the film
Commentaries and editorials

Analysis: Deja Vu,
All Over Again, And Again

by Bill Rudolph
NW Fishletter, January 31, 2014

I was cleaning out an old file cabinet last week and came across a treasure trove of old articles, studies, speeches, legal briefs, hate mail, and other assorted information hiding there since the mid-1990s, when I first starting covering the fish beat. After a morning spent in a random read through all that verbiage, I was struck by how many of those old fish fights are still going on, though some may be on hold for a few more years, until all that extra BPA funding runs out.

In those dark days of nearly total ignorance on my part of the workings of the Columbia River fish recovery industry, I wrote several feature stories about the situation on the Columbia with ESA-listed Chinook. At the time, I was assistant editor of the Alaska Fisherman's Journal, and some of my readers caught some of those fish (a lot more than their own, in fact). I had had a previous life where I had preyed on them, myself.

My first story was still lurking in that old file. It began in a bar in Skamokawa, across the river from Astoria, when one rainy afternoon, I interviewed a NMFS technician who followed sea lions and seals around in the estuary, which introduced me to a whole new universe.

Soon I was shuttling between the warring camps of fish modelers, trying to get some straight answers. It was the FLUSH (Fish Leaving Under Several Hypotheses) modelers from states and tribes versus the UW's BPA-funded CRiSP (Columbia River Salmon Passage) for supremacy of the hearts and minds of regional researchers. And reporters.

The two camps debated the values of flow, spill and barging fish, without the benefit of much survival data at all. In the end, both estimated inriver survivals that were too low, but FLUSH results were much lower--around 20 percent. (Today's inriver survivals are double that and more.) The CRiSP model was calibrated with meager, but brand new PIT-tag data, and it estimated juvenile fish survival about twice the FLUSH results. Since then, CRiSP has been retooled by NMFS and is now called COMPASS--still an important part of the agency's BiOp survival modeling effort.

CRiSP modelers said FLUSH results were affected by years like 1973 when trash buildup at Lower Granite Dam was so bad that descaling and later mortality of juvenile salmon was estimated at 50 percent at that project alone. Since the mortality was not related to the long travel times that the FLUSH model identified as the basis for mortality, the CRiSP modelers said it had skewed the parameter incorrectly. Furthermore, they said the FLUSH analysis reflected hydro conditions that no longer existed.

That's the same criticism used by federal scientists and others to knock the analysis by today's state and tribal modelers who have been proselytizing for another boost in spill at dams. About twice as many fish are already passing the dams via spillways and surface weirs than the spill boosters claim, according to the most recent Corps' data.

As more PIT-tag detectors were built at downstream dams and more data was collected, the FLUSH model was consigned to the dustbin of history, and the high level of mortality that FLUSH modelers said originally was supposed to occur in the pool behind Lower Granite Dam, was not found there, or in fact, anywhere in the hydro system. It's now hypothesized to occur somewhere north of Vancouver Island.

The region is lucky the PIT-tag came along when it did, and NMFS scientists saw it as a tool to finally get some answers to dam survival questions. But early on, some other stakeholders actually tried to keep the PIT-tag research from happening. ODFW led the charge, with then-assistant fisheries chief Doug DeHart explaining to reporters that the new-fangled research jeopardized the survival of the very fish they were supposed to be saving.

Witness one dog-eared piece of evidence from my old files, an old, slippery and faded fax sent to me by the Columbia River Alliance, back in May of 1994. It was a copy of a Portland Daily Journal of Commerce story that reported ODFW, along with Idaho, Washington and the lower Columbia tribes, had asked NMFS to deny its own researchers a permit to continue a two-year study involving PIT-tagged fish.

Oregon had claimed the fish would suffer increased mortality two ways--first, from the lowered spill levels necessary to conduct the study (it called for eliminating spill at three lower Snake dams), and secondly, the state claimed that many fish would die from the sampling and tagging process. I actually found that 1994 letter from ODFW to NMFS in my files, and it mentioned another bugaboo--the much vaunted issue of "long-term delayed and indirect mortality that would occur after release of smolts," which still remains an unproved hypothesis from states and tribes to this very day.

Only about 21,000 upriver spring Chinook showed up in 1994, about half the number expected, with 2,400 wild ones heading for the Snake. Only 600 jacks showed up, decimated by the El NiƱo from the year before. True to form, the upriver return in 1995 didn't even hit 10,000 Chinook.

This year 227,000 are expected. What's different? The ocean changed for the better, and right on cue. Just before the 2000 BiOp was released, jack counts jumped from about 1,200 in 1998 to 9,800 in 1999, signaling better days ahead. The upriver spring Chinook return shot from 38,000 in 1999 to 179,000 in 2000, and more than 416,000 in 2001.

