15 years there was nothing as bad as last year, as far as just seeing dead bodies of sockeye," said Mahovlich, a fish biologist for the Muckleshoot Tribe, which helps manage the sockeye run.

But he was even more startled by the final picture that emerged late in the year: As many as 200,000 sockeye, roughly half the run, had disappeared somewhere between the Locks and their spawning grounds in streams beyond the lake.

The mystery of the missing sockeye has scientists puzzled and worried, as they try to decipher the fate of a cherished run that passes through the heart of Seattle. So far, scientists are focusing their suspicion on abnormal water temperature. And they worry that climate change could make it more than a freak occurrence.

Already there are signs that this wasn't a one-time event. In the past 34 years, three of the four years with the biggest disparity in fish numbers between the Locks and the spawning grounds have been since 2000. And recent research shows Lake Washington has warmed over the past three decades, driven in part by rising air temperatures that could be a symptom of global warming.

But for the salmon, 2004 was the worst by far. It caught the attention of researchers and convinced them that the drop wasn't just a figment of imprecise counting methods.

"I'm afraid it's not one freak year," said Eric Warner, another Muckleshoot fisheries biologist. "I think it's probably a hint of things to come."

Good start goes bad

When the sockeye started passing through the Locks last June, the season was looking good. People hired to count salmon climbing the fish ladder reported enough fish that a sockeye fishing season was launched.

Then, in August, dead salmon began to turn up in larger numbers. Concerned, the state tested some of the dead salmon for signs of disease but found no indications of a major outbreak.

In October, when the fish were expected to arrive in their spawning streams, the magnitude of the problem emerged. In Bear Creek, where as many as 40,000 sockeye spawned a generation earlier, just 1,500 arrived. In the Cedar River, where most of the sockeye usually return, scientists kept waiting for a surge of fish but instead saw only a steady, low flow.

By winter, the fish disappearance, combined with hot weather the previous summer and an absence of a disease outbreak, led researchers to suspect high water temperatures in the Lake Washington Ship Canal. It's narrow and shallow, and sockeye have to run it to reach the cool, deep waters of Lake Washington.

Sunny, warm weather last June and July had driven water temperatures unusually high. Ship-canal water hit 68 degrees Fahrenheit on June 21, two weeks earlier than ever recorded. On July 18, it reached 71.6 degrees.

For salmon, water temperature above 68 degrees is dangerous. It can stress them out, making them more vulnerable to disease. At 77 degrees, water temperature can kill them outright.

Warm water was suspected in the disappearance last year of sockeye in British Columbia. As many as 1.3 million sockeye failed to show up at spawning grounds on the Fraser River, one of the most important commercial salmon runs in the province. It was a catastrophe that led to government inquiries and fingerpointing over tribal fishing rights.

Causing anxiety

The Lake Washington sockeye disappearance hasn't produced such controversy. For example, no one has blamed problems on overfishing. But it is causing anxiety about the state of the sockeye, the city's largest salmon run.

Chinook salmon, which pass through the Locks mainly in August, didn't appear to suffer the same effects as the sockeye, said Warner, the Muckleshoot biologist. They may not be as temperature sensitive.

While the sockeye aren't native to Lake Washington — they were introduced in the 1930s — the fish have cultural and economic importance.

The Muckleshoot, Tulalip and Suquamish tribes have rights to catch the sockeye, which are centerpieces in ceremonies as well as a source of food and income. For non-Indians, a good run draws flotillas of recreational anglers to Lake Washington in July. Tourists flock to the Locks, partly to watch sockeye salmon wriggle and splash their way past huge viewing windows.

Now tribal and state scientists are preparing studies to track the sockeye and follow clues about what might be happening to them.

One study will tag salmon with sensors to measure the water temperatures they are enduring. Another would use underwater radio receivers to determine how many actually make it through the ship canal. Yet another is aimed at getting a more accurate estimate of how many fish make it from the Locks to the spawning grounds.

"There's just a lot of 'Maybe this is it,' and guesses," said Steve Foley, the lead Lake Washington salmon biologist for the state Fish and Wildlife Department. "We're looking this year to try to get some answers."

Checking the temperature

But water temperature remains the prime suspect. And some worry that it could become worse.

Evidence has linked rising air temperatures in the Seattle area to increased surface-water temperatures in Lake Washington.

A 2004 study by University of Washington and King County researchers found that average summer surface-water temperatures have risen by about 4 degrees over a 35-year period, with air temperatures the strongest influence.

Last year's sockeye problems show "exactly why that study is relevant," said Michael Brett, a University of Washington scientist who co-wrote the study.

And for fish scientists, the warming trends leave them wondering if what happened last year to the sockeye may become the norm.

"If it's a one-in-30-years event, then it's not very important. If it starts happening with frequency, then it's a critical thing," said Jim Ames, sockeye-program manager for the state Fish and Wildlife Department.

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

--------------------------------------------------------------------------------