Modelling crop traits and fluxes under multiple abiotic stressors

Agriculture is the largest consumer of freshwater, accounting for approximately 70% of the total global usage. As the human population continues to grow, demand for water will be exacerbated by a changing climate and shifting temperature and precipitation regimes. Dynamically modelling crop physiological function will be crucial to optimising crop management strategies. In this study we synergise hyperspectral and thermal remotely-sensed data to model plant traits and water fluxes in spring wheat (Triticum aestivum) in growth chambers within a controlled environment experiment under water and/or nitrogen stress conditions. Results showed that plants which had first received nitrogen fertiliser and were subsequently droughted presented the lowest water fluxes, and the lowest leaf chlorophyll content and photosynthetic capacity (Vcmax) values. Partial least squares regression (PLSR) analysis of hyperspectral reflectance data revealed key wavelengths sensitive to six different plant traits and fluxes (including relative water content, leaf nitrogen, stomatal conductance), with strong correlations between measured and modelled values (R2 = 0.84; p<0.001, 0.60; p<0.001, and 0.65; p<0.001, respectively). By incorporating optical reflectance data into a modified surface energy-balance model to incorporate the changing optical properties of the leaves under stress, we increased the accuracy of modelled water fluxes against leaf porometry measurements during abiotic stress (R2 = 0.46; p<0.01 and R2 = 0.61; p<0.001, for the original and improved transpiration model respectively). This work points to the importance of considering the influence of stressors on crop fluxes and traits both in isolation and combined. The novel integration of optical and thermal remote sensing techniques paves the way for the improved dynamic modelling of crop physiological function.