Effects of pH and seasonal temperature variation on simultaneous partial nitrification and anammox in free-water surface wetlands

Journal of Environmental Management 110 (2012) 103e109

Contents lists available at SciVerse ScienceDirect

Journal of Environmental Management journal homepage: www.elsevier.com/locate/jenvman

Effects of pH and seasonal temperature variation on simultaneous partial nitrification and anammox in free-water surface wetlands Yuling He a, Wendong Tao a, *, Ziyuan Wang b, Walid Shayya c a

Department of Environmental Resources Engineering, College of Environmental Science and Forestry, State University of New York, 1 Forestry Dr., Syracuse, NY 13210, USA School of Environment, Beijing Normal University, 19 Xingjiekouwai St., Beijing 100875, China c College of Agriculture and Technology at Morrisville, State University of New York, 134 Marshall Hall, Morrisville, NY 13408, USA b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 11 January 2012 Received in revised form 28 May 2012 Accepted 13 June 2012 Available online 2 July 2012

Design considerations to enhance simultaneous partial nitrification and anammox in constructed wetlands are largely unknown. This study examined the effects of pH and seasonal temperature variation on simultaneous partial nitrification and anammox in two free-water surface wetlands. In order to enhance partial nitrification and inhibit nitrite oxidation, furnace slag was placed on the rooting substrate to maintain different pH levels in the wetland water. The wetlands were batch operated for dairy wastewater treatment under oxygen-limited conditions at a cycle time of 7 d. Fluorescence in situ hybridization analysis found that aerobic ammonium oxidizing bacteria and anammox bacteria accounted for 42e73% of the bacterial populations in the wetlands, which was the highest relative abundance of ammonium oxidizing and anammox bacteria in constructed wetlands enhancing simultaneous partial nitrification and anammox. The two wetlands removed total inorganic nitrogen efficiently, 3.36e3.38 g/m2/d in the warm season with water temperatures at 18.9e24.9
C and 1.09e1.50 g/m2/d in the cool season at 13.8e18.9
C. Plant uptake contributed 2e45% to the total inorganic nitrogen removal in the growing season. A seasonal temperature variation of more than 6
C would affect simultaneous partial nitrification and anammox significantly. Significant pH effects were identified only when the temperatures were below 18.9
C. Anammox was the limiting stage of simultaneous partial nitrification and anammox in the wetlands. Water pH should be controlled along with influent ammonium concentration and temperature to avoid toxicity of free ammonia to anammox bacteria. Ó 2012 Elsevier Ltd. All rights reserved.

Keywords: Constructed wetland Simultaneous partial nitrification and anammox Dairy wastewater pH effect Temperature effect Nitrogen removal

1. Introduction Constructed wetlands have been designed to remove nitrogen mainly by nitrification and denitrification. Nitrogen removal from ammonium-rich wastewater in constructed wetlands using the nitrificationeheterotrophic denitrification process is usually restricted by limited availability of oxygen and organic carbon (Kadlec and Wallace, 2009; Vymazal, 2007). Nitrification is a twostep oxidation process including aerobic ammonium oxidation to nitrite (partial nitrification) by ammonium oxidizing bacteria (AOB) and nitrite oxidation by nitrite oxidizing bacteria (NOB). Anaerobic ammonium oxidation (anammox) was discovered in the mid-1990s (van de Graaf et al., 1996), which uses nitrite to oxidize ammonium under anaerobic conditions. When partial nitrification and anammox are coupled, only about one half of ammonium needs to be

* Corresponding author. Tel.: þ1 315 470 4928; fax: þ1 315 470 6958. E-mail address: [email protected] (W. Tao). 0301-4797/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jenvman.2012.06.009

oxidized to nitrite first and anammox bacteria use the remaining ammonium as electron donors for autotrophic denitrification of the nitrite produced (van der Star et al., 2007; Kartal et al., 2010). Compared with the nitrificationedenitrification process, the partial nitrificationeanammox process requires 60% less of oxygen and eliminates the requirement for organic carbon. The partial nitrificationeanammox process is typically used to treat warm wastewater with high ammonium concentrations such as sludge digester liquor and landfill leachate. Partial nitrification and anammox have been coupled for ammonium removal either in two reactors in series or simultaneously in one oxygen-limited reactor (van der Star et al., 2007; Kartal et al., 2010). Simultaneous partial nitrification and anammox (SNA) has been employed recently in free-water surface wetlands receiving ammonium-rich wastewater under oxygen-limited conditions. Dong and Sun (2007) modified a subsurface flow wetland to enhance SNA in a following free-water surface wetland. Tao and Wang (2009) found that vegetation and basic rooting substrate produced slightly aerobic and basic conditions in free-water surface wetlands, which

