Similarities among small watercourses based on multiparameter physico-chemical measurements.

With the introduction of the Water Framework Directive, the relative importance of smaller waterways increased. This statement is particularly true for Hungary, where water-quality monitoring of most smaller rivers only began 12 years ago. Due to their large number, and the lack of historical data concerning their state, systematic monitoring is a challenge. In the current study, 101 creeks are characterized on the one hand by 13 physico-chemical quality parameters (pH, electric conductivity, chloride ion concentration, dissolved oxygen, oxygen saturation, biochemical oxygen demand, chemical oxygen demand, total organic carbon, ammonium nitrogen, total inorganic nitrogen, total nitrogen, orthophosphate and total phosphorus), on the other hand by their watershed's relief, land use, and point sources' pollution indicators. Euclidean distance between water bodies (henceforth WBs) is calculated according to normalized physico-chemical monitoring values. They are grouped into clusters using the hierarchical clustering method. Watershed characteristics are used to explain the clustering via linear discriminant analysis. The investigation revealed that the main driver of cluster group creation is related to human impact: diffuse agricultural and point-source pollution. The first of the three clusters involved water bodies with low or no human impact; the second cluster contained those with medium-level anthropogenic disturbance, while waters with high pollution values formed the third cluster. Mean distance between heavily polluted waters was 1.5 times higher than that between those showing no or low disturbance, meaning that pristine waters are more similar to one another than polluted ones. The current number of samples per river is twice as high in cluster 1 as in cluster 3, revealing that there is room for optimization of the monitoring system. This contribution uses Hungary as a case study. [ABSTRACT FROM AUTHOR]