Microbial nitrogen removal pathways in integrated vertical-flow constructed wetland systems

Accepted Manuscript Microbial nitrogen removal pathways in integrated vertical-flow constructed wetland systems Yun Hu, Feng He, Lin Ma, Yi Zhang, Zhenbin Wu PII: DOI: Reference:

S0960-8524(16)30082-7 http://dx.doi.org/10.1016/j.biortech.2016.01.106 BITE 16017

To appear in:

Bioresource Technology

Received Date: Revised Date: Accepted Date:

19 November 2015 25 January 2016 28 January 2016

Please cite this article as: Hu, Y., He, F., Ma, L., Zhang, Y., Wu, Z., Microbial nitrogen removal pathways in integrated vertical-flow constructed wetland systems, Bioresource Technology (2016), doi: http://dx.doi.org/ 10.1016/j.biortech.2016.01.106

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Microbial nitrogen removal pathways in integrated vertical-flow constructed wetland systems Yun Hu a,b, Feng He a,*, Lin Ma a, Yi Zhang a, Zhenbin Wu a a

State Key Laboratory of Freshwater Ecology and Biotechnology, Institute of

Hydrobiology, Chinese Academy of Sciences, Wuhan 430072, China b

University of Chinese Academy of Sciences, Beijing 100049, China

*

Corresponding author. Tel.: +86 27 68780832.

E-mail address: [email protected]

Abstract Microbial nitrogen (N) removal pathways in planted (Canna indica L.) and unplanted integrated vertical-flow constructed wetland systems (IVCWs) were investigated. Results of, molecular biological and isotope pairing experiments showed that nitrifying, anammox, and denitrifying bacteria were distributed in both down-flow and up-flow columns of the IVCWs. Further, the N transforming bacteria in the planted IVCWs were significantly higher than that in the unplanted ones (p < 0.05). Moreover, the potential nitrification, anammox, and denitrification rates were highest (18.90, 11.75, and 7.84 nmol N g−1 h−1, respectively) in the down-flow column of the planted IVCWs. Significant correlations between these potential rates and the absolute abundance of N transformation genes further confirmed the existence of simultaneous nitrification, anammox, and denitrification (SNAD) processes in the IVCWs. The anammox process was the major N removal pathway (55.6–60.0%) in the IVCWs. The results will further our understanding of the microbial N removal mechanisms in IVCWs. Keywords: Integrated vertical-flow constructed wetland; Nitrogen removal pathway; Isotope pairing technique; Nitrogen transformation functional gene; Anammox

1. Introduction Owing to the advantages of cost-effective and ease-operation, constructed wetlands (CWs) are widely used to treat a variety of wastewaters such as agricultural wastewater (O’Neill et al., 2011), landfill leachate (Lavrova & Koumanova, 2010), and domestic sewage (Sani et al., 2013). CWs are divided into three categories according to their

hydrology and flow paths: surface flow constructed wetlands (SFCWs), horizontal subsurface-flow constructed wetlands (HSSF-CWs), and vertical-flow constructed wetlands (VFCWs) (Vymazal, 2007). HSSF-CWs can provide suitable conditions for denitrification due to their predominantly anoxic-anaerobic, while VFCWs can only provide beneficial conditions for nitrification (Vymazal, 2005). In order to improve the treatment efficiency, a hybrid system integrated vertical-flow constructed wetland (IVCW) was developed, which combines a down-flow and an up-flow VFCW in series (Wu et al., 2008). In recent years, IVCW has received increasing attention, and owing to its high purification efficiency, has become one of the main CW types in China (Wu et al., 2008). Previous researches have primarily focused on optimizing the operational parameters of IVCWs in order to enhance nitrogen (N) removal efficiency (Tao et al., 2010; Chang et al., 2012). However, the underlying N removal pathways in IVCWs are still unclear. N removal processes in CWs are complex and varied, including assimilation by plants and microorganisms, adsorption by substrate, sedimentation of organic N, ammonia volatilization, ammonification, nitrification, and denitrification (Vymazal, 2007). Among these, traditional denitrification once was thought to be the main process of permanent N removal. However, there also exists an unconventional pathway—anaerobic ammonium oxidation (anammox)—in which ammonium is directly oxidized to N2 under anaerobic conditions (Mulder et al., 1995). Recently, anammox has received considerable attention because it is an alternative N removal

