Effectiveness of biodiversity indicators varies with extent, grain, and region

The use of indicator taxa for conservation planning is common, despite inconsistent evidence regarding their effectiveness. These inconsistencies may be the result of differences among species and taxonomic groups studied, geographic location, or scale of analysis. The scale of analysis can be defined by grain and extent, which are often confounded. Grain is the size of each observational unit and extent is the size of the entire study area. Using species occurrence records compiled by NatureServe from survey data, range maps, and expert opinion, we examined correlations in species richness between each of seven taxa (amphibians, birds, butterflies, freshwater fish, mammals, freshwater mussels, and reptiles) and total richness of the remaining six taxa at varying grains and extents in two regions of the US (Mid-Atlantic and Pacific Northwest). We examined four different spatial units of interest: hexagon (∼649 km2), subecoregion (3800–34,000 km2), ecoregion (8300–79,000 km2), and geographic region (315,000–426,000 km2). We analyzed the correlations with varying extent of analysis (grain held constant at the hexagon) and varying grain (extent held constant at the region). The strength of correlation among taxa was context dependent, varying widely with grain, extent, region, and taxon. This suggests that (1) taxon, grain, extent, and study location explain, in part, inconsistent results of previous studies; (2) planning based on indicator relationships developed at other grains or extents should be undertaken cautiously; and (3) planning based on indicator relationships developed in other geographic locations is risky, even if planning occurs at an equivalent grain and extent.