Performance of tidal and non-tidal mangrove constructed wetlands in treating maricultural wastewater

Mariculture intensification has led to an increase in residual bait and culture metabolites, and consequently, an increase in the content of pollutants such as nitrogen and phosphorus in the aquaculture area, inhibiting the sustainable development of the aquaculture industry (Islam & Tanaka 2004). Moreover, the discharged aquaculture wastewater increases concentrations of nitrogen and phosphorus, leading to eutrophication in water bodies, causing algal blooms, imbalances, and disorders in the aquaculture ecosystem (Han et al. 2016; Lu et al. 2018). Therefore, exploring pollution control technologies that can adapt efficiently to aquaculture wastewater is essential for maintaining sustainable development of the marine aquaculture industry and the ecological environment. In addition, mariculture wastewater purification technologies are mostly limited by salinity and the presence of low concentrations of nitrogen and phosphorus. Constructed wetlands (CWs), which have a wide range of applications, have been widely used in the treatment of municipal sewage, aquaculture sewage, and other wastewaters as an engineering system that fits into natural green water treatment technologies (Wu et al. 2008; Zhi & Ji 2012; Ge et al. 2019).

CWs can be implemented on the basis of the corresponding physical, chemical, and biological processes through a synergy of substrates, plants, and microorganisms to promote the recycle and reuse of nutrients in sewage, and have a good removal effect on organic matter, suspended particulate matter, nitrogen, phosphorus, and other pollutants (Xiong et al. 2011; Avellán & Gremillion 2019; Li et al. 2019a). Two emerging techniques pertaining to treatment wetlands include free water surface flow constructed wetlands (FWS CWs) and subsurface flow constructed wetlands subdivided into horizontal flow (HF) and vertical flow (VF) wetlands (Dotro et al. 2017). Tidal flow (TF) constructed wetlands, though in exploratory stages, represent a new intensified CW system in which wastewater acts as a passive pump to repel and draw oxygen from the atmosphere into the substrate atmosphere where the biofilm is rapidly oxygenated (Chang et al. 2014; Liu et al. 2014; Li et al. 2015).

To treat saline aquaculture wastewater, salt-tolerant plants need to be selected for CWs. Mangrove species are a good choice because they have been reported to have a high salt tolerance and a higher effect in purifying fresh water sources (McKee et al. 2007; Bastakoti et al. 2019). A comparative study by Leung et al. (2016) found that the nutrient and toxic pollutant removal rates of mangroves (Aegiceras corniculatum and Bruguiera gymnorrhiza) were significantly higher than those of non-mangrove freshwater wetland plants (Acorus calamus, Canna indica, and Phragmites australis) on river water pollutants in industrial wastewater. Another study based on a CW with three mangrove plants (S. caseolaris, A. corniculatum, and K. candel) reported removal efficiencies of approximately 70% of organic matter, 50% of total nitrogen (TN), 60% of ammonia nitrogen (NH₃-N), 60% of total phosphorus (TP), and 90% of coliforms from sewage (Yang et al. 2008). Microorganisms are believed to play a crucial role in removing or transforming organic and inorganic pollutants, and perform almost 90% of N removal in CW wastewater treatment systems (Li et al. 2018; Yang et al. 2018). With changes in the surrounding environment, the bacterial community structure and the number of bacteria with nitrogen removal functions also change (Cao et al. 2017). Based on the differences in the bacterial environment, tidal and non-tidal CW operations which affect the functional bacteria are critical for exploring the N purification performance.

Previous studies for the removal efficiency of pollutants in CWs focused on plant optimization, matrix material selection and other aspects for short-term improvement of wastewater treatment, however, there is a clear lack of comparative studies on efficiency and sustainable long-term CWs system operation, especially the tidal and non-tidal CWs. Therefore, in this study, one tidal and two non-tidal CWs (HF and FWS) planted mangrove A. corniculatum were set up, and comprehensively evaluated the nitrogen and phosphorous removal ability to treat simulated marine wastewater. Furthermore, the composition of microbial communities and nitrogen transformation pathway in sediment of the tidal and non-tidal CWs are also investigated.

