Convergent-divergent selection for seed production and forage traits in orchardgrass .3. Correlated responses for forage traits

Most orchardgrass (Dactylis glomerata L.) cultivars are developed in a two-stage selection protocol. Stage 1 involves selection for agronomic traits among spaced plants and Stage 2 involves selecting among a limited number of clones based on their polycross progeny performance, usually in sward plots. The objective of this study was to quantify genetic progress in sward-plot forage yield and related traits of four orchardgrass populations subjected to two cycles of recurrent phenotypic selection for agronomic traits of spaced plants, Selection was conducted at four locations, using either a regional approach (convergent-divergent) or selection for local adaptation at one location. Selection criteria varied among locations, depending on the local environment and each breeder's goals. Convergent-divergent selection led to a 7 and 8% increase in forage yield of the MO2 and WO11 populations, respectively, averaged over all evaluation locations. Local selection did not lead to forage yield improvement for any population. Both selection methods led to decreased rust reaction in at least two of the four populations and later maturity in all four populations. The improvements in forage yield appeared to be due partly to populational buffering among progeny of plants selected at different Locations, a phenomenon not possible within the local populations because of single-location selection. These results indicate that selection for agronomic traits of spaced plants in orchardgrass can lead to improved sward-plot forage yield across a wide range of environments when multiple-location selection is applied to some germplasms.