SQUAMATA

Rivas, Gilson A., Molina, César R., Ugueto, Gabriel N., Barros, Tito R., Barrio-Amorós, César L. & Kok, Philippe J. R., 2012, 3211, Zootaxa 3211, pp. 1-64 : 37-48

Order SQUAMATA

Ameiva ameiva . —Considered a polytypic species with 10 subspecies, four of them occurring in Venezuela. The validity of these subspecies was questioned by Vanzolini (1986). Avila-Pires (1995) followed Vanzolini noting that a thorough study of the variation in this species is necessary.

Ameiva bifrontata . — Peters and Donoso-Barros (1970) recognized four subspecies of A. bifrontata , two of them from Venezuela, the nominal form on the mainland and A. b. insulana restricted to Los Testigos islands. The main difference presented by Ruthven (1924) to distinguish between the populations was the discontinuity of the brachials and antebrachials in the Los Testigos form. We do not believe that this represents more than a clinal variation and until more extensive and detailed studies of the insular form are made, we prefer to treat it as a synonym of the nominal mainland subspecies.

Ameiva provitaae . —This species is a possible synonym of A. bifrontata and, until now, was not included in any taxonomic study. The characters thought to be unique to diagnose this species (such as a red snout, see García-Pérez 1995a) are present in different populations of A. bifrontata (GU, unpubl. data). Additional data are needed to clarify the status of this species.

Anolis species. —The taxonomic status of the genus Anolis sensu lato has been unstable and in constant change. Molecular data ( Nicholson 2002) indicate that within Anolis there is a distinct clade ( Norops ) morphologically characterized by having the autotomy septa lying anterior to anteriorly directed transverse process ( Etheridge 1960). However, the other clade ( Anolis ) is not monophyletic and, although the name Norops is recognized by several authors, we refrain from doing so as this would render the remaining anoles a paraphyletic group. The monophyletic status of Anolis (including Norops ) has been recently reassessed based on osteology, internal anatomy, chromosomes, molecular data, allozymes, and immunology ( Poe 2004).

Anolis biporcatus . — The species is reported for the first time from the Maracaibo Lake Basin based on MBLUZ 992 , a specimen collected in a gallery forest along the Palmar river ( Km 75 on the road Maracaibo-Machiques , Zulia state, Venezuela). Previous records in Venezuela are from the Mountain Range of Perijá and the Mountain Range of Mérida ( Rivas Fuenmayor & Barrio-Amorós 2003) .

Anolis bonairensis . — Meek (1910) reported a specimen ( FMNH 2684, Fig. 7) as A. alligator from Las Aves Archipelago. GU examined this specimen, which indeed agrees in most diagnostic characters with A. bonairensis . We acknowledge the presence of this species in Las Aves Archipelago, but stress that it needs to be formally confirmed by additional material.

Anolis chrysolepis vs Anolis nitens . —The nomenclatural history of these species has been controversial (see Myers & Donnelly 1997, Myers 2008, Cadle 2009 for further details). We agree with these authors and no longer recognize the name Anolis nitens . D’Angiolella et al. (2011) reviewed the Anolis chrysolepis species group based on morphology and molecular data and elevated all subspecies of A. chrysolepis to species status. The species present in Venezuela is thus A. planiceps (type locality Caracas, Venezuela).

Anolis eewi . — Vanzolini and Williams (1970) synonymised this species with A. chrysolepis . Gorzula and Señaris (1999) noted that the species is distinguished from A. chrysolepis by shorter legs. Myers and Donnelly (2008) synonymised the species with A. chrysolepis again. We believe that Anolis eewi is a valid species, but until new data are obtained, A. eewi is not included in the list presented herein.

Atractus tamessari . —This recently described species ( Kok 2006) was known only from Kaieteur National Park, western central Guyana. In the description of Atractus surucucu from Roraima State in Brazil, Prudente and Passos (2008) cited two specimens of A. tamessari from Venezuela in their list of material examined (MHNLS 15124 from Uei-tepui and MHNLS 5950 from “El Dorado-Santa Elena de Uairen”). Since Prudente and Passos (2008) erroneously reported A. tamessari as having 17 dorsal scales rows (instead of 15) the exact identity of these museum specimens is doubtful. Only one of these individuals was available to us, MHNLS 15124, which is a juvenile female of an Atractus species having 15 dorsal scales rows that fits well the description of A. tamessari . We thus formally report the presence of A. tamessari for the first time in Venezuela and expand its known distribution ca. 130 km SWW. The specimen listed as A. tamessari from “El Dorado-Santa Elena de Uairen” by Prudente and Passos (2008) should be better examined to ascertain its identity.

