|dc.description.abstract||The performance of ten spring wheat varieties in respect to yield and yield components was analysed in a split-split plot design with location (Chirundu, Unza and Mpongwe representing Region I, II and III of Zambia respectively) as the main plot, planting dates (mid-April, mid-May,
mid-June and mid-July) as the sub plot and variety as the sub-subplot. Five heat tolerant lines and five local varieties were used. The planting dates represented a wide range of temperature regimes.
This was aimed at simulating the growing conditions of the crop when planting is delayed. The aim was to establish the extent of yield loss with delayed planting and also to determine which of the yield components are greatly affected with delayed planting and how strongly these yield components influence yield.
The varieties did not differ significantly in yield across location and planting dates. Locations had a significant influence on yield. Chirundu in the hot Zambezi valley gave the lowest average yield of the three locations. The locations and planting dates caused differences in number of spikelets/ear,number of tillers/square meter, thousand-kernel weight, plant height, leaf area index and harvest index and were lower at Chirundu compared to the other locations. The optimum planting date for Unza and Mpongwe was May while Chirundu it was April. Delayed planting resulted in low yields
and low expression of yield components. The effect of delayed planting was more pronounced in
Chirundu. The heat tolerant lines showed a slight edge in yield than the local varieties. The heat tolerant varieties showed superior performance in number of spikelets/ear, number of tillers/square
meter and plant height.
A stepwise multiple regression analysis identified number of spikelets/ear as the most important trait explaining variation in yield and the number of tillers/square meter and harvest index showing a marginal effect. When location and planting dates combinations were taken as environments in all locations July planting gave the most negative value of the environmental index. This indicated that this location-planting date combination was the most adverse environment for wheat production.
The stability analysis using Eberhart and Russell (1966) method, considering the regression coefficient, mean square deviation from regression and mean advantage over grand mean showed the heat tolerant lines tend to be much more stable than local varieties.
The overall indication of this study was that wheat breeders could deliberately select for heat tolerance with respect to yield stability. In this process targeting number of spikelets/ear and number of tillers/square meter may result in improved yields and stability of yield performance under even harsh temperature regimes, thus ensuring that good yields are obtained even when planting is delayed. The growing of heat tolerant varieties would encourage wheat production among both the small scale and large-scale farmers in the hot areas, which are well endowed with water resources.||en_US