Use of an Optical Sorter to Select for Kernel Color in Winter Wheat (Triticum aestivum L.) Breeding.

Breeding tools are needed to differentiate within segregating wheat populations for white kernels.  We evaluated an optical sorter to select white kernels.  The sorter was used on six F₂ populations, each originating from a cross of different hard white and red winter wheat parents.  Prior to initial planting, samples were categorized as red-sorted (rs), and white-sorted (ws) and separately maintained throughout the study.  Each seed category was planted to plots replicated at two and three South Dakota locations in 2010 and 2011, respectively.  After each sort and harvest, random samples of 300 kernels from each category were stained with a KOH solution to help visually determine the percent white kernels.  In year-1, prior to planting, the proportion of white F₃ kernels in categories were; population 1 = 26% rs, 42% ws; population 2 = 3% rs, 14% ws; population 3 = 12% rs, 32% ws; population 4 = 17% rs, 32% ws; population 5 = 6% rs, 17% ws; and population 6 = 8% rs, 27% ws.  For population 1 after year-1 harvest, the proportions of white F₄ kernels were 41% rs and 50% ws at one location and 37% rs and 44% ws at another.  After sorting within categories again in year-2, population 1 samples had proportions of white kernels of 33% rs and 74% ws.  After harvest in year-2, F₅ samples of population 1 had proportions of white kernels of 36% rs and 79% ws at one location, 47% rs and 66% ws at a second, and 46% rs and 66% ws at a third location. The sorter effectively separated white kernels within the original populations, and for population 1, it enhanced the separation of white kernels after an additional sort and harvest.  Differences in sort effectiveness are likely attributable to genotype population differences and environment.