Abstract
Five varieties of wheat differing considerably in genetic origin and in milling and baking qualities were crossed in diallel to determine if certain specific combinations would be outstanding in yield, and, if high yields are adversely associated with quality. An F1 was crossed with 4 of the 5 varieties to assess its use in determining general combining ability. Components of yield were studied in parents and compared with hybrids to assess predictive value of the parental components. The 14 hybrids ranged in yield from 28% below to 26% above the better parent. Three of the hybrids were significantly higher in yield than the higher yielding parent and 1 hybrid yielded significantly less than the lower yielding parent of the cross. The F1 tester crosses substantiated the results obtained from averaging specific combinations and evidence was presented indicating that a suitable F1 could be of value in deter-mining general combining ability. The high quality variety ''Pembina'' proved to be an excellent combiner for yield and the other high quality variety in the program, ''Canthatch,'' proved to be a very poor combiner for yield. Although each of the components of yield were significantly correlated with yield, components of yield could not be used to predict parental performance. Contrary to previous reports, a significant positive correlation (r =. 684) was obtained between yield and protein. A significant positive correlation (r =. 760) also was obtained between yield and dough development time. The variety ''Pembina'' appeared to have good combining ability for all quality characteristics. Hybrids including ''Pembina'' as a parent were at least equal to ''Pembina'' in farinograph absorption and were relatively superior to other hybrids in the carrying capacity of flour as indicated by loaf volume measurements. Conversely Canthatch was a very poor combiner for all quality characteristics.