Headwater stream condition and nutrient runoff: Relating SWAT to empirical ecological measures in an agricultural watershed in Pennsylvania.

Managing nonpoint sources of nutrients and sediments is the primary challenge for improving conditions in the Susquehanna-Chesapeake basin. Aquatic macroinvertebrates are widely used indicators of stream ecological integrity, but the relationship between nutrient runoff and macroinvertebrate response remains indistinct. Logistical and financial hurdles hinder collection of high-resolution empirical nutrient data, but landscape-based models like the Soil and Water Assessment Tool (SWAT) offer a more practical approach. Nutrient runoff was simulated with SWAT for a small, upland, agricultural Pennsylvania watershed. Three levels of ecological assessment were used to interpret SWAT results. Macroinvertebrate communities (intensive) were sampled at 14 sites and described using an Index of Biotic Integrity (IBI). Biological integrity was moderately degraded in many reaches. The Stream-Wetland-Riparian (SWR) Index (rapid) and landscape metrics (remote) also indicated prevalent agricultural stressors. Baseflow nitrate grab samples, collected once per season, showed no significant relationship with IBI score. Thirty spatiotemporal scales of nutrient data were extracted from SWAT for phosphorus, nitrate, and organic nitrogen. Best subsets regression was performed on IBI scores using SWAT, land cover, and SWR variables. Results were significant (p < .001) with high R2 values (84.8 and 86.2), signifying a negative relationship between instream nutrient concentration and IBI score. This study demonstrates the viability of SWAT as an alternative to in-field nutrient sampling, the value of spatiotemporal scale in model outputs, and the importance of site condition variables in relating nutrients to stream ecological health.