pathway at a low oxygen level and C/N ratio (Du et al., 2014). CWs could offer low dissolved oxygen (DO) concentration conditions for the anammox process (Zhu et al., 2010), and the anammox process has been successfully operated in HSSF-CWs (Wang & Li, 2011). IVCWs are characterized by predominantly anoxic-anaerobic conditions (Tao et al., 2010), which was favorable for anammox process. Liao et al. (2008) deduced that there is an approach demanding no carbon can deoxidize nitrite (NO2−) to N2 in IVCWs under C/N ratio below 1. Hence, verified the presence of anammox under low C/N ratio and investigated the relevant N removal pathways in IVCWs might provide theoretical basis for improving N removal efficacy. N removal pathways have been estimated by mass balance method (Wu et al., 2013), stable isotope trace technique (Coban et al., 2015a), absolute abundance of functional gene (Wang et al., 2014), and microbial activity (Zhu et al., 2011). However, there is few information on the quantitative relationships between the abundance of N transformation functional genes (amoA, nxrA, anammox bacterial 16S rRNA, nirK, nirS, and nosZ) and potential microbial activity to integrated analysis N removal pathways in IVCWs. The removal of N in CWs is due primarily to microbial activity (Faulwetter et al., 2009), and plant is a prominent parameter that affects the microbial community in CWs (Picard et al., 2005). The main objective of the present study was to investigate the microbial N removal pathways in planted and unplanted IVCWs. We measured the potential anammox and denitrification activity using isotope pairing technique, and

determined the abundance of N transformation functional genes using molecular biological methods (quantitative PCR, qPCR). It is expected to reveal the microbial N removal mechanisms in IVCWs.

2. Materials and methods 2.1. Design and operation of the IVCWs Six lab-scale IVCWs (2 sets × 3 replicates), each with a down-flow Polyvinylchlorid (PVC) column (diameter, 16 cm; height, 100 cm) in series with an up-flow column of the same size, and a working volume of 12 L, were built up in this study. Each down-flow and up-flow column was filled with gravel (φ = 5–8 mm; porosity = 0.4) to a depth of 15 cm, followed by an upper layer of washed river sand (φ = 0–4 mm; porosity = 0.5) 45 cm deep. These IVCWs were divided into two groups: control (W1: no plants) and planted (W2: 50 plants m-2) (Fig. S1). Canna indica L. were transplanted into the down-flow and up-flow column of W2. The IVCWs were located near Donghu Lake in Wuhan, China. The mean (± standard deviation; s.d.) air temperature was 29.5 ± 3.8 °C. Before the start of the experiment, the IVCWs were fed with domestic wastewater obtained from the Shahu wastewater treatment plant (Wuhan City, Hubei Province, China) for better growth of the plants and microorganisms. The IVCWs were then fed with synthesized wastewater for 8 batches to attain steady-state conditions. The influent wastewater was synthesized with ammonium chloride (NH4Cl) in tap water, and the influent conditions were as follows: pH, 7.3 ± 0.1; COD, 15.0 ± 0.8 mg/L; total nitrogen (TN), 15.4 ± 0.2 mg/L; ammonium (NH4+), 15.0 ± 0.2 mg/L; nitrate (NO3−), 0.4 ± 0.1

mg/L; NO2− not detected. The corresponding effluent concentrations are shown in Fig. S2. The systems were operated in batch mode, each batch was held for a hydraulic retention time (HRT) of 6 d. Inflow water was stored in a plastic tank, added to the top of the down-flow columns, and gravity-drained to the up-flow columns. Steady-state was assumed that the difference of TN and NH4+ removal efficiency were less than 5% for two consecutive batches (Fig. S3). The maximum TN and NH4+ removal efficiencies in the unplanted IVCWs were 70.3% and 79.8%, which were 87.5% and 92.5% for the planted ones. The DO concentrations in the IVCWs are shown in Fig. 1. 2.2. Sample collection At the end of batch 8, substrate samples were collected from the top (10 cm), middle (20 cm), and bottom (45 cm) sections of both the down-flow and up-flow columns of the IVCWs. The three samples were mixed well and stored at –20 °C for further use. 2.3. DNA extraction and qPCR assay Before DNA extraction, approximately 50 g of the substrate sample mixtures were vigorously shaken at 225 rpm for 1h in 50 mL sterile centrifuge tubes in order to release the attached microorganisms into the liquid phase. After centrifuging at 10000 × g for 15 min, the precipitate was collected from the liquid phase. Next, total genomic DNA was extracted using the E.Z.N.A.® Soil DNA Kit (Omega Biotek, USA). The extracted genomic DNA was detected by 1.5% agarose gel electrophoresis. qPCR assay was performed to quantify the anammox bacterial 16S rRNA; as well as

