Environ Sci Pollut Res DOI 10.1007/s11356-014-3575-3 TREATMENT OF POLLUTION IN CONSTRUCTED WETLANDS: FROM THE FUNDAMENTAL MECHANISMS TO THE FULL SCALE APPLICATIONS. WETPOL 2013
Microbial nitrogen transformation in constructed wetlands treating contaminated groundwater Oksana Coban & Peter Kuschk & Naomi S. Wells & Gerhard Strauch & Kay Knoeller
Received: 30 June 2014 / Accepted: 8 September 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Pathways of ammonium (NH4+) removal were investigated using the stable isotope approach in constructed wetlands (CWs). We investigated and compared several types of CWs: planted horizontal subsurface flow (HSSF), unplanted HSSF, and floating plant root mat (FPRM), including spatial and seasonal variations. Plant presence was the key factor influencing efficiency of NH4+ removal in all CWs, what was illustrated by lower NH4+-N removal by the unplanted HSSF CW in comparison with planted CWs. No statistically significant differences in NH4+ removal efficiencies between seasons were detected. Even though plant uptake accounted for 32–100 % of NH4+ removal during spring and summer in planted CWs, throughout the year, most of NH4+ was removed via simultaneous nitrification-denitrification, what was clearly shown by linear increase of δ15N-NH4+ with decrease of loads along the flow path and absence of nitrate (NO3−) accumulation. Average yearly enrichment factor for nitrification was −7.9‰ for planted HSSF CW and −5.8‰ for FPRM. Lack of enrichment for δ15N-NO3− implied that other processes, such as nitrification and mineralization were superimposed on denitrification and makes the stable isotope Responsible editor: Robert Duran O. Coban (*) : N. S. Wells : K. Knoeller Department of Catchment Hydrology, UFZ - Helmholtz Centre for Environmental Research, Theodor-Lieser-Str. 4, 06120 Halle/Saale, Germany e-mail: [email protected] P. Kuschk Department of Environmental Biotechnology, UFZ - Helmholtz Centre for Environmental Research, Permoserstraße 15, 04318 Leipzig, Germany G. Strauch Department of Hydrogeology, UFZ - Helmholtz Centre for Environmental Research, Permoserstraße 15, 04318 Leipzig, Germany
approach unsuitable for the estimation of denitrification in the systems obtaining NH4+ rich inflow water. Keywords Nitrogen . Ammonium . Constructed wetland . Nitrification . Denitrification . Isotope fractionation
Introduction Ammonium (NH4+) is one of the major toxic compounds as well as a critical long-term pollutant in marine environments (Ip et al. 2001), landfill leachate (Mangimbulude et al. 2012), some factories (Harrington and McInnes 2009; Havens et al. 2001), and groundwater (Siljeg et al. 2010). As it is toxic to fish and causes eutrophication of lakes, rivers, and wetlands, NH4+ contamination poses a serious environmental problem. Constructed wetlands (CWs) have been widely used in wastewater treatment due to their low energy requirements and easy operation (Garcia et al. 2010). They are appropriate to remove a wide range of contaminants, in particular nitrogen (N) compounds. The main processes that can affect N removal in CWs are plant uptake and microbial assimilation or reduction to inert dinitrogen gas (N2). NH4+ is the dominant form of N in sewage/effluent affected systems. When NH4+ enters a CW, it can be taken up by plants, immobilized into the organic material, or oxidized to nitrate (NO3−) by microorganisms. In the latter case, the produced NO3− can further undergo denitrification under anaerobic conditions, which is the stepwise reduction of NO3− via nitrite (NO2−) to nitrous oxide (N2O) and N2 using organic carbon as an electron acceptor. Over the last decade, there is also evidence that some microorganisms can reduce NH4+ directly to N2 using NO2− or NO3− as an electron donor in a process known as anammox (anaerobic ammonium oxidation) (Lee et al. 2009). However, the quantification of these processes can be difficult due to a
Environ Sci Pollut Res
