Aquaculture products are key protein source in the global food production, which are increasing constantly in recent decades (FAO, 2020). They are mainly from marine water culture, inland freshwater culture and capture fisheries. In China, inland freshwater culture is the most important farming model, with a total area of 5,116.32 thousands ha and production of 30.14 million tons in 2019 (Yu et al., 2020). In the intensive inland freshwater farming, high-quality freshwater inflow was required. In turn, discharge of effluents can contaminate adjacent water bodies, which causes many environmental issues.

Like livestock farming and husbandry, large amounts of waste in aquaculture are generated, such as uneaten feed, feces, and dissolved material, which may be released into surface water and cause water pollution (Sindilariu et al., 2009). The pollution problem has aroused widespread concern as it exerts an adverse effect to aquatic ecological integrity (Turcios and Papenbrock, 2014). Excessive input nitrogen (N) and phosphorus (P) in aquatic ecosystems can promote the growth of algae and other phytoplankton, and subsequently cause dissolved oxygen (DO) depletion, water quality deterioration, and even death of aquatic animals (Bhateria and Jain, 2016; Wang et al., 2021). In China, the discharge standard of aquaculture effluents has been issued to prevent the harmful environmental effect (SCT9101-2007) (Ministry of Agriculture and Rural Affairs, 2007). However, existing methods for the treatment of aquaculture effluents, such as bio-filter and physical filter, require high investment and increase the cost of farming, which is beyond the reach of most of small producers. Therefore, it is highly necessary to seek effective method with a relatively low cost to address sustainability issues of inland fish farming.

Constructed wetland (CW) is one of the most promising methods for dealing with aquaculture effluents as it possesses some advantages, such as low cost, ecological benefit and ease of operation and maintenance (Tepe, 2018; Hang Pham et al., 2021). When integrated with special matrix and aquatic plants, CW can effectively remove many types of contaminants from wastewater, including organic matter, N, P, suspended solids, heavy metals and pathogenic microorganism through the coordinated action of biological, physical and chemical processes (Almuktar et al., 2018; Wang et al., 2021). Meanwhile, substrate microorganisms (e.g. nitrifying bacteria and denitrifying bacteria) play a pivotal role in conversion of nutrients, especially N, in wastewater (Liang et al., 2003; Li et al., 2018).

In aquaculture, CW has been applied to treat the discharged effluents from farms (Schulz et al., 2003; Sindilariu et al., 2007), or as a water filter unit to connect with recirculating aquaculture system (Zhang et al., 2010; Shi et al., 2011). According to surface area, the CW is classified into lab-scale CW (< 10 m²) (Saeed and Sun, 2013; Jesus et al., 2017), pilot-scale CW (10-100m²) (Papaevangelou et al., 2016; Uggetti et al., 2016) and commercial-scale CW (> 100 m²) (Tilley et al., 2002; Shi et al., 2011) in research or farming practice regarding aquaculture. Early study reported that a hybrid free water surface-horizontal subsurface flow CW (surface area of 5 m²) could remove 95% of total inorganic N and 32-71% of P in effluents from milkfish (Chanos chanos) pond (Lin et al., 2002). Horizontal subsurface flow CW (28 m²) effectively reduced the total suspended solids (TSS), chemical oxygen demand (COD), total nitrogen (TN) and total phosphorus (TP) concentrations in effluents form the flow-through trout fish farm (Chazarenc et al., 2007). Vertical flow CW (80 m²) for treatment of aquaculture ponds water proved to be very effective and the reduction amounted to 61.5% of ammonia nitrogen ( NH 4 + -N), 68% of nitrate nitrogen ( NO 3 − -N) and 20% of P (Li et al., 2007). Shi et al. (2011) built a combination CW (221.0 m²) to purify brackish effluent from commercial recirculating and super-intensive shrimp farming systems, and it exhibited good removal of TN (67%), TAN (71%) and TSS (66%). It is apparent that the most of these studies were conducted at lab-scale and pilot-scale systems. However, there was relatively little published research regarding practical application of a commercial-scale CW in aquaculture of China, and thus, its treatment performance needs to be further evaluated.

Based on the issues addressed above, we built a commercial-scale integrated CW system (2,400 m²) with vertical subsurface flow, which connected with fish ponds with a total surface area of 19,200 m² to form a recirculating pond system. The objective of this study was to evaluate the performance of the CW system to manage the pond water quality with no water exchange during the farming period. Further, we examine whether the system improved the quality of effluents to comply with the discharge standard at the end of farming.

