Nitrous oxide and methane production and consumption at five full-size denitrifying bioreactors treating subsurface drainage water.

Nitrate (NO3-) removal in denitrifying bioreactors is influenced by flow, water chemistry, and design, but it is not known how these widely varying factors impact the production of nitrous oxide (N2O) or methane (CH4) across sites. Woodchip bioreactors link the hydrosphere and atmosphere in this respect, so five full-size bioreactors in Illinois, USA, were monitored for NO3-, N2O, and CH4 to better document where this water treatment technology resides along the pollution swapping to climate smart spectrum. Both surface fluxes and dissolved forms of N2O and CH4 were measured (n = 7-11 sampling campaigns per site) at bioreactors ranging from <1 to nearly 5 years old and treating subsurface drainage areas from between 6.9 and 29 ha. Across all sites, N2O surface and dissolved volumetric production rates averaged 1.0 ± 1.6 mg N2O-N/m3-d and 24 ± 62 mg dN2O-N/m3-d, respectively, and CH4 production rates averaged 6.0 ± 26 mg CH4-C/m3-d and 310 ± 520 mg dCH4-C/m3-d for surface and dissolved, respectively. However, N2O was consistently consumed at one bioreactor, and only three of the five sites produced notable CH4. Surface fluxes of CH4 were significantly reduced by the presence of a soil cover. Bioreactor denitrification was relatively efficient, with only 0.51 ± 3.5 % of removed nitrate emitted as N2O (n = 48). Modeled indirect N2O emissions factors were significantly lower when a bioreactor was present versus absent (EF5: 0.0055 versus 0.0062 kg N2O-N/kg NO3-N; p = 0.0011). While further greenhouse gas research on bioreactors is recommended, this should not be used as an excuse to slow adoption efforts. Bioreactors provide a practical option for voluntary water quality improvement in the heavily tile-drained US Midwest and elsewhere.

Paired denitrifying bioreactors with wide orientation for increased drainage flow capacity.

Denitrifying bioreactors are a conservation drainage practice for reducing nitrate loads in subsurface agricultural drainage. Bioreactor hydraulic capacity is limited by cross-sectional area perpendicular to flow through the woodchip bed, with excess bypass flow untreated. Paired bioreactors with wide orientations were built in 2017 in Illinois, USA, to treat drainage from a relatively large 29 ha field. The paired design consisted of: a larger, Main bioreactor (LWD: 6.1 × 18.3 × 0.9 m) for treating base flow, and 2) a smaller, Booster bioreactor (7.8 × 13.1 × 0.9 m) receiving bypass flow from the Main bioreactor during periods of high flow. Over three years of monitoring, the paired bioreactor captured 84-92% of the annual drainage discharge which demonstrated an expanded cross-sectional area could improve bioreactor flow capture, even for a large drainage area. However, the paired bioreactors removed 6-28% of the annual N load leaving the field (1.8-5.6 kg N ha-1 removed; 52-161 kg N), which was not a notable improvement compared to bioreactors treating smaller drainage areas. The design operated as intended at low annual flow-weighted hydraulic retention times (HRTs) of usually ≤2 h, but these short HRTs ultimately limited bioreactor nitrate removal efficiency. Daily HRTs of <2 h often resulted in nitrate flushing. The Main bioreactor had higher hydraulic loading as intended and was responsible for the majority of flow captured in each year although not always the most nitrate mass removal. The Booster bioreactor provided better nitrate removal than the Main at HRTs of 3.0-11.9 h, possibly due to its drying cycles which may have liberated more available carbon. This new design approach tested at the field-scale illustrated tradeoffs between greater flow capacity (via increased bioreactor width) and longer HRT (via increased length), given a consistent bioreactor surface footprint.

Phosphorus removal in denitrifying woodchip bioreactors varies by wood type and water chemistry.

