An evaluation of the use of the nature conservancy vegetation classification for mapping bird distribution at chincoteague national wildlife refuge

Publication Date


Journal or Book Title

Natural Areas Journal


A study was conducted to evaluate the use of the 1995 Nature Conservancy vegetation classification (25 plant alliances recognised) for mapping bird distributions on Chincoteague National Wildlife Refuge, Virginia. 328 point counts were conducted in 10 plant alliances and, using goodness-of-fit tests, the hypothesis that bird species were randomly distributed among plant alliances was tested. Using cluster analysis, similarity in bird species composition among plant alliances was also tested. Of 56 bird species that were abundant enough for analysis, 53 were non-randomly distributed among plant alliances, and of these, 9 bird species selected only one plant alliance. Bird species composition was similar among the 3 tree-dominated plant alliances (Pinus taeda evergreen wetland forest alliance, P. taeda-Quercus nigra mixed evergreen-deciduous forest alliance and Pinus taeda evergreen woodland alliance), as well as among the 3 medium and short grass-dominated alliances (Spartina alterniflora, Ammophila breviligulata and Spartina patens). In contrast, both shrub-dominated alliances (Myrica cerifera and Baccharis halimifolia-Iva frutescens) and one grass-dominated alliance had bird species compositions that were very dissimilar to all other plant alliances. It is suggested that some bird species are habitat specific at the plant alliance level, however, the correspondence between bird species distributions and plant alliances depended partially on the plant alliance under consideration. It is concluded that the Nature Conservancy plant alliances are useful for mapping bird distributions, but that in areas in which the vegetation has not been classified, standard correlational approaches to analysing bird habitat relationships would be equally effective.







This document is currently not available here.