Published online 5 July 2005
Published in J Environ Qual 34:1270-1276 (2005)
DOI: 10.2134/jeq2005.0008
© 2005 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
TECHNICAL REPORTS
Bioremediation and Biodegradation
Maximum Rates of Nitrate Removal in a Denitrification Wall
Louis A. Schippera,b,*,
Gregory F. Barklec and
Maja Vojvodic-Vukovica
a Landcare Research NZ Ltd, Private Bag 3127, Hamilton, New Zealand
b Now at Department of Earth Science, University of Waikato, Hamilton, New Zealand
c Aqualinc Research Ltd, Private Bag 14041 Enderley, Hamilton, New Zealand
* Corresponding author (L.schipper{at}Waikato.ac.nz)
Received for publication January 11, 2005.
 |
ABSTRACT
|
|---|
Denitrification walls are constructed by mixing a carbon source such as sawdust into soils through which ground water passes. These systems can reduce nitrate inputs to receiving waters by enhancing denitrification. Maximum rates of nitrate removal by denitrification need to be determined for design purposes. To determine maximum rates of nitrate removal we added excess nitrate (50 mg N L–1) to a trench up-gradient of a denitrification wall during a 9-d trial. Bromide (100 g L–1) was also added as a conservative tracer. Movement of nitrate and bromide was measured from shallow wells and soil samples were removed for measurements of denitrification, carbon availability, nitrate, and other microbial parameters. Rates of nitrate removal, determined from the ratio of NO3–N to Br and ground water flow, averaged 1.4 g N m–3 of wall d–1 and were markedly greater than denitrification rates determined using the acetylene block technique (average: 0.11 g N m–3 of wall d–1). These nitrate removal rates were generally lower than reported in other denitrification walls. Denitrification rates increased when nitrate was added to the laboratory incubations, indicating that despite large nitrate inputs in the field, denitrification remained limited by nitrate. This limitation was partially attributed to nitrate predominantly moving through zones of greater hydraulic conductivity or in the mobile fraction of the ground water and slow diffusion to the immobile fraction where denitrifiers were active.
|
INTRODUCTION
|
|---|
DENITRIFICATION WALLS have been tested as a means to remove nitrate from shallow ground water (Robertson et al., 2000; Schipper and Vojvodic-Vukovic, 2001). In these systems, particulate organic matter is incorporated into subsoil below the water table to enhance denitrification. Organic matter serves two purposes: first to reduce the oxygen concentration of the ground water by stimulating aerobic respiration, and second to provide a carbon source to denitrifying bacteria. Oxygen removal is necessary because denitrifying bacteria use oxygen as a terminal electron acceptor in preference to nitrate. Denitrifying walls have proved successful for nitrate removal in a wide range of locations including Canada (Robertson and Cherry, 1995; Robertson et al., 2000), New Zealand (Schipper and Vojvodic-Vukovic, 1998, 2000, 2001), and Australia (Fahrner, 2002). However, not all denitrification walls successfully remove nitrate from ground water. Schipper et al. (2004) showed that the majority of ground water passed underneath a denitrification wall installed in coarse sands with thin layers of silt and clay. We showed that mixing of aquifer material and sawdust under saturated conditions led to a twofold decrease in hydraulic conductivity.
Reported rates of nitrate removal for these different denitrification walls have varied widely due to large differences in nitrate inputs, and the variety and proportion of incorporated carbon substrates (Table 1). While a number of other laboratory studies have reported rates of nitrate removal for different carbon sources, these studies are often relatively short-term experiments under artificial conditions and so cannot be easily applied at the field scale.
For denitrification walls to be more widely adopted, the designers of such systems require information on upper rates of nitrate removal for a wide range of ground water flow rates and nitrate loads. Schipper and Vojvodic-Vukovic (2000) demonstrated that denitrification rates were sufficient to account for nitrate removal in a denitrification wall in New Zealand. However, nitrate concentrations in this wall were always close to zero due to complete consumption of nitrate, so denitrification was primarily limited by nitrate concentration (Schipper and Vojvodic-Vukovic, 1998). Consequently, the upper rates of nitrate removal by denitrification could not be determined.
