Abstract There is strong demand for new Fusarium head blight (FHB)‐resistant small grain crop varieties that accumulate less of the mycotoxin deoxynivalenol (DON). It is difficult to breed for FHB resistance because it is a quantitative trait, and phenotyping is tedious and subjective. Implementation of high‐throughput, objective phenotyping methods would advance breeding success in developing FHB resistant varieties. Optical sorting uses a camera to rapidly and efficiently separate healthy grains from Fusarium ‐infected grains. The sorter does not rely on potentially biased human observations but rather utilizes a calibrated camera to autonomously and reproducibly classify and separate the grains based on their appearance. We evaluated the ability of the optical sorter to reduce DON in closely related wheat ( Triticum aestivum L.) breeding lines with various combinations of scab resistance genes. The sorter did an excellent job of reducing DON content (10%–22%). However, this did not translate into consistently reduced DON with each cycle of selection. Combinations of resistance genes in a small subset of genotyped lines did not affect the ability of the sorter to reduce DON. Previous research has shown that Fhb1 frequency increases with cycles of selection, which was confirmed in this study. Furthermore, the frequencies of resistance quantitative trait loci (QTL) 1A and 4A Neuse also increased with just one cycle of selection. However, the frequency of 3B Massey decreased. Results suggest the optical sorter can be an effective tool for selecting grain with reduced DON content and more FHB resistance QTL for breeding purposes.
Morris et al. (Fri,) studied this question.