Data recorded from annual tagging and regular tag resightings of southern elephant seals at Marion Island, permitted the investigation of temporal variation in the terrestrial haulout cycle of this species, and estimation of participation levels in the little understood resting haulout phase, that could be related to age, sex, and sexual status. Primigravid females moulted later, and were less inclined to rest, than nulligravid females of the same age, but moulted earlier than mature females, among which the mean moulting dates of respective age-classes were practically the same. It is proposed that earlier implantation among primigravid females, as a function of their returning to sea and regaining condition before parous females, resulted in the earlier mean breeding haulout date of primiparous females. The mean haulout dates of breeding and moulting males were negatively and postively correlated with ascending age, respectively. More than half of the surviving individuals of each immature age- and sex-class were observed to haul out to rest, with participation levels generally consistent from year to year, indicating chat the autumn-winter haulout represents more than simple random haulout events. It is suggested that elephant seals are faced with a trade-off between maximising their time spent foraging, and gaining valuable experience at terrestrial functioning, in their immature years. / Dissertation (MSc (Zoology))--University of Pretoria, 2006. / Zoology and Entomology / unrestricted

The southern elephant seal (SES), Mirounga leonina, was intensively harvested during the 18th and 19th centuries, though never reduced to the levels seen for the northern species (Mirounga angustirostris). Although a number of putative populations occurring within the species’ circumpolar distribution in the Southern Ocean have been genetically assessed, no data was previously available for the Marion Island population. This study integrates Marion Island into the broader database by genetically profiling individuals with 9 microsatellite DNA loci (n = 73) and a single mitochondrial DNA (mtDNA, n = 68) locus corresponding to hypervariable region I (HVRI) of the non-coding displacement loop (D-loop). These data were then combined with existing haplotype datasets from five island populations, namely Heard Island, Peninsula Valdés, Macquarie Island, South Georgia, Elephant Island, and Sea Lion Island breeding colonies, and with comparable microsatellite typing data from four populations, namely Peninsula Valdés, South Georgia, Elephant Island and Sea Lion island, respectively, permitting inter-population level comparisons. Genetic variation of the Marion Island population was high for both microsatellite and mtDNA and consistent with levels previously reported for the other populations, with the exception of Peninsula Valdés (Argentina) where diversity levels are low. Forty polymorphic sites defined 44 mtDNA haplotypes from 68 Marion Island individuals. Of the 44 sequence haplotypes, three were shared with Sea Lion Island, one with Elephant Island, two with Heard Island and one with Macquarie Island. From the microsatellite data, it was found that Marion Island, like most other SES populations, had no private alleles. The one exceptional population is that at Sea Lion Island which has several private alleles at two loci. Marion Island was significantly differentiated from each of the other breeding colonies included in the study based on FST analyses for both microsatellite and mtDNA data. The magnitude of genetic differentiation between Marion Island and the South Georgia, Sea Lion and Elephant Islands was somewhat higher than that previously reported when the latter three islands were compared, but considerably less than the differentiation found between Marion Island and either Peninsula Valdés or Macquarie Island. Though the two markers showed similar trends with respect to population structuring, the pairwise differentiation at microsatellite loci was an order of magnitude lower than that of mtDNA, suggesting more frequent male-mediated gene flow between putative populations than female-mediated gene flow. Higher male dispersal was also confirmed by migration rate estimates from the microsatellite data compared to estimates from the mtDNA locus. These data are consistent with the earlier interpretation that most putative island populations show moderate levels of differentiation not directly related to geographic distance, while the mainland population in Argentina and the population at Macquarie Island stand out as being highly differentiated from the rest. / Dissertation (MSc)--University of Pretoria, 2008. / Zoology and Entomology / unrestricted

While studies of dispersal note significant fidelity to natal site and to the site of first reproduction, few consider fidelity to other sites, and none have done so systematically. This study examined fidelity to all terrestrial sites within the study area during the course of its life, by a migratory marine predator, the southern elephant seal, Mirounga leonina. It also attempted to assess the role played by the winter haulout in terms of site fidelity. Finally it examined the influence of possible deterioration in spatial memory over time on site fidelity. The data used in this study were generated by a long-term mark-recapture programme conducted at subantarctic Marion Island. Although immature elephant seals of both sexes return to the vicinity of their natal sites, they appear to avoid popular breeding beaches, returning closest to the site previously used. At the first reproductive haulout, however, females return closer to their natal site than any other site, while males, although hauling out in the vicinity of their natal site, haul out closest to sites used in the year prior to the first breeding haulout. Subsequently, adults of both sexes breed closest to the breeding haulout of the previous year and moult closest to the moult haulout of the previous year. While males show greater site fidelity during the breeding season, there is no difference in site fidelity during the moult. Primiparous females show greater levels of site fidelity if recorded in the study site as an immature animal during either the winter or the moult haulouts. Also in female elephant seals, lower site fidelity is associated with an increase in the duration of period of absence from a site, and a lower number of visits to a site. Various factors related to site familiarity, social factors and anthropogenic disturbance may be responsible for the lack of strict site fidelity that is evident. Dispersal patterns may differ between the sexes due to differences in their life history. While purpose of the winter haulout by immature seals may be to increase familiarity with haulout sites, and thus site fidelity, the moult haulout also plays a role. / Thesis (PhD)--University of Pretoria, 2013. / Zoology and Entomology / Unrestricted

Thesis (PhD)--University of Pretoria, 1992. / gm2014 / Zoology and Entomology / unrestricted

