Great hydroelectric dams like the one above the Grand Canyon were once championed as the ultimate way to generate clean power. Yet we now know they wreak ecological destruction downstream. But can opening the dam undo the damage? Matt Kaplan reports.

In April of 1996, floodwaters raged through the Grand Canyon's Colorado River. Water rushed up beaches usually reserved for campsites, small trees drowned and rapids vanished from sight. For a week, the entire river was transformed into a turbulent monster.

Awe inspiring as the flood was, it was not a natural event. The water had been released on purpose from an upstream dam to reproduce the sort of flood the Grand Canyon would have experienced every year during winter and spring rains, before the Colorado River was dammed in 1962.

The reason for releasing 1,300 cubic meters per second over the course of a week was not nostalgia. Rather, it was an attempt to better understand the role floods play in maintaining the environment of the Grand Canyon. Now a second flood has been scheduled for this winter. This time, say researchers, they'll take what they learned from the 1996 flood and make a really good job of repairing much of the damage done to the Grand Canyon by the dam. That's if shortsighted politics doesn't prevent the flood happening at all, forcing research to grind to a halt.

Dams have been on the Colorado river for nearly a century. However, there were no dams upstream of the Grand Canyon until 1962, when the Glen Canyon Dam was built. At the time, the idea of damming the river above the canyon didn't alarm anyone: it was a dam-building era and flood control seemed like a good thing. Besides, the dam would generate cheap, clean hydroelectricity.

But by the late 1980's, Glen Canyon Dam's devastating long-term effects had become apparent. The Grand Canyon was suffering from a serious dearth of new sediment. River rafting guides found that beaches they were used to camping on had eroded to half their original size.

Meanwhile the National Park Service had noticed that Colorado River's largest fish, the Pike Minnow, had completely disappeared from the river system. The humpback chub, a fish native to and found only in the Colorado River system was seriously endangered.

At the same time, carnivorous fish like brown and rainbow trout had spread through the entire river system. These fish were alien species that had been introduced downstream every year by the fishing industry. Over time, the industry found they didn't need to stock the river as much. Today they don't stock it at all.

It wasn't just water species that were changing. A river tree called the tamarisk had spread up the whole river, ousting native plants.

The reasons for the geological changes were easy to identify. Before the dam was constructed, river water in the canyon carried so much sediment that the river actually had a cloudy rust color, giving it the name Colorado.

But with the dam in place the water was forced to come to a halt before flowing into the Grand Canyon. There over 90 per cent of the river's sediment was dropped behind the dam's massive stone walls, never to enter the canyon. Not only that, but rather than water flow alternating between high floods and a trickle, the flow rate was kept at a low steadily rate, so low that it couldn't pick up and circulate sediment from the river floor. Today the river runs crystal clear and baby humpback chub, who use cloudy water to hide from predators, are an easy target. Meanwhile the carnivorous trout can see easily to hunt down their prey.

The lack of sediment also explains the beach erosion. Beaches are dependent upon a continuous supply of fresh fine sand into the system. Without this able to pass through the damn, the only source are the small tributary rivers flowing unimpeded into the Colorado below the dam.

The dam has caused other geological problems. By limiting the river's flow rate to a steady trickle all year round, the dam ended the Canyon's seasonal floods. Rapids, which depend upon floodwaters to clear debris out of their white-water generating boulders, became choked.

This isn't only bad for river rafters. The stop and start in the water flow makes it hard for fish to move up and downriver. Researchers think this adversely affects chub, who like constant and turbulent waters, but trout thrive in clear and placid environments. "Chub decline over the past 10 years correlate with increases in the trout population, and both of these developments seem to link to changes in the dam's operation" says aquatic ecologist Michael Yard of the US Geological Survey.

In 1989, under growing public concern and pressure from environmental groups, the US Bureau of Reclamation sent a task force of researchers into the canyon to see what could be done. The result was an environmental impact statement highlighting a multitude of problems in the Grand Canyon ecosystem. In the statement, researchers hypothesised that sediment, deposited from tributary rivers, was collecting along the Colorado's bottom due to the restricted water flow caused by the dam. They followed this hypothesis with a stunning and at the time revolutionary idea. Why not try to reverse some of the decline by opening the dam for a short time and using flood waters to re-circulate this river bottom sediment? In April of 1995 a desperate Bureau gave the go ahead.

