Abstract
Highlands, hydrographic systems and coastal areas have been hypothesised to form corridors across the hyperarid Sahara desert in North Africa, allowing dispersal and gene flow for non-xeric species. Here we aim to provide a genetic test for the trans-Saharan corridor model, and predict the location and stability of ecological-corridors, by combining phylogeography and palaeoclimatic modelling. The model was the Psammophis schokari (Schokari sand racer) group, fast-moving and widely distributed generalist colubrids occurring mostly in arid and semiarid scrublands. We combined dated phylogenies of mitochondrial and nuclear markers with palaeoclimatic modelling. For the phylogeographic analysis, we used 75 samples of P. schokari and P. aegyptius, and Bayesian and Maximum-Likelihood methods. For the ecological models, we used Maxent over the distribution of P. schokari and West African lineages. Models were projected to past conditions (mid Holocene, Last Glacial Maximum and Last Inter-Glacial) to infer climatic stable areas. Climatic stability was predicted to be mostly restricted to coastal areas and not spatially continuous. A putative temporary trans-Saharan corridor was identified in Eastern Sahara, with a more stable one along the Atlantic coast. Six parapatric lineages were identified within P. schokari, four occurring in North Africa. These likely diverged during the Pliocene. The Tamanraset River might have been a vicariant agent. African lineages may have experienced further subsequent diversification during the late Pleistocene. The main P. schokari refugia were probably located along the northern margins of the Sahara, allowing its North-to-South colonization. Trans-Saharan corridors seem to have played a role in P. schokari biogeography, allowing colonization of central Saharan mountains and Sahel. Some might have worked as refugia, and even the most stable corridors may have sections working as filters, depending on each climatic phase. We expect the use of trans-Saharan corridors to decrease for more mesic species or with less dispersal capabilities.
Highlights
Numerous geological and climatic events have affected the geographic and biological diversity of North Africa in the last few million years (Le Houérou 1997; Fabre 2005)
Genetic identity of specimens was coherent with prior morphology-based species assignment: specimens identified as Psammophis schokari and P. aegyptius formed two reciprocally monophyletic groups in the concatenated phylogenetic Maximum Likelihood (ML) and BI trees (Fig. 2)
The nuclear DNA did not exhibit a clear separation in the network analysis, with a shared basal haplotype in c-mos, and some RAG2 schokari haplotypes being closer to aegyptius haplotypes than to the remaining schokari alleles (Fig. 2)
Summary
Numerous geological and climatic events have affected the geographic and biological diversity of North Africa in the last few million years (Le Houérou 1997; Fabre 2005). Geological events include the opening of the Mediterranean to the Atlantic 7-9 million years ago (Ma), the subsequent closure 6 Ma and re-opening 5.3 Ma (MSC, Krijgsman et al, 1999), recurrent episodes of desiccation and refilling in the Red Sea area (Girdler 1991; Bosworth et al, 2005), marine transgressions (Tawadros 2011), or the Atlas mountains uplift (de Jong, 1998) These had climatic repercussions, but the most wide-ranging climatic event was a shift from tropical to arid environments around mid-Miocene (Zachos et al, 2001) that eventually led to the appearance of the Sahara desert between 7 Ma and 2.5 Ma (Schuster et al, 2006; Swezey 2009). Arid phases allowed range expansions of xeric taxa (Arnold et al, 2008; Kissling et al, 2016; Leaché et al, 2017; Pook et al, 2009), later broken during humid phases (Metallinou et al, 2015; Pook et al., 2009)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
