Effects of European land use on contemporary tree-climate relationships in the northeastern United States: Implications for predictive models

Much of our understanding of the climatic controls on tree species distributions is based on contemporary observational datasets. For example, forest inventory analysis (FIA) and other spatial datasets are used to build correlative models of climate suitability for plant taxa for use in environmental niche models. More complex dynamic models rely on species interactions, physiological processes, and competition, among other processes, that are also parameterized against contemporary data. However, as much as a quarter of the forested region in the upper Midwestern United States may be considered novel relative to pre-settlement baselines (Goring et al. submitted). Hence, modern surveys or even long-term datasets may represent only a portion of the ecological or climate space taxa might occupy. Using gridded datasets of pre-settlement vegetation for the northeastern United States from Town Propritor Suveys and the Public Land Survey, we examine the effects of European land-use conversion – logging, agricultural conversion and re-establishment – on climate-vegetation relationships.

We show that in regions where land-use change is climatically biased, such as conversion to agriculture along the prairie-forest boundary, impacts on the realized climatic niches for various tree taxa can be significant. Improving predicted distributions of taxa is critical for planning and mitigating the effects of widespread shifts in forest composition resulting from climate change. Using pre-settlement data can improve our understanding of the potential niches occupied by major forest taxa, improving the predictive abilities of environmental niche and mechanistic models.