Denitrifying woodchip bioreactors are a practical nitrogen (N) mitigation technology but evaluating the potential for bioreactor phosphorus (P) removal is highly relevant given that (1) agricultural runoff often contains N and P, (2) very low P concentrations cause eutrophication, and (3) there are few options for removing dissolved P once it is in runoff. A series of batch tests evaluated P removal by woodchips that naturally contained a range of metals known to sorb P and then three design and environmental factors (water matrix, particle size, initial dissolved reactive phosphorus (DRP) concentration). Woodchips with the highest aluminum and iron content provided the most dissolved P removal (13±2.5 mg DRP removed/kg woodchip). However, poplar woodchips, which had low metals content, provided the second highest removal (12±0.4 mg/kg) when they were tested with P-dosed river water which had a relatively complex water matrix. Chemical P sorption due to woodchip elements may be possible, but it is likely one of a variety of P removal mechanisms in real-world bioreactor settings. Scaling the results indicated bioreactors could remove 0.40 to 13 g DRP/ha. Woodchip bioreactor dissolved P removal will likely be small in magnitude, but any such contribution is an added-value benefit of this denitrifying technology.

Efficacy of heated tourmaline in reducing biomass clogging within woodchip bioreactors.

Woodchip bioreactors can effectively remove waterborne nitrates from subsurface agricultural drainage and prevent the eutrophication of receiving water, but rapid biofilm growth can severely reduce water flux and denitrification efficiency of this practice within a few years. Tourmaline minerals with thermal excitation could generate reactive oxygen species which would inhibit bacterial growth. In this study, laboratory scale woodchip bioreactors were set up to test the anti-clogging and denitrification efficiency of heated woodchips with tourmaline, heated woodchips without tourmaline, and unheated woodchips. The results showed that the heated tourmaline treatment could reduce the clogging and optimize the nitrate removal rate (47.6 g N/m3/day) under all three hydrologic retention times tested (1, 4, and 8 h). Dissolved oxygen and pH values fluctuated with the removal rate and temperature change, while temperature was identified as the key factor impacting the tourmaline treatment. The heated tourmaline treatment had the lowest biofilm growth (lowest DNA concentration), while the 16S rRNA and a higher abundance of nirS-, nirK-, and nosZ-encoding denitrifying bacteria (based on qPCR) confirmed the higher denitrification efficiency of the heated tourmaline treatment.

Relative performance of different data mining techniques for nitrate concentration and load estimation in different type of watersheds.

The increasing availability of water quality datasets has led to a greater focus on hydrologic and water quality analysis, thus requiring more efficient and accurate modelling methods. Data mining techniques have been increasingly used for water quality analysis and prediction of the concentration and load of nitrogen pollutants instead of more traditional simulation methods. In this study, we tested the multilayer perceptron (MLP), k-nearest neighbor (k-NN), random forest, and reduced error pruning tree (REPTree) methods, along with the traditional linear regression, to predict nitrate levels based on long-term data from six watersheds with different land-use practices in the midwestern United States. Both the concentration and load results indicated that REPTree had the best performance, with an R2 of 0.61-0.85 and a relative absolute error of <75.8%. The different watershed types, however, influenced the performance of the data mining methods, where all four methods showed a higher accuracy for urban dominant watershed and lower accuracy for agricultural and forest watersheds. Out of these four methods, classification tree methods (REPTree and RF) performed better than cluster methods (MLP and k-NN) for agricultural and forested watersheds. Our results indicated that both the data structure based on the dominant land use and type of algorithmic method should be carefully considered for selecting a data mining method to predict nitrate concentration and load for a watershed.

Evaluation of fly ash pellets for phosphorus removal in a laboratory scale denitrifying bioreactor.

