The salmon were dying and nobody knew why.
About 20 years ago, ambitious restoration projects had brought coho salmon back to urban creeks in the Seattle area. But after it rained, the fish would display strange behaviors: listing to one side, rolling over, swimming in circles. Within hours they would die — before spawning, taking the next generation with them. In some streams, up to 90 percent of coho salmon were lost.
“To be running into these sick fish was fairly astonishing,” said Jenifer McIntyre, now a toxicologist and professor at Washington State University who is part of a team that, years later, has finally solved the mystery of the dying salmon around Puget Sound. “In those early years, we debated intensely, what could be the cause of this?”
The team’s findings were published Thursday in the journal Science.
The investigation began with a forensic examination. Was it a metal or some other chemical in the water? Nothing they could find. A problem with the temperature? No. Perhaps lack of oxygen? The salmon looked as though they were suffocating, but they had plenty to breathe. There was no evidence of disease or pesticide exposure. But the connection to rain and the lack of any other explanation led Dr. McIntyre and her team to focus on runoff from roads.
Partnering with a local fish hatchery run by the Suquamish Tribe, they decided to put the theory to the test, exposing fish to a mixture they created of chemicals they knew to be in roadway runoff, like heavy metals and hydrocarbons from motor oil. But the salmon were unaffected, even at surprisingly high concentrations.
The scientists decided to try again with the real stuff, actual runoff. Luckily for them, a downspout from an elevated road emptied into the parking lot at the Northwest Fisheries Science Center, where some of the team members worked. On a rainy day in 2012, they filled stainless steel containers with a translucent dark liquid coming out of the spout. This time, the salmon exhibited the bizarre symptoms and promptly died.
“What is it in that mixture?” Dr. McIntyre recalled thinking. “This is just water that’s on the road, it’s what we tramp through in our rain boots.” It must be something that people don’t regularly measure, she figured.
Enter Edward P. Kolodziej, an environmental engineer and chemist at the University of Washington. His lab used a machine called a high resolution mass spectrometer to compare the chemical composition of highway runoff with that of water collected from two urban creeks where the salmon were dying. The samples shared chemicals related to tire particles. So the team brewed up a test concoction by soaking shredded tire tread in water. The salmon died.
They were getting closer to the answer, but the tire water still contained more than 2,000 chemicals. To solve the mystery, they had to identify the specific culprit. Dr. Kolodziej and other researchers painstakingly narrowed the field, separating the tire solution into different chemical combinations and then testing them on fish. With a Venn diagram-type approach, they got their list down to 200 chemicals. Whenever they identified one that was known in the literature to be toxic to fish, they purchased it and tried out that individual chemical.
“We’d almost take bets in lab as to whether or not the chemical that we thought might be doing it would kill the fish,” Dr. Kolodziej said. “And it never did.” Not the flame retardants. Not the plasticizers. Not a bunch of others that you’ve never heard of.
“We were stuck,” Zhenyu Tian, a research scientist who conducted many of the analyses, said.
“Depressed,” Dr. Kolodziej said.
Then a Ph.D. student, Haoqi Nina Zhao, suggested a new way to separate out chemicals that led to a prime suspect. But they couldn’t test it, because they didn’t know what it was.
“It’s almost like you have a fingerprint,” Dr. Tian said. “But you really don’t know who this is, because in your database, this fingerprint doesn’t exist.”
Dr. Tian’s “aha!” moment came one morning. Guessing that the mystery chemical had transformed from a substance originally added to the tire, he looked for a compound whose carbon and nitrogen molecules matched, ignoring the oxygen and hydrogen, since the latter are more likely to be altered when a chemical transforms. In an Environmental Protection Agency report on tire rubber, he found a match: an antioxidant called 6PPD.
The researchers ordered the smallest amount they could, about a pound of purple pellets. When they oxidized the substance, the resulting chemical looked just like the one they had worked so hard to isolate from the tire water. It was time to test this version, 6PPD-quinone, on the salmon.
“I find it incredibly sad to watch fish die,” Dr. Kolodziej said. “You’re just watching these fish struggle. And yet you’re happy you understand why.”
The killer was the 6PPD-quinone from the tires in the roadway runoff.
“The analysis that they did is really amazing,” said Nancy Denslow, a professor and director of aquatic toxicology at the University of Florida who was not affiliated with the study. She also praised the large number of authors. “It’s wonderful to see big groups of people coming together to solve problems,” she said. “Group science is fantastic.”
Their answer raises so many questions, however, that Dr. McIntyre, the toxicologist who watched disoriented salmon in creeks 15 years ago, now has even more work to do.
She has forthcoming research about how roadway runoff affects some other species of fish (not nearly as dramatically, but there are still consequences). The team is in conversations with the tire industry and hopes manufacturers will be willing to look for a replacement preservative. The scientists are concerned about broader health impacts from the chemicals in tires, including on humans, especially because tires are often recycled to make artificial turf for sports fields. “It seems to me that there could be inhalation of those finer particles,” Dr. McIntyre said. “Now you’ve got that leaching happening on the lung tissue.”
While chemicals have always surrounded us (plants themselves are chemical factories), within the last hundred years, humans have been making them synthetically. “We’ve been synthesizing them kind of faster than we can keep up,” Dr. Kolodziej said.
“I think the vast majority of those are fine, but there are bad actor chemicals floating around out there,” he said. “And it’s a long, slow and difficult process to identify them.”