complexity of the wetland systems concerning flow dynamics, root exudation, substrates (gravel), and the mosaic of aerobic and anaerobic zones within the root zone of the plants (Kadlec 2000; Martin and Reddy 1997). For investigations of N cycle in CWs, a wide range of methods is applied, such as physicochemical measurements, molecular biology techniques, isotope tracking using 15N labels, and the 15/14N natural abundance of various N compounds. However, none of these techniques is satisfactory as a result of high spatial and temporal variations in the N transformation processes (Groffman et al. 2006). The natural abundance of the rare stable isotope of N, 15N, is now used widely in research on N cycling in pathways and ecosystems. Variations in 15/14N among samples reflect N isotope fractionations. These occur because bacteria discriminate against heavy isotope, causing the residual pool to become progressively enriched with 15N as the reaction progresses (Robinson 2001). Once the degree of isotopic discrimination that occurs during a process is known (i.e., the enrichment factor), stable isotope ratios can be used in concert with concentration data to quantify fluxes. While the microbial transformation processes produce isotope fractionation, other processes such as microbial immobilization, mineralization, and plant uptake as well as dilutions and hydrological transport do not cause any changes in stable isotope abundances (Kendall and McDonnell 1999). Currently, the utilization of process-related stable isotope fractionation effects in many parts of the N cycle is of high interest. However, environmental N pools reflect a mixture of a variety of sources and processes (Hauck 1973), making interpretation of N isotope signatures challenging. A quantitative assessment of denitrification to the overall NO3− losses can be made only when nitrification is assumed not to influence 15/14N in wetland systems (Lund et al. 2000; Sovik and Morkved 2008). Therefore, this method is not suitable for estimating denitrification contribution to NO3− removal in CWs treating contaminated water, where NH4+ and/or organic N concentration in inflow are high and the system is partially aerobic (Sovik and Morkved 2007). There is still lack of investigations on N fractionation of both NH4+ and NO3− in CWs which receive inflow with high NH4+ concentrations. Moreover, the influence of factors such as the presence of plants and substrate, i.e., various types of CWs on N transformations was not examined as well. Thus, there is a need for a better knowledge of the effect of such processes as plant uptake, microbial assimilation, mineralization, and nitrification on N isotopic fractionation in concert with denitrification in different types of CWs. Reinhardt et al. (2006) quantified nitrification, mineralization, and denitrification by interpretation of the natural isotopic composition of NO3− as well as NH4+. However, the information was limited as samples were taken only from inflow and outflow. Advances in stable isotope methods allow
getting accurate measurements at very low concentrations. Measurements of the 15/14N natural abundance of both NO3− and NH4+ were used to identify spatial variations in N transformations at a fine spatial resolution throughout the CWs with different substrates. Thus, the objectives of this study were to use the stable isotope composition of multiple N species in order to (i) compare pathways of NH4+ removal in three different types of CWs and (ii) investigate spatial and seasonal variations in N transformation processes in CWs.
Materials and methods Constructed wetland design The constructed wetlands in the study site were built as a part of CoTra (Compartment Transfer) project in 2007 at Leuna Megasite near Leipzig, Germany. Leuna has been a major chemical manufacturing site since the beginning of the 20th century, and accidental spills, improper handling, and damages due to heavy bombing during World War II have left the underlying groundwater highly contaminated (Martienssen et al. 2006). The constructed wetlands consisted of stainless steel basins (5 m×1.1 m×0.6 m) filled with three plant and substrate combinations. Two systems were horizontal subsurface flow (HSSF), one planted with common reed (Phragmites australis) and the second served as control and remained unplanted. They were filled with gravel (grain size 2– 3.2 mm) up to a height of 50 cm and the water level was set to 40 cm, resulting in a vadose zone of 10 cm. The third system was constructed as a hydroponic floating plant root mat (FPRM), with P. australis only supported by the densely woven root bed (no gravel) and maintained at a water depth of 30 cm. The precipitation was measured daily by a nearby weather station, and the inflow and outflow volumes were recorded using a flow meter. The theoretical hydraulic