Influência do uso e cobertura do solo sobre a ictiofauna de riachos das cabeceiras do Alto Rio Paranapanema

Abstract

Human activities have generated unprecedented impacts on aquatic biodiversity. Thus, knowing the impacts generated by these activities is extremely important, as fish from the Neotropical regions represents approximately 1\5 of the world's fish species, or perhaps 10% of all vertebrate species on the planet. Agricultural activities cause changes in the landscape and, depending on land use, they can reduce permeabilization and percolation, increase runoff, increase in sedimentation, fragmentation of forests, and the entry of pollutants into the watercourse. These effects cause changes in the stream structure and, consequently, affect the fish communities. However, different types of land use do not promote identical changes in aquatic systems, so the effects on fish communities will be differentiated. Understanding the direct and indirect effects that affect the structure of fish assemblages is of great relevance, moreover, in studies that include several factors: land-use and land-cover (LULC), stream structure, geographic and spatial (watercourse distance). These types of studies promote the expansion of knowledge to support conservation measures. The present study aims to answer whether there is a relationship between land use for agricultural activities and fish assemblages and to test whether these relationships are dependent on the spatial scale. The study was conducted in the southeastern of the Rain Forest in an extension of approximately 240 km from the Upper Rio Paranapanema Basin. Sampling was carried out in 30 streams of headwaters that include stretches of 2nd to 5th order. Satellite images were acquired to carry out the classification of land-use and land-cover. Redundancy analyzes (RDA) were applied to test the relationship between land use and composition and paths analyses to test direct and indirect relationships between land use and the richness of stream fish species. The results showed that there is a relationship between land use and the composition and richness of fish species, but it is dependent on the spatial scale adopted. The analysis of partition of variance and redundancy showed that in the catchment scale most of the variation in LULC was explained together with by geographic and spatial factors indicating intercorrelation between these two sets of variables. There was a relationship between local environmental variables and LULC on the catchment scale. The RDA test showed that there is a relationship between composition and local environmental variables and spatial factors. There is an indirect relationship between the richness of fish species and land use, but it depends on the type of use practiced. The species were associated with the type of land use. Generalist species are more common in streams whose catchment has higher anthropic pressure. Our work indicates that the hierarchical concept is valid to identify the environmental factors that affect the composition of the fish assemblages. This study highlights the importance of catchment management in the preservation of stream fish. The work also indicates that agriculture has a greater indirect (negative) effect on the fish richness and that different types of land use indirectly affect the fish richness of streams, as the land use on the watershed scale contributes to the alteration of the stream structure, which consequently affects the assembly of stream fish.