Where did Arctic-Alpine mosses survive in a frozen Europe? Insights from a multispecies coalescent analysis

Arctic-Alpine species in Europe exhibit a strikingly disjunct distribution between Fennoscandia and the high mountain ranges of the South. During the cold stages of the Pleistocene, these areas were covered almost entirely with a continuous ice sheet. Three hypotheses were formulated to explain the survival of organisms in Arctic-Alpine ranges during the ice-ages: (i) the ‘nunatak’ hypothesis, which suggests in-situ survival in micro-refugia within the ice sheet; (ii) the ‘tabula rasa’ hypothesis, which proposes the total eradication of populations within the ice sheet and subsequent recolonization from refugia outside the main glaciated area; and (iii) the ‘out-of-Europe’ hypothesis, according to which all European populations went extinct so that Europe was completely back-colonized from allochtonous populations, and in particular, North American ones. These hypotheses were tested here using the Arctic-Alpine bryophyte flora as a model. We used species distribution models to generate prior information on extant and Last Glacial Maximum effective population sizes. We then confronted these predictions to observed genetic data derived from DNA sequence variation at 3-4 unlinked loci for a representative population sampling of 3 species (Amphidium lapponicum, Timmia austriaca and Timmia bavarica) within their Holarctic range. More precisely, we employed Approximate Bayesian Computation to compare the observed patterns of genetic structure and diversity with those expected under the competing scenarios of ‘tabula rasa’, ‘nunatak’ and ‘out-of-Europe’ hypotheses.