Y. He et al. / Journal of Environmental Management 110 (2012) 103e109

2. Materials and methods 2.1. Free-water surface wetlands Two free-water surface wetlands (CWS and CW) were constructed in a greenhouse in Syracuse, New York, USA on 5 February 2010 as depicted in Fig. 1. The wetlands were rectangular tanks lined with ethylene propylene diene monomer pond liner. Each tank had internal flat dimensions of 0.42 m by 0.45 m. The wetlands had 0.20 m of sandy loam as rooting substrate. Segments of common reed (Phragmites australis) roots were transplanted to the wetlands on 15 February and 18 March 2010. Water level in the wetlands was gradually increased until 17 May 2010 while new shoots were developed. Electric arc furnace slag (effective size of 5 mm) was put on the rooting substrate of CWS (5.4 L each) and CW (2.2 L) between 7 June and 12 July 2010 to buffer water pH. More slag was put on the rooting substrate of the two wetlands (1.5 L each) on 31 August 2010 to raise water pH. More slag (0.5 L) was added again to the rooting substrate of CWS on 14 December 2010 to maintain higher pH values. The free-water surface wetlands were operated by a batch mode with a cycle time of 7 d. The wetlands were discharged through a ball valve 10 mm above the rooting substrate. Dairy wastewater

Common reeds

Pre-treated dairy wastewater

20 cm

Slag Effluent Free water surface wetland (CW)

Sandy loam

pretreated in partial nitrificationeanammox biofilters was fed to the wetlands starting from 5 July 2010. Organic carbon-free, pretreated synthetic wastewater (NH4Cl dissolved in tap water) was fed to the wetlands to enrich anammox bacteria and AOB during the initial five months before 5 July 2010. Treatment performance and the discussion hereinafter are based on the operational results between 5 July 2010 and 10 March 2011 when dairy wastewater was treated. The free-water surface wetlands were a part of a treatment train to remove the high concentrations of ammonium from the dairy wastewater. The wetland influent was rich in ammonium and had relatively low concentrations of nitrite and nitrate (Fig. 2). The ammonium concentrations in the influent fed to the two wetlands were higher than that (39e266 mg N/L) in the dairy wastewater treated in earlier constructed wetlands (Kadlec and Wallace, 2009; Knight et al., 2000). Hydraulic loading rate and nitrogen loading rate to the wetlands were 22.1 mm/d and 5.9 g/m2/d, respectively. Because of infestation of common reeds, the wetlands were closed on 10 March 2011.

2.2. Field measurements and chemical analyses Influent and effluent were measured weekly for dissolved oxygen, redox potential, temperature, and pH with portable meters. Influent and effluent volumes were measured with a graduated cylinder. Influent and effluent samples were analyzed weekly with a QuickChem 8500 series automatic flow injection analyzer (Lachat Instruments, Loveland, CO, USA) using the phenolate method for ammonium and hydrazine reduction method for nitrate and nitrite (APHA et al., 1998). NHþ 4 eN concentration determined with the phenolate method represents the total ammonia concentration (NHþ 4 eN þ NH3eN) in the water samples. Total inorganic nitrogen   (TIN) was calculated as the sum of NHþ 4 eN, NO2 eN and NO3 eN. Free ammonia concentration was calculated with measured total ammonia concentration, temperature and pH (Camargo Valero and Mara, 2010). Nitrogen was not measured in January and early February 2011 due to logistical reasons. COD was determined occasionally, following Standard Method 5220 (APHA et al., 1998). Plant average height, stem diameter and stem number were recorded to estimate the growth of plant biomass when a significant change was observed. Different sizes of stems (total 14) were randomly sampled from the wetlands on 7 October 2010. After

400

80 Ammonium

Nitrite

Nitrate

350

70

300

60

250

50

200

40

150

30

100

20

50

10

0

0

Nitrite and nitrate concentration, mg N/L

were favorable to SNA. Tao et al. (2011) examined the influence of ammonium loading rate on SNA in free-water surface wetlands. In single constructed wetlands, NOB compete with AOB for dissolved oxygen and with anammox bacteria for nitrite. Therefore, NOB should be inhibited for enhancement of SNA. Temperatures higher than 30
C and pH between 7.8 and 8.5 have been used to inhibit NOB and promote AOB in bioreactors (Bae et al., 2001; Hellinga et al., 1998; Park et al., 2007; van der Star et al., 2007). Anammox has an optimal pH of 7.5e8.0 and optimum temperature around 30e40
C (Strous et al., 1997; van de Graaf et al., 1996; van der Star et al., 2007). Unlike traditional bioreactors, constructed wetlands as natural wastewater treatment systems are subject to seasonal temperature variations. To date, it remains unclear about the effects of pH and seasonal temperature variation on the entire SNA process in single constructed wetlands. This study tracked nitrogen removal and operational conditions in two pilot-scale free-water surface wetlands enhancing SNA for dairy wastewater treatment. Furnace slag was utilized in the freewater surface wetlands as a low-cost method to buffer pH of wetland water. The main objectives of this study were: 1) to examine the effectiveness of SNA for nitrogen removal from dairy wastewater in free-water surface wetlands; and 2) to evaluate the effects of pH and seasonal temperature variation on SNA in freewater surface wetlands. Microbial community composition and the contribution of plant assimilation to nitrogen removal were also investigated in the wetlands.

