Solid-phase microextraction coupled with gas chromatography-mass spectrometry (SPME-GC-MS) was used to analyze two triazine (atrazine and simazine) and three chloroacetamide herbicides (acetochlor, alachlor, and metolachlor) in water samples from a midwest US agricultural drainage ditch for two growing seasons. The effects of salt concentration, sample volume, extraction time, and injection time on extraction efficiency using a 100-μm polydimethylsiloxane-coated fiber were investigated. By optimizing these parameters, ditch water detection limits of 0.5 μg L-1 simazine and 0.25 μg L-1 atrazine, acetochlor, alachlor, and metolachlor were achieved. The optimum salt concentration was found to be 83% NaCl, while sample volume (10 or 20 mL) negligibly affected analyte peak areas. The optimum extraction time was 40 min, and the optimum injection time was 15 min. Results indicated that atrazine levels in the ditch water exceeded the US maximum contaminant level for drinking water 12% of the time, and atrazine was the most frequently detected among studied analytes. Solid-phase microextraction methods were successfully developed to quantify low levels of herbicides in tile-fed drain water by gas chromatography-mass spectrometry.

In situ denitrification walls and biofilters made of wood chips are being implemented as innovative technologies for the removal of nitrates in tile drainage water from farms to reduce pollution of surface waters and the hypoxia problem in the Gulf of Mexico. Although fairly effective in removing nitrates, not much is known about the effectiveness of the biofilters in removal of herbicides, pesticides, and antibiotics in the drainage water. Using weathered wood chips obtained from an in situ denitrification wall, four common pollutants tested sorbed strongly to wood chips in the following order: enrofloxacin > monensin A > atrazine > sulfamethazine. Of the four chemicals tested, enrofloxacin was found to desorb the least by water extraction. The apparent hysteresis index for atrazine was found to be lower than that for enrofloxacin and sulfamethazine indicating greater sorption–desorption hysteresis for atrazine than enrofloxacin and sulfamethazine. Consecutive steps of water desorption and organic solvent extraction indicated that more than 65% of the sorbed atrazine, 70% of sulfamethazine, 90% of enrofloxacin, and 80% of monensin A were retained in wood chips. Results of this study showed that wood chip denitrification walls or biofilters have an added benefit in retaining herbicides and antibiotics and therefore can act as a barrier to reduce pollution of surface water and groundwater.

The research of the environmental fate of pesticides has demonstrated that applied compounds are altered in their molecular structure over time and are distributed within the environment. To assess the risk for contamination by transformation products (TP) of the herbicides flufenacet and metazachlor, the following four water body types were sampled in a small-scale catchment of 50 km² in 2015/2016: tile drainage water, stream water, shallow groundwater, and drinking water of private wells. The TP were omnipresent in every type of water body, more frequently and in concentrations up to 10 times higher than their parent compounds. Especially metazachlor sulfonic acid, metazachlor oxalic acid, and flufenacet oxalic acid were detected in almost every drainage and stream sample. The transformation process leads to more mobile and more persistent molecules resulting in higher detection frequencies and concentrations, which can even occur a year or more after the application of the parent compound. The vulnerability of shallow groundwater and private drinking water wells to leaching compounds is proved by numerous positives of metazachlor-TP with maximum concentrations of 0.7 μg L⁻¹ (drinking water) and 20 μg L⁻¹ (shallow groundwater) of metazachlor sulfonic acid. Rainfall events during the application period cause high discharge of the parent compound and lower release of TP. Later rainfall events lead to high displacement of TP. For an integrated risk assessment of water bodies, the environmental behavior of pesticide-TP has to be included into regular state-of-the-art water quality monitoring.

