Keywords: Water quality, Species and food webs, Fishes, Nearshore habitat, Freshwater habitat, Toxic contaminants, Contaminants of emerging concern, Salish Sea Currents magazine, Salmonids, Stormwater, Species of concern

The search for why large numbers of spawning coho salmon have been dying in Puget Sound's urban streams goes as far back as the 1980s and culminated this year with the discovery of a previously unidentified chemical related to automobile tires. We offer a detailed timeline for the discovery. 

1980s and ‘90s

Strange behaviors seen in coho salmon

Biologists working in the watersheds came to realize that a large proportion of adult coho salmon were dying before they could spawn. First seen among coho in Bellingham’s Whatcom Creek and later in “restored” streams in Seattle, unusual behaviors, documented by observers, included erratic swimming and disorientation, leading to rapid death for the fish.


Researchers confirm unusual mortality in restored streams

One early investigation of Longfellow Creek in West Seattle found that 86 percent of the adult females died before spawning. That high rate of mortality along with observations in other streams led to the recognition that restoring an urban stream is not enough. One must also pay attention to the water chemistry, which can be degraded with the addition of stormwater, according to lead author Nat Scholz of NOAA’s Northwest Fisheries Science Center. 


High mortality rate linked to stormwater runoff

Researchers came to understand that high rates of mortality among coho could be expected after a heavy rainfall, especially following a dry spell. Typically, coho will hang out near the mouth of a stream and then move upstream when flows increase. “Whether salmon are dying from exposure to a single contaminant or a mixture of contaminants is not yet known,” states an article written by Sarah McCarthy and other NOAA researchers.


Motor vehicles gain attention, as other sources are ruled out

An eight-year study involving fish returning to a half-dozen Seattle-area streams ended with the conclusion that coho exposed to stormwater did not die from disease pathogens, poor physical condition or common household pesticides, nor were they killed by high temperatures, low oxygen or ammonia in the water.

“The weight of evidence therefore suggests that adult coho salmon are unusually vulnerable to the toxic effects of one or more chemical contaminants, most likely delivered to urban spawning habitats via stormwater runoff,” states a report by Scholz and other researchers. “The rapid progression of the syndrome and the specific nature of the symptoms are consistent with acute cardiorespiratory toxicity.”

A separate study of watershed conditions pointed to a greater risk of mortality in areas dominated by impervious surfaces. “While not definitive, our results reinforce the parsimonious explanation that coho deaths are caused by one or more contaminants originating from motor vehicles,” states a report by Blake Feist of the Northwest Fisheries Science Center and other researchers.


Direct exposure to roadway runoff implicates motor vehicles

Juvenile coho placed in roadway runoff died within 12 hours of exposure, adding new evidence that motor vehicles were to blame. The study, led by Jen McIntyre of Washington State University’s Puyallup Research and Extension Center, also found that filtering the road runoff through a column of soil (sand, compost and bark) removed the deadly chemicals and allowed the coho to survive.


Mortality can’t be linked to well-known toxics in cars

Adult coho exposed to roadway runoff died rapidly in a study led by Julann Spromberg with NOAA’s Northwest Fisheries Science Center. In contrast, the fish did not die when placed in an “artificial stormwater” of hydrocarbons and metals, typical of road runoff. As a result of the study, researchers realized that they needed to look for other toxic chemicals associated with motor vehicles.


Search narrows with identity of chemicals in fish tissues

Sophisticated analytical equipment, including high-resolution mass spectroscopy, was brought to bear on the problem of searching for the specific chemical or chemicals that kill coho among thousands of compounds found in roadway runoff. This “non-target approach” was first taken in a study led by Bowen Du of the University of Washington’s Center for Urban Waters in Tacoma.

Using extraction techniques, organic chemicals were first separated out of roadway runoff as well as from tissues of coho exposed to runoff. Mass spectroscopy, which identifies compounds by molecular weight, revealed chemicals that had passed from stormwater into fish tissues. The study confirmed the success of this new approach, identified several suspect chemicals — including some from tires — and concluded that “many novel or poorly characterized organic contaminants exist in urban stormwater runoff and exposed biota.”

In a separate study led by Feist, researchers measured coho mortality rates at 51 spawning sites and developed a watershed map showing coho vulnerability for streams throughout the entire Puget Sound region in relation to the road network. The land-use analysis further implicated motor vehicles as a source of one or more chemicals toxic to coho.


Tires become prime suspect; blood chemistry gains attention

To develop a “chemical signature” for the lethal compounds, researchers compared chemicals found in toxic stormwater to chemicals released from automobiles in a study led by Katherine Peter of the UW’s Center for Urban Waters in Tacoma. Using high-resolution mass spectroscopy, they found that a solution made from ground-up tire treads turned out to be most similar, chemically, to components in water samples that had killed coho, when compared to windshield wiper fluid, automatic transmission fluid, power steering fluid, used motor oil, used gear oil and antifreeze. The list of suspect chemicals — actually groups of chemicals — was narrowed substantially.

In a separate study led by McIntyre, coho and chum salmon were exposed to the same highway runoff. Coho developed the now-familiar erratic behaviors of toxic exposure, accompanied by radical changes in blood chemistry. Chum seemed immune to whatever was affecting the coho. This confirmed the stark difference between species, as observed in the streams, and caused researchers to look closely for physiological differences.


Juvenile coho prove to be good stand-ins for adults

Juvenile coho exposed to highway runoff showed symptoms like those seen in adult coho, as described in a study led by Michelle Chow of the University of Washington’s School of Aquatic Fisheries. Most importantly, the study revealed that changes in both behavior and blood chemistry progressed among juveniles the same as in adults. Those findings opened the door to further studies using smaller juvenile coho — more abundant and available year-round — to represent toxic effects on all age groups. The study also showed that the effects are irreversible, suggesting that a single rainstorm that washes chemicals off a roadway could doom a large percentage of coho — adult or juvenile — swimming in nearby streams.


Deadly compound found to be the product of chemical reaction

In a multi-step process, researchers separated chemicals from tires into constituent groups, which were then tested for toxicity in juvenile coho. Step-by-step, separation and testing narrowed the list of possible compounds that could be responsible for coho deaths in the study led by Zhenyu Tian of the UW’s Center for Urban Waters in Tacoma. Chemicals in the remaining toxic mixture were then identified through high-resolution mass spectroscopy. It turns out that the tire-related chemical most responsible for coho deaths is actually not one of the many chemicals added to tires. Rather it is a chemical produced when ozone reacts with a compound infused in rubber to protect tires against ozone damage.


A returning Coho Salmon at the Suquamish Tribe's Grovers Creek Hatchery. Photos: K. King/USFWS (CC BY-NC 2.0)

Environmental engineers and chemists at the University of Washington Tacoma have identified a mysterious compound implicated in the deaths of large numbers of coho salmon in Puget Sound. The chemical is linked with a rubber additive commonly used in tires and is thought to kill more than half of the spawning coho that enter the region's urban streams every year. 


Stacked tires. Photo: Kool Cats Photography. (CC BY-NC 2.0)

Modern automobile tires are a complex mixture of chemicals, all used together in different ways to give tires their structure and properties, including riding comfort, safety and long life. Chemicals from tire wear particles are now thought to be responsible for the deaths of large numbers of coho salmon returning to spawn in Puget Sound streams. 

About the author: Christopher Dunagan is a senior writer at the Puget Sound Institute.

Implementation Strategies

Sponsored by:

United States Environmental Protection Agency logo

Creative Commons License
Story text available under a Creative Commons Attribution-NonCommercial-NoDerivs license: CC BY-NC-ND 4.0.