Variation in lucerne seed yields in relation to genotype and intensity of pollination

Abstract

Variation in the seed yields of twelve lucerne lines was studied at two South Australian sites between 1967 and 1970. At Tintinara, the lines were sown in drills and not irrigated. At Adelaide, plants were spaced on a 90 X 90 cm grid and irrigated in early summer. Each line, and also individual genotypes at Adelaide, yielded consistently in all harvests and at both sites. The Australian cultivar Hunter River, African, and Spanish line, gave the highest yields. Nine plants with the most consistent yields, and representing a range from low to high yields, were selected from the Hunter River and Spanish lines in the Adelaide experiment, were clonally multiplied: and exposed to different intensities of pollination in a third experiment. The pollination intensities were total exclusion of bees, exposure to casual visits from ambient bees, and exposure to a low bee density and a high bee density within pollination cages. Seed yields were negligible when bees were excluded. They were larger when exposed to casual bee visits, and larger still but similar at low and high bee densities in cages. There were large yield differences between genotypes, but there was no significant interaction between genotypes and the three intensities of pollination involving bees. There were large genotypic differences in the pathways to seed production, through raceme and floret development, with large differences in potential seed production, but these were not correlated with actual seed production. Even in the most efficient genotypes, at the highest pollination intensity, only one quarter of the potential seed yield was realised.