Application of multivariate techniques in conservation planning frameworks

Over the last few years systematic conservation planning has gained greater focus in freshwater ecology. As most conservation plans target multiple biodiversity features, this is inherently a multivariate problem. Apart from the use of multivariate modeling techniques to estimate taxa distributions, we present three applications of classification and ordination techniques in a conservation framework. Highly unique sites - thus of high conservation value - can be identified using the multivariate distance to a group centroid or their nearest neighbors. Systematic conservation planning algorithms, such as bootstrapped heuristics or simulated annealing, deliver multiple solutions to meet conservation targets. Cluster analysis and non-metric multidimensional scaling help to select complementary conservation plans out of a pool of possible solutions. When taxa or other biodiversity features cannot be extrapolated across the landscape, multivariate environmental similarity patterns are linked directly to species patterns to create environmental surrogates that can be used for spatial prioritisation. We will demonstrate how these three approaches can enhance applied conservation planning schemes using invertebrate data from Victoria, Australia and the Yukon Territory, Canada.