Removal and fate of arsenic in the rhizosphere of Juncus effusus treating artificial wastewater in laboratory-scale constructed wetlands

Removal and fate of arsenic in the rhizosphere of Juncus effusus treating artificial wastewater in laboratory-scale constructed wetlands

Ecological Engineering 69 (2014) 93–105 Contents lists available at ScienceDirect Ecological Engineering journal homepage: www.elsevier.com/locate/e...

2MB Sizes 10 Downloads 187 Views

Ecological Engineering 69 (2014) 93–105

Contents lists available at ScienceDirect

Ecological Engineering journal homepage: www.elsevier.com/locate/ecoleng

Removal and fate of arsenic in the rhizosphere of Juncus effusus treating artificial wastewater in laboratory-scale constructed wetlands Khaja Zillur Rahman a,∗ , Arndt Wiessner b , Peter Kuschk b , Manfred van Afferden a , Jürgen Mattusch c , Roland Arno Müller a a

Centre for Environmental Biotechnology, UFZ−Helmholtz Centre for Environmental Research, Permoserstrasse 15, 04318 Leipzig, Germany Department of Environmental Biotechnology, UFZ−Helmholtz Centre for Environmental Research, Permoserstrasse 15, 04318 Leipzig, Germany c Department of Analytical Chemistry, UFZ−Helmholtz Centre for Environmental Research, Permoserstrasse 15, 04318 Leipzig, Germany b

a r t i c l e

i n f o

Article history: Received 24 October 2013 Received in revised form 30 January 2014 Accepted 29 March 2014 Keywords: Arsenic removal Artificial wastewater Constructed wetland Juncus effusus Mass balance Rhizosphere

a b s t r a c t The deposition, fate and distribution of arsenic (As) under dynamic redox conditions within the rhizosphere of helophytes in treatment wetlands are still poorly understood. For this purpose, long-term experiments were carried out in specially designed laboratory-scale constructed wetland reactors treating artificial domestic wastewater containing As (200 ␮g As l−1 ) in order to investigate the key aspects of As immobilization, to identify the main As removal pathway by using a mass balance approach and to assess the role of different sulfate (SO4 2− ) concentrations on As mass retention. The results with a highly efficient As mass retention (>92%) indicated a better performance under C-deficient and oxidized conditions (Eh ∼324–795 mV) regardless to the SO4 2− concentration in the inflow wastewater. An elevated SO4 2− concentration (25 mg S l−1 in the inflow) facilitated high As-retention (>90%) under C-surplus and microbial dissimilatory SO4 2− reducing condition (Eh ∼−225–−149 mV) within the root-near environment of the rhizosphere in constructed wetlands. Mean pH in a range of 6.6–7.7 might be favoring the immobilization of As but a comparatively low pH (3.9–5.9) within the root vicinity might enhance plant uptake. In general, higher As concentrations were exhibited by the plant roots (90–315 mg As kg−1 dry wt) as compared to the shoots (3.5–3.8 mg As kg−1 dry wt). Nearly 3.5-fold higher As concentrations within the roots from the experimental reactor as compared to the roots collected from control reactor clearly indicated that a higher amount of As was retained, accumulated, adsorbed, metabolized to other forms on root surface and/or translocated into the roots of Juncus effusus, where organic C and SO4 2− were abundant. Based on As mass balance calculation, the reactor with the highest SO4 2− loading was found to be retained nearly 85% of the total As mass input. Out of which only <1% of the total inflow As mass was sequestered or translocated into the plant shoots, 42.2% was accumulated/recovered within the plant roots, 17.2% was entrapped or deposited within the sediments of the gravel bed, 16.2% was recovered in the pore water and 15.3% was flushed out as outflow. The remaining 9% was considered as unaccountable, which might be released due to volatilizations or lost due to various unknown reasons. A 5-fold higher SO4 2− concentration within the reactor might facilitate lower pH (3.9–5.9) and consequent remobilization caused a higher amount of free or exchangeable As in the pore water (16.2%), that probably resulted in a higher As uptake (42.2%) by the plant roots as compared to the roots from the control reactor (only 13%). The findings demonstrate the deposition and fate of As within the rhizosphere, which are of high importance for an efficient treatment of wastewater containing As under constructed wetland conditions. © 2014 Elsevier B.V. All rights reserved.

∗ Corresponding author. Tel.: +49 341 235 1019/179 940 9506; fax: +49 341 235 1830. E-mail address: [email protected] (K.Z. Rahman). http://dx.doi.org/10.1016/j.ecoleng.2014.03.050 0925-8574/© 2014 Elsevier B.V. All rights reserved.

94

K.Z. Rahman et al. / Ecological Engineering 69 (2014) 93–105

1. Introduction Arsenic (As) is a toxic metalloid which can pollute water, soil, crops and the environment at large, ultimately affecting human health (Zhao et al., 2010). More than 245 minerals contain As, and the principal source of As is geological. However, human activities such as mining, pesticide application, and burning of fossil fuels also cause As pollution (Sharma and Sohn, 2009). In recent years, there has been an increasing contamination of water, soil and crops by this metalloid in many regions of the world (Tripathi et al., 2007), particularly in some countries of southern Asia (Meharg, 2004). It is therefore very important to choose appropriate methods to control As in the environment. Several treatment technologies have been applied for the removal of As from contaminated waters, such as coagulation/filtration, ion exchange, lime softening, adsorption on iron oxides or activated alumina, reverse osmosis etc. (Zouboulis and Katsoyiannis, 2002). Constructed wetlands are low-energy based ‘green’ technologies that have been increasingly applied in wastewater treatment since the mid-1980s (Sun and Saeed, 2009) and have considerable potential to remove metals and metalloids, including As (Buddhawong et al., 2005; Rahman et al., 2008a; Ye et al., 2003). Wetland plants have been shown to play important roles in constructed wetlands to remove As from wastewater (Rahman et al., 2008a, 2011; Singhakant et al., 2009). Complex interactions of As under redox gradient (both micro- and macro) conditions have already been investigated in different laboratory-scale horizontal subsurface-flow constructed wetlands treating an artificial wastewater (Rahman et al., 2008a). The rhizosphere of constructed wetlands offers specific macroand micro gradients of redox conditions enabling the development of highly diverse microbial consortia capable of different beneficial redox reactions (Bezbaruah and Zhang, 2004; Liesack et al., 2000; Wiessner et al., 2005b). Particularly due to the release of oxygen and organic carbon at the same time by the roots of helophytes into the rhizosphere, spatial and temporal micro-scale gradients of oxygen concentrations and redox states are established close to the root surfaces. These conditions enable the development of microbial mats and layers of functionally different microorganisms which simultaneously realize multiple interactive processes like nitrification, denitrification, mineralization of organic carbon, methanogenesis, reduction and oxidation of several sulfur and As compounds on a small spatial scale (Bezbaruah and Zhang, 2004; Darrah et al., 2006; Rahman et al., 2008b; Wiessner et al., 2005b). Recently, the application of a specially designed laboratoryscale constructed wetland (Kappelmeyer et al., 2002) in order to evaluate micro-gradient processes within the near-root environment of the rhizosphere was shown to be useful (Wiessner et al., 2005a,b). The behavior of metals in aquatic systems is complex and may include interactions among or between the major wetland compartments, above-ground plant parts, roots, litter, biofilms, soil, and water (Kadlec and Knight, 1996). Volatilization of metals into a gaseous phase occurs with mercury, selenium, and arsenic to a lesser degree. Dissolved metals can adsorb onto particles, or exist complexes to inorganic and organic ligands, or be present in solution in the free-ion state. The adsorption and co-precipitation of As on hydrous oxides of Fe, Mn or Al oxides and Fe sulfides is an important sink for As immobilization (Jacks et al., 2002). Moreover, dissimilatory reduction caused by iron- and sulfate-reducing bacteria is widely considered the primary mechanism responsible for the rapid As reduction and release observed in anaerobic environments (Islam et al., 2004; Kirk et al., 2004). Based on the characteristic of metal hyperaccumulation in plants, suitable and sustainable remediation strategies could be

developed (Rai et al., 1995). Fitz and Wenzel (2002) proposed that hyperaccumulators may enhance metal solubility in the rhizosphere via root exudation, consequently increasing plant metal uptake. Larios et al. (2012) showed that the plants accumulated extremely high amounts of total As in their tissues which varied depending on the part of the plant, with roots accumulating the most As in all the studied plants (up to 1400 mg kg−1 dry wt). In the context of constructed wetlands, García et al. (2010) reported that the direct uptake and accumulation of As in plants appears to play a very minor role in As removal. The same conclusion was drawn by Singhakant et al. (2009), who reported that only 0.5–1% of the total As input was accumulated in plant tissues. However, there are also studies indicating that wetland plants have a remarkable effect on As retention (Rahman et al., 2011; Sasmaza and Obek, 2009). First results of laboratory-scale investigations showed the transformation processes and redox dynamics of As-species particularly in the near-root environment of the Juncus effusus in model constructed wetlands. Changes in dynamic redox conditions and re-oxidation of reduced sulfur into other S species (e.g. S0 , SO4 2– ) caused a total sulfur enrichment and a consequent As remobilization within the rhizosphere in this study (Rahman et al., 2008b). But it is necessary to investigate the role of organic C, S and pH on As retention within the root vicinity and enhanced plant uptake under different redox conditions. Differences between corresponding inflow and outflow data of total As indicated remarkable amounts of As immobilization within the rhizosphere. Therefore, in addition to the investigations of As-removal efficiency and dynamics of As-species (Rahman et al., 2008b), it is necessary to deepen the understanding of the deposition, fate and mass balance of As within the micro-scale root zone environment of the rhizosphere in treatment wetlands. However, Rahman et al. (2011) in another study showed the fate and distribution of As along the flow path of a laboratory-scale horizontal subsurface-flow constructed wetland. But the knowledge regarding the accumulation and mass balance of As under the micro-scale gradient of redox conditions and the role of C, S and pH within the rhizosphere of helophytes is still insufficient. In the past, little attention has been paid and virtually no information is available until now that directly addresses the removal, fate and plant uptake of As under dynamic redox conditions within the rhizosphere of constructed wetland. In the context of our research work in this field, several investigations like fundamental aspects and mechanisms of As fixation, influences of dynamic redox conditions on As biotransformation processes, stability of As within the planted and unplanted wetlands, bioaccumulation and uptake of As in plant biomass, sludge sediment analysis, mass balance of As and S, etc., were carried out in different laboratory-scale constructed wetlands (Rahman et al., 2008a,b; Rahman et al., 2011). In principle, the major objectives of this study were i) to investigate the fate, accumulation and distribution of As under redox dynamic conditions on a micro-scale gradient within the rhizosphere; ii) to use a mass balance approach to identify the main As removal pathway under “ideal flow” conditions; and iii) to assess the role of organic C, pH and different SO4 2− concentrations on As mass retention within the rhizosphere of helophytes in treatment wetlands. By using Juncus effusus, long-term experiments in a specially designed laboratory-scale constructed wetland treating an artificial wastewater containing As (200 ␮g l−1 ) were performed to evaluate all these processes. The results of this study may help to better understand the key aspects of As immobilization within the plant root-zone of the rhizosphere under constructed wetland conditions and to optimize management practices for maximum As retention in different wetland compartments (shoots, roots, sediments etc.) through a complete mass balance analysis.

