The US National Vegetation Classification (NVC) promises to provide a fine-grained community classification that is sensitive to subtle variation in environment and biodiversity and backed by quantitative plot data that allow both accurate description/prediction of vegetation and assessment of degree of fit with described types. These units have the potential to serve as targets for management and restoration. Because they represent the range of variation across the type, they are more representative and reliable than traditional reference sites. In the Carolinas we are developing a cyber infrastructure based on the NVC and supporting vegetation plots to guide management and restoration of natural lands. The first step is use of an extensive database of plots (>7000) to validate and quantitatively describe the units within the classification. We examine available plots using multivariate methods, fit them to extant types, propose formal, plot-based descriptions, revisions, and additions, identify types that appear not represented and search out possible examples. Tools are provided for predicting which types should be expected at a particular site. This information is used in tools for selecting species to plant and for assessing the successional trajectory of monitored sites.
Results/Conclusions
For assessment of the NVC in the Carolinas, we selected vegetation groups where we had broad coverage. In North Carolina thirteen Piedmont riparian forest types are recognized. In an assessment of these forests in the Carp Fear Basin we recognized 7 forest types of which 3 matched established types and 7 did not. Insular community types such as those of rock outcrops or non-alluvial wetlands tend to be less predictable owing to chance events of immigration. In a study of non-alluvial wetlands of the southern Appalachians 127 plots were parsed into 12 heterogeneous types, in contrast to the 27 NVC types currently reported. Fire-maintained longleaf pine vegetation was once widely distributed across the Coastal Plain. Across the Carolinas 52 associations had been recognized. Our examination led to recognition of 57 of which 28 were essentially unchanged, 11 had complex relationship to old types and 17 were new. Thus far we have assigned 3395 plots to 441 associations. With 503 types currently attributed to the Carolinas, it appears that most will be represented. However, with 128 presently represented by a single plot, coverage appears uneven and considerable field work will be needed before the NVC is adequately documented for the Carolinas.