ESTIMATION OF MAIZE GENETIC COEFFICIENTS FOR THE CERES-MAIZE MODEL USING NON-DESTRUCTIVE METHODS

Abstract

The Crop Environment Resource Synthesis-Maize (CERES-Maize) mechanistic model, included in the Decision Support System for Agrotechnology Transfer (DSSAT), is a useful and powerful tool that simulates the growth and yield of maize in different environments. The qualitative and quantitative information provided to the CERES-Maize model guarantees reliability in the simulations obtained. However, it requires a lot of information such as soil characteristics, daily climate, crop characteristics and management, as well as six genetic coefficients. This research assessed a non-destructive methodology for estimating the six required parameters (genetic coefficients): P1, P2, P5, G2, G3 and PHINT, based on the maize physiology, measured from the Growing Degree Days (GDD) base 10.  An experiment was established at the experimental station of the International Maize and Wheat Improvement Center (CIMMYT) in Tlaltizapan, Morelos, Mexico, where 27 white maize hybrids and 14 yellow maize hybrids were manually sown in an irrigation conservation tillage system. Once the simulations of maize growth and yield were obtained with CERES-Maize model, the genetic coefficients were calibrated using the Generalized Likelihood Uncertainty Estimation (GLUE).  After calibration of the six genetic coefficients for all hybrids,  an average values of P1, G2 and G3 were within the typical range,  while  P2 and P5  were greater than  the typical range, and PHINT was below typical range.  However,  the simulation model showed  good performance after calibration, according to the average R2 of 0.9809 and 0.9730 between  measured and simulated yields for white and yellow hybrids respectively. The coefficients estimated in this study can be used in the CERES-Maize model to simulate maize yields in different regions of the country for the hybrids used here.