retention time (assuming no water loss) was 6.88 days for the gravel-based CWs. Inflow water was pumped in from adjacent contaminated groundwater. Main characteristics of inflow water during the period of investigations are shown in Table 1. Sampling procedure Inflow and outflow water samples as well as pore water samples were collected every 2 weeks from July 2012 through June 2013, with exception of winter season (November 2012– March 2013), making a total of 14 sampling days. Water from the filter was pumped using a peristaltic pump (REGLO digital, Ismatec) from several points distributed along the flow path (1, 2.5, 4 m) and the depth of the filter (0.2, 0.3, 0.4 m for HSSF CW; 0.3 m for FPRM). The temperature and redox
Environ Sci Pollut Res Table 1 Mean (n=37) concentration of organic and inorganic contaminants of groundwater used as the wetland’s influent in Leuna (2012– 2013) Contaminant
Mean concentration [mg L−1]
Standard deviation
Benzene Methyl-tert-butyl ether (MTBE) NH4+-N NO2−-N NO3−-N Total organic carbon (TOC) Chemical oxygen demand (COD) The 5-day biological oxygen demand (BOD5)
4.1 0.4 23.4 0.2 b.d.l. 17.5 45.0 21.0
4.0 0.4 5.0 0.2 – 7.0 22.0 14.0
b.d.l. below detection limit
potential were measured on-site using a flow-through cell equipped with a redox electrode (Pt/Ag+/AgCl/Cl−; Sentix ORP, WTW, Germany). pH was measured in situ using a SenTix41 electrode with pH 537 Microprocessor (WTW, Weilheim, Germany). The samples for NH4+, NO2−, and NO3− concentrations were transferred into 25-mL brown glass bottles and measured in the field laboratory. An additional 50 ml of sample were collected for NO3− isotope analysis, and 0.5–2 L of sample (dependent on NH4+ concentration) for NH4+ isotope analysis was acidified to pH 2 using 98 % sulfuric acid. All samples were stored at 4 °C until analysis. Laboratory experiments The site-specific enrichment factor for denitrification was measured by placing 40 g of homogenized samples of gravel and roots from different depths and distances into serum bottles and filling with 0.12 L of inflow water. The experiment was running in three replicates. Fifteen milliliter of 3-(Nmorpholino) propanesulfonic acid (MOPS) buffer was added to the final concentration of 50 mM. Then bottles were flushed with N2 gas and sealed. Samples were preincubated for 24 h at 20 °C in the dark with shaking in order to remove residual oxygen. Subsequently, organic carbon (equimolar mixture of lactate, acetate, benzoate, and ethanol) and NO3− were added to each bottle to bring them up to a final concentration of 7 mg NO3−-N L−1. Twenty milliliter of sample was taken at 1, 2, and 3.5 h after substrate addition with a syringe, filter with 0.25 mm and stored at +4 °C until analysis. Sampling points were chosen based on preliminary experiment, where 90 % of NO3− was removed after 3.5 h. The site-specific enrichment factor for nitrification was determined in an analogous way. One hundred grams of homogenized samples of gravel and roots from planted HSSF CW were placed in 500-ml Schott bottles with 0.3 L of inflow water. The experiment was running in three
replicates. Samples were loosely covered by screw caps to allow air access and incubated at 20 °C in the dark with shaking. Subsequently, NH4+ was added to the final concentration of nearly 150 mg NH4+-N L−1; 0.025–0.1 L of sample dependent on NH4+ concentration was taken with a syringe immediately after substrate addition, and then again after 44, 75, and 100 h. Concentration of NH4+ was measured immediately, and for isotope measurements, samples were acidified to pH 2 by 98 % sulfuric acid and stored at +4 °C until analysis. It should be noted that these site-specific enrichment factors were measured using only substrate from the HSSF CW because only this system had consistent N removal throughout the year, and therefore, the enrichment factors were the most representative for the microbial N transformation processes occurring in CWs. Given that the HSSF CW had maximum N removal in summer, the incubation temperature for the calculation of enrichment factors under simulated in situ conditions in this system was chosen based on summer average air temperature (+20 °C). In order to avoid microbial growth and therefore, any changes in the enrichment factors, samples were incubated immediately and the adding of substrate and sampling was conducted without delay as described above.
