Abstract

ABSTRACT To predict the mineralization of nitrogen(N) in organic amendments in paddy fields, we applied data obtained from incubation experiments under waterlogged conditions by a hierarchical Bayesian model using acid-detergent-soluble organic N (ADSON) as a model input. Each of 20 organic amendments (6 cattle manure composts, 3 swine manure composts, 3 poultry manure composts, 2 oil cakes, 2 rice brans, 1 fish meal, 1 sludge fertilizer, and 2 mixed commercial organic fertilizers) was mixed with paddy soil, and the N mineralization curves obtained after 1, 2, 4, 8, and 12 weeks of waterlogging were incorporated into the model to obtain values of the model parameters k i , α1, and Q 10. All model parameters converged with a mean absolute error (MAE) of 0.044 g N kg−1, a root mean square error (RMSE) of 0.060 g N kg−1, and a model efficiency (EF) of 0.96. These values were comparable to or slightly better than those of a previous model for upland soil conditions. Most of the observed values were within the 95% Bayesian prediction interval, indicating that the model performed well with the observed data. Although it is generally believed that livestock manure composts decompose more slowly than organic fertilizers such as oil cake, the livestock manure composts had a large value of the rate constant k i , presumably because of the initial rapid decomposition of the small amounts of organic N present in them. In paddy fields, the soil conditions are aerobic between the application of organic amendments and water entry. In such a case, we propose the use of the upland field model to predict the mineralization of organic N during the period of aerobic soil conditions and of this study’s model during the period of waterlogging, subtracting the amount of already mineralized N from the parameters N in and ADSON. The model devised here is expected to be useful for fertilization design in organic agriculture and in the reduction of the use of inorganic fertilizer.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call