xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200409Ka 4500
controlfield tag 001 002317649
007 cr bnu|||uuuuu
008 100903s2009 flua ob 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003122
Population genetics of Antarctic seals
h [electronic resource] /
by Caitlin Curtis.
[Tampa, Fla.] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains 117 pages.
Dissertation (Ph.D.)--University of South Florida, 2009.
Includes bibliographical references.
Text (Electronic dissertation) in PDF format.
ABSTRACT: I developed and tested a protocol for determining the sex of individual pinnipeds using the sex-chromosome specific genes ZFX and ZFY. I screened a total of 368 seals (168 crabeater, Lobodon carcinophagus; 159 Weddell, Leptonychotes weddellii; and 41 Ross, Ommatophoca rossii) of known or unknown sex and compared the molecular sex to the sex assigned at the time of collection in the Ross and Amundsen seas, Antarctica. Discrepancies ranged from 0.0% 6.7% among species. It is unclear, however, if mis-assignment of sex occurred in situ or in the laboratory. It also is possible, however, that the assigned morphological and molecular sex both are correct, owing perhaps to developmental effects of environmental pollution. I sequenced a portion (ca 475 bp) of the mitochondrial control region of Weddell seals (N = 181); crabeater seals (N = 143); and Ross seals (N = 41). I resolved 251 haplotypes with a haplotype diversity of 0.98 to 0.99.Bayesian estimates of theta from the program LAMARC ranged from 0.075 for Weddell seals to 0.576 for crabeater seals. I used the values of theta to estimate female effective population sizes (NEF), which were 40,700 to 63,000 for Weddell seals, 44,400 to 97,800 for Ross seals, and 358,500 to 531,900 for crabeater seals. Weddell seals and crabeater seals had significant, unimodal mean pairwise difference mismatch distributions (p = 0.56 and 0.36, respectively), suggesting that their populations expanded suddenly around 731,000 years ago (Weddell seals) and around 1.6 million years ago (crabeater seals). Both of these expansions occurred during times of intensified glaciations and may have been fostered by expanding pack ice habitat. Autosomal microsatellite based NEs were 147,850 for L. Weddellii, 344,950 for O. rossii, and 939,600 for L. carcinophagus. I screened one X-linked microsatellite (Lw18), which yielded a larger NE estimate for O. rossii than the other two species.Microsatellite NE estimates are compared with previously published mitochondrial NE estimates and this comparison indicates that the Ross seal may have a serially monogamous system of mating. I find no sign of a recent, sustained genetic bottleneck in any of the three species.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
Advisor: Earl McCoy, Ph.D.
t USF Electronic Theses and Dissertations.
Population Genetics of Antarctic Seals by Caitlin Curtis A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Biology College of Arts and Sciences University of South Florida Major Professor: Earl McCoy, Ph.D. Stephen A. Karl, Ph.D. Henry Mushinsky, Ph.D. Brent Stewart, Ph.D. Valerie Harwood, Ph.D. Date of Approval: July 17, 2009 Keywords: ZFX, ZFY, Leptonychotes weddellii, m itochondrial DNA, microsatellites, Y chromosome Copyright 2009, Caitlin Curtis
Acknowledgments As with the majority of large-scale pr ojects, this work could not have been completed without the assistance of many people. I would like to extend the lionÂ’s share of thanks to Dr. Stephen A. Karl, my major advisor, who brought me a new understanding of population genetics and did so with enthusiasm and a purple marker. Dr. Brent Stewart provided samples and advi ce. Thank you to Drs. Earl McCoy, Henry Mushinsky and Valerie Harwood for serving on my committee. This project involved a great deal of laboratory analyses which c ould not have been accomplished without the support and friendship of fellow graduate st udents, including Dr. Anna Bass, Dr. Ken Hayes, Dr. Emily Severance, Tonia Schwartz, Andrey Castro, and Cecilia Puchulutegui. Dr. Scott Lynn provided unflagging support and a dvice, regardless of the hour of day or relevancy of the question. Jose and Corinne Bello provided logistical support in the form of a place to work and an unlimited supply of chips, printer paper and lounge music. Cecilia Puchulutegui and Hugo Montiel provi ded asado when my plate was empty. Marc Dahl gave me a good shove when I needed it and Bob and Janis Gallo demonstrated how it all can come together, pref erably with a good bottle of Ch ateauneuf-du-Pape. I thank my parents, Nancy Curtis and Al Kott, and my brother Chris Curtis. In life, as in biology, sometimes the most important driving forces are not the largest or most obvious. Size doesnÂ’t always matter. I give thanks to and for my son Julian, who led me by the hand to a renewed enthusiasm for biology when I be gan to wonder where mine had gone, and my daughter Brune, who taught me about hope when I thought it was lost.
i Table of Contents List of Tables iii List of Figures iv Abstract v Chapter 1: Introductory Remarks 1 Chapter 2: Sexing Pinnipeds with ZFX and ZFY Loci 6 Introduction 6 Methods 8 Results and Discussion 11 Acknowledgements 17 Chapter 3: Pleistocene Population Expansions of Antarctic Pack-Ice Seals 21 Introduction 21 Methods 26 Samples 26 Genetic Methods 26 Sequence and Population Analyses 27 Results 29 MtDNA Variation 29 Effective Population Size 30 Demographic History of Antarctic Pack-Ice Seals 31 Discussion 32 Acknowledgements 42 Chapter 4: Autosomal and Sex-linked Patterns of Genetic Partitioning Among Three Species of Antarctic Seals 49 Introduction 49 Methods 55 Samples 55 Laboratory methods 56
ii Autosomal and X-linked Microsatellites 56 Y Chromosome Sequences 57 Data Analyses 59 Results 61 Microsatellites 61 Y Chromosome Variation 63 Discussion 64 Acknowledgements 78 Chapter 5: Conclusion 84 References 88 Appendices 113 Appendix 1: Among specie s variability in the ZFX gene 114 Appendix 2: Among species variability in ZFY gene 115 Appendix 3: DBY8 Sequence variable sites 116 Appendix 4: UTY11 Sequence variable sites 117 About the Author End Page
iii List of Tables Table 2.1 Primer sequence, annealing temp erature, and fragment size for loci used in ZYX / ZFY screening 18 Table 2.2 Summary of sex de termination in pinnipeds 19 Table 3.1 Genetic diversity and fe male effective popul ation sizes in pack ice seals 43 Table 3.2 Estimation of sudden populati on expansion in Antarctic seals 44 Table 4.1 The number of individuals ( N ), expected ( Hexp) and observed ( Hobs) heterozygosities, and number of alleles ( A ) seen at microsatellite loci in three Antarctic pack-ice seals 79 Table 4.2 Autosomal F estimate (Xu and Fu 2003) and the genetically effective population size ( NEA = F4 and SD) for microsatellie loci surveyed in pack-ice seals 80 Table 4.3 Estimates of gene tically effective population sizes and the ratio of NE to census size (in parentheses) for maternally, paternally, and biparentally inherited loci in pack-ice seals based on F (Xu and Fu 2003) or as estimated using the program LAMARC (Kuhner 2006). 81 Table 4.4 Observed and expected microsatel lite heterozygosities of seals. 82
iv List of Figures Figure 2.1. Map of the Antarctic contin ent showing area from which samples were collected 20 Figure 3.1. Map of the Antarctic contin ent showing area from which samples were collected 45 Figure 3.2a. Mismatch distribution for Weddell seals 46 Figure 3.2b. Mismatch distribu tion for crabeater seals 47 Figure 3.2c. Mismatch distribution for Ross seals 48 Figure 4.1. Map of the Antarctic contin ent showing area from which samples 83 were collected
v Population Genetics of Antarctic Seals Caitlin Curtis ABSTRACT I developed and tested a protocol for determin ing the sex of individual pinnipeds using the sex-chromosome specific genes ZFX and ZFY I screened a total of 368 seals (168 crabeater, Lobodon carcinophagus ; 159 Weddell, Leptonychotes weddellii ; and 41 Ross, Ommatophoca rossii ) of known or unknown sex and compared the molecular sex to the sex assigned at the time of collection in the Ross and Amundsen seas, Antarctica. Discrepancies ranged from 0.0% Â– 6.7% among species. It is unclear, however, if misassignment of sex occurred in situ or in the laboratory. It also is possible, however, that the assigned morphological and molecular sex both are correct, owing perhaps to developmental effects of environmental pollution. I sequenced a portion (ca 475 bp) of the mito chondrial control regi on of Weddell seals ( N = 181); crabeater seals ( N = 143); and Ross seals ( N = 41). I resolved 251 haplotypes with a haplotype diversity of 0.98 to 0.99. Bayesian estimates of from the program LAMARC ranged from 0.075 for Weddell seals to 0.576 for cr abeater seals. I us ed the values of
vi theta to estimate female effective population sizes ( NEF), which were 40,700 to 63,000 for Weddell seals, 44,400 to 97,800 for Ross seals and 358,500 to 531,900 for crabeater seals Weddell seals and crabeater seals had signifi cant, unimodal mean pairwise difference mismatch distributions ( p = 0.56 and 0.36, respectively), sugg esting that their populations expanded suddenly around 731,000 years ago (We ddell seals) and around 1.6 million years ago (crabeater seals). Both of these expans ions occurred during times of intensified glaciations and may have been fost ered by expanding pack ice habitat. Autosomal microsatellite based NEs were 147,850 for L. Weddellii 344,950 for O. rossii, and 939,600 for L. carcinophagus I screened one X-linked mi crosatellite (Lw18), which yielded a larger NE estimate for O. rossii than the other two sp ecies. Microsatellite NE estimates are compared with pr eviously published mitochondrial NE estimates and this comparison indicates that the Ross seal may ha ve a serially monogamous system of mating. I find no sign of a recent, sustained geneti c bottleneck in any of the three species.
1 Chapter 1: Introductory Remarks Population size and endurance may be posit ively correlated, as larger populations often maintain larger amounts of genetic vari ability allowing for continual adaptation to changing environmental and biotic conditions (e.g., Hansson and Westerberg, 2002; Reed and Frankham 2003). Quantifying po pulation sizes, however, may be difficult in species inhabiting logistically inaccessible aquatic Anta rctic pack-ice and fast-ice habitats. In addition, simple counts of i ndividuals may not accurately reflect the numbers of individuals contributing genetic information to subsequent genera tions, thus the long term genetic variability ma intained within a populati on. Effective population size (NE), as defined by Sewall Wright (1931) describes the number of individuals in an ideal population that would show the same dispersi on of allele frequencies as the observed population. Estimates of NE are often different (usually lo wer) than census size due to large variances in individual reproductiv e success, population size changes across generations, non-random systems of mating (e.g., polygeny and polyandry), and unequal sex ratios. When consideri ng evolutionary processes, NE is more important than a population count because NE represents the actual numbers of individuals that commonly contribute to each generation over the long te rm. Here, I estimate genetic effective
2 population sizes of three of the four phocid ca rnivores that live in the seasonal fast ice and pack ice habitats of the western Amundsen and Ross seas in west Antarctica. These seals occupy important niches in the not oriously short Antarctic marine food web. Due to their isolation around the Antarcti c continent, Antarctic phocid seals have large population sizes and have persisted relatively free from anthropogenic disturbances. In addition, they are closely related, yet employ varying long -term mating strategies. My research focused on three species of Anta rctic phocid seals. The Weddell seal, Leptonychotes weddellii, (Lesson 1826) named after the British sealing commander Sir James Weddell, tends to inhabit the land-fa st ice surrounding the An tarctic continent. Adults are brown, lighter ventrally, and mottle d with large darker and lighter patches, which tend to be silvery white on the ventra l surface. Males can grow up to 2.9 m in length, whereas females may reach about 3.3 m, weighing around 400-600 kg (Jefferson et al. 1993). Weddell seals have a varied di et, consisting primarily of fish and cephalopods such as squid and octopi (Plotz et al. 1991), and are known to forage occasionally on Antarctic krill (Euphausia superba). Conversely, Weddells and most other seals become prey items of orcas (Orcinus orca). Leopard seals (Hydrurga leptonyx) are also known to feed on pups and su badults. Males use their large canine teeth to maintain breathing holes in the i ce, and actively defend the three-dimensional surrounding territory (Kaufman et al. 1975, Bartsh et al. 1992), though prolonged scraping at the ice may wear teeth down to th e pulp cavity over time, leading to mortality
3 (Stirling 1969). Weddell seals have a mode rately polygynous breeding system, in which males may mate with as many females as may share his breathing hole (Gelatt et al. 2000). The Weddell seal is known for its very deep dives which may reach 700m, and may stay underwater for more than 60 minutes (Kanatous et al. 2002). Such deep dives typically involve foraging sessions (Mitani et al. 2004), as well as searching for cracks in the ice sheets that can lead to new breathing holes. The seal is thought to be able to remain submerged for such long periods of time in part due to low levels of aerobic lipidbased muscle metabolism (Kanatous et al. 2002). The crabeater seal, Lobodon carcinophagus (Hombron and Jacquinot 1842), uses its highly specialized multilobed teeth to strain Antarctic krill (Euphausia superba), which form the majority of its diet (Laws 1977). Despite its name, the crabeater seal does not feed on crab. Dark yellowish brown to silvery grey, which is lighter ventrally, they have slender bodies and long muzzles (Jefferson et al. 1993). Crabeater seals are somewhat sexually dimorphic in size, with males up to 2.6 m and females 2.8 m (Laws et al. 2003). In late summer (i.e., post-breed ing season) when surveys have been conducted, they are generally found near the outer edges of the pack ice. Crabeater seals are thought to be serially monogamous, whereby one males mates with a single female within a breeding season, though not necessarily the same female between seasons (Stirling 1983). Population size has been esti mated at 7 to 15 million individuals (Laws 1977, Erickson and Hanson 1990) and the ci rcumpolar population appears to be
4 panmictic with no indication of geographi cally localized breeding groups (Davis et al. 2000, 2008). Relatively little is known about the Ross seal, Ommatophoca rossii (Gray 1844). Ross seals reach more than 2.4 m and 204 kg (Jefferson et al. 1993), making them the smallest of the Antarctic phocids, though females are usually slightly la rger than males. Countershaded, they are dark gr ey dorsally and silvery grey underneath, often with brown or reddish brown stripes on the neck, sides and chest (Jefferson et al. 1993). They appear to be solitary, and utilize the stable ice floes on the exterior of the pack ice to molt, while they give birth on the more densely pack ed interior pack ice (Splettstoesser et al. 2000, Stewart 2007). Though they have a circum polar distribution, hi gher numbers may be found in the Ross Sea, King Haakon VII Sea, and perhaps portions of the western Weddell Sea (Stewart 2007). Ross seals may spe nd much of the rest of their time to the north of the pack ice, alone in the open wa ter (Stewart 2007). The mating system of this species is not known. My research comprises four main elemen ts, which have been organized into three chapters. Chapter one (Sexing Pinnipeds with ZFX and ZFY Loci) focuses on developing and testing a protocol for genetically determ ining the sex of pinnipe ds in the laboratory through developing and employing sex-chromoso me specific genetic markers located in the zinc-finger protein regions of the X and Y chromosomes ( ZFX and ZFY, respectively). Presence of the Y chromosome specific ZFY marker, as determined by
5 presence of a PCR amplicon of the correspond ing size, indicated a male tissue sample, whereas absence of the ZFY marker indicated a female sample. Laboratory confirmation of the sex of each seal sample was critical to the sex-specific analyses in later chapters. Genetic data were compared to the se x of the animals as it was described in situ upon collection. Chapter two (Pleis tocene Population Expansions of Antarctic Pack-Ice Seals) uses mitochondrial DNA sequence data to esti mate the current dist ribution of genetic variation among the three species of seals, which is in turn used to estimate female effective population size and long term popul ation stability and demography. Chapter three (Autosomal and sex-linked patterns of genetic partitioning among three species of Antarctic seals) used nine autosoma l DNA microsatellites and one X-linked microsatellite to estimate the current dist ribution of genetic va riation among the three species of seals, which in turn was used to estimate to tal effective population sizes. Autosomal and X-linked effective population sizes were compared to the mitochondrial estimates from Chapter two, as well as previously published census estimates. The microsatellite data were also used to look for evidence of pa st genetic bottleneck events. Y chromosome data is presented in this chapter.
6 Chapter 2: Sexing Pinnipeds with ZFX and ZFY Loci Introduction The ability to accurately and reliably iden tify the sex of free-ranging animals is essential for estimating sex ratios, ca tegorizing behavioral observations, and understanding almost every aspect of an an imalÂ’s life history. For many species, it may be relatively easy to distinguish between a dult males and females if they are sexually dimorphic in size or color. Distinguishing subadult or juvenile males from females, however, often can be challenging. Even a dult sex can be difficult to determine in animals like the phocid pinnipeds that live in pack ice or fast ice hab itats of the Antarctic and exhibit little to no sexual dimorphism. This difficulty is even more pronounced when individuals are viewed from a distance with no direct physical examination. In phocid pinnipeds, male genitalia are internal and, cons equently, the only clue to sex in otherwise sexually monomorphic species is the presence (male) or absence (female) of a ventral penile opening. For animals with chromoso mal sex determining mechanisms, molecular sex determination has the potential to une quivocally determine the genetic sex of individuals. This approach can circumvent many of the diffi culties in identifying sex of
7 animals in the field and articularly useful in secretive species wher e only traces of the individual such as blood, ha ir, or scat are available. Two sex-chromosome specific genes, ZFX and ZFY are zinc-finger homologues located on the X and Y chromosomes, respecti vely (Pecon-Slattery and OÂ’Brien 1998). Because it is typically located outside the pseudoautosomal region of the Y chromosome in eutherian mammals (Mardon and Page 1989, Page et al. 1987), ZFY only rarely recombines with ZFX (but see Pecon-Slattery et al. 2000), and together these genes have proven useful as a molecular method for de termining the sex of many mammalian species (e.g., felids, Pilgrim et al. 2005; canids, Lucchini et al. 2002; sea otters, Hattori et al. 2003; pinnipeds, ungulates, and ursids, Shaw et al. 2003; cetaceans Morin et al. 2005; prosimians and humans, Fredsted and Villesen 2004; and rodents, Marchal et al. 2003). Shaw et al. (2003) developed a PCR-based ZFY / ZFX assay applicable in a variety of mammals, including one pinni ped, the harbor seal (Phoca vitulina ), by using a single generic primer pair to simultaneously amp lify both homologues in a single PCR reaction, then verifying presence or absence of the amplicons on an agarose gel This approach, however, relies on differences in the size of the Xand Y-specific regions that may be subjected to PCR competition and lead to al lele dropout of the larger allele producing incorrect sex assignment. Although I have no em pirical indication that this is happening with the Shaw et al. (2003) primers, it is nonethel ess a well-documented phenomenon and one I wished to avoid (Piyamongkol et al. 2003, Sefc et al. 2003, Buchan et al.
