the film forum library tutorial contact

The Great Salmon Compromise by Ben Goldfarb High Country News, December 8, 2014

The Columbia Basin Fish Accords have funded $1 billion worth of
habitat restoration projects, but can they replace free-flowing rivers?

On an ice-blue afternoon in mid-May, I met Kat Brigham in a mall cafe on the Oregon shoulder of the Columbia River.

Brigham, 67, has dark, intelligent eyes, creased by laughter behind black-framed glasses, and she ranks among the eminences of Columbia fisheries management. A former commercial salmon fisherman who once captained an all-female crew, she's devoted her life to the fishing rights of her tribe, the Umatilla. No phase of the salmon lifecycle has escaped her advocacy: She's helped set harvest limits out in the Pacific Ocean, where salmon spend their adult lives, and argued for a basin-wide strategy for hatcheries, where 80 percent of the region's fish are born. "Our goal has always been to put fish back," she told me, hands clasped on the table. "We don't want to study them to death."

Brigham's primary weapon is a set of treaties that her tribe and others struck in 1855 with Isaac Stevens, governor of what was then called Washington Territory. The treaties ushered Northwestern tribes onto reservations, but also, crucially, preserved their right to fish at their "usual and accustomed" sites. Though Supreme Court rulings eventually upheld those rights, Indians were nonetheless harassed, arrested, even shot at for attempting to exercise them. Brigham remembers white protesters in the 1970s holding placards that read, "Save a Salmon, Can an Indian." "I can still see the picket lines that my grandfather and I walked through" on the way to tribal meetings, she said.

Angry white fishermen weren't the tribes' only adversaries. Inside the cafe, we spoke over the low background roar of water hurtling through the Dalles Dam, a nearly 2-mile-long concrete wall whose 1,780-megawatt capacity helps power a nearby Google data center. It was the Dalles that, on March 10, 1957, encased Celilo Falls in a tomb of slackwater. For millennia, tribal fishermen plied Celilo's roiling cascades and rapids, capturing salmon with spears and dipnets from teetering wooden scaffolds; the area was a tribal trading hub, known to historians as the "Wall Street of the West." Hours after the Dalles began operation, Celilo was gone.

The Dalles and other dams gave the Northwest some of the country's cheapest power, but they also cut off 55 percent of the Columbia Basin's fish habitat, with disastrous results. In 1991, following a petition from the Shoshone-Bannock Tribe, Snake River sockeye salmon were protected under the federal Endangered Species Act; today, a baker's dozen of salmon and steelhead stocks are listed. Overfishing and habitat destruction from mining, logging and development -- which bury spawning beds in sediment, strip banks of vegetation and raise water temperatures -- have taken their toll. Still, dams are the most visible culprits, and scientific groups like the American Fisheries Society maintain that salmon recovery can't be achieved without breaching the four big ones on the lower Snake River.

That conflicts with the federal government's salmon recovery assessment, or biological opinion, which claims the fish can be saved primarily through habitat restoration. And despite the dams' evident role in salmon declines, most Northwest tribes now support the government's strategy. On our table's faux-wood surface, Brigham slapped the salt and pepper shakers together. Imagine, she said, that they are a dam. "If fish only come this far, that's not meeting our treaty obligations," she told me. She tapped the table above her condiment dam. "If we want to exercise our culture, we need fish up here" -- in the upstream tributaries where her tribe has its "usual and accustomed" fishing sites.

To restore those degraded streams, in 2008 Brigham and other tribal leaders struck a deal -- called the Columbia Basin Fish Accords -- with the Bonneville Power Administration (BPA), the New Deal-era federal agency that sells the region's hydropower. The agreement distributes nearly $1 billion over a decade to around 200 tribal and state fish projects, from thousands of acres of habitat restoration to dozens of fish hatcheries; from ladders for lamprey to rubber bullets for salmon-eating sea lions. The Yakama took the largest share, at $343 million; the Umatilla, $179 million; the Colville, $223 million; and the Warm Springs, $86 million. The Shoshone-Bannock and Kalispel signed on later, and the states of Idaho, Montana and Washington received funding as well. The Columbia River Inter-Tribal Fish Commission (CRITFC), the fish and wildlife agency for which Brigham serves as secretary, received $96 million of its own.

