Adelges tsugae (hemlock woolly adelgid, HWA), an invasive insect native to Japan, was introduced to the United States in the early 1950s. HWA feeds on the ray parenchyma cells of Tsuga canadensis (eastern hemlock), causing defoliation and death. Since its introduction, HWA has spread throughout much of the range of T. canadensis, causing widespread mortality. Tsuga canadensis is considered a foundational species across much of its range due to its strong control over the microclimate. HWA spread to the southern Appalachian Mountains, reaching Great Smoky Mountain National Park in 2002. By 2006, HWA occurred in all watersheds in the Park. Currently, few trees survive outside of those treated with insecticide.
To examine post-adelgid changes in forest structure and composition, long-term vegetation monitoring plots were established in 2003, before the spread of HWA within the park, and resampled in 2017. These plots were distributed across five forest types: hemlock, montane cove, montane oak-hickory, acid hardwood, and northern hardwood. Plots with surviving treated T. canadensis trees were included as a separate type. Data were collected from overstory, sapling, seedling, and herbaceous strata.
Results/Conclusions
Nonmetric multidimensional scaling (NMDS) revealed unidirectional changes in both overstory and sapling species composition. The sapling strata of plots with greater pre-HWA T. canadensis importance values and lower R. maximum basal area generally exhibited greater compositional change between 2003 and 2017 and compositional changes were correlated with potential annual incident radiation, elevation, and percent slope. We observed less change in seedling species composition on plots with greater basal area of R. maximum, greater total sapling density, and lower pre-HWA T. canadensis importance. Herbaceous species composition in northern hardwood, acid hardwood, montane cove, and hemlock forests tended to converge through time. Environmental vectors indicated that elevation, sapling density, and R. maximum basal area were positively associated with greater changes in herbaceous species composition, while pre-HWA T. canadensis importance was negatively associated with changes in herbaceous composition.
Our results suggest that changes occurred across all vegetation strata. While we observed greater composition change in the seedling and sapling strata on plots with greater T. canadensis importance and less change on plots with greater R. maximum basal area, these relationships did not necessarily extend to the herbaceous layer, suggesting other variables, such as soil characteristics, may be more influential in determining species composition.