Ammonium concentration, mg N/L

104

Free water surface wetland (CWS)

Fig. 1. Experimental setup of two free-water surface wetlands operated in parallel at different water pH levels elevated by furnace slag.

Fig. 2. Nitrogen concentrations of the influent fed to the free-water surface wetlands.

Y. He et al. / Journal of Environmental Management 110 (2012) 103e109

cleaning, stem diameter and total height were measured. The plant samples were then dried at 80
C for 24 h and weighed. A regression equation was developed to estimate the above-ground plant biomass (Fig. S1):

Mp ¼ 3:5  H  1:76;

R2 ¼ 0:65

(1)

where Mp ¼ above-ground biomass of individual plants (dry weight), g; and H ¼ above-ground height of individual plants, m. 2.3. Fluorescence in situ hybridization Fluorescence in situ hybridization (FISH) was performed to identify microbial community composition. Sediment samples were taken in triplicate (top 20 mm) in the two wetlands on 1 December 2010. Biomass samples were immediately immersed in freshly prepared 4% (w/v) paraformaldehyde solution for 2 h. Microorganisms were detached from sediment particles by 1-min sonication (45 W, 50/60 Hz) 3 times (Schaule et al., 2000). Cell suspension was washed twice with 3 phosphate-buffered saline. The cell pellets were then re-suspended in 3 mL of 1 phosphatebuffered saline and 3 mL of absolute ethanol, and stored at 20
C. AOB and anammox bacteria were semi-quantified by FISH according to Amann et al. (1995). Group-specific 16S rRNA gene probes EUB338I labeled with FI, Nso190 labeled with 5HEX, and Amx368 labeled with Cy5 (Eurofins MWG Operon, Huntsville, Alabama, USA) were used to detect most eubacteria, b-proteobacterial AOB, and all anammox bacteria, respectively (Nielsen et al., 2009). Each sample was prepared in duplicate for hybridization. Hybridization was conducted for 2 h at 46
C. Following hybridization and washing, the slides were mounted in Citifluor (Citifluor Ltd., London, UK). FISH micrographs were captured digitally with a Zeiss LSM510 scanning confocal/two photon microscope system (Carl Zeiss, Inc., Germany). Relative abundance was quantified for AOB and anammox bacteria as area percentage of group-specific bacteria in total bacteria using Image-Pro Plus 6.0 software (Media Cybernetics, Inc., Bethesda, Maryland, USA). Total bacteria were quantified approximately with EUB338-positive and Amx368-positive areas.

105

difference between the means of the two wetlands if P  0.05. Correlation analysis was performed to test the effects of temperature on SNA and the trends of nitrogen removal rates over time, outputting the Pearson’s product moment correlation coefficient, r. 3. Results and discussion 3.1. Microbial processes for nitrogen removal The free-water surface wetlands had low redox potentials and dissolved oxygen concentrations (Table 1), indicating conditions that were suitable for coupling partial nitrification and anammox (Gaul et al., 2005; Tao et al., 2011; van der Star et al., 2007). Microorganisms in free-water surface wetlands may attach to sediment particles, plant stems, and litter, forming biofilms. Microgradients of oxygen throughout biofilms allow partial nitrification to occur in the outer layers and anammox in the inner layers of biofilms (Kartal et al., 2010). FISH analysis detected abundant anammox bacteria and AOB in the wetland sediment (Figures S2 and S3). On average, anammox bacteria and AOB together accounted for the majority of the bacterial populations (Table 2), indicating that SNA played a significant role in biological nitrogen removal in the free-water surface wetlands. Compared with previous studies that employed SNA in constructed wetlands (Chiemchaisri et al., 2009; Dong and Sun, 2007; Paredes et al., 2007), this study achieved substantially higher relative abundances of both AOB and anammox bacteria. The rest of the bacteria were possibly denitrifying bacteria and NOB. When organic carbon is available for denitrification, nitrite produced by AOB may be used for denitrification instead of anammox. Anammox bacteria are outcompeted by denitrifying bacteria at higher COD concentrations, being reported mostly between 237 mg/L and 1600 mg/L (Chamchoi et al., 2008; DapenaMora et al., 2007; Dong and Tollner, 2003; Molinuevo et al., 2009; Ruscalleda et al., 2008; Tang et al., 2010). The dairy wastewater in the present study had 70e149 mg/L of COD, indicating a low C/N ratio that could limit competition of denitrification with anammox for nitrite. Free ammonia is the real substrate of AOB and a high concentration of free ammonia inhibits NOB (Wiesmann, 1994; Chung

