Phylogeography of charrs (Salvelinus spp.) in Swiss and Greenlandic lakes
To investigate the phylogeography of charrs, mitochondrial DNA sequence variation of the entire control region was analysed in 95 Arctic charrs (Salvelinus alpinus) from Greenland and 191 Alpine charrs (Salvelinus umbla) from lakes in the Central Alps. This also allowed testing for possible genetic differences among sympatric and allopatric ecotypes of charrs found in the study areas. Previous studies analysing a 500 bp fragment of the mitochondrial DNA control region defined an Arctic, Acadian, Atlantic, Siberian and Beringean evolutionary lineage among the Holarctic distribution of charrs (Salvelinus spp.). Slightly increased genetic resolution was achieved in the present study by analysing the entire (approximately 1000 bp) control region. Median-joining haplotype networks revealed secondary contact between the Atlantic and Arctic lineage in Greenland with several haplotypes each. Rough estimations of the divergence time between these two lineages indicate a vicariance event in the mid Pleistocene before glacial retreat allowed them to (re)-colonize Greenland in the late Pleistocene or early Holocene. Geographical patterns of haplotype distribution were observed in Greenland, where two landlocked lakes showed solely Atlantic haplotypes and waterbodies accessible from the sea showed both, Atlantic and Arctic haplotypes. Since post-glacial rebound lifted Greenland’s landmass after the last glacial coverage, these findings could indicate a colonization of individuals belonging to the Atlantic lineage before the Arctic lineage whereupon these landlocked lakes were still accessible from the sea. Low genetic variability was observed among the samples from the Central Alps, where 95.8% of all the individuals shared haplotype ATL01. Interestingly, this haplotype is known to occur among the entire extent of the Atlantic lineage and was also found in Greenland and Norway in the present study. A rapid post-glacial expansion of an Atlantic population, probably experiencing a bottleneck effect during the last glacial period, could explain such distribution. Hardly any patterns of haplotype distribution among ecotypes were observed in both study areas. This is possibly caused by the slow mutation rate of the maternally inherited mitochondrial DNA paired with an evolutionary recent emerging of ecotypes in such geologically young post-glacial lakes. Nuclear genetic markers with a higher genetic resolution than mitochondrial DNA could be more suitable to assess genetic differences between ecotypes.