Toxic road runoff kills adult coho salmon in hours, study finds

A three-year-old adult coho makes its way through the Issaquah Salmon Hatchery. (Mike Siegel / The Seattle Times)
A three-year-old adult coho makes its way through the Issaquah Salmon Hatchery. (Mike Siegel / The Seattle Times)


By  Sandi Doughton, Seattle Times


A new study shows that stormwater runoff from urban roadways is so toxic to coho salmon that it can kill adult fish in as little as 2½ hours.

But the research by Seattle scientists also points to a relatively easy fix: Filtration through a simple, soil-based system.

“It’s basically … letting the Earth do what it does so well, what it has done for eons: cleaning things up,” said Julann Spromberg, a toxicologist for the National Oceanic and Atmospheric Administration and co-author of the report published Thursday in the Journal of Applied Ecology.

Scientists have long suspected that the mixture of oil, heavy metals and grime that washes off highways and roads can be poisonous to coho, but the study is the first to prove it.

The research got its start more than a decade ago, when habitat-restoration projects began coaxing a trickle of coho back to several urban streams in the Puget Sound area. But many of those fish died before they could spawn. And the deaths seemed to coincide with rainstorms that sent runoff surging through drainage pipes and into the waterways.

In some place, like Longfellow Creek in West Seattle’s Delridge area, up to 90 percent of females were killed.

“It was apparent that something coming out of those pipes was causing it,” Spromberg said.

She and her colleagues tried to reproduce the effect in the lab. But the artificial mixture of oil and other chemicals they concocted had no effect on the fish.

So their next step was to try the real thing: Actual runoff, collected at NOAA’s Northwest Fisheries Science Centerfrom a downspout that drains a Highway 520 onramp near Montlake.

“When we brought out the real urban runoff: Bang! They were down, they were sick, they were dead,” said co-author Jenifer McIntyre, a researcher at Washington State University’s Puyallup Research and Extension Center.

In experiments at the Suquamish tribal hatchery near Poulsbo, every coho exposed to the runoff died — some within a few hours, all within a day. Before death, the fish became lethargic, rolled around and swam to the surface as if gulping for air, McIntrye said.

The fact that actual runoff proved fatal while the scientists’ concoction did not underscores an unsolved mystery about which chemical or combination of chemicals are so toxic to the fish. It could be any number of compounds that weren’t part of the artificial brew, including byproducts of oil and gasoline combustion, chemicals released by tires or tiny particles from brake linings, Spromberg said.

 “We still need to keep looking at what exact compounds are involved.”

But whatever the chemical culprit, the scientists found it could be removed by passing the runoff through 55-gallon drums packed with layers of gravel, soil and compost. None of the fish exposed to the filtered stormwater died or fell ill.

“It was remarkable,” McIntyre said.

 The finding is a strong endorsement of rain gardens, grassy swales and other “green” alternatives to traditional drains and pipes designed to collect stormwater. The idea is instead to let the runoff percolate through the ground, as it did before so much of the area was paved and developed.

State regulations strongly encourage developments to use such approaches, according to the Washington Department of Ecology. A project called 12,000 Rain Gardens in Puget Sound is also promoting their use.

“We should be seeing more and more of these systems in the future,” McIntyre said.

Coho, which were once abundant throughout the Northwest, may be particularly vulnerable to toxic runoff because they spawn in the fall, prompted by seasonal rains. Habitat destruction, fishing and other factors almost certainly contributed to the species’ precipitous decline, Spromberg said.

Chum salmon, whose habitat and spawning seasons overlap those of coho in many places, don’t appear to be as affected by runoff — something the scientists plan to investigate this fall.

Perhaps the major limitation of the study is the small sample size. Only 60 coho were used in the experiments, 20 in each of two experimental and one control groups. The scientists were lucky to get that many, thanks to the cooperation of the Suquamish Tribe, McIntryre said.

Also, the urban runoff collected near Montlake was undiluted in the experiments and represents about the worst possible case: runoff from a busy highway in a big city, a DOE official who was not involved in the study pointed out.

“It’s great that the treatment gets rid of toxicity from this nasty stuff,” Karen Dinicola of DOE’s stormwater program wrote in an email. But it’s particularly challenging to retrofit urban-collection systems with greener alternatives, she said.

But the results of the research could help guide future development in rural watersheds where coho runs remain, the researchers said. And it can also be used to help inform urban-restoration projects as well, so fish aren’t lured back to appealing habitats, only to be clobbered by toxic runoff.

The researchers are preparing for their next round of studies, which will include tests to zero in on what is actually killing the coho.

The rain that soaked the region Wednesday also filled their runoff-collection barrels, Spromberg said.

“We only have one shot a year, when the fish come back and we can do the experiments and take the samples,” she said. “Hopefully, with this rain we’ll have more fish coming in soon.”

Puyallup Tribe Looking For Coho Family Tree


Mar 25th, 2014 Northwest Indian Fisheries Commission

The Puyallup Tribe of Indians is building a library of genetic material from coho salmon to better understand the different populations throughout the Puyallup River watershed.