Our objective was to determine upper rates of denitrification in an existing denitrification wall where nitrate concentration was not limiting. Over a 9-d period we continuously dosed the wall with a high concentration of dissolved nitrate and bromide. Bromide was used as a conservative tracer for nitrate. We measured denitrification rates and denitrifying enzyme activity (as a surrogate for the size of denitrifying population) in soil samples taken from within the wall below the water table. Nitrate removal rates from ground water were also estimated from the change in the ratio of nitrate to bromide concentrations and ground water flow rates. To determine factors controlling denitrification, we measured changes in microbial biomass, available carbon, and nitrate concentrations in well water and soil water samples.
|
MATERIALS AND METHODS
|
|---|
Study Site
The study was conducted at the Bardowie dairy farm (37°57' S, 175°27' E), Cambridge, North Island, New Zealand. The pastures on the farm are spray-irrigated with effluent from the nearby Hautapu Dairy Factory. The average nitrogen loading in 1999 was 700 kg N ha–1 yr–1 but has more recently declined to less than 500 kg N ha–1 yr–1. Schipper and Vojvodic-Vukovic (1998) described construction of this denitrification wall. Briefly, the denitrification wall was constructed in January 1996 by digging a trench (about 35 m long, 1.5 m deep, and 1.5 m wide) that intercepted shallow ground water. The soil excavated from the trench was mixed with 40 m3 of sawdust (Monterey pine, Pinus radiata D. Don) then returned to the trench. This shallow ground water had nitrate N concentrations ranging from 5 to 15 mg N L–1.
Nitrate and Bromide Dosing
The field trial was conducted for 9 d in September 2002 when the water table was 0.4 m below the soil surface. To increase the concentrations of nitrate in the denitrification wall we used holding tanks with a dosing system that continuously added dissolved nitrate and bromide into a shallow up-gradient trench to supplement the incoming ground water. Bromide was added as a conservative tracer. Before adding nitrate and bromide into the denitrification wall we conducted a preliminary tracer test to determine the direction and velocity of ground water flow. Rhodamine-WT dye was added to the dosing system and piezometers monitored for Rhodamine-WT color.
The dosing system consisted of three large tanks with a total volume of 7000 L. Potassium nitrate and potassium bromide was dissolved into the water to give concentrations of 50 mg N L–1 and 100 mg Br L–1. This solution was irrigated into a small dosing trench (approximately 3.4 m long by 0.4 m deep and 0.25 m wide) that had been dug at the up-gradient edge of the denitrification wall. The tracer solution was added through variable flow irrigation drippers into the trench at an average flow rate of 50 L h–1. In total, approximately 11400 L (570 g of NO3–N and 1140 g of Br) was applied over the 9 d.
Soil and Water Sampling and Analysis
To monitor transport of bromide and nitrate through the wall, we installed four wells to 1.5 m depth down-gradient of the dosing trench. Three wells (1, 2, and 3) were installed in a parallel row 0.8 m from the trench edge, and 1.0 m apart. A further well (4) was installed at 1.9 m down-gradient from the trench. The entire length of the wells was screened. These wells were sampled daily after purging 2 well volumes using a small-diameter tube and flushed syringe (60 mL). The ground water samples were collected at the same time soil samples were taken. Ground water samples were stored in plastic bottles and immediately returned to the laboratory (less than 30-min transit time) where they were stored at 4°C until analysis for nitrate and bromide. Nitrate was analyzed by standard autoanalyzer methods (Blakemore et al., 1987) and bromide analyzed by ion selective electrode (Metrohm, Herisau, Switzerland) calibrated with standards covering the range of measurements (Abdalla and Lear, 1975).