8 2005). I expanded on the work by Shaw et al. (2003) and created primers specifically targeted to separately amplify a portion of the ZFX or ZFY gene in pinnipeds. This allowed me to genotype each seal to determin e sex and to directly sequence one or both of these genes which could be used for fo rensic or species id entification purposes. Crabeater ( Lobodon carcinophagus ), Ross ( Ommatophoca rossii ), and Weddell ( Leptonychotes weddellii ) seals are phocid pinnipeds that live almost exclusively in fast ice or pack ice habitats ar ound the Antarctic Continent (R eeves and Stewart 2003, Reeves et al. 1992). Males and females of each species are similar in size and color and are not easily distinguished most of the time without close inspection for th e presence or absence of a ventral penile opening. Consequently a genetic method for determining or verifying sex may be useful to a variety of ecological studies in these species. Methods Tissue samples were collected from 168 free-living crabeater seals (L. carcinophagus ), 159 Weddell seals (L. weddellii), and 41 Ross seals (O. rossii) via remote darting or direct handling during the 2000 Antarctic Pack Ice Seal (APIS) scientific cruise (Decker et al. 2002, Solls et al. 2005). Although a ll three species are circum-polar, most samples were collected from the pack-ice zone of the eastern
9 Amundsen and Ross Seas, approximately 67 Â– 78 S, 129 Â– 180 W (Figure 2.1) and some L. weddellii samples came from McMurdo Sound, Antarctica. Of the crabeater seals, there were 71 field identified ma les, 56 females, and 41 sex unknown. For Weddell seals there were 90 males, 64 females, and 5 unknown and for the Ro ss seals there were 25 males and 16 females. I also assayed f our male and two female northern elephant seals and one male and one female California sea lion. Since samples from the latter two species were from captive individuals, I am highly confident of the true sex of the individuals. Most of the sampled Antarc tic phocids, however, were free ranging and many were sampled remotely by biopsy dart and not directly examined. I designed primers specifically target ing the last intron of the phocid ZFX and ZFY genes. To accomplish this, I used the tw o, previously published sets of nested generic ZFX and ZFY felid primers (Pecon-Slattery and OÂ’Brien 1998; Table 2.1) to simultaneously amplify both gene regions fr om a male crabeater seal. PCR products were cloned (Original TA Cloning Kit, Invitr ogen, Inc. Carlsbad, California, USA) and sequenced on an ABI 3730XL automatic seque ncer (Macrogen Inc., Seoul, Korea). Sequences were aligned using Sequencher software (GeneCodes Corp., Ann Arbor, MI) and compared to published ZFX and ZFY sequences in GenBank. From the aligned crabeater seal and GenBank sequences, locus-spec ific primers were desi gned to target the ZFX or ZFY loci separately (Table 2.1). A subset of male crabeater, Weddell, and Ross seals were amplified and sequenced at both loci. I also assayed ZFY and ZFX in one
10 northern elephant seal ( Mirounga angustirostris ) and one California sea lion (Zalophus californianus ). To determine the sex of individua ls, I set up separate PCR reactions for the ZFX and ZFY genes for all individuals and visuali zed the amplifications side by side on an agarose gel. There are four pos sible amplification patterns. If both ZFX and ZFY or just ZFY alone amplified, the individual was assigned male. If the ZFX but not the ZFY amplified the individual was assigned female. If neither locus amplified, the individual was classified as unresolved. It should also be noted that individuals that amplify for the ZFX locus and not the ZFY could be either females or non-amplifying males (i.e., false negative for ZFY ). The inclusion of a second male specific locus (e.g., SRY) can be helpful in confirming the results (Gilson et al. 1998). DNA was extracted from frozen tissue sa mples using standard phenol-chloroform techniques and/or using a DNeasy Tissu e Kit (Qiagen, Valencia, CA, USA). Amplification reactions generally were 25 l and contained 0.5 l total cell DNA, 1 X reaction buffer (Promega, Madi son, WI, USA), 2.0 mM MgCl2, 0.2 mM of each dNTP, 10 pmol of each primer, 6 mg BSA, and 1.25 U Taq polymerase (Promega, Madison, WI, USA). Thermocycling conditions were 95 C 2 min, 35 cycles of 95 C 1 min, annealing temperature (Table 2.1) 1 min, and 72 C 1 min, followed by a final extension at 72 C for 7 minutes. PCR products were visualized on a 2% agarose gel with ethidium bromide to assess quantity and fidelity of amplificati on and then purified using either Microcon Centrifugal Filter Units (Mil lipore Corp., Billerica, MA, USA) or QIAquick spin
11 columns (Qiagen, Valencia, CA, USA). A pproximately 100 ng of purified PCR product was directly sequenced in both directi ons on an ABI 3730XL automatic sequencer (Macrogen Inc., Geumcheon-gu, Seoul, Korea). Results and Discussion All five species successfully amplified fo r both loci. The length of the crabeater seal ZFY fragment, including the primer sequen ces, was 931 nucleotides (nt). The other species produced DNA fragments of similar length. The length of the crabeater seal ZFX fragment including the primer sequences wa s 1045 nt and the other species produced fragments of similar length. The ZFX fragment matches with 83% identity to the final ZFX intron of Bos taurus (Lawson and Hewitt 2002; GenBank accession AF241273) and my ZFY fragment was 79% identical to the fi nal intron of the Am ur leopard (AB211426). After removing segments for which reliable sequence data was not obtained from all individuals (usually at the beginning and end of the sequenc e), I was able to cleanly resolve 851 and 956 nucleotides of sequence for the ZFY and ZFX loci, respectively. The genotypic sex of nearly all seals (95.8%) agr eed with the sex assigned in the field (Table 2.2). Discrepancies between the field and laboratory assigned sex ranged from 0 to 6.7%. Conflicts betw een the laboratory and field a ssigned sex of an individual
12 can arise from either field sexing errors or an unidentified laboratory artifact. I am most confident in the laboratory assessment when both ZFX and ZFY amplifications produce strong, clear bands since non-homologous amplifi cation of an appropriate sized fragment is unlikely given the specificity of the primers. Furthermore, the homology of the fragment can be verified by DNA sequence or RFLP analysis. Any male that amplifies for the ZFX gene but fails to amplify for the ZFY gene would appear to be a female by my genetic tests. That the ZFX gene amplified indicates that the failure of ZFY amplification was not due to poor template quality or other general amplifications problems. Given that I saw no intra-specific variation and that th ese primers generally worked well in all three species I also believe that this type of locus specific artifact is unlikely. Nonetheless, I cannot definitively rule out that the eight individuals that were field identified as male but failed to amplify for the ZFY locus may indeed, be male. A field-identified female appearing to be male based on amplification of both the ZFY and ZFX loci, is a likely candidate for field miside ntification. Sex of indi viduals in the field was determined in one of three ways. Fi rst, many crabeater and Weddell seals were closely approached so that skin samples could be taken from the traili ng edge of their rear flippers while they slept. At this time, individuals were visually examined for the presence of a ventral penile opening or for distinctive scarring around the neck and fore flippers (suggesting male in cr abeater seals). Six of the inco rrectly identified individuals in Table 2.2 were sampled in this way, evenly split between the two possible
13 discrepancies (i.e., field male but genetic female and field female but genetic male). Second, a number of seals (particularly cr abeater seals which are more difficult to approach) were remotely sampled by biopsy da rting either using a crossbow or a hand held dart pole. The sex in these cases was assigned based on the quickly observed presence or absence of a penile opening or scarring on neck and near the fore flippers. Five of the total 15 incorrectly sexed individu als were sampled in this way and all but one of these five, a Weddell seal, were field ma les but genetic females. Finally, a smaller number of seals (13) were captured and an esthetized while samples were taken and telemetry instruments attached. These indivi duals were examined closely and assigned a sex by a wildlife veterinarian. On ly one individual field-identif ied as male but failing to amplify the ZFY gene (a crabeater) was among the examined individuals. This may indicate that, although rarely, fa lse negative amp lification of ZFY has occurred. Interestingly, three of the examined indi viduals field-identified as female (two crabeater seals and one Ross seal) produced strong ZFY (and ZFX ) amplifications. Given that I believe that false positive amplification is unlikely, it would seems logical that these individuals might have been misidentif ied in the field. To the contrary, however, given that these individuals received a clos e examination by a wildlife veterinarian, it also seems highly unlikely that the field id entification is wrong. Although I cannot fully validate either the field or la boratory methods for sex identific ation, it is possible that the observed discrepancy of a fiel d-identified female clearly being a genetic male does not
14 involve errors in either method. It has been observed that three nor thern elephant seals ( M. angustirostris ) in California clearly had secondary sexual char acteristics of adult males yet lack a penile opening leading to ambiguous sexual assignment (Stewart BS personal observation). Unfortunately, I did not have access or samples from these individuals to determine genetic sex. Presum ably, if I did, they would produce positive amplifications for ZFY Persistent organic pollutants (POPs) are nearly ubiquitous in the environment (Damstra, et al. 2002) and many of them (e.g., dithioth reitol, polychlorinated biphenyls, and tributyltin) are endocrinedisrupting chemicals. In vertebrates, some POPs act as estrogen mimics or androgen antagonists resul ting in genetically male individuals possess female physical characteristic s (as is seen here; Ayaki et al. 2005, Hayes et al. 2002, Penaz et al. 2005). Although the near absence of industrial development and its remoteness make the Antarctic appear to be an unlikely place for POPs this is clearly not the case. While studying sediment cores in McMurdo Sound, Ross Sea, Antarctica, Nigri et al. (2004) documented detectable levels of but yltins (i.e., TBT, DBT, and MBT) at six of eight surveyed sites. Buty ltins are commonly used in antifouling paints on large boats including the icebreakers th at visit McMurdo Sound. Nigri et al. (2004) thought that butyltins might be introdu ced into the sediment from paint chips rubbed off of icebreakers. One of their sampling site s in McMurdo Sound, Cape Armitage, had extremely high levels of butyltin Â“Â…only exceeded in very busy harboursÂ…Â” (Negri et al.
15 2004). Goerke et al. (2004) also documented a 30 Â– 160 fold biomagnification of several POPs in Weddell seals and s outhern elephant seals ( Mirounga leonina ). Although it is difficult to say at this time what, if any, effect Antarctic POPs are having on the health and sexual development of Antarctic pack-ice seals, there is a wealth of studies indicating that the presence of POPs in the environment can have long-lasting and dire consequences for wildlife. They also may have been contributing factors resulting in field sex identifications not agreeing with my genetic sex assignments. Regardless, being able to genetically assess the sex of free-ranging seals can provide a backup method for testing the veracity of visu al designations, allow sex determination of DNA samples when individuals are not handled or even sighted, and provide a key to the understanding of the impacts th at POPs might have in marine ecosystems. Furthermore, to fully assess the potential e ffects that POPs may be having on natural populations, it is necessary to have both genetic and mo rphological information along with some understanding of the indivi dual exposure to POPs. I found no intra-specific ZFY variation after sequencing 12 crabeater (GenBank accession number DQ493902), 10 Weddell ( DQ493904), or 10 Ross seals (DQ493903). There also was no intra-specific variation in ZFX genes of those speci es after screening four (DQ811091), two (DQ811093), and two (DQ811092) individuals, respectively. In addition, I sequenced ZFY and ZFX from one each of northern elephant seals ( Mirounga angustirostris ; ZFY Â– DQ493906, ZFX Â– DQ811095) and California sea lion ( Zalophus
16 californianus ; ZFY Â– DQ493905, ZFX Â– DQ811094). Among all five species there was considerable inter-specifi c variation at both the ZFY and ZFX genes (Appendix A and B). At the ZFX locus there were 50 variable nucleotide posi tions (~5%). Of those, 29 were transitions, 10 were transversion s, and three were deletions ( one 3 nt and two 4 nt). At ZFY 61 of the nucleotide sites we re variable (~7%). Of t hose, 33 were transitions, 13 were transversions, one had both a transiti on and a transversion, and one had both a deletion and a transition. There was a singl e 4 nt deletion in th e Weddell sequence and there were two 2 nt and two 3 nt deletions in the Z. californianus sequence. It appears that all species can be uni quely identified from all others through DNA sequence by at least three (crabeater seal at ZFY ) to at most 31 (California sea lion at ZFY ) sites. My assessment of intra-specific variation, how ever, was limited and species designations should be made on multiple sites or a larger number of individuals from throughout the range of the species. Nonethele ss, this level of va riation is likely to be useful in identifying species from forensic samples and for phylogenetic analyses. The results of this study have been thr eefold. First, I have indicated that the ZFX and ZFY loci are likely to be good nuclear mark ers for species iden tification. Second, I have demonstrated the utility of new ZFX and ZFY markers for assigning the sex of pinnipeds especially when access to the an imal is highly limited. Third, I uncovered several cases where the sex designated base d on morphology appears to be in conflict with molecular markers. In some of these cases, it likely is simply misidentification in the
17 field or technical artifact. Ot hers, however, may be indica ting that persistent organic pollutants are having significant impact on An tarctic fauna. Further targeted research, however, is needed before final conclusions about the likelihood and magnitude of the effect can be reached. Acknowledgments Phocid samples in this study were collect ed in Antarctica by Brent Stewart during a multidisciplinary research cruise to the Ross and Amundsen seas in 2000. I thank Claudia Rocha for laboratory assistance and 3 anonymous reviewers for helpful comments on earlier drafts of the manuscrip t. Funding for this study was provided by an American Museum of Natural History, Lern er-Gray grant to Caitlin Curtis and by the National Science Foundation grants OPP 98-16011 and OPP 98-16035 to Brent Stewart and DEB 98-06905 and DEB 0321924 to Stephen Karl.
18 Table 2.1. Primer sequence, annealing temperat ure, and fragment size for loci used in ZFX / ZFY screening. All amplifications were of the crabeater seal (Lobodon carcinophagus) Locus Primer Sequence (5' to 3') C Size (bp) a Reference Zinc-finger Y ( ZFY F) F: CCAAACAGGTGAGGGTGCACA 60 931 This Study Zinc-finger Y ( ZFY R) R: GTAATCACAGTCAGTACAGTGG Zinc-finger X ( ZFX F) F: TGAGGGCACATGAGTCCCACA 55 1045 This Study Zinc-finger X (ZF2RA) R: GGTGGTTGTGTAAACTTATCTT ZF1F F: ATAGATGAGTCTGCTGGC 48 Multip le Pecon-Slattery and OÂ’Brian 1989 ZF1R R: CGTTTCAAATCACTTGA ZF2F F: GGTGATTCCAGGC AGTAC 52 ~1200 Pecon-Sla ttery and OÂ’Brian 1989 ZF2R R: TGGTCAGCTTGTGGCTCTCCT a Size of the fragment in cludes the primer sequence.
19 Table 2.2. Summary of sex determination of pinnipeds. The number correct refers to agreement between the field and laboratory sex determinations and the numbe r incorrect refers to disagreement. Unkn own refers to individuals where sex was not assigned in the field and unres olved is the number of samples that did not reliably amplify with either ZFX or ZFY primers. Field Identified As Male Female S pecies Correct Incorrect Correct Incorrect Unknown Unresolved Total a L obodon carcinophagus 65 (94.2%) 4 (5.7%) 48 (96.0%) 2 (4.0%) 41 14 (8.3%) 168 L eptonychotes weddellii 86 (95.5%) 4 (4.5%) 57 (93.4%) 4 (6.6%) 5 5 (3.1%) 159 O mmatophoca rossii 25 (100%) 0 (0%) 14 (93.3%) 1 (6.7%) 0 1 (2.4%) 41 M irounga angustirostris 4 (100%) 0 (0%) 2 (100%) 0 (0%) 0 0 6 Z alophus californianus 1 (100%) 0 (0%) 1 (100%) 0 (0%) 0 0 2 T otal 181 (95.8%)8 (4.2%) 122 (94.6%)7 (5.4%) 46 19 (5.1%) 376 a Row sum may not equal total becaus e some individuals field-classified as unknown also were unresolved.