Nonpoint source phosphorus (P) and nitrogen (N) pollution from agriculture is a global concern. Planting a cover crop after harvesting annual crops such as maize may help mitigate nutrient transport risk to surface and groundwater. Few studies have focused on the impact of a winter rye cover crop on both surface runoff (SR) and tile drainage (TD) water quality. Here, we measured N and P losses in SR and TD from maize plots grown with and without a winter rye cover crop. Four plots (46 × 23 m) in northern New York, USA, equipped with automated SR and TD flow monitoring were planted with winter rye (Secale cereal) in 2016 and 2017 after maize silage harvest. Plots were managed as typical silage fields for dairy farms in the region and received fertilizer and manure applications. Dissolved reactive P (DRP), total P (TP), nitrate-N, total N (TN), and total suspended solids (TSS) loads were monitored from 4/7/16 to 6/29/17. Cumulative SR (volumetric depth equivalent) was 1.8-fold lower for rye compared to control plots. Although runoff and loading were variable, cumulative TSS, TP, and DRP losses were approximately 3-fold lower for rye plots compared to control. Cumulative TN and nitrate-N loads for TD were similar; however, cumulative TN loss for SR was lower for rye plots. Surface runoff was the main pathway of P loss (> 90% of DRP and TP loss) with > 90% of cumulative P exported from 2017 snowmelt events. Results suggest winter rye mitigated N and P transport risk in SR compared to the common practice of leaving maize silage fields bare after harvest.

Drainage filters using porous granular material constitute new innovative technologies for remediating phosphorus (P) from agricultural tile drainage water. In drainage filters where convective velocities are often high, we hypothesize that intragranular diffusion may affect solute transport depending on filter characteristics and flow rate. This was investigated for six drainage filter materials (Leca, Filtralite-P®, granulated limestone, crushed seashells, calcined diatomite earth (CDE), and a poorly ordered Fe oxide aggregate (CFH)) conducting a tritium (³H₂O) tracer experiment at low (0.26 cm h⁻¹), medium (23 cm h⁻¹), and high (41 cm h⁻¹) flux densities. The filter materials differed widely with respect to grain-size distribution (D ₅₀ from 1.6 to 3.3 mm), uniformity coefficient (1.7 to 2.2), particle density (1.75 to 2.76 g cm⁻³), bulk density (0.34 to 1.46 g cm⁻³), and water-filled porosity (0.39 to 0.73 cm³ cm⁻³). Measurements of specific surface area (SSA) included both SSABET and SSAEGME to ensure inclusion of the intragranular microporosity, not accounted by N₂-BET. SSA varied widely across methods and allowed the differentiation of filters according to the significance of the intragranular porosity. Tritium transport varied from approximately equilibrium transport at all flow rates in Leca, Filtralite-P®, and limestone, to progressive non-equilibrium transport as flow rate increased in Seashells, CDE, and CFH. In general, the filter materials were highly variable in hydro-physical properties. Filters with (approximately) equilibrium transport were, however, all characterized by low specific surface areas. The non-equilibrium transport was explained by an intragranular diffusion in filters with larger specific surface area (Seashells, CDE, and CFH).

Subsurface tile drainage systems with drainspacings of 15 m in 0.4 ha and 25 m in 3.2 ha wereinstalled at the farmers' field in 1986 and 1987,respectively, to study their effect on the reclamationof the coastal saline sodic clay soils. The system'sperformance in terms of the changing physical andchemical properties of the soil and rice yield wascontinuously monitored for a decade. Field datasuggested the possibility of adopting wider drainspacings and thus, drainage system with 35 and 55 mspacings was laid in 1997 in a 4 ha area. On theseinstallations the losses of NH₄ ⁺-N throughsub-surface drainage effluent were estimated. Thearea under 25 m drain spacing was the control with nocrops, fertilization and irrigation. Analysis ofwater samples collected daily for 10 days startingfrom 40 DAT from the drain laterals revealed thatthere were no trace of NH₄ ⁺-N in theeffluent from 15 and 25 m drain spacings. However,the effluent from 35 and 55 m spacings contained anaverage of 6.704 mg L⁻¹ and 4.205 mg L⁻¹ of NH₄ ⁺-N, respectively, before irrigation and2.438 and 1.650 mg L⁻¹ after irrigation. Themagnitudes of the losses of NH₄ ⁺-N duringthe crop season were 6.43 kg ha⁻¹ in 35 m spacingwith a drainage rate of 5.6 mm d⁻¹ and 2.14 kgha⁻¹ in 55 m spacing with a drainage rate of 3.5 mm d⁻¹. The rice yield was 6.5 Mg ha⁻¹ in15 m drain spacing where no ammonium losses throughsubsurface drainage effluent occurred. The rice yieldsunder 35 and 55 m drain spacings were 1.9 and 1.8 Mgha⁻¹, respectively. The poor yield was due tosignificant loss of ammonium form of nitrogen throughthe drainage effluent and lesser availability of totalnitrogen to the plants. The plant uptake of nitrogen in the unreclaimed area with 55 m spacing was half ofthat in the reclaimed area with 15 m spacing.