2.4. Data analysis Nitrogen removal performance of the wetlands was evaluated with equation (2):

MRR ¼ ðCi  Vi  Ce  Ve Þ=T=As

Table 1 Operational conditions of free-water surface wetlands.a Wetland CW

(2)

Influent

where MRR ¼ areal mass removal rate of ammonium or TIN, g N/m2/d; Ci, Ce ¼ concentrations of nitrogen in influent and effluent, respectively, mg N/L; Vi, Ve ¼ volumes of influent and effluent, respectively, m3; T ¼ cycle time, d; and As ¼ wetland surface area, m2. The contribution of plant uptake to nitrogen removal was calculated in equation (3):

Water pH Redox potential, mV Dissolved oxygen, mg/L Water temperature,
C Volume, L Ammonium, mg N/L Total inorganic N, mg N/L

Rp ¼ DM  2%=As =D

Water pH Redox potential, mV Dissolved oxygen, mg/L Water temperature,
C Volume, L Ammonium, mg N/L Total inorganic N, mg N/L

(3)

where Rp ¼ rate of plant assimilation for nitrogen, g N/m2/d; DM ¼ increase of plant biomass (dry weight), g; 2% ¼ nitrogen content in common reeds (Kadlec and Wallace, 2009); As ¼ wetland surface area, m2; and D ¼ # of days between two measurements of plant growth, d. One-way ANOVA was performed to test the differences in operational conditions and nitrogen removal performance between the two wetlands, giving P-values. There was a significant

Wetland CWS Effluent

Influent

Effluent

Warm season (July 5eSeptember 21) 7.64  0.18 7.00  0.62 7.70  0.10 153  8 199  37 147  8 1.72  0.68 1.22  0.38 1.22  0.88 26.9  4.6 21.5  3.8 26.3  3.9 29.3  0.4 11.7  4.6 29.3  0.4 267  5 329  92 267  5 272  4 352  97 272  4

7.45 175 0.90 21.7 9.3 377 428

      

0.57 29 0.39 3.5 4.0 100 115

Cool season 8.00  0.14 122  8 0.84  0.37 17.9  2.1 28.7  1.2 252  38 266  41

7.98 131 0.39 17.0 22.3 246 261

      

0.10 21 0.25 1.3 2.4 50 54

(September 21eMarch 10) 7.68  0.17 7.97  0.14 143  20 124  8 0.37  0.28 0.91  0.28 17.0  1.3 17.8  2.0 24.6  1.6 28.7  1.2 239  48 252  38 254  50 266  41

a Data in mean  standard deviation. n ¼ 17e24 for influent and 20e27 for effluent. The effluent concentrations were higher than the influent concentrations in the warm season because the effluent volumes were greatly lower than the influent volumes due to evapotranspiration.

106

Y. He et al. / Journal of Environmental Management 110 (2012) 103e109

Table 2 Relative abundance of AOB and anammox bacteria in surface sediment of free water surface wetlands treating dairy wastewater.a

Anammox bacteria AOB

Wetland CWS 18e26% 24e47%

Ranges of relative abundance based on 6e9 fields of view per slide.

