Site specific weed management in maize based on a UAV-derived prescription map

The effective weed management plays a crucial role in agricul- tural production, as weeds represent a major biotic stressor leading to sig- nificant reductions in crop yields. Precision agriculture offers potential solu- tions for facilitating the transition to low-chemical herbicide management systems, aligning with recent European strategies. This study was conducted on a maize field at the "Angelo Menozzi" exper- imental farm (Landriano, Italy). Forty-two plots were intentionally left un- treated without applying pre-emergence herbicides in order to achieve weed infestation levels a suitable for the experiment. The objective was to gener- ate a prescription map for site-specific treatment in post-emergence. An unmanned aerial vehicle (UAV) DJI Mavic 3 quadcopter, was used to conduct aerial surveys at 40 meters altitude during the V3-V4 phenological stage of maize, acquiring RGB and multispectral (560 nm, 650 nm, 730 nm, and 860 nm) images of the field plots. DVI vegetation index was calculated from these images to derive a weed infestation map. Post-emergence treat- ment thresholds were established based on weed pressure in different plots, resulting in a prescription map for site-specific herbicide distribution. Preci- sion application was operated with a boom-sprayer (iXter B16biXtra, Kver- neland) equipped with RTK-GPS and individually controlled nozzles. Initial findings indicate that implementing the prescription map in site- specific management plots, even under high weed density conditions, led to a significant 70% reduction in herbicide usage compared to uniformly man- aged plots. However, the impact on final crop yield remains uncertain, re- quiring further experimental cycles to refine the site-specific approach