Submitted to: Transactions of the ASABE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/15/2005
Publication Date: 2/15/2006
Citation: Wang, X., Harmel, R.D., Williams, J.R., Harman, W.L. 2006. Evaluation of EPIC for assessing crop yield, runoff, sediment and nutrient losses from watersheds with poultry litter fertilization. Transactions of the ASABE. 49(1):47-59. Interpretive Summary: Computer modeling tools are valuable for research and decision making regarding economic and environmental impacts of natural resource management alternatives. Models provide a cost-effective method to the expensive and difficult task of extensive data collection efforts for exploring management impacts. One such management alternative is the use of poultry litter fertilizer for crop production. Although modeling is a powerful tool, some data collection is necessary to adjust modeling procedures and evaluate model performance. In this study, the Environmental Policy Integrated Climate model version 3060 (EPIC3060) was used. Four years of data (2001-2004) collected from six cultivated experimental watersheds near Riesel, Texas, were used as input and evaluation data for the model. The study watersheds were fallow in 2001, cropped with corn in 2002 and 2003, and with winter wheat in 2004. Each of the six watersheds received an annual poultry litter application at target rates from 0 to 6 t/ac. Statistical tests showed that the annual crop yield, monthly runoff, and monthly sediment and nutrient loss values simulated with EPIC were not different from measured values. The lone exception occurred for soluble P losses on the control watershed. Overall, the results indicate that EPIC has the ability to accurately simulate corn and wheat yields, runoff, and sediment and nutrient losses for poultry litter application watersheds.
Technical Abstract: The EPIC3060 (Environmental Policy Integrated Climate version 3060) was calibrated and validated using four years (2001-2004) of data collected from six cultivated experimental watersheds near Riesel, Texas. The study watersheds were fallow in 2001, cropped with corn in 2002 and 2003, and followed by winter wheat in 2004. Poultry litter application rates from 0 to 15.1 Mg ha**-1 were randomly assigned to each of the watersheds. The monthly data of 2001-2002 from one watershed were used to calibrate the initial CN2, erosion control practice factor, RUSLE C factor coefficient parm(23), and phosphorus sorption ratio. EPIC was validated using the remaining data. The percent errors (PE) for average annual corn yield and average wheat yield across six watersheds were under 1%. The modeling efficiency (EF) was 0.96, and R**2 was 0.97 based on annual crop yields. EPIC successfully predicted surface runoff on annual, monthly, and daily basis for all watersheds with EF values larger than 0.5 and R2 larger than 0.7. The sediment, organic N and P, soluble P, and NO3-N losses predicted by EPIC were satisfactory, with EF values ranging from 0.59 to 0.87 and R2 from 0.63 to 0.92 based on annual comparisons. PE values for these environmental indicators were within 10%, except for organic N losses. EF values based on monthly comparisons were generally larger than 0.4 (in 25 out of 30 tests). Paired t-tests for annual crop yield, monthly runoff, and monthly sediment and nutrient losses showed that the EPIC simulated means were not significantly different from that of observed values at the significance level of alpha = 0.05, except for soluble P losses for watershed Y6. These statistical tests indicate that EPIC has the ability to accurately simulate corn and wheat yields, runoff, and sediment and nutrient losses for poultry little application watersheds.