1. This paper introduces key messages from a number of papers emanating from the Second International Symposium on Riverine Landscapes held in August 2004 in Sweden, focusing on river restoration. Together these papers provide an overview of the science of river restoration, and point out future research needs.
2. Restoration tests the feasibility of recreating complex ecosystems from more simple and degraded states, thereby presenting a major challenge to ecological science. Therefore, close cooperation between practitioners and scientists would be beneficial, but most river restoration projects are currently performed with little or no scientific involvement.
3. Key messages emanating from this series of papers are: The scope, i.e. the maximum and minimum spatial extent and temporal duration of habitat use, of species targeted for restoration should be acknowledged, so that all relevant stages in their life cycles are considered. Species that have been lost from a stream cannot be assumed to recolonise spontaneously, calling for strategies to ensure the return of target species to be integrated into projects. Possible effects of invasive exotic species also need to be incorporated into project plans, either to minimise the impact of exotics, or to modify the expected outcome of restoration in cases where extirpation of exotics is impractical.
4. Restoration of important ecological processes often implies improving connectivity of the stream. For example, longitudinal and lateral connectivity can be enhanced by restoring fluvial dynamics on flood-suppressed rivers and by increasing water availability in rivers subject to water diversion or withdrawal, thereby increasing habitat and species diversity. Restoring links between surface and ground water flow enhances vertical connectivity and communities associated with the hyporheic zone.
5. Future restoration schemes should consider where in the catchment to locate projects to make restoration most effective, consider the cumulative effects of many small projects, and evaluate the potential to restore ecosystem processes under highly constrained conditions such as in urban areas. Moreover, restoration projects should be properly monitored to assess whether restoration has been successful, thus enabling adaptive management and learning for the future from both successful and unsuccessful restorations.