Daily, three soil samples were taken from both 0 to 20 cm and 20 to 40 cm below the water table using a dutch auger. Samples were taken between 30 and 60 cm away from the dosing trench (i.e., between Wells 1 to 3 and Well 4). A subsample was immediately transferred to a pre-weighed plastic bottle containing 100 mL of KCl (2 M) and shaken. In the laboratory, the bottle was reweighed to obtain wet weight of soil, and the KCl extract filtered (Advantec 5C; Advantec MFS, Dublin, CA). The filtrate was analyzed for nitrate and ammonium by standard autoanalyzer techniques (Blakemore et al., 1987). Nitrate and ammonium concentration in the soil water was calculated using soil wet-weight and its soil moisture content determined following oven drying (24 h at 105°C).
Soil samples were analyzed for denitrification rates (with and without nitrate and glucose amendments) and denitrifying enzyme activity (DEA) on the day of collection. The remaining soil was stored at 4°C for subsequent analysis for microbial biomass and available carbon. Full methods are given below.
Denitrification rates in the wall were determined as previously described by Schipper and Vojvodic-Vukovic (2000). Soil samples (10 g fresh weight) were statically incubated in airtight bottles (100 mL) at 12°C (the same temperature as measured at 1 m in the wall during the experiment). The headspace was replaced with N2 gas and then acetylene (10 mL) added. Headspace samples were taken after 1 and 3 h and analyzed for N2O as described below. To determine whether denitrification was nitrate limited a further subsample was prepared as described above and nitrate (20 mL at 50 mg N L–1) was added. From the fifth day after the dosing experiment, subsamples were also incubated as above, with only glucose (200 mg C L–1) added to determine whether denitrification had become carbon limited.
The DEA was measured using a modified method of Tiedje et al. (1989). This method is a surrogate measure of the size of the population of denitrifying bacteria. Soil (10 g fresh weight) was weighed into airtight bottles (100 mL) and amended with a 20-mL solution containing glucose (0.2 g L–1), potassium nitrate (0.1 g N L–1), and chloramphenicol (0.12 g L–1). The headspace was flushed with nitrogen gas, and 10 mL acetylene added to inhibit reduction of N2O to N2. Bottles were incubated with shaking (200 rpm) at 28°C. After 15 and 75 min, headspace samples were removed and analyzed for N2O using a gas chromatograph (Model PU4410; Philips, Eindhoven, the Netherlands) equipped with an electron capture detector. Operating conditions were: an oven temperature of 60°C, injector temperature of 160°C, and detector temperature of 350°C. The carrier gas was 10% methane in argon at a flow rate of 40 mL min–1 through a 4-m column of Poropak Q.
To determine available carbon, soil (10 g fresh weight) was weighed into bottles (610 mL) and statically incubated at 25°C for 7 d, then a headspace sample was taken and analyzed for CO2 using an infrared gas analyzer (Series 225; Analytical Development Ltd Corporation, Hoddesdon, England). Microbial biomass carbon was determined by chloroform fumigation of soil (25 g fresh weight) followed by extraction with 100 mL of 0.5 M K2SO4 (Vance et al., 1987). Total organic carbon (TOC) in the K2SO4 extracts was determined using a total organic carbon analyzer (Model TOC-5000; Shimadzu, Kyoto, Japan). Nonfumigated controls were similarly extracted and analyzed for TOC. A kEC factor of 0.41 (the proportion of microbial biomass carbon that is extractable following fumigation) was used to convert the TOC flush to microbial biomass.
Biochemical activity and biomass were expressed on a volumetric basis using a dry bulk density of 0.87 Mg m–3 (Schipper and Vojvodic-Vukovic, 2000) after adjusting for water content.
