Development and evaluation of habitat models for herpetofauna and small mammals

Abstract

We evaluated the ability of discriminant analysis (DA), logistic regression (LR), and multiple regression (MR) to describe habitat use by amphibians, reptiles, and small mammals found in California oak woodlands. We also compared models derived from pitfall and live trapping data for several species. Habitat relations modeled by DA and LR produced similar results, averaging about 70% classification success of the trapping stations to the correct group (capture or noncapture habitat). Although more variables were included in DA (4-5) than in LR (2-3), those included in LR were typically a subset of those in DA. On average, MR habitat models accounted for 56% of the variation in the index of relative species abundance. The variables included in the MR models were seldom the same as those for DA and LR. Most differences between MR and the other two methods were related to differences in spatial scale: MR modeled habitat among grids, whereas DA and LR modeled habitat within grids. Habitat models for the same species differed between the trapping methods used. Live traps are most useful for describing general habitat relations of some small mammal species across large geographic areas, whereas pitfall traps are useful for intensive sampling of a larger portion of the vertebrate community within smaller geographic areas. Thus, the choice of trapping methods must be based on the study goals, biology of the species, and the spatial scale of study.