Signatories expected the deal not only to rejuvenate salmon and steelhead runs, but also to boost depressed tribal economies. The Warm Springs fisheries staff, department manager Brad Houslet told me, tripled to over 100 workers, nearly half of whom are tribal members. "A lot more of our people are doing the work, whereas before we had outsiders doing it," said Yakama Councilman Gerald Lewis. "And while fish runs aren't where they were in the past, they're starting to come up to numbers where our fishermen can actually go out and harvest."

There was, however, a catch: Signatories had to stop fighting the biological opinion, which the tribes had attacked in court for its failure to help fish. They also agreed not to advocate for dam breaching or increased spill -- water that's allowed to flow over dams, rather than through turbines, to help juvenile fish survive their trip downriver -- until the deal expired in 2018. "My reaction was that (the Accords) were bribes," said Michael Blumm, a professor at Lewis & Clark law school.

The deal's proponents vehemently disagree. This year, 2.3 million returning salmon and steelhead passed the Bonneville Dam, the most since counting began over 75 years ago -- evidence, they say, that the projects are working. "We wanted to prove that fish restoration and hydropower could coexist," Steve Wright, who served as BPA head when the deal was signed, told me. "And so far, that objective is being met."

With another round of negotiations pending, it's worth asking what the Accords have accomplished. Are the basin's fish truly closer to recovery? Or, by neutralizing tribal advocacy power, have the Accords enshrined the status quo, leaving us to tinker along creekbanks while the real obstacles remain? If we throw enough money at the problem, can we have our fish and our dams, too?

The Fish Accords are rooted in the West's most aggravating legal cycle. In 1992, newly compelled by the listing of sockeye, the National Oceanic and Atmospheric Administration (NOAA) prepared its first biological opinion, or BiOp, for the Columbia hydroelectric system -- an analysis that claimed the dams wouldn't jeopardize endangered fish. Idaho sued the feds the next year, and U.S. District Judge Malcolm Marsh struck down the plan. And so began the carousel: Dam operators prepare a BiOp; some combination of states and conservation groups file suit, with the tribes joining as friends of the court; a judge shoots down the BiOp; NOAA returns a few years later with a tweaked plan. Lather, rinse, repeat.

In 2005, Judge James Redden, Marsh's sharp-witted successor, finally tired of the routine. He ordered the federal government to rewrite its plan again, but this time with input from states and tribes. The new BiOp, he wrote, "must not be a secret process with a disastrous surprise ending."

Soon after, Brigham received an olive branch from Steve Wright. At first, she was suspicious: She had, after all, spent much of her life fighting the federal government, and the tribes had a compelling case that the dams were violating their hard-won rights. In 1999, Don Sampson, then head of the Inter-Tribal Fish Commission, told High Country News that the tribes were working on a "war plan," and that damages could run into the billions. "If the fish aren't there," he said, "then the treaty is broken."

The more Brigham and Wright talked, though, the more a ceasefire made sense. "Some of our non-Indian friends say, 'You always win in court!' " Brigham told me. But she insisted that wasn't true: Even the famous Boldt decision, the 1974 Supreme Court ruling that guaranteed the tribes half the harvest, had deprived them of some ceremonial fish. The litigants were technically winning the BiOp battle, but it didn't feel that way.

Returning fish to degraded streams would require extensive -- and expensive -- habitat restoration. BPA was already paying for plenty of that, as obligated by the Endangered Species Act, the tribal treaties, and the Northwest Power Act, a 1980 law that compelled the agency to mitigate the dams' damages to fish. But shifting budgets and bureaucracy constantly threatened funding, hampering more ambitious restoration. (That was true for the government, too: Among the court's longstanding critiques of the BiOp was that its habitat projects weren't "reasonably certain" to occur.) Accepting the deal, Brigham figured, would help her tribe ensure that its rights were honored. "We wanted to get projects on the ground, we wanted to make things go smoother, and we wanted results," she told me.

Though most of the tribe's Board of Trustees supported the Accords, the decision was a hard one. "There were some concerns that we were giving away rights," said Eric Quaempts, the Umatilla's director of natural resources. Other tribes hesitated, too. Initially, more Yakama voted against signing the Accords than in favor. Not enough voted to make the decision official, though, so the Yakama voted again; this time it passed.

