Utilizing High‐Throughput Phenotypic Data for Improved Phenotypic Selection of Stress‐Adaptive Traits in Wheat

Abstract

Efficient phenotyping methods are key to increasing genetic gain and precisely mapping genetic variation. Recent phenotyping developments have resulted in high‐throughput phenotyping platforms that utilize proximal sensing to simultaneously measure multiple physiological traits. However, there has been limited exploration of this high‐resolution, multiple phenotypic data. To address this, two wheat (Triticum aestivum L.) biparental populations were grown for 3 yr under two different treatments, drought and heat stress, at the International Maize and Wheat Improvement Center, Ciudad Obregon, Mexico. The lines were evaluated at multiple time points throughout the growing season with “Phenocart,” a portable field phenotyping platform that integrates precision GPS, spectral reflectance, and thermal sensors. Both normalized difference vegetation index (NDVI) and canopy temperature (CT) were correlated to final grain yield. We found that broad‐sense heritability (H2) and correlation to yield for both NDVI and CT had a regular pattern over the growing season. The maximum correlation and H2 existed during mid‐grain‐fill stage, while correlations were low for early‐ and late‐season measurements. We also found that the H2 of CT on a given day was a good indication of how well that dataset correlated to yield. In addition, the temporal NDVI data from heading to senescence was modeled to evaluate stay‐green and senescence differences between lines. Based on the repeatable correlations, high‐throughput phenotyping platforms can be used to assist with indirect selection through rapid collection of physiological measurements compared with direct selection for grain yield alone.