Introduction

Wetlands are one of the most important natural methane (CH4) emission sources (Chowdhury and Dick 2013), while aerobic methanotrophs have been widely accepted as a key contributor to the mitigation of methane emission from wetland ecosystems (Chowdhury and Dick 2013; Yun et al. 2015). However, recently, the occurrence of nitrite-dependent anaerobic methane oxidation (n-damo) process in natural wetlands has received increasing attention, due to its potential role of important methane sink (Hu et al. 2014). A few previous studies have documented the distribution of n-damo microorganisms in natural wetlands (Chen et al. 2015; Han and Gu 2013; Hu et al. 2014; Shen et al. 2015a, b; Wang et al. 2016). These studies suggested that n-damo organisms in natural wetlands could be influenced by sampling site (Shen et al. 2015a, b; Wang et al. 2016) and wetland soil depth (Hu et al. 2014; Shen et al. 2015a, b). Moreover, plant type can influence both methanogenic archaeal community and methanogenesis pathways (Jiang et al. 2010; Tian et al. 2012), which suggests that plant type might have a great influence on the methane production in wetland anoxic environment. Therefore, it can be hypothesized that plant type might also have a link with wetland n-damo microorganisms. However, to date, information on the influence of plant type on n-damo microorganisms in natural wetland is still very limited (Chen et al. 2015), although plant type is known to be an important factor shaping wetland microbial communities (Rietl et al. 2016; Lee et al. 2015; Lee and Kang 2016; Tang et al. 2011).

Constructed wetland (CW) systems have been a favorite option for the treatment of municipal and industrial wastewater (Bakhshoodeh et al., 2016; de la Varga et al. 2015; Uggetti et al., 2016; Zhang et al. 2015; Zhi and Ji 2014), and polluted surface waters (Guan et al. 2015; He et al., 2016; Tournebize et al. 2015). There have been several reports on aerobic methanotrophs in CW systems (DeJournett et al. 2007; Fausser et al. 2013; Niu et al. 2015), while the presence of CW n-damo organisms in such environments has not been addressed. Therefore, the main aim of the present study was to investigate the n-damo organisms in CW systems and the influence of plant type.

Materials and methods

Wetland description and sampling

Five full-scale vertical-flow CW systems (length 23 m, width 14.5 m, height 1.2 m) were constructed in subtropical Huizhou City (southern China) to treat heavily polluted river water at the retention time of 16 h. The mean annual precipitation and air temperature in local area were about 2200 mm and 22 °C, respectively. The average influent chemical oxygen demand (COD) and ammonia nitrogen (NH4 +-N) were 100 mg/L and 10 mg/L, respectively. All these CWs were filled with fine gravels (diameter 8–12 mm) with a layer depth of 90 cm, and planted with Cyperus papyrus (wetland I), Typha orientalis Presl (wetland II), Vetiveria zizanioides (wetland III), Canna indica L. (wetland IV) and Juncus effusus L. (wetland V), respectively. Before this study, these wetland systems had been continuously run for approximately three years. In the present study, in the central zone of each CW system, gravel particle samples in triplicate were collected from the depth of 20 cm (upper layer), 50 cm (middle layer) and 80 cm (lower layer). Samples (IU, IM and IL), (IIU, IIM and IIL), (IIIU, IIIM and IIIL), (IVU, IVM and IVL) and (VU, VM and VL) denote the samples from the upper, middle and lower layers in wetlands I, II, III, IV and V, respectively. Physico-chemical properties of waters at these three different wetland layers are shown in Table S1.

