Sirenian Conservation Genetics and Florida Manatee (Trichechus manatus latirostris) Cytogenetics

Material Information

Sirenian Conservation Genetics and Florida Manatee (Trichechus manatus latirostris) Cytogenetics
Kellogg, Margaret
Place of Publication:
[Gainesville, Fla.]
University of Florida
Publication Date:
Physical Description:
1 online resource (160 p.)

Thesis/Dissertation Information

Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Veterinary Medical Sciences
Veterinary Medicine
Committee Chair:
McGuire, Peter M.
Committee Members:
Reep, Roger L.
Lefebvre, Lynn W.
King, Timothy
Graduation Date:


Subjects / Keywords:
Alleles ( jstor )
Chromosomes ( jstor )
DNA ( jstor )
Dugongs ( jstor )
Haplotypes ( jstor )
Manatees ( jstor )
Microsatellites ( jstor )
Population genetics ( jstor )
Population structure ( jstor )
Species ( jstor )
Veterinary Medicine -- Dissertations, Academic -- UF
afrotherian, belize, conservation, cytogenetics, dna, dugon, dugong, endangered, florida, genetics, latirostris, manatee, manatus, microsatellite, mitochondrial, paenungulata, population, puerto, rico, sirenian, threatened, trichechus, zoofish
City of Gainesville ( local )
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Veterinary Medical Sciences thesis, Ph.D.


The threatened West Indian manatee is a slowly reproducing aquatic mammal, whose small, isolated populations are negatively impacted by habitat destruction and anthropogenic mortality. Long-term exploitation and small population sizes can lower genetic diversity, resulting in decreased fitness, reduced adaptation to environmental change, and potentially lead to extinction. Consequently, genetic studies, using microsatellite and mitochondrial DNA, were implemented to quantify the genetic diversity and identify unique populations or regions in need of protection. These studies will facilitate manatee conservation, management, and recovery efforts. The West Indian manatee is composed of the Florida (Trichechus manatus latirostris) and Antillean manatee (T. m. manatus) subspecies. The Florida and Antillean Puerto Rico manatees are managed as a single endangered population under the U.S. jurisdiction of the Endangered Species Act. A recent status review suggested that the species be downlisted to threatened, primarily due to the recovery of the Florida population. In this study, the Florida and Puerto Rico populations were determined to be genetically distinct, and in conjunction with the differing habitats, threats, and population sizes, it is recommended that each population be managed separately. Genetic studies of the Belize and Puerto Rico manatee populations detected reduced variation and subtle genetic structure, suggesting limited vagility and the potential for unique populations. The conservation of the subpopulations and delineation of corridors could maintain and potentially improve the low genetic diversity. Moreover, Belize is believed to be a source population, where adequate protection could lead to increased expatriation and repopulation of exploited regions. To improve sirenian population genetic analyses, dugong (Dugong dugon) and manatee microsatellite primers were compared and the most informative loci for each group were selected. A panel of 11 dugong and 13 manatee cross-species and species-specific markers obtained better results utilizing fewer primers, improving time and cost effectiveness. Finally, a cytogenetic study applied Zoo-FISH techniques to investigate the cross-species homology and evolutionary relationship of human and Florida manatee chromosomes. The clade Paenungulata was supported, linking manatees, hyraxes, and elephants and confirming their assignment to the Afrotheria super-order. Furthermore, Afrotheria, Xenarthra, or a combination of both was supported as the basal eutherian super-order. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis (Ph.D.)--University of Florida, 2008.
Adviser: McGuire, Peter M.
Electronic Access:
Statement of Responsibility:
by Margaret Kellogg.

Record Information

Source Institution:
Rights Management:
Copyright Kellogg, Margaret. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
LD1780 2008 ( lcc )


This item has the following downloads:

Full Text




2008 Margaret Elizabeth Kellogg 2


To my mother, Mary Sudholt Kellogg, for he r unwavering support and encouragement and instilling scholarly exce llence in all I do. 3


ACKNOWLEDGMENTS I would like to thank all the ex ceptional people who made this dissertation possible. My mentor, Dr. Peter McGuire has pr ovided steadfast support, encouragement, and expert editorial advice. I thank him for providing me with an exte nsive and multi-discipline scientific education. I will always consult the life lessons he has so generously bestowed upon me. I also greatly appreciate the numerous national and internat ional fieldwork and sc ientific conference opportunities. Mr. Robert Bonde has been extr emely munificent with his immense sirenian knowledge. I appreciate all of the time he spent discussing and editing my work. This dissertation would not be possi ble without the samples and co llaborations he has acquired throughout the past 30 years. I am grateful to have such wonderf ul mentors and role models. Dr. Kimberly Pause provided the foundation on which my population genetics knowledge is based. She has generously assisted with ever y aspect of this disse rtation and has been a wonderful friend. Mr. Sean McCann has enthusia stically contributed to this project, providing the assistance I needed to complete it. Ma ny thanks to Ms. Ginger Clark and the UF ICBR Genetic Analysis Laboratory for providing ex tensive technical and wr iting assistance and laboratory space for this project. Ms. Cathy Beck, Ms. Susan Butler, Mr. Jim Rei d, and the USGS Sirenia Project staff have been generous with their time and have grac iously included me in many wonderful manatee captures. I am very thankful to partake in such exciting work with the experts in the field. The USGS provided funding for this project and th e samples collected under the Sirenia Project permit (USFWS Wildlife Research Permit MA791721/4). I would like to thank my committee members, Drs. Roger Reep, Roberto Zori, Timothy King, and Lynn Lefebvre, for providing guidance and editorial advice. Drs. Charles Courtney, Ruth Francis-Floyd, and the UF Aquatic Animal H ealth Program have been extremely generous 4


in providing financial support for this pr oject and wonderful academic and fieldwork opportunities. I would like to thank Drs. R oberto Zori, Thomas Dennis, a nd Mr. Brian Gray, and the UF Cytogenetics Laboratory staff for providing copious technical support and for taking time out of their important work and busy schedules for the mana tee projects. I am also greatly indebted to Ms. Melanie Pate in Dr. John Ha rveys Laboratory for her knowledge, time, and equipment. She is a wonderful and patience teacher. Additiona lly, Ms. Linda Green and Ms. Diane Duke in the ICBR hybridoma core have provided expert assistance and technical skills. Lastly, I would like to thank Dr. William Farmerie and his laboratory for support, advice, and superior knowledge in the molecular field. I am grateful to the collabor ators that made this work possible. I thank Dr. Antonio Mignucci-Giannoni (Environmental Research) for the samples and his assistance with the Puerto Rico study. Dr. James Powell (The Sea to Shore Alliance) and Ms. Nicole Auil-Gomez (Wildlife Trust) provided samples and assist ance with the Belize study. Drs. Janet Lanyon, Damien Broderick, Jennifer Ovenden, and Ms. Helen Peereboom (University of Queensland) provided samples, primers, and statistical analysis for the dugong micros atellite study. Drs. Roscoe Stanyon (University of Florence), Sandra Bu rkett, and Mr. Gary St one (National Cancer Institute) provided the Zoo-FISH reagents, conducted cytogenetic experime nts, and a great place to stay in Italy! Additiona lly, I thank them for their immens e editorial assistance and superb scientific un derstanding. I am ever indebted to my friends and family for all of their encouragement and support during the last four years. I thank Mr. Charles Hunter, the love of my life, for his support, excellent editorial suggestions, and always providing a willing ear to listen and a keen mind to 5


discuss science (and for cooking dinner the last two months!). Many thanks to my Melbourne Beach Sudholt family, Grandma, Marty, Ginny, Lois, Richard, and James, for providing unending prayers, love, and the courage to keep going; they have been my support-system through this process. I thank Ms. Meagan Beresford, a wonderful friend, for her encouragement and unwavering faith in my work. Always follow your dreams; they will transpire! I thank my brother, Mr. E. Harrison Kellogg, for his kindness, support, and wi sdom beyond his age. Finally, I thank my mother, Ms. Mary S. Kellogg, whom this dissertation is dedicated to; she has instilled in me a deep love of learning, a spirit of pers everance (perfection), and a desire for academic excellence that were the foundation for this degree. Her unwavering support, encouragement, and faith have tremendously contributed to the co mpletion of this dissert ation. I thank her for inspiring the fortitude I needed to succeed! 6


TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4LIST OF TABLES................................................................................................................. ........10LIST OF FIGURES.......................................................................................................................11ABSTRACT...................................................................................................................................12 CHAPTER 1 INTRODUCTION................................................................................................................. .14The Order Sirenia.............................................................................................................. .....14Manatee Life History........................................................................................................... ...16West Indian Manatee Conservation........................................................................................19Conservation Genetics............................................................................................................21Molecular Markers..........................................................................................................23Measuring Mammalian Ge netic Differentiation.............................................................26Population Genetic Theory.....................................................................................................2 8Hardy-Weinberg Equilibrium..........................................................................................28Statistical Methods..........................................................................................................30Application of Conservation Genetics....................................................................................31Taxonomic Standings......................................................................................................31Habitat Fragmentation.....................................................................................................32Pedigree Reconstruction..................................................................................................32Forensic Applications......................................................................................................33Cytogenetic Studies............................................................................................................ ....34Marine Mammal and Manatee Ch romosome Investigations...........................................35Zoo-FISH.........................................................................................................................36Mammalian Zoo-FISH....................................................................................................37Microdissection...............................................................................................................3 9Sirenian Molecular Studies.....................................................................................................39Sirenian Population and Conservation Genetic Studies..................................................40Belize and Puerto Rico Ma natee Conservation Studies..................................................472 COMPREHENSIVE GENETIC INVESTIGATION RECOGNIZES EVOLUTIONARY DIVERGENCE IN THE FLORIDA ( Trichechus manatus latirostris ) AND PUERTO RICO (T. m. manatus ) MANATEE POPULATIONS AND SUBTLE SUBSTRUCTURE IN PUERTO RICO.................................................................49Introduction................................................................................................................... ..........49Materials and Methods...........................................................................................................52Sample Collection and DNA Extraction.........................................................................52Mitochondrial DNA Analysis..........................................................................................53 7


Microsatellite Analysis....................................................................................................54Statistical Analysis..........................................................................................................5 4Cytogenetic Analyses......................................................................................................57Results.....................................................................................................................................57Mitochondrial Sequence Analysis...................................................................................57Microsatellite Marker Analysis.......................................................................................58Cytogenetic Analyses......................................................................................................61Discussion...............................................................................................................................62Florida and Puerto Rico Mitochondrial DNA.................................................................62Florida and Puerto Rico Microsatellite Analysis.............................................................64Population Structure and the Environment......................................................................65Puerto Rico Genetic Divers ity and Geographic Division................................................66Preservation of the Manatee in Puerto Rico....................................................................693 CONSERVATION GENETICS OF THE ANTILLEAN MANATEE ( Trichechus manatus manatus) POPULATION IN BELIZE: THE LAST CARIBBEAN STRONGHOLD.....................................................................................................................76Introduction................................................................................................................... ..........76Materials and Methods...........................................................................................................80Sample Collection and DNA Extraction.........................................................................80Mitochondrial DNA Analysis..........................................................................................80Microsatellite Analysis....................................................................................................81Statistical Analysis..........................................................................................................8 2Cytogenetic Analyses......................................................................................................84Results.....................................................................................................................................84Mitochondrial Sequence Analysis...................................................................................84Microsatellite Marker Analysis.......................................................................................85Cytogenetic Analyses......................................................................................................87Discussion...............................................................................................................................87Source and Sink...............................................................................................................8 8Tracking...........................................................................................................................89Mainland versus Island Habitats.....................................................................................89Tropical versus Temperate Climates...............................................................................91Current Population and Future Directions.......................................................................914 CROSS-SPECIES COMPARISON OF AUSTRALIAN DUGO NG AND FLORIDA MANATEE MICROSATELLITE LOCI AND THE CHARACTERIZATION OF HIGHLY INFORMATIVE MARKER-PANELS................................................................100Introduction................................................................................................................... ........100Materials and Methods.........................................................................................................101 8


Results...................................................................................................................................103Amplification of Mana tee Loci in Dugong...................................................................104Amplification of Dugong Loci in Manatee...................................................................104Most Informative Markers: Dugong and Manatee Primers Combined.........................105Discussion.............................................................................................................................106Cross-Species Amplification.........................................................................................106Conservation Implications.............................................................................................1075 CHROMOSOME PAINTING IN TH E MANATEE STRONGLY SUPPORTS AFROTHERIA AND PAENUNGULATA..........................................................................112Introduction................................................................................................................... ........112The Florida Manatee......................................................................................................113Previous Cytogenetic Reports on Manatees..................................................................113Methods................................................................................................................................113Results...................................................................................................................................115Discussion.............................................................................................................................116Support for the Tethytheria and Paenungulata Assemblage..........................................118Branching Order in Afrotheria......................................................................................119The Root of the Eutherian Tree.....................................................................................119Conclusions...........................................................................................................................1216 CONCLUSION................................................................................................................... ..125Marine Mammal Population Genetics..................................................................................125Within Population Variation.................................................................................................127Population Bottlenecks..................................................................................................127Gene Flow in Island and Mainland Populations............................................................129Among Population Variation and Gene Flow......................................................................131Moderate Variation and Gene Flow..............................................................................131Panmictic P opulations...................................................................................................134Conclusions...........................................................................................................................136LIST OF REFERENCES.............................................................................................................139BIOGRAPHICAL SKETCH.......................................................................................................160 9


LIST OF TABLES Table page 2-1. Characteristics of the 15 polymorphic micr osatellite loci implemented on the Puerto Rico manatee (Trichechus manatus manatus) samples.....................................................722-2. Pairwise FST and RST values generated from a survey of 15 microsatellite loci from Trichechus manatus manatus in five geographic regions in Puerto Rico..........................723-1. Characteristics of the 16 microsate llite loci implemented on Belize manatee ( Trichechus manatus manatus) samples............................................................................953-2. Pairwise FST and RST values generated from a survey of 16 microsatellite loci in Trichechus manatus manatus in three geographic regions in Belize.................................954-1. Characterization of Dugong dugon and Trichechus manatus latirostris primers amplified on D. dugon samples.......................................................................................1084-2. Characterization of Dugong dugon and Trichechus manatus latirostris primers amplified on T. m. latirostris samples.............................................................................1104-3. Dugong dugon and Trichechus manatus latirostris marker-panel summaries....................1114-4. Sibling ( P(ID)Sib) and Hardy-Weinberg equilibrium (HW P(ID)) probability of identity values for the four study groups.......................................................................................1115-1. Number of segments homologous to hum an chromosome found in Afrotherian species...122 10


LIST OF FIGURES Figure page 2-1. Puerto Rico bathymetric map with location and mitochondrial haplotype assignment of captured or recovered manatees.........................................................................................732-2. Florida and Puerto Rico i ndividual neighbor-joining tree.....................................................742-3. Puerto Rico population unrooted neighbor-joining tree.......................................................752-4. Summary plot of q estimates generated by the seque ntial cluster analysis of the program STRUCTURE performed on the Florida and Puerto Rico T. manatus genotypes...........................................................................................................................753-1. Geographic map of Belize with the three st udy sites inset; Belize City Cayes, Northern and Southern Lagoon, and Placencia Lagoon....................................................................963-2. Summary plot of q estimates generated by the seque ntial cluster analysis of the program STRUCTURE performed on the Belize and Florida T. manatus genotypes...........973-3. Summary plot of q estimates performed on the q-sorted Florida group and Belize T. manatus genotypes, indicating two Belize clusters...........................................................973-4. Belize and Florid a individual un-rooted neighbor-joining tree.............................................983-5. Belize and Florida population un-rooted neighbor-joining tree............................................995-1. Examples of human chromosome hybridizations in the manatee.......................................1235-2. The manatee karyotype and ideogram.................................................................................124 11


Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy SIRENIAN CONSERVATION GENETI CS AND FLORIDA MANATEE ( Trichechus manatus latirostris ) CYTOGENETICS By Margaret Elizabeth Kellogg December 2008 Chair: Peter M. McGuire Major: Veterinary Medical Sciences The threatened West Indian manatee is a slowly reproducing aquatic mammal, whose small, isolated populations are negatively imp acted by habitat destruction and anthropogenic mortality. Long-term exploitation and small population sizes can lower genetic diversity, resulting in decreased fitness, reduced adaptatio n to environmental change, and potentially lead to extinction. Consequently, genetic studies, using microsatellite a nd mitochondrial DNA, were implemented to quantify the genetic diversity and identify unique populations or regions in need of protection. These studies will facilitate manatee conservation, management, and recovery efforts. The West Indian manatee is composed of the Florida (Trichechus manatus latirostris ) and Antillean manatee ( T. m. manatus ) subspecies. The Florida and An tillean Puerto Rico manatees are managed as a single endangered population under the U.S. jurisdic tion of the Endangered Species Act. A recent status review suggested that the species be dow nlisted to threatened, primarily due to the recovery of the Florida population. In this study, the Flor ida and Puerto Rico populations were determined to be genetically distinct, and in conjunction with the differing habitats, threats, and population sizes, it is recommended that each population be managed separately. 12


Genetic studies of the Belize and Puerto Rico manatee populations detected reduced variation and subtle genetic st ructure, suggesting lim ited vagility and the potential for unique populations. The conservation of the subpopulations and delineation of corridors could maintain and potentially improve the low genetic diversity. Moreover, Belize is believed to be a source population, where adequate protectio n could lead to increased e xpatriation and repopulation of exploited regions. To improve sirenian populat ion genetic analyses, dugong ( Dugong dugon) and manatee microsatellite primers were compared and the most informative loci for each group were selected. A panel of 11 dugong and 13 manatee cross-species and species-specific markers obtained better results utilizing fewer primer s, improving time and cost effectiveness. Finally, a cytogenetic study a pplied Zoo-FISH techniques to investigate the cross-species homology and evolutionary relati onship of human and Florida manatee chromosomes. The clade Paenungulata was supported, linking manatees, hyra xes, and elephants and confirming their assignment to the Afrotheria super-order. Furthe rmore, Afrotheria, Xenarthra, or a combination of both was supported as the basal eutherian super-order. 13


14 CHAPTER 1 INTRODUCTION The Order Sirenia Manatees and dugongs compose the Order Sireni a, and are large, herbivorous, obligate aquatic mammals. Early European explorers mistook these animals for mermaids. The order was named sirenia when authors confused mermai ds (half-women, half-fish) with the sirens of Greek mythology (half-women, half-birds). Sirens are sea nymphs who lured ancient mariners to their doom on dangerous rocks with mesmeriz ing songs. The name manatee was either derived from the Latin word manatus, meaning having hands, as their pectoral fins resemble human hands, or from the Carib word manati, meaning udder, and referring to the mammary glands located beneath their flippers. The name dugong was derived from the Malay word duyung, meaning lady of the sea. Evolving from terrestrial qua drupeds, the order Sirenia or iginated near Africa 45-50 million years ago (Mya), during the Middle Eocene. The most primitive sirenian known to date, Prorastomus (48 -37 Mya), was discovered in Jamaica. The four-legged amphibious creature possessed adaptations still present in modern si renians. These morphological characteristics include dense rib bones and an ancient fifth pr emolar, lost in the Cr etaceous by all other mammals. The most recent common ancestor to extant Sirenians, Protosiren (37-33 Mya) possessed a modern sirenian skeleton that lacked hind limbs (Domning 1981b; Domning 1982b; Domning 2001). The order is currently repres ented by five species in two families, Dugongidae and Trichechidae. In the family Dugongidae, the subfamily Dugonginae appear ed in the Oligocene (33.9-23 Mya). Dugongidae contai ns two modern species, the extant dugong, and the extinct Stellers sea cow. The dugong ( Dugong dugon) inhabits tropical and su btropical regions in the


Indian and Pacific Oceans. Its closest living relative, the Stellers sea cow ( Hydrodamalis gigas ) was hunted to extinction by 1768 (Stejneger 1887), only 27 years after its discovery by modern European humans. It lived in subpolar regions a nd fed on marine algae. Fossil records indicate that the adult sea cow was 25 feet in length, ha d flukes seven to eight feet long, and weighed more than 8,000 pounds. The other Sirenian family, Trichechidae, comprises the genus Trichechus speculated to have evolved from the extinct Ribodon genus (5.3-3.6 Mya). The two genera have supernumerary molars, an adaptation for a grass diet, which are replaced horizontally throughout life (Domning 1982b; Domning 2001). Trichechids originated in South American lagoons and expanded into the Caribbean in the Late Pliocene to Early Pleistocene. The family Trichechidae comprises three liv ing species of manatee, the West Indian, Amazonian, and West African manatee. The West Indian manatee, Trichechus manatus includes two subspecies, the Florida manatee, T. m. latirostris, and Antillean manatee, T. m. manatus The Florida manatee is loca ted throughout the coastal areas of the southeastern United States. The Antillean manatee is found in the Caribbean, Mexico, Central America and South America to the northeast coast of Brazil. West Indian manatees are distributed throughout rivers, estuaries, and marine environm ents. The Amazonian manatee, T. inunguis is the only exclusively freshwater trichechid, and is restricted to the Amazon basin. The West African manatee, T. senegalensis resides on the west coast of Africa from Senegal to Angola. Morphologically, manatees are similar. However, T. manatus tends to be larger and T. inunguis has a white pigmentation patch on its chest or ab domen and does not have nails on its pectoral fins. 15


Sirenias closest phylogenetic relatives are the orders Proboscidea, elephants, and Hyracoidea, hyraxes. Together, these orders form the clade, Paenugulata (De Jong & Zweers 1980; Kellogg et al. 2007). Sirenians and proboscideans diverged approximately 50-60 Mya (Rainey et al. 1984). Manatee Life History Manatees are large, long liv ed, and slowly reproducing a quatic mammals. They have a lifespan of up to 60 years. Upon adulthood, Florida manatees reach an average of 2.7 m (9 ft) and weigh between 400-545 kg (900-1,200 lb), with fe males of comparable age generally being larger. A maximum of 4.0 m (12 ft) and 1,772 kg (3,900 lb) has been recorded. Female manatees become sexually mature be tween the ages of 3-5 and produce a calf every 2-5 years. Gestation is about 13 months and a calf may remain dependant on its mother for two years. Newborn calves have an averag e mass of 27 kg (60 lbs) and are 1.2-1.37 m (4-4.5 ft) long. Due to the low reproductive rate of the species, the human inflicted mortality rate may exceed the populations ability to grow (Bossart 1999). The cow-calf pair is the main social unit for the species. Other social groups are tran sient and include mating herds and warm-water aggregations (Reynolds III & Odell 1991). Extant sirenians inhabit tropical and subtropi cal regions and are the only strictly aquatic herbivorous mammals. West Indian and West African manatees are adapted to shallow water in fresh, brackish, or marine environments. The Amazonian manatee is limited to fresh water habitats. Water depths between 0.9-2.1 m (3-7ft) ar e preferred and manatees rarely dive deeper than 6.1 m (20ft). Most manatees migrate seasonally, between winter warm-water sites and summer distribution areas. Track ing studies and photo identificati on have indicated that some Florida manatees annually mi grate long distances (Deutsch et al. 2003). Tracked manatees have ranged up to 850 km (Reid et al. 1991). 16


A diet constrained by low calorie aquatic plants may contribute to the species limited cold tolerance. Manatees cannot maintain sufficient body heat in cool water and are restricted to temperatures above 17-20C (Glaser & Reynolds III 2003). To optimize the limited vegetation calories and minimize energy expenditure, a low meta bolic rate has evolved. The metabolic rate is 15-22% of what is expected for similar si zed terrestrial animals (O'Shea & Reep 1990). Sirenians have evolved large body sizes to proce ss substantial quantities of food for sufficient energy production. The manatee diet consists of seagrasses in marine and estuarine systems, (Syringodium filiforme, Thalassia testudinum Halodule wrightii ) and various submerged and floating fresh water vegetation, ( Vallisneria americana Ceratophyllum demursum, Hydrilla verticillata Myriophyllum spicatum Ruppia maritime Potamogeton pectinatus Elodea canadensis ). The fibrous diet along with unintentionally consumed substrate quickly erodes the teeth and has selected for a supernumerary molar dental adaptation, otherwise known as hind-molar progression (Domning 1982b). The molars wear dow n and are stimulated to move forward in the jaw. Bone material from the socket in front of each tooth is replaced on the back of the tooth, effectively moving the teeth forwar d. New molars erupt in the back of the row approximately every year and manatees can produce an unlimite d number of teeth in their lifetime (Domning 1983). This form of molar replacement is seen on ly in one other species, the Nabarlek rock wallaby ( Peradorcas concinna ; Sanson 1980). The retention of fi ve premolars, as opposed to three in most herbivores, is a synapomorphic trait linking all present day sirenians (Domning 1994). Manatees have the capability to remember the precise locations of resources. An abnormally thick cerebral cortex is used for higher order information processing and long-term 17


memory storage (Purves et al. 2001; Reep & Bonde 2006). The size and layering patterns are comparable to carnivores and primates (Marshall & Reep 1995; Reep et al. 1989). The large cerebral cortex may be used to store the detailed locations of fresh water, aquatic plant beds, and winter warm water refuges. Manatees have strong winter site fidelity and predictable patterns of movement over great distances, indicating remarkable memoriza tion skills (U.S. Fish and Wildlife Service 2001). Manatees have poor vision, which is in part due to highly vascularized corneas (Bauer et al. 2003; Harper et al. 2005). Additionally, the majority of their time is spent in turbid water, which severely limits the use of eyesight. To compensate for reduced vision, manatee sensory systems have adapted to the aquatic environment. Instead of relying on eyesight, sensory tactile hairs on the face and body are used to identify vibrations and movements in the water column (Reep et al. 2002). This is analogous to the lateral line sens ory organ in fish. The manatee is one of the few mammals to have tactile hairs on the body. To feed easily on submerged vegetation, mana tees are negatively buoyant, sinking at less than 10 m, while positively buoyant at the surface (Kipps et al. 2002). Dense rib bones provide a ballast or weight, allowing them to remain submerged. The dense osteosclerotic rib bones lack marrow, except at the tips (Clifton et al. 2008). This is unusual, as most mammals have high concentrations of marrow in their long bones. Marrow is present in the manat ee skull, vertebrae, and sternu m. Hematology studies indicate that manatees have low numbers of white blood cells in comparison to domestic species. Manatees have both large and small lymphocytes monocytes, and basophils. They do not have typical neutrophils, but instead pseudoeosinophils or heterophils, whose granules stain pink with the Wright-Giemsa stain. Elephants and hyrax es also have cells with similar staining 18


characteristics, linking the species (Schalm 1965). Manatees have large but low numbers of red blood cells, prolonging the rate of oxygen diffusion during a dive (Medway et al. 1982). Despite the lack of marrow in the long bones, manatees have a superior immune system, wound healing, and repair process compared to other animals (Bonde et al. 2004). The manatee immune system appears highly developed to prot ect against the harsh marine environment and the effects of human-related injury (Bossart et al. 2002). They appear res ilient to natural disease and traumatic human-related injury (Bonde et al. 2004; Buergelt & Bonde 1983; White & Francis-Floyd 1990). In fact, Bossart points to the immune system as a reas on for their ability to survive in heavily polluted water (1999). Manatees are relatively immune to infectious agents, and show no clinical signs to infection by pseudorabies virus, San Miguel sea lion virus type 1, and equine encephalitis, porpoise and dolphin morbilliviruses (Duignan et al. 1995; Geraci et al. 1999). West Indian Manatee Conservation All three manatee species are considered vulnerable by the Inte rnational Union for Conservation of Nature and Natural Re sources (IUCN Thornback & Jenkins 1982). T. manatus and T. inunguis are listed in Appendix I and T. senegalensis is listed in Appendix II of the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES). The manatee was the first marine mammal to re ceive legal protection from hunting, granted by Florida in 1893. Additionally, the species received protecti on from the Endangered Species Preservation Act in 1967 and the Marine Mammal Protection Act in 1972. The U.S. Federal government granted the Florida manatee endanger ed status in 1973, due to excessive mortality and small population size (U.S. Fi sh and Wildlife Service 1979). Manatees have been hunted throughout their habitat, exterminating or severely reducing some populations. Records from 1542 reveal that indigenous people hunted Amazonian 19


manatees for their meat and hides to make sh ields, shoes, or canoes (Best 1984). From 19351954, approximately 200,000 Amazonian and West In dian manatees were poached in Brazil (Best 1982; Domning 1981a; Domning 1982a). It is estimated that 4,000-7,000 Amazonian manatees were killed per year. The carcasses were rendered for oil and meat, and the hides were used to make machine belts, pulleys, and hoses (Best 1984). Currently, the T. manatus population in Brazil has approximate ly 500 individuals and is cons idered critically endangered by the Brazilian Action Plan for Aquatic Mamma ls (IBAMA 2001; Lima 1997; Luna 2001). The hunted Florida manatee population wa s reduced gradually from the 16th century through the early 1900s, with the largest reduc tion in the 1800s by European settlers (Hartman 1974). Although no longer hunted, in 2006 417 Florida manatees died from natural and anthropogenic causes in an estimated populati on size of 3,000-4,000 individuals (Florida Fish and Wildlife Research Institute 2007b). Recent human-related manat ee mortality has been caused by: crushing in flood c ontrol structures (O'Shea et al. 1985), entanglement or ingestion of debris or fishing gear (Beck & Barros 1991; de Thoisy et al. 2003; Marmontel et al. 1997; Mignucci-Giannoni et al. 2000) blunt or sharp trauma fr om collision with water vessels (Marmontel et al. 1997), and habitat destruction and alteration (de Thoisy et al. 2003). Annually, direct and indirect human-related mo rtality can account for 33% of all Florida manatee deaths (Bossart 1999). In the late 1980s, more than 100 manatees were found dead each year in a population of approximately 1,200 i ndividuals. Many deaths were anthropogenic in nature, with boat strikes cau sing the most mortality (O'Shea et al. 1985). To decrease the number of human-related mortalities, manatee sanctuaries an d boat slow speed zones were implemented (U.S. Fish and Wildlife Service 199 6). Studies on life history and patterns of 20


seasonal movements and distribut ions provided information for management plans and habitat regulations (Deutsch 2000; Deutsch et al. 2003; O'Shea & Hartley 1995; Rathbun et al. 1990). Many manatee populations have remained small due to a variety of natural mortality. Natural mortality is most frequently due to ne onatal complications, cold weather stress (Bossart et al. 2003), and brevetoxicosis from red tide (Bossart et al. 1998). These events often lead to a large number of manatee deaths. In 1996, 150 manatees died from brevotoxicosis alone (U.S. Marine Mammal Commission 1996). Integrative approaches are n eeded to investigate, manage and monitor the extensively threatened Trichechus family. The conservation of manatees will require multi-discipline approaches including biological, ecological, and genetic data. In particular, conservation genetics can assist with id entifying populations, genetic substructure, and quantifying immigration. Conservation Genetics Conservation biology is the study of organisms, communities, and ecosystems directly or indirectly affected by human ac tivities (Soule 1985). Within the discipline of conservation biology, genetics has been a powerful tool to quan tify genetic variation an d protect and manage threatened species (Frankham et al. 2002). Molecular studies can as sist in protecting populations that are small or experiencing environmental change (Groves et al. 2002; Raven & Wilson 1992). A major tenet of conservation genetics is the preservation or improvement of genetic diversity to allow greater amplitude for adaptati on, speciation, or evolution. Reduced genetic diversity can result in a decrease of fecundity and survival and compromises a species ability to evolve and endure environmental change, potentia lly resulting in exti nction (Avise 2004). Genetic diversity has been linked to fitness, popu lation size, number of i nbred individuals, and 21


population persistence. Sma ll populations can suffer from inbreeding and reduced immune system variation, which increases susceptib ility to disease, e.g., the cheetah ( Acinonyx jubatus ; O'Brien 1994), fitness and physical def ects, e.g., the Florida panther ( Felix concolor coryi ; Roelke et al. 1993), and overall lower popul ation viability (Sherwin et al. 2000). Conservation genetics can identify unique popula tions and address issues of low variation. Substantial genetic variation, the degree varying from species to species, is essential for population vitality and persistence (O 'Brien 1994). An extended period of low population abundance, or near extinction of the population, must occur for a substantial loss of genetic variation. Bottlenecks result in se vere reductions in genetic varia tion due to loss of individuals, possibly resulting in inbreeding depression. During inbreeding depression, fitness may often decline in populations that recover more slowly or have reductions in heterozygosity. Low heterozygosity may be due to the buildup of mildly deleterious alleles a nd lethal heterozygous recessive alleles (Amos & Rubinsztein 1996; Westemeier et al. 1998). Delays in population recovery can cause further losses in fitness. Both theory (Nei 1975) and experiment (Miller & Hedrick 2001) have suggested that the impacts of severe bottlenecks may be minimized with rapid population recovery. Conservation genetics can also assist in forensic applications defining subspecies, understanding species biology, id entification of management units, and the resolution of taxonomic uncertainties. Management regulations of ten rely on classificati ons of subspecies or distinct population segments (DPS) for supervision or special protections. A DPS is defined as a reproductively isolated or evolutionary significant population. Subspecies inhabit a unique habitat or geographic range from other member s of the species and do not normally exhibit reproductive isolation. Therefore subspecies are distinguished by habitat and genetic differences 22


