96 J. Alan Holman and Jerry N. McDonald 



both S. dekayi and S. occipitomaculata occur in the Saltville area 

 today. 



Thamnophis sp., Gartersnake or Ribbonsnake 



Material. — Trunk vertebra: USNM 404745, from Unit W3, SV-1. 



Remarks. — Brattstrom (1967) showed that the vertebrae of Tham- 

 nophis are more elongate than those of the related genus Nerodia. It is 

 almost impossible to separate isolated vertebrae of the two species of 

 Thamnophis (T. sauritus and T sir talis) that occur in the vicinity of 

 Saltville today. 



DISCUSSION 



The known herptile fauna from Saltville has been divided into three 

 faunules on the basis of the depositional units from which the remains 

 were recovered (Table 1). The taxonomic composition and chronology 

 of these faunules can provide information about the duration of resi- 

 dency of the taxa, the depositional environment in which each was best 

 sampled, and the microhabitat of the respective taxa. 



Unit W3, the sorted stream channel bed load deposit found at SV-1 

 and SV-2, contained seven taxa including all identified specimens of 

 Notophthalmus cf. viridescens, Rana pipiens group, Nerodia sipedon, 

 Thamnophis sp., Elaphe cf. E. obsoleta, and Storeria sp. Only Bufo 

 woodhousei fowleri is found in W3 and other depositional units. The 

 stratigraphic nature of Unit W3 — silts, sands, and fine gravels, ranging 

 from well sorted and laminated deposits to "unsorted" masses (perhaps 

 mixed biogenically, as by trampling by large mammals) — indicates that 

 the member deposits were laid down by moderately to slowly moving 

 water, perhaps through several cycles of rise and fall. Fluctuations in 

 stream stage would have permitted periodic integration of the remains 

 of terrestrial vertebrates into the stream bed load, especially those taxa 

 that inhabited or periodically used the riparian zone. This might explain 

 the presence of terrestrial taxa, including most of the snakes, in the 

 fluvial deposits. The large amount of woody plant remains of uniform 

 size (< 50 mm) in Unit W3 at SV-2 strongly suggests fluvial sorting of 

 "sediments" of terrestrial origin. Alternatively, semi-aquatic or avian 

 predators or scavengers could have dropped the remains of terrestrial 

 prey in or near the stream during feeding. The possibility that large 

 mammals might have mixed units W2 and W3 at SV-1 while watering 

 or feeding has been considered. However, in view of the fact that the 

 composition of the herptile samples in Unit W3 at SV-1 and SV-2 is 

 remarkably similar and that the composition of W2 and W3 at SV-1 are 

 generally different, mixing of these two deposits must be considered 

 unsubstantiated at present. The herptiles of Unit W3 may, therefore, be 

 taken to represent a sampling of the Saltville Valley lotic and riparian 

 herpetofauna as of ca. 14,500 to 14,000 B.P. 



