Environmental cluster analysis as a tool for selecting complementary networks of conservation sites

Identifying networks of sites that efficiently represent the biotic diversity of larger regions is a prerequisite for effective, conservation planning. Recent approaches for systematic conservation planning rely on the idea of complementarity, i.e., the selection of networks of sites that complement each other in their species composition. A major obstacle in applying these methods is the need for detailed data on species distribution, which is usually not available. In this study, we test the hypothesis that cluster analysis based on environmental variables (rainfall, temperature, and lithology) can be used to identify sets of complementary sites that efficiently represent regional species diversity. The performance of this approach (which we term environmental cluster analysis, ECA) is evaluated using an extensive database of the flora of Israel as a test case. Our results indicate that the ECA performed significantly better than a random null model in representing regional floristic diversity. Moreover, sites representing regions (clusters) defined by the ECA were more efficient in capturing the floristic diversity of Israel than sites representing the floristic regions of the area. In contrast to our expectation, the efficiency of the ECA was particularly pronounced in the analysis of rare species. The main mechanism behind the superiority of the ECA was its ability to identify sets of sites characterized by high turnover of species, rather than individually rich sites. The overall results suggest that ECA may serve as an important tool for the identification of complementary networks of conservation sites.