FLAGSTAFF, Ariz. » Maintaining sandbars crucial to wildlife in the Grand Canyon would require more frequent high water flows from Glen Canyon Dam that coincide with natural flooding of Colorado River tributaries, a U.S. Geological Survey scientist said Tuesday.

A 2008 experiment meant to mimic natural flooding built up sandbars and beaches along a 277-mile stretch of the river, providing habitat for plants and animals, beaches for campers and protection for archaeological sites. But the sandbars largely were eroded six months later, compounded by fluctuating flows from the dam.

Ted Melis, deputy chief of the USGS' Grand Canyon Monitoring and Research Center, said Tuesday that the key to maintaining the sandbars is not simply manipulating the flows from the dam on the Arizona-Utah line. The frequency and timing of the flows would have to exceed the erosion that occurs between them, he said.

"The question still remains, can it be maintained over some long time period?" he said.

Interior Secretary Ken Salazar has called for more of the man-made floods. A plan on when and how such high-flow experiments should be conducted will be based partly on the USGS report released Tuesday.

Environmentalists have said that more seasonally adjusted steady flows are needed to preserve the sandbars, along with possibly augmenting sediment and using temperature control devices. Any change to dam operations likely would affect the amount of hydropower produced and requirements for water delivery to six states.

"What is so frustrating is they can fix this, but they won't," said John Weisheit, conservation director for the Moab, Utah-based conservation group Living Rivers. "Glen Canyon Dam is a cash-register dam, and they want it to make money. They're going to have to make a choice. The Grand Canyon or some kilowatt hours?"

Melis said that opportunities for high-flow experiments would occur on average once every two years, when sediment is deposited mainly through flooding of the Paria River and other tributaries. The sooner high flows are released, the better because sand becomes more coarse over time and lessens the chances that sandbars can be built efficiently, Melis said.

"Short of actually importing additional sand supplies, the best option we have right now is to proactively experiment with these inputs when they occur," he said.

The frequency, timing and duration of the manmade floods can be adjusted to get differing responses. Spring, for example, would be the best time to deposit sands inland to protect archaeological sites, the USGS report said. Conducting the high-flow experiment in March also reduced the chance that seeds of the nonnative tamarisk would successfully germinate and upped the survival rates of rainbow trout in their early stages of life.

Since the 1960s, Glen Canyon Dam has blocked 90 percent of sediment from the Colorado from flowing downstream, turning the once muddy and warm river into a cool, clear environment that helped speed the spread of extinction of fish species and pushed others near the edge.

Despite the erosion of the sandbars months after the 2008 high-flow experiment, the USGS said gains were made in the median elevation of sandbars and abundance of backwater habitat.