OREANDA-NEWS It's a cold, dreary day in early March, and Josefin Sundin is standing in one of the two aquarium rooms at the Ar Research Station on a remote corner of Gotland, a Swedish island in the Baltic Sea. "This is where it all happened," she says, while gazing around as if searching for fresh clues. Her colleague and friend Fredrik Jutfelt takes cellphone pictures.

Nine months ago, these two researchers triggered a scandal in Swedish science by accusing another friend and colleague of making up research supposedly done here. Now, they have returned to Gotland to discuss what happened—and how whistleblowing has taken over their lives. The station is deserted; the 2017 research season has yet to start. But the station manager, Anders Nissling, has made a pot of strong coffee and is happy to give a tour of the offices and laboratories where researchers come to study the creatures and ecosystems of the sea and a nearby lake.

At the heart of the case is a three-page paper that made headlines after it was published in Science* on 3 June 2016. It showed that, given a choice between a natural diet and tiny plastic fragments, perch larvae will consume the plastic "like teens eat fast food," as a BBC story put it. This unhealthy appetite reduced their growth and made them more vulnerable to predators. It was a dire warning, suggesting the plastic trash washing into rivers, lakes, and oceans was creating ecological havoc.

The study was also, Sundin and Jutfelt claim, "a complete fantasy." It was purportedly done at the Ar station in the spring of 2015 by Oona Lönnstedt, a research fellow at Sweden's Uppsala University (UU); her supervisor and only co-author, Peter Eklöv, did not work on the island. Sundin, a postdoc at UU, was working at the station at that time, too, and occasionally lent Lönnstedt a hand. But she saw no sign of a study of the scope and size described in Science.

Jutfelt, who like Sundin is Swedish but works as an associate professor at the Norwegian University of Science and Technology in Trondheim, also spent a few days at the station when the study supposedly took place, and didn't see it either. Lönnstedt wasn't even on the island long enough to do the study described in Science, the duo claims. Many other details were, well, fishy, they said, such as Lönnstedt's claim that part of the study's data was forever lost because her laptop was stolen 10 days after the paper was published.

A group of five aquatic ecologists and physiologists elsewhere in the world has helped Sundin and Jutfelt sort through a mounting pile of evidence and make their case that the work was fraudulent. But Lönnstedt and Eklöv have denied any wrongdoing. "Of course I did these experiments," Lönnstedt told Science this past December. She said the allegations were fueled by "jealousy" on Sundin's part. "If you compare my CV with her CV … then yeah, there is a big difference," she said. Lönnstedt is currently on leave from the university and didn't respond to requests this month for a follow-up interview; Eklöv has declined to answer questions altogether.

Last August, a panel charged by UU to conduct a preliminary inquiry dismissed the charges and suggested that Sundin, Jutfelt, and their colleagues had unfairly maligned Lönnstedt and Eklöv. But a second, in-depth investigation by a panel at the Central Ethical Review Board (CEPN) in Stockholm is ongoing, and an expert hired by that group recently delivered a more damning report that raised the possibility of fraud. CEPN is expected to issue a final statement in April.

The outcome may have an impact well beyond four lives and careers. Sweden is still recovering from the scandal around celebrity surgeon Paolo Macchiarini, who was fired last year for ethical breaches that his university, the Karolinska Institute in Stockholm, had initially dismissed. The case shook confidence in Swedish science and raised concerns about Swedish universities' ability to investigate their own researchers. If UU, too, bungled its investigation, as the whistleblowers in this case claim, it could bolster support for a plan released last month that would take misconduct investigations out of university hands and transfer them to a new government agency.

The case has raised a host of other issues as well. Dominique Roche of the University of Neuchâtel in Switzerland, one of five scientists supporting Sundin and Jutfelt, is critical of Science, which didn't issue a so-called editorial expression of concern about the paper until December 2016. Roche says the journal itself should have investigated the paper, which has racked up 36 citations. Others argue the case shows that the fields of ecology and evolution have been too slow to adopt the kind of transparent practices that build trust and help prevent misconduct.

"Accomplished, driven and outcome-focused, I have an excellent record as a reliable and highly productive employee," a CV published on Lönnstedt's personal website says. Lönnstedt earned her Ph.D. in marine biology at James Cook University (JCU) in Townsville, Australia, just 3 years ago. But several of her papers—including work showing how lionfish use their fins to send each other invitations for a collective hunt—have already attracted press attention. She has also explored how environmental problems such as ocean acidification, coral bleaching, and invasive species affect fish behavior. "She lives for research and is a very dedicated and ethical researcher," says Mark McCormick, a former supervisor at JCU and a co-author on more than 15 of Lönnstedt's papers.

After she returned to Sweden in 2014, Lönnstedt turned to a new threat: microplastics. The term refers to plastic particles smaller than 5 millimeters, which include the "microbeads" in skin scrubs and plastic detritus broken down by mechanical forces, sunlight, and weather.