20 Figure 2.1. Map of the Antarctic continent show ing the area where genetic samples were collected. Weddell Sea Ross Sea Ronne Ice ShelfCape Ponsett Amory Ice Shelf 500 Km 0 N 180 W 90 W 90 E 150 W Study Area Ross Ice Shelf Cape Adare Amundsen Sea McMurdo Station 120 W
21 Chapter 3: Pleistocene Population Expansions of Antarctic Pack-Ice Seals Introduction Population abundance and persistence are of ten positively correlated, as larger populations are thought to be more resilien t to ecological perturbations. Similarly, genetically variable pop ulations are generally more fit th an less variable ones presumably because greater genetic variability allows for adaptation to changing environmental and biotic conditions (e.g., Hansson and West erberg, 2002; Reed a nd Frankham 2003). A simple count of individuals in a populati on may not, however, accurately reflect the numbers of individuals that are actually cont ributing to the next generation and thus the long-term evolutionary potential of the popul ation. Wright (1931) de fined the genetically effective population size ( NE) as the number of individuals in an ideal population that would have the same rate of genetic drift as the observed population. When considering evolutionary processes, NE is more important than a population count because NE represents the actual num bers of individuals th at commonly contribute to each generation over the long term. Several different factors can conspire to result in NE being different from census population size (almost always smaller). Large vari ance in individual reproductive success, population size changes across generations, non-random systems of
22 mating (e.g., polygeny and polyandry), and unequ al sex ratios all work to reduce NE compared to census size (Hartl and Cl ark, 1997). Consequently, to know the evolutionary potential of a population an understanding of the genetic effective population size is needed first. Here, I estima te genetic effective population sizes of three of the four phocid carnivores that live in the seasonal fast ice and pack ice habitats of the western Amundsen and Ross seas in west An tarctica. These seals occupy important niches in the notoriously s hort Antarctic marine food web. Demography of Antarctic seals remains a critical but relati vely little understood element of the Antarctic mari ne ecosystem, though it is important to management models and conservation plans for the Southern Ocean. This is particularly true for crabeater seals ( Lobodon carcinophaga ), which have a large circumpolar population and feed exclusively on Antarctic krill (Euphausia superba ; e.g., Mori and Butterworth 2004). Owing principally to the remoteness and heavily ice-covered habitats of pack ice seals, there has been only de minimus anthropogenic exploitation and impact to their populations (e.g., Oritsla nd 1970, Stirling 1971, Laws et al. 2002). Abundance and population trend data for crabeater seals ha ve been specifically identified by the International Whaling Commission as importa nt, though deficient, when considering trophic interactions with Southern Ocean whale populations (International Whaling Commission 2005) and this is arguabl y true for all Antarctic seals. My research focused on three closely rela ted species of phocid carnivores that
23 have circumpolar Antarctic distributions, o ccupy distinct ecological niches, and have different mating tactics. Th e population of Weddell seals ( Leptonychotes weddellii ) has been estimated at around 730,000 to 800,000 (Laws 1977, Erickson and Hanson 1990) and typically inhabits the land-fast ice surrounding the Antarctic continent when breeding. They appear to be polygynous with males defending underwater territories and mating with up to five females during each breeding season (Gelatt et al. 2000). Regional populations apparently consist of separate geographically disjunct breeding colonies (Stirling 1969). Adult female seals appear to be philopatric to br eeding site and natal philopatry may also be strong (e.g., Stirling 1969). Relative to seals outside of the Antarctic, however, choice of breeding site s by Weddell seals is more variable (cf. Croxall and Hiby, 1983). Adjacent colonies may have substantial genetic exchange driven in part by intraspecific competiti on for prime space within or among breeding colonies (Hastings and Testa, 1998). Further, mo st estimates of site fidelity are based on tagging studies that generally reveal levels of exchange sufficient (i.e., on average, one migrant exchanged per generation) to geneti cally homogenize populations. There is some indication that at least dist antly separated colonies (i .e., on opposite sides of the continent) might be genetically differentiated (Davis et al. 2000, 2008). Those assessments, however, have indicated only sl ight genetic subdivision among breeding colonies and no genetic subdivi sion when individuals are mi xed in the pack ice in the Ross and Amundsen seas (Davis et al. 2008). But given the poten tial for philopatry and a
24 polygnous mating system, I would expect that NE in this species would be smaller than the other species where these attributes were not found. Crabeater seals are generally found near th e outer edges of the pack ice in late summer (i.e., post-breeding season) when su rveys have been conducted. Population size has been estimated at 7 to 15 million individuals (Laws 1977, Erickson and Hanson 1990) and the circumpolar population appears to be panmictic with no indication of geographically localized breeding groups (Davis et al. 2000, 2008). The abundance of crabeater seals was expected, by some, to ha ve increased substant ially following the decline and near collaps e of populations of krill-eating bale en whales in the Antarctic in the early to mid 20th century (e.g., Laws 1977, Smetacek and Nicol 2005). Between the 1920Â’s and the mid 1960Â’s, the total abundan ce of baleen whales dropped from an estimate of 22 million metric tons to approximately 2 million metric tons (Macintosh 1970). Reduced competition for krill presumably would have promoted an increase in the abundance of crabeater seals. The population si ze of crabeater seals might have declined in recent years, however, owing to resource competition from rebounding whale populations and the beginning of comme rcial harvesting of krill (Bester et al. 1995). Crabeater seals are thought to be serially monogamous with males mating with only a single female each mating season but not neces sarily the same female between seasons (Stirling 1983). Consequently, I expect that the combination of this breeding system with panmixia and an extremely large population size will result in a larger NE and less of a
25 difference between NE and census size. Much less is known about the ecolo gy or breeding biology of Ross seals ( Ommatophoca rossii ). They are usually found in d eeper pack ice where medium and large ice floes are common. Their circumpol ar abundance has been estimated at 131,000 to 220,000 (Laws 1977, Erickson and Hanson 1990) and there is no indication of geographic population subdivision (Davis et al. 2008). All three species, however, have evolved and existed in relative isolation and have experienced very little to no sustained hunting pressure. Historic population changes, therefore, should be a reflection of natural processes and not a result of anthropogenic disturbance. To better understand the ecology and breeding biology of these unique animals, I explored the degree of genetic variation in nuclear and mitochon drial loci of a relatively large number of seals sampled in the Ross and western Amundsen seas in west Antarctica from December 1999 through February 2000. Specifically, I have assa yed DNA sequence variation in sex-linked nuclear loci (both X and Y; Curtis et al. 2007), autosomal and sex-linked microsatellite loci (data not shown), and ma ternally inherited cytoplasmi c mitochondrial DNA. Here, I report the results from the matern ally inherited mitochondrial DNA.
26 Methods Samples Biopsy samples of skin were collect ed from 181 free-living Weddell seals 143 crabeater seals, and 41 Ross seals by remote darting or direct handl ing during the Austral summer from December 1999 through February 2000 (cf. Decker et al. 2002, Solls et al. 2005). Most samples were collected from the pack ice zone (i.e., not specific breeding colonies) of the eastern Ross and western Am undsen seas, between 67and 78 S latitude and 129 to 180 W (Figure 3.1) though a few samples from Weddell seals were collected from a long-studied co lony in Erebus Bay, McMurdo Sound. Sex of most individuals was determined and or confirmed in the laboratory using ZFX and ZFY genetic screening (Curtis et al. 2007). Tissue samples were e ither stored in ethanol or frozen at Â–80oC until DNA extraction. Genetic Methods Total cell DNA (tDNA) was extracted using standard phenol-chloroform techniques or with a DNeasy Tissue Kit (Qia gen, Valencia, CA, USA), and 5 L of DNA
27 was visualized on a 0.8% agarose gel with ethidium bromide to assess DNA quality. I amplified a roughly 475 nucleotide (nt) fragme nt of the 5Â’ portion of the mitochondrial control region using conserved, generic primer s as follows: forward primer TDKD (Slade et al. 1994): 5'-CCTGAAGTAGGAACCAGATG-3', reverse primer L15926 (Kocher et al. 1989): 5'-TCAAAGCTTACACCAGTCTTGTAAA CC-3'. A 50 L reaction contained 0.5 L tDNA, 1 X reaction buffer (Promga, Madison, WI, USA), 2.0 mM MgCl2, 0.2 mM of each dNTP, 10 pmol of each primer, 6 g BSA, and 1.25 U Taq polymerase (Promga, Madison, WI, USA). Thermocycling conditions were 94 C for 1 min, 35 cycles of 94 C for 1 min, 50 C for 1 minute, 72 C 1 min, and a final extension at 72 C for 7 min. PCR products were visualized on a 2% agarose gel with ethidium bromide to assess quantity and fidelity of amplificati on and then purified using either Microcon Centrifugal Filter Units (Mil lipore Corp., Billerica, MA, USA) or QIAquick spin columns (Qiagen, Valencia, CA, USA). A pproximately 100 ng of purified PCR product was used as template in the sequencing reac tion, and sequences in both directions were run on an ABI 3730XL automatic sequencer (Macrogen Inc., Geumcheon-gu, Seoul, Korea). Sequence and Population Analyses Forward and reverse sequences were a ligned using Sequencher software (Gene Codes, Ann Arbor, MI, USA), and edited manually when necessary. All aligned
28 sequences were analyzed using Arlequin vers. 3.01 (Excoffier et al. 2005) to estimate standard genetic indices, incl uding nucleotide diversity () and haplotype diversity ( h ; Nei, 1987). I estimated using the Bayesian, Markoff-chain, maximum-likelihood approach implemented in the program LA MARC (Kuhner 2006). All analyses used a general time reversible (GTR ) model of evolution chosen as the most likely by the Akaike Information Criteria within M ODELTEST v3.06 (Posada and Crandall 1998), two simultaneous searches with heating, a nd 100,000 steps with an initial burn-in of 10,000 steps. Analyses were repeated three tim es with different random number seeds to assess consistency. Assuming that the populatio ns are in equilibrium, female effective population sizes ( NEF) for the three specie were estimated by replacing in the equation,NEF 2 where is the mutation rate per site per generation. I evaluated evidence for historic populati on expansion using two methods. First, I used FuÂ’s FS value, which is primarily based on the differences between expected numbers of alleles (estimated through 10,000 computer simulations based on the observed pair-wise differences in my samp les) and observed numbers of alleles (Fu 1997). FS is particularly sensitive to past populat ion expansions, which typically generate large, negative numbers due to the predominance of new, rare haplotypes in the sample. I also analyzed the distribution of all pa ir-wise haplotype differences (mismatch distributions), and calculated the goodness of fit of the es timated distribution to that
29 predicted by a sudden expansion model usi ng 16,000 computer simulations (Rogers and Harpending 1992, Rogers 1995, Schneider a nd Excoffier 1999, Excoffier 2004) and associated raggedness index (Harpending 1994) as performe d in ARLEQUIN (Excoffier et al. 2005). Mismatch distributions tend to be unimodal, and smooth (i.e. wave-like) in populations that have undergone populati on size changes. Multimodal or random and rough distributions are characteristic of popul ations that have experienced long-term stability. The significance or goodness of f it of the observed data to the predicted distribution modeled for sudden expansion growth was assessed by using a sum of squares (SSD) method. When observed distri butions fit the sudden expansion model ( p 0.05), I estimated, using ARLEQUIN, the numbe r of generations si nce the expansion ( t ) from the peak of the distribution ( as t = 2 re-arranged from Li 1977) where is the rate of mutation per gene per generation. Results MtDNA Variation I obtained useable mitochondrial sequence from 365 seals, resulting in 251 unique
30 haplotypes. A 471 bp fragment of the cont rol region was aligned for 181 Weddell seals (GenBank accession numbers EU653156 Â– EU653238) a 472 nt fragment was aligned for 143 crabeater seals (EU653021 Â– EU653155) and a 481 bp segment was aligned for 41 Ross seals (EU653239 Â– EU653272). DNA fragments from Ross seals were truncated to 420 bp due to poor sequence quality and missing data in several seal s near the ends of the fragment. No polymorphic sites were obser ved in the truncated segments. There were large numbers of haplotypes in each species with 83 haplotypes id entified in the Weddell seals, with 49 of them (59.0%) found in only a single seal (singlet ons; Table 3.1). There were 33 haplotypes in the Ross seals, with 26 singletons (78.8%). Notably, there were 135 haplotypes in the crabeater seals, w ith 127 singletons (94.1%). Consequently, estimates of haplotype diversity were large a nd approaching 1.0 in all cases (Table 3.1). Effective Population Size All LAMARC analyses gave very similar results within species and unimodal distributions for the likelihood distributions indicating that I have the true probability density functions. The estimated sampling sizes (ESS) were all greater than 100 and often much larger and when combined, the ESSs were all greater than 200. The estimates of across the three separate runs within a sp ecies were within ~1% or less of the
31 combined median value. The median estimates ranged from 0.075 to 0.576 (Table 3.1). Using a mutation rate ( for the mitochondrial control re gion estimated from pinnipeds of 7.5 X 10-8 substitutions/site/year (Slade et al. 1998) and a conservative estimate of generation time of 9 years for each speci es (cf. Croxall and Hiby 1983, Bengtson and Laws 1985, Harding and Harkonen 1995, Hadley et al. 2006), the genetically effective female population sizes ranged from ~55,600 fo r Weddell seals to ~426,700 for crabeater seals, with Ross seals closer in numbe r to Weddell seals (~65,200; Table 3.1). Demographic History of Pack-Ice Seals All three seal species had significant FS values (i.e. p < 0.01) and all were less than -20.0 (Table 3.2), persuasive evidence of po pulation expansion. I also tested for population expansion using mismatch distribut ion estimates. Weddell and crabeater seals were not statistically significantly different from a unimodal mean pair-wise difference distributions ( p = 0.56 and 0.36, respectively, Table 3.2, Figu re 3.2), suggesting that they had previously increased rapidly in abunda nce. The Ross seal distribution was not unimodal ( p = 0.04, Figure 3.2) and does not indicate a ny substantial historic change in population size (Rogers and Harpending 1992, Slatkin and Hudson 1991). This trend was reinforced with HarpendingÂ’s raggedness values (Harpending 1994), which were
32 significantly different from ra gged distribution in Weddell and crabeater seals but not so in Ross seals ( p = 0.84, 0.82, and 0.05, respectively) The peak of the unimodal distribution () for Weddell seals was 5.74 corresponding to a population expansion 81,246 generations ago (731,210 years). The for crabeater seals was twice as large (12.59) suggesting that their numbers increased about 177,825 generations (1.6 million year s) ago. Both correspond to times of decreasing Antarctic temperatures and increasing seasonal ice extent (Petit et al. 1999). Discussion My assessment of genetic variation in the mtDNA control region indicates that Weddell, crabeater, and Ross seals all have levels of intraspecific haplotype and nucleotide diversity similar to, or greater th an, those reported for other marine mammals that have not been substantially harvested and that do not have strongly matrilineal population structuring (e.g., Dalebout et al. 2005, Malik et al. 2000, Westlake and OÂ’Corry-Crowe 2002). Haplotype diversity in th ese Antarctic seals is extremely high and indicates that populations of th ese species have been consiste ntly large for long periods.
33 My conclusions are conditioned, however, on a couple of caveats. First, mutation rates have consistently been difficult to estimate and likely have large margins of error. Consequently, if my estimate of mutation rate is off by as mu ch as an order of magnitude then my estimates of NEF and time since population expansion are also off by an order of magnitude. I think, however, that the mutati on rate that I used is reasonable and appropriate. I take th is rate from Slade et al. (1998) who estimated the mutation rate using mtDNA d-loop divergence and fossil reco rds of northern and southern elephant seals, the leopard seal, and the Weddell seal. As one of those is a sp ecies that I analyzed here and the others are clos ely related, I think that the estimate is not affected by phylogenetic peculiarities. Although the taxa used for the comparison were 4.5 million years divergent, at most, I also do not th ink that saturation had any effect. Slade et al. (1994) tested the relationship between geneti c distance and divergence time in these and other pinnipeds that diverged from 4.5 to 40 million years ago and found a linear relationship in all comparisons except for the mo st distantly related pairs of taxa for both transitions and transversions. A non-linear relati onship is characteris tic of saturation in the sequence data. Consequently, they concl uded that saturation wa snÂ’t present until 30 million years divergence. This means that the estimates of divergence with which the mutation rate was estimated are unlikely to be effected by homoplasy in the data. A second caveat concerns the estimate of generation time. Only a few studies have examined the reproductiv e characteristics of these sp ecies. Croxall and Hiby (1983)
34 reported an extensive survey on breeding a nd survival in Weddell seals and estimated that the average age of first reproduction was 4 Â– 5 years and after about 6 years of age 80% or more of the females have pups. Lifesp an in Antarctic seals is not generally more than 25 years and Croxall and Hiby found that over 80% of the females with pups were between the ages of 6 and 13 and the oldest reproducing female was 17. Harding and Harkonen (1995) estimated average of sexual ma turity in crabeater seals at between 3.7 and 5 years old and Bengtson and Siniff ( 1981) estimated it at 3.8 years. Hadley et al. (2006) reported average age of first reproduc tion in a long-studied population of Weddell seals in the Ross Sea (McMurdo Sound) at 7.6 years. Cons equently, I think that an estimate of generation time of 9 years is conservatively (i.e., underestimates NE) appropriate. If this were an underor overe stimate by even two years then my estimates of NEF would be around 20% lower (or higher) bu t, given the 95-percentile range, well within the estimate range of NE using 9 years. Given the above, I think that my estimates of NEF and time since population expansion are reasonable. Moreover, the re lative rank order of the estimates is not affected by errors in mutation rate or generation time estimates. Further, NEF values reflect the magnitude of th e breeding population over evol utionary time and may not correspond to contemporary values. It is im portant also to note that estimates of genetically effective population size using NEF 2 assume that the population under
35 consideration is in equilibrium. Given that I detected historic popul ation expansions for the crabeater and Weddell seals, this is unlikely to be true in these species. How far from equilibrium and what effect this might have on my estimates of NE, however, are unknown. It is reasonable to assume that the time since expansion has been relatively long (80,000 Â– 170,000 generations), the populat ions are converging on equilibrium levels. Nonetheless, that I st ill can detect the pop ulation expansion i ndicates that they have yet to completely reach equilibrium. My estimate of NEF for crabeater seals is around 426,700 females. The serially monogamous mating system of this species suggests that female and male long-term effective population sizes should be roughly e qual and thus a total genetically effective population size of less than 853,400 seals. This compares with previous visual census estimates of circumpolar abundance of 7 to 15 million (Laws 1977, Erickson and Hanson 1990). The Ross and Amundsen seas are key areas of ice habitat and a larger effective population size would be expected if the prior circumpolar es timates are accurate. In a review of 192 empirically derived estimates, Frankham (1995) showed that the ratio of NE to census size is often less than 0.5 and frequently less than 0.25. Considering just mammals (45 studies of 25 species), the NE to census size was 0.45 0.21. My estimate, however, is 0.057 Â– 0.122 (Table 3.1); up to an order of magnitude smaller than commonly seen. Avise et al. (1988), however, showed that NE could be two to three orders of magnitude lower th an census size estimates.
