To solve such problems, the engineers and scientists charged with managing the river have evolved a new approach: instead of trying to conquer nature, they are bold enough to think they can outfox it. Using a process called adaptive management, they hope to stave off or reverse damage done to the river and its life forms by manipulating flows from the dam.
Adaptive management, they say, rests on two core principles. First, complex systems are inherently unpredictable; it is impossible to know the consequences of various human actions. Second, the only way to manage complex problems is through a collaborative process in which everyone with a stake agrees to try new measures. When experiments fail, as some do, these stakeholders must stand ready to try something else based on common interests.
The technique is being tried in many ecosystems, including the Everglades, the Columbia River basin and the San Francisco Bay delta. But it is furthest along here at the Grand Canyon, said Dr. Carl Walters, a professor in the Fisheries Center at the University of British Columbia, who helped create the approach. "It's working there," he said, "because of a dedicated community of people willing to work together and a relatively simple institutional setting."
The Grand Canyon's adaptive managers — scientists and engineers who work for a variety of state and federal agencies — have their work cut out for them. The Colorado River ecosystem has changed radically since the Glen Canyon Dam was finished in 1963 with the aim of storing water for Colorado, New Mexico, Utah and Wyoming.
Until then, the river was notoriously erratic. In the spring, gathering snowmelt from its vast watershed could send 300,000 cubic feet of ice-cold water per second through narrow canyon walls — enough to cover a football field 10 feet deep every second. The flooding river carried enough sand to fill a football stadium to the rim every three hours. In summer, the water slowed to a trickle — perhaps 1,000 cubic feet per second — and warmed to 85 degrees. Great piles of driftwood accumulated along the river banks, along with tall sand dunes.
The 710-foot-high dam, at the northern edge of the Grand Canyon, is a plug holding back the world's longest reservoir — Lake Powell, extending 186 miles upriver. Water flowing through the dam is held to 31,000 cubic feet a second and never exceeds 48 degrees. It contains no sand, no driftwood and few nutrients.
Somehow no one realized four decades ago that the dam would profoundly alter the downstream ecosystem, said Dr. Barry Gold, chief of the Grand Canyon Monitoring and Research Center in Flagstaff, Ariz., where the river's adaptive management plans are coordinated. A lone civil engineer warned in 1971 that the river would eventually lose all its sand, he said, "but no one listened."
For a time, Dr. Gold said, the river seemed to hold up well. Eight species of native fish could still be found, their muscular bodies, leathery skin and tiny eyes adapted for life in the formerly erratic, silt-clouded river.
Wildlife officials stocked the river with nonnative rainbow trout, turning
the 15-mile stretch below the dam into a blue-ribbon trout fishery. Rafting and
camping exploded in popularity. Sandstorms disappeared. A nonnative tree called
tamarisk gave shelter to endangered songbirds like the Southwestern flycatcher.
The muddy waters turned into rippling sheets of emerald green with diamond
And the dam, which powered generators, proved to be a big moneymaker for the federal government. Because power demands varied, so did the amount of water allowed to pass through the turbines: from 5,000 to 25,000 cubic feet per second, in a single day. The flow sent sinuous tides pulsing down the river.
But by 1990, the dam's negative effects began to show, said Randall Peterson, manager of the federal Bureau of Reclamation's adaptive management division in Salt Lake City. The huge daily fluctuations in water flow were eroding the river channel and sandbars, making rafting difficult. Native fish were in decline. Wind had blown away high banks of sand covering archaeological sites. Rocky debris from side canyons was accumulating in the main channel, making the rapids more dangerous.
Alarmed, the federal Department of the Interior, which oversees the dam, decided in 1991 to curtail daily fluctuations. From then on, the amount of water released downstream could not vary more than 8,000 cubic feet per second in any one day. Power revenue fell $30 million a year. The daily pulses of water flattened out, making the river safer for rafting and fishing. A year later Congress passed the Grand Canyon Protection Act. Dr. Gold's research center was formed in 1995.
"We were asked to figure out how to use the dam to protect the river and associated resources, and make recommendations to stakeholders," Dr. Gold said. His center convened 25 interested parties, representing state and federal agencies, Indian tribes, environmental groups, recreation interests and power contractors, to discuss proposed solutions.