L’océan austral est un écosystème fragile dont la dynamique est influencée par des variations climatiques qui vont structurer la distribution spatio-temporelle des ressources. L’objectif de cette thèse était d’étudier les stratégies d’acquisition des ressources en proies mises en place par l’éléphant de mer austral face aux contraintes énergétiques (coûts du transport et coûts d’accès à la ressource) et temporelles (temps passé au fond d’une plongée et limite de plongée aérobie) auxquelles il est soumis. L’utilisation d’un ensemble d’enregistreurs de données déployés sur ces animaux a permis de reconstruire en trois dimensions leur plongée, mais également de calculer leur effort de nage, le nombre de proies rencontrées ainsi que leur dépense énergétique. Notre étude montre qu’à l’échelle d’une plongée, les éléphants de mer adaptent leur trajectoire, mais également le temps qu’ils passent au fond, en fonction du nombre de proies rencontrées. Pour des densités locales de proie importantes, ils passent plus de temps au fond, et concentrent leur recherche en zone restreinte, caractérisée par une diminution de la vitesse et une augmentation de la sinuosité horizontale. Au-delà de 550 m, le coût d’accès aux ressources devient supérieur aux coûts d’acquisition ; ils doivent alors faire face à un compromis entre l’accessibilité et la disponibilité en proies. À mesure qu’ils s’alimentent, ces phoques augmentent leur flottabilité, diminuant de surcroît leur dépense énergétique. Cette étude démontre également une structuration spatio-temporelle de cette dépense énergétique qui semble être liée au succès d’alimentation et donc à la distribution des ressources en proies. / The Southern Ocean is a fragile ecosystem whose dynamics are influenced by climate change that will structure the spatio-temporal distribution of resources. The objective of this PhD was to investigate the foraging strategies used by the southern elephant seal, under energetic (cost of transport and costs of access to the resource) and temporal (time at the bottom of a dive and aerobic dive limit) constraints. Using a set of animal-borne data loggers allowed us to reconstruct their three-dimensional path underwater, but also to calculate their swimming effort, the number of prey encountered and their energy expenditure. At the dive level, our study shows that elephant seals adapt their path, but also the time spent at the bottom, depending on the number of prey encountered. For high local prey density, they spent more time at the bottom, and concentrated their foraging effort in areas restricted search, characterized by a decrease in speed and an increase in horizontal sinuosity. Beyond 550 m, the cost of access to resources becomes greater than the cost of acquisition ; they must therefore deal with a trade-off between prey accessibility and availability. While feeding, these seals increase their buoyancy, reducing furthermore their energy expenditure. This study also shows a spatio-temporal structure of the energy expenditure that appears to be related to feeding success and therefore to prey resources distribution.

Holistic studies of mammalian life history factors and their consequences on population demography require an intensive, multifaceted field methodology and effort over long temporal scales. A 25-year longitudinal mark-recapture experiment on southern elephant seals, Mirounga leonina, at Subantarctic Marion Island provide such a foundation for demographic analyses and relevant methodology advancement. Two gaps in the methodology related to life history and population demographic research are, the absence of large samples of known mass individuals, and an inability to identify mother-pup relatedness. A novel three-dimensional photogrammetric technique is designed here that allows for mass estimation of large samples of southern elephant seals in the field. An effective temporary marking technique for unweaned pups is implemented that allows for identification of large samples of pups with known mothers prior to the maternal bond being severed at weaning. These known pups can then be marked with more robust tags and relatedness information is preserved long-term. Thus, mass estimates can now be applied as covariates in modelling analyses to address questions of, for example, maternal investment, kinship associated behaviour, and the consequences thereof on survival and reproductive parameters. The state change in the Marion Island southern elephant seal population from decrease to stabilisation/increase is shown to have resulted from improved survivorship in both juvenile and adult female age classes. Male seals of all ages did not indicate improved survivorship following the period of decline. The inflexion in survivorship is identified as 1994, whence improved survivorship of juvenile seals preceded that of young adult females. This inflexion in survivorship is postulated to have resulted in a population trend inflexion around 1998. Female southern elephant seals do not show evidence of actuarial senescence, but reproductive senescence is apparent after 12 years of age. A longterm reproductive cost (reduced breeding effort) is associated with early primiparity (age three) as compared with later primiparity (4- 5- or 6-year-old). The mean proportion of 3-year-old breeders has not increased after 1994 as has been hypothesized in previous studies. Contrary to previous assumptions, females do not as a rule breed every year. Annually interrupted breeding efforts are more common than consecutive breeding efforts. No difference in the proportions of interrupted versus uninterrupted breeding efforts was identified between periods of population decline and stabilisation/increase. Longevity as predicted by survival estimates exceeds the observed frequencies. This study provides unique longevity and fertility schedules for the species. The improved survivorship, reproductive senescence and breeding schedules of female southern elephant seals in this population provide groundwork for reevaluation of previous studies and their conclusions. The addition of relatedness and body condition information will allow for sophisticated multistate modelling of population demography in future studies. However, analytical procedures and techniques employed need to be meticulously designed and thoroughly thought through to avoid mis-interpretation of biological data. In addition to a multistate single species analytical approach, the importance of an ecosystem approach to species population demographic studies is highlighted through the augmenting of data on relevant potential drivers of population change, such as killer whales, Orcinus orca. / Thesis (PhD)--University of Pretoria, 2009. / Zoology and Entomology / unrestricted