Just beyond the high pass into Stanley, Idaho, at the end of a very long day, I ran out of gas. It was 1997, and I still remember the sound of the engine on my father’s thirsty old Bronco, first sputtering, and then dying. But I was somewhere around seven thousand feet in the Sawtooth mountains and had gravity on my side. I managed to keep rolling downhill for miles, taking the long curves in silence, fast enough at first, then slower, then slow. It was already getting dark, and when I finally coasted to a stop on the gravel shoulder, I heard, rather than saw, the river.
It was called, like so many others, the Salmon River. All across the 260,000-square-mile Columbia River watershed, stretching across seven states and one Canadian province, tens of millions of salmon had once coursed upstream from the Pacific Ocean each year, completing a typical life cycle of two to four years by breeding in the same freshwater streams where they were born. But in the 1860s, gold mining and timbering had begun to silt up the high mountain spawning beds, covering and suffocating their eggs. And by the 1880s, massive canneries at the Columbia’s mouth in Astoria, Oregon, were tinning a million salmon a year. The Columbia salmon population began to plummet. Between 1938 and 1979, the construction of eight huge hydroelectric dams on the Columbia and Snake Rivers deepened the collapse to extinction levels.
But there, in the river, was one of them. In the dusk, standing beside the Bronco, I heard a tremendous splash in the shallows. I ignored it, but moments later it happened again, and in disbelief I stalked slowly up the bank. A monster was there, in the curve of the shallows, a dark shape against the light gravel of the riverbed. At first I could only see the thrashing splashes of her tail, but as I approached, I began to make out the mottled flesh of a four-year-old sockeye hen, struggling to clear an area for a redd, or nest.
Against the odds, despite a century and a half of human interference, she had managed to swim 900 miles and climb 6,200 feet into the Sawtooths. The pure, oxygenated water and lack of predators here were ideal for reproducing. If a male had made the same journey, and could find her, they would send hundreds of thousands of tiny smolts downstream in the spring.
The final moments of life are never pretty. The sockeye’s skin was peeling away, its flesh lumpy and gray. Literally falling apart, the hen nevertheless summoned the energy to sweep away the silt that might choke her eggs. Every few minutes, she felt the spur of her DNA, and gave one more abrupt splash of the tail, one more heave.
Eventually an old RV came squeaking down the hill. The driver gave me a gallon of gas and, reluctantly, took my money.
Twenty years later, I jumped over a chain-link fence outside Riggins, Idaho, into a salmon hatchery on the banks of the Rapid River, a tributary of the Salmon. The facility was unimpressive, mostly corrugated sheds and a few pieces of heavy equipment scattered around. It was closed, barely—the posted visiting hours had just expired—but I had been hopping a lot of fences with my trout rod recently, and it felt Western-normal. Anyway, I really needed to see a $68,000 fish.
Back in 1875, when salmon runs first started to crash, the Smithsonian scientist and U.S. fish commissioner Spencer Fullerton Baird was hired by the Oregon legislature to draft a plan. He knew and identified the real problems: overfishing, the degradation of high mountain streams, and an overabundance of dams, many with no fish passage. But Baird, calculating that there was no political will to solve the underlying problems, proposed a cheaper solution: hatcheries. The salmon life cycle would be reproduced by scientists, the eggs fertilized and sown into rivers like wheat into fields. For just $15,000 a year in operating costs, he promised an unlimited harvest, and Oregon and the rest of the Columbia watershed set about building hatcheries that cranked out billions of fertilized eggs over the course of the next century.
It didn’t work. Salmon runs plummeted, despite the constant supply of artificially fertilized fish. Even when the first run of salmon was listed under the Endangered Species Act, in 1991, and the United States began spending tens of billions more on hatcheries and other salmon-restoration projects in the Pacific Northwest, the runs only marginally improved. While the threat of extinction has receded since 1983’s record low of 185,000 Columbia Chinook, the few million-fish runs have still been well below historic levels, and last year only 336,000 Chinook surged up a river that drains more water than all of France.
