- Email: [email protected]
Genetic potential for N2O emissions from the sediment of a free water surface constructed wetland
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
Available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/watres
Genetic potential for N2O emissions from the sediment of a free water surface constructed wetland Arantzazu Garcı´a-Lledo´ a, Ariadna Vilar-Sanz a, Rosalia Trias a, Sara Hallin b, Lluı´s Ban˜eras a,* a b
Molecular Microbial Ecology Group, Institute of Aquatic Ecology, Universitat de Girona, C/ Maria Aure`lia Capmany, 69, 17071 Girona, Spain Swedish University of Agricultural Sciences, Department of Microbiology, Box 7025, Uppsala 750 07, Sweden
article info
abstract
Article history:
Removal of nitrogen is a key aspect in the functioning of constructed wetlands. However,
Received 13 April 2011
incomplete denitrification may result in the net emission of the greenhouse gas nitrous
Received in revised form
oxide (N2O) resulting in an undesired effect of a system supposed to provide an ecosystem
22 July 2011
service. In this work we evaluated the genetic potential for N2O emissions in relation to the
Accepted 14 August 2011
presence or absence of Phragmites and Typha in a free water surface constructed wetland
Available online 30 August 2011
(FWS-CW), since vegetation, through the increase in organic matter due to litter degradation, may significantly affect the denitrification capacity in planted areas. Quantitative
Keywords:
real-time PCR analyses of genes in the denitrification pathway indicating capacity to
Constructed wetlands
produce or reduce N2O were conducted at periods of different water discharge. Genetic
Denitrification
potential for N2O emissions was estimated from the relative abundances of all denitrifi-
Genetic potential
cation genes and nitrous oxide reductase encoding genes (nosZ ). nosZ abundance was
N2O emission
invariably lower than the other denitrifying genes (down to 100 fold), and differences
qPCR
increased significantly during periods of high nitrate loads in the CW suggesting a higher
Vegetation
genetic potential for N2O emissions. This situation coincided with lower nitrogen removal efficiencies in the treatment cell. The presence and the type of vegetation, mainly due to changes in the sediment carbon and nitrogen content, correlated negatively to the ratio between nitrate and nitrite reducers and positively to the ratio between nitrite and nitrous oxide reducers. These results suggest that the potential for nitrous oxide emissions is higher in vegetated sediments. ª 2011 Elsevier Ltd. All rights reserved.
1.
Introduction
Constructed wetlands (CWs) are recognized as feasible alternatives for removal of nitrogen in wastewater from agricultural, industrial or municipal activities and have been exploited as secondary or tertiary treatment alternatives to promote water reuse (Reed et al., 1995; DeBusk and DeBusk, 2000). There are two types of CWs, the subsurface flow systems (SSF) and the free-water surface systems (FWS) that
mainly differ in the presence of a free water flow over the sediment surface. Water treatment in CWs involves processes driven by the sediment, the vegetation and its associated microbial communities (EPA, 2000). It is generally assumed that areas planted with emergent macrophytes positively affect the water restoration capacity of the wetland (Zhu and Sikora, 1995; Lin et al., 2002; Ibekwe et al., 2007). This is due to stimulation of microbial growth at either the epiphyton or the root surface, e.g. by influencing the oxygen conditions and
* Corresponding author. Tel.: þ34 972 418 177; fax: þ34 972 418 150. E-mail addresses: [email protected] (A. Garcı´a-Lledo´), [email protected] (A. Vilar-Sanz), [email protected] (R. Trias), [email protected] (S. Hallin), [email protected] (L. Ban˜eras). 0043-1354/$ e see front matter ª 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.watres.2011.08.025
5622
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
creating microenvironments with oxiceanoxic zones (Gersberg et al., 1986; Zhu and Sikora, 1995; Boudraa et al., 1999) or by the excretion of organic compounds from the roots (Prade and Trolldenier, 1990; Brix, 1997; Nguyen, 2003), as well as causing changes in hydraulic retention time (Whitney et al., 2003; Toet et al., 2005; Kjellin et al., 2007). All these processes influence nitrogen removal in CWs since they affect the nitrifying and denitrifying microorganisms, which are ultimately responsible for nitrogen transformations resulting in nitrogen removal (Johnston, 1991; Zhu and Sikora, 1995; Boudraa et al., 1999; Ko¨rner, 1999; Vymazal, 2001; Lin et al., 2002; Francis et al., 2007). The microbial process denitrification is especially interesting in CWs as it represents the net nitrogen loss from the system. Denitrifying bacteria use nitrate as an alternative electron acceptor, which is sequentially reduced to nitrogen gas. The first step is catalyzed by nitrate reductases, either a membrane-bound enzyme (Nar type) or a soluble periplasmic enzyme (Nap) (Zumft, 1997; Moreno-Vivian et al., 1999; Philippot, 2002). These two reductases are not exclusive, but can be found simultaneously in the same organism (Carter et al., 1995; Gregory et al., 2003; Roussel-Delif et al., 2005). Dissimilatory nitrite reductases (Nir) catalyze the second step of denitrification and two functionally equivalent enzymes have been described, a cytochrome cd1 type and a copper containing type, encoded by the nirS and nirK genes, respectively. They are mutually exclusive and have not been found in the same strain so far (Zumft, 1997), although different strains of the same species may contain different Nir genes (Coyne et al., 1989; Philippot, 2002). Nitric oxide reductases (Nor) catalyze the reduction of nitric oxide to nitrous oxide and the last step of denitrification is catalyzed by nitrous oxide reductases (Nos) which lead to the production of nitrogen gas (Zumft, 1997). All denitrifying genes described so far have been used as molecular markers for qualitative and quantitative studies of denitrifying bacteria in the environment. The expression of denitrifying genes is dependent on the presence of the respective enzymes substrates, i.e. sequential oxidized forms of inorganic nitrogen, and low oxygen concentration due to the facultative nature of this process (Tiedje, 1988). Most denitrifiers use organic compounds as electron donors and easily available carbon is another prerequisite. Emissions of the potent greenhouse gas N2O may be due to the activity of some denitrifying bacteria that are unable to perform the final step of denitrification, due to the lack of the nosZ gene (Wood et al., 2001; Kandeler et al., 2006; Vial et al., 2006; Jones et al., 2008; Abell et al., 2010; Philippot et al., 2011). Cheneby et al., 2004 compared denitrifying bacteria between non-planted and maize planted soil and found that nosZ lacking bacteria were dominant in the rhizosphere suggesting that plants can affect N2O emission by selecting for denitrifiers that do not have the capacity to reduce N2O. Plants increase easily available carbon compounds by root exudates and through the increase in organic matter due to litter degradation. In addition, plants cause radial oxygen gradients around the roots. These factors affect denitrifying bacteria, as well as other heterotrophic bacteria (Boudraa et al., 1999; Henry et al., 2008). Studies specifically directed to the analysis of denitrification in the rhizosphere have shown that changes in the activity, composition and the abundance of
denitrifying bacteria may be plant-specific (Cheneby et al., 2004; Henry et al., 2008; Ruiz-Rueda et al., 2009). In the sediment of planted areas the presence of decomposing litter may be an additional factor supporting and shaping the denitrifying community (Ingersoll and Baker, 1998). If emergent macrophytes in CWs also select for denitrifiers genetically incapable of reducing N2O is not known. If so, although promoting nitrogen removal in CWs, macrophytes potentially contribute to increase the N2O/N2 end-product ratio. We evaluated the genetic potential for N2O emissions of the sediment in relation to the presence or absence of the main plant species used in treatment wetlands, Phragmites australis and Typha latifolia. For that purpose, we determined the abundance of denitrifiers by quantitative real-time PCR (qPCR) of key functional genes in the denitrification pathway (narG, napA, nirS, nirK and nosZ ) in comparison to the total bacterial community size targeted by the 16S rRNA gene in the sediments of the Empuriabrava FWS-CWs at periods of different water discharge over the year.
2.
Material and methods
2.1.
Study site
The Empuriabrava FWS-CWs (Girona, NE Spain) were designed in 1998 as a tertiary treatment to increase the water quality of the effluent of a nearby located wastewater treatment plant (WWTP). The CWs are included in the natural preserved area of Els Aiguamolls de l’Emporda` and are designed to provide additional water to avoid excessive desiccation of the flooded area in summer (Ruiz-Rueda et al., 2007). Nitrogen is removed at the WWTP by a combined nitrificationedenitrification process using carrousel-type bioreactors and enhanced aeration, resulting in nitrate loads to the CWs changing between 251 and 1016 kg N per month. The Empuriabrava CWs consists of three parallel cells followed by a shallower lagoon and all measurements were conducted on treatment cell 3. The cell is planted with independent communities of reed (P. australis) and cattail (T. latifolia) showing an almost equivalent distribution of both plant species (49.3% of reed and 50.7% of cattail; Fig. 1). The average water depth in the cell was about 0.6 m and the sediment was water saturated during the sampling period. The sediment depths varied between 5 and 10 cm in non-vegetated areas and increased to 15e20 cm in planted surfaces. The location of the vegetation spots and the plant species were designed in the original project and planted shortly after construction of the CW (1998). Harvest of the aerial biomass is performed every one or two years during winter.
2.2.
Sampling and DNA extraction
Temperature, conductivity, oxygen and pH at the sampling locations were measured with a portable multiparametric probe (Yellow Spring Instruments 650MDS). Water samples (20 ml) from the water column in the wetland were collected and analyzed for nitrate, nitrite and ammonia concentration as described previously (Garcı´a-Lledo´ et al., 2011). Water samples (100 ml) were regularly also collected at the
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
5623
Fig. 1 e Scheme of vegetation stands and sampling locations in the third cell of the Empuriabrava FWS-CWs. Sampling plots are indicated as black dots with the following two-letter codes: SE e bulk sediment, TY e areas planted with Typha latifolia, PH e areas planted with Phragmites australis.
secondary effluent of the WWTP and at the end point of the treatment cells for analyses of pH, conductivity, biological oxygen demand (BOD), chemical oxygen demand (COD), oxygen saturation, and concentrations of nitrate, nitrite and ammonium, using conventional standard methods for wastewater analyses (APHA, 1998). Total water flow into the wetland system was automatically monitored daily. Sediment samples were collected May 2008, August 2008, March 2009 and July 2009 near the inlet and outlet areas. Three sampling plots of one square meter each were defined at the two areas. Sampling plot 1 was not covered with any emergent plant and is referred as bulk sediment (SE). Plot 2 consisted of an area covered with T. latifolia (TY) and Plot 3 was covered with P. australis (PH). Three sediment replicates were taken in each plot (Fig. 1). Sediment cores were obtained from planted and unplanted areas using a 2-cm-diameter methacrylate tube mounted in a manual sampler. The upper 3 cm of the sediment core were aseptically transferred to a container and chilled on ice for transportation. Once in the laboratory, sediment was completely homogenized and triplicates of 2 g aliquots were stored at 80 C until processed. Total nucleic acids were extracted using a modified CTAB protocol previously described for the simultaneous recovery of DNA and RNA from soils (Hurt et al., 2001). Purification of DNA was done with AllPrep DNA/RNA Mini Kit (Qiagen) according to the manufacturer’s instructions. DNA extraction and quality was checked with a 0.8% agarose gel. DNA concentration was quantified with a NanoVue Plus Spectrophotometer (GE Healthcare).
2.3.
Sediment chemical analyses
The content of TC and TN in the sediment were analyzed by combustion of dried samples at 975 C in a Perkin Elmer AE SeriesII equipped with a TCD detector. The results were evaluated using the K factor method with cystine (C6H12N2O4S2) as a standard. Duplicates were always performed for all chemical determinations. Due to the low nitrogen content of the sediment, in most cases near the detection limit, total Kjeldahl nitrogen (TKN) was also
measured using standard methods for sediments (APHA, 1998). Sediment pH was measured from 1/5 (dry weight/ volume) sediment samples homogenized with double distilled water.
2.4.
Quantitative PCR of functional genes
Bacteria involved in the denitrification processes were estimated by the quantification of key functional genes using quantitative real-time PCR. The 16S rRNA gene was also quantified to estimate the total amount of bacteria. Primers and thermal cycling conditions used for each reaction have previously been described for narG and napA (Bru et al., 2007) and for nirS, nirK, nosZ and 16S rRNA genes (Hallin et al., 2009). All qPCR reactions were performed on a Bio-Rad IQ5 thermal cycler (Bio-Rad Laboratories, Inc.) in a total volume of 20 ml, containing 1X Phire Hot Start II DNA Polymerase (Finnzymes Oy. Espoo, Finland), 1 mM of each primer, 103 ng ml1 of Bovine Serum Albumin (BSA) and 1 ng of DNA template. Results were analyzed using Bio-Rad IQ5 software. Standard curves were obtained using serial dilutions from 102 to 108 copies of linearized plasmids containing the respective functional genes. Controls without templates gave null or negligible values. To ensure that sediment samples did not have inhibitory effects on PCR performance an inhibitory test were run with all samples at the working concentration together with a known amount of circular plasmid. The measured threshold cycle (Ct) values were compared with those of a control of the plasmid mixed with water. Despite the use of highly diluted DNA extractions (down to 1 ng), inhibition effects could not be removed from one sample of a Phragmites covered sediment collected in May 2008 and three samples from sediments covered with Typha collected in August 2008 and March 2009. These samples have been removed from the statistical analyses.
2.5.