(Avise & Ball Jr. 1990; O'Brien & Mayr 1991). Another term, e volutionary significant unit, indicates genetically differentia ted populations within a specie s and has been criticized for ignoring behavioral adaptation, su ch as mating rituals or habi tat use, which can create a reproductive barrier before molecular changes occur (Crandall et al. 2000). Genetic factors that can be addressed in c onservation biology include the management of small populations, inbreeding and outbreeding depression, loss of genetic diversity, and the reduction of gene flow. Many wild populations have experienced substantial and long-term population losses. Glacier movement and str ong hunting pressure can cause marine mammal population reductions. For example, northern elephant seal harvesting reduced the species to approximately 10-30 individuals in 1860. Sma ll inbred populations can limit the effective population size (Ne) and in turn genetic diversity. The e ffective population size is the number of individuals in an idealized population that would have the same genetic properties as observed in a real population (Wright 1931). The effective population size is typically smaller than the actual population because of unequal sex ratios, fluctuation in population size s, and variation in progeny numbers. Not all individu als contribute to the next ge neration equally. For instance, highly inbred individuals are gene tically the same and count as fewer effective individuals to provide unique genotypes in the next generation. Molecular Markers Mitochondrial and microsatellite DNA Genetic markers typically used in conservati on genetics include chromosomes, allozymes, and mitochondrial and microsatellite DNA. Mitochondrial and mi crosatellite DNA are commonly implemented in populatio n genetics due to their variability, ease of use, and high mutation rate. Non-invasive techniques can be used to acquire small samples of tissues for polymerase chain reaction (PCR) DNA amplification. 23


Circular mitochondrial DNA (mtDNA) molecu les can assist in defining taxonomic relationships and can differentiate am ong populations within a species (Frankham et al. 2002). The maternally inherited mitochondrial genome provi des female migration and lineage patterns. The cytochrome b and displacement loop sequences in the control region are commonly used to distinguish haplotypes or sequence differences. Inferences are most accurate when the genetic diversity is statistically analyzed below the ordinal level. Haplot ypes outside of the order may or may not follow the pattern of muta tion observed within the lineage. Microsatellites, or short tandem repeats (STR), are nuclear polymorphic loci with 2-6 basepair (bp) repeat units. In most species, the markers provide enough variability to distinguish individuals from each other with high probability. Many loci are investigated to create a unique genotype for each individual. The majority of mi crosatellite loci are sel ectionally neutral and base-pair mutations are not constrained by functi onal demands of natural selection, allowing for more variation. Replication slippage is the primary mutationa l mechanism that creates microsatellites. During replication, the DNA polymerase enzyme may stutter, or the DNA strands may be displaced and realign incorrectly, leading to an insertion or dele tion of repeat units (Ellegren 2000). The stepwise mutation model (Ohta & Kimura 1973) is the simplest model and assumes that the microsatellite expands or contracts by one repeat unit each mutation. This gain and loss of repeats can lead to a homoplastic state, false equality of alleles base d on independent mutation to the same size. Therefore, individuals may be similar by state but not by descent. For this reason, statistics were adapted to address rela tedness based on the mutational mechanism to achieve a certain state. 24


Nuclear microsatellite DNA is inherited from both parents and participates in a recombining process. Both parents have two alleles and the offspri ng randomly inherits one allele from each parent at every microsatellite locus. In a breeding population, the combination of microsatellites wi ll become distinct from other populations. The high microsatellite mutational rate provides a contemporary perspective of the populations unique genetic signature. Statistical analyses are most appropriately conducted with individuals from the same species, geographic lo cation, and evolutionary time. The increased potential for homoplasy and hi gh mutational rates reduces the microsatellite phylogenetic signal throughout evolution. Comparisons across sp ecies compound this effect and are not recommended. Microsatellite and mitochondrial DNA comparisons Microsatellite and mitochondrial DNA data can be compared within a population, but may provide contrasting information. Mitochondr ial DNA provides information only from the matrilineal lineage and has a slower mutation ra te than that of microsatellites (Frankham et al. 2002). Because of these characteristics, mtDNA data characterize evolutionarily historical patterns of a population and can address phylogenetic relationships among species. Strict geographical partitioning of mt DNA lineages is often found in animal species with low or moderate dispersal abilities (Knoll & RowellRahier 1998). Many mari ne mammal populations have strong matrifocal genetic st ructure. Consequently, more popul ation structure is detected at mtDNA than at microsatellite DNA (Hoelzel et al. 2002). Microsatelli te data provide information on female and male dispersal among th e populations. Therefore, comparing the data from both markers may convey contrasting popula tion structure due to sex-biased dispersal and/or different evolutionary periods being addressed. 25


The construction of phylogenetic trees can assist in the analysis of populations or species. The construction of a mitochondrial tree to illustrate phylogenetic relationships can be performed within an order, using the most recent comm on ancestor to root the tree. Alternatively, microsatellites cannot accurately quantify distant evolution between species, and therefore it is inappropriate to include an ances tor or root the phylogenetic tr ee. Excessive homoplasy is found in microsatellites due to high variation and slipped-strand mis-pairing. The phylogenetic tree affinities of microsatellite alleles can be deduced only from the repeat number. This may be a poor indicator of the true evolutionary relationships among long-diverged alleles. The ultimate consequence of extended periods of evolution with high mutation is deterioration of the phylogenetic signal, limiting microsatellite tree anal yses to the recent genetic changes of a single species. Measuring Mammalian Genetic Differentiation Haplotype analysis Molecular studies can quantify genetic varia tion and assess the evol utionary pot ential of species with three main statistical measures. The first measure is the number of haplotypes (k ) in a population or species. When comparing th e number of haplotypes among populations and studies, a similar sample size, number of populations and base-pairs is needed. In contrast, the following two measures can address frequencie s and averages and can be compared across studies. The first is haplotype diversity ( h ), which is a function of the number and frequency of haplotypes within a sample, or the probability that two randomly chosen haplotypes will be different in the sample (Nei & Tajima 1981). The second is nucleotide diversity ( ), defined as the heterozygosity at the base pair level, or th e average number of nucleotide differences per site in pair-wise comparisons among DNA sequences (N ei 1987). In contrast to haplotype diversity, accounts for relationships among haplotypes. Depending on the loci and species surveyed, h is 26


above 0.5 for animal DNA and is typically in the range of 0.001-0.020. When comparing among studies, the number of populations and mark ers are variables that should always be addressed. Genotype analysis The most extensively implemented techni que for addressing genome-wide genetic diversity is the quantification of heterozygosity (H) and allelic di versity (A). However, pooling divergent populations in the analysis will inco rrectly inflate the observed diversity. Sample number, number of populations, ty pe of marker, and number of lo ci must be considered. The observed and expected heterozygosity (HO and HE) values are averaged over many loci to characterize genetic diversity for a population or species. Observ ed heterozygosity is the number of heterozygotes, having two alleles at a locus, divided by the total number of individuals sampled. Expected heterozygosity is the fractio n of a population that would be heterozygous if the population mated at random. The second measure, allelic diversity, is the mean number of alleles per locus averaged over multiple loci. The values are obtained with in a species or population and the results can be effectively compared acro ss species or populations. Heterozygosity and allelic diversity values can be compared over different variables and taxa. Microsatellite variation in threatened sp ecies is one of the most powerful and practical means currently available to quantify diversity. Allozymes, different alleles of a gene producing one enzyme and RFLPs, varia tion in DNA sequence detected by restriction enzyme cleavage, can also be used to measure these variables. Microsatellite genetic dive rsities do not strongly differ among placental mammalian orders (Garner et al. 2005). In large outbred placental mammal populations, the average microsatellite heterozygosity (Have) for polymorphic loci is 0.6-0.7 and the average number of alleles per locus (Aave) is 8.8 (DiBattista 2007; Garner et al. 27


2005). In comparison, endangered species can have up to one-half lower ge netic diversity than non-endangered species (Frankham et al. 2002). Compared to undi sturbed populations, hunted or harvested, and fragmented populations we re found to have reduced differentiation (Have = 0.50.6, Aave = 6.9; DiBattista 2007). A temporal effect was found in the genetic vari ation of historically or long-term disturbed populations. Populations with historical disturbances have lower diversity (Aave = 4.8) than recent disturbances (Aave = 7.8) and long-term disturbances have lower diversity (Aave = 6.5) than short-term disturbances (Aave = 8.2; DiBattista 2007). Many marine mammal populations experience similar genetic diversity losses from historical and long-te rm (~200 years) hunting pressures. Population Genetic Theory Basic theoretical principles create mathema tical models of idealized populations and can compare them to natural populations. This comparison provides information on the processes responsible for the detected genetic patterns (Gillespie 1998). Hardy-Weinberg Equilibrium Hardy-Weinberg equilibrium (HWE) is a law st ating that allele and genotype frequencies will reach equilibrium, defined by a binomial distri bution, in one generation and remain constant in large, randomly mating populations that e xperience no migration, selection, mutation, or nonrandom mating. HWE is an assumption applied in many models and analys es and provides the basis for detecting selection a nd the effects of inbreeding. The HWE assumptions include large populations, diploid organism s, random mating, sexual re production, non-overlapping generations, and negligible migr ation, mutation, and na tural selection. In a natural population, HWE expected and observed allele frequencies are calculated in a chi-squared analysis to determine if deviations are from chance or enduring processes (Frankham et al. 2002). 28


In general, large natural out breeding populations are in e quilibrium. The effects of migration, mutation, selection, a nd overlapping generations, alt hough violating the model, are minimal in these populations. Alternatively, ma rine mammal populations often violate the HWE assumptions with appreciable effects. For example, marine mammals are long lived and therefore have lengthy and overlap ping generations. Research has indicated that recently killed bowhead whales ( Balaena mysticetus ) still have harpoons in their bodies from the 1790s, which, along with analysis of amino acids, has es timated a life span of 211 years (George et al. 1999). Secondly, elevated marine mammal migration an d high dispersal capabilities may violate the model. Annually, adult male sperm whales ( Physeter macroscephalus ) migrate between the poles and tropical waters (Rice 1989a). Third, some marine mammal populations are extremely small, increasing the effects of mi gration, drift, and natu ral selection, such as the North Atlantic right whale ( Eubalaena glacialis ), numbering in the 300s (Waldick et al. 2002) and the Puerto Rico Antillean manatee ( T. m. manatus), estimated to have 250 individuals (Slone et al. 2006). Finally, many marine mammals exhibit non-rand om mating strategies. For example, dugongs ( Dugong dugon) defend mating territories. Dominant males may mate with all of the females in an area, creating a high proportion of gene tically similar offspring (Anderson 1997). Manatees deviate from the HWE assumpti ons. Manatees have long and overlapping generation times and most populations are small. They have the potential for long distance migrations between populations. The manatee ma ting system appears to be promiscuous and random, with females being surrounded by a number of males. However, if only the strongest male succeeds in fertilizing a female, a few domi nant males will contribute to the population, constituting a non-random mating system. 29


Statistical Methods F-statistics Numerous population studies address the parti tioning of genetic vari ation within and among subdivided populations (Wright 1951) Differentiation is directly related to the populations inbreeding coefficient among subpopulations. FST, the fixation index, is the effect of population sub-division due to inbreeding. FST is calculated from the rela tionship between heterozygosity and inbreeding. It is the probability that tw o alleles drawn randomly from a population are identical by descent. High rates of gene flow lo wer the probability, while low rates increase the probability. Small populations have more potentia l for inbreeding and usua lly genetically drift more rapidly than large populations. A value above 0.15 indicates significant divergence between the populations (Frankham et al. 2002). Computer simulations Computer simulations provide the means for addressing complex models with many interacting factors. Maximum parsim ony, maximum likelihood, Bayesian phylogenetic inference, and neighbor-joining trees are mathema tical methods used in computer programs to estimate phylogenies and c onstruct phylogenetic trees. Maximum parsimony is a non-parametric statistical method, in which no assumptions are made about the frequency distributions of the va riables being assessed. The method is ultimately used to create a phylogenetic tree. The corr ect tree topography has the least number of evolutionary changes between the character states. Parallel and convergent evolution can cause the method to be erroneous because it cannot decipher between the two processes. Secondly, maximum likelihood is a parametric statistical method, which makes assumptions about certain models of character evolution. Th is method must correctly account for differences in the rate of evolution among characters to estimate phyloge nies accurately. A third method, Bayesian 30


phylogenetic inference, uses maximum likelihood to create a posteri or distribution for a parameter generated by multiple alignments. It typically uses the Markov Chain Monte Carlo (MCMC) algorithm, which is a st ochastic, non-deterministic repe tition of algorithms, where the previous state is irrelevant for predicting the probability of subsequent stat es. Finally, neighborjoining is a bottom-up clustering method used in the composition of phylogenetic trees. A distance matrix comparing all the individuals can be made from a variety of measures to create the tree. Neis distance, Da, and Calvalli-Sforza and Edwards distance, Dc, are the most common measures for microsatellite analyses (Takezak i & Nei 1996). The tree is based on the minimumevolution criterion and the least total branch lengt h is preferred at each st ep of the algorithm. Application of Conservation Genetics Taxonomic Standings Statistical models are used to address evolutionary re lationships, resolve taxonomic uncertainties, and define management units. A re cognized taxonomically di stinct subspecies of the meadow jumping mouse ( Zapus hudsonius preblei ) was recommended for delisting from the U.S. Endangered Species Act (ESA) afte r morphological and phylogenetic comparisons determined it not to be significantly differe nt from the other four subspecies (Ramey et al. 2005). A second study using 21 microsatellite loci and 1380 mtDNA base pairs revealed that Z. h. preblei was genetically distinct and confirmed the original taxomic classification (King et al. 2006). Phylogeographical structuring of haplot ypes and multilocus genotypes using Bayesian inference and trees, were found for all five subspecies of Zapus hudsonius. Z. h. preblei appears to have been reproductively isol ated from the other subspecies for enough time to be on an independent evolutionary path. The study supported th e threatened subspecies status under the ESA. 31


Although Florida and Puerto Rico manatees share a mitochondrial haplotype, the nuclear data confirm that the two populations are distinct. Florida and Beli ze manatees also appear to be on reproductively isolated paths. These prelim inary data support the subspecies distinction between Florida and Antillean manatees. Habitat Fragmentation Molecular techniques can assess detrimental genetic effects in severely fragmented populations. Microsatellite mark ers addressed the effects of ha bitat fragmentation and loss on nine Florida black bear populations ( Ursus americanus floridanus ; Dixon et al. 2007). Small populations were isolated by ha bitat fragmentation, indicating inbreeding depression. Some Florida populations had extremely low genetic variation, although most were within the range of other North American black bears. Heterozygosity ranged from 0.270.71 in nine populations. The analyses compared genetic diversity and population size and found that maintaining bear populations above 200 individuals provided the most genetic vari ation. Additionally, managing the population as a metapopulation and creating or maintaining corridors assists in maintaining gene flow. Monitoring should be continued to in dicate whether a translocation of a genetically distinct bear is needed to rescue a population from negative inbreeding effects. The populations of Antillean manatees must also be monitored to prevent inbreeding. Many of these populations are sma ll and isolated on islands or co astal refuges. Large-scale coastal construction or dock build ing could easily fragment the habi tat. Specifically, to increase genetic diversity in Puerto Rico, it will be im portant to preserve the habitat linking the two putative subpopulations and perhaps provide or improve corridors. Pedigree Reconstruction Observing marine mammal reproduction to obtain information on mating success and lineages is difficult and logistically challengi ng. Additionally, females may mate with multiple 32


males. This lack of biological data and am biguous paternity can be addressed with molecular tools. Male reproductive success was analyzed in the Mexican pacific populations of humpback whales ( Megaptera novaeangliae ; Cerchio et al. 2005). Maximum likelihood analysis in the program CERVUS (Marshall et al. 1998), examined 13 microsatellite markers. A few highly represented males displaying su ccessful reproductive mating tactic s were expected. However, no male was assigned to more than three calves and most males sired only one calf. The lack of dominant males was attributed to previous whaling pressure s, creating a young male population with less competitive abilities. Genetic diversity values were r obust. The number of alleles ranged from 4-19 with an average of 10.1 and expected heterozygosity ranged from 0.338 to 0.889 with an average of 0.710. Similar studie s could assist with understanding manatee reproductive success. Forensic Applications Molecular techniques were used in a forensic fashion to identi fy species and sample origin of cetaceans in Asia. The only cetaceans reported to be harvested are Minke whales and dolphins. Sold in unregulated Japanese markets, Minkes are taken in Japans program for scientific research and small cetaceans are harves ted by the Japanese and incidentally in Korean fishing nets. Endangered species were suspected of being harvested. Whale products purchased in markets in Japan and the Republic of South Korea were PCR amplified at the sample collection locations, so that only amp lified products were removed from the countries (Baker et al. 2000). Microsatellite analyses dete rmined taxonomic species and at times distinguished regiona l subpopulations of purchased products. In the 655 products investigated, DNA sequencing identified 12 species or subspecies of whale, numerous dolphin species, sheep, and horse. Seven of the identified whale species are protected by internationa l whaling agreements dating back to at least 1989. Tissues of Minke 33


whale from the exploited North Pacific and the protected Sea of Japan were identified. Mix stock estimates and maximum-likelihood methods determined that 31% of the Minke samples originated from the protected area. These resu lts of undocumented/unrepo rted exploitation were then added to a model of population dynamics for the International Whaling Commission. Cytogenetic Studies Sirenians are phylogenetic outlier s and despite similarities in body shape, adaptations, and habitat, manatees and dugongs have no evolutionary relationship with the other major orders of marine mammals. The orders geographical e volution is not fully understood and phylogenetic relations are based on morphology, limited recent biochemical, and genetic evidence. Sirenians likely originated in Eurasia or Africa and spread into tropic al South America by the middle Eocene (45-50 million years ago; Domning 1994; Reynolds III & Odell 1991). Paleontological evidence suggests that si renians are members of the super-order Afrotheria, and are grouped with proboscidea (elephants) and hyr acoidea (hyraxes) in the clade Paenungulata (De Jong & Zweers 1980) The manatee Zoo-FISH stud y presented here provides support of the evolutionary relationships within the super-order. Further support of this grouping was provided by cross-species Zoo-FISH within the clade (Pardini et al. 2007). Additionally, the root of the ancestral eutherian tree has not been identified with certainty. Afrotheria, Xenarthra or a combin ation of the two super-orders, contains the most basal placental mammalian karyotype. These cla ssifications are complicated a nd could be resolved by using cytogenetic techniques. Chromosome diversity ca n assist in differentiating species Species usually differ in the number, shape, and/or banding patt ern within their chromosomes. For example, the Chinese and Indian muntjacs (barking deer) morphologically a ppear to be the same species and are managed as such. However, their chromosomes are dram atically different, the Chinese muntjac having 46 34


chromosomes and the Indian muntjac male s and females having 6 and 7 chromosomes respectively (Ryder & Fleisher 1996). Marine Mammal and Manatee Chromosome Investigations Previous marine mammal chromosome studies were restricted to conventional Giemsa stained chromosomes on a limited number of individuals (Arn ason 1974a; Arnason 1974b; Arnason & Benirschke 1973; Assis et al. 1988; Duffield et al. 1967; Lounghman et al. 1970; White et al. 1976b). These studies established the modal diploid number (2N), total number of chromosome arms (fundamental number), and rest ricted gross morphological features (size and centromere position). Conventional Giemsa solid-stained chromosome studies, completed on a limited number of individual manatees, established the chromo some number as 2N = 56 for the Amazonian manatee (Trichechus inunguis Lounghman et al. 1970) and 2N = 48 for the Florida manatee (White et al. 1976b). Following conventional Giemsa st aining, chromosome-banding procedures allowed for the identification of individual chromo some regions. Giemsa and trypsin staining, or GTG-banding, was used to create karyotypes a nd ideograms for the Amazonian manatee (Assis et al. 1988) and the Florida manatee (Gray et al. 2002). The divergence of the two species occurred from a fusion or fission rearrangement in the homozygous state involving four biarmed autosome pairs in T. manatus or eight pairs of acrocentric autosomes in T. inunguis An additional study was conducted on potentially hy bridized Amazonian manatees and coastal Brazilian West Indian manatees. The individuals had a mixture of Amazonian and West Indian physical morphological charac teristics. It had a T. inunguis mtDNA control region (T) haplotype but a diploid number of 2N = 50, intermedia te between the two species. The authors extrapolated that the number of chromosomes could result from an F2 backcross of an F1 hybrid 35


female breeding with a male T. manatus There are limited additional cytogenetic data published on Sirenians. Zoo-FISH While analysis of chromosome banding homo logies has been used extensively to investigate the evolutionary conservation of chromosomal segments and phylogenetic relationships among various mammalian species, es tablished interspecific banding comparisons are speculative and should be confirmed by di rect mapping of DNA seque nces to chromosomes (Bielec et al. 1997). Direct comparisons using in situ hybridization identifies cons erved chromosomal blocks or regions and differentiates between true a nd false chromosomal banding similarities. Chromosomal banding does not diffe rentiate between true phyloge netic (homologous) similarity and false or convergent (non-hom ologous) resemblance as seen by cross-species chromosome painting or Zoofluorescence in s itu hybridization (Zoo-FISH; Stanyon et al. 1995). Direct comparisons are useful in identifying large chromosomal blocks of conserved syntenies, especially for species that do not share any morphological or banding pattern similarity or that diverged more than 90 million years ago (Rettenberger et al. 1995). Phylogenetic studies and direct comparativ e mapping between vertebrate genomes have become possible through the deve lopment of comprehensive genome maps for humans and mice. Although the karyotype and ideogr am of the Florida manatee have been completed, species comparisons should be made using cross-species chromosome paints or manatee specific sequences. These comparative genomic techniques are being applied to br oad groups of animals to facilitate comparative genomic studi es at the chromosome level (Bielec et al. 1998; Chowdhary et al. 1998; O'Brien et al. 1997). Zoo-FISH is informativ e for intraspecific, within the order, and interspecific, outside the order, comparisons. A probe from a species with a 36


mapped genome, consisting of a complex mixture of sequences, is hybridized to metaphase chromosomes from a different species, identifying the corresponding chromosome regions in both species as being evolutionary "orthologues." Delineating homologous chromosomal segments among mammalian orders determines the order of genes and serves as the founda tion of comparative genetic maps (Bielec et al. 1998). Zoo-FISH studies help to identify conserved chromosomal and neighborin g segments, specific rearrangement patterns, major genes in mappoor species, and improve the understanding of the evolutionary processes of mammalian genomes through identifying phylogeny, ancestral genomic segments, and human race origins (Chowdhary et al. 1998). Comparative maps provide unique information for studying changes in genome organization and inferring the sequence of rearrangements occurring during mammalian e volution. Additionally, Zoo-FISH provides reliable information on linkage predictions and candidate diseas e trait and gene identification (Eppig & Nadeau 1995). Genes that are inherited as a single unit, i.e., lin ked, occur on the same chromosome. The nature of conserved chromosomal se gments represents a powerful suite of phylogenetic data for resolving the precise progression of mammalia n evolution (O'Brien et al. 1997). Ancestral chromosomal segment identificat ion will contribute to evaluating the human and manatee genomic divergence. Mammalian Zoo-FISH Comparative Zoo-FISH studies have delin eated extensively conserved chromosome segments in the karyotypes of humans and othe r distantly related species, including the great apes (Wienberg et al. 1990), pig (Rettenberger et al. 1995), cow (Chowdhary et al. 1996; Hayes 1995; Solinas-Toldo et al. 1995), cat (Rettenberger et al. 1995), horse (Raudsepp et al. 1996) and mink (Hameister et al. 1997). Published comparative mole cular cytogenetic data on marine 37


mammals are limited to the fin whale (Scherthan et al. 1994), harbor seal (Fronicke et al. 1997), and bottlenose dolphin (Bielec et al. 1998). Zoo-FISH comparisons between human (HSA) and harbor seal ( Phoca vitulina ) identified 31 conserved segments, 79% similarity, and covered the complete autosomal complement and the X ch romosome. Zoo-FISH comparisons between human and bottlenose dolphin ( Tursiops truncatus ) found that all HSA chromosome paints, except the Y probe, hybridized to corresponding chromosomes on Tursiops All dolphin chromosomes were painted except for the smallest submetacentric pair. Thirty-six segments of conserved synteny were identified (Bielec et al. 1998). Human and manatee chromosomes will be similarly compared to assess the evolutionary conservation between th e two distantly related species. Investigations of the We st Indian manatee genome address the composition of phylogenetic relationships. Hybridization of human ( Homo sapien ; HSA) and West Indian manatee ( Trichechus manatus latirostris) chromosomes directly compares the conservation of genomic evolution. Used as a template, the sequenced human genome establishes a preliminary map of the manatee karyotype. The comparative investigation evaluates the degree of syntenic conservation and chromosomal rearrangement. Evaluation of chromosomal rearrangements, such as inversions and transl ocations, can assist in discer ning genome organization (Eppig & Nadeau 1995). Within the conserved segments, human whole chromosome paints also serve as a reliable guide for manatee gene order. Zoo-FISH maps for mo del animals usually show a 97% agreement with detailed gene mapp ing data (Chowdhary & Raudsepp 2001). Manatee Zoo-FISH studies can be used to identify and quantify human conserved chromosome homology as well as evolutionary conservation and phyolge netic relationships within Afrotheria and Paenungulata. 38


Microdissection Micro-manipulated dissection or chromosome microdissection is a cytogenetic technique primarily used to identify rearranged segments of abnormal chromosomes. This technique physically isolates whole or part ial chromosomes using a microscopi c needle. Multiple copies of the isolated DNA are produced through polymeras e chain reaction (PCR) amplification. The product is used as a probe to id entify homologous segments. Micr odissection has been used to link chromosomal aberrations to cancer or i nherited genetic disorders and to identify irregularities exhibited by chromosome s of tumor cells (Dennis & Stock 1999). Varieties of studies use microdissection to produce probes from whole or partial chromosomes. Microdissected manatee probes can identify regions of homology with human chromosomes. This could support the results from the reciprocal experiments using human derived probes. Additionally, microdissected pr obes may clarify the position and orientation of homologous chromosome segments. West Indian manatee probes could be used to identify evolutionary divergence w ithin the order sirenia and with other related species. This technique could also be used to identify rearrangements found in any aberrant manatee karyotype. Sirenian Molecular Studies All extant species of Sirenia are considered vulnerable to extinction by conservation organizations. Many conservation efforts have b een made to protect and manage West Indian manatees. Throughout Florida and Puerto Rico, st eps have been taken to limit hunting and the number of anthropogenically rela ted deaths. Biological studies have illuminated life history characteristics and tracking studies have disclosed movement patterns and habitat use. Genetic studies have elucidated potentia l population relatedness and structur e and can assist in protecting unique populations. 39


West Indian manatees typically exist in sm all isolated populations. Small populations tend to have low levels of variation and can easil y become inbred, cyclically reducing genetic diversity. This can then reduce fecundity, surv ival, and a populations ability to cope with environmental change. Populations with low genetic diversity have a severely compromised ability to evolve and a high extinction risk. Molecular techniques can monitor these risks by quantifying inbreeding and geneti c variation in a population. Genetic markers may assist in identifying migr ation and defining management units within the species. Distinct or unique populations i nhabiting different environments should be managed as separate populations. Mana tee dispersal patterns and nata l area can be investigated. Knowledge of a populations geneti c signature may lead to the id entification of an individuals original population, espe cially in the Caribbean where diffe rent populations re side on separated islands. Reproductive success in male manatees is difficult to assess because multiple males are involved in mating herds. Micros atellites have the capability to decipher paternity and male reproductive success. Molecular markers may also assist in detecting hybridization of T. manatus and T. inunguis in Brazil. Finally, nuclear markers have been developed to assign manatee sex (Tringali et al. 2008a). This is helpful in the cas e of badly decomposed carcasses or field conditions in which gender may not be evident. Sirenian Population and Conservation Genetic Studies The first manatee genetic study was conducted on 59 Florida manatees using 24 allozyme loci, 10 of which were polymorphic with two or three alleles (McClenaghan & O'Shea 1988). Allozymes represent different alleles of a gene producing one enzyme. The allozymes had 0.3 % polymorphic loci and 0.0500 mean he terozygosity in Florida. The estimates were determined to be equivalent or higher than estimates for ot her terrestrial and marine mammals. Genetic homogeneity was found throughout the differentiated regions in Fl orida and was most likely due 40


to high rates of gene flow throughout the peninsula. No unique alleles were observed among the regions. The authors did hypothesize that Floridas genetic diversity was limited by decreased gene flow from other Antillean populations because of strong geographical ba rriers in the Straits of Florida and the northern Gulf of Mexico (Domning & Hayek 1986). The McClenaghan and OShea (1988) study sugges ted that manatees have equivalent or higher heterozygosity than other terrestrial and marine mammals, most likely due to the high gene flow across the region. Allozymes are non-neutral gene products, which may experience selection. The diverse variables of the manatee habitat (i.e., salinity, temp erature, and red tide) may cause the Florida manatee to have increased heterogeneity. Alternatively, this study may be biased due to the pooling of tissues and degrad ation of enzymes, whic h could increase the apparent number of alleles. The first DNA-based study used mitochondrial cytochrome b to analyze the Florida population (Bradley et al. 1993). All individuals were determ ined to have the same haplotype, although the sample size was small and only 225 base pairs were sequenced. Sirenias evolutionary relationship with proboscideans was supported by an altered amino acid locus that is invariant for 20 other mammal species. The following mitochondrial DNA study amp lified 410 bps of the control region displacement loop in eight West Indian manatee populations (Garcia-Rodriguez et al. 1998). A total of 15 haplotypes was identified in 86 individuals. The pooled populations of the T. manatus species had high diversity h = 0.839 and = 0.04. A strong division between populations was indicated with significant haplotype frequency shifts. Three lineag es were apparent: Florida and the West Indies; the Gulf of Mexico to the Cari bbean rivers of South Am erica; and the northeast Atlantic coast of South America. The highest p opulation diversity was in Guyana with suspected 41


hybrids, h = 0.857 and = 0.044. The lowest population dive rsity excluding Florida, which had no variation, was Puerto Rico with h = 0.530 and = 0.001. T. inunguis diversity values were = 0.005 and h = 0.875 (n = 16). Only one haplotype was identified in Florida (A01), possibly due to a recent colonization or bottleneck. Th is haplotype was also found in Puerto Rico, connecting Puerto Rico to Florida historically. In an effort to find a marker with a higher resolution of genetic va riation, microsatellite DNA primers were developed and ch aracterized (Garcia-Rodriguez et al. 2000). Development of microsatellite primers implemented two en richment techniques for nucleotide repeats; magnetic bead capture and nylon filters. From 61 microsatellite-bearing clones, eight polymorphic primer sets were identified in 50 Fl orida manatees. An additional three markers were polymorphic for Antillean and Amazonian ma natees, while a total of nine was polymorphic for the dugong. The overall level of heterozygos ity (0.410) and allelic diversity (A = 2-6, AAVE = 2.9) were low. The markers had some of the lowest levels of genetic diversity found in species-specific microsatellites, indicating lim ited intrinsic diversity, a founder effect, or bottleneck of evolutionary significance. In a subsequent study, 361 bps of the m itochondrial DNA control region addressed the genetic differentiation and geographical populat ion structure in the Amazonian manatee (Cantanhede et al. 2005). A total of 84 Amazonian ma natees represented 33 haplotypes ( = 0.0075, h = 0.909). The majority of haplotypes (n = 25) were found in one individual each, while only eight were found in multiple individual s. Amazonian manatees had more diversity than any of the three proposed West Indian ma natee clusters. The high diversity may be from pooling divergent populations or a large population size. Subse quently, no genetic bottlenecks were detected. 42


