On Measuring Microhabitat Affinities with Special Reference to Small Mammals

Abstract

Two different sampling regimes have been used to measure microhabitat utilization patterns in animal communities. The more common one employs a completely random arrangement of sampling stations with respect to habitat structure, and microhabitat affinities are inferred from correlations between animal density at a station and habitat features present nearby. An alternative method employs a stratified random arrangement of sampling stations. Microhabitat categories are defined a priori, predetermined sampling effort is expended in each microhabitat, and affinities are inferred from relative frequency of occurrence in each microhabitat. We discuss assumptions implicit in each method and compare microhabitat affinities of five California coastal sage scrub rodent species obtained by the two methods. Although there is some qualitative agreement between methods, we conclude that the stratified random regime has several advantages over the other for studies of microhabitat affinity. For other sorts of study, however, a completely random design may be superior.