In their Science paper, Lönnstedt and Eklöv claimed that European perch larvae—which are more vulnerable to pollution than adult fish—prefer to eat 0.09-millimeter polystyrene beads over a standard food, tiny Artemia brine shrimp. Experiments also showed that plastic-consuming larvae were less able to recognize chemical alarm cues when exposed to pike, a predator fish, and as a result were far more likely to end up in a pike's stomach. The findings might explain why the number of young perch entering the Baltic has been dropping, they wrote.

"I was quite impressed," says Chelsea Rochman of the University of Toronto in Canada, who penned a commentary praising the work's policy relevance in the same issue. Most previous research used higher doses of microplastics, says Rochman, which makes it easier to see effects but raises questions about real-world relevance; Lönnstedt and Eklöv used levels actually found in the environment. Rochman adds that past studies usually focused on cells, gene expression, or individuals. "This was one of the first to ask more ecologically relevant questions." She was not surprised to see it end up in Science. Five months after the study was published, Lönnstedt received a $330,000 grant for "future research leaders" from Formas, a Swedish funding agency, for her work on microplastics.

Sundin remembers the moment she began to read the paper. "I thought I was losing my mind," she says. "There was a description of this big experiment, and I had absolutely no recollection of it." She discussed it with Jutfelt, and both agreed that neither the logistical nor the scientific details added up. The same day, they began email and Skype discussions with the other five scientists, who they knew from conferences and fieldwork on Australia's Great Barrier Reef. "We're all pretty anal about good science," Jutfelt says.

Larval pike prey on perch hatching from translucent eggs. The disputed paper reported that eating tiny morsels of plastic makes perch more vulnerable to predation.
© Blickwinkel/Alamy Stock Photo
"We thought about whether we should let it slide, whether it was too much for us to take on," recalls Timothy Clark of the University of Tasmania in Hobart, Australia. "Whistleblowing is risky, it can affect your future employability," he says—a special risk for Sundin, who doesn't yet have a permanent job. "But I think she couldn't have lived with herself if she hadn't done it." The group also worried that attacking the study might suggest they aren't concerned about microplastics. They are, very much.

On 16 June 2016, the researchers sent the authors 20 questions about the Science paper. Four days later, they also asked the university to launch a preliminary investigation. The timeline was a central issue, they told the UU panel handling the inquiry. Lönnstedt began the microplastics study on 5 May 2015 and left the island on 15 May, they claimed; as part of the evidence, they provided a photo posted on Lönnstedt's Facebook account on 16 May showing her sipping champagne with a friend in Stockholm. Lönnstedt didn't return to Ar that month or the next, so they concluded there wasn't time for the study described in Science, which would have taken at least 3 weeks.

Many other things didn't add up. The study would have required the simultaneous use of 30 aquaria of 1 liter each. Jutfelt had taken a picture of Lönnstedt's laboratory setup, showing only 18 beakers; some held a different fish species, and only three could hold a liter. Sundin says she collected the juvenile pike used in the study at a bog 65 kilometers away on 30 April 2015 and gave Lönnstedt only a few—not enough for the study. If Lönnstedt got more pike on her own, Sundin says, it's not clear how she traveled to the bog (Lönnstedt doesn't drive), or why she didn't record the catch in a log book, as required.

Then there were the missing data. Science requires authors to place the data underlying papers on the journal's website or in a public repository. Lönnstedt and Eklöv had not done that, and after publication it became clear that they would not be able to provide all the data. Lönnstedt claims her laptop and backup drives were stolen from an unlocked car on 12 or 13 June 2016, just before Science asked her to correct the omission. Other backups didn't exist, Lönnstedt and Eklöv claim, because a university server malfunctioned. Lönnstedt says only 15% of the data is still missing; a table compiled by the whistleblowers suggests it's 78%.

In letters to the UU panel, Lönnstedt and Eklöv addressed the allegations and answered the whistleblowers' 20 questions. The setup with 18 beakers was a pilot study, they said; the experiments described in Science were housed elsewhere, but three photos documenting that were on the stolen laptop. The Facebook photo from Stockholm didn't prove anything, because Lönnstedt sometimes waited weeks to post. They did not address other issues, such as the pike collection, in detail.

The UU panel was satisfied. Lönnstedt and Eklöv had "thoroughly answered and explained" every issue in a "satisfactory and credible manner," the group wrote in a three-page report on 31 August. The missing data were partly the result of a misunderstanding, it said, adding, incorrectly, that "all necessary raw data has been freely available … for some time." The whistleblowers, the panel said, "appear to have a very strong desire" for a misconduct investigation, but most of their issues could have been aired in "normal scholarly discussion."

