How do indicator groups provide information about the relative biodiversity of different sets of areas?: on hotspots, complementarity and pattern-based approaches
Regional conservation evaluation requires the effective use of available surrogate information, such as that provided by indicator taxa, for estimating the biodiversity represented by candidate protected areas. A recent study demonstrated that accumulating areas individually species-rich for one group of indicator organisms generally did not result in a set containing the areas that were species-rich for other groups. However, the requirement for a species-rich set, and not just individually-rich areas suggests an alternative assessment based on complementary-areas methods, which find a set of areas such that each of the indicator species is represented at least once. A set of areas covering all indicator taxa is assumed to be generally biodiverse. One limitation of this approach is that the set of areas species-rich for an indicator group is likely to represent many other organisms only if the members of the group span a wide range of habitats or environments. This supports the sampling of environmental pattern itself as an alternative strategy for selecting a set of biodiverse areas. Because such a pattern can be inferred from an indicator group, this same sampling strategy may extend the predictive value of such groups. Further, environmental pattern may incorporate additional useful information; for example, abundance information for the indicator taxa may improve predictions when used in this framework. These pattern strategies use the indicator group in a way that does not depend on the usual complementarity criterion in which a set is to be found that represents each indicator species. An example suggests that the complementary-areas approach may not be the best general strategy for using indicator groups in biodiversity assessment. We conclude that the current complementarity paradigm should be replaced by a more general pattern-based approach which views `complementarity' not as a predictor, but as a general property to be predicted.