Today, about 300 million salmon are planted annually in the American Pacific Northwest (Alaska and Canada plant hundreds of millions more). They are grown in some 300 public fish hatcheries scattered around the watershed, many of them jointly operated by the state and the feds. Rapid River’s design was typical: a series of twelve long cement pens, all of which were stained with black mold and topped with rusting walkways. About 3 million tiny fish were spread out in front of me. In the spring they would be released directly into the river, or transported in tanker trucks and poured out upstream using long plastic chutes. Here they were still growing and waiting, circling slowly—a dark cloud of life.
Four out of five salmon in the Pacific Northwest are now born in hatcheries, and the vast majority will die. This has led the Northwest Power and Conservation Council, an organization that advises the region’s governments on environmental and energy matters, to estimate that for fish spawned in the lower river hatcheries, where success is relatively easy, it costs $66 to produce a harvested fish that navigates the trip home. In the middle regions of the watershed, where fewer fish survive the journey, it can cost up to $9,000 to bring a fish back. And the program to breed Chinook in the high mountains of Washington has been even less successful. The Council estimates that the true cost of bringing a single spawner back to that uppermost region is as much as $68,031.
In some cases, it may be even more. Two decades after I stopped by the Salmon River, for instance, just 157 sockeye made it back to those waters. Biologists determined that not one of them was hatchery-born—they were all wild fish, whose robust natural paternity and hardy stream upbringing had given them a distinct advantage in surviving the return trip to Idaho. The hatchery fish were weaker, raised in tubs and selected from a narrow genetic slice that lowered their adaptability to a changing climate. Even though they outnumbered the natural-born salmon by millions, the failure rate in this case was so high that the vast sums of money spent in upper Idaho had not purchased a single hatchery success.
The 157 wild sockeye ascending the Salmon River were far too valuable to leave in the stream to reproduce naturally. So, as is typical in the salmon-industrial complex, they were electrocuted and captured by biologists and federal “fish techs,” who trucked them in cold water tanks to another hatchery, and kept them alive in pools until their natural autumnal urges materialized and they were clubbed over the head and killed. The females were then cut open and stripped of their eggs, which were milted by a few males and later used to breed the next generation.
Even the remains of those 157 sockeye were too precious to waste. Historically, the massive annual salmon runs delivered a pulse of phosphorus and other ocean minerals to the interior of the continent, the fishes’ own flesh the means of transmission. Bears and birds had once spread the carcasses, and minerals, around; now Boy Scouts did it. There is a merit badge for fish conservation, and a troop of Idaho Scouts carried the 157 carcasses deep into the woods and left them in streambeds to mimic the conclusion of a salmon’s natural life cycle. Even death can be simulated.
At this point in history, no phase of the salmon’s life is left to nature. The spring trip undertaken by the Rapid River Chinook toward salt water would be even harder than the journey back up two or three years later, requiring them to slip past eight massive dams, past predators human and otherwise, all while avoiding wrong-turn tributaries and dam pools so slack they offered no guiding current, as well as their opposite: intentional “spills” of reservoir water, used for navigation, that can suddenly accelerate the river’s flow. There were irrigation withdrawals, turbine screens, sea lions, Superfund sites, and a thousand other obstacles—a journey so perilous fish biologists call it “running the gantlet.” Fewer than 1 percent of wild salmon return successfully to their natal waters; for hatchery-born fish, the figure is closer to .001 percent.
Now it was my turn to run the gantlet. I was headed downstream to my home in Portland, 430 road miles away, a hydrographic descent through the problems, pollutants, and bad decisions of the Columbia; through the delusion that we can control nature and reproduce the entire life cycle of a species. Every mile was a reminder of how, after a century and a half of human management, hubris, and greed, the tens of millions of fiercely strong wild salmon had become hundreds of thousands of narrow and devolved replicants. We had already killed the salmon, and our crime was written across the landscape.
