Skip to main content
ARS Home » Research » Publications at this Location » Publication #144951


item Dowell, Floyd

Submitted to: Cereal Chemistry
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/8/2003
Publication Date: 1/1/2004
Citation: Pasikatan, M.C., Dowell, F.E. 2004. High-speed segregation of high-and low-protein single wheat seeds. Cereal Chemistry. 2004. 81(1): 145-150.

Interpretive Summary: High protein content is preferred in wheat products such as pasta while low protein content is desirable for cakes and cookies. Wheat with higher protein commands higher price (about 5 cents more per bushel) in export markets. A specific protein content range is one of the goals of wheat breeding programs but breeders don¿t have a non-destructive method to rapidly screen single wheat kernels for protein content. We evaluated the potential of a commercial high-volume color sorter for this purpose. The sorter was equipped with near-infrared optical filters and sensors for this application. Wheat blends that approximated low- and high-protein early generation wheat populations were obtained. Wheat blends were passed through the sorter five times where each pass removed 10% of the wheat mass. The sorter moved the protein content toward higher, or lower, protein depending on the desired outcome. The sorter has potential to help breeders in shifting early generation wheat populations toward the target protein level.

Technical Abstract: Wheat breeders need a nondestructive method to rapidly sort high- or low-protein single kernels from samples for their breeding programs. For this reason, a commercial color sorter equipped with near-infrared filters was evaluated for its potential to sort high- and low-protein single wheat kernels. Hard red winter and hard white wheat cultivars with protein content >12.5% (classed as high-protein, 12% moisture basis) or < 11.5% (classed as low-protein) were blended in proportions of 50:50 and 95:5 (or 5:95) mass. These wheat blends were sorted using five passes that removed 10% of the mass for each pass. The bulk protein content of accepted kernels (accepts) and rejected kernels (rejects) were measured for each pass. For 50:50 blends, the protein in the first-pass rejects increased or decreased as much as 1%. For the accepts, each pass increased or decreased the protein content of accepts by about 0.1%, depending on wheat blends and whether low-protein or high-protein wheat was rejected. At most, two re-sorts of accepts would be required to move 95:5 blends toward the direction of the dominant protein content. The 95:5 and 50:50 blends approximate the low- and high-protein mixture of early generation wheat populations, and thus the sorter has potential to aid breeders in purifying samples for developing high- or low-protein wheat. Results show that sorting was partly driven by color and vitreousness differences between high- and low-protein fractions. A new background specific for high- or low-protein and fabrication of better optical filters for protein might help improve the sorter performance.