ABSTRACT:We assessed the effect of greenhouse gas-induced climate change, as well as the direct fertilization effect of CO 2 , on crop yields in Québec, Canada. Our methodology coupled the transient diagnostics of 2 Atmosphere-Ocean General Circulation Models (CGCM1 and HadCM3) to the DSSAT 3.5 crop simulation system to simulate current and future (2040-2069) crop yields for spring wheat, maize, soybean and potato grown in 8 agricultural regions of Québec. For the future (2040-2069), we predict significant yield increases for soybean, lesser increases for wheat, no significant change for maize, and yield decreases for potato. These yields, especially for soybean, are further increased when incorporating the CO 2 fertilization effect, but vary according to the crop, climate scenario and agricultural region. Similar trends have been found in comparable agricultural regions in the Northeastern USA and in Southern Finland. These results are useful for designing appropriate crop and farm management adaptation strategies in response to future climate change.KEY WORDS: Climate change · CO 2 increase · Canada · Crop yields
Resale or republication not permitted without written consent of the publisherClim Res 34: [105][106][107][108][109][110][111][112][113][114][115][116][117] 2007 GCMs (general circulation models) CGCM1 and HadCM3, as well as data from the crop model DSSAT 3.5 (Brassard 2003).
METHODS AND DATAThis study uses the coupled climate scenario-crop model approach in which present and future climate conditions, generated by the selected climate models, following different CO 2 and other GHG emission scenarios, are integrated as inputs into the different crop models so as to simulate crop growth, development and production. All model variables, other than weather (e.g. soil, cultivar and management), are held constant between present and future crop yield simulations. Present and future crop yields are then compared to evaluate the impacts of GHG-induced climate change on agriculture. The A-OGCM climate models used in this study are the coupled CGCM1 of the Canadian Centre for Climate Modeling and Analysis (CCCma) and the HadCM3 of the British Hadley Centre. The CGCM1 model is forced by the IS92a emission scenario, whilst the HadCM3 model is forced by both the SRES A2 and B2 emission scenarios (Houghton et al. 2001). The selection of the A-OGCMs was based on their relevance to Canadian conditions and their ability to provide daily climate diagnostics at a relatively fine scale. The crop models used were those of the Decision Support System for Agrotechnology Transfer (DSSAT) version 3.5, namely CERES, CROPGRO and SUBSTOR, suitable for use under Canadian conditions (Mahdian & Gallichand 1997).This study was conducted in 8 of the 13 agricultural regions of the province of Québec, Canada (Fig. 1). These regions are those of the Régie des Assurances Agricoles du Québec (RAAQ 2001a). The choice of study region was based on the importance of the cultivated acreages and the desire to have as complete a representat...