Conflicting relationships of Vipera walser inferred from nuclear genes sequences and mitochondrial DNA

The description of Vipera walser from the Northern Italian Alps as a new species (Ghielmi et al., 2016, Journal of Zoological Systematics and Evolutionary Research, 54, 161) was one of the most unexpected surprises of European herpetology in the 21st century. In mitochondrial (mt) DNA, it is closely related to a group of vipers only present in the Caucasus region and Northeastern Anatolia. However, its morphology is similar to the V. berus populations that inhabit nearby mountains in the Swiss-Italian Alps, which raises questions on its relationships and status. We thus sequenced five nuclear (nu) genes to determine the position of V. walser relative to V. berus and to the Caucasian/Northeastern Anatolian vipers in nuDNA. We also reanalyzed five previously sequenced mtDNA fragments. NuDNA markers recovered V. walser as closely related to Italian populations of V. berus and not to the Caucasian/Anatolian species, thus contradicting the mtDNA phylogeny. We checked that each of the five mtDNA fragments independently amplified by Ghielmi et al. (2016, Journal of Zoological Systematics and Evolutionary Research, 54, 161) produced individual gene trees compatible with the concatenated mtDNA phylogeny, thus excluding the hypothesis that NUMTs sequencing generated the mtDNA relationships reported by Ghielmi et al. (2016, Journal of Zoological Systematics and Evolutionary Research, 54, 161). Given the low level of nuclear differentiation between V. walser and the Italian population of V. berus, we argue that ancient admixture between V. berus and the ancestral population of V. walser is the most likely explanation for this case of cyto-nuclear discordance and we discuss the consequences of these results on the systematic status of V. walser.