Performance and application of the APSIM Nwheat model in the Netherlands

Abstract

APSIM Nwheat is a crop system simulation model, consisting of modules that incorporate aspects of soil water, nitrogen (N), crop residues, and crop growth and development. The model was applied to simulate above- and below-ground growth, grain yield, water and N uptake, and soil water and soil N of wheat crops in the Netherlands. Model outputs were compared with detailed measurements of field experiments from three locations with two different soil types. The experiments covered two seasons and a range of N-fertiliser applications. The overall APSIM Nwheat model simulations of soil mineral N, N uptake, shoot growth, phenology, kernels m −2, specific grain weight and grain N were acceptable. Grain yields (dry weight) and grain protein concentrations were well simulated with a root mean square deviation (RMSD) of 0.8 t ha −1 and 1.6 protein%, respectively. Additionally, the model simulations were compared with grain yields from a long-term winter wheat experiment with different N applications, two additional N experiments and regional grain yield records. The model reproduced the general effects of N treatments on yields. Simulations showed a good consistency with the higher yields of the long-term experiment, but overpredicted the lower yields. Simulations and earlier regional yields differed, but they showed uniformity for the last decade. In a simulation experiment, the APSIM Nwheat model was used with historical weather data to study the relationship between rate and timing of N fertiliser and grain yield, grain protein and soil residual N. A median grain yield of 4.5 t ha −1 was achieved without applying fertiliser, utilising mineral soil N from previous seasons, from mineralisation and N deposition. Application of N fertiliser in February to increase soil mineral N to 140 kg N ha −1 improved the median yield to 7.8 t ha −1 but had little effect on grain protein concentration with a range of 8–10%. Nitrogen applications at tillering and the beginning of stem elongation further increased grain yield and in particular grain protein, but did not affect soil residual N, except in a year with low rainfall during stem elongation. A late N application at flag leaf stage increased grain protein content by several per cent. This increase had only a small effect on grain yield and did not increase soil residual N with up to 40 kg N ha −1 applied, except when N uptake was limited by low rainfall in the period after the flag leaf stage. The economic and environmental optima in winter wheat were identified with up to 140 kg N ha −1 in February, 90 kg N ha −1 between tillering and beginning of stem elongation and 40 kg N ha −1 at flag leaf stage resulting in a median of 8.5 t ha −1 grain yield, 14.0% grain protein and 13 kg N ha −1 soil residual N after the harvest. The maximum simulated yield with maximum N input from two locations in the Netherlands was 9.9 t ha −1.