Predicting soil organic matter nitrogen mineralization

Abstract The objectives of this study were to develop a model for soil organic matter (SOM) N mineralization that required minimal inputs, to test the model using independent datasets, and to determine if the results could be used in adjusting commercial N fertilizer recommendations. A first‐order model was adjusted monthly for differences in soil temperature and moisture. A daily time step was used to allow N mineralization estimates from one date to another. Simulations began in the month where temperature exceeded 10 °C. Model predictions were compared with those in published studies in northern California and Illinois. In the first California study, model predictions were near observed values in four of five locations. In the second California study, model predictions were within the 95% confidence interval of those based on laboratory incubation corrected for field temperature and moisture regimes. In a third study, unfertilized corn ( Zea mays L.) yields were compared with seasonal N mineralization predictions plus N from the prior crop for 15 Illinois soils. The result was 32.0 kg corn kg −1 seasonal available N as compared with the typical value of 67.5 kg corn kg −1 commercial N fertilizer. Nitrogen mineralization predictions for Illinois soils were related to total N, suggesting that an equation could replace simulation modeling for a given soil–location combination. It is proposed that the N mineralization model, which has readily obtainable inputs (total N, bulk density, monthly weather), could be used in conjunction with field studies to refine commercial N fertilizer recommendations.