Pesticide trapping efficiency of vegetated filter strips (VFS) is commonly predicted with low success using empirical equations based solely on physical characteristics such as width and slope. The objective of this research was to develop and evaluate an empirical model with a foundation of VFS hydrological, sedimentological, and chemical specific parameters. The literature was reviewed to pool data from five studies with hypothesized significant parameters: pesticide and soil properties, percent reduction in runoff volume (i.e., infiltration) and sedimentation, and filter strip width. The empirical model was constructed using a phase distribution parameter, defined as the ratio of pesticide mass in dissolved form to pesticide mass sorbed to sediment, along with the percent infiltration, percent sedimentation, and the percent clay content (R(2) = 0.86 and standard deviation of differences [STDD] of 7.8%). Filter strip width was not a statistically significant parameter in the empirical model. For low to moderately sorbing pesticides, the phase distribution factor became statistically insignificant; for highly sorbing pesticides, the phase distribution factor became the most statistically significant parameter. For independent model evaluation datasets, the empirical model based on infiltration and sediment reduction, the phase distribution factor, and the percent clay content (STDD of 14.5%) outperformed existing filter strip width equations (STDD of 38.7%). This research proposed a procedure linking a VFS hydrologic simulation model with the proposed empirical trapping efficiency equation. For datasets with sufficient information for the VFS modeling, the linked numerical and empirical models significantly (R(2) = 0.74) improved predictions of pesticide trapping over empirical equations based solely on physical VFS characteristics.

Macropore flow results in the rapid movement of pesticides to subsurface drains, which may be caused in part by a small portion of macropores directly connected to drains. However, current models fail to account for this direct connection. This research investigated the interrelationship between macropore flow and subsurface drainage on conservative solute and pesticide transport using the Root Zone Water Quality Model (RZWQM). Potassium bromide tracer and isoxaflutole, the active ingredient in BALANCE herbicide [(5-cyclopropyl-4-isoxazolyl) [2(methylsulfonyl)-4-(trifluoromethyl)phenyl] methanone], with average half-life of 1.7 d were applied to a 30.4-ha Indiana corn (Zea mays L.) field. Water flow and chemical concentrations emanating from the drains were measured from two samplers. Model predictions of drain flow after minimal calibration reasonably matched observations (slope = 1.03, intercept = 0.01, and R(2) = 0.75). Without direct hydraulic connection of macropores to drains, RZWQM under predicted bromide and isoxaflutole concentration during the first measured peak after application (e.g., observed isoxaflutole concentration was between 1.2 and 1.4 mug L(-1), RZWQM concentration was 0.1 mug L(-1)). This research modified RZWQM to include an express fraction relating the percentage of macropores in direct hydraulic connection to drains. The modified model captured the first measured peak in bromide and isoxaflutole concentrations using an express fraction of 2% (e.g., simulated isoxaflutole concentration increased to 1.7 mug L(-1)). The RZWQM modified to include a macropore express fraction more accurately simulates chemical movement through macropores to subsurface drains. An express fraction is required to match peak concentrations in subsurface drains shortly after chemical applications.

Runoff volume and flow concentration are hydrological factors that limit effectiveness of vegetated filter strips (VFS) in removing pesticides from surface runoff. Empirical equations that predict VFS pesticide effectiveness based solely on physical characteristics are insufficient on the event scale because they do not completely account for hydrological processes. This research investigated the effect of drainage area ratio (i.e., the ratio of field area to VFS area) and flow concentration (i.e., uniform versus concentrated flow) on pesticide removal efficiency of a VFS and used these data to provide further field verification of a recently proposed numerical/empirical modeling procedure for predicting removal efficiency under variable flow conditions. Runoff volumes were used to simulate drainage area ratios of 15:1 and 30:1. Flow concentration was investigated based on size of the VFS by applying artificial runoff to 10% of the plot width (i.e., concentrated flow) or the full plot width (i.e., uniform flow). Artificial runoff was metered into 4.6-m long VFS plots for 90 min after a simulated rainfall of 63 mm applied over 2 h. The artificial runoff contained sediment and was dosed with chlorpyrifos and atrazine. Pesticide removal efficiency of VFS for uniform flow conditions (59% infiltration; 88% sediment removal) was 85% for chlorpyrifos and 62% for atrazine. Flow concentration reduced removal efficiencies regardless of drainage area ratio (i.e., 16% infiltration, 31% sediment removal, 21% chlorpyrifos removal, and 12% atrazine removal). Without calibration, the predictive modeling based on the integrated VFSMOD and empirical hydrologic-based pesticide trapping efficiency equation predicted atrazine and chlorpyrifos removal efficiency under uniform and concentrated flow conditions. Consideration for hydrological processes, as opposed to statistical relationships based on buffer physical characteristics, is required to adequately predict VFS pesticide trapping efficiency.

No abstract