Nitrogen removal in Myriophyllum aquaticum wetland

Research Article Received: 7 January 2016

Revised: 2 April 2016

Accepted article published: 11 April 2016

Published online in Wiley Online Library: 9 May 2016

(wileyonlinelibrary.com) DOI 10.1002/jsfa.7752

Nitrogen removal in Myriophyllum aquaticum wetland microcosms for swine wastewater treatment: 15N-labelled nitrogen mass balance analysis Shunan Zhang,a,b Feng Liu,a,b* Runlin Xiao,a,b Yang Hea,b,c and Jinshui Wua,b Abstract BACKGROUND: Ecological treatments are effective for treating agricultural wastewater. In this study, wetland microcosms vegetated with Myriophyllum aquaticum were designed for nitrogen (N) removal from two strengths of swine wastewater, and 15 N-labelled ammonium (NH4 + -N) was added to evaluate the dominant NH4 + -N removal pathway. RESULTS: The results showed that 98.8% of NH4 + -N and 88.3% of TN (TN: 248.6 mg L−1 ) were removed from low-strength swine wastewater (SW1) after an incubation of 21 days, with corresponding values for high-strength swine wastewater (SW2) being 99.2% of NH4 + -N and 87.8% of TN (TN: 494.9 mg L−1 ). Plant uptake and soil adsorption respectively accounted for 24.0% and 15.6% of the added 15 N. Meanwhile, above-ground tissues of M. aquaticum had significantly higher biomass and TN content than below-ground (P < 0.05). 15 N mass balance analysis indicated that gas losses contributed 52.0% to the added 15 N, but the N2 O flux constituted only 7.5% of total gas losses. The dynamics of NO3 − -N and N2 O flux revealed that strong nitrification and denitrification occurred in M. aquaticum microcosms, which was a dominant N removal pathway. CONCLUSION: These findings demonstrated that M. aquaticum could feasibly be used to construct wetlands for high N-loaded animal wastewater treatment. © 2016 Society of Chemical Industry Keywords: 15 N-labelled nitrogen; Myriophyllum aquaticum; swine wastewater; plant uptake; gas emission

INTRODUCTION

J Sci Food Agric 2017; 97: 505–511

∗

Correspondence to: F Liu, No. 644 the Second Yuanda Road, Furong District, Changsha, Hunan 410125, People’s Republic of China. E-mail: [email protected]

a Key Laboratory of Agro-ecological Processes in Subtropical Regions, Institute of Subtropical Agriculture, Chinese Academy of Sciences, Hunan 410125, People’s Republic of China b Changsha Research Station for Agricultural and Environmental Monitoring, Institute of Subtropical Agriculture, Chinese Academy of Sciences, Hunan 410125, People’s Republic of China c Graduate University of Chinese Academy of Sciences, Beijing 100039, People’s Republic of China

www.soci.org

© 2016 Society of Chemical Industry

505

In recent years, environmental pollution problems caused by swine production have created substantial concern.1,2 The discharge of untreated swine wastewater results in excessive nutrient transport to streams, rivers and lakes, in turn inducing eutrophication of surface water.3 Furthermore, groundwater pollution has also been shown to occur as a result of swine wastewater discharges.4 Given that China is one of the world’s largest pork producers, the treatment of swine wastewater is thus becoming a matter of urgency for the country. The swine industry is located largely in rural areas in China, where it is not feasible to build sewage treatment plants given the high construction and operation costs involved. However, constructed wetlands (CWs), based on the simple operation and low implementation costs, have been widely used to treat different types of wastewater, including domestic sewage, industrial wastewater, acid mine drainage and agricultural pollution.5 – 7 CWs are considered as complex bioreactors, where numerous physical, chemical and biological processes occur. Plant uptake, microbial nitrification–denitrification and soil adsorption are three main pathways for nitrogen (N) removal.8 – 10 Plant uptake can play an important role in removing N from wastewater in CWs,11 but decomposition of plant litter would release assimilated N to the environment. With gaseous products of NO, N2 O, and N2 , microbial

