Removal of antibiotics from urban wastewater by constructed wetland optimization

Removal of antibiotics from urban wastewater by constructed wetland optimization

Chemosphere 83 (2011) 713–719 Contents lists available at ScienceDirect Chemosphere journal homepage: Technical...

356KB Sizes 0 Downloads 42 Views

Chemosphere 83 (2011) 713–719

Contents lists available at ScienceDirect

Chemosphere journal homepage:

Technical Note

Removal of antibiotics from urban wastewater by constructed wetland optimization María Hijosa-Valsero a,⇑, Guido Fink b, Michael P. Schlüsener b, Ricardo Sidrach-Cardona c, Javier Martín-Villacorta a, Thomas Ternes b, Eloy Bécares a a

Facultad de Ciencias Biológicas y Ambientales, Universidad de León, Campus de Vegazana s/n, 24071 León, Spain Bundesanstalt für Gewässerkunde, Am Mainzer Tor 1, Koblenz D-56068, Germany c Instituto de Medio Ambiente, Universidad de León, C/La Serna 58, 24007 León, Spain b

a r t i c l e

i n f o

Article history: Received 23 November 2010 Received in revised form 1 February 2011 Accepted 1 February 2011 Available online 26 February 2011 Keywords: Antibiotics Constructed wetlands Urban wastewater Removal WWTP

a b s t r a c t Seven mesocosm-scale constructed wetlands (CWs), differing in their design characteristics, were set up in the open air to assess their efficiency to remove antibiotics from urban raw wastewater. A conventional wastewater treatment plant (WWTP) was simultaneously monitored. The experiment took place in autumn. An analytical methodology including HPLC–MS/MS was developed to measure antibiotic concentrations in the soluble water fraction, in the suspended solids fraction and in the WWTP sludge. Considering the soluble water fraction, the only easily eliminated antibiotics in the WWTP were doxycycline (61 ± 38%) and sulfamethoxazole (60 ± 26%). All the studied types of CWs were efficient for the removal of sulfamethoxazole (59 ± 30–87 ± 41%), as found in the WWTP, and, in addition, they removed trimethoprim (65 ± 21–96 ± 29%). The elimination of other antibiotics in CWs was limited by the specific system-configuration: amoxicillin (45 ± 15%) was only eliminated by a free-water (FW) subsurface flow (SSF) CW planted with Typha angustifolia; doxycycline was removed in FW systems planted with T. angustifolia (65 ± 34–75 ± 40%), in a Phragmites australis-floating macrophytes system (62 ± 31%) and in conventional horizontal SSF-systems (71 ± 39%); clarithromycin was partially eliminated by an unplanted FW-SSF system (50 ± 18%); erythromycin could only be removed by a P. australis-horizontal SSF system (64 ± 30%); and ampicillin was eliminated by a T. angustifolia-floating macrophytes system (29 ± 4%). Lincomycin was not removed by any of the systems (WWTP or CWs). The presence or absence of plants, the vegetal species (T. angustifolia or P. australis), the flow type and the CW design characteristics regulated the specific removal mechanisms. Therefore, CWs are not an overall solution to remove antibiotics from urban wastewater during cold seasons. However, more studies are needed to assess their ability in warmer periods and to determine the behaviour of full-scale systems. Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction Pharmaceuticals and personal care products (PPCPs) are substances of widespread use, frequently detected in natural surface water and groundwater (Kolpin et al., 2002). These compounds, originating from wastewater treatment plant (WWTP) effluents, reach environmental water bodies (Ternes, 1998). Indeed, conventional WWTPs have been observed to be unable to totally remove PPCPs (Joss et al., 2006). Their continued mass use and their design as biologically active molecules mean that their incomplete elimination by conventional treatment systems is a concern of unknown

⇑ Corresponding author. Address: Facultad de Ciencias Biológicas y Ambientales, Departamento de Química y Física Aplicadas, Universidad de León, Campus de Vegazana s/n, 24071 León, Spain. Tel.: +34 987291000x5158; fax: +34 987291945. E-mail addresses: [email protected] (M. Hijosa-Valsero), fi[email protected] (G. Fink), [email protected] (M.P. Schlüsener), [email protected] (R. Sidrach-Cardona), [email protected] (J. Martín-Villacorta), [email protected] (T. Ternes), [email protected] (E. Bécares). 0045-6535/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2011.02.004

consequences (Daughton and Ternes, 1999). One of the most worrying groups of substances is that of antibiotics. Their extensive and sometimes inadequate use has caused the appearance of bacterial strains resistant to many antibiotics (multidrug resistance) (Taubes, 2008). Alonso et al. (2001) state that antibiotic resistance genes have an environmental origin, sometimes as an antibiotic protective mechanism and sometimes with a different function. In fact, resistant microorganisms have been found in different ecosystems (Schwartz et al., 2003; Allen et al., 2009). Constructed wetlands (CWs) are alternative wastewater treatment systems, consuming little energy and with relatively low maintenance costs. Due to the high surface/equivalent-inhabitant ratio required to achieve wastewater quality parameters, CWs are only feasible in small urban communities. Recently, the suitability of CWs for the removal of some PPCPs (Matamoros et al., 2005) has been assessed. Nevertheless, their efficiency to specifically remove antibiotics is still unknown. Conkle et al. (2008) studied the elimination of sulfapyridine and sulfamethoxazole in a CW and Park et al. (2009) analysed sulfamethoxazole removal.