A phylogenetic analysis showed no differen ce among Amazonian manatee populations. A nested clade analysis indicated evidence of minimal and long-distance dispersal, but also restricted gene flow among some regions. In essence, the population behaves in a panmictic fashion, but because of the vastness of the range gene flow is reduced among isolated regions. This molecular signal is attributed to the season al migration patterns of distant dispersal during the flood season and the return to a few d eeper areas during the low-water season. The Amazonian manatee has been under protec tion for the last 40 y ears and calculations estimate a current population size of 445,000 female effective breeders (Nef). Cantanhede et al (2005) suggested that since 1960, the population ha s undergone a 95% increase in size due to reduced hunting. This is corroborated by calculations of life span, female reproductive age, and the number of calves produced each year. In 40 y ears, a single female could produce a lineage with 27 female offspring. Historical and current population size based on agrees with these estimates, increasing from 2.23 X 104 Nef ( 0 = 0.337) to 4.55 X 105 Nef ( = 6.565). The statistic uses sequence data to assess genetic difference, similar to FST. Since = 4Ne where is the mutation rate, the number of effective breeders can be calculated. The study also suggests that the genus Trichechus is most accurately viewed as four equally divergent lineages. The three T. manatus lineages mentioned in Garcia-Rodriguez et al. (1998) are as different from each other as they are from T. inunguis Diverse aquatic biogeographical habitats may have pr oduced the four distinct lineages. Next, a study was conducted using a 410 bp fr agment of the mtDNA control region on 330 individuals from the West Indian, Amazoni an, West African, and dugong species (Vianna et al. 2006). Individuals from 10 countries revealed 20 West Indian, 31 Amazonian, and five West African manatee haplotypes. The West Indian manatees had the highest nucleotide diversity ( = 43


0.038648), followed by the West African manatee ( = 0.019581), and the Amazonian manatee ( = 0.005353). The T. inunguis and T. manatus values are similar to those reported in GarciaRodriguez et al. (1998) and Cantanhede et al. (2005). Three haplotypes were observed in Puerto Rico (A01, A02, and B01) and Belize (A03, A04, and J01). The program BARRIER proposed two geographic obstructions to migr ation and potential population divisions. A geographic separation isolated the Domini can Republic and Puerto Rico from the other West Indian populations and another isol ated Guyana and Brazil. The T. manatus subspecies split into three distinct clusters with median-joining networks and neighbor-joining trees, displa ying a heterogeneous geographica l distribution. Cluster I was composed of Florida, Mexico, the Greater Antill es, Central America and the Caribbean coast of South America; Cluster II contained Mexico a nd Central America and the Caribbean coast of South America; and Cluster III was the northea stern coast of South America (Brazil and Guyana). These are comparable clusters to Garcia-Rodriguez et al. (1998). A positive and significant correlation was found between genetic and coastline geographical distances using a Mantel test, supporting the idea that manatees mi grate along the coasts an d rarely in pelagic ocean. The Florida manatee bottleneck was supported. The authors concluded th at the distribution of T. manatus is limited around the warm waters of the equator and is a re sult of a stepping stone model of expansion along shallow coastal waters. The coalescence time for the mtDNA lineag es suggest that the Amazon manatee is the most basal taxon followed by the West African a nd West Indian manatees. The mtDNA control region data suggest that the Amazonian manatee is the sister group to th e West Indian manatee cluster I. Therefore, the species would be para phyletic and not all of the descendants from the most recent common ancestor would be included in T. manatus This is contradictory to the 44


control region data in Cantanhede et al. (2005), where four equally divergent lineages were identified. Possibly the additi on of four West Indian haplot ypes related to the Amazonian haplotypes influenced the results. However, analyses of 615 bps of cytochrome b in the same study suggested that T. inunguis is the basal species and that T. manatus and T. senegalensis were derived from the same marine ancestor. This is in agreement w ith Domning (1994), making the marine species monophyletic. The authors then extrapolate that since T. inunguis is the most basal lineage, it may be the only surviving species of an ancient lineage adapted to fresh water. More analyses with additional characters are needed to determine the true relationship. Potential hybrids from T. manatus and T. inunguis were studied using the mtDNA control region, two microsatellite primers and cytogene tic techniques. At the mouth of the Amazon River in Brazil and throughout the Orinoco Rive r, eight potential hybri ds were identified. Interspecific hybrids resulting fr om at least two generation (F2) backcrosses between T. manatus and T. inunguis were documented. In a study of Australian dugongs, mitochondria l and microsatellite markers addressed migration and population structure (McDonald 2005). Of the 115 individuals analyzed, 52 haplotypes were identified, alt hough only 19 were found in more than one individual. The 492 bps indicated 59 variable sites, h = 0.96 and = 0.029. A strong separation at the now submerged Torres Straight land bridge indicated population division during the Pleistocene. The lack of homogenization betw een the two areas indicates female philopatry. The authors also implemented six manatee mi crosatellite primers developed by GarciaRodriquez et al. (2000) that were highly polymorphic on dugongs. The analyses failed to indicate any geographic population st ructure across Australia. An analysis of molecular variance 45


(AMOVA) indicated high connectivity and gene flow among the regions. An isolation-bydistance pattern was found, most lik ely due to the large spatial sc ale of Australia. High allelic diversity may be an indicator of the larger population (100,000 individual s) and great geographic area inhabited by dugongs. Additionally, analyses did not indicate that the population had endured a bottleneck, as suggested in the Florida manatee population. The lack of concordance between the nuclear and mtDNA results indicate male dispersal. Males travel between the two mtDNA assigned populations and provide the gene flow to homogenize the nuclear DNA. The two mtDNA lineages are interbreed ing in the overlapping areas and assist in producing th e isolation by distance pattern. The next study focused on manatee microsatelli tes. The microsatellite primers developed by Garcia-Rodriguez et al. (1998) were not variable enough to provide individual multilocus genotypes for the Florida manatee. Therefore, to increase the resolution, additional microsatellite markers were developed (Pause et al. 2007). Multiple protocols were used: magnetic-bead capture technology, an unenriched lib rary, and a biotinylat ed probe enriched library. A total of 10 polymorphic loci was identified with the number of alleles per locus ranging from 2-7 and an average of 4.2 allele s. Heterozygosity ranged from 0.321-0.680 with an average of 0.501. All primers amplified T. inunguis and all but one amplified D. dugon. To increase the arsenal of polymorphic mi crosatellite primers on Florida manatees, Tringali et al. ( 2008b), developed 18 additional primers. A PCR-based isolation of microsatellite arrays (PIMA) method was used. The average number of alleles per locus was 2.5 and ranged from 2-4. Average heterozygosity was 0.34 and ranged from 0.02-0.78. Lastly, microsatellite primers were de veloped for the dugong. A biotin-labeled oligonucleotide was used for library enrichme nt. A total of 32 polymorphic loci was 46


characterized but only 26 were reported (Broderick et al. 2007). The average number of alleles was 4.88 with a range of 2-10 alle les per locus. The average heterozygosity of 50 dugongs was 0.52, with a range of 0.12-0.84. Of the 26, 22 were polymorphic on one Florida manatee and six were polymorphic on an Asian elephant. Belize and Puerto Rico Manatee Conservation Studies In the following studies, further analyses will use mitochondrial and microsatellite DNA to gain an understanding of the West Indian manat ee. Belize and Puerto Rico manatee populations will be compared to the Florida population. Th e study will address nearly all of the genetic issues Frankham et al. (2002) identifies in c onservation biology, includ ing: species biology, forensic applications, inbreedi ng and outbreeding depres sion, loss of genetic diversity, reduction of gene flow, management of small populations defining management units, and the subspecies taxonomic distinction. Additionally, a dugong and Fl orida manatee cross-species microsatellite primer comparison was made to determine the most informative and efficient panel of markers to aid population studies for both families. The long life span, elusive behavior, isolated ha bitat, and the turbid environment of West Indian manatees make it difficult to obtain information on movement and mating patterns. Relative to the long lifespan of the animal, mark -recapture or satellite tr acking studies occur over short periods of time. Altern atively, genetic techniques can pr ovide information on measurable parameters such as the movement of individuals and population structure. Comprehensive longterm movement patterns can be inferred from patt erns of gene flow. Genetic fingerprinting can assist in the identification of indi viduals and parentage assignments. The amount of interbreeding between the ad jacent geographical lo calities of Belize, Florida and Puerto Rico will be quantified and the degree of genetic structuring within the 47


populations will be assessed. From these data, important ecological infe rences and management decisions can be made to enhance the conservation efforts for the West Indian manatee. 48


49 CHAPTER 2 COMPREHENSIVE GENETIC INVESTIGATION RECOGNIZES EVOLUTIONARY DIVERGENCE IN THE FLORIDA ( Trichechus manatus latirostris ) AND PUERTO RICO ( T. m. manatus ) MANATEE POPULATIONS AND SUB TLE SUBSTRUCTURE IN PUERTO RICO Introduction Manatees (Sirenia: Trichechidae), inhabit tr opical and subtropical wa ters and are the only obligate herbivorous aquatic mammals. The endangered West Indian manatee ( Trichechus manatus ) is found in freshwater rivers, estuarine, and marine environments. The West Indian manatee has two recognized subspecies dist inguished by morphological, biological, and ecological data (Domning 1994; Domning 2005; Do mning & Hayek 1986). The Florida manatee subspecies ( T. m. latirostris ) is restricted to the southeastern United States and Gulf of Mexico. Small populations of the Antillean manatee subspecies ( T. m. manatus ) exist in the West Indies, Caribbean, Mexico and Central a nd northeastern South Am erica despite severe past exploitation (Domning & Hayek 1986; Hatt 1933). Low reproduc tive rate and intrinsic population density make this species particularly vulnerable to hum an disturbance (Bossart 1999). Ongoing habitat loss and high mortality threaten the future of the populations. The West Indian manatee is recognized as a vulnerable taxon by th e International Union for Conservation of Nature and Natural Res ources (IUCN; Thornback & Jenkins 1982). The Florida and Puerto Rico populati ons are listed and managed toge ther as endangered under the authority of the Endangered Sp ecies Act (U.S. Fish and Wild life Service 1982). Sustained management and conservation efforts have resulte d in the recent recommendation of downlisting the populations to a threatened status (U.S. Fish and Wildlife Service 2007). This may not be advantageous as the threats, ha bitat, population sizes, and needed protections are different for the


two groups. Additionally, the exte nt of migration and breeding between the two populations has not been thoroughly examined (Garcia-Rodriguez et al. 1998). The Florida manatee population size is estimated to be 3300 individuals (Florida Fish and Wildlife Research Institute 2007b). The major causes of mortality in Florida include boat strikes (24%; O'Shea et al. 1985), perinatal (24%) and unknow n (29%) (based on 6338 necropsies conducted from 19742007; Florida Fish and Wildlif e Research Institute 2007a). The rate for each mortality cause has remained stable througho ut the last 30 years, while known mortalities have increased, possibly a reflecti on of increasing population size. The Puerto Rico population is es timated to be about 250 individuals with increasing threats and mortality (Slone et al. 2006). They are coastally marine, dependent on seagrass beds and sources of fresh water (river mouths, run-offs, coastal fresh water holes, water treatment plant outfalls, etc.) but rarely vent uring inside the ri vers (Mignucci-Giannon i 1989; Rathbun & Possardt 1986). Reported Puerto Rico manatee d eaths are increasing an average of 9.6% per year (SD = 16.9%), and were higher during th e last 20 years of the study, 1974-1995 (MignucciGiannoni et al. 2000). The major cause of mortality wa s due to human interaction, including net entanglement, boat strike, and poaching (52.2% ; 15.6% watercraft), natural (22.2%; 20.0% dependent calves), and unknown (25.6%). Recently, watercraft mortalities in Puerto Rico have almost reached the proportions seen in Florida. An assessment of the genetic health of the population would benefit c onservation efforts. Conservation genetics represents a powerful tool for the assessment and management of threatened species (Frankham et al. 2002). Molecular markers can re veal genetic distinctiveness among or between taxa or populations with subtle or undetectable morphological differentiation, which traditional conservation bi ology techniques cannot. Genetic examination of animals has 50


contributed to the understanding of population structure, individua l identification, life history traits, patterns of gene exchange and genealogical or evolutiona ry relationships (i.e. phylogeny; Avise 2004). Genetic diversity is associated with fitness, population size and persistence, and the number of inbred individuals. Reduced genetic variation in a species can decrease fecundity, the ability to evolve and endure environmental ch ange, and may eventually lead to extinction. Studies of endangered marine mammals, such as the Guadalupe fur seal (Weber et al. 2004) and the North Atlantic right whale (Waldick et al. 2002) have addressed lo w genetic variation and aided in conservation of genetically unique and endangere d populations. Previous studies of genetic diversity in Fl orida identified only one matrilineal haplotype, A01. The A01 haplotype was additionally identif ied in the Puerto Rico population along with two others, A02 and B01 (Garcia-Rodriguez et al. 1998; Rodriguez-Lopez 2004). It has been suggested that since the two popul ations contain the same haplot ype, they may be related and experiencing immigration. To date, low levels of diversity have been found in genetic studies of the West Indian manatee (Bradley et al. 1993; Garcia-Rodriguez et al. 1998; McClenaghan & O'Shea 1988; Vianna et al. 2006). Within Florida, no variation wa s identified in the mitochondrial (mt) cytochrome b or control region DNA (Bradley et al. 1993; Garcia-Rodriguez et al. 1998). Other manatee populations also have lim ited mitochondrial diversity, making it difficult to elucidate detailed geneflow within and between the mana tee populations. To improve the resolution of genetic investigations, more variable Florida manatee nuclear microsatellite markers were developed (Garcia-Rodriguez et al. 2000; Pause et al. 2007; Tringali et al. 2008b). These markers allow for examination of population gene tic units, evolutionary and taxonomic patterns, and forensic investigations to identif y regions of origin (Garcia-Rodriguez et al. 1998). The 51


robust marker panel successfully provided each analyzed Florida individual with a unique genotype. The Florida population was determined to have low genetic di versity, possibly due to a bottleneck or founder effect (e xpected heterozygosity = 0.501; average number of alleles = 4.2). A fundamental knowledge of the diversity w ithin Puerto Rico can help to determine the genetic health of the population and assess the fre quency of new recruits into the population. An integrative conservation appr oach that identifies and sust ains ecological processes and evolutionary lineages is needed to protect and manage the biodiversity present in these small populations. The analysis presented here uses mitochondrial control region haplotypes and multilocus microsatellite genotypes to examine the relationship between Florida and Puerto Rico manatees and to address the level of variation an d fine scale genetic structure within the Puerto Rico population. The identific ation and characterization of migration, colonization, and extinction processes (Avise 2004) will a ssist the conservati on of the populations. Materials and Methods Sample Collection and DNA Extraction Florida and Puerto Rico manatee blood and epidermis tissue were collected from recovered carcasses or during wild manatee health assessmen ts. Additional Puerto Rico samples were collected from manatees in the rescue and reha bilitation program. Puerto Rico manatee genomic DNA was isolated using QIAGEN s DNeasy Blood and Tissue kits (Valencia, California) for 115 animals, 52 males and 63 females. Florida tissue DNA extraction t echniques are described by Pause et al. (2007). From the Florida dataset, 96 individuals were randomly chosen, proportionally representing the four geographica lly imposed management units: Northwest, Southwest, Atlantic, and St. Johns River. 52


Mitochondrial DNA Analysis Primers from Garcia-Rodriguez et al. (2000) were used to amplified a 410 base pair portion of the mitochondrial DNA co ntrol region displacement loop in 81 Puerto Rican samples, while additional sequences where obtained from Vianna et al. (2006). In total, 115 sequences from Puerto Rico (22 North, 36 East, 37 South, 20 West coast animals) and 96 sequences from Florida were analyzed. The mitochondrial DNA control region was pol ymerase chain reaction (PCR) amplified with primers developed from regions of 100% homology between cow and dolphin sequences (heavy strand primer, CR-5, and light strand primer, CR-4; Palumbi et al. 1991; Southern et al. 1988). The PCR reaction conditions were as follows: 10 ng DNA, 1 x PCR buffer (10 mM TrisHCl, pH 8.3, 50 mM KCl, 0.001% gelatin; Sigma-Aldr ich, Inc., St. Louis MO), 0.8 mM dNTP, 3 mM MgCl2, 0.24 M of each primer, 0.04 units of Sigma Jump Start Taq DNA polymerase. PCR cycling profile: 5 min at 94C; then 35 cycles of 1 min at 94C, 1 min at 55C, 1 min at 72C; then 10 min at 72C. Amplified products were purified using the Qiaquick PCR purification kit (QIAGEN). DNA sequencing was accomplished in the DNA Sequencing Core at the University of Florida, Gainesville, FL with the BigDye terminator protocol developed by Applied Biosystems Inc. using fluorescently labele d dideoxynucleotides. To verify sequences, haplotypes were aligned with mana tee sequences located in GenBa nk using the default setting in SEQUENCHER 4.5 (Gene Codes Corporation, Ann Arbor, MI). Control region fragments were sequenced in the 5-3 heavy-strand orientation. Finally, representatives from each haplotype and any ambiguous sequences were sequenced in th e 3-5 direction to en sure the accuracy of nucleotide designations. The degrees of differentiation, FST and ST, between Florida and Puerto Rico and among Puerto Ricos geographic regi ons were calculated using ARLEQUIN 3.1 (Excoffier et al. 2005). 53


Estimates of sequence divergence used the Kimu ra 2-parameter genetic distance model (Jin & Nei 1990; Kimura 1980). The variance distributi on was based on haplotype frequencies alone; all haplotypes were treated as equally differentiated (FST). Lastly, Tajimas D of selective neutrality, the number of polymor phic sites, S, number of nucleotide substitutions, NS, the genetic diversity, h, and nucleotide diversity, were calculated (Nei 1987; Tajima 1993). Microsatellite Analysis A total of 15 microsatellit e loci (Garcia-Rodriguez et al. 2000; Pause et al. 2007) was used for the Puerto Rico samples. The remaining three polymorphic Florida manatee markers, TmaE4 TmaE26, and TmaH23 (Pause et al. 2007) were determined to be homozygous in Puerto Rico manatees. Isolated DNA was P CR amplified using: 14 ng DNA, 0.8 mM dNTPs, 1x Sigma PCR buffer (10 mM Tris-HCl, pH 8.3, 50 mM KCl, 0.001% gelatin), 0.04 units Sigma Jump Start Taq polymerase, 0.24 M each primer and BSA where needed (Table 2-1). MgCl2 concentrations were 3 mM, except for TmaH13, TmaKb60 and TmaSC5, which required 2 mM. Amplifications were carried out on a PTC-200 thermal cycler (MJ Research, Waltham, MA) using the following conditions: initial denaturing at 95C for 5 min, 35 cycles at 94C for 30 s, annealing temp for 1 min (Table 2-1), 72C fo r 1 min, final extension 10 min at 72C. All individuals were successfully amplified at a minimum of 14 loci. Fragment analysis was performed on an Applied Biosystems ABI 3730 Genetic Analyzer. GENEMARKER, version 1.5 (Soft Genetics, State College, PA), was used to analyze the microsatellite fragment data. The microsatellite data for Florida samples were ki ndly provided for use in this manuscript (Pause et al. 2007). A Microsoft Access database wa s used to store a llelic information. Statistical Analysis The level of polymorphism wa s estimated by the observed (HO) and expected heterozygosity (HE), polymorphic information content and the number of alleles per locus (A; 54


Table 2-1) using GENALEX 6 (Peakall & Smouse 2006). Departures from the expected genotypic frequencies in Hardy-Weinberg equili brium (HWE) were tested using the Markov chain method (dememorization 1000, batche s 100, iterations per batch 1000) in GENEPOP 3.4 (Raymond & Rousset 1995). Additionally, linkage disequilibrium was tested for non-random associations between alleles of different loci The Markov chain method was implemented and the P -values were adjusted using Bonferroni sequential correction for multiple comparisons (Rice 1989b). To assess overall genetic differentiation at the population level, GENALEX 6 calculated FST using the infinite alleles model and RST using the stepwise mutation model. Comparisons included Florida and Puerto Rico and Puerto Rico di vided into nine categories that could be interchangeably grouped: male (M), female (F), North (N), South (S), West (W), East haplotype A01 (EA), East haplotype B01 (EB) dry (December to May) and wet (June to November) season. Geographic locations were ba sed on sample recovery location, haplotypes, radio tracking data, usag e areas, and genetic differentiati on. The South group included the B01 individuals from the Southwest corner, as genetic differentiation test indicated that these were not significantly different. EA and EB were co mbined to east (E) for the male and female geographic comparisons to maintain sample sizes for proper comparisons. Cluster analysis using multi-locus genotypes The program STRUCTURE 2.2 (Pritchard et al. 2000) was used to identify the genetic relationship and putative ancestral source populations of Florida and Puerto Rico manatees, and the genetic subdivision within Puerto Rico. STRUCTURE, a model based clustering algorithm infers population structure by probabilistically assigning individuals without a priori geographic or ancestral knowledge to a specific number ( K ) of clusters (presumably populations). In determining the number of clusters, the algorithm attempts to minimize deviations from HardyWeinberg equilibrium. 55


Simulations were conducted using the admixtur e model, which assumes that individuals could have some propor tion of membership ( q) from each of K clusters, leading to the potential identification of recent immigrants. Multiple Ma rkov chains can delineate differences within populations; therefore three parall el chains were analyzed for K = {1}, with a run-length of 100,000 repetitions of Markov chain Monte Ca rlo, following the burn-in period of 10,000 iterations. The three values for the estimated ln( Pr ( X | K )) were averaged, from which the posterior probabilities were calculated. The K with the greatest posterior probability ( Pr 1.0000) was identified as the optimum number of subpopulations. If the ln( Pr ( X | K )) does not fluctuate strongly, the fewest number of clusters with the greatest Pr is correct. Individual assignment success was recorded as th e highest likelihood of assignment ( q), and the percentage of individuals in a cluster with q > 0.90 was calculated. A 0.90 a ssignment value indicates that the individual is highly assigned to the cluster, with little lik elihood it belongs to a different cluster. To test the relatedness and degree of Florida and Puerto Rico admixture, the two groups were analyzed together. A subsequent analysis was conducted with a clustered Florida group acting as a divergent population, since only one population was dete cted when Puerto Rico was analyzed alone. This allowed subtle Puerto Ri co population differentiation to be detected by the program. Un-rooted neighbor-joining trees Neighbor-joining trees based on individual and population genetic distances were used to visualize relationships among populations, subp opulations, and individuals. The Phylogeny Inference Package (PHYLIP; Felsenstein 2004) es timated genetic distances based on pairwise Calvalli-Sforza and Edwards chord distance, DC. DC is based on allele fr equencies and provides accurate microsatellite tree topology (Takezaki & Nei 1996). Trees were constructed by 56


comparing individual genotypes with or without a priori partitioning. Comparison of the Florida and Puerto Rico individual ge notypes was conducted with MICROSAT, version 1.5d (Minch et al. 1997), to create a distance matrix and NEIGHBOR in PHYLIP to produce the tree. To create population trees, a priori sub-grouped allele frequencies were subjected to the programs in PHYLIP, using 1,000 replicates with the bootstrap support values at the branching nodes of the tree. The lengths of the branches represented relative genetic distances Geographic localities were analyzed from Florida and Puerto Ric o. Due to the strong geographical haplotype structuring and genetic division identified in Puerto Rico, the genetic distances among the samples from the four geographic regions: N, S, E, and W were compared using a neighborjoining tree. Cytogenetic Analyses Giemsa-banded karyotype analyses have only previously been performed on the Florida subspecies ( T. manatus latirostris ). Therefore, to assess cyt ogenetic differences between the subspecies, banded karyotype anal ysis was performed on Puerto Rico animals. Sodium heparin vacutainers were used to collect blood and samples were transporte d as quickly as possible to the laboratory. The cytogenetic analysis followed protocols described by Gray et al. (2002). Results Mitochondrial Sequence Analysis Mitochondrial DNA sequences from the NCBI da tabase were compared with the 81 Puerto Rico samples sequenced for this study, and the previously sequenced Florida and Puerto Rico samples. Within the 115 Puer to Rico individuals, 35 A01, four A02, 75 B01, and one B02, a previously unidentified haplotype, were found ( = 0.00132; h = 0.48500). This indicates that within Puerto Rico, there is a moderate to low chance of randomly drawing two different haplotypes, but low nucleotide divergence amon g those haplotypes. Within Florida one 57


haplotype was observed (A01; = 0.00000). Adjusted (net) Florida and Puerto Rico mtDNA sequence divergence estimates were ST = 0.49596 (Kimura 2-parameter) and FST = 0.65696 ( P < 0.00001). Three polymorphic sites (0.73%) and three nucl eotide substitutions were identified in the four haplotypes. Within Puerto Rico, Tajimas D = -0.08330 ( P < 0.50200) was not significant ( P < 0.05) therefore; the null hypothesis of selective neutrality cannot be re jected. Analysis of molecular variance (AMOVA) of data genetic differentiation between th e strongly structured haplotypic regions in Puerto Rico was global ST and FST = 0.77967 and 0.81501 ( P < 0.00001), respectively. The five Puerto Rico regions ha d strong haplotype divisi on (Figure 2-1). The north shore was composed primar ily of the A01 haplotype with one B01 and one B02 individual identified. When the nuclear genotypes were su bmitted to assignment testing, the B01 individual was assigned to the south (B01) and the B02 i ndividual was assigned to the east (B01) group. The south shore was composed entirely of the B01 haplotype. A mixtur e of A01 and B01 were detected along the east coast. A01, B01, and the other closely related haplotype, A02, were located on the west coast. This mitochondrial DNA pattern suggests female specific site-fidelity on the north and south coasts and some m ovement to the east and west coasts. Microsatellite Marker Analysis The Puerto Rico population has low levels of nuclear polymorphism (HE = 0.447 (0.1730.708); HO = 0.454 (0.191-0.745); A = 3.9 (2-6)) as co mpared to the Florida population (HE = 0.480; A = 5.3) over 18 loci. Additional results for the Florida population can be found in Pause et al. (2008). Within Puerto Ric o, three loci deviated from Hardy-Weinberg equilibrium ( TmaE7 TmaE08, and TmaK01) even after a sequential Bonfe rroni adjustment. The deviation may be due to inbreeding, substr ucturing of the population (i.e., Wahlund effect), or the presence of null alleles. Two loci (E08 and E14) had evidence of null alleles due to a heterozyogote 58


deficiency. After 105 comparisons and a Bonfe rroni correction, linkage disequilibrium was not observed (overall = 0.05, P < 0.001). The inbreeding coefficient FIS was -0.004 overall and did not suggest inbreeding in the popul ation. Private alleles were de tected for Florida (16) and Puerto Rico (13) at low freque ncy. The error rate was determ ined by re-genotyping 11% of the individuals. All detected errors were due to inconsistencies in the PCR, fragment analysis, or scoring. Average error among all alleles was 16.6% due to one problematic Puerto Rico sample. No errors were found after that sample was removed from the analysis. Genetic differentiation among populations was estimated by using pairwise FST and RST comparisons. Pairwise FST and RST values for Florida and Puerto Rico were 0.163 and 0.119, respectively, and significant ( P < 0.001). The majority of values among the five geographic groups in Puerto Rico were lo w but significant (Table 2-2). The proportion of the genetic variance contained in the Puerto Rico population relative to the total genetic variance, FST, was 0.1008 ( P < 0.01), indicating moderate differentiation among the populations. Of the 10 pairwise FST estimates, nine were signifi cant. Moderate but highly significant (p 0.007) FST population structure was identified for N and EA (0.051) and W and EA (0.064). Of the 10 RST comparisons, five were si gnificant. Significant RST values indicating moderate to strong structure with W versus N (0.052), EB (0.074), S (0.099), and EA (0.111). It should be noted that the RST value of adjacent units N and EA was slightly over significance ( P < 0.053) with moderate differentiation (0.048). The only significant female FST comparison with moderate genetic differentiation was E and W (0.051). Significant female RST values were higher, including S and W (0.085) and E and W (0.096). Male FST values were low but significantly di fferent between N and E (0.027) and N and S (0.044). No male RST values were significant. 59


Significant male and female pairwise FST comparisons included FE and MS (0.047) and FE and MW (0.063). The significant male and female RST comparison FW and MS (0.138) indicated strong structure. Cluster analysis using multi-locus genotypes Bayesian methods in the program STRUCTURE assigned individuals to genetic clusters without a priori population designation. The Florida and Puerto Rico manatees analyses had similar fluctuating ln( Pr ( X | K )) estimates {K = 2-8} with K = 2 being the fewest number of clusters capturing the major stru cture in the data with little admixture (Pritchard & Wen 2004). The resultant K = 2 proportion of each Florida individual having ancestry in Florida was q = 0.986 and each Puerto Rico individual ha ving ancestry in Puerto Rico was q = 0.979 (Figure 24). Florida analyzed alone i ndicated four clusters (Pause et al. 2008). The Puerto Rico population analyz ed alone showed no delineation of population structure. The results remained mostly unchanged from 1-10 clusters with the lowest value at K = 1 cluster (-2576). Since Florida and Puerto Rico were determ ined to be genetically distant, Puerto Rico was run with a genetically similar Florida clus ter serving as a divergent group for comparison. This analysis produced K = 4 clusters (-2945), with Florid a individuals groupi ng as one cluster and the Puerto Rico population breaking into 3 clusters. Low q values were detected in each cluster, indicating admixtur e of these individuals. The first Puerto Rico cluster, q = 0.76, contained 41 individuals, 71 % from the N and E. No seasonal or haplotype grouping was detected. The first 18 individuals were 82% N and E with an average q = 0.91. Clusters 2 and 3 had low assignment values, q = 0.48 and 0.56, respectively, indicating reduced genetic signature or admixture. Cluster 2 contained 45 individuals and was mostly composed of S and W animals (78%) with 69% of its total membership collected during th e rainy season. It was mostly composed of B01 haplotypes 60


(80%) but did contain th e four A02 haplotypes in the W Puer to Rico population, three of which grouped together. The E animals in cluster 2 we re all B01. Cluster 3 contained 23 individuals with a majority of N and E individuals (65%) a nd 70% of its members be ing collected during the rainy season. Un-rooted neighbor-joining trees The neighbor-joining method is useful to ad dress evolutionary re lationships between populations (Nei 1996). The Flor ida and Puerto Rico indivi dual neighbor-joining tree represented a robust division be tween the two subspecies. All individuals were correctly assigned to their respective popul ations except for two Florida animals. These individuals are migrants or had genotypes similar to Puerto Rico As these are subspecies and have not been separated for a long evolutionary time, it is like ly that individuals may have similar genotypes by chance. Additionally, a large genetic distance separated the Florida and Puerto Rico groups (Figure 2-2). The Florida and Puerto Rico population neighbor-joining tr ee depicted a strong distinction between the two West Indian subgr oups with 100% bootstrap support (not shown). An individual Puerto Rico ge notype tree indicated mostly mixe d genotypes and some geographic substructure (not shown). The Puerto Rico population neighbor-joining tree specified two major sub-groups within the island, N cl ustering with E and S clustering with W (Figure 2-3). The S and W coasts were separated from the N and E coast with 100% bootstrap support. The N and E coasts were separated with 96% bootstrap support. Cytogenetic Analyses The Giemsa-banded karyotype analysis confir med that the Puerto Rico manatee has 48 chromosomes. This is in agreement with solid stained West Indian coastal Brazil manatees and banded Florida manatee chromosomes (Assis et al. 1988; Gray et al. 2002; Vianna et al. 2006). The banding pattern was an alogous to that observed in the Florida manatee. 61


