ABSTRACT

The effectiveness of assessing unreplicated F3 lines for yield in a wheat (Triticum aestivum L.) cross was determined by the regression and correlation between F3 lines and related F4 bulk means, and between F3 lines and related F5 family means. The correlation and regression coefficients obtained between F3 line and F4 bulk mean yield expressed as a percentage of adjacent control were 0.59** and 0.39**, respectively. The correlation and regression coefficients obtained between F3 line and F5 family mean yield expressed as a percentage of control were 0.56* and 0.39*, respectively. By growing a large sample of the progeny of each F3 genotype adjacent to a control, it was possible to discriminate among F3 lines for heritable quantitative differences.

Control plots of Triticum aestivum var. Manitou were grown adjacent to every plot of breeding material in three hard red spring wheat nurseries at the University of Manitoba. Simple correlations between yields were high (r =.88,.87 and.63) and significant (P.01) for control plots at 2.7-m (9-ft) centers but decreased rapidly to nonsignificance with increasing distance between plot centers. The data indicate that for the particular type of plot used, the yield of a control plot provides a good measure of the soil fertility in terms of the yielding ability of an adjacent plot.

Evaluation of the effect of selection in 3 different F 3 populations of hard red spring wheat was achieved for yield and breadmaking quality parameters by examining the subsequent performance of F 3 selections as populations of F 5 lines. For baking absorption the mean performance of the F5 populations was positively related to the performance of the F 3 selections in each of the 3 population-years . A significant relationship between the performance in F 3 and in F 5 was not obtained in any of the 3 years for bushel weight, flour ash, remix loaf volume, farinograph mixing tolerance index or for flour color grade . A high positive relationship between performance in F3 and F 5 was found in two out of three years for protein content, flour protein content and 1,000-kernel weight. Predictive ability for sedimentation value, flour yield, farinograph development time and blend loaf volume was inconsistent for different population-years .A positive relationship between the yield of F 3 plots and the mean yield of F 5 populations was found in only 1 year . In the other 2 years the relationship between F3 and F5 yields ranged from nonsignificant to low, negative and significant . The latter results were possibly related to the relatively narrow high yielding range of the total F 3 variability for yield which was sampled in each of those years . In all 3 population-years the highest yielding F 5 populations were derived from F 3 lines which were high yielding on a plot basis and also very high yielding relative to the yield of their adjacent control in the F 3 nursery .Broad sense heritabilities calculated for the F 5 generation support the thesis that selection between F 3-derived populations would be a worthwhile plant breeding procedure for many of the breadmaking quality parameters .

A new grain, Triticale, (a synthesized hexaploid species combining the genomes of Triticum durum and Secale cereale) was tested as the principal ingredient in a chick starter ration. On a pound-for-pound basis Triticale was approximately equal to hard spring wheat in nutritive value for chicks as judged by growth, efficiency of feed utilization and ration metabohzable energy. Supplementation of a 17.5 per cent protein, Triticale ration with glycine, DL-methionine and/or L-lysine showed lysine to be the most limiting amino acid for chicks. A similar study with a 21 per cent protein ration showed that both L-lysine and DL-methionine were limiting for chick growth. The results bring to the fore the potential importance of new, synthesized plant species to animal production and, simultaneously, present an evaluation of a grain from one of these new species as a feed ingredient for chickens.