On May 2, 2008, at Columbia Hills State Park, federal agencies and tribes signed the Accords. A 300-year-old petroglyph of a female Indian chief with giant oval eyes -- Tsagaglalal, She Who Watches -- gazed over the proceedings. "These Accords move the focus away from gavel-to-gavel management and toward gravel-to-gravel management," Wright intoned. Leaders from affected agencies and tribes scrawled signatures on a commemorative buckskin.

Every tribe, that is, but one.

When I visited her office, Rebecca Miles had just returned from a funeral. Miles, 42, is executive director of the Nez Perce Tribe; the funeral was for her great-uncle, the legendary Nisqually activist Billy Frank. Six thousand people, including Washington's governor and its two U.S. senators, attended his service.

Now Miles was back in Lapwai, Idaho, a reservation town of 1,100 halfway up Idaho's chimney, where a sign at a shuttered gas station displayed, instead of prices, the phrase "An Arm and a Leg." A bracelet of white beads spelling "BILLY" dangled from Miles' wrist. A lock of highlighted hair swept across her warm face. "Just to be in Uncle Billy's presence was something," she said. The first time Miles went to Washington, D.C., to advocate for her tribe's fishing rights, she bumped into Frank as he "hauled ass" through the halls of Congress. They'd hugged and hugged, Frank brushing off his entourage as they tried to whisk him along.

Frank's modus operandi was civil disobedience: He had been arrested more than 50 times for catching salmon at traditional grounds. Miles, in her own way, was just as rebellious. When she was elected chair of the tribal council in 2005, at the age of 31, she became the first woman -- and youngest person -- ever to hold the position.

At first, Miles was pigeonholed into social service committees; fisheries were the domain of men. But after the Nez Perce missed a couple of fish meetings in Portland, Miles volunteered; soon, she was traveling for policy summits, and earning the grudging respect of male colleagues. "Every time we opened a meeting, we'd hear speeches from men asking what the hell women were doing here," Miles recalled. "You'd see eyes rolling in the back of the room. It was like, 'OK, now that you're finished, can we get to work?' " Sometimes her two young sons, Tre and Ivory, slept beneath the tables.

Not long after Miles became the Nez Perce's lead fisheries negotiator, Wright approached her, as he had Brigham and other tribal leaders, about an agreement. At first, she favored the deal. Four endangered fish stocks -- Snake River steelhead, sockeye, and spring and fall chinook salmon -- spawned within Nez Perce territory. If Bonneville Power was going to keep these fish alive, Miles thought it would have to cooperate with the tribe's fisheries department, already hard at work on recovery projects. The seeds of a favorable arrangement had been planted; all that remained was to water them.

But the Nez Perce differed from its fellow tribes in important ways. For starters, it possessed better fish habitat. To explain why, Dave Johnson, program manager for the tribe's fisheries department, pointed to a map pinned to the wood-paneled wall of his Lapwai office. A circulatory system of blue tributaries ran through green blocks representing public lands -- the Selway-Bitterroot and Frank Church wildernesses, national forests galore. These protected lands are where many Snake salmon return to spawn, Johnson explained. "This is some of the best salmon habitat we've got left."

All that habitat, though, hasn't created prolific runs. "You just have to look at the populations that spawn in the wilderness areas," said Johnson, a Navajo whose nimbus of gray hair adds inches to his lanky frame. On the Middle Fork of the Salmon River, a Snake tributary that runs through the Frank Church, many creeks host fewer than 100 spring and summer chinook annually -- a tiny fraction of their capacity. "Those runs are barely holding their own, just above the flatline," Johnson said. "They're certainly not showing gangbusters improvement. That, to our mind, is the proof in the pudding."

In other words: Dams, not habitat, are Idaho's biggest problem. While juvenile fish born further downriver -- where the Umatilla, Yakama and Warm Springs reside -- contend with no more than four dams en route to the ocean, Idaho's fish traverse four additional dams on the lower Snake as they move downriver. (Most do, anyway: Around a third of the basin's smolts are transported around dams in barges.) "We've got double the challenges for mainstem migration," Greg Stahl, assistant policy director at Idaho Rivers United, told me.

Over the years, said Russ Kiefer, fisheries biologist at the Idaho Department of Fish and Game, agencies have gotten better at keeping young fish alive as they pass through dams -- for instance, by installing spillway weirs that keep smolts near the water's surface. But while survival has improved "at the concrete," some Snake River stocks have lagged nonetheless: Around one in every 100 juvenile spring chinook has returned as an adult over the last decade -- well below government objectives, which aim for a minimum of 2 percent to maintain stocks, and 4 percent to rebuild them. Wild steelhead have also come up short.

