Drones, DNA, and weather: A phase-oriented hybrid engine to predict sugar beet disease
A fungus that can wipe out up to 50% of a sugar beet crop may soon meet its match in a new generation of smart disease forecasting. A new study published in Phytopathology shows how combining drone imagery, weather data, and qPCR-based airborne spore monitoring can reveal where disease is present and what the pathogen is likely to do next—giving growers a critical edge in timing control measures.
Led by Facundo R. Ispizua Yamati of the Institute of Sugar Beet Research (IfZ) in Göttingen, Germany, the research focuses on Cercospora leaf spot, caused by Cercospora beticola. In field trials from 2020 to 2022, the team structured the epidemic into four biological phases—incubation, fructification, dissemination, and yield impact—to track the disease’s hidden life cycle.
“What excites me most is the seamless interlocking of high-tech sensors and the fundamental biology of how diseases develop,” said Ispizua Yamati. “We have demonstrated that pixels are not processes; by grounding machine learning in biological phases, our models move beyond seeing ‘spots’ on a leaf to actually interpreting the pathogenesis of the fungus.”
The study integrates mechanistic disease models, meteorological data, uncrewed aerial vehicle imagery, and molecular diagnostics into a single predictive framework. By combining these data streams into phase-specific hybrid models, the researchers reduced prediction error by up to 39%. Disease severity was best predicted using climate variables and drone-derived crop indices, while spore production and dispersal were linked to humidity, temperature thresholds, and wind variability.
“We built a smart system that works like a hybrid engine,” Ispizua Yamati said. “By combining these technologies, we can forecast risk much more accurately, moving us toward ‘precision medicine’ for crops.”
The findings also clarify how environmental conditions shape epidemics. “One of the key things we discovered is that the disease does not behave the same way all the time,” Ispizua Yamati said. Spore spread was favored by light, variable winds under conducive microclimates. Yield and sugar content declined with earlier disease onset and higher final severity, with losses reaching up to 0.0123 kg of root fresh weight per plant per severity point.
For plant pathologists and growers, the work highlights a shift toward data-driven, biology-informed disease management. By aligning fungicide applications with the actual life stages of the pathogen, this approach could reduce costs and limit unnecessary environmental impact. Read “Hybrid Modeling of Cercospora Leaf Spot Epidemiology: Integrating Mechanistic and Machine Learning Approaches Using Remote-Sensing and Environmental Data” to learn more—available open access in Phytopathology.
For over 100 years Phytopathology®, published by The American Phytopathological Society, has been the premier international journal for publication of articles on fundamental research that advances understanding of the nature of plant diseases, the agents that cause them, their spread, the losses they cause, and measures used to control them.
Journal
Phytopathology
Article Title
Hybrid Modeling of Cercospora Leaf Spot Epidemiology: Integrating Mechanistic and Machine Learning Approaches Using Remote-Sensing and Environmental Data
No comments:
Post a Comment