The material used for comparison (temporarily housed in the personal collection of the author) includes the following extant snakes distributed in Argentina [9], [19] and [20]: Boiruna maculata, Clelia rustica, Clelia bicolor, Phimophis vittatus, Helicops leopardinus, Liophis miliaris, Liophis anomalus, Liophis poecilogyrus, Lystrophis dorbignyi, Oxyrhopus rhombifer, Philodryas patagoniensis, Philodryas baroni, Thamnodynastes hypoconia, Bothrops alternatus, Bothrops ammodytoides, Epicrates cenchria, Eunectes notaeus. Anatomical terminology follows Auffenberg [6], Hoffstetter and Gasc [12], and Thireau [17].

Colubridae Oppel, 1811

Xenodontinae Bonaparte, 1845

Genus Clelia Fitzinger, 1826

Clelia sp.

The vertebrae exhibit a marked degree of ossification, suggesting that the specimen corresponds to a mature individual. The morphology of the vertebrae corresponds to the medio-posterior trunk region, since there are no hypapophyses or lymphapophyses, and because the hemal keel is robust (Fig. 2). The vertebra is higher than long (height: 7.2 mm), with the vertebral body short (length: 5.2 mm). It is distinct from Philodryas and Liophis because these two taxa have more elongate vertebral centra. The cotyle and condyle are circular shaped. A conspicuous lateral foramen is present below the neural arch, as usual in snakes. The paracotylar foramen is present and located lateral to the cotyle. The hemal keel is transversely wide, typical of the medio-posterior trunk region, thus defining two subequal concavities together with the lateral margins (margo lateralis). The neural spines in the trunk region of Clelia are high and craniocaudally unexpanded. In contrast to Philodryas [5 (Fig. 3)] and Liophis, the tip of the neural spine forms both a cranial and caudal prominent expansion, which is distinctive of Clelia. The prezygapophyseal processes are robust (‘obtuse shape’ of Auffenberg [6]) and laterally oriented, in contrast to the condition present in various other xenodontines and non-colubrid cenophidian snakes (e.g., Philodryas, Liophis, Bothrops), in which the prezygapophyseal processes are smaller and sharper and laterocaudally oriented (‘acuminate shape’ of Auffenberg [6]). The articular surface of the prezygapophyses is oval-shaped and the zygosphene is crenate-shaped, as in Clelia rustica.

The ribs of the MLP 98-XI-12-1 show morphology typical of colubrids, with a conspicuous tuberculiform process and separate articulate facets owing to a groove in the articulate head.

Because of the lack of a comprehensive survey of vertebral form of South American colubrids, it is a difficult task to assign isolated fossil vertebrae, even at a generic level.

The extant species of colubrid snakes present in the Buenos Aires province show a strong similarity in the trunk region of the vertebral column, which hinders their identification (e.g., most of the Argentine species of Liophis cannot be recognized from their trunk vertebrae). However, some vertebral features exhibit significative differences among South American xenodontines, which allows the taxonomic identification of isolated vertebrae [14]. The craniocaudal development of the neural spine, the short vertebral body, and the robust and laterally oriented prezygapophyseal processes are characters found together in Clelia rustica. However, trunk vertebrae are not a reliable basis that would allow specific allocation; thus MLP 98-XI-12–1 is only identified at genus level.

The level in which MLP 98-XI-12-1 was found is interpreted as formed under humid conditions [7]. The Argentine extant species of Clelia are distributed in humid environments, which is consistent with the paleoenvironmental inference previously made. The distribution of fossil material of colubrid snakes throughout the Cenozoic of Argentina is scarce, showing a significant hiatus from the Pliocene to the Upper Pleistocene [4] and [10].

The origin of the subfamily Xenodontinae is interpreted as having occurred in Asia–North America, arriving to Central and South America by means of dispersal events [8] and [18]. However, the age of these events is ignored, just because of the absence of a well-documented fossil record [11]. Thus, the discovery of the genus Clelia in the Early to Middle Pleistocene of southern South America suggests that the origin of the subfamily Xenodontinae and their later dispersion toward the rest of America should have taken place by or before the Pliocene period. This hypothesis is in agreement with the recent findings of undetermined Colubridae vertebrae from the Miocene of Patagonia [4].

The present record of an undetermined species of the genus Clelia in the Ensenadan age constitutes the oldest xenodontine genus reported up to now in South America. The specimen here described is the first mention of a colubrid snakes for the above mentioned age, representing a significant addition to our knowledge of the Cenozoic fossil snake fauna.