UAV multispectral imagery for assessing tree health in tart cherry orchards with a focus on powdery mildew detection

Unmanned aerial vehicle (UAV)-based multispectral imagery could provide a more efficient approach to pest scouting in tart cherry, a relatively low-input orchard crop. We evaluated whether vegetation indices derived from UAV-mounted multispectral imagery could detect powdery mildew (PM) in three commercial tart cherry orchard blocks in Utah during 2022 and 2023. Powdery mildew severity was surveyed repeatedly and compared with vegetation indices calculated from multispectral imagery collected during the same period. The optimized chlorophyll vegetation index (OCVI) showed the strongest overall relationship with PM severity, but correlations were only moderate when data were pooled across orchards and dates. Within individual orchards and sampling dates, vegetation indices were not consistently related to PM severity. Relationships were especially weak when analyses were restricted to observations below an estimated economic threshold of 20% severity, where early detection would be most useful for management decisions. Although UAV multispectral imagery was not reliable for early PM detection, it was useful for identifying broader spatial variability in tree health. Tree health, assessed using red-edge normalized difference vegetation index as an indicator of canopy chlorosis and vigor, varied substantially among orchards and was associated with canopy density and apparent soil electrical conductivity (ECa). In the most variable orchard, poorer tree health and greater chlorosis were concentrated in areas of lower canopy density and higher ECa, corresponding to finer-textured soils. These findings indicate that UAV multispectral imagery was not reliable for early detection of PM in tart cherry under commercial orchard conditions but was useful for characterizing orchard-scale variability in tree health related to canopy structure and soil properties.