Quantitative ecology reveals scale-dependent structural processes of the Antarctic benthos through time

Modern day benthic community composition and distribution in the Southern Ocean are influenced by a variety of environmental and evolutionary factors across different spatial scales. Depth, temperature, salinity, and sea ice impact community structure at large spatial scales (e.g., > 60 km). Far less is known about the interactions between taxa and environment at smaller scales, particularly due to the difficulties in collecting polar data. The study of biodiversity typically requires the direct collection of material, including at the larval stage. However, benthic sampling in the Southern Ocean is extremely challenging, and traditional sampling methods also cannot be deployed in regions of high topographic complexity. Advances in camera technology, such as towed camera systems, have now allowed for extensive observation in previously under-sampled areas. While photographs are single snapshots in time, ecological inferences can be made of the community as a whole, allowing us to make in situ observations of these otherwise difficult to sample ecosystems. In this thesis I explored ecosystem structure in the Weddell Sea across a number of spatial and temporal scales using seabed photographs and fossil collections. I first investigated coexistence mechanisms and competition for two solitary scleractinian cup coral populations at the centimetre scale on the rocky slopes of Powell Basin (~2000 m depth). Using Spatial Point Process Analyses I found that these corals likely produced larvae that settle near their parents, and that the one coral morph changed its reproductive behaviour when sympatric to the other morph. This dispersal plasticity likely allowed for the coexistence of ecologically similar taxa at small spatial scales, thereby enhancing alpha diversity. Next, I examined benthic function at the ecosystem level in two habitats in the Weddell Sea: the rocky slopes of Powell Basin and the shallow muddy bottoms (~450 m water depth) near the Antarctic Peninsula. On the rocky Powell Basin slope, I discovered a dense, speciose epibenthic community dominated by suspension and filter feeders. Bayesian Network Inference showed that no single group controlled the network structure. Unlike in very shallow Antarctic systems, predators, especially starfish, though present, had little ecological impact, showing that predator-prey interactions were not a structural factor in the Powell Basin. In contrast, the soft substrate community was less dense and exhibited limited interactions, potentially due to sampling limitations, and was mediated only by dropstones. Finally, I investigated large scale benthic ecosystem complexity in deep time, using a minimally time- and habitat-averaged fossil record. I focused on the latest Maastrichtian 66 million years ago, just before the Cretaceous-Paleogene mass extinction, a global event following which the modern Antarctic fauna evolved. I analysed fossils from the López de Bertodano Formation of Seymour Island and found increasing metacommunity complexity, possibly driven by climate warming, over the 4 million years leading up to the extinction. This study provided valuable context for understanding the sudden impact of the extinction, and the ecological changes that occurred during this global evolutionary shift.

Cambridge International and Newnham College Scholarship, administered by The Cambridge Trust. Research funding from Natural Environment Research Council (NERC) Independent Research Fellowship NE/S014756/1, awarded to Emily Mitchell.