Chemical analysis Analysis of inorganic ions (NH4+, NO2−, NO3−) was conducted using a photometer (Spectroquant® Nova 60, Merck) and the Merck quick tests (number 1.00683.0001 for NH4+, 1.14776. for NO2−, and 1.09713.0001 for NO3−). Analytical precision for NH4+-N was ±0.53 mg L−1, for NO2−-N ±0.003 mg L−1, and for NO3−-N ±0.11 mg L−1. For isotope analysis, NH 4 + was liberated from aqueous solutions (samples) as NH3 under basic pH and redissolved in a standard solution of sulfuric acid. The preparation of NH4+ followed the basic principle of a Kjeldahl distillation (SaezPlaza et al. 2013). The resulting ammonium sulfate was homogenized and weighed into tin capsules that were combusted in an elemental analyzer Vario ISOTOPE Cube (Elementar Analysensysteme GmbH, Hanau) connected to an isotope ratio mass spectrometer Isoprime 100 (Isoprime Ltd, Cheadle Hulm). For nitrification incubations, NH4+ isotopic composition was measured by oxidizing NH4+ to NO2− using a bromate solution, and then reacting with acetic acid buffered sodium azide to create N2O (Zhang and Altabet 2008). The 15/14 N of the resultant N2O was measured on an IRMS Delta V plus (Thermo Electron GmbH, Bremen) with a Gasbench II (Thermo Electron GmbH, Bremen). The 15/14N isotope ratio is expressed as delta (δ)-notation in per mil (‰) relative to the standard AIR. Each sample batch was run with international reference materials for calibration USGS25 (δ15N, −30.4‰), and USGS26 (δ15N, +53.7‰) (Adamsen and Reeder 1983),
Environ Sci Pollut Res
plus an internal standard (δ15N, 0‰). For both methods, the analytical precision was ±0.4‰. The NO3− isotope signature was determined using the denitrifier method (McIlvin and Casciotti 2011) and measured on an IRMS Delta V plus (Thermo Electron GmbH, Bremen) with a Gasbench II (Thermo Electron GmbH, Bremen). That method allowed for a measurement of δ15N by measuring N2O produced by controlled reduction of sample NO3− by the bacterial strain Pseudomonas chlororaphis (ATCC #13985) in 12-ml Exetainer vials. Analytical precision for δ15N was ±0.4‰. For calibration of N, the reference NO3− IAEA-N3 (δ15N, +4.7 ‰ AIR), USGS32 (δ15N, +180 ‰ AIR), and USGS 35 (δ15N, +2.7‰ AIR) were used.
Here, Mij is the contaminant mass flux in mg d−1 at distance i m and depth j cm, Cij is the contaminant concentration in mg L−1 at distance i m and depth j cm, Vin is the inflow rate in L d−1, Vout is the outflow volume in L d−1, Li is the distance from the inflow at point i m, and L is the length of the CW in m. Data analysis Differences between treatments and over time and distance and depth were determined using one-way analysis of variance (ANOVA). The statistical analysis was performed by IBM SPSS Statistics 21 software, and the differences were regarded as significant at p
Microbial nitrogen transformation in constructed wetlands treating contaminated groundwater Oksana Coban & Peter Kuschk & Naomi S. Wells & Gerhard Strauch & Kay Knoeller
Received: 30 June 2014 / Accepted: 8 September 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Pathways of ammonium (NH4+) removal were investigated using the stable isotope approach in constructed wetlands (CWs). We investigated and compared several types of CWs: planted horizontal subsurface flow (HSSF), unplanted HSSF, and floating plant root mat (FPRM), including spatial and seasonal variations. Plant presence was the key factor influencing efficiency of NH4+ removal in all CWs, what was illustrated by lower NH4+-N removal by the unplanted HSSF CW in comparison with planted CWs. No statistically significant differences in NH4+ removal efficiencies between seasons were detected. Even though plant uptake accounted for 32–100 % of NH4+ removal during spring and summer in planted CWs, throughout the year, most of NH4+ was removed via simultaneous nitrification-denitrification, what was clearly shown by linear increase of δ15N-NH4+ with decrease of loads along the flow path and absence of nitrate (NO3−) accumulation. Average yearly enrichment factor for nitrification was −7.9‰ for planted HSSF CW and −5.8‰ for FPRM. Lack of enrichment for δ15N-NO3− implied that other processes, such as nitrification and mineralization were superimposed on denitrification and makes the stable isotope Responsible editor: Robert Duran O. Coban (*) : N. S. Wells : K. Knoeller Department of Catchment Hydrology, UFZ - Helmholtz Centre for Environmental Research, Theodor-Lieser-Str. 4, 06120 Halle/Saale, Germany e-mail: [email protected] P. Kuschk Department of Environmental Biotechnology, UFZ - Helmholtz Centre for Environmental Research, Permoserstraße 15, 04318 Leipzig, Germany G. Strauch Department of Hydrogeology, UFZ - Helmholtz Centre for Environmental Research, Permoserstraße 15, 04318 Leipzig, Germany
approach unsuitable for the estimation of denitrification in the systems obtaining NH4+ rich inflow water. Keywords Nitrogen . Ammonium . Constructed wetland . Nitrification . Denitrification . Isotope fractionation