Nitrate and orthophosphate from agricultural activities contribute significantly to nutrient loading in surface water bodies around the world. This study evaluated the efficacy of woodchips and fly ash pellets in tandem to remove nitrate and orthophosphate from simulated agricultural runoff in flow-through tests. The fly ash pellets had previously been developed specifically for orthophosphate removal for this type of application, and the sorption bench testing showed a good promise for flow-through testing. The lab-scale horizontal-flow bioreactor used in this study consisted of an upstream column filled with woodchips followed by a downstream column filled with fly ash pellets (3 and 1 m lengths, respectively; both 0.15 m diameter). Using influent concentrations of 12 mg/L nitrate and 5 mg/L orthophosphate, the woodchip bioreactor section was able to remove 49-85% of the nitrate concentration at three hydraulic retention times ranging from 0.67 to 4.0 h. The nitrate removal rate for woodchips ranged from 40 to 49 g N/m3/d. Higher hydraulic retention times (i.e., smaller flow rates) corresponded with greater nitrate load reduction. The fly ash pellets showed relatively stable removal efficiency of 68-75% across all retention times. Total orthophosphate adsorption by the pellets was 0.059-0.114 mg P/g which was far less than the saturated capacity (1.69 mg/g; based on previous work). The fly ash pellets also removed some nitrate and the woodchips also removed some orthophosphate, but these reductions were not significant. Overall, woodchip denitrification followed by fly ash pellet P-sorption can be an effective treatment technology for nitrate and phosphate removal in subsurface drainage.

Characterization of fly ash ceramic pellet for phosphorus removal.

Phosphorus has been recognized as a leading pollutant for surface water quality deterioration. In the Midwestern USA, subsurface drainage not only provides a pathway for excess water to leave the field but it also drains out nutrients like nitrogen (N) and phosphorus (P). Fly ash has been identified as one of the viable materials for phosphorus removal from contaminated waters. In this study, a ceramic pellet was manufactured using fly ash for P absorption. Three types of pellet with varying lime and clay proportions by weight (type 1: 10% lime + 30% clay, type 2: 20% lime + 20% clay, and type 3: 30% lime + 10% clay) were characterized and evaluated for absorption efficiency. The result showed that type 3 pellet (60% fly ash with 30% lime and 10% clay) had the highest porosity (14%) and absorption efficiency and saturated absorption capacity (1.98 mg P/g pellet) compared to type 1 and 2 pellets. The heavy metal leaching was the least (30 µg/L of chromium after 5 h) for type 3 pellet compared to other two. The microcosmic structure of pellet from scanning electron microscope showed the type 3 pellet had the better distribution of aluminum and iron oxide on the surface compared other two pellets. This result indicates that addition of lime and clay can improve P absorption capacity of fly ash while reducing the potential to reduce chromium leaching.

Temperature and Substrate Control Woodchip Bioreactor Performance in Reducing Tile Nitrate Loads in East-Central Illinois.

Tile drainage is the major source of nitrate in the upper Midwest, and end-of-tile removal techniques such as wood chip bioreactors have been installed that allow current farming practices to continue, with nitrate removed through denitrification. There have been few multiyear studies of bioreactors examining controls on nitrate removal rates. We evaluated the nitrate removal performance of two wood chip bioreactors during the first 3 yr of operation and examined the major factors that regulated nitrate removal. Bioreactor 2 was subject to river flooding, and performance was not assessed. Bioreactor 1 had average monthly nitrate removal rates of 23 to 44 g N m d in Year 1, which decreased to 1.2 to 11 g N m d in Years 2 and 3. The greater N removal rates in Year 1 and early in Year 2 were likely due to highly degradable C in the woodchips. Only late in Year 2 and in Year 3 was there a strong temperature response in the nitrate removal rate. Less than 1% of the nitrate removed was emitted as NO. Due to large tile inputs of nitrate (729-2127 kg N) at high concentrations (∼30 mg nitrate N L) in Years 2 and 3, overall removal efficiency was low (3 and 7% in Years 2 and 3, respectively). Based on a process-based bioreactor performance model, Bioreactor 1 would have needed to be 9 times as large as the current system to remove 50% of the nitrate load from this 20-ha field.

Characterizing the Performance of Denitrifying Bioreactors during Simulated Subsurface Drainage Events.

