Perch larvae

A study suggesting that larvae of perch (above) are harmed by microplastics has drawn charges of research misconduct.

jpa1999/iStockphoto

Misconduct allegations fly in spat over paper on microplastics and fish larvae

When Fredrik Jutfelt and Josefin Sundin read a paper on a hot environmental issue in the 3 June issue of Science, the two researchers immediately felt that something was very wrong. Both knew Oona Lönnstedt,  the research fellow at Sweden's Uppsala University (UU) who had conducted the study, and both had been at the Ar research station on the island of Gotland around the time that Lönnstedt says she carried out the experiments, which showed that tiny particles called microplastics can harm fish larvae. Jutfelt, an associate professor at the Norwegian University of Science and Technology in Trondheim, and Sundin, a UU postdoc, believed there was no way that Lönnstedt had been able to carry out the elaborate study.

Less than 3 weeks later, the duo wrote UU that they had "a strong suspicion of research misconduct" and asked for an investigation. Their letter, initially reported by Retraction Watch in August, was cosigned by five scientists from Canada, Switzerland, and Australia, who hadn’t been at the research station but also had severe misgivings about the paper and who helped Sundin and Jutfelt build their case.

There was just a massive discrepancy between what we saw in that room and have pictures of, and what was presented in the paper.

Fredrik Jutfelt, Norwegian University of Science and Technology

This week, Science is publishing an “Editorial expression of concern” about the paper, because Lönnstedt and her supervisor at UU, Peter Eklöv, have been unable to produce all of the raw data behind their results. Lönnstedt says the data were stored on a laptop computer that was stolen from her husband's car 10 days after the paper was published, and that no backups exist.

The ultimate fate of the paper is still very unclear, however. On 31 August, a three-member expert panel at UU finished a preliminary investigation that cleared the accused of misconduct and recommended against a full investigation. The panel even berated Sundin and Jutfelt for blowing the whistle; most of their objections "come within the ambit of normal scholarly discussion, which could have been conducted directly with the authors of the article," it wrote.

But a second investigation, by Sweden's Central Ethical Review Board, is still ongoing. Science Editor Andrew Sugden says the journal initially wanted to wait for this board's verdict before taking action, but decided to publish the expression of concern because the board has taken longer than expected and readers needed to know data for the paper are missing. (Science requires raw data to be published on its website as supplementary online material or placed in an online archive such as the Dryad Digital Repository; the UU team failed to deposit all of its data when the paper was published.)

Both the accused and the whistleblowers say that they believe the ongoing inquiry will vindicate them, and Lönnstedt tells ScienceInsider that the authors of the letter to UU are "lying."

The paper, which received a lot of press attention, focused on plastic fragments of less than half a millimeter in size that result from the mechanical breakdown of bags and other products. There’s increasing evidence that these microplastics collect in rivers, lakes, and oceans around the world, but so far, little is known about their effects on aquatic organisms and ecosystems. What Lönnstedt and Eklöv reported was alarming: They had exposed larvae of European perch maintained in aquaria at the research station to microplastics and found that they had decreased growth and altered feeding and behavior. Microplastics made the larvae less responsive to chemical warning signals and more likely to be eaten by pike in a series of predation experiments, the pair further reported. In an accompanying Perspective, Chelsea Rochman of the University of Toronto in Canada wrote that the study "marks an important step toward understanding of microplastics" and was relevant to policymakers.

Sundin says she was familiar with the study because she was at the research station between 8 April and 16 June of 2015; Lönnstedt tells ScienceInsider the experiments took place in April and May. Sundin says she even took care of Lönnstedt's fish while Lönnstedt was away for a weekend. In documentation provided to both investigations, Sundin and the other complainants provided a list of 11 problems with the Science paper, many partially relying on what she and Jutfelt—who visited the island from 5 to 8 May—observed up close.

They claim, for example, that the microplastics didn't arrive at the lab until early May, that Lönnstedt arrived at the lab on 4 May, and began exposing her the larvae the next day. She left on 15 May, the duo claims, which would not have allowed her to do the 3-week study described in the Science paper. The study, they say, would have required 30 1-liter beakers; Jutfelt and Sundin saw only 18 beakers, 15 of which were 0.6 liters in size. (Photos taken by them and a photographer of a UU magazine during the experiments back this up, the duo says.) They add that the full study would have used far more perch eggs than Lönnstedt had at her disposal. Nor were there enough pike for the predation experiments, the duo says; and Sundin says she should know because she collected the pike used for the study before Lönnstedt arrived. "There was just a massive discrepancy between what we saw in that room and have pictures of, and what was presented the paper," Jutfelt says.

Lönnstedt challenges many of these claims. She says that she was at the station “all of May.” Sundin and Jutfelt not seeing those experiments doesn't mean they didn't take place, she says. Lönnstedt also says she found enough perch eggs for the study, and caught extra pike herself in addition to those provided by Sundin.

Lönnstedt acknowledges that given the laptop theft, part of the raw data behind the experiment is missing. Because of glitches in UU's computer system, there is no backup of the data, she adds. Lönnstedt says the theft, which she wrote about on her Facebook account, occurred on the night of 12 June or very early on 13 June—less than 24 hours before a Science editor informed her that the journal had learned that she had not posted raw data for the experiments, and requested that she do so as soon as possible. "I totally understand" that Science is alerting readers to this problem, she says.  

Jutfelt calls the theft of the laptop just as questions about the data were arising, along with the backup problems, “a ridiculous coincidence. … It's a dog-ate-my-homework defense," he says.

Eklöv, Lönnstedt's co-author and supervisor, did not respond to Science's requests for comment; Jutfelt and Sundin say he did not visit the island when the study took place.

In the report of its "preliminary investigation," the UU panel sided with Lönnstedt. She and Eklöv had explained everything "in a satisfactory and credible manner," wrote the panel, which asked UU to "take diligent steps to restore the reputation of the accused." But the panel’s report didn't provide detailed rebuttals of the long list of problems provided by Sundin and Jutfelt, who say that the investigation was superficial.

Both sides say the case has taken a big personal toll. Jutfelt and Sundin say they have spent an enormous amount of time and energy preparing and documenting their case and agonized about blowing the whistle on a researcher they considered a friend; Sundin calls it "a very painful decision."

Lönnstedt says she feels the complaints are a "personal attack" that she believes are rooted in "jealousy." She also accuses her accusers of "hacking into my Facebook account" to find evidence. Jutfelt and Sundin acknowledge that they included posts from Lönnstedt to document her whereabouts at various times and her claim of the theft, but say the account was accessible to anyone at the time. "We also got legal advice from Uppsala University; they said if it's relevant to the case, it's not libel or slander," he says.

Much may now depend on the conclusions of an expert group on misconduct at Sweden's Central Ethical Board, which is doing its own, independent investigation. Jutfelt says he's hopeful because it appears that the group is "doing a more thorough job." Lönnstedt says she's not worried about the outcome. A spokesperson for the board says it is not clear when it will wrap up the inquiry.