Etude de l’adaptation des diatomées en Péninsule Antarctique au cours des derniers millénaires

Ancient sedimentary DNA (sedaDNA) is a molecular tool transferred from ecology to paleoceanography, now being applied in the Southern Ocean (SO). Important variations in the abundance of ecologically-important diatom genera (e.g., Chaetoceros, Thalassiosira and Fragilariopsis) have been described in SO sediment cores through microfossils counts. However, environmental studies have shown that many diatom species undergo dissolution during settling and are therefore not preserved in the sediment, which can be alleviated with sedaDNA. In a first step, we optimized both the extraction and sequencing library preparation protocols to ensure higher DNA yield, good inhibitor removal and low dimer concentrations for amplicon and shotgun sequencings. DNA sequencing will allow us to confirm and refine the composition of diatom communities preserved in this core, thanks to the reference databases of these species in the contemporary ocean, including baseline data from Tara Oceans. In particular, we are targeting the V9 region of the eukaryotic-specific small ribosomal subunit as well as a diatom-specific fragment within a region encoding the large RuBisCO subunit. These amplicons are widely used in studies of contemporary phytoplankton communities as well as in paleo genomics studies, producing a signal of the microbial diversity at a low sequencing cost. In a second step, sedaDNA will be analysed in a deep-sea core from the northern Antarctic Peninsula (432cm), covering the Late Holocene, and compared to the classical diatom counting approach as well as other proxies such as Highly Branched Isoprenoids for past sea-ice coverage. We propose a new approach to understand adaptative responses of diatoms communities to past environmental changes in the SO both through the analysis of species diversity variations over time and through a community functional perspective, targeting genes coding for key biological functions in the shotgun data such as ice-binding proteins coding genes.