The need to mitigate nitrate export from corn and soybean fields with subsurface (tile) drainage systems, a major environmental issue in the midwestern United States, has made the efficacy of field-edge, subsurface bioreactors an active subject of research. This study of three such bioreactors located on the University of Illinois South Farms during their first 6 mo of operation (July-Dec. 2012) focused on the interactions of seasonal temperature changes and hydraulic retention times (HRTs), which were subject to experimental manipulation. Changes in nitrate, phosphate, oxygen, and dissolved organic carbon were monitored in influent and effluent to assess the benefits and the potential harmful effects of bioreactors for nearby aquatic ecosystems. On average, bioreactors reduced nitrate loads by 63%, with minimum and maximum reductions of 20 and 98% at low and high HRTs, respectively. The removal rate per unit reactor volume averaged 11.6 g NO-N m d (range, 5-30 g NO-N m d). Multiple regression models with exponential dependencies on influent water temperature and on HRT explained 73% of the variance in NO-N load reduction and 43% of the variance in its removal rate. Although concentrations of dissolved reactive phosphorus and dissolved organic carbon in the bioreactor effluent increased relative to the influent by an order of magnitude during initial tests, within 1 mo of operation they stabilized at nearly equal values.

Navigating the socio-bio-geo-chemistry and engineering of nitrogen management in two illinois tile-drained watersheds.

Reducing nitrate loads from corn and soybean, tile-drained, agricultural production systems in the Upper Mississippi River basin is a major challenge that has not been met. We evaluated a range of possible management practices from biophysical and social science perspectives that could reduce nitrate losses from tile-drained fields in the Upper Salt Fork and Embarras River watersheds of east-central Illinois. Long-term water quality monitoring on these watersheds showed that nitrate losses averaged 30.6 and 23.0 kg nitrate N ha yr (Embarras and Upper Salt Fork watersheds, respectively), with maximum nitrate concentrations between 14 and 18 mg N L. With a series of on-farm studies, we conducted tile monitoring to evaluate several possible nitrate reduction conservation practices. Fertilizer timing and cover crops reduced nitrate losses (30% reduction in a year with large nitrate losses), whereas drainage water management on one tile system demonstrated the problems with possible retrofit designs (water flowed laterally from the drainage water management tile to the free drainage system nearby). Tile woodchip bioreactors had good nitrate removal in 2012 (80% nitrate reduction), and wetlands had previously been shown to remove nitrate (45% reductions) in the Embarras watershed. Interviews and surveys indicated strong environmental concern and stewardship ethics among landowners and farmers, but the many financial and operational constraints that they operate under limited their willingness to adopt conservation practices that targeted nitrate reduction. Under the policy and production systems currently in place, large-scale reductions in nitrate losses from watersheds such as these in east-central Illinois will be difficult.

Spatial variation in the bacterial and denitrifying bacterial community in a biofilter treating subsurface agricultural drainage.

Denitrifying biofilters can remove agricultural nitrates from subsurface drainage, reducing nitrate pollution that contributes to coastal hypoxic zones. The performance and reliability of natural and engineered systems dependent upon microbially mediated processes, such as the denitrifying biofilters, can be affected by the spatial structure of their microbial communities. Furthermore, our understanding of the relationship between microbial community composition and function is influenced by the spatial distribution of samples.In this study we characterized the spatial structure of bacterial communities in a denitrifying biofilter in central Illinois. Bacterial communities were assessed using automated ribosomal intergenic spacer analysis for bacteria and terminal restriction fragment length polymorphism of nosZ for denitrifying bacteria.Non-metric multidimensional scaling and analysis of similarity (ANOSIM) analyses indicated that bacteria showed statistically significant spatial structure by depth and transect,while denitrifying bacteria did not exhibit significant spatial structure. For determination of spatial patterns, we developed a package of automated functions for the R statistical environment that allows directional analysis of microbial community composition data using either ANOSIM or Mantel statistics.Applying this package to the biofilter data, the flow path correlation range for the bacterial community was 6.4 m at the shallower, periodically in undated depth and 10.7 m at the deeper, continually submerged depth. These spatial structures suggest a strong influence of hydrology on the microbial community composition in these denitrifying biofilters. Understanding such spatial structure can also guide optimal sample collection strategies for microbial community analyses.