Statistical Analysis
Relationships between nitrate and measures of denitrification (with and without nitrate amendment) and DEA were tested using linear regression (Genstat 6.0; VSN International, 2002). Where appropriate, values were log transformed. Regressions were performed using data from individual soil samples and also using the average daily values.
|
RESULTS
|
|---|
Tracers in Soil and Ground Water
Dosing of nitrate and bromide into the wall increased concentrations of nitrate and bromide in well water and increased nitrate in soil water (Fig. 1 and 2)
. Bromide concentrations were not measured in soil water due to interference of the ion selective electrode by the KCl (2 M) extractant. Nitrate N concentrations in Wells 1, 2, and 3 were consistently greater than in soil water as determined from KCl extracts (Fig. 1).
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 1. Nitrate concentrations in Wells 1, 2, 3, and soil water during the course of the trial. Error bars are one standard deviation.
|
|
Bromide and nitrate N concentration in Well 4 (1.9 m from the dosing trench) increased from Day 2 until Day 5, when the concentration of both ions reached a plateau (Fig. 2). Bromide concentrations reached a peak concentration of 55 mg L–1 (55% of dosing concentration). Bromide was added as a conservative tracer for nitrate suggesting that, in the absence of removal processes, nitrate N concentration should have been 27.5 mg N L–1 at Well 4 (55% of dosing concentration of 50 mg N L–1). However, nitrate N concentrations in Well 4 were about 15 mg N L–1 indicating a removal of about 12.5 mg N L–1. Daily N removal rates per volume of wall (g N m–3 d–1) can be estimated from this removal using Darcy's law:
 | [1] |
where 
is the porewater velocity (m d–1), A is the cross-sectional area conducting ground water (1 m2 x the wall porosity), 
[nitrate N] is the decrease in nitrate N concentration (converted to g N m–3 of ground water), and soil volume is volume of wall the nitrate travels through, that is, the distance (m) between the dosing trench and Well 4 x cross-sectional area in m2. The porewater velocity was calculated from bromide data at Well 4 to be 0.47 m d–1 using an analytical solution for one-dimensional solute transport for continuous injection of a tracer (Javandel et al., 1984). The porosity of the wall material averaged 46% (Schipper and Barkle, unpublished data, 1999). From Eq. [1], the nitrate N removal rate was calculated as 1.4 g N m–3 of wall d–1. This is likely an underestimate of N removal because it does not account for background nitrate inputs to the wall.
Bromide concentrations in Wells 1, 2, and 3 reached the 50% concentration of the added tracer before our initial sampling on the first day (data not shown) and consequently it was not possible to calculate flow rates or removal of nitrate.
Microbial Activity
There were no consistent temporal trends in denitrification or denitrifying enzyme activity as nitrate was added into the denitrification wall (Fig. 3)
. Denitrification rates without added nitrate averaged 4.4 ng N cm–3 h–1 (standard deviation = 4.5, n = 54), which is equivalent to 0.11 g N m–3 of wall d–1. The highest measured rate was 22.7 ng N cm–3 h–1 (0.54 g N m–3 of wall d–1), which was less than predicted by Eq. [1]. Denitrification rates with added glucose averaged 5.9 ng N cm–3 h–1 (standard deviation = 7.0, n = 30) and were not different from denitrification rates without amendments (P < 0.05, paired t test, Genstat 6.0). Denitrification rates measured with added nitrate averaged 10.3 ng N cm–3 h–1 (standard deviation = 13.8, n = 54) and on a daily basis were greater than measurements of denitrification without any amendments or where glucose was added (P < 0.01, paired t test, Genstat 6.0). This suggested that, despite relatively large inputs of nitrate, denitrification rates in the wall were still limited by nitrate but not by carbon. Denitrifying enzyme activity averaged 88 ng N cm–3 h–1 (standard deviation = 132, n = 54) and was greater than other measures of denitrification. Average microbial biomass and available carbon seemed to show a sinusoidal pattern but variability was high and there is no apparent explanation for this pattern (Fig. 4)
.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 3. (A) Denitrification rates (with range of amendments) and (B) denitrifying enzyme activity during the dosing trial determined in laboratory incubations. All units are per cm–3 of wall material. Errors bars are one standard deviation.
