Palaeoecological changes in populations of Antarctic ice-dependent predators and their environmental drivers

The Southern Ocean is undergoing rapid physical and biological changes that are likely to have profound implications for Antarctic fauna. One such change is a projected decline in the extent of the Antarctic sea ice field by the end of the century, and an associated shortening of the sea ice season. Fauna that are dependent on Antarctic sea ice for breeding or foraging habitat are likely to be vulnerable to climate change. The Weddell seal (Leptonychotes weddellii) and emperor penguin (Aptenodytes forsteri) both use Antarctic sea ice as a breeding platform, while the Adelie penguin (Pygoscelis adeliae) breeds on ice-free ground, but forages largely within the sea ice zone. In order to develop successful conservation plans for such at-risk taxa, an understanding of the likely impacts of climate change is essential. As the changes currently underway in the Southern Ocean represent a long-term, environmental regime shift (as opposed to a short-term fluctuation), it is prudent to consider the responses of species to similar climate regime shifts in the past. By examining changes in the abundance and distribution of taxa over tens of thousands of years and placing these changes into context with palaeoclimate records, valuable insights into species' long-term responses to environmental change can be gained. When considered in combination with contemporary ecology, information on historical trends can give a more complete picture of species' li ely responses to climate change in the future. For this thesis, changes in the abundance and distribution of Weddell seals, emperor penguins and Adelie penguins over millennia were examined using mitochondrial DNA and genome-wide SNPs. Genetic diversity and contemporary population structure were also assessed, in order to provide context for the historical findings and to better understand the resilience of these predators to environmental change. Two mitochondrial markers (the control region and cytochome b) were sequenced for extant colonies and radiocarbon dated sub-fossil remains of the three species. In total, 250 emperor penguins from nine colonies spanning the Ross Sea, Weddell Sea and East Antarctica, 96 Weddell seals from seven sites in East Antarctica, and 56 Adelie penguins from six colonies in East Antarctica, were sequenced. The East Antarctic Adelie penguin sequences were supplemented with existing datasets of 36 penguins from the Scotia Arc and 49 penguins from the Ross Sea, to allow for circum-Antarctic comparisons. Using mitochondrial DNA sequences, the population trends of all three species in East Antarctica during and since the last glacial maximum (LGM, 26 ‚Äö- 19.5 kya) were reconstructed using coalescent Bayesian skyline methods. Phylogenies were also constructed for emperor and Adelie penguins from around the continent in order to identify genetic lineages associated with past refugia. It was hypothesised that the sea ice-breeding Weddell seal and emperor penguin may have prospered during the glacial period, due to an increase in breeding habitat and reduced competition from other, less cold-tolerant predators, whereas Adelie penguins were likely to have been reduced in number. However, findings indicated that both Adelie and emperor penguin populations were smaller during the last glacial period than they are today, and populations expanded in East Antarctica (135-fold and 5.7-fold, respectively) during the glacial-interglacial transition and the Holocene. The timing and magnitude of the population expansions were different for the two penguin species, and appear to have had different environmental drivers. In the case of the Adelie penguin, expansion began ca. 14 kya, suggesting that deglaciation, and the associated increase in ice-free ground suitable for nesting, was the likely driver. Meanwhile, the emperor penguin expansion in East Antarctica occurred 4,000 years later, coincident with reductions in sea ice that may have led to more favourable foraging conditions. The phylogenetic analyses indicated that both penguin species were restricted to refugia during the LGM. In the case of the Adelie penguin, there was evidence for two refugia, while for emperor penguins there were probably three refugia. Results strongly suggested that both species had a refuge in the Ross Sea, possibly associated with a polynya in the region which may have provided an oasis within the sea ice field. The location of other refugia were unable to be identified based on mitochondrial data analysis. Despite the similar ecological niches and overlapping distributions of emperor penguins and Weddell seals, findings indicated the two meso-predators responded very differently to historical climate change. While emperor penguin numbers increased rapidly in the Holocene, the size of Weddell seal populations was unchanged. Emperor penguins appear to possess a greater capacity to adapt to environmental change than Weddell seals. Emperor penguins prospered during the Holocene warming while Weddell seals did not, most likely due to a higher dispersal ability (and hence gene flow among colonies), higher evolutionary rate and fine-scale differences in their preferred foraging locations. The vastly different climate change responses of two ecologically equivalent predators suggests that differing adaptive capacities and/or fine-scale niche differences play a major role in species' climate change responses. Meanwhile, the similarity in the responses of the ice-breeding emperor penguin and rock-breeding Adelie penguin to climate change was also surprising. Overall, the climate change responses of species are complex and may prove difficult to predict. The degree of dispersal among breeding colonies is an important component of understanding a species' likely climate change response. This is because dispersal facilitates range shifts, and increases adaptive capacity by facilitating gene flow among breeding sites, thus replenishing the gene pool with new, potentially adaptive alleles. Delimiting breeding populations is also crucial for conservation planning, so that populations can be managed at an ecologically relevant spatial scale. A total of 114 individuals from eight emperor penguin colonies from the Ross Sea, Weddell Sea and East Antarctica were genotyped using 20,005 genome-wide SNPs generated using restriction-site-associated DNA sequencing (RADSeq). Their genetic population structure was then assessed using STRUCTURE analysis, principal components analysis, and analyses of pairwise FST, revealing a total of six extant breeding populations among the eight emperor penguin colonies sampled. There was indications of ongoing gene flow between colonies located up to 550 km apart, providing evidence against strong philopatry in emperor penguins. This finding has important implications for forecasting studies for the species, which have previously considered each colony as an isolated unit. The demographic histories of emperor penguin breeding populations were assessed over the past 50,000 years using RAD-Seq loci in combination with mitochondrial DNA sequences, via the coalescent Bayesian skyline method. This approach allowed for comparisons of the timing and magnitude of population expansions in different locations. The genetic evidence strongly indicated that there was an emperor penguin refuge at Cape Roget in the Ross Sea. This colony, unlike all other emperor penguin populations in this study, did not experience a population bottleneck during the LGM, and also had significantly high genetic diversity consistent with an older, ancestral population. The Cape Roget colony is proximate to the proposed location of an LGM polynya offshore from the Adare Peninsula that could have supported a refuge population of emperor penguins, by providing foraging access amidst the extensive sea ice field. This region in the north-western Ross Sea may have been an important penguin breeding habitat for at least 50,000 years and should therefore be considered in any future management for conservation. While decadal monitoring studies provide invaluable data on the short-term environmental sensitivities of predator populations, given the long-term nature of projected climate change it is also prudent to consider the climate-driven responses of populations over longer time scales. This study suggests that interspecific differences, even for sympatric species with similar ecological niches, are likely to affect the future climate change responses of Southern Ocean marine predators and should be considered in future conservation plans. The findings also highlight the importance of protecting productive foraging grounds proximate to breeding locations, as well as the potential role of polynyas as future Southern Ocean refugia.