Abstract
Restoration of degraded coastal and estuarine habitats owing to human activities is a major global concern. In Puget Sound, Washington, U.S.A., removal of hard armor from beaches and intertidal zones has become a priority for state and local agencies. However, the effectiveness of these shoreline restoration programs for subtidal habitats and fish is unknown. We surveyed six restoration sites in Puget Sound over 2 years to evaluate associations between shoreline restoration and subtidal fish abundance. We measured the abundance of juvenile salmonids and forage fishes along armored, restored, and reference shorelines. Bayesian generalized linear models showed limited support for associations between shoreline restoration and these fishes in the 3-7 years since armor removal. Pacific herring were more abundant at reference shorelines; the shoreline effect for surf smelt varied by survey site. Shoreline restoration was not an important predictor of salmonid abundance; the best models for Chinook and chum salmon included predictors for survey site and eelgrass, respectively. The retention of survey site in several species' top models reveals the influence of the broader landscape context. We also found seasonal variation in abundance for chum salmon and surf smelt. Our results suggest that juvenile forage fish and salmonids in estuaries likely have unique responses to shoreline features, and that the positive effects of armor removal either do not extend into subtidal areas or are not detectable at local scales. To be most effective, coastal restoration programs should consider broader landscape patterns as well as species-specific habitat needs when prioritizing investments.