLDEP) Marine Mammal Pathobiology Lab, in St. Petersburg, Florida (see Table 2-

5). An animal brought in for necropsy was judged for good condition (necropsied

within 72 hours of death). From those animals in good condition, fecal samples were

collected and treated as mentioned above for captive and wild animals. The

reproductive tracts of males testiss, epididymis, and vas deferens) and females (ovary,

uterine horn, uterus and cervix) were collected. The ovaries were dissected so that

the length and width could be measured. Follicles and corpora lutea were counted

and their width measured using calipers to the nearest 0.1 mm as outlined in the

protocol utilized by Marsh et al. (1984) and Marmontel (1988). Tissue was stored in

10% buffered formalin. Individuals were assessed for reproductive state according to

one of the six categories mentioned above (pregnant female, juvenile/calf female,

cycling female, etc.). The phase of the cycling female was identified as luteal, with

CL present, or follicular, with no CL present.



Radioimmunoassay

Manatee fecal samples were removed from the freezer and freeze dried in a

lyophilizer (VirTis Co., Freezemobile 3, http://www.virtis.com/). The assays utilized

a solubilization and extraction step with 5 ml 100% ethanol and 5 ml citrate buffer pH

3.7. A dried fecal sample of 0.25 g and 10 ml of solution were solubilized on a

rotating mixer, at room temperature, overnight. This was centrifuged and 1 ml of the









supernatant was decanted. For the progesterone and estradiol assay 50 [tl of sulfatase

and glucuronidase enzymes (Sigma Chemical Company, http://www.sigma-

aldrich.com) were added to cleave side chains off the hormones being measured.

Five milliliters of phosphate buffer solution (PBS) were mixed together with the

enzymes, sulfatase 5,000 units type H-5: from helix pomatia and P-glucuronidase

50,000 units type H-5 from helix pomatia. This was incubated overnight at 4 C.

One milliliter of this solution was then extracted with 4 ml of ethyl ether, vortexed for

one minute, the aqueous phase snap frozen in a bath of methanol and dry ice, the

organic phase decanted and dried under air. This procedure was repeated for a double

extraction.

The technique utilized to measure hormone concentrations from fecal and

plasma samples in the Florida manatee was a double antibody 125I radioimmunoassay.

This was in a kit form and purchased from ICN Biomedicals, Inc.

(http://www.icnbiomed.com/). The hormones measured were 17p3-estradiol,

progesterone and testosterone. The protocols for each of the three hormones provided

with the kits were followed directly except for two aspects. First, a total counts

control tube was used with only the 125I radiolabeled hormone at the same amount as

given to all other tubes in the assay; and secondly, an amount of sample, different

than what they suggest, was added to the assay for each hormone. The amount of

sample used for each hormone assay was determined by dilution curves and where

the sample concentration fit on the standard curve. In general all of the assays

utilized a standard curve which included the following tubes in duplicate: total counts

(Tc), non-specific binding (NSB), baseline with no steroid (BO), plus 6 or 8 tubes









with increasing concentrations of standard steroid. Once the standard or sample

solutions were pipetted in their proper tubes, the radiolabel and first antibody were

added. This would then be vortexed and incubated at 370C for a specified time. The

second antibody or precipitate was then added and eventually centrifuged for 20 min.

The solution was decanted and the remaining pellet was counted for radiation levels.

The following paragraphs will describe the details of each technique and the

validating data for measuring 17p3-estradiol, progesterone and testosterone

concentrations from fecal and plasma samples. All samples were measured in

duplicate. A summary of some stumbling blocks over come in working out this

technique is included in the Appendix under Assay Trials & Tribulations.

17p-estradiol

The mean recovery of estradiol from the fecal samples was 41 7.6% SD.

The recoveries following extraction averaged 56 5.0% SD. The estradiol assay

utilized 450 [Il of solubilized fecal sample solution and 150 [l for the plasma

samples. The cross-reactivities of 17p-estradiol antiserum with other steroids were as

follows: 20% for estrone, 1.5% for estriol, 0.68% for 170c-estradiol, and less than

0.01% for all other steroids examined. The minimum concentration detectable was

890 pg/g for fecal samples and 67 pg/ml for plasma samples. Internal standards were

measured using concentrations of 0, 10, 30, 100, 300, 1000, and 3000 pg/ml of

estradiol each with 100 jl of the solubilized fecal solution. The resulting linear curve

was y = -7.1771 + 0.83105x where x equals the log concentration and y equals the

bound. This internal standard curve has a correlation coefficient of 0.99 with the

standard curve. A dilution curve of the solubilized fecal solution utilized 50, 100,









200 and 500 [il. The dilution curve was compared to the standard curve and a test for

homogeneity of the regression indicated the curves did not differ. The plasma

internal standard curve utilized the same concentrations of estradiol as the fecal

internal standards, however 150 [Il of plasma was added to each concentration. The

resulting linear curve was y = -79.778 + 1.1078x. This internal standard curve has a

correlation coefficient of 0.86363 with the standard curve. A plasma dilution curve

of 25, 50, 100, and 200 [l was compared to the standard curve. A test for

homogeneity of the regression indicated the curves did not differ. The 17p3-estradiol

interassay and intra-assay coefficients of variation were 4.6% and 2.1%, respectively,

for fecal and plasma samples together.

Progesterone

The progesterone assay employed fecal and plasma amounts of 200 [l and

300 [il, respectively. The recovery of added radiolabeled progesterone from fecal

samples averaged 37 + 12% SD and the extraction recoveries averaged 61 + 8.1%

SD. The minimum concentration of progesterone detectable was 4.7 ng/g from fecal

samples and 0.08 ng/ml from plasma samples. Progesterone antiserum cross-

reactivities with other steroids were as follows: 5.4% for 20ca-dihydroprogesterone,

3.8% for desoxycorticosterone, less than 1.0% for corticosterone,

17ca-hydroxyprogesterone, pregnenolone, androstenedione, and testosterone, and less

that 0.01% for all other steroids examined. Progesterone internal standards (0, 0.2,

0.5, 2, 5, 10, 25, and 50 ng/ml) were added to 100 [l of solubilized fecal solution.

The calculated linear curve was y = 254.31 + 0.939x and is parallel to the standard

curve with a correlation coefficient of 0.99. The progesterone dilution curve (50,









100, 200, 500 [tl) did not differ from the standard curve, as indicated by a test of

homogeneity of variance. The plasma internal standard used the same concentrations

of progesterone as for the fecal samples, however they were added to 300 [tl of

plasma. The curve generated the following linear equation: y = 584.44 + 0.99255x

and had a correlation coefficient of .99 with the standard curve. The dilution curve

for the plasma samples (150, 300, 450, 600 [tl) was found to be similar to the standard

curve with a test of homogeneity of variance. Interassay and intra-assay coefficients

of variation were measured for fecal and plasma samples together for progesterone:

18 and 2.1%, respectively.

