NITROGEN-FERTILIZED GRASS IN A SUBTROPICAL DAIRY SYSTEM .1. EFFECT OFLEVEL OF NITROGEN-FERTILIZER ON PASTURE YIELD AND SOIL CHEMICAL CHARACTERISTICS

The effects of 5 levels of nitrogen (N) fertiliser on pasture yield an d composition and soil chemical characteristics of Rhodes grass (Chlor is gayana) cv. Callide pasture under grazing and cutting in a subtropi cal environment of 800 mm annual rainfall were studied over 6 years. P asture received annual basal dressings of superphosphate (250 kg/ha) a nd potassium chloride (125 kg/ha for plots, 125 kg/ha.alternate year f or grazed paddocks). Urea was applied in 3 equal dressings in Septembe r, December, and February, at rates equal to 0, 150, 300, 450, and 600 kg N/ha.year. Stocking rate was 2 Holstein-Friesian milking cows/ha t hroughout the year, and separate areas of grazing oats (0.4 ha/cow), c racked grain (0.8 t/cow.year), and hay or silage were used to suppleme nt pasture. Under both cutting and grazing, pasture dry matter (DM) yi eld increased linearly with applied N to about 300 kg N/ha.year, with little further increase at higher levels. Under grazing there was evid ence of a decrease in yield at 600 kg N/ha.year, due to total death of the pasture following frosting in winter and the need for regeneratio n from seedlings in spring; this regeneration was slowed by the large amount of surface litter. Grass N content increased and phosphorus and potassium contents decreased with increasing levels of applied N. All 3 nutrients increased from year 1 to 6. Leaf content of grazed pastur e was highest during spring (>30% DM) and declined through to autumn ( <20% DM), with no consistent effect of N level. Leaf content was consi stently higher under cutting (>50%) but was not altered by level of ap plied N. Soil nitrate-N levels increased (P<0.05) with level of applie d N, from 4 to 42 mg/L at 0 and 600 kg N/ha.year, respectively. After 6 years of fertilisation at 300 and 450 N/ha.year, nitrate-N levels we re similar to those for 600 kg N/ha.year. Soil pH decreased (P<0.05) w ith applied N, by 0.15 and 0.28 units annually for 150 and 600 kg N/ha .year, respectively. We conclude that in this environment large respon ses in pasture growth occur under both cutting and grazing to levels o f applied N to about 300 kg N/ha.year, with little response beyond thi s level.