Review of : Using Model Selection and Model Averaging to Predict the Response of Chinook Salmon to Dam Removal and Forecasting the Response of Klamath Basin Chinook Populations to Dam Removal and Restoration of Anadromy Versus No Action

The Klamath Basin Restoration Agreement (KBRA) calls for the Secretary of the Interior to make a finding by 12 March 2012 whether four dams on the Klamath River should be removed. The two studies under review are intended to inform this decision. In particular, they addess the question whether and by how much removing the dams will increase the production of Klamath Basin Chinook salmon over the coming fifty years; the answer, reached by two different modeling approches, is “Probably, but it is hard to say how much.”Both models give highly uncertain estimates of future production of Klamath Basin Chinook, but that should be expected, and one of the strengths of these models is that they deal explicitly with uncertainty. Both models ignore a good deal of detailed information about the basin, which may make the models seem simplistic and unrealistic, and will upset the people who have developed the information. Both studies fail to take account of two important factors: global warming, and interactions between hatchery and naturally produced salmon. Models are tools of science, not science itself. The two models reviewed here incorporate the best available modeling approaches, but, in my assessment, to meet the standard of “best available science,” the models need to deal with the effects of climate change and interactions with hatchery fish.