作者: Grey W. Pendleton , Kimberly Titus , Eugene DeGayner , Craig J. Flatten , Richard E. Lowell
DOI: 10.2307/1400583
关键词: Geography 、 Statistics 、 Measure (data warehouse) 、 Analysis method 、 Ecology 、 Habitat 、 Selection (genetic algorithm) 、 Statistical hypothesis testing
摘要: We review compositional analysis for testing whether animals use habitats in proportion to their availability (i.e., random use). illustrate aspects of this method with goshawk habitat data collected southeast Alaska. Compositional addresses several inadequacies other selection approaches. As the appropriate summary measure analyses, we propose geometric mean ratio. However, caution must be used interpreting results based on analysis. It is not clear what definition method) best all cases, and hypothesis tests allow only narrow conclusions drawn without additional or assumptions.