A recent study provides evidence that some results of behavioural experiments with insects cannot be fully reproduced. So far, possible reproducibility problems have been little discussed in this context.
If an experiment is repeated under similar conditions, the results should be the same. In reality, the situation is often different — scientists speak of a “reproducibility crisis,” which affects different disciplines. A recent study by an eleven-member research team from the Universities of Münster, Bielefeld and Jena (Germany) has provided evidence that some results of behavioural experiments with insects cannot be fully reproduced. Nevertheless, at least half of the findings in the various experiments could successfully be reproduced. Depending on the different definitions and methods used to determine reproducibility, the non-reproducible results ranged from 17 to 42 per cent.
Reproducibility is studied intensively in biomedical research and in behavioural research on mammals. However, there are no comparable systematic studies on insects. The team led by behavioural biologist Prof Helene Richter from the University of Münster has now used a multi-laboratory approach to test the reproducibility of ecological insect studies. They conducted three different behavioural experiments. For each experiment, the researchers used a different insect species (turnip sawfly, meadow grasshopper and red flour beetle). They carried out all three studies in laboratories in Münster, Bielefeld and Jena and compared the results. The experiments examined the effects of starvation on behaviour in larvae of the turnip sawfly, the relationship between body colour and preferred substrate colour in grasshoppers and the choice of habitat in red flour beetles.
To the research team’s knowledge, the study is the first to systematically demonstrate that behavioural studies on insects can also be affected by poor reproducibility. This was particularly surprising as insect studies generally used comparatively large sample sizes and could therefore provide more robust results. However, reproducibility was higher compared to other systematic replication studies that were not carried out on insects. This suggested that reproducibility problems are less severe in insect studies than in other areas of science.
The results are of particular interest to scientists in behavioural biology and ecology, but also for all other disciplines in which behavioural experiments are carried out with animals. The deliberate introduction of systematic variations could improve reproducibility in studies with living organisms, the research team concludes.
