Support for the multi-taxa approach in biodiversity assessment, as shown by epigaeic invertebrates in an Afromontane forest archipelago

Abstract

Often a single indicator invertebrate taxon is used for assessing changing landscape patterns. However, we argue here against the exclusive selection of a single group. Covariation in diversity patterns of spiders, carabids, staphylinids and ants were compared in and between five highly naturally-fragmented Afromontane forest patches (size range: 3.5–25.2 ha) in South Africa. Significantly fewer individuals and species were captured in smaller forest patches (<6 ha) for most of the taxa, except Formicidae, where a higher number of species were captured in medium-sized patches (7–9 ha). When sampling effort was standardized, a higher diversity (rarefaction and Simpson's diversity index) was obtained in the smaller patches for Carabidae, Staphylinidae and Formicidae. The only significant positive correlation between taxa, in terms of numbers of species, was between Carabidae and Staphylinidae. The other taxa showed only weak positive correlations between species richness, or negative correlations. Multivariate techniques showed significant species turnover between patch assemblages for each taxon, and also showed that some taxa are more similar in assemblage-structure than others. An index of complementarity showed that species compositions of the selected taxa varied greatly between forest patches of different sizes. Our results support the multi-taxa approach, in conservation studies, even at the level of taxonomically-related groups sharing a common habitat stratum.