Pacific Northwest and California freshwater resources are key elements in the life history and ecology of Pacific salmon and steelhead listed as threatened or endangered under the Endangered Species Act. Risk to listed Pacific salmonid species can be assessed by evaluating the spatial and temporal co-occurrence of salmonid species at sensitive life stages and pesticide concentrations at levels that may elicit adverse effects. Understanding the potential for the co-occurrence requires knowledge of pesticide use patterns and application methods, pesticide properties that influence environmental fate, as well as landscape/land management, edaphic, and climatic factors that influence off-site movement into surface water. Reported here is the use of a passive sampling device to monitor selected current use pesticides in surface water on a continuous basis. Passive sampling devices (PSDs) were deployed continuously in 5 watersheds within the Pudding River subbasin, critical habitat for the Upper Willamette River Chinook and Steelhead ESUs, between June 2010 and October 2011 in order to characterize the temporal trends in surface water concentrations. PSDs were deployed in off-channel habitats preferred by juvenile salmonids. The majority of the monitoring results were well below EPA aquatic life
benchmarks, as well as levels of concern for listed salmonids. Using the EPA ecoregion framework watershed sensitivity to pesticide surface water loading was characterized. The Soil and Water Assessment Tool (SWAT) was used to assess the relationship between land management practices and PSD monitoring data collected in the Zollner Creek watershed. SWAT was evaluated under different parameterization scenarios representing increasing levels of local knowledge of the system in order to evaluate model performance in relation to average daily stream flow. Using spatially distributed precipitation data and incorporating engineered drainage features into model parameterization resulted in a satisfactory fit of average daily stream flow indicating satisfactory characterization of the watershed hydrology. SWAT was then used to simulate the fate of chlorpyrifos and trifluralin, the two most commonly detected pesticides in the PSD monitoring. The pattern of simulated time-weighted average (TWA) pesticide concentrations was similar to measured values. However, simulated pesticide TWA concentrations consistently underestimated measured values. The most likely source of this bias is underrepresented pesticide use practices.