NITROGEN-FERTILIZED GRASS IN A SUBTROPICAL DAIRY SYSTEM .3. EFFECT OFSTOCKING RATE ON THE RESPONSE TO NITROGEN-FERTILIZER

Two stocking rates, one as practised on farms (2 cows/ha) and the othe r 50% higher, were assessed for effect on pasture and milk yield respo nse to applied nitrogen (N) fertiliser (0-600 kg N/ha.year) for Holste in-Friesian cows grazing Rhodes grass (Chloris gayana) cv. Callide pas tures. Pastures were grazed in combination with grazing oats for winte r, with overall farm stocking rates of 1.17 and 1.37 cows/ha for the 2 treatments. Cows were maintained on these areas for 3 years. Cracked grain was given at 0.8 t/cow.year, and hay or silage supplements were given when green grass yield was <0.5 t dry matter (DM)/ha. The increm ental response (P<0.05) in milk yield to each kg increase in level of applied N was 4.93 kg/ha at 1.17 cows/ha and 1.64 kg/ha at 1.37 cows/h a. The amount of conserved forage fed at the high stocking rate increa sed (530 and 970 kg/ha.year at 1.17 and 1.37 cows/ha), and financial m argins over costs were reduced at the high stocking rate. The low milk response at the high stocking rate was associated with a low response in pasture growth. At <2 t pasture DM/ha on offer, incremental respon se to applied N declined, and there may have been an excessive loss of N through volatilisation in heavily grazed pastures. Milk yield per c ow was closely related to total pasture yield on offer (P<0.01), and t o leaf and stem yields (P<0.05). Relationships were stronger in summer and autumn than in spring. Over the full year, milk yield increased b y 1.24 kg/kg leaf DM or 0.24 kg/kg total pasture DM on offer. At the h igher stocking rate, surface soil (0-10 cm) concentrations of phosphor us and nitrate were higher than at the lower stocking rate. We conclud e that in areas of moderate rainfall (<1000 mm/year) in the subtropics , high stocking rates resulting in low pasture yields and exposed grou nd surface will be associated with low efficiency of use of applied N.