Testosterone

Amounts of 25 [tl for the fecal assay and 50 [tl for the plasma assay were

applied for measuring testosterone concentrations. The mean recovery of added

radiolabeled testosterone to fecal samples was 43 + 11% SD and 66 + 11% SD for

extracted recoveries. The minimum detectable concentration of testosterone from the

solubilized fecal solution was averaged at 160 ng/g and 2 ng/ml from plasma samples.

The identified cross-reactivities for the testosterone antiserum were as follows: 3.4%

for 5u-dihydrotestosterone, 2.2% for 5c.-androstane-3p3, 17p3-diol, 2.0% for 11-

oxotestosterone, less than 1.0% for 6p3-hydroxytestosterone, 5f3-androstane-3p3,

17p-diol, 5p-dihydrotestosterone, androstenedione, epiandrosterone, and less than

0.01% for all other steroids examined. The fecal testosterone internal standard (0,

0.1, 0.25, 0.5, 1, 2.5, 5, and 10 ng/ml) had a calculated linear curve ofy = 1673.6 +

1.5289x and a correlation coefficient of 0.99 with the standard curve. The dilution

curve of solubilized fecal solution (50, 100, 200, and 500 [tl) was found to be similar









to the standard curve by a test of homogeneity of variance. The plasma internal

standard utilized the same concentrations of testosterone as for the fecal samples and

the calculated linear equation was y = 189.16 + 0.78694. The curve was found to be

similar to the standard curve with a correlation coefficient of 0.99. A plasma dilution

curve (25, 50, 100, and 200 [il) was tested for homogeneity and found to be similar to

the internal standard. Interassay and intra-assay coefficients of variation were

measured for fecal and plasma samples together for testosterone: 20 and 2.9,

respectively.

Raw data from the gama counter (LKB-Wallac 1282 CompuGamma, LKB

Wallac, http://www.wallac.fi/) was initially analyzed utilizing Beckman ImmunoFit

EIA/RIA program, version 3.1 (Copyright 1989-1991, Beckman Instruments Inc.,

Microsoft Corp.). This program provided four statistical fits (logit-log fit, 4-

parameter logistic curve fit, linear regression, or cubic spline curve fit) for the

standard curve produced in each assay to adjust the initial raw data points to the

standard curve. Only one statistical curve fit was used for all assays measuring one of

the three hormones, which was dependent upon the consistency of the standard curves

run for all assays. The best statistical fit for 17p3-estradiol and testosterone was the

cubic spline; the 4-parameter logistic curve best fit the standards for the progesterone

assay.

The fecal steroid assay techniques were conducted in the United States

Geological Survey's Biological Resources Division (USGS-BRD) Florida Caribbean

Science Center (Gainesville, FL) in association with Dr. Tim Gross. These facilities

provided all the permanent equipment necessary to carry out these procedures. This









laboratory works under the radioactive materials license #09-25420-01 associated

with the US Dept. of the Interior.



Analysis of Hormonal Parameters

Analysis of hormone concentrations between different groups of manatees

based on location, sex and month utilized analysis of variance (ANOVA). Pair-wise

contrasts were tested by confidence intervals using the least significant differences

(LSD) value for the Tukey method. The SAS program, version 6.12, was used for its

univariate plot and general linear model (GLM) procedure to conduct these analyses

(SAS Institute Inc. 1989). The assumptions that need to be met for ANOVA are

normality, constant variance and no correlation between observations. The

appearance of non-normality can be caused by a lack of constant variance, otherwise

referred to as hetroskedasity. Normality of the hormone concentration data was

characterized by skewness, kurtosis, histograms, box plots, and a ShapiroWilk

normality test. The data set did not appear to be distributed normally and thus was

transformed to the natural log for analysis. Ideally, the values for skewness and

kurtosis should be zero, and not greater than five. Additionally, if the data is

symmetrical, implying that it is normally distributed, the mean should equal the

median. The values for these characteristics of normality are shown in Table 2-1.

The normal text headers indicate values before log transformation, and the bold text

headers indicate values after the data was log transformed. All of the female values

for skewness and kurtosis were brought to values below 3.5 after log transformation.

All of the mean and median values were also brought to values within one point of

each other, after log transformation. Log transformation of the data provided the best









improvement of the data towards normality; however, in all cases the ShapiroWilk

normality test did not indicate the data were normally distributed. The reason the

ShapiroWilk test did not indicate normality for the log transformed data is due to the

large number of low concentration values. This indication of non-normality would be

similarly indicated by other normality tests. Long periods of no estradiol or

progesterone fluctuations are the cause for these low concentration values.

Additionally, the low concentrations all have similar values because the minimum

concentration detectable by the fecal RIA was calculated and set at a given value for

each hormone assay, truncating the lowest concentrations. However, the remaining

concentration values, above the low minimum values, are normally distributed and

the ANOVA is flexible enough in its normality assumptions to accommodate for this.

Another option would have been to run a non-parametric analysis, however the non-

transformed data were still too non-normally distributed and the variance was not

constant.

For the comparisons in Experiment Four, an ANOVA was run using the Proc

GLM procedure in SAS. This analysis utilized the data transformed to the natural log

as mentioned above. Pair-wise contrasts were tested for by confidence intervals using

the least significant differences (LSD) value for the Tukey method.

The wild samples in Experiment Four, Figures 2-7 and 2-9 also include

hormone concentrations from necropsied manatees for the testosterone and

progesterone values. ANOVA and Tukey pair-wise contrasts indicated that male and

female concentrations for testosterone and progesterone were similar between the

wild and necropsy values (female testosterone p= 0.18, male testosterone p= 0.83,









female progesterone p= 0.20, male progesterone p= 0.74). Estradiol concentrations

between wild and necropsied manatees were statistically similar for males (p= 0.41),

but differed statistically for females (p= 0.0072). Therefore, the wild female estradiol

concentrations do not include values from necropsied manatees.




Results

Experiment One: Fecal Marker/Gut Transit Times

This was conducted as a pilot study to try to measure the transit time of

digesta in the gastrointestinal tract and to improve the efficiency of collecting fecal

samples from the captive manatees, by decreasing collection time and correctly

identifying from which manatee a fecal originated. It took six days from first feeding

for the corn to pass through the digestive tract of the Homosassa female. On the

seventh day post-feeding, once feeding was stopped, no corn was identified in the

feces collected. Utilizing the information from the Homosassa female improved

identification of collections made at the Living Seas, Epcot for 1 year from 3 male

manatees, Miami Seaquarium from 2 males and 2 females, and at SeaWorld Florida

from 5 females and 3 males. On average 550 g per day was enough to mark feces

from adult manatees. Utilization of the marker reduced misidentification of fecal

samples and decreased the time interval spent waiting for an individual to defecate.

