EVIDENCE THAT CASTANEA DENTATA PERSISTS ON XERIC, MID-ELEVATION SITES IN THE SOUTHERN APPALACHIANS.

Castanea dentata (American chestnut) persists today in the southern Appalachian forests of North Carolina as sparsely occurring sprouting root systems. The introduction of two exotic fungal diseases into North America, root rot (Phytophthora cinnamomi) in the early 1800s and chestnut blight (Cryphonectria parasitica) in the early 1900s, eliminated C. dentata as a dominant canopy species. Identifying and understanding the characteristics of sites where C. dentata persists could offer insights for management and restoration. Using a regional forest inventory and analysis (FIA) database, I studied the ecology of C. dentata on 831 forested sample plots in 21 mountainous counties of western North Carolina. The purpose of my study was to determine if the presence of C. dentata was associated with common topographic variables and the co-occurrence of other tree species that could be used in a regression model for evaluation of sites for potential management activities. I found that C. dentata occurred on only 3.5% of the sample plots. Its presence was directly associated with elevation >900 m, positively related to the cooccurring species Quercus prinus (chestnut oak) and Q. rubra (northern red oak), and negatively related to Liriodendron tulipifera (yellow-poplar). Logistic regression revealed poor performance of formulations with multiple significant biological variables (i.e., tree species) because of multicollinearity effects with elevation. Good model performance was achieved with a two-variable formulation using elevation and a weighted averages score derived from direct gradient analysis and ordination of moisture affinities of the 30 tree species in the study plots. An explanation for the effect of elevation on occurrence of C. dentata on sites >900 m is unknown, but the arborescent species scores suggest probable xeric to subxeric plot moisture regimes. Limitations of the FIA data for my study were many including few plots with C. dentata, lack of plot soil characteristics and history of disturbance from fire. Results from my study may have application to rank forest sites for investigation of biological control of C. parasitica through hypovirulence and for identifying stands for application of silvicultural practices to reduce environmental stress and increase survival of existing root systems. [ABSTRACT FROM AUTHOR]