36 Although few assessments of population ge netic subdivision have been made, there is little indication of extant barrier s to dispersal. In th e two published genetic studies that I am aware of, there were no indications of populati on subdivision around the Antarctic for crabeater seals (Davis et al. 2000, Davis et al. 2008). As such, I would expect my estimates of NE to be reasonably close to the visual census estimates. This is clearly not the case. Several fact ors can lead to a highly reduced NE relative to census size. In spite of the genetic data, there may be some level of undetected geographic subdivision among crabeater seal s, particularly between eas t and west Antarctica and perhaps other geographically disjunct areas like the Weddell Sea. Consequently, my estimates would then reflect the number of breeding seals in the immediately sampled and geographically defined area of the west ern Amundsen and Ross seas. In Erickson and HansonÂ’s (1990) survey of seal densities in Antarctica, they defined the Ross Sea as bounded by 130 W to 160 E which roughly en closes the study area. The population estimate for the Ross Sea crabeater seal was 1.3 million individuals (Erickson and Hanson 1990). My estimate of as many as 853,400 individuals then would give a NE to census size ratio of 0.657, which is more ty pical of wild animal populations. I have, however, no indication that the area I sa mpled constitutes a separate population. Although Davis et al. (2008) found no indication of popul ation subdivision, my data set is only a partial subset of theirs and may be genetically structured. If this were the case and my samples represent more than one sub-population, I would expect that my
37 estimates of NE would be overestimates. It might also be that the mitochondrial DNA variation is underestimating the effective population size (Bazin et al. 2006), though it is difficult to address the potential for this. Pe rhaps, it may be possible through examination of nuclear loci. The high levels of haplot ype diversity do suggest, however, the mtDNA is evolving in a neutral fashion for these thr ee species of Antarctic seals. There also is little indication that crabeater seals underwent a 4-10 fold population expansion in the last 30-50 years, as sugges ted by the Â‘krill surp lus hypothesisÂ’ (Laws 1977; Mori and Butterworth 2004). If this were true, I w ould expect considerab ly lower levels of nucleotide and haplotype dive rsity. There may not, however, have been enough time for such a recent population size ch ange to be reflected in the mtDNA data. Nonetheless, fluctuations in population size, high variance in female reproductive success, or recent expansion in population size all w ould result in an estimate of NE that is substantially lower that actual population size. Extrapolating total effective popula tion size in Weddell seals may be less straightforward than a simple doubling of female numbers, which would imply a long term 1:1 sex ratio. Given the polygynous mating syst em of this species, this is unlikely to be the case. My genetically effective popul ation size estimate is 55,600 females. Male effective population sizes may be lower and t hus the total genetica lly effective population size may be less than 111,200. Even so, this number is considerably lower than the circumpolar estimate of 730,000 by Laws (1977) and ~800,000 by Erickson and Hanson
38 (1990). The geographic range of my effec tive population size estimate is not clear, though both genetic studies to date (Davis et al. 2000, 2008) do suggest that it is possible that I have sampled a subpopulation. Therefor e, it may be that my estimate of effective population size reflects a regi onal geographic scale (i.e., eastern Ross and western Amundsen seas). Erickson and Hanson (1990) estimate that there were ~50,000 Weddell seals in the Ross Sea. If I was estimating the num ber of seals in just this area, the census size would then be smaller than the geneti cally effective size; an unlikely situation. Erickson and Hanson (1990), however, do caution that their population size estimates are very conservative and likely underestimate true abundances. My estimate of 111,200 also assumes a 1:1 sex ratio, which is most likely not the case and mine is an overestimate. My samples cover approximately 60o longitude, which is less than 20% of the circumference of Antarctica. If I assume that the distribution of Weddell seals around Antarctica is uniform and I have sampled a discrete subpopulation, that would put the total effective population size at ~556,000 individuals. A lthough still substantially smaller than the direct census estimate, the NE to census size ratio is 0.67 Â– 0.76 and similar to the estimate of ~0.50 for other mammals (Frankham 1995) Not enough information is known about Ro ss seals to speculate about their breeding system, however if a 1:1 sex rati o is assumed, total genetically effective population size may be ~130,400 reproducing indivi duals, which is very near the census estimate of ~130,000 to 220,000 made by Ericks on and Hanson (1990) and Laws (1977),
39 respectively, and would result in the highest NE to census size ratio of all three species. Census sizes, however, were estimates that th e authors speculated may be underestimates due to the unknown numbers of seals in the wa ter at any given time, even at times of peak haul out, and the fact that a significan t part of the range of the seals was unsurveyed. Even so, the ratio of the gene tically effective population size to census size for Ross seals is clearly larger than the other two species (Table 3.1). There are several factors or combinations of factors that can account for this. Unlike both Weddell and crabeater seals, Ross seals did not show indications of historical population size changes. With changing population size, the genetically eff ective size is better approximated by the harmonic mean of size across generations a nd as such is substa ntially reduced and influenced by the smallest population sizes. If Ross seal populations have been fairly stable relative to the other two species th at show population expa nsions, then I would expect them to have the largest NE to census size ratio. It is also possible that nonequilibrium states for the other two speci es are resulting in an underestimation of NE. I have, however, no data with this to support or refute this argument. A total of 11,414 seals were observed dur ing the multi-disciplinary cruise in 1999-2000 when the samples reported here were collected (Ackley et al. 2003). Of the 7,781 seals that were identified to speci es, 4,817 were crabeater (53.8%), 2,852 were Weddell (36.7%), 79 were Ross (1%), and 33 (0.4%) were leopard seals ( Hydrurga leptonyx ). Though Ross seals were seen less ofte n than Weddell seals, my estimate of
40 effective population size for Ross seals was high er than that for Weddell seals. This is likely due to differences in their primary habi tats and distributions. Ross seals primarily live in pack ice habitats, where the surveys were conducted, whereas Weddell seals mostly occur close to the coast in fast ice ha bitats, at least in summer. Ross seals may be more difficult to sight than Weddell seals. A historic population bottleneck of Weddell seals followed by population expansion also woul d have resulted in a smaller genetically effective population size and may account for so me of the difference. If Ross seals are monogamous, compared to the polygynous Weddell seal, then not withstanding other influences, this would also result in larger estimates of Ross seals. This should not, however, be reflected in the maternally inherited mtDNA. Crabeater and Weddell seals have apparently increased in abundance since prior glaciations. My mismatch analysis indicated that, hi storically, crabeate r seal experienced a population expansion (Table 3.2) approximately 1.6 million years ago. The Pliocene Â– Pleistocene boundary is ~1.8 million years ago and roughly marks the start of the latest ice age. At this time sea levels were lowe r than now, and ice volume was significantly larger. Temperatures oscillated between wa rmer and colder periods on an approximately 41,000 year cycle until about 1.0 million years ago when the cycle switched to 100,000 years. Correlative changes in fast ice and p ack ice extent and seasonal tenure might have provided larger and better habitats for br eeding and molting and consequently promoted range and population expansions. My analyses indicate that Wedde ll seals increased in
41 abundance around 731,000 years ago. Expansions of fast ice and pack ice habitat may have provided more breeding habitat for Ross seals but had little influence on pelagic, non-ice habitats where Ross seals appear to sp end most of each year foraging. This may explain the results of my analyses, whic h did not unambiguousl y indicate a sudden population expansion in the Ross seal. Antarctic seals are important predators in the remarkably short food web of Southern Ocean ecosystems. Unlike most ma rine mammals, Antarc tic seals have not been substantially affected directly by huma ns, and small local harvests during the past several decades could not explain the patterns of genetic variability that I observed or the derived estimates of popul ation history. My resu lts indicate that the NE for pack ice seals is generally lower than census size but on par with what is seen for other animals. There are considerable amounts of ge netic diversity (at least for mtDNA) in all three species indicating that, genetic ally, these species are healthy (Spielman et al. 2004). Two, if not all three, of the species I studied evidently increased s ubstantially in abundance in prehistoric times correlative with expanding ice habitat during times of increased glaciation. The earlier sp eculation that crabeater seal popu lations might have increased substantially with the reduc tions in krill-eating compe titors is not supported by my analyses. The rank order of my NE estimates does not coincide with those of previous direct counts surveys suggests that populat ion size of Ross seals is not adequately estimated by surveys as traditionally conducted.
42 Acknowledgements Skin samples from Antarctic seals were collected by Brent Stewart during a multidisciplinary research cruise to the Ross and Amundsen seas from December 1999 through February 2000. Funding for this study wa s provided by an American Museum of Natural History, Lerner-Gray Grant to Ca itlin Curtis and by the National Science Foundation grants OPP98-16011 and OPP 98-16035 to BSS and DEB98-06905 and DEB03-21924 to Stephen Karl. The research was authorized by research permits 976 under the United States Marine Mammal Pr otection Act and 2000-01 under The United States Antarctic Conservation Act. The research was approved by the Institutional Animal Care and Use Committee of HubbÂ’sSeaWorld Research Institute, which is registered as a Research Facility with th e United States Depart ment of AgricultureAnimal and Plant inspection Se rvice. I thank three anonymou s reviews for extensive and thoughtful comments that greatly improved this manuscript. This is HIMB contribution No. 1334.
43 Table 3.1. Genetic diversity and female effective population sizes in pack i ce seals. Number of indi viduals (N), number of haplotypes (n) number of singleton haplotypes ( n singletons), percent singletons (%), haplotype diversity (h) mean pair wise differences between sequences () with standard deviation (SD), Ba yesian estimate of diversity () with upper and lower 95 percentile, range of female eff ective population size estimated for (EF), and the ratio of genetically effective size to census size NEN Species Parameter Weddell seal crabeater seal Ross seal N ( n ) 181 (83) 143 (135) 41 (33) n singletons (%) 49 (59.0%) 127 (94.1%) 26 (78.8%) h SD 0.98 0.04 0.99 0.01 0.99 0.08 SD 0.012 0.006 0.027 0.014 0.020 0.011 (95 percentile) 0.075 (0.055 Â– 0.085) 0.576 (0.484 Â– 0.718) 0.088 (0.060 Â– 0.132) NEF 55,600 426,700 65,200 N E N 0.139 Â– 0.153 0.057 Â– 0.122 0.592 Â– 1.0
44 Table 3.2. Estimation of sudden population expans ion in Antarctic seal s. Shown are FuÂ’s Fs test, sum of squared deviation (and significance) for the mismat ch distribution, HarpendingÂ’s ragge dness index (and significance), and number of generations (and years) before pres ent when the expansion occurred. Species Parameter Weddell seal crabeater seal Ross seal Fu Fs ( p ) -24.99 (0.00) -24.09 (0.00) -20.13 (0.00) SSD (Mismatch p ) 0.002 (0.56) 0.001 (0.36) 0.006 (0.04) Raggedness ( p ) 0.007 (0.84) 0.002 (0.82) 0.017 (0.05) 5.74 12.59 Â— Generations (years) 81,246 (731,210) 177,825 (1,600,424) Â—
45 Figure 3.1. Map of the Antarctic continent show ing the area where genetic samples were collected. Weddell Sea Ross Sea Ronne Ice ShelfCape Ponsett Amory Ice Shelf 500 Km 0 N 180 W 90 W 90 E 150 W Study Area Ross Ice Shelf Cape Adare Amundsen Sea McMurdo Station 120 W
46 Figure 3.2a. Mismatch distribution for Wedde ll seals. Bars represent observed distribution of pair-wise differences among samples and the line s hows the distribution modeled for sudden population growth. Note that the Yaxis scales differ among graphs. Conformance to the sudden growth model can only be rejected for Ross seals. 0 500 1000 1500 2000 2500 024681012141618202224262830 Pairwise Differences
47 Figure 3.2b. Mismatch distribu tion for crabeater seals. Bars represent observed distribution of pair-wise differences among samples and the line s hows the distribution modeled for sudden population growth. Note that the Yaxis scales differ among graphs. Conformance to the sudden growth model can only be rejected for Ross seals. 0 200 400 600 800 1000 1200 024681012141618202224262830 Pairwise Differences
48 Figure 3.2c. Mismatch distribution for Ross seal s. Bars represent obs erved distribution of pair-wise differences among samples and the line shows the distribution modeled for sudden population growth. Note that the Yaxis scales differ among graphs. Conformance to the sudden growth model can on ly be rejected for Ross seals. 0 25 50 75 100 125 150 024681012141618202224262830 Pairwise Differences
49 Chapter 4: Autosomal and Sex-linked Patterns of Genetic Partitioning in Three Species of Antarctic Seals Introduction The level of genetic varia tion in a population (or speci es) is shaped primarily by the forces of mutation, drift, selection and migration, and can be an important clue to the evolutionary trends in po pulation size, demography a nd long-term stability. While mutation serves to generate genetic variati on, drift tends to counterbalance it by removing variation (Hartl and Clark 1997). In the abse nce of selection and migration, a mutationdrift equilibrium should be established. The parameter theta ( is the amount of neutral genetic diversity expected at equilibrium and equals 4 NE (for nuclear genes) where NE is the genetically effective population size and is the mutation rate per gene per generation. Theta is large when there is a la rge population size or high mutation rate (or both) and conversely, small with a small NE or slow mutation rate (or both). The genetically effective population size is usually less than census size ( NC), but can provide a more useful description of the actual num bers of individuals contributing to each generation over evolutionary time. Several factors conspire to reduce NE relative to NC
50 and include deviations from 1:1 sex ratios, any form of non-random mating, variance in reproductive success, fluctuations in population size from one generation to the next, and overlapping generations (Har tl and Clark 1997). The ratio, N ENC on average, is often less than 0.5, and frequently less than 0.25 (Frankham 1995), with average mammalian ratios slightly lower than 0.45. Antarctic seals provide an ideal syst em to capitalize on the relationship of and the neutral mutation rate to estimate effectiv e population size for several reasons. Due to their relative inaccessibilit y, Antarctic seals have robus t population sizes and have enjoyed a history relatively free from anth ropogenic disturbances. Four species of Antarctic seals, the crabeater (Lobodon carcinphagus), the Ross (Ommatophoca rossii), the Weddell (Leptonychotes weddellii) and the leopard seals (Hydrurga leptonyx) are more closely related to each other than to any other species, though the relative placement of the crabeater and Ro ss seals is debated (see Higdon et al. 2007, Arnason et al. 2006). All four species appear to have dive rged relatively recently, at three points in time between 4.3 and 7.1 million years ago (mya) (Higdon et al 2007), or at the most up to ~9 mya (Arnason et al. 2006). This makes direct DNA analysis of homologous regions possible, including cross-species am plification of polymorphic microsatellites (Galbusera et al. 2000). Though they are closely re lated, the three species have significantly different life hist ory characteristics and mati ng systems, including serial
