Model-based optimisation of nitrogen and water management for wheat-maize systems in the North China Plain

Excessive nitrogen fertiliser application and irrigation in the North China Plain leads to nitrate accumulation in sub-soil and water pollution. HERMES, a dynamic, process-oriented soil-crop model was used to evaluate the effects of improved nitrate and water management on nitrate leaching losses. The model was validated against field studies with a winter wheat (Triticum aestivum L.)-summer maize (Zea mays L.) double-cropping system. A real-time model-based nitrogen fertiliser recommendation (NFR) was carried out for one wheat crop within the rotation and compared to farmers' practice and soil mineral nitrogen (N-min) content-based fertilisation treatments. Consequences of varying irrigation and annual weather variability on model-based NFR and further model outputs were assessed via simulation scenarios. A best-practice simulation scenario with model-based NFR and adapted irrigation was compared to reduced N and farmers' practice treatments and to a dry and a wet scenario. Results of the real-time model-based NFR and the other treatments showed no differences in grain yield. Different fertiliser inputs led to higher nitrogen use efficiency (not significant) of the model-based NFR. Increasing amounts of irrigation resulted in significantly higher N leaching, higher N requirements and reduced yields. The impact of weather variation on model-based NFR was smaller. In the best-practice scenario simulation, nitrogen input could be reduced to 17.1 % of conventional farmers' practice, irrigation water to 72.3 % and nitrogen leaching below 0.9 m to 1.8 % and below 2.0 m soil depth to 0.9 % within 2 years. The model-based NFR in combination with adapted irrigation had the highest potential to reduce nitrate leaching.