Prospects and limitations for agricultural engineering in the development of sustainable weed control methods – examples from European research

Summary

This paper gives a brief review of the major achievements in European research on physical weed control methods for agricultural and horticultural crops. Most of the work has emerged from an increasing awareness and concern about pesticide consumption in many Western European countries. Also an increasing interest in organic farming has further pushed the development of more sustainable weed control methods. Generally, the research has been joint projects between engineers and agronomists and the results have so far revealed some prospects as well as limitations for non-chemical methods to become useful solutions, not only for the organic growers but also for the conventional ones. A number of investigations have focussed on rather simple low-tech mechanical weeding principles, such as harrowing, brush weeding, hoeing, torsion weeding, and finger weeding, aiming at describing the weeding mechanisms for a better optimisation of the usage. In some crops, such as transplanted vegetables, potatoes, maize, winter oil seed rape, and partly small grain cereals, mechanical weed control has been quite effective and may become a relevant alternative to chemical weed control. However, current mechanical methods generally work with low selectivity, as they do not distinguish between weed and crop plants when applied into the crop row. Attempts to change the constructions and materials of the weeding tools have not decisively improved the selectivity and more intelligent methods capable of controlling only the weeds are therefore needed. The first step in that direction was the introduction of electronic steering systems for automatic guidance of inter-row hoes. They are based on image analysis of the crop row, and the technology is considered to be a kind of platform for the development of more advanced systems for robotic weeding in the rows of row crops, such as sugar beets, maize, and most vegetables. However, recent studies have shown that such an ambition may be difficult to fulfil because weed plants growing right beside the crop plants are the most harmful ones in terms of suppressing the crop plants. Whether any computer-based system would be able to guide a cutting device with sufficient accuracy and speed to remove those weeds in a practical situation in row crops seems questionable. Hence, other projects have been started with a view to avoid this challenge, trying to look for less complicated methods with more short-termed prospects of being applicable in practice. Steaming the soil prior to crop establishment and in bands corresponding to crop rows appears to have some potential in that context since an almost complete intra-row weed control can be achieved without affecting the crop. However, as with most other thermal methods, high energy consumption is a key-issue that needs to be solved.