Statistical analyses
All statistical analyses were performed using SPSS for Windows 15.0 (SPSS, Inc). Measured gene abundances were log transformed in order to ensure a normal distribution of
5624
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
data, which was checked by KolmogoroveSmirnov and ShapiroeWilk tests. Differences in gene abundances or gene ratios were tested for the effects of planted vs. non-planted sediments, sampling time, location of sampling plots and the corresponding interactions by one-way ANOVA and Welch tests. Differences between means of either gene were analyzed using GameseHowell test (for unequal variance data) post hoc analyses for multiple comparisons. To test for effects of sampling period, type of sediment, position inside the treatment cell and interactions on denitrifying community, univariate general linear model (GLM) analyses were performed. Correlation analysis of microbial abundances with physical and chemical variables was performed using nonparametric Spearman’s correlation test. Paired samples t-test was chosen to compare differences between genes coding for enzymes catalyzing the same reaction (i.e. narG vs. napA). The significance level for all tests was 0.05.
3.
Results
3.1.
Characterization of the Empuriabrava FWS-CWs
During the studied period, the water inflow to the wetlands presented stable pH conditions, with values around 7.6. BOD was low, usually under 3 mg O2/l, except for occasional dates when higher concentrations were recorded (Fig. 2). COD varied from 30 to 121 mg O2/l, but no seasonal trends were observed in the variation. Ammonia concentrations were low, with slightly higher values between July and September (up to 4.5 mg NeNHþ 4 /l). In contrast, nitrate concentrations showed a higher variation, between 0.1 and 15.8 mg NeNO 3 /l, with concentration peaks occurring intermittently throughout the year due to changes in nitrogen removal efficiencies in the WWTP. The water flowing through the wetlands had strong seasonal variations, from 1500 m3/day (February) to almost 6000 m3/day (August). This large variation is due to the intense tourism in a nearby located touristic area that increases water consumption during summer. Physicochemical parameters of the water and sediment measured at the sampling locations are presented in Table 1. Water conductivity values varied from 2.7 (August 2008) to 7.8 mV/cm (March 2009). There was a significant negative correlation between conductivity and water temperature (Spearman’s correlation coefficient r ¼ 0.641, p < 0.001, results not shown). The higher conductivity values during winter are due to the infiltration of sea water in some of the wastewater network collectors, which is diluted during summer due to a higher wastewater volume to be treated at the WWTP. The water flow to the wetland varied significantly during the four sampling periods, receiving 2437 and 2202 m3/ day in May 2008 and March 2009, and 6007 and 4379 m3/day in August 2008 and July 2009, respectively. Hydraulic retention times (HRT) for treatment cell 3 were calculated during the sampling periods by independent measurements of the water flow, with values ranging from 4.2 to 15.5 days. Unfortunately, the water flow was unexpectedly cut for a short period in July 2009 and these conditions are not indicative of the average situation in the cell. To circumvent this problem, the monthly
average water flow to the wetland system was used in the statistical analyses. Differences in both water and sediment pH were not extremely large and values ranged from 7.2 to 8.7 and from 6.9 to 7.9, respectively. Water nitrate and ammonium concentrations at the sampling locations ranged between 0.01 and þ 1.24 mg NeNO 3 /l and 0.06 and 0.49 mg NeNH4 /l, respectively. Nitrite concentrations were not detectable. Positive correlations were obtained between conductivity and nitrate (r ¼ 0.350, p < 0.01) or ammonium (r ¼ 0.275, p < 0.05). Higher TC and TKN values were usually obtained in sediments collected from areas planted with P. australis ( p < 0.001). Nitrogen removal efficiencies (%) were calculated on the basis on the total nitrogen load and changes in the inlet and outlet concentrations in treatment cell 3 (Garcı´a-Lledo´ et al., 2011). Efficiencies were estimated to be 48 and 61% in periods of low water discharge (May 2008 and March 2009, respectively) and increased up to 71% when higher inflows entered the CWS (July 2009). Nitrogen removal efficiencies correlated positively with ammonia concentration (r ¼ 0.685, p < 0.01) and negatively with nitrate concentrations (r ¼ 0.640, p < 0.01) at the influent. Moreover, oxygen and temperature showed a negative (r ¼ 0.484, p < 0.01) and positive (r ¼ 0.703, p < 0.01) correlation to nitrogen removal efficiencies, respectively.
3.2.
Abundance of 16S rRNA and denitrification genes
The abundance of 16S rRNA genes ranged from 1.3 1011 to 5.8 1012 copies/g dw sediment and appeared to be fairly stable among samples and over time (Fig. 3; Supplementary Table 1). These numbers were always higher than those obtained for any of the functional genes. Mean values for all determined functional gene abundances were between 1 and 4 logs below the 16S rRNA gene. The narG and napA genes ranged from 8.8 108 to 7.1 1010 and from 1.2 109 to 6.4 1010 copies/g dw sediment, respectively. Mean napA gene numbers were significantly higher than narG according to a paired sample t-test (t ¼ 2.578, p < 0.05), but these differences were exclusively due to dominance of napA genes in bulk sediments in March 2009 and no differences were detected when the March samples were excluded (t ¼ 1.116, p > 0.05). Sediment samples showed between 4.9 107 and 6.9 109 copies/g dw sediment for nirS and from 1.6 109 to 2.1 1011 copies/g dw sediment for nirK, suggesting a significantly greater abundance of nirK-type denitrifiers (t ¼ 28.828, p < 0.001). Finally, nosZ ranged from 3.5 107 to 1.3 109 copies/g dw sediment, and were significantly lower than both nirS and nirK (t ¼ 5.184, p < 0.001 and t ¼ 51.770, p < 0.001, respectively).
3.3. size
Factors determining the denitrifying community
Univariate general linear model tests (GLM) were used to analyze the effect of time (sampling period), type of sediment, sampling position in the treatment cell and their interactions, as defined factors to explain differences in gene abundances. Highly significant effects ( p < 0.001) were found when
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
5625
Fig. 2 e Time variation for the 2008e2009 period of the water flow, concentrations of BOD and COD and ammonia and nitrate at the inlet of the constructed wetland. Sediment sampling dates are indicated with arrows.
sampling period and types of sediment were considered (Table 3). The abundance of the nirS gene was the only exception and was not influenced by the presence of vegetation. Additionally, the sampling position only had an effect on 16S rRNA and narG ( p < 0.05). All possible interactions between factors had no significant effect on gene abundances, except for nitrate reductase genes in some cases. Nevertheless, partial Eta squared values (%) for these interactions were rather low, indicating a low significance in the overall variance of the considered genes. Accordingly, in subsequent analyses only the sampling time and the type of sediment were considered. In all cases, the lowest abundance of all analyzed genes was found in samples obtained during August 2008, except for
the nitrate reductase gene narG (Fig. 3AeC). In addition, the presence of vegetation, especially of Phragmites, yielded a significant increase for all genes but napA and nirS according to a non-parametric Welch test (F > 3.671, p < 0.05) (Fig. 3DeF). Differences in the abundance of napA could only be assessed by a post hoc analysis and resulted in higher values in sediments also covered with Phragmites. The presence of Typha did not cause any clear influence on the gene abundances and, depending on the genetic marker considered, was similar to either the bulk sediment or the one with Phragmites. In order to analyze differences in the gene abundances according to physicochemical characteristics of the sediment and water, pair-wise correlation tests were performed. nosZ abundance, being indicative of capacity for N2O reduction,
5626
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
Table 1 e Chemical properties of the sediment and overlying water of the different sampling locations and periods. Data from sediments correspond to mean values and standard deviations of three replicate samples. Type of sediment
Date
Water characteristics
Temp ( C) Cond (mS/cm) O2 (mg/l) pH
Sediment properties
NeNO 3
(mg/l)
NeNHþ 4
(mg/l)
pH
TC (%)
TKN (%)
SE inlet TY inlet PH inlet
15-May-08 21-May-08 20-May-08
20.7 20.4 20.0
3.8 3.6 3.5
6.1 7.6 8.8
7.6 8.0 8.0
0.45 1.07 1.24
0.37 0.21 0.21
7.4 0.1 4.0 0.2 0.21 0.02 7.7 0.1 5.6 1.1 0.30 0.06 7.4 0.1 13.5 1.9 1.23 0.08
SE inlet SE outlet TY inlet PH inlet PH outlet
20-Aug-08 11-Aug-08 21-Aug-08 18-Aug-08 12-Aug-08
24.8 25.9 25.3 24.9 26.6
2.7 2.8 2.8 3.1 2.9
4.3 7.8 5.4 5.2 3.2
7.7 8.0 7.9 7.9 8.1
0.08 0.01 0.12 0.21 0.01
0.21 0.08 0.27 0.12 0.06
7.7 0.0 7.7 0.1 7.9 0.2 7.9 0.0 7.8 0.1
3.8 0.4 3.3 0.4 4.2 0.7 4.3 0.3 5.3 0.2
0.16 0.09 0.18 0.23 0.32
0.02 0.02 0.10 0.04 0.07
SE inlet SE outlet TY inlet TY outlet PH inlet PH outlet
9-Mar-09 3-Mar-09 11-Mar-09 4-Mar-09 10-Mar-09 5-Mar-09
12.6 12.2 13.3 11.1 14.0 10.0
7.1 7.8 7.2 7.8 7.2 7.6
6.1 7.8 5.5 6.8 5.6 5.7
8.4 8.3 8.1 8.3 8.3 8.3
0.26 0.40 0.11 0.33 0.19 0.57
0.19 0.25 0.17 0.27 0.11 0.22
6.9 0.1 7.2 0.1 7.0 0.1 7.0 0.1 7.4 0.1 7.3 0.4
4.2 0.2 3.9 0.2 4.9 0.4 4.3 1.4 9.8 0.8 8.7 0.2
0.07 0.13 0.16 0.16 0.90 0.54
0.04 0.05 0.08 0.16 0.11 0.22
SE inlet SE outlet TY inlet TY outlet PH inlet PH outlet
8-Jul-09 7-Jul-09 13-Jul-09 14-Jul-09 10-Jul-09 16-Jul-09
22.6 25.5 26.4 26.1 22.4 27.6
4.5 4.7 4.6 4.7 4.4 4.5
3.6 3.9 14.9 10.2 3.0 3.8
7.3 7.7 8.7 8.3 7.7 8.2
0.18 0.08 0.03 0.03 0.03 0.03
0.32 0.49 0.28 0.15 0.25 0.11
7.7 0.2 5.7 1.0 7.2 0.1 3.8 0.1 7.6 0.3 5.0 1.3 7.2 0.1 4.8 0.9 7.1 0.2 12.0 2.9 7.5 0.2 9.0 0.7
0.47 0.18 0.29 0.29 1.04 0.83
0.19 0.04 0.22 0.18 0.67 0.21
SE, bulk sediment; TY, areas planted with Typha latifolia; PH, areas planted with Phragmites australis.
correlated positively with sediment TC (r ¼ 0.661, p < 0.01), TKN (r ¼ 0.612, p < 0.01) and negatively with the water flow and COD (r ¼ 0.330, p < 0.05; r ¼ 0.277, p < 0.05, respectively; Table 2). For the other genes, both sediment carbon and nitrogen content also correlated positively and significantly with abundances and negative correlations were observed when the water flow was considered, except for narG. The nitrate content in the water was also an important factor that specifically correlated with nitrite reductase genes, nirS and nirK, and the nitrate reductase napA. Surprisingly, no significant correlations were found for any of the genes and BOD in the influent water or sediment pH, with the exception of water pH and napA.
3.4.
Relative abundance of denitrification genes
The variation in the relative proportion of genes coding for enzymes catalyzing different steps in the denitrification can be used to assess removal efficiencies in terms of potential accumulation of intermediates in the denitrification pathway. Variation of calculated ratios according to time (sampling period), type of sediment, sampling position in the treatment cell and the corresponding interactions was estimated by GLM (Table 4). The relative contribution of the functional genes (narG, napA, nirS, nirK and nosZ ) to the total bacteria population, represented by 16S rRNA did not vary according to the presence of vegetation ( p > 0.05). When sampling period was considered, significant variations could be found for the relative amount of the functional genes narG, napA and nosZ ( p < 0.05) (results not shown). Interesting results can be deduced when ratios between two functional genes are calculated. The ratio between nitrate and nitrite reductases (qnarG þ qnapA)/(qnirS þ qnirK ) varied from 0.2 to 5.4 and was
significantly higher in samples obtained in August 2008 independently of the presence of vegetation or not ( p < 0.05). TC and TKN in the sediment were negatively correlated with (qnarG þ qnapA)/(qnirS þ qnirK ) (r ¼ 0.366, p < 0.01 and r ¼ 0.261, p < 0.05, respectively). The ratio (qnirS þ qnirK )/ qnosZ was significantly influenced by the sampling period ( p < 0.05) and the interaction between sampling period and type of vegetation. The highest difference between the abundance of nitrite reductases (qnirS þ qnirK ) and nitrous oxide reductase (qnosZ ) was found in March and May, when higher nitrate concentrations were measured, and accounted for almost two orders of magnitude. When qnirK/qnosZ and qnirS/ qnosZ ratios were analyzed separately, a significant effect of sampling time was found for both of them ( p < 0.05), but the interaction between sampling time and vegetation was only found significant for the former ( p < 0.05). Relevant variables affecting the ratio between nitrite reducers and nitrous oxide reducers were temperature (r ¼ 0.332, p < 0.05), nitrate content in water (r ¼ 0.314, p < 0.05) and total carbon in sediment (r ¼ 0.309, p < 0.05). Finally, the (qnarG þ qnapA)/ qnosZ ratio was fairly stable between the four sampling periods ( p > 0.05) and no effect of the a priori defined factors was detected.
4.