In contrast to their Snake counterparts, mid-Columbia spring chinook have averaged over 3 percent returns since 2000. There's evidence that "delayed mortality" is partly responsible for the discrepancy: Though some 96 percent of smolts survive any individual dam, the accumulated stress of passing eight barriers, along with dodging predators and slogging through reservoirs, leaves Snake River fish more vulnerable to death once they reach the ocean.

For those reasons, supporting the BiOp was anathema to Miles. After years of disputing NOAA's salmon recovery plan, after years of arguing -- often victoriously -- that Bonneville Power was killing fish and breaking the law, the Nez Perce Tribe was being asked to accept the very hydroelectric system that helped jeopardize its fish in the first place. The deal may have signaled the end of "gavel-to-gavel management," as Wright put it, but the gavel was one of the best tools the tribes had. What's more, the BPA asked that tribes cease advocating for breaching the Columbia and Snake dams -- not only in court, but also in the court of public opinion. The Accords, Miles thought, were a muzzle. "One of my councilmen said, 'You mean we can't even wear sandwich boards?' " Miles recalled. Wright traveled to Lapwai hoping to change the council's mind, but found it steadfast; there would be no Accord with the Nez Perce.

When the triumphant agencies presented their Accords-packed BiOp before Judge Redden in the spring of 2008, the Nez Perce's erstwhile allies sat on the other side of the room, alongside BPA, the Bureau of Reclamation, and the Army Corps of Engineers, which operates the dams. Still, Miles was optimistic: Redden, she suspected, would find this version of the BiOp as insufficient as the last. Sure enough, in August 2011, he ordered the agencies to return to court in 2014 with longer-lasting mitigation measures -- and, crucially, to "consider whether more aggressive action, such as dam removal and/or additional flow augmentation . . . are necessary to avoid jeopardy." The Accords had failed to sway the person whose opinion mattered most.

"Judge Redden ruled strictly on the legality of the BiOp," Miles told me, her BILLY bracelet clicking as she jabbed the air. She leaned across her desk. "He did not rule on the fact that Steve Wright now had 20 more friends."

When you talk to Miles, it's easy to see the Accords as a ploy to marginalize the tribes, and the Nez Perce -- along with the state of Oregon, which also declined to sign -- as the only incorruptible actors. As Lewis & Clark's Michael Blumm put it, "Thank God for the Nez Perce."

Perhaps, though, the Nez Perce Tribe is simply as pragmatic as its peers. Its location left it uniquely exposed to the lower Snake dams, and the Accords would have eliminated its only weapon against the hydropower system. What's best for your fish depends on your place on the river.

And though the Accords may prevent the Lower Basin tribes from fighting for now, there's nothing stopping them from resuming their legal battle in 2018 if they feel the projects aren't working. "Removal might be the best thing for the tribes, but we know that's not going to happen anytime soon," JP Patt, a member of the Warm Springs Tribe and former executive director of the Inter-Tribal Fish Commission, told me. "We have to operate within the world we live in." Why not redouble habitat efforts and see what happens?

To see how the tribes are answering that question, I drove west from Lapwai through emerald wheat fields to Pendleton, Oregon, home of Brigham's Umatilla. Outside the tribe's gleaming, spaceship-like headquarters, I climbed into an SUV with Gary James, the tribe's fisheries program manager. James is energetic, with a salt-and-pepper goatee and a self-deprecating sense of humor. Once, in a meeting where James felt his decades of experience were being discounted by the assembled Ph.D.s, he rose to reveal an undershirt that read, "One fish, two fish, red fish, blue fish." "Doctor Seuss," he told them, "is the only doctor I need."

When James, who's white, became the Umatilla's first fisheries biologist in 1982, the job seemed impossible. To be a fisheries biologist, you need fish; to have fish, you need water. The Umatilla River had neither. Irrigated farms had drained its lower stretches; chinook and coho hadn't spawned there in over 50 years.

