An hour’s drive east of Eugene, Oregon, Quartz Creek pours down the flanks of the Western Cascades, across a widening valley and into the McKenzie River. One morning last August, I stood on a bridge spanning the creek and watched thunderheads boil up over a distant ridgeline, trying to wrap my head around how this place became itself.
One version of its story goes like this: Some 12 million years ago, the Earth’s crust thrust upward from beneath a volcanic plateau. The plateau buckled, forming the rough shape of a mountain range. Over millennia, rain and ice sculpted this shoulder of rock, carving narrow canyons into steep terrain and carting the eroded sediment downstream to deposit across gentler slopes. In these broad depositional valleys, like this section of Quartz Creek, water spread across the land to create wetlands laced with branching channels. Chinook salmon, bull trout and Pacific lamprey hatched and grew in the slow-moving water. Some migrated to the Pacific and returned years later to spawn, bringing nutrients from the sea to nourish riparian forests of cottonwood, fir and hemlock. Frequent windstorms, wildfires and landslides toppled trees, strewing them across the valley bottom. Beavers moved the wood into dams, forming ponds and redirecting water.
All this may sound like chaos, but the relentless flux sustained a kind of stability, preventing any single channel from becoming dominant and maintaining a mosaic of deep pools and turbulent confluences, sandy bars and gravel beds, fast flows and slow side-channels. The landscape supported an equally diverse range of biota, which, in turn, supported people. Native tribes and bands, including the Kalapuya, Mollala, and Warm Springs, lived year-round at lower elevations and came here in summer to fish, hunt, and gather huckleberries and hazel.
In the mid-1800s, Euro-American settlers arrived. By 1860, the U.S. government had forcibly removed local tribes — whose descendants belong to the Confederated Tribes of Grand Ronde, Confederated Tribes of Siletz Indians and Confederated Tribes of Warm Springs — to reservations, and settlers began harvesting the forests. Quartz Creek, like many area streams, presented an obstacle, with its swampy floodplain and unpredictable flows. So the newcomers dug drainage ditches, built berms and raised roadbeds. Like wool spun into yarn, the creek’s many threads began to draw together. In turn, the valley was remade: The concentrated flow deepened its channel, while the rest of the floodplain grew drier. Fewer downed trees reached the water, and accelerated currents flushed sediment through the valley, leaving scarce calm pockets for fish. Lamprey declined. Chinook salmon and bull trout vanished altogether.
Transforming Quartz Creek’s unruly flow into something closer to a ditch must have entailed a lot of labor, something I found myself pondering as I stood on the bridge that August morning and watched another group of humans hard at work. With the help of modern technologies — excavators, LiDAR, GPS mapping — they hoped to undo the efforts of their predecessors and remake the valley, again.
The project is the latest phase of a river restoration effort begun a decade ago in the McKenzie River watershed. It uses a new approach called Stage 0, which aims to turn canal-like channels back into dynamic wetland-stream complexes by regrading parts of a valley’s floor. It’s a bold process, requiring a tremendous disruption of the existing landscape. Studies of the long-term effects of this approach have not yet been completed, and the project’s scale is unprecedented. As geomorphologist Gordon Grant put it, “It’s a full-on field experiment.”
The McKenzie River is beloved by boaters, anglers and environmentalists alike, and many are uncomfortable with using it to test this highly invasive restoration technique. Others see it as a chance to repair some of the damage done to the watershed, before it’s too late.
The McKenzie River once supported some 110,000 Upper Willamette River chinook salmon, a threatened native fish integral to the tribal cultures and ecological health of the Columbia River Basin. Now, the McKenzie’s population is just under 2% of that historic abundance — and it’s the largest remaining wild population, our best hope for the species’ recovery. Without intervention, some analyses predict the population could be extinct by 2050. “We know we don’t know everything,” Elizabeth Goward, community engagement manager for the McKenzie River Trust, told me. “But if we don’t act now, we could lose this species.”
