Estimating yield response to temperature and identifying critical temperatures for annual crops in the Canadian prairie region

Susan M. Robertson, Scott R. Jeffrey, James R. Unterschultz, Peter C. Boxall
2013 Canadian Journal of Plant Science  
2013. Estimating yield response to temperature and identifying critical temperatures for annual crops in the Canadian Prairie region. Can. J. Plant Sci. 93: 1237Á1247. Historical yield and temperature data, by municipal district for the three prairie provinces, are used to estimate the marginal effect of exposure to specific temperatures in defined ranges during the growing season. Incorporating these non-linear temperature effects into the model improves yield forecasting for Canadian prairie
more » ... r Canadian prairie annual crops over models that use average temperatures or growing degree days. Critical maximum temperatures at which yields decline, calculated for winter wheat, spring wheat, durum wheat, barley, spring rye, fall rye, oats, canola and flax, range between 28 and 348C, depending on the crop. Additional critical minimum and maximum temperatures are estimated using the marginal effect of exposure to specific temperatures in defined ranges. Estimates of critical maximum temperatures and their marginal impact on yields are important for research and policy analysis on various issues and problems, including climate change, risk management instruments such as crop insurance, and development of heat tolerant crop varieties. Robertson, S. M., Jeffrey, S. R., Unterschultz, J. R. et Boxall, P. C. 2013. Estimation de la re´action du rendement a'la tempe´rature et identification des tempe´ratures critiques pour les cultures annuelles dans la re´gion des Prairies canadiennes. Can. J. Plant Sci. 93: 1237Á1247. On a recouru aux donne´es historiques sur le rendement et la tempe´rature dans les districts municipaux des trois provinces canadiennes des Prairies pour estimer l'effet marginal de l'exposition a'certaines tempe´ratures dans des plages bien de´finies, durant la pe´riode ve´ge´tative. Lorsqu'on inte'gre les effets non line´aires attribuables a'la tempe´rature au mode'le, on obtient de meilleures pre´visions de rendement que celles re´sultant des mode'les qui recourent a'la tempe´rature moyenne ou aux degre´s-jours de croissance pour les espe'ces annuelles cultive´es dans les provinces des Prairies canadiennes. Le maximum de tempe´rature a'partir duquel le rendement diminue a e´te´calcule´pour le ble´d'hiver, le ble´de printemps, le ble´dur, l'orge, le seigle de printemps, le seigle d'automne, l'avoine, le canola et le lin. Il se situe entre 28 8C et 34 8C, selon la culture. On estime d'autres minimums et maximums critiques en fonction de l'effet marginal qui re´sulte de l'exposition a'des tempe´ratures pre´cises, a'l'inte´rieur d'une fourchette de´finie. Estimer la tempe´rature maximale critique et son incidence marginale sur le rendement a de l'importance pour la recherche et l'analyse des politiques relatives a'divers dossiers ou proble'mes, notamment le changement climatique, les outils de gestion du risque comme l'assurance-re´colte et le de´veloppement de varie´te´s qui tole'reront la chaleur davantage. Mots clé s: Tempe´rature (critique), pre´vision du rendement Abbreviations: AMAPE, adjusted mean absolute percentage error; AVG, average air temperature; CRAIN, cumulative growing season rainfall; DEG, cumulative exposure to air temperature increments; GDD, growing degree day; ln, natural log; MAE, mean absolute error; MRAIN, monthly rainfall; PATEMP, partial aggregation of monthly cumulative exposure to air temperature increments; RMSE, root mean square error Can.
doi:10.4141/cjps2013-125 fatcat:poltb5m7ivaq3e5rjzrx53esui