et al., 2006). Several studies found that a free ammonia concentration higher than 0.2e1 mg N/L inhibits nitrite oxidation substantially and a concentration up to 50 mg N/L only decreases AOB activity slightly (Abeling and Seyfrid, 1992; Balmelle et al., 1992; Chung et al., 2006). Free ammonia concentration was mostly between 0.33 mg N/L and 10.8 mg N/L in the free-water surface wetlands, suggesting that NOB was inhibited while AOB activity was maintained. 3.2. Nitrogen removal performance This study attained higher nitrogen removal rates by enhancing SNA in free-water surface wetlands. The average TIN removal rates were 3.37 g N/m2/d in the warm season from 5 July to 21 September 2010 and 1.30 g N/m2/d in the cool season from 21 September 2010 to 10 March 2011 (Fig. 3). As Vymazal (2007) reviewed, free-water surface wetlands removed 0.68 g N/m2/d of TIN on average. As Kadlec and Wallace (2009) reviewed, only 20% of free-water surface wetlands could remove more than 1.51 g N/m2/d of total nitrogen. Recent studies that integrated partial nitrification and anammox in free-water surface wetlands removed 0.80e1.13 g N/m2/d of TIN (Dong and Sun, 2007; Tao and Wang, 2009; Tao et al., 2011). The wetland influent had low nitrite concentrations (Fig. 2). The þ influent NO 2 /NH4 ratio (0.02) was much lower than the stoichiometric ratio (1.32) for anammox (Strous et al., 1998), suggesting that nitrite was the limiting substrate of anammox bacteria. Consequently, NO 2 produced by AOB could be consumed immediately by anammox. A small nitrite accumulation rate (0.04e0.06 g N/m2/d on average) suggested that anammox was possibly the limiting stage of SNA in the free-water surface wetlands. Nitrate removal rates were low (0.03 g N/m2/d on average), which were likely restrained by the lack of bioavailable organic carbon for heterotrophic denitrifying bacteria. Therefore, the TIN removal rates were similar to the ammonium removal rates in the two wetlands (Fig. 3).

Ammonium removal rate, g N/m2/d

5

5 Wetland CWS

Wetland CWS Wetland CW

4 3 2 1 0

TIN removal rate, g N/m2/d

a

Wetland CW 19e38% 12e27%

Vanderzaag et al. (2008) measured ammonia volatilization from June to September in three free-water surface wetlands treating dairy wastewater. The ammonium concentration (284e314 mg N/L at inlet; 229e258 at outlet), pH (7.4e7.7 at inlet and 7.2e7.5 at outlet), and temperature (14.3e18.5
C at outlet) reported by Vanderzaag et al. (2008) were similar to those in the present study (Table 1). The monthly average ammonia volatilization fluxes determined by Vanderzaag et al. (2008) were 0.035 0.596 g N/m2/d, averaging at 0.322 g N/m2/d. Vanderzaag et al. (2008) measured ammonia volatilization using steady-state chambers at a constant airflow of 0.075 m3/m2/s. Volatilization of free ammonia out of wetland water could be enhanced by a constant airflow. The free-water surface wetlands in the present study were operated in a wind-free greenhouse. The effluent concentrations of free ammonia were similar in the warm and cool seasons, i.e., 7.2 and 7.3 mg N/L respectively in wetland CWS, and 2.5 and 3.8 mg N/L respectively in CW. It was hence estimated that ammonia volatilization flux in this study was less than 0.322 g N/m2/d, which accounted for less than 10% of the TIN removal rate in the warm season and less than 25% in the cool season. The average effluent concentrations of ammonium and TIN were much higher in the warm season than those in the cool season (Table 1). Nitrogen removal rate could be higher at a higher concentration. TIN removal in free-water surface wetlands enhancing SNA followed first-order reaction kinetics at 60 mg N/L (Tao and Wang, 2009) and zero-order kinetics at 90 mg N/L and 200 mg N/L (Tao et al., 2011). Therefore, the high ammonium and TIN removal rates in the warm season of the present study could be partially attributed to the effluent concentrations elevated by evapotranspiration. In field applications where the discharge permit limits or discharge standards are concentration-based instead of mass-based, additional treatment will be required. The measurements on plant growth in the wetlands (Table 3) showed that common reeds grew in terms of stem number and height until late September when temperature decreased sharply (Fig. 4). Plants began to die from late September due to low temperature and aphid infestation. Below-ground biomass was approximated to be the same as the above-ground biomass in the growing season (Gottschall et al., 2007). Nitrogen removal rate via net plant growth was subsequently estimated to be 0.40e0.64 g N/m2/d in wetland CWS and 0.08e1.51 g N/m2/d in wetland CW, accounting for 12e19% and 2e45% of TIN removal in CWS and CW in the warm season, respectively. Ammonium concentration was much higher than nitrate concentration in the

Wetland CW

4 3 2 1 0

Year 2010

Year 2011

Year 2010

Fig. 3. Seasonal variations of nitrogen removal in the free-water surface wetlands.