Discussion
The location of samples was a minor factor related to the abundance of denitrifying genes in the sediment of the treatment wetland. Instead, sampling period and type of sediment, mainly differing by nutrient availability, seems to be more important factors controlling the total abundance of denitrification genes in this system. Nevertheless, data from
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
5627
Fig. 3 e Mean values of the number of copies of the bacterial 16S rRNA and the functional genes, narG, napA, nirS, nirK and nosZ according to sampling period (AeC) and the type of sediment (DeF). Standard errors of the mean are indicated. Different letters above the bars indicate significant differences ( p < 0.05) between sampling period (left) or type of sediment (right).
qPCR must be examined carefully since bacteria may have more than one copy of the same gene per genome. Bacterial genomes can harbor up to 13 copies of the 16S rRNA gene (Fogel et al., 1999). The denitrification genes most often only exist in one copy, although two or three copies have been shown for all denitrification genes (Philippot, 2002; Jones et al., 2008, 2011). Despite these limitations, qPCR studies provide a realistic quantification of the size of the denitrification gene pool and ratios between gene pools can be used to infer community dynamics among samples of similar characteristics (Henry et al., 2006; Kandeler et al., 2006; Geets et al., 2007; Enwall et al., 2010). The relative abundance of functional gene densities in relation to the bacterial 16S rRNA gene abundance indicated
that the denitrifying and nitrate respiring bacteria in the sediment constituted a fraction of 1.6 0.4 and 1.2 0.7% of the bacterial communities, respectively. The values we obtained in the sediment of Empuriabrava FWS-CWs compared well with those found in the only wetland sediment characterized in terms of denitrifier abundances so far (Chon et al., 2011) and studies performed in soils (e.g. Henry et al., 2006; Kandeler et al., 2006; Hallin et al., 2009). The strong correlation between narG and napA genes and the fact that both have similar ratios to the 16S rRNA gene, suggests a high proportion of Proteobacteria in the sediment, since both narG and napA gene can be found in members of this phylum (Philippot and Hojberg, 1999; Roussel-Delif et al., 2005; Bru et al., 2007). Accordingly, quantitative analyses of bacteria at
5628
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
Table 2 e Spearman correlation coefficients (r) between environmental variables in the wetland sediment and water, influent water and abundances of the 16S rRNA and functional genes.
Wetland characteristics
Water Temp ( C) Cond (mS/cm) O2 (mg/l) RedOx (mV) pH NeNO 3 (mg N/l) NeNHþ 4 (mg N/l) Sediment TC (%) TKN (%) C:N pH
Influent characteristics
Water flow (m3/day) COD (mg O2/l) BOD (mg O2/l) NO 3 (mg N/l) NHþ 4 (mg N/l)
q16S rRNA
qnarG
qnapA
qnirS
qnirK
qnosZ
0.186 ns 0.356 ** 0.167 ns 0.114 ns 0.141 ns 0.308 * 0.179 ns
0.076 ns 0.174 ns 0.070 ns 0.043 ns 0.071 ns 0.205 ns 0.177 ns
0.282 * 0.407 ** 0.134 ns 0.002 ns 0.281 * 0.325 * 0.240 ns
0.230 ns 0.241 ns 0.102 ns 0.210 ns 0.073 ns 0.366 ** 0.200 ns
0.248 ns 0.387 ** 0.086 ns 0.124 ns 0.182 ns 0.351 ** 0.168 ns
0.139 ns 0.341 ** 0.113 ns 0.098 ns 0.195 ns 0.224 ns 0.155 ns
0.712 ** 0.653 ** 0.582 ** 0.251 ns 0.421 ** 0.376 ** 0.047 ns 0.243 ns 0.168 ns
0.544 ** 0.573 ** 0.546 ** 0.062 ns 0.160 ns 0.134 ns 0.078 ns 0.074 ns 0.120 ns
0.428 ** 0.395 ** 0.34 ** 0.219 ns 0.361 ** 0.254 ns 0.112 ns 0.285 * 0.235 ns
0.613 ** 0.539 ** 0.468 ** 0.143 ns 0.384 ** 0.346 ** 0.099 ns 0.282 * 0.123 ns
0.658 ** 0.600 ** 0.526 ** 0.234 ns 0.443 ** 0.356 ** 0.107 ns 0.310 * 0.230 ns
0.661 ** 0.612 ** 0.561 ** 0.225 ns 0.330 * 0.227 * 0.11 ns 0.182 ns 0.116 ns
ns, not significant; *p < 0.05; **p < 0.01.
the inlet and outlet of WWTPs in other studies show a clear dominance of Proteobacteria (Juretschko et al., 2002; Chouari et al., 2010; McLellan et al., 2010). The denitrifier community, assessed by the quantification of nirS and nirK genes, was dominated by NirK-type denitrifiers, which were up to 2 logs more abundant than the NirStype. This contrasts with previous studies in the same wetland using a PCR-TRFLP approach in which positive nirK PCR amplifications could not be obtained for any of the sediment samples (Ruiz-Rueda et al., 2007). This discrepancy is likely due to significant changes in the quality of the water entering the Empuriabrava FWS-CWs due to reconstructions in June 2007 (Garcı´a-Lledo´ et al., 2011). Even though the two nitrite reductases are functionally equivalent, denitrifiers harboring either nitrite reductase seem to show a preference for certain environments and are likely not under the same community assembly rules (Jones and Hallin, 2010). Recent studies in agricultural soils also suggested that the existence of the two types of nitrite reductase is due to differential niche preferences (Philippot et al., 2009; Enwall et al., 2010). However, the stability of the present physicochemical conditions of the influent to the Empuriabrava FWS-CWs and in the sediments prevents us from finding the environmental drivers
affecting the qnirS/qnirK ratio in this environment, other than that the sediment and water provided conditions that favor the NirK-type denitrifiers. In agreement with our results, the nosZ abundance in soils is often lower than that of other denitrifying genes (Henry et al., 2006; Hallin et al., 2009; Philippot et al., 2009). The nosZ gene encodes the final step of denitrification, which makes the net nitrogen removal from CWs all the way to N2 possible, but denitrifier communities with low nosZ to nir gene ratios can result in increased N2O/(N2 þ N2O) end-product ratios (Philippot et al., 2011). The lack of nitrous oxide reductase in some denitrifying bacteria was reported a decade ago when the Agrobacterium tumefaciens genome was sequenced (Wood et al., 2001) and recently it has been shown that approximately 1/3 of genome sequenced denitrifying bacterial isolates have this truncated pathway (Jones et al., 2008). The relatively low abundance of nosZ genes in the treatment wetland was especially pronounced in August 2008, coinciding with low nitrate levels and an increase in the nitrogen load in the form of ammonia to the FWS-CWs. In contrast to the other periods, the ammonium was also decreasing between the inlet and outlet of the different vegetation plots (Table 1). At the same time, the relative abundance of the
5629
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
Table 3 e Results of general linear model (GLM) analyses for the effect of the three factors established (sampling period, type of sediment and sampling location) and their interactions with the abundances of 16S rRNA and functional genes. Time
q16S rRNA qnarG qnapA qnirS qnirK qnosZ
Sed
F
p
Eta
F
p
Eta
F
p
Eta
23.21 4.27 6.76 11.13 14.42 10.48
*** * ** *** *** ***
0.65 0.25 0.35 0.47 0.53 0.45
16.49 11.93 10.04 2.87 10.96 14.52
*** *** *** ns *** ***
0.46 0.39 0.35 0.13 0.37 0.43
6.01 6.77 3.41 2.25 3.49 3.18
* * ns ns ns ns
0.14 0.15 0.08 0.06 0.08 0.08
Time Sed
q16S rRNA qnarG qnapA qnirS qnirK qnosZ
Loc
Time Loc
Sed Loc
Time Sed Loc
F
p
Eta
F
p
Eta
F
p
Eta
F
p
1.67 2.71 1.38 1.03 2.23 1.06
ns * ns ns ns ns
0.21 0.30 0.18 0.14 0.26 0.14
1.69 2.15 3.92 0.31 1.14 1.14
ns ns * ns ns ns
0.08 0.10 0.17 0.02 0.06 0.06
1.65 1.08 0.88 0.63 1.26 0.56
ns ns ns ns ns ns
0.08 0.05 0.04 0.03 0.06 0.03
1.11 2.69 3.00 0.84 0.27 2.54
Eta
ns ns * ns ns ns
0.08 0.18 0.19 0.06 0.02 0.17
ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001 Time, sampling period; Sed, type of sediment; Loc, sample location in the treatment cell. Eta, partial Eta Squared.
nitrate reductases, assessed by narG and napA genes, was higher compared to the other sampling occasions. These results suggest that the conditions favored bacteria capable of dissimilatory nitrate reduction to ammonium in addition to denitrifiers that do not have the complete denitrification pathway. Higher potential rates of nitrate reduction to ammonia were measured during August 2008 in bulk sediments and areas covered with P. australis thus confirming the previous results (Garcı´a-Lledo´ et al., 2011). Conditions that invariably will affect denitrification activity, such as temperature and lower nitrate to carbon ratios varied during the studied period. Relatively high nitrate loads to the Empuriabrava FWS-CW occurred intermittently
but frequently throughout the year. Although only two of these periods, May 2008 and March 2009, were sampled in this study, the increase in the measured gene abundances let us hypothesize that the probability of N2O production in the Empuriabrava FWS-CWs would increase in relation to the nitrate concentration. However, additional measurements in summer coinciding with the increase of nitrate at the inlet would be needed to confirm this hypothesis. Other studies conducted on CWs mesocosms have reported significantly higher N2O emissions during summer (Sovik and Klove, 2007). In the Empuriabrava FWS-CWs significant differences of the genetic potential for N2O emissions according to plant species were only detected when interactions with sampling
Table 4 e Results of general linear model (GLM) analyses for the effect of the three factors established (sampling period, type of sediment and sampling location) and their interactions with the ratios between the functional genes involved in denitrification process. Time
(qnarG þ qnapA)/(qnirS þ qnirK ) (qnirS þ qnirK )/qnosZ qnirS/qnirK qnirS/qnosZ qnirK/qnosZ (qnarG þ qnapA)/qnosZ
Sed
F
p
Eta
F
p
Eta
F
p
Eta
3.64 3.81 4.16 4.14 3.76 1.17
* * * * * ns
0.23 0.24 0.25 0.25 0.23 0.09
1.06 0.62 4.79 1.57 0.73 0.31
ns ns * ns ns ns
0.05 0.03 0.21 0.08 0.04 0.17
0.15 0.27 0.00 0.18 0.29 1.08
ns ns ns ns ns ns
0.00 0.01 0.00 0.01 0.01 0.03
Time Sed
(qnarG þ qnapA)/(qnirS þ qnirK ) (qnirS þ qnirK)/qnosZ qnirS/qnirK qnirS/qnosZ qnirK/qnosZ (qnarG þ qnapA)/qnosZ
Loc
Time Loc
Sed Loc
Time Sed Loc
F
p
Eta
F
p
Eta
F
p
Eta
F
1.91 2.88 0.88 1.55 2.91 1.29
ns * ns ns * ns
0.24 0.32 0.13 0.20 0.32 0.17
1.12 0.11 1.85 0.34 0.17 2.82
ns ns ns ns ns ns
0.06 0.01 0.09 0.02 0.01 0.13
1.15 2.09 0.29 0.68 2.17 0.27
ns ns ns ns ns ns
0.06 0.10 0.02 0.04 0.11 0.01
1.74 1.68 2.22 0.50 1.83 0.29
p ns ns ns ns ns ns
Eta 0.12 0.12 0.15 0.04 0.13 0.02
ns, not significant; *, p < 0.05; Time, sampling period; Sed, type of sediment; Loc, sample location in the treatment cell. Eta, partial Eta Squared.
5630
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
time were considered. However, according to the calculated (qnirS þ qnirK )/qnosZ ratio, the highest genetic potential would be found mainly in vegetated sediments where higher nitrogen availability occurred due to leaf litter decomposition. Others have measured higher N2O fluxes in mesocosms planted with P. australis when compared to other macrophytes and unvegetated sediments (Maltais-Landry et al., 2009). Nevertheless, the effect of planted areas for the production of N2O in wetlands has been poorly studied, and further research in this direction is still needed to fully understand the dynamics of constructed wetlands in relation to greenhouse gas emissions.
5.
Conclusions
The Empuriabrava FWS-CWs showed variations in physicochemical parameters analyzed both in water and sediment between the different periods. These conditions favored the maintenance of a denitrifying community that significantly changed according to nutrient concentration and the types of vegetation. The low abundance of nosZ genes compared with the other denitrification genes is an indicative of the genetic capacity of the system to potentially accumulate the N2O intermediary and the relative proportion of nosZ genes decreased during periods of high nitrate content to the wetlands. Overall, the quantitative data on denitrification gene abundances provide evidence of a high potential for nitrous oxide emissions along the entire sediment surface and in particular during periods of high nitrate loading. The vegetation effect was mainly detected in combination with sampling time and resulted in an increase of the potential for nitrous oxide emissions in vegetated areas. The increase in the nitrite to nitrous oxide reductase genes ratio has been related to the higher total carbon content in these sediments.
Acknowledgments A.G-L and A.V-S. are recipients of pre-doctoral grants from the Ministerio de Ciencia y Educacio´n and the Universitat de Girona, respectively. The authors thank the contributions of Anna Huguet, Jordi Sala and Lluı´s Sala for field analyses. This research has been funded by the Spanish Ministerio de Ciencia y Educacio´n (grant CGL2009-08338).