51 monogamy and polygyny. This allo ws the opportunity to make predictions about femaleonly NE estimates (mitochondrial estimates) in relation to biparentally based NE estimates (those inferred with autosomal microsatellites). Estimating NE in Antarctic pack ice seals is not only useful from a theoretical perspective, but population abundance is a critical, but poorly understood, element of Antarctic ecology. Species abunda nce is particularly important due to the uniquely short Antarctic food web and its implications on mana gement models and plans. For example, one of the species in my study, the crabeate r seal, feeds exclusiv ely on Antarctic krill (Euphausia superba), which is also the primary diet of the six species of baleen whales found in the Antarctic. An abundance or s hortage of krill, through natural causes or commercial overharvesting (e.g. Bester 1995) or dramatic population si ze fluctuations of baleen whale stocks might result in changes in crabeater seal abundan ce, or vice versa. The decline and near collapse of krill-eating baleen whale populations in the early to mid 20th century (Macintosh 1970) was hypothesized to have spawned a significant increase in crabeater seal abundance (Laws 1977, Mori and Butterworth 2004). In a 2005 report, the International Whaling Commission highlig hted the importance, and deficiency, of abundance and population trend data of the crabeater seal to the understanding of trophic interactions with Southern Ocean whale populations (International Whaling Commission 2005). My research focused on three species of Antarctic seals that have circumpolar
52 distributions and occupy distin ct niches. More is known a bout the Weddell seal than the other two species of seals, due to a br eeding population near Ross Island, McMurdo Sound, which has been extensively studied since its discovery in 1907 (Wilson 1907), and to annual tagging and census studies whic h have been carried out since 1969 (Stirling 1969, Cameron and Siniff 2004). Though they are found in pack ice, Weddell seals typically inhabit the landfast ice regions around the continent. The extent of fast-ice fluctuates annually due to freeze patterns, changing influence from wind and tides, and glacial movement (Ushio 2006). Breeding-age adults return in the Austral spring to breeding-birthing colonies situated in the fa st ice, to which they demonstrate a limited degree of philopatry. Davis et al. (2008) used a genetic appro ach to show some degree of genetic differentiation between populations around the continent and demonstrated a weak signal of isolation by distance. Copulation takes place under the ice (Cline et al. 1971). Males use their teeth to maintain breat hing holes in the ice, and they actively guard the 3-dimensional territory surrounding these holes (Kaufman et al. 1975, Bartsh et al. 1992). This allows males to mate with as many females as may share the breathing hole and to thwarting access to these fema les by competing males. Genetic studies indicate that males su ccessfully mate and produce pups with up to five females within a season (Gelatt et al. 2000). Successful males (i.e. thos e able to reproduce) have a mean seasonal reproductive success of 1.2 with a variance of 1.6 (Gelatt et al. 2000). Population size census estimates for this species range from 730,000 (Laws 1977) to
53 800,000 (Erickson and Hanson 1990) individual s. A mitochondrial DNA based female genetically effective population size ( NEF) estimate was ~55,600, suggesting as many as ~111,200 total individuals (Curtis, et al. 2009). While significantly lower than census estimates, this number may reflect a regiona l estimate encompassing roughly 20% of the continent (eastern Ross and western Amundsen seas), which was the approximate sample collection area (Curtis et al 2009). If this were the case, the total NE may be as high as ~556,000 individuals around the co ntinent and would give a 2 N EFNC ratio of 0.70 Â– 0.76, similar to the ~0.5 estimate for other mammals (Frankham 1995). It is also possible that mtDNA estimate represents the circumpolar population size but the mitochondrial DNA, per se is significantly underestimating NEF (Bazin et al. 2006). In any case, given the potential for philopatry (and higher potent ial for inbreeding) and a polygnous mating system, I would expect that the ratio of neutral, biparentally inherited nuclear DNA estimate of NE ( NEA) to 2 NEF would be smaller for Weddell seals than for the other species where these attributes are not pres ent. In other words, the Weddell seal N EA2 NEF ratio should be less than that for crabeater or Ross seals. Crabeater seals (L. carcinophagus) are distributed throughout the perimeter of the Antarctic continent, and young seals occasionally travel in large groups and sometimes for great distances. No significant level of population genetic subdivision has been
54 detected (Davis et al. 2000, 2008), suggesting a single, pa nmictic population likely due to high levels of gene flow around the continent. Crabeater se als utilize the floating pack ice for breeding, and are thought to be serially monogamous (Siniff et al. 1979). Population size has been estimated at 7 to 15 million individuals (Laws 1977, Erickson and Hanson 1990). An estimate of NEF was ~426,700, suggesting as many as ~853,400 total individuals (Curtis et al. 2009). Thus a 2 N EFNC ratio of 0.057 Â– 0.122; an order of magnitude smaller than commonly seen. Assumi ng crabeater seals are a single, panmictic population and are serially monogamous with a 1:1 sex ratio, I would expect N EA2 NEF 1.0. Comparatively little is k nown about the Ross seal ( O. rossii ) They appear to be solitary, and utilize the stable ice floes on the ex terior of the pack ice to molt, while they give birth on the more densely packed interior pack ice (Splettstoesser et al. 2000, Stewart 2007). Though they have a circum polar distribution, hi gher numbers may be found in the Ross Sea, King Haakon VII Sea, and perhaps portions of the western Weddell Sea (Stewart 2007). They may spend mu ch of the rest of their time to the north of the pack ice, alone in the open water (Stewart 2007). There is no indication of population genetic subdivision (Davis et al. 2008). The mitochondrial DNA based NEF, was ~65,200, suggesting as many as ~130,400 total individuals (Curtis et al. 2009), which is very similar to the circumpolar census estimates of 131,000 to 220,000 (Laws
55 1977, Erickson and Hanson 1990). Not enough info rmation exists to speculate on their mating system, nor the ratio of NEA to NEF. To gain a better understanding of the rela tionship of female genetically effective population sizes to the overall genetically effective population sizes within and among these three species, I compare and contrast NEA based on biparentally inherited microsatellites with previous ly published mitochondrial 2 NEF and census size estimates. In addition, I examined one X-linked micr osatellite and two Y-linked DNA sequences totaling approximately 740 nt. I also asse ss long-term population st ability by looking for signs of recent, sustai ned population bottlenecks. Methods Samples Tissue samples were collected fr om 214 free-living Weddell seals (L. weddellii), 175 crabeater seals (L. carcinophagus) and 41 Ross seals (O. rossii) via remote darting or direct handling during the Antarctic Pack Ice Seal (APIS) multidisciplinary cruise (Decker et al. 2002) in the Austral summer of 2000. A lthough all three species are circumpolar, most samples were collected from the pack-ice zone of the western Amundsen and eastern Ross seas, approximately 67 Â– 78 South and 129 Â– 180 West
56 (Figure 4.1). Some L. weddellii samples came from McMurdo Sound, in the western Ross sea. Collection date and location data was avai lable for all individuals. Sex and other biometric data were available for some but not all individuals. Se x of most individuals was determined or confirmed using ZFX and ZFY genetic screening (Curtis et. al., 2007). Samples were either stored in ethanol or frozen at Â–80 C until DNA extraction. Laboratory methods Autosomal and X-linked Microsatellites DNA was purified using standard phenol -chloroform techniques or using a DNeasy Tissue Kit (Qiagen, Valencia, CA), a nd 5 l of total cell DNA was visualized on a 0.8% agarose gel stained with ethidium bromide to assess DNA quality. Dinucleotide microsatellites were amplified us ing the primers designed by Davis et al. (2002), and forward primers were fluorescently labe led (Integrated DNA Technologies, Inc., Coralville, IA). Polymerase chain reaction (PCR) amplification products were visualized on a 2% agarose gel stained with ethidium bromide to assess efficiency and fidelity of amplification. Microsatellite PCR amplifica tions took place in th ree multiplex reactions pooling loci Lw18, Hl15, Hl16, and Hl20 (annealing temperat ure of 56 C), Lw10, Lw11,
57 and Lw16 (57 C), and Hl4, Lc5, and Lw8 ( 55 C). Lw18, Hl15, Hl16 and Hl20 were screened simultaneously on an ABI 3100 Gene tic Analyzer (Iowa State University Sequencing Facility), with the remaining six loci being pooled and screened separately. DNA amplifications were generally a 10 l reaction containing 0.5 l total cell DNA, 1 X reaction buffer (Promega, Madison, WI), 2.0 mM MgCl2, 0.2 mM of each dNTP, 10 pmol of each primer, 6 g bovine serum albumen, and 1.25 U Taq polymerase (Promega). Thermocycling conditions for microsatellites follow ed a standardized protocol of 94 C for 1 min, 30 cycles of 94 C for 45 sec, annealing temperature for 45 sec, and 72 C for 1 min followed by a final extension of 72 C for 7 min. Y Chromosome Sequences I amplified a portion of the 8th intron of the Y-linked DEAD-box Y gene (DBY8) using the mammalian primers DBY8F 5'Â–CCCCAACAAGAGAATTGGCTÂ–3Â’and DBY8R 5'Â–CAGCACCACCATAKACTACAÂ–3' (Hellborg and Ellegren 2004). I amplified approximately 560 nt of the 11th intron of the Y-linked Ubiquitously Transcribed Tetratricopeptide Repeat gene (UTY11) locus using the mammalian primers UTY11F: 5Â’Â–CATCAATTTTGTAYMAAT CCAAAAÂ–3Â’ and UTY11R: 5Â’Â– TGGTAGAGAAAAGTCCAAGAÂ–3Â’(Hellb org and Ellegren 2004). Amplifications were
58 as previously described except they were 50 l reactions. Thermocycling conditions for UTY11 followed a touchdown protocol of 94 C for 1 min and 35 cycles of 94 C for 1 min, an annealing temperature for 1 min star ting at 55 C and decreasing 0.25 C every cycle, and 72 C for 1 min followed by a final extension at 72 C for 7 min. Thermocycling conditions for DBY8 followed a similar touchdown pr otocol, except that all cycles except the final extension were 30 seconds and the touchdown annealing cycle started at 65 C. PCR products were visu alized on a 2% agarose gel stained with ethidium bromide to assess efficiency a nd fidelity of amplification. For all Y chromosome specific primer sets, two female samples were also amplified in each batch of male amplifications to serve as a negative control. PCR amplification products were purified using either Microcon Centrifugal Filter Units (M illipore Corp., Billerica, MA) or QIAquick spin columns (Qiagen, Inc., Vale ncia, CA) and quantified on an agarose gel stained with ethidium bromide. Approxima tely 100 ng of purified product was used as template in sequencing reactions and sequen ced in one or both directions on an ABI 3730XL sequencer (Macrogen Inc., Seoul, Sout h Korea or the University of South Florida Sequencing Facility, Tampa, FL). Due to the minimal amount of observed variation, sequences were sequenced in both directions only when there was any ambiguity or potential variable sites. Forward and revers e sequences were aligned and joined using Sequencher software (GeneCode s Corp., Ann Arbor, MI), and edited, when
59 necessary. Multiple sequences were aligned using Clustal W (v.1.83, Chenna et al. 2003). Data Analyses All loci were analyzed using Genescan so ftware (Applied Biosystems, Inc., Foster City, CA). Number of alleles and hete rozygosity were estimated using the Excel microsatellite toolkit v3.1 (Park, 2001). Ha rdy-Weinberg exact te st and deviation from equilibrium frequency and linkage disequilibri um estimated for all autosomal loci were done using GENEPOP v3.4 (Raymond a nd Rousset 1995). The program BOTTLENECK (Piry et al. 1999) was used to look for signs of recent bottlenecks (heterozygosity excess) within each species. It is most sensitive in detecting bottlenecks 2 NE Â– 4 NE generations in the past depending on th e severity of the bottleneck and the mutation rate of the loci being analyze d. This program calculates expected heterozygosity at mutation-drift equilibrium ( HEQ) based on the number of alleles at a locus and sample size using the infinite alleles model (IAM), the stepwise mutation model (SMM), and the two-phase model (TPM, an intermediary between the two). The HEQ estimates are averaged across loci and compared with the observed levels of heterozygosity, with a null hypothesis of no hete rozygosity excess. I did not use the IAM
60 model because the SMM and TPM are most appr opriate for evaluating microsatellite data (DiRienzo et al. 1994, Luikart and Cornuet 1998, Pi ry 1999). I allowed 95% single stepwise mutations and 5% multi-step mutati ons in the TPM, with a 12% variance among multiple steps, as recommended by Piry et al. (1999). I focused on the WilcoxonÂ’s signed rank test, which is suggest ed by the authors to be the mo st powerful with < 20 loci (minimum of four loci required). Locus specific F values were estimated for all species at the ten loci (Xu and Fu 2004) using the ThetaF program (H. Xu, pers. comm.). To estimate the effective population size ( NEA), the mean of F across all nine autosomal loci within each species was used as a surrogate for in the equation NEA = 4 where is the mutation rate per gene per generati on under the assumption of mutation-drift equilibrium. Because locus Lw18 is X-linked (Davis et al. 2002), the sex of all individuals was determined a priori through a combination of field and lab screening (Curtis et al. 2007) and females possessing a single Lw18 allele were designated as homozygotes. The equation NEX = 3 (Yu et al. 2002) was used to estimate the genetically effective X-linked population size ( NEX) for Lw18 using female samples only. Several microsatellie studies of mammals, including seals, have estimated or assumed mutation rates of 10-3 Â– 10-6 mutations per gene per gene ration (Schlotterer 2000, Palo et al. 2001, Kretzmann et al. 2006, see Ellegren 2000 for review). Here, I use10-5 to
61 estimate effective population size because it is the most frequently cited rate, is approximately in the middle of the cited ra nge, and faster or slower rates change the genetically effective population size linearly. Results Microsatellites All nine autosomal loci successfully amp lified product in each species. Table 4.1 shows the number of individuals, number of alleles, and observed and expected heterozygosity at each locus for each species. L. carcinophagus showed the highest number of alleles and the highest heterozygosity at most loci and L. weddellii had the lowest heterozygosity (though not always the lo west number of alleles) at most loci (Table 4.1). Allele sizes and numbers of alleles at each locus were consistent with those observed in the same species by Davis et al. (2002, Table 4.1). The mean sizes of alleles did not differ significantly in pairwise comp arisons among species (S tudentÂ’s one-tailed t-test, data not shown). Some loci did not amplify in ever y individual. All loci in all species were in Hardy-Weinberg genotypi c frequency equilibrium except for two in L. carcinophagus (Table 4.1). Two sets of lo ci were statistically linked in L. weddellii and O. rossii, and four sets of loci were linked in L. carcinophagus (data not shown). The
62 linked loci differed, however, among species; ther efore I include all av ailable loci in the analyses. I did not detect significant heteryoz ygosity excess in any species using BOTTLENECK (Piry et al. 1999). Under the SMM or the TPM, the Wilcoxon test showed no patterns of excess heterozygosity in any species. The data do suggest heterozygosity deficiencies in all three species under the SMM, though under the TPM, L. carcinophagus was not significant ( P = 0.29), while the other species were ( P = 0.00 Â– 0.01). Estimates of F vary widely across loci within a species and across species at a locus (Table 4.2). Locus Lc 5 had, by far, the lowest F estimates for all species. The autosomal arithmetic means of F were 5.91, 13.80, and 37.58 for L. weddellii O. rossii and L. carcinophagus respectively (Table 4.2). Using a mutation rate of 10-5 resulted in genetically effective population sizes of 147,850 to 939,600 across species (Table 4.3). For the X-linked microsatellite locus, Lw 18, no female homozygotes were observed in O. rossii, although they were seen in the other two species. For L. carcinophagus and O. rossii, the F estimates at Lw18 were larger than the arithmetic mean across autosomal microsatellites, though this was not the case for L. weddellii (Table 4.2).
63 Y Chromosome Variation Sequences from both Y specific primer pair s aligned with high sequence identities to Y-specific regions in the National Ce nter for Biotechnology Information human genome database. UTY11 aligned with 83% identity to Homo sapien Ubiquitous TPRmotif Protein Y isoform (UTY) gene (Gen ebank accession number AF265575, Shen et al., 2000). DBY8 aligned with 84% identity to a Y chromosome region encompassing part of a DEAD-box protein gene, exons 9 a nd 10 and the intron between them (Genbank accession number AC004474, Birren et al. 1999 unpublished). DE AD-box proteins, named for the presence of the amino acid sequ ence Â‘D-E-A-DÂ’ (motif II or the Walker B motif), play an important role in RNA me tabolism (Linder 2006). These high sequence identity alignments, coupled with sex-restricted amplifi cation (in males only), support that my sequences were from the Y chromosome. I sequenced a total of 224 male seal s for the DBY8 locus, including 26 O. rossii 88 L. carcinophagus and 110 L. weddellii, discovering five new haplotypes, two in L. weddellii, and two in L. carcinophagus (Genbank accession numbers FJ813489FJ813491) The two L. weddellii haplotypes differed by one G/C transversion, and the rare haplotype was shared by two, non-nuclear family individuals. These two individuals were genetically similar at the autosomal lo ci, but clearly had di fferent mothers (i.e., mtDNA haplotypes) and fathers (i.e., di fferent DBY8 sequences). The two L.
