Automated Single‐Kernel Sorting to Select for Quality Traits in Wheat Breeding Lines

Abstract

ABSTRACTAn automated single kernel near‐infrared system was used to select kernels to enhance the end‐use quality of hard red wheat breeder samples. Twenty breeding populations and advanced lines were sorted for hardness index, protein content, and kernel color. To determine whether the phenotypic sorting was based upon genetic or environmental differences, the progeny of the unsorted control and sorted samples were planted at two locations two years later to determine whether differences in the sorted samples were transmitted to the progeny (e.g., based on genetic differences). The average hardness index of the harvested wheat samples for segregating populations improved significantly by seven hardness units. For the advanced lines, hardness index was not affected by sorting, indicating little genetic variation within these lines. When sorting by protein content, a significant increase from 12.1 to 12.6% was observed at one location. Purity of the red samples was improved from ≈78% (unsorted control) to ≈92% (sorted samples), while the purity of the white samples improved from 22% (control) to ≈62% (sorted samples). Similar positive results were found for sorting red and blue kernel samples. Sorting for kernel hardness, color, and protein content is effective and based upon genetic variation.