Vegetation vulnerability to climate change in the Appalachians
Eastern U.S. forests have a long history of change related to climate and land use. As glaciers retreated and temperatures rose, tree species advanced from southern areas and may also have dispersed from isolated refugial populations. Recent changes include extensive deforestation by European settlers, fire suppression, and invasive pests and pathogens sucsh as chestnut blight and hemlock wooly adelgid. Although parts of the southeast have seen little or no warming in the 1900’s, climate change is expected to impact eastern forests in the coming decades and centuries. A critical question facing land managers in the Appalachian Landscape Cooperative (ALCC) is how climate change might influence the distribution of tree species and forest types of interest. We present an analysis focused on the ALCC and the Protected Area Centered Ecosystems (PACEs) of four National Park Service units within it: Upper Delaware Scenic and Recreational River (UPDE), Delaware Water Gap National Recreation Area (DEWA), Shenandoah National Park (SHEN) and Great Smoky Mountains National Park (GRSM). We used a combination of high resolution (800m) downscaled climate projections and correlative distribution models trained on Forest Inventory and Analysis field observations to assess the relative sensitivity of 40 tree species to projected changes in climate.
Most of the models were able to reliably differentiate between presences and absences in the current period and most of those were reasonably well calibrated. However, some species with small ranges such as Fraser fir and table mountain pine did not have enough observations to yield well calibrated models. At the scale of PACEs, suitability under future climate changed dramatically for some species (e.g. loblolly pine in GRSM) while changing relatively little for others (e.g. white oak in SHEN). Some species showed nonlinear changes in suitability (e.g. American beech in DEWA declining in the first half of the century while showing little change in the second half). In general the results agree with prior tree species modeling efforts in the eastern U.S. but provide improved depiction of topographically influenced climate gradients at spatial scales commensurate with eastern U.S. NPS units and other federal lands. In concert with Park based monitoring programs and field experiments these results can help inform climate adaptation priorities and actions in the coming decades.