ARS | Publication request: Improving Grain Breeding Programs Through NIR-Based Single-Kernel Sorting

Submitted to: ASABE Annual International Meeting
Publication Type: Abstract Only
Publication Acceptance Date: April 13, 2007
Publication Date: N/A

Technical Abstract: We developed automated visible and near-infrared (NIR) spetroscopy procedures and instrumentation to select kernels with specific hardness, protein, and color traits to enhance the development of scab resistant hard and soft wheat varieties. The system also shows potential to sort for other characteristics such as scab damage, vomitoxin levels, ergosterol levels, vitraeousness, sprout damage as measured by alpha amylase content or falling number, moisture content, selenium content, Karnal bunt-infected kernels, and waxy character. Our single kernel near-infrared system can sort single kernels based on specified properties at a rate of about one kernel/2s (500-1000g/day). We also have high-speed sorting technology that can sort visible defects at rates as much as 80,000 kernels/s (300 bu/hr). This technology is now used routinely for such applications as purifying red or white breeding lines, removing Karnal but-infected kernels during routine inspection for the APHIS national surveys, and selecting waxy seeds from segregating populations. While most of our work has been with wheat, we have also shown applications for proso millet, barley, rise, and sorghum. This poster will report the development of this NIR-based sorting system and our sorting accuracies.