Are insect populations truly declining, or are we relying on flawed data? This question lies at the heart of a recent controversy in the scientific community, where a prominent study on insect population trends has been flagged for concerns over its underlying database. In a move that has sparked debate, Science magazine has issued a permanent expression of concern for a 2020 meta-analysis that examined population patterns of freshwater and terrestrial insects. But here's where it gets controversial: the study, which has been cited over 820 times, was based on the InsectChange database, which critics argue contains “experimentally manipulated” data.
The issue first surfaced when researchers Laurence Gaume and Marion Desquilbet, along with eight colleagues, published a critique in Science in December 2020. They pointed out significant flaws in the meta-analysis, including the lack of weighting for studies based on geographic location, human impact, and insect diversity. And this is the part most people miss: the critics identified over 500 methodological and statistical errors in the InsectChange database, raising serious questions about its reliability. Despite these concerns, the study’s lead author, Roel van Klink, defended the work, arguing that disputes centered on inclusion criteria rather than fundamental flaws.
The controversy deepened when Science issued only a minimal erratum in response to the initial critique, prompting Gaume and Desquilbet to publish a more extensive critique in Peer Community Journal in 2024. They argued that the database’s biases render it unsuitable for estimating insect population changes, a claim supported by other critics who highlighted discrepancies in water quality assessments. But here’s the kicker: while Science has acknowledged the concerns, it has stopped short of retracting the paper, instead opting for a permanent expression of concern that urges readers to use updated versions of the data.
This decision has left Gaume and Desquilbet frustrated, with Gaume stating that Science is effectively ‘kicking the can down the road’ by not addressing the biased results that continue to influence public opinion. The debate raises broader questions about data integrity, peer review, and the responsibility of scientific journals to correct the record. What do you think? Is Science’s response adequate, or should the paper be retracted? Let us know in the comments below. For those interested in supporting independent oversight in scientific publishing, consider contributing to Retraction Watch or following their work on social media platforms like X, Facebook, LinkedIn, or Bluesky.