K.Z. Rahman et al. / Ecological Engineering 69 (2014) 93–105

95

compared with the control reactor R1, which was constantly fed with limited or very low inflow SO4 2− concentration. 2.2. Experimental conditions

Fig. 1. Schematic diagram of a specially designed laboratory-scale constructed wetland (planted fixed bed reactor–PFBR): (1) feeding storage tank; (2) pump; (3) glass vessel; (4) distribution chamber; (5) suction cylinder; (6) gravel bed; (7) recirculation pump; (8) magnetic valve; (9) on-line measurement; (10) outflow; (11) plants. (Adapted from Kappelmeyer et al., 2002).

2. Materials and methods 2.1. Experimental design: Planted fixed-bed reactors The experiments were carried out in three laboratory-scale planted-fixed bed reactors (R1, R2 and R3), which were established under conditions of complete mixing of the pore water by a permanent circulation. Since the internal flow conditions were comparable to an ideal mixed vessel, and therefore, macro-scale gradients of concentrations were equalized and the effects of the micro-gradient changes could be determined. The design and the principles of operation of the reactor (PFR – planted fixed bed reactor) have been described previously in detail (Kappelmeyer et al., 2002; Wiessner et al., 2005a,b). Briefly, the rhizosphere of the reactors was represented by a rooted gravel bed (particle size 2–4 mm) in a glass vessel of 28 cm diameter and a height of 30 cm. Each reactor was planted with five rush plants (Juncus effusus) with an initial total shoot number of 72, 84 and 80 in the reactors R1, R2 and R3, respectively. The reactors were closed tightly with a Teflon lid containing five circular openings through which the plants were grown in the gravel bed (Fig. 1). The glass reactors were covered to prevent algal growth. The free pore water volume in the planted beds amounted to 10 l and the hydraulic retention time was adjusted to 5 days. The reactors were placed in a greenhouse and operated under defined environmental conditions to simulate an average summer day in a moderate climate (Kappelmeyer et al., 2002; Wiessner et al., 2005b, 2008). The temperature was set to 22 ◦ C from 6 a.m. to 9 p.m. to simulate daytime and to 16 ◦ C at night. One lamp (Master SONPIA 400 W, Phillips, Belgium) was switched on during daytime as an additional artificial light source whenever the natural light fell below 1110 ␮mol m−2 s−1 (Wiessner et al., 2013). The main reason to use three model reactors operating in parallel was to investigate the influences of simultaneously increasing inflow feeding SO4 2− concentrations on the As mass retention, both under C-deficient and C-surplus conditions in the respective reactors. Obtained results from the reactors R2 and R3 were then

An artificial wastewater (Wiessner et al., 2005a) simulated a typical secondary effluent of domestic wastewater was used in this investigation. The inflow concentrations of the used ingredients were (in mg l−1 ): 204.9 CH3 COONa, 107.1 C6 H5 COONa, 117.8 NH4 Cl, 28 K2 HPO4 , 7 NaCl, 3.4 MgCl2 ·6H2 O, 4 CaCl2 ·2H2 O, 0.89/22.2/44.4/110.9 Na2 SO4 , 0.2 As(V) (Titrisol® As2 O5 in H2 O, Merck, Germany) and 1 ml l−1 of trace mineral solution which was adapted from Buddhawong et al. (2005). The resulting concentrations of the parameters were (in mg l−1 ): 340 chemical oxygen demand (COD), 122.3 total organic carbon (TOC), 30.8 ammonia-N, 5 phosphate-P, 0.2/5/10/25 sulfate-S and 0.2 As(V). These compounds were dissolved in deionized water and fed to the corresponding reactors in different experimental phases. Artificial wastewater was freshly prepared in every 3 days to prevent microbial degradation during storage and operation (Fig. 1). N2 gas was vigorously purged through and bubbled out of the liquid phase of the wastewater for approximately 20−25 min after each preparation of fresh feeding solution in order to remove any traces of dissolved oxygen. Each reactor was fed separately from a feeding glass bottle with a volume of 10 l. In order to keep an anoxic environment inside the feeding glass bottle containing the artificial wastewater, a continuous purging of nitrogen gas (N2 ) through the headspace of the feeding bottles was maintained throughout the whole operation period (Fig. 1). The reactors were operated for several months (nearly 4–5 months) using similar artificial wastewater (Wiessner et al., 2005a) but without any SO4 2− or As to establish biofilms and microbiological activity. Within these 4–5 months of normal operation, new shoots were grown from those five plants in each reactor. Prior to start of our experiment by using simulated wastewater with As and SO4 2− , the total numbers of healthy green shoots were counted as 231, 245 and 241 in R1, R2 and R3, respectively. The duration of the main experimental period was 340 days for all three reactors in this study. At the end of the experimental period of 340 days, the total shoot numbers were decreased down to a minimum of 180, 211 and 198 in R1, R2 and R3, respectively. The reactors were fed and run under five different experimental phases (phase I, II, III, IV and V) realized by varying SO4 2− concentrations in the simulated wastewater inflow solution. The operation conditions in all the phases within the reactors R1, R2 and R3 are listed in Table 1. The hypothesis behind different experimental phases was to investigate how the redox dynamics of As was influenced under C-deficient and oxidized condition with simultaneous increasing of SO4 2− concentrations in each reactors (phase I), under C-surplus and reducing condition with both constant (in R1) and simultaneous increasing and/or decreasing of SO4 2− concentrations (phase II, III, IV and V) in the reactors R2 and R3. Each phase had a sufficient duration to guarantee a representative number of samples that could be taken from each reactor. In experimental phase I (duration of first 83 days), the reactors were fed by a continuous inflow of As-contaminated artificial wastewater (200 ␮g As l−1 ) associated with SO4 2− concentrations of 0.2, 5 and 25 mg S l−1 but without organic C in the reactor R1, R2 and R3, respectively. The resulted molecular ratio of SO4 2− -S to As(V) varied as 1:1, 25:1 and 125:1 in the inflow wastewater of the reactors R1, R2 and R3, respectively. In experimental phase II (from day 83 to day144), the reactors were fed and operated under conditions of surplus C by adding organic C-sources (resulting a COD of ∼340 mg l−1 ) along with As(V) and only traces of SO4 2− (0.2 mg S l−1 ) in the simulated

96

K.Z. Rahman et al. / Ecological Engineering 69 (2014) 93–105

Table 1 Operation conditions within the model reactors R1, R2 and R3 accomplished by defined arsenic, organic carbon and sulfate inflow concentrations of the artificial wastewater (phase I–V). Reactor

Parameter

Unit

Experimental phases I

n

II

R1

As (V) SO4 2− -S TOC As (V) SO4 2− -S TOC As (V) SO4 2− -S TOC

␮g l−1 mg l−1 mg l−1 ␮g l−1 mg l−1 mg l−1 ␮g l−1 mg l−1 mg l−1

202 ± 8 0.2 ± 0.1 bdl 204 ± 5 5±1 bdl 198 ± 10 25 ± 3 bdl

12 12 12 12 12 12 12 12 12

205 0.2 120 201 0.2 124 199 0.2 126

R2

R3

± ± ± ± ± ± ± ± ±

6 0.1 4 5 0.1 5 10 0.1 7

n

III

10 10 10 10 10 10 10 10 10

196 0.2 118 199 5 120 201 25 118

± ± ± ± ± ± ± ± ±

4 0.1 10 8 1 9 7 5 11

n

IV

13 13 13 13 13 13 13 13 13

202 0.2 120 201 5 125 199 10 123

± ± ± ± ± ± ± ± ±

5 0.1 9 6 1 8 7 2 6

n

V

n

15 15 15 15 15 15 15 15 15

203 ± 6 0.2 ± 0.1 128 ± 4 201 ± 4 bdl 125 ± 5 204 ± 7 bdl 126 ± 6

6 6 6 6 6 6 6 6 6

bdl: below the detection limit (<1 mg TOC l−1 ; <0.1 mg SO4 2− -S l−1 ).

wastewater inflow solution of all three reactors. Similar running conditions were prevailing at the end of this phase and the resulted molecular ratio of SO4 2− -S to As(V) was 1:1 in the reactors. In phase III (from day 144 to 235), the same C-dosage and SO4 2− concentration as like in phase II was maintained in the control reactor R1 and only the SO4 2− concentration was increased to 5 and 25 mg S l−1 in the inflow feeding solutions of the reactors R2 and R3, respectively. During experimental phase IV (corresponding to day 235 to 315), conditions in all three reactors were similar to the inflow feeding solution of phase III operation, but only exception was the concentration of SO4 2− in reactor R3. Instead of a SO4 2− concentration of 25 mg S l−1 (as in phase III), the amount was decreased to 10 mg S l−1 in this particular phase IV in the reactor R3 (Table 1). The newly resulting molecular ratio of SO4 2− -S to As(V) was established as 50:1 under this experimental phase in reactor R3. Maintaining the same operating conditions as like in Phase IV, the supply of the inflow SO4 2− was completely stopped in reactor R2 and R3 (phase V) until the termination of the experiment on day 340. A comparative analysis was carried out in between the reactors under each circumstance and also within different operation phases. Plant transpiration represents 98% of the total water loss (Wiessner et al., 2005a) and was measured by balancing the inflow and the outflow amounts of water once a week. The flow balances were also used to control and adjust the inflow rate. All green shoots from a length of at least approximately 2 cm were counted once a month.