Appendix. Supplementary material Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.watres.2011.08.025.
references
Abell, G.C., Revill, A.T., Smith, C., Bissett, A.P., Volkman, J.K., Robert, S.S., 2010. Archaeal ammonia oxidizers and nirS-type denitrifiers dominate sediment nitrifying and denitrifying
populations in a subtropical macrotidal estuary. The ISME Journal 4 (2), 286e300. APHA, A.a.W., 1998. Standard Methods for the Examination of Water and Wastewater. American Public Health Association, Washington. Boudraa, A.O., Champier, J., Djebali, M., Behloul, F., Beghdadi, A., Bachand, P.A.M., Horne, A.J., 1999. Denitrification in constructed free-water surface wetlands: II. Effects of vegetation and temperature. Ecological Engineering 14, 17e32. Brix, H., 1997. Do macrophytes play a role in constructed treatment wetlands? Water Science and Technology 35, 11e17. Bru, D., Sarr, A., Philippot, L., 2007. Relative abundances of proteobacterial membrane-bound and periplasmic nitrate reductases in selected environments. Applied and Environmental Microbiology 73 (18), 5971e5974. Carter, J.P., Hsaio, Y.H., Spiro, S., Richardson, D.J., 1995. Soil and sediment bacteria capable of aerobic nitrate respiration. Applied and Environmental Microbiology 61 (8), 2852e2858. Cheneby, D., Perrez, S., Devroe, C., Hallet, S., Couton, Y., Bizouard, F., Iuretig, G., Germon, J.C., Philippot, L., 2004. Denitrifying bacteria in bulk and maize-rhizospheric soil: diversity and N2O-reducing abilities. Canadian Journal of Microbiology 50 (7), 469e474. Chon, K., Chang, J.-S., Lee, E., Lee, J., Ryu, J., Cho, J., 2011. Abundance of denitrifying genes coding for nitrate (narG), nitrite (nirS), and nitrous oxide (nosZ) reductases in estuarine versus wastewater effluent-fed constructed wetlands. Ecological Engineering 37 (1), 64e69. Chouari, R., Le Paslier, D., Daegelen, P., Dauga, C., Weissenbach, J., Sghir, A., 2010. Molecular analyses of the microbial community composition of an anoxic basin of a municipal wastewater treatment plant reveal a novel lineage of proteobacteria. Microbial Ecology 60 (2), 272e281. Coyne, M.S., Arunakumari, A., Averill, B.A., Tiedje, J.M., 1989. Immunological identification and distribution of dissimilatory heme cd1 and nonheme copper nitrite reductases in denitrifying bacteria. Applied and Environmental Microbiology 55 (11), 2924e2931. DeBusk, T., DeBusk, W., 2000. In: Kent, D.M. (Ed.), Applied Wetlands Science and Technology, second ed. CRC Press, Boca Raton, Florida (USA), pp. 241e279. Enwall, K., Throback, I.N., Stenberg, M., Soderstrom, M., Hallin, S., 2010. Soil resources influence spatial patterns of denitrifying communities at scales compatible with land management. Applied and Environmental Microbiology 76 (7), 2243e2250. EPA, 2000. Guiding Principles for Constructed Treatment Wetlands Providing for Water Quality and Wildlife Habitat. United States Environmental Protection Agency, Office of Water. Fogel, G.B., Collins, C.R., Li, J., Brunk, C.F., 1999. Prokaryotic genome size and SSU rDNA copy number: estimation of microbial relative abundance from a mixed population. Microbial Ecology 38 (2), 93e113. Francis, C.A., Beman, J.M., Kuypers, M.M., 2007. New processes and players in the nitrogen cycle: the microbial ecology of anaerobic and archaeal ammonia oxidation. The ISME Journal 1 (1), 19e27. Garcı´a-Lledo´, A., Ruiz-Rueda, O., Vilar-Sanz, A., Sala, L., Ban˜eras, L., 2011. Nitrogen removal efficiencies in a free water surface constructed wetland in relation to plant coverage. Ecological Engineering 37, 678e684. Geets, J., de Cooman, M., Wittebolle, L., Heylen, K., Vanparys, B., De Vos, P., Verstraete, W., Boon, N., 2007. Real-time PCR assay for the simultaneous quantification of nitrifying and denitrifying bacteria in activated sludge. Applied Microbiology and Biotechnology 75 (1), 211e221. Gersberg, R.M., Elkins, B.V., Lyon, S.R., Goldman, C.R., 1986. Role of aquatic plants in wastewater treatment by artificial wetlands. Water Research 20 (3), 363e368.
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
Gregory, L.G., Bond, P.L., Richardson, D.J., Spiro, S., 2003. Characterization of a nitrate-respiring bacterial community using the nitrate reductase gene (narG) as a functional marker. Microbiology 149 (1), 229e237. Hallin, S., Jones, C.M., Schloter, M., Philippot, L., 2009. Relationship between N-cycling communities and ecosystem functioning in a 50-year-old fertilization experiment. The ISME Journal 3 (5), 597e605. Henry, S., Bru, D., Stres, B., Hallet, S., Philippot, L., 2006. Quantitative detection of the nosZ gene, encoding nitrous oxide reductase, and comparison of the abundances of 16S rRNA, narG, nirK, and nosZ genes in soils. Applied and Environmental Microbiology 72 (8), 5181e5189. Henry, S., Texier, S., Hallet, S., Bru, D., Dambreville, C., Cheneby, D., Bizouard, F., Germon, J.C., Philippot, L., 2008. Disentangling the rhizosphere effect on nitrate reducers and denitrifiers: insight into the role of root exudates. Environmental Microbiology 10 (11), 3082e3092. Hurt, R.A., Qiu, X., Wu, L., Roh, Y., Palumbo, A.V., Tiedje, J.M., Zhou, J., 2001. Simultaneous recovery of RNA and DNA from soils and sediments. Applied and Environmental Microbiology 67 (10), 4495e4503. Ibekwe, A.M., Lyon, S.R., Leddy, M., Jacobson-Meyers, M., 2007. Impact of plant density and microbial composition on water quality from a free water surface constructed wetland. Journal of Applied Microbiology 102 (4), 921e936. Ingersoll, T.L., Baker, L.A., 1998. Nitrate removal in wetland microcosms. Water Research 32 (3), 677e684. Johnston, C.A., 1991. Sediment and nutrient retention by freshwater wetlands: effects on surface water quality. Critical Reviews in Environmental Control 21 (5), 491e565. Jones, C.M., Hallin, S., 2010. Ecological and evolutionary factors underlying global and local assembly of denitrifier communities. The ISME Journal 4 (5), 633e641. Jones, C.M., Stres, B., Rosenquist, M., Hallin, S., 2008. Phylogenetic analysis of nitrite, nitric oxide, and nitrous oxide respiratory enzymes reveal a complex evolutionary history for denitrification. Molecular Biology and Evolution 25 (9), 1955e1966. Jones, C.M., Welsh, A., Throback, I.N., Dorsch, P., Bakken, L.R., Hallin, S., 2011. Phenotypic and genotypic heterogeneity among closely related soil-borne N2- and N2O-producing Bacillus isolates harboring the nosZ gene. FEMS Microbiology Ecology 76 (3), 541e552. Juretschko, S., Loy, A., Lehner, A., Wagner, M., 2002. The microbial community composition of a nitrifyingedenitrifying activated sludge from an industrial sewage treatment plant analyzed by the full-cycle rRNA approach. Systematic and Applied Microbiology 25 (1), 84e99. Kandeler, E., Deiglmayr, K., Tscherko, D., Bru, D., Philippot, L., 2006. Abundance of narG, nirS, nirK, and nosZ genes of denitrifying bacteria during primary successions of a glacier foreland. Applied and Environmental Microbiology 72 (9), 5957e5962. Kjellin, J., Hallin, S., Worman, A., 2007. Spatial variations in denitrification activity in wetland sediments explained by hydrology and denitrifying community structure. Water Research 41 (20), 4710e4720. Ko¨rner, S., 1999. Nitrifying and denitrifying bacteria in epiphytic communities of submerged macrophytes in a treated sewage channel. Acta hydrochimica et hydrobiologica 27 (1), 27e31. Lin, Y.F., Jing, S.R., Wang, T.W., Lee, D.Y., 2002. Effects of macrophytes and external carbon sources on nitrate removal from groundwater in constructed wetlands. Environmental Pollution 119 (3), 413e420. Maltais-Landry, G., Maranger, R., Brisson, J., Chazarenc, F., 2009. Greenhouse gas production and efficiency of planted and artificially aerated constructed wetlands. Environmental Pollution 157 (3), 748e754.
5631
McLellan, S.L., Huse, S.M., Mueller-Spitz, S.R., Andreishcheva, E.N., Sogin, M.L., 2010. Diversity and population structure of sewage-derived microorganisms in wastewater treatment plant influent. Environmental Microbiology 12 (2), 378e392. Moreno-Vivian, C., Cabello, P., Martinez-Luque, M., Blasco, R., Castillo, F., 1999. Prokaryotic nitrate reduction: molecular properties and functional distinction among bacterial nitrate reductases. Journal of Bacteriology 181 (21), 6573e6584. Nguyen, C., 2003. Rhizodeposition of organic C by plants: mechanisms and controls. Agronomie 23 (5e6), 375e396. Philippot, L., 2002. Denitrifying genes in bacterial and Archaeal genomes. Biochimica et Biophysica Acta 1577 (3), 355e376. Philippot, L., Andert, J., Jones, C.M., Bru, D., Hallin, S., 2011. Importance of denitrifiers lacking the genes encoding the nitrous oxide reductase for N2O emissions from soil. Global Change Microbiology 17, 1497e1504. Philippot, L., Cuhel, J., Saby, N.P., Cheneby, D., Chronakova, A., Bru, D., Arrouays, D., Martin-Laurent, F., Simek, M., 2009. Mapping fieldscale spatial patterns of size and activity of the denitrifier community. Environmental Microbiology 11 (6), 1518e1526. Philippot, L., Hojberg, O., 1999. Dissimilatory nitrate reductases in bacteria. Biochimica et Biophysica Acta 1446 (1e2), 1e23. Prade, K., Trolldenier, G., 1990. Denitrification in the rhizosphere of rice and wheat seedlings as influenced by plant K status, air-filled porosity and substrate organic matter. Soil Biology and Biochemistry 22 (6), 769e773. Reed, S.C., Crites, R.W., Middlebrooks, E.J., 1995. Natural Systems for Waste Management and Treatment. McGraw-Hill, New York. Roussel-Delif, L., Tarnawski, S., Hamelin, J., Philippot, L., Aragno, M., Fromin, N., 2005. Frequency and diversity of nitrate reductase genes among nitrate-dissimilating Pseudomonas in the rhizosphere of perennial grasses grown in field conditions. Microbial Ecology 49 (1), 63e72. Ruiz-Rueda, O., Hallin, S., Baneras, L., 2009. Structure and function of denitrifying and nitrifying bacterial communities in relation to the plant species in a constructed wetland. FEMS Microbiology Ecology 67 (2), 308e319. Ruiz-Rueda, O., Trias, R., Garcia-Gil, L.J., Baneras, L., 2007. Diversity of the nitrite reductase gene nirS in the sediment of a free-water surface constructed wetland. International Microbiology 10 (4), 253e260. Sovik, A.K., Klove, B., 2007. Emission of N2O and CH4 from a constructed wetland in southeastern Norway. Science of the Total Environment 380 (1e3), 28e37. Tiedje, J., 1988. Ecology of denitrification and dissimilatory nitrate reduction to ammonium. In: Zehnder, A.J.B. (Ed.), Biology of Anaerobic Microorganisms. Wiley, New York, pp. 179e244. Toet, S., Bouwman, M., Cevaal, A., Verhoeven, J.T., 2005. Nutrient removal through autumn harvest of Phragmites australis and Thypha latifolia shoots in relation to nutrient loading in a wetland system used for polishing sewage treatment plant effluent. Journal of Environmental Science and Health, Part A 40 (6), 1133e1156. Vial, L., Lavire, C., Mavingui, P., Blaha, D., Haurat, J., MoenneLoccoz, Y., Bally, R., Wisniewski-Dye, F., 2006. Phase variation and genomic architecture changes in Azospirillum. Journal of Bacteriology 188 (15), 5364e5373. Vymazal, J., 2001. Types of constructed wetlands for wastewater treatment: their potential for nutrient removal. In: Vymazal, J. (Ed.), Transformations of Nutrients in Natural and Constructed Wetlands. Backhuys Publishers, Leiden, The Netherlands, pp. 1e93. Whitney, D., Rossman, A., Hayden, N., 2003. Evaluating an existing subsurface flow constructed wetland in Akumal, Mexico. Ecological Engineering 20 (1), 105e111. Wood, D.W., Setubal, J.C., Kaul, R., Monks, D.E., Kitajima, J.P., Okura, V.K., Zhou, Y., Chen, L., Wood, G.E., Almeida Jr., N.F.,
5632
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
Woo, L., Chen, Y., Paulsen, I.T., Eisen, J.A., Karp, P.D., Bovee Sr., D., Chapman, P., Clendenning, J., Deatherage, G., Gillet, W., Grant, C., Kutyavin, T., Levy, R., Li, M.J., McClelland, E., Palmieri, A., Raymond, C., Rouse, G., Saenphimmachak, C., Wu, Z., Romero, P., Gordon, D., Zhang, S., Yoo, H., Tao, Y., Biddle, P., Jung, M., Krespan, W., Perry, M., Gordon-Kamm, B., Liao, L., Kim, S., Hendrick, C., Zhao, Z.Y., Dolan, M., Chumley, F., Tingey, S.V., Tomb, J.F., Gordon, M.P., Olson, M.V., Nester, E.W., 2001. The
genome of the natural genetic engineer Agrobacterium tumefaciens C58. Science 294 (5550), 2317e2323. Zhu, T., Sikora, F.J., 1995. Ammonium and nitrate removal in vegetated and unvegetated gravel bed microcosm wetlands. Water Science and Technology 32 (3), 219e228. Zumft, W.G., 1997. Cell biology and molecular basis of denitrification. Microbiology and Molecular Biology Reviews 61 (4), 533e616.
Available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/watres
Genetic potential for N2O emissions from the sediment of a free water surface constructed wetland Arantzazu Garcı´a-Lledo´ a, Ariadna Vilar-Sanz a, Rosalia Trias a, Sara Hallin b, Lluı´s Ban˜eras a,* a b
Molecular Microbial Ecology Group, Institute of Aquatic Ecology, Universitat de Girona, C/ Maria Aure`lia Capmany, 69, 17071 Girona, Spain Swedish University of Agricultural Sciences, Department of Microbiology, Box 7025, Uppsala 750 07, Sweden
article info
abstract
Article history:
Removal of nitrogen is a key aspect in the functioning of constructed wetlands. However,
Received 13 April 2011
incomplete denitrification may result in the net emission of the greenhouse gas nitrous
Received in revised form
oxide (N2O) resulting in an undesired effect of a system supposed to provide an ecosystem
22 July 2011
service. In this work we evaluated the genetic potential for N2O emissions in relation to the
Accepted 14 August 2011
presence or absence of Phragmites and Typha in a free water surface constructed wetland
Available online 30 August 2011
(FWS-CW), since vegetation, through the increase in organic matter due to litter degradation, may significantly affect the denitrification capacity in planted areas. Quantitative
Keywords:
real-time PCR analyses of genes in the denitrification pathway indicating capacity to
Constructed wetlands
produce or reduce N2O were conducted at periods of different water discharge. Genetic
Denitrification
potential for N2O emissions was estimated from the relative abundances of all denitrifi-
Genetic potential
cation genes and nitrous oxide reductase encoding genes (nosZ ). nosZ abundance was
N2O emission
invariably lower than the other denitrifying genes (down to 100 fold), and differences
qPCR
increased significantly during periods of high nitrate loads in the CW suggesting a higher
Vegetation
genetic potential for N2O emissions. This situation coincided with lower nitrogen removal efficiencies in the treatment cell. The presence and the type of vegetation, mainly due to changes in the sediment carbon and nitrogen content, correlated negatively to the ratio between nitrate and nitrite reducers and positively to the ratio between nitrite and nitrous oxide reducers. These results suggest that the potential for nitrous oxide emissions is higher in vegetated sediments. ª 2011 Elsevier Ltd. All rights reserved.
1.
Introduction
Constructed wetlands (CWs) are recognized as feasible alternatives for removal of nitrogen in wastewater from agricultural, industrial or municipal activities and have been exploited as secondary or tertiary treatment alternatives to promote water reuse (Reed et al., 1995; DeBusk and DeBusk, 2000). There are two types of CWs, the subsurface flow systems (SSF) and the free-water surface systems (FWS) that
mainly differ in the presence of a free water flow over the sediment surface. Water treatment in CWs involves processes driven by the sediment, the vegetation and its associated microbial communities (EPA, 2000). It is generally assumed that areas planted with emergent macrophytes positively affect the water restoration capacity of the wetland (Zhu and Sikora, 1995; Lin et al., 2002; Ibekwe et al., 2007). This is due to stimulation of microbial growth at either the epiphyton or the root surface, e.g. by influencing the oxygen conditions and
* Corresponding author. Tel.: þ34 972 418 177; fax: þ34 972 418 150. E-mail addresses: [email protected] (A. Garcı´a-Lledo´), [email protected] (A. Vilar-Sanz), [email protected] (R. Trias), [email protected] (S. Hallin), [email protected] (L. Ban˜eras). 0043-1354/$ e see front matter ª 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.watres.2011.08.025
5622
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
creating microenvironments with oxiceanoxic zones (Gersberg et al., 1986; Zhu and Sikora, 1995; Boudraa et al., 1999) or by the excretion of organic compounds from the roots (Prade and Trolldenier, 1990; Brix, 1997; Nguyen, 2003), as well as causing changes in hydraulic retention time (Whitney et al., 2003; Toet et al., 2005; Kjellin et al., 2007). All these processes influence nitrogen removal in CWs since they affect the nitrifying and denitrifying microorganisms, which are ultimately responsible for nitrogen transformations resulting in nitrogen removal (Johnston, 1991; Zhu and Sikora, 1995; Boudraa et al., 1999; Ko¨rner, 1999; Vymazal, 2001; Lin et al., 2002; Francis et al., 2007). The microbial process denitrification is especially interesting in CWs as it represents the net nitrogen loss from the system. Denitrifying bacteria use nitrate as an alternative electron acceptor, which is sequentially reduced to nitrogen gas. The first step is catalyzed by nitrate reductases, either a membrane-bound enzyme (Nar type) or a soluble periplasmic enzyme (Nap) (Zumft, 1997; Moreno-Vivian et al., 1999; Philippot, 2002). These two reductases are not exclusive, but can be found simultaneously in the same organism (Carter et al., 1995; Gregory et al., 2003; Roussel-Delif et al., 2005). Dissimilatory nitrite reductases (Nir) catalyze the second step of denitrification and two functionally equivalent enzymes have been described, a cytochrome cd1 type and a copper containing type, encoded by the nirS and nirK genes, respectively. They are mutually exclusive and have not been found in the same strain so far (Zumft, 1997), although different strains of the same species may contain different Nir genes (Coyne et al., 1989; Philippot, 2002). Nitric oxide reductases (Nor) catalyze the reduction of nitric oxide to nitrous oxide and the last step of denitrification is catalyzed by nitrous oxide reductases (Nos) which lead to the production of nitrogen gas (Zumft, 1997). All denitrifying genes described so far have been used as molecular markers for qualitative and quantitative studies of denitrifying bacteria in the environment. The expression of denitrifying genes is dependent on the presence of the respective enzymes substrates, i.e. sequential oxidized forms of inorganic nitrogen, and low oxygen concentration due to the facultative nature of this process (Tiedje, 1988). Most denitrifiers use organic compounds as electron donors and easily available carbon is another prerequisite. Emissions of the potent greenhouse gas N2O may be due to the activity of some denitrifying bacteria that are unable to perform the final step of denitrification, due to the lack of the nosZ gene (Wood et al., 2001; Kandeler et al., 2006; Vial et al., 2006; Jones et al., 2008; Abell et al., 2010; Philippot et al., 2011). Cheneby et al., 2004 compared denitrifying bacteria between non-planted and maize planted soil and found that nosZ lacking bacteria were dominant in the rhizosphere suggesting that plants can affect N2O emission by selecting for denitrifiers that do not have the capacity to reduce N2O. Plants increase easily available carbon compounds by root exudates and through the increase in organic matter due to litter degradation. In addition, plants cause radial oxygen gradients around the roots. These factors affect denitrifying bacteria, as well as other heterotrophic bacteria (Boudraa et al., 1999; Henry et al., 2008). Studies specifically directed to the analysis of denitrification in the rhizosphere have shown that changes in the activity, composition and the abundance of
denitrifying bacteria may be plant-specific (Cheneby et al., 2004; Henry et al., 2008; Ruiz-Rueda et al., 2009). In the sediment of planted areas the presence of decomposing litter may be an additional factor supporting and shaping the denitrifying community (Ingersoll and Baker, 1998). If emergent macrophytes in CWs also select for denitrifiers genetically incapable of reducing N2O is not known. If so, although promoting nitrogen removal in CWs, macrophytes potentially contribute to increase the N2O/N2 end-product ratio. We evaluated the genetic potential for N2O emissions of the sediment in relation to the presence or absence of the main plant species used in treatment wetlands, Phragmites australis and Typha latifolia. For that purpose, we determined the abundance of denitrifiers by quantitative real-time PCR (qPCR) of key functional genes in the denitrification pathway (narG, napA, nirS, nirK and nosZ ) in comparison to the total bacterial community size targeted by the 16S rRNA gene in the sediments of the Empuriabrava FWS-CWs at periods of different water discharge over the year.
2.
Material and methods
2.1.
Study site
The Empuriabrava FWS-CWs (Girona, NE Spain) were designed in 1998 as a tertiary treatment to increase the water quality of the effluent of a nearby located wastewater treatment plant (WWTP). The CWs are included in the natural preserved area of Els Aiguamolls de l’Emporda` and are designed to provide additional water to avoid excessive desiccation of the flooded area in summer (Ruiz-Rueda et al., 2007). Nitrogen is removed at the WWTP by a combined nitrificationedenitrification process using carrousel-type bioreactors and enhanced aeration, resulting in nitrate loads to the CWs changing between 251 and 1016 kg N per month. The Empuriabrava CWs consists of three parallel cells followed by a shallower lagoon and all measurements were conducted on treatment cell 3. The cell is planted with independent communities of reed (P. australis) and cattail (T. latifolia) showing an almost equivalent distribution of both plant species (49.3% of reed and 50.7% of cattail; Fig. 1). The average water depth in the cell was about 0.6 m and the sediment was water saturated during the sampling period. The sediment depths varied between 5 and 10 cm in non-vegetated areas and increased to 15e20 cm in planted surfaces. The location of the vegetation spots and the plant species were designed in the original project and planted shortly after construction of the CW (1998). Harvest of the aerial biomass is performed every one or two years during winter.
2.2.
Sampling and DNA extraction
Temperature, conductivity, oxygen and pH at the sampling locations were measured with a portable multiparametric probe (Yellow Spring Instruments 650MDS). Water samples (20 ml) from the water column in the wetland were collected and analyzed for nitrate, nitrite and ammonia concentration as described previously (Garcı´a-Lledo´ et al., 2011). Water samples (100 ml) were regularly also collected at the
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
5623
Fig. 1 e Scheme of vegetation stands and sampling locations in the third cell of the Empuriabrava FWS-CWs. Sampling plots are indicated as black dots with the following two-letter codes: SE e bulk sediment, TY e areas planted with Typha latifolia, PH e areas planted with Phragmites australis.
secondary effluent of the WWTP and at the end point of the treatment cells for analyses of pH, conductivity, biological oxygen demand (BOD), chemical oxygen demand (COD), oxygen saturation, and concentrations of nitrate, nitrite and ammonium, using conventional standard methods for wastewater analyses (APHA, 1998). Total water flow into the wetland system was automatically monitored daily. Sediment samples were collected May 2008, August 2008, March 2009 and July 2009 near the inlet and outlet areas. Three sampling plots of one square meter each were defined at the two areas. Sampling plot 1 was not covered with any emergent plant and is referred as bulk sediment (SE). Plot 2 consisted of an area covered with T. latifolia (TY) and Plot 3 was covered with P. australis (PH). Three sediment replicates were taken in each plot (Fig. 1). Sediment cores were obtained from planted and unplanted areas using a 2-cm-diameter methacrylate tube mounted in a manual sampler. The upper 3 cm of the sediment core were aseptically transferred to a container and chilled on ice for transportation. Once in the laboratory, sediment was completely homogenized and triplicates of 2 g aliquots were stored at 80 C until processed. Total nucleic acids were extracted using a modified CTAB protocol previously described for the simultaneous recovery of DNA and RNA from soils (Hurt et al., 2001). Purification of DNA was done with AllPrep DNA/RNA Mini Kit (Qiagen) according to the manufacturer’s instructions. DNA extraction and quality was checked with a 0.8% agarose gel. DNA concentration was quantified with a NanoVue Plus Spectrophotometer (GE Healthcare).
2.3.
Sediment chemical analyses
The content of TC and TN in the sediment were analyzed by combustion of dried samples at 975 C in a Perkin Elmer AE SeriesII equipped with a TCD detector. The results were evaluated using the K factor method with cystine (C6H12N2O4S2) as a standard. Duplicates were always performed for all chemical determinations. Due to the low nitrogen content of the sediment, in most cases near the detection limit, total Kjeldahl nitrogen (TKN) was also
measured using standard methods for sediments (APHA, 1998). Sediment pH was measured from 1/5 (dry weight/ volume) sediment samples homogenized with double distilled water.
2.4.
Quantitative PCR of functional genes
Bacteria involved in the denitrification processes were estimated by the quantification of key functional genes using quantitative real-time PCR. The 16S rRNA gene was also quantified to estimate the total amount of bacteria. Primers and thermal cycling conditions used for each reaction have previously been described for narG and napA (Bru et al., 2007) and for nirS, nirK, nosZ and 16S rRNA genes (Hallin et al., 2009). All qPCR reactions were performed on a Bio-Rad IQ5 thermal cycler (Bio-Rad Laboratories, Inc.) in a total volume of 20 ml, containing 1X Phire Hot Start II DNA Polymerase (Finnzymes Oy. Espoo, Finland), 1 mM of each primer, 103 ng ml1 of Bovine Serum Albumin (BSA) and 1 ng of DNA template. Results were analyzed using Bio-Rad IQ5 software. Standard curves were obtained using serial dilutions from 102 to 108 copies of linearized plasmids containing the respective functional genes. Controls without templates gave null or negligible values. To ensure that sediment samples did not have inhibitory effects on PCR performance an inhibitory test were run with all samples at the working concentration together with a known amount of circular plasmid. The measured threshold cycle (Ct) values were compared with those of a control of the plasmid mixed with water. Despite the use of highly diluted DNA extractions (down to 1 ng), inhibition effects could not be removed from one sample of a Phragmites covered sediment collected in May 2008 and three samples from sediments covered with Typha collected in August 2008 and March 2009. These samples have been removed from the statistical analyses.
2.5.
