The analysis of genetic variation of Testudo hermanni in the Italian Peninsula and Sicily showed the existence of three main genetic clusters. Previous studies identified some genetic differentiation between populations from the Italian Peninsula and Sicily, and among populations within the peninsula [30,32]. However, the analysis of the geographic patterns of genetic variation applied here allowed us to identify marked sub-structuring within the southernmost region of the Italian Peninsula, and to define the geographic distributions of the distinct genetic lineages. In fact, we identified one genetic cluster spread in Sicily, one cluster spread from the Aspromonte massif to the Sila mountain chain, and one cluster spread throughout the rest of the Italian Peninsula. Furthermore, we identified two restricted areas of genetic admixture, one in the south of the Aspromonte massif, and one in the north of Calabria, corresponding to the Pollino massif. Overall, our results unveil a hotspot of genetic diversity for Testudo hermanni in southern Italy, in an area spanning from the Pollino massif to the Aspromonte massif, and suggest that the interplay between high topographic complexity and Pleistocene climate changes in this region triggered the formation of this hotspot.
The occurrence of genetic sub-structuring and distinct genetic clusters within the Italian Peninsula has been observed in several temperate species including amphibians [16,1722,23,25], mammals [14,18,20,41] and reptiles [21], among others [12]. In most of these species, the Calabrian region was identified as a hotspot of genetic lineages [20-23,16]. Within this region, major mountain areas are arranged along the north-south axis and are separated by lowland fluvial valleys. This topographic structure led glacio-eustatic sea-level oscillations to turn mountain massifs into paleo-islands [42-47]. In particular, these dynamics repeatedly insularized the Sila and Aspromonte massifs, heavily affecting the population structure of terrestrial animal species inhabiting these areas, and leaving detectable imprints in their current genetic structure [14,22,23,16]. This scenario is concordant also with the genetic structure that we identified for the Hermann’s tortoise in southern Italy. Indeed, looking at the distribution of the three genetic clusters, it is possible to identify at least three putative areas acting as Pleistocene refugia for Hermann’s tortoise populations: one located in the southern part of the Calabrian region, one located somewhere north of this area, and one located in Sicily. In the absence of molecular dating analyses, our hypothesis on the Pleistocene history Hermann’s tortoise populations in southern Italy should be taken with caution. However, support for this hypothesis comes also from the fossil record, which identified sites where populations survived during the Late Pleistocene to be located mainly in the southern part of the peninsula, between the Campania and the Calabrian regions [48]. Further investigations involving a higher number of markers (e.g. SNPs and/or nuclear sequence markers) could shed light on the demographic and evolutionary histories of the three lineages.
Our data clearly identified two areas of genetic admixture, one located in the northern and one in the southern edge of the Calabrian region. These areas, likely originating from secondary contacts among distinct lineages, closely match with areas of secondary contact and admixture observed in several other taxa [14,22,23,49]. Within these areas, gene flow between differentiated lineages boosted the level of population genetic diversity, leading to the comparatively high values of both heterozygosity and allelic richness observed (see Table 1). As a consequence, the whole Calabrian region emerge as a structured hotspot of intraspecific genetic variation for the Hermann’s tortoise, where both unique lineages and high levels of population genetic diversity are found.
Our results have remarkable implications for the management of the Hermann’s tortoise populations. Because of the widespread population decline, T. hermanni is considered as Near Threatened by the IUCN red list of threatened species at a global scale [50] and as Endangered in Italy [51]. However, we identified three unique evolutionarily significant units [52], two of them with narrow and endemic ranges. Assessments of their demographic consistence, as well as of the current threats to their populations have to be planned in the near future, in order to integrate the genetic information into the regional strategies for biodiversity conservation. Furthermore, identifying a hotspot of intraspecific genetic variation in a previously under-investigated region claims for a more detailed investigation on the status of populations inhabiting the hotspot, which might represent a valuable resource for the conservation and management of this species [5]. Indeed, intraspecific genetic variation provides populations with the potential to adapt to the ongoing changes in their biotic and abiotic environment [53,54]. At the same time, because of the link between genetic diversity and effective population size, these populations are less likely to be affected by the detrimental consequences of genetic drift and inbreeding depression [5,55-57]. Thus, Hermann’s tortoise populations from the southern part of the Italian Peninsula clearly represent a conservation priority for this species. Finally, the sharp genetic structure identified here, which define the proper geographic distribution of the distinct management units, provides valid support for more informed relocation programs of confiscated animals in the wild [30].