USING RIVER RESTORATION OPERATIONS TO TEST PREDICTIVE ECOHYDRAULIC MODELS: FISH AND INVERTEBRATE COMMUNITIES OF THE RHÔNE RIVER.

Despite strong investments dedicated to river physical restoration, it is still difficult to assess the ecological effects of restoration operations. This is largely due to the difficulty of organizing long-term scientific surveys that anticipate operations and allow comparisons of their effects over multiple years in multiple sites. It is even more difficult to assess how ecohydraulic models can predict actual biological responses to stream restoration. As other large rivers in Europe, the Rhône River has been regulated for navigation, irrigation and hydroelectricity since the 19th century. A restoration program of the river started officially in 1998. It has involved minimum flow increases in bypassed main channels, the modifications of the connections of several dozens floodplain channels with the main river and the dredging of their sediments. So far, four bypassed main-channel sections and 26 floodplain channels have been restored. We used the Rhône restoration project as a template for testing the predictive power of several ecohydraulic models. In particular, we developed models to predict how fish and macroinvertebrate communities should respond to the physical restoration of the river. Some of these were based on hydraulic microhabitat models, that predicted how observed microhabitat preference of fish and invertebrate taxa should result in community changes after restoration. Other models were based on observed correlations between community structure and the degree of connectivity of floodplain channels with the main river. Several years after restoration, important community changes were observed in th