Charuwan Mankong and Sakda Jongkaewwattana²

Chiang mai University, Chiang mai. Thailand.

1/ M.Sc. (Agriculture) Thesis

2/ Thesis advisor

……………………………………

This study focused on the response of planting dates and estimation of genetic coefficient of Thai hybrid maize varieties for CERES-Maize model. In addition, performance of the model in terms of growth and development as well as yield using the estimated genetic coefficients were also studied. The study consisted of 2 parts, the first part was field experiment conducted in order to investigate the response of maize to environmental factors in association with different planting dates. Data from this experiment were then used for the estimation of genetic coefficient for the CERES-Maize model. The second part of this study was model evaluation.

Field experiment was conducted at multiple cropping centre, Chiang mai University during March to December 2000. Split plot in RCB with 3 replications were applied in which main plot was planting dates (16 March, 2 May, 15 June, and 1 August) and sub-plot was 3 Thai hybrid maize varieties namely NSX 991003, DK 999 and NSX 982013.

Field experiment results showed that phenological stages i.e. silking date and phenological maturity date of all 3 maize varieties depend upon growing degree day (GDD). In addition, the estimation of each phenological stage of maize using GDD was found more accurate than using of number of days counted from planting. Planting dates demonstrated significant effect on some plant growth parameters, i.e. plant height, number of days to maximum leave dry weight, maximum leave dry weight, leave growth rate, and stem growth rate. These parameters showed an increasing trend along with late planting from 16 March and had maximum values at 1 August planting. However, planting date had no effect on number of days to maximum stem dry weight and maximum stem dry weight. All 3 maize varieties planted on 15 June produced lowest grain yield which were 914 kg/rai for NSX 991003, 1,778 kg/rai for DK 999, and 1,030 kg/rai for NSX 982013. Experiment results showed that grain yield increased with later planting. However maximum grain yield was obtained at 15 June planting, i. e. 1,064, 1,318, and 1,221 kg/rai for NSX 991003, DK 999, and NSX 982013 respectively. Planting date showed no effect on number of grains per row, 100 grain weight, and number of kernels per plant.

Using the estimated genetic coefficients as an input to run the model, it was found that the model can simulate phenological stage, i.e. silking date and phenological maturity date of all 3 varieties accurately with an estimation error of + 2 days. Moreover, the model simulated dry matter accumulation in the same fashion as observed data in corresponds with planting dates. However, the model over-estimated above ground dry matter, grain yield, harvest index, and single grain weight. The deviation of simulated grain yield and observed grain yield ranged between 23 – 44%. In contrast, the model under-estimated number of grain per m² and number of grain per row.

In conclusion, results of this study suggest that using of GDD to estimate phenological stage of Thai hybrid maize was more accurate than estimation from number of days counted from planting. The CERES-Maize which used the concept of GDD to estimate phenological stages demonstrated good simulation of phenological stages of all 3 maized varieties used in this study. Simulation results showed over-estimation of grain yield and some other parameters. This could be because of the assumptions of the model which was set that there was no lossed from pest, e.g. diseases, insect pest, and weed as well as other natural factors such as lodging. Thus simulated yield could be recognized as potential yield. Generally speaking, CERES- Maize model can be able to simulate maize growth and yield in response to planting dates reasonably.

charuwan