Using a Fish-Based Model to Assess the Ecological Status of Lotic Systems in Serbia

Abstract

Fish biotic indices have become common tools for measuring and monitoring the ecological status of aquatic ecosystems. In this study, we aimed to develop the first fish-based model for stream quality assessment in Serbia taking into account the regional specificity of the country. Fish samples were collected between 2003 and 2011 at 131 sampling sites. We employed a self-organizing map (SOM) in order to group samples into river types depending on the character of the habitat they came from. Next, the k-means cluster analysis classified samples into four groups, each describing a particular ecological condition. The indicator species were presented for each group based on their constancy and dominance. Gradients over the SOM map were sought for 17 fish community metrics. On the basis of the core metrics and selected indicator species, we proposed a fish-based index for the assessment of the ecological status of running waters in Serbia.