[1] Scientists Tell FPC To Dump Upriver/Downriver Comparisons
An independent review of the latest PIT-tag survival study from the Fish Passage Center has garnered some positive comments, though scientists who looked at it have recommended gutting one of its major elements--the upriver/downriver stock comparison--because of the many confounding factors in the analysis.
The joint report was released last week by the Independent Scientific Advisory Board, which looks at general questions involved in salmon recovery science, and the Independent Scientific Review Panel, which judges the scientific merit of proposals in BPA's F&W program.
The report was conducted in response to questions from the Northwest Power and Conservation Council. The council wanted an updated review before they OK any future funding for the 10-year-old FPC project, called the Comparative Survival Study (CSS).
Though the review authors said the CSS represents "a successful implementation of a large-scale tagging program," they took issue with the study's complicated analysis that related fish travel times, survival and instantaneous mortality rates to environmental variables like water travel time, spill, and Julian day.
"It would appear that that a concerted effort at developing more parsimonious models is necessary," they said, noting that the CSS report came up short on explaining assumptions and uncertainties.
The reviewers echoed earlier criticism from both NOAA Fisheries and BPA that called the comparison between Snake River and John Day stocks bogus, but they said it in a much nicer way.
"The observation that SARs [smolt-to adult returns] for wild chinook were only one-quarter that of similar downriver populations that migrated through fewer dams does not provide proof that the hydrosystem is responsible for this SAR difference; there are myriad alternative possible explanations," said the reviewers. "The comparisons are weak because of the limited number of sites compared and the multitude of confounding variables."
The reviewers also mentioned their own report from earlier this year that advised against trying to measure absolute latent mortality, which is a large part of the FPC's analysis of upriver/downriver productivity. The ISAB's April report said the focus should be on the total mortality of inriver migrants and transported fish.
In previous annual reviews, the independent panels had recommended that the CSS project incorporate more downriver sites in its analysis. Now they doubt if there are enough of them available to make a meaningful comparison.
The CSS project began 10 years ago, after the original ISAB gave it a provisional thumbs up in 1997. At that time, FPC sponsors planned to use downriver stocks as "controls." The science board's report said the proposal needed "substantial revision" before it could achieve enough scientific rigor for the scientists to endorse it as the basis for a long-term study.
But over the years, the study has continued, marked by steady controversy, but enough positive feedback from scientific review has kept it going. At one point, sponsors dumped two downriver hatchery stocks in the study because of such poor returns. They cited disease problems as the main culprit.
The FPC's report said that Snake River hatchery spring chinook and both wild and hatchery Snake steelhead show some benefits from being transported, but that wild chinook had no net benefit.
However, the reviewers said sample sizes for wild chinook were too small to draw any conclusions.
The ISAB did say that overall SARs for most groups of fish weren't sufficient for stock persistence, and far short of the NPCC's 2-percent minimum recovery objectives, but they noted that these goals may be revised soon and will likely be "more tailored to individual ESUs."
The reviewers did mention that these PIT-tag SARs are likely quite lower than those of the untagged populations in general, but they don't call for pulling the plug on the study.
"The assumption that tagged fish are typical of the untagged fish in their respective cohorts is difficult to test in any empirical fashion and is a bit worrisome," they said.
In earlier comments to critics, the CSS Oversight Committee supported continuation of the project. It acknowledged that Snake hatchery fish were helped more by transportation and showed relatively lower levels of differential mortality than wild stocks, but annual SARs between wild and hatchery stocks were highly correlated.
"In view of this high correlation," said the committee, "continuing the CSS time series of hatchery SARs will be important to augment wild chinook SAR information following future years of low escapements, in addition to providing valuable management information for the specific hatcheries."
But NMFS had said the same information could be gleaned from much smaller numbers of PIT-tagged fish or simply by comparing adult returns of clipped (hatchery) fish to unclipped fish. Their earlier analysis found that PIT-tagged wild fish returns were only about 60 percent of untagged fish, which made them question the CSS results claiming that Snake River wild chinook returns aren't meeting the 2-percent return-rate goals for the region.
The feds said fish size and estuary timing likely play important roles in survival. The feds also took issue with the report's conclusion that by using flows to reduce the travel time of certain stocks such as steelhead through the hydro system, mortality would go down by 5.6 percent.
The report's analysis that correlates survival with water particle travel time "is a classic example of a 'spurious correlation,'" according to the feds.
The FPC response said that NMFS had "mischaracterized the work," and that fish travel time is a function of other variables as well as water-particle travel time, which include average-percent spill and Julian day.
The ISAB/ISRP reviewers have tried to stay above the fray, noting that they "only briefly reviewed the comments from others on the CSS report in the context of our chapter reviews."
They added, "We do not address the comments point by point, and we did not re-analyze any of the CSS or commenters' specific data analyses due to the short time available for our review."
The ISRP side of the review says that the CSS project should be modified to reduce tagging efforts in the Columbia River called for in its upriver/downriver comparison because it doesn't satisfy scientific criteria.
But both groups of scientists agreed future work should investigate why PIT-tagged Snake River wild spring chinook have lower SARs than unmarked fish, and why the Snake wild spring chinook don't seem to benefit from barging.
They also recommended CSS parties submit their central results in a paper to a peer-reviewed publication.
Tony Grover, who heads the NPCC's F&W division, told NW Fishletter that his staff will recommend to Council members that all PIT-tag efforts associated with the upriver/downriver analysis not be funded in the future. However, he said if some of the tagging can be shown to be useful in other work, it should continue. -Bill Rudolph
The following links were mentioned in this story:
NW Fishletter 26, Jan. 21, 1997
Subscriptions and Feedback
Subscribe to the Fishletter notification e-mail list.
Send e-mail comments to the editor.
THE ARCHIVE :: Previous NW Fishletter issues and supporting documents.
NW Fishletter is produced by Energy NewsData.
Publisher: Cyrus Noë, Editor: Bill Rudolph
Phone: (206) 285-4848 Fax: (206) 281-8035
