Using Digital Terrain Modeling to Predict Ecological Types in the Balsam Mountains of Western North Carolina
Relationships between overstory composition and topographic conditions were studied in high-elevation (>1300 meters) forests in the Balsam Mountains of western North Carolina to determine whether models could be developed to predict the occurrence of number vegetative communities in relation to topographic variables (elevation, landscape position, surface geometry, slope gradient, aspect) were measured on 0.10-hectare plots located across a range of mounainous landforms. Analysis of species density and dominance using multvariate classification and ordination techniques indentified five ecological types. Multivariate discriminant analysis indicated that elevation and landscape position were the most important environmental variables associated with presence of ecological types. Landscapes in the study area > 1300 meters were classified into predicted ecological types by applying discriminant functions to digital terrain models, using a desktop computer-based Geographic Information System. Accuracy of the classification, evaluated with an independent data set, indicated 86 percent agreement with model predictions. This approach may offer a relatively quick method for analyzing and predicting with limited data the distribution of important ecological condictions over large areas. To explore potential applications of this approach, results were applied in the Balsam Mountains and nearby Nantahala Mountains to predict occurrence of potential habitat for the endangered Carolina northern flying squirrel (Glaucomys sabrinus coloratus) and to aid in making decisions about recovery efforts and field surveys for the species.