Images of Quartz Creek made in 2025 (left) and in 2026 (right) show how quickly the river basin was transformed following the Stage 0 project restoration. Sarah Koenigsberg/High Country News
BRIAN CLUER STARTED FLYING airplanes as a teenager. He spent hours in the cockpit, looking down at the land surrounding his Idaho hometown. So perhaps it’s no surprise that he became a fluvial geomorphologist, someone who studies how rivers and streams sculpt the Earth’s surface.
As a student, Cluer learned the principles established by pioneers in the field — including Luna Leopold, son of famed conservationist Aldo Leopold — who had studied river systems in the mid-Atlantic. They had concluded that undisturbed streams resembled meandering single channels, an archetype that would inform both restoration goals and the collective imagination: When we picture a pristine river, most of us see a winding ribbon of clear water between defined banks.
In the ’90s, an enterprising hydrologist named David Rosgen began championing a restoration method called Natural Channel Design (NCD), which used formulas based on these seminal studies and exhaustive site assessments to attempt to determine how an impaired stream could be reshaped into its natural stable form. At the time, federal regulations had begun allowing developers to “mitigate” negative environmental impacts by restoring habitats elsewhere, setting river restoration on track to becoming the multibillion-dollar industry it is today. NCD was widely embraced by private companies and public agencies, including the Environmental Protection Agency. Though Rosgen’s designs were intended to allow for some movement over time, his disciples often hardened banks with riprap and boulders, locking the channels in place. Some of these projects became infamous failures when floodwaters destroyed the fixed forms. Overall, Cluer said, many NCD projects fell short of their ecological promises.
“When you think about rivers in that way, you always get a wiggly, single-thread channel,” Cluer told me. During his decades of flying, he’d come to believe “there was something in nature much more wild and broad and spread out and chaotic and undefined than that.” He’d noticed remote river valleys that lacked defined channels and looked more like big wetlands. “That got my creative thoughts going,” he said.
In 2008, two geologists published a paper in Science affirming Cluer’s hunch. The authors revisited the mid-Atlantic streams studied by early geomorphologists and found they were not, as assumed, “natural.” Instead, they’d been shaped by mill dams, thousands of which were constructed across the region by Euro-American settlers beginning in the late 1600s. Though no longer visible, these dams had significantly altered valley floors. The single-channel archetype, the authors concluded, was an artifact of human manipulation.
Meanwhile, researchers in the Western U.S. and elsewhere were digging deeper into historic accounts and using new technologies, such as LiDAR, to better understand a given landscape’s particular history. Though circumstances differed from place to place, their conclusions were similar: Many precolonial streams likely looked less like winding ribbons and more like multi-threaded wetlands. Combining this growing heap of research with their personal experiences, Cluer and another fluvial geomorphologist, Colin Thorne, started developing an updated model of stream evolution. Theirs began not with a single channel but with a wet valley floor webbed with streams. They dubbed this new starting point Stage 0.
WHILE THORNE AND CLUER were developing their theoretical framework, Kate Meyer, a fish biologist then with the Forest Service, was tasked with improving fish habitat in the McKenzie watershed. She and her team sought to restore a tributary called Deer Creek. Like Quartz Creek, it had endured decades of logging and “stream cleaning,” a misguided effort in the 1960s and ’70s to improve river health by removing logs and debris. Wood was all but eliminated from the system, and berms confined the stream to a straightened, high-velocity channel — “essentially a firehose,” Meyer said.
Working with the McKenzie Watershed Council, Meyer’s team planned to restore the creek by adding wood to the channel, using logjams to slow water and trap sediment. “It was what everyone was doing at the time,” she said. But the results were consistently underwhelming. High-energy streams often washed out the wood, and even when logjams remained, sediment accumulation could take decades.
