Using models of canopy height and vegetation indices (VIs), drone-based remote sensing (RS) can allow for large-scale assessments of plant growth and nutrient status. The goal of this thesis was to assess drone-based RS in addressing research and production challenges in processing carrots and fresh-market tomatoes. In processing carrots, a two-year trial was conducted to investigate the effects of topdress N rate, number of topdresses, and timing of split applications on processing carrot production, as well as the potential for VIs to guide N topdress decisions. Yield and shoot biomass were found to increase with higher N rates. Splitting applications did not affect yield but increased shoot biomass and N uptake in a wet year. Both early and late split applications showed potential to increase N loss. VI-based sufficiency indices explained at most 66% and 29% of the variation in carrot root yield in 2019 and 2020, respectively, but explained greater variation on average than petiole sap nitrate (6%) and recommended N applications 26% less often. In fresh-market tomatoes, RS was integrated into a cover crop by N fertilizer rate experiment to compare RS measurements to manual measurements of plant height, leaf tissue N, and leaf chlorophyll meter (SPAD) readings. Crop surface model plant height estimates were good estimators of measured heights (R2 = 0.89-0.96), with comparable correlations to final yields and ability to resolve significant treatment differences. Foliar N identified more significant differences between treatments than SPAD or VIs. In both experiments, drone-based RS demonstrated the potential to detect relevant in-season plant treatment responses comparably to manual measurements, with possible advantages in scalability, cost, and resolution.
Read
