Haunted by the sight of disoriented coho salmon keeling over and dying before they could spawn, scientists have been searching for more than 20 years to find a deadly chemical thought to be prevalent in the urban streams of Puget Sound.

Now environmental chemists have painstakingly narrowed the suspects to a single mysterious compound found in automobile tires. Until now, the little-known compound was never recognized as a problem. Yet its presence in stormwater washing off roads is likely responsible for thousands of coho dying before their time. So far, nobody knows how the chemical might be affecting other fish and aquatic life.

Critical steps in the long search for this highly toxic substance involved the latest lab equipment, systematic planning and advanced techniques not even available when the search began, experts say. The final push for the finish line is described in a research article published today (Dec. 3) in the journal Science.

Tracing coho mortality back to a little-known chemical in tires revitalizes a question that biologists have been pondering for years: If an unknown chemical in stormwater can do so much damage to a well-known, beloved species like coho salmon, could that chemical or others be injuring or even wiping out lesser-known species?

“Once you get into the field and see fish dying, you feel a strong motivation to get this done,” said Zhenyu Tian, a University of Washington researcher and lead author of the paper. “This was an opportunity to answer a scientific question that will really make a difference.”

Zhenyu Tian, a research scientist at the Center for Urban Waters at UW Tacoma, holds a sampling pole, which is used to collect creek water for future tests. Mark Stone/UW

Tian was part of a team at the University of Washington Tacoma's Center for Urban Waters [Editor's note: The Center for Urban Waters is our parent organization and research partner] under the leadership of Ed Kolodziej, an associate professor in the UW’s Department of Civil & Environmental Engineering and co-senior author of the paper. The collaboration extended to other regional experts who had been working on the problem for years.

One day last December, the researchers were able to hold up a test tube with a small amount of magenta-colored crystalline substance at the bottom. They had successfully isolated the deadly compound from more than 2,000 tire-related chemicals, according to their analysis.

Like searching for a needle in a haystack, this relatively new “non-targeted” approach started with a full array of chemicals found in tires. Through chemical separation, toxicity testing and identification, the team narrowed the list of suspect chemicals. This type of analysis goes well beyond the traditional method of looking for specific chemicals with known toxicity. The non-targeted approach can identify chemicals not found on anyone’s list of dangerous compounds.

“Non-targeted analysis is cutting edge, especially in the environmental sciences,” noted Kolodziej. “It is based on letting the instrument tell you what is there, without making any assumptions.”

Although many chemicals go into manufacturing tires, the newly discovered chemical, called 6-PPD-quinone, is not one of them. It is, in fact, a chemical produced when ground-level ozone reacts with a chemical in tires called 6-PPD. That’s the compound put into tires to prevent ozone damage to the rubber.

It turns out that the toxicity of 6-PPD goes up more than 100-fold when converted to 6-PPD-quinone, which then gets released into stormwater with tiny particles that wear off of tire tread.

Tracing coho mortality back to a little-known chemical in tires revitalizes a question that biologists have been pondering for years: If an unknown chemical in stormwater can do so much damage to a well-known, beloved species like coho salmon, could that chemical or others be injuring or even wiping out lesser-known species?

Now that the identity of the tire chemical is known, scientists face a multitude of new questions:

• How is this compound affecting the tissues and metabolism of coho and potentially other aquatic species?
• Might the compound affect human health through exposure to athletic tracks and fields that use ground-up tires in artificial surfaces?
• Can a substitute chemical be found to protect tires without causing such devastating effects to the ecosystem?
• What kind of ecological damage has this chemical caused since it was added to tires decades ago?
• Will this discovery ultimately help to restore coho salmon to the many urban streams throughout Puget Sound and up and down the West Coast?

A deadly discovery

Symptoms of acute poisoning among coho salmon have been described in urban streams since the late 1980s. Crippled by something lurking in the water, coho will swim in circles or in random directions at the water’s surface, mouths wide open in a most unnatural manner. Eventually, the fish lose all control, roll over, and float downstream to a rapid death.

By the fall of 2001, such deaths were being reported in several urban streams, including Seattle’s Longfellow Creek. The removal of culverts and the restoration of stream channels had allowed salmon to find their way upstream and into areas that had gone unoccupied for decades. While such stream restorations were cause for celebration, some streams became an unintended trap for coho as they recolonized the waterway.

Once state and federal experts realized what was happening to the Puget Sound coho, they began to investigate this “pre-spawn mortality,” now called urban runoff mortality syndrome.

Nat Scholz, a marine zoologist with NOAA’s Northwest Fisheries Science Center, says he will never forget the first time he encountered the problem in Seattle’s Longfellow Creek. It was the fall of 2002. Scholz, just out of graduate school, had been surveying the stream with a colleague. They were waiting for rains to trigger the movement of the coho.

