Fox snake

Abstract: Successful species conservation is dependent upon understanding that the habitat preferences of species occur on multiple spatial scales; however, effective synthesis of findings between studies can be hindered by the use of different methods. We applied two habitat selection analyses to two sets of snake radiotelemetry data, which led to differences in results. We discuss perceived strengths and weaknesses of each approach and how they can be used in future studies of snakes. Compositional Analysis (CA) is a classification-based approach, while Euclidean Distance Analysis (EDA) is a distance-based approach. Each method detected non-random habitat selection on both the landscape and home range level, indicating that both Eastern Massasaugas (Sistrurus catenatus) and Eastern Fox Snakes (Pantherophis gloydi) made habitat selection choices at multiple spatial scales. Eastern Massasaugas used a mosaic of habitat types but were consistently associated with forest edge and scrub-shrub wetland, while Eastern Fox Snakes were most often associated with upland edge and old field habitat types. The two methods drew different conclusions, however, and complete understanding of habitat associations may require the use of both methods, taking into account the limitations of each. Compositional Analysis may be more applicable at the landscape level than EDA because of reliance of EDA upon random points, which leads to the identification of unused habitat types (e.g., open water) as preferred by Eastern Fox Snakes. However, at the home range level, EDA was better able to incorporate edge habitats into analyses, a habitat feature likely important for many snake species. We recommend that investigators adopt a combination of these methods, employing CA at the landscape level and EDA on finer scales.