Terrestrial Drift Fences with Pitfall Traps 11 



around and return to their original retreat, and subsequent forays might 

 be in other directions (C. R. Shoop, pers. comm.). Such behavior could 

 bias interpretations of direction preferences. 



Ecology of a species is probably the most important factor influenc- 

 ing the rates and patterns of capture. Home range size and migratory 

 movements are critical in certain species. For example, the mole sala- 

 mander, Amby stoma talpoidewn, has a life cycle in which adults, under 

 most conditions, characteristically migrate to a breeding pond during 

 winter and return to land in early spring. Juveniles exit the pond a few 

 months later (Patterson 1978; Semlitsch 1981). Therefore, any fence 

 placed parallel to the edge of the breeding pond will capture most, if not 

 all, salamanders moving through the area sampled. 



Southeastern crowned snakes, Tantilla coronata, or scarlet snakes, 

 Cemophora coccinea, may occur in the same habitat as A. talpoidewn. 

 However, individual home ranges in these two species of snakes are 

 independent of the orientation and distance to water, so that movement 

 primarily represents daily activity patterns. Although either species of 

 snake may be captured in relatively large numbers in pitfall traps 

 (Nelson and Gibbons 1972; Semlitsch et al. 1981), the drift fence will 

 only reveal that part of a population whose home ranges overlap the 

 fence system. Thus, whether or not a study species has a congregating 

 focal point as part of its life cycle will influence the effectiveness of the 

 technique in assessing population size. 



Similarly, use of drift fences around an aquatic habitat to monitor 

 terrestrial movement of turtles will result in capture of a higher propor- 

 tion of some species than others. The propensity of eastern mud turtles, 

 Kinosternon subrubrwn, to overwinter on land (Bennett et al. 1970) 

 means that they are more likely to migrate through the land-water inter- 

 face than is a more aquatic species such as the stinkpot, Sternotherus 

 odoratus. Undoubtedly, there are other subtle, important ecological dif- 

 ferences among species that affect their respective rates of capture in 

 similar manners. 



Certain false impressions about abundance, diversity or behavior of 

 animals in an area can be given by factors related to design of the tap- 

 ping system, in combination with the ecology of the species involved. 

 One of the most important design factors may be the spatial arrange- 

 ment of the drift fences (Fig. 1B,C,D). Distance of a fence from a critical 

 habitat, such as an aquatic breeding site or a terrestial denning or nest- 

 ing area, can greatly influence the number of captures of certain species. 

 The key factor is whether the fence intercepts the path of migration, or 

 other movement, of animals from one site to another. A potential 

 impact of fence placement can be readily seen in the disparity in 

 numbers of certain species that leave or enter from particular directions 



