Department of Life Sciences, University of Liège, 4000 Liège, BELGIUM.
The taxonomic composition of the cyanobacterial communities remained more or less constant throughout the Mid-Late Holocene climate optimum due to their well-developed adaptive capabilities. However, we cannot discard the possibility of bias by the resolution level provided by the 16S rRNA genes. Indeed, in some cases the short 16S sequences obtained by the methods used (i.e. DGGE) do not provide enough information for an accurate identification of the genotypes. The analysis of complete 16S sequences plus ITS could improve this identification. In addition, the conservation ratio of 16S rRNA sequences could be too high, and therefore the diversity underestimated. Recently, Stackebrandt & Ebers (2006) suggested that bacterial DNA-DNA hybridizations values of 70% (which correspond to the species concept) should correspond to 16S rRNA similarities about 98-99%. Thus, this high threshold implies that the marker is maybe, too conserved.
On the contrary in the uppermost sediments of Beak-1, coinciding with the last 2 or 3 decades, the cyanobacterial community structure changed and richness increased. This was in concordance with the increase of tychoplanktonic diatoms. These changes might reflect the first ecological responses of the lakes to the recent period of rapid warming observed in the Antarctic Peninsula (Verleyen et al., 2010a). However, they might also be related to natural lake succession processes. For example, aquatic moss communities become abundant during a period which precedes anthropogenic induced climate warming and slightly precede the shifts in the diatom communities. Hence, it is still uncertain whether these changes are part of the natural lake succession or rather related to climate warming.