Simulating planting date and cultivar effects on dryland maize production using CERESmaize model

Abstract

Maize farmers and extension agents in dry sudan savanna need information on how planting date and the choice of variety affect grain yield. This study was conducted to test the ability of model to predict maize yields under varying planting dates. Data on two open-pollinated maize cultivars (TSB-SR and TZE-COMP4) sown on different dates (June 29th, July 13th, July 21st and July 28th) in 2006 and 2007 at Azir (11° 01.820´ N, 12°37.714´ E; 441 m) and Damboa (11° 10.379´; 12° 47.145´E; 396 m) in the Sudan Savanna of Nigeria were used in running the model. Experimental data from Azir in 2006 was used to calibrate the model, while the data for 2007 at Azir 2006 and 2007 at Damboa were used for model validation. The model predicted days to anthesis at Damboa as reasonably well in both 2006 and 2007 (d-index >0.8), while at Azir, the prediction of days to anthesis was very poor in 2007. The match between predicted and observed grain yield were very good in 2007 at both locations. The root mean square error (RMSE) values for grain yield in 2007 were 431.5 and 226.5 kg ha-1 at Azir, and 799.5 and 611.5 kg ha-1 at Damboa for TZB SR and TZE COMP4, respectively, while the d-index values were all greater than 0.94. Generally, the model predicted decrease in grain yield with delay in planting date except for TZB-SR at Azir in 2006 where planting on July 13th gave higher yield than planting on June 29th. The grain yield values from the simulations suggested late June to early July as the optimum planting window for both varieties at both Azir and Damboa.