The mystery of the disappearing Lake Washington sockeye salmon, which began last year with thousands of the fish vanishing between the Ballard Locks and their spawning grounds, has taken a new turn.

This year, the fish aren't even making it to the Locks.

Scientists had expected about 398,000 sockeye to return from the ocean this season. But with the run nearly half over, they say they will be lucky if 100,000 fish make it back.

"We got slammed on one end one year and we got slammed on another end the other year," said Mike Mahovlich, a fish biologist for the Muckleshoot Tribe.

The poor return is a disappointment for tribal fishermen and sport anglers, whose hopes for a sockeye-fishing season this year are dashed. It's also a further puzzle for scientists trying to untangle what's happening to fish that spend most of their life out of sight.

Last year, elevated water temperature in the freshwater Lake Washington Ship Canal was the prime suspect when roughly half of an estimated 400,000 fish that made it through the Locks disappeared. Salmon are sensitive to high temperatures, and water in the canal was unusually warm in 2004.

But this year, the Pacific Ocean is the focus. The salmon are not returning from the salt water where most have spent the past two years, since leaving the fresh waters of the Lake Washington watershed where they were born.

"The ocean's kind of a mysterious entity," said Jim Ames, sockeye-program manager for the state Fish and Wildlife Department. "It's really a black box out there. First of all, we don't know where these fish go out in the ocean."

Sockeye biologists expected a bumper crop of salmon in Lake Washington this year, based on their counts of the young, ocean-bound sockeye that left the lake for the ocean in early 2003.

Even if the return reaches 100,000, it would be the lowest ocean-survival rate recorded for the sockeye, Ames said.

Even so, back-to-back years of low spawning numbers don't necessarily spell disaster for Lake Washington sockeye, Ames said. A hatchery helps prop up the fish population, and even a small number of spawning fish can produce a lot of juveniles if spawning conditions are good.

But concerns remain that rising temperatures in Lake Washington could pose long-term problems.

And a low return this year makes it less likely there will be enough offspring to have a fishing season in 2009, when sockeye born this year will return from their eventual voyage to the sea, Ames said.

Members of the Muckleshoot Tribe, one of three tribes with rights to catch the Lake Washington sockeye, won't catch the fish this year, even for ceremonial purposes or for their own food.

"That fish means a lot to the tradition and the culture of the tribe. It's very sad to see that the fish are not returning in the abundance that we had hoped," said Leo LaClair, of the tribe's fish commission.

Sport anglers also flock to Lake Washington on good sockeye years. To them, a canceled season is like Seattle losing Seafair, said Frank Urabeck, a spokesman for the conservation and sport-fishing group Trout Unlimited.

"It's a big disappointment, so I think there's a lot of pressure on the state and others in the scientific community to figure out what's going on," he said.

The mystery mirrors what happened earlier this year on the Columbia River, where less than half the expected spring chinooks returned to spawn. The fishing season on the river was canceled, and fisheries managers were left wondering where the fish had gone.

There are some early indications other sockeye runs this year may fall short. The earliest run on the Fraser River in British Columbia is on course to be less than half what it was supposed to be. But it's too soon to make predictions about the main runs on the Fraser, which come later.

Other Northwest salmon runs, however, aren't following the same path. The sockeye runs in Alaska's Bristol Bay are as expected and could reach more than 32 million returning fish. And unlike the spring chinooks on the Columbia, the summer chinooks are plentiful enough that a fishing season opened for the first time in decades.

Many factors, including shifts in ocean currents, a decline in a major food source or the spread of animals that eat salmon, can influence how salmon fare in the ocean, said Pete Lawson, a Newport, Ore., fisheries biologist for the federal Northwest Fisheries Science Center.

For example, a change in ocean currents that occurred in the north Pacific in 2003 would have been less favorable for salmon, according to a Northwest Fisheries memo written in May about the spring chinook run on the Columbia. But that alone wouldn't entirely explain what happened to the chinooks, Lawson said.

"We're trying to cast some light on it," Lawson said. "But it's a big place, and it's not easy to know what's going on."

Warren Cornwall: 206-464-2311 or wcornwall@seattletimes.com