Genetics of wheat starch B-granule content

Abstract

Two lines of hexaploid wheat were crossed and the basic generations of parent, F1, F2 and back-cross were sown in a controlled-environment chamber. FreshF1 and back-cross grains were generated, so the material could be handled either as the standard set of basic generations on a whole-plant basis, or as an extended set on an embryo or endosperm basis. The experiment was repeated. Mature grains were harvested and the starch particle size distribution was analysed in 3284 grains from 111 plants. Means and variances were partitioned into additive, dominance and interaction components. Grains from cross-pollinations had B-granule contents between parental values, rather than of the maternal parent, indicating an involvement of the grain genotype. Quantitative models based on endosperm genotype gave a better fit to the data than those based on embryo genotype. The difference in starch B-granule content between the parents was largely due to additive genes. Dominant genes were also indicated, with the first dose in the triploid endosperm having a large effect while the second dose had little or none. Non-allelic interactions were significant in the second experiment where the use of more types of backcross made them more detectable. There were also small and significant residual effects of the maternal plant in the first experiment, attributed to the vigour of the F1 mother plant and to the cytoplasm of Sunco. Narrow-sense heritability was low, between 0.05and 0.18 depending on the generation. Transgressive segregation was not found, suggesting that all alleles tending to increase the B-granule content were found in the Sunco parent and none in ME71. There was also no detectable heterosis in this character. The results show that breeding and selection for a low B-granule content should be possible but a further reduction will require new and complementary genes.