The verification and application of bioindicators: a case study of dung beetles in a savanna ecosystem

Abstract

Summary Bioindicators of habitat quality and environmental change must be identified quantitatively and tested independently to confirm their usefulness. We used the indicator value (IndVal) method, which combines measures of habitat fidelity and specificity, to assess the indicator responses of dung beetles (Coleoptera: Scarabaeidae) in Tembe Elephant Park, South Africa. The indicator responses were verified by sampling in different areas of the Park, 2 years after the responses were originally quantified. We postulated that terrestrial insect indicators with different combinations of habitat specificity and fidelity might fulfil different indicator roles. Indicator species with strong habitat specificity (characteristic species) are unlikely to provide information on the direction of ecological change despite high vulnerability. Rather, detector species that span a range of ecological states are likely to be better in this role. We used IndVal for selecting such detector species that indicate the direction of ecological change. Sets of species were found to be robust bioindicators, i.e. reliably characteristic across the habitat of which they were indicative. The suite of indicators was refined by discarding those with IndVals that varied significantly across years, thus improving the confidence in the final suite of species selected. By clearly responding to a change in habitat between two ecological states, detector species provided information complementary to that provided by characteristic species. The IndVal method enabled both the identification and testing of indicator (characteristic and detector) species. Because of its resilience to changes in abundance, IndVal is a particularly effective tool for ecological bioindication. We conclude that both characteristic and detector species are useful bioindicators of habitat quality and conversion. We propose that bioindicators that are categorized and verified in this way will have valuable application in the monitoring of habitat integrity.