This study demonstrated the successful use of a laboratory-scale baffled horizontal constructed wetland substituted with mixed organic media for zinc removal from high acidity (∼610 mg L−1 as CaCO3), sulfate-rich (∼1,300 mg L−1) wastewater. The wetland was planted with Typha latifolia. The mean zinc concentration in the influent was gradually increased from 0.56 ± 0.02 mg L−1 to 5.3 ± 0.42 mg L−1. The mean zinc concentration in the outflow was 0.22 ± 0.19 mg L−1, which accounted for 95% zinc removal throughout the study. However, total zinc uptake by the plants was 533 mg kg−1, accounting for only 1.2% of total zinc removal; therefore, major zinc retention occurred within wetland media (83%). The overall activity and specific sulfidogenic activity decreased at the end of the study to 1.43 mg chemical oxygen demand removed per mg of TVS per day and 0.60 mg sulfate reduced per mg of TVS per day, respectively. Additionally, 16S rRNA sequencing revealed major dominant phyla present: Firmicutes (36%), Proteobacteria (16%), Actinobacteria (8.8%), Planctomycetes (7.8%), Chloroflexi (3.5%), Acidobacteria (1.9%) and Fibrobacteres (1.5%).

  • Effective acidity removal with alkalinity generation (35–750 mg L−1 as CaCO3).

  • CW exhibited good zinc removal efficiency (92–95%), with >80% retention in media.

  • Zinc uptake by plants was 533 mg kg−1, accounting for only 1.2% of total zinc removal.

  • Specific sulfidogenic activity accounted for 0.60–0.83 mg sulfate reduced per mg of TVS per day.

  • Metagenomics revealed the dominance of Firmicutes (36%) and Proteobacteria (16%).

Graphical Abstract

Graphical Abstract
Graphical Abstract

The application of constructed wetlands (CWs) for water quality improvement has been practiced for the last five decades (Cirelli et al. 2007). CW technology is often regarded as a passive treatment option with minimum use of resources, energy and manpower, but with a larger area footprint than conventional chemical or biological treatment plants (Brix 1994). CWs are employed as the tertiary treatment units (as a polishing step) to conventional treatment plants in many cases. In recent years, there have been many developments and ongoing research focusing on CWs as the primary treatment units at pilot and full-scale levels (Younger & Henderson 2014). India, being a tropical country, offers an attractive opportunity for the implementation of CWs due to its prevailing warm climatic conditions, which is considered more favourable for the performance efficiency of CWs.

Metal contamination of water bodies poses a major environmental concern due to its toxic and non-biodegradable nature. Metal pollution from geogenic sources is minimal, whereas various anthropogenic sources such as electroplating, tanning, mining, beneficiation process, etc., have significantly elevated the metal concentration levels in water, soil and sediments at an alarming rate (Sahoo et al. 2017). Zinc is one such metal that is largely associated with acid mine drainage (AMD) from coal mining activities, steel processing and municipal solid wastes, resulting in toxic landfill leachate. The potential impacts of zinc pollution are bio-accumulation and bio-magnification in aquatic species and the acidity of waters (Gozzard et al. 2011).

Major metal removal pathways outlined in CWs are filtration, sedimentation, adsorption, precipitation (as oxides, hydroxides and sulfides), microbial oxidation/reduction and plant uptake (Wojciechowska & Waara 2011). Although the prime processes involved in dissimilatory sulfate reduction and subsequent metal removal in CWs are well studied, the role of organic media (as a carbon source), plant species, media longevity and identification of bacterial communities involved in the treatment of highly acidic zinc polluted wastewater in CWs have not been clearly characterized.

The importance of sulfate-reducing bacteria (SRB) in metal removal is well recognized in CWs dealing with sulfate-rich metallic wastewater, which, in fact, is governed by various environmental constraints such as pH, temperature, redox potential, nature of organic matter, organic carbon to sulfate ratio, sulfide concentration and the presence of competing microbes such as methanogens (Kaksonen & Puhakka 2007). Sulfidogenic oxidation of organic matter (electron donor) generates bicarbonate alkalinity and neutralizes the acidity of water, which is also necessary for the proper functioning of SRBs between pH 6 and 8. Several studies have highlighted the ability of SRBs to use various simple carbon sources (direct/indirect) such as animal wastes, agricultural waste matter (goat manure, spent mushroom compost, corn waste, rice waste, alfalfa, straw, peat and husk), alcohol (methanol and ethanol), lactate and formate (Choudhary & Sheoran 2012; Nielsen et al. 2019).

This study aims to evaluate the performance of horizontal CW for zinc removal from highly acidic and sulfate-rich wastewater. The role of plants and mixed organic media was investigated to understand the fate of zinc and its major removal routes. Biomass activity tests were conducted to determine maximum substrate utilization (or overall activity) and specific sulfidogenic activity. Metagenomics analysis was further performed to identify taxa and the abundance of microbial communities involved in biological processes occurring within the CW.