2.3. Sample collection and analysis Samples for total As, total organic carbon (TOC) and sulfate (SO4 2− ) analysis were taken once a week. The preservation technique of the collected samples for measuring the total As content and the analytical procedure of hydride generation atomic adsorption spectrometry (HG-AAS) have already been described by Daus et al. (2002) and Schmidt et al. (2004). The volatile As species were analyzed by gas chromatography/mass spectrometry (GC–MS, Shimadzu-GC-17A-Shimadzu GC-MS-Qp5000) with electron ionization and quadrupole analyzer, using the method described by Pantsar-Kallio and Korpela (2000). The analyses were performed isothermally at 50 ◦ C, and helium was used as a carrier gas. Due to the reactor design, the circulation flow represents the actual concentration of the pore water inside of the reactor and thus the pH and redox potential (Eh ) were controlled continuously in the circulation flow and the data were recorded online twice per hour (Fig. 1).

SO4 2− was analyzed by ion chromatography (DIONEX 100, columns AS4A-SC/AG4A-SC and CS12A/CG12A; Idstein, Germany) and a conductivity detector. TOC was analyzed by using a TOC analyzer (Shimadzu, TOC 600, Duisburg, Germany). After the termination of the experiment, plant biomass samples (shoots and roots) and sludge sediments were collected from each reactor in order to investigate the potential for As removal efficiency. Plant samples were sectioned into their shoot and root components after collecting them from each reactor. The roots were first thoroughly washed with tap water and then with deionized water to remove any gravel aggregate or sludge sediment. The plant shoots and roots collected from each reactor were freshly weighed, dried at 105–108 ◦ C for 3 days, allowed to cool, and then the dry weights were determined and the water content was calculated. These dry weights are used throughout the text unless otherwise specified. The dried samples were ground to a fine powder using a mortar and pestle under liquid nitrogen in order to obtain a homogeneous sample and then preserved in sealed plastic bottles for analysis. For the analysis of the total As concentration in plant biomass (shoots and roots), the homogenized powdered samples were digested by microwave extraction (PE Anton Paar GmbH, Graz, Austria). For digestion, 2.0 ml of digestion mixture (HNO3 :HCl = 4:1) were added to 0.5 g powder in a Teflon pressure bomb and heated to 260 ◦ C for 1 h. After the digest cooled down, it was filled with deionized water to a total volume of 10 ml, mixed and filtered using a 0.45 ␮m syringe filter (Satorius AG, Goettingen, Germany). The filtrate solution was analyzed for total As by using hydride generation atomic absorption spectrometry (HG-AAS) with a detection limit of 0.3 ␮g As l−1 . Acid blanks were analyzed in order to assess possible contamination. All analyses were performed in duplicate. Analysis of the sludge sediment samples was performed by the energy dispersive X-ray fluorescence (EDXRF) spectrometer XLAB 2000 (SPECTRO Instruments) running with the software package XLAB Pro 2.2. Collected sludge sediments from each reactor were dried at 105 ◦ C for 24 h using oven MA4O (Satorius, Germany) and were ground by means of an agate ball mill (Retsch). Well-ground sample material (1 g) was mixed with stearine wax (Hoechst, Germany) for XRF analysis as a binder in a ratio of 80:20 (w/w) and subsequently pressed at 150 MPa to pellets (with an internal diameter of 32 mm) and analyzed by means of energy dispersive Xray fluorescence analysis (EDXRF). The mean value of two replicates was calculated. The relative error of the method was 2–3%. Experimental results were evaluated statistically (e.g. mean, standard deviation, bar graphs error bar) by using spreadsheet program Microsoft Excel (Microsoft Corporation) in this study.

K.Z. Rahman et al. / Ecological Engineering 69 (2014) 93–105

97

2.4. As mass balance calculation After 340 days of the experiment, a complete mass balance of As was investigated in each reactor by considering the total As mass input, the total As mass output, and the total As retained in the plant biomass (the shoots and the roots), in the pore water and in the sludge sediments. The remaining (loss or gain) of the As mass from the mass balance calculation was considered to be unaccountable. The total As mass input and output in each reactor was calculated from the cumulative total As mass inflow and outflow during the whole operation time period (Fig. 2). A simple As mass balance was computed according to the following Eq. (1) (Singhakant et al., 2009): Asin = Asout + Asplant + Assed. + Aspw + Asunaccount

(1)

Where Asin is total As mass in the inflow (g); Asout is total As mass in the outflow (g); Asplant is total As mass in the plant biomass (g), including plant roots and shoots; Assed. is total As mass retained in the sludge sediment (g); Aspw is total As mass retained in the pore water or standing water (g); Asunaccount. is the total As mass that was unaccountable (g), including the loss or gain of As from the mass balance calculation. 3. Results 3.1. As removal efficiency Arsenic removal efficiencies in terms of total amount of the As mass inflow and the total As retention in all three corresponding reactors during the whole study period of 340 days is shown in Fig. 2. A total inflow mass of 140.0, 135.1 and 123.3 mg As was fed and a total of 72.8, 27.7 and 18.9 mg As mass was flushed out through the outlet of reactors R1, R2 and R3, respectively. Therefore, a cumulative total mass of 67.2, 107.4 and 104.4 mg As was retained, which resulted in nearly 48%, 80% and 85% of the total As mass retention in the corresponding reactors R1, R2 and R3, respectively (Fig. 3). During the stoppage period of SO4 2− -S supply (in phase V), no remobilization or loss of the total As mass was observed in the reactors, R2 and R3.

Fig. 2. Cumulative total As mass inflow and retention in the three laboratory-scale reactors (R1, R2 and R3) during the whole experimental period of 340 days.

Under highly oxidized conditions in phase I within the first 83 days of operation, the corresponding mean removal rate was calculated as 0.388 ± 0.012, 0.384 ± 0.015 and 0.347 ± 0.06 mg As per day and thus contributed to a mean As retention of 94%, 94% and 92%, in the reactors R1, R2 and R3, respectively (Table 2). Addition of organic C-sources in phase II (day 83–144) showed an immediate effect with a considerably high As concentration in the outflow, which resulted in a lower As retention of only 15, 28 and 61% in the reactors R1, R2 and R3, respectively. During the experimental phase III (from day 144 to day 235) under anaerobic condition

Table 2 Summary of the treatment performance in the reactors R1, R2 and R3 during the whole operational period of 340 days (phase I–V). Phase

Duration (day)

I

0–83

II

83–144

III

144–235

IV

235–315

V

297–340

n: number of samples.

Parameter

Inflow rate Outflow rate Removal n Inflow rate Outflow rate Removal n Inflow rate Outflow rate Removal n Inflow rate Outflow rate Removal n Inflow rate Outflow rate Removal n

Unit

mg/day mg/day % – mg/day mg/day % – mg/day mg/day % – mg/day mg/day % – mg/day mg/day % –

Reactor R1

R2

R3

0.411 ± 0.002 0.023 ± 0.012 94 12 0.415 ± 0.001 0.352 ± 0.093 15 10 0.411 ± 0.009 0.278 ± 0.028 32 13 0.410 ± 0.002 0.243 ± 0.021 41 15 0.412 ± 0.001 0.276 ± 0.038 33 6

0.408 ± 0.002 0.023 ± 0.016 94 12 0.408 ± 0.001 0.294 ± 0.091 28 10 0.404 ± 0.009 0.045 ± 0.025 89 13 0.404 ± 0.002 0.017 ± 0.011 96 15 0.403 ± 0.001 0.037 ± 0.020 91 6

0.379 ± 0.065 0.032 ± 0.014 92 12 0.364 ± 0.056 0.146 ± 0.066 61 10 0.326 ± 0.052 0.033 ± 0.010 90 13 0.393 ± 0.009 0.021 ± 0.013 95 15 0.397 ± 0.001 0.019 ± 0.003 95 6

98

K.Z. Rahman et al. / Ecological Engineering 69 (2014) 93–105

All three reactors showed an efficient TOC removal (>74%) during the experimental phases with the addition of organic C-sources. Interestingly, a highly efficient TOC removal in the range of 86–94% was observed in the reactors R2 and R3, as compared to control reactor R1 (with a range of 74–86%). 3.3. Redox (Eh ) and pH

Fig. 3. As mass retention capacity of the model reactors R1, R2 and R3 calculated as a percentage of the inflow total As mass that retained inside the reactors during the whole experimental period of 340 days.

Table 3 Removal efficiency of SO4 2− -S and TOC within the model reactors R1, R2 and R3 in different experimental phases (phase I–V). Reactor

Parameter

Removal efficiency (%) Phase I

Phase II

Phase III

Phase IV

Phase IV

R1

SO4 2− -S TOC SO4 2− -S TOC SO4 2− -S TOC

− − − − − −

− 75 − 86 − 89

− 86 73 94 84 88

− 74 68 93 71 89

− 75 − 92 − 93

R2 R3

along with varying SO4 2− concentration, the data demonstrated a comparatively much higher As retention than the previous phase II with a mean value 89% and 90% in R2 and R3, respectively and only 32% was retained in the control reactor R1 (Table 2). Traces (2–3 ␮g As l−1 ) of volatile As compound [gaseous arsine (AsH3 )] was found in this experimental phase but only in the reactor R3. Data in Table 2 also shows that at least 5% more As was retained within reactor R3 when SO4 2− concentration was 10 mg S l−1 (in phase IV) instead of 25 mg S l−1 (in phase III). 3.2. SO4 2− -S and TOC removal efficiency Table 3 shows the removal efficiency of SO4 2− -S and TOC in different experimental phases within the reactors R1, R2 and R3. Overall, a comparatively higher SO4 2− -S removal was observed in R3 (71–84%) in comparison to the reactor R2 (68–73%) under Csurplus conditions (phase II–V). As expected, no SO4 2− reduction was observed under C-deficient and oxidized conditions in experimental phase I.