Using colored corn made it possible to identify fecal samples from particular

manatees by feeding each manatee a specified color. Although manatees are

caprophagic, the amount of marker incidentally eaten from another manatee appeared

to be minimal.









Samples were collected at Miami Seaquarium to measure gut transit times

from 4 manatees, 2 males and 2 females. The duration in days for both gut transfer

time aspects, transit and retention, from all 4 manatees are found in Table 2-2. Table

2-3 also includes the lengths and weights of each manatee fed the colored corn. The

mean number of days for corn to first be seen in the fecal sample after first feeding, or

gut transit, was 7 days 0.37 SEM. The mean number of days for corn to leave the

digestive tract once identified in the fecal sample, or gut retention, was 8 days 0.37

SEM. This data compares closely with the female at Homosassa Springs for which

transit and retention times were initially measured as 6 and 7 days, respectively.

There did not appear to be any increase in transit or retention with an increase in

length or weight. No statistical analysis was conducted due to the small number of

animals used in this preliminary study.



Experiment Two: Fecal Handling

This study was conducted to determine if different aspects of handling fecal

samples in the field artificially changed the hormone concentrations measured (see

Table 2-4). A paired t-test analysis was conducted to determine if significant

differences from the normal condition were associated with alternate handling

procedures. None of these handling procedures produced results that were

significantly different from the normalss'. However, the small sample size of only

five individuals should be taken into consideration. The condition closest to a

statistically significant value was the dry condition with p = 0.10.









Experiment Three: Fecal Comparisons with Plasma and Tissue Samples

Plasma and fecal

Each of the plasma samples with a matching fecal sample or samples were

measured for 17p3-estradiol, progesterone and testosterone. A comparison between

the fecal and plasma samples is presented in Figure 2-1. The samples are each

identified by the manatee's name or identification number. The animals are grouped

by gender and reproductive status. In general, the plasma sample concentrations are

much lower than the fecal sample concentrations, as would be expected by the

accumulation of steroids in the liver, then gall-bladder, and subsequent release into

the duodenum (Honour 1984). There is great deal of variation in both the fecal and

plasma sample progesterone concentrations, between individuals, and in relation to

the different groups of reproductive status and gender. Elevated fecal hormone

concentrations are not necessarily associated with elevated plasma hormone

concentrations. Fecal estradiol concentrations are relatively low except for one high

concentration measured in an adult male. In comparison to fecal estradiol, greater

variation is present in the estradiol plasma sample concentrations. Measurements of

testosterone in both plasma and fecal samples do appear to track each other in relation

to differences in concentration, unlike estradiol and progesterone. Both plasma and

fecal testosterone concentrations are low except for concentrations from adult male

samples.

Figures 2-2 and 2-3 present the mean hormone concentrations for the different

reproductive stages by gender for plasma and fecal samples. Testosterone measured

in both plasma and fecal samples indicate higher concentrations in adult males

compared to male calves and females. Plasma estradiol has variable concentrations









that are similar across the reproductive groups in both males and females. The fecal

sample estradiol concentrations indicate a few female adult and pregnant samples

higher than the other reproductive groups except for one very high concentration in

an adult male. The plasma progesterone concentrations between the reproductive

groups are similar except for one sample from an adult female with a high

concentration measured. The fecal progesterone concentrations across the

reproductive groups have similar values with high variability. No clear biological

characteristics can be distinguished between the reproductive groups when looking at

the estradiol and progesterone measurements. In contrast, some of the adult male

plasma and fecal samples had much higher testosterone concentrations in comparison

to the other reproductive groups.

All plasma samples, including samples without matching fecal samples, were

averaged and graphed to compare male and female hormone concentrations (see

Figure 2-4). The symbols labeled F study and M study along the X-axis are the

female and male values, respectively, from a previous study (Francis-Floyd et al.

1991) that measured progesterone and estradiol plasma concentrations from 4

manatees. These four subjects included 2 non-pregnant females (age 5.2 and 6.6

years), a pregnant female and a mature, adult male. The paper refers to the non-

pregnant females as subadult and pre-pubertal; however, the more recent revised

size/age class definitions for manatees indicates that manatees age 3 years or greater

are reproductively mature. The female progesterone values from the Francis-Floyd et

al. study (1991) fall well within the range of values measured in this study and the

male progesterone concentrations from both studies are similar. However, both male









and female estradiol values from the Francis-Floyd et al. study (1991) are much lower

than the concentrations this study measured. In general, this study found the mean

concentration of female plasma progesterone to be higher than male concentrations,

although not statistically higher. Estradiol mean concentrations between male and

female manatees in this study were similar. Differences between male and female

plasma testosterone concentrations are easily identifiable with males having much

higher concentrations.

In comparison to the plasma samples, fecal hormone concentrations shown in

Figure 2-5 indicate a similar distinction between male and female testosterone

concentrations, but a great deal of overlap in values between male and female

progesterone and estradiol concentrations. The samples included in this graph,

labeled 'Female' and 'Male', are all of the fecal samples that were measured for each

of the three hormones. The similarity between males and females is more

pronounced for progesterone than for estradiol where the range of values is much

higher in males than females. Due to long periods of anestrus or lack of any apparent

hormonal fluctuations, in some cases several months, found in the longitudinal fecal

data (see figures 3-1 through 3-12) an additional comparison labeled "Above 2STD"

is included. This comparison contains mean values above two standard deviations

from the yearly average of each of the 12 captive females. The idea is to compare

female concentrations of estradiol and progesterone that would represent potential

values of follicular and luteal active phases. Although the means for concentrations

above two standard deviations in both estradiol and progesterone are higher than









those including potentially non active phases under 'Female', there is still a great deal

of overlap with the 'Male' values.

