Researchers at the USDA-ARS and Utah State University explored the potential of utilizing a UAV drone to predict Kernza seed yield during pre-harvest.

Abstract

Accurate estimation of seed yield in perennial grains like Thinopyrum intermedium (intermediate wheatgrass [IWG]) remains a challenge due to high interannual and spatial variability of this important trait. We evaluated the potential of before-harvest unmanned aerial vehicle (UAV)-derived multispectral vegetation indices to predict IWG seed yield under experimental conditions in northern Utah over three consecutive years (2022–2024). A global Random Forest model integrating all years was compared against year-specific models using standard metrics (R², root mean square error [RMSE], mean absolute error, and Lin’s concordance correlation coefficient). The global model outperformed individual-year models (R² = 0.813; RMSE = 24.3 kg ha). Chlorophyll-sensitive indices such as green normalized difference vegetation index and chlorophyll index green were ranked as the most influential predictors, while classic indices including normalized difference vegetation index and the ratio vegetation index showed limited predictive power. These findings highlight the effectiveness of pre-harvest UAV imagery for yield prediction in perennial grasses and support the use of multi-year model training frameworks.

Plain Language Summary

Perennial grain crops such as intermediate wheatgrass can provide food and protect soil, but their seed harvests change a lot from year to year, making it hard for farmers and breeders to predict yields. We tested whether images taken by a small drone just before harvest could be used to estimate seed yield in experimental fields in northern Utah over three growing seasons. We built computer models that linked the color and brightness of the crop canopy in the images to measure seed yields. The models that combined data from all 3 years were more accurate than models built for each year alone, and features related to leaf greenness were the best predictors of yield. Our results show that simple drone flights before harvest can give useful early estimates of seed yield in this new grain crop. This approach could help breeders select better lines and support management decisions in farmers’ fields.