Three mangrove CWs were separately constructed in three independent belts planted with 2-year-old A. corniculatum, transplanted from a greenhouse. All belts measured 5 m×0.6 m×0.5 m (length×width×height) and were located in the Institute of Urban Environment, Chinese Academy of Sciences in Xiamen, China (24°36′N, 118°3′E). The mean air temperature was 22.2 °C, ranging from 3 °C to 36 °C over the duration of the CW operation. The non-tidal mangrove CWs included HF mangrove wetland and farm constructed wetlands (FCWs). The HF mangrove wetland were divided into three parts: an influent zone (0.3 m×0.3 m×0.3 m, L×W×H) filled with clean cobble (20–30 mm in diameter), allowing subsurface flow; a treatment zone (4.1 m×0.3 m×0.3 m, L×W×H) consisting of two layers. The lower layer (150 mm) contained coal cinder, sawdust, and concrete blocks to improve the hydraulic conditions and the upper layer was composed of soil collected from mangrove wetlands to make the plant grow better. The effluent zone (0.6 m×0.3 m×0.3 m, L×W×H) was filled with clean concrete blocks (10–30 mm in diameter) to prevent blockages in the effluent. The inlet and outlet distances from the wetland bottom were 0.5 and 0.25 m, respectively. The influent and effluent zones in FWS mangrove wetland were the same as those in HF mangrove wetland; however, the treatment zone was filled with 300 mm soil collected from mangrove wetlands. The influent and treatment zones of the TF mangrove wetland were the same as in HF mangrove wetland, except for two outlets installed in the effluent zone; one outlet was installed at a distance of 0.5 m from the wetland bottom and the other at a distance of 0.25 m from the wetland bottom with a solenoid valve controlled by a timing switch to simulate tides (Figure 1).

The wastewater of the three CWs pumped from the distribution pool (1.5 m×1.5 m×1.6 m, L×W×H) and controlled by flowmeter with 60 L·h⁻¹ as batch-flow about intake 0.5 h and interval 1.5 h. The hydraulic loading rate of each wetland was 0.36 m³·d⁻¹. The influent wastewater was synthesized from tap water and the following components: potassium nitrate (21.75 mg·L⁻¹), ammonium chloride (7.75 mg·L⁻¹) and potassium dihydrogen phosphate (4.5 mg·L⁻¹). The synthetic input water was disposed of sea salt to make the wastewater salinity at about 20‰, and with designated TN concentration range of 5–10 mg·L⁻¹, the TP concentration ranged from 0.2 to 0.6 mg·L⁻¹, which was similar to the results of our measurement from real mariculture wastewater. Prior to the experiment, we monitored the start-up phase of the CWs over a half a year after cultivating mangrove. Afterwards, we carried out the experimental phase (365d) under the stable operation of the CWs.

Water samples were collected once a week from the three CWs over 365 days of operation. There were three sediment sample points along the length of constructed wetlands at 1 m, 3 m, and 5 m; and were collected after 193d (warmest month September) and 285d (coldest month January) of the operating CWs. The leaf samples were collected at the same time as the sediment samples and mature and complete leaves with similar size.

The collected water samples were measured using an Ultraviolet Spectrophotometer (UV6100, Mapada Instruments, China) according to Standard Methods (Water and wastewater monitoring and analysis method. 2002), alkaline potassium persulfate and potassium persulfate digestion of the water to measure TN and TP respectively. The filtered water was tested for inorganic forms of nitrogen: ammonium-nitrogen (NH₄⁺-N) (Nessler's reagent spectrometry method), nitrate-nitrogen (NO₃⁻-N) (UV-Spectrometer method), nitrite-nitrogen (NO₂⁻-N) (colorimetric method). The TN content in the leaves and sediment were determined by CNS Analysis (Vario Macro CHNS-O-CL, Germany). The TP content in the leaves and sediment was determined by ICP-OES (7000DV, Optima, PerkinElmer, USA). The characteristics of the sediment were measured by portable pH meter (Thermo Star A329). Extraction of DNA from 285d sediment sample used the FastDNA^TM Spin Kit (MP Biomedicals, USA). The bacterial community structure was assessed using the V3–V4 region of the 16S rRNA gene.

One-way analysis of variance (ANOVA) tests were performed using SPSS software (version 17.0) to identify significant differences between samples (at the 0.05 significance level). Sequencing of the microbes was based on the Ion S5^TM XL sequencing platform, and a small fragment library was constructed by single-end sequencing. R software (version 2.15.3) was used to analyse microbial diversity in the CWs. The Tax4Fun based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database was used for the metabolic functions and pathways of the nitrogen metabolism.

