A fundamental characteristic of natural river ecosystems is their high degree of variability over time and space. While this variability makes the work of restoration practitioners and scientists very difficult, it also is critical to a river’s recovery. The resilience of stream ecosystems or their ability to be restored and withstand new disturbances is closely linked to their dynamic nature.
Yet, the processes of setting goals, designing projects and monitoring restoration efforts have all been constrained by human discomfort with environmental change and unpredictability. I explore various ways in which we as scientists might move forward with fundamentally new approaches for goal-setting, design, and monitoring that are process-based and embrace environmental variability. Traditional monitoring approaches are reviewed and results from a national-level study of river restoration in the United States are described. Less than 10% of all restoration project records in the U.S. include a written description of project assessment. Interviews with project managers suggest a somewhat higher rate of assessment that typically focuses on channel form and other structural metrics. The scientific, socio-political, and practical implications of shifting toward more process-based approaches will be discussed.