Uso de distinção taxonômica e métricas de paisagem para avaliar comunidades de macroinvertebrados em córregos neotropicais

Abstract

To reduce diversity loss, it is essential to describe and predict how anthropogenic activities may influence the components of biodiversity. This challenge involves understanding how environmental variables influence diversity across different spatial scales and developing ways to measure these relationships in a fast, economic, and easily communicative way. In this study, we investigated whether simple landscape measures could explain the variability of a taxonomic distinctness index in macroinvertebrate communities in Neotropical region streams. For this purpose, we selected 39 streams. Fifteen streams were located in six protected areas (i.e., conservation units), five were located in a municipal reserve, and 19 were located in a region occupied by extensive monoculture of sugarcane or eucalyptus, banana plantations, and pastures. The results allowed us to generate seven hypotheses based on taxonomic distinctness in response to selected landscape metrics. Four of these hypotheses have previously been presented in the literature: i) natural environmental gradients can mask the disturbance effect on biodiversity; ii) some environmental gradients may not be strong enough to imprint responses on diversity indices; iii) studies evaluating multiples anthropogenic stressors may express environmental simplification resulting in a low response of indices; iv) the predictor variables used may not represent the important processes and mechanisms at appropriate scales for the evaluation of community; and v) the structure of macroinvertebrate communities responds to multiples factors in different scales, including local ones. The two additional hypotheses that we generated are the following: i) taxonomic groups assessed do not predictably respond to environmental gradients of anthropogenic disturbance in terms of taxonomic variability (indirectly phylogenetic) and ii) taxonomic refinement used in index quantification is not adequate to detect environmental gradients, of all the generated hypotheses the ones that if fit and explain of better form the gotten results are: hypotheses IV, v, vi and vii, amongst these, the two first ones corroborate with results of literature.