Unravelling the evolutionary history and future prospects of endemic species restricted to former glacial refugia

Abstract

The contemporary distribution and genetic composition of biodiversity bear a signature of species' evolutionary histories and the effects of past climatic oscillations. For many European species, the Mediterranean peninsulas of Iberia, Italy and the Balkans acted as glacial refugia and the source of range recolonization, and as a result, they contain disproportionately high levels of diversity. As these areas are particularly threatened by future climate change, it is important to understand how past climatic changes affected their biodiversity. We use an integrated approach, combining markers with different evolutionary rates and combining phylogenetic analysis with approximate Bayesian computation and species distribution modelling across temporal scales. We relate phylogeographic processes to patterns of genetic variation in Myotis escalerai, a bat species endemic to the Iberian Peninsula. We found a distinct population structure at the mitochondrial level with a strong geographic signature, indicating lineage divergence into separate glacial refugia within the Iberian refugium. However, microsatellite markers suggest higher levels of gene flow resulting in more limited structure at recent time frames. The evolutionary history of M. escalerai was shaped by the effects of climatic oscillations and changes in forest cover and composition, while its future is threatened by climatically induced range contractions and the role of ecological barriers due to competition interactions in restricting its distribution. This study warns that Mediterranean peninsulas, which provided refuge for European biodiversity during past glaciation events, may become a trap for limited dispersal and ecologically limited endemic species under future climate change, resulting in loss of entire lineages.

Keywords: Myotis escalerai; approximate Bayesian computation; bats; climate change; phylogeography; species distribution modelling.