Water quality evaluation through a multivariate statistical HJ-Biplot approach

Water quality is a sensitive topic of worldwide concern that is defined by a series of physical, chemical and biological characteristics. The complex nature for studying water quality requires finding simple models to identify the variables that influence it the most. Therefore, the use of multivariate analysis techniques will be of immense help to find relationships and conclusions which aid us to determine the state of water quality through biological, physical and chemical indicators. In this investigation, the following variables were evaluated: Temperature, pH, transparency, turbidity, nitrates, orthophosphates, phosphorus, total nitrogen, chlorophyll a, solar radiation, dissolved oxygen and microcystins. These determined water quality at the sample sites, Gamboa and Paraiso. The results obtained through the application of a two-way multivariate analysis method called HJ-Biplot reflect variables relationships of chemical, physical and biological compositions. Furthermore, results conform two clusters of sample points that satisfactorily match to the region seasons. Cluster 1 is characterized by the presence of the following variables: pH, transparency, chlorophyll a, oxygen and temperature. On the other hand, cluster 2 comprises the following variables: nitrate, orthophosphates, turbidity and P-total. They are all parameters that suffer variations in the rainy season, and that, in turn, can influence the presence of ciabobacteria with toxigenic potential. Furthermore, this study demonstrates that the multivariate statistical methods are valuable for interpreting complex data sets, specifically, for water quality monitoring network.