Mapping plant-scale variation in crop physiological traits and water fluxes

Nitrogen (N) is a vital plant element, affecting plant physiological processes, carbon and water fluxes and ultimately crop yields. However, N uptake by crops can vary over fine spatiotemporal scales, and optimising the application of N-fertiliser to maximise crop performance is challenging. To investigate the potential of spatially mapping the impact of N fertiliser application on crop physiological performance and yield, we leverage both optical and thermal data sampled from drone platforms and ground-level leaf measurements, across a range of different N, Sulphur (S) and sucrose treatments in winter wheat. Using leaf level hyperspectral reflectance data, leaf chlorophyll content was accurately modelled across fertiliser treatments via partial least squares regression (PLSR; R2= 0.93, P < 0.001). Leaf photosynthetic capacity (Vcmax) exhibited a strong linear relationship with leaf chlorophyll (R2 = 0.77; P < 0.001). Using drone-acquired MERIS terrestrial chlorophyll index (MTCI) values as a proxy for leaf chlorophyll (R2 = 0.76; P < 0.001), Vcmax was spatially mapped at the centimetre-scale. Thermal drone and ground measurements demonstrated that N application leads to cooler leaf temperatures, which led to a strong relationship with ground-measured leaf stomatal conductance (R2= 0.6; P < 0.01). Final grain yield was most accurately predicted by optical reflectance (MTCI, R2 = 0.94; P < 0.001). Precise retrieval of leaf-level crop performance indicators from drones establishes significant potential for optimising fertiliser application, to reduce environmental costs and improve yields.