Ammonia volatilization from drill sown rice bays

Abstract

Ammonia volatilization losses and other N transformations were studied in drill sown rice bays fertilized with urea at various times between permanent flooding (PF) and panicle initiation (PI). Ammonia loss was measured directly with flow chambers and indirectly through application of Freney et al.'s (1985) model. Both techniques indicated that ammonia volatilization was negligible from fields fertilized immediately before PF. Applying 100 kg urea-N ha−1 to floodwater one day after flooding significantly increased floodwater ammoniacal-N and urea-N content, however the concentrations fell rapidly over the following five days. Fertilizer-N dissolved in the floodwater was in the urea rather than the ammoniacal-N form, indicating slow hydrolysis until it moved into the soil. Floodwater on plots receiving urea one day after PF frequently had more than double the NO3-N concentration of plots fertilized before flooding. Applying up to 140 kg urea-N ha−1 at PI increased floodwater ammoniacal-N concentrations from almost zero to over 27 g m−3, but three days after fertilization there was less than 3 g m−3 present. Fertilization also increased NH4-N concentration in the top 40 mm of soil. Higher ammoniacal-N concentration at PI suggests higher urease activity. Floodwater pH at PI was low, with a mean daily maximum of 7.8 and this reduced ammonia loss to less than 1% of the applied N. The results indicate that volatilization from fields fertilized prior to PF is minimal because of the low floodwater pH and ammoniacal-N concentration, while low floodwater pH restricts volatilization from fields topdressed at PI.