Horizontal sub-surface CW

A horizontal sub-surface CW with dimensions of 95 cm length × 32 cm width × 55 cm height was constructed with top 5-cm freeboard and four baffles consisting of eight perforations (Φ 12 mm placed alternatively in an up-down manner along the length of the CW creating five compartments (or zones) as shown in Figure 1. Eleven tubes for sampling were provided on the front-facing panel (numbered 1–11), extending to the mid-width of CW. CW was placed in an outdoor environment (sheltered from rain) at the rooftop of the academic complex (Civil Department), IIT Guwahati. The zones near the inlet and outlet served as the influent and effluent flow buffering zones filled with coarse gravel beds (50 cm thick, 12.5 mm < Φ < 20 mm). The middle zones (A, B and C) were filled with the main reactive substrate bed, which comprised the bottom-most pea-sized gravel (5 cm thick, 10 mm < Φ < 12.5 mm), overlaid by layers of bamboo chips, cow manure and soil with a having thickness of 10, 30 and 5 cm, respectively. Cow manure and bamboo chips were procured from local units in Amingaon, Guwahati. The physicochemical and elemental analysis of the media is presented in our previous study (Singh & Chakraborty 2020). Two fresh saplings of Typha latifolia (40–50 cm, planting density of 25 plants per m2) were transplanted in the middle zones (A, B and C).

Figure 1

(a) Design configuration and (b) photographic image of CW set up.

Figure 1

(a) Design configuration and (b) photographic image of CW set up.

Synthetic wastewater composition and operational details

The zinc concentration in water bodies impacted with runoff from zinc mines and industrial effluents has been reported to typically vary between 0.03–2.0 mg L−1 (O'Sullivan et al. 2004; Kumari & Tripathi 2015). Many previous studies have also involved zinc removal in CW for influent zinc concentrations in the range 0.03–24 mg L−1 (Stein et al. 2007; Di Luca et al. 2011; Gandy et al. 2016; Bavandpour et al. 2018). Thus, synthetic wastewater was prepared using tap water, which mainly contained zinc and sulfate with mean concentrations of 5.15 ± 0.20 mg L−1 and 1,247 ± 296 mg L−1, respectively. The tap water used for feed preparation contained traces of zinc (0.07 ± 0.01 mg L−1) and sulfate (28.9 ± 1.56 mg L−1). The mean pH and acidity were 2.08 ± 0.11 and 610 ± 21 mg L−1 as calcium carbonate (CaCO3), respectively, adjusted by the addition of sulfuric acid. The CW was continuously loaded with synthetic sulfate-rich zinc wastewater using a single-channel peristaltic pump (Miclins, India) at a hydraulic loading rate and nominal hydraulic retention time (nHRT) of 0.031 m per day and 7 days, respectively. The overall pack porosity of the completely saturated bed was 50%. Initially, CW was fed with lower concentrations of zinc and sulfate to ensure proper growth of plants and development of microbes, and thereafter, pollutant concentrations were gradually increased (phase I–V). Table 1 describes the complete operational details and feed characteristics of CW during various phases.

Table 1

Operational details and feed characteristics of CW

Operational phaseHRT (days)Zinc (mg L−1)Acidity (mg L−1 as CaCO3)Sulfate (mg L−1)
Phase I 0.560 ± 0.016 – 132 ± 42.6 
Phase II 1.26 ± 0.079 86 ± 9.30 205 ± 2.89 
Phase III 2.46 ± 0.242 276 ± 3.80 295 ± 7.47 
Phase IV 3.56 ± 0.189 363 ± 7.02 961 ± 141 
Phase V 5.15 ± 0.197 610 ± 20.6 1,247 ± 296 
Operational phaseHRT (days)Zinc (mg L−1)Acidity (mg L−1 as CaCO3)Sulfate (mg L−1)
Phase I 0.560 ± 0.016 – 132 ± 42.6 
Phase II 1.26 ± 0.079 86 ± 9.30 205 ± 2.89 
Phase III 2.46 ± 0.242 276 ± 3.80 295 ± 7.47 
Phase IV 3.56 ± 0.189 363 ± 7.02 961 ± 141 
Phase V 5.15 ± 0.197 610 ± 20.6 1,247 ± 296 

Analysis of water samples

Influent and effluent samples were regularly collected on alternate days for over 181 days and immediately analyzed for pH (pH meter, Systronics), chemical oxygen demand (COD), acidity, alkalinity, total dissolved sulfide and sulfate according to APHA (2012). For zinc measurements, samples were collected in separate polystyrene specimen tubes and immediately acidified to pH<2 (using nitric acid (HNO3)), filtered using 0.45-μm filter and stored in a refrigerator below 4 °C until analysis. Zinc concentration was determined using an atomic absorption spectrophotometer (AAS) equipped with high-energy air-acetylene flame (SpectrAA 55B, Varian).

Analysis of plants

Plants were carefully harvested from the CW following the conclusion of the study and subjected to chemical analysis to estimate the zinc concentration. First, plants were thoroughly washed several times with tap water and then with distilled water to ensure complete removal of surface impurities, soil particles and debris. Then the plants were oven-dried at 80 °C for about 48 h and segregated into root and shoot parts. The dried plant parts were ground to obtain a homogenous powdered mixture (<2 mm) and about 1 g of homogenized samples were acid digested at 80 °C (Allen et al. 1974). The acid digested solution was filtered and diluted to 100 mL for zinc analysis in AAS. Similarly, the zinc estimation was also performed before transplanting plants in the CW and termed ‘unpolluted’.

Analysis of wetland media

At the end of the study, the wetland bed was discarded entirely and media samples from middle zones (A, B and C) were collected at different depths (0, 15, 30 and 50 cm). The media samples were air-dried and homogenized (<2 mm) and subjected to acid digestion procedure with repeated addition of concentrated HNO3 and hydrogen peroxide (30%, H2O2) as per method 3050 B in USEPA (1996) for the estimation of total zinc concentration. In addition, media samples were partitioned into various forms by using suitable extracting reagents at every step according to Tessier et al. (1979), where the release of metals from different fractions of the examined sample was determined.