Table 4 shows the summary of the mean pH and redox values (also the ranges) in different experimental phases within the reactors R1, R2 and R3. No addition of organic C-sources ensured a persistent aerobic condition (with an Eh ∼ 324–795 mV) inside the reactors during experimental phase I (first 83 days). Reducing conditions were established immediately after the addition of organic C-sources (122.3 mg TOC l−1 ) with a drop down of Eh value from an aerobic (721–795 mV) to anaerobic (−217–−144 mV) condition within all three corresponding reactors in experimental phase II (Table 4). But a simultaneous addition of SO4 2− with constant C-dosage resulted in a low but relatively stable redox potential value (Eh ∼ −225–−149 mV) within all three reactors in phase III–V. Mean pH value of 3.1 ± 1.2, 3.6 ± 0.6 and 3.2 ± 0.4 was observed at the beginning (in phase I) and after addition of C-dosage in phase II, the pH remarkably raised to a mean value 4.3 ± 0.4, 6.6 ± 0.7 and 5.9 ± 0.5 in the pore water of the corresponding reactors R1, R2 and R3, respectively (Table 4). Afterwards, a relatively steady pH-value was observed in the reactors during the remaining experimental phases. Interestingly, a decrease of nearly 2–3 pH units was recorded in R3 than in the reactor R2 during the phases III–V. 3.4. Concentration of total As in plant biomass (shoots and roots) After the termination of all experimental phases, the sampling and measurement of plant biomass (shoots and roots) in terms of fresh and dry weight, water content (%) and thereby the dry weight (in kg) of total shoots and roots in each reactor were carried out and all these measurements are shown in Table 5. We measured a relatively lower total root biomass of 0.93 kg (fresh wt) collected from reactor R3, as compared to the other two reactors R1 and R2 with 1.10 and 1.16 kg (fresh wt), respectively. Based on the water content analysis data from the plant biomass samples collected from the reactors R1, R2 and R3, the dry weight of the total shoots were 0.16, 0.18, and 0.14 kg, and the corresponding dry weights of the total roots were calculated as 0.20, 0.21, and 0.16 kg, respectively (Table 5). The analytical results of the total As concentrations in plant biomass are presented in Fig. 4. The mean total As concentration was measured as 3.88 ± 0.13, 3.62 ± 0.07 and 3.57 ± 0.08 5 mg As kg−1 (dry wt) in the shoots and was considerably very high in

Table 4 Summary of the average pH, redox (Eh ) values (mean ± SD) and ranges of the values (from minimum to maximum) in each reactor during the whole operational period (phase I–V). Experimental phases Reactor

Parameter

I

II

III

IV

V

R1

pH

3.1 ± 1.2 (1.1–5.0) 499 ± 203 (324–740) 3.6 ± 0.6 (3.2–5.1) 597 ± 88 (396–721) 3.2 ± 0.4 (2.8–4.1) 670 ± 88 (495–795)

4.3 ± 0.4 (3.4–4.8) 28 ± 162 (−168–307) 6.6 ± 0.7 (5.1–7.4) −97 ± 176 (−217–277) 5.9 ± 0.5 (5.1–6.5) −52 ± 139 (−144–259)

4.9 ± 1.4 (3.5–6.6) −187 ± 6 (−196–−179) 7.4 ± 0.1 (7.2–7.5) −217 ± 4 (−225–−213) 5.8 ± 0.2 (5.5–6.1) −195 ± 17 (−211–−149)

6.7 ± 0.1 (6.6–6.8) −182 ± 6 (−189–−169) 7.6 ± 0.1 (7.4–7.8) −219 ± 5 (−225–−210) 5.2 ± 0.5 (4.4–5.6) −201 ± 13 (−215–−172)

6.9 ± 0.1 (6.8–7.1) −191 ± 4 (−196–−188) 7.7 ± 0.1 (7.6–7.8) −216 ± 10 (−224–−202) 3.9 ± 0.1 (3.8–4.1) −199 ± 23 (−220–−166)

Eh (mV) R2

pH Eh (mV)

R3

pH Eh (mV)

K.Z. Rahman et al. / Ecological Engineering 69 (2014) 93–105

99

Table 5 Plant analysis data in terms of fresh and dry weight of shoots and roots samples, water content and total dry weight of collected shoots and roots biomass (in kg) from each reactor after the termination of the experiment. Reactor

R1 R2 R3

Shoots

Roots

Sample fresh wt (g)

Sample dry wt (g)

Water content (%)

Total fresh wt (kg)

Total dry wt (kg)

Sample fresh wt (g)

Sample dry wt (g)

Water content (%)

Total fresh wt (kg)

Total dry wt (kg)

1.393 2.130 2.236

0.512 0.689 0.860

63 68 62

0.421 0.575 0.361

0.156 0.184 0.137

11.440 7.620 9.817

2.029 1.352 1.728

82 82 82

1.100 1.160 0.930

0.195 0.207 0.164

Table 6 Distribution of total As in each reactor, which includes the total inflow and outflow mass of As and the distribution in different compartments (shoots, roots, sediment, etc.) over the whole operation period of 340 days. Data are given in mg As, and the values in parentheses are percentile amounts of total inflow As. Reactor

Total inflow (mg As)

Shoots (mg As)

Roots (mg As)

Sediment (mg As)

Pore water (mg As)

Unaccountable (mg As)

Outflow (mg As)

R1 R2 R3

140.0 (100%) 135.1 (100%) 123.3 (100%)

0.6 (0.4%) 0.5 (0.4%) 0.4 (0.3%)

17.6 (12.6%) 22.9 (17.0%) 51.6 (41.9%)

6.2 (4.4%) 30.9 (22.9%) 21.2 (17.2%)

6.5 (4.6%) 13.9 (10.3%) 20.0 (16.2%)

36.5 (26.1%) 39.1 (28.9%) 11.1 (9.0%)

72.8 (52.0%) 27.7 (20.5%) 18.9 (15.3%)

the roots as 90 ± 2.83, 110.5 ± 3.54, 315 ± 4.24 mg As kg−1 (dry wt) within the reactors R1, R2 and R3, respectively. On the contrary, the mean As concentration was found to be only 0.5 ± 0.14 and 2.25 ± 0.2 mg As kg−1 (dry wt) in the control shoot and root samples of J. effusus (collected from natural, non-contaminated source), respectively. 3.5. Concentration of the total As in the sediment The results from the collected sediments showed a mean total As concentration of 206 ± 5, 806 ± 9 and 941 ± 4 mg As kg−1 (dry wt) in the corresponding three reactors R1, R2 and R3, respectively (Fig. 4). In comparison, a mean As concentration of 3.2 ± 0.4 mg As kg−1 (dry wt) in the control sediment sample (collected from a natural, noncontaminated source) was shown to be nearly 251–294 times lower

than the concentration found in the sediments collected from the reactors R2 and R3. 3.6. Data for As mass balance Table 6 represents a summary of the total As mass deposition and distribution in the different wetland compartments (shoots, roots, sediments etc.) and a complete mass balance calculation within the reactors over the whole operation period of 340 days. Considering the inflowing As mass within the reactors as 100%, the percentile value of the outflowing As mass and retention was estimated. For instance, a total mass of 140.0 mg As (100%) was fed as an inflow, and a total mass of 72.8 mg As (52%) was flushed out of the reactor R1 as outflow. The recovered As masses in the pore water, within the shoots, the roots and in the sediments collected from this reactor (R1) were measured as 6.5, 0.6, 17.6 and 6.2 mg As, which resulted in a nearly 4.6%, 0.4%, 12.6% and 4.4% of the total inflowing As mass retention in these compartments, respectively. The remaining 36.5 mg As (nearly 26% of the total inflow As mass) was considered to be unaccountable or retained in uncounted sinks. Similarly, after calculating the total As mass in the plant shoots, roots and sediments of other two reactors, it was observed that nearly 17% and 42% of the total inflowing As mass were accumulated/recovered/concentrated within the roots and nearly 23% and 17% were entrapped or deposited within the gravel bed (as sediments) of the reactors R2 and R3, respectively (Fig. 5). Only 9% of the total inflow As mass was calculated as unaccountable within R3 in comparison with the reactor R2, where nearly 29% was considered to be retained in uncounted sink or termed as unaccountable. In all three reactors, only a very little amount (<1% of the inflow As mass) was translocated into the shoots of J. effusus in this experiment. 4. Discussion 4.1. As mass retention capacities of the model reactors

Fig. 4. Concentrations of total As in plant shoots (A), roots (B) and sediments (C) collected from the reactors R1, R2 and R3 (the values are the mean of two replicates and the error bars are standard deviations).

Overall, the total As mass retention capacities of the reactors R2 and R3 were much higher (80% and 85%, respectively) than the control reactor R1 (only 48%). In general, SO4 2− loading was much higher in the reactors R2 and R3, as compared to the reactor R1. Therefore, amount of SO4 2− is playing an important role in case of As mass retention within the root-near environment of the rhizosphere in treatment wetlands. A highly efficient As mass retention (>92%) indicated a better performance under C-deficient and oxidized conditions (phase I) regardless to the concentration of SO4 2− in the inflow

100

K.Z. Rahman et al. / Ecological Engineering 69 (2014) 93–105

This is also in agreement with the fact that under reducing conditions, Fe(III) is reduced to Fe(II) and resulting in the mobilization of some of the adsorbed As, particularly from sediments and the plant root-zone (Kneebone et al., 2002). Moreover, reduction of As(V) to more mobile As(III) and subsequent As(III) enrichment due to microbial activity might also facilitated As remobilization (Rahman et al., 2008b). Therefore, it was evident in this study that the addition of electron donor has a huge impact on As mobility within the rhizosphere of helophytes in treatment wetlands. Presumably there were several competing reactions including dissolution or desorption, precipitation or adsorption and remobilization occurring simultaneously under such conditions. The occurrence of sulfide (S2– ) under anoxic conditions may immobilize As due to its high affinity to sulfide minerals or the formation of As-sulfide minerals (Bostick and Fendorf, 2003). This might be the reason of highly efficient As mass retention (>89%) under anaerobic condition by varying SO4 2− concentrations in R2 and R3 (phase III; Table 2). Other authors also suggested that under reducing environments and in the presence of S and Fe, As can form insoluble sulfide compounds (Buddhawong et al., 2005; Singhakant et al., 2009a), such as orpiment As2 S3 , in which As is present as As(III) and arsenopyrite (FeAsS). Dissimilatory reduction caused by iron- and sulfate-reducing bacteria is widely considered as the primary mechanism responsible for the rapid As reduction and release observed in anaerobic environments (Islam et al., 2004). Recently, Mattes et al. (2010) provided more details of the wetland system described in Duncan et al. (2004), highlighting that not only sulfatereducing bacteria played a role in As removal, but iron-oxidizing bacteria also making a significant contribution. Addition of relatively lower SO4 2− concentration (10 mg S l−1 in phase IV) contributed to 5% more As retention within the reactor R3 as compared to the SO4 2− concentration in phase III (25 mg S l−1 ). From this, it can also be concluded that a very high SO4 2− loading does not necessarily improve As removal performances within the rhizosphere of constructed wetlands. Fig. 5. Mass balance of As estimated as a percentage of inflow total As mass that retained in different wetland compartments investigated in the model reactors R1, R2 and R3.