There was, Meyer knew, another way. In eastern Oregon, her colleagues had been trying a new approach: Rather than working to improve existing channels, they were getting rid of them altogether. The idea had arisen in 2002, when Paul Powers, another fisheries biologist with the Forest Service, visited an NCD restoration project in the Siuslaw National Forest that had recently been disrupted by a landslide. Initially, project leaders viewed this as a catastrophic setback: the dirt had ruined the channel they’d designed. But when Powers visited, he found that the slide had dispersed the stream across the valley bottom, creating slower flows, increased wetlands, and plentiful fish rearing habitat. He began trying to replicate this outcome at Whychus Creek in central Oregon, intentionally directing flows out of the channel and into the floodplain.
In 2014, Meyer attended a river restoration symposium where Cluer and Thorne presented their new paper on Stage 0. “It was a total epiphany moment, to see the concepts we were working with as practitioners described from the theoretical perspective,” she said. When Powers joined the Deer Creek team in 2016, he suggested they try it there. Meyer was excited, but also nervous. “I thought, ‘You mean we’re just going to bury the stream?’”
“We know we don’t know everything. But if we don’t act now, we could lose this species.”
They started small, dismantling levees and using the material to fill sections of the channel. The stream immediately spread across the floodplain, creating multiple slower-moving channels and deep pools. In the unfilled reaches, by comparison, little changed. When Meyer and Powers invited Cluer and Thorne to visit their projects, the researchers were stunned to see their theories enacted in practice. “I thought, ‘Oh my goodness, these people are actually doing it,” Thorne told me.
Emboldened by the support of Thorne and Cluer and the improvements in Deer Creek — in 2017, chinook salmon were found spawning in the creek for the first time since 1993 — Meyer and her team were eager to try a larger project. In 2018, they began work on the South Fork tributary of the McKenzie, planning to restore a 200-acre stretch to Stage 0 conditions. By then, practitioners like Meyer and Powers had developed a working methodology.
The first step, which may also be the most contentious, is to identify an appropriate site. Stage 0 is suitable for low-gradient, historically depositional valleys where streams can spread across their floodplains without disturbing infrastructure. But how much gradient is too much remains debated, and understandings of landscape histories are ever-evolving. Next, practitioners use clues such as relic wetlands or stands of old-growth trees to approximate the valley floor’s shape before it was altered by settlers. Using LiDAR, they map the precise present-day topography, compare it to the target shape, and create a grading plan. Fish are trapped and relocated downstream, and the river is temporarily diverted into a side channel. Then, operators use bulldozers and excavators to reshape parts of the valley floor, filling channels and removing levees. Logs and woody debris are arranged across the floodplain, some partially buried and others left to move freely. The wood both creates habitat and slows water, functions that are crucial as vegetation regrows. Lastly, the diversion is removed and the stream is released to disperse across the valley floor, where it begins the work of rebuilding the riverscape.
LIKE MOST RUTS, channels are hard to get rid of, and the process isn’t pretty. When I visited Quartz Creek in August, the landscape looked downright devastated. A muddy stream flowed alongside hundreds of acres of dusty soil strewn with dead wood: giant logs, tangled branches, heaps of slash. “People say, ‘This isn’t Stage 0, it’s Ground Zero — it looks like you nuked the place,” Thorne told me. But Lara Colley, a local resident and the floodplain restoration projects manager for the McKenzie Watershed Council, was smiling proudly when she met me on site. In an orange vest and hard hat, she fanned her arm toward a mostly empty staging ground. “They’re all gone!” she said, meaning the logs. Until a few weeks ago, some 6,700 logs and pieces of wood had been stacked here; now, they were distributed across the floodplain. Colley, who dressed as a log last Halloween, had amassed the wood from Forest Service and Bureau of Land Management lands where trees had been thinned for wildlife habitat or removed after recent wildfires. “I could’ve looked at a log and told you where it came from,” she said.
Since the Forest Service and the McKenzie Watershed Council began collaborating on Stage 0 projects in 2016, the Eugene Water and Electric Board (EWEB) and the McKenzie River Trust have joined the project’s leadership, and several other people and organizations have contributed to the effort. Each collaborator brings different perspectives and resources, Goward told me, enabling the work to persist despite recent federal layoffs and budget cuts. “I like to describe it as an ecosystem,” she said.
