Pesticide Loading Analysis in the Sacramento River Watershed
Project Contractor
Marty Williams, Waterborne Environmental, Inc.
williamsm@waterborne-env.com
SRWP Project Lead
Debra Denton, US EPA Region 9
denton.debra@epa.gov
Project Budget
Three phases of funding totaling $100,500
Project Schedule
September 14, 2005 – March 10, 2006
Project Background
The objective of this study was to develop a computer model that can be used to estimate pesticide loadings to the Sacramento River and its tributaries under varying pesticide application, weather, and agriculture management measure scenarios. Funding for all three phases of work for this project came from the Sacramento River Toxic Pollutant Control Program (SRTPCP), which is administered by the Sacramento Regional County Sanitation District and the U.S. Environmental Protection Agency (USEPA).
Project Description
Five chemicals (chlorpyrifos, diazinon, diuron, paraquat dichloride, and permethrin) were selected for analysis from a list of 22 chemicals identified by the Central Valley Regional Water Quality Control Board (CVRWQCB, 2005). The five chemicals were selected based on volume of use, toxicity, persistence in the environment, and the amount of chemical applied during the wet season.
A Geographical Information System (GIS) was used to construct approximately 43,000 model simulations representing unique combinations of soil, land use, and chemical use within the study area. Simulations were conducted using the U.S. Environmental Protection Agnecy’s (USEPA’s) Pesticide Root Zone Model (PRZM). Information about chemical use was obtained from the Pesticide Use Reporting (PUR) database. Environmental fate properties of the five chemicals were obtained from the USDA-ARS Pesticide Property database. Detailed land use data for the pesticide application sites in the 23 counties was also obtained from the PUR database. The land uses were associated with seven major categories for modeling: corn, fruit, grain, grape, grass, nut, and vegetable. Soil parameters were identified from the State Soil Geographic (STATSGO) database. Cropping dates for emergence, maturation, and harvest and other crop parameters for interception storage, maximum coverage, active root depth, aerial coverage, maximum canopy height were derived from USEPA’s Office of Pesticide Program’s standard “Tier 2” modeling scenarios. Simulations were conducted for 30-years of historical weather to evaluate runoff loadings under a range of potential low, moderate, and high rainfall events. The weather data was obtained from USEPA’s Center for Exposure Assessment Modeling (CEAM) for five meteorological stations within and around the watershed.
Simulations were conducted at the section level, which is the reporting level of the PUR database and has a resolution of 1 square mile. Edge-of-field predictions of pesticide runoff were aggregated (scaled-up) to the township scale (36 square miles) for visual presentation and to the county scale for tabular presentation. Model predictions were represented in terms of temporal probability of occurrence by calculating 50th and 90th percentile annual mass loadings.
Project Findings
Predicted pesticide loads were concentrated within nine counties, namely Butte, Colusa, Glenn, Sacramento, Solano, Sutter, Tehama, Yolo, and Yuba. Highest loadings occurred around the tributaries and streams of the major rivers within the study area. Predicted loads for diuron and paraquat dichloride were much higher than the remaining three with respect to percent of applied active ingredient. These two chemicals, along with diazinon, have the highest wet-season applications.
Areas of greatest uncertainty from a model setup standpoint relate to accurate knowledge and characterization of the field systems summarized by the PUR database and the fate and transport of each chemical under local conditions. This study should not be expected to predict accurate pesticide losses from individual fields. Rather, the study is best used to identify areas likely to contribute to pesticide loadings to aquatic systems. Areas predicted as having the highest sources may be candidates for more detailed analysis, monitoring, or mitigation (such placement of best management practices).