Fecal and necropsy tissue

A total of 22 female reproductive tracts was collected from necropsied

manatees. Of the female reproductive tracts collected 16 had matching fecal samples

collected and assayed for all three hormones. Table 2-4 contains the measurements

taken from each of the female reproductive tracts, body length and hormone

concentrations. Females are categorized as luteal when CL are present on the ovaries,

follicular when no CL are present, and pregnant when a fetus or embryo is detected

and the uterus is distended. The single lactating female was indicated as such on the

necropsy report. All calves and juveniles measured were lacking CLs and only 4 out

of 11 manatees in this category had any follicles larger than 2 mm. Figure 2-6 depicts

the mean hormone concentrations of male and female manatees from different

reproductive groups. Of the progesterone concentrations measured, pregnant females

have the highest mean, but calves with no CL, a follicular female with no CL and the

males have some values in the same range as the pregnant females. The highest

estradiol concentrations were measured in pregnant females, a follicular female and a

juvenile male. In comparison, the estradiol values were relatively low for adult

males, female calves, a lactating female, and a luteal female. A single adult male had

a particularly high testosterone value, and two other male values were slightly higher

than those concentrations measured in female manatees.









Experiment Four: A Comparison of Gender, Location, Age and Individuals

Gender and location

Samples that were collected from captive and wild manatees, both male and

female, at each of the locations are compared in figure 2-7. When comparing males

to females for a particular hormone, differences in scale are noted with an asterisk.

There are two aspects of the six graphs that are particularly important to note: 1) the

overall differences in hormone concentrations between males and females and 2) the

pattern of variation between locations.

Significant differences were present when comparing the mean hormone

concentrations by location, but the pattern of differences between locations was

consistent across the three hormones. Within the female locations the order from

highest concentration to lowest was SeaWorld (SWF), Wild, and Homosassa Springs

(HS), with SeaWorld and Homosassa being significantly different from each other,

but both had values that did not differ significantly from the Wild group, for all three

hormones. The order of male hormone concentrations across locations was Epcot and

SeaWorld with the highest means, followed by Lowry then Wild with the lowest

mean concentrations. For the mean values of testosterone and progesterone Epcot

and SeaWorld were statistically similar and both were significantly higher than

Lowry and Wild. Lowry and Wild were similar in hormone concentrations. The

mean estradiol concentrations for males indicated that Lowry, SeaWorld and Wild

were statistically similar and Epcot was different from the three other locations.

Across all three hormones the trend was for male manatees to have higher

concentrations compared to females. Male testosterone concentrations were more

than an order of magnitude higher than female concentrations across all locations.









The difference between males and females was not as immense for progesterone

concentrations across all locations. Male samples from the wild and Lowry had

progesterone values in the same range as female values, however, progesterone

concentrations from Epcot and SWF males were much higher than females and were

significantly increased over the other male locations. Male samples assayed for

estradiol concentrations from the Wild were within the same range as some female

samples, however the other male groups were higher and Epcot was significantly

higher.

Graphs shown in Figure 2-8 compare the three hormones for all male and

female fecal samples assayed. These samples include captive, wild and necropsied

animals. Across the three hormones males consistently have higher concentrations

measured. Of the samples that were assayed for each of the three hormones, ratios of

progesterone over testosterone, estradiol over testosterone and estradiol over

progesterone were compared between males and females. Only a ratio of

progesterone over testosterone indicates higher values for female manatees.

In Figure 2-8 Continued, the testosterone and progesterone over testosterone

graphs are duplicated and include arbitrarily determined threshold values that could

be used to test whether a fecal sample from an unknown manatee could correctly be

identified as male or female. The testosterone threshold value of 818 ng/g delineates

98.5% of females below the threshold and 86.7% of males above the threshold. In the

progesterone over testosterone a threshold value of 0.2 delineated 72.8% of females

above the threshold and 88.8% males below the threshold. In both cases values









below the threshold values have overlapping hormone concentrations between males

and females.

Age

Comparisons between different reproductive stages and age groups were made

using samples from wild and necropsy animals. Progesterone and testosterone values

from both wild and necropsy animals were combined for analysis; however, the

estradiol concentrations measured in wild and necropsy animals were statistically

different (at p = 0.016), so only wild estradiol values are included in the analysis of

age and reproductive status. Figure 2-9 presents the least square means (LSM) values

and the + standard error for the different age and reproductive groups measured for

the three hormones. No statistical differences were found for testosterone and

estradiol concentrations between the groups (p = 0.13 and p = 0.83, respectively).

Female juveniles had the highest estradiol concentrations and were statistically

different from calves and pregnant females (p = 0.017).

In addition to variance noted between locations, mentioned above, differences

in mean hormone concentrations were also found between individuals. Figures 2-10,

2-11, and 2-12 represent female LS means per individual at Homosassa, SeaWorld,

and males at Epcot and Lowry, respectively. Homosassa female estradiol

concentrations appear to comprise two groups, with Amanda and Rosie having

significantly higher concentrations than the other five females. This pattern,

however, does not hold for progesterone concentrations, where Betsy has the highest

values, Lorelei has the lowest, and the rest of the females are intermediate. Of the

SeaWorld females, Charlotte and Rita have concentrations significantly higher than









the other three females. Rita also has significantly increased progesterone

concentrations along with Stubbie. Charlotte and Georgia have the lowest

progesterone values. The testosterone LS means for each of the males at Epcot were

statistically different, with the highest values found for Gene, next Hurricane, and

then Chester. Both males at Lowry had statistically similar testosterone

concentrations.




Discussion

Experiment One: Fecal Marker/Gut Transit Time

Experiment One was a preliminary examination to measure the gut transit

time of the Florida manatee and determine the viability of using colored corn as a

means to mark fecal samples, improving fecal identification and collection efficiency.

Utilization of fecal markers has been well established in digestive studies looking at

the passage rates of digesta and fecal composition (Warner, 1981; Van Soest, 1982).

Both liquid and solid phase markers are available. There is special concern in feeding

a fecal marker to an endangered species, therefore the marker selected for this study

was colored corn. The corn is naturally digestible if it should be retained within the

digestive tract and the dye is the same coloring provided for human consumption

under the Food and Drug Administration. The corn is ground into small pieces and

has the appearance of grits. The fecal marker was successful in identifying the

manatee from which a fecal sample originated and was subsequently utilized in

several facilities to collect samples.









Measurement of digesta transit time was done on a gross, per day basis. The

results from the Homosassa female indicate a gut transit time of 6 days and gut

retention time of 7 days. This is comparable to data collected from 4 Miami

Seaquarium manatees that averaged a 7 day gut transit time and an 8 day gut retention

time. In hours this would range from 144 to 192 hrs with a median value of 168 hr.

Similar results were found for manatee gut transit time from Lomolino (1984) 6 days,

and Best (1981) 5 days.