The center found that conditions on the river were deteriorating alarmingly. Whole beaches had disappeared. Four species of native fish had become extinct. An Asian tapeworm appeared; it now infects most native fish that survive. Rainbow trout, now spawning naturally in the wild, increased their numbers sixfold, so that some parts of the river contained 17,000 trout per mile. Steadier flows apparently increased their survival rates but reduced their food resources, so they became smaller and thinner.
In the spring of 1996, Mr. Peterson said, researchers tried out their first big experiment using the Glen Canyon Dam. For one week, they released 45,000 cubic feet of water per second, using special spillways. They figured the high water would lift sand stored on the bottom of the river and deposit it onto beaches.
While the experiment looked like a huge success at first, it quickly went awry. A year later, most of the sand was gone. "We made a huge mistake," said Dr. Theodore Melis, a sediment expert at the research center. The sand that built the beaches, it turned out, had come not from the river bottom but from existing beaches and eddies. Then fluctuating flows continued to erode sand as before. Two different experiments in 1997 and 2000 also failed to make beaches or retain sand.
Meanwhile, the rainbow trout continue to proliferate, said Dr. Lew Coggins, a fisheries biologist at the center. As many as a million rainbow trout are now in the river, eating midges, plants and possibly a native fish called the humpback chub. Ten years ago, some 8,300 adult chub lived in the river; today there are only 2,100 large enough to spawn. Biologists worry that this may not be enough to sustain the population.
If each rainbow trout around the mouth of the Little Colorado River ate just one baby chub a day when the chub are most vulnerable, Dr. Coggins said, the chub might not survive. Moreover, a second introduced species, brown trout, is lurking in the area, and "they are voracious predators," he said.
To address these two problems — too many trout and not enough sand — scientists recently proposed a new experiment to the participating stakeholders.
"You can't make new beaches unless you have new sand to work with," said Dr. David Rubin, a sedimentologist at the United States Geological Survey in Santa Cruz, Calif.
In this area, the Colorado River's main source for new sand is the Paria River, 16 miles downstream from the dam. It can dump millions of tons of sediment into the bigger river during monsoon storms in the late summer and fall.
"Ideally we'd like to send a pulse of floodwater through the dam immediately after a monsoon rain," Dr. Rubin said, to transfer the new sand onto beaches. But dam regulations do not permit experimental floods until January, after most of the new sand is already lost. The alternative, he said, is to keep flows very low after a big storm to preserve as much new sand as possible and then to release a two-day flood in January to capture it along hundreds of miles of shoreline.
To reduce the rainbow trout, managers want to run high fluctuating flows through the Grand Canyon for three months. This would strand and dry out trout beds, killing eggs and some large adults, without harming most other fish, which tend to dwell in side channels. Scientists also plan to capture trout at the mouth of the Little Colorado and examine their stomachs for chub (the remains will be dumped downstream).
"This is a bold experiment that can be stopped after a few years if it turns out the chub population is not improving," said Dr. James Kitchell, a professor of zoology and director of the Center for Limnology at the University of Wisconsin. "Regardless of that, a reduction in trout should improve the trout fishery by reducing crowding and producing more trophy fish in coming years."
"Of course," he went on, "the best way to disentangle the puzzle is to repeat the experiment with high numbers of trout and low numbers of trout. But in this case, we may not have the luxury of proper experimental design. Time may be short for the chub.
"We need to create a sense of urgency and pursue the most direct, immediate actions required to reverse the decline. That's the essence of adaptive management. You learn from experience, design new experiments, test a hypothesis and go from there."
Mr. Peterson says adaptive management works only if interested parties endorse the experiments. Now, a majority is behind the plan to reduce trout and use the monsoon flows to provide new sand to make beaches; the most vocal opposition comes from 25 licensed trout guides who generate $20 million a year taking 30,000 anglers from Lee's Ferry to good fishing spots below the dam.
"We disagree there are too many trout," said Terry Gunn, owner of the Lee's Ferry Fly Shop. "And there's no good evidence that rainbow trout eat chub."
The collaborative process can never please everyone, Mr. Peterson said. Dr.
Gold agreed. "It's a rugby scrum," he said. But he added: "Just
as the river channel, which is not seen from the surface, shapes the way the
water flows, the leadership skills brought to an adaptive management program
shape the science and policy that make up the program."