PREDICTION OF NITROGEN-FERTILIZER REQUIREMENTS OF MAIZE IN SUBTROPICAL QUEENSLAND

Sixteen field experiments were conducted at 9 sites in the South Burne tt region of subtropical Queensland, to determine grain yield response of maize to fertiliser nitrogen (N) and to assess soil mineral N leve ls at sowing for predicting N requirement. At 6 sites, areas were eith er winter-cropped or bare-fallowed, resulting in different cropping hi stories immediately prior to summer maize. In each experiment, 4 rates of N (0, 38, 76, and 152 kg/ha) were applied, with an additional rate (304 kg/ha) at 3 sites that received supplementary irrigation. Immedi ately prior to sowing, soil samples for mineral N and moisture were ta ken from each 10-cm increment to a depth of 120 cm. Soil nitrate-N lev els (0-120 cm) before sowing were 16-100 kg N/ha (winter-cropped) and 65-167 kg N/ha (bare-fallowed). Application of N significantly (P<0.05 ) increased grain yield in 14 of the 16 experiments. Maximum grain yie lds in non-irrigated experiments ranged from 2.08 to 5.61 t/ha and ref lected profile-available water at sowing and rainfall during the growi ng season. Maximum yields in irrigated experiments ranged from 4.44 to 6.95 t/ha. The magnitude of the response was greater at winter-croppe d sites (relative yields 33-89%) than at fallow sites (82-100%). Relat ive grain yield was well correlated with nitrate-N in the 0-6 cm profi le (R2 = 0.74). There was also a good relationship between relative gr ain yield and nitrate-N at 0-10 cm depth (R2 = 0.64).