Biomass activity

Two kinds of assay studies (S1 and S2) were conducted to determine the specific maximum activity (overall activity) and specific sulfidogenic activity. A batch assay study was performed to assess the sulfidogenic activity of sulfur-reducing bacteria. The wet solid media (or biomass) was collected from the wetland and initial total volatile solids (TVS) was calculated as described in APHA (2012). The assays were conducted in 1,000 mL serum bottles with 1-L liquid volume under continuous stirring. About 8–10 g of media was added to each serum bottle to obtain a TVS concentration in the range of 1–1.24 g L−1. A blank assay (B) was also conducted in parallel without any substrate addition to determine COD exerted by the media. A synthetic substrate containing dextrose (as a simple carbon source) and sodium sulfate was used. Finally, distilled water (purged with nitrogen) was added up to the 1-L mark and the serum bottles were tightly capped. Nutrients (nitrogen, phosphorous and trace elements) were not supplemented to restrict the biomass growth during the test period.

In the S1 assay, only dextrose (1.4–1.9 g) was added to reach an initial COD concentration in the range of 1.5–2 g L−1. In the S2 assay, dextrose (COD 1.2 g L−1) and Na2SO4 (sulfate concentration of 1.7 g L−1) was added to achieve the COD/ ratio of 0.7. After the completion of each cycle, stirring was stopped and biomass was allowed to settle prior to withdrawing samples (about 5 mL) using a syringe, which was immediately filtered through Whatman 41 filter paper. COD and sulfate concentrations were measured initially at shorter intervals (0.25 h) and later at longer intervals (up to 6 h). After the first feeding, the supernatant of the serum bottle was decanted and refilled with the same feed, which constitutes the second feeding. Likewise, the procedure was repeated for the third feeding and, finally, the amount of TVS and dissolved sulfide concentration was determined at this stage. The biomass activity was estimated by the final TVS and slope of the linear portion of the total COD removed or sulfate reduced with time and calculated as (Singh & Chakraborty 2021):
formula
(1)

Microbial metagenomics

The microbial diversity and relative abundance of various microbial communities present in the CW were identified. A representative homogenized media sample from triplicate sampling was sent to Eurofins Genomics India Pvt. Ltd for 16S (V3–V4) ribosomal RNA (rRNA) sequencing using Illumina MiSeq platform. A commercially available soil kit (Nucleospin soil, Macherey-Nagel) was used to isolate and extract the metagenomic DNA from the media sample. The qualities of the isolated metagenomic DNA were quantified by loading 1 μL of sample into a Nanodrop spectrophotometer (NanoDrop One, ThermoScientific) for determining the A260/230 ratio. The first amplicon polymerase chain reaction (PCR) was set up using the isolated metagenomic DNA along with a bacterial 16S V3–V4 region-specific primer set. Forward and reverse primers used for the amplification of 16S rDNA targeting bacteria were GCCTACGGGNGGCWGCAG and ACTACHVGGGTATCTAATCC, respectively. The PCR product (3 μL) was resolved on 1.2% agarose gel at 120 V for approximately 60 min or until the sample reached three-quarters of the gel. The first amplicon generation was followed by Next-Generation Sequencing library preparation using Nextera XT Index Kit (Illumina Inc.). Quantitative Insights into Microbial Ecology pipeline (v1.8) was used to process raw reads. The sequencing data were deposited in the National Center for Biotechnology Information (https://www.ncbi.nlm.nih.gov/) in the bioproject PRJNA734092 as Sequence Read Archive with biosample accession numbers SAMN19471447 and SAMN19471448.

Zinc removal and water quality improvement

The variation recorded in influent acidity and effluent alkalinity with time during various operation phases is depicted in Figure 2(a). Mean effluent pH during phase I, II, III and IV were 7.01, 7.29, 7.13 and 7.09, respectively. However, during phase V, a substantial decrease in effluent pH was observed after 100 days of operation with mean effluent pH of 6.28. The immediate increase in pH (phase I–II) could be associated with the formation of carbonate ions from microbial oxidation of organic media and hydroxyl ion dissolution from organic media (Neculita et al. 2011). Therefore, the wetland demonstrated an overall increase in pH from inflow to outflow, and alkalinity declined from 750 to 35 mg L−1 (as CaCO3) in the effluent during phase V. The release of more soluble organics at the beginning of the CW increased the effluent COD, and therefore the average COD measured was 537 mg L−1 (phase I–IV), which decreased over time to about 111 mg L−1 (phase V). Reduction in the sulfate concentration was also observed, accounting for about 78% and 56% during phase (I–IV) and V, respectively. The production of the rotten egg-like odour was very prominent and evident due to sulfide generation during port sampling. However, total dissolved sulfide concentration measured in samples constituted <5 mg L−1, indicating the escape of gaseous hydrogen sulfide liberated during the sulfate reduction process by SRBs to the atmosphere from the CW.

Figure 2

(a) Acidity, alkalinity and (b) zinc profile of CW.

Figure 2

(a) Acidity, alkalinity and (b) zinc profile of CW.