artificial wastewater of all three reactors (see Fig. 2 and Table 2). Highly oxidized conditions (Eh ∼ 324–795 mV) were not favorable for microbial dissimilatory SO4 2− reduction and therefore no SO4 2− reductions or removal were found in the corresponding reactors (see phase I; Table 3). Hence, it can be concluded that the immobilization of As within the rhizosphere of helophytes was accomplished by mechanisms other than arsenic-sulfide precipitation (likely as As2 S3 ). Adsorption and/or concomitant coprecipitation of As, specifically with Fe(III) oxyhydroxides was the most probable reasons for As immobilization under such conditions (Bednar et al., 2005; Rahman et al., 2008b). Highly oxidized conditions due to plant root-mediated O2 release and re-oxidation of reduced As-species probably influenced higher sorption and precipitation reactions within the rhizosphere of Juncus effusus. Addition of organic C-sources probably caused an immediate mobilization of As within the reactors under limited SO4 2− and reducing conditions (Rahman et al., 2008b), and hence a comparatively low As removal efficiency was observed during experimental phase II in all three reactors (Table 2). Under such conditions, there was probably a shift in terminal electron acceptors from O2 to Fe(III) oxyhydroxides. Thereby, reductive dissolution of Fe(III) oxyhydroxides mediated by microbial activities probably contributed to a release of immobilized As fraction in the aqueous phase of all the corresponding reactors (Smedley and Kinniburgh, 2002).

4.2. Correlation of Redox (Eh ), pH and As mass retention The mobility and plant availability of many trace and toxic metals and metalloids in wetland soils are often governed by oxidation-reduction (redox) potential and associated pH in the rhizosphere (Gambrell, 1994), while the oxygen released from the roots probably influences the redox and pH of rhizosphere (Yang et al., 2010). Flooding conditions induce an enrichment of metals in soils surrounding the roots of wetland plants (Wright and Otte, 1999). Under C-deficient condition in this study, a very high redox potential (with an Eh ∼ 324–795 mV) were obtained inside the reactors, which clearly indicated an aerobic condition in the root-near environment of the rhizosphere. Aerobic conditions were unfavorable for SO4 2− reducing bacteria and therefore it can be concluded that adsorption and co-precipitation of As, specifically predominant and thermodynamically more stable As(V) with iron(oxy)hydroxide favored As retention under such oxic conditions. Reducing conditions play an important role for SO4 2− reduction, which requires a reducing environment and an electron donor. Under C-surplus condition along with available SO4 2− (in R2 and R3), the Eh values varied in between −149 and −225 that resulted in a higher SO4 2− removal (68–84%) and a highly efficient As mass retention (>89%) within the reactors in this study (phase III–IV; Tables 2–4). Different authors have reported different Eh values required by sulfate-reducing bacteria to thrive: less than −200 mV (Cabrera et al., 2006), less than −100 mV (Willow and Cohen, 2003), or between −150 and −200 mV (Tuttle, 1969). Rahman et al. (2008a) noted that microbial SO4 2− reduction was greater

K.Z. Rahman et al. / Ecological Engineering 69 (2014) 93–105

under redox potential values between −160 and −190 mV, and that higher SO4 2− removal contributed to higher removal of As. However, these conclusions were drawn based on their measured water quality parameters (Eh , TOC), without monitoring microbial community composition or function directly. Microorganisms can enhance the removal of As by mediating redox and precipitation processes (Lizama et al., 2011). Talukder et al. (2012) showed that under aerobic water management (Eh ∼ 135–138 mV), As uptake by the rice plant parts was significantly less as compared to anaerobic environment (Eh ∼ −41–−76 mV). The study has confirmed that anaerobic water management is the main reason for the high enhanced As uptake in rice. This was also in agreement in this study with a remarkably higher As uptake in the plant biomass (mainly roots) under a persistent anaerobic environment, even in the roots from the control reactor R1 with a concentration of 90 ± 2.83 mg As kg−1 (dry wt) (Tables 4 and 6; Fig. 4B). Several studies have shown that many factors affect the bioaccumulation (biosorption) of metals and metalloids in aquatic ecosystems. Among the physico-chemical factors, pH is possibly the most important (Gadd, 2009; Lizama et al., 2011) and plays a vital role for As mass retention within the rhizosphere. Persistent low pH value in all corresponding reactors might referred to as a rhizosphere acidification under oxidized conditions, which could have resulted from the effect of K+ uptake and release of H+ under conditions of low redox buffer capacity (Rahman et al., 2008b). Arsenic-induced root exudation might also a probable reason for a reduced pH in the rhizosphere. One study showed that the reduction in rhizosphere pH increases the chemical activity of most metals, thereby increasing As uptake by the ferns (Tu and Ma, 2004). Nevertheless, As removal is strongly dependent on the pH and these low pH in the reactors remarkably favored As removal (>90%, predominantly as arsenate) under this C-deficient oxic condition. In general, sulfate-reducing microorganisms do not grow well at pH values below 5.5 and prefer higher levels of alkalinity, with 6.6 being optimal (Govind et al., 1999). Therefore, rapid As removal processes took place within the rhizosphere by other processes than dissimilatory SO4 2− reduction. Addition of organic C-sources caused simultaneously a reducing condition inside the reactors and an increment of rhizosphere pH level due to buffering effect (phase II; Table 4). Rapid SO4 2− reduction by dissimilatory sulfate-reducers utilizing already enriched S-pool might produce alkalinity and triggered the increment of rhizosphere pH under such conditions. Better As mass retention in these experimental phases were observed in a highly consistent pH with an average pH value of 7.0 ± 0.5 and 5.7 ± 0.5 in R2 and R3, respectively (phase III–V, Tables 2 and 4). Changes in alkalinity can indicate changes in the speciation of As and sulfate-reducing bacteria activity, since these bacteria can also provide alkalinity to the water and affect its pH (Cohen, 2006). Lizama et al. (2011) in a review paper stated that the changes in the speciation of As can affect pH, as the oxidation of As(III) to As(V) decreases the pH value, whereas the precipitation of arseno-sulfides increases it. This was also demonstrated within the rhizosphere of Juncus effusus in this study. From a very high to a relatively lower or even limited SO4 2− loading might also be resulting in a decreasing tendency of rhizosphere pH, even though under C-surplus conditions (phase III–V in R3; Tables 1 and 4). In fact, it was very interesting to observe a comparatively lower mean pH in all thorough the experimental phases (I–V) of R3, specifically a decrease of nearly 2–3 pH units during the phases III–V as compared to the reactor R2, This lower pH in the rhizosphere might be a convincing reason for higher amount of re-mobilized As in the pore water (16.2%) as well as higher As concentration in the roots collected from the reactor R3 than the roots from R1 and R2 (Figs. 4 and 5). So, an increase of rhizosphere pH

101

might be favoring the immobilization of liable and exchangeable As fractions in the rhizosphere but a lower pH within the root vicinity of constructed wetlands might consequently enhance plant uptake. Therefore, a pH change may affect the bioavailability of As and may also play a vital role for plant As uptake within the rhizosphere. For example, Wells and Richardson (1985) reported a decrease in arsenate uptake in the moss Hylocomium splendens with increasing pH. In this moss, arsenate uptake was optimal at pH 5, where H2 AsO4 − was the dominant form in solution. As the pH increased to pH 8, where HAsO4 2− was the dominant anion, arsenate uptake decreased. Mukherjee and Kumar (2005) observed in the aquatic plant Pistia stratiotes that the maximum As uptake rates occurred at pH 6.5.

4.3. Accumulation of As in plant biomass Terrestrial plants are able to accumulate As to a substantial extent (Visoottiviseth et al., 2002) but survive the stress to differing degrees of vitality. During the whole experimental period (341 days), As was accumulated in different compartments of the plant biomass, mainly into plant shoots and roots. No remarkable differences can be found in terms of As translocation into the plant shoots of the corresponding reactors. In general, the mean As concentrations (dry wt) within the plant shoots of the reactors were 7-fold higher than the As concentration in the control shoot samples (Fig. 4A). The concentrations of total As in the plant roots were extremely higher than in the shoots in this study (Fig. 4). In general, total As concentrations exhibited in the roots were 23–88 times higher than to their shoots in all reactors. A vast majority of As were found to be fixed in/on the roots and only a very limited amount was translocated to the shoots. Buddhawong (2005) and other authors also reported similar facts when dealing with As and heavy metals in constructed wetlands. Roots continuously remained under direct exposure to As in both oxic and anoxic environment and translocation rate of As into shoots presumably depends on other factors and varies within different plant species. Carbonell et al. (1998) studied the As content in Spartina alterniflora and found As in the range of 0.80–1.77 mg kg−1 in shoots and 6.87–86.60 mg kg−1 in the roots. The much higher accumulation of As in the plant roots compared to the above-ground biomass (shoots) corresponded to the studies for Typha latifolia, Equisetum fluviatile, Triglochin palustre, and Sparganium sp. (Dushenko et al., 1995). However, Rahman et al. (2011) showed a translocation with a range of 12–20 times from the roots to the shoots when investigating the fate of As in a laboratoryscale horizontal subsurface-flow constructed wetland. The low As translocation from roots to shoots is probably due to the fact that arsenate is rapidly reduced to arsenite in the roots, followed by complexation with thiols and sequestration in the root vacuoles (Zhao et al., 2009). Iron plaque on root surfaces has been shown to control the uptake and transfer of As by rice (Liu et al., 2005). Several other studies have shown that roots accumulate more As than do shoots (Adhikari et al., 2011; Hozhina et al., 2001). Different reasons may explain why As remains mostly in plant roots, such as limited translocation of As from roots to shoots (Wang et al., 2002), the presence of Fe and S (Zhao et al., 2010), the effect of As speciation in the mechanism of translocation and its relationship to the phosphate transporter (Dhankher, 2005), and the formation of As(III)-phytochelation (PC) complexes in roots and subsequent sequestration in root vacuoles (Xue and Yan, 2011). However, the form of As that is translocated to shoots or how this translocation occurs is not known (Dhankher, 2005; Lizama et al., 2011). Further studies on these As species are therefore needed to investigate the mechanism of As uptake, translocation and accumulation,

