1 Institut de Syst  ematique,  Evolution, Biodiversit e, ISYEB - UMR 7205 - CNRS, MNHN, UPMC, EPHE, Mus eum national d’Histoire naturelle, Sorbonne Universit es, Paris France; 2 Institut Scientifique de Rabat, Rabat Maroc; 3 CEFE, UMR 5175 CNRS EPHE, Montpellier Cedex 5, France Phylogeographic patterns in North African water frog Pelophylax saharicus (Anura: Ranidae) VIOLAINE NICOLAS 1 ,ABDERRAHMANE MATAAME 2 ,PIERRE-ANDR  E CROCHET 3 ,PHILIPPE GENIEZ 3 and ANNEMARIE OHLER 1 Abstract We studied the phylogeography of the Sahara frog in North Africa. We widely sampled frogs from Morocco to Tunisia (195 individuals) and sequenced two mitochondrial (16S and CO1) and one nuclear (Rag1) genes. Our results confirm that Moroccan populations of Pelophylax saharicus are genetically distinct from Algerian ones. Specimens from Alger and Djelfa (central Algeria) are genetically closer to Moroccan specimens than to east Algerian ones, and the split between these two lineages may have occurred approximately 2.6 Mya. A similar pattern of differentiation was observed in several other species and was hypothesized to be linked to the formation of the fossil island called the ‘Edough Peninsula’ in eastern Alge- ria around 4.2 Ma and then to have been reinforced by Pleistocene climatic changes. At the Moroccan scale, we found a low level of genetic diversity and no clear phylogeographic pattern within P. saharicus. However, our SAShA analyses revealed a mixture of random and underdistributed haplo- types, which may indicate a complex population genetic or biogeographic history. Key words: Amphibian – Maghreb – Moulouya – phylogeography – taxonomy Introduction Paleoenvironmental data and climatic reconstructions show that throughout the Quaternary, North Africa regularly shifted from wetter to drier climatic conditions (DeMenocal 2004; Schuster et al. 2006). This climatic instability led to periodic modifica- tions of habitats, including expansion/contraction of the Sahara desert, emergence of more open habitats (savannas) or deep modifications of hydrographic networks (Rognon 1989; Le Hou erou 1997; Drake et al. 2011). In addition to the Sahara des- ert, which is known to have acted as a barrier preventing exchanges between north and central Africa and has promoted vicariant speciation (Douady et al. 2003; Ben Faleh et al. 2012), the Atlas Mountains also act as a barrier (Aulagnier and Theve- not 1986). Other barriers such as the arid corridor of the Mou- louya River valley are known to represent obstacles to population dispersal, especially for small vertebrates (Alvarez et al. 2000; Escoriza et al. 2006; Barata et al. 2008). Owing to their general low individual mobility, amphibian populations tend to be highly structured genetically over short geographical distances and retain high-resolution signals of his- torical events that generated current species distribution (Zeisset and Beebee 2008). They are thus good candidates for inferring biogeographic processes responsible for current pattern of biodi- versity distribution. The Sahara frog Pelophylax saharicus (Bou- lenger, 1913) is widely distributed (Fig. 1) from Morocco through Algeria and Tunisia to northern Libya and north-western Egypt. It is abundant where suitable aquatic habitats exists (from sea level to 2670 m asl) and is the most common amphibian of the Maghreb region (Donaire-Barroso et al. 2009; Beukema et al. 2013). The species thus represents an interesting species to better understand biogeographic patterns in North Africa. Pelophylax Fitzinger, 1843, is a genus of true frogs wide- spread in Eurasia, with a few species ranging into northern Africa. It was considered for a long time as belonging to the genus Rana Linnaeus, 1858, but recent molecular studies showed that it is a distinct lineage recognized as a distinct genus (Frost et al. 2006; Stuart 2008; Pyron and Wiens 2011; Frost 2014). Within this genus, the taxonomy of the Sahara frog has long been controversial. Hemmer et al. (1980) suggested from electro- phoretic studies of the albumin alleles that the North African Pe- lophylax might form a hybridogenetic system with diploid and triploid individuals of the hybrid taxon found in most popula- tions investigated from Morocco to eastern Algeria. This is at odd with subsequent studies based on allozymes which did not detect mixture of different genotypes in any of the populations sampled, in spite of a similar geographical coverage (Buckley et al. 1994, 1996; Arano et al. 1998) and with karyological data from Tunisia and eastern Algeria that did not detect any triploid individual (Amor et al. 2010a). Acoustic similarities between North African and Iberian water frogs led Steinwarz and Schnei- der (1991) to treat P. saharicus as a synonym of Pelophylax perezi, a systematic hypothesis clearly invalidated by their deep genetic divergence (Buckley et al. 1994; Pl€ otner et al. 2012). Consequently, P. saharicus is now treated as a valid species by most taxonomic authorities, and its sister species is P. perezi (Lymberakis et al. 2007; Beukema et al. 2013; Frost 2014). Across the Maghreb, P. saharicus is composed of two distinct clades, one occurring in Morocco and the other one in Algeria and Tunisia (Buckley et al. 1994, 1996; Arano et al. 1998; Har- ris et al. 2003). According to Arano et al. (1998), the geographi- cal limit between the two groups could be the Moulouya River corridor, but this has never been properly investigated. Accord- ing to these authors, these two clades should be recognized as different subspecies: P. saharicus saharicus (eastern clade) and P. saharicus riodeoroi (Salvador and Peris, 1975) (western clade). Within Tunisian P. saharicus, several morphotypes were recorded, but this morphological variation probably results from phenotypic plasticity in response to environmental factors, as lit- tle genetic variability was observed within this country and it was not partitioned across morphotypes (Amor et al. 2009, 2010a,b,c; Farjallah et al. 2012). The occurrence of different phenotypes within Moroccan P. saharicus has been described as well (Bons and Geniez 1996), which led some authors to suggest the existence of multiple taxa. However, preliminary genetic data based on allozymes and 12S gene sequencing of few specimens Corresponding author: Violaine Nicolas (vnicolas@mnhn.fr) Contributing authors: Abderrahmane Mataame (abdel_mat@hotmail.fr), Pierre-Andr e Crochet (pierre-andre.crochet@cefe.cnrs.fr), Philippe Geniez (Philippe.GENIEZ@cefe.cnrs.fr), Annemarie Ohler (ohler@mnhn.fr) Accepted on 25 December 2014 © 2015 Blackwell Verlag GmbH J Zoolog Syst Evol Res doi: 10.1111/jzs.12094