Researchers enter fraught debate over dams and salmon runs
A fish is injected with a tracker for a study.
Whether born in gravel streambeds or stainless-steel hatchery tanks, young salmon face a daunting obstacle course when they journey through the Lower Snake and Columbia rivers to the Pacific. Eight dams block their path in Washington state alone.
Some dams have bypass flumes and diversion screens to spare these youngsters from the powerhouse turbines. Above four dams, river managers collect the smolt and truck or barge them hundreds of miles. The fish are set free below the Bonneville Dam, the last barricade on the way to the sea. In 1993, for example, the U.S. Army Corps of Engineers transported 15 million juvenile fish while 2.8 million stayed in the river.
While some juvenile salmon avoid the turbines, they may still face dangers in the reservoirs behind each dam. It is thought they must swim twice as hard compared to the open river. They may face predators that can pick them off more easily in the slow current.
Returning adult runs are down from last century’s 16 million to 3 million or less—proof, say many environmentalists and fishery scientists, that the system is totally out of balance. Idaho Gov. Cecil B. Andrus, for example, advocates restoring the river current to a more natural flow, what environmentalists call an “in-river” solution.
Instead of trucking and barging, they advocate drawdowns. Move more water through the dam in the spring and summer to flush salmon through the reservoirs, they argue. The less time the fish spend in the reservoirs, the more likely they will survive the journey to the sea. Environmentalists also feel the salmon are healthier fish if they stay in the river. Salmon that are barged become so disoriented that they lose the ability to return upstream to spawn, say drawdown advocates.
But using water to flush salmon down the river means less for hydroelectric power, irrigation, commercial barge traffic and even recreation, says Nola Conway, a spokeswoman for the Army Corps of Engineers. The corps says modifying their dams for drawdowns would cost from $1.3 billion to $4.9 billion.
In the middle of the debate, UW and National Marine Fisheries Service (NMSF) researchers have come up with numbers that disturb long-held conclusions about what’s needed to help speed young fish down river.
Their work tracks survival rates in the river system. Using sophisticated radio transmitters inserted into the young fish, they found that most young salmon survive their time in one Snake River reservoir, a discovery that flies in the face of previous studies. Under these circumstances, the numbers have not only been unwanted, but there have been efforts to halt the research altogether.
At the center of the conflict is a two-year-old research effort by John Skalski, director of the UW’s Center for Quantitative Sciences, and the NMSF’s John Williams and Bob Iwamoto. They have been trying to assess the number of young fish that might survive the reservoirs and dams along the Lower Snake River. The Army Corps of Engineers has equipped four dams along that part of the river with automated fish counters. The study was funded by the corps and the Bonneville Power Administration.
“What we’ve gotten in two years in that area is more data than has been gathered on the whole river system during the past decade,” says Skalski.
Among the study areas is Lower Granite reservoir and dam. Located west of Pullman, the dam is the first obstacle young salmon encounter when they leave Idaho. The reservoir and dam have long been considered real trouble spots for the fish: Some estimates say that 30 percent of the fish don’t make it any farther. About half those deaths—15 percent—are supposed to occur in the reservoir.
At Lower Granite, a drawdown could lower the water 30 to 50 feet during the spring and summer when young fish migrate. After the drawdown, river managers might be able to manipulate the river system to raise the water level behind the dam. But the reservoir could remain at a low level until winter rains and snows refill it. To be effective, drawdowns could be needed at four, five or more dams in a row.
Those who advocate drawdowns feel it would be a wise use of water: After all, there is considerable room for improvement at Lower Granite reservoir, for example, if 10 to 15 percent of the fish die there.
But UW/NMFS results say the neighborhood isn’t as bad as everyone pictured. It could be that hardly any of the fish die in the reservoir.
Skalski and his colleagues found that between 10 and 15 percent of the fish died or disappeared as they traveled through the entire reservoir and the dam. But two previous studies measured a rate of 15 percent. Assuming those numbers are correct, then virtually none of the fish die in the reservoir.
A corkscrew flume diverts juvenile salmon away from the Little Goose Dam on the Lower Snake River.
“This is much higher survival than anyone thought,” Skalski says. “If it bears up under further scrutiny, it raises questions of how much better survival will be with drawdowns.”
These numbers depend on high technology for their accuracy. “PIT” tags—part radio transmitter, part computer and part batter—are inserted into the body cavities of a selected number of wild and hatchery fish. The signal from each PIT tag is distinctive to the fish bearing it. An individual fish that is detected at all four counters on the Lower Snake is recognized as one fish and not counted as four different survivors of the journey. Counters can detect fish bearing PIT tags even in water moving at 12 feet a second. No previous method of measuring fish passage has ever been as precise.
With multiple detectors, researchers can capture data at more points. If a survivor of the Lower Granite journey is missed by the counter at that dam, it might be detected at one or more of the next dams.
