Testing different CO2 response algorithms against a FACE crop rotation experiment.

Abstract A soil–crop–environment model was used to describe the combined effects of atmospheric carbon dioxide concentration [CO 2 ], temperature and precipitation on different agricultural crop species. Within this model, a set of algorithms describing CO 2 response to photosynthesis and crop water use efficiency, which differed in complexity and parameter requirements, was tested for its suitability to explain crop growth responses and soil moisture dynamics observed over 6 years in a crop rotation (winter barley, sugar beet and winter wheat) with two cycles under normal and elevated atmospheric CO 2 levels (FACE experiment; Weigel and Dammgen [46]). All algorithms were able to describe an observed increase in above-ground dry matter for all crops in the rotation, with Willmott's index of agreement (IoA) ranging between 0.93 and 0.99. Increasing water use efficiency with rising CO 2 was also successfully portrayed (IoA 0.80–0.86). A combination of a semi-empirical Michaelis–Menten approach, describing a direct impact of CO 2 on photosynthesis, and a Penman–Monteith approach with a simple stomata conductance model for evapotranspiration yielded the best simulations. This combination is therefore considered suitable for the description of yield responses to rising [CO 2 ] at the regional level. However, the performance of all tested algorithms was only marginally different at 550 ppm CO 2 .