Abstract

We established the statistical relationships between seasonal weather variables and average annual wheat yield (Hard Red Spring and Durum wheat: Triticum spp.) for the period of 1979–2016 for 296 rural municipalities (RMs) throughout six soil zones comprising the arable agricultural zone of Saskatchewan, Canada. Controlling climate variables were identified through Pearson’s product moment correlation analysis and used in stepwise regression to predict wheat yields in each RM. This analysis provided predictive regression equations and summary statistics at a fine spatial resolution, explaining up to 75% of the annual variance of wheat yield, in order to re-evaluate the climate factor rating in the arable land productivity model for the Saskatchewan Assessment and Management Agency (SAMA). Historical climate data (1885–2016) and Regional Climate Model (RCM) projections for the growing season (May–August) were also examined to put current climatic trends into longer-term perspective, as well as develop a better understanding of possible future climatic impacts on wheat yield in Saskatchewan. Historical trends demonstrate a decrease in maximum temperature and an increase in minimum temperature and precipitation throughout all soil zones. RCM projections also show a potential increase in temperatures and total precipitation by 5 °C and 10%, respectively. We recommended against a modification of the climate factor rating at this time because (1) any increase in wheat yield could not be attributed directly to the weather variables with the strongest trends, and (2) climate and wheat yield are changing more or less consistently across the zone of arable land, and one soil zone is not becoming more productive than another. View Full-Text