Shading effects on community composition and food web structure of a deforested pasture stream: Evidences from a field experiment in Brazil

Experimental manipulations in the field can evidence cause-effect relations from the manipulated variables. In this context, the aim of the present study was to investigate the effects of shading on stream community composition and food web structure. The experiment was conducted in two reaches of a pasture stream: one that remained open (control) and another that was covered with 75% factor shading cloth (treatment), sampled before and after 21 months. At each sampling, algae, macrophytes, plankton, meiofauna, macroinvertebrates, and fish were collected. All heterotrophic organisms were submitted to diet analysis and dietary data was used to calculate food web parameters. Community composition was evaluated through cluster analysis. A total of 7556 individuals of 148 taxa were identified. The control reach presented lower richness, abundance, and food web parameter values than the treated reach before shading cloth installation, whereas after the experiment the opposite was observed. Despite these differences, both reaches had changed over the experimental period for being under increased siltation as a result of long term land use effects, which resulted in higher similarity of community composition between periods than between treatments. This observation was corroborated by higher food web complexity before the experiment, with decreases in all food web parameters (except connectance) after 21 months, especially in the treated reach. Hence, decreases in community attributes and food web parameters after the experiment evidenced the effects of siltation, while the strongest decreases observed at the treated reach evidenced the effects of shading. Finally, we stress that the artificial shading itself does not promote allochthonous materials inputs (which add up to food web basis and promote habitat heterogeneity), so full-canopy riparian forest restoration is of fundamental importance for low-order streams. (C) 2013 Elsevier GmbH. All rights reserved.