Efeito da paisagem na ocorrência da ictiofauna em rios de primeira ordem da região sudeste do Brasil

Abstract

Objective: In this study we aimed to understand how extrinsic environmental factors measured in the watercourses and the surrounding landscape influence the ichthyofauna of first-order streams. Methods: Data were collected within the Corumbataí River Basin, São Paulo, southeastern Brazil, during the dry season of 2012. We sampled the ichthyofauna in 13 stretches of streams distributed across four river sub-basins. The stretches differed in relation to the presence/absence of riparian forest, the predominant type of matrix and the percentage of forest. Response variables were species richness and the occurrence of functional groups and explanatory variables include both local and landscape structures from the surrounding environment. Local variables comprised the following water quality and structural attributes: pH, temperature, conductivity, turbidity, flow rate, depth, width, type of substrate. Landscape variables included presence/absence of riparian vegetation, type of vegetation, type of matrix, percentage of forest and canopy cover. Results: A total of 268 individuals were recorded, which were distributed among 12 species. The landscape structure influenced the occurrence of functional groups in first-order streams, especially allochthonous-feeders, nektonic and hypoxia-intolerant species. The presence of riparian forest was the most important predictor. Species richness was negatively related to the presence of riparian vegetation, supporting the hypothesis that degraded landscapes lead to a reduction in diversity. Conclusion: The protection of riparian vegetation is critical to the maintenance of ichthyofauna diversity in first-order streams. The presence or absence of riparian vegetation differently affected the occurrence of species depending on their functional characteristics, particularly those related to the tolerance to hypoxia, source of alimentary items and the position in the water column.