He Built 640 Rock Dams in a Dead Desert Creek — Then Oregon Came After Him

In 2007, Scott Campbell started blocking Oregon's dry creeks with rock from his own hills. The former veterinarian had sold a chain of 750 clinics and bought a ranch where, without beavers, the creeks had turned into deep ditches. Engineers estimated restoration at a million dollars per mile. Campbell decided to make do with excavators and homemade dams, without any permits. Within a few years, a creek that used to run dry had become a pond holding 11 feet of water. But instead of recognition, Campbell got a $10,000 fine. The state declared his 1,000 dams a threat to nature. In 2015, his ranch faced a new threat. A massive wildfire had already burned more than 5,000 acres and was racing toward Camp Creek. That's where more than a hundred of Campbell's dams stood. Now they had to prove what they were really worth. The answer depended on what Campbell had done with this creek eight years before the fire. When he first walked Camp Creek on foot, there was nothing there to save. The ranch lay in the Sylvie's Valley between the small towns of Burns and John Day, and no owner had lived there for more than 50 years. Together with leases on federal land, it came to 140,000 acres, about 40,000 of them his own. The creeks he found looked like trenches. One channel had cut into the valley floor more than six feet deep, deeper than a grown man is tall. Where there should have been wet meadow, sagebrush and juniper grew. In spring, snowmelt filled these ditches, and within a few weeks, the water was gone, leaving nothing for summer. The answer to who dug these trenches lay in the journals of the Hudson's Bay Company. In the 1820s, its trappers took thousands of beavers out of the Sylvie's Basin. Campbell himself estimates that over the valley's history, about 200,000 animals were removed, and today, there may be about 200 here. That's his own estimate, not scientific data. Without beaver dams, the water ran fast and cut into the bed, and the deeper the ditch got, the faster the water ran. Campbell came to a simple conclusion. If the trenches appeared when the dams disappeared, then something that works like dams could bring the valley back. Beavers wouldn't do it on their own, because a bare ditch has no willow for them to eat and build with. He started with what other restorers were doing at the time. Wooden posts, branches, woven structures. The first big flood ripped these structures out and carried them downstream. The wood couldn't stand up to the force of water that had been gaining speed in these channels for decades. A few low rock barriers had been on the ranch since 2001, before his time, and they survived the flood. Campbell looked at them and decided to build with rock. The excavators started taking rock, gravel, and soil straight from the ranch's hillsides. Campbell turned the sloping channel into a staircase. There was one rule. One dam for every foot of elevation drop. On steep stretches, that meant more than 100 structures per mile, one roughly every 50 feet. The valley determined the size of each dam. An average one took about 200 cubic yards of material. 20 loaded dump trucks. The largest exceeded a thousand cubic yards. A hundred dump trucks in a single pile of rock across the creek. Each one had to be placed so that spring water went over the crest rather than bursting through the base. One mistake in the calculations, and a flood would take out the structure along with the ones below it. Campbell deliberately didn't make the dams watertight, and from the outside, it looked like a flaw. Water seeped through the rocks slowly, so the creek didn't stop, it slowed down. While it stood behind each step, it pushed down and outward into the gravel of the valley floor, and the meadow got wet from below, from under the roots. Sand and silt settled behind each step, and the creek bed gradually rose back toward the old floodplain. For a rancher, Here that meant something simple. Hay grew on spring moisture that stayed in the soil, and by early summer, Campbell no longer had to divert water from the river onto his fields. Camp Creek became the main proving ground. On three and a half miles, Campbell built more than a hundred dams. When the ponds filled, the water in places reached eleven and a half feet deep, as high as the ceiling in an old farmhouse, in a creek that used to dry up every summer. The price was startling to anyone in river restoration. Engineered projects cost more than a million dollars per mile. Campbell came in under $20,000 because the rock was right underfoot and the equipment was already on the ranch. That difference would later become his main argument to the state, and the main reason the state didn't believe him. By 2012, the difference was visible in photos. A 2002 photo of Camp Creek showed sagebrush on dry banks. Three and a half years after construction, greenery lined the water. A year after that, the valley looked like a meadow. Neighboring ranchers started coming by to see it. Campbell expanded the work to new creeks and didn't file a single permit application. According to him, he simply didn't know a permit was required. In the summer of 2014, Campbell was preparing a gathering at the ranch he called the Frontier Oregon Summit. 