Abstract
The aim of this study was to assess the spatio-temporal variation in water quality in the Grande River and the Ondas River, in the city of Barreiras, Bahia, Brazil. Water samples were collected at 11 points along the rivers, and eight physical-chemical parameters (electrical conductivity, pH, alkalinity, apparent and true color, turbidity, dissolved oxygen and biochemical oxygen demand) and three microbiological indicators (heterotrophic bacteria, total and thermotolerant coliforms) were analyzed. Spatio-temporal variation was assessed using the multivariate techniques of Principal Component Analysis/Factorial Analysis (PCA/FA) and Hierarchical Cluster Analysis (HCA). The results of the PCA/FA highlighted eight of the eleven parameters as the main ones responsible for the variations in water quality, with the greatest increase in these parameters being observed in the rainy season, especially among the points influenced by sewage discharges and by the influence of the urban area. The CA grouped the results from 11 points into three main groups: group 1 corresponded to points influenced by sewage discharges; group 2 grouped points with mainly urban influences; and group 3 grouped points in rural areas. These groupings showed the negative influence of urbanization and also statistically significant variations between the groups and periods. The most degraded conditions were in group 1, and the least degraded conditions in group 3. Assessment of the variations between the monitoring periods showed that rainfall had a significant impact on the increase or decrease in the parameters assessed, as a result of surface runoff linked to urbanization and increased river flow.