M. Hijosa-Valsero et al. / Chemosphere 83 (2011) 713–719

However, the optimization of CW design configuration for the removal of antibiotics has not yet been carried out. In the present work the efficiency of several mesocosm-scale CWs in removing antibiotics from urban wastewater was assessed and compared to that of a conventional WWTP. This removal could be due to sorption, plant uptake and partial or complete degradation processes. These CWs differed in some design parameters, namely, the presence or absence of plants, their species (Typha angustifolia or Phragmites australis) and the flow configuration and the presence or absence of soil matrix (i.e., floating macrophytes surface flow – FM-SF, free-water surface flow – FW-SF, free-water subsurface flow – FW-SSF – or conventional horizontal subsurface flow – SSF). These systems had been operating for more than 2 years and they had been assessed for the removal of non-antibiotic PPCPs (Hijosa-Valsero et al., 2010). To the best of our knowledge, this is the first time that such a variety of mesocosms has been screened to determine the best design characteristics of CWs for antibiotic removal. The compounds studied belonged to several groups: tetracyclines (chlortetracycline and doxycycline), b-lactams (amoxicillin and ampicillin), macrolides (clarithromycin, erythromycin and its metabolite anhydroerythromycin, oleandomycin, roxithromycin and azithromycin), lincosamides (lincomycin) and sulfonamides (sulfadoxine, sulfamerazine, sulfamethoxazole and its metabolite N-acetyl-sulfamethoxazole, sulfadimethoxine, sulfisoxazole, sulfametazine and trimethoprim). Trimethoprim is not actually a sulfonamide but its use is often associated to that of sulfamethoxazole. The veterinary antibiotic tiamulin was also studied.

2. Material and methods 2.1. Description of the systems The experiment took place in November 2009. In May 2007 seven mesocosm-scale CWs were set up in the open air inside the

facilities of the León WWTP (NW Spain). All CWs consisted of a fibreglass container (80 cm wide, 130 cm long, 50 cm high) with a surface area of approximately 1 m2. The CWs differed from each other in their design parameters, which are summarized in Fig. 1. The aerial part of the plants had been harvested before the experimental period, as it was a common practice to remove the biomass after the summer season in these experimental systems. However, the living roots remained inside the beds. The theoretical hydraulic retention time (HRT) values of tanks CW1, CW2, CW3, CW4, CW5, CW6 and CW7 were, respectively, 2.1, 3.3, 5.1, 6.1, 2.9, 2.5 and 2.6 d. The León WWTP is a conventional plant described in the Supplementary Material (SM) section. Urban wastewater coming from the primary clarifier of the WWTP was transferred to a homogenisation tank of 0.5 m3. All the CWs were fed with this homogenised wastewater at a continuous flow rate of 50 L d 1 (input load 50 mm d 1). 2.2. Sampling regime In November 2009, 48-h composite water samples were collected every 2 d during a 10-d period in the influent and effluent of the seven CWs and in the WWTP discharge point. These samples were used for the analysis of antibiotics in the soluble aqueous fraction (n = 5) and in the insoluble aqueous fraction (n = 1), i.e. substances adsorbed onto suspended solids (SS). In addition, conventional wastewater quality parameters (COD, BOD and TSS) were measured to characterise the wastewater (n = 3). Sludge grab samples were taken every 2 d (n = 5) from the biological reactor of the WWTP. All samples were collected in 1-L amber glass bottles, which were transported refrigerated (4 °C) to the laboratory, where they were processed within 24 h. During the sampling campaign the average temperature was 9.4 °C, with a minimum of 3 °C and a maximum of 14 °C. The accumulated rainfall during this 10-d period was 65 mm (data taken

Fig. 1. Schematic design characteristics of the CWs. Systems CW1 and CW5 had a water depth of 30 cm and plant growth was supported by 20 cm long and 10 cm diameter garden-net cylinders (4 cm pore size). Systems CW2, CW3 and CW4 had a 25 cm layer of free-water (FW) over a 25 cm layer of siliceous gravel (d10 = 4 mm). Systems CW6 and CW7 consisted of a 45 cm siliceous gravel (d10 = 4 mm) layer, through which a 40 cm water layer flowed. Vegetation coverage was 100% in the planted systems (HijosaValsero et al., 2010).

M. Hijosa-Valsero et al. / Chemosphere 83 (2011) 713–719

from the nearest weather station: La Virgen del Camino, León, Spain).

2.3. Material All solvents were HPLC-grade. Citric acid monohydrate C6H8O7H2O, mono-sodium hydrogen phosphate monohydrate NaH2PO4H2O and NaOH were analytical grade reagents. Antibiotic standards were purchased from Sigma–Aldrich (Seelze, Germany). Chlortetracycline has an epimeric form (4-epichlortetracycline). Both of them are present in environmental samples and were considered in the analytical method and in the quantification process. The internal standards (IS) for the analytical method demeclocycline (IS for tetracyclines), penicillin G (IS for penicillins), sulfamerazine-d4 (IS for sulfonamides), sulfamethoxazole-d4 (IS for sulfamethoxazol) and N-acetyl-sulfamethoxazole-d5 (IS for Nacetyl-sulfamethoxazol) were bought from Sigma–Aldrich and (E)-9-[O-(2-methyloxime)]-erythromycin (IS for macrolides and lincomycin) was synthesized following Schlüsener et al. (2003). Samples were filtered with 4.5 cm-diameter, 1.2 lm-pore size glass–fibre filters (Millipore, Billerica, MA, USA). Solid phase extraction (SPE) was performed with Strata SAX (55 lm, 70 Å, 500 mg, 6 mL) cartridges, Strata-X Polymeric Reversed Phase (33 lm, 500 mg, 6 mL) cartridges (Phenomenex, Torrance, CA, USA) and Oasis HLB (200 mg, 6 mL) cartridges (Waters, Milford, MA, USA). Diatomaceous earth was provided by Biotage (Isolute HM-N, Düsseldorf, Germany). To avoid the formation of complexes between tetracyclines and the organic matter present in the sample, a buffer solution was prepared containing NaH2PO4H2O 0.05 M and citric acid 0.05 M in water (Zhu et al., 2001). For the pressurized solvent extraction (PSE) an ASE 200 Dionex (Sunnyvale, CA, USA) was used. The PSE-solution consisted of methanol: citric acid 0.2 M (pH 4.7 adjusted with NaOH), 1:1 (v/ v), as described by Jacobsen et al. (2004).

