| Climate change is expected to drive major changes in agricultural production around the world, but estimates of the economic impact of these changes for Canadian agricultural production have been inconsistent. Most models use aggregate temperature data such as average temperature or growing degree days. This research shows that a novel approach that measures the marginal effect of exposure to specific temperatures in defined ranges improves yield forecasting. These novel temperature variables are incorporated into a production function to forecast yields for winter wheat, spring wheat, durum, barley, fall rye, oats, canola and flax. A spatial linear programming model in which gross margins are maximized is run for three scenarios: no climate change, a small increase in average temperature, and a large increase in average temperature. The model is calibrated to output from 2005 to 2010 and then run from 2011 to 2050. The model predicts that with a small increase in emissions, there will be a net increase in producer surplus to Canadian farmers, with wheat and canola dominating the landscape. This is similar to the current landscape; however, most crops migrate further north and west from their current range. As well, spring wheat acreage declines in favour of winter wheat, largely due to the higher yields for winter wheat. However, with a large increase in emissions, by 2050 the dominant crops in the landscape are barley and winter wheat, driven by changes in precipitation and temperature. The implications for Canadian agricultural production achieved by a spatially disaggregated model are a departure from the results of other modelling approaches and should be tested against a greater variety of behavioural assumptions and price conditions. Further study can help identify if crops other than those included here will become more prevalent. A major shift in the type of crops grown in the region would have implications for global food prices and food security. |
