| Summary: | Insufficient water is the environmental factor most limiting crop productivity in the semi-arid and dry subhumid regions of Saskatchewan. This study was undertaken to establish a framework for the development of winter wheat cultivars which are less sensitive to drought stress. Five winter wheat genotypes, 'Norstar' and 'Norwin' (products of previous breeding efforts) and three more recently selected 'CDC Kestrel-type' advanced Lines were grown in 17 field environments between 1989 and 1991. This study established that the CDC Kestrel-type lines have a higher yield potential and higher average yield than Norstar or Norwin, but they do not differ in yield under most dryland conditions. The yield advantage of the CDC Kestrel type Lines was associated with conditions that favoured the establishment of yield potential. Conversely, the elimination of the CDC Kestrel-type yield advantage was associated with conditions which suppressed the establishment of yield potential. Differences in rainfall, evaporative demand, soil water depletion, evapotranspiration and aerial biomass accumulation during four periods of crop development resulted in three temporal patterns of drought stress: (1) intermittent, (2) terminal and (3) low stress. The environmental effect on grain yield was due to crop water conditions during all development periods. Crop water conditions from heading to anthesis were particularly important to grain yield. Flag leaf water content was positively related to grain yield in the dryland trials. There was no association between aerial biomass at anthesis and grain yield. Both pre- and post-anthesis ET were of similar importance to grain yield. Soil water reserves were depleted by the time the crop had headed in the dryland trials. Consequently, flag leaf water content declined. The distribution of growing season rainfall determined the timing and intensity of drought. A relatively small genotype-environment (GE) interaction for grain yield resulted in a poor correlation between yield potential and yield under drought. A portion of the GE-interaction resulted from non-uniform yield response to improved crop water conditions and from highly erratic yields among environments. The cumulative effect of weather over the season was more important to GE-interaction than weather during any development period. Nevertheless, crop water conditions between heading and anthesis were important determinants of the non-uniform yield response. Genotype-environment interaction was judged to be unimportant since it was mainly a consequence of genotypic differences in yield potential rather than drought tolerance. The potential number of kernels spike<sup>-1</sup> determined genotypic differences in yield potential. Consequently, grain yield in well-watered environments was sink-limited. The response of grain yield to improved crop water conditions was also related to the expression of kernel number. Drought stress reduced the expression of yield potential through lower shoot numbers and lower kernels spike<sup>-1</sup> resulting in fewer kernels. There was a positive relationship between aerial biomass and kernel number. The framework established in this study for improving winter wheat performance in Saskatchewan suggests reducing post-anthesis stress on the photosynthetic source while maintaining sink capacity through the establishment of adequate yield potential early in the season. It is proposed that increased kernels spike<sup>-1</sup> on fewer tillers may provide a balance between sink and source limitations to yield. Due to the need to rapidly establish yield potential, a more conservative pre-anthesis ET and crop growth will not likely result in higher yields. These observations suggested that a water saver model is not appropriate for crops in this region.
|
|---|