|
|
View larger version (49K):
[in this window]
[in a new window]
|
Fig. 4. Changes in (A) microbial biomass and (B) carbon availability during the dosing trial. Units are per cm–3 of wall material. Errors bars are one standard deviation.
|
|
There were no significant relationships between denitrification rates (with or without nitrate amendments) and nitrate concentrations when comparing individual soil samples (data not shown). However, significant positive correlations were observed between daily-average denitrification (no nitrate amendment) and daily-average nitrate (r2 = 0.57, P < 0.05, n = 8) (Fig. 5)
and also between daily-average DEA (log10 transformed) and daily-average nitrate (r2 = 0.48, P < 0.05, n = 8) (not shown). No significant correlation was observed between daily-average denitrification (with nitrate amendment) and daily-average nitrate concentration.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 5. Relationship between daily-average nitrate and daily-average denitrification rate determined in laboratory incubations without nitrate amendments. Units are per cm–3 of wall material.
|
|
|
DISCUSSION
|
|---|
Nitrate Nitrogen Concentrations in Ground Water
There are at least two possible reasons for the difference in nitrate N concentration in ground water collected from the wells compared with concentrations in soil extracts. Spatial variations in hydraulic conductivity may have formed when the denitrification wall was first constructed due to incomplete mixing of sawdust and soil. Added water and tracers would travel preferentially along zones of higher conductivity to the wells. However, nitrate N concentrations would be lower in soil extracts as they are the average of concentrations in high and low zones of hydraulic conductivity. Similarly, added tracers may have been partitioned between mobile and less mobile or immobile phases of water. Immobile water is the fraction of water that is trapped within pores with small openings or dead-ends or water electrostatically bound to soil surfaces. Solute transport in ground water occurs primarily in the mobile fraction and the percentage of water in this domain can range between 40% in fine sandy loams (Clothier et al., 1992) to 90% in sandy subsoils (Oliver and Smettem, 2003). Nitrate injected into the wall would preferentially move in the mobile phase. The wells would also predominantly sample the mobile water in the short term and consequently nitrate N concentrations in wells closest to the dosing trench (Wells 1, 2, and 3) were similar to concentration applied in the trench (i.e., 50 mg N L–1). In contrast, nitrate N concentration in the immobile fraction would be lower because of the time required for diffusion from the mobile water fraction. Hence nitrate N concentrations in soil extracts would be lower because of the "dilution" effect from the less concentrated immobile water fractions.
Nitrate Removal and Limits
Nitrate N removal estimated from the nitrate to bromide ratio method (1.4 g N m–3 of wall d–1) was markedly greater than determined in laboratory measurements using acetylene blockage techniques (mean of 0.11 g m–3 of wall d–1 with a high of 0.54 g m–3 of wall d–1). This is in contrast to previous measurements at this site where we found denitrification rates were sufficiently high to account for observed losses of nitrate moving with ground water (Schipper and Vojvodic-Vukovic, 2000). Conversely, measurements of denitrification rates at a second denitrification wall underestimated observed rates of nitrate removal from ground water (Schipper et al., 2004). We reasoned this underestimation was due to our inability to sample representative soil volumes at the up-gradient edge of the wall where denitrification would be greatest (Schipper et al., 2004). Such "hotspots" of high rates of denitrification have been associated with patches of organic matter in surface soils (Parkin, 1987) and in aquifer material (Addy et al., 1999). In the current study, denitrification was limited by nitrate concentration rather than carbon, as shown by the increase in denitrification rates when the soils were amended with additional nitrate. The significant positive relationship between average daily nitrate N concentration and denitrification rates (without added nitrate) also suggests nitrate limitation. This limitation occurred despite dosing with relatively high concentrations of nitrate N. This apparent contradiction might be explained by the uneven distribution of nitrate throughout the wall either in zones of greater hydraulic conductivity or nitrate partitioning between mobile and immobile fractions. This uneven distribution would mean that zones within the wall would have low nitrate concentrations limiting denitrification. Laboratory measurements on samples combining zones of high and low nitrate concentrations with added nitrate would consequently give greater denitrification rates.
