The ecological restoration of large rivers needs science-based, predictive tools meeting public expectations: an overview of the Rhône project

International audience1. Effective environmental management needs models that reliably predict quantitative ecological changes as a function of restoration effort (e.g. cost) and meet expectations of stakeholders. Principal threats to large rivers are linked to human-caused modifications of discharge and morphology of channels and floodplains. However, comprehensive large-scale tests of the reliability of models pre- dicting ecological consequences of restoring these elements are still lacking.2. Following a governmental decision, water managers, local authorities and the ‘Compagnie Natio- nale du Rho^ne’ financed a scientific programme to develop, test and subsequently use predictive models to assess the restoration (particularly minimum flow increases and reconnections of flood- plain channels with the main channel) of eight regulated reaches of the French Rho^ne River. This approach was fostered by (i) the existence of local initiatives aimed at the ecological improvement of the Rho^ne; (ii) a history of interactions based on trust among stakeholders; and (iii) knowledge pro- vided by a large interdisciplinary research group that studied the Rho^ne for two decades before the programme started in 1998.3. This Special Issue synthesises the insights gained over recent decades of research during which four river reaches (total length 47 km) were restored since 1999. It contains 11 articles including this overview. One article relates physical habitats in the floodplain to river hydrology and morphology; five articles test predictive models linking changes in habitat conditions to changes in taxa abun- dance, community metrics and biological traits of macroinvertebrates and fish; and four articles address the effects of restoration in larger contexts (long-term community trends, optimisation of sampling strategies, social processes and bioindication).4. We describe the Rho^ne restoration project, explain the conceptual framework used to predict the effects of restoration on river biota and describe the contents of the Special Issue, the main results and their implications. 5. The Rho^ne restoration led to more lotic and diverse aquatic communities and renewed social links with the river. When reliable pre-restoration data are available, simple habitat models can be used to predict quantitative ecological changes as a function of restoration effort. The project illustrates the need to describe changes in hydraulic conditions in studies of physical river restoration and shows the effort required for a powerful assessment of restoration effects