Atractus univittatus / fuliginosus . —The original description of A. fuliginosus was based on a single specimen from “200 milles of Caracas”, Venezuela and the status of that species remains unclear (as well as the precise type locality). Atractus fuliginosus was later recorded from the northeastern part of the country ( Cerro Turimiquire), but that population has since been described as A. matthewi ( Markezich & Barrio-Amorós 2004) . Adding some more confusion, in their description of Atractus eriki , an Andean species, Esqueda et al. (2005a) mentioned a specimen of A. fuliginosus from San Juan de Los Morros, a locality lying in the Central Coastal Range with which they compared their new taxon. Surprisingly, these authors did not compare A. eriki with A. univittatus (even if it is listed in their material examined), despite the fact that both species possess 17-17-17 dorsal scale rows and that the localities provided for the A. univittatus specimens they examined (Guanare and Barinitas) are closer to the Venezuelan Andes than San Juan de Los Morros. It is not clear whether Esqueda and colleagues considered univittatus a synonym of fuliginosus , but it should be noted that if Coluber fuliginosus Hallowell 1845 and Rhabdosoma univittatus Jan 1862 are synonyms, the name fuliginosus would have priority. We unfortunately did not examine the holotype of A. fuliginosus and cannot confirm or infirm this possibility. No additional specimen of A. fuliginosus has been collected in recent years ( Roze 1966; Lancini 1979) and a revision of the type material of A. fuliginosus compared to A. univittatus is badly needed to resolve this issue.

Bachia flavescens complex. — Dixon (1973) recognized two species in the B. flavescens group, each with two subspecies: B. flavescens flavescens , B. f. schlegeli, B. monodactylus monodactylus and B. m. parkeri. Hoogmoed (1973) suggested that B. monodactylus is a synonym of B. cophias . Dixon (1973) distinguished the subspecies of each species based on “weak” and variable characters. According to Dixon (1973) few characters distinguish B. flavescens from B. monodactylus ( B. cophias sensu Hoogmoed 1973 ) and additional material should attest that the two species are conspecific. Avila-Pires (1995), concluded that B. flavescens is a monotypic, highly variable species. Kizirian and McDiarmid (1998) considered that the variation in what is known as B. flavescens is impressive and that the taxon should be considered as a species complex. We agree with Kizirian and McDiarmid (1998) and consider Bachia flavescens to be a species complex in need of revision.

Bachia heteropa ssp. — Bachia heteropa is a polytypic species composed of five subspecies, four of which are present in Venezuela. Populations of the genus Bachia in northern Venezuela, Trinidad, Tobago, Grenada, and the Grenadines were referred to six separate taxa (B. h. heteropa , B. h. alleni , B. h. trinitatis , B. h. lineata , B. h. marcelae and B. h. anomala ) ( Dixon 1973). Burt and Burt (1933) synonymised B. anomala (from El Mene, Falcón State) with B. lineata . Thomas (1965) considered B. trinitatis to be a subspecies of B. alleni . Later, Donoso-Barros (1968) considered B. marcelae to be a subspecies of B. lineata , retaining B. heteropa and B. alleni as separate species. The revision of the genus by Dixon (1973) provided a better understanding of the group. Dixon noted that there seems to be a progressive reduction in size and loss of prefrontal scales as well as toes from east to west. After examining most available specimens, Dixon (1973) considered all populations in northern Venezuela, Trinidad, Tobago, Grenada and Grenadines to be B. heteropa , which was the oldest name available for this group of lizards. Dixon considered the east to west cline an adaptation to more fossorial habits. Intermediate specimens between B. h. heteropa and B. h. trinitatis are known from eastern Venezuela but intermediates between B. h. lineata and B. h. heteropa are unknown although both subspecies have been recorded from nearby areas (B. h. heteropa from la Guaira, Vargas State and B. h. lineata from Distrito Federal, Caracas). More material is definitely needed to clarify the status of these subspecies. We consider that the significant amount of variation among populations of this taxon could indicate a complex of species, but a taxonomic revision of this taxon is out of the scope of this paper and we decide to treat B. heteropa as a monotypic taxon until this situation is resolved.

Bothriechis schlegelii . —This species was reported from Ureña (Pifano et al. 1950), a semi-arid area in the Táchira depression, Táchira State. Pifano et al. (1950) provided a photograph of an individual, but no catalogue number or name of a collection housing a voucher specimen. The species has been cited from a Colombian locality nearby ( Campbell & Lamar 2004), and despite the absence (loss?) of a voucher specimen we consider this species as belonging to the Venezuelan herpetofauna, stressing that a new voucher should confirm its presence. Bothriechis schlegelii usually inhabits tropical and mountain rainforest. It must be noted that Sandner-Montilla (1965) apparently also mentioned the species from the Amazonas State (as stated in Campbell & Lamar 2004). Campbell and Lamar (2004) considered that these records are likely erroneous and that they must be confirmed. We were unable to obtain a copy of Sandner-Montilla (1965), but we read a copy of the second edition of that booklet (Sandner- Montilla 1975), in which there is no mention of B. schlegelii from the Amazonas State. We assume that the previous records were indeed erroneous and were removed by the author in the second edition of his work.

The Bothrops asper -colombiensis-isabelae problem. —For many years a great deal of controversy has arisen about the proper name to be given to the Bothrops species inhabiting the lowlands of the Venezuelan Coastal Range. One of these names is B. colombiensis , used by numerous authors for a species present north of the Orinoco River ( Lancini 1979). The type locality, “ Republic of Colombia, within two hundred miles of Caracas, Venezuela ” could be any locality in northern Venezuela instead of Colombia because the closest border with Colombia is located 700 km SW (approximately 450 miles) from Caracas.

Sandner-Montilla (1979a) synonymised B. colombiensis with B. lanceolatus , the latter currently considered endemic to the island of Martinique, Lesser Antilles ( Campbell & Lamar 2004; Fenwick et al. 2009), which is 437 km off mainland Venezuela. The same author later described B. l. aidae ( Sandner-Montilla 1981), based on a specimen from the Colección de ejemplares vivos de la Familia Crotalidae del Serpentario del Instituto Venezolano de Ofidiología, Caracas (Collection of living specimens of the family Crotalidae of the Serpentarium of the Venezuelan Institute of Ophidiology, Caracas) from “Selvas de Guatopo, Estado Miranda ”. Sandner-Montilla (1990) also described B. l. nacaritae from “Cariprima, N. de Valencia". “Cariprima” is not found on any available map, but both Estado Miranda and Valencia are in the Coastal Range. Both subspecies were considered synonyms of B. atrox by Campbell and Lamar (2004), although they probably meant B. asper since the Coastal Range is outside the geographic distribution of B. atrox ( Campbell & Lamar 2004: 377, map 57), but inside the geographic distribution of B. asper ( Campbell & Lamar 2004: 372, map 56). In any case, we consider the populations from Valencia and Guatopo as B. asper rather than B. atrox , except the population from the cloud forest that is B. venezuelensis .

Johnson and Dixon (1984) considered B. colombiensis a synonym of B. atrox , but at least four species were involved under what they named B. atrox (as stated by Campbell & Lamar 2004). Sandner-Montilla (1979b) described an additional species from Guanare, in the foothills of the Venezuelan Andes : Bothrops isabelae (also based on living specimens kept in captivity under number “Bi No. 1”). Markezich and Taphorn (1993) in a study on the variation of some populations of Venezuelan Bothrops concluded that B. isabelae is not valid and synonymised it with B. atrox . Fenwick et al. (2009) in the first attempt to use both morphological and molecular evidence to study South American pit vipers, included sequence fragments of B. isabelae obtained from Wüster et al. (2002) (no specific locality was provided) in their analyses. In all analyses by Fenwick et al. (2009), B. isabelae falls together with B. atrox , with a seemingly low genetic distance. Nevertheless, Fenwick et al. (2009) included B. isabelae in their list of Bothrops species (therefore apparently recognizing it as a valid taxon).

Some authors (e.g. Lancini 1979; Navarrete et al. 2009) still use B. colombiensis to refer to the most common pit vipers that inhabit northern Venezuela without any justification. These authors also consider B. isabelae to be a valid species (see Navarrete et al. 2009). We do not consider B. colombiensis , B. isabelae , B. l. nacaritae, and B. l. aidae as valid taxa until comparative systematic analyses involving both morphological and molecular evidence from specimens from the type localities are conducted. We follow Campbell and Lamar (2004) in the use of B. asper and B. atrox for these populations of northern Venezuela.

Colubridae / Dipsadidae . —Two papers on the molecular phylogeny of the Colubroidea appeared recently, both suggesting changes to the current taxonomy of this group of snakes ( Hedges et al. 2009; Zaher et al. 2009). These papers support the recognition of the subfamilies Xenodontinae and Dipsadinae within the family Dipsadidae Bonaparte, 1838 (including also the new subfamily Carphophiinae ). Other changes suggested by these authors involving genera occurring in Venezuela are the resurrection of the genus Lygophis and the synonymisations of Lystrophis and Waglerophis with Xenodon , and of Erythrolamprus with Liophis . Curcio et al. (2009a) mentioned that the proposal of synonymising Erythrolamprus with Liophis is premature in view of the insufficient evidence provided by Zaher et al. (2009), proposing maintenance of the traditional recognition of both names as valid until new data become available based on a more comprehensive sampling with proper technical supporting evidence (e.g. inclusion of type species). Curcio et al. (2009a) also mentioned a similar problem with the resurrection of Lygophis , pointing out that such a taxonomic change should have been guided by reference to the type species of the genus, e.g. Herpetodryas lineatus Schlegel (= Liophis lineatus ), which was not sampled by Zaher et al. (2009). However, we consider the arguments provided by Zaher et al. (2009) for the recognition of Lygophis sufficient for the resurrection of that genus, which is adopted in our list.

Although Zaher et al. (2009) regarded Xenoxybelis as a synonym of Philodryas , we are not convinced by their arguments and prefer to wait for more analyses involving a larger number of species of the genus Philodryas . Vidal et al. (2000) also found Xenoxybelis nested within Philodryas . However, as mentioned by these authors, a more detailed phylogenetic analysis of the newly extended genus Philodryas might show the necessity of a partition of some Philodryas and the recovered monophyletic subunits.

Other results of Zaher et al. (2009) involve the allocation of most West Indian snakes to the tribe Alsophiini , a group also studied by Hedges et al. (2009) who proposed a new reclassification for these snakes. Hedges et al. (2009) supported some of the changes proposed by Zaher et al. (2009), one of them being the partitioning of Colubridae . According to their proposal, the following subfamilies and genera are now considered as belonging to the family Dipsadidae : Dipsadinae : Atractus , Dipsas , Imantodes , Leptodeira , Ninia , Plesiodipsas , Sibon and Urotheca . Xenodontinae : Thamnodynastes (Tachymenini) , Taeniophallus (Echinaterini) , Clelia , Oxyrhopus , Phimophis , Pseudoboa , Siphlophis (Pseudoboini) , Philodryas (Philodryadini) , Hydrodynastes (Hydrodynastini) , Hydrops , Helicops , Pseudoeryx (Hydropsini) , Liophis , Erythrolamprus , Umbrivaga and Xenodon (Xenodontini) . Dipsadidae incertae sedis: Enulius . Other genera currently remain in Colubridae .

Cercosaura phelpsorum . — Euspondylus phelpsi was originally described by Lancini (1968) on the basis of a single female supposedly from Cerro Jaua (corrected to Cerro Sarisariñama by Steyermark & Brewer-Carías 1976). In his revision of the genus Prionodactylus, Uzzell (1973) did not reallocate E. phelpsi to Prionodactylus because a double widened row of gular scales—one of the diagnostic characters of the genus Prionodactylus according to Uzzell—was not mentioned in the original description. In 1996, Myers and Donnelly described a new species, E. goeleti from Cerro Yavi, Amazonas State, which according to the authors appears to be closely related to E. phelpsi but differs from it by having a pale vertebral stripe distinctly broadening posteriorly (vs maintaining the same width), greenish white ventral colouration (vs slate gray with black speckling), and distinctly keeled dorsal scales (vs weakly keeled). These authors also corrected the name of the species from E. phelpsi to E. phelpsorum ( Myers & Donnelly 1996) . Gorzula and Señaris (1999) reallocated the species to the genus Prionodactylus without providing any argument, and proposed to synonymise P. goeleti with P. phelpsorum based on (1) the fact that all the scale counts given by Lancini fall within the variation given by Myers and Donnelly (1996), (2) the subjectivity of the difference between the weakly keeled scales of P. phelpsorum and the distinctly keeled scales of P. goeleti , and (3) the likely misinterpretation of the ventral colouration and shape of the middorsal stripe. Soon after, Mijares-Urrutia (2000) reported the holotype of P. phelpsorum lost and concluded that in the absence of a neotype and additional material allowing further comparisons P. goeleti should be removed from the synonymy of P. phelpsorum . Myers and Donnelly (2001) partly agreed with Gorzula and Señaris (1999) mentioning that although neither Euspondylus nor Prionodactylus have enough characters that uniquely define each genus, the allocation of E. phelpsorum in Prionodactylus “has the practical advantage of associating them in the same group with the similar-appearing Andean Prionodactylus vertebralis ”. However, Myers and Donnelly (2001) disagreed with Gorzula and Señaris (1999) stating: “In allocating the name Euspondylus goeleti to the synonymy of Prionodactylus phelpsorum, Gorzula and Señaris apparently confused ontogenetic variation with the kind of extensive intraspecific variation that is not correlated with age, sex or geography”. Myers and Donnelly (2001) insisted again on the differences in width between the vertebral stripe present in P. phelpsorum and that of P. goeleti , pointing out, with the support of photographs, that the vertebral stripe of the former species does not widen posteriorly but is equally narrow, as it is not the case in all known specimens of P. goeleti (less pronounced in juveniles). Myers and Donnelly (2001) considered the difference in ventral colouration as a secondary diagnostic character and noted: “But we took Lancini at face value that the throat and venter of the female specimen of phelpsorum were slate gray in life, and assumed that if the ventral ground colour in life had been pale he would have said so. If, however, he had misleadingly extrapolated ventral colour from the preserved specimen, it would be a different matter since female goeleti became light gray in preservative...”. Myers and Donnelly (2001) also noted that the two species belong to geographically remote populations, reinforcing their arguments. In her recent revision of the genera Cercosaura , Pantodactylus , and Prionodactylus, Doan (2003) concluded that P. goeleti shows variation in the size of the gular scales. One of the specimens examined by Doan exhibited a complete double row of gulars, another had no difference in size between the midventral gulars and the ventrolateral gulars, and another one had a double row widening posteriorly only. Doan (2003) also pointed out that this can be seen in the photographs provided by Myers and Donnelly (1996, 2001). Doan (2003) concluded: “I would be hesitant to decide on its generic status except for its obviously close relationship with P. nigroventris ... a species that possesses all the diagnostic characters of Prionodactylus . Based on these similarities, I retain this species in Prionodactylus ”. Doan (2003) also synonymised the genera Pantodactylus and Prionodactylus with Cercosaura and considered C. goeleti a junior synonym of C. phelpsorum .

We examined four museum specimens identified as Cercosaura phelpsorum , three from the type locality of C. goeleti (MHNLS 11135, MHNLS 11141−42, all collected at 2550 m elevation), and one individual from Cerro Huachamacari, a mountain located about 200 km SWW from Cerro Sarisariñama and about 230 km S from Cerro Yavi (MHNLS 12518, collected at 1700 m elevation), and an unidentified Cercosaura from Cerro Autana ( Fuentes & Rivas Fuenmayor 2000), a mountain located about 220 km NW from Cerro Sarisariñama and about 180 km SW from Cerro Yavi (MBUCV 7114, collected at 1450 m elevation). We could not find any pertinent morphological difference between these five preserved specimens and the description of C. goeleti . The degree of keeling is variable among them, the middorsal-scale counts varied between 38–42 (variation not related to geographical distribution), the widening of the vertebral stripe is always present and seems variable and related with age, from 3 (one juvenile) to 4–5 scales (four adults) in the inguinal region (as also highlighted by Myers & Donnelly 2001). Unfortunately ventral colour in life is unknown, but it is similar among all preserved specimens, except the juvenile. Because none of those specimens has the vertebral stripe exactly as described by Lancini (1968) in C. phelpsorum , the state of that stripe apparently remains the only obvious morphological difference between C. goeleti and C. phelpsorum . In our opinion, this single character is not sufficient to separate species and might well be considered as interpopulational variation. This is strengthened by the fact that “ C. goeleti ” has a much wider range than previously thought, and that if the species is present in Cerro Autana and Cerro Huachamacari there is no reason that it would be absent from Cerro Sarisariñama, which lies in the same biogeographic area. Our new distributional data are additional arguments to agree with Gorzula and Señaris (1999) and Doan (2003) concluding that because of the absence of convincing diagnostic characters between C. goeleti and C. phelpsorum the former should be considered a junior synonym of the latter. However, new collection efforts are needed in the Jaua-Sarisariñama area in order to better understand the inter-and intrapopulational variation of C. phelpsorum since the most recent herpetological expedition to Sarisariñama ( Barrio-Amorós & Brewer-Carías 2008) has been unsuccessful in finding the species.

Cnemidophorus lemniscatus ssp. —Clearly a complex of species. Reeder et al. (2002) placed C. l. splendidus and C. arenivagus as sister taxa, suggesting that a specific status of C. l. splendidus deserves re-evaluation. Markezich et al. (1997) previously stated: “...future work may well result in elevating C. l. splendidus and other forms to recognition as species also”. C. J. Cole (pers. comm.) is currently working at the recognition of C. l. splendidus as a full species.

Coluber mentovarius ssp. — A species formerly placed in the genus Masticophis, a genus that has been synonymised with Coluber several times ( Savage 2002). Recent phylogenetic data corroborate that Masticophis cannot be separated from Coluber , and thus the former must be treated as a junior synonym of the latter ( Utiger et al. 2005). Herein, we adopt these taxonomic changes although they have not been universally accepted ( Collins & Taggart 2008). There is considerable variation within each of the Venezuelan subspecies, but they still appear to be well differentiated and separated geographically. Our data (unpublished) suggest that C. m. centralis has a same pattern of geographic distribution as other lowland colubrid species, such as Drymarchon caudomaculatus and Thamnodynastes paraguanae , which could be used as models to test biogeographic hypotheses in northern South America. The population of this subspecies in Panama should be re-evaluated, and the populations in Colombia and Venezuela assigned to C. m. centralis might deserve full species status.

Crotalus maricelae . —This species was described from an intra-Andean valley in the Cordillera de Mérida (García- Pérez 1995b). Crotalus maricelae was diagnosed based on cephalic scutellation, colour pattern and reduced adult size ( García-Pérez 1995b, Vanzolini & Calleffo 2002; Campbell & Lamar 2004). The original description is very brief and the species is virtually indistinguishable from C. durissus cumanensis . Esqueda et al. (2001) suggested that C. maricelae is a nomen nudum and that this population does not deserve recognition because all diagnostic characters fall within the variation range of C. d. cumanensis. La Marca and Soriano (2004) considered Crotalus maricelae a valid taxon, using the same characters mentioned in the original description. This species is not included in the list and we follow Campbell and Lamar (2004) in considering this taxon a synonym of Crotalus durissus .

Crotalus pifanorum . —The taxonomy of this species has been controversial because of the imprecise original description and the lack of type specimens. Originally described from the Llanos in southern Guárico (Sandner- Montilla 1980), C. pifanorum has been considered a subspecies of C. durissus by Campbell and Lamar (2004). In an undergraduate unpublished dissertation, Pérez-Bidó (1992) compared pifanorum and cumanensis (karyotype, protein electrophoresis and microdermatoglyphs) and supported the recognition of C. pifanorum as a valid species. Several years later, the thesis was published ( Pérez et al. 1997), but without providing a list of specimens examined and their geographic provenance. Pending additional data we still consider C. pifanorum a subspecies of C. durissus .

Crotalus vegrandis . —The taxonomic history of C. vegrandis has been controversial as well. Originally described as a distinct species by Klauber (1941), the same author ( Klauber 1956) relegated it to a subspecies of Crotalus durissus . Nevertheless, this arrangement was not followed by all authors, and C. vegrandis is often treated as a valid species ( Lancini 1967; Peters & Orejas-Miranda 1970). In their recent revision, Campbell and Lamar (2004) treated C. vegrandis as a subspecies of C. durissus . We consider C. vegrandis a full species mostly based on the following morphological differences: smaller length, peculiar colour pattern composed of gray and/or reddish brown ground colour with large white flecks distributed as irregular rings (vs rhomboidal blotches in C. durissus ). Crotalus vegrandis has an isolated distribution in the lowlands of Maturín and Anzoátegui States, northeastern Venezuela.

Dipsas copei . —We follow Harvey (2009) in considering that D. copei is not a synonym of Leptognathus incertus (= D. incerta ) and that D. incerta could be a valid species related to D. praeornata as first noted by Amaral (1929 a, 1929b) and subsequently by Harvey (2009).

Dipsas cf. indica . —This species is tentatively included in the Pantepui region based on a specimen reported from La Neblina (1515 m) by McDiarmid and Paolillo (1988) and by McDiarmid and Donnelly (2005). We did not examine that individual and its exact identity thus remains unconfirmed. The maximum known elevation reported for D. indica is 1000 m (see Natera-Mumaw & Battiston 2008).

Dipsas peruana / D. latifrontalis . — Dipsas latifrontalis was originally described from Aricagua in the Venezuelan Andes ( Boulenger 1905) and later recorded from some localities in Ecuador and the Venezuelan Coastal Range ( Peters 1960; Esqueda et al. 2001). Harvey and Embert (2009) resurrected the name D. praeornata for the Venezuelan Coastal Range population and relegated D. latifrontalis to a synonym of D. peruana . The status of the Ecuadorian populations assigned to this species is unclear ( Cadle 2005). We provide here preliminary data on the specimen CVULA 7883 from San Isidro, Estado Barinas, 1400 m, considered the second known specimen of D. peruana ( Fig. 8) from Venezuela (along with the holotype of D. latifrontalis at the BMNH). The specimen has 15- 15-15 dorsal scales, 192 ventral scales, 109 subcaudal scales, 8 supralabials and 11 infralabials.

Dipsas pratti . —This species was described based on a female from Medellin, Colombia at an elevation of 1700 m ( Boulenger 1897). Four more species of Dipsas were later described ( Cope 1899; Prado 1940, 1941; Werner 1916) and were found to be juveniles of the same species (the differences reported only reflected ontogenetic changes) and therefore synonymised with D. pratti ( Peters 1960) . However, the species’ systematics and natural history still remain unknown. The species is currently known only from a few specimens from the Cordillera Central and the Cordillera Oriental in Colombia, where it was recently reported in the departments of Santander and Bolívar ( Moreno-Arias et al. 2006). Unfortunately these authors do not provide any information on the morphological variation and natural history. Recently, several specimens assigned to D. pratti were collected on the Venezuelan side of the Sierra de Perijá (Estado Zulia) at 1600 m. A redescription of this species including the Venezuelan specimens is in press (Barros et al. in press).

Epicrates maurus . —Considered a subspecies of E. cenchria by some authors (e.g. Lancini & Kornacker 1989; Perez-Santos & Moreno 1988) or full species by others (e.g. Chippaux 1987; Gorzula & Señaris 1999; Barrio- Amorós & Díaz de Pascual 2008). Recently, Passos and Fernandes (2008), on the basis of external morphology, osteology and hemipenial characteristics, showed that E. maurus is a distinct species.

Gekkonidae . — Kluge (1967) considered five subfamilies in that family. Sphaerodactylinae was recently reviewed and elevated to family status ( Sphaerodactylidae ) containing the New World genera Coleodactylus , Gonatodes , Lepidoblepharis , Pseudogonatodes and Sphaerodactylus , and the Old World genera Saurodactylus , Quedenfeldtia , Aristelliger , Teratoscincus , Pristurus and Euleptes ( Gamble et al. 2008a) . Gamble et al. (2008b) recovered another monophyletic trans-Atlantic gecko clade ( Phyllodactylidae ) containing, among others, the genera Phyllodactylus and Thecadactylus . The Venezuelan species of Hemidactylus (native and introduced) remain in the family Gekkonidae .

Gonatodes superciliaris . — Barrio-Amorós and Brewer-Carías (2008) described G. superciliaris as a Sarisariñamatepui endemic. The new species is morphologically indistinguishable from G. alexandermendesi ( Cole & Kok 2006) and G. superciliaris is likely a junior synonym of this species. However, since preliminary phylogenetic analyses do not clearly corroborate this fact (see Schargel et al. 2010) and in the absence of additional data we still include the species in our list.

Gonatodes vittatus ssp.— Roze (1956) described an insular taxon, G. vittatus roquensis , from El Gran Roque, archipiélago de los Roques, Venezuela. One of us (GR) noted that the holotype (MHNLS 1031) is an adult specimen of G. antillensis (MHNLS 1031, erroneously cited as MNHLS 10800). The remaining series of paratypes deposited in MHNLS (1032–1043) and MBUCV (MBUCV 1906–1909) are G. aff. vittatus . Thus, we do not consider roquensis as a valid subspecies, but wish to stress that these insular populations should be examined using molecular data to confirm their taxonomic identity.

Helicops hogei . —Here we recognize the taxon as distinct, but a reassessment of the species is suggested in order to support its validity (see Rossman 2002).

Hydrops martii . —We report for the first time H. martii in Venezuela from the conjunction of the Casiquiare arm and the Siapa river in southern Amazonas State, based on photographs taken by CBA in 1998 during a fluvial expedition in search of turtles ( Fig. 9) ( Barrio-Amorós & Narbaiza 2008). The specimen was not collected due to the lack of permits at that time. Based on the available photographs, the specimen has 173 ventrals, and the typical pattern of the species, consisting of black rings of 1–2 scales wide surrounded by an incomplete one scale wide ring and separated from the other by a reddish ring of 2–3 scales wide. The photographs agree well with other published illustrations of the species ( Bartlett & Bartlett 2003; Campbell & Lamar 2004). We consider the pictures, kept in the collection of Fundación Andigena (FA 184–85), as photographic vouchers.

Lampropeltis triangulum andesiana . —The occurrence of Lampropeltis triangulum in Venezuela was reported by Roze (1966) and subsequent workers ( Lancini 1979; Kornacker 1999) as the subspecies L. t. micropholis. Barrio and Navarrete (1999) clarified the subspecies identity of the Venezuelan specimens and considered them to belong to L. t. andesiana.

Leptophis ahaetulla ssp. —Three subspecies of L. ahaetulla have been recorded from Venezuela: L. a. ahaetulla , L. a. coeruleodorsus, and L. a. occidentalis ( Lancini 1979; Albuquerque 2009). Leptophis a. coeruleodorsus has a wide distribution throughout the lowlands of Venezuela, east of the Andes , north and south of the Orinoco River and is also present on Isla Margarita. Leptophis a. ahaetulla in Venezuela was formerly considered to represent a distinct subspecies, L. a. copei . Both taxa were shown to be identical by Albuquerque (2009), and the name ahaetulla has precedence over copei . In Venezuela Leptophis a. ahaetulla is known from few specimens collected in the southern part of the Amazonas State. Leptophis a. occidentalis is present in north and northwestern Venezuela west of the Andes . This last form is the most morphologically differentiated among the subspecies found in Venezuela. The taxonomic status of this subspecies is still poorly understood and some populations have enough unique features to deserve specific status. N. R. de Albuquerque (pers. comm.) is currently working on a systematic revision of Leptophis .

Liophis miliaris . —A specimen of L. miliaris from the Amazonas State ( Fuentes 2003) is herein re-identified as Liophis cf. zweifeli (see below) because, in addition to several colour pattern characteristics, it possesses a single apical pit in some dorsal scales and presents a low number of ventral scales (134). None of these characters has ever been found in any L. miliaris population (see account for L. zweifeli ). Liophis miliaris is known in Venezuela from Bolívar State ( Kornacker et al. 2002; Myers & Donnelly 2008).

Liophis zweifeli . —We consider this species distinct from the L. reginae complex. Liophis zweifeli differs from L. reginae (meristic and colour characters taken from Dixon 1983a) in having a salt-and-pepper dorsal pattern (vs dorsum with dense pale and dark paravertebral flecking in L. reginae ) and in having a higher number of subcaudal scales (69−88 vs 55−78 in L. r. reginae ). Both species are forest snakes. Liophis zweifeli is found mainly throughout the Cordillera de Mérida and the Coastal Ranges in Venezuela, although some specimens of L. cf. zweifeli are known from localities in southern Venezuela. That species has been previously confused with individuals of L. miliaris from Bolívar, south of the Orinoco River ( Kornacker et al. 2002). Dixon (1983a) recorded a specimen of L. zweifeli (as L. reginae zweifeli ) from Cerro Yapacana in central Amazonas, Venezuela. We examined an additional specimen from the same locality (EBRG 1156, cited in Fuentes 2003), which fits generally well the description of L. zweifeli , except that it has fewer subcaudal scales, although the very tip of the tail is missing (47 vs 78–86 fide Roze [1966] and 69–88 fide Dixon [1983a]). A re-evaluation of this taxon and a revision of the L. reginae species complex are needed to properly determine the status of this population in the state of Amazonas.

Oxyrhopus formosus / occipitalis . —We follow MacCulloch et al. (2009) and consider the specimens reported as O. formosus from Venezuela to be O. occipitalis .

Oxyrhopus melanogenys ( Fig. 10). — Lynch (2009) described Oxyrhopus vanidicus from Colombia. According to the original description, the new species only slightly differs in colour pattern from the very similar O. melanogenys . Lynch (2009) stated that some populations formerly assigned to O. melanogenys and O. aff. melanogenys from northern South America are Oxyrhopus vanidicus . We examined four specimens (EBRG 1536, 2069, MHNLS 11967, 11998) from Amazonas and Bolívar States in southern Venezuela, which fall within the variation reported for O. vanidicus . However, we could not find consistent differences between O. vanidicus and O. melanogenys . Pending additional evidence we consider all the specimens previously referred to O. aff. melanogenys and O. trigeminus in Venezuela as O. melanogenys (see also comments under Oxyrhopus trigeminus , and MacCulloch et al. 2009).

Proctoporus cephalolineatus . — García-Pérez and Yustiz (1995) described Proctoporus cephalolineatus , providing a very brief and incomplete description, and suggested that it is closely related to P. achlyens and P. shrevei ( P. luctuosus group). Doan and Schargel (2003) doubted that this species belongs to the P. luctuosus group and even rejected the fact that it belongs to Proctoporus because of the presence of prefrontal scales, but that it rather could belong to either Euspondylus or Pholidobolus . Including this species in the probably paraphyletic genus Euspondylus would bring more confusion. Pholidobolus is distributed far from the Venezuelan territory and is still poorly understood ( Hillis 1985). The single specimen of P. cephalolineatus could be an aberrant specimen of the geographically close P. inanis , a species from the Andes of Mérida recently described and now in the genus Riama ( Doan & Schargel 2003; Doan & Castoe 2005), in which case the name P. cephalolineatus would have priority.

Rena dimidiata . —This species very likely occurs in Venezuela, but is not included in our list because no voucher specimen exists. The type locality of R. dimidiata has been restricted to an area within Roraima State, Brazil (close to the border with Venezuela) by Peters & Orejas-Miranda (1970). The first mention of the species in Venezuela was apparently by Lancini (1979), who based the record on Orejas-Miranda (1966, 1969) and cited the species from the country with no voucher mentioned. Based on this report the species was later reported from Venezuela by several authors (e.g. Lancini & Kornacker 1989; La Marca 1997; Péfaur & Rivero 2000). Orejas-Miranda (1966, 1969), however, did not mention R. dimidiata from Venezuela, but in 1970, Orejas-Miranda in Peters and Orejas- Miranda (1970) provided the following distribution for the species: “Guianas, northern Brazil and southeastern Venezuela ”. Hoogmoed (1977) stated: “According to Orejas-Miranda (1967) the species occurs in a restricted zone in the border area of Brazil, Guyana and Venezuela, coinciding with the Rupununi savanna and its extensions into Brazil and Venezuela. In 1970 the area is given as being slightly more extensive by the same author [=Orejas- Miranda]". Hoogmoed did not explicitly report the species from Venezuela but tried to describe the general area south of Roraima given by Orejas-Miranda and also reported Orejas-Miranda´s 1970 extension of the distribution of R. dimidiata into Venezuela. Interestingly, there is no mention of R. dimidiata from Venezuela in that paper either. A revision of the family Leptotyphlopidae in Venezuela is needed, particularly to elucidate the taxonomic status of the Epictia goudotti - albifrons complex, which includes the population from Isla Margarita, currently placed in the synonymy of E. goudotti .

Tantilla melanocephala . — The species is reported from the Pantepui region based on a specimen identified by GR (MHNLS 12339) collected in Cerro Jaua, between 1600-1800 m elevation.

Thamnodynastes species. —Ten species of this genus are currently recognized in Venezuela. Although the taxonomy of this group has been largely confusing, resulting in occasional confrontations (e.g. Gorzula & Ayarzagüena 1996 vs Myers & Donnelly 1996) some recent contributions provide a better understanding of that difficult genus (e.g. Bailey & Thomas 2006).

Tretioscincus bifasciatus . —Formerly divided into two subspecies, T. b. bifasciatus from northern Caribbean Colombia and T. b. kugleri from Venezuela and Aruba. The main difference between the subspecies appears to be the colour pattern. However, the species seems to show considerable individual variation in the intensity of dorsolateral stripes and ventral spotting. Marcuzzi (1950a) examined specimens from Bobare, Lara, and El Pilar, Sucre and questioned the status of the subspecies T. b. kugleri , arguing that colour variation in a species, which appears to be highly variable in this case, is not recommended in the recognition of subspecies. We agree with and follow Marcuzzi in not recognizing T. bifasciatus kugleri as a valid taxon.