Microsatellite analyses of the Antarctic endemic lichen Buellia frigida Darb. (Physciaceae) suggest limited dispersal and the presence of glacial refugia in the Ross Sea region

In order to assess the origins of Antarctic lichens (local or long distance), we examined the population genetic structure of the endemic Antarctic lichen Buellia frigida across a latitudinal gradient of roughly 10° along the Transantarctic Mountains, Western Antarctica, using four microsatellite loci. All loci were highly polymorphic. Data were analysed as both biallelic (dikaryotic) and as haploid in order to determine whether different life-cycle phases could influence our interpretation of population structure. For biallelic data, allelic richness ( A ) ranged from 5.25 to 7.99 and measures of diversity suggested low levels of gene flow among most sites (e.g. F ST = G ST = 0.09–0.31; D EST = 0.03–0.7). For haploid data, allelic richness ( A ) ranged from 3.5 to 5.46, private allelic richness ( A r ) ranged from 0.81 to 2.05, Nei’s unbiased genetic distance ranged from 0.15 to 1.42 and Nei’s unbiased genetic identity ranged from 0.24 to 0.86 among locations. Two locations, the McMurdo Dry Valleys and Queen Maud Mountains, stand out as possible glacial refugia, with both having a high number of private alleles. Despite the high potential for wind-dispersed spores, it appears likely that successful colonisation in different areas is restricted. One possible explanation is that the combination of ice-free conditions and water availability occurs only during the short summer period when the prevailing wind patterns may influence dispersal pathways. Dispersal from the southernmost site (Queen Maud Mountains) appears particularly restricted and may be the result of dispersal barriers such as glaciers. We conclude that a combination of prevailing wind patterns and physical barriers restrict spore settlement and therefore dispersal and recruitment among regions.