Comment Judging Cost-Effectiveness of Management of Snake River Salmon: Response to Halsing and Moore PAUL H. WILSON, ∗ § HOWARD A. SCHALLER, ∗ CHARLES E. PETROSKY,† AND JOHN LOOMIS‡ ∗ Columbia River Fisheries Program Office, U.S. Fish and Wildlife Service, 1211 SE Cardinal Court, Suite 100, Vancouver, WA 98683, U.S.A. †Idaho Department of Fish and Game, 600 South Walnut Avenue, P.O. Box 25, Boise, ID 83707, U.S.A. ‡Department of Agricultural and Resource Economics, Colorado State University, Fort Collins, CO 80523, U.S.A. Introduction Halsing and Moore (2008) used Snake River spring/ summer Chinook salmon (Oncorhynchus tshawytscha) as an example to present a synthesis of biological and economic information to develop a cost-effectiveness tool for assessing management alternatives for threat- ened or endangered species. Although we believe that elements of their approach could be useful to prior- itize management alternatives and illuminate trade-offs between biological benefits and economic costs, we fear that their analysis may be of limited utility for Snake River anadromous salmonid management. Halsing and Moore used outdated, inferior models and parameter estimates to simulate the biological responses they used to rank cost-effectiveness of management alternatives, which de- pended on small differences between estimated popu- lation growth rates. They relied on a precision in esti- mated population growth rate unwarranted by the data and applied inconsistent economic analysis assumptions across scenarios, casting doubt on their cost-effectiveness findings. Biological and Management Modeling The passage model and range of differential transporta- tion mortality (D) estimates Halsing and Moore (2008) used in the analysis may mislead decision makers about the relative biological benefits of smolt transportation ver- sus in-river management options. They describe the pas- sage model they used to estimate the effects of in-river measures on second-year survival rates Columbia River Salmon Passage (CRiSP) as the “most data-driven” model for this research. In fact, it has been criticized for being §email paul_h_wilson@fws.gov Paper submitted July 23, 2008; revised manuscript accepted October 9, 2008. too complex and overparameterized, with too many the- oretical assumptions that lack empirical evidence, by 2 independent scientific review panels (Peters et al. 1998; ISAB [Independent Scientific Advisory Board] 2006). Nei- ther National Oceanic and Atmospheric Administration (NOAA) Fisheries (Zabel et al. 2008) nor U.S. Fish and Wildlife Service (Schaller et al. 2007) uses mechanis- tic predator-based models of juvenile salmon migration such as CRiSP; instead, they prefer predictive models that relate empirical estimates of survival rate and migra- tion velocity to environmental variables. Juvenile survival through the hydrosystem as modeled by CRiSP is notori- ously insensitive to river flow and spill at dams, compared with less complex, more empirical models (Peters et al. 1998). Newer, more rigorous estimates of differential trans- portation mortality (D) do not support the higher values of the range modeled by Halsing and Moore. Schaller et al. (2007), unlike Williams et al. (2005), account for interan- nual variation in sample size and estimate the mean of D over a 10-year period for wild Snake River spring/summer Chinook salmon (SRSSC) to range from 0.30 to 0.48, de- pending on point of transport. The mean D value for all fish, weighted by the proportions transported at each project, is 0.39. This suggests that little credibility should be given to the higher D values (0.7 and 1.0) modeled by Halsing and Moore. Overestimation of D leads to overes- timation of the biological and cost- effectiveness of tern removal and underestimation of the benefits of strategies that minimize or discontinue transportation. The degree to which and the mechanisms by which the hydrosystem causes latent mortality of juveniles in the estuary and early-ocean life stage (s e ) have been high- lighted as critical uncertainties (Kareiva et al. 2000; Wil- son 2003). Latent mortality may be explained, in part, 475 Conservation Biology, Volume 23, No. 2, 475–478 Journal compilation C  2009 Society for Conservation Biology. No claim to original US government works. DOI: 10.1111/j.1523-1739.2009.01170.x