Effects of elevated CO2 concentration and increased temperature on winter wheat: test of ARCWHEAT1 simulation model

Winter wheat (Triticum aestivum L., ev. Mercia) was grown in a controlled-environment facility at two CO₂ concentrations (targets 350 and 700 μmol mol^?1), and two temperature regimes (tracking ambient and ambient + 4°C). Observations of phenology, canopy growth, dry matter production and grain yield were used to test the ARCWHEAT1 simulation model. Dry-matter production and grain yield were increased at elevated CO₂ concentration (27 and 39%, respectively) and reduced at increased temperature (?16 and ?35%, respectively). ARCWHEAT1 substantially underestimated canopy growth for all treatments. However, differences in the facility environment from field conditions over the winter, indicated by the unusually rapid canopy growth observed in this period, meant that empirical model relationships were being used outside the conditions for which they were developed. The ARCWHEAT productivity submodel, given observed green area indices as inputs, overestimated the effect of CO₂ on productivity. An alternative, more mechanistic submodel of productivity, based on the SUCROS87 and Farquhar & von Caemmerer models, simulated observed crop biomass very closely. When these productivity simulations were inputed into the ARCWHEAT1 partitioning and grain-fill submodels, grain yield was predicted poorly, mainly as a result of the assumption that the number of grains is proportional to total growth during a short pre-anthesis phase. While yield was not correlated with growth in this phase, it was correlated with growth in longer pre-anthesis phases, indicating that ARCWHEAT1 could be improved by taking into account the contribution of earlier growth in determining yield.