80 invited guests, lawmakers, heads of business associations, the governor. On June 8, he sent Governor John Kitzhaber a memo with the expected outcomes. One point concerned his creaks specifically. Laws should allow environmental improvements quickly and without expensive, unnecessary permits. The governor replied that the guest list was excellent. Twelve days later, a letter came from Richard Whitman, the governor's advisor. Whitman had worked with state agencies for years and he wrote what he had heard from them. The ranch was doing work without even trying to get federal or state permits. The dams were blocking passage for red band trout and killing vegetation along the creeks. Whitman's position was stronger than it may have looked from the ranch. The rules on moving soil in state waters exist because for decades people filled in wetlands and wrecked rivers and then said they were improving them. The state had no way to tell a good dam from a bad one without seeing it. And here, on private land, stood hundreds of structures no inspector had ever seen. Kitzhaber didn't want to go while the issue was unresolved. That summer, the Department of State Lands signed a nine-page agreement with Campbell. Instead of civil penalties, he paid $10,000 to Oregon's school fund and committed to submitting a plan to correct the damage by January 2016. The dams stayed in place. The price was that the state, together with the ranch, documented every structure, and from that moment on, each one had either to allow fish passage or get an official exemption. The summit took place on June 22 and 23. A rare thunderstorm passed over the desert, but the tent Campbell had ordered held everyone. Guests were served goat barbecue. The governor flew in on his host's private plane, said little, and left early. Campbell didn't become any more restrained. On July 19, while the agreement was still being finalized, he wrote to the governor about the wildlife agency. If it got rid of two or three arrogant mid-level bureaucrats, eastern Oregon's economy would grow by 10 percent. Now the state had a description of every one of his dams, and Campbell had photos of a green valley. Neither side had measurements that would convince the other. Campbell let scientists onto the ranch. Oregon State University graduate student Caroline Nash and Forest Service hydrologist Gordon Grant installed instruments on Cottonwood Creek, weirs to measure flow, wells for groundwater, and cross-sections of vegetation. Part of the creek was deliberately left without dams as a control. In early 2015, Nash presented the first numbers for Camp Creek. Surface water had increased sevenfold. The band of streamside vegetation had roughly doubled in width. The profitability of the whole ranch had grown by 10% thanks to restoring a single creek. Campbell put his name on this report next to the scientists' names. Under the table was a footnote. There were no before measurements, so some numbers hadn't been verified in the field. Next to the questions about late-summer flow and groundwater, Nash put question marks. With these numbers, Campbell went to the legislature. The idea for the bill, according to its author, State Representative Brian Clem, was born at that same summit in June 2014. HB 3217 would have allowed a pilot program for these dams in the Malheur Lake Basin. Since 2010, Campbell had given Clem more than $26,000 for his campaigns. He told reporters that his own ranch gained nothing from the law and that he simply believed in these dams. In April, the committee approved the bill five votes to two. One of the two votes against came from Representative Ken Helm. Remember that name. On May 18, 2015, the bill was heard in the Senate. The Oregon Department of Fish and Wildlife came out against it. The agency evaluated rock dams in the Malheur Basin and stated that they wouldn't produce the effects of real beaver dams and would cause harm. That same day, Grant told senators about the observatory on Cottonwood Creek, where data was only just being collected. The House passed the bill, but it stalled in the Senate. Campbell's first attempt to change the rules came to nothing, and now his word would carry less weight than any agency number. A few weeks later, fire took up the argument on his ranch. In the summer of 2015, at the peak of fire season, a prescribed burn in eastern Oregon got out of control. The fire moved through dry sagebrush and grass and burned more than 5,000 acres, the area of a sizable city. Its front was rolling toward Camp Creek, where a few years earlier Campbell had built more than a hundred dams. It was the first creek where he had tested his method, and the first one he might lose. Seven years earlier, the floor of this valley had been the same dry strip of sagebrush and juniper as the hillsides around it, and fire would have run straight through it. Now, in August, there was a meadow on wet ground. The fire reached it, couldn't cross, and burned itself out. Acres of pasture beyond the valley and several buildings survived along with it. The mechanism is simple. Fire moves through dry fuel. Sedge growing out of wet soil in August is full of water, and the flames first have to boil off that water before the plant can burn. A grass fire doesn't have enough heat for that. Rock that Campbell placed to feed his cattle worked as a firebreak made of living plants. Here, I need to be honest about how we know this. The account of the fire was recorded by researcher Martin Doyle, and the ranch manager said in an interview with the Forest Service that in its former state, the valley wouldn't have stopped the fire, and it would have gone on for many more miles. I found no official report confirming the size of the fire or where it stopped. This is testimony from the people involved. The following summer, in August 2016, creeks on the ranch were flowing where they used to dry up every year. Redband trout were caught in the Sylvies River, in Camp Creek, and in Hay Creek, in stretches where the Fish and Wildlife Department's records had never documented them. Crayfish and freshwater mussels appeared. Campbell banned beaver trapping on the ranch, and the beavers began building their own dams right on top of his rock ones. The same trout the state had cited as evidence of harm in 2014 now became Campbell's main evidence. But all these observations came from the owners. What was needed was a measurement none of them controlled. By 2017, Campbell had a green valley, testimony about the fire, trout, and not a single piece of evidence the state would accept as its He told Forest Service researchers he