Other pathways of nitrate consumption are also possible, including nitrogen accumulation in organic matter or dissimilatory nitrate reduction to ammonium (DNRA). However, we monitored ammonium in soil extracts and well water during the course of the 9-d experiment (data not shown) but did not observe any increases as would be expected if DNRA was an important mechanism for nitrate consumption.
A further possible reason for the underestimation of nitrate removal by laboratory incubations compared with the field tracer study may have been due to limitations of the acetylene block approach. The acetylene block technique may underestimate denitrification due oxidation of the intermediate NO to nitrite in the presence of oxygen and acetylene (e.g., Bollmann and Conrad, 1997; McKenney et al., 1997). While denitrification in this study was measured in bottles flushed with nitrogen, even only very small amounts of oxygen that inadvertently enter incubation bottles can stimulate this oxidation (Murray and Knowles, 2004). Consequently, we consider denitrification rates measured by acetylene blockage were likely an underestimate and the values obtained from the nitrate to bromide tracer work were a better measure of the upper rate of nitrate removal in this 7-yr-old wall.
The nitrate N removal rate of 1.4 g N m–3 of wall d–1 was the same order of magnitude as rates measured in a number other denitrification walls (Table 1). Studies in Canada (Robertson et al., 2000) measured rates of between 0.7 and 2.6 g N m–3 of wall d–1 in walls constructed with 15 to 20% sawdust. They also measured much higher rates (5 to 30 g N m–3 d–1) in a bioreactor constructed with 100% organic matter, which would support a more active microbial population. Higher removal rates (15 g N m–3 of wall d–1) were measured in a denitrification wall with high nitrate inputs (>60 mg N L–1) and where soil temperatures regularly exceeded 30°C (Fahrner, 2002). Downstream nitrate concentrations exiting this wall exceeded 10 mg N L–1 at times so denitrification was probably not limited by nitrate concentrations. However, measurements were only made for the first year following construction and nitrate removal may decline with time as available carbon declines (e.g., Schipper and Vojvodic-Vukovic, 2001).
For assurance about the performance of denitrification walls we recommend direct testing in the field. Our estimate of maximum nitrate removal rate is probably conservative, because even after adding additional nitrate to the denitrification wall, nitrate levels were still limiting. Because of the multitude of possible designs, carbon substrates, and concentrations and operating conditions for denitrification walls, it will not be possible to derive one rate that will be applicable to all situations. It is also clear that laboratory tests will not necessarily be good predictors of denitrification rates in the field. We have previously demonstrated that ground water can short-circuit underneath denitrification walls, which was also not easily predicted from laboratory studies (Schipper et al., 2004). These differences in outcomes between laboratory and field testing demonstrate the need to test in situ at appropriate scales.
|
ACKNOWLEDGMENTS
|
|---|
Staff from Hautapu Dairy Factory, Fonterra, are thanked for their assistance in establishing and running the site. Dr. Art Gold, Dr. Graham Sparling, Mr. Trevor Webb, and anonymous reviewers are thanked for constructive comments. This work was funded under New Zealand Foundation for Research Science and Technology Contracts C09X0304 and LVLX0302.
|
REFERENCES
|
|---|
- Abdalla, N.A., and B. Lear. 1975. Determination of inorganic bromide in soils and plant tissues with a bromide selective-ion electrode. Commun. Soil Sci. Plant Anal. 6:489–494.