There did not appear to be any increase in transit or retention with an increase

in length or weight. These data suggest that the size of the animal does not play an

important role in the time required for digesta to traverse the gastrointestinal tract of

manatees; however, a more refined measurement of time, such as hours, may provide

greater detail. Other factors such as diet and fiber content also may be more

important. These factors were not included in the preliminary study because the

manatees were allowed free choice of the foods made available to them, which

included iceberg and romaine lettuce, sweet potatoes, carrots, apples, bananas, and

Purina @ monkey chow. The data does indicate that if fecal and plasma samples

were collected at the same time from a given manatee, the fecal sample would

represent hormone concentrations several days behind that of the plasma sample.

In comparison to other mammalian transit times, the manatee has an unusually

long retention time. The elephant has a mean transit time between 21 and 46 hours

(Rees 1982; Warner 1981). Horses, which are similar to manatees in utilizing

hindgut fermentation, have a mean gut transit time ranging from 28 to 38 hours

depending upon the diet fed. Ruminants such as cattle (Bos taurus) or buffalo









(Bubalus bubalis) have means that range 68.8 28.2 hrs SD and 94.8 + 3.3 hrs SD,

respectively. In fact, in the comparative review conducted by Warner (1981) the only

species that appear to come within the same range of manatees are the koala

(Phascolarctos cinereus) and sloths (the three-toed sloth, Bradypus tridactylus, and

the two-toed sloth, Choleopus didactylus).

In relation to the Florida manatee's gastrointestinal tract function and

structure, this long gut transit time allows manatees to be extremely efficient in their

digestion and absorption of plant material. Manatees have an immense large intestine

and cecum, where the majority of cellulolysis (83%) occurs (Reynolds & Rommel

1996). The size and length of the hindgut, as well as ridges (mainly in the colon) that

lie perpendicular to the passage of food also point to the slow rate of travel for

digesta. Manatees are comparable in their efficiency at digesting cellulose with

ruminants, such as sheep and cow, but are considerably more efficient than another

hindgut herbivore, the horse (Bum 1986; Reynolds & Rommel 1996).

This long gut transit time in manatees suggests a considerable time delay in

the hormonal values measured from fecal samples, compared to plasma samples,

which represent an immediate reflection of biological state. To make a comparison,

elephants, another hindgut herbivore, have a gut transit time of 1-2 days and the time

difference between serum and fecal values was 2 days (Rees 1982; Warner 1981;

Wasser et al. 1996). Horses have a gut transit time between 1-2 days and have a time

delay between fecal and serum samples of 20-26 hours (Palme et al. 1996; Warner

1981). The time delay in manatees will be dependent on the transit time from the bile

duct just past the duodenal ampulla where the steroids enter the digestive tract from









the liver, to the rectum (Reynolds & Krause 1982; Reynolds & Rommel 1996).

Because manatees are hindgut fermenting herbivores, this is where the majority of

digestion occurs, as mentioned above, and thus the slowest area for passage of digesta

(Bum 1986; Reynolds & Rommel 1996). This may mean that a time difference

ranging 5 days to a week may separate values measured from manatee plasma and

fecal samples. The peak excretion time in hours, of radiolabeled steroids

administered to several species, is compared by Schwarzenberger et al. (1996).



Experiment Two: Fecal Handling

The results of Experiment Two imply that no serious degradation or

metabolism occurred during the handling of fecal samples in the field until they were

frozen and subsequently assayed. As mentioned in the methods, measuring

testosterone concentrations was the best option at the time the study was conducted.

However, including measurements of estradiol and progesterone concentrations

would have provided a more complete data set.

Ideally, in addition to the Handling Experiment, a study administering

radiolabeled estradiol, progesterone and testosterone should have been done to

determine the specific metabolites excreted, as well as the proportion of metabolites

excreted in feces versus urine. Working with an endangered species has provided

limitations in the types of invasive studies that could be conducted. However, since

the beginning of this study several papers have been published utilizing the infusion

of radiolabeled steroids on endangered or threatened species such as the African

elephant (Wasser et al. 1996), Sumatran rhinoceros (Dicerorhinus sumatrensis)

(Heistermann et al. 1998), white rhinoceros (Ceratotherium simum) (Hindle &









Hodges 1990), and African wild dog (Lycaonpictus) (Monfort et al. 1997). These

endangered species are in addition to the many domestic or non-threatened exotic

species studied, including the domestic cat (Felis catus) (Brown et al. 1994; Shille

1990), baboon (Papio cynocephalus) (Wasser et al. 1994), slow loris (Nycticebus

coucang) (Perez et al. 1988), cotton-top tamarin (Saguinus oedipus) (Ziegler et al.

1989), Siberian polecat (Mustela eversmanni), North American river otter (Lutra

canadensis) (Gross 1992), sheep, horse and pig (Sus scrofa) (Palme et al. 1996), to

list just a few. Recent studies reviewing the data of fecal metabolites excreted across

species have shown that the majority of steroid metabolites are unconjugated.

Estrogens are principally excreted as estrone and/or 17a- or 17p3-estradiol. The

majority of fecal progestins are excreted not as progesterone, but as 5a- or

5p-reduced pregnanediones and hydroxylated pregnanes (Schwarzenberger et al.

1996).

The next possible step to take, barring the injection of radiolabeled steroids,

would be to utilize High Pressure Liquid Chromatography (HPLC) to identify

metabolites that are subsequently measured for their immunoreactivity in an assay

such as an enzyme immunoassay (EIA) or RIA. This was done for manatees,

however the sheet with identified peaks indicated by known eluted metabolites from

the HPLC was lost. Without the key, the assayed metabolites can not be identified.

However, the estrogen RIA did measure a single distinct peak shared by each of the

three females studied. The progesterone RIA measured a single peak shared by two

of the three females sampled, as well as 2 peaks from one female and a single peak

from another female not shared among the other females sampled. In total the









progesterone RIA measured 4 peaks between aliquots 65-80. The HPLC column

used was a Beckman (http://www.beckman.com/) ultrasphere ODS (C18) 4.6 mm X

25 cm, 5 |tm particle size. The mobile phase was 60% methanol and 40% distilled

water. Twenty microliters of the solubilized fecal solution were measured from each

of three female manatees. The flow was 1 ml/min at room temperature with 30 sec

fractions. The HPLC system used was a Beckman System Gold with software, pump

module 126, and UV detector module 168. The RIA technique utilized to assay the

aliquots from the HPLC was the older technique mentioned in Assay Trials &

Tribulations (see Appendix.)



