COMMENTED DISTRIBUTIONAL LIST OF THE REPTILES OF MAURITANIA (WEST AFRICA)

This is the first comprehensive review of the reptiles of Mauritania. It includes distributional information and comments. Mauritania harbors 86 species of reptiles belonging to 21 families. Among these families, Colubridae and Lacertidae are the most diverse, with 14 and 13 species respectively. Other families, such as Agamidae, Gekkonidae, Scincidae or Viperidae are also well represented. Among the 80 continental species, 47.5% are Saharan, 33.8% Afrotropical, 16.2% Sahelian and 2.5% Mediterranean. The marine turtles form another important group, with six species. Eight species are excluded from the country list because of old identification errors, there is not enough evidence of their presence or due to changes in political borders. Among the species expected to occur in Mauritania, at least nine may occur in Saharan environments, 13 in the Sahel savannah and two may have been introduced.


Introduction
Mauritania is one of the herpetologically least known countries of Africa.During the last twenty years, Morocco, West Sahara and even Mali have received considerable attention from herpetologists (Bons & Geniez, 1996;Schleich et al., 1996;Geniez et al., 2004).Mauritania was almost forgotten by naturalists after the French colonization and establishment of the current political borders.Most natural history information was produced before the sixties by French naturalists who collected from mainly coastal areas and populated places.Some expeditions took place during the beginning of the last century in what we know today as the Islamic Republic of Mauritania.Naturalists such as M. Chudeau, M. Audan, G. Bouet, T. Monod, and M. Murat collected most of the specimens that were subsequently studied by Pellegrin, Chabanaud, Angel, Villiers, etc.The first author that reported some collections from Mauritania was Moquard (1895), who wrote about the reptiles of Cap Blanc, in the former French West Africa.Subsequently, Pellegrin (1910) studied the specimens collected during an expedition directed by M. Gruve in 1908 between Saint Louis and Cap Blanc, and reported eight species.Chabanaud (1916Chabanaud ( , 1917) studied Audan's specimens collected during 1909, 1912and 1913. Also Boulenger (1919) contributed some information for the region, based on material deposited in the British Museum.However, he referred to "Mauritania" in a broad sense, probably including parts of West Sahara.He also studied some specimens collected by Audan in the Sahelian Trarza region.Böttger (1921) commented that the distribution of some reptiles found by him in the former Spanish colony of "Río de Oro", reached the Mauritanian Cap Blanc and further corrected one observation of Pellegrin about the coastal Stenodactylus species.Angel (1938Angel ( , 1939) ) studied the material collected by M. Murat during the expedition "D'études de la biologie des Acridiens", directed by M. Zolotarevsky in 1936 and 1937 along the central and northwestern part of the country.He described the subspecies Eryx muelleri subniger and reported information for another twelve species.
After these early contributions, Mauritania received more attention, mainly through the works of Villiers (1950Villiers ( , 1951Villiers ( , 1954Villiers ( , 1956Villiers ( , 1975)), who commented and described the distribution and/or some aspects of the natural history of at least 23 snakes from Mauritania.His snake checklist for Mauritania (Villiers, 1950) was the obligatory reference for subsequent revisions of the Mauritanian herpetofauna.Moreover, Dekeyser & Villiers (1956) completed the most thorough ecological and faunistical study performed to date in Mauritania.They analyzed the distribution by habitats and microhabitats of plant and animal species in the central area of Adrar, and reported the presence of many reptiles, some of which are very difficult to detect even today.After this relatively intense activity before Mauritania's independence in 1960, no other herpetological survey took place in the country, except those of Maigret (1975) and Le Toquin et al., (1980) on marine turtles.Nevertheless, more general taxonomic works included specimens from Mauritania (i.e Grandison, 1956Grandison, , 1961)).Already in the late seventies and eighties, the interest in African herpetology increased.Taxonomic revisions (Arnold & Levinton, 1977;Joger, 1980Joger, , 1981Joger, , 1984;;Salvador, 1982;Schätti, 1986;Hahn & Wallach, 1998) and biogeographical or distributional analyses (i.e.Maigret, 1983;Lambert, 1984;Hughes, 1983), included information on Mauritania and, furthermore, included specimens collected during earlier expeditions.These efforts led to the description of new species and contributed to the unravelling and/or identification of several taxonomic problems.Welch's (1982) list and Leberre's (1989) popular book were the more comprehensive lists published on Mauritanian herpetofauna at this time.But other studies in adjacent areas, for example Senegal (Böhme, 1978;Cissé & Karns, 1979) or Morocco (Bons & Geniez, 1996), served to illustrate how incomplete the herpetological knowledge of Mauritania was.In more recent times we have seen a steady increase in herpetological research in Mauritania.This was probably favoured by the rapid increase of knowledge of other West African countries (i.e.Bons & Geniez, 1996;Joger & Lambert, 1996;Schleich et al., 1996), and political stability.Thus, during the nineties and early 21 st Century, many naturalists contributed to the knowledge of the Mauritanian herpetofauna.Bons & Geniez (1996) provided some information for northern Mauritania; while Ineich (1997) compiled the most complete reptile list to date (33 species), as a result of his research along the coastal area; Arvy et al., (1996), Arvy & Dia (1997) and Arvy et al., (1997) increased our knowledge of turtles; many papers contributed with new discoveries and additional distributional information (Hasi et al., 1997;Lambert & Mullie, 1998;Donaire et al., 2000;Böhme, 2000;Böhme et al., 2001;Shine et al., 2001;Padial et al., 2002;Padial, 2003;Geniez & Luch, 2004;Aymerich et al., 2004).Moreover, some species were described from neighboring countries (Geniez & Foucart, 1995;Joger & Lambert, 1996;Mateo et al., 1998) and subsequently reported from Mauritania.Other species were described exclusively for Mauritania (Wallach, 2002;Padial, 2005;Geniez & Arnold, 2006).In addition, the systematics of some complex groups was analyzed (Mateo et al., 1998;Wilms & Böhme, 2001;Crochet et al., 2003).Finally Geniez et al., (2000;2004) and Chippaux (2001) added more herpetological information for Mauritania.Geniez et al.'s (2004) review was the most comprehensive herpetofaunistic list for northern Mauritania to date.
National species lists are of high relevance for conservation policy and scientific research (Padial & De la Riva, 2006).Biodiversity is managed at the political level, and international founding is prioritized for endangered species, biodiversity hotspots, areas of endemism, etc.In this context, the Mauritanian herpetofauna lies outside the scope of serious attention, since there is no scientific list at all [the first list for amphibians was recently published by Padial & De la Riva (2004)].The only available sources for general information on Mauritanian reptile diversity are on the web: the EMBL Reptile Database ( 2006) and the WCMC (2006).But they contain some outdated information and old mistakes.Whereas some countries around Mauritania have already been herpetofaunistically analyzed (see above), Mauritania urgently needs a reptilian species list.
My recent fieldwork in Mauritania and efforts in compiling and analyzing all available herpetological information from the country have led to this preliminary checklist.I have removed some erro-  neous records and also included taxonomic changes, new records, and a thorough literature revision.I also include a preliminary list of some species expected to occur.In fact, the reptile diversity in the country has been seriously underestimated and many areas need to be surveyed.Mauritania could play a very relevant role in conservation strategies and biogeographical analyses of the West African herpetofauna.I hope this list may facilitate all future observations on Mauritanian reptiles and provide a baseline for other herpetological studies and more general biodiversity research in the country.

