In this work we present the application of multivariate statistical analysis and Geographic Information Systems for the conduct of studies the variation of water quality in catchment predominantly agricultural. We conducted the study in the watershed of the Rio das Lontras river/Parana; being carried out with the aid of a multi-parameter probe, in ten spots defined upstream and downstream of the watershed, in four distinct seasons. For mapping land use and occupation, we used mosaicked images dating from the year 2014, from Google Earth. The definition of the areas of influence by the Thiessen polygons technique and multivariate statistical techniques, particularly the Redundancy Analysis (RDA), were used for the correlation research between the explanatory and explanatory variables. In the area of spring, located upstream of the watershed, the contribution to the variation in water quality became smaller, compared to the other monitoring spots. In spots located downstream, we found that the interference in water quality has become greater, due to the effects of diffuse pollution along the watercourse.
Read full abstract