Statistical analyses
All statistical analyses were performed using SPSS for Windows 15.0 (SPSS, Inc). Measured gene abundances were log transformed in order to ensure a normal distribution of
5624
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
data, which was checked by KolmogoroveSmirnov and ShapiroeWilk tests. Differences in gene abundances or gene ratios were tested for the effects of planted vs. non-planted sediments, sampling time, location of sampling plots and the corresponding interactions by one-way ANOVA and Welch tests. Differences between means of either gene were analyzed using GameseHowell test (for unequal variance data) post hoc analyses for multiple comparisons. To test for effects of sampling period, type of sediment, position inside the treatment cell and interactions on denitrifying community, univariate general linear model (GLM) analyses were performed. Correlation analysis of microbial abundances with physical and chemical variables was performed using nonparametric Spearman’s correlation test. Paired samples t-test was chosen to compare differences between genes coding for enzymes catalyzing the same reaction (i.e. narG vs. napA). The significance level for all tests was 0.05.
3.
Results
3.1.
Characterization of the Empuriabrava FWS-CWs
During the studied period, the water inflow to the wetlands presented stable pH conditions, with values around 7.6. BOD was low, usually under 3 mg O2/l, except for occasional dates when higher concentrations were recorded (Fig. 2). COD varied from 30 to 121 mg O2/l, but no seasonal trends were observed in the variation. Ammonia concentrations were low, with slightly higher values between July and September (up to 4.5 mg NeNHþ 4 /l). In contrast, nitrate concentrations showed a higher variation, between 0.1 and 15.8 mg NeNO 3 /l, with concentration peaks occurring intermittently throughout the year due to changes in nitrogen removal efficiencies in the WWTP. The water flowing through the wetlands had strong seasonal variations, from 1500 m3/day (February) to almost 6000 m3/day (August). This large variation is due to the intense tourism in a nearby located touristic area that increases water consumption during summer. Physicochemical parameters of the water and sediment measured at the sampling locations are presented in Table 1. Water conductivity values varied from 2.7 (August 2008) to 7.8 mV/cm (March 2009). There was a significant negative correlation between conductivity and water temperature (Spearman’s correlation coefficient r ¼ 0.641, p < 0.001, results not shown). The higher conductivity values during winter are due to the infiltration of sea water in some of the wastewater network collectors, which is diluted during summer due to a higher wastewater volume to be treated at the WWTP. The water flow to the wetland varied significantly during the four sampling periods, receiving 2437 and 2202 m3/ day in May 2008 and March 2009, and 6007 and 4379 m3/day in August 2008 and July 2009, respectively. Hydraulic retention times (HRT) for treatment cell 3 were calculated during the sampling periods by independent measurements of the water flow, with values ranging from 4.2 to 15.5 days. Unfortunately, the water flow was unexpectedly cut for a short period in July 2009 and these conditions are not indicative of the average situation in the cell. To circumvent this problem, the monthly
average water flow to the wetland system was used in the statistical analyses. Differences in both water and sediment pH were not extremely large and values ranged from 7.2 to 8.7 and from 6.9 to 7.9, respectively. Water nitrate and ammonium concentrations at the sampling locations ranged between 0.01 and þ 1.24 mg NeNO 3 /l and 0.06 and 0.49 mg NeNH4 /l, respectively. Nitrite concentrations were not detectable. Positive correlations were obtained between conductivity and nitrate (r ¼ 0.350, p < 0.01) or ammonium (r ¼ 0.275, p < 0.05). Higher TC and TKN values were usually obtained in sediments collected from areas planted with P. australis ( p < 0.001). Nitrogen removal efficiencies (%) were calculated on the basis on the total nitrogen load and changes in the inlet and outlet concentrations in treatment cell 3 (Garcı´a-Lledo´ et al., 2011). Efficiencies were estimated to be 48 and 61% in periods of low water discharge (May 2008 and March 2009, respectively) and increased up to 71% when higher inflows entered the CWS (July 2009). Nitrogen removal efficiencies correlated positively with ammonia concentration (r ¼ 0.685, p < 0.01) and negatively with nitrate concentrations (r ¼ 0.640, p < 0.01) at the influent. Moreover, oxygen and temperature showed a negative (r ¼ 0.484, p < 0.01) and positive (r ¼ 0.703, p < 0.01) correlation to nitrogen removal efficiencies, respectively.
3.2.
Abundance of 16S rRNA and denitrification genes
The abundance of 16S rRNA genes ranged from 1.3 1011 to 5.8 1012 copies/g dw sediment and appeared to be fairly stable among samples and over time (Fig. 3; Supplementary Table 1). These numbers were always higher than those obtained for any of the functional genes. Mean values for all determined functional gene abundances were between 1 and 4 logs below the 16S rRNA gene. The narG and napA genes ranged from 8.8 108 to 7.1 1010 and from 1.2 109 to 6.4 1010 copies/g dw sediment, respectively. Mean napA gene numbers were significantly higher than narG according to a paired sample t-test (t ¼ 2.578, p < 0.05), but these differences were exclusively due to dominance of napA genes in bulk sediments in March 2009 and no differences were detected when the March samples were excluded (t ¼ 1.116, p > 0.05). Sediment samples showed between 4.9 107 and 6.9 109 copies/g dw sediment for nirS and from 1.6 109 to 2.1 1011 copies/g dw sediment for nirK, suggesting a significantly greater abundance of nirK-type denitrifiers (t ¼ 28.828, p < 0.001). Finally, nosZ ranged from 3.5 107 to 1.3 109 copies/g dw sediment, and were significantly lower than both nirS and nirK (t ¼ 5.184, p < 0.001 and t ¼ 51.770, p < 0.001, respectively).
3.3. size
Factors determining the denitrifying community
Univariate general linear model tests (GLM) were used to analyze the effect of time (sampling period), type of sediment, sampling position in the treatment cell and their interactions, as defined factors to explain differences in gene abundances. Highly significant effects ( p < 0.001) were found when
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
5625
Fig. 2 e Time variation for the 2008e2009 period of the water flow, concentrations of BOD and COD and ammonia and nitrate at the inlet of the constructed wetland. Sediment sampling dates are indicated with arrows.
sampling period and types of sediment were considered (Table 3). The abundance of the nirS gene was the only exception and was not influenced by the presence of vegetation. Additionally, the sampling position only had an effect on 16S rRNA and narG ( p < 0.05). All possible interactions between factors had no significant effect on gene abundances, except for nitrate reductase genes in some cases. Nevertheless, partial Eta squared values (%) for these interactions were rather low, indicating a low significance in the overall variance of the considered genes. Accordingly, in subsequent analyses only the sampling time and the type of sediment were considered. In all cases, the lowest abundance of all analyzed genes was found in samples obtained during August 2008, except for
the nitrate reductase gene narG (Fig. 3AeC). In addition, the presence of vegetation, especially of Phragmites, yielded a significant increase for all genes but napA and nirS according to a non-parametric Welch test (F > 3.671, p < 0.05) (Fig. 3DeF). Differences in the abundance of napA could only be assessed by a post hoc analysis and resulted in higher values in sediments also covered with Phragmites. The presence of Typha did not cause any clear influence on the gene abundances and, depending on the genetic marker considered, was similar to either the bulk sediment or the one with Phragmites. In order to analyze differences in the gene abundances according to physicochemical characteristics of the sediment and water, pair-wise correlation tests were performed. nosZ abundance, being indicative of capacity for N2O reduction,
5626
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
Table 1 e Chemical properties of the sediment and overlying water of the different sampling locations and periods. Data from sediments correspond to mean values and standard deviations of three replicate samples. Type of sediment
Date
Water characteristics
Temp ( C) Cond (mS/cm) O2 (mg/l) pH
Sediment properties
NeNO 3
(mg/l)
NeNHþ 4
(mg/l)
pH
TC (%)
TKN (%)
SE inlet TY inlet PH inlet
15-May-08 21-May-08 20-May-08
20.7 20.4 20.0
3.8 3.6 3.5
6.1 7.6 8.8
7.6 8.0 8.0
0.45 1.07 1.24
0.37 0.21 0.21
7.4 0.1 4.0 0.2 0.21 0.02 7.7 0.1 5.6 1.1 0.30 0.06 7.4 0.1 13.5 1.9 1.23 0.08
SE inlet SE outlet TY inlet PH inlet PH outlet
20-Aug-08 11-Aug-08 21-Aug-08 18-Aug-08 12-Aug-08
24.8 25.9 25.3 24.9 26.6
2.7 2.8 2.8 3.1 2.9
4.3 7.8 5.4 5.2 3.2
7.7 8.0 7.9 7.9 8.1
0.08 0.01 0.12 0.21 0.01
0.21 0.08 0.27 0.12 0.06
7.7 0.0 7.7 0.1 7.9 0.2 7.9 0.0 7.8 0.1
3.8 0.4 3.3 0.4 4.2 0.7 4.3 0.3 5.3 0.2
0.16 0.09 0.18 0.23 0.32
0.02 0.02 0.10 0.04 0.07
SE inlet SE outlet TY inlet TY outlet PH inlet PH outlet
9-Mar-09 3-Mar-09 11-Mar-09 4-Mar-09 10-Mar-09 5-Mar-09
12.6 12.2 13.3 11.1 14.0 10.0
7.1 7.8 7.2 7.8 7.2 7.6
6.1 7.8 5.5 6.8 5.6 5.7
8.4 8.3 8.1 8.3 8.3 8.3
0.26 0.40 0.11 0.33 0.19 0.57
0.19 0.25 0.17 0.27 0.11 0.22
6.9 0.1 7.2 0.1 7.0 0.1 7.0 0.1 7.4 0.1 7.3 0.4
4.2 0.2 3.9 0.2 4.9 0.4 4.3 1.4 9.8 0.8 8.7 0.2
0.07 0.13 0.16 0.16 0.90 0.54
0.04 0.05 0.08 0.16 0.11 0.22
SE inlet SE outlet TY inlet TY outlet PH inlet PH outlet
8-Jul-09 7-Jul-09 13-Jul-09 14-Jul-09 10-Jul-09 16-Jul-09
22.6 25.5 26.4 26.1 22.4 27.6
4.5 4.7 4.6 4.7 4.4 4.5
3.6 3.9 14.9 10.2 3.0 3.8
7.3 7.7 8.7 8.3 7.7 8.2
0.18 0.08 0.03 0.03 0.03 0.03
0.32 0.49 0.28 0.15 0.25 0.11
7.7 0.2 5.7 1.0 7.2 0.1 3.8 0.1 7.6 0.3 5.0 1.3 7.2 0.1 4.8 0.9 7.1 0.2 12.0 2.9 7.5 0.2 9.0 0.7
0.47 0.18 0.29 0.29 1.04 0.83
0.19 0.04 0.22 0.18 0.67 0.21
SE, bulk sediment; TY, areas planted with Typha latifolia; PH, areas planted with Phragmites australis.
correlated positively with sediment TC (r ¼ 0.661, p < 0.01), TKN (r ¼ 0.612, p < 0.01) and negatively with the water flow and COD (r ¼ 0.330, p < 0.05; r ¼ 0.277, p < 0.05, respectively; Table 2). For the other genes, both sediment carbon and nitrogen content also correlated positively and significantly with abundances and negative correlations were observed when the water flow was considered, except for narG. The nitrate content in the water was also an important factor that specifically correlated with nitrite reductase genes, nirS and nirK, and the nitrate reductase napA. Surprisingly, no significant correlations were found for any of the genes and BOD in the influent water or sediment pH, with the exception of water pH and napA.
3.4.
Relative abundance of denitrification genes
The variation in the relative proportion of genes coding for enzymes catalyzing different steps in the denitrification can be used to assess removal efficiencies in terms of potential accumulation of intermediates in the denitrification pathway. Variation of calculated ratios according to time (sampling period), type of sediment, sampling position in the treatment cell and the corresponding interactions was estimated by GLM (Table 4). The relative contribution of the functional genes (narG, napA, nirS, nirK and nosZ ) to the total bacteria population, represented by 16S rRNA did not vary according to the presence of vegetation ( p > 0.05). When sampling period was considered, significant variations could be found for the relative amount of the functional genes narG, napA and nosZ ( p < 0.05) (results not shown). Interesting results can be deduced when ratios between two functional genes are calculated. The ratio between nitrate and nitrite reductases (qnarG þ qnapA)/(qnirS þ qnirK ) varied from 0.2 to 5.4 and was
significantly higher in samples obtained in August 2008 independently of the presence of vegetation or not ( p < 0.05). TC and TKN in the sediment were negatively correlated with (qnarG þ qnapA)/(qnirS þ qnirK ) (r ¼ 0.366, p < 0.01 and r ¼ 0.261, p < 0.05, respectively). The ratio (qnirS þ qnirK )/ qnosZ was significantly influenced by the sampling period ( p < 0.05) and the interaction between sampling period and type of vegetation. The highest difference between the abundance of nitrite reductases (qnirS þ qnirK ) and nitrous oxide reductase (qnosZ ) was found in March and May, when higher nitrate concentrations were measured, and accounted for almost two orders of magnitude. When qnirK/qnosZ and qnirS/ qnosZ ratios were analyzed separately, a significant effect of sampling time was found for both of them ( p < 0.05), but the interaction between sampling time and vegetation was only found significant for the former ( p < 0.05). Relevant variables affecting the ratio between nitrite reducers and nitrous oxide reducers were temperature (r ¼ 0.332, p < 0.05), nitrate content in water (r ¼ 0.314, p < 0.05) and total carbon in sediment (r ¼ 0.309, p < 0.05). Finally, the (qnarG þ qnapA)/ qnosZ ratio was fairly stable between the four sampling periods ( p > 0.05) and no effect of the a priori defined factors was detected.
4.
