Mitochondrial phylogeography and subspecies of the wide-ranging sub-Saharan leopard tortoise Stigmochelys pardalis (Testudines: Testudinidae) - a case study for the pitfalls of pseudogenes and GenBank sequences

The leopard tortoise (Stigmochelys pardalis) is the most widely distributed sub-Saharan tortoise species, with a range extending from the Horn of Africa all over eastern Africa to the Republic of South Africa, Namibia and southernmost Angola. Using 1938 bp of mitochondrial DNA (cyt b gene, partial ND4 gene plus adjacent tRNA genes) from a nearly range-wide sampling, we examined its phylogeographic structure and compared our findings with previously published GenBank sequences. We identified seven major clades that are largely parapatrically distributed. A few records of distinct haplotypes at the same locality or in close proximity could be the result of translocation of tortoises by man. The greatest diversity occurs in the south of the species' range, with five out of the seven clades. Testing for isolation-by-distance suggests that the observed phylogeographic structure is the result of restricted geographical gene flow and not of historical vicariance. This is in sharp contrast to wide-ranging thermophilic reptiles from the western Palaearctic, whose phylogeographic structure was significantly shaped by Pleistocene range interruptions, but also by earlier dispersal and vicariant events. Most cyt b sequences of S. pardalis from GenBank turned out to be nuclear pseudogenes, or to be of chimerical origin from such pseudogenes and authentic mitochondrial sequences, which argues for caution regarding uncritical usage of GenBank sequences. The recent revalidation of the two subspecies of S. pardalis was based on such a chimerical sequence that was erroneously identified with the subspecies S. p. babcocki. Furthermore, according to our data, the distribution of mitochondrial clades does match neither the traditional subspecies ranges nor the pronounced geographical size variation of leopard tortoises. We conclude that there is no rationale for recognizing subspecies within S. pardalis. © 2010 Blackwell Verlag GmbH.