We operated the CWs from summer to winter, and the changes in the sediment characteristics of the tidal and non-tidal CWs are shown in Table 1. The main N removal processes (nitrification and denitrification) in CWs can be influenced by the characteristics of sediments: the nitrification process can cease, and the denitrification process supported by organic matter can be hampered at pH<6.0 and pH>8.0, and the highest rate is observed at a pH range of 7.0–7.5 (Saeed & Sun 2012). In our results, as the running time increased, the pH of the three CWs decreased; the pH of the non-tidal FWS mangrove wetland and tidal TF mangrove wetland was closer to the optimum values. The moisture content (MC), soil organic matter (SOM), total carbon (TC), TN and TP of the sediment increased in all CWs; however, TC in the FWS mangrove wetland and TP in the HF mangrove wetland decreased (Table 1). These results could indicate an accumulation of organic matter and nutrients from the wastewater in the CW sediments, especially in the tidal TF mangrove wetland. It should be noted that sediment plays an important role in the process of CW treatment serves as a filter medium to retain nutrients from the wastewater useful for the formation of biofilm (Parde et al. 2021), which is consistent with the results of this study. However, the removal of TP was mainly caused by sediment adsorption in the CWs, and in the long term, TP was retained in the CWs which could pose a risk of sorbent release for a continuous supply of excessive phosphorus (Mendes et al. 2018). The highest sediment TP increase was observed in the tidal TF mangrove wetland, indicating that the operation mode of the tidal TF mangrove wetland improved sediment adsorption and early release of phosphorus into the treated wastewater compared with the non-tidal CWs.

The mangroves in the CWs grew well during the CW operation period, as shown in Table 2, and the leaf area (LA) and dry weight (DW) in the non-tidal HF mangrove wetland and tidal TF mangrove wetland were significantly higher than those in the non-tidal FWS mangrove wetland (p<0.05). Wetland plants can directly absorb nutrients in CWs, particularly in less nutrient-loaded wetland systems, and a high nutrient removal efficiency can be achieved. A comparative study of three wetland systems with loadings ranging from 0.44 to 1.83 g N m⁻² day⁻¹ demonstrated that up to 30% of N was removed by the plants (Peterson & Teal 1996), and an integrated vertical flow CW to purify low eutrophic water showed that plant uptake of N and P was 46.8 and 51.0%, respectively (Jiang et al. 2004). It is believed that the N and P in the CW plants are accumulated and stored in the leaves for a long time. Therefore, we analysed TN and TP contents in the leaves from the three CWs during the summer and winter. The TN and TP content in the leaves increased, and the highest content of both TN and TP was observed in the leaves from the non-tidal HF mangrove wetland. These observations indicate a significant uptake and storage of N and P in mangrove leaves from non-tidal HF mangrove wetland.

As shown in Figure 2, the average concentration of NO₃⁻-N, NH₄⁺-N, TN, TP from all CWs was significantly lower (p<0.05) than the inflow concentration. For the NO₂⁻-N due to the low concentration was difficulty removed, the average outflow concentration from the non-tidal HF mangrove wetland and tidal TF mangrove wetland was significantly lower than the inflow (p<0.05) but there was no significant difference between average outflow concentrations from non-tidal FWS mangrove wetland. The lowest outflow concentrations (p<0.05) of NO₂⁻-N, NO₃⁻-N, TN, and TP in non-tidal HF mangrove wetland were 0.01 (0–0.02), 1.26 (0.05–3.72), 1.92 (0.17–4.34), 0.20 (0.03–0.46) mg/L, respectively; and the lowest outflow concentration (p<0.05) of NH₄-N in tidal TF mangrove wetland was 0.22 (0–1.69) mg/L. Non-tidal FWS mangrove wetland had the highest outflow concentrations, except for NO₃^−_N.

Low concentrations of nitrogen and phosphorus have been associated with discharged maricultural wastewater (Brown et al. 1999). These observations are consistent with our investigations, where we employed low concentrations of nitrogen and phosphorus in the inflow settings. Low concentrations of nitrogen and phosphorus with salinity have been reported to be difficult to remove by CWs (Parde et al. 2021); however, in the current study, the outflow concentrations of the three CWs, except for NO₂⁻-N in the non-tidal FWS mangrove wetland, were lower than the inflow concentrations, indicating that all the non-tidal and tidal CWs could treat maricultural wastewater. In our study, the main forms of nitrogen were NO₃⁻-N and NH₄⁺-N, the NO₃⁻-N was highly removed by the non-tidal HF mangrove wetland compared with that removed by tidal TF mangrove wetland. The findings in our research are not similar to other findings reported by different researchers (Hu et al. 2014; Chand et al. 2021). The possible reason could probably result from the saline water, the removal of NH₄⁺-N and NO₃⁻-N through the nitrification process, as the denitrifying bacteria are more sensitive to salt compared with nitrifying bacteria (Dinçer & Kargi 1999), thus the tidal CW operation mode accumulate more salt that affects the denitrification process to remove NO₃⁻-N.