In aquaculture effluents, inorganic N is a key pollutant, primarily generated from residual feeds and feces (Piedrahita, 2003). Overloading N can cause eutrophication, algal blooms and water anoxia, which may further lead to abnormal changes in physiology and behavior of aquatic animals (Banerjee et al., 2021). Thus TN level is a common parameter used to assess water quality in aquaculture. A CW (9.5 m²) was used to treat effluents from a recirculating aquaculture system farmed Oreochromis niloticus, and there was a 95.5% removal rate for TN (Behrends et al., 1999). Similarly, Lin et al. (2002) built a hybrid CW (5 m²) to treat effluents of Chanos chanos farming, and TN concentration was significantly removed by the CW system. Zhang et al. (2010) used a farm-scale CW (320 m²) to improve water quality in Ictalurus punctatus farming, and the TN removal rate was 48%. In addition, TN removal rate was related to types of CW systems, such as, 12.4% in floating-bed CW, 64.7% in horizontal subsurface flow CW, and 23.0% in surface flow CW (Bai et al., 2020). In the present study, the TN removal rate was 36.85% during the fish farming period and 50.24% during the effluents discharge period, indicating the integrated CW system has a high removal rate for inorganic N. In addition, the average TN value was 4.92 mg/L at CW outlet during the effluents discharge period, which met the water quality of Class II in the freshwater sewage discharge standard of China (SCT9101-2007).

In aquaculture practice, NH 4 + -N, NO 2 − -N and NO 3 − -N are ubiquitous pollutants, and have received the most attention due to high toxicity on aquatic animals (Molayemraftar et al., 2022). High levels of NH 4 + -N and NO 2 − -N have been reported to inhibit growth performance, induce tissue damage or even death in aquatic animals (Ip and Chew, 2010; Kroupova et al., 2018). Chronic NO 3 − -N exposure reduces growth, increases susceptibility to hypoxia, disrupts endocrine system and affects the health status in fish (Monsees et al., 2017; Kellock et al., 2018; Isaza et al., 2021). It has been reported that CW was an efficient and cost-effective management practice to maintain reasonable concentrations of NH 4 + -N, NO 2 − -N and NO 3 − -N (Lin et al., 2010). Mahmood et al. (2016) structured a CW-commercial ponds of Macrobrachium rosenbergii system, and found that the 43.8% NO 3 − -N, 25.7% NH 4 + -N and 14.3% NO 2 − -N were removed. In a lab-scale closed recirculation aquaculture system, horizontal subsurface flow CW removed 99% NO 2 − -N and 82–99% NO 3 − -N, but had lower removal rate for NH 4 + -N (10%) (Hang Pham et al., 2021). In this work, the The CW system exhibited a high and stable removal rate for NH 4 + -N (62.93%), NO 3 − -N (59.66%) and NO 2 − -N (60.54%), and improved substantially water quality form the fish pond during the fish farming period. It is worth noting that the average NH 4 + -N concentration at CW outlet was below 0.5 mg/L during the effluents discharge period, meeting environmental quality standard of Class II for surface water in China (GB3838-2002). The data also indicated that the integrated CW system had a high removal efficiency for inorganic N in aquaculture effluents, which reduced aquaculture-included environmental contamination.

In CW system, N removal is a complex process, which may be related to sedimentation, uptake of aquatic plant, adsorption of media and biodegradation (Liu et al., 2019; Lu et al., 2020). N as an essential nutrient can be absorbed by aquatic plants, thus some aquatic plants are planted in CW to remove N. CW planted Thalia dealbata and Pontederia cordata displayed 66.44% and 68.56% TN reduction rate and 54.09% and 77.39% NH 4 + -N reduction rate (Liu, 2011). In a CW planted Canna indica, the removal rate for NH 4 + -N was 45.3-84.5% (Zhu et al., 2004). In this study, the Thalia dealbata, Iris germanica and Canna indica were planted into the CW to clean the effluents from fish pond, which reflected that the inorganic N removal was related to uptake of the aquatic plants. However, some researchers suggested that CW removed N primarily through ammonification, nitrification and denitrification, but not aquatic plants uptake (Uusheimo et al., 2018; Liu et al., 2019; Lu et al., 2020). In addition, the substrates of CW, such as pebbles and pelelith, have been reported to have strong adsorption on N (Wang et al., 2018). Previous study suggested that, in commercial-scale CW system, TN removal rate was often around 40%, of which limiting factors were organic carbon concentration and low DO level (Xie et al., 2018; Lu et al., 2020). In this study, we speculated the major limiting factor for N removal was the DO level (2.07 ± 0.29 mg/L in CW outlet), because heterotrophic microorganisms played an important role in N removal process (Friedland, 2004).

