Breeding Methodology in Wheat. I. Determination of Characters Measured on F 2 Spaced Plants for Yield Selection in Spring Wheat 1

Three crosses of spring wheat ( Triticum aestivum L.) involving the cultivar ‘Glenlea’ as a common female parent were used to evaluate the use of physiological and/or morphological parameters alone or in combination on F 2 plants as selection criteria to identify high‐yielding F 4 bulks. The F 2 generation was grown as spaced plants in the field under two types of environmental regimes: (1) a regime and (2) a normal regime. Stress‐free plants were given regular irrigation, optimum fertilization and complete weed control to minimize stress while the normal‐regime plants were given no special treatment. Two hundred plants of each cross were measured for physiological, morphological, yield‐component, and phenological traits. A random sample of 50 seeds from each of these plants was used to generate F 4 bulks. For each cross, groups of 45 F 4 bulks derived from stress‐free F 2 's and 75 F 4 bulks derived from normal F 2 's were tested for yield in the F 4 in a six‐replicate, partially balanced lattice design. The results indicated that single F 2 parameters which described source capacity, sink capacity or plant morphology all identified high‐yield potential in the F 2 ; however, at 15% selection intensity only 17 out of 53 high‐yielding F 4 bulks were retained. A multiple regression analysis was found to be better than a single parameter approach, retaining nearly half (24 out of 53) of the high‐yielding lines. A combined‐cross analysis indicated that yield, productivity, kernel number, peduncle length, and productivity per tiller were all significantly correlated with F 4 bulk yields. A multiple regression analysis of the combined crosses indicated that productivity and peduncle length were the only significant common parameters.