HEAT TOLERANCE FOR YIELD AND ITS COMPONENTS IN DIFFERENT WHEAT CULTIVARS

Twenty one diverse, standard and experimental cultivars of common spring wheat (Triticum aestivum L.) were tested for the effect of heat stress on phenology, yield and its components by growing the materials for 2 years under full irrigation during the hot summer (offseason), and the cool winter (normal) conditions. Heat tolerance was estimated for each variable by the 'heat susceptibility index' (S) which scales the reduction in cultivar performance from cool to hot conditions relative to the respective mean reduction over all cultivars. Genotypes differed significantly in S for yield and its components. The ranking of cultivars in S over the 2 years was consistent for yield, kernels per spike and kernel weight, but not for spike number. Of the three yield components, the greatest genotypic variation in S was expressed for kernels per spike. However, S for yield could not be simply attributed to S in a unique component across all cultivars. On the other hand, a general linear model regression of summer yield on its components revealed that the most important yield component affecting yield variation among cultivars under heat stress was kernel number per spike. Kernel number per spike was positively associated across cultivars with longer duration and greater stability of thermal time requirement from emergence to 'double ridge'. It is therefore concluded that kernel number per spike under heat stress is a reasonable estimate of heat tolerance in yield of wheat and that this tolerance is operative already during the first 2 to 3 weeks of growth.