Introduction Ammonium (NH4+) is one of the major toxic compounds as well as a critical long-term pollutant in marine environments (Ip et al. 2001), landfill leachate (Mangimbulude et al. 2012), some factories (Harrington and McInnes 2009; Havens et al. 2001), and groundwater (Siljeg et al. 2010). As it is toxic to fish and causes eutrophication of lakes, rivers, and wetlands, NH4+ contamination poses a serious environmental problem. Constructed wetlands (CWs) have been widely used in wastewater treatment due to their low energy requirements and easy operation (Garcia et al. 2010). They are appropriate to remove a wide range of contaminants, in particular nitrogen (N) compounds. The main processes that can affect N removal in CWs are plant uptake and microbial assimilation or reduction to inert dinitrogen gas (N2). NH4+ is the dominant form of N in sewage/effluent affected systems. When NH4+ enters a CW, it can be taken up by plants, immobilized into the organic material, or oxidized to nitrate (NO3−) by microorganisms. In the latter case, the produced NO3− can further undergo denitrification under anaerobic conditions, which is the stepwise reduction of NO3− via nitrite (NO2−) to nitrous oxide (N2O) and N2 using organic carbon as an electron acceptor. Over the last decade, there is also evidence that some microorganisms can reduce NH4+ directly to N2 using NO2− or NO3− as an electron donor in a process known as anammox (anaerobic ammonium oxidation) (Lee et al. 2009). However, the quantification of these processes can be difficult due to a
Environ Sci Pollut Res
complexity of the wetland systems concerning flow dynamics, root exudation, substrates (gravel), and the mosaic of aerobic and anaerobic zones within the root zone of the plants (Kadlec 2000; Martin and Reddy 1997). For investigations of N cycle in CWs, a wide range of methods is applied, such as physicochemical measurements, molecular biology techniques, isotope tracking using 15N labels, and the 15/14N natural abundance of various N compounds. However, none of these techniques is satisfactory as a result of high spatial and temporal variations in the N transformation processes (Groffman et al. 2006). The natural abundance of the rare stable isotope of N, 15N, is now used widely in research on N cycling in pathways and ecosystems. Variations in 15/14N among samples reflect N isotope fractionations. These occur because bacteria discriminate against heavy isotope, causing the residual pool to become progressively enriched with 15N as the reaction progresses (Robinson 2001). Once the degree of isotopic discrimination that occurs during a process is known (i.e., the enrichment factor), stable isotope ratios can be used in concert with concentration data to quantify fluxes. While the microbial transformation processes produce isotope fractionation, other processes such as microbial immobilization, mineralization, and plant uptake as well as dilutions and hydrological transport do not cause any changes in stable isotope abundances (Kendall and McDonnell 1999). Currently, the utilization of process-related stable isotope fractionation effects in many parts of the N cycle is of high interest. However, environmental N pools reflect a mixture of a variety of sources and processes (Hauck 1973), making interpretation of N isotope signatures challenging. A quantitative assessment of denitrification to the overall NO3− losses can be made only when nitrification is assumed not to influence 15/14N in wetland systems (Lund et al. 2000; Sovik and Morkved 2008). Therefore, this method is not suitable for estimating denitrification contribution to NO3− removal in CWs treating contaminated water, where NH4+ and/or organic N concentration in inflow are high and the system is partially aerobic (Sovik and Morkved 2007). There is still lack of investigations on N fractionation of both NH4+ and NO3− in CWs which receive inflow with high NH4+ concentrations. Moreover, the influence of factors such as the presence of plants and substrate, i.e., various types of CWs on N transformations was not examined as well. Thus, there is a need for a better knowledge of the effect of such processes as plant uptake, microbial assimilation, mineralization, and nitrification on N isotopic fractionation in concert with denitrification in different types of CWs. Reinhardt et al. (2006) quantified nitrification, mineralization, and denitrification by interpretation of the natural isotopic composition of NO3− as well as NH4+. However, the information was limited as samples were taken only from inflow and outflow. Advances in stable isotope methods allow
getting accurate measurements at very low concentrations. Measurements of the 15/14N natural abundance of both NO3− and NH4+ were used to identify spatial variations in N transformations at a fine spatial resolution throughout the CWs with different substrates. Thus, the objectives of this study were to use the stable isotope composition of multiple N species in order to (i) compare pathways of NH4+ removal in three different types of CWs and (ii) investigate spatial and seasonal variations in N transformation processes in CWs.