“We had set ourselves up for a routine survey,” Scholz said. “It was raining by the time we got there in the afternoon, and it didn’t take long to realize that we had a significant problem on our hands. There were fish in all stages. Some were swimming in circles and gaping. Some were tumbling. We had fish washing down between our legs.”

Scholz said it was shocking to see a fish kill of this magnitude in the absence of an obvious chemical spill of some sort.

“I was relatively new to toxicology,” he said, “but I knew this was going to be incredibly challenging. It is the kind of project that you want to start early in your career.”

Scholz knew, even then, that it could take years to find an answer. By 2011, NOAA researchers had studied tissues from dead or dying coho in numerous urban streams, comparing them to normal fish tissues. Published results showed a strong connection between dead coho and stormwater. Extensive testing found that the cause of death did not appear to be disease pathogens, poor physical condition or common household pesticides, nor were the fish killed by high temperatures, low oxygen or ammonia in the water.

Mounting evidence connected coho mortality in streams to locations with the greatest roadway runoff, suggesting that the toxic chemical or chemicals may be coming from motor vehicles.

Something in the stormwater

Jen McIntyre became fascinated with pre-spawn mortality after hearing Scholz speak on the subject to a group of scientists in 2003. As a graduate student in the University of Washington’s School of Aquatic and Fishery Sciences, she was troubled to learn that coho could die on their own doorstep before completing the essential act of reproduction.

“It seems completely unfair that they would go through so much in their lives and struggle their way to the ocean and back — to the very stream where they were born,” she said. “When you see every fish ending up dead, it definitely heightens the urgency.”

McIntyre, one of the Science paper's co-senior authors, studied toxic chemicals in the Lake Washington food web for her master’s degree. Her doctoral research focused on coho and the metal copper, which disrupts the sense of smell in coho, making it hard for them to avoid predation and find their way home. That research was key to a ban on copper in brake pads approved by Washington and California lawmakers in 2010.

Between her master’s and doctoral research, McIntyre worked for NOAA’s Northwest Fisheries Science Center as the first coordinator of the pre-spawn mortality studies.

“I can remember walking up Longfellow Creek and seeing a beautiful fish with its back out of the water,” she said. “I was creeping slowly right up to it, but it didn’t move. It was stiff but not dead. Its heart was still beating, but it was completely stiff (as in rigor mortis).”

It was disheartening to see a magnificent 24-inch coho die before its time, McIntyre said, and there must have been thousands like it throughout Puget Sound. In 2011, she was hired by Washington State University to continue her research into the effects of stormwater on salmon, this time at WSU’s Puyallup Research and Extension Center.

The work included collecting stormwater from a busy roadway in Seattle and measuring its effects on coho. One of her findings was that an unknown chemical or chemicals found in roadway runoff seemed to affect the heart or circulatory system of coho salmon but, oddly enough, not chum salmon.

When a preservative in vehicle tires called 6PPD reacts with ozone it is transformed into multiple chemicals, including 6-PPD-quinone. This chemical, which gets released into stormwater with tiny particles that wear off of tire treads, is the one researchers found that is responsible for killing coho salmon. Mark Stone/UW

Later studies showed lesser toxic effects on steelhead and Chinook, but no effects on chum or sockeye.

McIntyre also discovered that coho could survive without any apparent harm as long as the toxic stormwater was filtered through a column of soil before exposure. Whatever was in the water could be removed through simple filtration.

While filtration seemed to offer a way to protect coho from road runoff, it didn’t take long for engineers to figure out that it would take an enormous amount of money to filter all the stormwater going from roads into streams — even if one were to tackle only the worst problems.

“Even before we knew the identity of this chemical, I was saying that it will take more than treating stormwater to solve this problem,” McIntyre said. “Here we have a chemical that is really toxic to at least one organism that society cares about. It only makes sense to work toward source control.”

Source control involves keeping dangerous chemicals out of the environment. In this case, that would probably mean reformulating the rubber in tires to eliminate at least one toxic compound.

Chemists join the challenge

The Center for Urban Waters at the University of Washington Tacoma opened in 2010 with a sophisticated laboratory and experienced staff capable of analyzing samples related to research, environmental compliance and toxic cleanup. In 2014, environmental engineer Ed Kolodziej came on board as a UW faculty member looking for thorny environmental problems to explore.

“All of the conversations around here seemed to revolve around stormwater and roadway runoff and salmon,” he said. “To me, it was obvious that we needed to build our capabilities toward stormwater.”

Kolodziej had heard that coho salmon were dying from unknown chemicals in stormwater, and he wanted to help solve the problem, joining biologists and ecotoxicologists who had been working on the problem for years.

“We had the perfect lab to investigate the question,” said Kolodziej, who had studied contaminants in fish at the University of California, Berkeley, and later toxic chemicals in agricultural runoff at the University of Nevada, Reno.