3. Analytical methodology The presence of tetracyclines in urban raw and treated wastewater (soluble and SS fractions) and in WWTP sludge was determined. In addition, the presence of b-lactam antibiotics, macrolides, sulfonamides, lincomycin and tiamulin in the soluble fraction of urban raw and treated wastewater was assessed. Molecular structures and physico-chemical properties of the studied compounds and their ISs can be found in Table SM-1.

3.1. Sample preparation 3.1.1. Water samples (soluble aqueous fraction) A volume of 100 mL-influent water (primary effluent of the WWTP) and 200 mL-effluent water (from the mesocosm-scale CWs and the discharge point of the WWTP) was filtered through glass–fibre filters. Then, for the analysis of tetracyclines and b-lactams, each sample was spiked with 100 ng demeclocycline and 100 ng penicillin G (both IS). To avoid the interference of complexes formation, 7 g NaH2PO4H2O and 10.6 g C6H8O7H2O were added per litre of sample (this constitutes a buffering solution similar to that described by Zhu et al. (2001)). For the analysis of macrolides, sulfonamides, lincomycin and tiamulin, each sample was spiked after the filtration with 100 ng (E)-9-[O-(2-methyloxime)]erythromycin, 100 ng sulfamerazine-d4, 100 ng sulfamethoxazole-d4 and 100 ng N-acetyl-sulfamethoxazole-d5 (four IS) and no salts were added.


3.1.2. SS samples A variable volume of water (100–500 mL) was filtered through a glass–fibre filter until the filter got saturated. This volume depends on the nature of the sample. The filter was then spiked with 100 ng demeclocycline (IS). 3.1.3. WWTP-sludge samples A 1 L sample was decanted for 30 min and the supernatant was discarded. The precipitated sludge was lyophilized until a final weight of approximately 3 g. After that, only 500 mg of dry sludge were taken, which were spiked with 100 ng demeclocycline (IS). 3.1.4. PSE of SS and sludge samples In the case of SS samples, every filter was folded and rolled inside a cell of the PSE system. In the case of sludge samples, 500 mg sludge were placed inside the cell. In both cases, diatomaceous earth was used as filling material. The PSE programme consisted of a first extraction with 30 mL of PSE-solution for 10 min at 140 hPa and a second extraction of 30 mL of PSE-solution for 10 min at 140 hPa (adapted from Jacobsen et al., 2004). The recovery of every studied compound in the PSE step is shown in Table SM-2. The extracted liquid (about 40 mL) was diluted adding 360 mL of buffering solution of NaH2PO4H2O 0.05 M and citric acid 0.05 M (Zhu et al., 2001). This dilution had the double effect of complicating complex formation and of diminishing the methanol proportion in the sample, which is necessary for the subsequent SPE process. 3.2. SPE and sample concentration Once the previous steps were concluded, all samples underwent a SPE process. The tetracycline and b-lactam analysis of aqueous samples was made with Strata-X cartridges, whereas a Strata SAX and a Strata-X cartridge were connected in series for the tetracyclines analysis in SS and sludge samples. In both cases, the cartridges were previously conditioned with 1  2 mL n-heptane, 1  2 mL acetone, 3  2 mL methanol and 4  2 mL buffer solution (NaH2PO4H2O 0.05 M and citric acid 0.05 M). The analysis of macrolides, sulfonamides, lincomycin and tiamulin in water samples was performed with Oasis HLB cartridges previously conditioned with 1  2 mL n-heptane, 1  2 mL acetone, 3  2 mL methanol and 4  2 mL groundwater. The sample flow rate through the cartridges was adjusted to approximately 5 mL min 1. Strata-X and Oasis HLB cartridges were then dried for 60 min using gaseous nitrogen and eluted with 2  1 mL methanol and 4  2 mL methanol, respectively. The extract was evaporated until 100 lL under a gentle nitrogen stream. Then the vial was reconstituted to 1000 lL with Milli-Q water. Recoveries of this SPE step are shown in Table SM-3. 3.3. Chromatographic analysis The HPLC system consisted of a G1313A autosampler, a G1311A quaternary HPLC pump, a G1379A degasser (all Agilent, Waldbronn, Germany), a CTO-10A column oven and a SCL-10A system controller (all Shimadzu, Duisburg, Germany). The detection was performed on a API 4000 mass spectrometer (Appliedbiosystems, Foster City, CA, USA). Parameters such as declustering potential (DP), collision energy (CE) and cell exit potential were optimised in the auto-tuning programme of the Analyst 1.4.2 software. For all compounds two multiple reaction monitoring (MRM) transitions were considered for identification and quantification of the analytes. The limit of quantification (LOQ) was defined as a signal-to-noise ratio of 10:1. These ratios were taken from the chromatograms of the calibration.


