HUME-Maize: modelling silage maize yield formation across genotypes and scales

Accurate crop growth models are essential for predicting yield and evaluating adaptation potentials under climate change. As one of Europe’s most important forage crops, silage maize ( Zea mays L.) requires a model that captures genotype-specific growth traits and environmental variability. The current study presents HUME-Maize (Hannover University Modelling Environment), a new process-based model for silage maize, specifically adapted to conditions in Northern Europe. Model development relied on an extensive dataset from two field trials (2007–2008, 2021–2022) at five sites across Germany, involving two mid-early maize hybrids released 12 years apart. Calibration was based on the older hybrid and evaluated on both, testing the need for hybrid-specific parameter adjustments. Model scalability and predictive accuracy were assessed with a nationwide dataset covering 287 site-year combinations. The baseline model, parameterized using an older hybrid, reproduced biomass formation well ( d = 0.98 and 0.95). Application to a newer hybrid without adjustment revealed systematic differences in leaf development and yield formation associated with breeding progress. Hybrid-specific adjustments of physiological traits, e.g. leaf growth rate, specific leaf area and radiation use efficiency, substantially improved model performance, reflecting the enhanced radiation use efficiency of the newer hybrid. On the national scale, the model achieved low prediction errors for final yield (16 % for the older and 21 % for the newer hybrid). Drought and nitrogen stress responses remain areas for refinement. HUME-Maize thus provides a robust framework for simulating hybrid-specific silage maize growth, supporting yield prediction and physiological analysis across hybrids and environments.