the key functional genes amoA, nxrA, nirK, nirS, and nosZ using the primers AMX809F/AMX1006R, amoA1F/amoA2R, F1norA/R2norA, nirk876F/nirK1040R, nirScd3aF/nirSR3cd, and nosZ1F/nosZ1R, respectively (Table 1). qPCR was performed on a Roche Lightcycler 480 Real-Time PCR detection system (Roche Diagnostics, Meylan, France) using SYBR Green based detection. Each 20 µL reaction mixtures contained 10 µL SYBR Green qPCR master mix (Applied Biosystems, USA), 2 µL of template DNA, 1 µL each of the relevant forward and reverse primers, and 6 µL of sterile water (Millipore, USA). All the samples were run in triplicates. The protocol and parameters for each target gene are summarized in Table 1.The plasmid DNA of all target genes were manufactured by Sangon Biotech Company (Shanghai, China); these plasmids were diluted to a series of 10-fold concentrations and used to prepare standard curves. The standard curves for the target genes showed R2 value between 0.990–0.999, and the amplification efficiencies were 90.3–106.9%. The data obtained from qPCR were normalized to copies per gram of the substrate sample. 2.4. Potential nitrification rate (PNR) Chlorate inhibition method was used to measure the PNR (Sims et al., 2012). Five grams of fresh substrate was incubated in a 50 mL centrifuge tube containing 20 mL of phosphate buffer solution (PBS) (containing [in g/L] NaCl, 8.0; KCl, 0.2; Na2HPO4, 0.2; and NaH2PO4, 0.2; pH 7.3) and 1 mM (NH4)2SO4. After incubation, potassium chlorate (KClO3) was added at a final concentration of 10 mM in order to inhibit the nitrite oxidation. The tubes were incubated on a shaker at in situ temperature (29.5 °C) for 12 h.

Next, the nitrate was extracted with 5 mL of 2 M KCl and measured by a spectrophotometer at 540 nm, using N-(1-naphthyl) ethylenediamine dihydrochloride. The PNR was calculated from the slope of the nitrite accumulation curve. 2.5. Potential anammox (ANR) and denitrification rates(DNR) 15

N-labelled NH4+ and nitrate (NO3−) were used to measure the ANR and DNR (Zhu

et al., 2011). Three grams of substrate sample was transferred to a He-flushed, 12 mL glass vial (Exetainer, Labco, UK), and N2-purged sterile water was added. The resulting substrate slurries were then preincubated on a shaker at 200 rpm, and 29.5 °C for 38 h to remove the residual NOx− (NO3− + NO2−) and oxygen. Subsequently, 100 µL of each isotopic mixture, i.e., (1) 15NH4+ (15N atom% = 98), (2) 15NH4+ + 14NO3−, and (3) 14

NH4+ + 15NO3− (15N atom% = 98) was added to the vials at a final concentration of

100 µM N. Microbial activity in each vial was stopped at intervals of 0 h, 1 h, 2 h, 4 h, and 8 h by syringing 200 µL of a 7 M ZnCl2 solution. Next, 2 mL of the aqueous phase was collected using a gas-tight syringe and injected into fresh He-flushed vials for analysis of the dissolved nitrogen gas. The rates of potential anammox and denitrification, and their contributions to the generated N2 were calculated from the production of 29N2 and 30N2, measured by isotope ratio mass spectrometry (Gasbench II and Delta V Advantage, Thermo Finnigan, Germany) and by using the equations of Thamdrup and Dalsgaard (2002). 2.6. Statistical analysis One-way analysis of variance (ANOVA) was performed to compare the differences

between unplanted (W1) and planted (W2) IVCWs. Linear regression was performed to determine relationships between the potential rates and the functional genes. The statistical analyses were performed by SPSS version 18.0 (IBM, USA).

3. Results and discussion 3.1. N transformation functional genes The abundance of nitrifying, anammox, and denitrifying bacteria in the substrate samples was reflected by the number of copies of amoA, nxrA, anammox bacterial 16S rRNA, nirK, nirS, and nosZ genes (Table 2). The ammonia monooxygenase coding gene amoA and nitrite oxidase coding gene nxrA are often regarded as rate-limiting genes involved in nitrification process (Rotthauwe et al., 1997; Poly et al., 2008). In our study, the absolute abundance of amoA and nxrA in IVCWs were 8.51 × 104–1.27 × 105 and 1.81 × 104–2.58 × 104 copies/g, respectively—indicating the presence of nitrifying bacteria. Further, the number of copies of nxrA gene was lower than that of amoA. This result was in agreement with the previous study in tidal flow constructed wetlands (TFCWs) (Zhi & Ji, 2014), which showed that ammonia oxidizing bacteria oxidize NH4+–N to NO2−–N, thus supplying substrate to nitrite oxidizing bacteria for oxidation of NO2−–N to NO3−–N. The anammox 16S rRNA is often used as the molecular marker of transformation of NH4+–N and NO2−–N to N2 (Tsushima et al., 2007). The copy numbers of anammox 16S rRNA in the down-flow and up-flow columns were 3.51 × 10 7 and 3.38 × 107 in W1; and 5.58 × 107 and 4.38 × 107 in W2, respectively. These copy numbers in our