M. Hijosa-Valsero et al. / Chemosphere 83 (2011) 713–719

3.3.1. Tetracyclines and b-lactam antibiotics HPLC–MS/MS-positive electrospray ionization (ESI+) method A Luna C18(2) 150  2 mm, 3 lm column (Phenomenex, Torrance, CA, USA) was used with a mobile phase consisting of (A) ammonium acetate 10 mM and formic acid 0.1% in water and (B) formic acid 0.1% in methanol. The flow was established at 0.2 mL min 1. The mobile phase gradient is shown in Table SM-4. The detector worked with the ESI(+) ionization source and data acquisition was performed in MRM mode. MS parameters for the analysis, transitions m/z, CE, DP, retention time and LOQ of every studied compound are shown in Table SM-5. 3.3.2. Macrolides, lincomycin, sulfonamides and tiamulin HPLC–MS/ MS-atmospheric pressure chemical ionization (APCI) method A Luna Phenyl–Hexyl 150  2 mm, 3 lm column (Phenomenex, Torrance, CA, USA) was used with a mobile phase consisting of (A) ammonium acetate 10 mM in water and (B) methanol. The flow was established at 0.2 mL min 1. The mobile phase gradient is shown in Table SM-6. The detector worked under the APCI ionization mode and data acquisition was performed in MRM mode. MS parameters for the analysis, transitions m/z, CE, DP, retention time and LOQ of every studied compound are shown in Table SM-7.

observed: doxycycline (0.3 ± 0.2 lg g 1) was the most abundant substance, followed by chlortetracycline (13 ± 3 ng g 1). The consumption of antibiotics in Spain is relatively high. In 2006, the total amount of some prescribed pharmaceuticals was 117.5 t amoxicillin combined with b-lactamase inhibitor, 66.9 t amoxicillin, 14.7 t clarithromycin, 7.8 t sulfamethoxazole combined with trimethoprim, 3.9 t azithromycin, 2.9 t lincosamides, 1.6 t erythromycin or 0.8 t doxycycline (INSALUD, 2002; AEMPSDGFPS, 2006; INE, 2006). Antibiotic concentrations in wastewaters around the world are very variable and they depend on the regional prescription model, the sewage dilution and the presence of effluent discharges from pharmaceutical industries and cattle farms. A review carried out by Miège et al. (2009) and focusing on the presence and fate of PPCPs in WWTPs, found average influent concentrations of 650 ng L 1 clarithromycin, 110 ng L 1 erythromycin and 340 ng L 1 sulfamethoxazole, which are similar to the data observed in the soluble aqueous fraction in the present study (Table 2). Amoxicillin concentration in the WWTP influent was high and it should not be forgotten that it is one of the most prescribed antibiotics in Spain.

4.2. Removal efficiencies 4. Results and discussion 4.1. Characteristics of the influent wastewater and the WWTP sludge All the systems were fed with urban wastewater coming from the primary clarifier of the WWTP. Conventional wastewater control parameters (COD, BOD and TSS) of this influent water and the other sampling points are shown in Table 1. Table 2 summarizes antibiotic concentrations in the soluble fraction of water. Chlortetracycline, oleandomycin, roxithromycin, anhydroerythromycin, sulfadoxine, sulfamerazine, sulfisoxazole, sulfametazine and tiamulin were not detected in any of these samples. Probably, the veterinary antibiotic tiamulin was not found because the wastewater had an urban origin. Azithromycin appeared in all samples but, due to a problem with the chromatographic method, its concentration could not be quantified. Sulfadimethoxine was only detected in one of the five samples taken. The pharmeceuticals with the highest influent concentrations in the aqueous fraction were amoxicillin, ampicillin and sulfamethoxazole and its metabolite N-acetyl-sulfamethoxazole (Table 2). Nevertheless, other substances showed very low concentrations, like lincomycin (Table 2). On the other hand, only antibiotics belonging to the group of tetracyclines were analysed in the SS. In this case, all the studied compounds appeared in all the samples (Table 2). Doxycycline was the most abundant compound of this group in the influent. Similarly, as far as the WWTP sludge is concerned, only the tetracyclines were monitored. A parallel pattern to that of the SS was

Table 1 Mean concentrations and 0.95 confidence intervals for COD, BOD and TSS (n = 3) during the experimental period in the influent, the WWTP effluent and the effluents of all CWs. COD (mg L Influent WWTP effluent Typha-FM-SF (CW 1) Typha-FW-SF (CW 2) Typha-FW-SSF (CW 3) Unplanted-FW-SSF (CW 4) Phragmites-FM-SF (CW 5) Phragmites-SSF (CW 6) Unplanted-SSF (CW 7)

