The federal government is churning out a plan to boost flows for salmon on the Klamath River, but relief will come only from the river's main tributary rather than the Klamath itself.

Last week, the Trinity Management Council recommended sending a burst of water down the Trinity River beginning in less than two weeks. It is designed to spur fish to migrate, then ramp back down.

The impetus behind the plan is the 2002 fish kill, which killed 34,000 to 68,000 chinook salmon in the lower Klamath. A huge portion of the Trinity River run died before it ever reached the confluence of the two rivers.

"What the recommendation was is very similar to last year," said Doug Schleusner, executive director of the Trinity River Restoration Program.

In one day, beginning on Aug. 22, flows from Lewiston Dam would jump from 450 cubic feet per second to 1,650 cfs. After that, the flows would drop slowly to 450 cfs on Sept. 15.

The U.S. Bureau of Reclamation is working on a final plan, which is expected this week.

While Reclamation will slowly increase Klamath flows as well, there is no plan to exceed its regular schedule. In the event another fish kill begins, there is no dedicated amount of water that could be released from Iron Gate Dam on the Klamath, said bureau Deputy Area Manager Christine Karas.

Karas said in an emergency, Reclamation could ask dam owner PacifiCorp to release water from Iron Gate Reservoir, but it would have to be filled again from Upper Klamath Lake.

How that would happen while meeting the bureau's obligation to keep lake levels high enough for endangered suckers -- and after providing irrigation water to farms all summer -- is difficult to know.

"We're going to be really tight in September," Karas said.

Troy Fletcher, executive director of the Yurok Tribe, sees a dichotomy in the plan.

He said Reclamation denies its operations harm fish on the Klamath River, but is using Trinity River water purchased at taxpayer's expense to bump up flows for salmon. All the while more water is flowing to irrigate farms than is being allowed down the river, he said.

"It seems disingenuous," Fletcher said.

Irrigators have argued that a combination of warm water -- like that in Klamath reservoirs -- and crowded conditions probably caused the 2002 fish kill. They've argued that releasing more warm water won't help.

A U.S. Fish and Wildlife report last year pointed to a variety of factors leading to the fish kill, among them too little water. This year, California's Fish and Game Department attributed the salmon deaths largely to low flows -- the only tool available to prevent future die-offs.

Fish and Wildlife Service, the National Oceanic and Atmospheric Administration and Reclamation voted against a request from tribes at the Trinity Management Council meeting to put more water down the Klamath.

Fish and Wildlife is working on criteria for releasing Trinity water, and hopes to send that to the bureau this week. Fisheries supervisor Mary Ellen Mueller said how the release will affect the different runs of salmon isn't known. There was no monitoring last year -- it is being proposed this year -- but Mueller said there's no evidence Klamath fish strayed up the Trinity, a concern voiced by some biologists.