Materials and methods Constructed wetland design The constructed wetlands in the study site were built as a part of CoTra (Compartment Transfer) project in 2007 at Leuna Megasite near Leipzig, Germany. Leuna has been a major chemical manufacturing site since the beginning of the 20th century, and accidental spills, improper handling, and damages due to heavy bombing during World War II have left the underlying groundwater highly contaminated (Martienssen et al. 2006). The constructed wetlands consisted of stainless steel basins (5 m×1.1 m×0.6 m) filled with three plant and substrate combinations. Two systems were horizontal subsurface flow (HSSF), one planted with common reed (Phragmites australis) and the second served as control and remained unplanted. They were filled with gravel (grain size 2– 3.2 mm) up to a height of 50 cm and the water level was set to 40 cm, resulting in a vadose zone of 10 cm. The third system was constructed as a hydroponic floating plant root mat (FPRM), with P. australis only supported by the densely woven root bed (no gravel) and maintained at a water depth of 30 cm. The precipitation was measured daily by a nearby weather station, and the inflow and outflow volumes were recorded using a flow meter. The theoretical hydraulic retention time (assuming no water loss) was 6.88 days for the gravel-based CWs. Inflow water was pumped in from adjacent contaminated groundwater. Main characteristics of inflow water during the period of investigations are shown in Table 1. Sampling procedure Inflow and outflow water samples as well as pore water samples were collected every 2 weeks from July 2012 through June 2013, with exception of winter season (November 2012– March 2013), making a total of 14 sampling days. Water from the filter was pumped using a peristaltic pump (REGLO digital, Ismatec) from several points distributed along the flow path (1, 2.5, 4 m) and the depth of the filter (0.2, 0.3, 0.4 m for HSSF CW; 0.3 m for FPRM). The temperature and redox
Environ Sci Pollut Res Table 1 Mean (n=37) concentration of organic and inorganic contaminants of groundwater used as the wetland’s influent in Leuna (2012– 2013) Contaminant
Mean concentration [mg L−1]
Standard deviation
Benzene Methyl-tert-butyl ether (MTBE) NH4+-N NO2−-N NO3−-N Total organic carbon (TOC) Chemical oxygen demand (COD) The 5-day biological oxygen demand (BOD5)
4.1 0.4 23.4 0.2 b.d.l. 17.5 45.0 21.0
4.0 0.4 5.0 0.2 – 7.0 22.0 14.0
b.d.l. below detection limit
potential were measured on-site using a flow-through cell equipped with a redox electrode (Pt/Ag+/AgCl/Cl−; Sentix ORP, WTW, Germany). pH was measured in situ using a SenTix41 electrode with pH 537 Microprocessor (WTW, Weilheim, Germany). The samples for NH4+, NO2−, and NO3− concentrations were transferred into 25-mL brown glass bottles and measured in the field laboratory. An additional 50 ml of sample were collected for NO3− isotope analysis, and 0.5–2 L of sample (dependent on NH4+ concentration) for NH4+ isotope analysis was acidified to pH 2 using 98 % sulfuric acid. All samples were stored at 4 °C until analysis. Laboratory experiments The site-specific enrichment factor for denitrification was measured by placing 40 g of homogenized samples of gravel and roots from different depths and distances into serum bottles and filling with 0.12 L of inflow water. The experiment was running in three replicates. Fifteen milliliter of 3-(Nmorpholino) propanesulfonic acid (MOPS) buffer was added to the final concentration of 50 mM. Then bottles were flushed with N2 gas and sealed. Samples were preincubated for 24 h at 20 °C in the dark with shaking in order to remove residual oxygen. Subsequently, organic carbon (equimolar mixture of lactate, acetate, benzoate, and ethanol) and NO3− were added to each bottle to bring them up to a final concentration of 7 mg NO3−-N L−1. Twenty milliliter of sample was taken at 1, 2, and 3.5 h after substrate addition with a syringe, filter with 0.25 mm and stored at +4 °C until analysis. Sampling points were chosen based on preliminary experiment, where 90 % of NO3− was removed after 3.5 h. The site-specific enrichment factor for nitrification was determined in an analogous way. One hundred grams of homogenized samples of gravel and roots from planted HSSF CW were placed in 500-ml Schott bottles with 0.3 L of inflow water. The experiment was running in three
replicates. Samples were loosely covered by screw caps to allow air access and incubated at 20 °C in the dark with shaking. Subsequently, NH4+ was added to the final concentration of nearly 150 mg NH4+-N L−1; 0.025–0.1 L of sample dependent on NH4+ concentration was taken with a syringe immediately after substrate addition, and then again after 44, 75, and 100 h. Concentration of NH4+ was measured immediately, and for isotope measurements, samples were acidified to pH 2 by 98 % sulfuric acid and stored at +4 °C until analysis. It should be noted that these site-specific enrichment factors were measured using only substrate from the HSSF CW because only this system had consistent N removal throughout the year, and therefore, the enrichment factors were the most representative for the microbial N transformation processes occurring in CWs. Given that the HSSF CW had maximum N removal in summer, the incubation temperature for the calculation of enrichment factors under simulated in situ conditions in this system was chosen based on summer average air temperature (+20 °C). In order to avoid microbial growth and therefore, any changes in the enrichment factors, samples were incubated immediately and the adding of substrate and sampling was conducted without delay as described above.
