‘Intelligent traps’ and flower strips could help tame oilseed rape pest

Originally published on Farming Future Food

Digital pest traps using artificial intelligence proved useful tools in guiding an integrated pest management programme against pollen beetle, as part of a major field experiment in Germany.

Pollen beetles feed on the buds and flowers of oil seed rape, reducing the fitness of the plants and potentially reducing yield for spring crops. There is widespread resistance to broad-spectrum pyrethroid insecticides, which have been used as a control measure.

Researchers from Leibniz Centre for Agricultural Landscape Research examined the effects of different management approaches on populations of the pest, as well as the usefulness of a new digital yellow water trap which is linked to an AI model to detect, classify, and count pests.

Tweaking management techniques

The work formed part of the patchCROP project, in which a commercial farm field has been arranged like a checkerboard to compare different land management approaches.

In this case, the team used three pest techniques: a conventional plant protection program representing ‘business as usual’, a reduced management strategy where crop protection products were used at upper limit pest thresholds, and an additional reduced programme which also featured flower strips surrounding the patch to enhance natural pest control. They also compared the patches to large reference fields using conventional and reduced approaches.

Across the three years of the study, the researchers found there was no significant decrease in crop yield when comparing the conventional and two reduced programmes in 2021 and 2022, while for 2023, there was a significantly lower yield in the reduced patch, but not the patch using flower strips.

“The timely insecticide applications by weekly expert decisions likely helped keep yield losses to a minimum. Intensive insect pest monitoring [on reduced programme sites] prevented insecticide spraying,” the scientists wrote in Journal of Sustainable and Organic Agriculture.

The study did not prove the advantage of patch cropping systems, however, as yields were lower in the patches than in the reference large fields.

AI role could prove pivotal

Meanwhile, the team observed no significant difference in the number of pollen beetles caught between digital and traditional yellow water traps. However, the digital trap’s link to AI seemed to provide additional advantages. High sensitivity and precision, particularly as numbers of insects arriving rose, could accurately detect when the pest had reached a specific level of severity.

“The combination of digital yellow water traps and AI image classification packaged into a smartphone application is an efficient and human-centred technology. These innovative technologies can support farmers in the rigorous monitoring of pollen beetles, while simultaneously reducing the labour-intensive nature of conventional insect pest monitoring methods,” they wrote.

In smaller fields, flower strips proved their worth as a natural method of crop protection, they noted, and the use of the digital water traps using additional data sources and more sophisticated AI could contribute to the design of sustainable pest management strategies using such landscape features.

The results of the study could help inform the European Union’s Farm to Fork Strategy, which proposes reductions in the use of conventional chemical insecticides and a shift towards IPM approaches.