The variation in influent and effluent zinc concentrations during different phases was measured and its removal efficiency (%) is shown in Figure 2(b). Zinc removal was very steady throughout, and the average removal efficiency of 97% and 92% was achieved during phase (I–IV) and V, respectively. EPA (2002) recommends the permissible discharge limit for zinc as 2 mg L−1 and zinc concentration in the effluent was far below the mentioned limit throughout the study period. The effective removal of zinc in the present CW was comparable to the results of previous studies (Lim et al. 2003; Vymazal & Krása 2003; Lesage 2006). Lesage (2006) studied the removal of heavy metals present in industrial wastewater using CW microcosms with a mixture of gravel and straw (15%) and reported high zinc removal efficiency (>78%) at an influent zinc concentration of 1–10 mg L−1. In a similar study by Vymazal & Krása (2003), zinc removal of 97% was achieved in gravel-bed CW that operated as a secondary treatment for sewage and stormwater with a much lower influent zinc concentration of 0.2 mg L−1, and also low concentrations of other metals. The removal of dissolved metals as metal sulfides depends on pH, the solubility product of the specific metal sulfide and the concentration of the reactants. However, zinc has been widely reported to undergo co-precipitation reactions to form insoluble compounds with oxy(hydroxides) and carbonates and, in some cases, precipitation as zinc sulfide is also possible under oxygen-deficient conditions (Stein et al. 2007; Di Luca et al. 2011; DiLoreto et al. 2016). In addition, the sorption onto the surface of organic media could be a major source of zinc retention during the initial phases of reactor operation (Allende et al. 2011; Nielsen et al. 2019). Further, Gandy et al. (2016) demonstrated the negative effect of additional carbon (in addition to the organic media) on zinc removal. However, there was a significant decrease in organic carbon over a long operation time, which possibly affected the SRB activity and resulted in lower alkalinity generation, although the zinc removal remained unaltered. Thus, precipitation as metal oxides (or hydroxides) and surface sorption on organic media appeared more likely to be a major zinc removal pathway. Table 2 presents a brief comparative summary of the zinc removal efficiency as reported in earlier literature.

Table 2

Brief comparative summary of the zinc removal efficiency reported in earlier literature

Type of treatmentCarbon sourceOperational detailsPollutant concentration (mg L−1)Removal efficiency (%)References
Passive sulfate-reducing bioreactors (PSRB) Straw Filled with waste rocks, mine impacted wastewater and inoculum; at 5 °C for 162 days Zinc (0.52), cadmium (0.008) and sulfate (∼550) Zinc (>95), cadmium (∼100) and sulfate (5.3) Nielsen et al. (2019)  
Hybrid wetland columns (laboratory scale) Composted urban green waste (plant biomass) Typha domingensis planted hybrid CW system (vertical column followed by horizontal), pH = 2–3, intermittent dosing at hydraulic loading rate = 0.252 m d−1 Copper (4), iron (200), manganese (18), lead (2) and zinc (12) Copper (90), iron (77), manganese (27), lead (98) and zinc (75) Bavandpour et al. (2018)  
Sub-surface flow CW (pilot scale) Compost, wood chips and activated sludge Unplanted, HRT = 7.5–14.5 h, pH = 7.74 and semi-continuous carbon additions (brewery waste) Zinc (2.32) and sulfate (134) Zinc (68) and sulfate (13) Gandy et al. (2016)  
Mussel shell bioreactor (MSB) Mussel shells HRT = 2.2–0.37 days and pH = 3.4 Iron (1.9), aluminium (15.7), zinc (0.26), nickel (0.07) and sulfate (172.6) Iron (∼99), aluminium (∼99), zinc (>90) and nickel (>90) DiLoreto et al. (2016)  
Passive bioreactor (bench scale) Goat manure HRT = 10 days, inoculum (whey and fresh cow manure) and pH = 2.70–3.35 Iron (189), copper (22), zinc (21), cobalt (1.2), nickel (10), manganese (32) and sulfate (4132–4960) Iron (99), copper (99), zinc (100), cobalt (96), nickel (98), manganese (40) and sulfate (54) Choudhary & Sheoran (2012)  
Free water surface CW (full scale) Sewage Typha domingensis planted, flow rate = 100 m3 per day and pH=9–10 Iron (0.78–34), chromium (0.05–0.4), nickel (0.03–0.04) and zinc (0.03–0.09) Iron (98), chromium (90), nickel (59) and zinc (57) Di Luca et al. (2011)  
Sub-surface CW microcosms Sucrose Typha latifolia planted CW at 24 °C on 20th incubation day and pH=6.5–6.7 Zinc (24) and sulfate (67) Zinc (≥ 98) and sulfate (∼80) Stein et al. (2007)  
Three in-series surface flow ponds (pilot scale) Spent mushroom compost Typha latifolia planted in one wetland cell and flow rate=1.5 mL min−1 Lead (0.15), zinc (2.0) and sulfate (900) Lead (32), zinc (74) and sulfate (31) O'Sullivan et al. (2004)  
Sub-surface CW (laboratory scale) Cow manure and bamboo chips Typha latifolia planted CW, HRT=7 days and pH=2.08 Zinc (5.15) and sulfate (1247) Zinc (95) and sulfate (56) Present study 
Type of treatmentCarbon sourceOperational detailsPollutant concentration (mg L−1)Removal efficiency (%)References
Passive sulfate-reducing bioreactors (PSRB) Straw Filled with waste rocks, mine impacted wastewater and inoculum; at 5 °C for 162 days Zinc (0.52), cadmium (0.008) and sulfate (∼550) Zinc (>95), cadmium (∼100) and sulfate (5.3) Nielsen et al. (2019)  
Hybrid wetland columns (laboratory scale) Composted urban green waste (plant biomass) Typha domingensis planted hybrid CW system (vertical column followed by horizontal), pH = 2–3, intermittent dosing at hydraulic loading rate = 0.252 m d−1 Copper (4), iron (200), manganese (18), lead (2) and zinc (12) Copper (90), iron (77), manganese (27), lead (98) and zinc (75) Bavandpour et al. (2018)  
Sub-surface flow CW (pilot scale) Compost, wood chips and activated sludge Unplanted, HRT = 7.5–14.5 h, pH = 7.74 and semi-continuous carbon additions (brewery waste) Zinc (2.32) and sulfate (134) Zinc (68) and sulfate (13) Gandy et al. (2016)  
Mussel shell bioreactor (MSB) Mussel shells HRT = 2.2–0.37 days and pH = 3.4 Iron (1.9), aluminium (15.7), zinc (0.26), nickel (0.07) and sulfate (172.6) Iron (∼99), aluminium (∼99), zinc (>90) and nickel (>90) DiLoreto et al. (2016)  
Passive bioreactor (bench scale) Goat manure HRT = 10 days, inoculum (whey and fresh cow manure) and pH = 2.70–3.35 Iron (189), copper (22), zinc (21), cobalt (1.2), nickel (10), manganese (32) and sulfate (4132–4960) Iron (99), copper (99), zinc (100), cobalt (96), nickel (98), manganese (40) and sulfate (54) Choudhary & Sheoran (2012)  
Free water surface CW (full scale) Sewage Typha domingensis planted, flow rate = 100 m3 per day and pH=9–10 Iron (0.78–34), chromium (0.05–0.4), nickel (0.03–0.04) and zinc (0.03–0.09) Iron (98), chromium (90), nickel (59) and zinc (57) Di Luca et al. (2011)  
Sub-surface CW microcosms Sucrose Typha latifolia planted CW at 24 °C on 20th incubation day and pH=6.5–6.7 Zinc (24) and sulfate (67) Zinc (≥ 98) and sulfate (∼80) Stein et al. (2007)  
Three in-series surface flow ponds (pilot scale) Spent mushroom compost Typha latifolia planted in one wetland cell and flow rate=1.5 mL min−1 Lead (0.15), zinc (2.0) and sulfate (900) Lead (32), zinc (74) and sulfate (31) O'Sullivan et al. (2004)  
Sub-surface CW (laboratory scale) Cow manure and bamboo chips Typha latifolia planted CW, HRT=7 days and pH=2.08 Zinc (5.15) and sulfate (1247) Zinc (95) and sulfate (56) Present study 