201 ± 56 14 ± 12 24 ± 4 29 ± 6 17 ± 7 50 ± 18 44 ± 11 48 ± 2 16 ± 12



BOD (mg L 99 ± 18 6±4 12 ± 9 21 ± 15 7±4 22 ± 2 34 ± 16 41 ± 6 12 ± 13



TSS (mg L 71 ± 12 14 ± 12 9±5 18 ± 7 22 ± 3 28 ± 12 12 ± 3 27 ± 14 9±3



The antibiotic removal efficiency of the seven CWs and the WWTP was assessed considering influent and effluent loads (see SM) and assuming that evapotranspiration water losses were negligible, since it was autumn and the plants had been harvested. Removal efficiencies in the soluble fraction (Table 2) are described below related to every antibiotic group. Sulfamethoxazole removal efficiencies were calculated considering the sum of the concentrations of this compound and its metabolite (N-acetyl-sulfamethoxazole). The removal efficiencies in the SS fraction were not calculated, since the results would not be reliable, because of the low number of samples (n = 1). Tetracyclines: Doxycycline removal efficiencies (Table 2) were higher in SSF systems (CW6, CW7), the Typha-FW-SSF (CW3), the Typha-FW-SF (CW2), the Phragmites-FM-SF (CW5) and the WWTP (between 61 ± 38 and 79 ± 72%). b-Lactams: Amoxicillin and especially ampicillin showed extremely low removal efficiencies (Table 2). Only the Typha-FW-SSF (CW3) removed a small portion of amoxicillin (45 ± 15%) and the Typha-FM-SF (CW1) scarcely eliminated ampicillin (29 ± 4%). Molecules with a b-lactam ring are relatively stable at neutral pH-values, which are typical in the León WWTP influent (7.0–7.5), whereas acidic and mainly basic pH values contribute to their fast degradation. The low removal efficiencies observed in the present experiment and the neutral pH values recorded minimize the possibility of physico-chemical degradation. In spite of the presence of many bacterial strains endowed with b-lactamase enzymes in WWTPs (Szczepanowski et al., 2009) as well as in the environment (Schwartz et al., 2003; Allen et al., 2009), biodegradation was not an especially active process in the studied systems for the removal of penicillins. Macrolides and lincomycin: Clarithromycin had very low removal efficiencies (Table 2): the best system (50 ± 18%) was the unplanted FW-SSF (CW4). Erythromycin was only eliminated by the Phragmites-SSF, CW6 (64 ± 30%). Lincomycin was not removed by any treatment system. This is in agreement with the low removal efficiencies reported by other authors for lincomycin in WWTPs (Karthikeyan and Meyer, 2006: Watkinson et al., 2007). A curious fact was that of finding lower lincomycin concentrations in the influent than in the CW-effluents (Table 2); a matrix effect, desorption of previously accumulated matter in the CW or the breaking-down of lincomycin conjugates might cause this anomaly.

Table 2 (a) Water samples (soluble fraction). Mean concentrations (ng L 1) and 0.95 confidence intervals of the studied antibiotics in the influent, the WWTP effluent and the effluents of all CWs (n = 5). Mean removal efficiencies (%) and absolute errors are given between brackets for every treatment system. (b) Suspended solids. Tetracyclines concentrations (ng L 1 of filtered water). These data should be considered carefully, since n = 1. For chlortetracycline both epimers were considered. Influent

WWTP effluent

Typha-FW-SF (CW 2)

Typha-FW-SSF (CW 3)

Unplanted-FW-SSF (CW 4)

Phragmites-FM-SF (CW 5)

Phragmites-SSF (CW 6)

Unplanted-SSF (CW 7)

(a) Water samples (soluble fraction) Tetracyclines Doxycycline

180 ± 83

70 ± 30 (61 ± 38)

b-Lactams Amoxicillin

46 000 ± 14 000

30 000 ± 12 000 (35 ± 17)


690 ± 81

640 ± 170 (7 ± 2)

490 ± 46 (29 ± 4)

470 ± 160 (32 ± 11)

840 ± 120 (n.r.)

Macrolides Clarithromycin

250 ± 84

200 ± 60 (18 ± 9)

190 ± 10 (22 ± 8)

170 ± 31 (32 ± 12)

150 ± 16 (39 ± 14)


57 ± 26

Lincosamides Lincomycin


5 ± 1 (24 ± 9)

260 ± 110

160 ± 46 (60 ± 26)

Sulfonamides Sulfamethoxazole* N-acetylsulfamethoxazole* Sulfadimethoxine**


(b) Suspended solids Chlortetracycline Doxycycline

1500 (53 ± 30)

28 000 ± 3700 (39 ± 13)

25 000 ± 2500 (45 ± 15)

96 ± 35 (47 ± 27) 43 000 ± 6400 (7 ± 2)

69 ± 14 (62 ± 31) 33 000 ± 9000 (27 ± 12)

1000 ± 260 (n.r.) 120 ± 13 (50 ± 18)

220 ± 37 (11 ± 5)

170 ± 22 (31 ± 12)

170 ± 23 (32 ± 12)

51 ± 10 (n.r.)

23 ± 4 (n.r.)

88 ± 15 (n.r.)

98 ± 16 (n.r.)

27 ± 3 (n.r.)

110 ± 37 (80 ± 42) 12 ± 14

37 (99 ± 2)

6 (99 ± 2)

13 ± 1 (88 ± 23)

36 ± 7 (65 ± 21) <10 12

102 ± 95 (78 ± 34)

49 (98 ± 2)

9 ± 2 (92 ± 34) 44 18

250 (92 ± 6)

75 ± 49 (87 ± 41)

76 ± 13 (n.r.) 140 ± 54 (73 ± 35)

8±1 7 (99 ± 2)

19 ± 3 (82 ± 25) 10 10

100 ± 26 (n.r.)

250 ± 45 (n.r.)

23 ± 10

4 ± 1 (96 ± 29) 28 16

21 ± 3 (64 ± 30)

227 ± 79 (59 ± 30)

30 ± 7

43 000 ± 8600 (6 ± 2) 810 ± 230 (n.r.)

