2017 journal article
Carolina Vegetation Survey: an initiative to improve regional implementation of the US National Vegetation Classification
PHYTOCOENOLOGIA, 48(2), 171–179.
Purpose: The purpose of the Carolina Vegetation Survey (CVS) is to provide a framework for characterization of natural plant communities throughout North and South Carolina and adjacent US states. The resulting classification supports scientific interpretation of vegetation pattern, biodiversity inventory, biodiversity monitoring, conservation efforts, and identification of restoration targets. Application of the approach: The CVS classification approach will lead to a synthetic treatment of the vegetation of the Carolinas. Although regional in its scope, the approach is generalizable to other geographic regions. It will support further development of the US National Vegetation Classification (USNVC), providing a model for similar work in other regions, thereby leading to more rapid improvement and application of the USNVC. Main features and protocols: Our protocols were developed for use with a large database of vegetation-plot records inventoried using a consistent, published methodology. Plot sizes typically range from 100 to 1000 m2, although data from smaller subplots are also collected. Each record has a full list of vascular plant species and includes cover-class estimates and tallies of woody stems. Species concepts and nomenclature are regularly updated to a consistent standard. Supporting data include soil chemical and physical properties and other site attributes. Class definition procedures employ node-based agglomerative hierarchical algorithms, informed by ordination procedures and by a priori assignment of records to vegetation classes. Advantages and limitations: Classification protocols draw on widelyused, well-established procedures and algorithms. Typological resolution aims to conform to one or more of the lower levels of the USNVC hierarchy. A limitation is that most plots were located using preferential sampling, which has the potential for incorporating selection biases. However, this approach captures rare or unanticipated types that would otherwise be missed. To date CVS data collection has been restricted to natural communities and consequently cannot inform classification of semi-natural or cultural vegetation.