Constructed wetlands (CWs) are used to reduce the pesticide inputs from tile drainage or run-off to surface water. Their effectiveness appears variable and remains to be better characterized and understood. The aim of this study was to assess the influences of two hydraulic parameters (i.e., dynamics and water level) on the sorption process occurring in CWs. Then, two solid/liquid ratios were studied (1/1 and 1/5) to mimic the water level variation in the field, and two agitation speeds were used (none and gentle agitation) to simulate different water dynamics (stagnation and flow pass, respectively). Sorption kinetics and isotherms were obtained for four pesticides with contrasting properties. The pesticide adsorption coefficients were classified as follows: boscalid (BSC) > cyproconazole (CYP) > isoproturon (IPU) ∼ dimethachlor (DMT) at any ratio or agitation, in agreement with their water solubilities and K ₒw values. The effect of the solid/liquid ratio was evidenced for all conditions. Indeed, the adsorption equilibrium time was reached more quickly for the 1/1 ratio (24–72 h) than for the 1/5 ratio (96–120 h). In addition, the adsorption coefficients (K f ᵃᵈˢ) were larger for the 1/1 ratio (1.8–11.2 L kg⁻¹) than for the 1/5 ratio (1.0–5.9 L kg⁻¹). The agitation effect was more evidenced for the 1/5 ratio and for the more hydrophobic molecules, such as BSC and CYP, for which adsorption equilibrium time was never reached with agitation (>120 h), while it was reached at 96 h without agitation. Moreover, the K f ᵃᵈˢ values were larger with agitation than without agitation for BSC and CYP, whereas they were similar for the two agitations for IPU and DMT. Our results demonstrated that the hydrodynamic function of CWs could influence pesticide sorption with variable effects according to the molecular properties and consequently influence the mitigation effect of CWs throughout the year.

Modeling (MONERIS) studies allowed calculation of nitrogen (N) and phosphorus (P) emission into the Vistula and Oder basins (Poland), and facilitated estimation of N and P retention in these catchments in 1995–2015. In the discussion of results, data of other authors were used in order to get an insight into N (1880–2015) and P emission (1955–2015) into the Oder basin. Population growth and agricultural intensification were responsible for respective 5.3-fold and 3.5-fold increase in N and P emission into the Oder basin, with the maximum (135,000 tons N year⁻¹; 14,000 tons P year⁻¹) observed at the turn of the 1980s/1990s. Pro-ecological activities during the economic transition period (since 1989) covered various sectors of the economy including agriculture, environmental protection related to, e.g., construction of a large number of waste water treatment plants (WWTPs). Consequently, in 1985–2015, the emission into the Oder basin decreased from the abovementioned maxima to 94,000 tons N year⁻¹ and to 5000 tons P year⁻¹, whereas in 1995–2015, the emission into the Vistula basin decreased from 170,000 to 140,000 tons N year⁻¹ and from 14,200 to 10,600 tons P year⁻¹. In 1995–2015, groundwater, tile drainage, and WWTPs played a key role in N emission, while erosion, overland flow, WWTPs, and urban areas played a predominant role in P emission. The relative shares of nutrient emission pathways in overall N and P emission were considerably changing over time. Extreme weather conditions have a great impact on increased (floods) or decreased (droughts) nutrient emission; particularly, N emission is susceptible to variable weather conditions. In total, approximately 91,000 tons of N and 7600 tons of P were retained annually in the river basins.