95 ± 40 (n.r.)

43 ± 17

36 000 ± 11 000 (21 ± 9)

46 ± 39 (79 ± 72)

750 ± 200 (n.r.)

86 ± 18 (n.r.)

170 ± 53 (65 ± 32)

53 ± 16 (71 ± 39)

700 ± 160 (n.r.)

83 ± 15 (n.r.)

520 (67 ± 16)

<10 <10

44 ± 12 (75 ± 40)

81 ± 28 (n.r.)

27 ± 17

89 ± 20 (14 ± 5) <10 10

63 ± 15 (65 ± 34)

79 ± 12 (n.r.)

160 ± 59 (71 ± 34)

92 ± 19

100 ± 26 18 47

27 000 ± 4300 (42 ± 14)

65 ± 28 (n.r.)

360 ± 110


67 ± 48 (63 ± 53)

25 ± 12 32 (99 ± 2)

1.3 ± 1.1 (99 ± 85) 23 13

5 ± 1 (95 ± 27) <10 <10

M. Hijosa-Valsero et al. / Chemosphere 83 (2011) 713–719

Typha-FM-SF (CW 1)

Notes: (n.r.): Not removed. * The removal of sulfamethoxazole was calculated considering sulfamethoxazole and N-acetyl-sulfamethoxazole (its metabolite) concentrations. **

Sulfadimethoxine was only detected in one out of five samples.



M. Hijosa-Valsero et al. / Chemosphere 83 (2011) 713–719

Sulfonamides: Sulfamethoxazole was relatively easily removed by CWs (59 ± 30–87 ± 41%) and by the WWTP (60 ± 26%), with the planted SSF system, CW6 (87 ± 41%), and the FW-SSF systems, CW3 and CW4 (about 80%), showing the highest values (Table 2). Sulfadimethoxine seemed to reach high removal efficiencies in all the CWs (67 ± 16–99 ± 2%), but lower values in the WWTP (53 ± 30%); however sulfadimethoxine data should be considered carefully, since this compound only appeared in one out of five samples taken. Trimethoprim could not be removed by the WWTP but was eliminated by all the CWs (65 ± 21–96 ± 29%). In general, it was observed that the WWTP was only able to easily remove doxycycline and sulfamethoxazole from the soluble fraction of water. The removal ability of CWs depended on their configuration and on the compound. This fact had been previously reported for other PPCPs, like ibuprofen, carbamazepine, salicylic acid, galaxolide, tonalide and caffeine, in the same mesocosm-systems (Hijosa-Valsero et al., 2010). All CWs were capable of significantly removing sulfamethoxazole, like the WWTP (Table 2). In addition, all the studied CWs contributed to the elimination of trimethoprim. The removal of all the studied substances in CWs could be attributed either to degradation (physico-chemical or biological), sorption or plant uptake. The adsorption of a compound onto solids depends on the chemical nature of the substance and it is related to Kow and Koc values (Table SM-1); tetracyclines are believed to adsorb onto soils and sludge (Martínez-Carballo et al., 2007) depending on pH, ionic strength and presence of cations in the medium (Andreu et al., 2009; Liu et al., 2009). Conkle et al. (2010) observed that sorption was an important removal pathway for fluoroquinolone antibiotics in a wetland soil. Although plant uptake is a possible removal way (Stottmeister et al., 2003; Imfeld et al., 2009), especially for those substances with log Kow values of 1–3.5 (Dietz and Schnoor, 2001), it is not very probable in our study, since it was carried out during autumn (when plants are not fully active) and after harvesting. Macrophytes can contribute directly (uptake, adsorption, release of exudates, etc.) or indirectly (oxygen pumping towards the rhizosphere, biofilm growth around roots, etc.) to pollutant removal in CWs. It is necessary to point out that seasonal changes have been observed in the León WWTP and in the studied mesocosm-CWs for the removal of naproxen, ibuprofen, caffeine and methyl dihydrojasmonate, finding higher efficiencies in summer (Hijosa-Valsero et al., 2010). Since the present experiment was carried out during a cold season, better removal efficiencies would be expected in the studied systems in summer, at least for some antibiotics, because biodegradation processes in CWs are enhanced at warm temperatures (Truu et al., 2009). Park et al. (2009) studied the removal of sulfamethoxazole in a full-scale SF-CW, and found removal efficiencies of 30% in spring and 50% in summer. Given the relatively low antibiotic removal in the studied WWTP, important discharges of these compounds into the environment are expected. The released loads of pharmaceuticals were calculated taking into account effluent concentrations and the discharge flow of the WWTP (123,000 m3 d 1). These data are shown in Table 3. 4.2.1. Influence of CW design characteristics In order to study the effect of CW design configuration on the removal of antibiotics from the soluble fraction, statistical comparisons (Mann–Whitney U tests) between effluent concentrations of CWs which only differed in one design parameter were made with the software Statistica 7 (StatSoft, Tulsa, OK, USA). Differences were considered significant when p < 0.05. The influence of the plant species chosen was assessed by comparing the Typha-FMSF (CW1) and the Phragmites-FM-SF (CW5); finding out that the Typha-system obtained significantly lower effluent concentrations

Table 3 Calculated antibiotic loads released by the León WWTP (measured at its discharge point). Mean values and 0.95 confidence intervals are shown. Load (g d



