Predicting the Response of the Forage Grass Timothy to Climate Change in Canada.

The effect of the projected climate warming on crop growth can be explored using an approach combining stochastic weather generators and crop growth models. The stochastic weather generator AAFC-WG was used to generate synthetic weather data for two periods (1961-1990, 2040-2069). CATIMO (Canadian Timothy Model), a process-based model, simulates the growth and nutritive value of the perennial grass timothy (Phleum pratense L.). Our objectives were 1) to compare simulations with CATIMO from synthetic and observed weather data for the period 1961-1990 at five locations in Canada and 2) to predict the yield and the digestibility of neutral detergent fiber (dNDF) of timothy at five locations in Canada for the period 2040-2069  using CATIMO. Dates of growth onset and harvests, DM yield, and dNDF simulated by CATIMO were similar with both synthetic and observed weather data (NRMSE < 5%); this confirmed that synthetic weather data represent the observed weather conditions for timothy growth in diverse climatic regions in Canada. Based on the average of two emission scenarios (A1B and A2) and two general circulation models (HadGEM and CGCM3), it is expected that 1) dates of timothy growth onset in spring will be 12 to 39 days earlier, 2) DM yield will be increased by 140 to 670 kg ha^-1 at the first harvest but decreased by 50 to 860 kg ha^-1 at the second harvest, 3) dNDF will be decreased mostly at the second harvest (-11 to -35 g kg^-1 NDF), and 4) one additional harvest will be possible.