Urban and residential areas are covered by impervious surfaces such as concrete. During rainfalls and irrigations, these surfaces transfer runoff directly to surface water through urban sewage. Impervious surfaces are highly possible to be contaminated with pesticides, through direct treatments or through indirect deposits from surrounding areas (e.g., lawns) via wind- or water-aided movements. Although pesticides have been widely found in urban watersheds, their persistence and runoff potential from impervious surfaces have been seldom studied.

This dissertation includes a series of lab and field scale experiments to assess pesticide contamination on concrete surfaces, as well as different environmental and application parameters that may affect their runoff potential.

The results showed pesticide wash-off potential from concrete rapidly decreased after treatments, but low levels were constantly detected in the water from concrete with extended outdoor exposure. Immediately after treatments with pesticides in liquid formulations, concrete surfaces were equilibrated with 30 mL CaCl₂ solution for 10 min to mimic runoff, and <25% of applied pesticides were recovered. The wash-off percentage further decreased to <5% after 7-d. Higher wash-off potential was observed for solid formulated pesticides. The initial rapid loss of wash-off transferability was followed by a stage of sustained pesticide wash-off, which continued for over 112-d. The persistent wash-off potential was also observed for concrete repeatedly equilibrated with water once every 3-d. Most pesticides were still detectable after 10 times of washing-drying cycles.

To explain the observed wash-off patterns, sorption/desorption and degradation of permethrin on concrete were further studied using ¹⁴C-labeled permethrin. Although the sorption equilibrium was quickly achieved within 24-h, permethrin desorption continued for over 300-h, and around 20 % of applied ¹⁴C still remained on concrete after 300-h. If the desorption started 7-d after permethrin treatments, the amounts of desorbed permethrin decreased by 58 % compared the freshly spiked concrete.

Finally, field study was conducted to assess pesticide runoff following simulated and natural rainfalls. 80 % of runoff pyrethroids were found on suspended particles. Besides, a surface wiping method was developed to collect transferable particles from concrete and its potential to predict pesticide runoff prior to rainfall starts was demonstrated.

Pyrethroid and organophosphate pesticides have been found in numerous waterways throughout the United States. Monitoring studies within the San Diego Creek watershed in Orange County, California, confirmed that runoff from agricultural fields are a primary contributor to pesticide loads within these waterways. As a mitigation option for non-point pesticide pollution, constructed wetlands were investigated in this study to reduce pesticide loading in agricultural tailwaters. Multiple wetlands located along the San Joaquin River in the Stanilaus County were observed over two consecutive irrigation seasons. The wetlands were found to be very effective at reducing pyrethroid concentrations (52 - 100%) and moderately effective at reducing chlorpyrifos concentrations (52 - 61%) under two flow regimes, with loads reduced by 95 - 100% in the outgoing water. Vertical transport of pyrethroids was found to be negligible, and less than 10% of chlorpyrifos was found to leach below 16 cm in the wetland floors. A significant increase in removal efficiencies between seasons within one portion of the wetland was observed, and was attributed to a dramatic increase in vegetation and decreased flow rates. Pyrethroids overwhelmingly sorbed to suspended solids greater than 0.7 µm (62 - 93%), with resulting apparent partitioning coefficients (Kd a) ranging from 1.9 E104 to 3.1 E 105. The freely dissolved concentrations of permethrin represented approximately 26 - 39% of its total mass in water samples with a decreasing trend in concentrations toward the outlet, signaling a decrease in the bioavailable fraction in the outgoing flow. Pesticide concentrations mimicked organic carbon content of the deposited sediment particle fractions. The highest concentrations were associated with larger particles comprised of aggregates of organic and decomposed plant material, which are less susceptible to sedimentation. This observation helps explain why the sediment basin was not effective at removing pesticides from the tailwater. Effective wetland lengths (L1/2) necessary to reduce pesticide concentrations by 50% were estimated to be less than 100 m for cyhalothrin, cypermethrin and permethrin under low flow (0.03 m3s-1), but reaching 267 m under high flow conditions (0.07 m-3 s-1). The degradation studies indicate that the pesticides have the potential for persisting within the wetland system between irrigation seasons. While the half-lives of lambda-cyhalothrin, cypermethrin, esfenvalerate, and permethrin were less than 1 yr (r2 > 0.48) under anaerobic conditions, no detectable dissipation occurred for most pesticides in situ during the dry season. Bifenthrin was found to be relatively stable in all sediments, indicating its potential for prolonged persistence within the wetland systems. The results from this study indicate that constructed wetlands may act as a sink for most pyrethroids, and the removal is achieved through sedimentation and sediment trapping by vegetation and gravity. However, accumulation of pesticides over time, and the associated environmental risks of the accumulated pesticide residues, should be further understood if constructed wetlands are to be widely used as a management practice.