Because estimates of small mammal species richness and diversity are strongly influenced by sampling methodology, 2 or more trap types are often used in studies of small mammal communities. However, in most cases, all traps are placed at ground level. In contrast, we used Sherman live traps placed at 1.5 m in trees in addition to Sherman live traps and Mosby box traps placed on the ground to sample small mammals in pine stands in the Upper Coastal Plain of South Carolina. To determine the importance of placing traps in trees, we compared estimates of small mammal (primarily rodent) species richness and diversity based on data from all traps (ground and tree) with estimates based on data from ground traps only. Estimates of species richness based on data from ground traps only did not differ from estimates based on data from all traps. However, 4 other diversity indices (Simpson Index, Shannon-Wiener Index, Shannon Evenness Index, and Brillouin Index) based on data from both tree and ground traps were significantly greater than indices based on data from ground traps only. The increase in the diversity estimates when data from all traps were used was primarily due to the large number of southern flying squirrels (Glaucomys volans) captured in tree traps. When data from ground traps only were considered, the community was highly dominated by cotton mice (Peromyscus gossypinus), but, when data from all traps were considered, cotton mice and southern flying squirrels were co-dominant (567 and 580 individuals, respectively). Our data suggest that studies of forest small mammal communities which do not include tree traps are biased because one of the most common and potentially important species, the southern flying squirrel, is highly underrepresented. We recommend that future studies of forest mammal communities, particularly those designed to test the effects of forest management practices on small mammal communities, include arboreal traps.