Tetracyclines Doxycycline


b-Lactams Amoxicillin Ampicillin

3700 ± 1500 79 ± 21

Macrolides Clarithromycin Erythromycin

25 ± 7 8±3

Lincosamides Lincomycin

0.6 ± 0.2

Sulfonamides Sulfamethoxazole N-acetyl-sulfamethoxazole Trimethoprim

19 ± 6 11 ± 2 27 ± 32

of ampicillin than the Phragmites-system (Table 2). The effect of plant presence was studied by comparing the Typha-FW-SSF (CW3) to its unplanted equivalent (CW4), and the Phragmites-SSF (CW6) to its unplanted equivalent (CW7), respectively. In the case of FW-SSF systems, the unplanted system (CW4), more directly insolated and highly populated by microscopic Chlorophyta algae, coped significantly better with the removal of clarithromycin and trimethoprim (Table 2), whereas the planted system (CW3) was more successful for the removal of amoxicillin. On the other hand, the planted SSF-system (CW6) was more efficient than its unplanted equivalent (CW7), showing significantly lower effluent concentrations of erythromycin and trimethoprim (Table 2). All these results indicate that the presence of plants could play an important role in the removal of some antibiotics and it should be taken into account that this study was carried out in autumn and after harvesting the plants, so greater vegetal influences are to be expected in summer (when plants are more active). When comparing a soilless Typha-FM-SF system (CW1) to a gravel-bed Typha-FW-SF system (CW2), the soilless system obtained significantly lower trimethoprim effluent concentrations. However, when the soilless Phragmites-FM-SF system (CW5) was compared to the Phragmites-SSF system (CW6), the opposite was observed and the gravel-bed system (CW6) was significantly more efficient for the removal of erythromycin, sulfamethoxazole and trimethoprim. The influence of the flow type was assessed by comparing the Typha-FW-SF (CW2) and the Typha-FW-SSF (CW3), finding better removal abilities in the Typha-FW-SSF system, which obtained significantly lower effluent concentrations of trimethoprim (Table 2). Other flow configurations were contrasted by comparing the unplanted FW-SSF (CW4) to the unplanted SSF (CW7); the FWSSF configuration was significantly more appropriate for the removal of clarithromycin than the SSF system (Table 2). These results illustrate the implication of flow type and gravel configuration in the removal of antibiotics in CWs. Based on the trends reported by these field data, some dominant removal mechanisms can be indicated. Tetracyclines (chlortetracycline and doxycycline) are probably eliminated by means of adsorption/retention processes. However, their specific physico-chemical properties (Table SM-1) would cause slightly different behaviours inside CWs. Root-related biofilm, plant exudates or micro-environment modifications near plant tissues could also play a role in their removal. Trimethoprim and, to a lesser extent, sulfamethoxazole, are easily eliminated in the CWs and not in the WWTP. These compounds are probably degraded by microorganisms. Erythromycin and clarithromycin removal was very rare in general. Both substances have relatively high log Kow and log Koc values (Table SM-1). Erythromycin was only removed by a

M. Hijosa-Valsero et al. / Chemosphere 83 (2011) 713–719

planted SSF system (its elimination is presumably favoured by the presence of plants). Clarithromycin removal was only achieved in an insolated system with a long HRT, so photodegradation, algal interactions, sorption or a slow biodegradation could be its main elimination pathways. Amoxicillin and ampicillin removal in CWs was extremely low and somehow enhanced by plant presence. Acknowledgments This study was funded by the Spanish Ministry of Science through the Projects CTM2005-06457-C05-03 and CTM200806676-C05-03/TECNO and by the Junta de Castilla y León through the project LE037A10-2. MH-V kindly thanks a FPU fellowship from the Spanish Ministry of Education. The authors thank Juan Carlos Sánchez Sánchez for the maintenance of the systems. We thank Acciona Agua and Mancomunidad de Saneamiento de León y su Alfoz for their technical support. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.chemosphere.2011.02.004. References AEMPS-DGFPS, 2006. Uso De Antibióticos En España Con Cargo Al S.N.S. Expresado En Dosis Diarias Definidas Por 1.000 Habitantes Y Día Durante El Año 2006 (Use Of Antibiotics In Spain Paid By The National Health Service Expressed In Daily Defined Doses Per 1000 Inhabitants And Day During 2006) (accessed 09.09). Allen, H.K., Moe, L.A., Rodbumrer, J., Gaarder, A., Handelsman, J., 2009. Functional metagenomics reveals diverse b-lactamases in a remote Alaskan soil. ISME J. 3, 243–251. Alonso, A., Sánchez, P., Martínez, J.L., 2001. Environmental selection of antibiotic resistance genes. Environ. Microbiol. 3, 1–9. Andreu, V., Vazquez-Roig, P., Blasco, C., Picó, Y., 2009. Determination of tetracycline residues in soil by pressurized liquid extraction and liquid chromatography tandem mass spectrometry. Anal. Bioanal. Chem. 394, 1329–1339. Conkle, J.L., White, J.R., Metcalfe, C.D., 2008. Reduction of pharmaceutically active compounds by a lagoon wetland wastewater treatment system in Southeast Louisiana. Chemosphere 73, 1741–1748. Conkle, J.L., Lattao, C., White, J.R., Cook, R., 2010. Competitive sorption and desorption behavior for three fluoroquinolone antibiotics in a wastewater treatment wetland soil. Chemosphere 80, 1353–1359. Daughton, C.G., Ternes, T.A., 1999. Pharmaceuticals and personal care products in the environment: agents of subtle change? Environ. Health Persp. 107, 907– 938. Dietz, A.C., Schnoor, J.L., 2001. Advances in Phytoremediation. Environ. Health Persp. 99, 163–168. Hijosa-Valsero, M., Matamoros, V., Sidrach-Cardona, R., Martín-Villacorta, J., Bécares, E., Bayona, J.M., 2010. Comprehensive assessment of the design configuration of constructed wetlands for the removal of pharmaceuticals and personal care products from urban wastewaters. Water Res. 44, 3669–3678. Imfeld, G., Braeckevelt, M., Kuschk, P., Richnow, H.H., 2009. Review: monitoring and assessing processes of organic chemicals removal in constructed wetlands. Chemosphere 74, 349–362.


