The problem has been studied only a few years. Now, experiments at Grover's Creek Hatchery in North Kitsap have confirmed that polluted Storm water has the ability to kill adult coho before they can spawn.
This "pre-spawn mortality," as it is called, could pose a serious threat to the ongoing salmon populations in many urban areas, said Nathaniel "Nat" Scholz, a biologist with NOAA's Northwest Fisheries Science Center.
The problem was first suspected in Seattle's Longfellow Creek, which receives a rush of Storm water whenever it rains. Observers noticed that many of the female coho that made it home to their natal streams were dying before they could lay their eggs, often within a few hours of a rainstorm. Leading up to their deaths, researchers noticed that the fish seemed confused, often going in different directions and turning onto their sides while swimming.
An extensive forensic analysis ruled out everything but toxic chemicals, Scholz said. Further investigations revealed that the more polluted a stream became, the more likely the fish were to experience pre-spawning mortality. Up to 90 percent of the females were dying in some streams following a rainstorm.
At first, heavy metals were suspected. After all, it has been shown in laboratory experiments that copper compounds can destroy the olfactory sense of smell in salmon. Adults exposed to copper presumably can fail to home in on their natal streams, while juveniles exposed to copper become highly vulnerable to predators, according to previous studies by Scholz and his associates.
But adult coho exposed to 10 times the level of metals found in the toxic Storm water failed to show the characteristic behavior of the dying salmon in Longfellow Creek and other urban streams.
Meanwhile, other studies demonstrated that 65 percent of coho embryos exposed to this toxic Storm water had severe physical abnormalities, such as malformed fins, bleeding on the brain and swelling around the heart, according to Julann Spromberg, who discussed the findings at a recent meeting of the Kitsap Poggie Club, a local fishing group. Typically, the malformed fish die at an early age, she said.
Still searching for the mysterious, deadly compound, NOAA researchers formed a partnership with the Suquamish Tribe to use the tribe's Grover's Creek Hatchery, which rears coho and gets a fair number of returns in most years.
The researchers collected Storm water from a highway during major rains last fall. The water first collected after a four-day dry period appeared the darkest and dirtiest. As the rains continued, the collected water looked to be less ominous.
These different samples of Storm water were placed into small tanks with clean water placed into identical tanks. The salmon were then exposed to the water for an hour or two.
To the surprise of the researchers, all of the coho exposed to the Storm water showed the behavior they had come to expect. The fish bumped into the sides of their tanks, showing no sense of direction and keeling over on their sides.
"They couldn't even figure out how to turn around, they were so out of it," Spromberg said. "There was something severely wrong with them."
Even the highway runoff that seemed the cleanest after days of rain killed the fish. On the other hand, coho exposed to water from a clean stream suffered no ill effects.
Scholz said typical highway runoff contains an enormous number of different compounds, and it is extremely difficult and costly to narrow down which ones may be affecting the fish. Because death comes so quickly, the cause must be a physiological or metabolic pathway, not any kind of disease progression, he said.
Many tissues were taken from fish in the Grover's Creek experiments in hopes of finding a problem in the heart, gills or perhaps other essential organs. The method involves testing for genetic markers to determine which organs are under stress. Results are still pending.
"The fish are telling us what is going on, given the high rates of mortality across many streams," Scholz said. "But, scientifically, this is a tricky challenge. We have to look at target organs and try to figure out why they are dying."
Scholz said the answer is likely to be one of two things. Either the mystery compound is two or more known chemicals working together synergistically - which means together they are worse - or the mystery compound is a single chemical that has never been identified for its extreme toxicity.
"We don't have evidence for either one," Scholz said. "It could be an unmonitored chemical contaminant or a group of chemicals working synergistically. We have chemists at the Northwest Science Center looking at what they can find in tires."
Because the Kitsap Peninsula is a "transitional" area between urban and rural development, researchers would like to extend their studies into a variety of local streams where the runoff comes from different types of development. Some areas have infiltrated much of the Storm water into the ground - either through old-fashioned Storm water ponds or with rain gardens and other kinds of low-impact development.
"What we want to know from the NOAA side of things is whether you can reduce pollution loading sufficiently to protect the fish," Scholz said.
Jon Oleyar, who counts salmon in Kitsap streams to estimate populations, says he has noticed the effects of pre-spawning mortality in urbanized sections of streams, such as lower Clear Creek near Silverdale and lower Dogfish Creek near Poulsbo - and other salmon may be affected as well.
"I've seen it in chinook, and I've seen it in chum, too," he said, "but I don't see a lot of it."
Chris May, manager of Kitsap County's Surface and Stormwater Management Program, said he is following the toxicity studies closely as new strategies are planned to deal with the problem.
One step Kitsap County has taken is to acquire three high-efficiency street sweepers, which actually vacuum up road dust and debris to keep it from washing into ditches and ultimately streams. The machines sweep about 700 miles a year, most frequently near stream crossings and along shorelines, he said.
The SSWM program monitors the amount of toxic chemicals found in 200 tons or so of material that gets swept up each year.
May said he hopes that street sweeping, Storm water management and other efforts can help prevent the problems of pre-spawn mortality observed in portions of King County and other urban areas.