“The data behind how all these fish are related can give us a pretty clear picture of how many populations are actually here,” said Russ Ladley, resource protection manager for the tribe. “Are populations that have different run timings independent of each other, or do they interbreed?”

A winter coho is sampled for genetic material on the White River, a tributary of the Puyallup.
A winter coho is sampled for genetic material on the White River, a tributary of the Puyallup.

Winter run coho migrate through the Puyallup as late as February or even March while the earliest run fish are often seen as soon as July. “There isn’t much time when coho aren’t moving into the freshwater to spawn,” Ladley said.

“I would like to collect an adequate sample so we have a background from which to compare,” Ladley said. “I want to know, for example, if the late time coho we see in the White River are different from early coho we see there.”

“Thirty years ago the state Department of Fish and Wildlife sprinkle planted coho fry throughout the watershed, so I would like to find out if the fish are all the same or are still diverse,” he said.

Much of the Puyallup coho’s historic habitat has been degraded in the past century and is still disappearing, making an analysis of interrelationships vital. Coho salmon spend an extra year in freshwater as juveniles compared to other salmon species, making them more vulnerable to declines in freshwater habitat.

For example, low summer flows have been dropping throughout the watershed for decades. “Coho are their most vulnerable when we get to summer low flows,” Ladley said. “Despite a prohibition of new water withdrawals, we’ve seen a continual decline in summer flows because of unregulated wells being allowed to spread across the watershed.”

Low flows reduce the amount of habitat available for coho rearing and can cut fish off completely from valuable habitat. “When it comes down to it, fish need water to survive,” Ladley said.

“Currently, we see a fairly broad range of return timing and coho utilizing habitat from near sea level to 3,000 feet of elevation in Mount Rainier National Park,” Ladley said. “It will be interesting to learn if this is one homogenous stock or whether clear genetic differences exist.”

“This genetic data will give us a clearer picture of exactly how diverse they are, and hopefully give us information we can use to better manage the stock,” he said.

More information on the decline in salmon habitat in the Puyallup River watershed can be found at: and for all of western Washington, here:


For more information, contact: Russ Ladley, resource protection manager, Puyallup Tribe of Indians, (253) 845-9225. Emmett O’Connell, South Sound Information Officer, NWIFC, (360) 528-4304,

Filtered Stormwater Added to Annual Coho Salmon Experiment

WSU toxicologist Jenifer McIntyre and USFW biologist Steve Damm adjust the spouts for the barrels mimicking rain gardens for stormwater filtration.
WSU toxicologist Jenifer McIntyre and USFW biologist Steve Damm adjust the spouts for the barrels mimicking rain gardens for stormwater filtration.

Source: NWIFC

Rain gardens filter toxic chemical contaminants from stormwater before it flows into Puget Sound streams, but no one knows how well they protect the salmon that spawn in those streams.

That was this year’s question during the annual coho salmon stormwater experiment at the Suquamish Tribe’s Grovers Creek Hatchery.

Since 2011, tribal, federal and state scientists have been studying how salmon are affected by stormwater before they spawn.

In previous years, the research team found that adult coho were dying prematurely when they returned to spawn in urban watersheds throughout many areas of Puget Sound. Working with the tribe, the team hopes to better understand why stormwater runoff is so toxic, and also identify stormwater treatment methods that can effectively protect adult spawners.

At the hatchery this fall, scientists ran raw stormwater through four barrels of sand and compost. The barrels mimic the filtration that occurs when runoff is cleaned using various green stormwater infrastructure or low-impact development technologies. Unfiltered and filtered water were then placed into large tanks with fish to monitor their survival and observe their behavior. The team also exposed fish to clean well water from the hatchery.

“The Washington Department of Ecology recommends this kind of filtration technique for bioremediation and new low-impact development,” said Julann Spromberg, a NOAA toxicologist. ”We wanted to see how well it would work from the perspective of the fish – can we remove enough of the pollution from urban runoff to keep the coho spawners alive?”

Preliminary results show that this year was a success. Fish in the filtered water for 24 hours were alive and behaving normally, Spromberg said. In addition, the team conducted its regular experiment – exposing fish to hatchery well water and raw stormwater – and came up the expected results: the former survived, the latter did not.

“We don’t know exactly which contaminants are causing the fish to die, but we do know the bioretention filtration technique is effective,” Spromberg said.

The next step will be to take the filtered stormwater and the sand/compost mixture to try and determine what toxic components were filtered out, narrowing down what contaminants are causing salmon to die before they can spawn.

“I’m glad we’re able to provide a space for these folks to do this work,” said Mike Huff, Grovers Creek hatchery manager. “Anything we can do to support salmon survival benefits everyone.”

Partners in the project include the Suquamish Tribe, Environmental Protection Agency, Seattle Public Utilities, Kitsap County, National Oceanic and Atmospheric Administration, Washington State University, and U.S. Fish and Wildlife Service.