Bioresource Technology 186 (2015) 207–214

Contents lists available at ScienceDirect

Bioresource Technology

journal homepage: www.elsevier .com/locate /bior tech

Combined effect of erythromycin, tetracycline and sulfamethoxazoleon performance of anaerobic sequencing batch reactors

http://dx.doi.org/10.1016/j.biortech.2015.03.0430960-8524/� 2015 Elsevier Ltd. All rights reserved.

⇑ Corresponding author at: Istanbul Technical University, Civil EngineeringFaculty, Environmental Engineering Department, 34469 Maslak, Istanbul, Turkey.Tel.: +90 5057927480.

E-mail addresses: seaydin@itu.edu.tr, sevcan_aydn@hotmail.com (S. Aydin).

Sevcan Aydin a,⇑, Bahar Ince b, Zeynep Cetecioglu a, Osman Arikan a, E. Gozde Ozbayram a,Aiyoub Shahi a, Orhan Ince a

a Environmental Engineering Department, Istanbul Technical University, Maslak, Istanbul, Turkeyb Institutes of Environmental Sciences, Bogazici University, Bebek, Istanbul, Turkey

h i g h l i g h t s

� Combined effects of antibiotics onanaerobic SBRs was examined.� Anaerobic pre-treatment of antibiotic

combinations can be a suitablealternative.� Erythromycin, tetracycline and

sulfamethoxazole were partiallybiodegraded.� Dual effects of antibiotics are more

toxic than triple effects in theanaerobic SBRs.

g r a p h i c a l a b s t r a c t

a r t i c l e i n f o

Article history:Received 6 January 2015Received in revised form 3 March 2015Accepted 7 March 2015Available online 14 March 2015

Keywords:Anaerobic treatmentCombined effectSulfamethoxazoleErythromycinTetracycline

a b s t r a c t

The combined effects of erythromycin–tetracycline–sulfamethoxazole (ETS) and sulfamethoxazole–te-tracycline (ST) antibiotics on the performance of anaerobic sequencing batch reactors were studied. Acontrol reactor was fed with wastewater that was free of antibiotics, while two additional reactors werefed with ETS and ST. The way in which the ETS and ST mixtures impact COD removal, VFA production,antibiotic degradation, biogas production and composition were investigated. The effects of the ETSmixtures were different from the ST mixtures, erythromycin can have an antagonistic effect on sul-famethoxazole and tetracycline. The anaerobic pre-treatment of these antibiotics can represent a suitablealternative to the use of chemical treatments for concentrations at 10 mg/L of S and 1 mg/L of T; 2 mg/L ofE, 2 mg/L of T and 20 mg/L of S for the ST and ETS reactors respectively, which corresponds to min 70%COD removal efficiency.

� 2015 Elsevier Ltd. All rights reserved.

1. Introduction

Antibiotics, which are one of the most commonly used pharma-ceutical products, have a wide range of uses in both human andveterinary medicine. These active compounds cause damage to

existing enzyme systems are discharged to receiving bodies with-out or with a little, mineralization from treatment plants andaccumulated in soil and sediment day by day (Kümmerer, 2009).This accumulation may result in the proliferation of antibiotic-resistant-pathogens that threaten public health through includingchanges in the native microbial population of the ecosystem(Resende et al., 2014). Although concentrations of these antibioticsare relatively low in raw domestic wastewater (100 ng/L–6 lg/L)they can be significantly higher in hospital and pharmaceuticalindustry effluents, reaching the 100–500 mg/L level (Martín et al.,

Table 1Tested antibiotic concentrations.

Sulfamethoxazole(mg/L)

Erythromycin(mg/L)

Tetracycline(mg/L)

Stage 1 0.5 0.1 0.1Stage 2 5 0.2 0.2Stage 3 5 0.5 0.5Stage 4 10 0.5 0.5Stage 5 10 1 1Stage 6 15 1 1Stage 7 15 1.5 1.5Stage 8 20 1.5 1.5Stage 9 20 2 2Stage 10 25 2.5 2.5Stage 11 40 2.5 2.5Stage 12 40 3 3

208 S. Aydin et al. / Bioresource Technology 186 (2015) 207–214

2012). Additionally, antibiotics typically occur within some formsof mixtures that have antagonistic or synergistic effects. The effectsof mixtures of this nature are generally stronger than the impact ofthe individual components from which they are formed, even if allcomponents are only present in low concentrations that do notprovoke significant toxic effects (Aydin et al., 2015a; Ozbayramet al., 2014).