Discussion Temporal separation of the Florida and Puer to Rico manatee populations has resulted in considerable divergence at th e nuclear DNA level and separate evolutionary progression. Although a reproductive barrier is most likely extrinsic, the long standing division, minimal migration, and an apparent bottleneck in Florida has resulted in the accumu lation of considerable differences between the Florida and Puerto Rico populations. The extent of the Florida and Puerto Rico differentiation is substantial and supported by highly significant findings corresponding to subspecies taxonomic classification and mo rphological and craniometry distinctions (Domning & Haye k 1986; Mignucci-Giannoni 1996). The separation is supported by FST and RST pairwise values, multi-locus genotype analyses, and un-rooted microsatellite neighbo r-joining trees. From these results the Flor ida and Puerto Rico manatee should be considered, managed, an d conserved as two dist inct populations with little to no migration and diverg ent genetic makeup. The small, isolated Puerto Rico population should be managed to protect its genetic diversity, since divergen t individuals from Florida are unlikely to join the population. Additionally, significant genetic diversity statistics and strong assignment to the resident populations indicate that the Belize and Mexico population is also separate from the Puerto Rico population (data not shown). This suggests that there is little genetic relationship between these populations in the Western Caribbean. Florida and Puerto Rico Mitochondrial DNA Garcia-Rodriguez et al. (1998) and Vianna et al. (2006) also identifie d one haplotype in Florida (A01) and three of the haplotypes in Puerto Rico (A0 1, A02, and B01). Colonization and/or a population bottleneck we re proposed to explain the lim ited haplotype diversity found in Florida (Garcia-Rodriguez et al. 1998). The A01 haplotype in both populations suggests a common ancestor or historical rela tionship and is likely due to th e colonization of Florida from 62


the Greater Antilles approximately 12,000 ya (Domning 2005; Reep & Bonde 2006). To date, Florida does not contain the B01 haplotype, co rroborating the limited migration between the populations and the A01 colonizatio n of Florida. It is possible that A01 females on the North coast of Puerto Rico traveled further north and colonized Florida. A01 is unique to northern populations, found also in Mexico and the Dominican Republic. The effective number of mtDNA is 25% of nuclear DNA, indicating th at mtDNA is more sensitive to population bottlenecks. Therefore, A01 could have quickly drifted to fixation in Florida. A clear haplotype pattern was identified in Pu erto Rico, suggesting strong female sitefidelity. MtDNA is maternally inherited and is not reflective of ge ne flow from males. Single haplotypes were detected excl usively on the N and S coasts indicating minimal movement between subpopulations. The mixed haplotypes to the E and W suggest movement of A01 and B01 females into these regions, presumably from the bordering subpopulations. In fact, an A01male manatee captured on the we st coast as part of a radiot elemetry study traveled to the north coast for a short time and then returned to the west coast. On the other hand, a B01female captured during the same study traveled consisten tly back and forth between the west and south coasts of the island. Since A02 has not been f ound in any other population to date, it most likely evolved from A01 descendants in this subpopulation. The unique pattern of A01 and B01 haplotypes on the N and S coasts, respectively, may be explained by the reduced population size. Small, isolated sub-populations of single haplotypes may have persisted and grown. The S coast may have been founded by B01 animals that moved to or from the E and W coasts followed by diverg ence to the A01 haplotyp e. Alternatively, the two coasts may have been founded by the separate haplotypes. 63


Florida and Puerto Rico Microsatellite Analysis Genetic diversity statistics, genetic distan ces, and multilocus Bayesian assignment tests support a strong nuclear divergen ce between the Florida and Puer to Rico manatee populations. According to Frankham et al. (2002) an FST of 0.15 represents significant genetic differentiation. Other studies have also corr oborated this division, suggesting that an FST in the range of 0-0.05 indicates little differentiation; 0.05-0.15, mode rate differentiation; 0.15-0.25, great differentiation; and values above 0.25, very great genetic di fferentiation (Balloux & Goudet 2002; Hartl & Clark 1997). Th erefore, the significant FST value of 0.163 between Florida and Puerto Rico indicates great ge netic differentiation between the populations. The populations have apparently been separated for many years w ith little recent migration between them. In fact, Vianna et al. (2006), using the program BARRIER 2.2 (Manni et al. 2004), identified a gene flow barrier between Florida and Puer to Rico using mito chondrial DNA. FST was 1.4 times larger than RST, indicating that random drif t rather than mutation is responsible for the genetic differences (Frankham et al. 2002). Since Puerto Rico is a small population and Florida was likely founded by few i ndividuals from the Gr eater Antilles, drift could have quickly separated the allele frequencies of the two populations. The delineation of the populations as shown by the program STRUCTURE 2.2 supports the hypothesis that these two populations have little genetic similarit y, with no admixture or recent migration, and should be considered two genetic ally distinct groups (Figure 2-4). The high proportion of assignment and ancestry of Florida i ndividuals to Florida (0 .986) and Puerto Rico individuals to Puerto Rico ( 0.979) represents strong separatio n and minimal recent migration between the populations. The similar Puerto Rico genotypes assigned to Florida could be explained by chance, since the genotypes could be similar, immigration or descendants of recent immigrants, or sample processing errors. The individual and population tr ees from Puerto Rico 64


and Florida also indicated larg e genetic separation a nd little mixing between the two populations with 100% bootstrap support. Population Structure and the Environment The different habitats in Flor ida and Puerto Rico may influe nce the divergence observed in population structures. Although some were significant, the Florida genetic differentiation values did not reach the 0.05 cut-off value, indicating gr eat mixing and weak structure. The largest FST value was 0.033 and the largest RST value was 0.027 (Pause et al. 2008). Within Puerto Rico, substantial genetic differentiation was reached (e.g., FST 0.064 and RST of 0.111). Peninsular Florida has a temperate climate, with significant seasonal water temperature changes. These changes trigger manatee migra tion between warm-water refugia in the winter and highly nutritious marine ecosys tems in the warmer months (Rathbun et al. 1995b; Rathbun et al. 1990). Therefore, Florida manatees travel long distance s around the peninsula and breed with other individuals th roughout the population (Deutsch et al. 2003; Fertl et al. 2005; Weigle et al. 2001). This results in high ge ne flow and little genetic diffe rentiation or structure in the population. Alternatively, Puerto Rico is a marine envir onment with minimal seasonal climate change. The manatees are not required to tr avel long distances to obtain re sources or warm water for their survival. During the period of April 1992 to June 2006, 33 manatees were radio tagged and tracked by USGS-FISC personnel and their collaborators (Slone et al. 2006). Manatees were captured on the E coast or in the SW region. The majority of the tracked animals had restricted movement patterns, alternating between seagrass beds and local fresh water sources. Many animals remained in the immediate area of Guayanilla Bay in the SW, or around Ensenada Honda on the E coast. This limited movement decr eases gene flow and allows for the formation of subtle population structure. Weak seasonal differences were detected only by the Bayesian 65


clustering method. Two of the clusters were 70% composed of animals collected during the rainy season. Puerto Rico Genetic Diversit y and Geographic Division Recent meta-analyses of microsatellite data determined that demographically-challenged mammalian populations (HE of 0.60 and A = 6.17-6.59) have lower genetic variation than undisturbed, healthy populations (HE = 0.65 and A = 8.18; DiBattista 2007). Garner et al (2005) identified similar heterozygosity values (challenged HE = 0.5; healthy HE = 0.68). The Puerto Rico population had lower genetic diversity (HE = 0.447; A = 3.9) than disturbed populations experiencing pollution, harvesti ng, or habitat fragmentation. Expected heterozygosity values were 9% lower for Puerto Rico than Florida, indi cating less variation in Puerto Rico. The earliest accounts of Puerto Rico include Tainos, Caribs, and Spaniards using manatee meat as an important food source (Acost a 1590; Stahl 1883). In fact, the Pope declared manatees a type of fish, allowing Spaniards to consume them on Fridays (Reep & Bonde 2006). Manatee hunting and consumption continued un til a last record wa s registered in 1995 (Mignucci-Giannoni et al. 2000). The severely reduced genetic di versity in Puerto Rico is likely reflective of the small population size and long-term persecution. Geographical genetic structure The Puerto Rico population was subdivided by five geographically separated haplotype groups. The small area and documentation of mana tees traveling great di stances (Reep & Bonde 2006) would have likely caused a homogenous haplotype distributi on without habitat barriers. Instead, a strong haplotype separation was detect ed, indicating habitat barriers. The nuclear DNA subpopulation separation was not as severe, suggesting that animals do travel and breed throughout the population to some degree. 66


The majority of the geographical subpopulation pairwise FST and RST values were low, but significant and consistent with the haplotype pattern, suggest ing subtle but detectable geographical structure. Only two pairwise FST values were above the 0.05 moderate differentiation threshold. The N and EA mode rate divergence (0.051) was unexpected as the units are adjacent to one another. A large amount of recreational development has taken place in the NE region, and may have contributed to the separation of the subpopulations. The E and W were significantly different at a moderate leve l, most likely due to the large distances and barriers to movement between the two coasts. When calculating significant RST pairwise values, four had m oderate differentiation and six of the 10 were significant. Interestingly, the S and W RST value was highly divergent (0.099) and significant, while the FST value was not significant. This could suggest differentiation from mutation as opposed to divergence through allele frequency drift. When the population was analyzed by geogra phic location and sex, SM and EF values were significant and moderately divergent, sugg esting that only the SF (B01) are traveling and breeding with the E (B01) subgroup and only EM ar e traveling and breeding in the south. The A01 haplotype is not disseminated by the EF to the S, while the B01 haplotype is moved to the E from the SF. Large and significant RST values separated the WF from the SM and SF. This again suggests a lack of WF (A01, A02, or B01) traveling S and so neither A01 haplotype is incorporated into the S and only B01 is maintained. Males and females from within one geographical group did not show any significant differentiation, s upporting the genetic unity of the subpopulations. Strong division in the island was indicated by th e separation of the NE and SW areas. The distribution of individuals id entified through strandings, aerial surveys, and tracking studies 67


indicate that the S and E ar e the most inhabited areas on the island. The 1992-2006 tracking studies identified the W and SW and the E and NE coasts as heavy use areas (Slone et al. 2006). The rivers used by manatees to obtain fres h water include the Guanajibo, Guanica, and Guayanilla, on the W and S coasts, and Yauco, a nd the effluent of the Cape Hart Sewage Treatment Plant on the E coast, corresponding to the heavy use areas. A high incidence of movement between the W and SW coasts was observed from Guanajibo to Guanica, potentially between the fresh water sources. Regions known to have large populations, but not targeted in the tracking studies included Jobos Bay, San Juan, and Luquillo. No radio tagged manatees utilized these areas, which were not far from E and SW capture sites, respectively, supporting the theory that these manatees have small home ra nges and rarely travel to these nearby areas. Barriers to movement in Puerto Rico These analyses indicate subtle population struct ure, which may be influenced by migration patterns, lack of time for convergence, or poor hab itat areas. Few sightings take place in the NW (Rincn to Barceloneta) and in the SE municipality of Maunabo, which are deep close to shore (Figure 2-1, black bars). For example, aeria l surveys from 2002-2004 sighted many manatees along the S coast but none in Ma unabo. Seagrass beds were dete cted in these regions, but in waters deeper than 18 ft with hi gher wave energy. Manatees rema in in isolated coastal habits and are rarely seen in open or deep ocean (Lefebvre et al. 2001). Deep water represents poor manatee habitat and possibly hinders manatee movement. A positive and significant correlation was found between genetic and coas tline geographical distances using a Mantel test, supporting the idea that manatees migrate along the coasts and rarely move in pelagic waters (Vianna et al. 2006). Because manatees are dependent on freshwat er sources and shallow vegetation, they stay close to shore. Manatees feed a nd rest in shallow water and prefer water depths of three to seven 68


ft (0.9 to 2.1 m). Deep seagrass beds are unlikely to be foraged and may be located in areas with high-energy wave action. Another impediment to movement may be hi gh wave action on the NW coast of Puerto Rico. Strong waves and currents likely discour age manatees from traversing these waters. Powell et al. (1981) and Rathbun et al. (1985), using aerial survey te chniques, did not detect any animals from the Culebrina River (W coast) to the Manti River (centr al N coast) where high energy wave action is observed. The poor and unused manatee habitats appear to prevent a panmictic population from forming, but do not prevent all movement. Ma natees can travel great distances (Deutsch et al. 2003; Fertl et al. 2005) and have crossed open ocean (Alvarez-Alemn et al. 2007). A degree of movement throughout the island is corroborated by lo w genetic differentiation statistics and little geographic population structure detected by STRUCTURE. Preservation of the Manatee in Puerto Rico Due to the small number of manatees in Puerto Rico, the population must be actively monitored and managed. The population estimate of 250 (ranging from 150-360) is a cause for concern (Frankham et al. 2002). It has been suggested that a minimum of 50 effective breeders (10% of the population) is needed to prevent inbreeding depression needed for long term survival (Wright 1951). Puerto Rico is well below this th reshold. This implies that population levels in the upper hundreds to thousands ar e needed to maintain evolutionary potential (Franklin 1980; Lande 1995). Although no physiological or genetic implications of inbreeding have been identified to date, the population is small and highl y susceptible to detrimental genetic effects. Demographic and stochastic events can qui ckly reduce genetic va riation and population levels in groups with few individuals. Small populations also have reduced genetic diversity, which can negatively influence fitness (Roelke et al. 1993), increase susceptibility to disease 69


(O'Brien et al. 1983), and decrease population viability (Sherwin et al. 2000). Immigration is most likely low, limiting the potential genetic di versity that could supplement the population. The islands close to Puerto Rico (i.e., Cuba, Jamaica, Hispaniola) are thought to have small or no remaining manatee populations to provide additional individuals. The high genetic differentiation values suggest th at immigration is minimal from th e larger Florida, Belize, or Mexico manatee populations. Since manatees have a long generation time (Marmontel 1995; Rathbun et al. 1995a), special habitat requirements (Reynolds et al. 1995), and vulnerability to stochastic events such as cold stress and red tide (O'Shea et al. 1991), it is imperative that anth ropogenic threats that cause manatee mortality be monitored and reduced. Increased human impacts and the high rate of development throughout Puerto Rico have strongly affected the environment. Humans tend to colonize regions that are excellent manatee habitat, such as protected and shallow bays with access to fresh water. Watercraft traffic and human presence has increased in bays where ma natees previously sought food, freshwater, and protected areas for rest and giving birth. The Puerto Rico manatee population has a history of being hunted and is therefore wary of humans. Th ey may not utilize resour ces near areas of high human activity. Boat traffic increases proportionally to th e human population and poses the largest anthropogenic mortality threat to the Puerto Rico manatee population (MignucciGiannoni et al. 2000). If the fast pace of human coloni zation and habitat destruction continues in Puerto Rico, manatees may be left with little sustainable habitat and no place to go. To assist the survival of the Puerto Rico Antillean manatee populat ion, the Puerto Rico Manatee Recovery Plan (Rathbun & Possardt 19 86) must be updated and implemented by law enforcement, rescue and rehabilitation groups, and cooperative multiagency agreements to assess 70


and reduce threats. Specifically, boat and jet-ski speed and traffic in manatee use areas must be evaluated, regulated, and enforced to reduce mo rtality and encourage manatee utilization. Enforcement must be accompanied with a multimed ia outreach campaign to educate the boat and coastal community. The SE and NW corners of the island should remain open to allow gene flow and promote genetic diversity between the subpopulations. Additionally, scientific research on life history traits and recurrent population surveys are needed to monitor the population status in the quickly changing Puerto Rico envir onment. An active and continued, science, conservation, and management effort will help to ensure the preservation of the Puerto Rico manatee population. 71


Table 2-1. Characteristics of the 15 polymorphic microsatellite loci implemented on the Puerto Rico manatee (T. m. manatus) samples Locus name Tm( C ) BSA A NE PIC HO HE TmaA02 56 2 1.319 0.407 0.209 0.242 TmaE1 55 + 5 2.323 1.065 0.571 0.570 TmaE02 58 2 1.766 0.625 0.418 0.434 TmaE7 56 + 5 1.996 0.841 0.420 0.499 TmaE08 60 5 1.581 0.714 0.205 0.368 TmaE11 58 5 3.419 1.355 0.745 0.708 TmaE14 56 + 5 1.840 0.837 0.369 0.457 TmaF14 58 2 1.427 0.476 0.295 0.299 TmaH13 60 4 1.984 0.843 0.527 0.496 TmaJ02 62 3 1.754 0.714 0.500 0.430 TmaK01 58 4 1.849 0.722 0.640 0.459 TmaKb60 62 6 1.923 0.808 0.464 0.480 TmaM79 54 + 2 1.209 0.315 0.191 0.173 TmaSC5 60 5 2.264 0.933 0.636 0.558 TmaSC13 56 4 2.129 0.921 0.613 0.530 Optimized annealing temperature ( Tm ), BSA requirement (0.4 mg/mL), number of alleles ( A ), effective number of alleles ( NE), polymorphic information content ( PIC ), and the observed and expected heterozygosity (HO and HE) for the Puerto Rico T. manatus manatus population Table 2-2. Pairwise FST (above diagonal) and RST values (below diagonal) generated from a survey of 15 microsatellite loci from T. m. manatus in five geographic regions in Puerto Rico. Italics indicate st atistically significant values Geographic region North East A01 East B01 South West North 0.051 0.025 0.031 0.038 East A01 0.048 0.029 0.037 0.021 East B01 0.009 0.014 0.030 0.064 South 0.030 0.016 0.007 0.015 West 0.052 0.111 0.074 0.099 72


73 Figure 2-1. Puerto Rico bathym etric map with location and mitochondrial haplotype assignment of captured or recovered manatees Black bars represent water depths of 100 fathoms close to shore


74 Figure 2-2. Florida (orange) and Pu erto Rico (blue) individual neighbor-joining tree, depicting nearly 100% correct assignment of individuals to their re sident population and a large genetic distance between the populations (arrow)


Figure 2-3. Puerto Rico un-root ed neighbor-joining tree, depicti ng a separation of the South and West from the North and East (100% suppor t) and a North and East separation (96% support) Figure 2-4. Summ ary plot of q estimates generated by the seque ntial cluster analysis of the program STRUCTURE performed on the Florida and Puerto Rico T. manatus genotypes 75


76 CHAPTER 3 CONSERVATION GENETICS OF THE ANTILLEAN MANATEE ( Trichechus manatus manatus) POPULATION IN BELIZE: THE LAST CARIBBEAN STRONGHOLD Introduction The West Indian manatee ( Trichechus manatus) is an endangered aquatic mammal found throughout the southeastern United States, Caribbea n, and Central and South America. All West Indian populations are classified as vulnerable to extinction (1982) by the International Union for Conservation of Nature (IUCN 2007). Two subs pecies are recognized, the Florida manatee ( T. m. latirostris) and the Antillean manatee ( T. m. manatus) The Antillean ma natee population in Belize, located south of Mexico and north and east of Guatemala, is the largest and most stable population in the subspecies range (Auil et al. 2007). Early legislat ion limited harvesting and encouraged conservation of the im periled population throughout Belize. Extensive hunting decimated many of the An tillean manatee populations and continues in many countries even today (Lefebvre et al. 2001; Smethurst & Nietschmann 1999). The historical Mayan people consumed a large amount of manatee meat and utilized many tissues in ceremonial activities, as indicated by examination of bone mi ddens at archeological sites (Gann 1911; Thompson 1939). Records from the voyage s of Christopher Columbus introduced European society to the Caribbean manatee, and during the 1700s and 1800s, the Spaniards and indigenous people severely exploited the sp ecies for sustenance. By 1936, the population decline was so severe that Belize introduced Ma natee Protection Ordinances to preserve the population (McCarthy 1986). Currently, the mana tee is listed as endangered by the Belize Wildlife Protection Act of 1981, Part II, Section 3(a) (Auil 1998). A Manatee Recovery Plan was written in 1998, requesting information on habita t use and movement patterns to aid in the development of conservation policies for the protection of the population.


Direct threats, a long generation time, a nd environmental impacts from the burgeoning human population limit manatee population growth. Manatees inhabit highly developed coastal regions, including fresh water rivers, brackish lagoons, and marine habitats (Lefebvre et al. 2001). Therefore, anthropogenic harm and mortalities are a re sult of watercraft collision, development and habitat destruction, incidental net and fishing gear entanglements, and recently, tourism activities (Auil et al. 2007). Increasing re sidential and industria l coastal development can quickly destroy manatee habitat through dred ging, agricultural contamination, sewage and industrial effluents, and mangrove and seagrass destruction. Harm to seagrass, a major food source for manatees, occurs through pollution, sedimentation, siltation, and secondarily through mangrove destruction. This environmental damage has a direct impact on manatee food sources and thus population size. A dditionally, despite the 1930s pr otections, imprudent poaching continues in Belize to this day (Bonde & Potter 1995). Causes of manatee deaths in Belize are similar to those imposed on most manatee populations. A study from 1996-1998 id entified perinatal (32%) a nd human related mortalities involving watercraft (16%) and poaching (20%) as the most common causes of death (Auil 1998). Calves made up 44% of the moribund manatees evaluated. A more recent study identified watercraft (17%) as the most comm on cause of death, aside from undetermined (20042007; Auil et al. 2007). Belize manatee abundance and distribution has been evaluated using extended-area aerial survey methods developed by Packard (1985). The greatest number of individuals counted during an aerial survey was 338 in 2002 (Auil 2004). High use areas were determined to be the Cayes adjacent to Belize City, the Belize River, Southern Lagoon, Placencia Lagoon, Corozal Bay, Indian Hill Lagoon, and the Port Honduras area including Deep River and Seven Hills 77


Lagoon. A five-year survey (1997 and 1999-2002) of manatee abundance in Belize indicated an overall slightly negative trend in the number counted per year, although variation in seasons and survey routes make conclusions difficult (Auil 2004). Capture of wild manatees for health assessm ent began in 1997 in an effort to provide information on the health status, age class, reproduction, distributi on, and genetic structure of the Belize population. Since that time, 118 unique anim als (50% female and 50% male) have been captured in 213 health assessments (55% female a nd 45% male). The study locations include the high use areas of Belize City Ca yes, a reef lagoon system off Belize City, the Southern Lagoon system, two large brackish inland lagoons, and more recently Placencia Lagoon, located 75 miles south of the Southern Lagoon system. The Belize City Cayes (BCC) are utilized by the tourist industry and experience heavy boat traffic from Belize City, the largest port in the country. Sight seeing, fishing, SCUBA diving, and recreational watercra ft activities take place in the marine waters surrounding the Cayes. The majority of carcass samples used in this study were recovere d near the Belize River mouth, an area of high watercraft activity, or so uth of the river, possibly caught in the ocean current. The Southern Lagoon system includes Southern (SL) and Northern (NL) Lagoons, collected by Main Creek, a narrow 2km waterway. It is located in a remote region, less developed than the Cayes. Travel between th e Belize City Cayes and Southern Lagoon can be accomplished using the Sibun or Belize Rivers from Belize City or through the Bar River connecting Southern Lagoon with th e Caribbean coast (Figure 3-1). In 1989, Southern Lagoon had the highest aerial survey count in Belize, with 55 manatees (O'Shea & Salisbury 1991). While the population is not heavily affected by watercraft, it is 78


potentially impacted by salinity changes due to incr eased rainfall, influencing food abundance for the manatee. Placencia Lagoon (PL) is a long, narrow lagoon 24km in length and 1-2 m deep, protected by the Placencia Peninsula. Plac encia is a popular tourist destination, with a large amount of coastal development negatively a ffecting the environment. Manatees here are also affected by agricultural runoff and shrimp farm discharge, as the sub-aquatic vegetation quantity and possibly quality has changed. Since the Belize manatee population is larg est in the Wider Caribbean (Auil 2004; Quintana-Rizzo & Reynolds III 200 7), it is postulated that Belize could provide genetic diversity and potentially assist in the r ecovery of populations to the sout h. The adjacent Antillean manatee population in Mexico, which is al so protected, is the second larg est. Little is known of the abundance or distribution of mana tees in the other Central American populations, although they are considered elusive and thought to be rare (Quintana-Rizzo & Reynolds III 2007). Manatees can travel great distances and could easily travel to and utilize resources among surrounding countries. In fact, a radio tagged individual wa s documented to travel from Mexico to Stann Creek, near PL, where the tag broke free (Morales-Vela et al. 2007). It is therefore essential that the manatee population in Belize be protected to allow for growth and supplementation of other populations. Future studies of genetics and immi gration analyses could identify the degree of movement and breeding occurri ng with other populations. Previous mitochondrial DNA studies identified three haplotypes in Belize, A03, A04, and J01, but could not address fine scale population structure (Vianna et al. 2006). Presented here are genetic analyses evaluating movement of manatees in the Belize City Cayes and Southern Lagoon system, high use areas of Belize. Mitoch ondrial and microsatellite DNA were used to elucidate the genetic diversity, re latedness, and population structur e of three regions in Belize. 79


Subpopulations that are determined to be diverg ent should be protected to preserve the unique genetic diversity. Additionally, corridors should be maintained to encourage the exchange of diversity among regions. The Belize Antillean populat ion will also be compared to the Florida population ( T. m. latirostris) to address the genetic relations hip between the subspecies. Materials and Methods Sample Collection and DNA Extraction Manatee blood and/or epidermi s tissue were collected from recovered carcasses or during wild manatee health assessments. Genomic DNA was isolated usi ng QIAGENs DNeasy Blood and Tissue kits (Valencia, California). Flor ida manatee sample DNA extraction techniques are described by Pause et al (2007). Within the Florida datase t, 96 individuals were randomly chosen, proportionally representing the four demographically imposed management units: Northwest, Southwest, Atla ntic and St. Johns River. Mitochondrial DNA Analysis Primers and PCR parameters from Garcia-Rodriguez et al (1998) amplified a 410 base pair portion of the mitochondrial DNA control re gion displacement loop for 116 individuals, 101 live captures and 15 carcasses. Mitochondrial DNA is maternally inherited and reflects the movement patterns of only females. The mitochondrial DNA control region was polymerase chain reaction (PCR) amplified with primer s developed from 100% homology of cow and dolphin sequences (heavy stand primer: CR-5 and light strand primer CR-4) after Southern et al. (1988)and Palumbi et al. (1991). The PCR reaction condition s were as follows: 10 ng DNA, 1 x Sigma PCR buffer (10 mM Tris-HCl, pH 8.3, 50 mM KCl, 0.001% gelatin; St. Louis, MO), 0.8 mM dNTP, 3 mM MgCl2, 0.24 M of each primer, 0.04 units of Sigma Jump Start Taq DNA polymerase. PCR cycling profile: 5 min at 94C; th en 35 cycles of 1 min at 94C, 1 min at 55C, 1 min at 72C; then 10 min at 72C. Amplifie d products were purified using the Qiaquick PCR 80


purification kit (QIAGEN). DNA sequencing was accomplished in the DNA Sequencing Core at the University of Florida with the BigDye term inator protocol develope d by Applied Biosystems Foster City, CA, using fluor escently labeled dideoxynucleotides. To verify sequences, haplotypes were aligned with mana tee sequences located in GenBa nk using the default setting in SEQUENCHER 4.5 (Gene Codes Corporation, Ann Arbor, MI). Control region fragments were sequenced in the 5-3 heavy-strand orientation. Finally, a repres entative from each haplotype and any ambiguous sequences were sequenced in th e 3-5 direction to en sure the accuracy of nucleotide designations. The degree of differentiation, FST and ST, between Florida and Belize and between the BCC and SL groups in Belize were calculated using ARLEQUIN 3.1 (Excoffier et al. 2005). Estimates of sequence divergence used the Kimu ra 2-parameter genetic distance model (Jin & Nei 1990; Kimura 1980). The variance distributi on was based on haplotype frequencies alone; all haplotypes were treated as equally differentiated (FST). Lastly, Tajimas D of selective neutrality, the number of polymor phic sites, S, number of nucleotide substitutions, NS, the genetic diversity, h, and nucleotide diversity, were calculated (Nei 1987; Tajima 1993). Microsatellite Analysis A total of 16 polymorphic microsatellite primers (Garcia-Rodriguez et al. 2000; Pause et al. 2007) was PCR amplified from 122 individuals. The study included 88 from the SL system (12 NL, 76 SL), 21 from BCC, 2 from PL and 11 carcasses. The Southern Lagoon system was analyzed both without division and separated as SL and NL. Three NL individuals were also captured once in SL. Two of these were identified in NL multiple times. PCR conditions were 14 ng DNA, 0.8 mM dNTPs, 1x Sigma PCR buffer (10 mM Tris-HCl, pH 8.3, 50 mM KCl, 0.001% gelatin), 0.04 units Sigma Jump Start Taq polymerase, and 0.24 M each primer. MgCl2 concentrations were 3 mM, except for TmaH13 TmaKb60 and TmaSC5 81


which required 2 mM. BSA was added as indicated in Tabl e 3-1. Amplifications were carried out on a PTC-200 thermal cycler (MJ Research ; Waltham, MA) using the following conditions: initial denaturing at 95C for 5 mi n, 35 cycles at 94C for 30 s, annealing temp for 1 min (Table 3-1), 72C for 1 min, final extension 10 min at 72C (Pause et al. 2007). Fragment analysis was performed on an Applied Biosystems ABI 3730 Genetic Analyzer. GENEMARKER, version 1.5 (Soft Genetics, State College, PA) was used to analyze the microsatellite fragment data. All individuals amplified at 14 or more loci. The Fl orida data were kindly provided for use in this manuscript (Pause et al. 2008). A Microsoft Access (Micro soft Corp., Redmond, WA) database was developed for storage of allelic information. Statistical Analysis The level of polymorphism wa s estimated by the observed (HO) and expected heterozygosity (HE) and the number of allele s per locus (A) using GENALEX 6 (Peakall & Smouse 2006). Departures from the expected genotypic frequencies in Hardy-Weinberg equilibrium (HWE) were tested using the Mark ov chain method (dememorization 10000, batches 100, iterations per batch 5000) in GENEPOP 4.0 (Raymond & Rousset 1995). Additionally, linkage disequilibrium was tested for non-random a ssociations between allele s of different loci. The Markov chain method was used and the P -values were adjusted using Bonferroni sequential correction for multiple comparisons. To assess ov erall genetic differentiation at the population level, GENALEX 6 calculated FST using the infinite alleles model and RST using the stepwise mutation model. Comparisons included Belize and Florida, and NL, SL, and BCC. The analyses within Belize may be biased due to the large proportion of SL system individuals (67.8%). Comparisons with PL were limited, due to the small sample size. 82


Cluster analysis using multi-locus genotypes The program STRUCTURE 2.2 (Pritchard et al. 2000) was used to identify the genetic subdivision within Belize and the genetic relationship and putati ve ancestral source populations of Belize and Florida manatees. STRUCTURE, a model based clustering algorithm, infers population structure by probabilistically assigning individuals without a priori geographic or ancestral knowledge to a specific number ( K ) of clusters (presuma bly populations). In determining the number of clusters, the algorithm attempts to minimize deviations from HardyWeinberg equilibrium. Simulations were conducted using the admixtur e model, which assumes that individuals could have some propor tion of membership ( q) from each of K clusters, leading to the potential identification of recent immigrants. Multiple Ma rkov chains can delineate differences within populations, therefore three parall el chains were analyzed for K = {1}, with a run-length of 100,000 repetitions of Markov chain Monte Ca rlo, following the burn-in period of 10,000 iterations. The three values for the estimated ln( Pr ( X | K )) were averaged, from which the posterior probabilities were calculated. The K with the greatest posterior probability ( Pr 1.0000) was identified as the optimum number of subpopulations. Individual assignment success was recorded as the highest likelihood of assignment ( q) and the percentage of individuals in a cluster with q > 0.90 was calculated. In the Florida and Belize STRUCTURE analysis, an increasing number of clusters were identified in Florida. To a llow subpopulations to be determin ed, Belize was analyzed with a genetically similar Florida cluster. Contras ting a diverse cluster w ith the Belize population allows subtle structure to be elucidated. 83


