Species Habitat Modeling Under Future Climates

As climates shift and climate change intensifies during the 21^st century, many species will need to migrate to new regions that will better support their survival and persistence. Understanding these potential impacts of climate change on tree and bird habitats can help forest managers plan for the future and mitigate some impacts of climate change. To address this need, we developed a modeling framework that integrates forest inventory records and environmental data, to predict the current habitat associations for individual tree and bird species in the eastern United States. These models are used to provide information on possible changes in habitat suitability under future climate scenarios.

Approach

Our modeling framework proceeds in multiple stages. First, we predict habitat suitability in contemporary and future climates (DISTRIB model), then we use historical migration rates and generation times to estimate which areas are most likely to be colonized by species (SHIFT model). The DISTRIB model, developed in the Eastern United States, predicts habitat suitability using regression-based machine learning techniques. The relationship between species-suitable habitats and environmental conditions can be predicted to non-inventoried landscapes or under scenarios of climate change. While locations of newly suitable habitats are informative to forest managers, understanding the likelihood of the species to colonize these habitats adds realism to the predictions. The SHIFT model uses current abundance, historical migration rates and propagule generation times of species to compute end-of-century migration potential for each species.

Maps, statistics, and summaries for individual species of trees and birds are provided in the Climate Change Atlas.

In addition to our work on eastern species, we have expanded our efforts to now include species in the Western United States, as well as the Canadian portion of species ranges, by merging inventories from Canada and the United States. This research is being applied to expand the scope further to include North American species habitats and migration potential for the updated Silvics of North America program.

Our habitat suitability models use several climate metrics to predict individual species relative abundance. While not all environmental data sets are important predictors within the model, the information can provide useful context to interpreting model outputs.

StoryMap: Climate Change Pressures in the 21st Century: Shifts in Growing Degree Days, Plant Hardiness Zones, and Heat Zones

The SHIFT model was modified to estimate the spread of Emerald Ash Borer (EAB), an invasive insect responsible for killing millions of ash (Fraxinus) trees throughout much of the Midwestern USA. By modeling the risk of spread for EAB in Michigan, Minnesota and Ohio to account for the insect's flight characteristics (Insect Flight Model) and external agents that enable EAB to "hitch a ride" (Insect Ride Model), forest managers have used this information to develop plans aimed to reduce or monitor further spread of the insect in other states.