Though many tribal members wanted to use their treaty rights to sue the irrigators, James and a consultant named Ed Chaney instead used the treaty as leverage to negotiate a giant water swap. They convinced farmers to help lobby for pumps, constructed by the Bureau of Reclamation and powered by BPA electricity, that would draw water from the Columbia and dump it into irrigation canals, allowing farms to leave water in the Umatilla for fish. Chaney twisted some arms to secure a couple million juvenile salmon for restocking, and the runs rebounded. This year, almost 6,000 spring chinook returned to spawn; as James and I cruised along the Umatilla, we spotted a wader-clad fisherman scrambling up a bank toting a fat, bloodied salmon.

The Umatilla River deal taught the tribe the value of negotiating, and years later influenced its choice to sign the Accords. This wasn't Nez Perce land, where wilderness habitat abounded. The Walla Walla region is so invested in agriculture that its minor league baseball team, the Sweets, is named after an onion. As we drove, James pointed out farming's impacts on habitat -- streams with concrete banks and straightened channels, purged of every stick and boulder. "I don't blame the Nez Perce -- when all of your tributaries are above eight dams, you're up against the wall," James said. "But we're on different playing fields."

To show me habitat in action, James took me to Meacham Creek, a cerulean tributary that wends down a lush mountain pass on its way to the Umatilla River. Decades ago, the Union Pacific railroad straitjacketed the stream with dikes to protect adjacent tracks from flooding. Meacham became a fast, warm sluice with no floodplains, bends or woody debris to shelter young fish. Sometimes the creek got so high that the main channel became a standing wave, washing juvenile salmon downstream.

In 2011, James and his department began restoring seven miles of Meacham Creek, a years-long process that will cost up to $6 million. Along one particularly degraded stretch, they knocked down the dikes and plugged the main channel with a logjam, bumping the watercourse back into its natural meanders and onto the floodplain. They dug pools and planted trees and installed dead wood. What was most remarkable, I thought as we scoured the water for signs of spawning, was how unremarkable everything appeared. Around the margins, you could just see the project's seams -- wire fencing around seedlings, bulldozer scars. Still, the creek looked like a creek. The pale oval of a steelhead spawning nest, or redd, glimmered at the tail end of a new pool.

As of last spring, Accords projects had protected or improved 175,000 acres of habitat, an area about twice the size of Seattle, making over 1,100 miles of disconnected spawning grounds accessible for fish. And "when you have this pot of money, it becomes easier to get grants (from other agencies)," said the Warm Springs' Brad Houslet. "It gives them certainty that projects will get off the ground."

Before the Accords, James told me, beaming, it would have been foolhardy to attempt such a thorough, long-term project as Meacham Creek. "We used to not even think about work of this magnitude."

Not even the staunchest dam-breaching advocate would deny that projects like Meacham Creek help fish. What conservationists dispute is whether the BiOp approach can meet recovery goals. "I spent years improving habitat," Chaney, James' collaborator in restoring the Umatilla, told me. But above the lower Snake dams, where pristine spawning grounds are plentiful, Chaney said that habitat is a red herring. "Fiddling with some creek that's been overgrazed . . . how can people take that seriously?"

Chaney, a towering man with a gleaming pate and a trim white goatee, is one of the few environmentalists willing to roast the Accords. He operates his one-man nonprofit, the Northwest Resource Information Center, from his home in Eagle, Idaho; a black cat named Winston patrols the cluttered kitchen table, threatening to upset reams of paper. Chaney's pale blue eyes pop cartoonishly when he's outraged. And he's often outraged: "Bonneville taking advantage of these tribes just gives me heartburn," he told me.

Although Chaney once consulted for the Umatilla and other tribes, he said he's now on Bonneville's blacklist. He's one of the agency's most vocal critics, particularly over its outsized influence on salmon science and policy. Through the Accords, he said, "they've taken one of the hole cards for Snake River salmon out of the picture. No national sympathy for tribes need get in the way of killing fish."

So it was that Chaney declared war on the deal, filing a library's worth of letters and formal complaints to the BPA, the Energy Department, the U.S. Inspector General, and a bevy of governors and state agencies. His grievances come down to this: Shouldn't it be illegal to make states and tribes accept the BiOp, a document that itself doesn't meet the law? One of the few replies came from the office of Washington's attorney general: a form letter thanking him for his support.

For Chaney, the chief problem with the Accords is that BPA was already committed to many of its projects. Sixty percent of the money did go to new and expanded programs that wouldn't have happened otherwise -- for example, $50 million to restore Pacific lamprey, which aren't endangered. (In exchange, signatories agreed not to petition for a listing.) The rest of the funding, however, went to projects that had already gained approval through the Northwest Power and Conservation Council, the regional body that guides Bonneville's salmon spending. "Basically, BPA said, 'We'll do what we're already doing, and you guys will go away,' " Chaney told me.