IVCWs were relatively higher than those reported in the previous studies elsewhere (Wang & Li, 2011; Wang et al., 2015), such as HSSF-CWs, TFCWs, and multimedia biofilters. This is probably due to the predominantly anaerobic conditions in the IVCWs. The IVCW comprised of a down-flow and an up-flow column in series, which maintained an aerobic-anoxic- anaerobic-anoxic-aerobic pattern along the direction of water flow (Fig. 1). Further, the concentrations distribution of dissolved oxygen in the IVCWs favored nitrification in the upper aerobic layer, while anammox could occur in the deeper anoxic- anaerobic zones of the down-flow and up-flow columns. Wang and Li (2011) showed that the anammox process was successfully established and operated steadily in new CWs after operating for 3 months. We, therefore, concluded that the operating time of 3 months in our study, as well as the operating parameters of our IVCW, were suitable for the anammox process to be established. The nirK, nirS, and nosZ genes are often used as markers to study denitrification process (Henry et al., 2004; Henry et al., 2006; Kandeler et al., 2006). Our results showed that the absolute abundance of these genes were 2.22 × 105–5.25× 105, 1.18 × 10 6–2.17 × 10 6, and 3.48 × 10 5–5.77× 10 5 copies/g, respectively. This result was consistent with those of previous studies, which reported that the nirS gene was more abundant than the nirK and nosZ (Kandeler et al., 2006; Chen et al., 2014). The higher nirS-type denitrifiers indicated that they played a dominant role in nitrite reduction and were major contributors to NO gas in the IVCWs. The final step of the denitrification process is encoded by the nosZ gene, which reduces N2O to N2. The abundance of nosZ

was lower than that of other denitrifying genes, probably because the nirS and nirK gene products provided substrate to the nosZ gene for the final denitrification step. The copy numbers of the N transformation functional genes in W2 were significantly higher than the copy numbers in W1 (p < 0.05). This can be explained by the fact that plants provide an oxygenated zone around the roots, which promotes the growth of nitrifying bacteria (Zhu & Sikora, 1994). Moreover, plants influence denitrification by providing organic carbon through rhizodeposition (Lin et al., 2002). Wang and Li (2011) found that plants also have favorable effects on the enrichment of anammox bacteria. This indicated that Canna indica L., used in our study, promoted the growth of N transforming bacteria. Further, the nitrifying, anammox, and denitrifying bacteria were distributed in both down-flow and up-flow columns of the IVCWs, being slightly more abundant in the down-flow columns. DO concentrations in the down-flow columns (0.55–2.2 mg/L) were also slightly higher than that in the up-flow columns (0.08–1.48 mg/L), and this was probably the reason for this phenomenon. Taken together, measuring the absolute abundance of N transformation genes helped to quantify the nitrifying, anammox, and denitrifying bacteria in the IVCWs. However, it should be noted that the mere presence of these genes does not represent the actual functional processes. Therefore, we also examined the potential activities of nitrification, anammox, and denitrification, as described below. 3.2. Potential activities and abundance of nitrifying, anammox, and denitrifying bacteria

The potential nitrification activities, characterized by PNR, of the bacteria in the IVCWs are shown in Fig. 2a. The highest PNR (18.90 nmol N g−1 h−1) was recorded in the down-flow column of W2, followed by the up-flow column of W2 (16.36 nmol N g−1 h−1), and the down-flow and up-flow columns of W1 (14.83 and 14.31 nmol N g−1 h−1, respectively). Moreover, regression analysis of the data strongly indicated a positive association of PNRs with the copy numbers of amoA gene (R2 = 0.811, p < 0.001; Fig. 3a). These results indicated that ammonia oxidation bacteria played a dominant role in the nitrification process in IVCWs. To determine the ANR and DNR, 15N isotope pairing experiments on the substrate were performed at in situ temperature. The incubations amended with only 15NH4+, no significant accumulation of either 29N2 or 30N2 was detected in the substrate of W1 and W2, indicating that total environmental 14NOx− had been consumed during the 38 h preincubation period. When both 15NH4+ and 14NO3− were added, 29N2 accumulated in every substrate sample, 30N2 accumulation could still not be detected (Fig. S4). These results revealed that the anammox process occurred in both W1 and W2. Further, significant activities of the anammox and denitrification processes were also observed, in presence of both 14NH4+ and 15NO3− incubation (Fig. 4). The ANRs in the down-flow and up-flow columns were 9.29 and 8.46 nmol N g−1 h−1, respectively, in W1; and 11.75 and 11.05 nmol N g−1 h−1, respectively, in W2. The DNRs in the down-flow and up-flow columns were 6.98 and 6.74 nmol N g−1 h−1, respectively, in W1; and 7.84 and 7.53 nmol N g−1 h−1, respectively, in W2. Further, 55.6–60.0% of the generated N2 was