Experiment Three: Fecal Comparisons with Plasma and Tissue Samples

Comparisons of steroid concentrations between plasma and fecal samples in

Experiment Three indicate that adult male manatees have higher concentrations of

testosterone than females or immature males, and that these higher levels are

similarly expressed in both sampling methods, feces and plasma. Estradiol and

progesterone concentrations, however, indicate much more variation. The fact that

higher estradiol or progesterone plasma concentrations are not similarly indicated by

higher fecal concentrations and vice versa in Figure 2-1 is partially explained by the

results of Experiment One. Time delay due to gut transit of the fecal hormones may

be the cause of discrepancies in estradiol and progesterone between fecal and plasma

concentrations. Female fluctuations in estradiol and progesterone may highlight this

time difference in concentrations by 2 days (as in elephants, Wasser et al. 1996) or

longer. However, this does not explain the variation in male estradiol and









progesterone concentrations. Differences between individuals may be another

important factor contributing to the variation.

Unlike testosterone, little biological distinction can be made between the

different reproductive groups utilizing estradiol or progesterone concentrations from

either plasma or fecal samples. A small number of animals and a subjective means of

determining an adult female's reproductive state (i.e. pregnant or cycling) from her

appearance may contribute to the inability of plasma samples to distinguish different

reproductive states in Figure 2-2. Plasma samples comparing male versus female

progesterone concentrations in Figure 2-4 do reflect a trend of higher female values.

In addition, the female estradiol plasma concentrations in relation to males are higher

than the fecal estradiol concentrations.

Unfortunately, knowing the physical reproductive state of the animal from

their reproductive tracts does not improve the ability to identify different reproductive

stages from hormone concentrations. The testosterone concentrations measured from

necropsied manatees shown in Figure 2-6 are similar to the other reproductive groups

except for one very high male value. The estradiol concentrations for the females

may indicate a trend for pregnant and follicular animals to have higher values.

The difficulties of identifying different reproductive groups from fecal

progesterone concentrations may be related to other predominant types of metabolites

excreted, as mentioned above (5uc- or 5f3-reduced pregnanediones and hydroxylated

pregnanes). Progesterone may not be the most appropriate metabolite to measure.

The fluctuations that naturally occur with estradiol and progesterone in females may

contribute to the inability to distinguish between males and females. However, the









mean concentration above 2 STD representing active luteal or follicular phases in

figure 2-5 still do not indicate sample values that much greater than in males. It may

be that male manatees do express relatively high concentrations of fecal estradiol

and/or progesterone. High amounts of estradiol have been measured in mature

stallions and boars (Bamberg et al. 1986; Palme 1994; Palme & Mostl 1993; Raeside

1978/1979; Schwarzenberger et al. 1996; Velle 1966). In fact, stallions have a daily

output of estrogens that are several fold higher than non-pregnant mares.

Ultimately, Experiment Three was not designed to collect serial fecal and

plasma samples that would be most appropriate to resolve some of the discrepancies

between plasma and fecal sample values. Plasma samples were only collected

opportunistically and individuals were sampled cross-sectionally. Neither was this

study intended to collect numerous reproductive tracts to review manatee

reproductive anatomy. The overall goal of this research was as a longitudinal study

of fecal hormone concentrations measured from identified individuals. Longitudinal

fecal collections would allow for a qualitative description of hormonal fluctuations

within an individual that would indicate estrous cycles and seasonal patterns.

Although it is not known whether the most appropriate fecal metabolites for the

Florida manatees are being measured, the hormone concentrations measured may still

provide meaningful fluctuations relative to an individual's estrous cycle and seasonal

activity levels. This has been demonstrated before in elephants. Elephant

progesterone concentrations correctly identified the length of the elephant estrous

cycle as -15 weeks, whereas 5c.-reduced metabolites, 5c-pregnane-3,20-dione

(5ca-DHP) and 5u-pregnane-3-ol-20-one (5ca-P-3-OH), are actually the major









circulating progestins (Brannian et al. 1988; Heistermann et al. 1997; Hess et al.

1983; Hodges et al. 1997; Hodges et al. 1994; Plotka et al. 1988).



Experiment Four: A Comparison of Gender, Location, Age and Individuals

In Experiment Four it was not unexpected to find that captive manatees have

different mean hormone concentrations between facilities and that captive animals

averaged higher concentrations than wild manatees (see Figure 2-7). Variables such

as diet and possible environmental stressors can profoundly affect hormone

concentrations. In general, depending upon how well a species reacts to captive

environments, regular availability of nutritious food, types of food, veterinary care,

and elimination of predatory and other natural threats of the animal's well being can

allow reproduction to occur at a much younger age and more frequently than in the

wild. It was anticipated that Homosassa Springs, in comparison to the other captive

facilities, would have lower mean hormone concentrations due to hay in the animals'

diet. This diet adds considerable bulk of undigested hay fibers to fecal mass and

directly affects the assay because this technique measures hormones on a per gram

basis.

Across all three hormones the trend was for male manatees to have higher

concentrations. This is unusual for estradiol and progesterone concentrations in

comparison to other mammalian hormone patterns. The expectation would be for

adult female estradiol and progesterone concentrations to be higher in comparison to

males. A possible explanation, as mentioned above, would be that male manatees do

produce higher fecal concentrations of sex steroids, this may be normal or artificially

induced by a captive environment. Other explanations for high male hormone









concentrations could be that something is artificially increasing the measurements of

male estradiol and progesterone such as cross-reactivity of antibodies with androgen

metabolites not previously tested, or it is possible progesterone and estradiol

metabolites are not the most appropriate for monitoring the adult female manatee

reproductive cycle.

The data from Figure 2-8 indicate that most males may be distinguished from

females by their testosterone concentrations and females may be separated from the

remaining males utilizing the progesterone/testosterone ratio. In the future, if another

progestin is discovered to be more appropriate, the larger value will only enhance the

separation of the male ratio from the female ratio. However, the current technique is

not refined enough to distinguish between different ages or reproductive groups, as

highlighted by Figure 2-9 and the studies of plasma samples and fecal samples from

necropsied animals.

Differences between individual mean hormone concentrations across the

period of a year or more are denoted in Figures 2-10, 2-11, and 2-12. The variation

between mean values may be reflective of reproductively active versus inactive

periods. As seen in Figures 3-1 through 3-12 certain animals displayed long periods

with no hormonal fluctuations or concentrations below detection levels (i.e. Stubbie

and Lorelei), in comparison to others with fluctuations throughout the sampling

period (Rosie and Amanda). An additional factor that may contribute to the variance

between individuals is differences in steroid recovery. As mentioned in Assay Trials

& Tribulations (see Appendix) steroid recovery within an individual was found to be

consistent, but not between individuals. Ideally, measurements of steroid recoveries









would have been made and calculated to normalize the variance between animals, but

the cost was prohibitive. This variance should be taken into consideration when

referring to manatees as a group, but should not interfere with the analysis of

longitudinal data from a single manatee.




