Meeting Abstract

P2.25  Thursday, Jan. 5  Challenges of modeling the environments of animals: Features of geospatial datasets bias predictions of thermal heterogeneity WANG, Congwen; SEARS, Michael W*; Bryn Mawr College; Bryn Mawr College msears@brynmawr.edu

A pressing challenge for ecologists is to predict the potential ranges of species in light of changing climates. One approach has been to model the niches of organisms using georeferenced climate, topography, and land cover as drivers. A potential problem for this approach is that the spatial resolution of input data is often much more coarse than the spatial resolution of habitats used by organisms. Further, this approach also assumes that all habitats are homogeneous within each spatial unit used as model input. Despite these issues, such models are becoming common tools for ecologists. We have begun to construct more realistic models that explicitly incorporate landscape features into classical biophysical models of organisms. Using simulated landscapes, we examined the effects of the spatial resolution of input data, fractal dimension of elevational features, and topographic relief on the thermal heterogeneity of surface temperatures of organisms. We found that thermal heterogeneity decreased as spatial data become more coarse, increased with the fractal dimension of the landscape, and decreased with reduced topographic relief. The implication of these simulations is that estimates of the thermal heterogeneity of habitats used to predict the ranges of species are often underestimated, likely producing underestimates of potential ranges.