Genetic analyses and selection for grain yield of common bean under sole crop and maize/bean intercrop.

摘要

The objective of this research was to assess breeding strategies that will result in genotypes with improved bean (Phaseolus vulgaris L.) yield for maize (Zea mays L.)/bean intercropping. Segregating populations and advanced lines of determinate and indeterminate bush beans were grown in sole crop and maize/bean intercrop at Elora and Woodstock. Bean grain yield and other agronomic traits were measured.;Interaction of genotype and cropping system was consistently found for grain yield and pods per plant. Few differences were noted between cropping systems for genetic variances and heritabilities of yield and yield-components. The influence of genotype x environment interaction, on the other hand, was substantially greater in intercrop than in sole crop.;Pods per plant was the trait affected most by intercropping. However, comparisons of realized heritability for yield and harvest index, based on selection in F;Early generation testing was found to be a reliable method for selecting among bean populations in both cropping systems based on grain yield and yield-related traits. In the presence of a genotype x cropping system interaction in a combined analysis over three environments, 14.3% of 63 genotypes were found to be superior in both sole crop and intercrop, when a selection pressure of 25% was applied in each cropping system. This underscores the need for some yield evaluation of breeding materials in intercrop. When the superior lines identified in an initial evaluation trial, conducted at Elora in 1994, were studied the following year for selection response in two locations, it was found, with respect to sole crop, that direct selection for yield was twice as effective as indirect selection in intercrop in each of the locations. In intercrop it was found that in one location response from direct selection was also twice as large as that from indirect selection in the alternative system, but in the other location there was a reversal of this trend. This dependence of response on location was attributed to genotype x environment interaction. Thus, yield testing in intercrop can be more effective if multiple sites are used.