Predicting Preferential Flow and Mitigating Agrochemicals in the Vadose Zone

Abstract

Nutrients and herbicides leach through drainage tiles to water bodies and degrade their quality. Denitrifying bioreactors reduce the nitrate from these tiles. In this dissertation, the removal of nitrate and atrazine by the denitrifying bioreactors is explored. Additionally, the preferential transport of herbicides in structured soils is examined and simulated. In a 3-year field study, nitrate removal was examined in six bioreactors at three field locations. On average, bioreactors removed 50% of the nitrate. In each site, there was a critical hydraulic retention time (HRT), above which nitrate was removed entirely. Below this critical HRT, there was a linear relationship between nitrate removal and HRT, which varied with temperature and site locations. Next, removal of atrazine by the woodchip (W) and woodchip amended with 50% biochar (WB) was studied in laboratory bioreactors in four HRTs. The first-order rate of the removal of atrazine for W and WB bioreactors were 0.007/h and 0.024/h, respectively. On average, biochar amendment increased atrazine removal by 40%. We determined that removal of atrazine was abiotic and hydroxyatrazine was the main degradation product in the anaerobic bioreactors. In addition, we investigated the effect of biochar on the removal of nitrate by the woodchip bioreactors with 0%, 12.5%, 25%, and 50% biochar amendment. We measured nitrate, nitrite, carbon dioxide, and nitrous oxide and found that biochar, exposed to atmospheric oxygen, reduced nitrate removal by 9 to 13% while increasing carbon dioxide production. We, thus, concluded that biochar acted as an electron acceptor. Lastly, we tested a parameter efficient preferential flow model (PFM) to estimate the breakthrough of atrazine and 2,4-D in structured soils and compared it with field measurements from multiple samplers. The PFM divides soil profile to the top distribution zone which delivered the chemicals to the conveyance zone below, where several flow paths exist. Using chloride tracer, we estimated hydrological parameters of PEM. Then, by including adsorption and degradation rates, PFM successfully simulated the preferential flow of herbicides with R2 varying from 0.59 to 0.99.