With just the punching of a few buttons the dam opened up its power plant and its non-energy generating outlet tubes to full capacity and, at the cost of two million dollars in lost electricity revenue, dumped 1,300 cubic meters per second of water into the canyon for a week. This was just under half the rate at which the river used to flood, but such a flood hadn't been seen in years. The National Park Service monitored the length of the Canyon and warned people entering the river area to be careful.

Geological researchers were hoping the flood might re-circulate fine sediment at the bottom of the river. Ecologists hoped the flood might also sweep away young trout, who aren't adapted for flood conditions, and clear the way for a chub recovery by reconstructing their native habitat. "It was an enormous science experiment," says Robert Webb, hydrologist at the US Geological Survey.

In the weeks after the floodwaters cleared, researchers were amazed by what they saw. The flood seemed to have restored the Canyon to a near-pristine state. Beaches that had been rocky and barren of sand turned into beautiful sandy hills and floodwaters brought river morphology back to the way it was in the river's pre-dam days. This was most clearly seen in the way the river's overall shape changed.

Canyon rivers like the Colorado, under natural conditions, have fast moving water along the river bottom that picks up sediment and distributes it in the form of beaches at its sides. This creates a characteristic V shape. With the introduction of dams, water slows down in the river and the V shape becomes less well defined since sediment is not as actively picked up from the bottom. Under dam conditions the river had degraded from a V to a U shape, something that would never have happened in the good old days. Historic huge floods are the reason why the Grand Canyon is a giant V shaped canyon.

But the 1996 flood pushed the river morphology back in time, shifting its form from a U to a V in just a matter of days. Beaches were reworked and sand bars were rebuilt, but the most impressive changes were seen in the rapids. The debris and sediment that had been choking them off was gone. "A lot of rapids were becoming un-runnable to the river rafters," says Webb, "the flood cleared a lot up." One in particular, the canyon's largest rapid, Lava Falls, saw a startling transformation. A logjam of debris that had been constricting its white water for over a year had been dislodged, increasing the width of the rapid by an average of 5 metres.

However, while all of the flood's positive effects have lasted with remarkable resilience since 1996, not everything worked the way the scientists hoped. Research lead by David Ruben and David Topping of the US Geological Survey's Grand Canyon Research Center in Flagstaff, Arizona, recently published in the journal Eos (reference?) shows that the badly eroded sandbars were not reconstructed from river-bottom sediment. Instead, the sand was coming from the base of the same sanbars. "Sand bars got smaller and higher," comments Topping. "It was a bit like using your credit card to bump up your bank account." The same applied to the rebuilt beaches. Sand bars and beaches are geologically similar except that a beach is above the water level and a bar just below it.

The flood's effect on endangered species was similarly lukewarm. While the chub were not harmed in any significant way by the flood, neither were the competing trout. And the tamarisk invasion was actually exacerbated by the floodwaters as they served to spread its seeds all along the canyon's banks.

The most important finding was that the basic idea behind the flood - that sediment accumulating at the bottom of the river could suddenly be used to rebuild beaches was possible. But this discovery was tempered by the fact that there was not nearly as much fine sediment on the river bottom as researchers' calculations had suggested there would be.

A possible explanation for this was that occasional dam releases during the 1980's and early 90's had flushed through fine sediment in the system. To researchers' surprise, it seemed possible that these rather small flows could pick up sediment and carry it downstream (eventually out of the canyon). But the water level was simply too low for the sediment to reach the tops of sand bars and beaches. "The only fine-grained material we got was fine sand, and that was just the small amount that we could draw out of eddies in the river. There just wasn't any left in the channel" says Kaplinski.

Other surprises were buried in the reams of data the flood generated. Photographs and measurements of beach and sandbar size suggested that most of the building was done at the start of the flood. After that, floodwaters seemed if anything to erode the beaches away again. Floodwaters are a "double edged sword" adds Topping. They are the only force that can get sediment from the river bottom and use it to build beaches, but like any fast flowing water, they also irrevocably move sediment downstream and eventually out of the Canyon. "Most of the major changes in the river's bars and beaches were made within the first 48 hours of the week-long flood," says Kaplinski.

