By Ricardo Javier, Second Year Masters, MSU Plant Sciences & Plant Pathology
My project is about using aerial phenotyping to assess the plant growth and health in winter wheat breeding lines. The main focus is to estimate grain yield, plant biomass and disease severity using unmanned aerial vehicle (UAV) imagery. The disease measured in this study is Fusarium head blight (FHB), a disease with some recent outbreaks in Montana.
The 2024 data collection for plant biomass and grain yield was performed in the Arthur H. Post Agronomy Farm in Four Corners, while the FHB severity data was collected in the Southern Agricultural Research Center in Huntley near Billings. The images were taken with a M200 drone with a Multispectral RedEdge-P sensor. These images were assembled in orthomosaics in the 6 different light bands, through which a value of each plot of the field was extracted. The different values were used to calculate vegetation indices (VI’s) to estimate the plant health and growth. These VI’s will be compared using machine learning (ML) models to the real values of grain yield, plant biomass and FHB severity. The idea is to measure how accurate the ML models are in predicting these traits using VI’s. The use of aerial phenotyping coupled with ML models can help the Winter Wheat Breeding Program to improve selection accuracy and rate of variety release, focused in productivity and disease resistant phenotypes.
**Note: Ricardo presented his project during the 2025 March Madness competition.