A NOVEL APPLICATION OF MAXENT RESPONSE FUNCTIONS TO IDENTIFY ABIOTIC AND BIOTIC DETERMINANTS OF AMERICAN EEL DISTRIBUTION IN MID-ATLANTIC U.S. RIVERS
In freshwater systems, species’ sorting from a regional species pool to the local community occurs via a series of hierarchical filters. For fishes, three classes of filters thought to be most important are: (1) physical habitat and channel morphology (hereafter “landscape variables”), (2) hydrology, and (3) biotic interactions. Although each class of filter has been shown to be important in some systems, direct comparisons of their relative importance are rare. Using the American eel (Anguilla rostrata) as a model organism, we quantified the influences of landscape, hydrologic, and biotic variables on historical eel distributions within six Mid-Atlantic drainages. We measured the agreement (or overlap) between probability densities for observed eel “presence” sites and “background” habitat. This approach stemmed from the MaxEnt response function methodology where the response curve (predicted presences/background) is highest when minimal overlap between presence and background probability density functions is observed. Similarity of background versus presence sites between filter classes was also assessed with multidimensional scaling. Results suggest that biotic interactions, as indicated by functional traits comparisons among species, may be uniquely important to historical eel distributions.
Taylor Woods (Primary Presenter/Author), Virginia Commonwealth University, firstname.lastname@example.org;
Daniel McGarvey ( Co-Presenter/Co-Author), Virginia Commonwealth University, email@example.com;