Trout Unlimited: Independent Analysis Shows The Federal System For Assessing Extinction Risks Of Snake River Salmon Is Seriously Flawed
2/3/2000 -- -- The federal government's assessment of the extinction risks of Snake River salmon and management options for recovery of the species is flawed because it contains seven critical errors, according to officials with Trout Unlimited, the national conservation organization.
Trout Unlimited officials said the errors are contained within the National Marine Fisheries Service's Cumulative Risk Initiative (CRI), the federal method for assessing extinction risks and management options for the Snake River salmon. They were discovered as a result of an independent scientific analysis that Trout Unlimited commissioned.
"These errors allow the federal government to overestimate the time we have to save the Snake River salmon and underestimate the impacts of dams on their survival. The study is flawed and if it is not corrected, it will result in the federal government making flawed decisions concerning what is the best course to take to prevent the wild Snake River salmon from becoming extinct," Jeff Curtis told members of the Federal Caucus gathered in Portland at the first of a dozen regional public hearings on federal efforts to save the salmon.
Curtis, the Western Conservation Director of Trout Unlimited, said Dr. Gretchen Oosterhout, a former reliability engineer with Hewlett Packard, who is now a system analyst and principal of Decision Matrix, Inc, did the analysis of the CRI. A well-respected engineer and analyst, Oosterhout has also recently done work for the Multi-Species Framework Process, the World Forestry Center and Portland General Electric.
The seven errors range from using unrealistically low salmon numbers - one fish or fewer - to determine if a species is functionally extinct, to using outdated timeframes to measure existing salmon populations, to the use of highly questionable methods in determining pre- and post-hydro system mortalities of migrating Snake River salmon.
"I don't know if these errors were intentional, but I do know they allow the federal government to pretend we have longer to save the salmon than we do and to point to sources other than the dams for the high mortality of young Snake River salmon. Whether you support dam removal or not, it is in everyone's best interest to make sure that any decision the federal government makes is based on good science. Unfortunately, some of the information the National Marine Fisheries Service is currently using to make decisions concerning the fate of the Snake River salmon is not good science," Curtis said.
The seven critical errors in the CRI include:
- The CRI sets the "extinction bar" far too low.
The CRI considers a stock of salmon functionally "extinct" when there is one fish or fewer returning to spawn. Other studies - including NMFS' own - have used numbers ranging from 15 to 300 fish for extinction definitions. Obviously, the time remaining until extinction - and thus for federal delay - is far greater if the definition of "extinct" is just one fish. In terms of sustainable populations, however, one fish just won't do.
- The CRI ignores the fact that the slide toward extinction accelerates as salmon populations decrease.
Risk analysis is about trends, not averages. It stands to reason that if you have 100 spawning fish returning one year and 90 the next followed by 60 and 25 after that, you're heading toward extinction at a compounding rate of speed. The CRI assumes an average rate of decline.
- The highly questionable CRI models kill off four times as many fish before they hit the hydro system than the best available data suggest.
We have linked these two because their policy implications are connected. The CRI models - by utilizing questionable analysis methods and disputed data - have shifted the mortality for spring/summer chinook from the time they enter the ocean after passing through the hydro system to the first year of life - prior to reaching the hydro system. Therefore, this shift lets the hydro system and the various "techo-fixes" like barging and trucking juvenile fish around the dams off the hook by assuming - contrary to the available empirical data - that four times as many fish are dying before they're subjected to the stresses of the hydro system than after. This inexplicable flip-flop from pre- to post-dam mortality takes the blame off of hydro and puts it onto habitat.
- The CRI ignores post-1990 data, even though such data are readily available.
The CRI calculates its extinction timeframes using numbers up to - but not after - 1990, even though more recent data are available. As pointed out in #2, trends leading to extinction accelerate as populations decrease. By eliminating more recent and lower return numbers, the result is a slower timeframe to extinction, again allowing more time for delay at the expense of the salmon.
- The CRI did not subject itself to peer review.
The CRI's predecessor, PATH, included state, tribal and U.S. Fish & Wildlife scientists in its process, resulting in higher-quality, more credible and peer-reviewed work product. The CRI remains an in-house project.
- The CRI does not use textbook risk assessment methods.
The CRI is supposed to be a risk initiative. Yet, it uses almost no standard risk assessment tools. Standard risk assessment consists of analyzing trends, sources of error, and uncertainties; the CRI analyzes none of these.
Trout Unlimited is North America's leading coldwater conservation organization, dedicated to the conservation protection and restoration of trout and salmon fisheries and their watersheds. The organization has more than 110,000 members in 472 Chapters in North America, including 8,000 members in the Pacific Northwest. The full Trout Unlimited-sponsored analysis of the CRI can be accessed at www.tu.org.