Forestry supporters seize on error in study to falsely claim logging doesn’t raise bushfire risk

The study, published in May in the journal Fire, claimed logging had made forests in Tasmania more flammable, but was withdrawn last week at the request of its authors after a major error was discovered.

Industry groups and Tasmania’s resources minister, Guy Barnett, used the retraction to assert there was a scientific consensus showing logging did not make forests more flammable. Experts told Guardian Australia that this was wrong, and scientific evidence exists confirming logging could increase the risk of dangerous fires.

Three scientists at the University of Tasmania had examined satellite images of Tasmania’s Huon Valley before and after bushfires in early 2019.

The study concluded old-growth forests and plantations were less likely to burn, and that keeping old forests intact and allowing eucalypt forests to mature beyond a usual 40 to 90-year cycle could also help reduce fire.

But the researchers used publicly available geographic data from the Tasmanian government to differentiate plantations from other forests, and this information was incorrectly categorised.

“We all make mistakes. I find it devastating to be part of,” said Prof Jamie Kirkpatrick, who co-authored the retracted study and is the most senior scientist among the researchers.

“I have written 333 refereed papers and this is the only one I have had to retract.”

The paper’s withdrawal prompted a flurry of statements from pro-forestry politicians, the logging industry and the professional association representing foresters.

The Liberal senator for Tasmania, Jonathon Duniam, said the Greens and environment groups “used this error-ridden paper for their own political attack on Australia’s world-leading, sustainable forest industries”, and that they should now admit they were wrong.

The Institute of Foresters of Australia (IFA) said it was seeking a public apology from the University of Tasmania “over the standard of the university’s review process that has resulted in this publication”.

On Monday, the Australian Forest Products Association (AFPA) issued a joint media release with the Tasmanian Forest Products Association with the title “Fake bushfire research thrown out by independent journal”.

The groups claimed the journal had retracted the paper after being contacted by third parties. This is disputed by both Kirkpatrick and the journal’s editor. Kirkpatrick said the suggestion the research was fake was “slanderous” and the paper had not been “thrown out” but withdrawn at the request of scientists.

Prof Alistair Smith, Fire’s editor-in-chief based at the University of Idaho, told Guardian Australia: “The retraction was made at the request of the authors of the original paper after they were alerted to an error. This is an excellent example of research integrity by the authors.”

Smith said science was driven by data and scientists showed their integrity when they retracted studies after being faced with data that countered conclusions. “This is what sets apart scientists from special interest groups. As such, all I can do is applaud the integrity of the authors as being of the highest standard,” he said.

Ross Hampton, the chief executive of AFPA, said he would write to the University of Tasmania as well as to media outlets he said had produced “a slew of media articles” on the back of the research.

A University of Tasmania spokesperson said the researchers had withdrawn the paper after checking the errors. “This was the right thing to do. The university will review the matter.”

Forestry industry supporters used the error to dismiss research that found logging could contribute to increased fire risk.

Hampton claimed “the scientific consensus is that there is no causal link between timber harvesting in Australia and overall increased bushfire intensity”.

Tasmania’s resources minister, Guy Barnett, also said “contemporary scientific consensus indicates that native forest harvesting does not exacerbate bushfires”, while the Tasmanian Labor primary industries spokesman, Shane Broad, said the paper had been used to deliver “unwarranted attacks” on the forestry profession.

Scientists said political and industry claims did not reflect the weight of peer-reviewed studies.

“I think the balance of evidence still suggests that logging makes forests more flammable,” Kirkpatrick said.

Prof David Lindenmayer, a forest ecologist at Australian National University who has researched the effects of logging on bushfire, also said the “scientific consensus is actually the other way to what the industry is saying”.

“Of course the industry wants to deny it,” he added.

In February, several climate and forest scientists signed an open letter calling for an end to native forest logging, in part because it could increase the risk of fire. The briefing document to support that letter did not include the retracted paper among its sources.

In a comment article published in the journal Nature Ecology and Evolution, in May, several senior forest scientists, including Lindenmayer, wrote there was now “compelling evidence” that logging native forests made bushfires more severe.

Lindenmayer told Guardian Australia: “The industry is deluded if it says there’s no relationship [between logging and the severity of bushfires]. Of course the industry would make that call and create doubt in people’s minds, but the empirical work is very strong.”

He said forests that had regenerated from previous logging tended to be more dense. Logging and thinning operations also left flammable debris on the forest floor.

A series of studies, including some of his own, showed “unequivocally” a link between forest flammability and logging, he added.

Prof Ross Bradstock, a bushfire risk expert at the University of Wollongong, said the evidence for the impacts of logging on fires was mixed and dependent on a wide array of factors, including the severity of the fires, the landscape, type of forest and the time since the areas were logged.

But he said he disagreed there was any scientific consensus showing no effect.