Discussion
The location of samples was a minor factor related to the abundance of denitrifying genes in the sediment of the treatment wetland. Instead, sampling period and type of sediment, mainly differing by nutrient availability, seems to be more important factors controlling the total abundance of denitrification genes in this system. Nevertheless, data from
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
5627
Fig. 3 e Mean values of the number of copies of the bacterial 16S rRNA and the functional genes, narG, napA, nirS, nirK and nosZ according to sampling period (AeC) and the type of sediment (DeF). Standard errors of the mean are indicated. Different letters above the bars indicate significant differences ( p < 0.05) between sampling period (left) or type of sediment (right).
qPCR must be examined carefully since bacteria may have more than one copy of the same gene per genome. Bacterial genomes can harbor up to 13 copies of the 16S rRNA gene (Fogel et al., 1999). The denitrification genes most often only exist in one copy, although two or three copies have been shown for all denitrification genes (Philippot, 2002; Jones et al., 2008, 2011). Despite these limitations, qPCR studies provide a realistic quantification of the size of the denitrification gene pool and ratios between gene pools can be used to infer community dynamics among samples of similar characteristics (Henry et al., 2006; Kandeler et al., 2006; Geets et al., 2007; Enwall et al., 2010). The relative abundance of functional gene densities in relation to the bacterial 16S rRNA gene abundance indicated
that the denitrifying and nitrate respiring bacteria in the sediment constituted a fraction of 1.6 0.4 and 1.2 0.7% of the bacterial communities, respectively. The values we obtained in the sediment of Empuriabrava FWS-CWs compared well with those found in the only wetland sediment characterized in terms of denitrifier abundances so far (Chon et al., 2011) and studies performed in soils (e.g. Henry et al., 2006; Kandeler et al., 2006; Hallin et al., 2009). The strong correlation between narG and napA genes and the fact that both have similar ratios to the 16S rRNA gene, suggests a high proportion of Proteobacteria in the sediment, since both narG and napA gene can be found in members of this phylum (Philippot and Hojberg, 1999; Roussel-Delif et al., 2005; Bru et al., 2007). Accordingly, quantitative analyses of bacteria at
5628
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
Table 2 e Spearman correlation coefficients (r) between environmental variables in the wetland sediment and water, influent water and abundances of the 16S rRNA and functional genes.
Wetland characteristics
Water Temp ( C) Cond (mS/cm) O2 (mg/l) RedOx (mV) pH NeNO 3 (mg N/l) NeNHþ 4 (mg N/l) Sediment TC (%) TKN (%) C:N pH
Influent characteristics
Water flow (m3/day) COD (mg O2/l) BOD (mg O2/l) NO 3 (mg N/l) NHþ 4 (mg N/l)
q16S rRNA
qnarG
qnapA
qnirS
qnirK
qnosZ
0.186 ns 0.356 ** 0.167 ns 0.114 ns 0.141 ns 0.308 * 0.179 ns
0.076 ns 0.174 ns 0.070 ns 0.043 ns 0.071 ns 0.205 ns 0.177 ns
0.282 * 0.407 ** 0.134 ns 0.002 ns 0.281 * 0.325 * 0.240 ns
0.230 ns 0.241 ns 0.102 ns 0.210 ns 0.073 ns 0.366 ** 0.200 ns
0.248 ns 0.387 ** 0.086 ns 0.124 ns 0.182 ns 0.351 ** 0.168 ns
0.139 ns 0.341 ** 0.113 ns 0.098 ns 0.195 ns 0.224 ns 0.155 ns
0.712 ** 0.653 ** 0.582 ** 0.251 ns 0.421 ** 0.376 ** 0.047 ns 0.243 ns 0.168 ns
0.544 ** 0.573 ** 0.546 ** 0.062 ns 0.160 ns 0.134 ns 0.078 ns 0.074 ns 0.120 ns
0.428 ** 0.395 ** 0.34 ** 0.219 ns 0.361 ** 0.254 ns 0.112 ns 0.285 * 0.235 ns
0.613 ** 0.539 ** 0.468 ** 0.143 ns 0.384 ** 0.346 ** 0.099 ns 0.282 * 0.123 ns
0.658 ** 0.600 ** 0.526 ** 0.234 ns 0.443 ** 0.356 ** 0.107 ns 0.310 * 0.230 ns
0.661 ** 0.612 ** 0.561 ** 0.225 ns 0.330 * 0.227 * 0.11 ns 0.182 ns 0.116 ns
ns, not significant; *p < 0.05; **p < 0.01.
the inlet and outlet of WWTPs in other studies show a clear dominance of Proteobacteria (Juretschko et al., 2002; Chouari et al., 2010; McLellan et al., 2010). The denitrifier community, assessed by the quantification of nirS and nirK genes, was dominated by NirK-type denitrifiers, which were up to 2 logs more abundant than the NirStype. This contrasts with previous studies in the same wetland using a PCR-TRFLP approach in which positive nirK PCR amplifications could not be obtained for any of the sediment samples (Ruiz-Rueda et al., 2007). This discrepancy is likely due to significant changes in the quality of the water entering the Empuriabrava FWS-CWs due to reconstructions in June 2007 (Garcı´a-Lledo´ et al., 2011). Even though the two nitrite reductases are functionally equivalent, denitrifiers harboring either nitrite reductase seem to show a preference for certain environments and are likely not under the same community assembly rules (Jones and Hallin, 2010). Recent studies in agricultural soils also suggested that the existence of the two types of nitrite reductase is due to differential niche preferences (Philippot et al., 2009; Enwall et al., 2010). However, the stability of the present physicochemical conditions of the influent to the Empuriabrava FWS-CWs and in the sediments prevents us from finding the environmental drivers
affecting the qnirS/qnirK ratio in this environment, other than that the sediment and water provided conditions that favor the NirK-type denitrifiers. In agreement with our results, the nosZ abundance in soils is often lower than that of other denitrifying genes (Henry et al., 2006; Hallin et al., 2009; Philippot et al., 2009). The nosZ gene encodes the final step of denitrification, which makes the net nitrogen removal from CWs all the way to N2 possible, but denitrifier communities with low nosZ to nir gene ratios can result in increased N2O/(N2 þ N2O) end-product ratios (Philippot et al., 2011). The lack of nitrous oxide reductase in some denitrifying bacteria was reported a decade ago when the Agrobacterium tumefaciens genome was sequenced (Wood et al., 2001) and recently it has been shown that approximately 1/3 of genome sequenced denitrifying bacterial isolates have this truncated pathway (Jones et al., 2008). The relatively low abundance of nosZ genes in the treatment wetland was especially pronounced in August 2008, coinciding with low nitrate levels and an increase in the nitrogen load in the form of ammonia to the FWS-CWs. In contrast to the other periods, the ammonium was also decreasing between the inlet and outlet of the different vegetation plots (Table 1). At the same time, the relative abundance of the
5629
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
Table 3 e Results of general linear model (GLM) analyses for the effect of the three factors established (sampling period, type of sediment and sampling location) and their interactions with the abundances of 16S rRNA and functional genes. Time
q16S rRNA qnarG qnapA qnirS qnirK qnosZ
Sed
F
p
Eta
F
p
Eta
F
p
Eta
23.21 4.27 6.76 11.13 14.42 10.48
*** * ** *** *** ***
0.65 0.25 0.35 0.47 0.53 0.45
16.49 11.93 10.04 2.87 10.96 14.52
*** *** *** ns *** ***
0.46 0.39 0.35 0.13 0.37 0.43
6.01 6.77 3.41 2.25 3.49 3.18
* * ns ns ns ns
0.14 0.15 0.08 0.06 0.08 0.08
Time Sed
q16S rRNA qnarG qnapA qnirS qnirK qnosZ
Loc
Time Loc
Sed Loc
Time Sed Loc
F
p
Eta
F
p
Eta
F
p
Eta
F
p
1.67 2.71 1.38 1.03 2.23 1.06
ns * ns ns ns ns
0.21 0.30 0.18 0.14 0.26 0.14
1.69 2.15 3.92 0.31 1.14 1.14
ns ns * ns ns ns
0.08 0.10 0.17 0.02 0.06 0.06
1.65 1.08 0.88 0.63 1.26 0.56
ns ns ns ns ns ns
0.08 0.05 0.04 0.03 0.06 0.03
1.11 2.69 3.00 0.84 0.27 2.54
Eta
ns ns * ns ns ns
0.08 0.18 0.19 0.06 0.02 0.17
ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001 Time, sampling period; Sed, type of sediment; Loc, sample location in the treatment cell. Eta, partial Eta Squared.
nitrate reductases, assessed by narG and napA genes, was higher compared to the other sampling occasions. These results suggest that the conditions favored bacteria capable of dissimilatory nitrate reduction to ammonium in addition to denitrifiers that do not have the complete denitrification pathway. Higher potential rates of nitrate reduction to ammonia were measured during August 2008 in bulk sediments and areas covered with P. australis thus confirming the previous results (Garcı´a-Lledo´ et al., 2011). Conditions that invariably will affect denitrification activity, such as temperature and lower nitrate to carbon ratios varied during the studied period. Relatively high nitrate loads to the Empuriabrava FWS-CW occurred intermittently
but frequently throughout the year. Although only two of these periods, May 2008 and March 2009, were sampled in this study, the increase in the measured gene abundances let us hypothesize that the probability of N2O production in the Empuriabrava FWS-CWs would increase in relation to the nitrate concentration. However, additional measurements in summer coinciding with the increase of nitrate at the inlet would be needed to confirm this hypothesis. Other studies conducted on CWs mesocosms have reported significantly higher N2O emissions during summer (Sovik and Klove, 2007). In the Empuriabrava FWS-CWs significant differences of the genetic potential for N2O emissions according to plant species were only detected when interactions with sampling
Table 4 e Results of general linear model (GLM) analyses for the effect of the three factors established (sampling period, type of sediment and sampling location) and their interactions with the ratios between the functional genes involved in denitrification process. Time
(qnarG þ qnapA)/(qnirS þ qnirK ) (qnirS þ qnirK )/qnosZ qnirS/qnirK qnirS/qnosZ qnirK/qnosZ (qnarG þ qnapA)/qnosZ
Sed
F
p
Eta
F
p
Eta
F
p
Eta
3.64 3.81 4.16 4.14 3.76 1.17
* * * * * ns
0.23 0.24 0.25 0.25 0.23 0.09
1.06 0.62 4.79 1.57 0.73 0.31
ns ns * ns ns ns
0.05 0.03 0.21 0.08 0.04 0.17
0.15 0.27 0.00 0.18 0.29 1.08
ns ns ns ns ns ns
0.00 0.01 0.00 0.01 0.01 0.03
Time Sed
(qnarG þ qnapA)/(qnirS þ qnirK ) (qnirS þ qnirK)/qnosZ qnirS/qnirK qnirS/qnosZ qnirK/qnosZ (qnarG þ qnapA)/qnosZ
Loc
Time Loc
Sed Loc
Time Sed Loc
F
p
Eta
F
p
Eta
F
p
Eta
F
1.91 2.88 0.88 1.55 2.91 1.29
ns * ns ns * ns
0.24 0.32 0.13 0.20 0.32 0.17
1.12 0.11 1.85 0.34 0.17 2.82
ns ns ns ns ns ns
0.06 0.01 0.09 0.02 0.01 0.13
1.15 2.09 0.29 0.68 2.17 0.27
ns ns ns ns ns ns
0.06 0.10 0.02 0.04 0.11 0.01
1.74 1.68 2.22 0.50 1.83 0.29
p ns ns ns ns ns ns
Eta 0.12 0.12 0.15 0.04 0.13 0.02
ns, not significant; *, p < 0.05; Time, sampling period; Sed, type of sediment; Loc, sample location in the treatment cell. Eta, partial Eta Squared.
5630
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
time were considered. However, according to the calculated (qnirS þ qnirK )/qnosZ ratio, the highest genetic potential would be found mainly in vegetated sediments where higher nitrogen availability occurred due to leaf litter decomposition. Others have measured higher N2O fluxes in mesocosms planted with P. australis when compared to other macrophytes and unvegetated sediments (Maltais-Landry et al., 2009). Nevertheless, the effect of planted areas for the production of N2O in wetlands has been poorly studied, and further research in this direction is still needed to fully understand the dynamics of constructed wetlands in relation to greenhouse gas emissions.
5.
Conclusions
The Empuriabrava FWS-CWs showed variations in physicochemical parameters analyzed both in water and sediment between the different periods. These conditions favored the maintenance of a denitrifying community that significantly changed according to nutrient concentration and the types of vegetation. The low abundance of nosZ genes compared with the other denitrification genes is an indicative of the genetic capacity of the system to potentially accumulate the N2O intermediary and the relative proportion of nosZ genes decreased during periods of high nitrate content to the wetlands. Overall, the quantitative data on denitrification gene abundances provide evidence of a high potential for nitrous oxide emissions along the entire sediment surface and in particular during periods of high nitrate loading. The vegetation effect was mainly detected in combination with sampling time and resulted in an increase of the potential for nitrous oxide emissions in vegetated areas. The increase in the nitrite to nitrous oxide reductase genes ratio has been related to the higher total carbon content in these sediments.
Acknowledgments A.G-L and A.V-S. are recipients of pre-doctoral grants from the Ministerio de Ciencia y Educacio´n and the Universitat de Girona, respectively. The authors thank the contributions of Anna Huguet, Jordi Sala and Lluı´s Sala for field analyses. This research has been funded by the Spanish Ministerio de Ciencia y Educacio´n (grant CGL2009-08338).
