Abstract
The West Asian stripe-necked terrapin Mauremys caspica is widespread throughout the Middle East-a region for which only few phylogeographic studies are available. Due to landscape alteration, pollution and intensification of water management, M. caspica is increasingly threatened. However, genetic diversity among and within populations is poorly known, impeding the identification of management units. Using a nearly rangewide sampling, we analyzed 14 microsatellite loci and mtDNA sequences in order to gain insight into the population structure and history of M. caspica. In agreement with a previous study, we found two clusters of mitochondrial haplotypes, with one cluster distributed in the east and the other in the west of the range. However, our microsatellite data suggested a more pronounced geographical structuring. When null alleles were coded as recessive with structure 2.3.2, three clusters were revealed, with one cluster matching roughly the range of the western mitochondrial cluster, and the composite ranges of the two other microsatellite clusters correspond to the distribution of the eastern mitochondrial cluster. Na < ve structure analyses without correction for null alleles were congruent with respect to the two eastern microsatellite clusters, but subdivided the western cluster into two units, with an additional geographical divide corresponding to the 'Anatolian diagonal'aEuro"a well-known high mountain barrier impeding exchange between western and eastern taxa. In na < ve analyses, the westernmost microsatellite cluster (from Central Anatolia) is quite isolated from the others, and its distinctness is also supported by fixation indices resembling the values among the other three clusters. One of the two eastern clusters is distributed in the Caucasus region plus Iran, and terrapins from Saudi Arabia and Bahrain constitute the second eastern cluster, supporting the view that these endangered populations are native. Coalescent-based analyses of our microsatellite data reveal for all four clusters bottlenecks 4,000-20,000 years ago, suggesting that climatic fluctuations of the Late Pleistocene and Holocene played an important role in shaping current genetic diversity. We propose that each of the four identified clusters, including the Central Anatolian one, should be treated as a distinct management unit. The presence of non-native terrapins in the animal trade of Bahrain highlights the danger of genetic pollution of the endangered Arabian populations. Further sampling is needed to elucidate the situation in southern and central Iran and Iraq. Our results confirm that genetic data do not support the validity of any of the three morphologically defined subspecies of M. caspica, and we propose that their usage be abandoned.