Even worse, to Chaney, is the way the Accords, and BPA generally, have constrained science. Witness the controversial issue of spill. Years of data suggest that spilling more water over dams, rather than passing it through turbines, improves juvenile survival by expediting the smolts on their journey downstream. In 2005, Judge Redden ordered the Army Corps to ramp up spill at four dams; today, thanks largely to Redden's order, BPA spills some 30 to 40 percent of water at any given hydroelectric project. Of course, water that doesn't generate power comes at a price: A 2013 power council report claimed that spill and various other operations intended to help fish have cost BPA $3 billion in power sales since 1978.

But are the dams spilling enough water? Last year, the state of Oregon, the Nez Perce and others called for an experiment that would boost spill beyond current levels over 10 years to see if it helped fish. More spill has its own perils: Send too much water cascading over a dam, and you risk filling the river with enough dissolved gas to harm fish. To keep smolts from getting the bends, gas limits in dam tailraces are currently set at 120 percent, meaning that dissolved gas can't exceed normal water conditions by more than 20 percent. "No matter where you believe the threshold is, the higher the gas levels, the closer you are to harm," says Bill Maslen, director of BPA's Fish and Wildlife program.

When scientists at the U.S. Fish and Wildlife Service and other agencies crunched the numbers on the Oregon spill proposal, however, they found some wiggle room. Their models suggested that jacking up allowable spill to 125 percent could get Snake River spring chinook and steelhead within range of the coveted 4 percent adult returns needed for recovery. The experiment, says the Nez Perce's Dave Johnson, would be "the big swing. I don't know what else we've got in our bag of tricks."

Additional spill, however, isn't in the BiOp, which means Accords signatories can't push for it. When the Power Council discussed the experiment at a meeting in December 2013, Idaho council representative Bill Booth told the room that, since the Accords committed his state to support the BiOp, he had to vote against sending the proposal to an independent science board for review. In one of his many filings, Chaney called the moment a "smoking gun" that demonstrated the Accords' influence on science.

Booth's objections notwithstanding, the spill experiment did end up before the science board, which ruled that, while the idea had merit, the proposal's design wasn't yet up to scientific snuff. Its backers are working on a refined version, but any plan that would cost BPA an additional $110 million per year faces long odds. "The data on spill is clear: Here's a path, short of breaching, that could recover these populations," said Steve Hawley, author of Recovering a Lost River. "The federal agencies are simply refusing to try it, and in some cases they're not even letting scientists present the data."

Still, for environmentalists, more spill would be a consolation prize. Their primary objective has always been breaching. But the decades-long campaign to remove the Snake's dams has never gotten very far, even as dams on Oregon's Sandy River and Washington's Elwha and White Salmon have been torn out. Those vanquished dams had one thing in common: They outlived their economic usefulness. By the same logic, conservationists are resting their hopes on the math of a man named Jim Waddell.

Fifteen years ago, Waddell, an Army Corps lifer, moved to Oregon to work at the agency's Walla Walla district just as it was completing a $30 million study on breaching the lower Snake dams. As Waddell reviewed the final drafts, which recommended keeping the dams, he realized that the Corps' estimate for their hydropower production seemed too high. He asked uncomfortable questions, received evasive answers. A few years later, the Corps sent him back to Atlanta.

But Waddell is persistent, and in 2013, now living in Port Angeles, Washington, he tackled the numbers again. His new investigations seemed to confirm his suspicions. The Snake dams didn't pay for themselves. Hydropower, he said, had indeed been overestimated, and the alleged operation costs were far too low. Other benefits -- farmers barging their wheat to market, which would be impossible without the dams -- had been exaggerated, too, since barges could be replaced by rail. And if the dams were breached, recreation opportunities, like boating and fishing, would be more valuable than the Corps claimed. Add it all up, said Waddell, and breaching today would be worth at least $158 million a year. "It's no longer about the salmon-lovers," Waddell told me. "It's about fiscal responsibility. We've got to make some choices."

