Project description

The common datura (Datura stramonium) is an important agricultural weed and also contains toxic secondary plant constituents, the tropane alkaloids scopolamine and atropine. These can lead to acute symptoms of poisoning in humans and animals. Maximum levels for tropane alkaloids in food have been in force in the European Union since September 2022. Maximum levels have long been set for datura seeds in animal feed.

The origin of contamination of harvested crops with tropane alkaloids is essentially due to the presence of datura in the fields at the time of harvest. Datura seeds and plant parts are also harvested and are mixed with the harvested crop before being processed. The usual cleaning of the harvested crop to remove foreign matter is not sufficient, as the plant sap released during the combine harvest already contaminates the harvested crop. The late-germinating, heat-loving datura develops particularly well in summer crops, partially escapes conventional weed control measures and often leads to late weed infestation.

Sustainable strategies are therefore essential in order to provide safe and high-quality feed and food. The method of using drones to detect datura was developed as part of the StopDatura project and tested under practical conditions in 2022 and 2023. The drones were equipped with a high-resolution RGB camera and flown at medium altitude (approx. 30 to 60 metres) over soya bean fields.

The focus of this project is on the soya bean crop. Two approaches are being pursued:

• Strategies for sustainable plant protection involving new technologies (drone technology) are being developed to minimise the infestation of agricultural crops with datura.
• Methods for analysing datura alkaloids in harvested crops are being evaluated and established in order to prevent contaminated harvested crops from entering the processing cycle

Benefits of the project

Aerial images taken from low altitudes by drones have considerable potential for weed detection. This is demonstrated by the results, which show that it is possible to detect datura in soya bean fields before harvest. If datura occurs, it can be precisely removed from the field before harvest (manually/mechanically). The method used to recognise datura is based on machine learning and can be continuously improved by using new images. The knowledge of areas infested with datura can also be used to adapt crop protection or crop rotation planning (e.g. no cultivation of an endangered crop) in the following year. Infestation maps from a region can also be used to prevent the spread of datura to clean areas. The method can also be used to create application maps for site-specific herbicide application to control datura and other weeds ("spot spraying").

Further information

Common datura - Information from AGES on pests - AGES