For EWEB — a public utility that supplies drinking water from the McKenzie River to 200,000 people in the Eugene metropolitan area — stream restoration is part of protecting water quality. “Quartz Creek has always been our chocolate milk,” said EWEB’s Water Resources Supervisor Susan Fricke. During high flows, the creek often ran brown with sediment, taxing EWEB’s filtration systems. Spreading the flow across the floodplain will allow sediment to drop out before it reaches the mainstem, saving the utility the cost and chemicals associated with removing it. “We consider the river part of our infrastructure,” Fricke said during my visit to Quartz Creek. “Preventing the problem is so much better than dealing with it later — we’re helping protect our future selves.”
“I like to describe it as an ecosystem.”
The natural resources department of the Confederated Tribes of Warm Springs has provided feedback on the project. “Stage 0 offers a holistic view of restoring river wetland corridors that reflects the Tribes’ goals for creating sustainable fisheries populations,” tribal fisheries biologist Logan Bodiford said in an email. “Our hope is that this project will better enable tribal members to exercise their treaty rights and access culturally significant resources.”
The design for Quartz Creek was led by Meyer, who left the Forest Service in 2025 to co-found a restoration consulting company. Franklin-Clarkson Timber Co., the private timber company that owns most of the land around the creek, provided access to it via a 50-year stewardship easement. The National Oceanic and Atmospheric Administration funded most of the $9.5 million project with a $7.6 million grant (made possible by the Infrastructure Investment and Jobs Act). And the actual dirt moving and log placement was done by Haley Construction, a family-run heavy construction company based in Lebanon, Oregon, which also holds the state record for hauling the longest log, a 135-foot behemoth.
With less than three months of dry weather to complete the project and so many moving parts — 20-some crew members, 11,200 cubic yards of slash, acres of dirt, a flowing river, all those logs — running the construction site was not easy. “It’s like directing an orchestra, getting everyone working together in a timely fashion,” Randy Haley, co-owner of Haley Construction and son of its founder, told me. “And in harmony!” Ashley Haley, project manager and Randy’s daughter, added, laughing. Restoration projects are demanding, she said, but “it’s very rewarding to do something that benefits the community and wildlife.” Many on the Haley crew agreed. “One man recently retired after 47 years of working for us,” Randy told me, “but he comes back to work every summer just to be part of these projects.”
When Randy’s parents started out in 1958, they worked primarily in timber, building roads and hauling logs. Over the past seven decades, the company’s focus has mirrored shifting societal priorities: The Haleys have put in bridges, built dams, taken out dams, and, for the past 35 years, undertaken an increasing number of river restoration projects. “You have to be able to adapt to changing needs, to reinvent yourself,” Randy told me.
“I thought, ‘You mean we’re just going to bury the stream?’”
Today, Haley Construction uses its logging expertise to repair some of that industry’s damage. “The knowledge of how to work with wood, in forests and around waterways, all that now lends itself to floodplain restoration,” Ashley told me. “But we can’t condemn the loggers,” Randy added. “They were doing a job they believed was right, at the time.”
Watching the Haleys’ machines scrape at a bald valley floor that had, not long ago, been a green riparian corridor, I found it hard not to wince. It didn’t help that as far as I could see in most directions, the mountainsides were cloaked in standing dead trees: More than 173,000 acres surrounding Quartz Creek were severely burned in the 2020 Holiday Farm Fire.
