Fishing for Answers in the Willamette River
It began more than a decade ago. Little fish with crooked backs were found
living in a stretch of the Willamette River near the town of Newberg, just
south of Portland. The little fish, relatives of minnows and most native to
the river, got people worrying about what could cause such deformed skeletons
in fish. Were they deformed by unseen poisons that could endanger human health
as well? A brand new $47 million water treatment plant sat almost idle a little
ways downstream on the Willamette River at Wilsonville. People refused to drink
the water. What was in the Newberg Pool?
The Oregon Department of Environmental Quality had spent years searching for
toxic chemicals that would explain the fish deformities. Their studies confirmed
that more than half of some species sampled in the Newberg Pool had kinked or
crooked backbones, a much higher frequency of deformities than they had found
in fish living farther upstream. But the DEQ studies were not able to find any
significant difference in chemical contamination between the Newberg Pool and
sites they’d sampled farther upstream. The water quality was the same, but the
fish were definitely different. What was going on in the Newberg Pool?
With more questions than answers, concern grew in the communities along the
lower Willamette River. In 2001, state representative Mae Yih of Albany and
legislators from the Portland Metro area called on Oregon State University
to find out what was causing fish deformities in the Newberg Pool. “This was
a researchable question, not a mystery of nature,” said Larry Curtis, head
of OSU’s Department of Environmental and Molecular Toxicology, “and I knew
we had the science and technology to find the answer.”
Skeletal deformities can be caused by many different things — toxic chemicals,
infectious diseases, hybridization and more. So instead of focusing strictly
on water quality, Curtis engaged OSU faculty from many different disciplines
across the Agricultural Experiment Station to examine the question from several
“Each person came to the study with their own theories of what could be causing
these skeletal deformities,” Curtis said. “The collaboration across disciplines
made it possible to do a comprehensive study that cross-examined evidence from
many different perspectives.”
With $500,000 from the Oregon Watershed Enhancement Board and two years to
complete their report, the scientists went to work in 2002. Almost immediately,
they discovered surprises and new challenges.
Doug Markle, a professor of fisheries and wildlife at OSU, had known about
the Newberg Pool deformities for several years. He had searched the preserved
specimen collections at OSU and beyond to see if such deformities had ever
been observed in the past. They had. His search took him to the Smithsonian
Institution in Washington, D.C., where he found three northern pikeminnows
that had been collected from the Willamette River by a scientific expedition
“We had x-rays taken of the three preserved fish,” Markle said, “and one of
them had skeletal deformities exactly like what we’re seeing now in the Newberg
Next, Markle’s team headed to the field to determine the distribution of deformed
fish in the Willamette River. Sampling fish along the 150-mile stretch between
Newberg and Corvallis, they confirmed DEQ findings that the Newberg Pool had
twice the number of fish deformities than upstream near Corvallis. But by sampling
in between, they found that the distribution of deformed fish was not a simple
“We found another hotspot about halfway between Newberg and Corvallis, near
the Wheatland Ferry,” Markle said. “But between the two hotspots — Wheatland
and Newberg — we found a very low incidence of deformed fish.”
Although they found no geographic pattern of distribution, they did find a
biological pattern. “We found that fish species with the highest incidence
of spinal deformities were broadcast spawners,” Markle explained. “That means
they release their eggs into the water and let the current disperse them. The
young, newly hatched fish swim into shallow water to avoid being eaten by larger
predatory fish. Other fish species rear their young in nests further offshore,
and we found that those species have a far lower rate of spinal deformities.”
Markle’s team also found that among the broadcast spawners, such as northern
pikeminnow and chiselmouth chub, young fish had a higher incidence of spinal
deformities than their older relatives. And they found that fish that hatched
earlier in the season, from May to mid-July, also had higher rates of skeletal
“Because we know that skeletal deformities can be caused by many chemicals
— organophosphates and heavy metals, for example — water chemistry was a large
part of the study,” said Curtis, an expert in environmental toxicology. But
characterizing water chemistry in a body as large as the Willamette River would
be a challenge.
“The Willamette River is not an homogenous pool of unchanging water,” explained
Kim Anderson, director of OSU’s Food Safety and Environmental Stewardship Program.
“Its chemistry changes in pulses, currents and seasonal differences that can
concentrate some chemicals and dilute others. If you take a sample in the afternoon
you could miss a pulse of chemicals released in the middle of the night.”
To further complicate matters, the chemists needed precise measurements of
only the dissolved contaminants that could be absorbed by fish or humans. Measurements
of total contamination in the river would blur their results. So Anderson’s
team designed and built sampling equipment that functioned like surrogate fish,
simulating how a living creature would absorb dissolved contaminants from the
river. Suspended in the water column, the model fish absorbed dissolved contaminants
24 hours a day, 7 days a week, from May through mid-July, when fish deformities
were found to be most prevalent.
One surrogate fish, essentially a thin plastic tube filled with a kind of
vegetable oil, absorbed organophosphates in the same way fatty tissues do in
fish or humans. The other surrogate fish looked like a plastic milk jug lid
and was filled with a resin that would absorb heavy metals.
The chemists knew that the contaminants they were trying to track could be
so dilute as to be undetectable using laboratory methods specified by the U.S.
Environmental Protection Agency (EPA). So, Anderson’s team developed new analytical
methods that were 1,000 times more sensitive than the EPA methods, that could
detect chemical compounds diluted to fractions of one part per trillion.
Anderson put that tiny fraction into perspective. “One part per trillion is
one second in 32,500 years,” she explained. “Or, if you put one orange golf
ball on the football field at OSU’s Reser Stadium and added one trillion white
golf balls, then the stadium would need to be one mile tall.”
