Evaluation Of Nitrogen And Phosphorus Losses By Surface Runoff With Experiment And Simulation In Huaibei Plain

Nitrogen (N) and phosphorus (P) losses by surface runoff from croplands have been one of the major sources of the agricultural non-point source pollution. Experimental and modeling research on N and P losses by surface runoff have important theoretical significance and application value for understanding the transport mechanism of N and P as well as reducing agricultural non-point source pollution.N and P losses by surface runoff under different cropping systems and fertilization modes were evaluated based on the field experiment under natural rainfall. Then, the EPIC model was assessed with the field experimental data for evaluating the effect of tillage and fertilization practices on sendiment and nitrate and dissolved P (DP) losses. The SCS model was modified by incorporating the variations of amount and time of antecedent rainfall, and the power function was investigated to predict N and P losses and sediment with runoff.The conclusions and innovations obtained in this study are as follows:1. Cropping systems significantly influenced surface runoff, sediment, and N and P losses. Wheat-corn, wheat-cotton, and wheat-soybean reduced surface runoff of24%,26%, and58%, and sediment of15%,48%, and58%, respectively, compared with wheat-fallow. The lowest concentrations of nitrate, dissolved N (DN) and total N (TN) were from the wheat-corn; however, the lowest N and P losses were from the wheat-soybean reducing nitrate loss by6%, DN loss by31%, TN loss by44%, DP loss by60%, and total P (TP) loss by63%, compared with wheat-corn. The losses in peak events accounted for>64%for ammonium,58%for nitrate, and41%for DN of the total losses occurring during the3-year experimental period, and the least temporal variations of the P concentrations and losses were observed from wheat-soybean. Particular N and DP are the dominant forms of N and P in runoff from croplands, respectively. Therefore, selecting wheat-soybean as the main double cropping system appears to be a practical method for controlling runoff, sediment, and N and P losses from croplands in Huaibei Plain.2. Fertilization modes affected N and P losses by surface runoff. No-fertilizer reduced TN concentration by20%, TN loss by30%, DP loss by20%, and TP loss by18%, for corn. For soybean, No-fertilizer increased runoff of13%, and decreased yield of23%, losses of nitrate, DP, and TP, and concentrations of nitrate, DN, TN, DP, and TP. Side-dressing decreased runoff and sediment, and increased concentrations of nitrate, DN, and TN. Therefore, to reduce corn fertilization rate for corn and cotton appears to be a practices for reducing N and P losses in Huaibei Plain.3. Assuming the influencing extent of antecedent precipitation on runoff decreases daily following exponential function as antecedent precipitation is below the threshold, the antecedent moisture condition was updated and the SCS method was modified. The modified SCS method increased the computational accuracy, and successfully predicted surface runoff on a daily basis for all cropping systems. For example, the the modeling efficiency was<0.4for the original SCS method and0.4-0.8for the modified SCS method, and the modified SCS method increased the R2by24%, compared with the original SCS method.4. Runoff forecast method was developed with rainfall classification, weather forecast, and runoff simulation. The daily runoff was predicted by EPIC model with daily precipitation data during the25years, and then the frequency of runoff occurrence was calculated according to daily rainfall intensity range and1-day antecedent rainfall. The runoff forecast method was approved to forecast effectively the occurrence of runoff in Huaibei Plain.5. Runoff, sediment, nitrate loss, and DP loss were predicted more accurately with EPIC model than the modified SCS method and the power function method. However, power functions were proved to be used for the estimation of sediment, DN loss, PN loss, and TN loss. The simulation result of EPIC model showed that no tillage and residue cover reduced sediment from wheat-fallow, and losses of nitrate and DP increased with fertilization rates for wheat-corn and wheat-soybean.