contributed by anammox in the IVCWs; this value is higher than that reported by previous studies on anammox in SFCWs (Erler et al., 2008) and VFCWs (Zhu et al., 2011). The significant anammox contribution in IVCWs probably relies on the combined effects of abundant NH4+, low C/N ratio and the operational conditions in hybrid systems. Refer to Coban et al. (2015b), a low C/N ratio less than 2 facilitates anammox. In our study, the C/N ratio in the influent wastewater was about 1. In addition, the PNRs were higher than the ANRs and DNRs, and the DNRs were lower than the ANRs in IVCWs, indicating that ammonia oxidation is the main pathway to supply NO2− for anammox in IVCWs. These results further confirmed that anammox played an important role in N removal in IVCWs. The ANRs showed a similar tend with the copy numbers of anammox 16S rRNA gene (Fig. 2b), thus reflecting a positive correlation between them (R2 = 0.689, p < 0.001; Fig. 3b). This indicated that the anammox 16S rRNA gene regulated anammox rates in IVCWs. The DNRs also exhibited a similar variations trend with the absolute abundance of nirK, nirS, and nosZ (Fig. 2c). To delve deeper into the quantitative relationship between denitrifying genes and DNRs in the IVCWs, the absolute abundance of nirK, nirS, and nosZ genes were used as candidate variables in the regression analysis. The variation of the absolute abundance of nirS gene was positively associated with DNR (R2 = 0.872, p < 0.001; Fig. 3c). Therefore, the nirS gene is directly associated with the denitrification process.

The potential microbial activities and absolute abundance of N transformation functional genes were higher in planted, than in unplanted IVCWs. Moreover, the TN removal efficiency of planted IVCWs (87.5%) was also higher than that of the unplanted ones (70.3%). This result indicated that the plants positively affected the growth of the N transforming bacteria and enhanced their N removal efficiency. A similar observation was made in SFCWs by Gagnon et al. (2007). In addition, Faulwetter et al. (2009) reported that plants promote enrichment of microorganisms that are responsible for pollution removal. 3.3. Microbial N removal pathways in the IVCWs Based on the distribution of the N transforming bacteria, and the quantitative relationships between the N transformation functional genes and potential microbial activities, we deduced the microbial N removal pathways in the IVCWs (Fig. 5). The abundance of amoA gene positively correlated with the nitrification rates, because its gene product was primarily involved in the conversion of NH4 + to NO2− under aerobic conditions. At the same time, ammonia oxidization provided NO2− as substrate for anammox and denitrification. The anammox 16S rRNA regulated the anammox rate, regulating the oxidation of NH4+ and NO2− to N2 under anoxic- anaerobic conditions. The nirS gene determined the denitrification rate and regulated the consumption of NO2−, which was supplied by the nitrification process. Therefore, we concluded that nitrification, anammox, and denitrification were all involved in and contributed to NH4+ removal. Ammonia oxidation was the major NH4+ transformation pathway in the

aerobic zones, while anammox and denitrification were the predominant NH4+ removal pathways in the anoxic- anaerobic zones of IVCWs.

4. Conclusion In this study, quantitative relationships between the abundance of N transformation functional genes and potential microbial activities showed that the amoA, anammox 16S rRNA, and nirS genes determined N removal rates, as well as the nitrification, anammox, and denitrification (SNAD) processes in IVCWs. Anammox process was the major N removal pathway in IVCWs. Moreover, the presence of plants enhanced the growth of the N transforming bacterial. Therefore, the results of this study provide a foundation stone for future attempts at developing stable and efficient N removal processes in IVCWs. Acknowledgments This work was financially supported by the National Science and Technology Support Program of the 12th Five-Year Plan, China (No. 2012BAJ21B03-04), the National Natural Science Foundation of China (51178452), the Major Science and Technology Program for Water Pollution Control and Treatment of China of the 12th Five-Year Plan (No. 2012ZX07101007-005), and the Hubei Province Science Foundation for Youths (2014CFB282). We thank Profs. Guibing Zhu for he help with the analytical methods of anammox and denitrification activity (Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences). We also thank Drs. Yafen Wang, Qiaohong Zhou, Biyun Liu, Enrong Xiao, Dong Xu, Junmei Wu and Ms. Liping

Zhang for the experimental help (Institute of Hydrobiology, Chinese Academy of Sciences). Appendix A. Supplementary data Supplementary Material containing 6 pages, with 4 figures. References 1.

Attard, E., Poly, F., Commeaux, C., Laurent, F., Terada, A., Smets, B.F., Recous, S.,

Le Roux, X., 2010. Shifts between Nitrospira- and Nitrobacter-like nitrite oxidizers underlie the response of soil potential nitrite oxidation to changes in tillage practices. Environ. Microbiol. 12 (2), 315–326. 2.