102

K.Z. Rahman et al. / Ecological Engineering 69 (2014) 93–105

considering both the water and the sediment and taking into account the relationship with Fe, S and phosphate. Our results with a mean concentration of 90 ± 2.8, 110.5 ± 3.5 and 315 ± 4.3 mg As kg−1 within the roots of Juncus effusus collected from the three corresponding reactors were definitely not surprising under an ideal-flow condition, also depending on various loading conditions and exposure time. Nearly 3-fold higher As concentrations as compared to the roots from the reactor R2 and 3.5-fold higher from the control reactor R1 clearly indicated a higher amount of As retention, accumulation, adsorbed, metabolized to other forms and/or translocated into the root vicinity of Juncus effusus in the reactor R3, where organic carbon and SO4 2− loading were abundant. Nearly 5-fold lower C/S ratio (TOC: SO4 2− S) within the reactor R3 might facilitate lower pH and consequent remobilization caused free or exchangeable As in the pore water and hence a higher concentration of As in roots as compared to the reactor R2 was found. Higher enrichment of total sulfur due to the re-oxidation of reduced S2– to oxidized S-species like elemental sulfur (S0 ), SO4 2− etc. might also cause remobilization of As (Rahman et al., 2008b) and hence a higher As-concentration in the roots of the reactor R3 was established. But interestingly, the dry weight of the total root biomass in reactor R3 was comparatively lower than the dry weight of the roots collected from other two reactors R1 and R2 (Table 5). Therefore, a higher As accumulation or uptake by the plant roots might have some effect on plant root growth under such long-term investigation. The results of the study by Favas et al. (2012) revealed high bioaccumulation levels of As in several species at a magnitude much higher than the concentration in the surrounding water. The highest concentrations of As were found in the submerged species Callitriche lusitanica (2346 mg As kg−1 dry wt), but the measured concentrations in the other emergent plants were significantly lower, even in the rhizomes/roots, such as T. latifolia (4.17 mg As kg−1 dry wt) and J. effusus (14.4 mg As kg−1 dry wt). The sorption and bio-concentration of As were measured in Schoenoplectus californicus and Typha angustifolia in a pilot-scale constructed wetland receiving wastewater inflows containing As at potentially hazardous levels (Sundberg-Jones and Hassan, 2007). Moreover, it has already been shown that roots and shoots As concentrations significantly increased with increasing As application rates to the rooting medium of a wetland ecosystem (Mkandawire and Dudel, 2005; Sundberg-Jones and Hassan, 2007). The root vicinity of Juncus effusus was continuously exposed to an artificial wastewater contaminated with As under idealized flow conditions, which might enhance bioaccumulation of As and the uptake by the roots in this investigation. The increase of As uptake attributed to iron plaque on root surfaces has been reported in rice and Spirodela polyrhiza (Liu et al., 2005; Rahman et al., 2008c). At the root plaque interface, siderophores or phytosiderophores exuded by microbes or roots may complex with Fe(III) and mobilize Fe-bound arsenate, taken up through phosphate co-transporters, which may lead to simultaneous uptake of Fe and arsenate (Liu et al., 2005). This phenomenon has also been observed for other metals such as Cu and Zn (Ye et al., 2001). Fitz and Wenzel (2002) reported that root-induced changes in the rhizosphere altered its chemical composition and facilitated As uptake by Pteris vittata L (Chinese brake fern), the first-known arsenic hyperaccumulator. Root architecture and physiology, rootinduced changes in water and nutrient availability, root exudates, and fungal and bacterial associations (Gahoonia and Nielsen, 2003) are all components of the dynamic rhizosphere system and probably facilitated higher As retention within the roots in this study. The formation of iron plaque on the macrophyte roots has a high affinity for As, tending to have a higher affinity for As(V) than As(III) (Chen et al., 2005), and might also be a probable reason for

the higher As concentration within the roots. The ability to carry oxygen from the air down to its stem and discharge it in the rhizosphere through the roots (Brammer and Ravenscroft, 2009) creates oxidized micro-environments around the roots in which iron is oxidized and precipitated to form a coating (Liu et al., 2006). This oxidation in the rhizosphere may also lead to As precipitating on root surfaces. The presence of available sulfur due to a high SO4 2− loading might also be a reason for the formation of iron plaque on the root surfaces of Juncus effusus in this study. Nevertheless, more investigations on the effects of O2 release through the plant-roots, changing of redox and pH for the formation of Fe plaque on root surfaces and hence altering the bioavailability of As in the rhizosphere of constructed wetlands are needed. 4.4. Distribution of As in sediment A big variety of suspended and biofilm-fixed microorganisms on plant-root surfaces along with surplus organic C produced sludge sediments, which were collected from the reactors after the termination of the experiments. Analytical results of the collected and dried sediments showed a wide variation with a mean concentration of 206 ± 5, 806 ± 9 and 941 ± 4 mg As kg−1 (dry wt) in R1, R2 and R3, respectively (Fig. 4C). Clear evidence of intensive dissimilatory SO4 2− reduction, highly efficient TOC removal, high As adsorption and/or precipitation in the sludge sediment of the reactor R3 suggested higher microbial activities within the rhizosphere in presence of higher SO4 2− loading as compared to other two reactors R1 and R2. 4.5. As mass balance Calculation of total As mass balance was carried out for all the experimental reactors at the end of our investigations. Since the reactors were closed tightly enough and well-controlled systems, we were able to obtain a quantitative mass balance of total As by expressing its distribution as percentage of the total mass loaded into the reactors. It was observed that the reactors R2 and R3 were performing with a much better As deposition than the control reactor R1 with limited SO4 2− loading. A substantially higher amount of As mass was accumulated in the roots than sludge sediments in reactor R3 with a higher SO4 2− loading. Based on the concentration of As in the plant biomass (shoots and roots) and sludge sediments of the reactor R3, we calculated a 42% of total inflow As mass were accumulated in the roots of Juncus effusus and 17.2% were deposited within the sediment (Fig. 5). A high level of As in the roots of R3 might be related to a higher SO4 2− loading (i.e. lower C/S ratio), consequently a lower pH and formation of elemental sulfur (S0 ) within this reactor, hence a greater remobilization of As in the solution for plant-root uptake. This was demonstrated by a comparatively higher As retention in the pore water (16.2%) and lower retention (9%) in unaccountable sink than the reactors R1 and R2 (Table 4; Fig. 5). The accumulation or deposition of As in the sediments occurred probably due to the formation of insoluble precipitates like As2 S3 via abiotic processes or, more probably, driven by the high level of microbial activity and biotic processes incorporated to the organic matter content within the rhizosphere. Along with sludge sediments, plant roots were also therefore considered to be the primary sink for the sequestration of As in the rhizosphere, which agrees with an As accumulation range of 44–49% within the roots of Juncus effusus investigated in the planted horizontal subsurfaceflow constructed wetlands (Rahman et al., 2011) and partially agrees with many other studies of treatment wetland systems (Sundaravadivel and Vigneswaran, 2001). Several studies have identified aquatic plants with high As content: Lagarosiphon major

K.Z. Rahman et al. / Ecological Engineering 69 (2014) 93–105

(300 mg As kg−1 dry wt) (Brooks and Robinson, 1998), Egeria densa (>1000 mg As kg−1 dry wt) (Robinson et al., 1995), C. demersum (>1000 mg As kg−1 dry wt) (Robinson et al., 1995), and Lemna gibba (1021.7 ± 250.8 mg As kg−1 dry wt) (Mkandawire et al., 2004). Therefore, the accumulation of As depends on the type of plant (Lizama et al., 2011; Zhao et al., 2010), and the potential of some aquatic plants to accumulate As has already been well demonstrated. This potential of aquatic plants to accumulate As supports their possible use in phytoremediation of As-contaminated water (Xue and Yan, 2011). Li et al. (2011) suggested that most of As uptake was accumulated in 9 wetland species root tissues rather than on root surfaces or in shoot tissues. The root tissue is therefore the main barrier for As transport in wetland plants. These might be critical considerations for improving the efficiency of As removal from wastewater in full-scale constructed wetland systems. Only <1% of total inflow As mass were accumulated into the plant shoots (Fig. 5). This result clearly indicated the minor role by the plant shoots in terms of As uptake and showed a very little contribution to the overall As mass balance. Standing or pore water accounted for nearly 5%, 10% and 16% of total inflow As mass in the three respective reactors with an increasing SO4 2− concentrations. Enrichment of total sulfur due to the re-oxidation of reduced S2– to oxidized S-species (S0 , SO4 2− etc.) might cause remobilization of As and hence a higher percentile As mass can be found in the pore water of the reactor R3 with a higher SO4 2− loading than the other two reactors R1 and R2. Nearly a total of 52%, 21% and 15% of the inflow As mass were passed through the reactors R1, R2, and R3, respectively and collected at the outlet. The larger amounts of As (52% of the total inflow As mass) were flushed out through the outlet of the control reactor R1 in comparison to the reactor R3 (15%), which indicated a highly stable bonding of As with the attached biofilms within plant root zone of the rhizosphere associated with higher SO4 2− loading. Therefore, the higher the SO4 2− loading as well as an enhanced dissimilatory SO4 2− reduction in the reactor (Rahman et al., 2008b), the better will be the As binding capacity within the root-near environment of the rhizosphere in constructed wetlands. Only traces (2–3 ␮g l−1 ) of inorganic volatile arsine (AsH3 ) were measured in this study, which resulted in a very small amount of As mass that was left out of the systems due to volatilization. Nearly 26%, 29% and 9% of total As mass in these three reactors were accumulated into unaccountable sink, which could be due to other microbial reactions, adsorption/precipitation to other unknown sink or even lost due to volatilization. Some of the retained or trapped As might have been bio-transformed by plant-root activity and associated microbes to other unidentified organic compounds, which prevented the calculation of a complete As mass balance in this study. From the previous results (Buddhawong, 2005), it was found that the gravel material could not absorb As mass in substantial amounts from the solution. Therefore, this kind of gravel itself had no remarkable impact on the mass balance of As in constructed wetland systems. Fractional analysis of deposited sediment might also be necessary to investigate different forms of As that were retaining within the rhizosphere of helophytes in treatment wetlands. Singhakant et al. (2009) reported about the forms of the retained As by using sequential fractionation, which could indicate As complexation with Fe and Mn on the media surface of 31–38% and As trapping into the media of 42–52% of the total As. However, more attention should be given to the accumulation of As in different wetland compartments and to avoid potential toxic effects the accumulated As could pose to the wetland plants.

