The interdependence of ammonia volatilization and denitrification as nitrogen loss processes in flooded rice fields in the Philippines

Summary The relative importance of ammonia volatilization and denitrification as loss processes following the application of urea to flooded rice by the traditional method was assessed at four sites with different characteristics in the Philippines. The effect of reducing ammonia loss on denitrification and total N loss was also studied. The total N loss was determined by a ¹⁵N-balance method and ammonia volatilization was assessed by a bulk aerodynamic method following the application of urea to small plots (4.8×5.2 m). As run-off was prevented and leaching losses were negligible, the denitrification loss was assessed as the difference between total N loss and ammonia loss. When urea was broadcast into the floodwater at transplanting, the ammonia loss varied from 10% to 56% of the applied N. Loss was smallest at Aguilar where wind speeds were low and the greatest at Mabitac where floodwater pH values and temperatures were high and the winds were strong. The ammonia loss was reduced at all sites by incorporating the urea into the soil by harrowing. However, the reduction achieved varied markedly between sites, with the largest reduction (from 56% to 7% loss of the applied N) being observed at Mabitac. The total N lost from the basal application into the floodwater ranged from 59% to 71% of the applied N. Incorporating the urea by harrowing reduced the total N loss at two sites, increased the total N loss at the third site, and had no effect at the fourth site. The denitrification losses ranged widely (from 3% to 50% of the applied N) when urea was broadcast into the floodwater at the four sites. The denitrification loss was low when the ammonia loss was high (Mabitac) and high when the ammonia loss was low (Aguilar). Reducing ammonia losses by incorporating the urea into the flooded soil resulted in increased denitrification losses at three of the sites and appeared to have no effect on denitrification at the fourth site. The results show that reducing the ammonia loss by incorporating urea into the soil does not necessarily result in reduced total N loss, and suggest that the efficiency of fertilizer N will be improved only when both N-loss processes are controlled simultaneously.