Abstract
Crop production is highly sensitive to climate. It is affected by long-term trends in average rainfall and temperature, inter-annual climate variability, shocks during specific phenological stages and extreme weather events. As climate changes, crop production strategies must change too. Field trials were conducted at Nuclear Institute for Agriculture and Biology, Faisalabad-Pakistan, on wheat to assess nutrient and water productivity in irrigated semiarid conditions of Faisalabad. The treatments were six dates of sowing (DOS) (20th October, 30th October, 10th November, 20th November, 30th November and 10th December) with five nitrogen levels (N-levels), i.e., (0, 60, 120, 180 and 240 kg ha(-1)). CSM-CERES wheat model under the umbrella of DSSAT (4.6) was used to assess the impact of changing climate on wheat production. Model was then evaluated on the basis of data collected during field experiments. Model's performance was evaluated by computing different statistical variables (d, R-2). Intergovernmental Panel on Climate Change's (IPCC), Representative Concentration Pathways (RCPs) were used to assess the climatic changes in the near, the middle and at the last of the century (2030, 2050 and 2090). Data collected during field experiment showed that biological and grain yields were increased up to 10th November DOS and then a decreasing trend was started up to 10th December DOS. A statistically significant (p <= 0.05) interaction was observed between DOS and N-levels. DOS and N-levels interactive affect showed significantly higher biological and grain yields at 180 and 240 kg N ha(-1), respectively, when crop was sown on 10th November, while significantly lower yield at 0 kg N ha(-1) with sowing date 20th October. The model results indicated that N-levels have not much significant effect on wheat yield under changing climate scenarios RCPs. But the changes in DOS showed significant results under these RCPs in irrigated conditions. The DOS 30th November with 180 kg N ha(-1) will performs better in 2030, 2050 and 2090 than other DOS as predicted by the model. Model predicts the lowest yields in the early date of sowing, i.e., 30th October, 10th November, etc. But in later DOS model showed significantly higher yields for 2030, 2050 and 2090.