A conceptual framework for selecting and analyzing stressor data to study species richness at large spatial scales

In this paper we develop a conceptual framework for selecting stressor data and analyzing their relationship to geographic patterns of species richness at large spatial scales. Aspects of climate and topography, which are not stressors per se, have been most strongly linked with geographic patterns of species richness at large spatial scales (e.g., continental to global scales). The adverse impact of stressors (e.g., habitat loss, pollution) on species has been demonstrated primarily on much smaller spatial scales. To date, there has been a lack of conceptual development on how to use stressor data to study geographic patterns of species richness at large spatial scales. The framework we developed includes four components: (1) clarification of the terms stress and stressor and categorization of factors affecting species richness into three groups-anthropogenic stressors, natural stressors, and natural co variates; (2) synthesis of the existing hypotheses for explaining geographic patterns of species richness to identify the scales over which stressors and natural covariates influence species richness and to provide supporting evidence for these relationships through review of previous studies; (3) identification of three criteria for selection of stressor and covariate data sets: (a) inclusion of data sets from each of the three categories identified in item 1, (b) inclusion of data sets representing different aspects of each category, and (c) to the extent possible, analysis of data quality; and (4) identification of two approaches for examining scale-dependent relationships among stressors, covariares, and patterns of species richness-scaling-up and regression tree analyses.