This fire, it turned out, played a key role in advancing the Stage 0 work in the watershed. For one, it provided ample available logs. It also made the work more palatable to some onlookers. “It’s easier to bring heavy equipment into a scorched valley, harder to drive a bulldozer into a beautiful second-growth forest,” said Goward. But perhaps most significantly, the fire illuminated one of Stage 0 restoration’s most compelling co-benefits. The blaze burned through the 200-acre restoration site on the South Fork of the McKenzie, at the time the largest Stage 0 project ever implemented. Preliminary observations indicate that while unrestored areas suffered uniform, severe burning, the restored region burned in patches, allowing wildlife to take refuge during the blaze and the forest to recover more quickly. In some parts of the restored reach, the wide expanse of water functioned as a fire break. “We didn’t expect this to be part of fire resiliency,” Fricke said, “But it was.”
Over the course of my afternoon at Quartz Creek, the thunderheads had drifted closer, and as we left the job site, they dropped fat shadows on the ravaged valley floor. Fricke pointed to a line in the dust: bobcat prints. The trail led to the water’s edge, then vanished.
WHILE IT CAN BE HARD TO WATCH the construction of a Stage 0 project, the concept’s appeal can be equally hard to resist. Faced with the mess we’ve made of ecological relations, who hasn’t longed for a fresh start? This approach, down to its nomenclatures (Stage 0 or, as it is sometimes called, “valley reset”), seems to promise just this: an opportunity to return to the beginning — to before the beginning. A chance to shake the Etch-a-Sketch and start anew.
Advocates and critics alike caution against this framing. “We don’t expect to put everything back to the way it was before Lewis and Clark,” Thorne told me. “What we’re doing is empowering nature — by which I mean birds, amphibians, trees, plants, bacteria, everything — to get to work on the riverscape again, to be able to make and remake it continuously.” The result, advocates believe, will be an increased diversity of habitats and biota that will strengthen the watershed’s resilience to new climate extremes. “Will it come out like it did before? Probably not,” Thorne said. “It’s a different world now, a different river, a different catchment.”
Critics argue that historic Stage 0 landscapes aren’t just impossible to recreate, but, in most places, likely never existed at all. David Rosgen, now 84 and still involved in implementing NCD projects around the country, believes the web-like stream networks described as Stage 0’s starting point occurred only in extremely low-gradient valleys and deltas. In those landscapes, he told me, Stage 0 restoration can work. “But a good idea applied as a universal solution is a bad idea,” he said. He believes that places like Quartz Creek and the South Fork of the McKenzie are too steep to have ever maintained wetland-stream complexes, and thinks these rivers would be most stable and ecologically beneficial as meandering channels.
With its emphasis on allowing natural processes to shape streams, Stage 0 runs counter to Rosgen’s method. But Gordon Grant, who recently retired from a 40-year career as a research hydrologist with the Forest Service’s Pacific Northwest Research Station, sees similarities in the rush of enthusiasm for each. “There’s a particular bandwagon effect that seems to associate with restoration,” he told me.
“It’s a different world now, a different river, a different catchment.”
Grant spent his career studying how Western Cascades streams respond to logging, dams and climate change. “If there is any river system on Earth I have any claim to even modestly understand, it is the McKenzie,” he told me. So when he learned of the Stage 0 work there, he began looking into it. “I’ve never tried to restore a river, and have nothing but admiration for those who do,” he said. “But I don’t necessarily worship at the same church as the restoration community.”
A self-proclaimed science geek, Grant describes the landscapes being created on the McKenzie as “novel geosystems.” “There’s nothing in the history of these creeks that looks like that,” he said. “I like experiments; it’s how we learn things, how we get better.” But experiments, he told me, warrant careful study before widespread implementation. From “stream cleaning” to NCD projects gone wrong, there’s no shortage of cautionary examples, he said. “So let’s stand back and ask: What potential risks are being set into motion here?”
Grant, who authored a paper titled “When do logs move in rivers?” is especially concerned about what a severe flood might do to the large wood used in these
projects. Mobilized logs can wreak havoc on infrastructure — bridges, dams, docks, embankments — and endanger boaters and swimmers. Though Stage 0 projects are designed with grate-like logjams intended to trap wood within the restored reach, the systems are far from foolproof, especially in high-energy mountain stream systems like those in the McKenzie watershed. “The potential for mischief has not been fully reckoned with,” Grant said.