Research, however, is about trial and error, and the canyon's geologists are using everything they learned from the 1996 flood to guide the next one. They now know that the next flood need only be about two days long for it to achieve its goals. They also know that the flood must follow a period of several months when the Glen Canyon Dam is kept at a very low output, so that sediment from downstream tributaries can build up.

This winter, researchers are going to try to use sediment injected into the system through natural floods on the Paria River and other tributaries that join the Canyon below the dam. To do this, they'll hold water releases from the dam at a very low rate all autumn. The intention is then to follow this period of sediment accumulation with a short, man-made, flood to do the beach and bar building.

Although researchers agree a manmade flood is the best way to fix the canyon's geology, they don't agree on the best timing. To ecologists, running the flood in winter seems to have a second advantage, of occurring during a time period when some invaders to the Grand Canyon area will be adversely affected. One of these invaders, the tamarisk river tree, will not be able successfully germinate seeds spread by the flood waters as it did in 1996 when rising waters coincided with the temperatures needed for the plant to reproduce. Although the tamarisk won't be driven out completely by the flood, it should be hit. But the discussion is complicated. The tamarisk, while possibly competing with native plants in the area, also helps an endangered bird, the South Western willow flycatcher, by providing it with a superb nesting habitat. In spite of their name, the birds seem to prefer tamarisk over willows. "Is a species automatically bad if it is non-native?" asks Webb.

As for the trout, aquatic ecologist Michael Yard is keen to follow up the winter flood with a series of high flows in the spring, when the trout are reproducing. These wouldn't be floods - they would go through the turbines instead of through the floodgates, but they would flush through more sediment that would usually be allowed from geological considerations. These flows should hit the trout during their sensitive spawning period, enabling chub, to hopefully recover. But geologists don't think repeated spring floods are a good idea, because they could erode away the fresh deposits. "It is possible that these spring flows could undo everything the winter flood builds," says Kaplinski.

Agreement between scientists is hard enough but the team that makes the decision on whether the dam should be opened includes representatives from Native American tribes, fisherman, river rafters, environmentalists, and of course power contractors. This team, more formally known as the Glen Canyon Dam Adaptive Management Program, recommends to the federal government how the dam should operate.

Not everyone on the team is backing the flood. The State of Colorado, whose rivers feed the Canyon, may try to fight any decision to flood in the courts, citing concerns over water conservation. This isn't backed by scientific thinking, but it's still not clear how a US administration with the stated aim of making the country more energy-independent will react to this pressure. Still Adaptive Management Program Manager Randy Peterson thinks sense will prevail. "We believe we can keep law suits out of the picture," he says.

The good news is that power companies who share in the dam's operation have agreed to the flood, despite the fact that they'll incur greater losses than in 1996. According to Peterson, the companies won't just lose around 2 million dollars during the flood itself. During the autumn, the reduced flow to save sediment will allow only a small portion of the hydroelectric generator to be used, at a cost of roughly 8 million dollars. If the smaller anti-trout flows go ahead in the spring, the company might be able to recoup some of its losses by running more water through than would normally be allowed (because of concerns over losing sediment) at that time of the year.

At the back of everyone's mind is the possibility that this flood could set a precedent. Jeff Mount, director of the Watershed Center at the University of California, Davis is watching the events particularly closely. "California has more than 1400 dams that provide electricity, water and flood control to 34 million people. Every one of those dams has altered sediment distribution. Dealing with this is going to be tough. We need experiments like the ones they are doing at the Grand Canyon to give us the courage to try our own," he says.

While this flood, like that of 1996, is still an experiment, everything that has been learned is making its chances of being a real scientific success much more likely. If the researchers are right and a few months restraint followed by a two day torrent can undo years of ecological destruction, environmentalists may clamour for floods to become a regular fixture, and not just at the Grand Canyon.

But perhaps even more complex is the scenario that will unfold if the researchers are wrong. If managing sediment injected into the system through tributary rivers does not work then researchers will have to turn their attention to another possibility - dredging the sediment that is caught behind the dam and dumping it on the other side, then flushing it through with a manmade flood like those being tried today. One concern is that this sediment is even more contaminated by non-native species than the Canyon. Geologists may have to consider the expensive option of bringing sediment in from elsewhere.

Ultimately many researchers think it's naive to imagine that huge power consumption can come without some environmental cost - whether that's to the geology or the ecosystem. In the end the only sure way to undo the damage may be to decommission the dam.