“There is no other way to get this in-river survival information,” Skalski says. The researchers feel these numbers better reflect actual survival and not the “minimum survival” estimates that have been generated in the past, NMSF’s Williams says.
Some previous studies, for example, counted tagged fish at three or more places in the river, starting at Lower Granite, then added up the total and called it the minimum survival. Those researchers didn’t have a way to consider how many fish might go undetected. If they weren’t spotted and counted, they were tallied as dead.
The previous studies also failed to consider the number of fish that survived the pool at Lower Granite but subsequently died between the dam and the count’s downstream finish line. Their conclusion that only 70 percent survive the Lower Granite reservoir and dam represents the very lowest estimate of survival. “We’re saying the actual survival is much higher than that,” Williams said.
The UW/NMFS data has caused divisions in the fishery research community. Jim Anderson, a UW School of Fisheries colleague of Skalski’s, is ready to incorporate the results into a project that uses computers to track the myriad of data about salmon migration in the Snake and Columbia. But programmers operating other computer models—notably the one used by the fisheries agencies of three states and by tribal fisheries managers—don’t plan to do so. “They are choosing to ignore new data. If they continue to do so, they run the risk of losing credibility,” Anderson says.
Bert Bowler, the Columbia River policy coordinator with the Idaho Fish and Game Department, says the new results don’t necessarily diminish the value of drawdowns because, among other things, no one has actually done a drawdown as it should be done.
Bowler says any dam, such as the Lower Granite, must be modified so young fish can spill through without facing the turbines during the drawdown. If water levels remain low, there need to be other changes so that fish returning as adults can get past the dam.
Bowler might be interested in using some of the techniques devised by Skalski and the NMFS researchers if changes in the dam structures were in place. For now, he says, these researchers are causing harmful delays.
“We need to make changes, drastic ones, to have an impact on declining runs,” he says. “Some of these scientists are into micromanagement. They want all these studies competed, reach by reach, before we do anything. It would be nice, but we can’t afford to wait.”
Jim Karr, director of the UW’s Institute for Environmental Studies, happens to believe sophisticated analysis of various parts of our rivers is crucial. Nevertheless, he says, he can understand the frustration.
“Scientists always want more money for studies,” Karr says. “Politicians want more studies to delay having to make decisions. People who’ve made up their minds about the problem, and feel they know the solution, don’t want to wait until tomorrow. They want it done now.”
Those favoring drawdowns believe enough study has been done and point in particular to a 1981 report prepared for the federal government by Carl Sims and Frank Ossiander. The paper is based on data collected in the 1970s that showed there is a statistical link between both flow and travel time and flow and survival for young salmon from the Upper Snake River.
But the Sims-Ossiander study is based on sampling done between 1973 and 1979 by different groups of people collecting information in different manners. These differences were not taken into account when interpreting the data, critics say.
Anderson has a 20-member UW team that works on the CRiSP (Columbia River Salmon Passage) model of what’s happening to fish in the river system. They include data from Sims and Ossiander even though there are problems that must be taken into account. For one thing, conditions were dismal all along the river system for two years in that decade, Anderson says. One of those years, the fish also had to contend with a lot of trash in front of one of the dams. These conditions probably caused more fish than usual to die, but Sims and Ossiander related deaths to flow only.
Anderson says managers also “run” the river differently than in the ’70s. Survival rates might have been higher—even in low-flow situation—if today’s bypass systems and methods of handling fish had been in place. In addition, sampling methods also have improved since the ’70s.
But the latest findings are still tentative. Williams expressed dismay that the UW/NMFS research was hailed by some as “the nail in the coffin of drawdowns.”
He says the results, which were still preliminary this fall, are based on the best available information and assumptions, but those can change. The work, for example, relies on turbine mortality information from two previous studies of Lower Granite Dam. If those numbers should be wrong, then there could be more deaths in the reservoir, he says.
“There are reasons to use this information with care,” he says.
There are also reasons to approach drawdowns with care, Skalski adds. As Idaho Fish and Game’s Bowler said, drawdowns have yet to be tested.
Then, too, using drawdowns at more than one dam—and possibly doing drawdowns while spilling water—will require retrofitting dams. “Even if drawdowns are a good concept, they could be poorly applied until dams are modified correctly,” Skalski says.
“Whatever approaches are tried, if we don’t measure it, we won’t know what works and what doesn’t,” Skalski says. He admits he is growing cynical. For the third year in a row, there are some who are trying to halt the UW/NMFS study, even though no other project comes close to giving such detailed information of smolt survival.
The researchers can only speculate on how—or even if—their results will be used.
“All we can do is provide the data,” Williams says. “If they ask for our recommendations on how to use it, we’ll tell them. Even then, they don’t have to do what we suggest.”