- Addy, K.L., A.J. Gold, P.M. Groffman, and P.A. Jacinthe. 1999. Ground water nitrate removal in subsoil of forested and mowed riparian buffer zones. J. Environ. Qual. 28:962–970.[Abstract/Free Full Text]
- Blakemore, L.C., P.L. Searle, and B.K. Daly. 1987. Methods for chemical analysis of soils. Sci. Rep. 80. New Zealand Soil Bureau, Lower Hutt, New Zealand.
- Bollmann, A., and R. Conrad. 1997. Enhancement by acetylene of the decomposition of nitric oxide in soil. Soil Biol. Biochem. 29:1057–1066.[CrossRef]
- Clothier, B.E., M.B. Kirkham, and J.E. McLean. 1992. In situ measurement of the effective transport volume for solute moving through soil. Soil Sci. Soc. Am. J. 56:733–736.[Abstract/Free Full Text]
- Fahrner, S. 2002. Groundwater nitrate removal using a bioremediation trench. Honours thesis. Univ. of Western Australia, Perth.
- Javandel, I., C. Doughty, and C. Tsang. 1984. Groundwater transport: Handbook of mathematical models. Water Resources Monogr. Ser. 10. Am. Geophys. Union, Washington, DC.
- McKenney, D.J., C.F. Drury, and S.W. Wang. 1997. Reaction of nitric oxide with acetylene and oxygen: Implications for denitrification assays. Soil Biol. Biochem. 61:1370–1375.
- Murray, R.E., and R. Knowles. 2004. Trace amounts of O2 affect NO and N2O production during denitrifying enzyme activity (DEA) assays. Soil Biol. Biochem. 36:513–517.[CrossRef]
- Oliver, Y.M., and K.R.J. Smettem. 2003. Parameterisation of physically based solute transport models in sandy soils. Aust. J. Soil Res. 41:771–788.[CrossRef]
- Parkin, T.B. 1987. Soil microsites as a source of denitrification variability. Soil Sci. Soc. Am. J. 51:1194–1199.[Abstract/Free Full Text]
- Robertson, W.D., D.W. Blowes, C.J. Ptacek, and J.A. Cherry. 2000. Long-term of performance of in situ reactive barriers for nitrate remediation. Ground Water 38:689–695.
- Robertson, W.D., and J.A. Cherry. 1995. In situ denitrification of septic-system nitrate using reactive porous media barriers: Field trials. Ground Water 33:99–111.
- Schipper, L.A., G.F. Barkle, M. Vojvodic-Vukovic, J.C. Hadfield, and C.P. Burgess. 2004. Hydraulic constraints on the performance of a groundwater denitrification wall for nitrate removal from shallow groundwater. J. Contam. Hydrol. 69:263–279.[Medline]
- Schipper, L.A., and M. Vojvodic-Vukovic. 1998. Nitrate removal from ground water using a denitrification wall amended with sawdust: Field trial. J. Environ. Qual. 27:664–668.[Abstract/Free Full Text]
- Schipper, L.A., and M. Vojvodic-Vukovic. 2000. Rates of nitrate removal from groundwater and denitrification in a constructed denitrification wall. Ecol. Eng. 14:269–278.
- Schipper, L.A., and M. Vojvodic-Vukovic. 2001. Five years of nitrate removal, denitrification and carbon dynamics in a denitrification wall. Water Res. 35:3473–3477.[Medline]
- Tiedje, J.M., S. Simkins, and P.M. Groffman. 1989. Perspectives on measurement of denitrification in the field including recommended protocols for acetylene based methods. Plant Soil 115:261–284.[CrossRef]
- Vance, E.D., P.C. Brookes, and D.S. Jenkinson. 1987. An extraction method for measuring soil microbial biomass C. Soil Biol. Biochem. 19:703–707.[CrossRef]
- VSN International. 2002. Genstat 6.0. VSN Int., Hempstead, UK.
TOP
ABSTRACT
INTRODUCTION