Even at this extraordinary level of laboratory precision, the chemists found
that most of the target chemicals in the Willamette River were still below
their limits of detection. They examined river sediments for persistent organic
toxins, and the few they were able to detect were far below any minimum exposure
levels set for human health. They even removed the ovaries from a sample of
pikeminnows just before spawning to see if there were traces of toxins in the
eggs that could disrupt bone development in young fish. They found nothing
that would explain the high rates of fish deformities in the Newberg Pool.
Still, the researchers challenged themselves. Perhaps, they argued, there
were other chemicals in the river that could contribute to fish deformities
besides the target compounds they were monitoring. Or perhaps one chemical,
that by itself was harmless, could become toxic in combination with trace amounts
of other seemingly harmless chemicals.
OSU toxicologist Jeffrey Jenkins took up the challenge.
Because you can’t analyze river water for every possible chemical contaminant
mixture, Jenkins designed a test to expose fish in the laboratory to organic
contaminants concentrated from river water. Jenkins and his team collected
water samples from the Newberg Pool and other locations in the Willamette,
then concentrated the water into extracts that intensified the dose of existing
contaminants. In the laboratory, they reared fish in these various water extracts
and examined them for spinal deformities or other irregularities.
Jenkins’ team found that very few fish, two percent or less in each water
extract, developed spinal deformities similar to those found in river fish.
Finding no differences among the various water extracts, Jenkins found no evidence
that unknown elements in the river water could be linked to fish deformities.
Examining every angle, the OSU researchers were not able to find a connection
between water chemistry and fish deformities. But the concern expressed by
people living along the lower Willamette was as much about their own health,
and their doubts about the new water treatment plant, as it was about fish.
Was Willamette River water safe to drink? Jenkins took his research further
to consider how much risk to human health was posed by chemical pollutants
in the Willamette River.
It’s impossible to calculate the risk from chemicals you can’t even detect.
So Jenkins examined data from water samples collected by the U.S. Geological
Survey from upstream tributaries of the Willamette. He hoped to find measurable
amounts of pesticides before they were diluted by the water in the main stem
river. Testing for the occurrence of nearly 100 different components of pesticides,
he detected the presence of fewer than half, all at extremely low concentrations.
Assuming these chemicals could make their way to the main stem, Jenkins calculated
what the concentration of those pesticides would be in the Willamette at the
point where river water entered the disputed water treatment plant at Wilsonville.
Jenkins compared his calculations with drinking water standards from EPA and
the World Health Organization. The calculated pesticide levels fell far below
the allowable limits, a thousand to a million times — in one case a billion
times — below the accepted limits for drinking water.
The Willamette River water quality was passing every test. So what was causing
the fish deformities?
Michael Kent, director of the Center for Fish Disease Research at OSU, had
cultured tissues from the deformed fish, searching for evidence of bacteria
or viruses. He had found nothing. But when he examined the skeletal tissue
of deformed fish in relation to x-rays that Markle provided, he found tiny
parasites wedged into the bones of the little fish.
Parasites live complicated lives. The most prevalent parasite Kent’s team
found was a kind of fluke that begins its life inside a snail. Later it’s released
into the water where it infects young minnows taking shelter in the shallows.
The fluke parasite drills into the bones of newly hatched fish, then encases
itself in a cyst that disrupts normal bone development. Minnows with deformed
backs swim slowly and awkwardly. They’re more visible and therefore more likely
to be eaten by birds, the ultimate host for the reproductive stage of the parasite’s
life cycle. Inside the bird, the parasite reproduces and its offspring are
excreted, finding their way back into snails, beginning the cycle again.
The biologists confirmed the presence of snails in the shallow water where
young broadcast-spawning minnows seek shelter.
Kent examined the tissues of hundreds of deformed fish collected from four
sites between the Newberg Pool and Corvallis. In more than 80 percent of the
fish he found two kinds of parasites attached to the bone deformities.
One question lingered: were parasites actually causing the deformities, or
were deformed fish simply more vulnerable to parasitic infection? Kent tested
the question in the lab, exposing healthy laboratory-born minnows to infected
snails. The laboratory fish developed curved spines, split vertebrae and other
skeletal deformities to the same degree as those found in the Newberg Pool.
Kent demonstrated that it was the presence of parasites when and where a fish
was born that was the strongest predictor of skeletal deformities.
The evidence undeniably pointed to parasites. Other possibilities, such as
water chemistry, infectious disease and toxic contamination, had been carefully
examined but could not explain the deformities.
“Such lack of evidence from any single investigation would seem to have little
significance,” Kent said, “but the negative results from so many lines of investigations
served to strengthen the positive results we found with parasites.”
Discovering the link between parasites and fish deformities answered some
questions and prompted more. Why, for example, is there a higher incidence
of parasitic infections in the Newberg Pool than elsewhere in the Willamette
basin? Doug Markle speculated that perhaps there are more snails in the Newberg
Pool or fish with a greater susceptibility to parasites. In any case, unlike
toxic chemicals, the parasites pose little or no risk to human health. Cooking
or freezing will kill the parasites in infected fish.
On time and on budget, the OSU scientists delivered their report in July,
2004, with a public hearing in Wilsonville to explain to the community what
they had found. Both the supporters and the detractors of the new water-treatment
plant were pleased with the report’s results.
“That’s a sign we’ve been an honest broker of information and produced a balanced
report,” Curtis said. “There are new questions that would be interesting to
pursue, but this wasn’t an invitation for open-ended research. We were called
on to answer one question — what’s causing fish deformities in the Newberg
Pool? — and we answered it.”