Chang, J.J., Wu, S.Q., Dai, Y.R., Liang, W., Wu, Z.B., 2012. Treatment

performance of integrated vertical-flow constructed wetland plots for domestic wastewater. Ecol. Eng. 44, 152–159. 3.

Chen, Y., Wen, Y., Zhou, Q., Vymazal, J., 2014. Effects of plant biomass on

denitrifying genes in subsurface-flow constructed wetlands. Bioresour. Technol. 157, 341–345. 4.

Coban, O., Kuschk, P., Wells, N.S., Strauch, G., Knoeller, K., 2015a. Microbial

nitrogen transformation in constructed wetlands treating contaminated groundwater. Environ. Sci. Pollut. Res. 22 (17), 12829–12839. 5.

Coban, O., Kuschk, P., Kappelemeyer, U., Spott, O., Martienssen, M., Jetten,

M.S.M., Knoeller, K., 2015b. Nitrogen transforming community in a horizontal subsurface-flow constructed wetland. Water Res. 74, 203–212.

6.

Du, R., Peng, Y.Z., Cao, S.B., Wu, C.C., Weng, D.C., Wang, S.Y., He, J.Z., 2014.

Advanced nitrogen removal with simultaneous Anammox and denitrification in sequencing batch reactor. Bioresour. Technol. 162, 316–322. 7.

Erler, D.V., Eyre, B.D., Davison, L., 2008. The contribution of anammox and

denitrification to sediment N2 production in a surface flow constructed. Environ. Sci. Technol. 42, 9144–9150. 8.

Faulwetter, J.L., Gagnon, V., Sundberg, C., Chazarenc, F., Burr, M.D., Brisson, J,

Camper, A.K, Stein, O.R., 2009. Microbial proceeses influencing performance of treatment wetlands: A review. Ecol. Eng. 35, 987–1004.

9.

Henry, S., Baudoin, E., López-Gutiérrez, J.C., Martin-Laurent, F., Brauman, A.,

Philippot, L., 2004. Quantification of denitrifying bacteria in soils by nirK gene targeted

real-time PCR. J. Microbiol. Meth. 59, 327–335.

10. 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. Appl. Environ. Microbiol. 72 (8),

5181–5189. 11. 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. Appl. Enviro. Microbiol. 72 (9), 5957–5962. 12. Lavrova, S., Koumanova, B., 2010. Influence of recirculation in a lab-scale vertical flow constructed wetland on the treatment efficiency of landfill leachate. Bioresour.

Technol. 101, 1756–1761. 13. Liao, X.S., He, F., Xu, D., Tao, M., Wu, Z.B., 2008. Influence of low C/N on nitrification and denitrification in wetlands. China Environmental Science, 28 (7), 603–607 (In Chinese). 14. 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. Environ. Pollut. 119, 420–423. 15. Mulder, A., Vandegraaf, A.A., Robertson, L.A., Kuenen, J.G., 1995. Anaerobic ammonium oxidation discovered in a denitrifying fluidized-bed reactor. FEMS Microbiol. Ecol. 16 (3), 177–183. 16. O’Neill, A., Foy, R.H., Phillips, D.H., 2011. Phosphorus retention in a constructed wetland system used to treat dairy wastewater. Bioresour. Technol. 102, 5024–5031. 17. Picard, C.R., Fraser, L.H., Steer, D., 2005. The interacting effects of temperature and plant community type on nutrient removal in wetland microcosms. Bioresour. Technol. 96, 1039–1047. 18. Poly, F., Wertz, S., Brothier, E., Degrange, V., 2008. First exploration of Nitrobacter diversity in soils by a PCR cloning-sequencing approach targeting functional gene nxrA. FEMS Microbiol. Ecol. 63 (1), 132–140. 19. Rotthauwe, J.H., Witzel, K.P., Liesack, W., 1997. The ammonia monooxygenase structural gene amoA as a functional marker: molecular fine-scale analysis of natural ammonia-oxidizing populations. Appl. Environ. Microbiol. 63 (12), 4704–4712.

