Abstract
Climate change effects on UK winter wheat grain yield are complex: warmer temperature, negative; greater carbon dioxide (CO(2)) concentration, positive; but other environmental variables and their timing also affect yield. In the absence of long-term experiments where temperature and CO(2) concentration were manipulated separately, we applied the crop simulation model Sirius with long-term daily meteorological data (1892-2016) for Rothamsted, Hertfordshire, UK (2007-2016 mean growing season temperature 1.03°C warmer than 1892-1991), and CO(2) concentration over this period, to investigate the separate effects of historic CO(2) and weather on simulated grain yield in three wheat cultivars of the modern era. We show a slight decline in simulated yield over the period 1892-2016 from the effect of weather (daily temperature, rainfall and sunshine hours) at fixed CO(2) (294.50 ppm, 1892 reference value), but a maximum 9.4% increase when accounting for increasing atmospheric CO(2) (from 294.50 to 404.21 ppm), differing slightly among cultivars. Notwithstanding considerable inter-annual variation, the slight yield decline at 294.50 ppm CO(2) over this 125-year period from the historic weather simulations for Rothamsted agrees with the expected decline from temperature increase alone, but the positive yield trend with actual CO(2) values does not match the recent stagnation in UK wheat yield.