While the Accords have reduced the constituency for breaching, at least temporarily, they haven't ended the courtroom battles. This spring, a coalition of environmental groups led by Earthjustice sued over the newest BiOp, released in January 2014. (Oregon joined the suit in October.) The suit was a bitter disappointment to Terry Flores, director of Northwest River Partners, which represents regional business interests -- including utilities that buy power from Bonneville and helped pay for the Accords through rate hikes. Fifteen to 20 percent of residential energy bills in the Northwest now go toward fish and wildlife, Flores told me. "It was challenging for us, looking at the price tag."

Still, Flores' group went along with the Accords in hopes that they would lead to delisting, saving utilities money in the long run. "If we can get the habitat benefits that we anticipate, we can get a 'no jeopardy' opinion from the judge," Flores said. Though she's heartened by improving salmon returns, the litigation continues. "We're reserving judgment," she concluded. "We still want to see the fruits."

Biologists like Gary James urge "listening to the fish" -- letting the runs be the ultimate arbiter of what's working. But in a watershed bigger than France, it's difficult to isolate one project's impact. "You can look at suites of actions and get a better indication, but there's a lot of variability," said BPA's Maslen. The agency's data so far supports the restoration efforts, he added. "Whether it's redds where new habitat is now available, or greater fish production where we have increased habitat complexity, what we invariably see is a fairly rapid response."

Indeed, every week this summer and fall seemed to bring new fishy bounty -- Okanagan sockeye, Clearwater coho, Hanford Reach chinook. In 2013, over a million fall chinook passed the Bonneville Dam for the first time since counting began in 1938; this year, more than 600,000 sockeye traversed the dam, another record. To conservation groups, the spikes are evidence that spill is working; to the agencies, it's a validation of the BiOp. Both sides credit favorable conditions in the Pacific Ocean. "Our analysis shows that if you implement these actions (listed in the BiOp), it's enough to avoid jeopardy," said Bruce Suzumoto, senior policy advisor for NOAA's West Coast division. "We've done a lot in the river, and we're getting good returns."

But while endangered Snake River spring chinook and steelhead may be "avoiding jeopardy" -- the BiOp's euphemism for staving off extinction -- environmentalists say they're not meaningfully recovering. Before the dams, Snake River spring/summer wild chinook returns numbered over 100,000; during the '90s, that number fell below 10,000, then rebounded to around 25,000 in 2010, where it's since hovered. As long as populations hold steady, said Greg Stahl of Idaho Rivers United, the feds can claim success -- even if delisting remains a dream. "We're still so far from recovery it seems like public relations spin to talk about them improving," said Stahl.

In a sense, the "avoid jeopardy" standard is a legal manifestation of a condition called shifting baselines syndrome, the long-term ecological amnesia that causes each successive generation to accept its own degraded present. Sure, 2.3 million fish passed over the Bonneville Dam this year, but 16 million used to migrate up the Columbia annually; it is proof of our reduced standards -- and the Endangered Species Act's low bar -- that we celebrate a fraction of historical runs. And shifting baselines have management implications. "What we have is a prevention of extinction policy, rather than a policy that achieves real recovery," Rod Sando, former head of the Idaho Department of Fish and Game, told me. "Recovery would mean managing dams in a different way" -- with more spill, or by breaching them altogether.

But for all the teeth-gnashing, baselines can shift in another direction: Things can improve without us noticing. Do kids catching salmon behind the Taco Bell in Pendleton realize there was a time, just three decades ago, when no water flowed in the Umatilla River? Your view depends on your historical frame. Go back far enough, and present-day runs seem tiny. But if you pull too far back, you risk missing recent progress.

That progress, at least in terms of numbers, has been aided by fish hatcheries. Snake River fall chinook, for instance, owe their survival to a Nez Perce hatchery on the Clearwater River that annually produces 5.5 million smolts. (Though the Nez Perce didn't sign the Accords, the tribe receives millions of BPA dollars from other sources.) While conservationists worry that hatchery-raised salmon dilute the finely tuned genes of wild stocks, tribes are among their most enthusiastic proponents, pumping tens of millions of juvenile salmon into the Columbia Basin every year.

The tribes' reliance on hatcheries is understandable: To them, salmon aren't just sustenance, they're key to cultural identity. In 2007, Eric Quaempts, the Umatilla's natural resources director, redesigned the tribe's land-management strategy around the preservation of traditional foods, including salmon, that are intimately connected to the Umatilla's ceremonies and creation story. If fish aren't abundant enough to harvest, the Umatilla and others risk losing their sense of self. And when natural spawning has been all but wiped out, fish have to come from somewhere. "The tribes would be glad to downsize or eliminate hatcheries if the fish started replacing themselves," said Gary James. "Mother Nature does it the best, but the tribes want fish."

