Effects of time of urea application on combine-sown Calrose rice in south-east Australia. II. Mineral nitrogen transformations in the soil-water system

Abstract

A combine-sown crop of Calrose rice was grown on an alkaline self-mulching, grey clay soil. Prilled urea was broadcast onto the dry soil surface or into the floodwater at one of four different times: sowing, before permanent flood, after permanent flood or panicle initiation. Within 24 days of fertilisation at sowing, levels of extractable N in the control and fertilised treatments were not significantly different. The rapid disappearance of fertiliser N was associated with rapid loss of NO-3 from the top 20 cm, and this loss appeared to be due to denitrification rather than leaching. The notable difference between fertilisation before and after permanent flood was the distribution of the fertiliser N in the soil and water. Fertiliser applied to the dry soil before flooding was carried deeper into the soil, while fertiliser applied a few hours later into the floodwater was located in the water and top 2 cm of soil. Therefore the potential for loss by ammonia volatilisation and nitrification/denitrification was probably higher in the latter treatment. Fertiliser N topdressed at panicle initiation was also located in the floodwater and top 2 cm of soil. However, fertilisation at this later growth stage was not associated with a yield decline when compared with fertilisation before permanent flood (Part I). Presumably this reflected the greater ability of the plants to compete with the processes leading to fertiliser N loss at this later growth stage, together with a lower ammonia volatilisation potential due to the influence of greater canopy cover on various factors influencing ammonia volatilisation. The data suggested that the efficiency of fertilisation at sowing and before permanent flood might be improved further by deep placement of the fertiliser or by use of a nitrification inhibitor.