Biological treatment processes have been reported to effectivelytreat some of the products of pharmaceutical manufacturingwastewaters (Shi et al., 2014). Although activated sludge treatmentprocesses have been used to treat these types of wastewaters, theycan be unsuitable when the COD levels exceed 1500 mg/L. The highCOD of the wastewaters makes them a favorable alternative foranaerobic biotechnology. Anaerobic digestion also uses less energy,has a lower sludge yield and lower nutrient requirements, ischeaper to implement, uses less space, and offers improved biogasrecovery (Chelliapan et al., 2006; Oktem et al., 2008).

Various anaerobic treatment systems have been used to treatsingle antibiotic compounds (Sreekanth et al., 2009; Auerbachet al., 2007). Although research in this area is promising, the num-ber of experimental studies that have investigated the combinedeffects of antibiotics on the anaerobic treatment of pharmaceuticalwastewaters is scarce. For example, Massé et al. (2000) investi-gated the presence of individual and combined antibiotics on theanaerobic digestion of swine manure slurry in SBRs. Their resultsindicated that the presence of 55 mg Carbadox/kg, 110 mgTylosin/kg and 16 mg Penicilin/kg combinations in manure slurriesdid not have a noticeable adverse effect on methane production.However, the presence of individual penicillin and tetracycline(550 mg/kg) in manure slurries reduced methane production by35% and 25%, respectively. Furthermore, Álvarez et al. (2010)observed the significant inhibition of antibiotic mixtures, includingoxytetracycline and chlortetracycline on the anaerobic digestion ofpig manure. Similarly, Beneragama et al. (2013) demonstrated theindividual oxytetracycline (OTC) and combined OTC and cefazolin(CFZ) at concentrations of 30, 60 and 90 mg/L resulted in a 70.3%,68.6% and 82.7%, 70.3% reduction methane production respec-tively. Aydin et al. (2015a) also found that various tetracycline, sul-famethoxazole and erythromycin combinations resulted in aremarkable synergistic effect on anaerobic digestion. This studyrevealed that antibiotic mixtures does have an adverse effect onhomoacetogenic bacteria and methanogens and that this may haveinhibitory effects on propionate (e.g. Syntrophobacter species,Pelotomaculum species) and butyrate-oxidizing syntrophic bacteria(e.g. Syntrophomonas spp. and Syntrophospora spp.), resulting inunfavorable effects on methanogenesis.

This study focused on the tetracycline, sulfamethoxazole anderythromycin antibiotics that are most commonly used in humanand veterinary medicine in Turkey. Erythromycin is among themain representative of the macrolide group of antibiotics forclinical use. An important difference between erythromycin andother macrolides, such as clarithromycin and roxithromycin, isthe sensitivity of erythromycin to pH. Tetracycline are also thecheapest classes of antibiotics available today, making them attrac-tive for use in developing countries that have limited health carebudgets. Erythromycin and tetracycline prevent bacteria fromgrowing by binding irreversibly to the 50S and 30S bacterialrRNA subunits respectively (Tenson et al., 2003). Sulfamethoxa-zole is one of the most frequently detected sulfonamide group ofantibiotics in wastewater treatment systems. It achieves an inhibi-tory effect through two main methods: it inhibits the synthesis ofnucleic acids and/or inhibits glutamic acid ability to permeate thebacterial wall, which prevents folic acid synthesis from being suc-cessful (McDermott et al., 2003). They are also active against abroad spectrum of Gram-positive and Gram-negative bacteriaincluding species of the genus Syntrophobacter, Pelotomaculum,

Syntrophomonas spp., Syntrophospora spp, Streptococcus,Staphylococcus, and Clostridium (Le-Minh et al., 2010).

This research aimed to develop an understanding of the tripleeffects of sulfamethoxazole–erythromycin–tetracycline (ETS) anddual effects of sulfamethoxazole–tetracycline (ST) on the perfor-mance of anaerobic sequencing batch reactor systems (SBRs)throughout a year operation. The chronic joint effects of ETS andST on the COD removal efficiency, VFA production, antibioticremoval, biogas production and composition in SBRs wereinvestigated.