Conclusions

This fecal RIA technique currently can not distinguish clearly between

different reproductive groups (i.e. adult vs. calves, pregnant vs. non-pregnant), with

the use of either plasma or tissue samples. Further refinement of the RIA will be

necessary to correctly identify the excreted metabolites and establish the parameters

defining reproductive groups and mature versus immature manatees. As mentioned

above, 17p3-estradiol is the predominant metabolite excreted in other species

suggesting that this metabolite may be an important indicator for manatees.

Progesterone may not be the major fecal metabolite, as reflected by other species, but

as in the elephant, may still indicate important aspects of female reproductive

function. Use of high-pressure liquid chromatography (HPLC) should be a simple

means of identifying the major metabolites present in feces and urine, if the

utilization of radiolabeled steroids is not feasible.

Corroboration with a greater number of plasma or urine samples now that

manatees are being trained for collections, collected more frequently than once or

twice a week, and from animals with known reproductive status will be key

components of future research. An intensive period of daily plasma (or as frequently

as possible) or urine samples in conjunction with fecal samples will be necessary to






73


identify the lag time between current hormones in the blood stream and steroids

excreted in feces.

Currently this technique has been properly validated in the lab and should be

sufficient in describing qualitative changes found within an individual manatee, to

document hormonal fluctuations related to the estrous cycle and seasonal changes.

Data related to groups of manatees will reflect variation between individuals but large

hormone fluctuations may be identified.












Table 2-1 Values of normality characteristics, before and after log transformation of fecal hormone concentrations.

E-F E-F E-M E-M P-F P-F P-M P-M T-F T-F T-M T-M
Skewness 8.87 1.21 3.38 -.24 4.32 -.77 .63 -1.82 10.47 -1.98 .73 -2.56
Kurtosis 90.28 .88 14.15 -1.24 27.28 -.28 .57 2.20 132.63 3.23 .14 8.29
Mean 19220.36 7.44 50742.74 9.34 230.24 4.21 452.45 5.17 276.42 4.88 17855.99 9.15
Median 444.44 6.10 20142.77 9.91 122.12 4.80 448.40 6.11 208.84 5.34 16399.81 9.71
Shapiro .0001 .0001 .0001 .0001 .0001 .0001 .0001 .0001 .0001 .0001 .0001 .0001
Wilk P

The headings across the top refer to estradiol for females (E-F), estradiol for males (E-M), progesterone for females (P-F),
progesterone for males (P-M), testosterone for females (T-F), and testosterone for males (T-M). The normal text heading indicates
values before transformation and the bold headings indicate values after log transformation.













Table 2-2 Gut transit and retention time of colored corn fed to adult manatees.


GUT TRANSIT TIME


TIME (days)


GUT RETENTION TIME

TIME (days)


NEW COLOR OLD COLOR OLD COLOR NEW COLOR
Newton (MS-Tm-9305) 7 7 6 8
Phoenix (MS-Tm-9304) 9 8 8 10
Romeo (MS-Tm-5701) 6 6 8 8
Juliet (MS-Tm-5801) 6 7 8 8


7.0 days


8.0 days


TIME LINE: 0 OLD COLOR -- NEW COLOR
(fed prior to study to mark fecals) (fed for Experiment One)


-- OLD COLOR
) (fed to resume marking fecals as prior to study)


Transit times (in days) it takes the fecal marker to get from first ingestion, to first be seen in the feces. Retention times (in days) it
takes the fecal marker from first visible traces in feces until last visible traces in feces.


Manatee


Average













Table 2-3 Lengths and weights of gut transit time manatees.

MANATEE LENGTH (cm) WEIGHT (kg)

Newton (male) N/A 409
Phoenix (female) 300 909
Romeo (male) 325 818
Juliet (female) 326 1136


















Table 2-4 Handling study normalized hormone concentrations.

MANATEE NORMAL WATER 30min WATER 2hrs AIR 30min


AIR 2hrs FREEZE
IMMEDIATE


FREEZE/ FREEZE/
THAW 3x THAW 5x


Ariel 100.00 67.00 88.00 88.00 72.00 92.00 130.00
HS-8602
Star 100.00 216.00 231.00 247.00 238.00 205.00 193.00 435.00
HS-8701
Amanda 100.00 126.00 94.00 121.00 130.00 145.00 114.00 183.00
HS-8601
Betsy 100.00 88.00 95.00 104.00 86.00 88.00 174.00 97.00 140.00
HS-9002
Rachel 100.00 101.00 97.00 75.00 101.00 100.00 83.00 203.00
HS-9005


DRY











Progesterone


Female





00 0 E E



0* 6
[]C


Male


* Fecal Prog.
] Plasma Prog.


:



* *


Estradiol


-6




-2-

- 2
I0


* Fecal Est. F
L] Plasma Est.


El E

E l

*I *LI
go go a


emale
S I

F =

* L


Male


F-1-



l i
0 0
.,_0

ge


Testosterone


- 3500
- 3000

- 2500
- 2000
- 1500

- 1000
- 500
-0


160
140
120
100
80
60
40
20
0


Figure 2-1 Matching fecal and plasma concentrations of progesterone, estradiol,
and testosterone for individual manatees.


120000 -

100000 -

80000 -

60000 -

40000 -

20000 -
0 -


7000
6000
5000
4000
3000
2000
1000
0


Fecal Test. Female n Male _
E] Plasma Test.

-
00 Q
: D <



0 [ E
-_













Progesterone


Female


Male


3500


3000


2500


2000


1500


1000


500


0


Female


Male


Female


Male


Figure 2-2 Mean plasma concentrations of progesterone, estradiol and testosterone that have matching fecal
samples, grouped by reproductive state. The bars indicate mean values + SEM and the squares indicate
individual animal values. The number of individuals in each group are as follows: Adult female = 7,
Pregnant = 1, Juvenile = 2, Adult male = 3, and Calf= 3.


Estradiol


Testosterone


NI
^^w \N


x c


<.















Progesterone


Female Male


Estradiol


Female


Testosterone


Male


Female


Male


100000 -



80000 -


60000 -



40000 -


20000 -



0


'r "r

NI^
N<< s


7000


6000


-0 5000
00

o 4000


2 3000


S2000


3.

-- +


Figure 2-3 Mean fecal concentrations of progesterone, estradiol and testosterone that have matching plasma
samples, grouped by reproductive state. The bars indicate mean values SEM and the circles indicate
individual animal values. The number of individuals in each group are as follows: Adult female = 8,
Pregnant = 5, Juvenile = 2, Adult male = 3, and Calf= 4.


