Monday, August 2, 2010 - 1:50 PM

OOS 2-2: Potential changes in transpiration with shifts in species composition following the loss of eastern hemlock in southern Appalachian riparian forests

Chelcy Ford, USDA Forest Service and James M. Vose, USDA Forest Service Southern Research Station.


Eastern hemlock (Tsuga canadensis (L.) Carr.) is declining throughout much of its range in the eastern US due to hemlock woolly adelgid (HWA) infestation. The loss of this foundation species will impact the hydrologic cycle in these systems. To estimate the impact on the hydrologic budget, we quantified transpiration over six years for T. canadensis, and over three years for co-occurring species Acer rubrum, Betula lenta, and Rhododendron maximum using sapflow probes. These three species represent the likely woody species that will dominate the trajectory of succession following the loss of eastern hemlock. In areas where R. maximum is a dominant component of the shrub layer, regeneration of overstory tree species is severely restricted. We developed relationships between transpiration and climate for all species. Given the loss of T. canadensis from the ecosystem, we modeled implications on transpiration from two succession scenarios: one in which hemlock is lost from the canopy and a R. maximum subcanopy results, and one in which eastern hemlock is replaced in the canopy by A. rubrum and B. lenta.


Transpiration was shown to decline since 2004 for T. canadensis, and no such decline was observed for the other species. We found that with the loss of hemlock leaf area, light levels in the subcanopy increased almost 17-fold, and we estimated that R. maximum would increase transpiration by over 4-fold. Although R. maximum transpiration increased, this increase was not enough to make up for the loss of T. canadensis’ contribution to transpiration. For example, estimated sap flow of R. maximum during days 142–285 in 2005 was found to be 4.78 g s-1 while sap flow of T. canadensis under healthy conditions during several days of the growing season in 2004 was 14.78 g s-1. By contrast, if species composition shifted from T. canadensis to 100% A. rubrum, 100% B. lenta, or a 50/50 mixture of the two species, stand sap flow increased by 38%, 71%, and 55% respectively. Although actual post-mortality scenarios are uncertain, the loss of T. canadensis will result in changes in the structure and function of this ecosystem.