Modelling the mineralization and leaching of nitrogen from crop residues during three successive growing seasons

A relatively simple model of the arable nitrogen cycle is presented containing modules to predict leaching, mineralization-immobilization turnover and crop grwoth and nitrogen uptake. The crop growth model was first calibrated against two of the sets of crop data from a given data-set (see McVoy et al., Ecological Modelling, this issue) and then the whole model tested aginst data from three experiments (of up to three yearsʹ duration) given optimal nitrogen and one where no nitrogen was given. Simulations of the data on the lightest soil and one experiment without nitrogen agreed particularly well with measured values. On other occassions the simulations of soil mineral N failed to respond to applications of fertilizer in the same way as the measurements. Possible reasons for this failure are discussed. The model also estimated dry matter and grain production, and nitrogen uptake and partitioning well at all sites. There is evidence that the model lags behind the field, however, in estimating immobilization and re-mineralization. Despite these differences the model may well find use in helping to provide fertilizer advice because of the cheapness of its predictions compared with taking measurements.