Year 2011

Y. He et al. / Journal of Environmental Management 110 (2012) 103e109

107

Table 3 Variation of plant growth in free water surface wetlands. Date in 2010

11 Jun 19 Jul 9 Aug 1 Oct 7 Oct 8 Nov 14 Dec

Wetland CW

Wetland CWS

Above-ground height, cm

Green stem number

Above-ground biomass, g

Above-ground height, cm

Green stem number

Above-ground biomass, g

94 157 158 150 110 102 96

44 91 108 122 73 57 38

66.6 337.3 402.2 422.5 151.2 102.2 60.2

94 148 128 125 120 93 92

34 49 69 118 89 72 66

51.5 165.6 184.8 306.0 215.3 107.5 95.3

which could be attributed to the decreasing plant biomass and influent ammonium concentration (Table 3 and Fig. 2). 3.3. Effects of seasonal temperature variation Temperatures higher than 30
C are optimal for anammox and favor AOB over NOB in bioreactors (Hellinga et al., 1998; van der Star et al., 2007). Water temperature in the free-water surface wetlands varied between 13.8
C and 24.9
C (Fig. 4), below the optimal temperatures. Therefore, both ammonium and TIN removal rates (excluding plant uptake) correlated positively with temperature in the two free-water surface wetlands (Fig. 5). Nevertheless,

28

9.0

24

8.5 8.0

20

Effluent pH

Effluent temperature, oC

two wetlands, suggesting that ammonium was the primary nitrogen form for plant uptake (Kadlec and Wallace, 2009; Vymazal, 2007). Therefore, the effects of pH and temperature variation on SNA were discussed hereinafter with ammonium and TIN removal rates excluding plant uptake of ammonium. Ammonium and TIN removal rates excluding plant uptake had weak correlations with time or number of batch operation cycles in the cool season (r ¼ 0.47 to 0.70). The total ammonium and TIN removal rates of wetland CW also had weak correlations with time (r ¼ 0.52 and 0.50 respectively). The total ammonium and TIN removal rates of wetland CWS appeared to correlate with cycle number in the cool season (r ¼ 0.80 and 0.82 respectively),

16 12 8 4

7.5 7.0 6.5 6.0 Wetland CWS

Wetland CWS

5.5

Wetland CW

Wetland CW

5.0

0

Year 2010

Year 2010

Year 2011

Year 2011

Fig. 4. Dynamics of water pH and temperature in the free-water surface wetlands.

6 Wetland CWS Wetland CW

5

TIN removal rate, g N/m2/d

Ammonium removal rate, g N/m2/d

6

4 CWS: r = 0.80

3 2

CW: r = 0.73

1 0

Wetland CWS Wetland CW

5 4

CWS: r = 0.76

3 2

CW: r = 0.73

1 0

10

15

20 25 Temperature, oC

30

10

15

20 25 Temperature, oC

30

Fig. 5. Relationships between microbial nitrogen removal (excluding plant uptake) and temperature in the free-water surface wetlands.

108

Y. He et al. / Journal of Environmental Management 110 (2012) 103e109

there were insignificant correlations between microbial nitrogen removal and temperature either in the warm season at temperatures of 18.9e24.9
C or the cool season at temperatures of 13.8e18.9
C (r ¼ 0.06e0.49). These statistical analyses implied that SNA was not affected by the seasonal temperature variation in the free-water surface wetlands unless there was a temperature change of more than 6
C in the range of 13.8e24.9
C. The relative insensitivity of SNA to seasonal temperature variation could be attributed to the slow growth of anammox bacteria (Strous et al., 1998; van der Star et al., 2007).

Startup Fund. We thank Dr. Arvydas Matiukas for allowing us to use his confocal core and providing training and technical services.