64 carcinophagus haplotypes differed by one A/G tran sition, and the rare haplotype was observed in a single individual. The DBY8 sequence, excluding primer sequences, was 141 nt in all three species. A ll three seal species differed by at least one base pair at DBY8 (Appendix 1). I sequenced a total of 196 male seal s for the UTY11 locus, including 25 O. rossii 78 L. carcinophagus, and 93 L. weddellii discovering four haplotypes (Genbank accession numbers FJ813492-FJ813494). Ther e were two haplotypes in L. weddellii differing by one A/G transition. No variation was found in O. rossii or L. carcinophagus at either locus. When comparing the seque nces among species, there are eleven variable sites, including six transiti ons, two transversions, and three indels (Appendix 1). Discussion My assessment of genetic variation in autosomal microsatellites suggests that Weddell, Ross, and crabeater seals have levels of intraspecific variation similar to those reported for other large mammals (Paetkau et al. 1995, Simonsen et al. 1998). The levels of diversity also fit into the range of published estimates for pinnipeds (Table 4.4), though direct comparisons should be made with caution due to differences in experimental design and populat ion sampling among studies. At the low end of the rage,
65 the critically endangered Mediterranean monk seal ( Monachus monachus ) whose population size has declined to an estimate maximum of 430 indivi duals divided between two remaining populations, has published hete rozygosity estimates ranging from 0.23 to 0.35 (Pastor et al. 2007). A similarly low level of he terozygosity is seen in the Hawaiian monk seal, Monachus schauinslandi (0.48) which has a current population size estimated of 1,247 seals (Schultz et al. 2008). Of the species surveyed for this st udy, the Weddell seal has the lowest heterozygosity (Table 4.4). Similar levels were reported for northern European harbor seals, Phoca vitulina, (Goodman 1998, Table 4.4), whose population was nearly halved in 1988 due to a severe outbreak of phoc ine distemper virus (Goodman 1998), though it appears to have rebounded. Antarctic seals are thought to be minimally anthropogenetically impacted and the five lo west heterozygosity levels seen are from northern hemisphere seals (Table 4.4). Given the remoteness of the Antarctic, it seems unlikely that the Weddell seal has experien ced significant recent population reductions unless they were naturally caused. An asse ssment of mtDNA in Weddell seals (Curtis et al 2009) did, however, indicate that the Weddell seal went through a population expansion approximately 731,000 years ago possi bly due to glaciation induced increases in fast and pack ice around Antarctica. None theless, given that microsatellites mutate very quickly, I expect that th ere has been sufficient time in the ~81,000 generations since the population expanded for several new alleles to arise. High heterozygosities, similar
66 to those in the crabeater seals tended to be associated with seals of high population abundance (but see Twiss et al. 2006). Antarctic fur seals ( Arotocephalus gazella ) are estimated to number 4 Â– 7 million and (S cientific Committee on Antarctic Research, 2006) and have observed heterozygosities of 0.81, comparable to the crabeater frequency of 0.78. Interestingly, Curtis et al. (2009) showed that the crab eater seal went through a similar population size expansion. This expa nsion, however, occurred much earlier (i.e., ~1.6 mya) and it is possible that any loss of heterozygosity would have been regained. Heterozygosity for the Ross seal is high, and falls between that seen in the Weddell and crabeater. Interestingly, the ra nk order of heterozygosities ma tches the rank order of the genetically effective population size s estimated from mtDNA (Curtis et al. 2009) and autosomal microsatellites (see below). I found no evidence suggesting that any of the species in my study underwent a recent population bottleneck (data not shown) or at least not one in the past 2 NE to 4 NE generations, the reported dete ction range of the statistical analyses A slow, gradual population decline would be unlik ely to result in a statistically signifi cant heterozygosity excess, and would not likely be detected using this type of analysis. Assuming NEs of 111,200 for L. weddellii, 130,400 for O. rossii, and 853,400 for L. carcinophagus (Curtis et al. 2009), and an average age at reproduction (i.e., generation time) of 9 years (Curtis et al. 2009), the timeframe I examined w ould roughly correspond to 2 Â– 4 mya for L. weddellii 2 Â– 5 mya for O. rossii and 15 Â– 30 mya for L. carcinophagus If the average
67 age at reproduction or NE estimates are off, this time frame would shift proportionally forward or backward in time, but the detec tion a bottleneck would not be effected. The BOTTLENECK analysis does indicate heterozy gosity deficiencies (not shown) in all three species under the SMM and for two sp ecies in the TPM, which suggest the populations may have experien ced a population expansion at some point in the recent past (Luikart and Cornuet 1998). This is co nsistent with the previous mtDNA estimates (Curtis et al. 2009) and the time of expansions are well within the above time frames. Mitochondrial DNA analysis, however, did not unambiguously reveal a population expansion for O. rossii (Curtis et al. 2009) It is possible that if this species underwent a significant expansion, it was smaller in scale than the others, or that it took place in a time range undetectable by mitochondrial si gnals. Nevertheless, as expected, these populations of Antarctic seals ha ve not experienced dramatic declines in population sizes and are more likely to be increasi ng in size over evolutionary time. Before I turn to estimates of genetically effective population si ze, I would like to address several caveats. It is possible that my genetic resu lts include ascertainment bias in the microsatellite loci which can artificially deflate heterozygosity and F. The microsatellite markers that I used were origin ally isolated from the three, closely related species, L. carcinophagus (Lc5), L. weddellii (Lw10, Lw11, Lw16, and Lw18) and the leopard seal, Hydrurga leptonyx (Hl4, Hl15, Hl16, and Hl20) The leopard seal is believed to be the sister taxa of the Weddell seal. It has b een shown that microsatellite
68 loci can show reduced heterozygosity and numbers of alleles when applied to species from which they were not originally isolat ed (i.e., cross species applications, Hutter et al. 1998). They also tend to have shorter pure re peat regions (i.e., are interrupted, Amos et al. 2003, Vowels and Amos, 2006) and shorter overa ll allele lengths in non-focal species (Amos et al. 2003). I do not, however, think that a ssertainment bias is significantly affecting the analyses. One-tailed student t-test (data not show n) did not indicate statistically significant differe nces in the number of allele s in crossversus non-cross species applications. Of th e 30 locus by species comparis ons, six are non-cross species applications. Of these six, the lowest relative per-locus allelism is seen in five of them and in all cases, the highest allelism is seen in cross-species applications. The same pattern generally holds for heterozygosity leve ls, as well (Table 4.1). Much of this pattern is due to the crabeater seal showing overall higher levels of variation then the other two species likely due to a larger population size. Even so, if the crabeater seal is excluded the highest levels of heterozygosity and alle lism are still seeing in the cross species applications. A second caveat concerns the mutation rate As with all populat ion size estimates, those based on microsatellite diversity are reliant on mutation rates. Unlike most DNA sequences, however, microsatellite mutation rate estimates may vary by up to four orders of magnitude. Mammalian microsate llite mutation rates range from 10-2 to 10-6 (Schlotterer 2000). Other studies using microsatellie have es timated or assumed mutation
69 rates of 10-3 Â– 10-6 mutations per gene per gene ration (Schlotterer 2000, Palo et al. 2001, Kretzmann et al. 2006. See Ellegren 2000 for review ). For my analyses, I used a mutation rate in the middle of this range (10-5, Table 4.3). It is impor tant to note that if my mutation rate estimate is o ff by an order of magnitude the NE estimates would also be off by an order of magnitude. This woul d not, however, affect the rank order of population estimates among the species. I do, ho wever, believe that the mutation rate I am using is a good approximation of the aver age mutation rate acro ss the microsatellite loci in my study and is one that is commonly seen in large mammals. Estimates of F vary widely across loci within a species and across species at a locus (Table 4.2). Locus Lc05 ha d, by far, the lowest overall F estimates. In Le weddellii and O. rossii this is likely due, in part, to a sm all number of alleles and very low heterozygosity levels. This ca nnot, however be the case for L. carcinophagus where Lc05 had the lowest F but 12 alleles. In this case, the low F was probably driven by highly unequal allele frequencies wher e one of the 12 alleles was se en at a frequency of 60%, and the next highest frequency was 1%. In general, however, single loci are poor estimators of and the accuracy of the overall estimation of NE is proportional to the number of loci (Felsenstein 2005). As such, most of the remainder of the discussion will focus on estimates using all autosomal loci combined. The rank order of microsatellite NEAs are in agreement w ith the mitochondrial
70 estimates for the three species (Curtis et al. 2009), with L. weddellii showing the smallest effective population size, followed by O. rossii and L. carcinophagus the largest (Table 4.3). This is something of a surpri se because the census estimates for O. rossii typically are lower than those of L. weddellii (Table 4.3). Even so, th e agreement between the two types of markers (and, generally, at each micros atellie locus, individually) indicates that this is likely to be a genuine, biological at tribute of the species. It is possible that the census sizes accurately reflect the true populat ion sizes, but a sustained reduction in size occurred for L. weddellii lowering NEA for this species. I did not, however, recover any indication that any of the species have under gone bottlenecks (data not shown). It is also interesting to note that O. rossii has a larger overall NEA estimates than the Weddell seal as well as on a locus-by-locus ba sis. As with the mtDNA estimate, NEA for the Ross seal is nearly as large if not larg er than the census size (Curtis et al. 2009); an unlikely outcome. This supports the statement made by Curtis et al. (2009) that Ross seal population sizes may be significantly undere stimated by traditional census survey techniques. While the rank order of NE estimates remains consis tent between mitochondrial and microsatellite analyses, the estimates of NEA were not the same as 2 NEF. For the most part, however, NEA and NEF are fairly similar and both much lower than the census size. For L. carcinophagus, NEA = 3939,600 and N EA2 NEF = 0.91 (Table 4.3) indicating that both
71 types of markers agree that the NE for this species is ~1,000,000 individuals. The per locus estimates of NEA are also fairly consistent and i ndicate an effective size of about 1,000,000 individuals (Table 4.2). This is, how ever, significantly lower than the published circumpolar census estimates of 7 to 15 million individuals (Laws 1977), and results in an N EANC ratio of 0.06 to 0.13, much lower th an typically observed in mammals (Frankham 1995). There is a similar pattern for Le weddellii with an N EA2 NEF of 0.75 and N EANC = 0.185. There are several factors that result in NE being lower than NC. In the Curtis et al. (2009) study, it was suggested that mtDNA might be underestimating the true NE for these species. Given the general ag reement of the nuclear and mitochondrial results, this is unlikely. Furt her, since mtDNA estimates only the female effective size, a 1:1 sex ratio was assumed and NEF was doubled to approximate the full genetically effective population size. For the crabeater se al, this appears to have been a supportable assumption. If the sex ratio was not 1:1, then I would expect NEF > NEA. Although these two estimates are not identical they differ by only very little and ar e likely well within each otherÂ’s range of deviation. The Weddell seal is know to be poly gynous and as such, it is expected that N EA2NEF < 1.0 because, although all females can mate each season, not all mature males
72 can. Here, again the estimates are fairly sim ilar, although the nuclear estimate is larger than the mitochondrial. There is no questi on that Weddell seals are polygynous so an explanation for this reversed results must lie elsewhere. It is po ssible, as mentioned previously, that the mtDNA can be underestimating NEF. I doubt that this is generally the case given the similarity between the marker s results and an expected pattern in the crabeater seal. Any underestima tion would also have to be species specific since the mtDNA and microsatellite estimates for L. carcinophagus are remarkable similar as expected. It could be that the mutation rate I chose for the mi crosatellites is too slow and thus is overestimating NEA. Changing the mutation rate from 1.0 X 10-5 to 2.0 X 10-5 lowers NEA for Weddell seals to 73,900. This then would make NEA < NEF, as anticipated. This also, however, would result in N ENC = 0.092, a value nearly an order of magnitude lower than is typically seen in mammals (Frankham 1995, but see Avise et al. 1988). This change, however, would also have to appl y to the other species and would reduce NEA for the crabeater seal to approximately half of 2 NEF estimated from mtDNA where NEA is expected to, and does, equal 2 NEF. It would also further reduce the census size to genetically effective size ratio in crabeaters making it even more unlike what is seen in mammals. It is possible that the microsatellites are overestimating NEA in L. weddellii There is considerable variation among the per-locus estimates of F (0.01 Â– 28.34, Table 4.2), although most of the loci have values less than 10.0. It is possible that some form of
73 diversifying selection is ac ting on some loci but not ot hers (Kashi and Soller 1999, Barker et al. 2009 ). If I remove th e locus with the largest F value (Lw10), NEA = 77,800 (Table 4.2) for the Weddell seal and is now about 70% less then the value of NEF, as would be expected from a polygynous sp ecies. There is, however, a corresponding reduction in the genetically effective and cen sus size ratio. Even so, I think that the NEA estimated excluding Lw10 is likely more accurate (reduced data, Table 4.3). As with the mitochondrial estimate, the autosomal NE for Weddell seals is lower than the published census size estimates of 730,000 (Laws 1977) to ~800,000 (Erickson and Hanson 1990), and results in an N ENC ratio of 0.10 Â– 0.11, also lower than typically observed. Given that population subdivisi on was noted in samples separated by more than 700 km (Davis et al. 2008), it is possible that I am sampling a regionally subdivided population, and therefore the NEA and NEF may be regional estim ates encompassing only the western Amundsen and eastern Ross seas. In their 1990 survey, Erickson and Hanson estimated that there were ~50,000 Weddell seals in the Ross Sea. If I were estimating the number of seals in just this area, the census size would then be smaller than the genetically effective size; an unlikely s ituation. Erickson and Hanson (1990), however, do caution that their population size estimat es are very conservative and likely underestimate true abundances. My samples cover approximately 60o longitude, which is less than 20% of the circumfere nce of Antarctica. If I assu me that the distribution of
74 Weddell seals around Antarctica is uniform a nd I have sampled a discrete subpopulation, that would put the total eff ective population size at ~739,250 individuals, which roughly corresponds to the direct census estimates. The Ross seal autosomal estimate of near ly 350,000 is enigmatic for two reasons. First, NEA for the Ross seals is 2.6 times larger than 2 NEF. Second, NEA is larger then the census size by a factor of ~1.6 (Table 4.2), an unlikely result. As with the Weddell seal, there are several factors that could account for this. Again, mitochondrial estimates may simply be underestimates of NE (Bazin et al. 2006), though this would not explain why autosomal NE exceeded the published census estimates (Erickson and Hanson 1990). Unlike both the Weddell and crabeater seals, O. rossii did not show a clear sign of population expansion with mitochondrial DNA (Curtis et al. 2009). It did, however, show a heterozygosity deficit in BOTTLEN ECK, suggesting that they may have experienced a more recent population expans ion, detectable with faster mutating microsatellite markers, but not yet reflect ed in the mitochondrial DNA. A sudden, recent population expansion, however, would re sult in a smaller, not larger, NEA. This also does not, however, explain why NEA is so much larger than NC. As with the Weddell seal, I believe that some of my data are obscuring th e accurate results. The per locus estimates of F seen in the Ross seal also have a cons iderable range of values (0.04 Â– 57.53, Table 4.2). Most loci, however, i ndicate small values of F and are generally less than 11.0
75 (Table 4.2). If I remove the two autosomal loci with the largest F values (Lw10 and Lw11, 57.53 and 34.18, respectively), NEA = 116,000 and is only slightly smaller than NEF. This would also mean that both NEA and NEF give genetically effective to census size ratios similar to other mammals of 0.53 a nd 0.59, respectively. If these new estimates are accurate I could then conclude that the Ross seal, like the crabeater, likely has a serially monogamous mating system with an even sex ratio. Genetically effective population size ( NEX) estimated with the single Xchromosome marker (Lw18) were roughly in line with or slightly higher than both mitochondrial and autosomal estimates of NE for L. weddellii and L. carcinophagus (Table 4.2). As mentioned ear lier, caution should be used with inferences based on a single locus. The NEX of O. rossii however, was over 5 million, which is more than an order of magnitude higher than the NEA or 2 NEF. I am at a loss to adequately explain why this estimate is so different from all othe r estimates and believe that it is anomalous. Forces, such as selection, tend to have a greater impact removing va riation in autosomes than in X chromosomes and hitchhiking may be a more powerful force on the X chromosome (Betancourt et al. 2004). Using a total of 133 microsatellites, Kauer et al. (2002) found statistically elevated averag e heterozygosity on the X chromosome ( H = 0.75) than autosomes (0.62) in African Drosophila populations, and this trend has been seen in other studies as well (e.g. Andol fatto 2001, Payseur and Nachman 2002). Until more X-linked microsatellites or other X-li nked markers are surveyed, I cannot know if
76 this is a quirk of a single locus or characteristic of the chromosome. I found a limited intraspecific variation fo r Weddell seals and crabeater seals at the Y-linked DBY8 and UTY11 loci. There wa s, however, some intraspecific variation within Weddell seals at th e UTY11 locus. Due to their polygynous mating system, however, male L. weddellii should have a reduced effective population size than females, and I expected to find reduced le vels of male diversity among L. weddellii relative to mitochondrial diversity as well as to L. carcinophagus which is not the case. Overall, however, the level of variation is quite low and no statistica l analysis of the data is possible I did not find any variation in O. rossii Y chromosome loci, most likely due to small sample size and the apparent slow evol utionary rate for this marker. Given the higher diversity patterns (than L. weddellii ) shown with both mitochondrial and microsatellite analyses for the Ross seal, I would expect further screening with Y chromosome markers to reveal genetic variat ion. I expect that further screening with these and other Y chromosome markers would reveal more genetic variation within all three species, which could render them useful in paternity analys is. These loci can, however, be used to identify samples to the species level and for future Y chromosome phylogenies of pinnipeds a nd higher order mammals. In summary, my autosomal NEA estimates were roughly in agreement with previous 2 NEF estimates with mitochondrial DNA for L. weddellii and L. carcinophagus though both are less than census es timates and in the case of L. carcinophagus quite
77 significantly. This may suggest that some le vel of yet undetected circumpolar population structure exists for this species and my estimates are for a localized area. O. rossii NEA was higher than that estimated with mitoc hondrial DNA, as well as being greater than census size estimates. The latter, however, is unlikely to be true and non-neutral forces on two loci may have resulted in anomalously high F values. If I remove these loci from the analysis, the Ross seal NEA estimate is less than the NC and nearly equal to 2 NEF. Similarly, when I removed one locus with an exceptionally large F value from the analysis of the L. weddellii autosomal estimates of NE were, as expected for a polygynous species (i.e., lower than 2 NEF). I find that an approach integrating genetic markers with different modes of inheritance (biparental, maternal, and paternal) as well as varying mutation rates is useful in focusing on evolut ionary trends occurring at varying points in the evolutionary history of a species. Concordance among markers and expectations based on inheritance mode can more powerfully reveal biologically meaningful demographic patterns.
78 Acknowledgements This research was funded by an American Museum of Natural History LernerGray grant to Caitlin Curtis, NSF O PP 98-16011 and OPP 98-16035 grants to Brent Stewart, and NSF DEB 98-06905 and DEB 03-21 924 grants to Stephen Karl. I would like to thank Hongyan Xu for generously providi ng statistical analysis software for theta estimations.
79 Table 4.1. The number of individuals ( N ), expected ( Hexp) and observed ( Hobs) heterozygosities, and number of alleles ( A ) seen at microsatellite loci in three Antarctic p ack-ice seals. Observed hetero zygosity values in bold were statistically significant de ficits ( p 0.01). Species Locus Le. weddellii Hl4 Hl15 Hl16 Hl20 Lc5 Lw8 Lw10 Lw11 Lw16 Lw181 Average SD N 188 184 192165143184141142 186 79(108) 163.3 28.3 Hexp 0.26 0.59 0.630.630.040.290. 880.81 0.78 0.68 0.56 0.27 Hobs 0.24 0.61 0.650.630.040.290. 920.82 0.80 0.62 0.56 0.28 A 7 22 87341113 10 10 9.5 5.4 Lo. carcinophagus N 141 147 15413514014515386 148 65(126) 137.5 19.9 Hexp 0.81 0.90 0.900.840.650.910. 690.91 0.94 0.91 0.85 0.10 Hobs 0.78 0.93 0.87 0.540.54 0.870.700.86 0.88 0.85 0.78 0.14 A 11 22 17101218917 28 17 16.1 5.9 O. rossii N 41 40 412218404040 41 15(23) 34.6 9.5 Hexp 0.52 0.65 0.680.820.110.690. 920.89 0.82 0.95 0.71 0.25 Hobs 0.51 0.63 0.680.730.000.630. 900.85 0.83 1.00 0.68 0.28 A 4 4 1011261913 12 19 10.0 6.0 1 This locus is X-linked so the number of females and total numb er (in parentheses) are listed se parately and the heterozygosity represents only females. The total number of individua ls regardless of sex was used in calculati ng the average and stan dard deviation of t he number of individuals, as well as the numbe r of alleles seen at each locus.
80 Table 4.2. Autosomal F estimate (Xu and Fu 2003) and the genetically effective population size ( NEA = F4 and SD) for microsatellie loci surveyed in pack-ice seals. The mutation rate ( ) was assumed to be 10-5 per gene per generation. Le. weddellii Lo. Carcinophagus O. rossii Locus FNEA FNEA FNEA Hl04 0.205,00010.54263,5001.0827,000 Hl15 1.7443,50042.301,057,5002.5463,500 Hl16 2.2255,50036.72918,0003.1278,000 Hl20 2.2756,80015.68392,00011.10277,500 Lc05 0.013002.5062,5000.041,000 Lw08 0.266,50047.091,177,3003.4786,800 Lw10 28.34708,5003.5288,00057.531,438,300 Lw11 10.30257,50052.321,308,00034.18854,500 Lw16 7.89197,300127.593,189,80011.12278,000 Lw181 3.36112,00044.341,478,000161.25,373,300 Arithmetic mean SD 5.91 9.15 37.58 38.78 13.80 19.50 1 This locus is X-linked so the effective population index is theta divi ded by three times the mutati on rate and estimated using only female samples. The values for Lw18 are not incl uded in the averages or standard deviations. able 4.3.
81 Table 4.3. Estimates of genetically eff ective population sizes and the ratio of NE to census size (in parentheses) for maternally, paternally, and biparentally inheri ted loci in pack-ice seals based on F (Xu and Fu 2003) or as estimated using the program LAMARC (Kuhner 2006). Values for the autosomal microsatel lites were calculated using the arithmetic mean of F across all nine loci. Microsatellite Mitochondrial X-linked3 Autosomal3 Autosomal4 Y-linked Species Census Size22NEF NEX NEA Reduced data yes/no L. Weddellii 800,000 111,200 (0.14) 112,000 (0.14) 147,850 (0.19) 77,800 (0.10) Yes L. Carcinophagus 15 million 853,400 (0.07) 1,478,000 (0.10) 939,600 (0.06) N/A Yes O. rossii 220,000 130,400 (0.59) 5,373,300 (24.45) 344,950 (1.57) 116,000 (0.53) No 1 Based on from Curtis et al. 2009. Estimated from mitochondrial DNA were doubled assuming that the sex ratio is 1:1. 2 These are the largest of the publis hed estimates and are from Erickson and Hanson (1990) and Laws (1977). 3 Mutation rate was assumed to be 10-5. 4 Single loci with puta tively aberrantly large F values were removed (see text).