Un-rooted neighbor-joining trees Un-rooted neighbor-joining trees based on i ndividual and population genetic distances were used to visualize relationships among populations, subpopulations, and individuals. The Phylogeny Inference Package (PHYLIP) estimated genetic distances based on pairwise CalvalliSforza and Edwards chord distance, DC (Felsenstein 2004). DC is based on allele frequencies and provides accurate microsatellite tree topology (Takezaki & Nei 1996). Trees were constructed by comparing individu al genotypes with or without a priori partitioning. Comparison of the Belize and Florida i ndividual genotypes was conducted with MICROSAT, version 1.5d (Minch et al. 1997), to create a distance matr ix and Neighbor in PHYLIP to produce the tree. To create popul ation trees, a priori sub-gro uped allele frequencies were subjected to the programs in PHYLIP. S upport values were determined by 1,000 bootstrap replicates, indicated at the branching nodes of the trees. The lengths of the branches represented relative genetic distances. Fl orida and the NL, SL, and BCC Be lize regions were analyzed. Cytogenetic Analyses Giemsa-banded karyotype analyses have only previously been performed on the Florida subspecies ( T. manatus latirostris ; Gray et al. 2002). Therefore, to assess cytogenetic differences between the subspecies, banded ka ryotype analysis was performed on Belize manatees. Sodium heparin vacutainers were used to collect blood and samples were transported as quickly as possible to the la boratory. The cytoge netic analysis followed protocols described by Gray et al. (2002). Results Mitochondrial Sequence Analysis The A04 (4) and J01 (15; 79%) haplotypes we re identified in the Belize City Cayes individuals. The Southern Lagoon system had A04 (52; 65%), J01 (22) and the only A03 (6) 84


haplotypes. Placencia had one A04 and one J01 i ndividual. Recovered carcasses consisted of A04 (9) and J01 (6) haplotypes. MtDNA sequence divergence estimates were h = 0.030602 and = 0.53430. Within Belize, Tajimas D = 4.11764 ( P < 1.0000) was not significant ( P < 0.05) therefore; the null hypothesis of selective neutrality cannot be re jected. Genetic differentiation estimates between BCC and SL were FST = 0.07818 ( P < 0.04505) and ST = -0.03632 ( P < 0.71171). Twenty-eight polymorphic sites (6.8%) and nucleotide subs titutions were identified in the three haplotypes. Within Florida one haplotype was observed (A01; = 0.00000). Belize and Florida mtDNA sequence divergence estimates were ST = 0.30208 (Kimura 2-parameter) and FST = 0.62640 ( P < 0.00001). Microsatellite Marker Analysis The sixteen nuclear microsatellite mark ers had lower levels of polymorphism (HE = 0.455 (0.238-0.755); HO = 0.455 (0.157-0.745); A = 3.4 (2-6); Table 3-1) than the Florida population over 18 loci (HE = 0.480; A = 5.3). Additional results for the Florida animals can be found in Pause et al. (2008). In Belize, Tma Kb60 and Tma Sc13 had evidence of null alleles due to a heterozygote deficiency. The null alleles in t hose loci may have caused the deviation from Hardy-Weinberg equilibrium, even after a se quential Bonferroni ad justment. After 120 comparisons and a Bonferroni correction, lin kage disequilibrium was observed between Tma E1 and TmaE14 (overall = 0.05, P < 0.05). The inbreeding coefficient FIS was 0.012 overall, suggesting slight inbreeding in th e population. Private al leles were detected for Florida (23) and Belize (15) at low frequency. The error rate was determined by re-genotyping 11% of the individuals. No errors were detected. Genetic differentiation among populations was estimated using pairwise FST and RST comparisons, which were significa nt, although low (Table 3-2). FST and RST between the BCC 85


and SL system were, 0.029, 0.038, respectively and significant. Pairwise FST and RST values for Belize and Florida were 0.141 and 0.082, respectively, and significant ( P < 0.001). Cluster analysis using multi-locus genotypes The STRUCTURE Bayesian assignment test identif ied two highly divergent populations, Belize (98.1%) and Florida (98.6%), with little admixture, ln(PD)AVE = -5804 (Figure 3-2). The Belize assignment percentages were reduced due to two indi viduals, TMBZ-004 and 099, with partial assignment to Florida (33.0% and 41.1 %). A value above 40% indicates that the individuals parents may have or iginated in Florida. TMBZ-099 was an A04 male from SL. STRUCTURE identified only one population when Belize was analyzed alone, ln(PD)AVE = 3011.7. However, when Belize was analyzed with a q-clustered Florida group, two Belize clusters were detected ln(PD)AVE = -3552.53 (Figure 3-3). Florida animals were strongly assigned to Florida (97.1%). The Belize clusters had lowe r assignment values, 83.3%, and 79.9% and contained 52 and 35 SL individuals, respectively. The first Belize cluster contained the majority of the NL inhabitants (83.3%). Th e second cluster containe d the majority of the BCC residents (80.9%). Again, TMBZ-004 and 099 were assigned to Florida, reducing the Belize assignment values. TMBZ-004 was weakly assigned to Florida (42.4 %), while TMBZ099 was strongly assigned to Florida (82.6%), indicating a similar ge notype by chance or a strong genetic relationship to the chosen Florida group. Un-rooted neighbor-joining trees The individual neighbor-joining tree identified Belize and Florida as two separate populations (Figure 3-4). Within the tree, two Belize groups cont ained three Florida animals. One of those clusters included the two Belize i ndividuals that were a ssigned to Florida by STRUCTURE. The genetic distance separating the Beli ze and Florida populations was modest. In the Belize individual neighbor-joini ng tree, the BCC individuals we re in groups of two to five 86


scattered among the SL individuals (Figure 3-4). The population tree id entified strong division between NL and SL/BCC (100%). SL and BCC were also separated 90% (Figure 3-5). Cytogenetic Analyses The Giemsa-banded karyotype analysis c onfirmed that the Belize manatee has 48 chromosomes. This is in agreement with solid stained West Indian coastal Brazil manatees and banded Florida manatee chromosomes (Assis et al. 1988; Gray et al. 2002; Vianna et al. 2006). The banding pattern was an alogous to that observed in the Florida manatee. Discussion In a study of Caribbean manatee populations, OShea and Salisbury (1991) concluded that Belize remains one of the last strongholds for th e species in this part of the world. Suitable habitat and reduced poaching has made Belize th e largest manatee populati on in the region (Auil 2004). Furthermore, until recently the human population has remained small, limiting environmental destruction and habitat fragmentation. Minimal genetic difference was detected with in Belize, indicating migration and breeding throughout the subpopulations. The STRUCTURE program identified two groups, SL with BCC and SL with NL, when Belize was analyzed with a genetically divergen t Florida cluster. Additionally, SL and BCC manatees were less genetically differentiated than NL and BCC manatees, even though the later are in closer pr oximity to one another. This suggests that animals traveling from BCC to the Southern Lagoon system (SL and NL) spend more time breeding in SL than in NL and may use the Bar Ri ver to a greater extent than the Sibun or Belize Rivers to the north. High hu man activity may discourage Sibun and Belize River use and the rivers should be protected as manatee habitat to allow geneflow between BCC and NL. NL and SL were not statistically diffe rent from each other, indicati ng that these lagoons comprise a breeding population. 87


The BCC and SL subpopulations had disparate propor tions of haplotypes. The majority of haplotypes in the BCC was J01 (79%) and A04 in SL (65%), possibly in dicating female sitefidelity and reduced movement between the BCC a nd SL. Haplotypes have been analyzed in Florida and from Mexico to Brazil, and the A03 haplotype has only been found in SL (Vianna et al. 2006). The lack of A03 in other regions suggests that A03 females remain in the population in which they were born, although more samples n eed to be analyzed from BCC and the rest of the region. The haplotype may have spontaneously mutated from A04 and has since remained in isolation in SL. The individual TMBZ-099 was a male 41% relate d to Florida. Males are sighted in the summer at breaks in the Belize Barrier Reef al ong the BCC, most likely utilizing the habitat seasonally, or waiting for receptive females. Th ese males are thought to tr avel a great distance, potentially allowing for movement between coun tries and a mechanism for geneflow between subspecies (Self-Sullivan et al. 2003). Radio-tagged males were also more likely to move away from SL (Powell et al. 2001). Source and Sink The sheltered environment, extensive food, and fresh water provide excellent habitat for Belize manatees. The country has a wide shelf, with a large, relatively shallow reef lagoon (>125m) and no appreciable current or tidal flow. The barrie r reef runs parallel to the coastline for 220-250km, providing bogues and lagoons as manat ee sanctuaries and fe eding grounds (McField et al. 1996). Shallow, calm water is ideal habitat for manatees to forage on superficial seagrass beds and travel in search of res ources or reproductiv e opportunities. The highly suitable Belize habitat allo ws the population to expand, potentially repopulating the surrounding countries. However, poaching could remove expatriated animals before they reproductively contribute to the population. A genetic st udy of other populations 88


could assess whether individua ls are traveling an d reproducing outside of Belize. The populations outside of Belize are small and may require supplementation to persist (O'Shea & Salisbury 1991). Tracking A total of 42 manatees has been radio track ed using a VHF, GPS, or UHF transmitter (Auil et al. 2007). One male tagged with a radio transmitter in SL tr aveled north to Mexico and returned one month later. Manatees tagged in Mexico have tr aveled directly to SL (~200km) and Placencia (~50km further south), in dicating possible knowledge of the coastline and travel across country boundaries. An animal tagged in Mexico traveled directly to SL through the Bar River to participate in mating herds (Auil et al. 2007) This type of lengthy migration, possibly searching for fresh water, reproductive opportunities, and high-quality food, is common in other manatee populations, such as Florida. Of the four manatees tagged in the BCC, three stayed within a 15-mile radius of the capture site. The majority of manatees tagged in the Southe rn Lagoon system remained in SL and had strong site-fidelity to the la goon system. SL has an upwelling spring, manatee hole, which helps support the large population found there and its av ailable fresh water may attract manatees traveling along the coast. The spring is a 10m depression with upwelling water that can reach high temperatures (33 C). Manatees (up to 20 at a time) actively utilize this depression for thermoregulation when the temperatur e in the lagoon drops below 26C (Auil et al. 2007). Only a few individuals traveled to NL, perhaps because SL has better-suited resources than NL, with more quantity and variation of vegetation (Auil et al. 2007). Mainland versus Island Habitats The Belize manatee population structure refl ects the Florida manatee population. These populations are located on the mainland of con tinents, allowing manatees to travel great 89


distances as compared to island geography where travel is more restricted. Mainland manatee habitat is typically uninterrupted habitat w ith few barriers and protected waters, allowing increased movement and geneflow and re ducing population structure. The average heterozygosity and number of allele s are lower in Belize than those estimated for Florida. This is likely due to long-term hunting pressures and the smaller population size of the Belize manatees. Alternatively, manatees on the island of Puerto Rico cont end with narrow continental shelves and steep drop-offs into deep ocean trenches. The North-West and South-East coasts of Puerto Rico have extremely deep water close to shore. This represents poor manatee habitat, with high wave action and deep seagrass beds. Slight reproductive barriers were identified, allowing for genetic differentiation and subpopulati ons to form within Puerto Rico (Kellogg et al. 2008). A meta-analysis of microsatellite data de termined that demographically-challenged mammalian populations affected by historical or long-term harvesting, fragmentation, or pollution have lower genetic variation (HE of 0.5 to 0.6 and A = 6.9) than undisturbed, healthy populations (HE = 0.6 to 0.7 A = 8.8; DiBattista 2007; Garner et al. 2005). Belize and Puerto Rico have much less genetic diversity than the reported average de mographically-challenged populations. This may be intrinsic to the populations, or reflect the severe persecution. Belize and Puerto Rico also have similar genetic di versity, although Belize ha s a slightly higher expected heterozygosity and lower number of alleles. Belize is a larger population allowing for greater genetic diversity however historical and long-term e xploitation may have severely reduced that diversity. The Beli ze population will need significant evolutionary time to attain pre-harvesting diversity levels. 90


Tropical versus Temperate Climates Florida manatees are at the northern limit of their range and must seek warm-water springs or basins during the winter (Deutsch et al. 2003; Fertl et al. 2005). In the warm months, they can travel extensively in search of marine seag rass and reproduction opportunities. The migration causes extensive mixing and little genetic diffe rentiation to occur w ithin the population. Alternatively, studies indicate that precipitation le vels affect Belize manatee distribution (Gibson 1995). The tropical climate in Belize fluctuates little and generally, manatees do not migrate for thermoregulation. In the dry months, animals are more abundant in rivers most likely to drink freshwater. During wet months, individuals are plentiful in the cayes, were food is more nutritious and fresh water is not difficult to find on the surface or from flowing rivers. Like Florida, the large amount of admixture produces little genetic differe nce between the Belize regions. Current Population and Future Directions Although Belize is the largest Antillean population, it is still under pressure from anthropogenic threats. From 1977 to 1991, the popul ation size appeared stable and calves were consistently 7% of the popul ation, indicating goo d recruitment and a healthy, possibly increasing, population. However, a more recent study (1997-2002) found an overall negative trend in manatee numbers. The population may be declining from escalating anthropogenic threats, such as development, agricultural run-off and boat traffic (Auil 2004). The long generation time, changing environmental hazards, and close proximity to human activity could quickly erode Belize manatee genetic diversity a nd ultimately the population size. Moreover, the genetic diversity within Belize is signifi cantly lower than dem ographically-challenged mammalian populations or the Florida population, potentially due to long-term and severe persecution. 91


Improving the reporting, recovery, and examina tion of injured or dead manatees could provide additional distribution, life history, and cause of death data. Enforcement of laws and increasing protections in the impacted regions could reduce the number of anthropogenic deaths. The construction of additional formal sanctuaries, habitat with no-entry or no-fishing areas, is needed. All nets, except for cast nets, should be banned to prevent entanglement. Furthermore, manatee tours have economic value (about $80 per person) and the growing tourism industry could physically harm or disrupt manatee behavior if not conducted properly, making regulations necessary for this i ndustry (Auil 2004). No wake zones and water vessel operation sp eed zones in shallow manatee habitat are needed to decrease boat strikes. A study in 1991 indicated that boat strikes were rare (O'Shea & Salisbury 1991). However, studies beginning eight years later identified boat scars on more than half (55.6%) of the individuals caught off Belize City, while onl y 16.8% of the individuals in Southern Lagoon had scars at the initial capture (Auil et al. 2007). The increasing incidence of boat strikes near Belize City indi cates the direct harm increase d boat traffic could have on the population. The evaluation and reduction of ha bitat degradation is critical to the survival of this elusive aquatic mammal. Pollution and effluents from industry and development should be highly monitored and regulated. Damage to seagrass beds by shrimp trawlers and shrimp farms must be monitored and reduced. The nitrif ication from shrimp farm effl uent can cause extensive algae blooms affecting seagrasses. The alga provides food for a marine snail, which is an intermediate host of a respiratory fluke ( Cochleotrema cochleotrema) known to parasitize manatee lungs. Manatees in Placencia Lagoon can carry large load s of the fluke and present at capture with mucoid discharge, possibly re lated to an infection (Auil et al. 2007). Additionally, the algal 92


bloom compromises the quality and growth of th e seagrass. The relatio nship of nitrification, snail, and algae densities on seag rass, and increased manatee parasi te loads must be investigated in this region. Enforcement of guidelines and regulati ons by Belize and neighboring countries government and institutions is essential to the pr eservation of manatee populations. Several local non-governmental organizations ha ve strong local and national manatee education programs. These should be empowered to continue educat ing the Belizean peopl e (O'Shea & Salisbury 1991). Collaboration among the governments of Gu atemala, Honduras, and Belize is needed to reduce poaching in the Gulf of Honduras. Sinc e manatees were over-harvested in Guatemala (Janson 1980), poaching has increased in southe rn Belize rivers by Guatemalan hunters (Bonde & Potter 1995; McCarthy 1986; Wright et al. 1959). In 1995, investigations documented a large amount of hunting in Port Honduras and identified 11 separate butchering sites where the meat is most likely transported to Guatemala to be sold (Bonde & Potter 1995) There are similar reports of poaching along the Mexi can border, even the area is pr otected by both countries. To assist conservation, DNA analysis of the suspected meat could identify the country of origin and illegal cross-border transport. A dditional actions in Belize are needed to limit this activity. To avoid further loss of genetic diversity in Be lize manatees it is essential to preserve and increase the genetic variation a nd number of individuals. An in crease in the population size may also assist in the repopulation of other countries. Quantifyi ng the degree of migration and diversity using genetic tools is recommended to clarify the role Belize plays in maintaining the subspecies in the Wider Caribbean. The foresight of the Belize government and the actions by Belize residents and organizations has allowed the Belize manatee popul ation to become the largest in the Wider 93


Caribbean, potentially expatriating to other coun tries. However, the population is increasingly threatened by human activities. Proper prot ections and continued m onitoring are needed to ensure the sustainability and expansion of the Belize Antillean manatee population. 94


Table 3-1. Characteristics of the 16 microsat ellite loci implemented on Belize manatee ( T. m. manatus) samples Locus name Tm BSA A NE PIC HO HE TmaA02 56 2 1.55 0.54 0.308 0.355 TmaE1 55 + 4 2.631 1.123 0.608 0.62 TmaE02 58 2 1.424 0.474 0.264 0.298 TmaE7 56 + 4 1.471 0.623 0.298 0.32 TmaE08 60 4 2.967 1.123 0.705 0.663 TmaE11 58 5 4.087 1.482 0.76 0.755 TmaE14 56 + 4 2.296 1.022 0.554 0.564 TmaE26 58 5 1.536 0.648 0.377 0.349 TmaF14 58 3 1.679 0.613 0.393 0.404 TmaH13 60 + 2 1.541 0.536 0.355 0.351 TmaJ02 62 2 1.93 0.675 0.512 0.482 TmaK01 58 2 1.637 0.578 0.529 0.389 TmaKb60 62 6 2.061 0.899 0.407 0.515 TmaM79 54 + 3 2.449 0.971 0.631 0.592 TmaSC5 60 4 1.611 0.708 0.413 0.379 TmaSC13 56 3 1.312 0.429 0.157 0.238 Optimized annealing temperature ( Tm ), BSA requirement (0.4 mg/mL), number of alleles ( A ), effective number of alleles ( NE), polymorphic information content ( PIC ), and the observed and expected heterozygosity (HO and HE) for the Belize T. manatus manatus population Table 3-2. Pairwise FST (above diagonal) and RST values (below diagonal) generated from a survey of 16 microsatellite loci in T. m. manatus in three geographic regions in Belize. Statistically signif icant values are in italics Geographic region NL SL BCC NL 0 0.042 SL 0 0.026 BCC 0.021 0.033 95


Figure 3-1. Geographic map of Beli ze with the three study sites inse t, A) Belize City Cayes, B) Northern (blue) and Southern (green) Lagoon, and C) Placencia Lagoon 96


Figure 3-2. Summ ary plot of q estimates generated by the seque ntial cluster analysis of the program STRUCTURE performed on the Belize and Florida T. manatus genotypes Figure 3-3. Summ ary plot of q estimates performed on the q-sorted Florida group and Belize T. manatus genotypes, indicating two Belize clusters 97


Figure 3-4. Belize (green) and Florida individual un-rooted ne ighbor-joining tree, depicting a genetic separation between the populations (blue bar) 98


Figure 3-5. Belize and Florida un -rooted neighbor-joining tree, de picting a separation of the NL from BCC and SL (100% support) and a BCC and SL separation (90% support) 99

PAGE 100

100 CHAPTER 4 CROSS-SPECIES COMPARISON OF AUSTRALIAN DUGONG AND FLORIDA MANATEE MICROSATELLITE LOCI AND THE CHARACTERIZATION OF HIGHLY INFORMATIVE MARKER-PANELS Introduction The Australian dugong ( Dugong dugon) and Florida manatee ( Trichechus manatus latirostris) are species in the Order Sirenia, threat ened by anthropogenic mortality and habitat degradation. Long generation times and small, fragmented populations make this order highly susceptible to human exploitati on. Currently, all extant sire nian species are considered vulnerable to extinction on a globa l scale (International Union fo r Conservation of Nature and Natural Resources (IUCN 2007)). The Florida ma natee, located throughout the waters of the southeastern United States, is listed as federa lly endangered and estimated to have a population of approximately 3300 individuals (USFWS 2007). Because of their slow reproductive rate, annual mortality may exceed the populations ab ility to produce a sufficient number of new recruits (Bossart 1999). Dugongs are found in th e tropical Indian and Western Pacific Oceans, with the greatest population concentration in th e marine waters of northern Australia. A 2001 aerial survey estimated 14,061 2,314 dugongs in the To rres Straight region. The urban coast of Queensland, Australia, sustains a smalle r population of approximately 2200 dugongs (Marsh 2006). Conservation genetics is a useful tool to evaluate and monitor threatened species (Frankham et al. 2002). Detailed information on the pr esent genetic status of threatened populations can assist with the development of comprehensive long-term management and protection plans. Molecular genetic studies can identify breeding populations, track migration of individuals and populations, and assist in modeling adult survival and reproductive rates.

PAGE 101

Reduced genetic diversity in a species can decrease fecundity, compromise the ability to evolve or endure environmental change, and may u ltimately result in extinction (Avise 2004). Microsatellites, short tandem repeats of nuclear DNA, are highly polymorphic codominant markers used to determine the genetic state of small populations (Cerchio et al. 2005; Coltman et al. 2007; Dixon et al. 2007), especially those with limited genetic variation. Typically, identification of microsatellites for each new species demands considerable time, effort, and cost. Cross-species comparisons can identify polymorphic loci, so that fewer speciesspecific markers are needed for robust studies (Chbel et al. 2002; Huang et al. 2005; Maudet et al. 2004; Nguyen et al. 2007). To aid conservation, dugong (Broderick et al. 2007) and Florida manatee (GarciaRodriguez et al. 2000; Pause et al. 2007) microsatellite markers have been developed for population and pedigree analyses. In an effort to increase the numb er of available primers, this study assesses the cross-species transferability and effici ency of the dugong and manatee primers. Using this information, the most effective marker-panels are compiled, in which identification of individual animals is achieved for each population. Materials and Methods The species-specific and cross-species amplif ication and polymorphism of 25 manatee and 32 dugong microsatellites was tested across four study groups; i) manatee samples amplified with manatee primers, ii) dugong samples amp lified with dugong primers, iii) dugong samples amplified with manatee primers and iv) manat ee samples amplified with dugong primers. Of those loci that amplified, the overall usefulness, such as polymorphic information content, probability of identity, and e ffective number of alleles were assessed among 98 dugongs collected from individuals on the northeast coast of Australia and 91 manatees representative of the four current management units identifie d in the state of Florida (USFWS 2007). 101

PAGE 102

A total of 30 dugong and 21 manatee polym orphic microsatellites was examined on dugong and manatee samples. Development and am plification of the loci followed protocols previously described (Broderick et al. 2007; Garcia-Rodriguez et al. 2000; Pause et al. 2007). Additionally, five microsatel lite loci were employed; DduE06, DduD02 DduG06 and DduH13, (Broderick et al. 2008) and TmaH23 used the same conditions described in Pause et al. (2007). Polymerase chain reaction (PCR) parameters used for cross-species amplification were the same as those published, with slight modi fication to annealing temperature, Tm, and MgCl2 concentration. The MgCl2 concentration for the dugong prime r-manatee sample set was 3mM. The MgCl2 concentration for the manatee prime r-dugong sample set was 2mM, except for TmaK01 (1.5mM) and TmaKb60 (3mM). The number of alleles ( A ), observed and expected heterozygosities ( HO and HE) and adherence to Hardy-Weinberg equili brium (HWE) were assessed using ARLEQUIN, version 3.1 (Schneider et al. 2000). The effective number of alleles (NE; Kimura & Crow 1964) was calculated using POPGENE, version 1.32 (Yeh & Boyle 1997). MICROCHECKER, version 2.2.3 (Van Oosterhout et al. 2004) tested for the presence of null a lleles at a 95% confidence interval. Linkage disequilibrium was analyzed by GENEPOP, version 3.2 (Raymond & Rousset 1995) and the polymorphic information cont ent (PIC) was tested using CERVUS (Kalinowski et al. 2007) GENECAP (Wilberg & Dreher 2004) calculate d the probability of identity ( PID), which is the probability of two individuals drawn at random fr om a population will have the same genotype at the loci assessed (Paetkau & Strobeck 1994), an d a related more conservative statistic for calculating P(ID) among siblings ( P(ID)sib; Evett & Weir 1998). As w ild populations consist of both related and unrelated individuals, the actual probability that a pair of individuals in a given population will have the same genotype depends on the degree of relatedness in that population 102

PAGE 103

and therefore lies between the two extremes of P(ID) and P(ID)sib (Waits et al. 2001). Sample size is a critical parameter in sample-resample studi es. The probability that two genotypes match by chance among n samples is approximately 1 (1 -P(ID))n (Evett and Weir 1998, p 243) and is known as the shadow effect (Mills et al. 2000). The shadow effect is exacerbated in large populations because the number of pairwise comparisons increases exponentially with sample size. The most informative dugong and manatee marker s were selected based on PIC scores and NE. Combinations of the highest-ranking loci were analyzed until PID values indicated that all individuals in the population have unique ge notypes. Documented population sizes may be underestimated (Marsh et al. 2004) and intrapopulational breeding through long distance travel is possible (Fertl et al. 2005). Therefore, for the dugong and manatee P(ID) analyses inflated population sizes of 20,000 and 4,500 were used, respectively. For comparison, P(ID) values were also calculated for the most informative speciesspecific primers. The number of primers that achieved similar P(ID) values as the cross-species and sp ecies-specific panels are reported. Results The transferability of dugong and manatee derived microsatellites confirmed the conservation of primer sites in the sirenian ge nome. Twenty-six dugong primers were used with the dugong samples and 17 were used with the manatee samples. Thirteen of those were used on both species. Twelve manatee primers were used with the dugong samples and 18 were used with the manatee samples. Seven of those were used on both species. Of the 25 manatee primers tested, 23 ( 92%) produced PCR product in the dugong. Of those that amplified, 11 (48%) were polymorphi c. Of the 32 dugong microsatellite primers tested, 27 (84%) yielded PCR produc t in the manatee. Of those that amplified, 17 (53%) were polymorphic. The resultant combined primer valu es for each species are reported in Tables 4-1 103

PAGE 104

and 4-2. The mean heterozygosity ( HE), number of alleles ( A ), polymorphic information content (PIC), and effective number of alleles ( NE) are reported for various combinations of dugong and manatee primers (Table 4-3). The most informative dugong and most informative manatee panels achieved a similar PID as the combined panels. The PID for each study group and the most informative marker-panels are reported in Table 4-4. The level of variability at each speciesspecific microsatellite locus in this study was comparable to those formerly reported. Amplification of Manatee Loci in Dugong All 11 polymorphic manatee microsatellite s used with the dugong samples were determined to be in Hardy-Weinberg equilibrium. The number of alleles per locus ranged from 2 to 10. The single locus observed hetero zygosities ranged from 0.071 to 0.734. Linkage disequilibrium was not identified after 55 pairwise comparisons. No evidence of null alleles was observed. The mean proportion of individuals typed was 0.84. The eight manatee primers that amplifie d both dugong and manatee samples displayed more variation in the dugong samples. Manatee primers, TmaE4, TmaE7 and TmaKb60 identified larger A NE, and PIC in the dugong than in the manatee samples (Tables 4-1 and 4-2). TmaKb60 identified considerably more alleles in the dugong than in the manatee samples (7 vs. 3). Amplification of Dugong Loci in Manatee The 17 polymorphic dugong microsatellites us ed with the manatee samples were determined to be in Hardy-Weinberg equilibriu m after a sequential Bonf erroni correction. The number of alleles per locus ranged from 2 to 5 and the single locus obs erved heterozygosities ranged from 0.044 to 0.593. Linkage disequilibrium was not identified after 136 pairwise comparisons. Evidence of null alleles was observed in DduC09 and DduF07. The mean proportion of individuals typed was 0.982. 104

PAGE 105

The 13 dugong primers that amplified both species identified equal or fewer alleles in the manatee than in the dugong (Tables 4-1 and 4-2). The dugong primers, DduE08, DduF06, and DduF07, identified equal A and greater NE than in the dugong samples. A greater PIC was observed for DduE08 and DduF07 in the manatee samples. Most Informative Markers: Dugong and Manatee Primers Combined Dugong samples The 37 polymorphic dugong and manatee loci were sorted by PIC and NE. Overall, the dugong primers were more informative than the manatee primers in the dugong samples (Table 4-3). The top 11 loci produced a P(ID)sib estimate of 5.15 x 10-05, in which unrelated individuals could be identified in a sample of N=1.9 x 104. After 55 pairwise comparisons, linkage disequilibrium was observed between TmaA04 /DduB01, TmaA09/DduB02, and TmaA09/DduE04. When compared to the dugong primers alone, the mean PIC and NE increased by 0.200 and 1.329, respectively. In comparison, it required two additional (N=13) dugongspecific primers to obtain a similar PID estimate ( P(ID)sib 2.93 x 10-05; HW P(ID) 7.68 x 10-12), although less informative results were achieved at the other parameters. The resultant mean values for all of the tested marker-panels were lower than those for the combined primer set. Manatee samples The 35 dugong and manatee loci that amplified manatee samples were sorted by PIC and NE. Overall, the manatee primers were more informative the than dugong primers in the manatee samples tested (Table 4-3). The top 13 loci produced a P(ID)sib estimate of 1.39 x 10-04, in which unrelated individuals could be id entified in a sample of N = 7000. Linkage disequilibrium was not observed after 78 pairwise comparisons. Th e mean PIC and effective number of alleles increased by 0.21 and 0.88, respectively, when compared to the manatee primers alone. In comparison, it required two additional (N=15) ma natee-specific primers to obtain a similar P(ID) 105

PAGE 106

estimate ( P(ID)sib 1.54 x 10-4; HW P(ID) 4.77 x 10-9), although less informative results were achieved at the other parameters. The resultant m ean values for all of the tested marker-panels were lower than those for the combined primer set. Discussion The most informative markers from the manatee and dugong primer sets produced more sensitive panels than the speci es-specific primers alone in bo th the dugong and manatee samples (Table 4-3). These primer-panels incorporat ed the fewest markers (decreasing cost and improving time effectiveness) while identifying individuals within the current population size estimates. When compared to the most inform ative species-specific primers, the combined primer sets produced higher mean heterozygosity, number of alleles, effective alleles, and polymorphic information content, and needed fewer primers to achieve similar PID values. The Hardy-Weinberg disequilibrium observed in the dugong and manatee samples likely resulted from finite population sizes and the increased pot ential for sampling related individuals in our dataset. Cross-Species Amplification Overall, the manatee primers performed better than the dugong primers in the cross-species studies. Therefore, when the dugong and mana tee primer-sets were combined, the dugong samples had the greatest improvement and fe wer loci were needed to obtain a lower PID. Of the 13 dugong primers that amplified in both species, three had a higher NE and two had a higher PIC in the manatee samples as compared to th e dugong samples. Of the nine manatee primers that amplified both species, six performed equa lly or better for all parameters in the dugong samples. The dugong has a larger population sizes and greater habitat dist ribution and diversity than the West Indian manatee. These circumstan ces can lead to an increase in genetic diversity (DiBattista 2007; Garner et al. 2005). 106

PAGE 107

Conservation Implications The characterization of the most informativ e marker-panels for the Australian dugong and Florida manatee greatly enhances the genetic tools for conservation of sirenians around the world. Molecular markers can identify individuals and provide information on life history, disease processes, and population sizes in conservation studies. Pedigr ee analyses can detect successful breeders and assist with predicti ng annual reproductive ra tes (although, markers for pedigree studies should take into account the effect of null allele s). Genetic studies can also assist in identifying intrapopulational differentia tion and genetically uni que groups of animals that would benefit from increased protection. Although the primer-sets presen ted here may not be optimal for all dugong and manatee populations, these panels are valuable during the in itial phase of genetic analyses. For example, recent sightings of dugongs thought to be regionally extinct in Okinawa, Japan have raised concerns for managers. The implementation of highly informative markers could help identify whether the population is unique and isolated or if individuals are traveling from adjacent populations. Microsatellite studi es of the Antillean, Amazonian and West African manatees in Central and South America and West Africa could shed light on the connectivity and relatedness among the populations. Identification and prot ection of source populations could lead to increased expatriation of individu als. Additionally, movement pr obabilities, adult survivorship, and reproductive rates could assist w ith population status modeling (Tringali et al. 2008b). The employment of more robust markers in sireni an population genetics studies should greatly facilitate conservation efforts for the rec overy of dugongs and manatees around the world. 107

PAGE 108

Table 4-1. Characterization of Dugong dugon and Trichechus manatus latirostris primers amplified on D. dugon samples No Locus name Tm (C) A HO HE PIC NE 1 DduB02 58 10 0.802 0.797 0.764 4.82 2 TmaKb60 62 7 0.734 0.79 0.757 4.645 3 DduC05 58 8 0.765 0.788 0.753 4.622 4 DduD08 58 6 0.593 0.789 0.751 4.643 5 DduE04 58 6 0.78 0.772 0.734 4.313 6 TmaA04 54 7 0.733 0.762 0.721 4.124 7 DduB01 58 7 0.737 0.747 0.705 3.899 8 DduG11 58 8 0.691 0.73 0.691 3.65 9 TmaA09 54 10 0.649 0.662 0.629 2.927 10 DduG12 58 9 0.645 0.646 0.613 2.795 11 DduE09 58 6 0.677 0.641 0.596 2.76 12 DduH04 58 5 0.582 0.632 0.584 2.697 13 DduC09 58 6 0.612 0.575 0.54 2.339 14 DduC03 58 4 0.598 0.586 0.528 2.399 15 TmaE7 58 5 0.538 0.607 0.526 2.522 16 DduH09 58 5 0.459 0.551 0.491 2.212 17 TmaE11 54 6 0.532 0.509 0.484 2.026 18 DduA12 58 6 0.531 0.515 0.479 2.05 19 DduH02 58 3 0.602 0.552 0.458 2.216 20 DduE08 58 3 0.531 0.526 0.436 2.097 21 TmaE08 54 3 0.466 0.48 0.43 1.915 22 DduC11 58 5 0.433 0.489 0.429 1.948 23 DduE11 58 3 0.51 0.462 0.414 1.849 24 DduB05 58 4 0.542 0.509 0.407 2.026 25 DduA01 58 4 0.453 0.445 0.387 1.792 26 DduA07 58 3 0.485 0.45 0.38 1.809 27 TmaE26 54 2 0.457 0.499 0.373 1.985 28 DduF07 58 4 0.175 0.411 0.372 1.692 29 DduD11 58 4 0.435 0.392 0.363 1.64 30 TmaK01 56 4 0.433 0.353 0.307 1.54 31 TmaE4 58 4 0.325 0.317 0.293 1.459 32 DduF11 58 4 0.281 0.289 0.274 1.404 33 DduF06 58 2 0.245 0.216 0.192 1.274 34 DduE03 58 2 0.216 0.21 0.187 1.264 35 TmaE14 61 2 0.195 0.198 0.177 1.244 36 DduG10 58 2 0.211 0.19 0.171 1.233 37 TmaA01 54 2 0.071 0.068 0.066 1.073 108

