Methodological Approach

There are difficulties, often unacknowledged, in seeking fine-scale taxonomic resolution of isolated skeletal elements of amphibians and reptiles. Our identifications are based on the reliable diagnostic features of the fossil material available. Whenever possible we minimized the use of geographic parsimony to refine our identifications, in part because of the antiquity of the material and our uncertainty regarding the distribution (or taxonomic evolution) of amphibian and reptile lineages in the middle Pleistocene. More importantly, the use of geography to refine taxonomic identifications prohibits or limits the use of the identified taxa in any kind of environmental or zoogeographic reconstructions. Failure to recognize this fact can result in circularity in construction of hypotheses of the impacts of climatic or environmental change on faunal dynamics. The only taxonomic assumptions we make that employ geography are those on the continental scale. Where these assumptions are made, they are explicitly stated.

Snakes are represented by a greater number of fossil specimens (117) than any other herpetofaunal group from Porcupine Cave. We recognize three major groups of snakes in the Porcupine Cave collection; all are colubroids, and some specimens cannot be identified beyond Colubroidea. To determine regional position within the vertebral column, we used vertebral characters discussed by LaDuke (1991a). Only precloacal vertebrae were identified beyond Colubroidea. Vertebrae preserving diagnostic features of Natricinae and Viperidae were identified to those taxa. Other non-Natricine colubrids were also recovered from Porcupine Cave, but it is difficult even to identify definitively from which major colu-brid clade these specimens were derived. For example, it can be difficult in some cases to differentiate isolated vertebrae of Xenodontinae from those of some species of Colubrinae (Cadle, 1987; Holman, 2000). These specimens are referred to throughout this chapter as "non-Natricine Colubridae." Specimens are not identified to genus or species. We impose this limitation primarily because we are unable to replicate the work of others who allocated isolated vertebrae to genus or species. As a result, the incomplete nature of many specimens, combined with our methodological approach, potentially leads to an underrepresentation of the snake diversity in Porcupine Cave.

The discovery of discrete vertebral characters to separate major clades of snakes remains a significant challenge for pa-leoherpetologists. A necessary prerequisite to such a discovery is a more thorough understanding of the morphological and serial variation within extant snake lineages. The few completed studies on vertebral variation in snakes (Hoffstetter and Gayrard, 1965; Thireau, 1967; Gasc, 1974; Smith, 1975;

LaDuke, 1991b; Moon, 1999) lay an important foundation for such research, but more work must be done.

0 0

Post a comment