82 Table 4.4. Observed and expected microsatellit e heterozygosities of seals. Number of individuals (N ), observed heterozygosity ( HO), expected heterozygosity ( HE), and number of loci ( A ) used in the study. Studies without pub lished observed heterozygosity were not included in this table. SpeciesLocationN HOHEA Citation Mediterranean monk seal (Monachus monachus) Eastern Mediterranean~600.230.3224 Pastor et al. 2007 Mediterranean monk seal (Monachus monachus) Western Mediterranean~600.350.3824 Pastor et al. 2007 Harbor seal (Phoca vitulina) Strandings along Dutch coast2040.340.3427 Riijks et al. 2008 Hawaiian monk seal (Monachus schauinslandi) Hawaiian archipelago24090.480.498** Schultz et al. 2008 Harbor seal (Phoca vitulina) Western Atlantic, Northern Europe~9720.50?7Goodman 1998 Weddell seal (Le. weddellii) Western Ross Sea, Antarctica~1630.560.5610This study Leopard seal (Hydrurga leptonyx) Bird Island, Antarctica250.640.6624 *Davis et al. 2002 Weddell seal (Le. weddellii) Big Razorback Island, Antarctica960.660.6724 *Davis et al. 2002 S. elephant seal (Mirounga leonina) Faukland Islands2630.660.669 Fabiani et al. 2004 Ross seal (O. rossii) Western Ross Sea, Antarctica~350.680.7110This study Leopard seal (Hydrurga leptonyx) Bird Island, Antarctica210.700.7218 Davis et al. 2000 Weddell seal (Le. weddellii) East coast and west coast Antarctica1580.720.7318 Davis et al. 2000 Gray seal (Halichoerus grypus) North Rona, Scotland3090.76?up to 11Poland and Pomeroy 2008 Gray seal (Halichoerus grypus) North Rona, Scotland~4000.76?up to 11 Twiss et al. 2006 Crabeater seal (Lo. carcinophagus) Western Ross Sea, Antarctica~1380.780.8510This study Crabeater seal (Lo. carcinophagus) RV Nathaniel B Palmer, 1994250.790.8324 *Davis et al. 2002 Antarctic fur seal (Arctocephalus gazella) Isle de Crozet (in AA circle)~1020.800.826Kingston and Gwilliam 2007 Subantarctic fur seal (Arctocephalus tropicalis) Isle de Crozet (in AA circle)~1020.810.826Kingston and Gwilliam 2007 Crabeater seal (Lo. carcinophagus) Around Antarctic continent490.810.8718 Davis et al. 2000 Harp seal (juveniles) (Phoca groenlandica) Northern United States650.83?12 Kretzmann et al. 2006 *We estimated means of HO and HE from published data in Davis et al. 2002. **A total of 154 were screen ed in the study (Schultz et al. 2008), we refer here to the data of the eight polymorphic loci.
83 Figure 4.1. Map of the Antarctic continent showing the area from which the samples were collected.
84 Chapter 5: Conclusion The major importance of this dissertation stemmed from estimating effective population sizes through severa l types of genetic markers, those with maternal inheritance (mitochondrial DNA), biparent ally inherited markers (autosomal microsatellites) and sex-linked markers. In addition, these markers encompassed both sequence fragments as well as microsatellite s, which have differing modes and rates of mutation. This allowed for several independent lines of investigati on with the available samples, to the extent that is possible with genetic analyses on the same sample set. Concordance between expectati ons and markers can be a powerful tool in revealing biologically meaningful demographic patterns and discord between markers of varying inheritance patterns may provi de insight into sex-specifi c processes including mating systems and long term effective sex ratios. Despite an enormous screening effort, however, I was unable to find Y chromosome sp ecific markers with e nough variability to allow any sort of meaningful statistical anal yses, so sex-linked estimations of effective population size were limited to a single X-li nked microsatellite. It is likely that developing a Y chromosome specific microsat ellite library would yi eld more variable markers for this purpose. When my data are considered together, it seems likely that Weddell seal effective population size is between 111,200 and 147,850 indi viduals, and refl ects a regional
85 population estimate in the western Amundsen a nd eastern Ross seas. It seems reasonable to postulate this to be the case, as ge netic subdivision has be en noted in samples separated by more th an 700 km (Davis et al. 2008). If I assume th at the distribution of Weddells is uniform around the continent, a nd roughly extrapolate th is regional estimate, it gives an estimate of 556,000 to 739,250 animals, which approaches the census size estimates of about 730,000 (Laws 1977) to 800,000 (Erickson and Hanson 1990). The mitochondrial data suggest th at these seals underwent a rapid population expansion roughly 731,000 years ago, a time of expanding ice habitat and lowe ring sea levels. Populations may have become more robust and expanded their ranges as new and better fast ice habitat became availa ble for breeding and molting. The effective population size estimat e for crabeater seals 853,400 and 939,600 individuals. This number is significantly lower than published census estimates of 7 to 15 million individuals (Laws 1977, Ericson a nd Hanson 1990). Given their migratory behavior and the present lack of evidence of genetic sub-structuri ng around the continent (Davis et al. 2008), it is possible that this reflects a circumpolar estimate, in which case there exists a large discrepancy between census and genetic estimates. Further investigation may reveal some level of undetected geographic subdivision among crabeater seals, particular ly between east and west An tarctica and perhaps other geographically disjunct areas like the Weddell Sea in whic h case our estimate may be limited to a portion of the perime ter of continent. Both Weddell seals and crabeater seals
86 appear to have undergone rapid population expansions during times of expanding ice habitat. One may postulate that dramatic, l ong term reductions in ice habitat off the coast of Antarctica may have an invers e effect on their population sizes. Ross seal effective population sizes vari ed significantly between mitochondrial estimates (130,400 individuals) and autoso mal estimates (344,950 individuals), though when single loci with putatively aberrantly large F values were removed the autosomal estimate was closer to that of the mito chondria (116,000 individuals), close to the circumpolar census estimates of 131,000 to 220,000 individuals (Laws 1977, Erickson and Hanson 1990). Due to the large discrepenc ies between the mitoc hondrial, autosomal, and X-linked microsatellite data, this speci es effective population size warrants further investigation. The Ross seal did not show a ny signs of rapid populat ion expansion, as did the crabeater and Weddell seals. It is interesting to note th at the rank order for the three species remained consistent with both mitochondrial and au tosomal markers, and that Ross seal effective population sizes appear to be larger than Weddell seals, which is not consistent with published census data. This may be due to greater genetic subdivision in Weddell seals, whereby Weddell seal effectiv e population size estimates reflect a more regional estimate (that of the western Amundsen and eastern Ross seas) whereas Ross seals reflect a more circumpolar estimate. Conversely, it is possibl e that Ross seals are undercounted using traditional census techniques. These data and effective population size estimates are good starting points, but
87 caution should be used with respect to these estimates and their application in management strategies. Ross seal estimates, based on 41 samples, may not have captured a true picture of the genetic variation in the population. The analyses of more samples in this species is critical, and this is arguab ly true for all the sp ecies in the study. Additionally, an equal sampling effort fr om equidistant point s around the continent would help to discern circumpolar estim ates for the species in this study. In addition to more samples, future st udies should employ a greater number of both autosomal microsatellites (ideally more than 20 variab le, unlinked microsatellites) and X-linked microsatellites, to reduce the variation ar ound the mean estimates of As research into pinniped biology continues, gene ration times of phocids should be revisited and confirmed. A better understanding of ge neration times may alte r effective population size estimates. All other factors remaining constant, decreasing generation times will increase effective population si ze estimates and vice versa. Future research into Ross seal biology is important, and may help explain why Ross seal populations appear to have remained stable through the Pleistocene epoc h, while the other two phocid seals in this study appear to have rapidly expanded during th is time. It is possible that behavioral traits specific to the Ross seal may expl ain this difference, perhaps the Ross seals propensity to use the open ocean to a greater extent, causing them to be less influenced by expanding ice habitat. Food source may be another consideration, as invertebrates, the primary food source of Ross seals, may have been a factor lim iting their expansion.
88 References Ackley SF, Bengtson JL, Boveng P, Castellin i M, Daly KL, Jacobs S, Kooyman GL, Laake J, Quetin L, Ross R, Siniff DB, St ewart BS, Stirling I, Torres J, Yochem PK (2003) A top-down, multidisciplinary study of the structure and function of the packice ecosystem in the eastern Ross Sea, Antarctica. Polar Record 39 219-230. Amos W, Hutter CM, Schug MD, Aquadro CF (2003) Directional evolution of size coupled with ascertainment bias for vari ation in Drosophila microsatellites. Molecular Biololgy and Evolution 20 (4), 660-662. Andolfatto P (2001) Contrasting patterns of X-linked and autosomal nucleotide variation in Drosophila melanogaster and Drosophila simulans Molecular Biology and Evolution, 18 (3), 279-290. Arnason U, Gullberg A, Janke A, Kullberg M, Lehman N, Petrov EA, Vainola R (2006) Pinniped phylogeny and a new hypothesis fo r their origin an d dispersal. Molecular Phylogenetics and Evolution, 41 (2), 345-354.
89 Avise JC, Ball RM, Arnold J (1988) Current versus historical population sizes in vertebrate species with high gene flow a comparison based on mitochondrialDNA lineages and inbreeding theo ry for neutral mutations. Molecular Biology and Evolution, 5 331-344. Barker JS, Frydenberg J, Gonzalez J, Davies HI, Ruiz A, Sorensen JG, Loeschcke V (2009) Bottlenecks, population differe ntiaton and apparent selection at microsatellite loci in Australian Drosophila buzzatti. Heredity (2009 electronicpublication ahead of print). Bartsh SS, Johnston SD, Siniff DB (1992) Terr itorial behavior and breeding frequency of male Weddell seals (Leptonychotes weddellii) in relation to age, size, and concentrations of serum te stosterone and cortisol. Canadian Journal of Zoology, 70 680-692. Bazin E, Glemin S, Galtier N (2006) Popul ation size does not influence mitochondrial genetic diversity in animals. Science, 312 (5773), 570-572. Bengtson JL, Laws RM (1985) Trends in crabeater seal age at maturity Â– an insight into
90 Antarctic marine interactions. Pp. 669-675. In: Antarctic nutrient cycles and food webs (Siegfried WR, Condy PR, Laws RM eds.). Springer-Verlag, Berlin. Bengtson JL, Siniff DB (1981) Reproductive aspects of female crabeater seals ( Lobodon carcinophagus ) along the Antarctic Peninsula. Canadian Journal of Zoology 59 91-102. Bester MN, Erickson AW, Ferguson JWH (1995) Seasonal change in the distribution and density of seals in the pack ice off Princess Martha Coast, Antarctica. Antarctic Science, 7 357-364. Betancourt AJ, Kim Y, Orr HA (2004) A pse udohitchhiking model of X vs autosomal diversity. Genetics 168 2261-2269. Buchan JC, Archie EA, Van Horn RC, Moss CJ, Alberts SC (2005) Locus effects and sources of error in noninvasive genotyping. Molecular Ecology Notes, 5 680-683. Cameron MF, Siniff DB (2004) Age-specific su rvival, abundance, and immigration rates of a Weddell seal (Leptonychotes weddellii) population in McMurdo Sound,
91 Antarctica. Canadian Journal of Zoology 82 601-615. Chenna R, Sugawara H, Koike T, Lopez R, Gibson TJ, Higgins DG, Thompson JD (2003) Multiple sequence alignment w ith the Clustal series of programs Nucleic Acids Research 31 (13), 3497-3500. Cline DR, Siniff DB, Erickson AW (1971) U nderwater copulation of the Weddell seal. Journal of Wildlife Management, 52 216-218. Croxall JP, Hiby L (1983) Fecundity, survival and site fidelity in Weddell seals, Leptonychotes weddelli The Journal of Applied Ecology 20 19-32. Curtis C, Stewart BS, Karl SA (2007) Sexing pinnipeds with ZFX and ZFY loci. Journal of Heredity, 98 280-285. Curtis C, Stewart BS, Karl SA (2009). Pl eistocene population expansions of Antarctic pack-ice seals. Molecular Ecology, in press. Dalebout ML, Robertson KM, Frantzis A, Engelhaupt D, Mignucci-Giannoni AA, Rosario-Delestre RJ, Baker CS (2005) Wo rldwide structure of mtDNA diversity among Cuvier's beaked whales ( Ziphius cavirostris ): implications for threatened
92 populations. Molecular Ecology, 14 3353-3371. Damstra T, Barlow S, Bergman S, Ka vlock R, Van Der Kraak G (2002) Global assessment of the state-of-the-science of endocrine disruptors. World Health Organization publication WHO/PCS/EDC/02.2 (visited last June 22, 2006) http://www.who.int/ipcs/publications/new _issues/endocrine_disruptors/en/index.h tml Davis CS, Stirling I, Strobeck C (2000) Geneti c diversity of Antarcti c pack ice seals in relation to life history characteristics. In: Antarctic ecosystems: models for a wider ecological understanding (eds. Davison W, Howard-Williams P, Broady P), pp 56-62. Christchurch, Ne w Zealand: Caxton Press. Davis CS, Gelatt TS, Siniff D, Strobeck C (2002) Dinucleotide microsatellite markers from the Antarctic seals and th eir use in other Pinnipeds. Molecular Ecology Notes, 2 (3), 203Â–208. Davis CS, Stirling I, Strobeck C, Coltman DW (2008) Population structure of icebreeding seals. Molecular Ecology, 17 3078-3094.
93 Decker D, Stewart BS, Lehman N (2002) Ma jor histocompatibility complex class II DOA sequences from three Antarctic seal species verify stabilizing selection on the dog locus. Tissue Antigens 60 533-537. Di Rienzo A, Peterson AC, Garza JC, Va ldes AM, Slatkin M, Freimer NB (1994) Mutational processes of simple sequen ce repeat loci in human populations. Proceedings of the National Academy of Science USA, 91 3166-3170. Ellegren H (2000) Microsatellite mutations in the germline: implications for evolutionary inference. Trends in Genetics, 16 (12), 551-558. Erickson AW, Hanson MB (1990) Continenta l estimates and population trends of Antarctic ice seals. In: Antarctic ecosystems, ecol ogical change and conservation (eds. Kerry KR and Hempel G), pp. 253264. Berlin and Heidelberg: SpringerVerlag. Excoffier L (2004) Patterns of DNA sequence diversity and genetic structure after a range expansion: lessons from the infinite-island model. Molecular Ecology 13 853-864.
94 Excoffier L, Laval G, Schneider S (2005) Ar lequin ver. 3.0: An integrated software package for population genetics data analysis. Evolutionary Bioinformatics Online 1 47-50. Fabiani A, Galimberti F, Sanvito S, Hoelzel AR (2004) Extreme polygyny among southern elephant seals on S ea Lion Island, Faukland Islands. Behavioral Ecology 15 (6), 961-969. Felsenstein, J (2005) Accuracy of calescent likelihood estimates: Do we need more sites, more sequence, or more loci? Molecular Biology and Evolution 23 691-700. Frankham R (1995) Effective popu lation size/adult population si ze in wildlife: a review. Genetics Research 66 95-107. Fredsted T, Villesen P (2004) Fast and reli able sexing of prosimian and human DNA. American Journal of Primatology, 64 345-350. Fu Y (1997) Statistical test s of neutrality of mutati ons against population growth, hitchhiking and b ackground selection. Genetics 147 915-925.
95 Galbusera P, van Dongen S, Matthysen E (2000) Cross-species amplification of microsatellites in passerine birds. Conservation Genetics, 1 (2), 163-168. Gelatt T, Davis C, Cameron M, Siniff D, Strobeck C (2000) The old and the new: integrating population ecology and populat ion genetics of Weddell seals. In: Antarctic ecosystems: models fo r a wider ecological understanding. (eds. Davison W, Howard-Williams P, Broady P), pp. 63-70. Christchurch, New Zealand: Caxton Press. Gilson A, Syvanen M, Levine K, Banks J (1998) Deer gender determination by polymerase chain reaction: validation study and application to tissues, bloodstains, and hair forensic samples from California. California Fish and Game 84 159-169. Goerke H, Weber K, Bornemann H, Ramdor S, Pltz J (2004) In creasing levels and biomagnification of persistent organic pollutants (POPs) in Antarctic biota. Marine Pollution Bulletin 48 295-302. Goodman SJ (1998) Patterns of extensiv e genetic differentiation and variation among European Harp Seals (Phoca vitulina vitulina) revealed using DNA microsatellite
96 polymorphisms. Molecular Biology and Evolution, 15 (2), 104-118. Hadley GL, Rotella JJ, Garrott RA, Nichols JD (2006) Variation in probability of first reproduction of Weddell seals. Journal of Animal Ecology 75 1058-1070. Hansson B, Westerberg L (2002) On the co rrelation between hetero zygosity and fitness in natural populations. Molecular Ecology 11 2467-2474. Harding KC, Harkonen T (1995) Estimating mean age of sexual maturity in the crabeater seal ( Lobodon carcinophagus ). Canadian Journal of Fisheries and Aquatic Sciences 52 2347-2352. Harpending HC (1994) Signature of anci ent population growth in a low-resolution mitochondrial DNA mismatch distribution. Human Biology, 66 591-600. Hartl DL, Clark AG (1997) Principles of population genetics 3rd ed. pp. 289-294. Sinauer Associates, Inc, Sunderland Massachusetts.