Quantitative PCR assay

DNA of the attached biomass on gravel particles was extracted using the commercial Powersoil DNA extraction kit (Mobio Laboratories, USA). The number of n-damo bacteria was estimated using the NC10-specific primer pair qP1F (5′-GGGCTTGACATCCCACGAACCTG-3′)/qP1R (5′-CGCCTTCCTCCAGCTTGACGC-3′) (Ettwig et al. 2009; Wang et al. 2012) with an ABI 7500 FAST (Applied Biosystems, USA) in 20-µL reaction mixture. The reaction mixture was composed of 2× SYBR Green PCR master mix (10 µL), template DNA (2 µl), and 10 µM qP1R and qP1R primers (each 0.8 µl), using the same amplification conditions as previously reported in the literature (Wang et al. 2016). Standard curves with a range between 103 and 108 gene copies/mL were obtained with serial dilutions of a known number of plasmid DNA harboring the target gene. The amplification efficiency and coefficient (r 2) were 98% and 0.995, respectively. The significant difference (P < 0.05) in n-damo bacterial abundance among CW samples was checked using one-way analysis of variance (ANOVA) followed by Student–Newman–Keuls test.

pmoA Clone library analysis

To construct clone libraries, a nested approach was applied to amplify n-damo pmoA gene, using A189_b/cmo682 and cmo182/cmo568 as first- and second-step primer sets, respectively (Liu et al. 2015; Luesken et al. 2011; Wang et al. 2012). The amplification program was as follows: 94 °C for 4 min; 40 cycles of 1-min denaturation at 94 °C, 1-min annealing at 56 °C, and 1.5-min elongation at 72 °C; and a final 10-min elongation step at 72 °C. The obtained n-damo pmoA sequences were deposited in the GenBank database under accessions KX579125–KX579464 and were grouped into operational taxonomic units (OTUs) with 95% similarity (Chen et al. 2015). Shannon index was further calculated with the MOTHUR program (Schloss et al. 2009). Neighbor-joining phylogenetic tree of the obtained wetland n-damo pmoA gene sequences and their references from GenBank was built with the MEGA 6.0 software (Tamura et al. 2013). Moreover, weighted Unifrac distance between wetland samples was calculated using R library GUniFrac, and Environment Clustering analysis was conducted based on weighted Unifrac distance using R (version i386, 3.3.0).

Results

Wetland n-damo bacteria abundance

The number of n-damo bacteria is usually assessed by determining their 16S rRNA gene count (Chen et al. 2015; Liu et al. 2015; Wang et al. 2012, 2016). In this study, the abundance of n-damo microorganisms in the middle layer sample from wetland II was below qPCR detection limit. The copy number of n-damo bacterial 16S rRNA gene in other CW samples ranged between 3.81 × 107 and 2.44 × 109 copies per gram dry gravel (Fig. 1). No significant difference in n-damo community abundance was observed in wetland I (P > 0.05), while significant difference occurred in other wetlands (II, III, IV and V) (P < 0.05). In wetland II, sample IIU had significantly higher n-damo community abundance than sample IIL (P < 0.05). In either wetland III or V, the upper layer samples displayed much less n-damo community abundance than the middle and lower layer samples (P < 0.05); however, an opposite trend was found in wetland IV. These results illustrated the different vertical change trends for n-damo community abundance in CW systems. In addition, at a given wetland layer depth, the five CW systems with different plants showed the significant difference in n-damo community abundance (P < 0.05).

Fig. 1
figure 1

Number of n-damo bacterial 16S rRNA gene in CW samples. Error bars represent standard deviation of mean (n = 3). Different letters above the columns illustrate significant differences among samples (P < 0.05)

Full size image

Wetland n-damo community diversity

In this study, sample IIM was not successfully amplified with a nested approach, corroborated by qPCR result. A total of 340 chimera-free n-damo pmoA gene sequences were retrieved from other CW samples (Table 1). Each n-damo pmoA library comprised 16–29 sequences and 1–7 OTUs. The Good’s coverage estimator (≥85%) illustrated that n-damo pmoA OTUs of each CW sample were well captured. Moreover, the wetland pmoA Shannon diversity index varied from 0 to 1.34. The evident vertical change of n-damo community diversity could be observed in each CW system. However, these CW systems displayed the different vertical change trends. For an example, with the increasing wetland depth, pmoA Shannon diversity increased in wetland III but decreased in wetland V. In both wetlands I and IV, the middle layer sample had the lowest n-damo community diversity, while the lower layer sample had the highest one. In addition, at a given wetland layer depth, the evident difference of n-damo community diversity could be found in all five different CW systems.

