Patterns of density dependence affect coho salmon population response to restoration

Streams supporting low abundance salmon populations are often targeted for restoration, yet evidence for population-scale increases in fish abundance following stream habitat treatments are rare. In order to characterize the influence of density dependence on the fish responses to restoration, we fit a series of Ricker stock-recruit models to coho salmon life cycle monitoring data from seven streams within two Intensively Monitored Watersheds (IMWs). We found strong evidence for density dependence overall, but the strength of density dependence varied considerably among locations and across life stages. In particular, we observed a strong contrast in patterns of density dependence in the Lower Columbia IMW (WA) compared to the Hood Canal IMW (WA). We observed consistently strong density dependence in the Lower Columbia IMW, and an increase in abundance following restoration in Abernathy Creek appeared to result from release from density dependent constraints on productivity at the parr to smolt stage. By contrast, we did not detect a fish response to restoration in either Big Beef or Little Anderson creeks in the Hood Canal IMW, both characterized by weak density dependence. Furthermore, a modelling exercise indicated large spawner abundances probing habitat capacity limitations increased the likelihood of detecting a fish response to stream restoration, compared to smaller spawner abundances. Taken together, these results suggest that management strategies that test juvenile capacity limits by saturating the spawning grounds with adults give the greatest opportunity to observe the increases in salmon population abundance that motivate stream restoration efforts in the Pacific Northwest.