20. Sani, A., Scholz, M., Bouillon, L., 2013. Seasonal assessment off experimental vertical-flow constructed wetlands treating domestic wastewater. Bioresour. Technol. 147, 585–596. 21. Sims, A., Horton, J., Gajaraj, S., McIntosh, S., Miles, R.J., Mueller, R., Reed, R., Hu, Z.Q., 2012. Temporal and spatial distributions of ammonia-oxidizing archaea and bacteria and their ratio as an indicator of oligotrophic conditions in natural wetlands. Water Res. 46, 4121–4129. 22. Tao, M., He, F., Xu, D., Li, M., Wu, Z.B., 2010. How artificial aeration improved the sewage treatment of an integrated vertical-flow constructed wetland. Pol. J. Environ. Stud. 19 (1), 181–189. 23. Thamdrup, B., Dalsgaard, T., 2002. Production of N2 through anaerobic ammonium oxidation coupled to nitrate reduction in marine sediments. Appl. Environ. Microbiol. 68 (3), 1312–1318. 24. Tsushima, I., Kindaichi, T., Okabe, S., 2007. Quantification of anaerobic ammonium-oxidizing bacteria in enrichment cultures by real-time PCR. Water Res. 41 (4), 785–794. 25. Vymazal, J., 2005. Horizontal sub-surface flow and hybrid constructed wetlands systems for wastewater treatment. Ecol. Eng. 25 (5), 478–490. 26. Vymazal, J., 2007. Removal of nutrients in various types of constructed wetlands. Sci. Total Environ. 380 (1–3), 48–65. 27. Wang, L., Li, T., 2011. Anaerobic ammonium oxidation in constructed wetlands

with bio-contact oxidation as pretreatment. Ecol. Eng. 37, 1225–1230. 28. Wang, C., Zhao, Y.C., Xie, B., Peng, Q., Hassan, M., Wang, X.Y., 2014. Nitrogen removal pathway of anaerobic ammonium oxidation in on-site aged refuse bioreactor. Bioresour. Technol. 159, 266–271. 29. Wang, H.L., Ji, G.D., Bai, X.Y., 2015. Enhanced long-term ammonium removal and its ranked contribution of microbial genes associated with nitrogen cycling in a lab-scale multimedia biofilter. Bioresour. Technol. 196, 57–64. 30. Wu, Z.B., et al., 2008. Integrated Vertical-flow Constructed Wetland. Science Press, Beijing, China, pp. 9–11 (In Chinese). 31. Wu, H.M., Zhang, J., Wei, R., Liang, S., 2013. Nitrogen transformations and balance in constructed wetlands for slightly pollution river water treatment using different macrophytes. Environ. Sci. Pollu. Res. 20, 443–451. 32. Zhi, W., Ji, G.D., 2014. Quantitative response relationships between nitrogen transformation rates and nitrogen functional genes in a tidal flow constructed wetland under C/N ratio constraints. Water Res. 64, 34–41. 33. Zhu, T., Sikora, F.J., 1994. Ammonium and nitrate removal in vegetated and unvegetated gravel bed microcosm wetlands. Water Sci. Technol. 32 (3), 219–228. 34. Zhu, G.B., Jetten, M.S.M., Kuschk, P., Ettwig, K.F., Yin, C.Q., 2010. Potential roles of anaerobic ammonium and methane oxidation in the nitrogen cycle of wetland ecosystems. Appl. Microbiol. Biotechnol. 86 (4), 1043–1055. 35. Zhu, G.B., Wang, S.Y., Feng, X.J., Fan, G., Jetten, M.S.M., Yin, C.Q., 2011.

Anammox bacterial abundance, biodiversity and activity in a constructed wetland. Environ. Sci. Technol. 45 (23), 9951–9958.

Captions Fig.1 DO concentration of water at the sampling sites during 8 batches. Six sampling ports at different depths 10, 20 and 45 cm in both down-flow and up-flow columns were labeled as S1, S2, S3, S4, S5, and S6 along the direction of water flow. Values are means (Error bars are indicate s.d. (n = 24)). Fig. 2 Variations of potential activity and the absolute abundance of functional genes in the substrate samples from the down-flow and up-flow columns of the IVCWs. Values are means (Error bars indicate s.d. (n = 3)). Fig. 3 Relationship between PNR & copy numbers of AOB amoA gene (a), ANR & copy numbers of anammox 16S rRNA gene (b) and DNR & copy numbers of nirS gene (c) in the IVCWs. Fig. 4 Production of 29N2 and 30N2 in incubation of substrate from the down-flow and up-flow columns of the IVCWs. “ra” means the ratio of anammox to total N2 production. Values are means (Error bars indicate s.d. (n = 3)). Fig. 5 Microbial nitrogen removal pathways in the IVCWs. Fig. S1 Schematic representation of the IVCWs. Arrows indicate water flow direction. Fig. S2 Effluent concentrations of the IVCWs during 8 batches. Values are means (Error bars are indicate s.d. (n = 3)).

Fig. S3 The proportion of TN and NH4+–N removal efficiencies in the IVCWs during 8 batches. Values are means (Error bars are indicate s.d. (n = 3)). Fig. S4 Examples of concentrations of 29N2 and 30N2 in samples amended with 15N labeled on ammonia and nitrate, separately. Case A and case B serve for negative and positive control respectively.

