TESTING THE ROOT-GROWTH SUBROUTINE OF THE CERES-WHEAT MODEL FOR 2 CULTIVARS OF DIFFERENT CYCLE LENGTH

The objective of this study was to test the root growth subroutine from the CERES-Wheat model during a growing season. The study involved observation and analysis of root growth and distribution throughout the growing season, and their simulation, using two wheat cultivars of different cycle length and two water regimes. The model accurately predicted crop development and yield, but overpredicted root depth by 90 cm at terminal spikelet and by 50 cm at booting, anthesis and mid grain-filling period. In addition, the model underpredicted root mass, presumably because of the way it handles root-shoot partitioning. There was good agreement between observed and simulated values of root-length density and soil water content (R2 = 0.66 and 0.86, respectively). The value of the specific root-length parameter in CERES-Wheat lies closer to data reported from a winter wheat crop experiment than to data from spring wheat reported in this and other studies. The model's ability to predict these attributes is discussed.