Study Designs and Tests for Comparing Resource Use and Availability

Publisher Website
Google Scholar
Abstract
We classified study designs for comparing resource (food, habitat) use and availability into 3 basic types. Design 1 permits investigation of resource selectivity only at the population level because individual animals are not identified. Designs 2 and 3 measure use by individuals and thus all examination of the variations in resource selection strategies. Resource availabilities are measured for each individual in Design 3 but not in Design 2. Graphical plots illustrating individual selection are recommended for data resulting from Designs 2 and 3 to assess variability and possible sex or age differences. We recommended a method for determining the number of random points required to bound the probable error in estimating resource availability proportions simultaneously, rather than individually. Four problem areas in the use of statistical methods for evaluating resource selectivity are identified: dependencies among observations, the misuse of the Chi-square goodness-of-fit test when availabilities are estimated, tests that do not control experimentwise error rates, and the sensitivity of tests to the subjective inclusion or exclusion of resources.