| Fertilizer and water are the two factors restricting the agricultural production on dryland and particularly, they are the key factors restricting the improvement of rain-fed agricultural productivity on the Loess Tableland. Research on the relationship between water and fertilizer factors and how they affect crop growth and yield is important for the agricultural productivity of dryland.Based on the features of farmland on the Loess Tableland, farmland evapotranspiration (ET) characteristics and the crop production status under different fertilization treatments were studied. Crop growth, water use, and production potential under different conditions of water and fertilizer were simulated in combination with DSSAT model and other methods. The main conclusions are as follows:1. Analyses showed that the growth stages of winter wheat, in terms of the mean daily ET rate, were in the order of heading>jointing>filling>reviving> ripening>overwintering. The daily ET on farmland was 5.69 and 4.52 mm at heading stage and jointing stage, respectively. The averaged daily ET rate was decreased to 2.10 mm/d at grain filling stage due to soil water deficit. The monthly ET from April to June was higher than that in other months, being a total of 399.78 mm, and its amounts in the three months accounted for 42.05%, 27.48%, and 11.07%, respectively. The three months were the fastingium period during the whole growth period.2. Field experiments were conducted to investigate the effects of combined application of fertilizer N and P on wheat yield and water use efficiency in different years of precipitation. Results showed that 1) under the experimental conditions, both yield and water use efficiency with high N and P treatments were the highest in contrast to control treatment (CK). Wheat yield was increased by 68.37% and 62.95% in dry year and normal year and water use efficiency, 36.72% and 31.21%, respectively. 2) In different years of precipitation, the treatments had the order of high N > low N > CK for all of leaf weight, corn straw weight, grain weight, and 1000-grain weight. The five different treatments, for grain yield and water use efficiency, were in the order of high N and high P > high N and low P > low N and low P > low N and high P > CK.3. Estimated the potential productivity of winter wheat in the Changwu experimental area by using potential budget formula.The results showed that: the photo-temperature potential productivity of winter wheat from 1985 to 2004 slightly fluctuated annually. The values were between 6978.90 and 8999.08 kg/hm2 and the relative variability was 7.00%. The values of the precipitation potential productivity deviated from the photo-temperature potential productivity ranged from 284.23 to 1911.43 kg/hm2 and the relative variability was 8.91%. The photo-temperature potential productivity and precipitation potential productivity in normal year were higher than those in dry year and were increased by 112.4 and 1724.99 kg/hm2, respectively. The water meeting index, Wdm was 0.91 and 0.74 in normal and dry years, respectively. Because of the larger amount of precipitation in normal year, crop water supply can basically meet crop water demands. In normal year, water deficit rate was as low as 2.46% and decreased yield rate was 9.35%. In dry year, however, water deficit rate was as high as 7.29% due to significantly reduced rainfall, directly leading to a significant reduction in production. The current production decrease rate was as much as 25.60%.4. Winter wheat growth, soil water dynamics, and potential productivity were simulated using DSSAT model. Results showed that DSSAT model had better simulation results for biological yield, water use, and potential productivity. The simulated and measured results for biomass and LAI in different fertilizer treatments were basically consistent and the RMSEs for LAI in the three treatments were 0.56, 0.59, and 0.65, respectively. The RMSEs for biomass under the three treatments were 1291.57, 1681.64, and 1809.36, respectively, and the degree of coincidence between the simulated and measured results was relatively higher. The differences between simulated and measured results at different soil depths under different N fertilizer treatments were small and the relative error ranged from 0.0037 to 0.0051 cm3/cm3, indicating a higher degree of coincidence. The comparative study of the simulated, estimated, and measured values for the production potential of winter wheat from 1986 to 2009 revealed that compared with the simulated and measured values, the estimated values were generally high, indicating a poor result in the description of crop production potential. In comparison with potential budget formula, however, DSSAT model presents a better simulation result, with the values computed by the model being close to the measured values with high yield. |
PDF Full Text Request