Chemical analysis Analysis of inorganic ions (NH4+, NO2−, NO3−) was conducted using a photometer (Spectroquant® Nova 60, Merck) and the Merck quick tests (number 1.00683.0001 for NH4+, 1.14776. for NO2−, and 1.09713.0001 for NO3−). Analytical precision for NH4+-N was ±0.53 mg L−1, for NO2−-N ±0.003 mg L−1, and for NO3−-N ±0.11 mg L−1. For isotope analysis, NH 4 + was liberated from aqueous solutions (samples) as NH3 under basic pH and redissolved in a standard solution of sulfuric acid. The preparation of NH4+ followed the basic principle of a Kjeldahl distillation (SaezPlaza et al. 2013). The resulting ammonium sulfate was homogenized and weighed into tin capsules that were combusted in an elemental analyzer Vario ISOTOPE Cube (Elementar Analysensysteme GmbH, Hanau) connected to an isotope ratio mass spectrometer Isoprime 100 (Isoprime Ltd, Cheadle Hulm). For nitrification incubations, NH4+ isotopic composition was measured by oxidizing NH4+ to NO2− using a bromate solution, and then reacting with acetic acid buffered sodium azide to create N2O (Zhang and Altabet 2008). The 15/14 N of the resultant N2O was measured on an IRMS Delta V plus (Thermo Electron GmbH, Bremen) with a Gasbench II (Thermo Electron GmbH, Bremen). The 15/14N isotope ratio is expressed as delta (δ)-notation in per mil (‰) relative to the standard AIR. Each sample batch was run with international reference materials for calibration USGS25 (δ15N, −30.4‰), and USGS26 (δ15N, +53.7‰) (Adamsen and Reeder 1983),
Environ Sci Pollut Res
plus an internal standard (δ15N, 0‰). For both methods, the analytical precision was ±0.4‰. The NO3− isotope signature was determined using the denitrifier method (McIlvin and Casciotti 2011) and measured on an IRMS Delta V plus (Thermo Electron GmbH, Bremen) with a Gasbench II (Thermo Electron GmbH, Bremen). That method allowed for a measurement of δ15N by measuring N2O produced by controlled reduction of sample NO3− by the bacterial strain Pseudomonas chlororaphis (ATCC #13985) in 12-ml Exetainer vials. Analytical precision for δ15N was ±0.4‰. For calibration of N, the reference NO3− IAEA-N3 (δ15N, +4.7 ‰ AIR), USGS32 (δ15N, +180 ‰ AIR), and USGS 35 (δ15N, +2.7‰ AIR) were used.
Here, Mij is the contaminant mass flux in mg d−1 at distance i m and depth j cm, Cij is the contaminant concentration in mg L−1 at distance i m and depth j cm, Vin is the inflow rate in L d−1, Vout is the outflow volume in L d−1, Li is the distance from the inflow at point i m, and L is the length of the CW in m. Data analysis Differences between treatments and over time and distance and depth were determined using one-way analysis of variance (ANOVA). The statistical analysis was performed by IBM SPSS Statistics 21 software, and the differences were regarded as significant at p