Zinc concentration in plants and media

The zinc uptake in plants and its parts is shown in Figure 3(a), expressed as mg zinc per kg dry weight (DW) of the plant. Unpolluted plants exhibited much lower zinc concentration in roots (9.3 mg kg−1) as well as shoots (7.1 mg kg−1). The zinc concentration accumulated in the roots of T. latifolia was 105–143 mg kg−1, whereas a much lower zinc uptake was observed in the shoots (41.9–101 mg kg−1). Higher zinc concentration in roots revealed good metal extraction capability of plants from the surrounding media. In contrast, lower zinc concentration in shoots explained the low translocation ability, which is considered as the defence mechanism of plants against metal, indicating good phyto-stabilization ability. Many studies have also reported similar values of zinc accumulation in the roots (50–350 mg kg−1) and shoots (15–200 mg kg−1) of T. latifolia (Ben Salem et al. 2017).

Figure 3

Zinc accumulation in (a) plant parts and (b) media with its partitioned fractions. DW represents dry weight of the plant's parts.

Figure 3

Zinc accumulation in (a) plant parts and (b) media with its partitioned fractions. DW represents dry weight of the plant's parts.

Figure 3(b) depicts the zinc concentration in media and its partitioned fractions (or forms) at various depths of each zone. Fractions F1, F2, F3, F4 and F5 represent exchangeable, carbonate-bound, iron oxide and manganese oxide-bound, organics and sulfide-bound, and residual (or inert) fraction, respectively (Tessier et al. 1979). It was observed that most zinc appeared in fractions bound to organic, sulfide, iron- and manganese- oxide, while some appeared in residual fractions. The highest zinc concentration was found in the media of zone A (near inlet), followed by B and C. Mean zinc concentration of about 124 mg kg−1, 70.5 mg kg−1 and 58 mg kg−1 in the F3 fraction, whereas about 99.6 mg kg−1, 44.3 mg kg−1 and 77.2 mg kg−1 in F4 fraction was found in zone A, B and C, respectively. Thus, these results indicated that zinc was mostly immobilized and, therefore, would not leach or back-release under ambient environmental conditions. For quality check, the sum of metal concentrations obtained for five fractions was compared with the total metal concentrations measured at various zone depths. For mean recovery (%), the ratio [(F1+F2+F3+F4+F5)Zn: total zinc] was considered and the obtained values were 97.3 ± 1.45%, 91.4 ± 8.97% and 93.9 ± 4.08% for zone A, B and C, respectively.

The fate of zinc was evaluated by assessing its removal pathways in wetland media and plants. The average zinc retained in the media was estimated from the product of the average zinc concentration (mg kg−1) and respective mass of media (kg) added in each of the zones. Zinc uptake by the plants was obtained from the product of total zinc uptake in plants (shoots and roots, mg kg−1) and its biomass yield (kg). Ignoring precipitation and transpiration, the metal mass (mg) in effluent and influent was calculated by taking into account average effluent (or influent) zinc concentration and total volume that exited (or entered) the CW (assuming constant flow rate) during each feeding cycle. It was found that a total of 7.29 g zinc entered CW during the experimental period and 0.379 g of zinc came out in effluent, while about 6.07 g and 0.085 g accumulated in media and plants, respectively. Therefore, a considerable amount of zinc (about 0.762 g) was untraced, which could be due to an error in sampling/analysis. Figure 4 shows the mass balance for zinc and it clearly indicates media as the major storage site for zinc (83%), whereas the studied plants showed very negligible contribution (1.2%) in zinc uptake. In a similar study by O'Sullivan et al. (2004) using T. latifolia, about 96% of the zinc removal was achieved in surface flow wetland, of which 94% was retained in the substrate while only 0.3% was measured in plant shoots.

