Thornton, Victor. MS Thesis
Species’ distributions are influenced by abiotic and biotic factors but direct comparison of their relative importance is difficult, particularly when working with complex, multi-species datasets. Here, we present a flexible method to compare abiotic and biotic influences at common scales. First, data representing abiotic and biotic factors are collected using a combination of geographic information system, remotely sensed, and species’ functional trait data. Next, the relative influences of each predictor variable on the occurrence of a focal species are compared. Specifically, ‘sample’ data from sites of known occurrence are compared with ‘background’ data (i.e. pseudo-absence data collected at sites where occurrence is unknown, combined with sample data). Predictor variables that may have the strongest influence on the focal species are identified as those where sample data are clearly distinct from the corresponding background distribution. To demonstrate the method, effects of hydrology, physical habitat, and co-occurring fish functional traits are assessed relative to the contemporary (1950 – 1990) distribution of the American Eel (Anguilla rostrata) in six Mid-Atlantic (USA) rivers. We find that Eel distribution has likely been influenced by the functional characteristics of co-occurring fishes and by local dam density, but not by other physical habitat or hydrologic factors.