Abstract
Rivers are dynamic and complex systems that constantly change their composition from sources to deltas. This is due to the influence of a set of variables controlled by hydro-litho-eco-atmospheric processes and anthropoclimate pressures which are, in turn, influenced by the catchment attributes. River water, compared to other environmental media, respond faster to disturbances, which are extensively and immediately reflected by the chemical composition of its waters. The paper explores the possibility to transpose the concept of ecological regime shift to river chemistry, by means of theoretical thoughts and a practical application to the case of the Tiber River (central Italy). Compositional Data Analysis (CoDA), robust Principal Component Analysis (PCA) and score-distance graphs were used to investigate data variability and the interlinks between response and forcing variables. The findings outline mechanisms and factors influencing the river self-restoring ability at a basin-wide scale, providing a better comprehension of the circumstances controlling the water system resilience, one of the major and most urgent challenges for the future of mankind.