PAGE 109

The number of alleles ( A ), observed and expected heterozygosity ( HO and HE), polymorphic information content ( PIC), and effective number of alleles ( NE) are reported. The top 11 primers achieved a PID estimate for accurate i ndividual identification 109

PAGE 110

Table 4-2. Characterization of Dugong dugon and Trichechus manatus latirostris primers amplified on T. m. latirostris samples No Locus name Tm (C) A HO HE PIC NE 1 TmaSC5 8 0.612 0.771 0.731 4.283 2 TmaE11 8 0.681 0.669 0.625 2.992 3 TmaE7 4 0.512 0.672 0.601 3.014 4 DduE06 58 5 0.556 0.641 0.579 2.761 5 TmaKb60 3 0.648 0.654 0.576 2.864 6 DduE08 58 4 0.593 0.629 0.57 2.669 7 TmaE1 6 0.562 0.604 0.559 2.502 8 TmaE14 5 0.54 0.611 0.531 2.53 9 DduF07 58 4 0.292 0.574 0.482 2.331 10 TmaE08 4 0.527 0.572 0.48 2.319 11 DduE04 58 3 0.527 0.502 0.403 1.998 12 TmaK01 4 0.584 0.458 0.403 1.835 13 TmaE02 3 0.484 0.488 0.383 1.944 14 DduD08 58 4 0.453 0.42 0.377 1.717 15 TmaJ02 4 0.407 0.413 0.367 1.697 16 DduC05 58 3 0.42 0.473 0.365 1.887 17 DduE09 58 2 0.407 0.468 0.357 1.87 18 TmaM79 2 0.56 0.468 0.357 1.87 19 TmaSC13 3 0.433 0.426 0.338 1.734 20 DduA12 58 4 0.433 0.366 0.335 1.571 21 DduC09 58 2 0.113 0.418 0.329 1.709 22 TmaH13 3 0.389 0.357 0.326 1.55 23 DduF06 58 2 0.374 0.401 0.319 1.663 24 TmaF14 3 0.352 0.385 0.318 1.621 25 DduH04 58 3 0.389 0.362 0.317 1.562 26 DduG06 58 2 0.356 0.386 0.31 1.622 27 TmaA02 2 0.352 0.357 0.292 1.551 28 DduA11 58 2 0.308 0.332 0.276 1.493 29 DduF11 58 2 0.286 0.306 0.258 1.436 30 TmaE4 2 0.264 0.262 0.226 1.352 31 TmaH23 55 2 0.129 0.146 0.134 1.169 32 DduA07 58 2 0.121 0.133 0.124 1.153 33 DduD02 58 3 0.132 0.125 0.117 1.141 34 TmaE26 2 0.1 0.115 0.108 1.13 35 DduB05 58 3 0.044 0.043 0.043 1.045 The number of alleles ( A ), observed and expected heterozygosity ( HO and HE), polymorphic information content ( PIC), and effective number of alleles ( NE) are reported. The top 13 primers achieved a PID estimate for accurate i ndividual identification 110

PAGE 111

Table 4-3. Dugong dugon and Trichechus manatus latirostris marker-panel summaries Dugong samples A HO HE PIC NE DP 4.920 0.523 0.548 0.501 2.598 MP 4.727 0.467 0.477 0.433 2.315 AP 4.892 0.506 0.518 0.472 2.457 ID 6.615 0.655 0.677 0.635 3.338 IP 7.636 0.710 0.739 0.701 3.927 Manatee samples A HO HE PIC NE DP 2.940 0.341 0.387 0.327 1.743 MP 3.778 0.452 0.468 0.409 2.109 AP 3.370 0.398 0.429 0.369 1.930 IM 4.133 0.510 0.527 0.459 2.287 IP 4.690 0.548 0.603 0.533 2.619 Dugong primers (DP), manatee primers (MP), all dugong and manatee primers together (AP), the most informative dugong (ID) and manatee (IM) specific primer s, and the most informative primer-panels for the dugong and manatee samples co mbined (IP). Averages are reported for the number of alleles ( A ), observed and exp ected heterozygosity ( HO and HE), polymorphic information content ( PIC), and effective number of alleles ( NE) Table 4-4. Sibling ( P(ID)Sib) and Hardy-Weinberg equilibrium (HW P(ID)) probability of identity values for the four study groups; dugong and manatee primers PCR amplified on dugong and manatee samples Primer set PID Dugong samples Manatee samples Dugong primers Sib 2.70E-04 8.90E-04 HW 4.69E-09 4.93E-07 Manatee primers Sib 7.10E-04 9.03E-05 HW 7.24E-09 1.62E-09 Most informative primer-set Sib 5.15E-05 1.39E-04 HW 1.24E-11 2.75E-09 111

PAGE 112

112 CHAPTER 5 CHROMOSOME PAINTING IN THE MANA TEE STRONGLY SUPPORTS AFROTHERIA AND PAENUNGULATA Introduction Recently the molecular based approaches of super-ordinal grouping of extant eutherians (Afrotheria, Euarchontoglires, La urasiatheria, and Xenarthra) has gained popularity (Murata et al. 2003; Murphy et al. 2001b; Springer et al. 2003). However, one of the four proposed superorders, Afrotheria, is controvers ial because it unites morphologically distinct species of African placentals (golden moles, tenrecs, otter shrews, elephant shrews, aardvarks, hyraxes, elephants, and sirenians). Within Afrotheria, sirenians, elephants, and hyraxes form a clade called Paenungulata. There is little morphological or paleontological evidence that provides support for Afrotheria (Stanhope et al. 1998). A movable snout was hypothe sized as a synapomorphic trait, but this feature is apparently not homologous across different afroth erian lineages (Whidden 2002). More recently, it was proposed that aspects of placentation could provide a synapomorphy for this assemblage (Carter et al. 2006; Carter et al. 2004). Some outstanding issues in higher eutherian phylogenomics include the exact root of the placental tree, the relationships within the super-ordinal clade La urasiatheria (moles, hedgehogs, shrews, bats, cetaceans, ungulates, pangolins, and carnivores), and resolving the tricho tomy of sirenians, elephants, and hyraxes (Murphy et al. 2004). Sirenia and Hyracoidea are the tw o afrotherian orders remaini ng to be investigated with molecular cytogenetic techniques. In this paper, the chromosome painting of the Florida manatee ( Trichechus manatus latirostris ) is reported. These data s hould be a valuable addition to our understanding of afrotherian relationshi ps and the eutherian ancestral karyotype.

PAGE 113

The Florida Manatee The endangered Florida manatee is a subs pecies of the West Indian manatee ( Trichechus manatus ) in the order Sirenia. Sirenians are ofte n considered phylogenetic outliers. Despite similarities in adaptations, habitat, and body shap e, they have no evolutionary relationship with the other orders of marine mammals. Extant si renians are the only herbivorous marine mammals and live in fresh, brackish, or marine hab itats dispersed along tr opical and subtropical environments. Previous Cytogenetic Reports on Manatees Solid stained chromosome studies were comp leted on a limited number of individual manatees, establishing the chromosome number as 2N = 48 for the Florida manatee (White et al. 1976a; White et al. 1977) and 2N = 56 for the Amazonian manatee ( Trichechus inunguis ; Lounghman et al. 1970). Following solid staining, chro mosome-banding procedures allowed for the identification of individua l chromosome regions. Giemsa and trypsin staining, or GTGbanding, was used to create karyotypes a nd ideograms for the Florida manatee (Gray et al. 2002) and the Amazonian manatee (Assis et al. 1988). Comparisons of chromosome painting data prov ide an independent test of the contrasting hypotheses on mammalian evolution and phylogeny. The research presented here clarifies the phylogenetic position of the manatee and tests th e validity of the radical taxonomic assemblage known as Afrotheria. The results are then comp ared to other chromosome painting data in Afrotheria. In light of the findings, the relatio nships within Afrotheria and the alternative organizations of the ancestral eutherian karyotype are assessed. Methods Chromosome preparations of a male Florida manatee (Trichechus manatus latirostris, TMA) were obtained from peripheral blood mono nuclear cells (PBMCs) and primary fibroblast 113

PAGE 114

cartilage cell culture. Cells were cultured in RPMI 1640 (Hyclone) supplemented with 20% fetal bovine serum (FBS), L-glutamine (0.01%) and gentamicin (25 g/ml). PBMCs were incubated in-vivo using phytohemagglutinin (PHA, 0.25 mg/mL) as a mitotic stimulant for 72 to 96 hr at 36C in 5% carbon dioxide, 95% air, and 100% re lative humidity. Routine procedures were used for chromosome preparati ons. We followed the chromosome nomenclature as previously published (Gray et al. 2002) pairing and grouping chromo somes by banding patterns, relative lengths and morphology. Human chromosome paints were obtained as previously described by chromosome flow sorting followed by degenerate oligonuc leotide primed PCR amplification (Stanyon et al. 1999; Telenius et al. 1992). Paints were labeled with either biotin-dUTP, digoxigen-dUTP (both from Roche Applied Science) or Sp ectrum Orange-dUTP (Vysis). Interspecific in-situ hybridizations of Florida manat ee chromosomes with human probes were performed with 300 to 500 ng of each biotin-labeled probe, 10 g of human Cot-1 DNA and 5 g of ssDNA. The mixture was prec ipitated and dissolved in 13 l of hybridization mixture (formamide 50%, dextran sulfate 10%, 2 SSC). Direct labeling with Spectrum Orange followed a Nick Translation protocol (Vysis) using 1 g of each amplified human DNA probe, 0.2 mM Spectrum Orange and 25 g each of human and manatee Cot-1 DNA (Applied Genetics Laboratories, Inc.). The mixture was precipitated and dissolved in 10 l distilled water. Approximately 300 ng of probe from this mixture were dissolved in 10.5 l Hybrizol VII (QBIOgene) and 0.75 g each of human and manatee Cot-1 DNA. The labeled probe mixture was denatured at 80C for 10 min and reannealed at 37C for 90 min before hybridization. Slides were aged at 37C for 30 min followed by dehydration in a room temperature 70, 80, 90, and 100% ethano l series. The DNA wa s denatured in 70% 114

PAGE 115

formamide/2 SSC, at 65C for 90 s, and que nched in an ice-cold ethanol series. Hybridization was carried out in a humidity cham ber at 37C for five days. Post-hybridization washes followed standard procedures at 40C. Biotin detection was performed with avidinconjugated FITC (Vector) for 45 min at 37C. Counterstaining was performed with DAPI (0.8 ng/ l) for 10 min and the slides were mounted with antifade (100 mg p-phenylenediamine in 80 ml glycerine, 20 ml PBS, pH 8). Analyses were performed under a Zeiss Axi ophot 2 or Axioskop fluorescence microscope coupled with a CCD camera (Photometrics), and images were captured with the Smart Capture software (Digital Scientific Inc.). Results Examples of human chromosome paints (H SA) hybridized to manatee (TMA) metaphase chromosomes are shown in Figure 5-1. Synteny was found intact in nine (4, 5, 6, 9, 11, 14, 17, 18, and 20) of the 22 human autosomal and X ch romosomes (Figure 5-2). Two hybridization signals were evident on separa te manatee chromosomes for te n human chromosomes (1, 7, 8, 10, 12, 13, 15, 16, 21, and 22). The human 19 paint hybridized to three TMA chromosomes (2, 12, and 14). Human chromosomes 2 and 3 were highly fragmented in the manatee genome and painted four and five chromosome s, respectively (Table 5-1). Du e to the small signals involved and the quality of the metaphase s, it was more difficult to assi gn the hybridization pattern for these two chromosomes. Human chromosome pa int 12 provided three signals on TMA 7, most likely due to an inversion. Chromosome paints with pericentromeric signals on both arms of the same chromosome were considered as one signa l. Centromere areas on the manatee karyotype were not hybridized. The Y chromosome was the only human probe that failed to provide a signal in the manatee. Alt ogether, the human autosomal ch romosome paints and the X chromosome paint delimited a total of 44 homol ogous segments in the manatee genome. Human 115

PAGE 116

chromosome paints hybridized to 20 (15 unique) segments in the manatee genome: 1/15, 1/19, 2/3 (twice), 3/7 (thrice), 3/ 13, 3/21, 5/21, 7/16, 8/22, 10/12 (twice), 11/20, 12/22 (thrice), 14/15, 16/19, and 18/19. Discussion The painting map of the manatee genome wa s compared with results published on other Afrotheria taxa: aardvark, elephant, elephant shrew, and golden mole (Fronicke et al. 2003; Robinson et al. 2004; Svartman et al. 2004; Yang et al. 2003). An assessment of the associations found in each taxa ar e shown in Table 5-1. All species have eight associations in common (1/19, 3/21, 5/21, 7/16, 10/12, 12/22, 14/15, and 16/19). Five of these associations are considered ancestral to all eutherians by mo st proposals (3/21, 7/16, 12/22 twice, 14/15, and 16/19). It appears that the associations 1/19 an d 5/21 can be used to link afrotherian species (Fronicke et al. 2003; Robinson et al. 2004; Svartman et al. 2004; Svartman et al. 2006). These associations provide cytogenetic s upport, in agreement with molecula r studies, that Afrotheria is a natural clade. New chromosome painting data in Xenarthra (a nteaters, sloths, and armadillos) are also informative towards the ancestral eutherian karyotype. Of the four species studied, Tamandua tetradactyla Choloepus didactylus, C. hoffmanii and Dasypus novemcinctus (Svartman et al. 2006; Yang et al. 2006), only the anteater has a 1/19 association. It is not likely that this association is homologous to Afrotheria, because the anteater has the most highly rearranged karyotype known in Xenarthra (Svartman et al. 2006). The manatee data indicate that the associ ation 10/12/22 is most likely ubiquitous throughout Afrotheria. A comb ination HSA10p/12p/22q and a single HSA10q were found in the aardvark and elephant karyotypes (Fronicke et al. 2003; Yang et al. 2003). An apparently identical association was later found in th e elephant shrew and golden mole (Robinson et al. 116

PAGE 117

2004). The question is whether this association is a third cytogenetic landmark for the Afrotheria clade, or instead should be considered part of the ancestral eutherian karyotype. The entire 10/12/22 association appears to be present in clades I, Afrotheria, and IV, Laurasiatheria, only partially present in clade II, Xenarthra (10/12), and absent in clade III, Euarchontoglires (primates, rabbits, rodents, tree shrews, and flying lemurs). Carnivores have a homologous 10/12/22 association to Afrotheria, as demonstrated by reciprocal chromosome painting (Graphodatsky et al. 2002; Nie et al. 2002). Eulipotyphla (shrews, solenodons, moles, hedgehogs, and Nesophontes ) also have the 10/12/22 association (Yang et al. 2006; Ye et al. 2006). Chromosome painting data in Xenarthra show that a 10/12 association is present in the armadillo ( D. novemcinctus ; Svartman et al. 2006). To date, the 10/ 12 association has been found in three of the four eutherian mammal clades. Yet, there is no reciprocal painting in Xenarthra to prove that the 10/12 association is truly homologous to that found in Afrotheria. Several hypotheses can be developed with different implications if Afrotheria or Xenarthra is considered basal. If Afrotheria is basal, th e occurrence of 10/12/22 in clades I and IV would suggest that this association is part of the ancestral eutherian karyotype with a subsequent, independent loss in clades II and III. The occurrence of the 10/12/22 association in clades I and IV could be considered a phylogenetic link. Al ternatively, the association could have been independently acquired in the two clades. If Xenarthra is basal, this association could have originated in Afrotheria and was then lost in clade III. Association 3/13 was found in the manatee, elep hant and elephant shrew. However, there are no reciprocal painting data between human and manatee or human and elephant shrew. Therefore, it is not possible to confirm that the 3/13 association is homologous (involves the same segments of both chromosomes 3 and 13). In view of the afrotherian molecular data, this 117

PAGE 118

association was independently derived in the Macroscelidae (elephant shrews) and Paenungulata phylogenetic lineages (Murphy et al. 2004). Support for the Tethytheria and Paenungulata Assemblage Before the advent of molecular studies, so me morphologists placed sirenians, elephants, and hyraxes under Ungulata. Elephants and sire nians were grouped together in Tethytheria, while hyraxes were placed in Phenacodonta along with perissodactyls (McKenna 1975). Results in molecular studies are incons istent and fail to resolve th e Paenungulate trifurcation (Murphy et al. 2004) and some data do not supp ort Tethytheria (Amrine-Madsen et al. 2003; Murphy et al. 2001a; Murphy et al. 2001b; Waddell & Shelley 2003). M itochondrial genome analyses do support Tethytheria, but exclude Hyracoidea (Murata et al. 2003). SINE insertion data produced incongruent phylogenetic relationships within Paenungulata, most likely due to a rapid divergence from a highly polymorphic last common ancestor (Liu & Miyamoto 1999). The chromosome mapping data strongly support Tethytheria (Sirenia and Proboscidea) and implies support for the clade P aenungulata (Sirenia, Proboscidea, and Hyracoidea). There appear to be four derived a ssociations linking elephants wi th manatees: 2/3, 3/13, 8/22, and 18/19. HSA 4/8p was not present in the mana tee and may represent a derived trait of Paenungulata. Both publications on the elephant indicate that this associat ion is also lacking (Fronicke et al. 2003; Yang et al. 2003). It is possible that the 4/8 associati on went undetected in our study, as well as in elephants. Although, the widespread occurr ence of the 4/8 association in all mammals, outside of elephants and most pr imates, lends credence to its inclusion in the ancestral eutherian karyotype. It would be useful to test these hypotheses with rock hyrax chromosome painting data. 118

PAGE 119

Branching Order in Afrotheria The branching order within Afrotheria has no t reached a consensus. Some authors have viewed Macroscelidae, the elephant shrews, as th e most basal and early divergent order within Afrotheria (Murphy et al. 2001a; Springer et al. 1999). However, Murphy et al (2001b) placed the triumvirate of sirenians, elephants, and hyraxes (Paenungulata) as basal, verified by additional molecular data (Kullberg et al. 2006; Murata et al. 2003; Springer et al. 2003). It is difficult to determine which order is most basal because sirenians and elephants, like other afrotherian species, have fairly derived karyotypes. According to Robinson et al (2004), associations 2/8, 3/20, and 10/17 link elephant shrews, golden moles/tenrecs and aardvarks. Only the association 2/8 is pr esent in all three. Recently, the association 2/8 wa s also found in anteater ( T. tetradactyla ), sloth ( Choloepus didactylus), and pangolin ( Manis javanica ; Svartman et al. 2006; Yang et al. 2006). Associations 3/20 and 10/ 17 are lacking in golden moles/tenrecs. Murphy et al (2004) proposed that the associations 3/20 and 10/17 were probably lost in golden moles/tenrecs. No reciprocal painting was done in elephant shrews or golde n moles/tenrecs and it is therefore unknown if these associations are actually homologous. There is weak cytogenetic ev idence linking elephant shrews and golden moles/tenrecs. An alternate h ypothesis might be a sister relationship between aardvarks and elephant shrews. Perhaps a ra pid divergence in elephant shrews, golden moles/tenrecs, and aardvarks also resulted in limited phylogenetic signals for these chromosome associations. The Root of the Eutherian Tree Although the super-order assemblies appear we ll established, the most basal position on the eutherian tree has not been determined with certainty (Delsuc et al. 2004; Murphy et al. 2001a; Murphy et al. 2001b). Afrotheria and Xenarthra are the two oldest eutherian clades and 119

PAGE 120

probably emerged from the Southern Hemisphe re in excess of 100 million years ago (Eizirik et al. 2001; Springer et al. 2004). Molecular dating and bioge ography have provided evidence that crown-group Eutheria may have their most recent common ancestry in the Southern Hemisphere (Gondwana; Springer et al. 2004). The other two clades (Lau rasiatheria and Euarchontoglires) can be grouped as Boreoeutheria (Liu et al. 2001). There are currently three hypotheses for the root of the euther ian tree. Most discussions from molecular studies place emphasis on either Afro theria or Xenarthra as the most basal clade (Douady et al. 2002a; Murphy et al. 2001b). A third hypothesis st ates that the ancestral eutherian karyotype is a combination of both clad es. This hypothesis cannot be completely ruled out and is preferred in some studies (Douady et al. 2002b; Kriegs et al. 2006). However, the suite of derived chromosomal associations found in all studied Afroth eria argues against the hypothesis that a combination of the two clades is basal to the eutherians. Recently, a report on retroelements gives suppor t for the hypothesis that Xenarthra is the sister group to all other placentals (Nishihara et al. 2005). Indeed, new cytogenetic comparisons show that the proposed ancestral eutherian karyotype is essent ially conserved in Xenarthra, specifically in the two-toed sloth ( Choloepus hoffmanii ; Svartman et al. 2004). These two studies should be given attenti on because both take into consider ation rare genomic events in which convergence is particularly limited. The conserved xenarthran karyotype may well be indicative of their phylogenomic position among euther ians. However, an essential point is that all reconstructions of the ancestral eutherian karyotype are preliminary until a relevant outgroup is studied with chromosome painting. A taxonomically rich array of species supported by appropriate out-groups is vita l to the reconstruction of ma mmalian genome evolution. The deficiency of comparative chromosome painting data between eutherians and marsupials is a 120

PAGE 121

severe limitation on attempts to delineate the ma mmalian ancestral genome. The analyses of other afrotherians, xenarthrans, and marsupi als may clarify these unresolved questions. Conclusions The chromosome painting data presented here leave little doubt that Tethytheria is a clade within Afrotheria and implies support for the Paenungulata assemblage. Recent retroposon data also confirmed Paenungulata, but could not resolve the phylogenetic relationships among elephants, sirenians and hyraxes (Liu & Miyamo to 1999). It is generally appreciated that characters with high evoluti onary rates provide good phylogene tic resolution. Afrotherian karyotypes demonstrate high ra tes of chromosome evolution and numerous derived interchromosomal rearrangements link elephants and manatees. It is therefore likely that additional chromosome painting in rock hyraxes could sh ed light on the divergence sequence and resolve the Paenungulata trichotomy. 121

PAGE 122

122Table 5-1. Number of segments homologous to hu man chromosome found in Afrotherian species Species 2n 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 AEK 48 1 2 1 1 1 1 2 2 1 2 1 2 1 1 1 2 1 1 2 1 1 2 manatee 48 2 4 5 1 1 1 2 2 1 2 1 2(4) 2 1 2 2 1 1 3 1 2 2 golden mole 30 1 2 1 1 1 1 2 2 1 2 1 2 1 1 1 2 1 1 2 1 1 2 elephant shrew 26 1 2 4 1 1 1 2 2 1 2 1 2 1 1 1 2(3) 1 1 2(3) 1 1 2 aardvark 20 1 2 1 1 1 1 2 2 1 2 1 2 1 1 1 2 1 1 2 1 1 2 elephant 56 4 4 5 3 1 1 2 2 1 2 3 2 2 1 2 2 1 1 3 1 2 2 The taxa in the first, left column: AEK = ancestral eutherian karyotype [14, 16, 40]. The second column list the 2n, diploid numbers for each species and the remaining columns refer to signals found for each human chromosome. The number in brackets refers to higher number of hybridization signals due to pericentric inversions

PAGE 123

123 Figure 5-1. Examples of hybridizations in th e manatee a) human 12, b) human 13, c) human 14 in green and 15 in red d) human 17 in green and 18 in red

PAGE 124

Figure 5-2. The karyotype of the manatee is shown to the left and the color-coded ideogram to the right (modified from Gray et al. 2002). Manatee chromosomes are numbered below and human chromosome homology is shown laterally 124

PAGE 125

125 CHAPTER 6 CONCLUSION Marine Mammal Population Genetics Genetic techniques have assisted in quantif ying and comparing the genetic diversity and dispersal ability of marine mammals. Some marine mammals have global distribution, like sperm whales, while others are more restricted, like the Puerto Rico manatee. The extent and pattern of dispersal has a large effect on population structure. Hi gh mobility and less restrictive aquatic boundaries are conducive for large, panm ictic marine mammal po pulations. However, most marine mammal species deve lop fine-scale population structur e due to social relationships and resource availability (Hoelz el 1998). Multiple genetic stoc ks in the same geographic area, especially with high human activity, like fisherie s, can lead to difficulties in monitoring and conserving specific lineages. Th is complicates the identifica tion of management units and temporal, spatial, and genetic factors must be taken into account for the conservation of a population. Many marine mammal species have similar demographic characte ristics, including generally large body size, hydrodynamic body shap es, modified appendages and various thermoregulatory adaptations. Additionally, sirenians and cetaceans have long life spans and generation times, producing limited offspring in their lifetime and investing a great deal of time and energy into their young. Marine mammals ha ve different mechanisms for reproduction and parturition. Breeding sites and suitable habi tat are limiting factors that affect population structure. Many species travel to one location to breed and then disperse gr eat distances to feed. This may restrict the gene flow, as only certain groups breed together. Resources, predation, and thermal factors affect reproduc tive population boundaries.

PAGE 126

Environmental and anthropogenic factors str ongly affect manatee and other marine mammal population structure and often limit genetic diversity. For example, direct harvesting and habitat destruction are anth ropogenic disturbances that often result in substantial reductions in population size and loss of genetic variability (genetic bot tlenecks). The Caribbean monk seal ( Monachos tropicalis ), the Japanese sea lion ( Zalophus japonicus), and Stellers sea cow ( Hydrodamalis gigas ) were hunted to extinction and many populations remain dangerously small, like the Northern Atlantic right whale ( Eubalaena glacialis ; Hoelzel et al. 2002). Additionally, Pleistocene glacial events may have limited marine mammal habitat and affected the population structure, as seen in the hooded s eal, harbor porpoise, an d manatee populations (Coltman et al. 2007; Garcia-Rodriguez et al. 1998; Tolley et al. 2001; Vianna et al. 2006). The following discussion will examine and compare the findings of marine mammal genetic investigations with the manatee results discussed previously. To understand better genetic diversity and populati on structure in the aquatic e nvironment, the mechanism of population decline, severity, r ecovery, and current genetic state of other marine mammal populations will be discussed and compared to the Belize, Florida, and Puerto Rico manatee populations. The average mammalian mitochondrial (mtDNA) haplotype diversity is h 0.5 and nucleotide diversity, = 0.001-0.020, depending on the loci a nd species surveyed (Frankham et al. 2002). The Florida manatee population has no m itochondrial diversity and Puerto Rico has low diversity ( h = 0.48500, = 0.00132). Belize has more variation ( h = 0.53430, = 0.030602), due to a high number of nucleotide differences among haplotypes. Disturbed populations (harvested or fragmented) were dete rmined to have lower microsatellite genetic diversity (HE = 0.60; A = 6.17-6.59) than non-disturbed populations (DiBattista 2007). The 126

PAGE 127

manatee populations surveyed here have lower expected heterozygosity and allelic diversity, respectively, than the average disturbed mammalian population: Florida (0.48, 5.3), Belize (0.455, 3.4), and Puerto Rico (0.447, 3.9). The Fl orida and Belize populati ons appear to have high gene flow and little genetic separation amo ng geographic regions. Meanwhile, Puerto Rico has strong mitochondrial DNA structure and more nuclear subdivision than the other populations. Within Population Variation Population Bottlenecks Long-term hunting pressures have severely a ffected manatees and other marine mammal species. For example, the polygynous northern elephant seal populati on was reduced to approximately 10-30 individuals from 1810-1860 (Hoelzel et al. 1993). In 1922, elephant seals received protection and have recovered from the bottleneck to 150,000 individuals (Stewart et al. 1994). Historical diversity was determined to be extremely high in 179bp of the mtDNA control region. In five pre-bottleneck bone samp les, four haplotypes were found with high h (0.9) and ( 0.0065). The high variation may be an artifact due to the representation of multiple generations in the analysis. In 149 contemporary samples, one population with low genetic diversity was indicated. In 300bp, only two divergent haplotypes were found, h = 0.41 and = 0.0066 (Weber et al. 2002). Consistent with the severity of the population bottleneck, the loss of haplotypes in the northern elephant seal is substantial. Although the population rapidly increased, a severe bottleneck, natural climatic cycles, or multiple population crashes as a consequence of persistent harvesting by native peoples before modern exploitation in 1810 may have limited the genetic diversity. Additionally, genetic diversity in highly polygynous sp ecies is strongly affected by a limited number of reproducing 127

PAGE 128

males and respectively the decr eased heterozygosity. The eff ective population size is lowered when all offspring from a harem are related. Manatees are not polygynous, but severe bottle necks, persistent e xploitation, and a long generation time have reduced the diversity to the levels observed in the elephant seals. The Florida population had lower and the Puerto Ri co populations had equivalent diversity. Once reduced, a long generation time can limit the sp eed with which population size, and more importantly, genetic diversity can recover. Th e manatee studies inves tigated an additional 110bp of the same marker, which is a significant incr ease when analyzing sequence data, giving strong support to the results of the manatee studies. In an analogous example, Hawaiian monk seals ( Monashus schauinslandi ) were hunted to near extinction in the 1800s. An estimated popul ation of 50 animals surv ived the bottleneck. Unfortunately, after a partial recovery to 3,000 individuals, a second decline occurred from 1950-1970, reducing the population by 50%. The first 303bp of the control region and 56bp of the proline tRNA gene were amplified in the mitochondrial genome. The combined 359bp iden tified three haplotypes in 50 individuals ( = 0.0006), with 86% of the individuals having one haplotype (Kretzmann et al. 1997). This indicates low variation and a hom ogenous population due to no genetic subdivision and a severe reduction of individuals and ge netic diversity. A slow recovery and additional population declines reduced the diversity further. Another Hawaiian monk seal study analyzed 27 microsatellite loci developed in related species (Kretzmann et al. 2001). Cross-species microsatellite application usually leads to less detected polymorphism than studies with speciesspecific primers. Only three of the primers were polymorphic, each having two alleles, H = 0.396. Reflectiv e of the mtDNA data, the five 128

PAGE 129

Hawaiian Islands populations indicated little to no nuclear genetic st ructure. Nevertheless, three primers are not typically statistically significant to obtain accura te genetic diversity values. The Belize, Florida, and Puerto Rico manat ee populations also have reduced variation at 15 loci. This may be due to climatic or an thropogenic bottlenecks or is intrinsic to the populations. Both the Florida and monk seal stud ies looked at n = 50 individuals. However, only one haplotype was detected in Florida and the seals had th ree. Belize and Puerto Rico manatees have three haplotypes, but mo re individuals were investigated ( 115), increasing the likelihood of identifying an additi onal haplotype. Bones from historical pre-bottleneck manatee specimens may contain additional haplotypes and should be investigated. Gene Flow in Island and Mainland Populations Molecular markers provide information on th e amount and extent of gene flow among populations and habitats. The Be lize and Florida manatees live in mainland habitat and have less genetic geographical structure than the manatees in the island hab itat of Puerto Rico. The island has areas of very deep water close to shor e, producing poor manatee habitat, and potential barriers to movement. Similarly, molecular an alyses of mainland a nd island populations of southern elephant seals identified strong differences in the genetic diversit y between the adjacent mainland and island habitats (Hoelzel et al. 2001). Mainland seal nucleotide diversity was low (0.003), suggesting a female founding event or recent bottleneck. One mainland maternal lineage was indicated, since two of the haplotypes are derivable from the third by a singl e base-pair change. Alternativ ely, nucleotide diversity was high in the 24 island haplotypes (0.023), possi bly due to extensive immigration from the surrounding network of islands. The island and mainland haplotypes were not derivable from each other and therefore maximum likelihood and neighbor-joining trees indicated no female migration between the populations. 129

