Asian carp in Minnesota? If you net one, yes. But DNA won't tell

April 09, 2013

Were the Asian carp here or weren't they?

In 2011, alarms rang as researchers reported finding signs of silver carp well into Minnesota waters.

Then, last week, a new study revealed no evidence of the invasive fish's DNA anywhere in the the state, including those same waters.

So are the fish here?

The short answer is that no one can say for certain.

The longer answer provides a glimpse into the evolving science of environmental DNA -- or "eDNA" -- and its limits.

DNA from the potentially devastating fish was discovered above the Coon Rapids Dam on the Mississippi River and in the St. Croix River just below the St. Croix Falls, Wis., dam, it was reported last year.

Silver and bighead carp, referred to collectively as Asian carp, are feared because they can wreak havoc on the food chain.

The Coon Rapids finding was especially confounding because it had previously seemed unlikely that silver carp, despite their notorious leaping ability, could have surmounted so many dams above their known haunts closer to Iowa so quickly.

The new study contained rigorous controls to avoid false positive results, researchers said. In the past two years, studies have suggested that eDNA techniques could lead researchers to falsely conclude silver carp DNA was present in water samples.

The new results suggest that might have happened in Minnesota in 2011.

But it also suggests that controlling for such false positives could reduce the sensitivity of the technique so much as to render it useless.

Still, the potential for eDNA is so powerful that refining it is essential, scientists say.

At present, the most reliable technique to determine where a fish lives is to net one. In the case of Asian carp, that involves hiring commercial netters, a major operation.

By contrast, eDNA could allow scientists to detect whether carp are present and in what densities and of what age by analyzing a cup of water skimmed from the surface.


When DNA is mentioned in criminal prosecutions today, it's often more persuasive than a smoking gun, even better than a fingerprint.

But the DNA of silver carp detected in 2011 -- or believed to be detected -- isn't the same as finding a strand of hair at a crime scene.

The science of eDNA, developed at the University of Notre Dame, has sought to detect silver and bighead carp since 2009.

The technique seeks to isolate fragments of DNA -- cells from scales, urine or mucous -- shed by Asian carp. They shed more such cells than many fish, so they're believed to be good candidates for scooping a pail of water and finding signs of their DNA.

But the techniques used in the past didn't actually match the DNA found in the water with known sequences of an Asian carp's genome. (The entire genetic code has not been mapped for either Asian carp.) Instead, scientists looked for segments of DNA that matched the size, basic chemistry and behavior of an Asian carp.

But with thousands of organisms living in the Mississippi River and elsewhere, it's possible that a fragment of DNA shed by, say, a common carp, could appear to "match" a part of silver carp DNA. It's even possible that part, but not all, of the DNA could match. Either way, that's a false positive result.

"EDNA would be a bad way to tell a chimp from a human being, without sequencing," said Peter Sorensen, director of the Aquatic Invasive Species Research Center at the University of Minnesota and one of the organizers of the new study.

By "sequencing," Sorensen was referring to a rigorous protocol developed by scientists at the Army Corp of Engineers to avoid false positives: Sequence 240 base pairs of the DNA found in samples. If all 240 base pairs match a known sequence of Asian carp DNA, the odds of a false positive drop to less than 1 percent.

"That's a fingerprint," Sorensen said. "Otherwise, you have something like a description: white male with brown hair. That could be me. But it could also be a lot of people."

2011 VS. 2012

The 2011 study, done by a private company, didn't employ that DNA sequencing regimen, Sorensen said.

The most recent study, which examined water samples taken in 2012, did. And it found nothing in Minnesota.

Even a positive result wouldn't necessarily mean live fish are present. Scientists have imagined any number of scenarios in which Asian carp DNA might be present without live fish. For example, a bird migrating north in the spring could feast on a carp in Arkansas and deposit its cells in Minnesota days later when it defecates.

Sorensen said one reason he supported the new study, which was conducted with the Minnesota Department of Natural Resources and the U.S. Geological Survey, was to see the effect of the rigors of DNA sequencing. Rather than always second-guess whether a positive result was from a fish or a bird, he wanted to first eliminate the possibility that a positive result was false.

The new study found numerous DNA samples that would have tested positive under the old regimen. But when they were broken down further and fully sequenced, none matched Asian carp.

"The DNA sequencing brings it to another level," he said. "One of the things that came out of this study was we found many pieces of DNA that have the same size as many silver carp fragments. The sequencing was crucial in eliminating them. ... If you look at it as a test of a test, it was very illuminating."

No one can subject the 2011 water samples to the new rigors because those samples weren't preserved. That's why it's impossible to know whether the 2011 results truly indicated the presence of silver carp DNA at all, regardless of the source.

Researchers affiliated with the new study said it was the most rigorous test yet given to eDNA methods.

But was the technique too rigorous?


Researchers used another control: They conducted identical tests on waters known to be infested with Asian carp.

The waters below Lock and Dam No. 19 in Keokuk, Iowa, are heavily infested with both bighead and silver carp.

When tested for silver carp, 68 percent of the samples tested positive after the rigors of sequencing. So the test could work. However, those same results reveal that 32 percent of the time, the test produced a false negative.

In the case of bighead carp, the testing, after the DNA sequencing, was unable to detect their presence.

That's a big fat carp-sized false negative.

"I thought this was possible, both the possibility of false negatives and false positives," Sorensen said. "I was a little surprised at how clear this was in the data. ... In the case of the bighead carp for certain, it makes me wonder what the value is. We know the fish were there."

The reasons for the false negatives aren't clear. Sorensen said he can't rule out that something in the water is masking the presence of DNA, or that researchers were testing the wrong part of the water column, or that bighead DNA degrades rapidly.

"The technique is less than perfect, and that's one of our primary findings," Sorensen said. "It still has enormous promise, but we're very early on in this technique, and it might take 10 years before it can be useful in detecting the presence or absence of carp."

Practically speaking, the $300,000 study means one thing: eDNA won't be considered anything more than a research tool for the time being.

"I had not formed an opinion of that before these results," said Steve Hirsch, director of the DNR's Ecological and Waters Division. "I wasn't sure what to make of the eDNA tests from 2011. The 2012 tests raised a lot of clarity. ... Our effort is going to be focused on netting."

Hirsch, Sorensen and others involved in the study said they expect some skepticism for a $300,000 study that concluded a technique was flawed.

Here's how Sorensen responded to that: "The federal government is spending millions for eDNA sampling in the Illinois River and the Great Lakes. As a taxpayer, I feel spending $150,000 of state money (and $150,000 in federal funds) to evaluate this technique is money well spent."