wanted to keep building dams, but kept running into permits. The answer came from where no one was looking. From orbit. Since 1984, Landsat satellites had passed over the Sylvies Valley every summer and photographed it, long before Campbell bought the ranch. The before data missing from his report had been sitting in an archive all along. In 2018, Mark Hausner of the Desert Research Institute, together with Nash, Grant and others, published an analysis. They took four of the ranch's creeks – Flat Creek, Jack Andy Creek, and the upper and lower parts of Camp Creek. For each one, they compared two numbers for every year from 1984 to 2017 – how much rain and snow fell, and how green the strip along the water was in June, July, and August. Before 2007, these numbers moved together. A dry year meant a pale summer. A wet year meant a green one. The researchers drew a line through those years and looked at where the years from 2012 to 2017, after construction, would fall. On all four creeks, they fell apart from the line. The greenery no longer depended so heavily on the sky. The authors explained that the structures had raised groundwater, and the plants along the creeks were getting water they hadn't had before. For vegetation, this was now a proven result, not the owner's word. That same year, the very people Campbell had signed his first numbers with published a study that sided with his critics. The conservation group Water Watch and the State Water Department had long argued that water held on the ranch stays on the ranch, while neighbors downstream get nothing. Nash built a model that accounted for every drop in the meadow and the channel. It was tested on data from the John Day River Basin rather than the Sylvie's, but the question was the same. A raised creek did increase groundwater and the overall store of moisture. But that moisture's contribution to late summer flow turned out to be so small it couldn't be measured, because the plants transpire most of it. Nash's conclusion was blunt. If the goal is more water in the rivers in August, it's better to try something else. So, by 2018, the answer had come together in two halves. The first was Campbell's argument. The rock made the valley greener and you can see it from space. The second was his opponent's argument. Science didn't confirm the promise of more water for everyone downstream. And the 2014 agreement was still waiting for each of his dams. The 2014 agreement hadn't gone anywhere. In 2019, Campbell filed a request for 160 exemptions from fish passage requirements on four creeks, Payne, Jack, Andy, Flat, and Jump. Each dam on them had either to allow fish passage or get an official ruling that passage wasn't needed. Two more Two more attempts to change the law, in 2019 and 2020, also failed to become law. The fourth attempt passed. Bill HB 2298 allowed landowners to build such structures on small streams in closed basins east of the Cascades without a permit to move soil, provided the work was finished by July 1, 2031. Among its chief sponsors was Ken Helm, the same representative who had voted against it in 2015. On May 21, 2021, Governor Kate Brown signed the law. For Campbell's old dams, nothing changed. In January 2022, the Fish and Wildlife Commission considered a request for ten exemptions on Jump Creek and postponed its decision, asking for more data. The ranch's lawyer wrote that they had been waiting for approval for more than four years. Campbell had chosen Jump Creek for restoration himself. Back in 2009, before the first dams on it, the ranch had asked the department to consider stocking red band trout in the creek. At the end of August 2022, department inspectors went out to Jump Creek. Above the ranch, the creek was dry. The photos show a bare channel in August, with no water a fish could live in. Biologists walked the sampling points and recorded the result. No trout found at any of them. The water department also found something that worked against Campbell. The ranch was storing water from Jump Creek without a right to it, and that violation had to be corrected. On October 14, 2022, the Commission approved all ten exemptions. The reasoning was written up dryly. Passage through these structures would provide no meaningful benefit to native migratory fish. Eight years after Whitman's letter, the state officially acknowledged on one of the creeks that there were no fish there to block. On that same creek, it found Campbell holding Holding water he had no right to. Both sides turned out to be partly right. Nash's presentation includes two photos of Camp Creek taken from the same spot. In the first, from September 1, 2002, the banks are dry and covered in sagebrush, and the water is somewhere at the bottom of a ditch you can't see from the road. In the second, from September 10, 2012, ten years and more than a hundred rock steps later, the same valley is covered in meadow, with water visible through it. Today, the Forest Service counts about 640 structures on 18 miles of creeks on the ranch. In 2017, Campbell himself put the number at about 350, and in 2014, Whitman was told about 1,000, so everyone was counting by their own rules. The new law allows dams like these to be built until 2031. Campbell fought for this law for six years and it went to everyone except his own old dams. Every one of them still lives under the 2014 agreement. After that, the state will have to decide what to do with them next. The satellite shows that the valley has gotten greener. The model shows that this doesn't bring more water downstream. Who has the right to go first on someone else's creek? A man with an excavator or an agency that needs years to measure the results is still undecided. And in Montana, another millionaire bought the worn-out Flying D Cattle Ranch and gave up on cows. Ted Turner released bison onto that land, and what they did to the grass, the water, and the value of the land itself? None of his neighbors saw coming.