Phylogeographic structure and ecological niche modelling reveal signals of isolation and postglacial colonisation in the European stag beetle

Lucanus cervus (L.), the stag beetle, is a saproxylic beetle species distributed widely across Europe. Throughout its distribution the species has exhibited pronounced declines and is widely considered threatened. Conservation efforts may be hindered by the lack of population genetic data and understanding of the spatial scale of population connectivity. To address this knowledge gap this research details the first broad scale phylogeographic study of L. cervus based on mitochondrial DNA (mtDNA) sequencing and microsatellite analysis of samples collected from 121 localities across Europe. Genetic data were complemented by palaeo-distribution models of spatial occupancy during the Last Glacial Maximum to strengthen inferences of refugial areas. A salient feature of the mtDNA was the identification of two lineages. Lineage I was widespread across Europe while lineage II was confined to Greece. Microsatellites supported the differentiation of the Greek samples and alongside palaeo-distribution models indicated this area was a glacial refuge. The genetic endemism of the Greek samples, and demographic results compatible with no signatures of spatial expansion likely reflects restricted dispersal into and out of the area. Lineage I exhibited a shallow star like phylogeny compatible with rapid population expansion across Europe. Demographic analysis indicated such expansions occurred after the Last Glacial Maximum. Nuclear diversity and hindcast species distribution models indicated a central Italian refuge for lineage I. Palaeo-distribution modelling results also suggested a western refuge in northern Iberia and south-west France. In conclusion the results provide evidence of glacial divergence in stag beetle while also suggesting high, at least on evolutionary timescales, gene flow across most of Europe. The data also provide a neutral genetic framework against which patterns of phenotypic variation may be assessed.