More than 98% of entomology papers contain so little species information on the insects being studied that they are essentially impossible to replicate, according to a survey of more than 550 articles published in 2016.
Laurence Packer, an entomologist at York University in Toronto, Canada, and his colleagues examined every paper published in 2016 in nine major entomology journals published by the United Kingdom’s Royal Entomological Society, the Entomological Society of America and the Entomological Society of Canada. Less than 2% of the papers included three key pieces of information: a description of how an insect was identified; evidence that biological samples had been documented and placed in a repository; and a reference to a recent definition of that species. Two-thirds of the papers did not contain any of these things. The work was published last month in Insect Conservation and Diversity1.
“The way taxonomic data is treated in most papers is so sloppy, it’s equivalent to saying, ‘Statistics were done by a statistician and the results were significant,’ without saying how the results were obtained,” says Packer. “That’s clearly unacceptable.”
Without reliable taxonomic information, it is impossible to be sure which exact species a researcher studied. That makes it difficult for other scientists to replicate their work. Packer says that biological pest control — using certain insect species to limit populations of undesirable organisms — has probably suffered as a result. Researchers studying the topic often get conflicting results from tests, which he suspects is at least partly due to incorrectly identifying the species used as biocontrol agents. “Without good species identification, you can’t even check what caused the different results,” he says.
Fixing the problem
Packer and his colleagues recommend five ways that scientists can improve the situation: include an insect’s order and family in the article title, abstract or keywords; clearly state how they identified a species; name and provide contact information for the person who made the identifications; cite the literature on which the identifications are based, including the most recent definition of a species; and store examples of all species in a named repository.
Packer says that other entomologists have been generally supportive of his team’s suggestions, although some have baulked at the amount of work it would require to implement them. The recommendation to store every study subject, in particular, raises logistical issues. “Providing a place to house all those samples could get expensive, and we would have to be careful about endangered species,” says Jim Hardie, director of science at the Royal Entomological Society in St Albans, UK, although using DNA as a barcode to identify species could help to keep the costs down.
Sydney Cameron, an entomologist at the University of Illinois at Urbana-Champaign and co-editor-in-chief of the journal Insect Systematics and Diversity, says that Packer’s study starts an important conversation, which has been largely ignored across biology. “That conversation needs to involve museums, as well as journals and scientists, so that everyone is on the same level,” she says.
At a minimum, she says, journals should require authors to explain clearly how they identified the species that they studied. “In our journal, we have clear statements about what is required,” she says. “Maybe we should go back and look at those and see if we have said enough.”
Sign up for the daily Nature Briefing email newsletter
Stay up to date with what matters in science and why, handpicked from Nature and other publications worldwide.