Throughout the one-year operational period, the removal rate of NO₂⁻-N in the CWs showed a marked difference (Figure 3). The non-tidal HF mangrove wetland had a stable and high average removal rate, followed by the tidal TF mangrove wetland, whereas the lowest removal rate was that of the non-tidal FWS mangrove wetland, with mean values of 88.4, 65.3, and 11.4%, respectively, indicating that nitrites were barely removed during the investigation period by the non-tidal FWS mangrove wetland setup. Contrary to the non-tidal system, the removal rate of NO₂⁻-N in the tidal TF mangrove wetland was unstable over the studied period. It could be concluded that there was no clear trend for NO₂⁻-N removal efficiency.

The removal rate of NO₃⁻-N in the non-tidal HF mangrove wetland was high, with a mean value of 80.5%; the average value for the non-tidal FWS mangrove wetland was 64.6%; whereas the low value reached 61.3% for the tidal TF mangrove wetland (Figure 3). From Figure 3, it can be observed that the NO₃⁻-N removal rate in the non-tidal HF mangrove wetland was relatively stable and decreased from summer to winter. Additionally, in the non-tidal FWS mangrove wetland, the NO₃⁻-N removal efficiency was higher in summer than in winter, whereas a great fluctuation and downward trend for NO₃⁻-N removal at the later stage in winter was observed in the tidal TF mangrove wetland. For NH₄⁺-N removal, all the CWs had evident removal effects, as shown in Figure 3. The tidal TF mangrove wetland had the highest mean NH₄⁺-N removal rate (90.6%), followed by non-tidal HF mangrove wetland (81.4%) and non-tidal FWS mangrove wetland (68.7%). From these results, it is clear that no evident seasonal variation affects the removal rate of NH₄⁺-N for the tidal TF mangrove wetland when compared to the performance of non-tidal HF mangrove wetland for NH₄⁺-N which fluctuated widely in summer, and that of non-tidal FWS mangrove wetland which had a rising trend in winter but decreased in summer. The TN removal rate in non-tidal HF mangrove wetland was the highest, followed by tidal TF mangrove wetland and non-tidal FWS mangrove wetland (79.7, 60.1, and 56.6%, respectively) (Figure 3). HF mangrove wetland had a relatively stable TN removal rate; however, the TN removal rate fluctuated violently in TF mangrove wetland, sometimes higher than HF mangrove wetland and sometimes lower than the FWS mangrove wetland, especially after the temperature fell in the winter. The TN removal rate of FWS mangrove wetland showed a distinct rise–fall trend. In addition, the TN removal efficiency of all CWs was higher in summer than in winter, and the TN removal rate of TF mangrove wetland and FWS mangrove wetland had a seasonal variation, with higher values in summer and lower values in winter.

