Large Scale Designs and Analytical Tools for Rigorous Evaluation of Habitat Enhancement Actions in Three Columbia Basin Watersheds

Substantial efforts have been undertaken to benefit ESA listed salmon and steelhead populations in the Pacific Northwest, but successful demonstrations of population-scale benefits of these actions are rare.  While research has shown some improvements in specific phases of salmonid life history due to management actions, much stronger cause-and-effect relationships between management actions and population response must be established to assess the effectiveness of regulatory and restoration actions in salmon recovery.  For most populations of ESA listed salmonids in the Pacific Northwest, current recovery strategies rely heavily on restoring freshwater habitat for spawning and rearing.  The rationale for this approach arises from the relative ease with which habitat restoration actions can be conceptualized and implemented, and the intuitive assessment that anthropogenic disturbance of watersheds must be a primary limiting factor of these populations’ dynamics.  Recently, the concept of Intensively Monitored Watersheds has arisen as one means of demonstrating the benefit of habitat actions on fish populations.  The basic premise being to concentrate the monitoring and treatment actions such that populations, or major portions thereof, can be treated and assessed, thereby generating sufficiently large effect sizes to be detected and attributed to the restoration actions implemented.  To generate sufficient effect sizes requires a concentrated effort of habitat restoration projects since individual habitat management actions generally do not directly impact population processes – their direct effect is to modify physical or biological habitat condition at too fine a scale to generate population level responses.  To generate the direct cause-and-effect relationships between the habitat actions and fish population process response requires a rigorous experimental design that incorporates the space and time scales of the predictor and response variables.  An adaptive management framework is necessary for the successful implementation and design of these projects; the framework forces the development of predictive, process models, affords the application of interim performance metrics for program modification, and is the basis for application of learning within and between project watersheds.