Evaluation of the Ceres-maize model in the simulating maize (Zea Mays L)growth, development and yield at different planning dates and nitrogen rates in a subtropical environment of Zambia

Abstract

Crop simulation models can accurately predict yield with a priori knowledge of the soil properties and crop management practices. Generation of new information through traditional agronomic research practices is not sufficient to meet the needs for new agro-technologies and they are generally season specific, expensive and time consuming. Crop simulation models have not been widely used in sub-Sahara Africa due to lack of knowledge and under the local condition very little work has been done to evaluate the DSSAT (Decision Support System for Agrotechnology Transfer) v4.5 CERES (Crop Environmental Resource Synthesis Model) -maize model. The objective of this study was to evaluate the performance of CERES-maize model in simulating the effect of date of planting, nitrogen fertilizer and root-zone soil water profile on growth and yield of maize (Zea mays L.). A split-plot field experiment with three replicates was conducted at the Agricultural Field Station (15° 24' S, 28° 20' E; 1,261 m asl), University of Zambia, Lusaka during the 2013/2014 on Ustic Isohyperthermic Paleausalf. The main plot was assigned to date of planting (PD) (24th November [PD1], 8th December [PD2] and 22nd December [PD3]) and subplots were assigned to nitrogen application rates (112 and 168 kg N ha"1) under rain-fed condition. Plant analysis data was observed at vegetative and reproductive stages. Treatments effects were analyzed using Analysis of Variance and mean separation by Least Square Difference using GenStat version 16. Date of planting significantly (P < 0.05) affected number of ears at anthesis, dough, and maturity stages, aboveground biomass and grain yield. Nitrogen application significantly affected vegetative and reproductive stages at P < 0.05. Grain yield production varied from 7.6 to 10.7 ton ha'1 based on the results of a 14-day delay in planting date which significantly decreased the biomass production. Generalized Likelihood Uncertainty Estimation (GLUE) within DSSAT v4.5 was used to compute the genetic coefficients using phenological stages, grain yield and yield components so that the observed data compared well with simulated outputs. Phenological stage deviation from the observed were from -4.0% to 14.0%. Grain yield and yield components were also accurately simulated, with prediction deviations ranging from -8.0 % to 45.0 % for the three planting dates. Grain yield root mean square error (RMSE) and normalized RMSE were 1.8 ton/kg and 21.4%, respectively. The model accurately simulated tops (above ground biomass) weight (d-stat=0.96) and vegetative weight (d-stat=0.93) with reasonable accuracy. The leaf area index (LAI), leaf weight and stem weight were simulated with less accuracy due to poor values of forecasting efficient (-3.17 - -0.65) and d-stat (0.52 - 0.59). The LAI low coefficients of determination were due to poor performance of the model. Simulation of soil root water availability demonstrated that substantial potential yield may have been lost due to water stress under rain-fed conditions especially for the second and third date of planting. The CERES-maize model can be used to accurately predict planting date, above ground biomass and grain yield under the local condition with reasonable accuracy. This study recommends that future studies should focus on evaluating and validating the CERES-maize model using maize and other field crops under the three agro-ecological regions of Zambia.