A widely publicized study, which claimed to demonstrate that plastic microbeads used in some cosmetic products can cause behavioral impairments to developing fish, has been retracted by the journal Science amid accusations of scientific misconduct.
The retraction follows a 21 April 2017 report by an expert group at Sweden’s Central Ethical Review Board (CEPN) which concluded that the authors of the study, Oona Lönnstedt and Peter Eklöv of Uppsala University, were “guilty of scientific dishonesty”.
The paper, first published on 3 June 2016, purported to investigate the effects on perch larvae reared in water with varying concentrations of plastic (polystyrene) microbeads. In general, their results appeared to show that environmentally significant concentrations of these particles could affect the development of perch behaviorally in such a way as to make them more likely to be killed by predators. A March 2016 Science news feature described Lönnstedt and Eklöv’s study:
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.
The retraction is the final chapter in a long and contentious process that began when whistleblowers working at the research station where Lönnstedt and Eklöv claim to have conducted their study cast doubt on whether the study was ever actually performed as described:
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. [Josefin] 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.
[Fredrik] 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.
Jutfelt and Sundin, along with five other scientists, sent formal request in writing to Uppsala University, asking for an investigation into the paper. In it, they said:
We have identified a number of potentially critical flaws regarding the execution and reporting of this study, which include: (1) missing data (wrongly stated in the paper as being available in the supplementary materials and in the Uppsala institutional repository); (2) inconsistencies in the sample sizes reported and the microplastic exposure concentrations used; (3) issues with the statistical design and analyses; and, most worryingly, (4) large disparities in the way the experiment has been reported by the authors compared with the reports of eye witnesses. These issues, many of which should have been identified at the peer-review stage, bear directly on the validity and reproducibility of the results presented in the paper.
The lack of raw data, attributed by the authors to a computer theft, forced Science to issue a 9 December 2016 “editorial expression of concern” which stated:
The authors have notified Science of the theft of the computer on which the raw data for the paper were stored. These data were not backed up on any other device nor deposited in an appropriate repository. Science is publishing this Editorial Expression of Concern to alert our readers to the fact that no further data can be made available, beyond those already presented in the paper and its supplement, to enable readers to understand, assess, reproduce, or extend the conclusions of the paper.
Uppsala University did conduct a preliminary investigation of the study using outside experts at this time but — controversially — their report argued that the issues raised, especially regarding data, had been adequately resolved and did not merit further sanctions. Simultaneously, however, the case was taken up by the Swedish Expert Group for Misconduct in Research, which appointed another researcher to look into the study. After an in-depth look at the complaints and Lönnstedt’s and Eklöv’s responses, the expert group concluded:
In view of the lack of ethical approval, the essential absence of original data for the experiments reported in the article, and the widespread lack of clarity concerning how the experiments were conducted, it is the opinion of the Expert Group that the article in Science should be recalled.
The full Uppsala University investigation is still pending, but prior the study’s 3 May 2017 retraction, the authors themselves endorsed calls for its removal as well, while not admitting fault, as reported in an Uppsala University press release:
In their letter to the journal, the researchers explain their request to retract the article by stating that science has to rest on solid ground and that the results of their study, even if correct, will not be trusted as long as a suspicion of misconduct remains. For this reason, the researchers request to retract the published research article.
In their retraction notice, published on 3 May 2017, Science wrote:
Although the authors have told Science that they disagree with elements of the Board’s report, and although Uppsala University has not yet concluded its own investigation, the weight of evidence is that the paper should now be retracted. In light of the Board’s recommendation and a 28 April 2017 request from the authors to retract the paper, Science is retracting the paper in full.