Comparison of contemporary and fossil diatom assemblages from the western Antarctic Peninsula shelf

During the past ten years, the Antarctic Peninsula has been identified as the most rapidly warming region of the Southern Hemisphere and it is important to place this warming in the context of the natural climate and oceanographic variability of the recent geological past. Many biological proxies, such as marine diatom assemblages, have been used to determine Southern Ocean palaeoceanographic conditions during the Late Quaternary, however, few investigations have attempted to link observations of modern floras with the fossil record. In this study we examine a suite of modern austral spring (December 2003) and summer (February 2002) surface water samples from along the western Antarctic Peninsula (WAP) continental shelf and compare these to core-top, surface sediment samples. Using detrended correspondence analysis (DCA) and principal component analysis (PCA) of diatom abundance data we investigate the relationship of contemporary diatom floras with the fossil record. This multivariate analysis reveals that our modern assemblages can be divided into three groups: summer southern WAP sites, summer northern WAP sites, and spring WAP sites. Sea surface temperature (SST) is an important environmental variable for explaining seasonal differences in diatom assemblages between spring and summer, but sea surface salinity (SSS) is more important for understanding temporally-equivalent regional variations in assemblage. Our summer diatom samples are more reminiscent of early season assemblages, reflecting the unusually late sea ice retreat from the region that year. When the modern assemblages are compared to the fossil record, it is clear that most of the important diatoms from the summer assemblage are not preserved into the sediments, and that the fossil record more closely reflects spring assemblages. This observation is important for any future attempts to quantitatively reconstruct palaeoceanographic conditions along the WAP during the Late Quaternary and highlights the need for many more such studies in order to address longer timescales, such as interannual variability, in the context of the fossil record.