Submitted to: Crop Science
Publication Type: Peer reviewed journal
Publication Acceptance Date: 12/27/2007
Publication Date: 7/15/2008
Publication URL: hdl.handle.net/10113/18253
Citation: Souza, E.J., Griffey, C., Kweon, M., Guttieri, M.J. 2008. Sources of Variation for Long-Flow Experimental Milling. Crop Science. 48(4):1432-1440. Interpretive Summary: Large experimental mills are used by wheat quality councils for evaluation of cultivars prior to release. The industry trusts the evaluations because they accurately predict how a specific sample would perform if milled on a commercial flour mill. We evaluated the genetic reproducibility of these tests. The question is simple how well does a single sample of a variety represent the performance of that variety in all situations. We found that if a new cultivar is paired with a known check that large differences between the new variety and the check can be detected reliably with a single milling. However milling two or more samples is required to detect differences of less than ½ a percentage point in flour yield. Differences in milling curves and break flour yield were less reliable than straight grade flour yield.
Technical Abstract: Flour milling quality of new wheat cultivars routinely is measured on long-flow experimental flour mills with multiple break and reduction rolls that produce many mill streams. To determine the relative contributions of genotypes, production years, and error to the variation in long-flow milling traits, we evaluated grain characters, milling characteristics, and soft wheat flour quality in two sets of cultivars in Virginia state trials for multiple years using the Allis-Chalmers long-flow flour mill at the Soft Wheat Quality Laboratory. In one study, 11 cultivars were evaluated for 5 years. Variation due to error, which was based on the cultivar x year interactions, was half the size of genotypic variance for straight-grade and break flour yield. In the second study, 27 cultivars were evaluated for 2 years. Genotype was a more important source of variance for milling traits than error: Error variances were approximately one-half of the genotype variances for straight grade flour yield and break flour yield. Variation in cookie diameter due to genotype was significant in both studies, but much smaller than variation due to error. Except for break flour yield and flour protein in the 2-year study, year effects were important for all traits in both studies. Therefore, evaluation of long-flow milling characteristics of new cultivars should be compared to a check cultivar produced in a similar year and environment. One such paired evaluation may be sufficient to detect large differences in milling performance between an experimental and a check cultivar, but two or more evaluations may be needed to identify differences of less than 5 g kg-1 in straight grade flour yield. Other milling and flour characteristics tended to have smaller ratios of genotype to error variance, suggesting that levels of replication required to detect small differences in those traits may be even greater than for straight grade flour yield.