Analysis of historical and recent Ross Sea (Antarctica) molluscs datasets: benchmarking of existing knowledge to assess future changes

In polar areas, where the structure of marine communities and biodiversity are potentially threatened by climate-driven changes, it is really important to count on the availability of historical datasets to analyse the species diversity and distribution patterns with the final aim of creating a reference baseline to measure any possible change in the future. Historical dataset, in general, offer the widest geographical coverage for any give area and, very often, contains sampling stations that have never been re-sampled, standing as the unique source of information for a given site. In Antarctica, as well as in many other areas, molluscs are one of the best-studied taxa, especially because they have a carbonate structure, which allow the classification even in the case where the body is found dead (Linse et al., 2006). In fact, all past Antarctic expeditions collected mollusc samples and this group, therefore, is a very good candidate to investigate and understand trends and patterns of marine biodiversity in the Southern Ocean. Aim of this thesis is the creation of a complete dataset about all distributional data about Mollusca for the Ross Sea by combining the SOMBASE, a very large dataset includes information about several historical expeditions that occurred between the 1898 and 1968, and several datasets from other recent expeditions that have been made available from different sources. This new large dataset spans for over 100 years of samples and includes more than 700 discrete sampling events and it will be used to assess the robustness of our knowledge baseline about Antarctic molluscs diversity and distribution in the Ross Sea. In particular, the thesis aims at modelling Ross Sea molluscs’ species richness as function of different environmental variables (e.g. depth, temperature, etc.) as well as of different sampling gears. All the above analyses will be performed by using appropriate statistical analyses such as Generalized Linear Model (GLM), Generalized Linear Mixed Model (GLMM), extrapolation and rarefaction techniques, in order to disclose all the possible information contained in the dataset. Overall, the information achievable by analysing this dataset is of relevance in establishing a robust reference baseline to measure any future change in molluscs’ diversity, richness and distribution in this region. Moreover, these results will help determining, in the future, where are the areas with sampling gaps to be filled and the most useful sampling gears to be used.