INE (Spanish Statistics National Institute), 2006. (accessed 09.09). INSALUD. Sistema De Codificación De Principios Activos Y Dosis Diarias Definidas Del INSALUD (Code System Of Active Principles And Daily Defined Doses Of The Spanish Health National Institute). 2nd Ed. Madrid, 2002. (accessed 09.09). Jacobsen, A.M., Halling-Sørensen, B., Ingerslev, F., Hansen, S.H., 2004. Simultaneous extraction of tetracycline, macrolide and sulfonamide antibiotics from agricultural soils using pressurised liquid extraction, followed by solid phaseextraction and liquid chromatography–tandem mass spectrometry. J. Chromatogr. A 1038, 157–170. Joss, A., Zabczynski, S., Göbel, A., Hoffmann, B., Löffler, D., McArdell, C.S., Ternes, T.A., Thomsen, A., Siegrist, H., 2006. Biological degradation of pharmaceuticals in municipal wastewater treatment: proposing a classification scheme. Water Res. 40, 1686–1696. Karthikeyan, K.G., Meyer, M.T., 2006. Occurrence of antibiotics in wastewater treatment facilities in Wisconsin, USA. Sci. Total Environ. 361, 196–207. Kolpin, D.W., Furlong, E.T., Meyer, M.T., Thurman, E.M., Zaugg, S.D., Barber, L.B., Buxton, H.T., 2002. Pharmaceuticals, hormones and other organic wastewater contaminants in US streams, 1999–2000: a national reconnaissance. Environ. Sci. Technol. 36, 1202–1211. Liu, H., Zhang, G., Liu, C.Q., Li, L., Xiang, M., 2009. The occurrence of chloramphenicol and tetracyclines in municipal sewage and the Nanming River, Guiyang City. China. J. Environ. Monitor. 11, 1199–1205. Martínez-Carballo, E., González-Barreiro, C., Scharf, S., Gans, O., 2007. Environmental monitoring study of selected veterinary antibiotics in animal manure and soils in Austria. Environ. Pollut. 148, 570–579. Matamoros, V., García, J., Bayona, J.M., 2005. Behavior of selected pharmaceuticals in subsurface flow constructed wetlands: a pilot-scale study. Environ. Sci. Technol. 39, 5449–5454. Miège, C., Choubert, J.M., Ribeiro, L., Eusèbe, M., Coquery, M., 2009. Fate of pharmaceuticals and personal care products in wastewater treatment plants–conception of a database and first results. Environ. Pollut. 157, 1721–1726. Park, N., Vanderford, B.J., Snyder, S.A., Sarp, S., Kim, S.D., Cho, J., 2009. Effective controls of micropollutants included in wastewater effluent using constructed wetlands under anoxic condition. Ecol. Eng. 35, 418–423. Schlüsener, M.P., Bester, K., Spiteller, M., 2003. Determination of antibiotics such as macrolides, ionophores and tiamulin in liquid manure by HPLC–MS/MS. Anal. Bioanal. Chem. 375, 942–947. Schwartz, T., Kohnen, W., Jansen, B., Obst, U., 2003. Detection of antibiotic-resistant bacteria and their resistance genes in wastewater, surface water, and drinking water biofilms. FEMS Microbiol. Ecol. 43, 325–335. Stottmeister, U., Wießner, A., Kuschk, P., Kappelmeyer, U., Kästner, M., Bederski, O., Müller, R.A., Moormann, H., 2003. Effects of plants and microorganisms in constructed wetlands for wastewater treatment. Biotechnol. Adv. 22, 93–117. Szczepanowski, R., Linke, B., Krahn, I., Gartemann, K.H., Gützkow, T., Eichler, W., Pühler, A., Schlüter, A., 2009. Detection of 140 clinically relevant antibiotic resistance genes in the plasmid metagenome of wastewater treatment plant bacteria showing reduced susceptibility to selected antibiotics. Microbiology 155, 2306–2319. Taubes, G., 2008. The bacteria fight back. Science 321, 356–361. Ternes, T.A., 1998. Occurrence of drugs in German sewage treatment plants and rivers. Water Res. 32, 3245–3260. Truu, M., Juhanson, J., Truu, J., 2009. Microbial biomass, activity and community composition in constructed wetlands. Sci. Total Environ. 407, 3958–3971. Watkinson, A.J., Murby, E.J., Costanzo, S.D., 2007. Removal of antibiotics in conventional and advanced wastewater treatment: implications for environmental discharge and wastewater recycling. Water Res. 41, 4164–4176. Zhu, J., Snow, D.D., Cassada, D.A., Monson, S.J., Spalding, R.F., 2001. Analysis of oxytetracycline, tetracycline, and chlortetracycline in water using solid-phase extraction and liquid chromatography–tandem mass spectrometry. J. Chromatogr. A 928, 177–186.