3.4. Effects of pH

Abeling, U., Seyfrid, C.,1992. Anaerobiceaerobic treatment of high strength ammonium wastewater e nitrogen removal via nitrite. Water Sci. Technol. 26, 1007e1015. Amann, R., Ludwig, W., Schleifer, K.H., 1995. Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiol. Rev. 59, 143e169. APHA, AWWA, WEF, 1998. Standard Methods for the Examination of Water and Wastewater, twentieth ed. American Public Health Association (APHA), American Water Works Association (AWWA), and Water Environment Federation (WEF), Washington, DC. Bae, W., Baek, S., Chung, J., Lee, Y., 2001. Optimal operational factors for nitrite accumulation in batch reactors. Biodegradation 12, 359e366. Balmelle, B., Nguyen, K., Capdeville, B., Cornier, J.C., Deguin, A., 1992. Study of factors controlling nitrite build-up in biological processes for water nitrification. Water Sci. Technol. 26, 1017e1025. Camargo Valero, M.A., Mara, D.D., 2010. Ammonia volatilisation in waste stabilisation ponds: a cascade of misinterpretations? Water Sci. Technol. 61 (3), 555e561. Chamchoi, N., Nitisoravut, S., Schmidt, J.E., 2008. Inactivation of ANAMMOX communities under concurrent operation of anaerobic ammonium oxidation (ANAMMOX) and denitrification. Bioresour. Technol. 99, 3331e3336. Chiemchaisri, C., Chiemchaisri, W., Junsod, J., Threedeach, S., Wicranarachchi, P.N., 2009. Leachate treatment and greenhouse gas emission in subsurface horizontal flow constructed wetland. Bioresour. Technol. 100, 3808e3814. Chung, J., Shim, H., Park, S., 2006. Optimization of free ammonia concentration for nitrite accumulation in shortcut biological nitrogen removal process. Bioprocess Biosyst. Eng. 28, 275e282. Dapena-Mora, A., Fernandez, I., Campos, J.L., Mosquera-Corral, A., Mendez, R., Jetten, M.S.M., 2007. Evaluation of activity and inhibition effects on anammox process by batch tests based on the nitrogen gas production. Enzym. Microb. Technol. 40, 859e865. Dong, Z., Sun, T., 2007. A potential new process for improving nitrogen removal in constructed wetlands e promoting coexistence of partial nitrification and ANAMMOX. Ecol. Eng. 31, 69e78. Dong, X., Tollner, E.W., 2003. Evaluation of anammox and denitrification during anaerobic digestion of poultry manure. Bioresour. Technol. 86, 139e145. Gaul, T., Marker, S., Kunst, S., 2005. Start-up of moving bed biofilm reactors for deammonification: the role of hydraulic retention time, alkalinity and oxygen supply. Water Sci. Technol. 52, 127e133. Gottschall, N., Boutin, C., Crolla, A., Kinsley, C., Champagne, P., 2007. The role of plants in the removal of nutrients at a constructed wetland treating agricultural (dairy) wastewater, Ontario, Canada. Ecol. Eng. 29, 154e163. Hellinga, C., Schellen, A., Mulder, J., van Loosdrecht, M., Heijnen, J., 1998. The SHARON process: an innovative method for nitrogen removal from ammoniumrich waste water. Water Sci. Technol. 37, 135e142. Jaroszynski, L.W., Cicek, N., Sparling, R., Oleszkiewicz, J.A., 2011. Importance of the operating pH in maintaining the stability of anoxic ammonium oxidation (anammox) activity in moving bed biofilm reactors. Bioresour. Technol. 102, 7051e7056. Jung, J.Y., Kang, S.H., Chung, Y.C., Ahn, D.H., 2007. Factors affecting the activity of anammox bacteria during start up in the continuous culture reactor. Water Sci. Technol. 55, 459e468. Kadlec, R.H., Wallace, S.D., 2009. Treatment Wetlands, second ed. CRC Press, Boca Raton, Florida. Kartal, B., Kuenen, J.G., van Loosdrecht, M.C.M., 2010. Sewage treatment with anammox. Science 328, 702e703. Knight, R.L., Payne, V.W.E., Borer, R.E., Clarke, R.A., Pries, J.H., 2000. Constructed wetlands for livestock wastewater management. Ecol. Eng. 15, 41e55. Molinuevo, B., Garcia, M.C., Karakashev, D., Angelidaki, I., 2009. Anammox for ammonia removal from pig manure effluents: effect of organic matter content on process performance. Bioresour. Technol. 100, 2171e2175. Nielsen, P.H., Daims, H., Lemmer, H., 2009. FISH Handbook for Biological Wastewater Treatment: Identification and Quantification of Microorganisms in Activated Sludge and Biofilms by FISH. IWA Publishing, London, UK. Paredes, D., Kuschk, P., Koser, H., 2007. Influence of plants and organic matter on the nitrogen removal in laboratory-scale model subsurface flow constructed wetlands inoculated with anaerobic ammonium oxidizing bacteria. Eng. Life Sci. 7, 565e576. Park, S., Bae, W., Chung, J., Baek, S., 2007. Empirical model of the pH dependence of the maximum specific nitrification rate. Process Biochem. 42, 1671e1676. Ruscalleda, M., Lopez, H., Ganigue, R., Puig, S., Balaguer, M.D., Colprim, J., 2008. Heterotrophic denitrification on granular anammox SBR treating urban landfill leachate. Water Sci. Technol. 58, 1749e1755.

