Simulating winter wheat yields under temperate conditions: exploring different management scenarios

This paper describes a methodology for analysing management strategies to find best agronomic practices using a crop simulation model (CERES-Wheat). The study area is the estate of Imperial College at Wye, in the Stour Catchment, Kent, UK, an area highly suited to winter wheat production. The model is validated using historic crop performance data. Yield responses to differing sowing rates (range 200-450 seeds m(-2)), sowing dates and rates of nitrogen application (between 100 and 220 kg ha (-1)) with soil types of medium to heavy texture were simulated under water-limited conditions using historical daily weather data. In model validation, observed yields ranged between 6.9 and 7.4 t ha(-1), while simulated ranges were between 6.9 and 7.8 ha(-1). The RSMD of the difference was small (0.24 t ha(-1)) and non-significant. Optimum management practices (in terms of planting date, seed density and nitrogen application) were thereby defined. Also, simulations of potential yield (i.e. yield with no water and nutrient stress) were run for comparison. Results of this study reveal that the calibrated and validated CERES-Wheat model can be successfully used for the prediction of wheat growth and yield under conditions appropriate to Western Europe. (C) 2001 Elsevier Science B.V. All rights reserved.