Behavior‐specific habitat models as a tool to inform ungulate restoration

Abstract

AbstractAcross North America, many ungulate species that experienced historic population declines and range contractions are now broadly distributed across their native ranges after the implementation of successful restoration programs. The use of translocation continues to serve as an important restoration tool and is often informed through habitat models used to identify potential translocation areas based on biotic and abiotic landscape characteristics. Within the context of partially migratory wildlife populations, resident and migrant population segments can select for varying habitat characteristics, yet these population segments are often pooled when building habitat models. We used a large spatial dataset collected from eight bighorn sheep (Ovis canadensis) populations to build separate winter and summer resource selection models for migrant and resident animals with the purposes of (1) characterizing differences in seasonal selection patterns between resident and migrant population segments and (2) generating broad spatial predictions of bighorn sheep habitat to inform future translocations across western Montana, USA. Lastly, we used the model to inform two potential management scenarios, first to establish a new population in an unoccupied area and second to expand the distribution of existing populations through intra‐mountain translocations. Selection patterns were generally similar among migrants and residents, especially in winter. Summer selection patterns varied between the two behaviors, with migrants selecting for higher elevations and residents selecting for lower elevations. Throughout the western Montana prediction area, bighorn sheep habitat was centered around mountain areas in all seasons. In the first management scenario, our model predicted that adequate resident and migratory bighorn sheep habitat existed in the restoration area, thus justifying the use of either resident or migrant source populations in translocation. In the second management scenario, our model predicted that there was broad potential for translocations into unoccupied areas adjacent to the current bighorn sheep distribution in western Montana. The behavior‐specific approach to predicting bighorn sheep seasonal habitat captures the specific habitat characteristics of multiple migratory behaviors and may help to inform targeted and effective translocation programs.