Lessons Learned While Integrating Habitat, Dispersal, Disturbance, and Life-History Traits into Species Habitat Models Under Climate Change

摘要

We present an approach to modeling potential climate-driven changes in habitat for tree and bird species in the eastern United States. First, we took an empirical-statistical modeling approach, using randomForest, with species abundance data from national inventories combined with soil, climate, and landscape variables, to build abundance-based habitat models for 134 tree and 147 bird species. We produced lists of species for which suitable habitat tends to increase, decrease, or stay the same for any region. Independent assessments of trends of large trees versus seedlings across the eastern U.S. show that 37 of 40 species in common under both studies are currently trending as modeled. We developed a framework, ModFacs, in which we used the literature to assign default modification factor scores for species characteristics that cannot be readily assessed in such models, including 12 disturbance factors (for example, drought, fire, insect pests), nine biological factors (for example, dispersal, shade tolerance), and assessment scores of novel climates, long-distance extrapolations, and output variability by climate model and emission scenario. We also used a spatially explicit cellular model, SHIFT, to calculate colonization potentials for some species, based on their abundance, historic dispersal distances, and the fragmented nature of the landscape. By combining results from the three efforts, we can create projections of potential climate change impacts over the next 100 years or so. Here we emphasize some of the lessons we have learned over 16 years in hopes that they may help guide future experiments, modeling efforts, and management.