Table 1 Diversity of n-damo pmoA gene in CW samples
Full size table

Wetland n-damo community structure

Figure 2 displays the phylogenetic relationship of the major OTUs (including at least two pmoA sequences from all CW samples) with their relatives reported in Genbank database. The pmoA sequences could be assigned into two clusters (clusters a and b). The pmoA sequences in cluster a were grouped with those from marine estuary and freshwater lake sediments and paddy soils. The sequence members in cluster b could be affiliated with the pmoA sequences from a variety of ecosystems, such as forest soil, sludge, and mangrove, river and lake sediments. All of the pmoA sequences from CW samples IIIU, IIIM, IVM and VU were distributed in cluster a, while those from samples IU and IM were only detected in cluster b. The pmoA sequences from samples IL, IIL, IVU and IVL mainly existed in cluster a, while those from samples IIU, IIIL, VM and VL dominated in cluster b. These results illustrated the vertical change of CW n-damo community structure. In this study, the comparison of CW n-damo community structure was further performed with Environment Clustering Analysis (Fig. 3). The two or three samples from each CW system were not closely grouped together, confirming the remarkable vertical change of n-damo community structure in each CW system. Moreover, at a given wetland layer depth, the samples from various CW systems were also clearly separated.

Fig. 2
figure 2

Phylogenetic tree of the major CW n-damo pmoA OTUs and their reference sequences from GenBank. The digit in parentheses represents the sequence number in the same pmoA OTU in a given CW sample. Numbers at the nodes show the levels of bootstrap support based on neighbor-joining analysis of 1000 resampled datasets. The values less than 50 are not displayed. The bar represents 2% sequence divergence

Full size image
Fig. 3
figure 3

Environment Clustering Analysis of CW n-damo communities

Full size image

Discussion

Several previous reports illustrated the vertical shift in soil n-damo community abundance in natural wetlands (Hu et al. 2014; Shen et al. 2015a, b), yet information on n-damo community abundance in CW system is still lacking. In this study, the evident vertical variation of n-damo community abundance was observed in CWs planted with Typha orientalis Presl, Vetiveria zizanioides, Canna indica L. and Juncus effusus L., but not in that with Cyperus papyrus. Moreover, to date, only Chen et al. (2015) suggested the influence of plant type on n-damo community abundance in Mai Po wetland coastal sediments. In the present study, the vertical change trend for n-damo community abundance differed remarkably in CWs with various plants. In addition, at a given wetland layer depth, n-damo community abundance also differed in these CW systems. Vegetation type could have a considerable impact on CW n-damo community abundance. Therefore, CW n-damo community abundance could be influenced by both layer depth and vegetation type.

So far, only Shen et al. (2015a) showed the vertical shift in soil n-damo community diversity in a natural wetland, while the present study provided the evidence that n-damo community diversity changed with layer depth in CWs. Moreover, only Chen et al. (2015) suggested the influence of plant type on sediment n-damo community diversity in an intertidal wetland. In this study, the vertical change trend for n-damo community diversity differed in CWs with various plants. At each layer depth, the evident difference of n-damo community diversity could be found in these CW systems. These results suggested that CW n-damo community diversity could depend on both wetland layer depth and vegetation type.

Several previous studies suggested the vertical change of soil n-damo community structure in natural wetland ecosystems (Hu et al. 2014; Shen et al. 2015a, b). Moreover, Chen et al. (2015) suggested that vegetation type could drive the change of sediment n-damo community structure in natural wetland. In this study, the results of both phylogenetic analysis and environment clusters analysis illustrated the vertical shift in CW n-damo community structure. In addition, the result of environment clusters analysis also indicated that, at each layer depth, n-damo community structure differed in CW systems with various plants, which suggested the profound influence of vegetation type on CW n-damo community structure. Therefore, both layer depth and vegetation type could influence CW n-damo community structure.

Conclusions

The vertical changes of n-damo community abundance, diversity and structure could occur in CW system. CW systems with different plants could have distinct n-damo community structure. CW n-damo community could be influenced by both wetland layer depth and vegetation type.