nitrification–denitrification is considered as permanent N removal pathway in CWs. To understand the degree to which the individual processes of plant uptake, microbial nitrification–denitrification, and soil adsorption contribute to N removal is a key question for optimizing the design of full-scale CWs.12 Stable isotope analyses have been used increasingly to discern N transformation pathways and rates in recent years.13 For example, 15 N isotope mass balance analysis identified the contributions of plant assimilation, microbial activity and sedimentation processes to N removal in CWs.12 Swine wastewater is characterized by high levels of ammonia nitrogen (NH4 + -N), suspended solids and organic carbon. High concentrations of NH4 + -N inhibit plant growth and lead to the

www.soci.org development of NH4 + -N toxicity syndrome.14,15 The toxicity of NH4 + -N to aquatic plants has limited the extensive use of CWs for swine wastewater treatment.16 Myriophyllum aquaticum is a perennial floating or submerged plant, and grows well in shallow gullies and wetlands, especially in nutrient-rich water.17 Myriophyllum aquaticum has a high potential to assimilate N from water and soil.18 Nevertheless, to date, few studies have reported on the effectiveness of M. aquaticum for N removal in the treatment of swine wastewater with high NH4 + loading. In the present study, M. aquaticum-based wetland microcosms were constructed to investigate N removal from swine wastewater using the 15 NH4 + -based isotopic tracing method. The main aims were: (i) to analyse N uptake and the distribution in the tissues of M. aquaticum; (ii) to determine the dynamics of NO3 − -N concentrations and N2 O emissions to investigate nitrification–denitrification processes in N removal; and (iii) to identify the dominant NH4 + -N removal pathway in M. aquaticum-based wetland.

MATERIALS AND METHODS Setup of wetland microcosms The wetland microcosms consisted of organic glass containers with dimensions of 50 cm × 40 cm × 50 cm (length × width × height). Air-dried rice soil was placed in each container as the substrate, with an approximate depth of 5 cm. The soils used in the study were loam Ultisols (USDA Soil Taxonomy) with a soil pH of 6.7, total nitrogen (TN) of 1.9 g kg−1 , soil organic carbon (SOC) of 21.9 g kg−1 , and sand, silt and clay content of 32.6%, 41.1% and 26.3%, respectively. Each container was planted with M. aquaticum transplanted from a constructed drainage ditch. The selected M. aquaticum specimens were similar in age and size, with a plant density of approximately 500 rhizomes m−2 . The initial fresh biomass in each container ranged from 1103.5 to 1190.0 g m−2 . Experimental operation The experiment was carried out in a greenhouse during the period 1–21 October 2013. Throughout, air temperature in the greenhouse was in the range of 19.3–35.4 ∘ C. Myriophyllum aquaticum was first pre-cultured for 7 days to allow for adaptation to the microcosms. Subsequently, 15 L swine wastewater was added to each M. aquaticum microcosm, keeping the water depth at 7.5 cm. Two N-loading levels of low-strength swine wastewater (SW1) and high-strength swine wastewater (SW2) were selected in this experiment. SW1 was prepared by diluting SW2 with distilled water at a 1:1 (v/v) ratio. SW2 was raw swine wastewater collected from an anaerobic lagoon, which contained 439.4 mg L−1 NH4 + -N, 0.5 mg L−1 NO3 − -N, 494.9 mg L−1 TN and 34.3 mg L−1 TP, with a pH value of 7.8. To investigate the fate of NH4 + -N in the M. aquaticum microcosm, 150 mg 15 N-labeled ammonium sulfate (20.27 atom% 15 N excess) was added to the M. aquaticum microcosm with SW1. In order to eliminate isotope effects, the SW1 treatment included six replicate M. aquaticum microcosms, with or without the addition of 15 N (each in triplicate). Three replicates were also arranged for the SW2 treatment. Throughout the experimental period, water evaporation losses were supplemented with distilled water 1 day before sample collection.

