Species Distribution Modeling: A Biosocial Approach

Abstract

Species distribution models (SDMs) are a method for predicting the spatial distribution of a target species. While previous literature highlights predominantly biophysical variables in developing prediction probabilities, few studies incorporate variables that consider the role of human influence on the target species’ distribution. This study aimed to explore the predictive capabilities of socioenvironmental variables on the development of an SDM for the known West Nile virus bridge vector, the Culex pipiens mosquito. Three maximum entropy models were produced using mosquito sampling data from Forsyth County, North Carolina, during the 2013 breeding season. The first model incorporated eight biophysical variables; the second model incorporated thirteen socioenvironmental variables; and the final model combined both biophysical and socioenvironmental variables to produce a biosocial model. Results show that the predictive capabilities of the social variables outperform both the biophysical variables alone or in combination with the socioenvironmental variables. The results of this research should engender further consideration to the applicability of socioenvironmental variables for SDMs where direct human influence could affect species distribution.