Development of novel technological approaches for a reliable crop characterization under changing environmental conditions

The expansion of world population requires the development of new strategies and tools for agriculture. Extensive breeding and agronomic efforts over the past 50 years have been responsible for tripling cereal yields, while advances in grain quality have been less evident. Continuing advances in the techniques available to breeders offer the potential to increase the rate of genetic improvement aiming to develop resilient crop and better (more resource use efficient) varieties. Plant breeders want to be able to phenotype large numbers of lines rapidly and accurately identify the best progeny. For this purpose, different methodological approaches have been proposed to evaluate these traits in the field: (1) proximal (remote) sensing and imaging, (2) laboratory analyses of samples, and (3) lab-based near-infrared reflectance spectroscopy analysis in the harvestable part of the crop. However, near-infrared reflectance spectroscopy-based field evaluation of yield and grain quality is currently a real option. Thus the development of new technological approaches, such as the use of hyperspectral imaging sensors or near-infrared reflectance spectroscopy under field conditions may be critical as a phenotypic approach for efficient breeding as well as in field management of crops. This article reports the description of the CropYQualT-CEC project funded by the H2020-MSCA-RISE program. This project pursues the main objective of generating a common solid knowledge basis within the context of precision agriculture and digital farming. Further, within the project context, the article also provides a case study in which prediction models for total grain protein content, based on the reflectance spectrum of wheat canopies, are presented. Measurements were performed at around anthesis, using a full range near-infrared reflectance spectroscopy field spectrometer. Several models explaining >60% of grain protein variance in field trials illustrate the predictive capacity and robustness of this methodology for inferring grain quality traits well in advance of harvest.