A model simulation analysis of soil nitrate concentrations—Does soil organic matter pool structure or catch crop growth parameters matter most?

Three different soil organic matter (SOM) submodels were tested within the framework of the soil–plant–atmosphere model Daisy. The three submodels were: the original Daisy SOM module (OLD) with relatively non-dynamic humus pools, a recalibrated Daisy SOM module (STD) with the same pool structure as the original, but with a more rapid turnover of the active humus pool, and a newly developed SOM model structure (CNSIM), with inclusion of a soil microbial residuals pool of relatively rapid turnover, and a relatively recalcitrant added organic matter pool, producing a larger and more sustained residual N mineralisation. Furthermore, two different parameterisations of the catch crop submodules, differing in grass growth and N assimilation dynamics, were tested and the relative influence of SOM module or catch crop growth module on the simulated variables assessed. The simulations were carried out with data from a field experiment with four mixed cropping systems and compared to measured results of crop production, N uptake and soil nitrate concentration. The cropping sequence was 3 years of grassland (cut or grazed) followed by 3 years of spring cereals with ryegrass as a catch crop and two levels of fertiliser application. Independently of the SOM module, plant production and nitrogen uptake for cereals were simulated well. The dynamics of the added organic matter (AOM) and SOM of the two Daisy submodules were nearly identical, whereas the CNSIM submodule built much more nitrogen into the AOM pools, especially during the pasture years. During the period with spring barley, the CNSIM module simulated similar amounts of AOM as the other modules. In general, the simulated nitrate concentrations at 100cm depth were higher than the measured values, but the changed dynamics in the CNSIM simulations resulted in even higher overestimation of the nitrate concentration than the two other modules. The choice of catch crop submodule had a considerable effect on nitrate concentration and therefore the potential for nitrate leaching, possibly overshading more futile differences produced by the different SOM submodules. The simulations show the importance of applying appropriate intercrop submodels when the model is used for simulating rotations with intercropping of grass-clover or undersown catch crops.