The spatial, temporal and structural distribution of Antarctic seafloor biodiversity

Biodiversity is nature's most valuable resource. The Southern Ocean contains significant levels of marine biodiversity as a result of its isolated history and a combination of exceptional environmental conditions. However, little is known about the spatial and temporal distribution of biodiversity on the Antarctic continental shelf, hindering informed marine spatial planning, policy development underpinning regulation of human activity, and predicting the response of Antarctic marine ecosystems to environmental change. In this thesis, I provide detailed insight into the spatial and temporal distribution of Antarctic benthic macrofaunal and demersal fish biodiversity. Using data from the George V shelf region in East Antarctica, I address some of the main issues currently hindering understanding of the functioning of the Antarctic ecosystem and the distribution of biodiversity at the seafloor. The focus is on spatial biodiversity prediction with particular consideration given to previously unavailable environmental factors that are integral in determining where species are able to live, and the poor relationships often found between species distributions and other environmental factors. Food is a fundamental requirement of life, influencing the distribution of all animal species, and for species living at the seafloor below the photosynthesis zone, this food derives from surface primary production. In Chapter 2, I present an interdisciplinary approach for modelling the redistribution of food-particles from the ocean-surface to the seafloor, combining satellite-data with an ocean-model, particle tracking, and diatom abundances from sediment grabs for validation. I show that different aspects of the estimated seafloor food-availability link directly to the abundance and richness of key Antarctic seafloor macrofauna observed from camera still-images. I then combine observations from the seafloor, bathymetry, and food-availability estimates to produce a (validated) predictive map of the distribution of important habitat-forming suspension feeders on the George V shelf (Chapter 3). Using a similar approach, in this chapter I also predict strong responses in the abundance of suspension feeders to changes at the ocean surface caused by a major glacier calving event in 2010. Biodiversity has many different attributes. Aside from the abundance of habitat-forming fauna, which is an important proxy for biodiversity, the spatial distribution of single species is also important in determining community structure. However, many species are rare which is why researchers have historically grouped species together based on taxonomic or functional similarity before modelling. Joint-species distribution models can aggregate species based on similarity in their responses to environmental factors, allowing prediction of the spatial distribution of multiple species, including rare species, with higher confidence than other more commonly used statistical methods, and with fewer assumptions regarding associations between grouped species. I use joint-species distribution models for mapping the distribution of diversity and community structure in benthic macro-invertebrates (Chapter 4) and demersal fish (Chapter 5). In chapter 4, I also analyse species-level data of benthic macro-invertebrates that are a-priori aggregated into higher level taxonomic and functional groups of species, and show that aggregating species into higher level groupings leads to increased modeluncertainty (Chapter 4). Comparing patterns in the spatial distribution of demersal fish and benthic invertebrates, I then show that communities of benthic species can be described along four broad habitat types characterised mainly by depth and slope of the seafloor, namely shallow-flat, shallow-steep, deep-flat, and deepsteep environments, with food-availability additionally influencing these habitat distinctions (Chapter 5). In this chapter I conceptualise knowledge about community structure and interactions among functional groups and key environmental drivers in the Antarctic benthic ecosystem in a qualitative network topology, test how differences in the environmental setting affect ecological structure, and validate the dynamic network model results with the mapped distributions of fish and macroinvertebrates, revealing insight into ecosystem functioning. This work shows that the large scale spatial, temporal and structural distribution of Antarctic benthic biodiversity is mainly influenced by depth and slope of the seafloor and by the availability of surface-derived food, which links the seafloor strongly to environmental processes at the ocean-surface. Inclusion of the food-availability models results in more accurate mapping of the spatial and temporal distribution of Antarctic marine biodiversity. The high confidence in the spatial predictions of benthic biota means that the representativeness of future Marine Protected Areas can be better assessed and possible changes in the benthic ecosystem due to environmental or human pressures can be better detected and understood. Future predicted increases in primary production will likely result in higher abundances of benthic suspension feeders and alter community composition, but patterns are variable, even within regions. The advances presented in this thesis can form the basis for future work to map the distribution of Antarctic benthic biodiversity on a circumpolar scale, identify habitats and species assemblages critical for conservation, quantify the total biomass of Antarctic benthic communities, estimate the contribution of Antarctic benthic communities to the sequestration of atmospheric carbon, and estimate how these communities will change in the future based on climate projections. In Chapter 6, I review and discuss how technological and collaborative advances, including those outlined in this thesis, allow us to predict marine biodiversity unlike ever before. The spatial and temporal scales at which predictions are now possible, and the confidence in the predictions themselves, help us to better assess management and policy decisions from local to global scales. These recent developments also allow exploring new ways to predict biodiversity in the future, for the mutual benefit of marine ecosystems and humanity.