The effects of superphosphate (0, 42, 125, and 375 kg ha-1 year-1), grass species (tall fescue, perennial ryegrass, and phalaris) and grass seeding rate (0, 1.1, 4.5, and 17.9 kg ha-1) on the productivity and botanical composition of grass-white clover pastures were investigated from 1967 to 1970 on a previously unfertilized site at Armidale, New South Wales. In all years, total pasture yields were increased by each additional increment of superphosphate applied. During the establishment year, the addition of 375 kg ha-1 superphosphate increased clover dry matter yield by more than 3500 kg ha-1, whereas grass production was low (< 1500 kg ha-1) on all treatments. Thereafter, the increases in grass yields at each level of superphosphate were the equal of, or exceeded, the clover responses, which were mainly linear throughout the experiment. Neither grass seeding rate nor grass species had a major effect on total pasture productivity. Their effects on the relative yields of the grass and clover components were most evident at the highest rate of superphosphate. At SP0, grass and clover percentages declined from 1967 to 1970 and at SP125 the proportions of grass, clover and weeds were about equal in all years. At SP375, the strong overall trend from clover-dominance in 1967 to grass-dominance in 1970 was most rapid in treatments sown with 4.5 or 17.9 kg ha-1 grass and least rapid in phalaris-white clover pastures. In 1969 and 1970, when most pastures at SP375, had reached the grass-dominant stage of their development, within-year fluctuations in grass and clover content were least in the fescue-white clover pastures.
Yield and botanical composition of four dryland pastures at Ashley Dene Research Farm over 8 years
