USING WILDLIFE COMMUNITIES TO IMPROVE VEGETATION CLASSIFICATION FOR CONSERVING BIODIVERSITY

Determining which vegetation types of organisms perceive similarly and classifying these types into groups that function as similar habitats are necessary steps toward expanding the focus of conservation strate gies from single species to ecosystems. Therefore, the methods used to determine these habitat classifications are crucial to the successful design and implementation of these conservation strategies. Typically , this process has been accomplished through best professional judgeme nt. We used quantitative techniques to group vegetation types into hab itats based on the occurrence of breeding wildlife species (n = 420) i n Oregon. After calculating faunal similarities among all regional veg etation types (n = 130), we used cluster analysis to group vegetation types into wildlife habitats. We classified the original 130 vegetatio n types into 30 wildlife habitat types that we believe function simila rly. We tested this classification to assess whether vegetation types could be correctly classified into habitat types based on wildlife spe cies composition. Discriminant analysis correctly classified 95% of th e vegetation types into their wildlife habitat types, strengthening ou r confidence in this approach. This approach for classifying habitat t ypes allows consistent development of conservation strategies at coars e resolutions and aids in identifying vegetation types where additiona l biodiversity surveys are needed. Finally, this approach can be refin ed continuously as the precision of vegetation mapping and our underst anding of organism-habitat associations improve.