97 Hastings KK, Testa JW (1998) Maternal and birth colony effects on survival of Weddell seal offspring from McMurdo Sound, Antarctica. Journal of Animal Ecology 67 722-740. Hattori K, Burdin AM, Onuma M, Suzuki M, Ohtaishi N (2003) Sex determination in the sea otter (Enhydra lutris) from tissue and dental pulp using PCR amplification. Canadian Journal of Zoology, 81 52-56. Hellborg L, Ellegren H (2004) Low levels of nucleotide diversity in mammalian Y chromosomes. Molecular Biology and Evolution, 21 (1), 158-163. Higdon JW, Bininda-Edmonds ORP, Beck RM D, Ferguson SH (2007) Phylogeny and divergence of the pinnipeds (Carnivora: Mammaila) assessed using a multigene dataset. BMC Evolutionary Biology, 7 216. Hutter CM, Schug MD, Aquadro CF (1998) Microsatellite variation in Drosophila melanogaster and Drosophila simulans: a reciprocal test of the ascertainment bias hypothesis. Molecular Biology and Evolution, 15 (12), 1620-1636. International Whaling Commission Report, Annex G (2005) Document SC/57/021.
98 Report of the Sub-committee on In -depth Assessment (IA). p.9. Jefferson TA, Leatherwood S, Webber MA (1993) FAO Species Identification Guide, Marine Mammals of the World United Nations Food and Agricultural Organization. Rome. 320 p. Kanatous SB, Davis RW, Watson R, Polase k L, Williams TM, Mathieu-Costello O (2002) Aerobic capacities in the skeletal muscles of Weddells seals: key to longer dive durations? The Journal of Experimental Biology 205 3601-3608. Kashi, Y, Soller M (1999) Functional Roles of Microsatellites and Minisatellites. In: Microsatellites: Evolution and Applications Edited by Goldstein and Schlotterer. Oxford University Press. Kauer M, Zangerl B, Dieringer D, Schlot terer C (2002) Chromosomal patterns of microsatellite variability contrast shar ply in African and non-African populations of Drosophila melanogaster Genetics, 160 247-256. Kaufman GW, Siniff DB, Re ichle R (1975) Colony behavior of Weddell seals, Leptonychotes weddellii, at Hutton Cliffs, Antarctica. Rapport et Proces-Verbaux
99 des Reunions du Conseil International Pour LÂ’Exploration de la Mer, 169 228246. Kingston JJ, Gwillium J (2007) Hybridizati on between two sympatrically breeding species of fur seals at Iles Cro zet revealed by genetic analysis. Conservation Genetics, 8 (5), 1133-1145. Kocher TD, Thomas WK, Meyer A, Edward s SV, Paabo A, Villablanca FX, Wilson AC (1989) Dynamics of mitochondrial DNA evol ution in animals: amplification and sequencing with conserved primers. Proceedings of the National Academy of Sciences, USA 86 6196-6200. Kretzmann M, Mentzer L, DiGiovanni R, Leslie MS, Amato G (2006) Microsatellite diversity and fitness in st randed juvenile harp seals (Phoca groenlandica). Journal of Heredity, 97 (6), 555-560. Kuhner, MK (2006) LAMARC 2.0: maximum likelihood and Bayesian estimation of population parameters. Bioinformatics 22 768-770. Laws RM (1977) Seals and whales of the southern ocean. Philosophical Transactions of
100 the Royal Society, London, B, 279 81-96. Laws RM, Baird A, Bryden MM (2003) Size and growth of the crabeateater seal, Lobodon carcinophagus ( Mammalia: Carnivora), Journal of Zoology, 259 103108. Laws RM, Baird A, Bryden MM (2002) Age estimation in crabeater seals ( Lobodon carcinophagus ). Journal of Zoology 258, 197-203. Lawson LJ, Hewitt GM (2002) Comparison of substitution rates in ZFX and ZFY introns of sheep and goat related species supports the hypothesis of male-biased mutation rates. Journal of Molecular Evolution 54 54-61. Li W-H (1977) Distribution of nucleotide differences between two randomly chosen cistrons in a finite population. Genetics 85 331-337. Linder P (2006) Dead-box proteins: a family affairactive and passive players in RNP remodeling. Nucleic Acids Research 34 (15), 4168-4180. Lucchini V, Fabbri E, Marucco F, Ricci S, Boitani L, Randi E (2002) Noninvasive
101 molecular tracking of colonizing wolf ( Canis lupus ) packs in the western Italian Alps. Molecular Ecology 11 857-868. Luikart G, Cornuet J-M (1998) Empirical eval uation of a test for identifying recently bottlenecked populations from allele frequency data. Conservation Biology, 12 (1), 228-237. Macintosh NA (1970) Whales and dr ill in the twentieth century. In Antarctic ecology vol. 1 (eds. Holdgate, MW) pp 194-212. Acad emic Press, New York, New York. Malik S, Brown MW, Kraus SD, White BN (2000) Analysis of mitochondrial DNA diversity within and between North and South Atlantic right whales. Marine Mammal Science 16 45-558. Marchal JA, Acosta MJ, Bullejos M, de la Guardia RD, Sanchez A (2003) Sex chromosomes, sex determination, and sex-linked sequences in Microtidae. Cytogenetic and Genome Research 101 266-273. Mardon G, Page DC (1989) The sex-determining region of the mouse Y chromosome encodes a protein with a highly ac idic domain and 13 zinc fingers. Cell 56 765-
102 770. Mitani Y, Watanabe Y, Sato K, Cameron MF, Naito Y (2004) 3D diving behavior of Weddell seals with respect to prey accessibility and abundance. Marine Ecology 281 275-281. Mori M, Butterworth DS (2004) Considera tion of multispecies interactions in the Antarctic: a preliminary m odel of the minki whale Â– bl ue whale Â– krill interaction. African Journal of Marine Science 26 245-259. Morin PA, Nestler A, Rubio-Cisneros NT, Robertson KM, Mesnick SL (2005) Interfamilial characterization of a region of the ZFX and ZFY genes facilitates sex determination in cetaceans and other mammals. Molecular Ecology 14 32753286. Negri AP, Hales LT, Battershill C, Wolff C, Webster NS (2004) TBT contamination identified in Antarctic marine sediments. Marine Pollution Bulletin, 48 11421144. Nei M (1987) Molecular E volutionary Genetics. Colu mbia University Press.
103 Oritsland T (1970) Sealing and seal research in the sout h-west Atlantic pack-ice, September-October 1964. Pp. 367-376, In: Antarctic Ecology, Vol. 1 (Holdgate MW, ed.). Academic Press. Paetkau D, Calvert W, Stirling I, Strobeck C (1995) Microsatellite analysis of population structure in Canadian Polar Bears. Molecular Ecology, 4 (3), 347-354. Page DC, Mosher R, Simpson EM, Fisher EMC, Mardon G, Pollack J, McGillivray B, de la Chapelle A, Brown LG (1987) The sex-determining region of the human Y chromosome encodes a finger protein. Cell, 51 1091-1104. Palo JU, Makinen HS, Helle E, Stenman O, Vai nola R (2001) Microsat ellite variation in ringed seals (Phoca hispida) : genetic structure of hi story of the Baltic Sea population. Heredity 86 609-617. Park SDE (2001) Trypanotolerance in West African Cattle and th e Population Genetic Effects of Selection. Ph.D. th esis. University of Dublin. Pastor T, Garza JC, Aguilar A, Tounta E, Androukaki E (2007) Genetic diversity and
104 differentiation between the two remaining populations of the cr itically endangered Mediterranean monk seal. Animal Conservation, 10 (4), 461-469. Payseur BA, Nachman MW (2002) Natural se lection at linked sites in humans. Gene, 300 31-42. Pecon-Slattery J, OÂ’Brien SJ (1998) Pa tterns of Y and X Chromosome DNA Sequence Divergence During the Felidae Radiation. Genetics, 148 1245-1255. Pecon-Slattery J, Sanner-Wachter L, OÂ’Brie n SJ (2000) Novel gene conversion between X-Y homologues located in the nonrecomb ining region of the Y chromosome in Felidae (Mammalia). Proceedings of th e National Academy of Science, USA, 97 5307-5312. Petit JR, Jouzel J, Raynaud D, Barkov NI, Bar nola J-M, Basile I, Bender M, Chappellaz J, Davis M, Delaygue G, Delmotte M, Kotlyakov VM, Legrand M, Lipenkov VY, Lorius C, Ppin L, Ritz R, Saltzman E, Stievenard M (1999) Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica. Nature 399 429-436. Pilgrim KL, McKelvey S, Riddle AE, Schwartz MK (2005) Felid sex identification based
105 on noninvasive genetic samples. Molecular Ecology Notes, 5 60-61. Piry S, Luikart G, Cornuet J-M (199 9) BOTTLENECK: a computer program for detecting recent reductions in the effectiv e population size using allele frequency data. Journal of Heredity, 90 502-503. Piyamongkol W, Bermudez MG, Harper JC, Da gan W (2003) Detailed investigation of factors influencing amplification efficiency and allele drop-out in single cell PCR: implications for preimplantation genetic diagnosis. Molecular Human Reproduction 9 411-420. Plotz J, Ekau W, Reijnders PJ H (1991) Diet of Weddell seals Leptonychotes weddellii at Vestkapp, Eastern Weddell Sea (Antarctic a), in relation to local food supply. Marine Mammal Science 7 (2), 136-144. Poland VF, Pomeroy PP, Twiss SD, Graves JA (2008) A fine scale study finds limited evidence of kin clustering in a grey seal colony. Marine Mammal Science 24 371-387. Posada D, Crandall KA (1998) MODELTEST: testing the model of DNA substitution.
106 Bioinformatics 14 817-818. Raymond M, Rousset F (1995) GENEPOP (versi on 1.2): population genetics software for exact tests and ecumenicism. Journal of Heredity 86 248-249. Reed DH, Frankham R (2003) Correlation be tween fitness and genetic diversity. Conservation Biology 17 230-237. Reeves RR, Stewart BS (2003) Marine Mamm als of the World: An Introduction. In: WalkerÂ’s Marine Mammals of the World (Nowak RM ed). Baltimore MD: The Johns Hopkins University Press: 1-64. Reeves RR, Stewart BS, Leatherwood S (1992) The Sierra Club Handbook of Seals and Sirenians San Francisco, CA: Sierra Club Books. Riijks JM, Hoffman JI, Kuiken T, Osterhaus ADME, Amos W (2008) Heterozygosity and lungworm burden in harbor seals (Phoca vitulina). Heredity, 100 587-593. Rogers A (1995) Genetic evidence fo r a Pleistocene population explosion. Evolution 49 608-615.
107 Rogers AR, Harpending H (1992) Population gr owth makes waves in the distribution of pairwise genetic differences. Molecular Biology and Evolution 9 552-569. Schlotterer C (2000) Evolutionary d ynamics of microsatellite DNA. Chromosoma, 109 (6), 365-371. Schultz JK, Baker JD, Toonen RJ, Bowen BW (2008) Extremely low genetic diversity in the endangered Hawaiian monk seal (Monachus schauinslandi). Journal of Heredity, 100 (1), 25-33. Scientific Committee for Antarctic Research (2 006) Proposal to de-lis t Antarctic fur seals as specially protected species. Antarctic Treaty Consultative Meeting, 2006. http://www.scar.org/treaty/atcmxxix/atcm29_wp039.pdf Schneider S, Excoffier L (1999) Estimati on of demographic parameters from the distribution of pairwise di fferences when the mutation rates vary among sites: Application to human mitochondrial DNA. Genetics 152 1079-1089. Sefc KM, Payne RB, Sorenson MD (2003) Micr osatellite amplific ation from museum feather samples: effects of fragment ation size and template concentration on genotyping errors. Auk 120 982-989.
108 Shaw CN, Wilson PJ, White BN (2003) A reliable molecular method of gender determination for mammals. Journal of Mammology 84 123-128. Shen P, Wang F, Underhill PA, Franco C, Yang WH, Roxas A, Sung R, Lin AA, Hyman RW, Vollrath D, Davis RW, Cavalli-Sfor za LL, Oefner PJ (2000) Population genetic implications from sequence vari ation in four Y chromosome genes. Proceedings of the National Academy of Science USA 97 (13), 7354-7359. Simonsen BT, Siegismund HR, Arctander P (1998) Population structure of African buffalo inferred from mtDNA sequences a nd microsatellite loci: high variation but low differentiation. Molecular Ecology, 7 225-237. Siniff DB, Stirling I, Bengston JL, Reichle RA (1979) Social and reproductive behavior of crabeater seals (Lobodon carcinophagus) during the austral spring. Canadian Journal of Zoology, 57 2243-2255. Slade RW, Moritz C, Heideman A (1994) Multiple nuclear-gene phylogenies: applications to pinnipeds and comp arison with a mitochondrial DNA gene phylogeny. Molecular Biology and Evolution 11 341-356.
109 Slade RW, Moritz C, Hoelzel, AR, Burton HR (1998) Molecular Population Genetics of the Southern Elephant Seal Mirounga leonina Genetics 149 1945-1957. Slatkin M, Hudson RR (1991) Pa irwise comparisons of mitochondrial-DNA sequences in stable and exponentially growing populations. Genetics 129 555-562. Smetacek V, Nicol S (2005) Polar ecosystems in a changing world. Nature 437 362368. Solls S, Stewart BS, Lehman N (2005) Conservation of MHC class I sequences among carnivores. Tissue Antigens 65 283-266. Splettstoesser J, Gavrilo M, Field C, Harri son P (2000) Notes on Antarctic wildlife: Ross Seals, Ommatophoca rossii and emperor penguins, Aptenodytes forsteri New Zealand Journal of Zoology 27 137-142. Spielman D, Brook BW, Frankham R (2004) Most species are not driven to extinction before genetic factors impact them. Proceedings of the National Academy of Science 101 15261-15264.
110 Stewart BSS (2007) Current status of the Ross seal (Ommatophoca rossii) : a specially protected species under A nnex II (Appendix I). XXX Antarc tic Treaty Consultative Meeting, Scientific Committee on Antarctic Research, May 2007. Stirling I (1969) Ecology of the Weddell seal in McMurdo Sound, Antarctica. Ecology, 50 573-586. Stirling I (1969) Tooth wear as a mo rtality factor in the Weddell seal, Leptonychotes weddellii, Journal of Mammalogy, 50 (3), 559-565. Stirling I. (1971) Population aspects of Weddell seal ha rvesting of McMurdo Sound, Antarctica. Polar Record 15 653-667. Stirling I (1983) The evolution of mating systems in pinnipeds. In: Advances in the study of mammalian behavior (eds. Eisenberg JF, Kl eiman DG), pp. 489-527. The American Society of Mammalogists, Sp ecial Publication No. 7, Shippensburg Pennsylvania.
111 Twiss SD, Poland VF, Graves JA, Pomeroy PP (2006) Finding fathers: spatio-temporal analysis of paternity a ssignment in grey seals (Halichoerus grypus). Molecular Ecology, 15 1939-1953. Ushio S (2006) Factors affecting fast-ice break-up frequency in Lutzow-Holm Bay, Antarctica. Annals of Glaciology 44 (1), 177-182. Vowles EJ, Amos W (2006) Quantifying ascer tainment bias and species-specific length differences in human and chimpanzee microsatellites using genome sequences. Molecular Biology and Evolution 23 (3), 598-607. Westlake RL, OÂ’Corry-Crowe GM (2002) M acrogeographic structur e and patterns of genetic diversity in harbor seals ( Phoca vitulina ) from Alaska to Japan. Journal of Mammalogy 83 1111-1126. Wilson EA (1907) Mammalia. Pgs 1-66 in National Antarctic Expedition 1901 Â– 1904, natural History II. Zoology. British Museum, London. Vol 2, pt 1.
112 Wright S (1931) Evolution in Mendelian populations. Genetics 16 97-159. Xu H, Fu Y-X (2004) Estimating effectiv e population size or mutation rate with microsatellites. Genetics, 166 555-563. Yu N, Fu Y-X, Li W-H (2002) DNA polymorphi sms in a worldwide sample of human X chromosomes. Molecular Biology and Evolution, 19 2131-2141.
114 Appendix 1. Among species variability in the ZFX gene. Variable site position is relative to GenBank sequence (see text for accession numbers). Variable Site 11222222233334444444444444555555556666666677777889 66155788902590000001133779466678890166668823688663 78419139766273456786705066201234537445670161706784 Crabeater CAATGGTTTTGCTAATACCTGCCCGTATACAGCGGTTTAACACTGTTTAT Ross ..........................G.................C.CGG. Weddell .....CA..............T....G---..G...........C..... Elephant ....A....C.T.............CG.................CC.... Sealion AGGC...CC.A.G----TGCA.GTC.G...GA.AAC----TGTCC....C
115 Appendix 2. Among species variability in ZFY gene. Variable site position is relative to GenBank sequences (see text for accession numbers). Variable Site 0000000011111122222222333333333444444555556666666666667777778 1246778915789901133367445556777467789266680001112244681144682 4269015346230940312751291231347711390778940890181412310169727 Crabeater TGGGTCCAACTGCTGGGATTTATAATTAAGCCAGAATGCACAATGAAAATCGTTCTCCTGC Ross GCAA..T........A......CG..........T.....T.................... Weddell .T.A.........C.A......C...............T.T..----.T..A......... Elephant ...A....T......A......C....G...TT......GT......G........T.... Sealion ..A---.C.TCAG.AAA--CCCC.---.GTT..A.--A..TGG....G.AT.ACGC.TAAT
116 Appendix 3. DBY8 Sequence Variable Sites. SITE NUMBER Haplotype 1336369 L. weddellii common GA T C L. weddellii alternate GA T G O. rossii AT T C L. carcinophagus common GT T C L. carcinophagus alternate GT C C
117 Appendix 4. UTY11 Sequence variable sites. SITE NUMBER Haplotype 26 85 101102103104125140199254280 281 387391465 L. weddellii common C T T A C T T G G T Â— Â— A G T L. weddellii alternate C T T A C T T G G T Â— Â— A A T O. rossii T Â— T A C T T A A C A T G G C L. carcinophagus T T Â— Â— Â— Â— A A G T A T A G C
118 About the Author Caitlin Curtis completed this research as a portion of her requirement for the degree of Ph.D in Biology at the University of South Florida. She is interested in population genetics and evoluti on, especially in marine orga nisms. She is currently working on genetic tracking of Florida mana tees, and population genetics of smalltooth sawfish and spiny lobsters with the Florida Fish and Wildlife Research Institute.