2. Methods

2.1. The experimental approach

The experimental set-up consisted of three anaerobic SBRs ofidentical dimensions and configurations. The anaerobic SBRs wererun in a daily ‘‘fill and draw’’ mode using a synthetic substrate mix-ture that included volatile fatty acids, glucose and starch (Aydinet al., 2014; Aydin et al., 2015b). The mixture was formulated toresemble the wastewater produced by the pharmaceutical indus-try, as per the solution described in a study conducted by Aminet al. (2006). The operation of the anaerobic SBRs included astart-up period of approximately 90 days to allow acclimationand establishment of steady-state conditions. In the steady-statecondition, the SBRs exhibited a stable performance, with an aver-age effluent soluble COD of 92 ± 19 mg/L, which corresponded toa COD removal of around 95% and an average biogas productionof 1247 ± 3 mL/day. The concentration of the influent antibioticswere gradually increased through successive phases, each lastingfor 30 days, until the metabolic collapse of the anaerobic batchreactors, which was inferred when there was no COD removal orbiogas production in the SBRs (Aydin et al., 2015b). The antibioticconcentrations used at each stage are provided in Table 1. Duringthe operation, the ST and ETS reactors operated effectively untilthe 10th (360 days operation) and 12th stages (420 days opera-tion) respectively. The concentrations of antibiotics used werebased on the inhibition levels of antibiotics reported byCetecioglu et al. (2013) and Aydin et al. (2015a). Daily antibioticsdosing was stopped after the total collapse of the ST and ETS reac-tors, which were then operated for a further 30 days in order toobserve whether the reactors recovered. A third anaerobic SBR thatwas fed free of antibiotics was operated for the entire research per-iod under identical conditions, and this served as the controlreactor.

The evaluation of the performance of the anaerobic SBRs waspredominantly based on measurements of soluble COD, and vola-tile fatty acid (VFA) concentrations determined both in the influentand effluent streams. These measurements were supplementedwith parallel daily measurements of the biogas production and

S. Aydin et al. / Bioresource Technology 186 (2015) 207–214 209

composition that were compiled through assessing main fractionssuch as CH4, CO2 and H2.

2.2. Operation of anaerobic sequencing batch reactor systems

The three anaerobic SBRs, which had a liquid volume of 1.5 L,were inoculated using granular sludge from an anaerobic contactreactor that is used to treat raki and fresh grape alcohol wastewa-ters. The anaerobic SBRs were operated in 24-h cycles that con-sisted of 10 min feeding, 23 h 40 min reaction, 1 min settling and9 min liquid withdrawal. The three anaerobic reactors were ini-tially operated to reach a steady state at an OLR of 2.5 kg COD/m3 day at which point daily antibiotic additions commenced.Throughout the operation, a hydraulic retention time (HRT) of2.5 days and a solids retention time (SRT) of 30 days were used.Reactor temperatures of 35 ± 2 �C and continuous mixing at90 rpm were maintained. Stable operation was reached on the90th day of the operation of the reactor. The amount of mixedliquor volatile suspended solid (MLVSS) was fixed at 5000 mg/L.The composition of synthetic wastewater constituted1160 mg COD/L starch, 750 mg COD/L glucose, 135 mg/L CODsodium acetate, 183 mg/L COD sodium butyrate and 272 mg/LCOD sodium propionate (Aydin et al., 2015b). The trace elementsolution, which was adopted from a previous study by Aydinet al. (2014) and Aydin et al. (2015b). To sustain operational stabil-ity, the pH of the SBRs were adjusted on a daily basis from 6.8 to7.2 through the addition of 1000 mg/L CaCO3 alkalinity.

2.3. Analytical methods

Duplicate samples were collected on a daily basis from theinfluent and effluent streamlines of the SBRs and chemical analysesprocesses, such as soluble COD, alkalinity, total solids (TS), totalvolatile solids (TVS), were performed in accordance with standardmethods (APHA, 2005). A Milligas Counter (Ritter Digital Counter,U.S.A.) was used to monitor the biogas production in the SBRs.The gas compositions and VFA concentrations were measuredusing gas chromatographs with a flame ionization detector(Perichrom, France and Agilent Technologies 6890N, USA, respec-tively). The column used was Elite FFAP (30 m � 0.32 mm). Theset point of the oven and maximum temperature of the inlet were100 �C and 240 �C respectively. Helium gas was used as a carriergas at a rate of 0.8 mL/min.

2.4. Erythromycin, tetracycline and sulfamethoxazole assay

Erythromycin, sulfamethoxazole and tetracycline assay was per-formed using a Shimadzu high-performance liquid chromatography(HPLC) instrument (Schimadzu LC-10 AD) that was equipped withUV detector (UV–Vis Detector, SPD 10-A) by injecting sample solu-tions onto a C18 analytical column. Prior to use, the solvents weredegassed through sonification in a transonic ultrasonic bath(ELMA D-78224, Singen/Htw). All results were analyzed by thesystem software, LC Solutions (Schimadzu Scientific InstrumentsInc., MD, USA).