The two free-water surface wetlands had similar influent pH values (P ¼ 0.99), and significantly different effluent pH values (P ¼ 0.01) because more furnace slag was put on the rooting substrate of wetland CWS. Coincidently, wetland CWS with higher effluent pH values had a higher AOB abundance in relative to wetland CW with lower effluent pH values (Tables 1 and 2) since the optimum pH values to favor AOB over NOB are between 7.8 and 8.5 (Bae et al., 2001; Hellinga et al., 1998; Park et al., 2007). The two wetlands had pH values mostly in the optimal range for anammox bacteria, pH 7.5e8.0 (Strous et al., 1997; van de Graaf et al., 1996). However, free ammonia concentration in wetland CWS was increased by the higher pH values to 2.3e10.8 mg N/L, which could inhibit anammox. Free ammonia concentrations at 1.7e8.3 mg N/L have been reported to inhibit anammox (Jaroszynski et al., 2011; Jung et al., 2007). Free ammonia concentration in wetland CW was usually between 0.33 mg N/L and 6.8 mg N/L. Consequently, a lower relative abundance of anammox bacteria was detected in wetland CWS than CW (Table 2). Despite the significant difference in effluent pH, there were insignificant differences in ammonium and TIN removal rates between the two wetlands throughout the entire operational period (P ¼ 0.28 and 0.29) and in the warm season (P ¼ 0.57 and 0.84). Nevertheless, the two wetlands had significantly different ammonium and TIN removal rates in the cool season from 21 September 2010 to 10 March 2011 (P ¼ 0.05 and 0.03) as shown in Fig. 3. This suggested that temperature was the primary factor influencing SNA in the wetlands. 4. Conclusions This study demonstrated that simultaneous partial nitrification and anammox could be employed under oxygen-limited conditions in free-water surface wetlands for efficient nitrogen removal from dairy wastewater. A seasonal temperature increase of more than 6
C can significantly improve ammonium and TIN removal via the simultaneous partial nitrification and anammox process. A pH between 7.5 and 7.8 favors partial nitrification over nitrite oxidation, but pH should be controlled along with ammonium concentration and temperature to avoid toxicity of free ammonia to anammox bacteria. Furnace slag, a waste product, can be used as a solid-phase pH buffer for enhancement of simultaneous partial nitrification and anammox in free-water surface wetlands. Plant assimilation could account for up to 45% of nitrogen removal in the free-water surface wetlands treating ammoniumrich dairy wastewater in the growing season. Acknowledgments This study was supported in part by USDA Cooperative Forest Research, State Graduate Assistantships, and SUNY ESF Faculty

Appendix A. Supporting information Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jenvman.2012. 06.009. References

Y. He et al. / Journal of Environmental Management 110 (2012) 103e109 Schaule, G., Griebe, T., Flemming, H.-C., 2000. Steps in biofilm sampling and characterization in biofouling cases. In: Flemming, H.-C., Szewzyk, U., Griebe, T. (Eds.), Biofilms: Investigative Methods & Applications. Technomic Publishing Company, Inc., Lancaster, PA, USA, pp. 1e22. Strous, M., Van Gerven, E., Zheng, P., Gijs Kuenen, J., Jetten, M.S.M., 1997. Ammonium removal from concentrated waste streams with the anaerobic ammonium oxidation (anammox) process in different reactor configurations. Water Res. 31, 1955e1962. Strous, M., Heijnen, J.J., Kuenen, J.G., Jetten, M.S.M.,1998. The sequencing batch reactor as a powerful tool for the study of slowly growing anaerobic ammoniumoxidizing microorganisms. Appl. Microbiol. Biotechnol. 50, 589e596. Tang, C., Zheng, P., Wang, C., Mahmood, Q., 2010. Suppression of anaerobic ammonium oxidizers under high organic content in high-rate anammox UASB reactor. Bioresour. Technol. 101, 1762e1768. Tao, W., Wang, J., 2009. Effects of vegetation, limestone and aeration on nitritation, anammox and denitrification in wetland treatment systems. Ecol. Eng. 35, 836e842.

109

Tao, W., Wen, J., Huchzermeier, M., 2011. Batch operation of biofilter e free water surface wetland series for enhancing nitritation and anammox. Water Environ. Res. 83, 541e548. van de Graaf, A.A., der Bruijn, P., Robertson, L.A., Jetten, M.S.M., Kuenen, J.G., 1996. Autotrophic growth of anaerobic ammonium-oxidizing microorganisms in a fluidized bed reactor. Microbiol. (Reading, UK) 142, 2187e2196. van der Star, W.R.L., Abma, W., Blommers, D., Mulder, J.W., Tokutomi, T., Strous, M., Picioreanu, C., van Loosdrecht, M.C.M., 2007. Startup of reactors for anoxic ammonium oxidation: experiences from the first full-scale anammox reactor in Rotterdam. Water Res. 41, 4149e4163. Vanderzaag, A.C., Gorden, R.J., Burton, D.L., Jamieson, R.C., Stratton, G.W., 2008. Ammonia emissions from surface flow and subsurface flow constructed wetlands treating dairy wastewater. J. Environ. Qual. 37, 2028e2036. Vymazal, J., 2007. Removal of nutrients in various types of constructed wetlands. Sci. Total Environ. 380, 48e65. Wiesmann, U., 1994. Biological nitrogen removal from wastewater. Adv. Biochem. Eng. Biotechnol. 51, 113e154.