ANALYSIS OF HABITAT SELECTION STUDIES WITH MULTIPLE PATCHES WITHIN COVER TYPES

Current statistical methods are inadequate for evaluation of the relat ion between spatial pattern of the landscape and observed patterns of habitat use by individuals or populations. For example, traditional ha bitat selection analysis methods do not use information about the size and distribution of the several patches of each cover type that may e xist within the study area. Statistical tests are presented for hypoth eses about disproportional use of cover types and patches within cover types, These tests require that use of individual patches is recorded , as well as the size of individual patches. Different designs are con sidered in which there are (1) single or multiple samples of use, and (2) equal or unequal habitat availability. Formulas for calculating Ty pe II statistical errors of the tests are presented and Monte Carlo si mulation is used to assess the accuracy of the formulas and to check t he Type I error rates of the proposed test statistics. With adequate s ample sizes, Type II error formulas can be a useful tool for planning of habitat selection studies. An example analysis is presented of a hy pothetical study. of habitat selection by ring-necked pheasants (Phasi anus colchicus) in a Midwestern landscape. The proposed tests also rep resent a contribution toward bringing together concepts of landscape e cology and wildlife habitat selection.