103

5. Conclusions This study reveals that wetland plants possess high As concentrations on roots and a very low As translocation from roots to the shoots. Redox potential and pH within the root-zone are greatly influencing As deposition and plant uptake. An increase of rhizosphere pH (alkalinity) might be favoring the immobilization of liable and exchangeable As within the root vicinity and a decrease of pH might cause remobilization of As and enhance plant (root) uptake. Sulfate loading with an elevated SO4 2− concentration is also playing an important role, which facilitates an efficient SO4 2− reduction and TOC removal as well as a higher As retention under microbial dissimilatory SO4 2− reducing condition within the rhizosphere. Consequently, high enrichment of total sulfur due to the re-oxidation might cause remobilization of As and enhance plant-root mediated As uptake. Nevertheless, the higher the SO4 2− loading, the better will be the As binding capacity within the root-near environment of the rhizosphere in constructed wetlands. Plant shoots (Juncus effusus) are playing a minor role in terms of As uptake and contributing very little to the overall As mass balance. Roots and sludge sediments are considered to be the primary sinks for As retention in the rhizosphere of constructed wetlands. Accumulation of a small portion of As to unaccountable sinks and/or a probable loss due to volatilization or other unknown reasons demands further research under such strict conditions on a priority basis in order to protect the surrounding environment from any toxic effects of volatile As compounds. Moreover, different physico-chemical properties on the root surface, storage characteristics in root tissues of wetland plants and precipitation/adsorption/retention of different As species clearly requires further investigation. Future studies should also intend to focus on exploring the patterns of microbial ecology and their interactions, associated As toxicity on plants and microbial biomass and volatilization of As under dynamic redox conditions within the rhizosphere of constructed wetland system. Acknowledgments The work was funded by a grant from the German Federal Ministry of Education and Research under the International Postgraduate Study in Water Technology (BMBF-IPSWaT) program and by grants from the Helmholtz-Centre for Environmental Research–UFZ, Leipzig, Germany. The authors would like to thank Kerstin Puschendorf, Ines Mäusezahl, Karsten Marien, Jürgen Steffen, Reinhard Schumann, and Uwe Kappelmeyer for their outstanding technical and analytical support. References Adhikari, A.R., Acharya, K., Shanahan, S.A., Zhou, X., 2011. Removal of nutrients and metals by constructed and naturally created wetlands in the Las Vegas Valley, Nevada. Environ. Monit. Assess. 180, 97–113. Bednar, A.J., Garbarino, J.R., Ranville, J.F., Wildeman, T.R., 2005. Effects of iron on arsenic speciation and redox chemistry in acid mine water. J. Geochem. Explor. 85, 55–62. Bezbaruah, A.N., Zhang, T.C., 2004. pH, redox, and oxygen microprofiles in rhizosphere of bulrush (Scirpus validus) in a constructed wetland treating municipal wastewater. Biotechnol. Bioeng. 88 (1), 60–70. Bostick, B.C., Fendorf, S., 2003. Arsenite sorption on troilite (FeS) and pyrite (FeS2 ). Geochim. Cosmochim. Acta 67, 909–921. Brammer, H., Ravenscroft, P., 2009. Arsenic in groundwater: a threat to sustainable agriculture in South and South-east Asia. Environ. Int. 35, 647–654. Brooks, R.R., Robinson, B.H., 1998. Aquatic phytoremediation by accumulator plants. In: Brooks, R.R. (Ed.), Plants that Hyperaccumulate Heavy Metals: Their Role in Archaeology, Microbiology, Mineral Exploration, Phytomining and Phytoremediation. CAB Int., Wallingford, pp. 203–226. Buddhawong, S., 2005. Constructed Wetlands and their Performance for Treatment of Water Contaminated with Arsenic and Heavy Metals. University of Leipzig, Germany, Ph.D. Thesis.

104

K.Z. Rahman et al. / Ecological Engineering 69 (2014) 93–105

Buddhawong, S., Kuschk, P., Mattusch, J., Wießner, A., Stottmeister, U., 2005. Removal of arsenic and zinc using different model wetlands systems. Eng. Life Sci. 5 (3), 247–252. Cabrera, G., Pérez, R., Gómez, J.M., Ábalos, A., Cantero, D., 2006. Toxic effects of dissolved heavy metals on Desulfovibrio vulgaris and Desulfovibrio sp. strains. J. Hazard. Mater. 135, 40–46. Carbonell, A.A., Aarabi, M.A., DeLaune, R.D., Gambrell, R.P., Patrick Jr., W.H., 1998. Arsenic in wetland vegetation: availability, phytotoxicity, uptake and effects on plant growth and nutrition. Sci. Total Environ. 217, 189–199. Chen, Z., Zhu, Y.G., Liu, W.J., Meharg, A.A., 2005. Direct evidence showing the effect of root surface iron plaque on arsenite and arsenate uptake into rice (Oryza sativa) roots. New Phytol. 165, 91–97. Cohen, R.R.H., 2006. Use of microbes for cost reduction of metal removal from metals and mining industry waste streams. J. Clean. Prod. 14, 1146–1157. Darrah, P.R., Jones, D.L., Kirk, G.J.D., Roose, T., 2006. Modelling the rhizosphere: a review of methods for ‘upscaling’ to the whole-plant scale. Eur. J. Soil Sci. 57, 13–25. Daus, B., Mattusch, J., Wennrich, R., Weiß, H., 2002. Investigation on stability and preservation of arsenic species in iron rich water samples. Talanta 58, 57–65. Dhankher, O.P., 2005. Arsenic metabolism in plants: an inside story. New Phytol. 168, 503–505. Duncan, W.F.A., Mattes, A.G., Gould, W.D., Goodazi, F., 2004. Multi-stage biological treatment system for removal of heavy metal contaminants. In: Rao, S.R., Harrison, F.W., Kosinski, J.A., Amaratunga, L.M., Cheng, T.C., Richards, G.G. (Eds.), The Fifth International Symposium on Waste Processing and Recycling in Mineral and Metallurgical Industries. Canadian Institute of Mining, Metallurgy and Petroleum, Hamilton, Ontario, Canada, pp. 469–483. Dushenko, W.T., Bright, D.A., Reimer, K.J., 1995. Arsenic bioaccumulation and toxicity in aquatic macrophytes exposed to gold-mine effluent: relationships with environmental partitioning. Aquat. Bot. 50, 141–158. Favas, P.J.C., Pratas, J., Prasad, M.N.V., 2012. Accumulation of arsenic by aquatic plants in large-scale field conditions: opportunities for phytoremediation and bioindication. Sci. Total Environ. 433, 390–397. Fitz, W.J., Wenzel, W.W., 2002. Arsenic transformations in the soil–rhizosphere–plant system: fundamentals and potential application to phytoremediation. J. Biotechnol. 99, 259–278. Gadd, G.M., 2009. Biosorption: critical review of scientific rationale, environmental importance and significance for pollution treatment. J. Chem. Technol. Biotechnol. 84, 13–28. Gahoonia, T.S., Nielsen, N.E., 2003. Phosphorus (P) uptake and growth of a root hairless barley mutant (bald root barley, brb) and wild type in low- and high-P soils. Plant Cell Environ. 26, 1759–1766. Gambrell, R.P., 1994. Trace and toxic metals in wetlands – a review. J. Environ. Qual. 23, 883–891. García, J., Rousseau, D.P.L., Morató, J., Lesage, E., Matamoros, V., Bayona, J.M., 2010. Contaminant removal processes in subsurface-flow constructed wetlands: a review. Crit. Rev. Environ. Sci. Technol. 40, 561–661. Govind, R., Yong, W., Tabak, H., 1999. Studies on bio-recovery of metals from acid mine drainage. In: Leeson, A., Allemans, B. (Eds.), Bioremediation of Metals and Inorganic Compounds. Battelle Press, Columbus, OH, p. 37. Hozhina, E.I., Khramov, A.A., Gerasimov, P.A., Kumarkov, A.A., 2001. Uptake of heavy metals, arsenic, and antimony by aquatic plants in the vicinity of ore mining and processing industries. J. Geochem. Explor. 74, 153–162. Islam, F.S., Gault, A.G., Boothman, C., Polya, D.A., Chatterjee, D., Lloyd, J.R., 2004. Direct evidence of arsenic release from Bengali sediments due to metal-reducing bacteria. Nature 430, 68–71. Jacks, G., Bhattacharya, P., Routh, J., Martin, M.T., 2002. Arsenic cycling in a covered mine tailings deposit. In: Schulz, H.D., Hadeler, A. (Eds.), Geo. Proc. Wiley Publication, Northern Sweden, pp. 303–309. Kadlec, R.H., Knight, R.L., 1996. Treatment Wetlands. CRC Press, Boca Raton, pp. 893. Kappelmeyer, U., Wießner, A., Kuschk, P., Kästner, M., 2002. Operation of a universal test unit for planted soil filters–planted fixed bed reactor. Eng. Life Sci. 2, 311–315. Kirk, M.F., Holm, T.R., Park, J., Jin, Q., Sanford, R.A., Fouke, B.W., Bethke, C.M., 2004. Bacterial sulfate reduction limits natural arsenic contamination in groundwater. Geology 32, 953–956. Kneebone, P.E., O’Day, P.A., Jones, N., Hering, J.G., 2002. Deposition and fate of arsenic in iron- and arsenic-enriched reservoir sediments. Environ. Sci. Technol. 36, 381–386. Larios, R., Martínez, R.F., LeHecho, I., Rucandio, I., 2012. A methodological approach to evaluate arsenic speciation and bioaccumulation in different plant species from two highly polluted mining areas. Sci. Total Environ. 414, 600–607. Li, H., Ye, Z.H., Wei, Z.J., Wong, M.H., 2011. Root porosity and radial oxygen loss related to arsenic tolerance and uptake in wetland plants. Environ. Pollut. 159, 30–37. Liesack, W., Schnell, S., Revsbech, N.P., 2000. Microbiology of flooded rice paddies. FEMS Microbiol. Rev. 24, 625–645. Liu, W.J., Zhu, Y.G., Hu, Y., Williams, P.N., Gault, A.G., Meharg, A.A., 2006. Arsenic sequestration in iron plaque, its accumulation and speciation in mature rice plants (Oryza sativa L.). Environ. Sci. Technol. 40, 5730–5736.