At first, the way home tracked the course of the Clearwater River, Lewis and Clark’s old route westward, the road a hash of aggressive logging trucks and middle-aged men barely in control of their Harleys. Coming around a bend at Orofino, I almost choked at my first sight of the Dworshak Dam, a reclining concrete berm more than 3,000 feet long and so high—at 717 feet, the third-tallest dam in America—that it was considered impossible to add a fish ladder, a kind of watery ramp that salmonids can ascend. (We speak of salmon, but since the last ice age there has been a wide family of salmonids, including the mighty spring and summer Chinook, the elusive summer steelhead, tiny westslope cutthroat trout and leaping rainbows, and the landlocked kokanee, all cousins whose differences in size and habits dissolve the moment you slide a knife down their spines, laying open a nearly identical network of pearlescent ribs wrapped around an organ cavity, proof that these species all come from a common ancestor and have adapted themselves to thrive in every habitat from the vast ocean to mountain streams so small that they disappear by July.)
When the Dworshak Dam was built, the 60,000 steelhead that hatched upriver each year had simply been written off. Altogether, some 40 percent of the original spawning habitat in the Columbia Basin has been lost in this way, cut off by big dams and little culverts, regarded as the necessary cost of flood control, agriculture, and cheap electricity. The proposed solution was always the same: build a “mitigation hatchery” to pump more fish into the damaged system. The Dworshak Dam’s inevitable handmaiden was the largest hatchery in the Pacific Northwest, a complex that now helps release 5 million hatchlings into Idaho rivers every year.
This was less an attempt to restore salmon than an effort to alleviate the concerns of fishermen like me while claiming to address the broader crisis. The ultimate purpose of salmon “restoration” efforts, in other words, is to keep anyone from noticing that salmon are not being restored. The dirty secret is that only about 2 percent of spending on “restoring” salmon actually flows directly into restoring natural salmon runs. The rest is spent on mitigating losses by managing the river—its dams and infrastructure projects, even its navigation features such as buoys and docks—and by “supplementing” rivers year-round with fish.
In Lewiston, Idaho, the placid Clearwater entered the fast-moving Snake River, and I resupplied with food and gas. The city is 465 river miles from the ocean, yet Lewiston has a Coast Guard Auxiliary station and is technically an oceangoing port with access to the Pacific. Connecting the city to the sea, so that grain barges might float all the way down to the terminals near Portland, had required the final, most destructive phase in the taming of the Columbia watershed, when the Snake was handcuffed by four hydroelectric dams in the 1960s and ’70s. A cold desert river full of oxygen, with swift white water and natural resting pools, had been transformed into a series of warm, stagnant reservoirs.
That night, I camped outside my car at a marina full of flat-water party boats. I paid a small fee for the privilege, thus supporting one of thousands of small businesses that make a living off dams and reservoirs. Jobs, careers, even whole towns have grown up around this infrastructure. Most dams are required to spill extra water at certain times of year, to help baby salmon slip down the gantlet, but water in the West is growth, farming, real estate, and direct revenue. A shipping executive in Portland once described to me how lowering the water level in the Snake by a single inch, to help salmon, forced him to cut a full ton of wheat from each barge. “That means more trips and more pollution,” he told me. More wheat would be carried in trains and trucks, raising emissions and prices. The effects rippled through everything.
After a night in the marina, and some fruitless casting for trout around the silent moorage at dawn, I followed the Snake northwest into Washington. I passed the Lower Granite Dam and then, hours later, the Little Goose Dam, where I sat under the midday sun, in the singed brown landscape, and watched the Snake thunder out of eight spillways, the plumes so large they clouded the sky, the sound so loud that my chest vibrated from a quarter mile away. Like all the Snake dams, Little Goose is low and ugly, with none of the soaring ambition of the 1930s W.P.A. dams. All four Snake dams together provide just 5 percent of the region’s electricity, yet the violence done to the river is boggling. It wasn’t like natural white water or a waterfall, but rather the ejecta blasted out of an industrial forge.
The locks beside the dam began to move. “Locking through” is a nautical ritual, and I expected to see a shipment of wheat or some crucial industrial product pass by. But when the two heavily reinforced steel gates finally finished swinging open, only a little red tug emerged, pushing a puny barge labeled juvenile salmon.
