APEX model simulation of runoff and sediment losses for grazed pasture watersheds with agroforestry buffers

摘要

Buffers have been found to reduce non-point source pollution (NPSP) from watersheds. Hydrologic simulation models assist in predicting the effects of buffers on runoff and sediment losses from small watersheds. The objective of this study was to calibrate, validate and simulate runoff and sediment losses and compare buffer effects on NPSP losses relative to control watersheds (no buffer) for seven years. The experimental design consists of four watersheds under pasture management which were monitored from 2002 through 2008; two with agroforestry buffers (AgB 100 and AgB 300) and two control watersheds (CW 400 and CW 600). Pasture areas included red clover (Trifolium pretense L.) and lespedeza (Kummerowia stipulacea Maxim.) planted into fescue (Festuca arundinacea Schreb.) while the agroforestry buffer area included Eastern cottonwood trees (Populus deltoids Bortr. ex Marsh.) planted into fescue. The APEX model was calibrated from 2002 to 2005 and was validated from 2006 to 2008. The r (2) and NSE values for the calibration and validation periods of the runoff varied from 0.52 to 0.78 and 0.50 to 0.74, respectively. The model did not predict sediment loss very well probably due to insufficient number of measured events and low measured sediment loss. The measured runoff was 57% higher for CW watersheds compared to AgB watersheds. The measured sediment loss was 95% higher for CW watersheds compared to AgB watersheds. After calibrating and validating the model, it was run for long-term scenario analyses for 10 years from 1999 to 2008. Simulated buffer width had a significant influence on runoff. Simulated runoff decreased by 24% when the buffer width was doubled compared to losses associated with the measured buffer width. Simulated runoff from the CW watersheds was 11% higher with double stocking density (relative to measured density) compared to AgB watersheds with double stocking density. With half stocking density (relative to measured density), the AgB watershed had 18% lower runoff compared to CW. Results from this study imply that establishment of agroforestry buffers on grazed pasture watersheds reduce runoff and sediment losses compared to control watersheds without buffers.