Liu, W.J., Zhu, Y.G., Smith, F.A., 2005. Effects of iron and manganese plaques on arsenic uptake by rice seedlings (Oryza sativa L.) grown in solution culture supplied with arsenate and arsenite. Plant Soil 277, 127–138. Lizama, K.A., Fletcher, T.D., Sun, G., 2011. Removal processes for arsenic in constructed wetlands. Chemosphere 84, 1032–1043. Mattes, A.G., Evans, L., Glasauer, S., 2010. Biologically based treatment system for removal of high concentrations of heavy metal and arsenic contaminants from landfill leachate–a look at a system after 10 years of full time operation. In: 12th International Conference on Wetland Systems for Water Pollution Control, Venice, Italy. Meharg, A.A., 2004. Arsenic in rice—understanding a new disaster for South-East Asia. Trends Plant Sci. 9, 415–417. Mkandawire, M., Dudel, E.G., 2005. Accumulation of arsenic in Lemna gibba L. (duckweed) in tailing waters of two abandoned uranium mining sites in Saxony, Germany. Sci. Total Environ. 305, 81–89. Mkandawire, M., Taubert, B., Dudel, E.G., 2004. Capacity of Lemna gibba L. (duckweed) for uranium and arsenic phytoremediation in mine tailing waters. Int. J. Phytorem. 6, 347–362. Mukherjee, S., Kumar, S., 2005. Arsenic uptake potential of water lettuce (Pistia Stratiotes L.). Int. J. Environ. Stud. 62, 249–258. Pantsar-Kallio, M., Korpela, A., 2000. Analysis of gaseous arsenic species and stability studies of arsine and trimethylarsine by gas-chromatography–mass spectrometry. Anal. Chim. Acta 410, 6570. Rahman, M.A., Hasegawa, H., Ueda, K., Maki, T., Rahman, M.M., 2008a. Arsenic uptake by aquatic macrophyte Spirodela polyrhiza L.: interactions with phosphate and iron. J. Hazard. Mater. 160, 356–361. Rahman, K.Z., Wiessner, A., Kuschk, P., Mattusch, J., Kästner, M., Müller, R.A., 2008b. Dynamics of arsenic species in laboratory-scale horizontal subsurface-flow constructed wetlands treating an artificial wastewater. Eng. Life Sci. 8, 603–611. Rahman, K.Z., Wiessner, A., Kuschk, P., Mattusch, J., Offelder, A., Kästner, M., Müller, R.A., 2008c. Redox dynamics of arsenic species in the root-near environment of Juncus effusus investigated in a macro-gradient-free rooted gravel bed reactor. Eng. Life Sci. 8 (6), 612–621. Rahman, K.Z., Wiessner, A., Kuschk, P., van Afferden, M., Mattusch, J., Müller, R.A., 2011. Fate and distribution of arsenic in laboratory-scale subsurface horizontalflow constructed wetlands treating an artificial wastewater. Ecol. Eng. 37, 1214–1224. Rai, U.N., Sinha, S., Tripathi, R.D., Chandra, P., 1995. Wastewater treatability potential of some aquatic macrophytes: removal of heavy metals. Ecol. Eng. 5, 5–12. Robinson, B.H., Brooks, R.R., Outred, H.A., Kirkman, J.H., 1995. The distribution and fate of arsenic in the Waikato river system, North Island, New Zealand. Chem. Spec. Bioavail. 7, 89–96. Sasmaza, A., Obek, E., 2009. The accumulation of arsenic, uranium, and boron in Lemna gibba L. exposed to secondary effluents. Ecol. Eng. 35, 1564–1567. Schmidt, A.C., Reisser, W., Mattusch, J., Wennrich, R., Jung, K., 2004. Analysis of arsenic species accumulation by plants and the influence on their nitrogen uptake. J. Anal. At. Spectrom. 19, 172–177. Sharma, V.K., Sohn, M., 2009. Aquatic arsenic: toxicity, speciation, transformations, and remediation. Environ. Int. 35, 743–759. Singhakant, C., Koottatep, T., Satayavivad, J., 2009. Fractional analysis of arsenic in subsurface-flow constructed wetlands with different length to depth ratios. Water Sci. Technol. 60, 1771–1778. Singhakant, C., Koottatep, T., Satayavivad, J., 2009a. Enhanced arsenic removals through plant interactions in subsurface-flow constructed wetlands. J. Environ. Sci. Health A 44, 163–169. Smedley, P.L., Kinniburgh, D.G., 2002. A review of the source, behavior and distribution of arsenic in natural waters. Appl. Geochem. 17, 517–568. Sun, G., Saeed, T., 2009. Kinetic modelling of organic matter removal in horizontal flow reed beds for domestic sewage treatment. Process Biochem. 44, 717–722. Sundaravadivel, M., Vigneswaran, S., 2001. Constructed wetland for wastewater treatment. Environ. Sci. Technol. 31 (4), 351–409. Sundberg-Jones, S.E., Hassan, S.M., 2007. Macrophyte sorption and bioconcentration of elements in a pilot constructed wetland for flue gas desulfurization wastewater treatment. Water Air Soil Pollut. 183 (1-4), 187–200. Talukder, A.S.M.H.M., Meisner, C.A., Sarkar, M.A.R., Islam, M.S., Sayre, K.D., Duxbury, J.M., Lauren, J.G., 2012. Effect of water management, arsenic and phosphorus levels on rice in a high-arsenic soil–water system: II arsenic uptake. Ecotoxicol. Environ. Saf. 80, 145–151. Tripathi, R.D., Srivastava, S., Mishra, S., Singh, N., Tuli, R., Gupta, D.K., Maathuis, F.J.M., 2007. Arsenic hazards: strategies for tolerance and remediation by plants. Trends Biotechnol. 25, 158–165. Tu, S., Ma, L.Q., 2004. Comparison of arsenic uptake and distribution in arsenic hyperaccumulator Pteris vittata L. and non-hyperaccumulator Nephrolepis exaltata L. J. Plant Nutr. 27, 1227–1242. Tuttle, J.H., 1969. Microbial sulfate reduction and its potential utility as an acid mine water pollution abatement procedure. Appl. Microbiol. 17, 297–302. Visoottiviseth, P., Francesconi, K., Sridokchan, W., 2002. The potential of Thai indigenous plant species for the phytoremediation of arsenic contaminated land. Environ. Pollut. 118, 453–461. Wang, J., Zhao, F.J., Meharg, A.A., Raab, A., Feldmann, J., McGrath, S.P., 2002. Mechanisms of arsenic hyperaccumulation in Pteris vittata. Uptake kinetics, interactions with phosphate, and arsenic speciation. Plant Physiol. 130, 1552–1561.

K.Z. Rahman et al. / Ecological Engineering 69 (2014) 93–105 Wells, J.M., Richardson, D.H.S., 1985. Anion accumulation by the moss Hylocomium splendens: uptake and competition studies involving arsenate, selenate, selenite, phosphate, sulphate and sulphite. New Phytol. 101, 571–583. Wiessner, A., Gonzalias, A.E., Kästner, M., Kuschk, P., 2008. Effects of sulphur cycle processes on ammonia removal in a laboratory-scale constructed wetland planted with Juncus effusus. Ecol. Eng. 34, 162–167. Wiessner, A., Kappelmeyer, U., Kaestner, M., Schultze-Nobre, L., Kuschk, P., 2013. Response of ammonium removal to growth and transpiration of Juncus effusus during the treatment of artificial sewage in laboratory-scale wetlands. Water Res. 47, 4265–4273. Wiessner, A., Kappelmeyer, U., Kuschk, P., Kästner, M., 2005a. Sulphate reduction and the removal of carbon and ammonia in a laboratory-scale constructed wetland. Water Res. 39 (19), 4643–4650. Wiessner, A., Kappelmeyer, U., Kuschk, P., Kästner, M., 2005b. Influence of the redox condition dynamics on the removal efficiency of a laboratory-scale constructed wetland. Water Res. 39 (1), 248–256. Willow, M.A., Cohen, R.R.H., 2003. PH, dissolved oxygen, and adsorption effects on metal removal in anaerobic bioreactors. J. Environ. Qual. 32, 1212–1221. Wright, D.J., Otte, M.L., 1999. Wetland plant effects on the biogeochemistry of metals beyond the rhizosphere. Biol. Environ. Proc. R. Ir. Acad. 99B, 3–10.

105

Xue, P.Y., Yan, C.Z., 2011. Arsenic accumulation and translocation in the submerged macrophyte Hydrilla verticillata (L.f.) Royle. Chemosphere 85, 1176–1181. Yang, J.X., Ma, Z.L., Ye, Z.H., Guo, X.Y., Qiu, R.L., 2010. Heavy metal (Pb, Zn) uptake and chemical changes in rhizosphere soils of four wetland plants with different radial oxygen loss. J. Environ. Sci. (China) 22, 696–702. Ye, Z.H., Lin, Z.Q., Whiting, S.N., de Souza, M.P., Terry, N., 2003. Possible use of constructed wetland to remove selenocyanate, arsenic, and boron from electric utility wastewater. Chemosphere 52, 1571–1579. Ye, Z.H., Whiting, S., Qian, J.H., Lytle, C.M., Lin, Z.Q., Terry, N., 2001. Trace element removal from coal pile leachate by an Alabama 10-year old constructed wetland. J. Environ. Qual. 30, 1710–1719. Zhao, F.J., Ma, J.F., Meharg, A.A., McGrath, S.P., 2009. Arsenic uptake and metabolism in plants. New Phytol. 181, 777–794. Zhao, F.J., McGrath, S.P., Meharg, A.A., 2010. Arsenic as a food chain contaminant: mechanisms of plant uptake and metabolism and mitigation strategies. Annu. Rev. Plant Biol. 61, 535–559. Zouboulis, A.I., Katsoyiannis, I.A., 2002. Removal of arsenates from contaminated water by coagulation-direct filtration. Sep. Sci. Technol. 37 (12), 2859–2873.