RESUMO
Predicting how anthropogenic activities may influence the various components of biodiversity is essential for finding ways to reduce diversity loss. This challenge involves: a) understanding how environmental factors influence diversity across different spatial scales, and b) developing ways to measure these relationships in a way that is fast, economical, and easy to communicate. In this study, we investigate whether landscape and bioclimatic variables could explain variation in biodiversity indices in macroinvertebrate communities from 39 Atlantic Forest streams. In addition to traditional diversity measures, i.e., species richness, abundance and Shannon index, we used a taxonomic distinctness index that measures the degree of phylogenetic relationship among taxa. The amount of variation in the diversity measures that was explained by environmental and spatial variables was estimated using variation partitioning based on multiple regression. Our study demonstrates that taxonomic distinctness does not respond in the same way as the traditional used in biodiversity studies. We found no evidence that taxonomic distinctness responds predictably to variation in landscape metrics, indicating the need for the incorporation of predictors at multiple scales in this type of study. The lack of congruence between taxonomic distinctness and other indices and its low predictability may be related to the fact that this measure expresses long-term evolutionary adaptation to ecosystem conditions, while the other traditional biodiversity metrics respond to short-term environmental changes.