Appendix. Supplementary material Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.watres.2011.08.025.
references
Abell, G.C., Revill, A.T., Smith, C., Bissett, A.P., Volkman, J.K., Robert, S.S., 2010. Archaeal ammonia oxidizers and nirS-type denitrifiers dominate sediment nitrifying and denitrifying
populations in a subtropical macrotidal estuary. The ISME Journal 4 (2), 286e300. APHA, A.a.W., 1998. Standard Methods for the Examination of Water and Wastewater. American Public Health Association, Washington. Boudraa, A.O., Champier, J., Djebali, M., Behloul, F., Beghdadi, A., Bachand, P.A.M., Horne, A.J., 1999. Denitrification in constructed free-water surface wetlands: II. Effects of vegetation and temperature. Ecological Engineering 14, 17e32. Brix, H., 1997. Do macrophytes play a role in constructed treatment wetlands? Water Science and Technology 35, 11e17. Bru, D., Sarr, A., Philippot, L., 2007. Relative abundances of proteobacterial membrane-bound and periplasmic nitrate reductases in selected environments. Applied and Environmental Microbiology 73 (18), 5971e5974. Carter, J.P., Hsaio, Y.H., Spiro, S., Richardson, D.J., 1995. Soil and sediment bacteria capable of aerobic nitrate respiration. Applied and Environmental Microbiology 61 (8), 2852e2858. Cheneby, D., Perrez, S., Devroe, C., Hallet, S., Couton, Y., Bizouard, F., Iuretig, G., Germon, J.C., Philippot, L., 2004. Denitrifying bacteria in bulk and maize-rhizospheric soil: diversity and N2O-reducing abilities. Canadian Journal of Microbiology 50 (7), 469e474. Chon, K., Chang, J.-S., Lee, E., Lee, J., Ryu, J., Cho, J., 2011. Abundance of denitrifying genes coding for nitrate (narG), nitrite (nirS), and nitrous oxide (nosZ) reductases in estuarine versus wastewater effluent-fed constructed wetlands. Ecological Engineering 37 (1), 64e69. Chouari, R., Le Paslier, D., Daegelen, P., Dauga, C., Weissenbach, J., Sghir, A., 2010. Molecular analyses of the microbial community composition of an anoxic basin of a municipal wastewater treatment plant reveal a novel lineage of proteobacteria. Microbial Ecology 60 (2), 272e281. Coyne, M.S., Arunakumari, A., Averill, B.A., Tiedje, J.M., 1989. Immunological identification and distribution of dissimilatory heme cd1 and nonheme copper nitrite reductases in denitrifying bacteria. Applied and Environmental Microbiology 55 (11), 2924e2931. DeBusk, T., DeBusk, W., 2000. In: Kent, D.M. (Ed.), Applied Wetlands Science and Technology, second ed. CRC Press, Boca Raton, Florida (USA), pp. 241e279. Enwall, K., Throback, I.N., Stenberg, M., Soderstrom, M., Hallin, S., 2010. Soil resources influence spatial patterns of denitrifying communities at scales compatible with land management. Applied and Environmental Microbiology 76 (7), 2243e2250. EPA, 2000. Guiding Principles for Constructed Treatment Wetlands Providing for Water Quality and Wildlife Habitat. United States Environmental Protection Agency, Office of Water. Fogel, G.B., Collins, C.R., Li, J., Brunk, C.F., 1999. Prokaryotic genome size and SSU rDNA copy number: estimation of microbial relative abundance from a mixed population. Microbial Ecology 38 (2), 93e113. Francis, C.A., Beman, J.M., Kuypers, M.M., 2007. New processes and players in the nitrogen cycle: the microbial ecology of anaerobic and archaeal ammonia oxidation. The ISME Journal 1 (1), 19e27. Garcı´a-Lledo´, A., Ruiz-Rueda, O., Vilar-Sanz, A., Sala, L., Ban˜eras, L., 2011. Nitrogen removal efficiencies in a free water surface constructed wetland in relation to plant coverage. Ecological Engineering 37, 678e684. Geets, J., de Cooman, M., Wittebolle, L., Heylen, K., Vanparys, B., De Vos, P., Verstraete, W., Boon, N., 2007. Real-time PCR assay for the simultaneous quantification of nitrifying and denitrifying bacteria in activated sludge. Applied Microbiology and Biotechnology 75 (1), 211e221. Gersberg, R.M., Elkins, B.V., Lyon, S.R., Goldman, C.R., 1986. Role of aquatic plants in wastewater treatment by artificial wetlands. Water Research 20 (3), 363e368.
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
Gregory, L.G., Bond, P.L., Richardson, D.J., Spiro, S., 2003. Characterization of a nitrate-respiring bacterial community using the nitrate reductase gene (narG) as a functional marker. Microbiology 149 (1), 229e237. Hallin, S., Jones, C.M., Schloter, M., Philippot, L., 2009. Relationship between N-cycling communities and ecosystem functioning in a 50-year-old fertilization experiment. The ISME Journal 3 (5), 597e605. Henry, S., Bru, D., Stres, B., Hallet, S., Philippot, L., 2006. Quantitative detection of the nosZ gene, encoding nitrous oxide reductase, and comparison of the abundances of 16S rRNA, narG, nirK, and nosZ genes in soils. Applied and Environmental Microbiology 72 (8), 5181e5189. Henry, S., Texier, S., Hallet, S., Bru, D., Dambreville, C., Cheneby, D., Bizouard, F., Germon, J.C., Philippot, L., 2008. Disentangling the rhizosphere effect on nitrate reducers and denitrifiers: insight into the role of root exudates. Environmental Microbiology 10 (11), 3082e3092. Hurt, R.A., Qiu, X., Wu, L., Roh, Y., Palumbo, A.V., Tiedje, J.M., Zhou, J., 2001. Simultaneous recovery of RNA and DNA from soils and sediments. Applied and Environmental Microbiology 67 (10), 4495e4503. Ibekwe, A.M., Lyon, S.R., Leddy, M., Jacobson-Meyers, M., 2007. Impact of plant density and microbial composition on water quality from a free water surface constructed wetland. Journal of Applied Microbiology 102 (4), 921e936. Ingersoll, T.L., Baker, L.A., 1998. Nitrate removal in wetland microcosms. Water Research 32 (3), 677e684. Johnston, C.A., 1991. Sediment and nutrient retention by freshwater wetlands: effects on surface water quality. Critical Reviews in Environmental Control 21 (5), 491e565. Jones, C.M., Hallin, S., 2010. Ecological and evolutionary factors underlying global and local assembly of denitrifier communities. The ISME Journal 4 (5), 633e641. Jones, C.M., Stres, B., Rosenquist, M., Hallin, S., 2008. Phylogenetic analysis of nitrite, nitric oxide, and nitrous oxide respiratory enzymes reveal a complex evolutionary history for denitrification. Molecular Biology and Evolution 25 (9), 1955e1966. Jones, C.M., Welsh, A., Throback, I.N., Dorsch, P., Bakken, L.R., Hallin, S., 2011. Phenotypic and genotypic heterogeneity among closely related soil-borne N2- and N2O-producing Bacillus isolates harboring the nosZ gene. FEMS Microbiology Ecology 76 (3), 541e552. Juretschko, S., Loy, A., Lehner, A., Wagner, M., 2002. The microbial community composition of a nitrifyingedenitrifying activated sludge from an industrial sewage treatment plant analyzed by the full-cycle rRNA approach. Systematic and Applied Microbiology 25 (1), 84e99. Kandeler, E., Deiglmayr, K., Tscherko, D., Bru, D., Philippot, L., 2006. Abundance of narG, nirS, nirK, and nosZ genes of denitrifying bacteria during primary successions of a glacier foreland. Applied and Environmental Microbiology 72 (9), 5957e5962. Kjellin, J., Hallin, S., Worman, A., 2007. Spatial variations in denitrification activity in wetland sediments explained by hydrology and denitrifying community structure. Water Research 41 (20), 4710e4720. Ko¨rner, S., 1999. Nitrifying and denitrifying bacteria in epiphytic communities of submerged macrophytes in a treated sewage channel. Acta hydrochimica et hydrobiologica 27 (1), 27e31. Lin, Y.F., Jing, S.R., Wang, T.W., Lee, D.Y., 2002. Effects of macrophytes and external carbon sources on nitrate removal from groundwater in constructed wetlands. Environmental Pollution 119 (3), 413e420. Maltais-Landry, G., Maranger, R., Brisson, J., Chazarenc, F., 2009. Greenhouse gas production and efficiency of planted and artificially aerated constructed wetlands. Environmental Pollution 157 (3), 748e754.
5631
McLellan, S.L., Huse, S.M., Mueller-Spitz, S.R., Andreishcheva, E.N., Sogin, M.L., 2010. Diversity and population structure of sewage-derived microorganisms in wastewater treatment plant influent. Environmental Microbiology 12 (2), 378e392. Moreno-Vivian, C., Cabello, P., Martinez-Luque, M., Blasco, R., Castillo, F., 1999. Prokaryotic nitrate reduction: molecular properties and functional distinction among bacterial nitrate reductases. Journal of Bacteriology 181 (21), 6573e6584. Nguyen, C., 2003. Rhizodeposition of organic C by plants: mechanisms and controls. Agronomie 23 (5e6), 375e396. Philippot, L., 2002. Denitrifying genes in bacterial and Archaeal genomes. Biochimica et Biophysica Acta 1577 (3), 355e376. Philippot, L., Andert, J., Jones, C.M., Bru, D., Hallin, S., 2011. Importance of denitrifiers lacking the genes encoding the nitrous oxide reductase for N2O emissions from soil. Global Change Microbiology 17, 1497e1504. Philippot, L., Cuhel, J., Saby, N.P., Cheneby, D., Chronakova, A., Bru, D., Arrouays, D., Martin-Laurent, F., Simek, M., 2009. Mapping fieldscale spatial patterns of size and activity of the denitrifier community. Environmental Microbiology 11 (6), 1518e1526. Philippot, L., Hojberg, O., 1999. Dissimilatory nitrate reductases in bacteria. Biochimica et Biophysica Acta 1446 (1e2), 1e23. Prade, K., Trolldenier, G., 1990. Denitrification in the rhizosphere of rice and wheat seedlings as influenced by plant K status, air-filled porosity and substrate organic matter. Soil Biology and Biochemistry 22 (6), 769e773. Reed, S.C., Crites, R.W., Middlebrooks, E.J., 1995. Natural Systems for Waste Management and Treatment. McGraw-Hill, New York. Roussel-Delif, L., Tarnawski, S., Hamelin, J., Philippot, L., Aragno, M., Fromin, N., 2005. Frequency and diversity of nitrate reductase genes among nitrate-dissimilating Pseudomonas in the rhizosphere of perennial grasses grown in field conditions. Microbial Ecology 49 (1), 63e72. Ruiz-Rueda, O., Hallin, S., Baneras, L., 2009. Structure and function of denitrifying and nitrifying bacterial communities in relation to the plant species in a constructed wetland. FEMS Microbiology Ecology 67 (2), 308e319. Ruiz-Rueda, O., Trias, R., Garcia-Gil, L.J., Baneras, L., 2007. Diversity of the nitrite reductase gene nirS in the sediment of a free-water surface constructed wetland. International Microbiology 10 (4), 253e260. Sovik, A.K., Klove, B., 2007. Emission of N2O and CH4 from a constructed wetland in southeastern Norway. Science of the Total Environment 380 (1e3), 28e37. Tiedje, J., 1988. Ecology of denitrification and dissimilatory nitrate reduction to ammonium. In: Zehnder, A.J.B. (Ed.), Biology of Anaerobic Microorganisms. Wiley, New York, pp. 179e244. Toet, S., Bouwman, M., Cevaal, A., Verhoeven, J.T., 2005. Nutrient removal through autumn harvest of Phragmites australis and Thypha latifolia shoots in relation to nutrient loading in a wetland system used for polishing sewage treatment plant effluent. Journal of Environmental Science and Health, Part A 40 (6), 1133e1156. Vial, L., Lavire, C., Mavingui, P., Blaha, D., Haurat, J., MoenneLoccoz, Y., Bally, R., Wisniewski-Dye, F., 2006. Phase variation and genomic architecture changes in Azospirillum. Journal of Bacteriology 188 (15), 5364e5373. Vymazal, J., 2001. Types of constructed wetlands for wastewater treatment: their potential for nutrient removal. In: Vymazal, J. (Ed.), Transformations of Nutrients in Natural and Constructed Wetlands. Backhuys Publishers, Leiden, The Netherlands, pp. 1e93. Whitney, D., Rossman, A., Hayden, N., 2003. Evaluating an existing subsurface flow constructed wetland in Akumal, Mexico. Ecological Engineering 20 (1), 105e111. Wood, D.W., Setubal, J.C., Kaul, R., Monks, D.E., Kitajima, J.P., Okura, V.K., Zhou, Y., Chen, L., Wood, G.E., Almeida Jr., N.F.,
5632
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 6 2 1 e5 6 3 2
Woo, L., Chen, Y., Paulsen, I.T., Eisen, J.A., Karp, P.D., Bovee Sr., D., Chapman, P., Clendenning, J., Deatherage, G., Gillet, W., Grant, C., Kutyavin, T., Levy, R., Li, M.J., McClelland, E., Palmieri, A., Raymond, C., Rouse, G., Saenphimmachak, C., Wu, Z., Romero, P., Gordon, D., Zhang, S., Yoo, H., Tao, Y., Biddle, P., Jung, M., Krespan, W., Perry, M., Gordon-Kamm, B., Liao, L., Kim, S., Hendrick, C., Zhao, Z.Y., Dolan, M., Chumley, F., Tingey, S.V., Tomb, J.F., Gordon, M.P., Olson, M.V., Nester, E.W., 2001. The
genome of the natural genetic engineer Agrobacterium tumefaciens C58. Science 294 (5550), 2317e2323. Zhu, T., Sikora, F.J., 1995. Ammonium and nitrate removal in vegetated and unvegetated gravel bed microcosm wetlands. Water Science and Technology 32 (3), 219e228. Zumft, W.G., 1997. Cell biology and molecular basis of denitrification. Microbiology and Molecular Biology Reviews 61 (4), 533e616.