PAGE 130

However, little genetic separation wa s seen between the two habitats (FST = 0.025) with four microsatellite loci. Ma inland and island heterozygosity was 0.5750 and 0.6241 while allelic diversity was 5.1 and 5.4, respectively. The sim ilar microsatellite he terozygosity and low genetic diversity value (FST) represents frequent male migr ation and breeding, homogenizing the nuclear DNA. Alternatively, no female migration between the two popul ations was indicated, since the mtDNA haplotypes were unrelated. The use of four loci lack st atistical power and may have biased the result. The two habitats also indicated morphologi cal differentiation. The mainland pups are larger, possibly due to more abundant resources, while the island animals have a significantly larger number of vibrissae (perhaps due to different foraging habitat). The non-related mainland and island haplotype s indicated that unrel ated females founded the populations. However, a low FST value suggests a contemporary relationship between the populations through male migration. In contrast, the Florida manatee population shares the A01 haplotype with the Puerto Rico island population, but the FST between the populations is high, suggesting strong differentiation and little contemporary migration. The mainland and island habitats produce similar genetic signals in the seals and manatees. The Puerto Rico island habitat has more mtDNA diversity than the closest mainland population, Florida. The smaller nucleotide diversity in the Florida mainland environment may be due to a founding event, limited supplementation from surrounding populations, or human persecution exterminating haplotypes. Additionally, Puerto Rico may have more diversity from Caribbean and South America immigration and/or geograp hic barriers within the island, allowing divergence over time. Belize and Florida show less population structure than the island habitat of Puerto Rico, possibly due to uninterrupted ha bitat and large dispersal capabilities. Florida 130

PAGE 131

manatees follow the seals mainland trend of having larger body sizes, although in the manatee this is due to adaptations for thermo-regulation. Among Population Variation and Gene Flow Molecular markers can test relatedness, gene flow, and population structure among geographic regions to understand better population differentiation. Harvesting, habitat selection, fragmentation, colonization, and/or genetic drif t can generate intraspecific structure among populations. Moderate Variation and Gene Flow Florida and Puerto Rico manatees are mana ged together as one population. Although the same haplotype (A01) is found in both populations, its presence does not support a contemporary genetic relationship. Therefore, the degree of nuclea r relatedness was tested to determine whether they are the same population or should be managed independently. Furthermore, the mtDNA and microsatellite results were compared to address male and female movement patterns. In a similar study, thr ee Steller sea lion ( Eumetopias jubatus ) populations were tested for migration and relatedness (Hoffman et al. 2006). An mtDNA study of 238 bp, 145 haplotypes in 1,500 individuals, indicated separa te eastern, western, and Asian groups in the North Pacific (Baker et al. 2005). Strong barriers to gene flow we re observed among the three stocks and diversity values within the individual rookeries were high ( h = 0.9164; = 0.00967). The high variation is consistent with ot her marine mammals that have not encountered severe bottlenecks from exploitation. A subsequent nuclear DNA study detected two homogenous populations as opposed to three, combining the western and Asian groups Over 700 animals were analyzed at 13 microsatellite loci (Hoffman et al., 2006). The average number of alleles was 7.9 with a range of 131

PAGE 132

4-11. Expected heterozygosity ranged from 0.237-0.843. Slatkins linearized FST (Slatkin 1995) genetic distance matrixes and trees identified two stocks with weak intraspecific structure. Cavalli-Sforza and Edwards chord distance, Dc (Cavalli-Sforza & Edwards 1967), generated a second genetic distance matrix and tree again id entifying two distinct cl ades of eastern and western stocks with short branch lengths and low levels of differentiation. The resultant phylogenetic tree topography from the two distance matrixes agreed, indicating robust analyses, although bootstrap significance values were not reported. Populations with close geographic proximity clustered together. A Mantel test examined associati ons between Slatkins linearized FST and straight-line geographic distance among rooke ries. An isolationby-distance pattern was found when the stocks were analyzed together. Isolation-by-distance o ccurs in subdivided populations via random genetic drift when subpopulat ions exchange genes at a rate dependent upon the geographical distance. No correlation between genetic and geographic distance was found when stocks were analyzed individually, indicat ing little subdivisi on within stocks. A Bayesian cluster analysis using STRUCTURE 2.1 (Pritchard et al. 2000) tested genetic relationships without a priori designation of geographical local ity. Again two weakly separated clusters were identified, similar to the tree groupings, but with some individuals not grouping with their natal areas. In summary, the mtDNA data specified three stocks, strong female sight fidelity, and a significant east-west split. The east-west separa tion is also found in harbor seals (Westlake & O'Corry-Crowe 2002) and sea otters (Cronin et al. 1996) and possibly represents a historical structure imposed by the Pleisto cene glacier. Conversely, the micr osatellite data indicated two populations and less east-wes t differentiation. The STRUCTURE plot and neighbor-joining trees showed minimum differentiation between the tw o clusters, suggesting genetic mixing. The 132

PAGE 133

combined results from the mtDNA and microsatel lite markers indicate ma le dispersal and strong philopatry in females. Females may stay with in one of the three populations while the males move considerably between the western and Asia n regions and moderately between the east and west. Only large tables with pair-wise comparisons among every population were presented. Reporting the averaged FST and RST values between the stocks woul d have assisted in quantifying the genetic differentiation. Steller sea lions inhabit the North Pacific Rim with a nearly consistent distribution. The historical population was estimated at a quarter of a million animals. The contemporary population has decreased to a little over 100,000, with the smallest population around 13,000 (Calkins et al. 1999). Although the decline is dramatic, these numbers are much larger than other populations of endangered mammals. The potential for a large effective population size and intact genetic variation is high. All sea lion genetic diversity estimates were higher than those for manatees. Sea lions have large populations, continuous habitat, and a high reproduc tive rate, while West Indian manatees have small, patchy, and often isolated populations. Belize and Florida are the largest West Indian manatee populations, numberi ng approximately 1,000 and 3,000, respectively. Small and erratically spaced populat ions lead to fewer effective breeders, less genetic variation, and limited dispersal and supplem entation capabilities. The nucle ar division identified between the Florida and Puerto Rico manatees was gr eater and more conclusive than the sea lion population division. Both mtDNA and nuclear markers support a ge ographical population division within the sea lion and Puerto Rico populations. Additionally, the sea lion and Puerto Rico manatee populations have similar male/female movement patte rns. Females express strong site fidelity as 133

PAGE 134

indicated by the mtDNA divisions, while the nuclear DNA indicates less of a population separation, suggesting male dispersal among the populations. Because Florida has no mtDNA diversity, the nuclear and mtDNA comparison cannot be made. Microsatellite FST and neighborjoining trees identified high gene flow and low ge netic differentiation within the Florida, Belize, and Steller sea lion populations. This is most lik ely due to the continuou s habitat and lack of barriers to dispersal. Panmictic Populations Molecular markers assist in evaluating gene flow among su b-populations or geographical regions within a population. For example, microsatellites detected high amounts of geneflow in the Florida manatee population. Similarly, the worl ds hooded seal populations appear to breed in a panmictic manner, although, the two populations and four breeding herds are geographically separated by large distances (Coltman et al. 2007). Over 900 bp were sequenced containing cyt b, tRNA, and the control region. A total of 105 haplotypes was observed in 123 individuals, with only 12 identified more than once. Th e variation was high, w ith haplotype diversity approaching 1.0 in all populati ons and nucleotide diversity be ing 0.023. The authors reported that = 0.023 was low, although the same value was c onsidered high in southern elephant seals and values of 0.00967 and 0.00660 were reported as lo w in Steller sea lion and northern elephant seal studies, respectively (Baker et al. 2005; Hoelzel et al. 2001; Hoelzel et al. 1993). The authors may have compared to an unidentified population or used a different scale of relative values. The high values suggest rapid growth allowing the development of many haplotypes. The mtDNA haplotype tree indicated limited geographical structur ing and an even diffusion of haplotypes. In the same study, microsatellite markers i nvestigated 300 hooded s eals at 13 polymorphic loci. The heterozygosity ranged from 0.29-0.91 (Have = 0.68) and the number of alleles ranged 134

PAGE 135

from 5-15 (Aave = 12). A high level of genetic vari ation was observed in both pooled and individual subpopulations. Pair-wise comparisons of genetic differentiation (FST) were at most 0.0009, representing little to no differentiation. A Bayesian cluster analysis using STRUCTURE was performed and only one populati on was identified. Although considerable genetic variation was detected with the nuclear and mtDNA markers, all of the variation wa s within, rather than among the subpopulations. The limited diversity among the herds may be due to recent re-colonization from one population after the last glacial period. A study of harbor porpoises ( Phocoena phocoena ) also showed a genetic and geographic re-colonization pattern that may have been influenced by the last glacial epoch (Tolley et al. 2001). The high heterozygosity is expected for species that breed on packed ice, as the unstable habitat does not facilitate natal site-fid elity or highly polygynous mating systems. Each hooded seal population was recommended for protection to conserve the variation and allow for the development of genetic differences. The study lacked analyses that would have in creased the robustness of the results. A neighbor-joining tree indicating one population would have strength ened the hypothesis of little geographical structuring. A dditionally, the results from STRUCTURE could have been improved with the addition of genotypes from a divergent population for comparison, as was done with the manatee analyses. If the data we re available, a regression of FST and straight-line geographic distance could have detected differentia tion in the more distant populations. The results of the hooded seal study corroborate th e structure detected in Florida manatees. Florida manatees also move throughout their range, encouraging a high degree of breeding among the members of the population and limiti ng diversity. Similarly, only one Florida manatee population was identified by STRUCTURE. The Florida population al so had a low global 135

PAGE 136

FST value (0.00735). The manatee FST is not as low as the seals (0.0009) because some east-west manatee substructure is detected in the winter. The high seal he terozygosity indicates a large and genetically stable population while, the redu ced manatee diversity may reflect the limited population sizes. Conclusions Overall, marine mammal populations have low diversity and limited population structure, influenced by bottlenecks, habitat degradati on, founder effects, glacia l events, anthropogenic mortality, and specialized mate se lection and breeding strategies. Florida, Belize, and Puerto Rico manatees have lower mtDNA and microsat ellite genetic diversity than many marine mammal populations. West Indian manatee populations have endur ed historical and long-term persecution. Most of the populations are small, erratically distri buted, and isolated. Manatees must remain close to the shore for food and fres hwater and few individua ls have been sighted traveling in pelagic waters, limiting dispersal and gene flow. Additi onally, manatees lack specific breeding and calving grounds, unlike other marine mammal populations, which could bring diverse individuals together, although in Florida long dist ance dispersal and breeding are observed. The high degree of gene flow a nd migration in Florida and Be lize manatees is possibly due to regularly distributed resources and few gene flow barriers in the habitat. Therefore, the population has not been partitioned into str ongly divergent subpopulations. The Florida individuals have only one haplot ype, produced by a bottleneck or founder event, which may have also reduced the founding nuclear diversity. The three Puerto Rico haplotypes are derive d from one another, producing low nucleotide diversity. Perhaps the founding females were se parated by barriers and genetically diverged. Matrilineal site fidelity has remained strong in Puerto Rico, effectively separating the island at 136

PAGE 137

the north-west and south-east regions. Limited fresh water and poor habitat may have created subtle genetic subdivision at mtDNA and microsat ellite markers. More division is seen with mtDNA than microsatellite markers, suggesting female site fidelity and male dispersal. Similarly, the Belize population has different ha plotype proportions in the Drowned Cayes and Southern Lagoon. Many marine mammal populations have experi enced strong harvesting pressure and other anthropogenic threats. Sirenians are still hunted throughout the world and recently other factors, such as water vessel strikes and habitat destru ction, have increased mo rtality. The serious anthropogenic threats experienced by sirenians may be due to overlapping habitat use. Manatees inhabit the coastal waters that are ideal for human colonization. Recent protections have assisted in slowly ameliorating these threats. Howe ver, high mortality has forced the manatee populations to remain small and grow slowl y, potentially stunting genetic diversity. A similarity was seen between the mainland /island populations of elephant seals and manatees. Greater diversity was identified in the island habitats, owing to a greater capacity for divergence and more immigration potential from genetically differentiated populations. As mentioned, many marine mammal populations have been strongly in fluenced by recent glacial epochs. Northern mana tee populations may have been reduced by the Wisconson glacial period. When the glacier receded, manatee habit became available in the northern limits of their range. This allowed a series of founding events, with manatees using th e Caribbean islands as stepping-stones. Alternatively, if manatees were presen t in the northern limits, the population size may have decreased with the cold water and th en expanded as the glacier melted. In either scenario, the founding Florida and Puerto Rico manatee population was most likely small and may have contained limited genetic diversity. 137

PAGE 138

The reported manatee diversity was lo wer than other marine mammal and demographically-challenged populations. Nevertheless, small mana tee populations with reduced diversity continue to thrive and grow. No significant immunological or reproductive barriers have been identified to date that would suggest a lack of fitness. Ho wever, immunological and functional genetic studies could address immunological function and aid in identifying imperiled populations or manatees with reduced fitness. These studies could also assist overall manatee health assessment and improve treatment. Antillean manatee populations must be protecte d from harm and persecution to allow for genetic diversity to increase. The degree of genetic diversity is correlated with a number of individuals and overall health in a population. The genetic diversity within these populations should be monitored to detect changes over time. Furthermore, additional West Indian manatee populations should be genetically analyzed to identify cross-border breeding and populations with reduced variation or risk for inbreeding. The hybrid zone betw een the Amazonian and West Indian manatee ranges should be analyzed with nuclear mark ers to facilitate management of the species. Ultimately, population viability will depend on the quality of habitat and reduction of anthropogenic threats. Cooperation of legislation, edu cation, and enforcement entities are needed to ensure the sustainability of manatee populations for the future. 138

PAGE 139

139 LIST OF REFERENCES Acosta J (1590) Historia natural y moral de las Indias, en que se traten las cosas notables del cielo, y elementos, metales, pl antas y animales de ellas: y los ritos y ceremonias, leyes, govierno y guerras de los Indios Sevilla, Espaa. Alvarez-Alemn A, Powell JA, Beck CA (2007) First report of a Florida manatee documented on the North Coast of Cuba. SireNews (Newsletter of th e IUCN/SSC Sirenia Specialist Group) 47, 9-10. Amos W, Rubinsztein DC (1996) Microsatellit es show mutational bias and heterozygote instability. Nature Genetics 13, 390-391. Amrine-Madsen H, Koepfli KP, Wayne RK, Sp ringer MS (2003) A new phylogenetic marker, apolipoprotein B, provides compelling evidence for eutherian relationships. Molecular Phylogenetics and Evolution 28, 225-240. Anderson PK (1997) Shark bay dugongs in summer. I: Lek mating. Behaviour, 134, 433-462. Arnason U (1974a) Comparative ch romosome studies in cetacea. Hereditas 77, 1-36. Arnason U (1974b) Comparative chro mosome studies in pinnipedia. Hereditas, 76, 179-225. Arnason U, Benirschke K (1973) Karyotypes and ideograms of sperm and pygmy sperm whales. Hereditas 75 67-74. Assis MFL, Best RC, Barros RMS, Yonena gayassuda Y (1988) Cytogenetic study of Trichechus inunguis (Amazonian manatee) Revista Brasileira De Genetica 11, 41-50. Auil NE (1998) Belize Manatee Recovery Plan. UNDP/GEF Coastal Zone Management Project. Caribbean Environment Programme, Kingston, Jamaica. Auil NE (2004) Abundance and distribution trends of the West Indian manatee in the coastal zone of Belize: Implications for conservation Master's Thesis, Texas A&M University, College Station, TX. Auil NE, Powell JA, Bonde RK, Andrewin K, Ga lves J (2007) Belize Conservation Programme 10 year summary to Liz Claiborne Art Orte nberg Foundation. Wildlif e Trust, New York, NY. Avise JC (2004) Molecular Markers, Natural History, and Evolution, Second Edition Sinauer Associates, Inc., Sunderland, MA. Avise JC, Ball Jr. RM (1990) Principles of gen ealogical concordance in species concepts and biological taxonomy. Oxford Survey of Evolutionary Biology 7, 45-67.

PAGE 140

Baker AR, Loughlin TR, Burkanov V et al. (2005) Variatio n of mitochondrial control region sequences of Steller sea lions : The three-stock hypothesis. Journal of Mammalogy 86, 1075-1084. Baker CS, Lento GM, Cipriano F, Palumbi SR (2000) Predicted decline of protected whales based on molecular genetic monitoring of Japanese and Korean markets. Proceedings of the Royal Society of London Se ries B-Biological Sciences 267 1191-1199. Balloux F, Goudet J (2002) Statis tical properties of population differentiation estimators under stepwise mutation in a finite island model. Molecular Ecology 11, 771-783. Bauer GB, Colbert DE, Gaspard JC, Littlefield B, Fellner W (2003) Underw ater visual acuity of Florida manatees ( Trichechus manatus latirostris ). International Journal of Comparative Psychology, 16, 130-142. Beck CA, Barros NB (1991) The impact of debris on the Florida manatee. Marine Pollution Bulletin 22, 508-510. Best RC (1982) Seasonal breeding in the Amazoni an manatee, Trichechus inunguis (Mammalia: Sirenia). Biotropica 14, 76-78. Best RC (1984) The aquatic mammals and reptiles of the Amazon. In: The Amazon. Limnology and Landscape Ecology of a Migh ty Tropical River and its Basin (ed. Sioli H), Netherlands. Bielec PE, Gallagher DS, Womack JE, Busbee DL (1998) Homologies between human and dolphin chromosomes detected by he terologous chromosome painting. Cytogenetics and Cell Genetics 81, 18-25. Bielec PE, Riggs PK, Womack JE, Busbee DL ( 1997) A high resolution GBG-banded karyotype of the Atlantic bottlenose dolphin, Tursiops truncatus: Generation of an ideogram, and NOR localization by fluorescence in situ hybridization. Cytogenetics and Cell Genetics 78, 6-11. Bonde RK, Aguirre AA, Powell JA (2004) Manatees as sentinels of marine ecosystem health: Are they the 2000-pound canaries? EcoHealth 1, 255-262. Bonde RK, Potter CW (1995) Manatee butcheri ng sites in Port Honduras. USGS, Florida Integrated Science Center, Sirenia Project, Gainesville, FL. Bossart GD (1999) The Florida manatee: On the verge of extinction? Journal of the American Veterinary Medical Association 214 1178-1183. Bossart GD, Baden DG, Ewing RY, Roberts B, Wr ight SD (1998) Brevet oxicosis in manatees ( Trichechus manatus latirostris) from the 1996 epizootic: Gross, histologic, and immunohistochemical features. Toxicological Pathology 26, 276. 140

PAGE 141

Bossart GD, Ewing RY, Lowe M et al. (2002) Viral papillomatosis in Florida manatees ( Trichechus manatus latirostris ). Experimental and Molecular Pathology 72, 37-48. Bossart GD, Meisner RA, Rommel SA, Ghim SJ, Je nson AB (2003) Pathological features of the Florida manatee cold stress syndrome. Aquatic Mammals 29, 9-17. Bradley JL, Wright SD, McGuire PM (1993) The Florida manatee cytochrome-b DNA sequence. Marine Mammal Science 9, 197-202. Broderick D, Ovenden J, Slade R, Lanyon JM (2007) Characterization of 26 new microsatellite loci in the dugong ( Dugong dugon). Molecular Ecology Notes 7, 1275-1277. Broderick D, Ovenden J, Slade R, Lanyon JM (2 008) Isolation and ch aracterization of 32 microsatellite loci in the Dugong ( Dugong dugon). Molecular Ecology Notes (Submitted) Buergelt C, Bonde R (1983) Toxopl asmic meningoencephalitis in a West Indian manatee. Journal of the American Vete rinary Medical Association 183 1294-1296. Calkins DG, McAllister DC, Pitcher KW, Pendleton GW (1999) Steller sea lion status and trend in Southeast Alaska: 1979-1997. Marine Mammal Science 15, 462-477. Cantanhede AM, Da Silva VMF, Farias IP et al. (2005) Phylogeography and population genetics of the endangered Amazonian ma natee, Trichechus inunguis Natterer, 1883 (Mammalia, Sirenia). Molecular Ecology 14, 401-413. Carter AM, Blankenship TN, E nders AC, Vogel P (2006) The fe tal membranes of the otter shrews and a synapomorphy for Afrotheria. Placenta 27, 258-268. Carter AM, Enders AC, Kunzle H, Oduor-Okelo D, Vogel P (2004) Placentation in species of phylogenetic importance: The Afrotheria. Animal Reproduction Science 82-83, 35-48. Cavalli-Sforza LL, Edwards AWF (1967) Phylog enetic analysis models and estimation procedures. American Journal of Human Genetics 19, 233-257. Cerchio S, Jacobsen JK, Cholewiak DM, Falc one EA, Merriwether DA ( 2005) Paternity in humpback whales, Megaptera novaeangliae : Assessing polygyny and skew in male reproductive success. Animal Behaviour, 70, 267-277. Chbel F, Broderick D, Idaghdour Y, Korrida A, McCormick P (2002) Characterization of 22 microsatellites loci from th e endangered Houbara bustard ( Chlamydotis undulata undulata ). Molecular Ecology Notes 2, 484-487. 141

PAGE 142

Chowdhary BP, Fronicke L, Gustavsson I, Schertha n H (1996) Comparative analysis of the cattle and human genomes: Detection of ZOO-FISH and gene mapping-based chromosomal homologies. Mammalian Genome 7, 297-302. Chowdhary BP, Raudsepp T (2001) Ch romosome painting in farm, pet and wild animal species. Methods in Cell Science 23, 37-55. Chowdhary BP, Raudsepp T, Fronicke L, Schert han H (1998) Emerging patterns of comparative genome organization in some mammalia n species as revealed by Zoo-FISH. Genome Research 8, 577-589. Clifton KB, Yan J, Mecholsky JJ, Reep RL ( 2008) Material propertie s of manatee rib bone. Journal of Zoology 274, 150-159. Coltman DW, Stenson G, Hammill MO et al. (2007) Panmictic population structure in the hooded seal ( Cystophora cristata ). Molecular Ecology Notes 16, 1639-1648. Crandall KA, Bininda-Emonds ORP, Mace GM, Wa yne RK (2000) Considering evolutionary processes in conservation biology. Trends in Ecology & Evolution 15, 290-295. Cronin MA, Bodkin J, Ballachey B, Estes J, Patton JC (1996) Mitochondrial-DNA variation among subspecies and populations of sea otters ( Enhydra lutris ). Journal of Mammalogy 77, 546-557. De Jong WW, Zweers A (1980) Confirmacao da relacoao entre peizes-bois, hyraxes e elefantes, pro meio do estudo da proteina das lentes dos olhos. Acta Amazonica 10, 897902. de Thoisy B, Spiegelberger T, Rousseau S et al. (2003) Distribution, habitat, and conservation status of the West Indian manatee Tr ichechus manatus in French Guiana. Oryx, 37, 431436. Delsuc F, Vizcaino SF, Douzery EJP (2004) Influe nce of Tertiary paleoenvironmental changes on the diversification of South American mammals: a relaxed molecular clock study within xenarthrans. Bmc Evolutionary Biology 4, 13. Dennis TR, Stock AD (1999) A molecular cytogenetic study of chromosome 3 rearrangements in small cell lung cancer: Consistent in volvement of chromosome band 3q13.2. Cancer Genetics and Cytogenetics 113, 134-140. Deutsch CJ (2000) Winter movements and use of warm-water refugia by radio-tagged West Indian manatees along the Atlantic coast of the United States, p. 131. Florida Power and Light Company and U. S. Geological Survey, Gainesville, FL. 142

PAGE 143

Deutsch CJ, Reid JP, Bonde RK et al. (2003) Seasonal movements, mi gratory behavior, and site fidelity of West Indian manatees along th e Atlantic Coast of the United States. Wildlife Monographs 1-77. DiBattista JD ( 2007) Patterns of genetic varia tion in anthropogenically impacted populations. Conservation Genetics 9, 141-156. Dixon JD, Oli MK, Wooten MC et al. (2007) Genetic consequences of habitat fragmentation and loss: The case of the Florida black bear ( Ursus americanus floridanus ). Conservation Genetics, 8, 455-464. Domning DP (1981a) Distributi on and status of manatees Trichechus spp. near the mouth of the Amazon River, Brazil. Biological Conservation 19, 85-97. Domning DP (1981b) Sea cows and sea grasses. Paleobiology, 7, 417-420. Domning DP (1982a) Commercial exploitation of manatees Trichechus in Brazil c. 1785-1973. Biological Conservation 22, 101-126. Domning DP (1982b) Evolution of ma natees A speculative history. Journal of Paleontology 56, 599-619. Domning DP (1983) Marching teeth of the manatee: Its special ad aptation to an abrasive diet has enabled this aquatic mammal to outdo the dugong. Natural History 92, 8-11. Domning DP (1994) A phylogenetic analysis of the Sirenia. Proceedings of the San Diego Society of Natural History 29, 177-189. Domning DP (2001) The earliest kn own fully quadrupedal sirenian. Nature, 413, 625-627. Domning DP (2005) Fossil Sirenia of the West Atlantic and Caribbean re gion. VII. Pleistocene Trichechus manatus Linnaeus, 1758. Journal of Vertebrate Paleontology 25, 685-701. Domning DP, Hayek LC (1986) Interspecific an d intraspecific morphological variation in manatees (Sirenia: Trichechus ). Marine Mammal Science 2, 87-144. Douady CJ, Catzeflis F, Kao DJ, Springer MS, Stanhope MJ (2002a) Molecular evidence for the monophyly of tenrecidae (mammalia) and the timing of the colonization of Madagascar by Malagasy tenrecs. Molecular Phylogenetics and Evolution 22, 357-363. Douady CJ, Chatelier PI, Madsen O, et al. (2002b) Molecular phylogenetic evidence confirming the Eulipotyphla concept and in support of hedgehogs as the sister group to shrews. Molecular Phylogenetics and Evolution 25, 200-209. Duffield DA, Ridgway SH, Sparkes RS (1967) Cy togenetic studies of 2 species of porpoise. Nature 213, 189-190. 143

PAGE 144

Duignan PJ, House C, Walsh MT et al. (1995) Morbillivirus in fection in manatees. Marine Mammal Science 11, 441-451. Eizirik E, Murphy WJ, O'Brien SJ (2001) Mo lecular dating and bioge ography of the early placental mammal radiation. Journal of Heredity 92, 212-219. Ellegren H (2000) Evolution of the avian sex ch romosomes and their role in sex determination. Trends in Ecology & Evolution 15, 188-192. Eppig JT, Nadeau JH (1995) Comparativ e maps: The mammalian jigsaw puzzle. Current Opinion in Genetics & Development 5, 709-716. Evett IW, Weir BS (1998) Interpreting DNA Evidence: Statistical Genetics for Forensic Scientists Sinauer Associates, Inc., Sunderland, MA, USA. Excoffier L, Laval G, Schneider S (2005) Arlequin ver 3.0: An integrated software package for population genetics data analysis. Evolutionary Bioinformatics Online 1, 47-50. Felsenstein J (2004) PHYLIP (Phylogeny Inferen ce Package) version 3.6, p. Distributed by the author. Department of Genome Sciences, University of Washington, Seattle, WA. Fertl D, Schiro AJ, Regan GT et al. (2005) Manatee occurrence in th e northern Gulf of Mexico, west of Florida. Gulf and Caribbean Research 17 69-94. Florida Fish and Wildlife Research Institute (2007a) Manatee Carcass Salvage Program St. Petersburg, FL, USA. Accessed 26 February 2007. Available from Florida Fish and Wildlife Research Institute (2007b) Manatee Synoptic Surveys St. Petersburg, FL, USA. Accessed 13 February 2007. Available from ures/view_article.asp?id=15246 Frankham R, Ballou JD, Briscoe DA (2002) Introduction to Conservation Genetics Cambridge University Press, Cambridge, UK. Franklin IR (1980) Evolutionary change in small populations. In: Conservation Biology: An Evolutionary-Ecological Perspective (eds. Soule ME, Wilcox BA), p. 135. Sinauer Associations, Sunderland, MA. Fronicke L, MullerNavia J, Romanakis K, Scherthan H (1997) Chromosomal homologies between human, harbor seal ( Phoca vitulina ) and the putative ancestral carnivore karyotype revealed by Zoo-FISH. Chromosoma 106, 108-113. Fronicke L, Wienberg J, Stone G, Adams L, Stanyon R (2003) Towards the delineatio n of the ancestral eutherian genome or ganization: Comparative geno me maps of human and the 144

PAGE 145

African elephant ( Loxodonta africana ) generated by chromosome painting. Proceedings of the Royal Society of London Series B-Biological Sciences 270, 1331-1340. Gann TWF (1911) Exploration carried-out in British Honduras during 1908-1909. Liverpool University Annals of Ar chaeological Anthropology 4, 72-87. Garcia-Rodriguez AI, Bowen BW, Domning DP et al. (1998) Phylogeography of the West Indian manatee ( Trichechus manatus): How many populations and how many taxa? Molecular Ecology 7, 1137-1149. Garcia-Rodriguez AI, Moraga-Amador D, Farmerie WG, McGuire PM, King TL (2000) Isolation and characterizati on of microsatellite DNA marker s in the Florida manatee ( Trichechus manatus latirostris ) and their application in selected sirenian species. Molecular Ecology 9, 2161-2163. Garner A, Rachlow JL, Hicks JF (2005) Patterns of genetic diversity and its loss in mammalian populations. Conservation Biology 19, 1215-1221. George JC, Bada J, Zeh J et al. (1999) Age and growth estimates of bowhead whales (Balaena mysticetus) via aspartic acid racemization. Canadian Journal of Zoology-Revue Canadienne De Zoologie, 77, 571-580. Geraci JR, Harwood J, Lounsbury VJ (1999) Marine mammal die-o ffs: Causes, investigations, and issues. In: Conservation and Management of Marine Mammals (eds. Twiss Jr. JR, Reeves RR). Smithsonian Institution Press, Washington, DC. Gibson J (1995) Managing manatees in Belize Master's Thesis, University of Newcastle upon Tyne, Newcastle, UK. Gillespie JH (1998) Population Genetics-A Concise Guide The Johns Hopkins University Press, Maryland. Glaser K, Reynolds III JE (2003) Mysterious Manatees University Press of Florida, Gainesville, FL. Graphodatsky AS, Yang F, Perelman PL et al. (2002) Comparative molecular cytogenetic studies in the order Carn ivora: Mapping chromosomal rearrangements onto the phylogenetic tree. Cytogenetic and Genome Research 96, 137-145. Gray BA, Zori RT, McGuire PM, Bonde RK (2002) A first generation cyt ogenetic ideogram for the Florida manatee ( Trichechus manatus latirostris ) based on multiple chromosome banding techniques. Hereditas 137, 215-223. Groves CR, Jensen DB, Valutis LL et al. (2002) Planning for biodivers ity conservation: Putting conservation science into practice. Bioscience 52 499-512. 145

PAGE 146

Hameister H, Klett C, Bruch J et al. (1997) Zoo-FISH analysis: The American mink (Mustela vison) closely resembles the cat karyotype. Chromosome Research 5, 5-11. Harper JY, Samuelson DA, Reep RL (2005) Corn eal vascularization in the Florida manatee (Trichechus manatus latirostris) and thr ee-dimensional reconstruction of vessels. Veterinary Ophthalmology 8, 89-99. Hartl DL, Clark AG (1997) Principles of Population Genetics Sinaur Associates Inc., Sunderland, MA. Hartman DS (1974) Distribution, status, and conservation of the manatee in the United States, p. 246. U.S. Fish and Wildlife Service, National Fish and Wildlife Laboratory, Contract No. 14-16-0008-748. Hatt RT (1933) A manatee collected by the American Museum Congo Expedition, with observations on the recent manatees. BULLETIN OF THE AMERICAN MUSEUM OF NATURAL HISTORY 66: 533-U80 Hayes H (1995) Chromosome painting with human chromosome-specific DNA libraries reveals the extent and distribution of conser ved segments in bovine chromosomes. Cytogenetics and Cell Genetics 71, 168-174. Hoelzel AR (1998) Genetic stru cture of cetacean populations in sympatry, parapatry, and mixed assemblages: Implications for conservation policy. Journal of Heredity 89 451-458. Hoelzel AR, Campagna C, Arnbo m T (2001) Genetic and morphometric differentiation between island and mainland southern elephant seal populations. Proceedings of the Royal Society of London Series B-Biological Sciences 268, 325-332. Hoelzel AR, Goldsworthy SD, Fleisher RC (2002) Population genetics. In: Marine Mammal Biology; an Evolutionary Approach (ed. Hoelzel AR), pp. 325-352. Blackwell Science, Oxford. Hoelzel AR, Halley J, Obrien SJ et al. (1993) Elephant seal genetic variation and the use of simulation-models to investigate historical population bottlenecks. Journal of Heredity 84, 443-449. Hoffman JI, Matson CW, Amos W, Loughlin TR, Bickham JW (2006) Deep genetic subdivision within a continuously distribu ted and highly vagile marine mammal, the Steller's sea lion ( Eumetopias jubatus). Molecular Ecology 15, 2821-2832. Huang Y, Tu J, Cheng X et al. (2005) Characterization of 35 novel microsatellite DNA markers from the duck ( Anas platyrhynchos ) genome and cross-amplifi cation in other birds. Genetics Selection Evolution 37, 455-472. 146

