Abstract
This study evaluated space‐time variation of water quality data sets for the Wadi Andlou Kasserine, Tunisia, using multivariate analytical techniques including cluster analysis, principal component analysis (PCA) and correlation analysis (CA). The present study showed that values of temperature (T), electrical conductivity (EC), biochemical oxygen demand (BOD), chemical oxygen demand (COD), chloride, phosphate, sulfate, fecal coliform, Streptococcus, Salmonella spp., and Enterococci were significantly higher in the dry season compared to samples taken in the wet season. Concurrently, pH, dissolved oxygen (DO), turbidity, suspended matter (SM), NH4+ presented values significantly lower in dry season than surveys in rainy season. The following parameters BOD5, COD, SO42−, Cl−, SM, Salmonella spp. did not meet the Tunisian standard NT 106 for surface water quality. Analysis of pollutant concentrations via PCA showed that the two first components accounted for 87.34 and 93.17% of the total variance for dry and rainy season sampling, respectively. The first component included T, pH, EC, SM, SO42−, Cl−, NH4+, turbidity, COD, BOD, PO43−, Enterococci, fecal coliform and streptococcus, Salmonella spp., and total bacteria. The second included DO and NO3−. Cluster analysis identified the five sampling sites into three groups in the rainy and dry season based on seasonal differences and various pollution levels caused by physicochemical and bacteriological properties. The results suggest that future monitoring can include fewer sampling sites, and stratified random sampling plans can save time and resources.
The spatial and temporal patterns of water quality in Wadi Andlou are controlled by seasonal variations and discharges from the pulp and paper industry. This study shows the effectiveness of these techniques (PCA and CA) for getting better information about the water quality and the design of a monitoring network for effective management of water resources.