Figure 4

Mass balance of zinc in CW.

Figure 4

Mass balance of zinc in CW.

Results of biomass activity

The microbial biomass activity test was conducted during acclimatization (phase I–IV) and towards the end of the study. The specific maximum activity was determined from the maximum slope of the S1 assay, which represents the maximum substrate removal (or overall activity). The specific sulfidogenic activity was determined from the S2 assay. The amount of sulfate reduced corresponding to COD metabolized by SRBs was calculated as each mole of sulfate reduced corresponds to 64 g COD. After acclimatization of CW, the overall activity and specific sulfidogenic activity obtained were 3.19 mg COD removed per mg of TVS per day and 0.83 mg sulfate reduced per mg of TVS per day, respectively. However, a significant reduction in the overall biomass activity was observed at the end of the study, which could be due to the deterioration in the wetland performance where effluent pH slowly started to decline, thus affecting the growth (or functioning) of microbes. The overall activity and specific sulfidogenic activity reduced to 1.43 mg COD removed per mg of TVS per day and 0.60 mg sulfate reduced per mg of TVS per day, respectively, as shown in Figure 5.

Figure 5

Biomass activity results (a) during acclimatization and (b) end of the study in CW.

Figure 5

Biomass activity results (a) during acclimatization and (b) end of the study in CW.

Identification of taxa and abundance of microbial communities

The 16S rRNA sequencing run yielded about 60,000 sequences per sample. The Archaea kingdom consisted of about 22%, whereas the remaining (78%) consisted of the bacterial kingdom, indicating that most of the bacterial taxa present in the media sample were targeted and identified. At the phylum level, archaea identified were Euryarchaeota (18.6%) and Crenarchaeota (2.9%). The metagenomics of the bacterial flora revealed the major dominant phyla included Firmicutes (36%), Proteobacteria (16%), Actinobacteria (8.8%), Planctomycetes (7.8%), Chloroflexi (3.5%), Acidobacteria (1.9%) and Fibrobacteres (1.5%). Other additional phyla identified (<1%) were Bacteroidetes (0.51%), Chlamydiae (0.36%), Verrucomicrobia (0.22%), Spirochaetes (0.19%), Nitrospirae (0.18%) and Chlorobi (0.13%). Many authors have also revealed the presence of similar bacterial groups in AMD-impacted streams and sediments (Fan et al. 2016; Ly et al. 2019). The most abundant classes detected were Clostridia (20%), Bacili (16%), Alphaproteobacteria (9.3%), Planctomycetia (7.0%), Actinobacteria (6.4%) and Deltaproteobacteria (5%), while some less detected classes (<5%) were Methanobacteria (3.4%), Miscellaneous Crenarchaeota Group (2.9%), Anaerolineae (2.2%), TG3 (1.5%) and Acidimicrobiia (1.4%). The relative abundance of acid-tolerant Gammaproteobacteria and Betaproteobacteria were similar. At the order level, most abundant orders were Clostridiales (19%), Bacillales (14%), Rhizobiales (7.1%) and Actinomycetales (6.4%). The most abundant family identified were Clostridiaceae (11%) and Planococcaceae (9.3%) followed by Hyphomicrobiaceae (4.2%) and Micromonosporaceae (4.0%), while many remained unclassified (17%). The relative abundance of different microbial communities at the phylum and genus level is represented in the form of pie chart diagrams shown in Figure 6. The abundance of Firmicutes and Proteobacteria have been reported in extreme environmentally stressed conditions, such as hypersaline sediments and heavy metal-contaminated soils (Emmerich et al. 2012). Therefore, the dominance of Firmicutes, Proteobacteria, Actinobacteria, Planctomycetes and Chloroflexi indicated their ability to adapt and grow in an acidic metal-rich environment.

Figure 6

Relative abundances (%) of dominant lineages at phylum and genus levels in media samples from CW.

Figure 6

Relative abundances (%) of dominant lineages at phylum and genus levels in media samples from CW.

Some less dominant genera of sulfur-metabolizing and sulfur-reducing bacteria (≤ 1%) identified were Desulfobacca, Desulfomonile, Desulfarculus, Desulfovibrio, Geobacter, Sulfuricurvum, Desulfosporosinus, Desulfococcus, Desulfomicrobium, Desulfotomaculum and Desulfitobacter. Thus, their presence confirms the occurrence of the SRB activity in the CW.

The study exhibited effective implementation of laboratory-scale sub-surface CW designed for the treatment of sulfate-rich, zinc polluted acidic wastewater. Alkalinity generation and metal removal were observed immediately after the start-up of the CW. This could possibly be due to the intrinsic property of the organic media that released sufficient hydroxyl and carbonate-bound surface ions, which aided pH neutralization and attributed zinc removal (95%) via precipitation and adsorption within the complex matrix of organic media. The mass balance analysis for zinc revealed media as the major sink of zinc retention (83%). In comparison, the zinc uptake by T. latifolia was minimal (1.2%) and thus, it does not significantly contribute to the zinc removal process. Biomass activity results indicated the existence of microbial activity involved in the sulfate reduction process. However, overall substrate utilization and specific sulfidogenic activity reduced to 1.43 mg COD removed per mg of TVS per day and 0.60 mg sulfate reduced per mg of TVS per day, respectively, at the end of the study. Metagenomics results showed the major dominant phyla involved in the remediation of acidic metallic wastewater were Firmicutes (36%), Proteobacteria (16%), Actinobacteria (8.8%), Planctomycetes (7.8%), Chloroflexi (3.5%), Acidobacteria (1.9%) and Fibrobacteres (1.5%), which included many acid-tolerant metal metabolizing groups of microbes.

The authors would like to acknowledge Mechanical Workshop, IIT Guwahati, for the fabrication of CW. We are incredibly thankful to Dr Subrat Kumar Mallick, Dr Sachin Kumar Tomar and Mr Chandra Bhanu Gupt for setting up and installing the CW. We also acknowledge Eurofins Genomics, Bangalore, India, for their service in metagenomic analysis.

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.

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

Allen
S. E.
,
Grimshaw
H. M.
,
Parkinson
J. A.
&
Quarmby
C.
1974
Chemical Analysis of Ecological Materials
.
Blackwell Scientific Publications
,
Oxford, UK
.
Allende
K. L.
,
Fletcher
T. D.
&
Sun
G.
2011
Enhancing the removal of arsenic, boron and heavy metals in subsurface flow constructed wetlands using different supporting media
.
Water Science and Technology
63
(
11
),
2612
2618
.
https://doi.org/10.2166/wst.2011.533
.
APHA
2012
Standard Methods for the Examination of Water and Wastewater
, 22nd edn.
American Public Health Association/American Water Works Association/Water Environment Federation
,
Washington DC, USA
.
Bavandpour
F.
,
Zou
Y. C.
,
He
Y. H.
,
Saeed
T.
,
Sun
Y.
&
Sun
G. Z.
2018
Removal of dissolved metals in wetland columns filled with shell grits and plant biomass
.
Chemical Engineering Journal
331
,
234
241
.
https://doi.org/10.1016/j.cej.2017.08.112
.
Ben Salem
Z.
,
Laffray
X.
,
Al-Ashoor
A.
,
Ayadi
H.
&
Aleya
L.
2017
Metals and metalloid bioconcentrations in the tissues of Typha latifolia grown in the four interconnected ponds of a domestic landfill site
.
Journal of Environmental Sciences
54
,
56
68
.
https://doi.org/10.1016/j.jes.2015.10.039
.
Brix
H.
1994
Use of constructed wetlands in water-pollution control – historical development, present status, and future perspectives
.
Water Science and Technology
30
(
8
),
209
223
.
https://doi.org/10.2166/wst.1994.0413
.
Choudhary
R. P.
&
Sheoran
A. S.
2012
Performance of single substrate in sulphate reducing bioreactor for the treatment of acid mine drainage
.
Minerals Engineering
39
,
29
35
.
https://doi.org/10.1016/j.mineng.2012.07.005
.
Cirelli
G. L.
,
Consoli
S.
,
Di Grande
V.
,
Milani
M.
&
Toscano
A.
2007
Subsurface constructed wetlands for wastewater treatment and reuse in agriculture: five years of experiences in Sicily, Italy
.
Water Science and Technology
56
(
3
),
183
191
.
https://doi.org/10.2166/wst.2007.498
.
Di Luca
G.
,
Maine
M.
,
Mufarrege
M.
,
Hadad
H.
,
Sánchez
G.
&
Bonetto
C.
2011
Metal retention and distribution in the sediment of a constructed wetland for industrial wastewater treatment
.
Ecological Engineering
37
(
9
),
1267
1275
.
https://doi.org/10.1016/j.ecoleng.2011.03.003
.
DiLoreto
Z.
,
Weber
P.
,
Olds
W.
,
Pope
J.
,
Trumm
D.
,
Chaganti
S.
,
Heath
D.
&
Weisener
C.
2016
Novel cost effective full scale mussel shell bioreactors for metal removal and acid neutralization
.
Journal of Environmental Management
183
,
601
612
.
https://doi.org/10.1016/j.jenvman.2016.09.023
.
Emmerich
M.
,
Bhansali
A.
,
Lösekann-Behrens
T.
,
Schröder
C.
,
Kappler
A.
&
Behrens
S.
2012
Abundance, distribution, and activity of Fe (II)-oxidizing and Fe (III)-reducing microorganisms in hypersaline sediments of Lake Kasin, southern Russia
.
Applied and Environmental Microbiology
78
(
12
),
4386
4399
.
https://doi.org/10.1128/AEM.07637-11
.
EPA
2002
Standards for Effluent Discharge Regulations
.
General Notice No. 44. of 2003. http://faolex.fao.org/docs/texts/mat52519.doc (accessed 17.04.19)
.
Fan
M.
,
Lin
Y.
,
Huo
H.
,
Liu
Y.
,
Zhao
L.
,
Wang
E.
,
Chen
W.
&
Wei
G.
2016
Microbial communities in riparian soils of a settling pond for mine drainage treatment
.
Water Research
96
,
198
207
.
https://doi.org/10.1016/j.watres.2016.03.061
.
Gandy
C. J.
,
Davis
J. E.
,
Orme
P. H.
,
Potter
H. A.
&
Jarvis
A. P.
2016
Metal removal mechanisms in a short hydraulic residence time subsurface flow compost wetland for mine drainage treatment
.
Ecological Engineering
97
,
179
185
.
https://doi.org/10.1016/j.ecoleng.2016.09.011
.
Gozzard
E.
,
Mayes
W.
,
Potter
H.
&
Jarvis
A.
2011
Seasonal and spatial variation of diffuse (non-point) source zinc pollution in a historically metal mined river catchment, UK
.
Environmental Pollution
159
(
10
),
3113
3122
.
https://doi.org/10.1016/j.envpol.2011.02.010
.
Kaksonen
A. H.
&
Puhakka
J. A.
2007