506

Sampling and analysis Water samples were collected on days 0, 1, 4, 7, 10, 14, 17 and 21. 150 mL swine wastewater was collected from each M.

wileyonlinelibrary.com/jsfa

S Zhang et al.

aquaticum microcosm using a 50 mL syringe. Water samples were then immediately returned to the laboratory for analysis within 24 h. Following filtering through 0.45 μm membranes, NH4 + -N and NO3 − -N concentrations were measured using an automatic flow injection analyser (Fia-star 5000, Foss Tecator, Sweden). TN was measured using the alkaline potassium persulfate oxidation method,19 with oxidized NO3 − -N analysed via the flow injection analyser. Water samples for isotopic composition of TN were first freeze-dried to a solid state and then analysed by mass spectrometry (Thermo Scientific MAT 253, USA). Water pH and dissolved (DO) were determined in situ using a portable multi-parameter meter (SG68-ELK-ISM, Mettler-Toledo, Switzerland). N2 O gas samples were collected between 9:00 and 9:30 am on days 1, 4, 7, 10, 14, 17 and 21. N2 O fluxes were measured using a closed-chamber method. The chambers (50 cm length × 40 cm width × 10 cm height), composed of organic glass, were fitted within the groove of the microcosms. When sampling, a chamber was placed on the groove (sealed with a water lock) and vegetation was included within the chamber. For subsequent correction of N2 O fluxes, air temperatures in the chambers were gauged before gas sampling. For each M. aquaticum microcosm, two gas samples were collected from the headspace of the chamber into evacuated 12 mL vials (Labco, UK) at 0 and 30 min for N2 O analysis. Meanwhile, another two gas samples were collected into evacuated 100 mL vials for 15 N2 O enrichment analysis. N2 O concentrations of gas samples were measured using a gas chromatograph (7890A, Agilent, USA) equipped with an automatic sample injector system and an electron capture detector (ECD), with high-purity N2 as the carrier gas. Detailed calculation methods for N2 O fluxes are described in Li et al.20 To determine 15 N enrichment of N2 O, 44 [N2 O]/45 [N2 O] mass ratios were measured using a mass spectrometer (MAT 253, Thermo Scientific, USA). About 100 g fresh soil was sampled from each M. aquaticum microcosm with a cylindrical soil auger, on days 0, 1, 7, 14 and 21. Impurities in the soil were first removed from each soil sample, which were then evenly mixed. Soil water content was gravimetrically measured by drying a portion of the fresh soil at 105 ∘ C for 48 h. NH4 + -N and NO3 − -N in the soil samples were extracted using a 2 mol L−1 KCl solution. The suspensions were first filtered with qualitative filter papers, and then used for analysis of NH4 + -N and NO3 − -N content using a Fia-star 5000 flow injection analyser (Foss Tecator, Sweden). Soil TN content was determined using the semi-micro Kjeldahl digestion method, and transformed NH4 + -N was measured via the automatic flow injection system. Mass spectrometry was used to determine 15 N enrichment of TN in soil (Thermo Scientific MAT 253, USA), after NH4 + -N was converted into nitrogen gas (N2 ). At the end of the experiment, M. aquaticum was harvested from the incubation containers, and soil washed off using distilled water. Because the adventitious roots of M. aquaticum formed on rhizomes in the water column or in the soil,21 the plant tissues under the water surface were used as below-ground parts, and the rest as above-ground parts. The fresh biomass of different parts of the plant was first weighed, and then all plant samples were oven-dried at 70 ∘ C until constant weight was obtained. The dry biomass of each sample was recorded for the calculation of relative growth rate (RGR). Detailed calculation methods for RGR are described in Zheng et al.22 The dried plant tissues were ground to a powder and filtered using a 1 mm sieve. The TN content in plant tissues was measured using the automatic flow injection analyser (Fia-star 5000, Foss Tecator, Sweden), after digestion in an H2 SO4 –H2 O2 solution. A Thermo Scientific MAT 253 mass


