Background/Question/Methods   It has been long since ecology is predominantly descriptive with regards to community structure and dynamics. In recent years, however, many ecologists have conducted studies involving experimental manipulation in order to evidence cause-effect relations from the manipulated variables. The aim of the present study was to evaluate, using an experimental approach, how shading from an artificial coverage influences the community structure. The experiment was conducted on two stretches (30 meters each) of a pasture stream; one that remained opened (control), and another that was covered with 75% shade factor netting. The two stretches were sampled before coverage installation (August 2007), and then re-sampled 21 months after the procedure. Macroalgae and macrophytes were sampled using the square method, and the percentage of stream bottom coverage by each producer was estimated; macroinvertebrates were sampled with Surber and D nets, plankton with 45 µm mesh net, meiofauna with a PVC tube, and fish with two electrofishing passes.

Results/Conclusions   Overall, 7,310 individuals belonging to 108 taxa were identified. Before coverage installation, the opened control stretch presented lower species richness (62), abundance (902), and producer coverage (83%) when compared to the stretch to be covered (78 species; 2,215 individuals; 96%). After the experimental period, the covered stretch presented lower richness (47) and abundance (1,564) than the opened one (58 species; 2,629 individuals), what may have resulted from the lower producer coverage percentage (46%control, 18%covered), indicating the occurrence of changes at the bottom levels after the experiment. In addition, a MDS with cluster analysis has revealed that overall community structure was more similar between periods than between experimental treatments, since both stretches were under the same abiotic influences and highly increased siltation. With the help of a network analysis, we have verified that 43 taxa (as larger fishes like Hoplias malabaricus and Cyphocharax vanderii and sensible invertebrates, like Anacroneuria) were missing in both control and covered stretches whereas 21 taxa (as tolerant fish as Corydoras aeneus and mollusks like Biomphalaria and Melanoides) were registered, while few taxa (such as Poecilia reticulata and Melanoides) became dominant. Some other species occurred in all samples and showed not to be affected by coverage or siltation, as Americabaetis, Caenis, Cricotopus, and Simulium, this latter being the dominant one. Thus, these results have demonstrated that, in a relatively short period, the coverage has actually affected the stream community structure, though the siltation effects were of greater importance to explain the observed changes.