Use of Latitude-Adjusted Elevation in Broad-Scale Species Distribution Models

Abstract

Using readily available spatial data and GIS, we developed a method to adjust elevation by latitude using an estimate of the elevation of alpine treeline across a latitudinal gradient. Latitudeadjusted elevation accounted for the influence of latitude on relationship between elevation and annual monthly maximum temperature, which demonstrates the ability of this method to apply elevation as a model predictor variable across latitudinal ranges. Elevation is particularly useful for predicting coarse-grained distribution patterns of many species because elevation integrates influences of climate, physiognomy, and vegetative cover into a single measurement. However, problems arise with use of elevation as a predictor across wide latitudinal gradients because climate and biotic distributions tend to respond to increases in elevation and latitude similarly. Latitude-adjusted elevation can be used to extrapolate species-distribution models beyond the latitudinal extent of data availability. Using this method we extrapolated a model predicting wolverine (Gulo gulo) habitat across a large region.