NITROGEN CONTRIBUTED BY GRAIN LEGUMES TO RICE GROWN IN ROTATION ON THE CUNUNURRA SOILS OF THE ORD IRRIGATION AREA, WESTERN-AUSTRALIA

The uptake of nitrogen (N) by dry season rice following wet season crops of soybean (for grain or green manure), green gram, Sesbania cannabina (a native legume), a cereal (sorghum or dryland rice for grain), or bare fallow, was studied for 3 cropping cycles over 4 years. The work was done on Cununurra clay (0.04% N) at Kimberley Research Station near Kununurra, W.A., in the Ord Irrigation Area. Stubbles were returned to the soil except in the first cycle when (excluding the green manure treatment) all tops were removed from the plots of maturity. There was a 12-month bare fallow period between the first and second cycles. Dry season rice was drill-sown with or without 100 kg ha-1 of N applied as urea at permanant flooding. Soybean, green gram and Sesbania crops accumulated 290-360, 80-130 and 110-180 kg N ha-1, respectively, in the tops at maturity. An average of about 40 kg N ha-1 was present in the stem bases and roots (0-20 cm depth). Estimates of nitrogen fixation based on 15N dilution measurements ranged from 65-72% of total plant N when the legumes were grown after 12 months fallow, to 93-95% when they were grown immediately following dry season rice. Fertiliser N at 25 kg ha-1 applied presowing (''starter'' N) had no significant effect on legume N yield at maturity. N returned in leaves, stems and hulls averaged 30, 50 and 80 kg N ha-1 for green gram, soybean and Sesbania, respectively. Rice grain yields and N uptake at maturity were generally highest after Sesbania and lowest after a wet season cereal crop. Differences among treatments were small and related to the quantity of N returned in residues. On average, 11% of the N in the residues was recovered in the tops of the following rice crop. Rice yields increased over the 4-year period, but mean increases were similar for legume and non-legume treatments. The average apparent recovery of N applied as urea to dry season rice at permanent flooding was 76%. The inclusion of a soybean cash crop in the rotation offers the possibility of a marginal reduction in the need for N fertiliser.