Novel solution to plant pathogen detection

Some nematodes are beneficial for plants, but others damage roots and spread plant viruses. Pathogenic nematodes can result in significant crop losses of 14 to 35% of total yield.

The impact of nematodes largely depends on the specific species present within crop soils and whether the correct treatment for managing the pest is used. As such, it is important for farmers to be able to identify the nematode species in their soils to inform their management decisions. However, the correct classification of nematodes requires either expertise or specialist equipment, which can be costly.

A collaboration between researchers at the University of Sheffield, led by Po Yang, a Professor of Pervasive Intelligence in the School of Computer Science, Dr Matthew Back at Harper Adams University, and Dr Nasamu Musa from industrial partner ADAS, has led to the development of an alternative system for nematode detection.

NemaDetect AI can classify five types of plant-parasitic nematodes with 96.4% average precision. An online tool offers a quick and affordable alternative to the detection methods currently available and has been adopted by a variety of agricultural stakeholders.

The service enables farmers to upload images and have the nematode classification returned, allowing them to accurately target damaging nematode pests with appropriate crop treatments and reduces pesticide overuse. The technology has the potential to reduce costs, improve yields, support food security and lower the risk of pesticide resistance development.