That mindset -- that a fish's genetics and ESA status are less important than its availability -- has led to some strange conflicts. In the Klickitat Basin, the Yakama and the state of Washington operate one of the Northwest's largest hatchery programs, churning out coho, steelhead, and spring and fall chinook with the help of several million dollars in annual Accords funding. Sometimes those fish stray across the Columbia and into the Hood River, a tributary co-managed by the Warm Springs Tribe and the state of Oregon. Rod French, a district fish biologist with the Oregon Department of Fish and Wildlife, co-authored scientific papers documenting the deleterious genetic effects of hatchery fish; his findings helped set policy in Oregon, which uses hatcheries more sparingly than it used to. I asked French what happens to tribally produced hatchery fish that are caught by his department on the Hood River. His answer was swift: "They're eliminated."

Before we visited Meacham Creek, James took me to see the Accords in action at the Umatilla River's Three Mile Dam. At its base, several dozen muscular chinook had been routed into a pen, where they finned in lazy loops. As we watched from a metal scaffold, the far end of the pen began to glide like the wall of a trash compactor, shunting the now-thrashing fish into a sort of dumbwaiter, which rose level with the platform on which we stood. The elevator door opened and the fish spilled into a new tank, its water laced with an anesthetic that made them easy to handle. A team of techs hauled each fish onto a table to record its length and sex, and then slid it into one of two waterslide-like chutes leading to waiting tanker trucks.

While some of the fish would become hatchery broodstock, others would be trucked to a new facility, where they would wait for weeks until they were ready to spawn, at which time they would be driven to tributaries that hadn't held fish in a century, and then released. "They might be thinking, 'Where am I, how did I get here?' " James said with a chuckle. "But by then, they're so close to spawning, that's probably all that's going through their minds."

Three Mile Dam to me seemed a perfect microcosm of Columbia River management. The goal of restoring salmon, and thereby honoring tribal fishing rights, was vital, even noble. Still, I couldn't help but wonder what the fishermen of Celilo Falls would think if they could see the bizarre means. Salmon corralled like cattle at a concrete wall, doped up, then whizzed away in trucks.

In a sense, there's a third shifting baseline on the Columbia -- one of weirdness, of increasingly intensive human meddling. There are fish in the river, fish that look and taste like salmon; but the fish's experience, and our experience of them, is so altered that they could almost be different animals. No longer do salmon leap waterfalls en route to gravelly spawning grounds; instead, to be a salmon is to slide down a chute into a truck, to be milked at a hatchery.

More than perhaps any creature, salmon epitomize modern wildlife management. We are willing to bend over backwards, to the point of comedy, to recover species we cherish: We captive-breed black-footed ferrets; we shoot barred owls to save spotted owls; we patiently teach whooping cranes to migrate behind aircraft. Yet coexistence occurs strictly on our terms -- and there is always at least one term left non-negotiable. We spend millions on wildlife crossings over highways, yet would never close the highways themselves; we relocate imperiled trees to help them weather climate change without daring to retool our carbon-based economy. In the Columbia Basin, the dams, and their power, are the inviolable condition, the infrastructure that fish and managers must turn cartwheels to accommodate. We will give salmon everything, except what we don't want to give.

"I hope we don't ever turn around and say, 'We should have done more,' " Rebecca Miles told me. Still, when it comes to salmon, Miles has experience taking the long view. As a child, her father and brother left her at home when they went on their salmon-gathering expeditions; the fishing grounds, which often simmered with tension, were no place for a girl, they said. Miles would wait up until 3 a.m. for her family to get home, sometimes with as many as 50 salmon, and eagerly cut and pack fish until the sun rose.

Today, at the same sites where federal agents once thrust rifles in the faces of Miles' relatives, Nez Perce fishermen camp alongside whites. Her own sons would never be kicked off the river. Now that she's executive director of the Nez Perce, she's no longer involved in the day-to-day machinations of fisheries policy, and she seemed both grateful and a little sorry to be out of the game. "I used to look at Billy Frank and wonder where he got the energy," she told me, touching her bracelet. "But right now, after coming from his service, I really feel like anything can be accomplished."