Prioritizing barrier removal to improve functional connectivity of rivers

1. Freshwater systems are severely impacted by connectivity reduction due to the construction of dams and weirs. The breach of this longitudinal connectivity imperils freshwater fish species world-wide. There is thus an increasing need for numerical tools that help decision-makers correctly allocate resources to prioritize restoration actions. 2. This study provides a methodology for prioritizing the removal of barriers. It is based on spatial graphs, which represent structural units as nodes and relationships between nodes as links, and uses habitat suitability modelling (Boosted Regression Trees) to weight nodes. To exemplify the application of this procedure, we used the Tagus River network and evaluated the impact of the dams (29 built between 1928 and 2004) on the occurrence of each of two fish species (Iberian barbel Luciobarbus bocagei - representing large potamodromous fish; and southern Iberian chub Squalius pyrenaicus - representing small water-column residents) and on the combination of both. 3. Results show that dam construction on the Tagus was responsible for a 48.4-54.4% reduction in river connectivity for different fish species. Actions to promote connectivity in just seven of the dams would increase connectivity by 35.0-37.2%. 4. The removal of a single barrier chosen through prioritization had a greater overall connectivity increase than the random removal of seven barriers. 5. Synthesis and applications. The proposed prioritization method, using spatial graphs and habitat suitability modelling, makes it possible to model the impact of the removal or placement of an insurmountable barrier on the overall functional connectivity of a river network, facilitating resource allocation and minimizing the impact of new barrier implementation.