It was carrying tankloads of hatchery fish down below the dam to prevent their deaths. If left to swim downstream by themselves, some of the sprats would get lost and overheated in the stagnant reservoirs behind each dam; despite screens, others would be ground up in the turbines. Those lucky enough to navigate the turbine chutes could still be crushed by the intense water pressure and later die from the bends, like a scuba diver, or expire from over-oxygenation. The Army Corps of Engineers maintains that nearly all the salmon survive this passage, but the pulverized smolts I saw at the Little Goose, pooled neatly below the dam, seemed to belie the claim.
Just beyond the Little Goose, the red barge tied up again and waited, guarding its cargo of baby clones against reality. I was itching to move on. The dam site, like many, features a Corps of Engineers parking lot where you can camp long-term. A few battered RVs, encircled with lawn chairs and generators and engine parts, littered the asphalt. They were occupied by the financially insecure, people trapped by economic brutality, unshaven old men with broken fishing rods propped against the railings of platforms overlooking the river. I felt an uncontrollable urge to flee before I became one of them.
Hours later, I passed the last and lowest dam on the Snake, Ice Harbor, visible right from the highway. The Columbia itself then finally slipped into view. The mother river was wide and slow out there, in a section called the Hanford Reach. Its clean gravel beds and sluggish currents once attracted the so-called June Hogs, a special run of monstrous Chinook. A century ago they were the biggest fish on the whole river, sometimes caught at 100 pounds. The June Hogs still appear at Hanford, but in greatly reduced numbers, and in greatly reduced size, bred down by the evolutionary pressure of a century of trophy hunting and warmer, more polluted oceans. Today, a salmon of 60 pounds would make the front page of the Oregonian.
In 1942, the federal government chose the Hanford Reach for its first plutonium production facility, where the fuel for Fat Man and the majority of the next 60,000 atomic bombs would be refined. Decades of high nuclear gamesmanship then left the reach with 53 million gallons of radioactive waste, which, over time, seeped out of the storage tanks at the facility and contaminated 100 square miles of groundwater. The Department of Energy now spends $2.1 billion a year—almost 10 percent of its annual budget—on cleaning up Hanford, which in turn has created tens of thousands of jobs and converted three nearby farm towns into the sprawling Tri-Cities metropolis. Almost half of the local county’s tax revenue now comes from cleanup-related projects. The truth was visible in the car dealerships, food franchises, and boat-storage facilities wrapped along the highway: we manage the river for human uses, and have always chosen industry over salmon. We were only ever pretending we could have both.
As I pulled out of Pendleton, Oregon, a city dominated by its famed 17,000-seat rodeo arena, my phone advised me that I wouldn’t need to make any turns for another two hundred miles. The Columbia settled into a mostly straight run westward, marking the border between Oregon and Washington. Headed to my final stop, in Portland, I flew over a hot and dusty plain. There was something sinister in covering huge stretches of river at seventy miles per hour, the Columbia right alongside, wide and gleaming in the high desert flatscape. What had taken Lewis and Clark a month took me a single afternoon.
Unlike the Snake, where dams were mostly hidden away in remote canyons, the Columbia’s human alterations were openly displayed to the neighboring highway. I passed along Lake Wallula, a deliberate misnomer for the wide reservoir of water backed up behind the McNary Dam. An hour later, I approached the John Day River, and again the pattern repeated: a slack “lake” had replaced the river, and a dam offered a broad cascade of turbine-whitened plumes, below which flowed a narrower, wilder river, the water briefly surging through its historic bed. Two dozen little aluminum boats were scattered downstream, hunting August fish trapped by the concrete wall.
Four volcanoes soon arose in the west, white-topped markers of the Cascade Range. Pods of fishing boats marked the entrances of cold rivers, first the Yakima, where the Yakima Indian reservation had been experimenting with a pneumatic “fish cannon” to shoot adult salmon over one dam, and then the Deschutes, where a $100 million concrete straw had been installed to suck baby fish from a reservoir and spurt them out downstream. These were paradoxical rivers, precious cold headwaters in the desert, the fish runs kept barely alive with billions in work-arounds and techno-quackery.