J Sci Food Agric 2017; 97: 505–511

Nitrogen removal in Myriophyllum aquaticum wetland microcosms

www.soci.org

spectrometer was used to measure the 15 N composition of TN in plant tissues. Statistical analysis The percentage of plant N uptake from swine wastewater (Nfw %) or other sources (Nfs %) was respectively calculated using Eqns (1) and (2): P × 100 (1) Nfw % = W Nfs % = 100 − Nfw %

(2)

where P is atom% 15 N excess in above-ground or below-ground tissues of M. aquaticum and W is atom% 15 N excess in 15 N-labelled wastewater. Given that plant uptake, soil storage and microbial removal were the three main N removal processes in CWs,23 the 15 N mass balance analysis in M. aquaticum microcosm was expressed as in Eqn (3), which is similar to that described in a published paper.24 The amount of N removal by microbial nitrification and denitrification was calculated using Eqn (4): Qi = Qf + Qpu + Qsa + Qv + Qen

(3)

Qen = Qi − Qf − Qpu − Qsa − Qv

(4)

where Qi is the initial 15 N load in wastewater, Qi = 150 mg; Qf is the final 15 N load in wastewater, which was calculated by multiplying C f (the final N concentration in wastewater at day 21) by V (the volume of wastewater in M. aquaticum microcosm, V = 15 L); Qpu is the TN uptake by the plant; Qsa is TN accumulation in the soil; Qv is ammonia volatilization, which was neglected in this study because pH values were in the range of 5.5–7.9 in swine wastewater; and Qen is the estimated TN removal in nitrification and denitrification processes. All statistical analyses were performed using SPSS11.5 (SPSS Inc., Chicago, IL, USA). The normal distribution of variables was checked through a one-sample Kolmogorov–Smirnov test. Differences of all parameters were compared for significance using Duncan tests. All significance tests were conducted at the 0.05 level (P < 0.05).

RESULTS

J Sci Food Agric 2017; 97: 505–511

wastewater during this period. Concentrations of NO3 − -N in the two categories were in the range of 0.4–75.3 mg L−1 , with an initial increase up to maximum values on day 10 or 14, and a subsequent decrease until day 21. Variations in TN concentrations were similar to those of NH4 + -N. Approximately 88% of TN was removed on day 21. Most TN removal occurred during the first 7 days, accounting for 57.6–74.7% of TN removal. Plant N uptake and soil absorption Figure 3 illustrates RGR, fresh biomass and TN content of M. aquaticum. For SW1 and SW2, RGR in all above-ground tissues of M. aquaticum was significantly higher than that in below-ground tissues (P < 0.05). After a 21-day growth period, the total fresh biomass of M. aquaticum in SW1 and SW2 had developed to 2483.5 and 2350.0 g m−2 , respectively. An above-ground fresh biomass of 1293.2 g m−2 was obtained in microcosms with SW1, which was slightly higher than 1246.4 g m−2 in SW2. However, the below-ground fresh biomass in SW1 (1190.3 g m−2 ) was significantly higher than that in SW2 (1103.6 g m−2 ) (P < 0.05). The highest TN content of 50.9 g kg−1 and lowest content of 18.6 g kg−1 were respectively observed in the above-ground and below-ground tissues of the SW1 microcosms. In the case of SW1 and SW2, TN content in all above-ground tissues of M. aquaticum was significantly higher than in below-ground tissues (P < 0.05).



507

Changes in pH, DO and N concentrations For two strengths swine wastewater (SW1 and SW2), the temporal trends of pH and dissolved oxygen (DO) were similar over the whole incubation period (Fig. 1). The highest pH values of 7.8 for SW1 and 7.9 for SW2 were observed on days 1 and 7, respectively, and the lowest values of 5.5 (SW1) and 6.6 (SW2) on day 18. DO concentrations first decreased with time, reaching minimum values on day 4 of 1.4 and 0.8 mg L−1 for SW1 and SW2, respectively. DO concentrations then increased with time, with maximum DO concentrations of 4.5 mg L−1 (SW1) and 3.8 mg L−1 (SW2) observed on day 14. However, the average DO concentration of 3.3 mg L−1 for SW1 was higher than that of 2.4 mg L−1 for SW2. Figure 2 shows dynamic changes in N concentrations for SW1 and SW2. During the experimental period, NH4 + -N concentrations decreased with increasing incubation time, from 217.1 to 2.7 mg L−1 in the case of SW1, and from 439.4 to 3.5 mg L−1 in the case of SW2. Rapid NH4 + -N removal occurred during the first 10 days, with 87.9–93.5% NH4 + -N removal for both categories of