Before the returns improved, few people involved in the Basin's salmon recovery effort were already aware of the huge role ocean conditions played in the opera. In 1996, the UW's Jim Anderson, chief architect of the CRiSP model, wrote a letter to the Army Corps of Engineers defending his model from attacks by FLUSH modelers. He said that critics attacked CRiSP because its results were different from theirs. "They seem to conclude that since smolt survival and adult survival indices are different, that CRiSP is wrong. They ignore decadal climatic factors that have a major influence on survival. Many in the fisheries and oceanographic communities understand the significance of the decadal scales variation. The Salmon Managers seem to disregard ocean and climatic processes."

Anderson had even hired an oceanographer to study processes that were occurring outside the Basin. Curt Ebbesmeyer, a private consultant, who has since become famous for his work with ocean currents and yellow duckies spilled from container ships, and later tracking drift wreckage from the Fukushima disaster, developed that work and came up with something he called the PNI, or Pacific Northwest Index, that correlated such factors as rainfall, snow depth at Mt. Rainier, the size of Willapa Bay oysters, and other temperature and precip data that led him to conclude that wetter, colder regimes seemed to persist for several decades at a time, interspersed with warmer, drier conditions.

Ebbesmeyer also noticed the transition from winter to spring off the mouth of the Columbia River occurred a month sooner than it did around 1900. If fish emerge from the river before the spring transition has occurred, he expected juvenile mortality to be increased. This could have major implications for hatchery releases and fish barging schedules. Fish barged downstream in the spring reach the estuary three weeks ahead of their river riding counterparts, he told a group of scientists at a conference in Nanaimo in the fall of 1996 that I attended. (You can read about it in NW Fishletter 20, one of the first I produced as editor.)

Guess what? Since then, NMFS has found, from years of PIT-tag return data, that early barged salmon and steelhead don't do very well, and have adjusted their barging strategy accordingly.

Another paper presented at that Nanaimo meeting described a multi-decadal cycle that its principal authors dubbed the Pacific Decadal Oscillation. When it appeared in print a year later, it was still not obvious the PDO was about to shift--"the last two reversals correspond with dramatic shifts in salmon production regimes in the North Pacific Ocean. This climate pattern also affects coastal sea and continental surface air temperatures, as well as streamflow in major west coast river systems, from Alaska to California."

And then, all of a sudden, it did, and return rates were suddenly four to ten times better, and the years-long arguments over how much water to spill over dams seemed rather beside the point. And it still does today, with the recent clamor for even more spill that will somehow, magically, return more fish.

I came across several fat folders that held the detritus from the mid-90's debates over spill, when NMFS was developing total dissolved gas standards for dam operations. For reasons of personal sanity, I had pretty much blocked out this period of my career.

One of those documents was the testimony of Canadian gas bubble expert, Larry Fidler, before a Senate subcommittee in June 1995 that was looking at NMFS' spill policy. Fidler, who had written a book on gas bubble disease and fish in 1988, had been part of an independent panel of experts convened by NMFS to help develop their policy, but the agency ended up using few of its recommendations.

In his remarks, Fidler mentioned a 1995 incident where most (85 percent) of the test fish held in 4-meter pens below Ice Harbor Dam died after being exposed to TDG levels of 128 percent. (That's only 3 percent higher than what CSS spill proponents are calling for today.) He said this showed that the NMFS' contention that fish could dive to avoid high gas levels was wrong (gas levels decline about 10 percent per meter of depth).

But that wasn't all. Fidler said the monitoring program designed in 1995 to check smolts for gas bubbles was flawed. He called it an "unfocused collection of attempts by various agencies to sample fish without an understanding of the appropriate locations and conditions under which to assess the impacts of DGS [Dissolved Gas Saturation] on the Columbia and Snake River juvenile and adult salmonids."

Fidler also noted that hydrostatic pressures in the bypass systems where fish routed before sampling could reduce or eliminate signs of GBT in fish before they were monitored. Without the expert panel's input, he said the monitoring effort was so poorly designed and implemented, the chance to understand effects of dissolved gas on fish in the Columbia and Snake Rivers had been lost.

It's the very same monitoring program run by the Fish Passage Center today, which typically finds little if any evidence of gas bubble disease in smolts, even when TDG levels are above legal limits (with a waiver, currently 120 percent in tailraces).

Fidler also used his soapbox to criticize NMFS for not letting the expert panel review a risk assessment of spill by states and tribes that concluded spilling up to 125 percent TDG produced greater survival than would occur if fish passed through turbines. He said the document had numerous flaws and underestimated the effects of dissolved gas on both adult and juvenile salmonids.

But we'll never know just what those experts found wrong, since the panel became another victim of the politicized science that ran the river in the good old days, though NMFS, a memo to Congress and Oregon water quality scientists also found the assessment "seriously flawed."

Columbia Basin politics and science still work in tandem--witness the traction gained by the latest spill proposal from one state and one tribe in recent months. It should never have got this far. Like my old headline from 1995 said, "The big spill--we may be killing the run in order to save it."

Bill Rudolph
Analysis: Deja Vu, All Over Again, And Again
NW Fishletter, January 31, 2014

See what you can learn

learn more on topics covered in the film
see the video
read the script
learn the songs
discussion forum
salmon animation