The Columbia started to slow and broaden again, the slack water created by the 1957 Dalles Dam, which had buried the old Celilo Falls, a set of rapids where coho and Chinook had been caught with dip nets for thousands of years. The salmon had once fueled trade and year-round prosperity for Native people across the region. One member of the Lewis and Clark expedition marveled at the bounty on display in one native village, remarking upon the numerous piles of “pounded Sammon Stacked up on the Shores.”
In 1996, the historian Richard White argued that the Columbia was so heavily interdicted by dams, hatcheries, canals, irrigation, and navigation schemes that it was no longer a natural river, but a hybrid of nature and technology. He called it the “organic machine.” I had been struggling on this drive to consider the river in these terms, but where was the nature? The “river” was one continuous series of slack-water pools, a canal through a desert. I had not seen a rock, or even a ripple of natural white water, for a hundred miles.
At dusk, I finally pulled to a halt at the Bonneville Dam. Bonneville, actually two different dams linked by an island, is the lowest and largest of the Columbia dams, and the source of the electricity that lights the lamp above my desk in Portland. I thought of a cluster of wind farms I had noticed earlier along the highway. There were thousands of turbines in the region now, including thousands more on the far side of the river, in Washington. Those windmills could generate, under the right conditions, so much energy that the Bonneville Power Administration would scale back electricity generation at several dams in response.
The alternative-energy revolution was going to bust the utility model, where the river was tamed to “do some work for me,” as Woody Guthrie sang. And if the power system was vulnerable to collapse, so was the entire set of political priorities on the river. In the dark, late that night, I drifted into my driveway in Portland and fell asleep in my own bed, dreaming of vested interests undone, of surprised faces at the Bonneville and John Day and Ice Harbor dams, the turbines motionless, the water running evenly down through a machine gone silent.
It is tempting to dream of blowing everything up, of restoring natural flows, closing all the hatcheries, and returning to some long-lost Eden. Langdon Cook, the Seattle-based nature writer, fantasized to me briefly about one solution that would quickly restore salmon runs. “Ban all fishing,” he said. In the entire Columbia basin. The problem, Cook noted, is that this measure would enrage commercial fishermen, who would unite with their allies, the Native American tribes who were more than happy to see the rivers filled with hatchery fish. Those tribes could and did fight such cases all the way to the Supreme Court, where they had a habit of winning.
One of the most militant and absurdly powerful special interests in the Pacific Northwest is my own cohort, the recreational anglers. At the slightest hint of a threat to the “put-and-take” system now in place, in which hatcheries supply weekend anglers with sport and table fish, they harass their legislators mercilessly. In 2014, the Three Rivers Sportsman’s Alliance protested the removal of dams on several Oregon rivers at the annual Portland Native Fish Society dinner, circling the venue in pickups towing fishing boats. The members of the Alliance believed, correctly, that without dams there would be no justification for the hatcheries that restocked the Sandy and other convenient rivers on a yearly basis. Greg Osburn, the Alliance’s director, sent me an old scientific study—the one hatchery advocates always cite—that claimed that hatcheries had successfully restored runs up in Idaho. Given that zero hatchery fish had actually reached the Sawtooths in 2017, it was woefully out of date. Moreover, most of Osburn’s supporters were not anglers themselves, he told me, but local businesspeople interested in profiting from the fishermen: fishing guides; bait shop proprietors; boat dealers; gas station, restaurant, and motel operators; and “lakefront” (reservoir) property owners. It was a list of short-term and shortsighted beneficiaries.
If we can’t blow up the machine, however, perhaps we can still tune it. Guido Rahr, president of the Wild Salmon Center in Portland, told me that human involvement on the Columbia was too complex to disentangle all at once. “You’ve got to dial everything up and down,” he said. That means adjusting hatchery outputs, rather than banning all hatcheries; altering dam flows by degrees to favor salmon, rather than demolishing all of the dams themselves. It means picking specific places to rebuild diverse and wild populations, rather than pumping hundreds of millions of de facto clones (and hundreds of millions of dollars) into the entire river.
