Remember the Third Dimension: Terrain Modeling Improves Estimates of Snake Home Range Size
Studies of snake movement frequently report home range sizes, and these data are key to our understanding of snake ecology. However, despite the fact that many snakes occupy landscapes with topographic relief, snake home range sizes are consistently reported as planimetric (two-dimensional) estimates. We investigate the capacity for planimetric area measurements to underestimate snake home range sizes, and we explore how this may confound understanding of snake ecology. We use radiotracking data to estimate home ranges for Crotalus mitchellii and C. ruber in a topographically-variable landscape, model surface terrain in the estimated home ranges using digital elevation data and widely-available GIS software, and compare planimetric and topographic (three-dimensional) area measurements for these home ranges. The topographic measurements exceeded planimetric ones, on average, by 14% for C. mitchellii (8% for females, 19% for males) and 9% for C. ruber (10% for females, 8% for males); this suggests that terrain modeling can provide considerably truer estimates of snake home range sizes. We discuss how such increased precision might benefit snake ecology research, and we discuss recent papers that might have concluded differently had topographic home range measurements been used. Terrain modeling has become relatively simple with the advent of modern GIS software, and snake ecologists could begin using it routinely.Abstract
Construction of a TIN surface model for a snake home range. (A) Minimum convex polygon home range overlaid atop digital elevation contours. (B) Delaunay triangulation for spot elevations extracted from contours. (C) Three-dimensional rendering of triangulation, based on known X, Y, and interpolated Z at triangle vertices; each triangle has constant slope angle and aspect across its face.
Home ranges (minimum convex polygons) of radiotracked Crotalus mitchellii. Each appears atop a hillshaded rendering of a localized TIN surface model, with the point locations from which it was derived. Common scale (see upper left panel) and orientation (up is north) are used throughout. MCP areas are reported as “[planimetric area] vs. [topographic area]”, with percent increases computed as (t-p)/p * 100, where t = topographic area and p = planimetric area.
Home ranges (minimum convex polygons) of radiotracked Crotalus ruber. Each appears atop a hillshaded rendering of a localized TIN surface model, with the point locations from which it was derived. Common scale (see upper left panel) and orientation (up is north) are used throughout. MCP areas are reported as “[planimetric area] vs. [topographic area]”, with percent increases computed as (t-p)/p * 100, where t = topographic area and p = planimetric area.
Contributor Notes
Associate Editor: E. Schultz.