High-Throughput Phenotyping using Digital and Hyperspectral Imaging Derived Biomarkers for Genotypic Nitrogen Response.

The development of crop varieties with higher nitrogen use efficiency (NUE) is crucial for sustainable crop production. Combining high-throughput genotyping and phenotyping will expedite finding novel alleles for breeding crop varieties with higher NUE. For phenotyping, digital and hyperspectral imaging techniques can efficiently evaluate growth, biophysical and biochemical performances of plant populations by quantifying canopy reflectance response. In this study, these techniques were used to derive automated phenotyping of indicator biomarkers; biomass and chlorophyll levels, corresponding to different nitrogen (N) levels. A detailed description and associated challenges and considerations required for digital and hyperspectral imaging are provided, with application to delineate N response in wheat. Computational approaches for spectrum calibration and rectification, plant area detection, and derivation of vegetation index analysis are presented. We developed a novel vegetation index with higher precision to estimate chlorophyll levels, underpinned by an image processing algorithm that effectively removed background spectra. Digital shoot biomass and growth parameters were derived enabling the efficient phenotyping of wheat plants at vegetative stage, obviating the need for phenotyping till maturity. Overall, results suggest value in the integration of high-throughput digital and spectral phenomics for rapid screening of large wheat populations for N response.