9th International Deer Biology Congress

AUTHORS: Jon D. Heale, Lisa A. Shipley, Daniel H. Thornton – Washington State University, School of the Environment; Paul M. Meyers, U.S. Fish & Wildlife Service

ABSTRACT: Columbian white-tailed deer (Odocoileus virginianus leucurus) were one of the original 78 species to be granted federal protection under the Endangered Species Preservation Act of 1966, pre-dating the Endangered Species Act of 1973. Columbian white-tailed deer originally inhabited river valleys and surrounding foothills dominated by shrubs [1]. By the early 1900s, due to habitat loss and unregulated harvest Columbian white-tailed deer were extirpated throughout most of their historic range, with two remnant distinct population segments (DPS) remaining [2]. These DPSs are geographically isolated, separated by about 320 km of mostly unsuitable habitat (Figure 1). The lower Columbia River population occupies diked bottomlands and islands in the lower Columbia River in southwest Washington and northwest Oregon, and the Douglas County population occupies oak woodlands in the Umpqua River Basin in Douglas County, Oregon [3]. In 2003, after 25 years of active management, the Douglas County DPS was removed from the Endangered Species Act. However, the lower Columbia River population only recently improved from endangered to threatened, a designation achieved in 2016. Improvements in their conservation status were driven by harvest restrictions, acquisition of critical lands and habitat restoration. Yet, because habitat along the Columbia River is fragmented, translocations may be the most effective means of re-establishing Columbian white-tailed deer throughout their historic range.
We examined habitat selection and the effect of current vegetation management (i.e., cattle grazing and mowing) by adult female deer at 3 sites in the lower Columbia River population. From 2013-2018 GPS location data was collected for deer translocated to Ridgefield National Wildlife Refuge (n = 16) and non-translocated resident deer at the Julia Butler Hansen Refuge (n = 14). We fit generalized linear mixed models to estimate habitat characteristics influencing selection. Selection declined as distance to concealment cover increased (P