High Priests Debate NMFS Extinction "Science"

In the Middle Ages, the Church hosted immense gatherings to debate arcane points of theology with no real world application, other than the often-bloody political fallout arising from theological disputes.  On Wednesday, March 29, 2000, the high priests of our state religion, environmentalism, gathered in Seattle for a workshop co-sponsored by the environmentalist group American Rivers and the Northwest Marine Fisheries Service's Northwest Science Center.  The focus of the discussions concerned highly-complex mathematical analyses of extinction risk for Northwest salmon populations produced by the National Marine Fisheries Service's "Cumulative Risk Initiative".  Conservation biologists and others argued about the appropriate means of estimating extinction risk, with NMFS getting pressure to further exaggerate extinction risks.

American Rivers, of course, is actively promoting removal of the Snake River Dams, using a propaganda campaign of lies and distortions.  Journalist Cyrus Noë recently pointed out that "it is poor judgment or worse for NMFS to cooperate officially with an organization that engages in fundraising by using deceptive and blatantly false salmon recovery information".  (Clearing Up, Mar. 20, 2000, at 6)  American Rivers turned to University of Washington law professor William H. Rodgers, Jr. to "moderate" the workshop.  Professor Rodgers, another high priest, is fond of such notions as the idea that traditional notions of private property are undermined by the new science of ecology and that Pacific Northwest tribes are legally entitled to 1840s levels of salmon in the rivers.

Like the Medieval convocations, the participants debated issues in the abstract, untethered by actual scientific data.  Both NMFS and its critics begin with the products of the infamous PATH process, a collection of guesses about historic salmon run sizes that is the product of at least seven layers of undocumented assumptions.  NMFS candidly acknowledges that this information, which it optimistically characterizes as "data", is likely to contain "extremely large" errors.  

On this foundation of sand, both NMFS and its critics run the "data" through immensely complicated mathematical models to produce estimates of the possibility that individual stocks will go extinct.  These estimates then become fodder for the American Rivers propaganda gristmill.  The models, like Medieval theology, exist in the abstract, untethered to provable reality.  

One could, for example, run the models on historical data to predict the risk of extinction, and then see if populations analyzed actually did go extinct.  No one did.  One could run the models on populations that are not listed under the Endangered Species Act, to compare the extinction risks to listed stocks.  No one did.  (There are a few scientists out there who are trying to match population theories to real experimental data, such as one team's isolated flour beetle populations.) 

Neither NMFS nor its critics deign to produce the information made legally relevant by the NMFS' Endangered Species Act listings:  the probability that a listed group will go extinct.  NMFS suggests that it cannot do so because it does not have enough data.  

But if you do model listed salmon groups (collections of salmon populations called "Evolutionarily Significant Units" or ESUs) as they actually exist, constantly straying into one another and re-colonizing areas where local populations depart, with the data we do have, you find that extinction risk is very low indeed.  The only analysis of which I am aware, conducted by John Emlen in 1994, found virtually no chance that Snake River spring/chinook salmon would go extinct in one hundred years.  Obviously, when a listed group of fish is made up of a number of runs, possibility that the listed group goes extinct is at least as small as the possibility that the healthiest single stock goes extinct, if not smaller. 

 Thus, for example, NMFS predicts that the Johnson Creek stock, part of the "Snake River spring/summer chinook ESU", has only a 3% chance of going extinct in 100 years.  So long as there are Johnson Creek fish, there are Snake River spring/summer chinook, even if many other runs are gone.  Probably dozens of unlisted critters have a 3% chance of going extinct in 100 years.  Maybe humans do.  

Perhaps the most serious flaw in NMFS' analyses is that they are based on data from 1980 to 1995, a period of time that may have contained the worst ocean conditions for salmon in 500 years.  Plugging even real data from this time period into a computer model predicting 100 years in the future assumes, in effect, that the terrible ocean conditions we have had will persist for 100 years in the future.  That seems unlikely.  And by cutting the data off in 1995, recent upturns in salmon runs are ignored, again producing an unreasonably pessimistic picture.  Ultimately, the sophisticated models do little more than dress up whatever trend is in the data you choose to analyze.   

I don't mean to be tough on the NMFS scientists, as they are the closest thing to real scientists working in government extinction science.  They've still got the courage to point out that their models predict that Lower Snake dam removal will not recover Snake River salmon, even when they make every available assumption to bolster the benefits of dam removal.  Indeed, the NMFS scientists suggest that only life stages of salmon where reductions in mortality will have much effect on salmon population growth rates are in the first year of life, before the salmon come anywhere near the dams, and in the estuary, where the birds eat them.

To corroborate their conclusions, they cite a recent study of 22 different runs of spring/summer chinook salmon that must pass from one to nine dams.  The study failed to demonstrate any statistically significant relationship between the number of dams passed and the population sizes.  But don't look for American Rivers to broadcast that result.  Instead, the orthodox view is that the extinction risks are so high we must take out the dams whether or not it will recover the salmon, just to "mitigate risk".

The saddest thing about all of this is that it would not be terribly difficult to produce sensible estimates of extinction risk, an obvious first step toward getting most of the listed salmon "species" off the Endangered Species List entirely.  But no one in the Clinton/Gore Administration or the environmental movement is ever going to spend a dime to do that.  It's up to the citizens of the Pacific Northwest and their representatives, elected and otherwise, to push the issue.  

© James Buchal, March 30, 2000

You have permission to reprint this article, and are encouraged to do so. The sooner people figure out what's going on, the quicker we'll have fish in the rivers.