In-season variable rate nitrogen recommendation for wheat precision production supported by fixed-wing UAV imagery.

Remote sensing data from optical canopy sensors has been successfully used for precision nitrogen (N) management. This study aimed to explore the potential of multispectral camera mounted on fixed-wing unmanned aerial vehicle (UAV) in guiding in-season N topdressing for regional wheat production. Wheat plot and field experiments were conducted with multiple cultivars and N rates during 2017–2019. Manual sampling and UAV photography were carried out at wheat key growth stages. A modified sufficiency index algorithm (MSIA), driven by UAV spectral data, was developed and used for variable rate N recommendation (VRNR). Firstly, data from plot experiment was used to determine the local optimum N rate ( N OPT ) for attaining the maximum available yield, and to quantify the relationships between vegetation index and above ground biomass, plant N uptake (PNU), and grain yield. The results suggested that local N OPT was 283.5 kg ha⁻¹, and red-edge soil adjusted vegetation index (RESAVI) yielded the highest accuracy in constructing PNU monitoring model (R² = 0.78) and dynamic curves of UAV spectrum (R² > 0.95). Secondly, using UAV spectral data of field experiment, the calculated real-time PNU and sufficiency index, along with N OPT , were imported into MSIA for in-season VRNR. Compared to local empirical N fertilization, MSIA gave accurate N recommendation rate according to wheat growth status, with harvest index and nitrogen agronomic efficiency averagely increased by 3.1% and 11.9%, respectively under 15.4% reduced N input and undiminished yield return. This study supplies a new approach to implement VRNR for regional wheat production supported by UAV imagery. [ABSTRACT FROM AUTHOR]