Figure 1. Dynamic changes in pH and DO concentration in wastewater in microcosm wetlands vegetated with M. aquaticum: (a) pH; (b) DO. The error bars represent the standard deviations (n = 3). The error bars in Figs 2–5 share the same meaning.

www.soci.org

S Zhang et al.

Figure 3. Relative growth rate (RGR) biomass and TN content in the tissues of M. aquaticum: (a) RGR; (b) biomass; (c) TN content. Different lower-case letters represent significant differences at the 0.05 level (P < 0.05). The lower-case letters in Fig. 4 have the same meaning.

Table 1.

15 N enrichment in the tissues of M. aquaticum 15 N-labelled

wastewater Figure 2. Dynamic changes in nitrogen concentration in wastewater in microcosm wetlands vegetated with M. aquaticum: (a) NH4 + -N; (b) NO3 − -N; (c) TN.

The results of 15 N tracer analysis showed that uptake of N from swine wastewater by M. aquaticum was mainly distributed in above-ground parts, with these accounting for 67.7% of total plant N uptake (Table 1). NH4 + -N and NO3 − -N content in microcosm soil covered a range of 2.3–40.0 and 1.6–59.4 mg kg−1 in the case of SW1 and SW2, respectively (Fig. 4). Soil NH4 + -N adsorption decreased significantly with time (P < 0.05). The highest adsorption values of NH4 + -N occurred on day 1. The maximum NO3 − -N content in soil was observed on day 7, with this being significantly higher than on days 1, 14 and 21 (P < 0.05). It was clear that soil NO3 − -N content decreased in a stepwise manner with time over 7–21 days.

atom% 15 N excess Nfw % Nfs %

4.0943

Above ground tissues 2.7707 67.7 32.3

Below-ground tissues 1.6286 39.8 60.2

Nfw % is the percentage of N in tissues of M. aquaticum assimilated from wastewater; Nfs % is the percentage of N in tissues of M. aquaticum from other sources, including soil N and plant N before placing in the wetland system.

fluxes, but a maximum N2 O flux of 155.48 mg N m−2 d−1 was observed on day 14. 15 N enrichment in the N2 O pool increased quickly with time during the initial phase, reaching the highest value of 3.94 on day 10 (Fig. 5b). Enrichment of 15 N2 O decreased gradually after day 10. 15

508

N2 O emission N2 O fluxes in M. aquaticum microcosms with SW1 were in the range of 0.2–82.9 mg N m−2 d−1 , with maximum N2 O fluxes occurring on day 10 (Fig. 5a). Compared to the microcosms with SW1, microcosms with SW2 showed a similar trend in N2 O emission


N mass balances analysis At the end of the incubation period, the quantities of 15 N in wastewater and soil were 12.6 and 23.4 mg, which accounted for 8.4% and 15.6% of the initial 150 mg 15 N in M. aquaticum microcosms with SW1, respectively (Table 2). 15 N uptake by M. aquaticum was 36.0 mg 15 N, making up 24.0% of the total 15 N amount. Mass


J Sci Food Agric 2017; 97: 505–511

Nitrogen removal in Myriophyllum aquaticum wetland microcosms

Figure 4. Dynamic changes in soil NH4 + -N and NO3 − -N content in microcosm wetlands vegetated with M. aquaticum: (a) NH4 + -N; (b) NO3 − -N.

balance analysis showed that 78.0 mg 15 N disappeared from the M. aquaticum microcosm, which indicated that 15 N loss through gas was equal to 52.0% of the added 15 N.

