Multi‐scale assessment of greater sage‐grouse fence collision as a function of site and broad scale factors

Abstract

AbstractPrevious research in Europe and North America suggested grouse are susceptible to collision with infrastructure, and anecdotal observation suggested greater sage‐grouse (Centrocercus urophasianus) fence collision in breeding habitats may be prevalent. However, no previous research systematically studied greater sage‐grouse fence collision in any portion of their range. We used data from probability‐based sampling of fences in greater sage‐grouse breeding habitats of southern Idaho, USA, to model factors associated with collision at microsite and broad spatial scales. Site‐scale modeling suggested collision may be influenced by technical attributes of fences, with collisions common at fence segments absent wooden fence posts and with segment widths >4 m. Broad‐scale modeling suggested relative probability of collision was influenced by region, a terrain ruggedness index (TRI), and fence density per square km. Conditional on those factors, collision counts were also influenced by distance to nearest active sage‐grouse lek. Our models provide a conceptual framework for prioritizing sage‐grouse breeding habitats for collision mitigation such as fence marking or moving, and suggest mitigation in breeding habitats should start in areas with moderate‐high fence densities (>1 km/km2) within 2 km of active leks. However, TRI attenuated other covariate effects, and mean TRI/km2 >10 m nearly eliminated sage‐grouse collision. Thus, our data suggested mitigation should focus on sites with flat to gently rolling terrain. Moreover, site‐scale modeling suggested constructing fences with larger and more conspicuous wooden fence posts and segment widths <4 m may reduce collision. © 2012 The Wildlife Society.