Selecting indicators based on biodiversity surrogacy and environmental response in a riverine network: Bringing operationality to biomonitoring.

An efficient indicator group should fulfill operational and performance-related criteria, including reasonable taxonomic knowledge, costs, response to environmental changes and strong proxy-capacity for biodiversity groups. However, in the real world many trade-offs are involved in the selection of an indicator group, and a single group rarely satisfies all criteria. We investigated cross-taxon congruence of assemblage composition patterns using both quantitative and qualitative data between adult dragonflies and aquatic insects (midges, Ephemeroptera-Plecoptera-Trichoptera (EPT), beetles and entire aquatic insects assemblage). Also, we tested whether environmental and spatial variables were important drivers for cross-taxon congruence. Finally, we developed a set of guiding criteria that avoid arbitrariness in the selection of the best indicator group. We sampled adult dragonflies and aquatic insects in 41 streams and rivers along a riparian vegetation loss gradient in a Neotropical riverine network. We used Procrustes analyses to verify surrogacy between groups and the association of each group with environmental and spatial predictors. The criteria used involves both operational and performance aspects to select an indicator group. Our results showed that adult dragonflies were weakly congruent with the entire aquatic insects assemblage and aquatic insect subgroups were moderate (beetles) to strongly (EPT and midges) congruent with the entire assemblage. Comparisons between EPT, midges and beetles were also weakly congruent, with the exception of midges-EPT. The association between assemblage patterns and overall environmental predictors was significant for all groups, while with spatial patterns, only midges and entire assemblage showed significant results. Numerical resolution slightly improved the congruence results. Incidence data could be used instead of abundance, due to higher congruence values compared to abundance data. Based on cost-benefit, EPT was selected as the best indicator group for monitoring the effects of riparian vegetation loss on aquatic biodiversity, and its use could be viable in biomonitoring programs. [ABSTRACT FROM AUTHOR]