WHEAT YIELDS FOR 50 YEARS AT SWIFT CURRENT, SASKATCHEWAN IN RELATION TO WEATHER

Half a century of wheat yield and weather records at Swift Current in southwestern Saskatchewan were analyzed to determine the response of wheat (Triticum aestivum L.) to changing weather patterns. Weather at Swift Current has undergone subtle but significant changes over the past 50 yr. Earlier years had disturbed conditions: hot, dry periods alternating with cool, wet ones resulting in yield fluctuations ranging from crop failures to maximum values. More recently the weather has been quiet: dry and cool but less variable from year to year. The resulting conditions were more favorable for near-normal but less variable yields. Simple precipitation-based yield–weather models developed two decades ago no longer apply, because temperature and precipitation patterns are currently out of phase relative to earlier conditions. A factorial yield–weather model was used to explain the complex relationship. This involved the summation of the product of several quadratic functions of various weather elements. Those elements considered were precipitation, maximum and minimum temperatures, global radiation estimated from duration of bright sunshine, evaporation from a buried pan, and time as an indicator of advancing technology. One function contained a term for the antecedant crop condition. The most important elements were precipitation for the summer-fallow period and for May, June and August; maximum temperatures for June and July; and global radiation for May. Advances in technology would seem to have very little influence on wheat yield trends after weather trends were accounted for. The model accounted for 73% (r = 0.854) of the yield variability and provided realistic functions for explaining the curvilinear influence of individual weather elements on wheat yield. The model is of a form that is readily adaptable for assessing, at any time during the crop development period, the influence of past and current weather on future expected yield. This could be useful for interpreting weather data in terms of crop production in weather and crop condition surveillance programs.