Evolutionary and demographic correlates of Pleistocene coastline changes in the Sicilian wall lizardPodarcis wagleriana

Abstract

AbstractAimEmergence of coastal lowlands during Pleistocene ice ages might have provided conditions for glacial expansions (demographic and spatial), rather than contraction, of coastal populations of temperate species. Here, we tested these predictions in the insular endemic Sicilian wall lizardPodarcis wagleriana.LocationSicily and neighbouring islands.MethodsWe sampled 179 individuals from 45 localities across the whole range ofP. wagleriana. We investigated demographic and spatial variations through time using Bayesian coalescent models (Bayesian phylogeographic reconstruction, Extended Bayesian Skyline plots, Isolation‐with‐migration models) based on multilocusDNAsequence data. We used species distribution modelling to reconstruct present and past habitat suitability.ResultsWe found two main lineages distributed in the east and west portions of the current species range and a third lineage restricted to a small area in the north of Sicily. Multiple lines of evidence from palaeogeographic (shorelines), palaeoclimatic (species distribution models), and multilocus genetic data (demographic and spatial Bayesian reconstructions) indicate that these lineages originated in distinct refugia, located in the north‐western and south‐eastern coastal lowlands, during Middle Pleistocene interglacial phases, and came into secondary contact following demographic and spatial expansions during the last glacial phase.Main conclusionsThis scenario of interglacial contraction and glacial expansion is in sharp contrast with patterns commonly observed in temperate species on the continent but parallels recent findings on other Mediterranean island endemics. Such a reverse expansion–contraction (EC) dynamic has been likely associated with glacial increases of climatically suitable coastal lowlands, suggesting this might be a general pattern in Mediterranean island species and also in other coastal regions strongly affected by glacial marine regressions during glacial episodes. This study provides explicit predictions and some methodological recommendations for testing the reverseECmodel in other region and taxa.