Apart from N, P level is another important parameter evaluating water quality. P can promote algal growth and has a significant impact on microcystins production by Microcystis aeruginosa in aquatic environment (Dai et al., 2016). High concentration of P deteriorated not only water quality (Smith et al., 1999), but caused microcystins accumulation though enhancing Microcystis biomass (Wang et al., 2010). CW has the potential to treat P in aquaculture effluents. For example, CW as a recirculation filter in large-scale shrimp aquaculture was effective in removal of TP (65%) (Tilley et al., 2002); a laboratory-scale CW markedly removed the TP (average reduction rate 31%) from Litopenaeus vannamei farming water (Hang Pham et al., 2021). In line with previous studies, our data showed that the integrated CW system effectively removed TP with an average removal rate of 65.95% during the fish farming period. Interestingly, our data also displayed that the average TP concentration was 0.09 mg/L at CW outlet, which lower than the threshold value (0.5 mg/L) in the freshwater effluents discharge standard Class II of China (SCT9101-2007). P removal in CW is a manifold process including physical, chemical and biological forces (Sindilariu et al., 2007). It also depends on the ecological situations, type of CW and planted macrophytes (Kumar and Dutta, 2019). P as an essential mineral can be absorbed by macrophytes in CW system. Various macrophytes possesses different uptake capacity of P, such as 48.1% for Canna indica and 76% for Thalia dealbata (BU et al., 2010; Yang et al., 2021). In addition, the P removal was related to adherence capability of a range of filter media in CW (Vohla et al., 2011; Wu et al., 2015).

Like TN and TP, COD concentration is a key monitoring parameter during aquaculture effluents discharge period. It is also used to evaluate organic pollutant in aquatic environment. Numerous studies suggested that CW had a strongly positive impact on the removal of organic matter with 59.7–89.0% COD reduction rate (Tuszyńska and Obarska-Pempkowiak, 2008). In the current study, the COD concentration was also clearly reduced by CW system, with 28.37–42.79% reduction rate. Meanwhile, the average COD concentration was 6.84 mg/L, below the threshold value (15 mg/L) in freshwater effluents discharge standard of China (SCT9101-2007). In CW, the COD can be removed by sedimentation and filtration suspended solids (Hang Pham et al., 2021). However, our data showed that the COD removal rate of sedimentation was 5.88-7.33% which was much lower than that of CW, reflecting the COD may be removed mainly via filtration of CW media. On the other hand, biodegradation by aerobic and anaerobic microorganisms was also an important removal mechanism of organic matter (Kumar and Dutta, 2019). In the aerobic biodegradation process, oxygen transfer may be a limiting factor for COD removal. Thus, the low COD reduction in this study may be related to DO level (2.07 ± 0.29 mg/L at CW outlet).

In the study, we built a commercial-scale vertical subsurface flow CW connected with fish ponds, which run stably during the fish farming and effluents discharge periods. During the fish farming period, the integrated CW system had high and stable removal efficiency for TN of 24.93–43.72%, TP of 61.92–72.18%, NH 4 + -N of 56.29–68.63%, NO 3 − -N of 56.66–64.81% and NO 2 − -N of 56.42–64.19%. During the effluents discharge period, average value of TN, TP and COD, three key parameters for effluents detection, was 4.92 mg/L, 0.09 mg/L and 6.84 mg/L, respectively, which met the water quality of Class II in freshwater effluents discharge standard of China (SCT9101-2007). Finally, this study evidently demonstrated that application of CW was an environmental sustainable sewage treatment strategy in intensive fish culture system.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

BL: Conceptualization, writing—original draft preparation, software; RJ: methodology, investigation validation, formal analysis, resources, YH: data cu-ration, writing—review and editing; JZ: visualization, supervision, project administration, funding acquisition. All authors read and approved the final version of the manuscript.

This study was supported by Central Public-interest Scientific Institution Basal Re-search Fund, CAFS (2020TD60), earmarked fund for CARS (CARS-45-20), and National Key R&D Program of China (2019YFD0900305).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.