600



500



0 400
40


300



200


1000


0














Progesterone


U
U


A














B

female male


Figure 2-4 A comparison between male and female plasma concentrations of progesterone, estradiol, and testosterone. The bars
with + SEM represent serum concentrations comparing male and female manatees from the current study where the female n =
18 and the male n = 4. The unfilled circles are individual values from this study. Bars with the same letters are statistically
similar and bars with different letters are statistically different at the 0.05 significance level. The shapes labeled by F study and
M study indicate female and male concentrations measured in a previous study by Francis-Floyd et al. (1991). The error bars for
the females aged 5.2 and 6.6 years represent 95% confidence interval.


4& &3


Estradiol


Testosterone













Progesterone


Estradiol


Testosterone


0







* 0



* I


, N, ,

v < i


700000 1--


600000 -


500000 -


400000 -


300000 -


200000 -


100000 -


0


6000


5000 -


4000 -


3000 -


2000 -


Figure 2-5 Comparison between male and female fecal concentrations of progesterone, estradiol and testosterone.
The circles represent individual values that contribute to the mean. The bars are mean values SEM for male and female
samples measured for all three hormones. The square is the mean value SEM of concentrations 2STD above each captive
manatee's individual mean of all samples collected, as indicated by unfilled circles. The Female n = 123, the Male
n = 96 and the Above 2STD n = 12.


*








0
0 o
0






ELI


1000


0


60000 -


50000 -


40000 -


30000 -


20000 -


10000 -


0 -


0
0

0
0



I






0


-













Table 2-5 Necropsied female manatee measurements of reproductive tracts, body length
and hormone concentrations.

ID/Name MSW MSW MSW Adair MSW MSW N95-612 MSW
96219 96160 9644 SWF- 96177 96171 SWF TM 96173
8638B 9513-B
Uterine Body 14.5 22.0 16.1 X 4.8 4.9 X 21.8
Length cm
Uterine Horn 32.5 32.5 23.4 X 14.1 12.1 X 25.3
A Length cm
Uterine Horn 17.6 30.0 20.0 X 12.7 12.2 X 25.8
B Length cm
Ovary A 15.1 13.5 14.2 10.4 10.3 6.6 X X
Length cm
Ovary A 14.2 9.8 7.0 9.3 3.6 4.6 X X
Width cm
Ovary B 13.0 12.5 10.5 12.5 5.6 6.6 X X
Length cm
Ovary B 12.0 10.7 8.8 7.3 3.3 4.8 X X
Width cm
# Follicles 38 22 1 19 0 0 X X
> 2mm
Largest 19.1 6.0 6.5 17.5 X X X X
follicle mmnun
# CL > 2mm 14 0 3 17 0 0 X X
Largest CL 9.2 X 6.3 6.5 X X X X
mm
Body length 297 320 285 X 224 215 233 274
cm
Comments Adult Adult Adult Adult Lg calf Lg. Lg. calf Adult
Luteal Follicular luteal luteal calf Cut into Lactating
phase dissected no several No
prior to uterus pieces ovaries
measure- only before
ment ovaries measure-
ments
Hormone No No No
values
Progesterone 474.43 55.97 156.03 380.72 252.09
ng/g
Estradiol pg/g 22363.29 572.91 444.44 444.44 8317.88
Testosterone 279.77 102.78 95.50 307.07 183.21
ng/g











Table 2-5 Continued


ID/Name MSW MSE MSW MSE MSW MSW MSW MSW
9608 9713 96244 9723 96161 9609 9714 96245
Uterine Body 4.9 12.6 11.8 4.2 6.3 6.6 14.5 22.7
Length cm
Uterine Horn 10.4 15.4 10.8 17.5 7.0 9.7 19.6 33.7
A Length cm
Uterine Horn 10.7 11.9 13.0 17.9 6.1 10.2 23.5 33.6
B Length cm
Ovary A 5.8 8.8 9.3 6.9 7.3 4.7 11.2 14.1
Length cm
Ovary A 6.2 6.5 6.6 6.0 2.7 3.6 10.0 13.8
Width cm
Ovary B 6.9 8.0 8.9 7.7 X X 10.0 11.3
Length cm
Ovary B 7.1 4.2 5.9 6.9 X X 5.3 11.0
Width cm
# Follicles 0 0 33 45 0 0 23 0
> 2mm
Largest X X 8.9 9.5 X X 10.4 X
follicle mmnun
# CL > 2mm 0 0 0 0 0 0 24 52
Largest CL X X X X X X 9.1 6.5
mm
Body length 170 232 198 241 211 216 292 364
cm
Comments calf Lg. calf Lg. Calf Lg. calf Lg. calf Adult Adult
calf Cut into One Luteal Luteal
several ovary phase
pieces missing.
before Surface
measure of ovary
-ments. is
Missing degraded
one
ovary.
Juv. or
young
adult
Hormone No No
values
Progesterone 516.08 332.88 263.77 493.96 2.34 26.55
ng/g
Estradiol pg/g 3285.74 9925.19 4753.98 444.44 444.44 452.96
Testosterone 312.63 139.50 143.20 95.38 141.09 184.58
ng/g










Table 2-5 Continued


ID/Name MSW MSW MSW MEC MEC MEC
96221 9742 9734 9723 9625 9715
Uterine Body 25.6 23.2 20.2 10.5 5.4 6.8
Length cm
Uterine Horn A 44.0 36.2 38.8 24.0 17.1 13.4
Length cm
Uterine Horn B 33.8 20.9 35.7 21.6 13.9 12.2
Length cm
Ovary A Length cm 17.4 10.8 15.1 10.7 6.6 9.4
Ovary A Width cm 11.9 8.6 12.3 14.0 10.2 9.9
Ovary B Length cm 13.8 13.1 X 9.4 7.8 9.5
Ovary B Width cm 8.8 8.4 X 8.5 5.7 7.1
# Follicles > 2mm 91 84 20 48 9 15
Largest follicle mm 10.6 8.5 11.6 11.3 9.5 11.7
# CL > 2mm 55 28 29 9 0 0
Largest CL mm 8.6 9.8 10.0 4.6 X X
Body length cm 338 274 280 285 232 247
Comments Adult Adult Adult, Adult. Lg calf Juv.
Pregnant Pregnant Pregnant. Luteal Many Many
With Ovary B has Uterus hemorrha hemorrha
placenta been cut, has been gic gic
follicles and cut structures structures
CL only before
counted measure-
from ovary ments
A
Hormone values No
Progesterone 689.52 330.93 490.38 189.58 272.52
ng/g
Estradiol pg/g 55489.40 18917.92 20498.36 2427.56 6780.11
Testosterone ng/g 106.96 80.00 80.00 145.23 149.94




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs