UROPELTIDAE Müller, 1832

Pyron, Robert Alexander, Ganesh, Sumaithangi Rajagopalan, Sayyed, Amit, Sharma, Vivek, Wallach, Van & Somaweera, Ruchira, 2016, A catalogue and systematic overview of the shield-tailed snakes (Serpentes: Uropeltidae), Zoosystema 38 (4), pp. 453-506 : 471-472

UROPELTIDAE Müller, 1832

TYPE GENUS. — Uropeltis Cuvier, 1829 by subsequent designation of Fitzinger (1843).

INCLUDED GENERA. — Brachyophidium , Melanophidium , Platyplectrurus , Plectrurus , Pseudoplectrurus , Teretrurus , Rhinophis , and Uropeltis .

DIAGNOSIS. — Uropeltids are distinguished from all other amniotes by a unique condition of the occipito-vertebral joint ( Baumeister 1908; Hoffstetter 1939; Williams 1959). In all species examined thus far (including species from Melanophidium , Platyplectrurus , Plectrurus , Rhinophis , Teretrurus , and Uropeltis ), the second vertebra articulates directly with the occipital condyle. These elements exhibit simple matching convexity and concavity, with no notching or indentation, and there is no intervening odontoid process. The neural arch of the atlas is always present and articulates directly with the neural arch of the second vertebra and the convexity of the occipital condyle. Furthermore, Uropeltidae differs significantly from its sister lineage, Cylindrophiidae + Anomochilus , in the following characteristics of visceral anatomy: longer trachea (mean 31% of SVL vs 27%), shorter right lung (23% vs 29%), more posterior orifice of the left lung (32% vs 26%), larger snout-heart interval indicating a more posterior heart (32% vs 27%), larger kidney-vent interval indicating more anterior kidneys (20% vs 13%), larger heart (4.3% vs 3.4%), smaller liver (23% vs 29%), larger right kidney (7.8% vs 5.1%), more anterior right kidney (84% vs 90%), larger left kidney (7.5% vs 4.9%), more anterior left kidney (88% vs 92%), more anterior junction of the systemic arches (–2% vs 0.4%), and more anterior gallbladder (68% vs 74%).

PHYLOGENETIC DEFINITION. — Includes the Most Recent Common Ancestor (MRCA) of Melanophidium wynaudense and Uropeltis ceylanica , and all descendants thereof, and all species more closely related to U. ceylanica than to Cylindrophis ruffus ( Fig. 1 ).

DISTRIBUTION. — Peninsular India and Sri Lanka, primarily in the southern Western Ghats of India and southwestern and central Sri Lanka, but with a few species in the Eastern Ghats and northern Western Ghats of India, and northern Sri Lanka ( Smith 1943; Rajendran 1985). Many species are highly restricted geographically, and known only from a few specimens and localities.

DESCRIPTION

Uropeltids are small snakes (generally <500 mm total length), usually fossorial though occasionally surface active, that feed primarily on earthworms and other invertebrates ( Smith 1943; Rajendran 1985). Many species of Rhinophis and Uropeltis have hypertrophied anterior-trunk musculature to facilitate burrowing (occasionally over a meter underground), with associated fusion of the skull elements (see Gans et al. 1978; Rieppel & Zaher 2002). Many species (especially of Rhinophis and Uropeltis ) have a highly specialized tail presenting a keratinized disk or plate with keels or projections, generally considered to be of unknown function ( Smith 1943; Rajendran 1985). In most species, a polygonal ocular shield covers the eye. In some taxa ( Platyplectrurus , Plectrurus , and Teretrurus ), the supraocular and postocular scales are separated, and the eye is distinct. As fas as is known, all uropeltids are viviparous ( Smith 1943; Rajendran 1985). All species have four supralabials, with the second and third in contact with the nasals, no internasals, and no loreal (with rare individual variation; see Constable 1949). A temporal is present in some taxa ( Brachyophidium , Platyplectrurus , and Teretrurus ), separating the parietal from the fourth supralabial, the latter two of which are in contact in the remaining taxa. The anal and subcaudals are divided, and the tail is short (c. 2-8% SVL; see data in Rajendran 1985; Table 2). Little is known about hemipenial morphology ( Smith 1943; Constable 1949). The organ in Melanophidium is short and thick, lacking spines, with the sulcus spermaticus winding through a series of long, convoluted folds ( Smith 1943). In Uropeltis grandis , it is longer and more slender, and covered with fine spines ( Smith 1943). In Rhinophis lineatus Gower & Maduwage, 2011 , it is moderately long (c. 4 mm), slender, and subcylindrical, with curved spines covering the distal third, with the first two-thirds smooth, and a shallow, smooth sulcus spermaticus ( Gower & Maduwage 2011). In R. dorsimaculatus Deraniyagala, 1941 , it is similarly long, slender, and subcylindrical, with fine spines covering most of the asulcate surface, and a smooth sulcate surface with shallow, smooth sulcus spermaticus ( Gower & Wickramasinghe 2016).

REMARKS

Some previous authors considered Uropeltidae to form a clade with Aniliidae , Cylindrophiidae , and Anomochilidae Cundall, Wallach & Rossman, 1993 , called Anilioidea (see Lee & Scanlon 2002; Conrad 2008; Gauthier et al. 2012), but this is rejected by most large-scale molecular analyses (see Wilcox et al. 2002; Gower et al. 2005; Wiens et al. 2012; Pyron et al. 2013a). From a biogeographic perspective, monophyly of these four families also seems extremely unlikely. The clade would have to be>150Ma to explain their distribution via Gondwanan vicariance. However, Alethinophidia has been dated to c. 100Ma ( Pyron & Burbrink 2012). Given the relatively young age of the four families, a complex route of dispersal from South America to India during the Cenozoic, with no relicts in intervening areas would be needed to account for their biogeographic distribution if they formed a single clade. Rather, large-scale convergence in aspects of cranial and verte- bral morphology related to burrowing (see Wiens et al. 2010) likely explains the morphological similarities of Aniliidae to Cylindrophiidae , Anomochilidae and Uropeltidae .

Figures

FIG. 1. — Molecular phylogeny of Uropeltidae Müller, 1832 based on Bayesian inference of 5.248bp of DNA-sequence data from 6 genes. Numbers at nodes represent posterior probabilities.