PAGE 147

IBAMA (2001) Plano de Ao de Mamferos Aquticos do Brasil. Verso II. Ministrio do Meio Ambiente, Instituto Brasileiro do Meio Ambi ente e dos Recursos Naturais Renovveis, Braslia. IUCN (2007) 2007 IUCN Red List of Threaten ed Species. Accessed 19 February 2008. Available from Janson T (1980) Discovering the mermaids. Oryx 15, 374. Jin L, Nei M (1990) Limitations of the evoluti onary parsimony method of phylogenetic analysis. Molecular Biology and Evolution 7, 82-102. Kalinowski S, Taper M, Marshall T (2007) Re vising how the computer program CERVUS accommodates genotyping error increases success in paternity assignment. Molecular Ecology Notes, 16, 1099-1006. Kellogg ME, Burkett S, Dennis TR et al. (2007) Chromosome painting in the manatee supports Afrotheria and Paenungulata. BMC Evolutionary Biology 7. Kellogg ME, Pause KC, Bonde RK et al. (2008) Comprehensive genetic investigation recognizes evolutionary divergence in the Florida ( Trichechus manatus latirostris ) and Puerto Rico ( T. m. manatus ) manatee populations and subtle substructure in Puerto Rico. In Prep Kimura M (1980) A simple method for estimating e volutionary rate of base substitution through comparative studies of nucleotide sequences. Journal of Molecular Evolution 16, 111120. Kimura M, Crow JF (1964) The number of alleles that can be maintained in a finite population. Genetics 49, 725-738. King TL, Eackles MS, Young CC (2006) Micr osatellite DNA markers for assessing phylogeographic and populati on strucutre in Preble's meadow jumping mice ( Zapus hudsonius preblei ) and cross-amplification among neighbouring taxa. Molecular Ecology Notes, 6, 670-673. Kipps EK, McLellan WA, Rommel SA, Pabst DA (2002) Skin density and its influence on buoyancy in the manatee ( Trichechus manatus latirostris ), harbor porpoise ( Phocoena phocoena), and bottlenose dolphin (Tursiops truncatus ). Marine Mammal Science 18 765-778. Knoll S, Rowell-Rahier M (1998) Distribution of genetic variance and isolation by distance in two leaf beetle species: Oreina cacaliae and Oreina speciosissima Heredity 81, 412421. 147

PAGE 148

Kretzmann MB, Gemmell NJ, Meyer A (2001) Microsatelli te analysis of population structure in the endangered Hawaiian monk seal. Conservation Biology 15, 457-466. Kretzmann MB, Gilmartin WG, Meyer A et al. (1997) Low genetic variability in the Hawaiian monk seal. Conservation Biology 11 482-490. Kriegs JO, Churakov G, Kiefmann M, et al. (2006) Retroposed elements as archives for the evolutionary history of placental mammals. Plos Biology, 4, 537-544. Kullberg M, Nilsson MA, Arnason U, Harley EH, Janke A (2006) Housekeeping genes for phylogenetic analysis of eutherian relationships. Molecular Biology and Evolution 23, 1493-1503. Lande R (1995) Mutation and conservation. Conservation Biology 9, 782-791. Lefebvre LW, Marmontel M, Reid JP, Ra thbun GB, Domning DP (2001) Status and biogeography of the West Indian manatee. In: Biogeography of the We st Indies: Patterns and Perspectives, Second Edition (eds. Woods CA, Sergile FE), pp. 425-474. CRC Press, Boca Raton, FL. Lima RP (1997) Peixe-Boi Marinho (Trichechus manatus): Distribuio, status de conservao e aspectos tradicionais ao l ongo do litoral nordeste do Brasil Master's Thesis, Universidade Federal de Pernambuco, Brazil. Liu FGR, Miyamoto MM (1999) Phylogenetic asse ssment of molecular and morphological data for eutherian mammals. Systematic Biology 48, 54-64. Liu FGR, Miyamoto MM, Freire NP, et al. (2001) Molecular and mo rphological supertrees for eutherien (placental) mammals. Science, 291, 1786-1789. Lounghman WD, Frederic FL, Herald ES ( 1970) The chromosome of a male manatee ( Trichechus inunguis ). International Zoo Yearbook 10, 151-152. Luna FO (2001) Distribuio, Status de Conservao e Aspectos Tradicionais do peixe-boi marinho (Trichechus manatus manatus) no litoral norte do Brasil Master's Thesis, Universidade Federal de Pernambuco, Brazil. Manni F, Guerard E, Heyer E (2004) Geographic patterns of (genetic, morphologic, linguistic) variation: How barriers can be dete cted by using Monmonier's algorithm. Human Biology, 76, 173-190. Marmontel M (1995) Age and reproducti on in female Florida manatees. In: Population Biology of the Florida Manatee (eds. OShea TJ, Ackerman BB, Percival HF), pp. 98. National Biological Service Information and Technology Report 1, Washington, D.C. 148

PAGE 149

Marmontel M, Humphrey SR, O'Shea TJ (1997) Population viability analysis of the Florida manatee ( Trichechus manatus latirostris ), 1976-1991. Conservation Biology, 11, 467481. Marsh H (2006) Dugong dugon. In: 2007 IUCN Red List of Threatened Species Accessed 19 February 2008. Available from Marsh H, Lawler I, Kwan D et al. (2004) Dugong distribution and abundance in Torres Strait. Australian Fisheries Management Authority Report #R 01/0895 45. Marshall CD, Reep RL (1995) Manatee cerebral co rtex: Cytoarchitecture of the caudal region in Trichechus manatus latirostris Brain, Behavior and Evolution 45, 1-18. Marshall TC, Slate J, Kruuk LEB, Pemberton JM (1998) Statistical confidence for likelihoodbased paternity inference in natural populations. Molecular Ecology 7, 639-655. Maudet C, Beja-Pereira A, Zeyl E et al. (2004) A standard set of pol ymorphic microsatellites for threatened mountain ungulates ( Caprini, Artiodactyla ). Molecular Ecology Notes 4, 4955. McCarthy TJ (1986) The gentle giants of Be lize. Part II: Dist ribution of manatees. Belize Audubon Society Bulletin 18, 1-4. McClenaghan LR, O'Shea TJ (1988) Genetic variability in the Florida manatee ( Trichechus manatus ). Journal of Mammalogy 69, 481-488. McDonald B (2005) Population genetics of dugongs around Aust ralia: Implications of gene flow and migration Doctorial Dissertation, James Cook University, Townsville, Australia. McField M, Wells S, Gibson J (1996) State of the Coastal Zone report. Coastal Zone Management Programme. Government of Belize, Belize City, Belize. McKenna MC (1975) Toward a phylogenetic classification of the Mammalia. In: Phylogeny of the Primates (eds. Luckett WP, Szalay FS), pp. 21-46. Plenum, New York. Medway W, Bruss ML, Bengtson JL, Black DJ (1982) Blood chemistry of the West Indian manatee ( Trichechus manatus). Journal of Wildlife Diseases 18, 229-234. Mignucci-Giannoni AA (1989) Zoogeography of marine mammals in Puerto Rico and the Virgin Islands Master's Thesis, University of Rhode Island, Kingston, RI. Mignucci-Giannoni AA (1996) Marine mammal strandings in Puer to Rico and the United States and British Virgin Islands Doctoral Dissertation, Universi ty of Puerto Rico, Mayagez, Puerto Rico. 149

PAGE 150

Mignucci-Giannoni AA, Montoya-Ospina RA, Jimenez-Marrero NM et al. (2000) Manatee mortality in Puerto Rico. Environmental Management 25, 189-198. Miller PS, Hedrick PW (2001) Purging of inbreeding depression and fitness decline in bottlenecked populations of Drosophila melanogaster. Journal of Evolutionary Biology 14, 595-601. Mills L, Citta J, Lair K, Schwartz M, Ta llmon D (2000) Estimating animal abundance using noninvasive DNA sampling: Promise and Pitfalls. Ecological Applications 10, 283-294. Minch E, Ruiz-Linares A, Goldstein DB, Fe ldman M, Cavalli-Sforza LL (1997) MICROSAT, version 1.5d: A computer program for calcula ting various statistics on microsatellite allele data. Standford Un iversity, Stanford, CA. Morales-Vela B, Padilla-Saldvar JA, Reid J, Butler S (2007) First records of long-distance manatee movements between Mexico and Belize. Sirenews 47 12-14. Murata Y, Nikaido M, Sasaki T, et al. (2003) Afrotherian phylogeny as inferred from complete mitochondrial genomes. Molecular Phylogenetics and Evolution 28, 253-260. Murphy WJ, Eizirik E, Johnson WE et al. (2001a) Molecular phylogene tics and the origins of placental mammals. Nature, 409, 614-618. Murphy WJ, Eizirik E, O'Brien SJ et al. (2001b) Resolution of the early placental mammal radiation using Bayesian phylogenetics. Science, 294, 2348-2351. Murphy WJ, Pevzner PA, O'Brien SJ (2004) Mammalian phylogenomics comes of age. Trends in Genetics, 20, 631-639. Nei M (1975) Molecular Population Genetics and Evolution North-Holland Publishing Company, Amsterdam, Netherlands. Nei M (1987) Genetic distan ce and molecular phylogeny. In: Population Genetics and Fishery Management (eds. Ryman N, Utter F), pp. 193-223. University of Washington Press, Seattle, WA. Nei M (1996) Phylogenetic analysis in molecular evolutionary genetics. Annual Review of Genetics, 30 371-403. Nei M, Tajima F (1981) Polymorphism detectable by restriction endonucleases. Genetics 97, 145-163. Nguyen T, Genini S, Bui L et al. (2007) Genomic conservation of cattle microsatellite loci in wild gaur ( Bos gaurus ) and current genetic status of this species in Vietnam. BMC Genetics 6, 77. 150

PAGE 151

Nie WH, Wang JH, O'Brien PCM et al. (2002) The genome phylogeny of domestic cat, red panda and five mustelid species revealed by comparative chromosome painting and Gbanding. Chromosome Research 10, 209-222. Nishihara H, Satta Y, Nikaido M et al. (2005) A retroposon analys is of Afrotherian phylogeny. Molecular Biology and Evolution, 22, 1823-1833. O'Brien SJ, Cevario SJ, Martenson JS et al. (1997) Comparative gene mapping in the domestic cat ( Felis catus ). Journal of Heredity, 88, 408-414. O'Brien SJ, Mayr E (1991) Bureaucratic mi schief recognizing endangered species and subspecies. Science 251 1187-1188. O'Brien SJ, Wildt DE, Goldman D, Merril CR, Bush M (1983) The cheetah is depauperate in genetic variation. Science 221, 459-462. O'Brien ST (1994) Guilds, structure and dynamics in the forest on Barro Colorado Island, Panama Doctorial Dissertation, University of Virginia, Charlottesville, VA. O'Shea TJ, Beck CA, Bonde RK, Kochman HI, Odell DK (1985) An analysis of manatee mortality patterns in Florida. Journal of Wildlife Management 49, 1-11. O'Shea TJ, Hartley WC (1995) Reproduction and early -age survival of manatees at Blue Spring, Upper St. Johns River, Florida. In: Population Biology of the Florida Manatee (eds. OShea TJ, Ackerman BB, Percival HF), pp. 157-170. National Biological Service Information and Technology Report 1. O'Shea TJ, Rathbun GB, Bonde RK, Buergelt CD, Odell DK (1991) An epizootic of Florida manatees associated with a dinoflagellate bloom. Marine Mammal Science 7, 165-179. O'Shea TJ, Reep RL (1990) Encephalization quotient s and life-history traits in the Sirenia. Journal of Mammalogy 71, 534-543. O'Shea TJ, Salisbury JS (1991) Belize a last stronghold for the manatees in the Caribbean. ORYX 25, 156-164. Ohta T, Kimura M (1973) Model of muta tion appropriate to estimate number of electrophoretically dete ctable alleles in a finite population. Genetical Research 22, 201204. Packard JM (1985) Development of manatee aer ial survey techniques. Manatee Population Research Report 7. Technical Report N o. 8-7, p. 68. Florida Cooperative Fish and Wildlife Research Unit. University of Florida, Gainesville, FL. Palumbi S, Martin A, Romano S et al. (1991) Simple Fools Guide to PCR, version II University of Hawaii, Honolulu, HI. 151

PAGE 152

Pardini AT, O'Brien PCM, Fu B et al. (2007) Chromosome painting among Proboscidea, Hyracoidea and Sirenia: Support for Paenungulata (Afrotheria, Mammalia) but not Tethytheria. Proceedings of the Royal Society B-Biological Sciences 274, 1333-1340. Pause KC, Kellogg ME, Beck CA et al. (2008) Genetic diversity and population structure of the Florida manatee ( Trichechus manatus latirostris). Submitted Pause KC, Nourisson C, Clark A et al. (2007) Polymorphic microsatellite DNA markers for the Florida manatee ( Trichechus manatus latirostris). Molecular Ecology Notes 7, 10731076. Peakall R, Smouse PE (2006) GENALEX 6: gene tic analysis in Excel. Population genetic software for teaching and research. Molecular Ecology Notes 6, 288-295. Powell JA, Belitsky DW, Rathbun GB (1981) St atus of the West Indian manatee (Trichechus manatus ) in Puerto Rico. Journal of Mammalogy 62, 642-646. Powell JA, Bonde RK, Aguirre AA et al. (2001) Biology and moveme nts of manatees in Southern Lagoon, Belize. In: The 14th Biennial Conference on the Biology of Marine Mammals Vancouver BC, Canada. Pritchard JK, Stephens M, Donnelly P (2000) In ference of population structure using multilocus genotype data. Genetics 155, 945-959. Pritchard JK, Wen W (2 004) Documentation for structure software: Version 2. Department of Human Genetics, University of Chicago, Chicago, IL, USA. Purves D, Augustine GJ, Fitzpatrick D et al. (2001) Neuroscience, p. 681. Sinauer Associates, Inc., Sunderland, MA, USA. Quintana-Rizzo E, Reynolds III J (2007) Region al management plan for the West Indian manatee ( Trichechus manatus). Caribbean Environment Programme, United Nations Environment Programme, CEP Technical Report. Rainey WE, Lowenstein JM, Sarich VM, Magor DM (1984) Sirenian molecular systematics including the extinct Steller's sea cow ( Hydrodamalis gigas ). Naturwissenschaften 71 586-588. Ramey RR, Liu HP, Epps CW, Carpenter LM, Weha usen JD (2005) Genetic relatedness of the Preble's meadow jumping mouse ( Zapus hudsonius preblei ) to nearby subspecies of Z. hudsonius as inferred from variation in cr anial morphology, mitochondrial DNA and microsatellite DNA: Implicati ons for taxonomy and conservation. Animal Conservation 8, 329-346. 152

PAGE 153

Rathbun GB, Carr N, Carr T, Woods CA (1985) The distribution of manatees and sea turtles in Puerto Rico, with emphasis on Rooseve lt Roads Naval Station, Springfield, VA. Rathbun GB, Possardt E (1986) Recovery plan for th e Puerto Rico population of the West Indian (Antillean) manatee ( Trichechus manatus manatus Linnaeus, 1758), p. 18. U.S. Fish and Wildlife Service, Atlanta, GA. Rathbun GB, Reid JP, Bonde RK, Powell JA (1995a) Reproduction in free-ranging Florida manatees. In: Population Biology of the Florida Manatee (eds. OShea TJ, Ackerman BB, Percival HF), pp. 135-156. National Biolog ical Service Inform ation and Technology Report 1, Washington, DC. Rathbun GB, Reid JP, Bonde RK, Powell JA (1995b) Reproduction in free-ranging Florida manatees. In: Population Biology of the Flor ida Manatee (Trichechus manatus latirostris). National Biological Servi ce, Information and Technology Report 1 (eds. O'Shea TJ, Ackerman BB, Percival HF), pp. 135-156. Rathbun GB, Reid JP, Carowan G (1990) Distri bution and movement patterns of manatees ( Trichechus manatus ) in northwestern peninsular Florida. Florida Marine Research Publications 48, 1-33. Raudsepp T, Fronicke L, Scherthan H, Gustavss on I, Chowdhary B (1996) Zoo-FISH delineates conserved chromosomal segments in horse and man. Chromosome Research 4, 218-225. Raven PH, Wilson EO (1992) A fifty-y ear plan for biodiversity surveys. Science, 258, 10991100. Raymond M, Rousset F (1995) GENEPOP (version 1.2): Population genetics software for exact tests and ecumenicism. Journal of Heredity 86, 248-249. Reep RL, Bonde RK (2006) The Florida Manatee: Biology and Conservation University Press of Florida, Gainesville, FL. Reep RL, Johnson JI, Switzer RC, Welker WI (1989) Manatee cerebral cort ex: Cytoarchitecture of the frontal region in Trichechus manatus latirostris Brain, Behavior and Evolution 34, 365-386. Reep RL, Marshall CD, Stoll ML (2002) Tact ile hairs on the postcranial body in Florida manatees: A mammalian lateral line? Brain, Behavior and Evolution 59, 141-154. Reid JP, Rathbun GB, Wilcox JR (1991) Distribu tion patterns of individua lly identifiable West Indian manatees ( Trichechus manatus) in Florida. Marine Mammal Science 7, 180-190. Rettenberger G, Klett C, Zechner U et al. (1995) ZOO-FISH anal ysis: Cat and human karyotypes closely resemble the puta tive ancestral mammalian karyotype. Chromosome Research 3, 479-486. 153

PAGE 154

Reynolds III JE, Odell DK (1991) Manatees and Dugongs Facts on File, Inc., New York, NY, USA. Reynolds JE, Szelistowski WA, Leon MA ( 1995) Status and conservation of manatees Trichechus manatus manatus in Costa Rica Biological Conservation 71, 193-196. Rice DW (1989a) Sperm whale. Physeter macrocephalus Linnaeus, 1758. In: Handbook of marine animals. Vol. 4 (eds. Ridgway SH, Harrison R), pp. 177-233. Academic Press, London. Rice WR (1989b) Analyzing tabl es of statistical tests. Evolution, 43, 223-225. Robinson TJ, Fu B, Ferguson-Smith MA, Yang F (2004) Cross-species chromosome painting in the golden mole and elephant-shrew: Support for the mammalian clades Afrotheria and Afroinsectiphillia but not Afroinsectivora. Proceedings of the Royal Society of London Series B-Biological Sciences 271, 1477-1484. Rodriguez-Lopez MA (2004) Phylogeography of manatees (Trichechus manatus) in Puerto Rico: A management tool for species conservation Master's Thesis, Universidad Metropolitana, San Juan, Puerto Rico. Roelke ME, Martenson JS, Obrien SJ (1993) Th e consequences of demographic reduction and genetic depletion in the endangered Florida panther. Current Biology, 3, 340-350. Ryder OA, Fleisher RC (1996) Ge netic research and its appl ications in the zoos. In: Wild Mammals in Captivity: Principles and Techniques eds. Kleiman DG, Allen ME, Thompson KV, Lumpkin S). Univers ity of Chicago, Chicago, IL. Sanson GD (1980) The morphology and occlusion of molariform cheek teeth in some Macropodinae (Marsupialia, Macropodidae). Australian Journal of Zoology Supplementary Series 28 341-365. Schalm OW (1965) Veterinary Hemotology, 2nd ed Lea and Febiger, Philelphia. Scherthan H, Cremer T, Arnason U et al. (1994) Comparative chromo some painting discloses homologous segments in distantly related mammals. Nature Genetics, 6, 342-347. Schneider S, Roessli D, Excoffier L (2000) Ar lequin ver. 2.001: A software for population genetics data analysis. (ed. Genetics and Bi ometry Laboratory). University of Geneva, Switzerland. Self-Sullivan C, Smith GW, Packard JM, LaCamma re KS (2003) Seasonal occurrence of male Antillean manatees ( Trichechus manatus manatus ) on the Belize Barrier Reef. Aquatic Mammals 29, 342-354. 154

PAGE 155

Sherwin WB, Moritz CM, (2000) Managi ng and monitoring genetic erosion. In: Genetics, Demography and the viability of fragmented populations (eds. Young A, Clarke G), pp. 9-34. Cambridge University Press, Cambridge, UK. Slatkin M (1995) A measure of popul ation subdivision based on micros atellite allele frequencies. Genetics, 139 457-462. Slone DH, Reid JP, Bonde RK, Butler SM, Stit h BM (2006) Summary of West Indian manatee ( Trichechus Manatus ) tracking by USGS-FISC Sirenia Pr oject in Puerto Rico. USGSFISC, Gainesville, FL. Smethurst D, Nietschmann B (1999) The distribution of manatees ( Trichechus manatus) in the coastal waterways of Tortuguero, Costa Rica. Biological Conservation 89, 267-274. Solinas-Toldo S, Lengauer C, Fries. R (1995) Comparative genome map of human and cattle. Genomics 27, 489-496. Soule J (1985) Glossary for Horticultural Crops New York. Southern SO, Southern PJ, Dizon AE (1988) Molecular characterization of a cloned dolphin mitochondrial genome Journal of Molecular Evolution 28, 32-42. Springer MS, Amrine HM, Burk A, Stanhope MJ (1999) Additional support for Afrotheria and Paenungulata, the performance of mitochondrial versus nuclear genes, and the impact of data partitions with heterogeneous base composition. Systematic Biology 48, 65-75. Springer MS, Murphy WJ, Eizirik E, O'Brien SJ (2003) Placental mammal diversification and the Cretaceous-Tertiary boundary. Proceedings of the National Academy of Sciences of the United States of America, 100, 1056-1061. Springer MS, Stanhope MJ, Madsen O, de Jong WW (2004) Molecules consolidate the placental mammal tree. Trends in Ecology & Evolution 19 430-438. Stahl A (1883) Fauna de Puerto-Rico. Clas ificacion sistmatica de los animales que corresponden esta fauna y catlogo del ga binet zoolgica del doctor A. Stahl en Bayamn, p. 248, Boletin Mercantil, Puerto Rico. Stanhope MJ, Madsen O, Waddell VG et al. (1998) Highly congruent molecular support for a diverse superordinal clade of endemic African mammals. Molecular Phylogenetics and Evolution 9 501-508. Stanyon R, Arnold N, Koehler U, Bigoni F, Wie nberg J (1995) Chromosomal painting shows that marked chromosomes in lesser apes and Old World monkeys are not homologous and evolved by convergence. Cytogenetics and Cell Genetics 68, 74-78. 155

PAGE 156

Stanyon R, Yang F, Cavagna P et al. (1999) Reciprocal chromo some painting shows that genomic rearrangement between rat and mouse proceeds ten times faster than between humans and cats. Cytogenetics and Cell Genetics 84, 150-155. Stejneger L (1887) How the great northern sea-cow ( Rytina ) became exterminated. American Naturalist, 21 1047-1054. Stewart BS, Yochem PK, Huber HR et al. (1994) History and present status of the northern elephant seal population. In: Elephant Seals: Population Ecology, Behavior, and Physiology (eds. LeBoeuf BJ, Laws RM), pp. 29-48. University of California Press, Berkeley, CA. Svartman M, Stone G, Page JE, Stanyon R (2004) A chromosome painting test of the basal Eutherian karyotype. Chromosome Research, 12, 45-53. Svartman M, Stone G, Stanyon R (2006) The ancestral Eutherian karyotype is present in Xenarthra. Plos Genetics 2, 1006-1011. Tajima F (1993) Measurement of DNA polymorphism. In: Mechanisms of Molecular Evolution. Introduction to Molecular Paleopopulation Biology (eds. Takahata N, Clark AG), p. 359. Japan Scientific Societies Press, Sinaue r Associates, Inc, Tokyo, Sunderland, MA. Takezaki N, Nei M (1996) Genetic distances and reconstruction of phylogenetic trees from microsatellite DNA. Genetics 144, 389-399. Telenius H, Carter NP, Bebb CE et al. (1992) Degenerate oligonucleotide-primed PCR General amplification of target DNA by a single degenerate primer. Genomics 13, 718725. Thompson JE (1939) Excavations at San Jose, British Honduras. Carnegie Institution of Washington Publication 506, 1-292. Thornback J, Jenkins M (1982) The IUCN Mammal Red Data Book. Part I. Threatened Mammalian Taxa of the Americas and the Australasian Zoogeographic Region (Excluding Cetacea). IUCN, Gland, Switzerland. Tolley KA, Vikingsson GA, Rosel PE (2001) Mitochondrial DNA sequence variation and phylogeographic patterns in harbour porpoises (Phocoena phocoena ) from the North Atlantic. Conservation Genetics 2, 349-361. Tringali MD, Davis MC, Rodriguez-Lopez MA et al. (2008a) Simultaneous use of the Xand Y-chromosome genes SMCX, SMCY, and DBY for sex determination in the Florida manatee ( Trichechus manatus latirostris ). Marine Mammal Science 24, 218-224. 156

PAGE 157

Tringali MD, Seyoum S, Carney SL et al. (2008b) Eighteen new polym orphic microsatellite markers for the endangered Florida manatee, Trichechus manatus latirostris. Molecular Ecology Resources 8, 328-331. U.S. Fish and Wildlife Service (1979) Es tablishment of manatee protection areas. Federal Register 44:60962 U.S. Fish and Wildlife Service (1982) Endangered and Threatened W ildlife and Plants (50CFR 17.11 and 17.12). U.S. Government Printing Office, Washington, DC. U.S. Fish and Wildlife Service (1996 ) Florida Manatee Recovery Plan, ( Trichechus manatus latirostris ), Second Revision, p. 160. U.S. Fish and Wildlife Service, Atlanta, GA. U.S. Fish and Wildlife Service (2001 ) Florida Manatee Recovery Plan ( Trichechus manatus latirostris ), Third Revision. United States Fish and Wildlife Servic e, Atlanta, GA. U.S. Fish and Wildlife Service (2007) West Indian manatee ( Trichechus manatus ) 5-Year Review: Summary and Evaluation, United States Fish and Wild life Service, Atlanta, GA, USA. U.S. Marine Mammal Commission (1996) A nnual Report to Congress, Bethesda, MD. USFWS (2007) West Indian Manatee ( Trichechus manatus) 5-Year Review: Summary and Evaluation, United States Fish and W ildlife Service, Atlanta, GA, USA. Van Oosterhout C, Hutchinson WF, Wills DPM, Shipley P (2004) MICRO-CHECKER: Software for identifying and correcting ge notyping errors in microsatellite data. Molecular Ecology Notes 4, 535-538. Vianna JA, Bonde RK, Caballero S et al. (2006) Phylogeography, phyl ogeny and hybridization in trichechid sirenians: Implications for manatee conservation. Molecular Ecology 15, 433-447. Waddell PJ, Shelley S (2003) Evaluating placen tal inter-ordinal phylogenies with novel sequences including RAG1, gamma-fibrinoge n, ND6, and mt-tRNA, plus MCMC-driven nucleotide, amino acid, and codon models. Molecular Phylogenetics and Evolution 28 197-224. Waldick RC, Kraus SS, Brown M, White BN (2002) Evaluating the effects of historic bottleneck events: An assessment of microsatellite variab ility in the endangered, North Atlantic right whale. Molecular Ecology 11, 2241-2250. Weber DS, Stewart BS, Garza J, Lehman N ( 2002) An empirical genetic assessment of the severity of the northern ele phant seal population bottleneck. Current Biology 10, 12871290. 157

PAGE 158

Weber DS, Stewart BS, Lehman N (2004) Gene tic consequences of a severe population bottleneck in the Guadalupe fur seal ( Arctocephalus townsendi ). Journal of Heredity 95, 144-153. Weigle BL, Wright IE, Ross M, Flamm R (2001) Movements of radio-tagged manatees in Tampa Bay and along Florida's west coast 1991-1996. Florida Marine Research Institute Technical Report TR-7 1-156. Westemeier RL, Brawn JD, Simpson SA et al. (1998) Tracking the l ong-term decline and recovery of an isolated population. Science, 282, 1695-1698. Westlake RL, O'Corry-Crowe GM (2002) Macrogeographic structure and patterns of genetic diversity in harbor seals ( Phoca vitulina ) from Alaska to Japan. Journal of Mammalogy, 83, 1111-1126. Whidden HP (2002) Extrinsic snout muscul ature in Afrotheria and Lipotyphla. Journal of Mammalian Evolution 9 161-184. White JR, Francis-Floyd R (1990) Ma natee biology and medicine. In: Handbook of Marine Mammal Medicine: Health, Disease and Rehabilitation (ed. Dierauf LA). CRC Press, Boca Raton, FL. White JR, Harkness DR, Isaacks RE, Duffield DA (1976a) Some studies on Blood of Florida manatee, Trichechus manatus latirostris. Comparative Biochemistry and Physiology APhysiology, 55, 413-417. White JR, Harkness DR, Isaacks RE, Duffield DA (1976b) Some studies on blood of the Florida manatee, Trichechus manatus latirostris. Comparative Biochemistry and Physiology aPhysiology, 55, 413-417. White JR, Harkness DR, Isaacks RE, Duffield DA (1977) Trichechus manatus latirostris (manatee). Order: Sirenia. Family Trichechidae. In: An Atlas of Mammalian Chromosomes (eds. Hsu TC, Benirschke K), p. 496. Springer-Verlag, New York, NY. Wienberg J, Jauch A, Stanyon R, Cremer T (1990) Molecular cytota xonomy of primates by chromosomal in situ suppression hybridization. Genomics 8, 347-350. Wilberg MJ, Dreher BP (2004) GENECAP: A progr am for analysis of multilocus genotype data for non-invasive sampling and captu re-recapture population estimation. Molecular Ecology Notes, 4, 783-785. Wright AC, Romney DH, Arbuckl e RH, Vial VE (1959) Land in British Honduras. Report of the British Honduras Land Survey Team, pp. 1-327. Colonial Resources Publication, London, UK. Wright S (1931) Evolution in Mendelian populations. Genetics 16, 0097-0159. 158

PAGE 159

Wright S (1951) The genetical structure of populations. Annals of Eugenics 15, 323-354. Yang F, Alkalaeva EZ, Perelman PL et al. (2003) Reciprocal chromosome painting among human, aardvark, and elephant (superorder Af rotheria) reveals the likely eutherian ancestral karyotype. Proceedings of the National Acade my of Sciences of the United States of America 100, 1062-1066. Yang FT, Graphodatsky AS, Li TL et al. (2006) Comparative genom e maps of the pangolin, hedgehog, sloth, anteater and human revealed by cross-species chromosome painting: further insight into the ancestral karyotype and genome evolution of eutherian mammals. Chromosome Research 14, 283-296. Ye JP, Biltueva L, Huang L et al. (2006) Cross-species ch romosome painting unveils cytogenetic signatures for the Eulipotyphla and evidence for the polyphyly of Insectivora. Chromosome Research 14, 151-159. Yeh FC, Boyle T (1997) Population genetic analysis of co-domin ant and dominant markers and quantitative traits. Belgian Journal of Botany 129 157. 159

PAGE 160

BIOGRAPHICAL SKETCH Margaret E. Kellogg was born in Boulder, CO, in 1981. She spent her early years in Ward, a small mountain town, and Tucson, AZ, before m oving to the plains outside of Boulder. She graduated from Niwot High School in 1999. She ear ned her B.S. in microbiology with minors in chemistry and plant molecular and cellular biolo gy from the University of Florida in 2003. While at the University of Florida, Margaret worked with Jacqueline Wilson in the U.S. Virgin Islands, investigating co ral reef fish behavioral intera ctions. She also worked in a research laboratory under the tutelage of Dr. Karen Koch, studying genomic DNA cell-wall mutations in maize. She began working with manatees in 2003, volunteering under Dr. Iskande Larkin, studying reproductive hormones. In 2004, she was accepted to graduate school under the mentorship of Dr. Peter McGuire in the College of Veterinary Medicine. She was a teaching assistant for the University of Florida, College of Veterinary Medicine La rge Animal Anatomy (2004) and SEAVET I (2008) courses. She has participated in many manat ee captures, health asse ssments, and necropsies throughout the state of Florida and Belize. She has attended many workshops incl uding Applied Conservation Genetics hosted by the U.S. Fish and Wildlife Service at th e National Conservation Training Center Shepherdstown West Virginia and taught by Drs. Fred Allendorf and Ti m King, leading experts in the conservation genetics field. She has received the University of Florida Named Presidential and Gritner Fellowships. Additionally, she was awarded th e Bonde/Reep Manatee Conservation Prize for her conservation genetics work with manatees.