Mark Sherwood of the Native Fish Society suggests shifting funds and priorities away from navigation and sport fishing and toward habitat restoration, enforcement, and monitoring. Rewilding the Columbia so that natural systems protect and expand the runs on their own would be no more expensive than the current system, according to Sherwood. “It’s a wash,” he told me, “but with all these other benefits, like clean water and restored wildlife.” Eventually, a healthy Columbia with big natural runs of salmon could generate $600 million in annual economic activity and add 20,000 jobs to the region.
Needless to say, the Trump Administration and its retrograde allies in the West will do nothing toward this future. But scaled against nature, Trump is brief. Salmon are adaptable, and patient, their four-year life cycle giving them twenty-five generations in a century, twenty-five chances to start over, and perhaps even adjust to a warming climate. Scaled against natural time, everything is possible.
As August turned to September, the sun dropped and the first cold freshet of autumn rainwater flowed out of the mountains and tickled the noses of the salmon waiting off Astoria, where the Columbia meets the Pacific Ocean. I’d fished them in the sea there, where the Chinook were called “brights,” because they came aboard gleaming with blue-green oceanic vitality, and the smaller coho glittered like silver nuggets. I’d caught them in the river too, where within hours of reentering the fresh water they had already begun to lose that color, their dullness the first sign of the fatal weakening that would accompany them upstream, toward the smell of some natal tributary, chilly, pure, free of predators, an ancestral homeland, even if, for four out of five, that ancestral stream was being channeled through a hatchery trap.
By October, a few coho and even fewer Chinook would make a right turn at the Willamette River, on the outskirts of Portland. They would swim upstream, through downtown, and some would turn left at a suburban gusher called Johnson Creek, which flows through a landscape of auto yards, tract housing, and homeless encampments. If they veered left again from Johnson, they would reach a rivulet called Crystal Springs Creek, near my own house.
Salmon had died out there in the 1850s, when the first mill dam went up. That dam was removed in the 1980s, but the salmon did not come back. Nor did they return during a decades-long effort on behalf of hatchery experts to plant sprats in the stream. But a few years ago, in 2014, a couple of strays—literally two—made the turn from Johnson Creek into Crystal Springs. The female was first spotted carving a redd in front of one of my neighbors’ houses. She had no reason to be there—she could not have been spawned in Crystal Springs. But the fish had colonized it anyway; not everything is determined by DNA. We must all have a taste for wandering, for finding a niche in new territory, for surviving changes. The male coho, long and scarred, his flesh beginning to decay, followed her around the weeds, patrolling.
Reed College, the city of Portland, the county, the state, the feds, and several salmon-recovery organizations spent the next two years tearing up the park through which Crystal Springs flows. Heavy machines ripped out narrow culverts and removed retaining walls. For half a mile the whole watercourse was de-channelized, so that the spring water could bend through the park, alternating slow curves and fast runs with rippling currents. Native wetlands were carefully replanted along the banks by teams of biologists and botanists. Dead trees were imported on trucks, ground smooth with machine tools to prevent children from getting splinters, individually numbered, and then lowered into the stream, creating artisanal eddies. Two million dollars later, the park reopened with a celebration featuring a giant fiberglass Chinook.
I go down there sometimes, hunting the fish with nothing but my eyes. My son joins me, often with a friend or two, and we wander the new boardwalk, looking. We follow the spring downstream, through the lawns, across roads, and past braying children. Once, in an eddy, my son found a bleached white carcass bobbing in the reeds. It was a coho—more bones than flesh. I had to wonder whether Boy Scouts, or professional “fish techs,” had intentionally planted this carcass in the shallows. Maybe it was a drop of phosphor for the ecosystem, or a biochemical signpost saying, Turn left here.
I haven’t seen a live salmon, yet. Sometimes I even walk the bank at night, just listening to the water as it forces its way down toward the sea. I don’t know how long I’ll have to wait. But something is coming.