Triplicate samples were collected from the biomass, influent andeffluent streamlines of the ETS and ST reactors on the 20th day ofevery antibiotic stage. The preparation of samples was performedduring our previous study (Cetecioglu et al., 2013). Quantificationbased on peak areas, was performed by external standard calibra-tion. The external standards used for the quantification of the com-pounds were sulfamethoxazole D3, erythromycin and tetracycline.Seven point calibration curves (0.5–100 ppb) were generated usinglinear regression analysis. The calibration standards were measuredthree times on a random basis. The linarites, as qualified by the linear

correlation coefficient, r2 and seven point calibration curves of thecompounds, also represented a very good fit, r2 > 0.99.

Erythromycin’s gradient elution was applied using (A) 32 mMpotassium phosphate buffer by dissolving 5.57 g dipotassiumhydrogen phosphate in 1000 mL water adjusted with concentratedphosphoric acid to pH 8.0 and a mixture (B) of acetonitrile/metha-nol (75/25). Gradient was run with 33% B from 0 to 28 min and33–45% B from 28 to 60 min, before being post-run with 33% Bfor 10 min. Erythromycin was detected at 215 nm. The flow ratewas 1.0 mL/min (Deubel and Holzgrabe, 2007). Sulfamethoxazolewas eluted with 0.1% formic acid in acetonitrile (solvent A) and0.1% formic acid in water (solvent B). The mobile phase com-position was changed as follows: A:B::5:95 at the offset rising to30:70 from 0 to 7 min. Equilibration was then performed from 7to 8.5 min at 30:70 and then returned to 5:95 from 8.5 to10 min. The flow rate was 0.7 mL/min. Detection was carried outat 270 nm (Karcı and Balcıoglu, 2009). Tetracycline was elutedwith a 74% 0.1 M okzalic acid and 25% metanol:asetonitril (1:1.5)solution which was delivered at a flow rate of 1 mL/min.Detection was carried out 367 nm (Yuan et al., 2010). The limitof detection (LOD) and limit of quantification (LOQ) of the ery-thromycin, tetracycline and sulfamethoxazole measurement were0.58, 0.56 and 0.52; 1.94, 1.85 and 1.75 ng/mL, respectively.

2.5. Statistical analysis

Statistical analyses were conducted in MINITAB (2013, USA). Todetermine the statistical significance of the inhibition of the antibi-otic mixtures (ETS and ST), the COD removal efficiencies of theASBRs were compared using a one-way ANOVA test. After that, astudent’s T-test was conducted. Significant differences were deter-mined at p < 0.05.

3. Results and discussion

3.1. Performance of ASBRs

3.1.1. pHThe pH levels were generally stable (pH 6.8–7.2) at all stages of

the control reactor (Fig. S1). However, the pH in the ST reactordropped to 5.9 at Stage 8 and it wasn’t until Stage 10 of the opera-tion of the ETS reactor that the pH dropped to a similar level of 5.6.The reduction in pH in both reactors was due to the rapid produc-tion of VFA created from reduced methanogenic activity and alsoincreased acetogenesis process. From the pH data (see also VFAand biogas data below) it can be assumed that the metabolicprocesses differed between Stages 1–8 of the ST reactor andStages 1–10 of the ETS reactor. Existing research indicates that ashort contact time between high antibiotic concentrations and bio-mass has a positive impact on organic acid production (acidogens)which results in faster growth kinetics and a better rate of adapta-tion to reduced pH than the methanogens (Ma et al., 2013). Areduction in pH can also be caused by an accumulation of organicacids due to the failure of methanogens to convert the organicacids to methane. An example of this can be observed in a studythat was carried out by Ahring et al. (1995), in which the research-ers reported that the organic acid production effects a decrease inpH, which prevents acetate and hydrogen from using methano-genic archaea.