She and Eklöv lashed out at the whistleblowers, questioning their motivations and tactics. Lönnstedt alleged to Science that the critics obtained her Facebook photos by "hacking" into her account. (Sundin and Jutfelt respond that they wouldn't know how to do that, and say that Lönnstedt's posts were visible to anyone at the time.) Eklöv wrote to the panel that for Sundin to criticize the study, after having helped Lönnstedt in the lab the year before, was "very unethical" and said the use of private photos as evidence was "highly distasteful."

Even before the UU review was complete, Sundin and Jutfelt had made use of their right to request that an expert panel on misconduct at CEPN investigate the matter as well. Lacking expertise in aquatic ecology, that group hired Stockholm University zoomorphologist Bertil Borg to investigate. He drew very different conclusions than the UU panel.

Borg's 19-page report, delivered 23 February, says the accused haven't provided satisfying answers to many questions, and uses words like "strange," "serious," and "quite remarkable" to describe the remaining issues. The timeline remains a crucial problem, Borg wrote. Although Lönnstedt claimed she was in Gotland at least through 20 May 2016, she had not produced tickets, photos, or emails proving that. There were no lab notebooks, and it was "very grave" that Sundin, Jutfelt, and two other witnesses reported that the experiments never took place, Borg said.

Borg also stumbled on a new problem. In the paper, Lönnstedt and Eklöv wrote that they had obtained an ethical permit for the study; they told the UU panel that it arrived 2 weeks after the study began. But Borg discovered that the permit was granted more than a month after the experiments ended, and that it was for a different study design and a different field station. (Lönnstedt and Eklöv blamed miscommunication.) "This renders their credibility questionable in other areas as well," Borg wrote. He stopped short of saying that fraud had occurred, however, saying, "the conclusion is simply that the suspicions of deceit cannot be denied."

Lönnstedt and Eklöv questioned Borg's independence in a response sent to CEPN by a law firm. Borg has co-authored papers with two scientists with connections to Sundin and Jutfelt, it notes. They also questioned Borg's expertise and criticized several of his findings as well as the witnesses' credibility, but they did not provide new evidence that the study described in Science happened.

Borg's ambiguous conclusion left Sundin and Jutfelt anxious. "We might get our reputation back as honest researchers, or the story is that we throw dirt around and we're just jealous of their nice publication," Jutfelt says. Borg declined to talk to Science. Clark says Borg may want the expert panel to make the final decision. "Putting myself in his shoes, he doesn't want to send somebody to the chopping block if he isn't 100% sure," he says. But CEPN's head of office, Jörgen Svidén, says the panel, scheduled to meet on 4 April, will be definitive about whether misconduct has occurred. "It will be yes or no."

Whatever the outcome, the contrasts between the two reports have triggered debate about whether the preliminary university investigation was thorough enough. The panel's three members—two of whom came from universities other than UU—have declined to discuss their work, pending the CEPN decision. But Roche believes "it was completely botched." In general, university investigations in Sweden tend "to sweep things under the carpet," adds Bengt Gerdin, a surgery professor emeritus at UU whose damning investigation of Macchiarini—now substantiated—was dismissed by Karolinska in 2015.

One problem is that Swedish universities have a "mishmash" of regulations and definitions of misconduct, Svidén says. In a report to the Swedish government in late February, a group led by UU literature professor Margaretha Fahlgren proposed letting a new government agency, the Research Misconduct Board, handle all investigations. Gerdin says that would be "a step forward."

Roche says journals should also serve as a safeguard. The suspect paper might not have been published if Science had first checked whether Lönnstedt and Eklöv had posted their data, he says. And once Science learned of the missing data, Roche adds, it should have started its own investigation. He's also annoyed that the editorial expression of concern is very hard to find on Science's website, and does not appear on the PDF version of the paper. "It should be very visible, front and center," he says.

Science Deputy Editor Andrew Sugden agrees about the note's visibility. "That's a problem we're going to fix," he says. He also acknowledges that the data omission slipped through. But Science, which follows misconduct guidelines set by the Committee on Publication Ethics, has no choice but to rely on institutional investigations, he says; "We're simply not geared up to do that. We're not investigators." After UU dismissed the case, Sugden says it made sense to await CEPN's report before taking further action.

The University of Tasmania's Clark shared another lesson from the experience in a Nature op-ed last month. Studies of animal behavior and other research that relies primarily on human observations should routinely be videotaped, he wrote, to show that the studies actually took place. "If extreme athletes can use self-mounted cameras to record their wildest adventures during mountaintop blizzards, scientists have little excuse not to record what goes on in lab and field studies," Clark wrote.

The four central characters, meanwhile, agree on one thing: The affair has been extremely stressful. "I am on sick leave for depression until further notice," Lönnstedt wrote to CEPN in December, adding that she would not be able to meet with Borg. "From my perspective, the past six months have been a true psychological terror." Sundin says she's exhausted as well. She never expected whistleblowing to become an all-consuming, 9-month job. "We just want this nightmare to end," adds Jutfelt, "so that we can go back to doing proper science again."