Biodiversity in the Anthropocene : quantifying assemblage level change

Abstract

Humans are altering ecosystems at alarming rates, but we do not fully know how anthropogenic pressures are affecting local biodiversity. This knowledge gap poses problems for understanding ecological systems and for undertaking effective conservation initiatives. The aim of this thesis was to increase our understanding of local scale biodiversity change by focusing on assemblages, which are groups of species from the same taxonomic group that live in the same place. I focused on assemblage size, structure and species composition, with the end-goal of making informed suggestions on how to monitor change effectively. First, I asked whether sampling approach was highly influential when estimating species richness and composition of Trinidadian freshwater fish assemblages. I compared assemblage composition of two very different data collections: a systematic survey endeavour and the historical museum records of The University of West Indies. Both datasets provided comparable estimates of species richness, and they both contained mostly the same species. Where they did differ in species lists, they did so among the locally uncommon or specialist species. Next, I asked how closely changes in assemblage size measured in the currencies of numerical abundance and biomass relate. My analysis showed that neither assemblage size quantified in terms of numerical abundance, nor in terms of biomass, is systematically changing across systems, but that change in one currency usually predicted change in the other currency. Thirdly, I asked whether there is evidence for other directional changes across assemblages. There was no evidence of changes in the dominance structure or dominant species body size. There was, however, evidence from increasing numbers of rare species entering assemblages and driving up local species richness, as well as a suggestion of increased turnover among dominant species. Local assemblages therefore are displaying measurable change, but not all metrics capture these changes.