Mangrove wetland systems can serve as filtration and precipitation systems to contribute significantly in the removal of nutrients and organic matter from wastewater because of their high productivity and large demand for nutrients (Wu et al. 2008). During the monitoring period, the nitrogen removal rate in the non-tidal HF mangrove wetland was higher than that in the non-tidal FWS mangrove wetland, as was demonstrated for NO₂⁻-N. Recent studies have shown that the N removal of non-tidal FWS mangrove wetland was influenced by climate variability: it increased with increasing temperatures and decreased at lower temperatures (Steidl et al. 2019). It was observed that the non-tidal FWS mangrove wetland and tidal TF mangrove wetland showed a higher percentage of nitrogen removal in summer than in winter, and the removal of nitrogen by tidal TF mangrove wetland fluctuated when compared with the non-tidal FWS mangrove wetland; therefore, tidal TF mangrove wetland was more efficient in the preliminary experiment than in the later stages of the experiment. However, the non-tidal HF mangrove wetland nitrogen removal rate was relatively stable and showed no visible seasonal alterations. Thus, it was verified that the main removal mechanisms for nitrogen could be explained by the fact that wastewater entering the CWs could be intercepted in the substrate of the CWs, and nitrogen removal was carried out through nitrification/denitrification reactions by the bacteria present in the substrate (Selvamurugan et al. 2011). Temperature can influence microbial activity; for example, at low temperatures, inhibition reactions can occur, explaining the observed instability and variation in the N removal for non-tidal FWS mangrove wetland and tidal TF mangrove wetland during the winter. In contrast, the non-tidal HF mangrove wetland had the most noteworthy steadiness of N removal and elevated N content in both the sediment and plant leaves at desirable working temperatures during the summer. These results demonstrate that high temperature favours the nitrification-denitrification process to release ammonia, nitric oxide, nitrogen dioxide, and nitrous oxide into the working environment, enhancing their ultimate uptake by the planted mangroves for growth. In our experiment, it appears that the NO₂⁻-N and NO₃⁻-N removal efficiencies were low, while the NH₄⁺-N removal rate was higher for tidal TF mangrove wetland than for non-tidal FWS mangrove wetland; however, this observation contradicts a previous report where tidal TF mangrove wetland resulted in a high nitrite and nitrate removal because of the periodic aerobic and anaerobic stages necessary for nitrification and denitrification processes (Wu et al. 2015; Tan et al. 2019). The oxygen transport capacity by the mangrove (the plant on CWs) was ignored in this study. The tidal CW operation is the most suitable habitat for mangrove growth as A. corniculatum has developed a rhizosphere that is conductive to oxygen exchange (Li et al. 2019b). The oxygen discharged from roots to the rhizosphere provided an aerobic environment for microorganisms which enhance the nitrification. However, the denitrification process could be limited by less anaerobic conditions.

The TP removal efficiency for the mangrove CWs was low and considerably affected by rain. The mean values for phosphorus removal in the non-tidal system were 40.8% (HF mangrove wetland), 16.3% (FWS mangrove wetland), and 19.2% (TF mangrove wetland). As shown in Figure 4, the TP removal rate of the non-tidal HF mangrove wetland was high, with relatively lower fluctuations among the CWs, ranging from 17.6 to 75.5%. The tidal TF mangrove wetland and non-tidal FWS mangrove wetland TP removal rates ranged from 0 to 48.4% and 0 to 47.1%, respectively. Generally, there were no observable effects associated with specific seasons that influenced TP removal efficiency for all tidal and non-tidal CWs.

The phosphorus removal mechanisms of the CWs mainly depended on substrate adsorption; therefore, within the CWs, the phosphorus removal rate decreased over an extended period of time (Gao et al. 2019). Phosphorus is a pivotal nutrient for plants in CW ecosystems (Nandakumar et al. 2019), and plants can take up phosphorus from the sediment that absorbs it from the influent. It was found in the current study that non-tidal HF mangrove wetland was effective in removing phosphorus, and the probable explanations are as follows: (i) the composition of the sediment, comprising coal cinder, sawdust, and concrete blocks, influenced the absorption process; (ii) consequently, the flow of wastewater through the substrates was enhanced by adsorption; and (iii) the uptake of phosphorus by plant leaves increased, explaining the reduction of phosphorus in the sediment and high TP content in the leaves. In case of tidal TF mangrove wetland (based on non-tidal HF mangrove wetland), the unsatisfactory phosphorus removal was possibly due to the release and leaching of dissolved phosphorus from the dry/wet sediments due to the physical disruption of the soil structure and desorption of the surfaces substrate and/or the death and lysis of soil microbial biomass due to rapid and drastic increase in water potential during rapid rewetting, with consequent increase in turgor or plasmoptysis (Brödlin et al. 2019; Cui et al. 2019). Thus, when the wastewater flushes the sediment, it releases the phosphorus into the environment. In addition, the TP content in the leaves from TF mangrove wetland did not increase, thereby contributing to phosphorus in the sediment. An unstable and low TP removal and increased TP content in the sediment (Table 1) which was observed for the non-tidal FWS mangrove wetland, could be explained by the extensive growth of algae which decreased the TP content in the water, while an accumulation of dead algae could promote phosphorus release in the sediment during CW operation (Zhu et al. 2013).