Runoff losses of suspended sediment, nitrogen, and phosphorus from a small watershed in Korea.

Nutrients and sediments in runoff lead to the degradation of water quality of lakes and streams. The development of schemes to mitigate such degradation requires a characterization of the underlying transport processes. The objectives of this study were to develop annual and seasonal load-discharge relationships for suspended sediment (SS), total nitrogen (TN), and total phosphorus (TP) losses from a small mixed land use watershed and to use these relationships to explicate the annual and monthly patterns of losses of these species. Data from 1996 to 2004 were used to develop load-discharge relationships for SS, TN, and TP at the HP#6 watershed, a subwatershed of the Balhan reservoir watershed located in Bongdam-myun and Paltan-myun, Gyeonggi-do, Korea. Standard least squares curve fitting and S-estimation procedures were used to fit power functions to the data collected over this time period. The fitted load-discharge relationships are indicative of seasonal variations in SS and TN and of TP losses from HP#6. The exponents of the fitted power functions for TN and TP in the fall, for TP in summer season, and for SS in all seasons are >1, indicating that the concentrations of these species increase as flow rate increases. Most of the SS, TN, and TP transported in runoff left the watershed between April and September; thus, cost-efficient strategies can be established by focusing on this period. Further study of the seasonal variations is required for a better characterization of seasonal losses of SS, TN, and TP in runoff from the HP#6 watershed.

Stream transport of herbicides and metabolites in a tile-drained, agricultural watershed.

The occurrence of metabolites of many commonly used herbicides in streams has not been studied extensively in tile-drained watersheds. We collected water samples throughout the Upper Embarras River watershed [92% corn, Zea mays L., and soybean, Glycine max (L.) Merr.] in east-central Illinois from March 1999 through September 2000 to study the occurrence of atrazine (2-chloro-4-ethylamino-6-isopropylamino-s-triazine), metolachlor 12-chloro-N-(2-ethyl-6-methylphenyl)-N-(methoxy-1-methylethyl) acetamide], alachlor [2-chloro-N-(2,6-diethylphenyl)-N-(methoxymethyl) acetamide], acetochlor [2-chloro-N-(ethoxymethyl)-N-(2-ethyl-6-methylphenyl) acetamide], and their metabolites. River water samples were collected from three subwatersheds of varying tile density (2.8-5.3 km tile km(-2)) and from the outlet (United States Geological Survey [USGS] gage site). Near-record-low totals for stream flow occurred during the study, and nearly all flow was from tiles. Concentrations of atrazine at the USGS gage site peaked at 15 and 17 microg L(-1) in 1999 and 2000, respectively, and metolachlor at 2.7 and 3.2 microg L(-1); this was during the first significant flow event following herbicide applications. Metabolites of the chloroacetanilide herbicides were detected more often than the parent compounds (evaluated during May to July each year, when tiles were flowing), with metolachlor ethanesulfonic acid [2-[(2-ethyl-6-methylphenyl)(2-methoxy-1-methylethyl)amino]-2-oxoethanesulfonic acid] detected most often (> 90% from all sites), and metolachlor oxanilic acid [2-[(2-ethyl-6-methylphenyl)(2-methoxy-1-methylethyl)amino]-2-oxoacetic acid] second (40-100% of samples at the four sites). When summed, the median concentration of the three chloroacetanilide parent compounds (acetochlor, alachlor, and metolachlor) at the USGS gage site was 3.4 microg L(-1), whereas it was 4.3 microg L(-1) for the six metabolites. These data confirm the importance of studying chloroacetanilide metabolites, along with parent compounds, in tile-drained watersheds.