PHOSPHORUS-FERTILIZER COMPARED UNDER GRAZING AND MOWING - RESPONSE

The response to phosphatic fertilizer of a pasture based on perennial ryegrass, subterranean clover and phalaris was assessed over 4 years. The pasture was established on a previously unfertilized area. Single superphosphate was applied at five rates. In addition there was an unf ertilized treatment. The highest rate of fertilizer, expressed as elem ental phosphorus (P), was reduced from 100 kg/ha in years 1 and 2 to 6 0 and 40 kg/ha in years 3 and 4 respectively. Each year total pasture drymatter (DM) production was measured during a 6-month growing season from early winter to late spring under four systems of defoliation: m own monthly (M1), mown every 2 months (M2), mown every 3 months (M3) o r set stocked with 1 year old sheep (G). Grazed plots were stocked at 10, 14 or 18 sheep/ha in 3 of the years, and at 8.75, 12.25 or 15.75 s heep/ha during the remaining year. The production of mown pasture gene rally decreased with increasing frequency of cutting, and was always l ess than the production of grazed pasture. Mean values for M1, M2, M3 and G were 2.85, 4.35; 5.44 and 6.86 t DM/ha respectively. The absolut e and marginal responses to fertilizer (kg DM/kgP) were always greater for the grazed treatments. This suggests that data from mowing trials seriously underestimate the fertilizer response of grazed systems. Th e amount of fertilizer required to reach a given proportion of potenti al yield response did not differ between the systems in the first 3 ye ars, but in year 4, more fertilizer was required by the grazed system to reach a given proportion of potential yield (P < 0.01. Strategies f or correcting the response curves of the mown treatments are considere d.