PS 87-70
Application of a new tolerance-based model to understand climate-related stress on US forests over the coming century

Friday, August 14, 2015
Exhibit Hall, Baltimore Convention Center
Jean Lienard, Department of Mathematics, Washington State University Vancouver, Vancouver, WA
John Harrison, School of the Environment, Washington State University Vancouver
Nikolay Strigul, Department of Mathematics and Statistics, Washington State University Vancouver, Vancouver, WA

Although it is widely recognized that climate change will require a major spatial reorganization of forests on the landscape, our ability to predict what this will look like has been quite limited. One fundamental problem has been our limited ability to scale up individual plant traits and growth/mortality characteristics to the ecosystem level due to ecosystem biocomplexity, including numerous non-linear functional relationships and feedback loops between different organisms. Current modeling efforts to predict future distribution of forested ecosystems as a function of climate include species distribution models (for precise, local scale predictions) and potential vegetation climate envelope models (for coarse-grained, large scale predictions). In this work we bridge these approaches by considering an intermediate level of complexity, using stand-level tolerances such as drought, shade and waterlogging tolerance. 


It is well established that environmental stressors increase mortality of intolerant trees. We demonstrate  that these effects propagate to the landscape level and that forest tolerances are explicitly shaped by climate. This discovery allows the development of a Tolerance Distribution Model (TDM), a novel quantitative tool to assess the impact of climatic changes at the biome level.  We first demonstrate that the shade, drought, and waterlogging tolerances of forest stands are strongly correlated with climate and edaphic characteristics in the conterminous USA. We then develop, apply and evaluate a TDM, that uses plant community tolerance indices to predict spatial distributions of forest ecosystems. Finally, using 17 climate change models, we identify forested ecosystems vulnerable to anticipated climate change, with a focus on drought-related stress. For this application in US, we show in particular that high elevation areas, Midwest US and Northeast US are at high risk under future climate. This approach is suitable for both small and large scales while maintaining a link to ecologically relevant climate-related stressors, and can be more broadly applied to understand climate effects on biome distribution.