Early generation selection for milling quality in five bread wheat crosses

A milling quality assessment protocol using a 10 g seed sample was developed. Its ability to discriminate between wheats differing in milling quality and its application to early generation selection were evaluated. The procedure reliably and accurately discriminated between the varieties Sunco, Hartog and Sunelg for the flour parameters yellow pigment content, Kent-Jones colour grade and yield. The procedure was applied to progeny of F2 single plant selections from five bread wheat crosses, and heritabilities and simulated response to selection measured in the F3 generation. Medium to high heritabilities were obtained for yellow pigment content and colour grade, while those for flour yield were generally lower and varied considerably between crosses. Results of simulated selection versus values for control varieties indicated that the protocol would result in positive response to selection for milling quality. The outcome for some crosses, however, was influenced by the comparatively high values for the control varieties, whereas the use of a lesser standard gave improved response to selection. The best simulated selection results were obtained from the Suneca/Sunfield cross in which values for both parents exceeded those of the standard (Hartog). However, there was no transgressive segregation for improved milling quality in any cross. Phenotypic correlations between milling parameters and grain yield were generally non-significant indicating that the sequence of trait selection (i.e. grain yield or milling quality) would have minimal influence on the other attribute. The data indicate that response to selection for milling quality would result from application of this protocol to the early generations of a wheat breeding program.