Fig.1 DO concentration of water at the sampling sites during 8 batches. Six sampling ports at different depths 10, 20 and 45 cm in both down-flow and up-flow columns were labeled as S1, S2, S3, S4, S5, and S6 along the direction of water flow. Values are means (Error bars are indicate s.d. (n = 24)).

Fig. 2 Variations of potential activity and the absolute abundance of functional genes in the substrate samples from the down-flow and up-flow columns of the IVCWs. Values are means (Error bars indicate s.d. (n = 3)).

Fig. 3 Relationship between PNR & copy numbers of AOB amoA gene (a), ANR & copy numbers of anammox 16S rRNA gene (b) and DNR & copy numbers of nirS gene (c) in the IVCWs.

Fig. 4 Production of 29N2 and 30N2 in incubation of substrate from the down-flow and up-flow columns of the IVCWs. “ra” means the ratio of anammox to total N2 production. Values are means (Error bars indicate s.d. (n = 3)).

Fig. 5 Microbial nitrogen removal pathways in the IVCWs.

Captions Table 1 Primers and protocols of target genes used in qPCR analysis. Table 2 Absolute abundance of nitrogen transformation genes in the substrate from the down-flow and up-flow columns of the IVCWs.

Table 1 Primers and protocols of target genes used in qPCR analysis. Target

Primer

gene

Primer sequence

Amplific qPCR program

Referenc

(5’-3’)

ation

(40 cycles)

e

3 min at 95 °C, 15

Tsushim

s at 95 °C, 20 s at

a et al., 2007

size (bp) anamm

AMX8 GCCGTAAACGATGG

ox 16S

09F

rRNA

AMX1 AACGTCTCACGACA

57 °C and 30 s at

066R

CGAGCTG

72 °C

amoA

amoA

GGGGTTTCTACTGG

(AOB)

1F

nxrA

nirK

nosZ

GCACT

3 min at 95 °C, 15

Rotthau

TGGT

s at 95 °C, 20 s at

we et al.,

amoA

CCCCTCKGSAAAGC

57 °C and 30 s at

1997

2R

CTTCTTC

72 °C

F1nor

CAGACCGACGTGT

491

A

3 min at 95 °C, 15

Attard et

GCGAAAG

s at 95 °C, 20 s at

al., 2010

R2nor

TCCACAAGGAACG

57 °C and 30 s at

A

GAAGGTC

72 °C

nirK87 ATYGGCGGVCAYG 6F

nirS

282

323

165

GCGA

3 min at 95 °C, 15

Henry et

s at 95 °C, 20 s at

al., 2004

nirK10 GCCTCGATCAGRTT

57 °C and 30 s at

40R

RTGGTT

72 °C

nirScd

GTSAACGTSAAGGA

3aF

RACSGG

413

3 min at 95 °C, 15

Kandele

s at 95 °C, 20 s at

r et al.,

nirSR3 GASTTCGGRTGSGT

57 °C and 30 s at

2006

cd

CTTGA

72 °C

nosZ1 F

WCSYTGTTCMTCG ACAGCCAG

nosZ1

ATGTCGATCARCTG

58 °C and 30 s at

R

VKCRTTYTC

72 °C

251

4 min at 94 °C, 45

Henry et

s at 94 °C, 45 s at

al., 2006

Table 2 Absolute abundance of nitrogen transformation genes in the substrate from the down-flow and up-flow columns of the IVCWs. Target gene

W1-down (copies/g) amoA (AOB) (8.82 ± 0.33) × 104 nxrA (1.81 ± 0.10) × 104 anammox 16S (3.51 ± 0.25) × rRNA 107 nirK (2.87 ± 0.03) × 105 nirS (1.28 ± 0.11) × 106 nosZ (3.97± 0.19) × 104 a Values are means (± s.d., n = 3).

W1-up (copies/g) (8. 51 ± 1.50) × 104 (1.74 ± 0.10) × 10 4 (3.38 ± 0.15) × 10 7 (2.22 ± 0.45) × 10 5 (1.18 ± 0.09) × 10 6 (3.48 ± 0.99) × 10 4

W2-down (copies/g) (1.27 ± 0.11) × 105 (2.58 ± 0.28) × 104 (5.58 ± 0.55) × 107 (5.25 ± 1.64) × 105 (2.17± 0.32) × 106 (5.77± 0.91) × 104

W2-up (copies/g) (1.11 ± 0.02) × 10 5 (1.93 ± 0.24) × 10 4 (4.38 ± 0.83) × 10 7 (3.64 ± 0.03) × 10 5 (1.90 ± 0.03) × 10 6 (5.13 ± 1.31) × 10 4

Highlights • Isotope pairing technique and molecular biological methods were combined. • Microbial N removal pathways and rates in IVCWs were estimated. • Simultaneous nitrification, anammox and denitrification processes existed in IVCWs. • Anammox contributed 55.6–60.0% to nitrogen removal in IVCWs.