Sulfate reduction based bioprocesses for the treatment of acid mine drainage and the recovery of metals
.
Engineering in Life Sciences
7
(
6
),
541
564
.
https://doi.org/10.1002/elsc.200720216
.
Kumari
M.
&
Tripathi
B. D.
2015
Efficiency of Phragmites australis and Typha latifolia for heavy metal removal from wastewater
.
Ecotoxicology and Environmental Safety
112
,
80
86
.
https://doi.org/10.1016/j.ecoenv.2014.10.034
.
Lesage
E.
2006
Behaviour of Heavy Metals in Constructed Treatment Wetlands
.
PhD thesis
,
Ghent University
,
Ghent, Belgium
.
Lim
P. E.
,
Mak
K.
,
Mohamed
N.
&
Noor
A. M.
2003
Removal and speciation of heavy metals along the treatment path of wastewater in subsurface-flow constructed wetlands
.
Water Science and Technology
48
(
5
),
307
313
.
https://doi.org/10.2166/wst.2003.0337
.
Ly
T.
,
Wright
J. R.
,
Weit
N.
,
McLimans
C. J.
,
Ulrich
N.
,
Tokarev
V.
,
Valkanas
M. M.
,
Trun
N.
,
Rummel
S.
,
Grant
C. J.
&
Lamendella
R.
2019
Microbial communities associated with passive acidic abandoned coal mine remediation
.
Frontiers in Microbiology
10
,
1955
.
https://doi.org/10.3389/fmicb.2019.01955
.
Neculita
C. M.
,
Yim
G.-J.
,
Lee
G.
,
Ji
S.-W.
,
Jung
J. W.
,
Park
H.-S.
&
Song
H.
2011
Comparative effectiveness of mixed organic substrates to mushroom compost for treatment of mine drainage in passive bioreactors
.
Chemosphere
83
(
1
),
76
82
.
https://doi.org/10.1016/j.chemosphere.2010.11.082
.
Nielsen
G.
,
Coudert
L.
,
Janin
A.
,
Blais
J. F.
&
Mercier
G.
2019
Influence of organic carbon sources on metal removal from mine impacted water using sulfate-reducing bacteria bioreactors in cold climates
.
Mine Water and the Environment
38
(
1
),
104
118
.
https://doi.org/10.1007/s10230-018-00580-3
.
O'Sullivan
A. D.
,
Murray
D. A.
&
Otte
M. L.
2004
Removal of sulfate, zinc, and lead from alkaline mine wastewater using pilot-scale surface-flow wetlands at Tara Mines, Ireland
.
Mine Water and the Environment
23
(
2
),
58
65
.
https://doi.org/10.1007/s10230-004-0040-4
.
Sahoo
P. K.
,
Tripathy
S.
,
Panigrahi
M. K.
&
Equeenuddin
S. M.
2017
Anthropogenic contamination and risk assessment of heavy metals in stream sediments influenced by acid mine drainage from a northeast coalfield, India
.
Bulletin of Engineering Geology and the Environment
76
(
2
),
537
552
.
https://doi.org/10.1007/s10064-016-0975-2
.
Singh
S.
&
Chakraborty
S.
2020
Performance of organic substrate amended constructed wetland treating acid mine drainage (AMD) of North-Eastern India
.
Journal of Hazardous Materials
397
,
122719
.
https://doi.org/10.1016/j.jhazmat.2020.122719
.
Singh
S.
&
Chakraborty
S.
2021
Bioremediation of acid mine drainage in constructed wetlands: aspect of vegetation (Typha latifolia), loading rate and metal recovery
.
Minerals Engineering
171
,
107083
.
https://doi.org/10.1016/j.mineng.2021.107083
.
Stein
O. R.
,
Borden-Stewart
D. J.
,
Hook
P. B.
&
Jones
W. L.
2007
Seasonal influence on sulfate reduction and zinc sequestration in subsurface treatment wetlands
.
Water Research
41
(
15
),
3440
3448
.
https://doi.org/10.1016/j.watres.2007.04.023
.
Tessier
A.
,
Campbell
P. G.
&
Bisson
M.
1979
Sequential extraction procedure for the speciation of particulate trace metals
.
Analytical Chemistry
51
(
7
),
844
851
.
https://doi.org/10.1021/ac50043a017
.
USEPA
1996
Method 3050B: Acid Digestion of Sediments, Sludges, and Soils
.
Environmental Protection Agency
,
Washington, DC, USA
.
Vymazal
J.
&
Krása
P.
2003
Distribution of Mn, Al, Cu and Zn in a constructed wetland receiving municipal sewage
.
Water Science and Technology
48
(
5
),
299
305
.
https://doi.org/10.2166/wst.2003.0336
.
Wojciechowska
E.
&
Waara
S.
2011
Distribution and removal efficiency of heavy metals in two constructed wetlands treating landfill leachate
.
Water Science and Technology
64
(
8
),
1597
1606
.
https://doi.org/10.2166/wst.2011.680
.
Younger
P. L.
&
Henderson
R.
2014
Synergistic wetland treatment of sewage and mine water: pollutant removal performance of the first full-scale system
.
Water Research
55
,
74
82
.
https://doi.org/10.1016/j.watres.2014.02.024
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY-NC-ND 4.0), which permits copying and redistribution for non-commercial purposes with no derivatives, provided the original work is properly cited (http://creativecommons.org/licenses/by-nc-nd/4.0/).