Effect of population structure on protein-yield improvements in spring wheat (Triticum aestivum L. em Thell)

In a study designed to develop a more efficient breeding method for concurrent protein-yield improvements in wheat (Triticum aestivum L. em. Thell), 7 base populations [2 F2's, 1 intermated F2 (IF2) and 4 partial backcross (PBC) populations] developed from biparental crosses involving 2 Canadian hard red spring (CHRS) and 2 Canadian utility (CU) wheat cultivars were evaluated in Winnipeg, Manitoba, Canada. The IF2 and PBC populations were generated for comparison with conventional F2 populations and to determine which of the 4 methods of population development would provide a more efficient means of producing potentially superior genetic recombinants. Parameters pertaining to means, variances, correlations, heritabilities and frequencies of desirable and undesirable progenies were used to evaluate the limitations to genetic gain that may be expected from selection for GY and GPC in F2, IF2, CHRS-PBC and CU-PBC populations. Analysis of protein and yield data from 105 S1 lines derived from each of the 7 populations showed the CU-PBC's to have the highest grain yield (GY) and the lowest grain protein concentration (GPC) means; and the CHRS-PBC's, the lowest GY and the highest GPC means. The F2 and IF2 populations were intermediate for both characteristics. Populations developed from the same biparental cross did not differ significantly with respect to the majority of genetic parameters. However, desirable progenies combining high GY with high GPC were more frequent in the CU-PBC, and least frequent in the CHRS-PBC populations. The observed superiority of the CU-PBC populations appeared to be related to the advantage the system has in preserving the genetic integrity of a proven cultivar, while adding desirable genetic factors from another cultivar, thus capitalizing on introgression and upgrading simultaneously.