Methods
Fieldwork was performed between April and August 2002.Specimens were collected only when they belonged to species with taxonomic problems or when they were not previously reported for the country.Common and easily identifiable species were collected by hand, photographed and released.Some specimens were found dead along the road or killed by local people.Specimens were sacrificed with injection of nicotine, fixed in 10% formalin, and preserved in ethanol 70º.They are deposited either in the Centro de Estudios Tropicales, Sevilla, Spain (CET) or the Museo Nacional de Ciencias Naturales, Madrid, Spain (MNCN).Tissue samples were deposited in the British Museum of Natural History (London).I tried to review and contrast all published literature about the herpetofauna of Mauritania.Nevertheless, it is possible that some localities, records or minor references may have been overlooked.In the text, scientific names are followed by the distribution and/or comments about the species.Localities are ordered by "wilayas" (main administrative divisions of the country, see Fig. 1).The wilayas are underlined.Localities are ordered alphabetically, and are followed by their respective references or the initials JMP in cases where I found the species in the respective locality.The Appendix compiles the coordinates of the localities extracted from the literature or visited by the author.Nevertheless, for some localities the coordinates could not be found.For general toponymy I followed the Carte Générale of Mauritania (1:2500000) (IGN, 1993).Some names may change when translated from French to Hassaniya or viceversa, hence, spelling of some localities can vary in different cartographies.
Agama castroviejoi Padial, 2005 DISTRIBUTION: Adrar plateau, road from Atar to Tidjikja (Padial, 2005); road from Atar to Chinguetti via Ebnou Pass (Brito, 2003); 44 km from Chinguetti on the road to Adrar, 19 km before the Nouatil Pass (Geniez & Arnold, 2006).COMMENTS: I assign the recent record of Agama boueti for the Adrar, near the type locality of A. castroviejoi, (Brito, 2003) to A. castroviejoi, a recently described species very similar to A. boueti and endemic to this mountain range.
Agama impalearis Böttger, 1874 DISTRIBUTION: Adrar: Amogjar pass, c. 50 km East of Atar (Brito, 2003); Road from Atar to Chinguetti (JMP).COMMENTS: Chabanaud (1917) cited Agama colonorum for "Mauritanie saharienne" based on a single individual collected by Audan.Despite Le Berre's (1989) distribution map for A. impalearis, where this species appears to occur widely across the country, no concrete records were known up to then.Joger (1979) noted that this species did not occur in Mauritania and that the nearest record was Sequia el Hamra (West Sahara).Nevertheless, Brito (2003) recently confirmed the presence of this species in the Adrar.I also observed some gregarious specimens perched on rocks in the Adrar plateau between Adrar and Chinguetti.These may correspond to the Southern form of A. impalearis, but no specimen could be captured for confirmation.

COMMENTS: Willms & Böhme (2001) elevated
Uromastyx acanthinura dispar to full species and Uromastyx acanthinura flavifasciata and U. mailensis were considered subspecies of the former.Assignation of specimens to species or subspecies of the Uromastyx acanthinura group have been commonly based on coloration and pholidosis (Mateo et al., 1998;Wilms & Böhme, 2001;Brito, 2003;Geniez et al., 2004).But some of the described subspecies occur in simpatry, which is not compatible with the definition of subspecies (Mayr, 1942).Hence, either some cryptic species occur in sympatry or it is a case of intraspecific polymorphism.
Nevertheless, the later authors mentioned significant differences between specimens of L. narirostris and L. boueti.Trape (2002) used the combination L. boueti.Under this author's view, the species shows differential characters [following Hahn & Wallach's, (1998) observations] that support its specific status.

TYPHLOPIDAE
Typhlops etheridgei Wallach, 2002 DISTRIBUTION: Adrar: between Atar and Choum (Wallach, 2002).COMMENTS: this species is endemic for Mauritania and only known from the type locality (between Atar and Choûm along Route Nationale 1, a distance of ca.85 air km through the Amsâga region, western Adrar Region, west-central Mauritania).
Hence, this species is expected to occur also in northern Mauritania.

