Density-dependent estuarine habitat relationships of wild subyearling Chinook Salmon

ABSTRACT Objective While habitat limitations of anadromous salmonids in freshwater have received much attention, few studies have documented the habitat-based limiting factors during residence of juveniles in estuaries. Here, we evaluate the strength of density dependence in juvenile Chinook Salmon Oncorhynchus tshawytscha in the tidal delta of the Skagit River (Washington) using stock–recruit models and population summaries of average size, timing, and residence and the spillover of fry (i.e., rapid out-migration of the smallest life history type) into Skagit Bay. Methods We used multi-model inference based on long-term population monitoring data (1995–2018) in freshwater, tidal delta, and nearshore marine habitats associated with the Skagit River to examine the contribution of density-dependent and density-independent metrics on these population metrics. We examined multiple stock–recruit functions for two life stage transitions: (1) subyearling out-migrants to delta residents and (2) out-migrants to preharvest adults. The first transition is the estuary rearing life stage, while the second is the entire period of marine residence. In addition, we examined the effects of various measures of population abundance on average body length measured in the tidal delta, the timing of delta entry of 50% of the cohort, cohort residence time, and percentage of fry captured along shorelines of Skagit Bay relative to environmental factors, such as the timing and magnitude of peak flow, average temperature, and amount of available tidal wetland habitat. Results The stock–recruit models with strongest support indicated that density of tidal delta subyearlings was strongly limited by out-migrants in the river (river fry and river parr) when out-migrations were large and that preharvest adult returns were strongly limited by the density of fry in the tidal delta. Density dependence was also reflected in average body length and timing in the tidal delta as well as spillover into Skagit Bay, in combination with environmental variables. Average length of subyearlings in the tidal delta declined as river migrant fry and temperature increased, and length notably increased with increasing tidal delta wetland area influenced by habitat restoration projects. Timing of tidal delta entry declined (occurred earlier) with increasing river fry and temperature, and timing increased (occurred later) as the date of maximum river discharge during tidal delta rearing increased. However, cohort residence time, as estimated by dates of 50% emigration in-river compared to the tidal delta, was sensitive only to environmental metrics. The proportion of nearshore catch comprising fry increased as river fry and maximum daily discharge during tidal delta rearing increased. Conclusions Our study results indicate that rearing limitations in the tidal delta are important drivers of density-dependent productivity, habitat use, life history variation, and growth for Chinook Salmon, combined with annually variable environmental conditions. Comparison of stock–recruit models also strongly suggests that these density-dependent limitations have strong carryover effects that persist throughout the marine life stage. These findings provide empirical support for restoration of tidal delta habitats to alleviate density-dependent constraints on estuarine nursery populations.