In 2000, the Corps tapped Dave Fuller and Pat Braaten to lead an experimental effort to help that agency comply with the Endangered Species Act and boost the species' prospects in the upper river. Scientists have long suspected that channelization and bank stabilization on the Lower Missouri, and dams and reservoirs on the Upper Missouri, were responsible for the species' decline. But the details -- the exact whys and hows -- were still uncertain.

Braaten had been working downstream in Missouri and Kansas on another native fish study. He grew up fishing in Minnesota and by high school knew that he wanted to be a fisheries biologist. His mother remembers him cutting open his catches to see what the fish were eating. "Now, I get paid to do that," he chuckles.

Fuller's road to fisheries work was less direct. He started out in engineering, following in his father's footsteps. But the idea of working with fish lingered, and in the early '90s, he moved from New York to Bozeman, Mont., chasing dreams of blue-ribbon trout streams. "I told my folks the move was for school, but it was for fishing," he confesses. After taking a year off to do Westslope cutthroat trout habitat work, he decided to switch careers and eventually ended up in Fort Peck, studying sturgeon, catfish, suckers and chubs.

The two biologists soon discovered they had much in common. "We hit it off really well," Braaten says. Both in their early 40s, they dress alike -- T-shirt, ball cap, jeans, boots -- share a love for bird-hunting and the outdoors, and even speak in the same cadence, their voices equal parts dirt and drawl. But it's their differences that make them a good team: "I'm more the thinker, the idea man," says Braaten. Fuller is "extremely good at making things happen in the field."

The Corps' plan when the two teamed up at Fort Peck was to help restore key aspects of the river's ecology that had been lost to dams, and see how the fish responded. The Fish and Wildlife Service, which drafts the biological opinion telling the Corps what recovery measures to implement, recognized that the fundamental aspect of all rivers is the seasonality of flow. If sturgeon did well before people altered things, then adjusting the dams' releases to better mimic natural conditions would, in theory, enhance spawning and juvenile development for pallid sturgeon and other native fish species.

They suspected that water temperature might be limiting reproduction and the survival of hatchlings. At Fort Peck, the dam releases water from the reservoir's chilly bottom, making the river much colder than it would be naturally. The Fort Peck Flow Modification Project, or "Flow Mod," would address this with two test releases of warm surface water over the spillway in 2001 and 2002, followed by spills once every three years if reservoir levels allowed.

At the time, this sort of thinking was taking hold on dammed rivers worldwide. In the 1990s, says Brian Richter, a freshwater conservation expert with The Nature Conservancy, scientists learned a lot about how important natural fluctuations, temperatures, and the like, were in maintaining river systems and native species. In the last decade or so, this knowledge has begun to influence management, with numerous experiments under way to "renaturalize" rivers without removing dams. The Grand Canyon has become a laboratory for some of the most high-profile of these experiments in the U.S. (see sidebar page 15). But there are plenty of obstacles: funding, politics, the limitations of existing infrastructure -- and on the Missouri, nature itself.

As it turned out, Flow Mod was not so much born as stillborn. Low water levels in the reservoir behind the dam prevented the Corps from conducting test spills in 2001, 2002 and 2003. The region was suffering from severe drought, making it technically impossible to spill since the reservoir was 30 to 50 feet below the spillway.

The delays had a silver lining, though. They allowed Braaten and Fuller to collect baseline data on the pallid population under normal dam operations, so that when the spills finally came, they would have something with which to compare the fish's response.

They also began investigating the mystery of the pallids' missing offspring. By sampling the stomachs of potential predators, they eliminated the possibility that ravenous fish, not dams, were to blame. This brought them to the most popular theory: that the distance newly hatched pallids, or "free embryos," need to drift while they develop into larvae and gain the strength to swim and forage exceeds the amount of free-flowing river between Fort Peck Dam and Lake Sakakawea -- the vast reservoir that Garrison Dam created downstream. Scientists thought the drifting embryos might be reaching the reservoir's slack water too early and getting gobbled by lake fish or sinking into the reservoir's oxygen-poor bottom layers and suffocating.

In 2004, Braaten and Fuller released thousands of hatchery-reared embryos into a side channel near Culbertson, Mont. They followed them downstream and re-collected what they could at different time intervals. The results were deflating. "It looked like even if we got pallids to move up the Missouri, there wouldn't be enough drift distance to make a difference," Braaten says.

But the release was the biologists' first crack at things, and because it took place in a side channel rather than the Missouri mainstem -- the far more complicated and dynamic part of the river where embryos naturally drift -- the results, says Braaten, didn't mean much.

This new information, however, renewed interest in modifying Intake Diversion Dam 70 miles up the Yellowstone to allow fish passage, opening up more than 150 miles of new habitat to native fish. If sturgeon used the passage, the added mileage might give embryos enough time to develop before hitting Lake Sakakawea.

In the meantime, Braaten and Fuller examined drift in the mainstem Missouri. In 2007, they released new findings: Immediately after hatching, pallid embryos drift for up to 14 days. Braaten's models showed that the slowest embryos would travel between 159 and 230 miles before they developed. But there are only about 211 miles of free-flowing river between Fort Peck and Lake Sakakawea. If spawning took place near or in the Milk, a turbid tributary that joins the Missouri 11 miles below Fort Peck, the slowest drifters might settle out of the current before hitting the reservoir. Most, however, would end up smack in the middle of it.

That same spring, Fuller discovered that two females he'd been tracking up and down the lower Yellowstone had spawned -- the first documented instances of pallid spawning in the Upper Basin. It was a big step forward, one that allowed researchers to cross spawning off the "reasons-for-recruitment-failure" list.

Later that year, thanks in part to Fuller's spawning discoveries on the Yellowstone, the pair's disheartening drift results on the Missouri, and the region's relentless drought, which continued to preclude spills from Fort Peck, the Corps obtained the legal authority it needed to partner with the Bureau of Reclamation to work on Intake, which the Bureau operates. It helped that the fish Fuller followed that spring had gone all the way to Intake, says recovery chief Jordan. "Folks could now say, 'OK, we stopped a migrating, reproductive female from going up the Yellowstone (past Intake).' " On top of that, there were Braaten's drift models, and an earlier study proving that juvenile pallids stocked above Intake were alive and well, suggesting the river there was suitable habitat. "You had three independent studies that showed there is hope," says Jordan, that opening up habitat on the Yellowstone would help pallids become self-sustaining between Fort Peck and Lake Sakakawea.

Besides, both agencies were worried that long-term flow modifications at Fort Peck would require a dam retrofit that could cost as much as $500 million -- 10 times the estimated cost of Intake improvements. Engineers were also concerned that Fort Peck's spillway wasn't strong enough to withstand large spills.