“Working collaboratively provides unique opportunities,” he said, “and it was a real joy to work with these strong toxicology groups and to have the tools to work with.”

Researchers at UW Tacoma, UW, and WSU collaborated to discover a chemical that kills coho salmon in urban streams before the fish can spawn. Shown here at Longfellow Creek, an urban creek in the Seattle area, are Ed Kolodziej (left), UW; Jenifer McIntyre (right), WSU; and Zhenyu Tian (background), Center for Urban Waters at UW Tacoma. Mark Stone/UW

By 2018, researchers at the Center for Urban Waters had conducted a series of experiments comparing chemicals in stormwater to those found in fish tissues. The focus shifted to tires when researchers discovered that a solution made from ground-up tire treads was chemically similar to components in toxic stormwater.

The next step would be an extraordinary challenge: Chemists would seek to separate the deadly chemical or chemicals, still a complete mystery, from an untold number of other compounds found in tires.

This approach is called non-target analysis, because one makes no assumptions about the chemicals that might be found. Traditional methods involved testing for individual chemicals, typically from a long list of known toxicants. But the tried-and-true approach would not work if the chemical were truly unknown.

At each step in the process, separated solutions are tested for toxicity. Those found to be toxic undergo further separation and testing in a process called “effect-directed analysis.” Eventually only one or a few chemicals remain.

Unraveling the mystery

Zhenyu Tian, who was hired by Kolodziej in 2017, had the right skills, ingenuity and perseverance to move the project to completion, though it would take 2 1/2 years of strenuous work, as Kolodziej described it.

As a graduate student at the University of North Carolina at Chapel Hill, Tian had used the non-target approach to analyze soils at a toxic-waste site before and after they underwent biological treatment. He was able to isolate and identify unusual toxic compounds produced during the bioremediation of hydrocarbons, which had triggered cleanup at the site.

At the Center for Urban Waters, Tian tackled the project with gusto, motivated by the knowledge that coho salmon were still dying in the streams while researchers sought out a cause.

For Tian, taking the job in Washington meant a temporary long-distance relationship with his wife, who had taken a good job in New York City. What followed was a lot of phone conversations, online video calls and an occasional rendezvous in distant cities before the pandemic. Being apart also meant that Tian would have few distractions from his work.

“I felt I needed to get this done, and I believed I was the best person to do it,” he said.

The chemical brew used for the analysis came from running water through a mixture of ground-up rubber from nine different kinds of tires, some old and some new. For these experiments, tiny juvenile coho were exposed to test the toxicity after each step of separation.

The original toxic mixture contained more than 2,200 chemical “features,” a term used in mass spectrometry to mean chemicals and their fragments. Through separation techniques, using different methods over and over in trial-and-error fashion, the toxic fraction was narrowed to 1,355 features, then to 659, 225, 24 and then just four. From the final four, the chemical formula of the toxic compound was found to be C18H22N2O2.

Searching in references from tire manufacturers, Tian could find no mention of such a chemical used to make tires. After all this work over two years, was the mystery to remain unsolved?

“This was a question that kept me up at night,” Tian said. “I kept wondering what it could be. Then suddenly, one morning last December, I had a thought go through my mind: This was probably a transformation product!”

He went back to the list of tire chemicals and found a chemical formula similar to the one he had derived, but it had two extra hydrogen atoms and no oxygen atoms. The exchange of hydrogen for oxygen in one of the tire chemicals had transformed it into the toxic compound that had remained a mystery for so long.

“It was kind of like magic,” Tian said. “I knew this had to be the parent chemical.”

The parent chemical, called 6-PPD, is used in rubber formulations to reduce the damage from ozone, which is related to automobile exhaust. In fact, 6-PPD sucked up ozone (a form of oxygen) so well that it turned into the highly toxic transformation product, 6-PPD-quinone.

While the toxicity of 6-PPD has been reported, it appears that 6-PPD-quinone has not been studied, Tian said. In fact, the limited references that he found had misidentified the transformation product as a chemical with a different molecular structure.

To verify the laboratory identification, the research team purchased 6-PPD from an industrial supplier and converted it to 6-PPD-quinone. Effects were measured on coho exposed to the pure compound at various concentrations to produce a standard dose-response curve. Toxic effects from the commercially derived product were consistent with what had been seen with tire leachate and roadway runoff.

To evaluate environmental effects, concentrations of 6-PPD-quinone were measured in samples of roadway runoff collected from various locations in and around Seattle and Los Angeles. Based on the dose-response curve, the researchers concluded that every sample of runoff, as it washes off the roads, contained enough 6-PPD-quinone to kill every coho that comes into contact with it. That’s exactly what had been observed for years in the urban streams of Puget Sound.