Author
 |
NOLAN, BERNARD - Us Geological Survey (USGS) |
|
Malone, Robert - Rob |
|
BARBASH, JACK - Us Geological Survey (USGS) |
|
Ma, Liwang |
|
Shaner, Dale |
|
Submitted to: Pest Management Science
Publication Type: Peer Reviewed Journal Publication Acceptance Date: 7/30/2014 Publication Date: 9/11/2014 Citation: Nolan, B.T., Malone, R.W., Barbash, J., Ma, L., Shaner, D.L. 2015. Data worth and prediction uncertainty for pesticide transport and fate models in Nebraska and Maryland, United States. Pest Management Science. 71:972-985. Interpretive Summary: Few studies have quantified the transport of both pesticides and metabolites in multiple agricultural settings of the United States. We used the Root Zone Water Quality Model (RZWQM) to predict the transport of the pesticide metolachlor and its metabolites, metolachlor ethane sulfonic acid (ESA) and metolachlor oxanilic acid (OXA), at field sites in Nebraska (NE) and Maryland (MD). Comparing model performance with measured field data showed generally satisfactory results for the NE site. Our results showed significant loss of metolachlor at NE by surface runoff, which could impact stream ecology. We were able to reliably estimate soil input variables for RZWQM such as bulk density. But other variables were less reliably estimated such as those affecting 1) metolachlor degradation and sorption to soil and 2) water flow through root and worm channels (macropores). However, our results show that the macropore and pesticide input variables were more important for the RZWQM predictions of metolachlor and ESA concentrations in subsurface soil. This disparity suggests that modeling could benefit from direct measurement of pesticide and macropore properties. This research will help model developers, model users, and agricultural scientists more clearly understand transport of pesticides and their metabolites in soil, and for science agencies such as USGS to determine data needs to more effectively use models to estimate agricultural impacts on water resources. Technical Abstract: Few studies have attempted to quantify mass balances of both pesticides and degradates in multiple agricultural settings of the United States. We used inverse modeling to calibrate the Root Zone Water Quality Model (RZWQM) for predicting the unsaturated-zone transport and fate of metolachlor, metolachlor ethane sulfonic acid (ESA), and metolachlor oxanilic acid (OXA) at field sites in Nebraska (NE) and Maryland (MD). Calibrated Koc values were 187–375 cm3 g-1 for metolachlor and 14–75 cm3 g-1 for ESA and OXA, and calibrated half-lives were 5–24 d for metolachlor and 107–200 d for ESA and OXA. Computed Kd values for metolachlor (3.2–4.1 cm3 g-1) and OXA (0.2–0.3 cm3 g-1) were similar to independently measured Kd values of these compounds (3.8 and 0.9 cm3 g-1, respectively). The index of agreement (IA) was > 0.5 for both sites and all observation groups except for pesticide degradates and nitrate at MD; however, IA = 0.54–0.66 for degradates at MD when the 0.52-m lysimeter data were excluded. Model fit to metolachlor, ESA, and OXA data at NE generally was satisfactory (IA = 0.68–0.85). Simulated mass balances showed significant loss of metolachlor at NE by surface runoff (112 g ha-1 yr-1), which could impact stream biota. Maximum predicted metolachlor concentration at NE (0.007 mg/L) was 7x greater than the chronic aquatic life benchmark for invertebrates. Calibrated macropore parameters (average diameter = 0.98–1.3 mm) were consistent with flow through former plant root channels rather than large openings typical of structured clay soils. Soil hydraulic parameters were more reliably estimated than pesticide and macropore parameters based on the available data. However, the latter were more important to the prediction uncertainty of maximum metolachlor and ESA concentrations. The disparity suggests that the modeling could benefit from direct measurement of field dissipation rates and macropore properties. |
