(go to NO.99 Contents)

Association between grain yield and other traits: Simple correlations were computed between grain yield and other traits of the ten genotypes for each sowing date across the two seasons. Grain yield consistently and significantly correlated with biomass, grains/m2, spikes/m2, grain growth rate and biomass growth rate (Table 2). However, the magnitude of the correlation coefficients varied with different sowing dates. Harvest index and thousand kernel weight were significantly correlated with the grain yield of the optimum and late sowings but not with that of the early sowing. In contrast, grains/spike showed significant correlation with the grain yield of the early and optimum sowings but not with that of the late sowing. The above results suggest that biomass, grains/m2, spikes/m2, grain growth rate,biomass growth rate and vegetative growth rate are important traits under early heat stress conditions. The importance of biomass and spikes/m2 confirmed results reported under similar conditions (Ishag and Badredin 1996; Reynolds et al. 1998). On the other hand, harvest index, thousand kernel weight and grain filling duration might be useful predictors of the grain yield for the late sowing date together with biomass, grains/m2, grain growth rate and biomass growth rate. Many reports found a close association between grain yield and grains/spike and suggested grains/spike as a selection criterion for heat stress tolerance (Shpiler and Blum 1991; He and Rajaram 1994). Under the conditions of the current study, the time when the grains/spike is determined, especially for the late sowing date, occurs usually during the coolest parts of the season.
The association of heat stress intensity of grain yield with that of other traits: The heat stress intensity of yield between the optimum and early sowings was strongly associated with that of biomass, grain/m2, grain growth rate, biomass growth rate and vegetative growth rate (Table 2). No such associations were found with harvest index, thousand kernel weight and grain filling duration. In the case of late sowing, heat stress intensity of grain yield was associated with that of biomass, grains/m2, grain growth rate and biomass growth rate but was not strong as in the case of the early sowing except for grain growth rate. On the other hand, stress intensity of harvest index and thousand kernel weight were significantly associated with that of grain yield of the late sowing date.

The strong associations between stress intensity of grain yield with that of biomass, biomass growth rate and vegetative growth rate reinforce the importance of biomass production when early heat stress is concerned. Genotypes that fail to produce sufficient biomass under the early heat stress conditions suffer more than those that are able to keep their biomass production level high. A typical example is the two contrasting genotypes Condor and VYT #3 97/98. The former showed a great reduction in its biomass production and hence low yield, while the latter was able to produce more biomass, and hence more yield, under the early sowing condition (Data not shown). Grains/m2 also proved to be important trait under both heat stress conditions, but more important under early heat stress conditions. Our results on grain growth rate and grain filling duration demonstrated that although grain filling duration might be important for late sowing, grain growth rate was more closely related with grain yield under all growing conditions, suggesting the importance of this trait under all heat stress conditions (Wardlaw and Moncur 1995; Zahedi et a!. 2003). Grains/spike, spikes/m2 and vegetative growth rate might be useful traits for the early heat stress. On the other hand, harvest index and thousand kernel weight were shown to be more important under late heat stress. Potential selection criteria: Improving wheat yield under stressed environments is a difficult task compared with favorable environments. Under such condition, multiple selection criteria should be applied to accelerate the development of high yielding and stable genotypes.

Based on the results of associations of yield and its HSI with other related traits, a number of traits can be identified as selection criteria under both early and late heat stress conditions. Biomass, grains/m2,spikes/m2 and vegetative growth rate could be used as selection criteria under early heat stress. Harvest index, thousand kernel weight and grain growth rate together with biomass and grains/m2 could be potential selection criteria under late heat stress condition.

A number of physiological traits were proposed as selection criteria under heat stress conditions (Reynolds et al. 1998, 2001) including canopy temperature depression, stomatal conductance and membrane thermostability. Assessment of wheat genotypes by combining these traits and those identified in this study could considerably improve the efficiency of developing heat tolerant cultivars.

Currently, the performance of promising lines is verified by testing in different locations in all major wheat producing areas in the Sudan. This seems to be useful to study the reaction of those genotypes in a specific environment, but it does not give ideas about how such genotype will react when grown early or late in the season. Ali and Ageeb (1994) mentioned that late planting offers a mean to select for late heat tolerance more efficiently than normal planting, and it has proved to be useful in identifying easily measurable traits associated with yield under heat stress conditions of Egypt and the Sudan. Tolerance for heat is required at both early and late stages of crop growth in the Sudan due to the short cool period during the season. Therefore, it might be important to conduct trials in different sowing dates at least using the most promising lines in few representative locations so that it can be cost-effective and selecting tolerant cultivars using traits identified in this study.

<--Back | -->Next

(go to NO.99 Contents)