Abstract
Primary production is a key feature for the functioning of aquatic ecosystems. The trophic state of streams reflects this primary production and is influenced by local limnological variables and the level of integrity of the watershed. In this study, the habitat integrity index (HII) and the physical, chemical, and biological parameters of tropical streams were sampled in two sub-basins, in order to understand the functioning of these ecosystems. We hypothesized that the limnological variables and the environmental integrity should determine the primary production in the streams, and that due to the similar land use in the sub-basins (i.e. agriculture and livestock) the streams would show similar limnological variables. To identify how the streams differed in relation to their limnological characterization and the HII, we performed Principal Component Analysis (PCA). The association between the variables was assessed by Pearson's correlation analysis. To identify the difference between the two basins, the Student's Test was performed. The best predictors of primary production were determined using multiple regression analysis with the Akaike selection criteria. The concentration of chlorophyll-a in the surface water indicated that the tropical streams sampled have, in general, low concentrations of nutrients, except at some more urban points. The variables that differed the most among the streams were blue-green cells/ml, pH, conductivity, and width of the riparian forest. Only redox potential and pH differed between the two watersheds. Temperature, blue-green cells/ml and width of the riparian forest were the variables that best predicted the primary production of phytoplankton in the watercourses in both watersheds. In this study, we emphasize the importance of temperature in aquatic productivity, particularly in the face of climate change which, along with deforestation, is increasing the amount of surface area that receives sunlight in the water courses, resulting in increased primary productivity and eutrophication.

This publication has 41 references indexed in Scilit: