Seed weight in Lolium perenne as affected by interactions among seeds within the inflorescence

Abstract

Little is known about interactions between seeds within the inflorescence of perennial ryegrass (Lolium perenne L.) and their effect on final seed weight. These relations were investigated in two glasshouse experiments using two genotypes, by reducing the number of seeds in an inflorescence at anthesis. In the first experiment, entire spikelets were removed to study the relations between seeds in different spikelets. The effects on seed dry weight were not proportional to the number of spikelets removed. In one genotype, removal of two-thirds of the spikelets increased seed dry weight of the remaining seeds by 15%. In the other genotype, such treatment did not increase seed dry weight. In the second experiment, investigating seed interactions within a spikelet, either two proximal seeds or two central seeds were maintained in a spikelet by removing the other ovules in combination with no or 75% shading. Shading by 75% reduced seed dry weight by about 10%. In the unshaded treatment, seed dry weight was not affected by ovule removal. Under shading, the central seeds in a spikelet were about 12% heavier if they grew alone, in contrast to the proximal seeds. The effects of a reduction in seed number on seed dry weight were not related to the final nitrogen concentration of seeds. These experiments show that assimilate partitioning and the relationship between seeds in the inflorescence of perennial ryegrass are already largely determined at anthesis; increasing seed yield by manipulations after anthesis is not feasible.