Discussion
Some of the species cited in the literature or in the Internet databases should be removed from the lists.For example, Psammodromus algirus (Linnaeus, 1766) ranges from southern France to South Morocco (Bons & Geniez, 1996).The record for Mauritania without locality (see Welch, 1982) is surely an error and I therefore exclude this species from the list.Saurodactylus brosseti Bons & Pasteur, 1957, cited for the Banc d'Arguin (Mahe, 1985in Geniez & Bons, 1996), was surely mistaken for Tropiocolotes tripolitanus (Geniez & Bons, 1996).However, S. brosseti may occur in northern Mauritania since it has been cited for Zemmour and Tindouf (Geniez et al., 2004).Ptyodactylus hasselquistii hasselquistii (Donndorff, 1798) was cited for Mauritania without locality (Welch, 1982).In a revision of this group, Heimes (1987) did not cite this species for the country and the nearest record in his distribution map corresponds to P. h.ragazzi from central Mali.Although the presence of Ptyodactylus is likely, I prefer to exclude this genus from the list until there is additional evidence of its presence in the country.Chabanaud (1917) cited Tarentola delalandii for Saharan Mauritania, a species that is restricted to the Canary Islands.Although we cannot discard an accidental reintroduction of this species in coastal Mauritania, I prefer to exclude this species from the list until confirmation.The record for Typhlops punctatus (Leach, 1819) by Wallach (2002) from the dry savanna of southern Mauritania is a mistake (Wallach, pers. com.).This species has not been recorded for the country to date, although its presence is to be expected because it occurs in northern Senegal (Roux-Estève, 1974).Malpolon monspessulanus and Macroprotodon cucullatus (Geoffroy St. Hilaire, 1827) were cited by Boulenger (1919) based on collections from Trarza made by Audan (see Villiers, 1950), we consider these records as erroneous, as did Villiers (1950) when he tentatively assigned these records to southern Morocco.The presence of Naja haje (Linnaeus, 1758) is not confirmed although it is listed in the EMBL reptile database.Villiers (1950) considered the presence of this species very likely.It occurs in Senegal and Mali, and Chippaux (2001) also included Mauritania within the range of this species without giving any concrete record.Bons & Geniez (1996) and Geniez et al. (2004) cited this species for two localities in West Sahara, very close to the Mauritanian border.The absence of any concrete record for Mauritania is most likely due to a lack of sampling.Pelusios castaneus (Schweigger, 1812) is cited in the EMBL reptile database, but we were not able to locate any concrete record or reference for this species in Mauritania and, therefore, I tentatively exclude this species from the list.Mabuya quinquetaeniata (Lichtenstein, 1823) is listed in the WCMC web site for Mauritania without a concrete reference for corroboration.Nevertheless, this species inhabits adjacent areas in Mali (Joger & Lambert, 1996) and its presence in Mauritania is expected.
The reptile fauna of Mauritania is currently composed of 86 species belonging to 21 families (Table 1).Lacertidae and Colubridae are the most diverse families, and Agamidae, Gekkonidae, Scincidae or Viperidae are also well represented.Among the 80 continental species, 47.5% are Saharan, 33.8% Afrotropical, 16.2% Sahelian and 2.5% Mediterranean.The marine turtles form another important group, with six species.These data contrast with the 53 species (of which only 48 are valid) reported in the most complete and useful Internet database for reptiles (EMBL Reptile Database, 2006).Another database (WCMC, 2006) offers a list of 73 species (of which 65 are valid).The herpetological diversity of Mauritania is in urgent need of a thorough revision, as evidenced by the panorama of reptile diversity reported here.
In summary, more than one hundred reptiles may be present in Mauritania.These data confer on Mauritania a good position in reptile diversity.For example, Joger & Lambert (1997) reported 107 species for Mali, and 48 have been cited for Western Sahara (Geniez et al., 2004).However, we are far from having a complete panorama of the reptile diversity in Mauritania.Most records come from a handful of scattered localities and vast areas remain unexplored.For example, huge Saharan areas in the northern and northeastern parts of the country, and great part of the Sahel savannas in the south, lack any record of reptiles.Moreover, these savannas are highly endangered due to intense wood harvesting (mainly for firewood) and intense agropastoral use.These factors, together with long droughts have quickly transformed the environment.Therefore, most efforts should be oriented toward the study of diversity in the Sahel.The study of isolated wetlands in the Sahara, occupied by a mixture of Saharan and relict Sahelian elements, is also of high relevance if we are to understand the desert/savannah dynamics driven by climatic changes.

Table 1 .
-Number and percentage of species for each family of Mauritanian reptiles.Tabla 1.-Número y porcentaje de especies para cada familia de reptiles en Mauritania.