Assessing conservation regionalization schemes: Employing a beta diversity metric to test the environmental surrogacy approach

Aim: Systematic conservation planning often involves the application of a regionalization scheme, which is assumed to delineate distinct ecological communities of target species. Commonly, such schemes are constructed using environmental surrogates; however, their effectiveness with regard to community delineation has been questioned in previous studies. Here, we aim to assess multiple environmental regionalization schemes' ability to delineate avian communities and to evaluate these schemes against a regionalization scheme built using species data directly. Location: British Columbia, Canada. Methods: We employed a beta diversity metric using community data from the BC Breeding Bird Atlas in multiple analysis of similarity (ANOSIM) tests, to assess the ability of a number of environmental regionalization schemes to delineate species turnover. We also developed a new species-based scheme using kriged local beta diversity values and a thematic resolution optimized through ANOSIM testing, which was then evaluated against the previously tested schemes. Results: All regionalization schemes delineated significant patterns in community structure, with the Bird Conservation Regions performing most similarly to the species-based regionalization. Regionalizations that required regions to be spatially contiguous outperformed non-contiguous regionalizations. Increasing thematic resolution (the number of regions within a regionalization) improved a regionalization's overall performance; however, regional redundancy also increased. Main conclusions: We argue that environmental regionalizations can function as effective alternatives to species-based regionalizations, particularly in areas with poor availability of species data. Also, we conclude that spatially contiguous regionalizations are superior to non-contiguous ones for delineating distinct communities. Lastly, we demonstrate how thematic resolution represents a trade-off between overall regionalization performance and regional redundancy, and how differing thematic resolutions can be employed depending upon the goals of the user.