Ophiodes

Abstract: The taxonomic status of the genera included within the anguid lizard subfamily Diploglossinae has been uncertain. A new character, the structure of the osteoderms, has been studied in hope of clarifying this situation. The history of the literature, structural detail, and taxonomic importance of osteoderms are reviewed. On the basis of osteoderm structure, two genera (Celestus, Sauresia) are recognized in addition to Diploglossus, Wetmorena, and Ophiodes. A possible history of the evolutionary trends and movements of the recognized genera is presented, with special reference to the situation in the West Indies. LIZARDS OF THE FAMILY ANGUIDAE are widely distributed in the New World; one subfamily, the Diploglossinae, occurs from southern Mexico into southern South America and the West Indies. Within this subfamily are presendy recognized (Meszoely 1970: 119) the three genera Diploglossus Weigmann, Wetmorena Cochran, and Ophiodes Wagler. The subfamily is especially well represented in the West Indies, and indeed the genus Wetmorena is restricted to the Hispaniolan south island (sensu Williams 1961). Grant (1944) suggested that "scale structure" might be useful in showing relationships between Jamaican species of "Celestus" (for usage of this name, see beyond). Since Jamaica and Hispaniola have the largest number of species of diploglossines of any of the Antillean islands, and since various authors have attempted to arrange these and other species into some sort of phylogenetic relationships (Grant 1940, Schwartz 1970), we felt that perhaps a careful study of the osteoderm structure of the Antillean members of the subfamily might be revealing. Such an assumption was in large part based upon Grant's paper noted above; but we quickly learned that he had confused two quite different elements-the decorations of the epidermal scales and the structure of the underlying osteoderms. Thus our original premise (namely, that osteodermal sculpturing and decorations might reveal relationships) had to be discarded. Nevertheless, osteoderms do reveal strong evolutionary trends and a possible history of the subfamily. A satellite problem has been that of the genera Celestus and Diploglossus (no one has seriously questioned the validity of Ophiodes, nor do we). The usage of one or both of these names has been excellently summarized by Myers (1973: 15-16). Briefly two genera were separated: Celestus Gray (type-species striatus of uncertain provenance) was distinguished from Diploglossus (type-species fasciatus from South America) in that the latter had sheathed (and supposedly retractile) claws whereas the former had the claws exposed. This character was seemingly correlated with the presence of a single prefrontal in contact with the frontal scale anteriorly. The picture was further complicated (or clarified) in that Underwood (1959) reviewed both Antillean and many continental species and felt that, since darlingtoni (hereinafter we deliberately refrain from using binomials) from Hispaniola occasionally has 3 scales in anterior contact with the frontal, and millepunctatus from Isla Malpelo off the Pacific coast of South America and microblepharis from Jamaica have the claws half-sheathed, these characters were not truly distinctive of two biological entities (-genera). In addition, he noted that there was a concordance in the presence of the nasal-rostral contact in the mainland species in general, and the absence of this contact in many Antillean species. Brongersma (1946) has previously noted that the Cuban delasagra had sheathed claws and that it could thus by definition not be included in Celestus (sensu stricto). Underwood showed that the Puerto Rican pleei also had sheathed claws. In addition, sheathed claws occur in sepsoides from Hispaniola and Wetmorena haetiana from the same island. Everything considered, Underwood felt that there was no constant set of characters which would absolutely separate all Celestus from all Diploglossus and that these two generic entities 58 BIOTROPICA 9(1) 58-72 1977 This content downloaded from 207.46.13.72 on Sat, 06 Aug 2016 05:39:56 UTC All use subject to http://about.jstor.org/terms should be merged (under the prior name Diploglossus). In addition, sepsoides, which had been proposed as a separate genus Sauresia Gray and had been accepted as the sole member of that genus by Cochran (1941), was also merged with Diploglossus by Underwood, although Wetmorena was retained as distinct. Later authors (Schwartz 1964, 1970, 1971; Thomas 1971; Myers 1973) when dealing with either Antillean or continental diploglossines have followed Underwood, although at least Myers showed some discontent with Underwood's arrangements. Meszoely (1970: 93) also followed Underwood while dealing primarily with North American fossil anguids. Inadvertently and erroneously he attributed to Underwood the statement that sepsoides may occasionally have 3 "prefrontals," a condition that we have not observed in the species. Conversely, Peters and Donoso-Barros (1970) used both Celestus and Diploglossus for the continental Neotropical diploglossines and distinguished between them (p. 2) on the basis of claws retractile (Diploglossus) or not (Celestus). Let us unequivocally state at this point that there is absolutely no evidence for retractility of the claws in these lizards. The junior author has handled literally hundreds of diploglossines from the West Indies, including most species, thus both with and without a claw sheath. The claws are in no way retractile, but the claws may be sheathed as a permanent condition, a condition which is species-constant. Since there is suLch confusion as to which species belong to which genus (assuming that Celestus and Diploglossus are differentiable in some way), it seemed a possibility that osteoderm structure (not the decorations or superficial sculpturing which is minimal and virtually non-existent) might be a key to unlock this puzzle. Accordingly, we have examined specimens of all taxa presently referred to Diploglossus (including "Celestus" and "Sauresia"), W'etuorena, and three species of Ophiodes. We are muLch less familiar with the continental diploglossines than with the Antillean members of the stubfamily; fortunately, there are more species in the islands (19) than on the mainland (13, including Isla Malpelo). The Antillean species are the following (arranged by islands): Cuba (delasagra Cocteau); Hispaniola (agasepsoides Thomas, costatus Cope, curtissi Grant, darlingtoni Cochran, viarcanoi Schwartz and Inchavistegui, sepsoides Gray, stenurus Cope, warrent Schwartz, WVetmorena haetiana Cochran); Ptuerto Rico (pleei DuLmeril and Bibron); Jamaica (barbouri Grant, crusculus Garman, duquesneyi Grant, fowt,leri Schwartz, hewardi Gray, microblepharis Underwood, occiduus Shaw); Montserrat (montisserrati Underwood). Of these species, costatus has a subspecies on Navassa Island and crusculus a subspecies on the Cayman Islands (Little Cayman, Cayman Brac) in addition to the primary islands with which they are associated above. The continental species (including the Pacific Isla Malpelo) are: atitlanensis Smith, bilobatus O'Shaughnessy, bivittatus Boulenger, cyanochloris Cope, enneagrammus Cope, fasciatus Gray, lessonae Peracca, millepunctatus O'Shaughnessy, montanus Schmidt, monotropis Kuhl, montisilvestris Myers, rozellae Smith. In addition to these nominal species striatus Gray was described from the "West Indies"; Schwartz (1964: 10) suggested that the holotype and only known specimen (with the frontal contacting three scales anteriorly) is likely not Antillean but rather originated from the continental mainland. However, Boulenger (1885: 289) had suggested that striatus and occiduus were synonymous; we will discuss this question later in the pres-