A preliminary study to predict net nitrogen mineralisation in a flooded rice soil using anaerobic incubation

Abstract

Complementary field and laboratory studies were conducted to determine whether laboratory measurements of net nitrogen (N) mineralisation under anaerobic conditions could be used to predict field rates in a flooded soil and N uptake by a rice crop. The laboratory experiment consisted of measurements of ammonium accumulation at 10, 20, 30, and 40¦C for 7, 14, and 28 days of anaerobic incubation. There was no accumulation of ammonium at 10¦C, but increasing ammonification rate at temperatures of 20¦C was observed, except for a slower rate at 40¦C after 14 days. Two models were tested on the data: a zero-order reaction in which rate of mineralisation was a linear function of temperature; a first-order reaction in which net N mineralisation rate was a proportion of a depleting pool of labile organic N. In the second model, the rate was also linearly related to temperature. Both models fitted the laboratory data well (R2 = 0.94 and 0.97, respectively), but the second model accounted better for mineralisation at 40¦C for the 28-day incubation. These models were then run, using daily mean temperatures over a rice-growing season, to predict net mineralisation in the field. The predictions were compared with measured net N mineralisation in a flooded soil and N uptake by the crop measured throughout the season in the field from which the incubated soil was sampled. Net N mineralisation and crop uptake increased throughout the season, reaching maximum values of 115 and 111 kg N/ha at maturity. The zero-order and first-order models both predicted net N mineralisation accurately until the middle of the season, after which the zero-order model overestimated net N mineralisation but the first-order model predicted the reduction in the rate of net N mineralisation with reasonable accuracy. The close agreement between the laboratory incubations and field measurements of net mineralisation and crop N uptake suggest that incubation tests may provide useful information for including in a model to assist rice growers' decisions about N fertiliser.