Selection rankings of woody species for white-tailed deer vary with browse intensity and landscape context within the Central Hardwood Forest Region

Abstract

White-tailed deer (Odocoileus virginianus) selection of woody species can be influenced by deer densities, food availability, and local and regional landscape characteristics. Determining selection rankings across varying regional landscapes is important to the management of both deer and forests. However, these regional-scale rankings are currently lacking. Here, we develop selection rankings for 63 species within woodlots across the southern, central, and northern regions of Indiana by counting the number of available and browsed twigs by species along transects. We then classified species into five selection classes: highly selected, slightly selected, neutral, slightly avoided, and strongly avoided. Some species displayed consistent classification across regions, including selection for greenbrier (Smilax spp.), hackberry (Celtis occidentalis), white ash (Fraxinus americana), and white oak (Quercus alba) and avoidance of spicebush (Lindera benzoin) and pawpaw (Asimina triloba). For the 16 species that exhibited significant regional variation in selection, we modeled the probability of a twig being browsed using explanatory variables measured at the woodlot scale, and within 200, 500, 1000, and 1500 m buffers around woodlots. Browsing intensity within a focal woodlot was more often associated with increased selection of an individual species than other explanatory variables. Nonetheless, both woodlot and landscape-scale variables influenced selection of woody browse by deer. In general, factors that increased browsing opportunities (e.g., increased density of non-avoided twigs, increased forest edge density, and increased woodlot edge) increased the selection of an individual species, while those that decreased browsing opportunities (e.g., increased non-native stem density and increased avoided twig density) decreased selection. Our selection classifications for common species in the Central Hardwood Forest Region highlight species that may be at risk of being negatively affected by severe browsing, while simultaneously identifying species that decrease foraging habitat quality. Our results also provide evidence that local and regional conditions drive differences in selection; and thus, require evaluation prior to management. More broadly, we showcase that multiple variables and spatial scales can affect selection of individual browse species by deer and merit consideration by researchers when studying the effects of browsing on forest ecosystems.