Seed production trials in Canterbury have shown (1) strong negative relationships between plant density and seed yields per plant, and (2) plant density ranges above which high plant numbers failed to compensate for low yields per plant and below which high yields per plant failed to compensate for low plant numbers. The net result was that seed yields per unit area tended to plateau at this density range, with no further gains being made by increasing plant numbers. It is concluded that the function of seeding rate is to provide sufficient viable seeds to attain this density. Optimum density ranges for 'Grasslands Tama' Westerwolds ryegrass and 'Grasslands Nui' perennial ryegrass were 300 to 400 plants/m, and for 'Grasslands Matua' prairie grass 100 to I30 plants/m. The value of superphosphate and potassic fertilizers for seed production was doubtful, but nitrogen increased seed yields regardless of time of application. Work in Canterbury concentrated on the time of N application and showed that nitrogen for seed production was most efficient when applied just after stem elongation had commenced. Grazing grass-seed crops reduced seed yields because of an indirect negative effect on seedhead size and individual seed weights, both of which proved to be more important than seedhead numbers in determining seed yields. Keywords: Seed production, grasses, plant density, nitrogen, stem elongation, grazing, seedhead size, seed weight.
RYEGRASS SEED CROPS
