Modeling the effects of genotypic and environmental variation on maize phenology: The phenology subroutine of the agmaize crop model

The predictive ability of process-based maize models is predicated on accurate prediction of phenology in terms of maize hybrids (genetics), abiotic factors affecting maize development (environment), and genotype × environment interactions. This chapter describes the phenology routines of a new process-based maize model, AgMaize, being implemented in the Decision Support System for Agrotechnology Transfer (DSSAT). It evaluates the model using a large number of diverse datasets and compares the outcome of the evaluation to predictions by the existing maize phenology model in DSSAT: the CERES-Maize phenology model. The life cycle of maize in AgMaize is divided into three phases: the pre-flowering, the flowering, and the post-flowering or grain-filling period. Except for germination, rate of development during the pre-flowering phase is based on the relationship between rate of leaf appearance and temperature. Predictions of phase duration by CERES-Maize were similar to those by AgMaize when the models were calibrated by modifying genotype coefficients.