Publications des agents du Cirad


How accurately do crop models simulate the impact of CO2 atmospheric concentration on maize yield and water use ? [O-2223-02]

Durand J.L., Delusca K., Boote K.J., Lizaso J., Manderscheid R., Rosenzweig C., Jones J., Weigel H.J., Ruane A.C., Anapalli S.S., Ahuja L., Basso B., Baron C., Bertuzzi P., Ripoche D., Biernath C., Priesak E., Deryng D., Ewert F., Gaiser T., Gayler S., Heilein F., Kersebaum K.C., Kim S.H., Müller C., Nendel C., Ramirez J., Tao F., Timlin D., Waha K., Twine T.E., Wang E., Webber H., Zhao Z., Rötter R.P., Srivastava A., Seidel S.. 2015. In : Our Common Future under Climate Change. International scientific conference Abstract Book 7-10 July 2015. Paris, France. Paris : CFCC15, p. 315-315. Our Common Future under Climate Change, 2015-07-07/2015-07-10, Paris (France).

Methods and Results Given the incertitudes on the climate change impacts on C4 crops, projections of regional maize production remain speculative. Assessment of the impacts of atmospheric CO2 concentration ([CO2]) on crop yield and resources uses using mechanistic models becomes increasingly important. Free Air [CO2] Enrichment (FACE) studies offer data to test and improve model quality. The objective of this work by the AgMIP Maize group was (i) to test multiple maize models for [CO2] responses against data gathered from a FACE study under two water regimes carried out in Germany during 2007 and 2008, and (ii) to pave the way to potential model modifications so as to improve their simulations of crop responses to [CO2]. The Experiment combined two [CO2] levels with two watering regimes. Yield, leaf area, soil water content and [CO2] levels were recorded both years, 2008 only exhibiting significant water deficit. After a preliminary calibration based on non limiting water conditions and under ambient [CO2] treatments of both years, a blind simulation was undertaken for the other treatments: High [CO2] (550 ppm) 2007 and 2008, both watering regimes, and DRY AMBIENT 2007 and 2008. Secondly, with full growth and yield data along with soil moisture data of all treatments, improvements of simulation results were attempted. Changes made to the models have been documented and submitted for further analysis. The results revealed: minimal [CO2] impacts with low variations among " uncalibrated " models except for the dry season of 2008 where the observed drought impact was simulated by the majority of models; most models caught but underestimated the CO2 impact on crop water status, leaf area, grain number and yield; the CO2 effect on transpiration was generally properly simulated, transpiration per leaf area decreasing but green leaf area duration increasing at 550 ppm [CO2]. As more data from FACE experiments become available, it will be highly desirable to replicate this exercise in order to come up with more robust conclusions on these responses and to improve model response to CO2. (Texte intégral)

Documents associés

Communication de congrès

Agents Cirad, auteurs de cette publication :