Measured and predicted mineralization of clover green manure at low temperatures at different depths in two soils

Predictive models of the temporal mineralization pattern of organic residues may help in development of strategies to synchronize N mineralization with the crop demand and minimize off-season losses. In the present investigation, two double first-order models with temperature as a driving variable were tested against data on decomposition and N mineralization, respectively, in two field experiments with green manure. On 15 November 1984, mesh bags with red clover (Trifolium pratense L.) shoot material were placed at five depths (0–30 cm) on a sandy-loam and a loam site in south-eastern Norway. 167 days after burial, 73% of the initial clover nitrogen remained on the surface, 62% at 5-cm depth, and 56% at 30-cm depth. The differences among buried samples largely persisted throughout the experimental period (1.5 years). The decomposition rate slowed down appreciably after day 270, when the amount of N in buried bags averaged 33% of the initial N. The effect of site was small and varied during the experiment. The decay model, which was derived from laboratory incubations, predicted the initial observations of remaining clover material fairly well. Later, predicted and measured values diverged because recalcitrant residues decomposed more extensively in the field than in the laboratory. The N mineralization model was tested against net N mineralization from white-clover (T. repens L.) green manure ploughed down in late October. The course of the net N mineralization was well described when disregarding an over-prediction (6–12% of applied clover N), which may be due to N losses not accounted for in the model. The predictions were sensitive to the kind of function applied for correction of decay rates at temperatures below 0° C. The results showed that decomposition of clover green manure is rapid, even at temperatures below 5° C. N-rich plant material, therefore, should be worked into the soil as late as possible in the autumn or, preferably, remain on the soil surface until spring in order to reduce the probability of N losses.