Effects of salinity gradients on benthic invertebrate and diatom communities in a German lowland river

The Lippe is a lowland river located in the Western part of Germany and has been heavily impacted by coal mining activities ever since. Although mining activities significantly decreased during the last decades, the associated discharge of salt-enriched mine water into the river still poses a persistent threat to the local benthic invertebrate and diatom communities. To analyze the effect of salt pollution on invertebrate and diatom species, biological and chemical data were compiled for this study from a publicly available database. Changes in the community composition due to increased salt concentrations were explored by Non-Metrical-Multidimensional Scaling. Indicator species and salinity thresholds for single species and communities were identified using the method TITAN (Threshold Indicator Taxa Analysis). The method is an analytical approach to detect changes in frequency and abundance of species along an environmental gradient by combining the methods of change point analysis (nCPA) and indicator species analysis. The obtained salinity preferences and individual and community thresholds were compared to the literature and existing salinity classifications. For both diatoms and benthic invertebrates, Non-Metrical-Multidimensional Scaling showed a clear split between samples of high and low salinities. Significant salinity thresholds were determined for 50 invertebrate and 58 diatom species of which 23 respectively 18 species were described as 'reliable' indicators according to the specifications given by Baker and King (2010). A majority of salt-tolerant indicator organisms were invasive species. For both organism groups, major changes in community composition were detected at a conductivity value exceeding 900 μS/cm. A reduction of the average salinity to below this threshold may have positive effects on the overall species richness and the persistence of sensitive taxa in the river Lippe. Individual and community thresholds may however be data-dependent to a certain degree and subjected to fluctuations considering the potential interdependencies between salinity and additional physico-chemical and environmental parameters (e.g. water temperature, lime content).