A common goal in land-management planning is to describe the relationship between management actions, vegetation and wildlife habitat conditions for large landscapes. Achieving this goal can be challenging because ecological processes of disturbance (natural and anthropogenic) and succession affect vegetation composition and structure, which subsequently affects current and future habitat conditions for wildlife. Habitat suitability is often used as a surrogate for demographic response by wildlife to vegetation change, yet this assumed relationship is rarely evaluated. I developed habitat suitability models for 10 wildlife species: American woodcock (Scolopax minor ), cerulean warbler (Dendroica cerulea), Henslow’s sparrow (Ammodramus henslowii), Indiana bat (Myotis sodalis), northern bobwhite (Colinus virginianus), ruffed grouse (Bonasa umbellus), timber rattlesnake ( Crotalus horridus), wood thrush (Hylocichla mustelina), worm-eating warbler (Helmitheros vermivorus), and yellow-breasted chat (Icteria virens). I described habitat suitability by an empirical or assumed relationship between habitat quality and resource attributes on a relative scale that ranges from 0 (not suitable habitat) to 1 (highly suitable habitat). The models were spatially explicit and included habitat and landscape relationships based on the best available empirical data and expert opinion.
I validated HSI models of breeding habitat suitability for wood thrush and yellow-breasted chats, two species on opposite ends of a forest succession gradient, by evaluating the association between wildlife demographic response and HSI values. First, I evaluated the statistical significance of HSI models as a predictor of three different demographic responses, within-site territory density, site-level territory density and nest success. I demonstrated a statistically significant link between HSI values and all three types of demographic responses for the yellow-breasted chat and site-level territory density for the wood thrush. Second, I evaluated support for models based on HSI values, individual suitability values, measured habitat or landscape attributes (e.g., tree age, tree species, ecological land type), and management treatments (e.g., even-aged and uneven-aged forest regeneration treatments) for each demographic response using model selection. Models containing HSI values were, in general, more supported than models containing only landscape attributes or management treatments for all three types of wildlife response. The assumption that changes in habitat suitability represent wildlife demographic response to vegetation change is supported for these two models; however, differences in species ecology may contribute to model significance and model selection uncertainty.
I then simulated future forest conditions and evaluated habitat suitability 10, 50, and 150 years from present as part of the Hoosier National Forest (HNF) plan revision. Three important patterns emerged over time with respect to forest composition and structure: white oak increased in area of dominance under all alternatives, the proportions of red oak species and maples were affected by the intensity of forest disturbance via harvest and fire, and in the absence of anthropogenic disturbance or wildfire, the HNF was dominated by late-successional vegetation conditions. Without tree harvest or prescribed fire, early successional wildlife species were greatly reduced within 50 years. Tree harvest benefited early successional species without greatly affecting habitat suitability for late successional species. These patterns are consistent with current knowledge of vegetation and wildlife response to management. The HNF planning team used the information on vegetation conditions and habitat suitability to select Alternative 5, which utilized a focal area to concentrate harvest activities, as the preferred alternative. By incorporating ecological processes of disturbance and succession while retaining the resolution necessary for evaluating wildlife habitat suitability, this modeling approach contributed knowledge to the planning process and was a valuable tool for communicating differences among alternatives to stakeholders. As such, this approach serves as a template for successful planning on forested lands and will be of value to land managers, resource planners, and ecologists.