The Use of Water Erosion Prediction Project to Model Runoff and Erosion in No-Till Dryland Agroecosystems.
Kate Norvell1, Neil Hansen1, James Ascough2, and Greg Butters1. (1) Colorado State University, Colorado State University, Dep. Soil And Crop Sci. C127 Plant Sci. Bldg, Fort Collins, CO 80523-1170, (2) USDA-ARS, USDA-ARS-NPA ASRU, 2150 Centre Ave., Bldg. D, Suite 200, Fort Collins, CO 80526
Runoff and erosion in dryland agroecosystems represent important and significant sources of water and nutrients losses, which can lead to sustainability and environmental concerns. Dryland systems are water limited and cannot spare water losses that could be captured with correct or improved management practices. The use of simulation models can add to the understanding of the impact of management decisions on soil-water storage. The objective of this study was to expand the understanding of runoff and erosion in no-till dryland agroecosystems in a semi-arid environment beyond the two years of the observed runoff and erosion data in Stratton, CO using the Water Erosion Prediction Project (WEPP) model. A detailed sensitivity analysis was performed on model inputs for the single storm and continuous versions of WEPP. In terms of soil properties, the WEPP model predicted runoff and sediment loss were most sensitive to effective hydraulic conductivity (Keff). Even after parameterization of observed runoff events, the model was unable, in general, to make accurate predictions for individual storms in the single storm and continuous mode of the model. However, average predictions of runoff and sediment loss were comparable to yearly totals observed at the study site. Consistent with observations the forecast model predictions clearly show that corn is the source of major runoff in both no-till (NT) rotations. Despite high runoff values in corn the model predictions for the rotation average of NT wheat-corn-fallow (WCF), NT wheat-corn-millet (WCM) and tilled wheat-fallow (WF) were similar at 30 mm, 28 mm, and 33 mm of annual runoff, respectively. To put the runoff numbers in context an estimate was made to quantify the amount of precipitation that was not used by the crop in the form of evapotranspiration (ET). By using the forecast model simulation from this study, long-term average yields from the study site, and crop specific yield and ET relationships developed in Colorado it was determined that runoff makes up about 20 % of these losses from both cropping systems. It can be assumed the remaining losses are attributed to evaporation. Despite small overall differences in runoff, large differences in sediment loss were predicted, with the greatest loss in tilled WF, 1.9 tons ha-1 yr -1, and 0.71 tons ha-1 yr -1 and 1.4 tons ha-1 yr -1 for WCF and WCM respectively. The model predictions suggest that a change from a tilled WF system to a NT system has a minor impact on runoff, but does decrease sediment loss by 0.5 – 1.2 tons ha-1 yr -1. Therefore, the sensitivity analysis determined Keff to be the parameter most sensitive to runoff and erosion. Despite, parameterization of Keff in single storm and continuous mode, the WEPP model was still unable to accurately predict runoff and erosion on an individual storm basis. Long-term predictions made by the WEPP model seem reasonable for the system.