Modelling the nitrogen dynamics of maize crops – Enhancing the APSIM maize model

• Published in: European journal of agronomy 2018-10, Vol.100, p.118-131
• Main Authors: Soufizadeh, S., Munaro, E., McLean, G., Massignam, A., van Oosterom, E.J., Chapman, S.C., Messina, C., Cooper, M., Hammer, G.L.
• Format: Article
• Language: English
• Subjects: Ecophysiology; Maize; Modelling; Nitrogen; Simulation
• ISSN: 1161-0301; 1873-7331
• DOI: 10.1016/j.eja.2017.12.007

•An improved approach to modelling nitrogen dynamics in maize crops is proposed.•The approach uses N content per unit leaf area as a key driver of N demand.•Comprehensive model testing indicated improved predictions of crop N attributes.•The model is suited for simulation studies of crop adaptation (G*M*E).•Further, it can be used to explore the genetic underpinning of complex traits. Crop growth simulation models require robust ecophysiological functionality to support credible simulation of diverse genotype × management × environment (G × M × E) combinations. Most efforts on modeling the nitrogen (N) dynamics of crops use a minimum, critical, and maximum N concentration per unit biomass based empirically on experimental observations. Here we present a physiologically more robust approach, originally implemented in sorghum, which uses the N content per unit leaf area as a key driver of N demand. The objective was to implement the conceptual framework of the APSIM sorghum nitrogen dynamics model in APSIM maize and to validate the robustness of the model across a range of G × M × E combinations. The N modelling framework is described and its parameterisation for maize is developed based on three previously reported detailed field experiments, conducted at Gatton (27°34′S, 152°20′), Queensland, Australia, supplemented by literature data. There was considerable correspondence with parameterisation results found for sorghum, suggesting potential for generality of this framework for modelling crop N dynamics in cereals. Comprehensive model testing indicated accurate predictions at organ and crop scale across a diverse range of experiments and demonstrated that observed responses to a range of management factors were reproduced credibly. This supports the use of the model to extrapolate and predict performance and adaptation under new G × M × E combinations. Capturing this advance with reduced complexity compared to the N concentration approach provides a firm basis to progress the role of modelling in exploring the genetic underpinning of complex traits and in plant breeding and crop improvement generally.