Combining molecular and fossil data to infer demographic history of Quercus cerris: insights on European eastern glacial refugia

AIM: Phylogeographical studies of Eastern Mediterranean species are rare. We aim to fill a gap in the current understanding of the role of Eastern Mediterranean glacial refugia, and their connections with other refugia across Europe. To this end, we studied the genetic diversity distribution and genetic structure of the modern population of Quercus cerris in relation to its Quaternary demographic history and to more ancient events. LOCATION: Mediterranean Basin; Italian, Balkan, Anatolian peninsulas. METHODS: A total of 192 populations were genotyped with six polymorphic chloroplast microsatellites, and the genetic diversity and differentiation of the populations were evaluated. The geographical structure of genetic variation was analysed with a Bayesian clustering method using baps 5.2. The demographic history of Q. cerris was explored by an approximate Bayesian computation procedure using diyabc 2.0. To reconstruct the past distribution of Q. cerris, we also considered the chronology and geographical distribution of fossil records. RESULTS: Thirty‐five haplotypes were found, three of which (together) were found in 71.82% of individuals. Bayesian analysis resulted in three genetically and geographically distinct clusters: a Western group, a Central group, and an Eastern group. The approximate Bayesian computation analysis, together with fossil data, showed a possible bottleneck leading to the divergence of the Eastern and Central populations in the Early Pleistocene (Gelasian). The split into two groups of populations in the Italian and Balkan Peninsulas, respectively, was probably caused by a marked population contraction during a glacial phase of the Middle Pleistocene. MAIN CONCLUSIONS: This study provides information on the potential role of Eastern Europe and the Near East as refugia and as a source for ancient westward range expansions in the Mediterranean region. Our study covers a remarkable gap in European oak phylogeography, showing a putative eastern origin of Q. cerris and the presence of large amounts of genetic diversity in this region.