When I visited his office last winter, Grant stood at a whiteboard and, in a valiant attempt to explain fluvial dynamics to a journalist with a thin physics background, painstakingly drew out basic equations: Q (flow) = Velocity x Depth x Width. I did my best to follow along, but my eyes were drawn to a poster taped to the side of a file cabinet: an outline of a face with the words “Bang Head Here.” Still, I understood enough to get Grant’s point. Floodwaters are shockingly powerful, capable of lifting enormous logs and tossing them downriver as if they were pool toys. When the last major flood hit the McKenzie, in 1996, Grant was there to watch. “The stream you visit at low-flow, moderate-flow, even big winter flow, is nothing like what you see in an extreme flood,” he said. “And a 100-year flood means each year we have a 1 out of 100 chance it will happen. That’s a significant risk.”
There are, of course, risks involved with leaving things as they are — extinctions, worsening wildfire impacts, diminished water quality. Which risks are acceptable, Grant pointed out, “has a lot to do with who’s sitting at the table.” On this project, he told me, “it’s mostly people trying to make the world better for fish.”
A FEW DOORS DOWN from Grant’s office, I met Rebecca Flitcroft and Brooke Penaluna, two research fish biologists with the Forest Service’s Pacific Northwest Research Station. Though salmon drive the Pacific Northwest’s restoration economy, Flitcroft told me, “The biggest gap in the literature on Stage 0 is actually around the question of: What does this do for fish?” Fish are notoriously difficult to monitor, so researchers often assess impacts by measuring changes to habitat. “The assumption is: if you build it, they will come,” Flitcroft said.
So far, research suggests that Stage 0 projects can, in fact, build it. A study of
17 sites across Oregon and Washington found that Stage 0 restoration increased low-
velocity rearing habitat, broadened the wetted area of valley floors by several factors, and increased the overall production of macroinvertebrates, essential components of salmonid food webs.
These shifts may benefit not only salmon but Pacific lamprey, a native fish of particular cultural significance to Indigenous communities. Historically abundant in the Columbia River Basin, lamprey were an important food source for many regional tribes, but their populations have declined dramatically. Lamprey share many habitat needs with salmon and, like salmon, provide vital ecosystem benefits. Larval lamprey filter-feed in river sediments for up to 10 years, purifying water and cleaning gravel beds in the process; adults transport marine nutrients to freshwater creeks. Until recently, however, lamprey have received little attention from non-Native conservationists.
Not all the findings of the Stage 0 study were glowing. Salmon need cold water, and researchers found that temperatures tended to rise after restoration. Sediment composition shifted from coarse to fine — which can be great for lamprey, but can clog salmon gills and fill in the gravel beds needed for spawning. eDNA analysis showed increases in overall aquatic biodiversity, which includes not only native species but invasives. “When you open up a channel, you open it up to everybody,” Penaluma said.
Still, the biggest question is what happens in the long term. Though a Stage 0 project can be constructed in just a few months, the real work of restoring the river begins only after the excavators depart and the water returns. Will these sites cool as shade trees regrow? Will the composition of sediment shift? Will the logjams stay put?
“The assumption is: if you build it, they will come.”
Luke Whitman leads the Oregon Department of Fish and Wildlife’s effort to monitor changes in Upper Willamette River chinook populations over time. Immediately after the South Fork Stage 0 project was completed in 2018, he told me, the number of spawning beds skyrocketed from 44 in 2018 to 272 in 2019. By 2025, the count had fallen to 58. “They’re still slightly above pre-restoration numbers, but the higher levels haven’t been maintained the way we’d hoped,” Whitman said. The reason is unclear, but he suspects it’s due to the Cougar Dam, which lies upstream of the restoration area and prevents scour flows, the floods that historically rearranged sediment and vegetation. “I don’t think we’re getting enough water to keep some of the new channels active, to keep moving things around,” Whitman told me. Still, he believes Stage 0 is a worthwhile experiment. “We’ve got to get creative wherever we can on the McKenzie. Not just with restoration, but with dam management, too,” he said. “Wherever we can take a shot, we should try.”
