Forecasting the Response of Klamath Basin Chinook Populations to Dam Removal and Restoration of Anadromy Versus No Action

Two alternative actions are being evaluated in the Klamath Basin: 1) a No Action Alternative (NAA) and 2) removal of four mainstem dams (Iron Gate, Copco I, Copco II, and J.C. Boyle) and initiation of habitat restoration in the Klamath Basin under a Dam Removal Alternative (DRA). The decision process regarding which action to implement requires annual forecasts of abundance with uncertainty under each of the two alternatives from 2012 to 2061. I forecasted escapement for both alternatives by constructing a life-cycle model (Evaluation of Dam Removal and Restoration of Anadromy, EDRRA) composed of: 1) a stock recruitment relationship between spawners and age 3 in the ocean, which is when they are vulnerable to the fishery, and 2) a fishery model that calculates harvest, maturation, and escapement. To develop stage 1 of the model under NAA, I estimated the historical stock recruitment relationship in the Klamath River below Iron Gate Dam in a Bayesian framework. To develop stage 1 of the model under DRA, I used the predictive spawner recruitment relationships in Liermann et al. (2010) to forecast recruitment to age 3 from tributaries to Upper Klamath Lake, which is the site of active reintroduction of anadromy. I also modified the spawner recruit relationship under DRA to include additional spawning capacity between Iron Gate Dam and Keno Dam. In order to facilitate the comparison of the two alternatives, I used paired Monte Carlo simulations to forecast the levels of escapement and harvest under NAA and DRA. Median escapements and harvest were higher in DRA relative to NAA with a high degree of overlap in 95% confidence intervals due to uncertainty in stock-recruitment dynamics. Still, there was a 0.75 probability of higher annual escapement and a 0.7 probability of higher annual harvest by performing DRA relative to NAA, despite uncertainty in the abundance forecasts.