The functions and behaviors of bacteria have an impact on the performance of CWs and therefore, the composition of bacterial community in the sediment was determined. The bacterial community showed a diversity of community structure by using Principal Component Analysis (PCA) (Figure 5). In the PCA dimension reduction analysis the functional composition of the samples is more similar, the distance between them in the dimension reduction diagram is closer. The result showed that in the non-tidal HF mangrove wetland and tidal TF mangrove wetland was similar, however, in the non-tidal FWS mangrove wetland the community structure was more different from the rest (Figure 5). The main reason of the observed differences was attached to the sediment composition materials which were the same in the non-tidal HF mangrove wetland and tidal TF mangrove wetland in contrast to their counterpart.

The compositions of bacterial community in the three CWs were further examined at the principal component analysis pylum and genus levels shown in a heatmap (Figure 6). The changes in the dominant bacterial community structure may represent microbial succession and competition within different operating modes. Proteobacteria, Chloroflexi were the main phyla in the tidal TF mangrove wetland. Proteobacteria is usually dominant in wastewater treatment reactors and contains a large number of functional strains involved in the degradation phosphorus, and nitrogen, etc. (Liao et al. 2013; Chang et al. 2019). Actinobacteria, Acidobacteria, Gemmatimonadetes were the dominant phylum in the non-tidal HF mangrove wetland. Firmicutes, Bacteroidetes and Cyanobacteria were the main dominant bacteria in non-tidal FWS mangrove wetland. Some Actinobacteria members could participate in phosphate uptake and nitrogen removal in the CWs (Jia et al. 2019). Acidobacteria, Firmicutes, Bacteroidetes, and Actinobacteria are significant for contaminant decomposition and are widespread phyla in CWs (Rampuria et al. 2020). Cyanobacteria was involved in nitrogen-fixing microorganism.

Figure 7 demonstrates the potential N-metabolism functional genes related to the nitrogen cycle in tidal and non-tidal CW sediments. It has been seen that the nitrification and denitrification is the vital foundation for the nitrogen removal mechanism in CWs, especially the denitrification is a key process of removing nitrogen (Li et al. 2018; Yang et al. 2019; Chen et al. 2021). In this study, in the transformation of NH₄⁺-N to NO₂⁻-N (nitrification), the pmoA-amoA, pmoB-amoB, and pmoC-amoC genes were higher in the non-tidal HF mangrove wetland and the hao gene was higher in the non-tidal FWS mangrove wetland. During the oxidation of NO₂⁻-N to NO₃⁻-N, narH, narY, nxrB, narG, narZ, and nxrA were higher in the non-tidal HF mangrove wetland. For denitrification, NO₃⁻-N reduction played a critical role in nitrogen transformation (Tan et al. 2021), and the relative abundance of narI, narV, narH, narY, nxrB, narG, narZ, and nxrA genes in reduction of NO₃⁻-N to NO₂⁻-N was higher in the non-tidal HF mangrove wetland, and napA and napB genes were higher in the tidal TF mangrove wetland. In the process of NO₂-N reduction to NO, the nirK gene was higher in the non-tidal HF mangrove wetland and the nirS gene was higher in the non-tidal FWS mangrove wetland; in the process of NO reduction to N₂O, the norB and norC genes were dominant in the non-tidal HF mangrove wetland; in the process of N₂O reduction to N₂, the NosZ gene was higher in the tidal TF mangrove wetland. According to the results, the non-tidal HF mangrove wetland more genes abundance related in the nitrogen removal, which could explain high efficiency removal of nitrogen by the microbial genes in the HF mangrove wetland.

In the present study we have demonstrated that the non-tidal and tidal mangrove CWs would have an effect in treating the salinity maricultural wastewaters. Our experimental results have shown that the non-tidal HF mangrove wetland with A. corniculatum mangrove plant has a high stability and high removal efficiency of both nitrogen (88.4% NO₂⁻-N, 80.5% NO₃⁻-N, 81.4% NH₄⁺-N and 79.7% TN) and phosphorus (40.8% TP) which could have significant implications in the treatment of saline wastewater. Moreover, we have revealed that the non-tidal HF mangrove wetland had a high nitrogen removal rate, which was related to the microbial community structure and quantitative associations of N-metabolism genes. Given the treatment efficiency and operating requirement, we conclude that the non-tidal HF mangrove wetland planted with mangrove A. corniculatum CW has application value in the treatment of maricultural wastewater.

This research was supported by the STS project from Science and Technology Department in Fujian [2021T3016, 2019T3023].

All relevant data are included in the paper or its Supplementary Information.