Land use at the reach scale as a major determinant for benthic invertebrate community in Mediterranean rivers of Cyprus

Abstract The European Water Framework Directive (WFD 2000/60/CE) requires the quantification of anthropogenic pressures for evaluating the risk of failing environmental objectives when setting river basin management plans and for biological data interpretation when assessing ecological status. Many studies emphasized the close relationship between biological communities (e.g., fishes and invertebrates) and land use. The need for evaluating the impact of non-natural land use at different spatial scales is often stated, and relatively small spatial scales can be particularly relevant to understand biotic community changes. Abiotic descriptors derived for the reach scale can thus be used to evaluate the success of restoration measures, and their selection and testing is urgent for river management purposes. In this context we describe a new index, which allows quantification of land use modification at the reach scale, the ‘Land Use Index - reach’ (LUIr). The LUIr was developed on the basis of data gathered in the field with the CARAVAGGIO method (river habitat survey technique). A full description of the index is presented. In addition, jointly with a series of environmental descriptors developed for the quantification of general or specific anthropogenic stressors ( sensu WFD), the effectiveness of the new index (LUIr) in describing environmental gradients as perceived from benthic invertebrates has been evaluated for Mediterranean rivers in Cyprus. The multiple regression analysis performed (all subset regression and relative importance) revealed that among the set of environmental descriptors the LUIr resulted to be the factor best explaining the variation of biological metrics. Additionally, the analysis highlighted that the biological metrics more influenced by land use at the catchment level (LUIc) are those estimating the overall diversity and are consequently more influenced by factors at larger scale. The descriptor derived by physico-chemical parameters was in a few cases selected by regression analysis as the most important factor in explaining variation of biological metrics. The descriptors able to detect variation of other habitat features (i.e., habitat modification, habitat diversity and lentic–lotic character) resulted as weak predictors in the models. Our results support the conclusion that the LUIr index can be a useful and cost-effective tool to interpret invertebrate community variation. Moreover, as urgently requested by WFD, LUIr can be used to set and check effectiveness of measures for the improvement of aquatic ecosystems, particularly for Mediterranean streams where agricultural land use is one of the main anthropogenic pressures.