Ex-judge to investigate controversial marine research

An Australian university has launched an investigation into the research record of a discredited scientist it educated, as findings by academics who supervised her doctoral training are challenged.

James Cook University said it has appointed an external panel to look for evidence of misconduct in the research conducted by marine biologist Oona Lönnstedt between 2010 and 2014, when she was undertaking PhD studies at the Queensland institution.

The university said the panel’s as yet unidentified members include “eminent academics with expertise in field work, marine science and ethics” and a former federal court judge.

The university announced the investigation a year ago after Dr Lönnstedt had been found guilty of fabricating data underpinning a study in her native Sweden, following her departure from JCU.

That study, published in 2016 in the journal Science, was retracted in 2017. Formal concerns have also been raised over data missing from three of the 15 papers Dr Lönnstedt co-authored while at JCU.

A spokesman for the university said the panel would report by the end of February “or such other date as agreed with the university”, and that JCU intended to release the findings “subject to legal and privacy considerations and any other obligations”.

Meanwhile, seven researchers who exposed Dr Lönnstedt’s misconduct have now cast doubt on a theory of ocean acidification that was championed by her PhD supervisors and featured in her doctoral thesis. The researchers say that their paper, published in Nature on 8 January, “comprehensively and transparently” refutes research findings that rising carbon dioxide levels in the oceans will make small coral reef fish easy pickings for predators.

Their three-year study harnessed more than 900 fish from six species, in attempts to replicate studies including JCU research published over the past decade. The team could not reproduce experimental findings that the carbon dioxide concentrations predicted by the end of the century would jeopardise small species’ sustainability by making them hyperactive, predictable and oblivious to predators.

Lead author of the three-year study, Timothy Clark, said it was the first time anybody had attempted to replicate the “profound” effects of carbon dioxide on fish behaviour outlined in dozens of studies conducted by “a small group of researchers”. He said the team had taken great care to match the conditions of the previous studies while improving the methodology “to maximise transparency and minimise the potential for experimenter biases”.

“We have exhausted all reasonable methodological avenues to explain the disparity in our findings compared with previous studies,” said Dr Clark, an aquatic physiologist at Victoria’s Deakin University. “The global scientific community deserves to understand how it is possible to achieve such remarkably different findings when addressing the same question.”

Five of the studies the team scrutinised were led by JCU reef researcher Philip Munday. Some were co-authored by marine ecologist Mark McCormick, who recently left JCU, and the University of Saskatchewan’s Douglas Chivers. The trio were Dr Lönnstedt’s PhD supervisors and have collaborated on at least six papers about the impacts of elevated carbon dioxide on fish behaviour.

Dr Lönnstedt contributed to three of these papers, including as lead author of an article published as part of her thesis. Professor Munday stressed that her work had not been included in the Nature replication study, and said he had no reason to suspect misconduct in any of the research examined.

He said he was not surprised that Dr Clark’s team had been unable to replicate his findings because it had used different methodologies. “You can hardly say you’ve repeated something if you’ve gone and done it in a different way,” he said.

Professor Munday cited differences in the species, developmental stages and environmental backgrounds of the fish used in his experiments. “Since then we’ve learned a lot about the environmental factors that might mitigate some of these effects on behaviour,” he said.

He said more than 70 studies had shown that elevated carbon dioxide levels could affect the behaviour of temperate as well as tropical fish. “There’s a lot of evidence, not just from my group. There’s been work on salmon, sharks [and] eels from different locations, by different groups, measured in different ways.”

john.ross@timeshighereducation.com