IN THE LAST WEEKS OF 2025, heavy rains drenched the Western Cascades. In Deer Creek, powerful flows moved wood throughout the restored reach. “It was both very exciting, because it was the most change we’ve seen on any project yet, and at the same time concerning, because we ended up with longer stretches where wood moved out,” Meyer told me. Without logjams to slow water in these places, the river could start down-cutting into a single channel again. In an ideal system, the deluge that washed out the logs would also bring in wood from the surrounding forest. But here, where the catchment has been logged for decades, downed wood is scarce, and the trees are much smaller.
“Originally, we thought it would just take one intervention and then you could walk away forever,” Meyer said. But river systems are nested inside larger systems, and many of the processes involved — wood and gravel recruitment, flood scouring — are still impacted by logging and dams. Restoration alone can’t fix all the processes, Meyer told me. “So we need to acknowledge that, and think more about long-term stewardship where we monitor and manage these sites over time.”
Thorne agreed. “There aren’t any one-and-dones for rivers,” he said. But he cautions against rushing into action. It’s hard to break a Stage 0 project, he said. “It can be rearranged, the wood and sediment moved around, but nature will repair it over time.” After a moment, he added, “Or it won’t. And the creek will be set on a different trajectory than the one we had in mind.”
IN LATE JANUARY, I returned to Quartz Creek with Goward. Bright sun slipped between clouds, warming the day to an unseasonable 60 degrees. Snow covered distant peaks, but only a dusting powdered the nearer ridgelines.
I stood again on the bridge spanning the creek. The valley still looked disheveled, piled with tangles of logs and mounds of slash, but the dusty wasteland of last August was replaced with flows of clear water. Braided streams curled across the valley, parting around logjams and lapping at mounds of newly deposited sand.
Watching the creek, I thought of something Gordon Grant told me. He’d turned from the equations scrawled on his whiteboard and said, “You can’t model something like Quartz Creek.” Because of the high-energy flows and the complexity introduced to the system — the unprecedented amount of wood, the intersecting paths of flow — “it’s beyond the capacity of our hydraulic computational fluid dynamic models. … It’s unpredictable.” He’d meant the words as a warning, which they certainly are. But they’re also a promise. Unpredictability, after all, is another word for possibility.
Humans can’t control the outcome of a Stage 0 project any more easily than we can disentangle ourselves from it. Instead, the process requires that people participate alongside a host of other actors — trees and rain, stones and fish, beavers and mayflies — allowing unforeseeable interactions to shape the future river. Here, vulnerability and hope are entwined.
“You can’t model something like Quartz Creek.”
Except for a few scattered firs, nearly all the trees stood leafless. Some had lost their foliage for winter, but most were dead. With snowpack lingering at a record-breaking low, the next wildfire season was already
looming.
After the Holiday Farm Fire, Goward told me, locals were devastated: “People looked around and thought, ‘This place will never be the same again.’” Below us, the stream rippled over gravel, and I could hear the pebbles — pulverized bits of the volcanic plateau that once lay here — clinking against one another. At the water’s edge, blades of new grass emerged through heaps of slash. “Everything around us is changing,” she said. “What we’re trying to do is restore the river’s ability to change with it.”
Sarah Koenigsberg is a filmmaker, photographer, and science communicator whose work is grounded in the West.
This story is part of High Country News’ Conservation Beyond Boundaries project, which is supported by the BAND Foundation and the Mighty Arrow Family Foundation.
We welcome reader letters. Email High Country News at editor@hcn.org or submit a letter to the editor. See our letters to the editor policy.
This article appeared in the June 2026 print edition of the magazine with the headline “The way of the river.”
Republish our articles for free, online or in print, under a Creative Commons license.