DISCUSSION

J Sci Food Agric 2017; 97: 505–511

Figure 5. Dynamic changes in N2 O fluxes and 15 N2 O enrichment in microcosm wetlands vegetated with M. aquaticum: (a) N2 O fluxes; (b) 15 N O enrichment. 2

respectively.33 In the present study, M. aquaticum microcosms with high-strength swine wastewater (SW2) had high removal rates of 99.2% for NH4 + -N and 87.8% for TN. Compared to SW2, more rapid removal of NH4 + -N and TN occurred in low-strength swine wastewater (SW1). A similar study over 2 months of operation reported that 78.3% of TN could be removed by duckweed from swine lagoon liquid with an initial TN concentration of 196 mg L−1 .2 Our observed N removal data indicated that M. aquaticum microcosms had higher efficiency in terms of N removal from swine wastewater than duckweed systems. In wetland systems, NH4 + -N removal occurs primarily through plant uptake, soil adsorption, volatilization of ammonia (NH3 ), and microbial nitrification and denitrification processes.8,9,34 The measured wastewater pH values were in the range of 5.5–7.9 throughout the experimental period, and NH3 volatilization contributed little to NH4 + -N removal in M. aquaticum microcosms because NH3 volatilization is not significant when the pH value is below 8.0.10 Based on 15 N mass balance analysis, N uptake by M. aquaticum accounted for 24.0% of the added 15 N. Myriophyllum aquaticum grows rapidly and can reach remarkable heights and biomass within a short period. In the present study, after 21 days of growth, the fresh biomass of M. aquaticum in SW1 and SW2 treatment increased from 1190.0 and 1103.5 g m−2 , to 2483.5



509

High NH4 + -N concentrations often inhibit the growth of aquatic plants in CWs,25 and the selection of wetland plant is important for high NH4 + -loading swine wastewater treatment. Sagittaria latifolia, Juncus effusus, Typha latifolia and Schoenoplectus validus were reported to be intolerant of NH4 + -N concentrations exceeding 200 mg L−1 ,26 – 29 and the tolerance levels of NH4 + -N for Cyperus alternifolius ranged from 147 to 236 mg L−1 (Table 3),28,30 Myriophyllum aquaticum may have a higher NH4 + -N tolerance ability than the above-mentioned plants, which was reflected by fast growth of M. aquaticum in the microcosms with SW2 (equalling about 400 mg L−1 NH4 + -N concentration). In addition, a demonstration project in the field demonstrated that M. aquaticum can survive in swine wastewater containing up to 200 mg L−1 NH4 + -N.31 Plant species have an important impact on N removal performance in wetland systems.32 CWs for treatment of NH4 + -rich dairy wastewater have been planted with Schoenoplectus validus, Phragmites australis, Glyceria maxima, Baumea articulata, Bolboschoenus fluviatilis, Cyperus involucratus and Zizania latifolia; mean NH4 + -N and TN removal rates recorded during an experiment over a period of 124 days were in the range of 65–92% and 59–90%,

www.soci.org

www.soci.org

Table 2.

S Zhang et al.

15 N mass balance in M. aquaticum microcosms 15 N amount (initial)

15 N mass (mg)

Wastewater (final)

150.0 ± 0.0

Soil (final)

12.6 ± 1.1 8.4 ± 0.6

Percentage of initial 15 N load (%)

23.4 ± 1.6 15.6 ± 0.9

Plant (final) 36.0 ± 3.8 24.0 ± 2.3

Gas loss 78.0 ± 4.4 52.0 ± 3.1

Values given represent mean ± SD (n = 3).

Table 3. Characteristics of aquatic plants for treating NH4 + -rich livestock wastewater Plant species Myriophyllum aquaticum Schoenoplectus validus Typha latifolia Sagittaria latifolia Juncus effusus Cyperus alternifolius a

Tolerance level of NH4 + -N (mg L−1 ) >200