The study was undertaken to model the impact of climate change on wheat production under arid and semiarid conditions of Punjab, Pakistan. The data required to run the model was collected from field experiment conducted at four sites viz. Agronomic Research Area, University of Agriculture Faisalabad, Adaptive Research Farm Gujranwala, Regional Agricultural Research Institute Bahawalpur and Barani agricultural Research Institute Chakwal. Measurement of crop growth and environmental variables were made to establish the causes underlying the variation in grain yield associated with cultivar and varying environment.
- Cropping system model CSM-CERES-Wheat (V 4.0.2) was used to simulate the growth, development and yield of wheat crop sown at different location under adequate water and nutrient supplies. The model predicted phonological development of all cultivars reasonably good and RMSE between observed and simulated days to maturity was 1.82, 1.0, 1.6 and 3.82 days at Bahawalpur, Faisalabad, Gujranwala and Chakwal sites respectively.
- The difference among stimulate and observed grain yields were 469, 102, 673 and 1045 kg ha-1 at Bahawalpur, Faisalabad, Gujranwala and Chakwal sites respectively while the difference in observed and simulated Total Dry Matter (TDM) accumulation values were 338, 260 and 729 kg ha-1 during calibration of model with six cultivars at each location.
- The CSM-CERES-Wheat model was validated for grain yield using data from other experiments conducted during 2008 at same location and treatments. The model simulated the grain yield very closely to the observed data. The average error between simulated and observed values 9.3% and values for RMSE were ranged from 185 to 1226 kg ha-1 among all sites while Mean Percentage Difference was ranged from 2.7 to 13.6.
- Crop duration had no effect at all location with increase in CO2 concentration from 360 ppm to 550 ppm. However TDM would increase with elevated CO2 concentration.
- Increase in CO2 concentration from360 to 550 ppm had significant effects on grain yield at all locations. These effects were considerably higher (3.82 %) more yield at Faisalabad as compared to 3.12, 2.92 and 2.32 % at Bahawalpur, Gujranwala and Chakwal with 550 ppm CO2.
- The results showed that increase in temperature shortened the crop duration from planting to physiological maturity, retarded growth and decreased yield up to 15 % under 30C increase in temperature as compared to current at all locations.
- Change in rainfall (3 and 6 % increase) had no effect on crop phenology, while minor difference were observed in grain yield and TDM production at Bahawalpur, Faisalabad and Gujranwala
- where crop was under canal irrigated conditions. However, at Chakwal site where crop was dependent on rainfall increase in rainfall has significant effects.
- Optimization of sowing date and cultivar selection with seasonal analysis tool showed that Sehar ans AS-2002 cultivar will be the most efficient cultivars at Faisalabad and Gujranwala sites but Iqbal-2002 will give better yield at Bahawalpur site. Cultivar Inqilab and Bhakkar will be efficient at Chakwal location that will give better yield and maximum monetary returns at 30C increase in current temperature.
Based on the results, the following conclusion can be made
- Testing of CSM-CERES-Wheat model and application in this study confirmed that this model could be acceptable for use as a research tool in the variable agro-environment of Punjab (Pakistan). The results suggested that the model can be used to guide alternate ways of improving wheat production in Pakistan. However, some model inputs for Punjab need to be determined including the genetic coefficients of various wheat varieties and the minimum data set for soils and weather data as a whole country.
- Climate change analysis indicated strong of temperature increases on wheat production in the Punjab. The yield will be substantially decreased with increasing temperature up to 30C in the region. Under the present scenarios, it seem that average grain yield will decreased by 15 % beyond 2050. However, the variability of yield at various locations is likely to increase significantly.