3.1.2. COD removalEfficient COD removal was observed during Stage 1 in the ST

reactor and Stages 1–3 in the ETS reactor: Soluble COD in the efflu-ent was reduced from an initial COD concentration of 2500 mg/L atthe beginning of each cycle to 92 ± 19 mg/L, corresponding to an

210 S. Aydin et al. / Bioresource Technology 186 (2015) 207–214

efficiency that exceeded 95% (Fig. 1). Similar COD removal wasobserved in the control reactor for the entire monitoring period.It is worth noting that synthetic substrate consists of organic com-pounds that are naturally biodegradable. Based on the results ofsimilar studies that used single substrate or substrate mixtures,these compounds would be totally removed and the low levels ofsoluble COD detected in the effluent is essentially residual solublemicrobial products that were generated during the biochemicalreactions (Cetecioglu et al., 2013; Aydin et al., 2015b).

Soluble COD removal efficiency was not significantly (p > 0.05)affected by the addition of ST mixtures during Stage 2 to Stage 3in the ST reactor influent (days 90–150). The first remarkable effectof ST mixtures on the reactor performance was detected duringStage 4 on day 155. A substantial increase to 480 ± 51 mg/L inthe soluble COD concentration in the effluent from the ST reactorwas observed during Stage 4, while the soluble COD concentrationof the effluent from the control reactor was 64.2 ± 38.5 mg/L(p < 0.05). Consequently, the performance of the reactor decreasedsubstantially after Stage 9 with the addition of the ST mixtures(30 mg/L) between the 330th and 360th days (Aydin et al.,2015b). Contrary to this result, Sponza and Demirden (2007) foundthat the COD removal efficiency decreased from 87% to 68% whensulfamerazine concentration was increased from 10 mg/L to90 mg/L. This contradiction could be due to variations in theanaerobic reactor types, the inoculum, operating conditions orantibiotic combinations used in the studies.

Semi-continuous ETS dosing of 11 mg/L at Stage 4, 12 mg/L atStage 5 and 17 mg/L at Stage 6 did not seem to exert a noticeableeffect on the overall COD removal. A significant (p < 0.05) increasein the effluent soluble COD concentration to 600 ± 15 mg/Lcorresponding to overall COD reduction of 75% in the ETS reactorwas observed during Stage 7, while the effluent soluble COD con-centration of the control reactor was 82 ± 21.2 mg/L. ETS mixtureswere increased to 46 mg/L during the following operation phase(Stage12), resulting in a significant decrease in the performanceof the reactor. The soluble COD value in the effluent increased tomore than 2000 mg/L, corresponding to an overall COD reductionof only 10% after the 420th day.

Antibiotics dosing was stopped at the end of Stage 12 in the ETSreactor and Stage 10 in the ST reactor in order to observe any pos-sible recovery in the performance of the reactor. However, themetabolic activity of the biomass could not be re-activated toinduce noticeable substrate utilization, and the operation of theST reactor was terminated on day 390 and the ETS was terminatedon day 450.

Fig. 1. COD removal efficiency

Previous research on the influence that antibiotics have onanaerobic treatment systems has not been conclusive. For example,Cetecioglu et al. (2013) observed that tetracycline had a significanteffect on anaerobic systems; however, Hu et al. (2011) reportedthat tetracycline had no influence on similar systems.Furthermore, Shimada et al. (2008) reported the strong inhibitionof anaerobic SBRs in the presence of tylosin (from the same groupof erythromycin), while in similar experiments Chelliapan et al.(2006) observed negligible effects of the addition of tylosin. Also,Loftin et al. (2005) observed that sulfamethoxazole significantlyinhibited COD removal in anaerobic batch systems, but Sponzaand Demirden (2007) observed a much less substantial impacton anaerobic process following the addition of sulfamerazine (sul-fonamide antibiotics). However, there is a lack of informationavailable that assesses the effect of ST and ETS on anaerobic treat-ment systems.

3.1.3. Biogas productionBiogas production was monitored during all stages of the opera-

tion of the reactor, with a particular focus on the assessment ofmethanogenic activity. Fig. 2 illustrates the biogas produced duringall stages of operation of the ETS and ST reactors. Biogas generationreduced significantly (p < 0.05) during Stage 4 for the ST reactor(86 ± 2%) and Stage 6 for the ETS reactor (87 ± 3%). The results ofthis study indicate that biogas production was correlated withthe efficiency with which COD was removed in the ETS and STreactors (Aydin et al., 2015b). Meanwhile, the amount of biogasproduced in the control reactor was almost constant throughoutthe operational time (1247 ± 3 mL/day).

Methane yield (L methane produced per g COD removed) can bea useful parameter for the purposes of assessing the performanceof an anaerobic reactor. Fig. 3 shows that the mean, steady-statemethane yield was relatively constant during Stages 1–5 of opera-tion of the Control and ETS reactors, indicating an average specificmethane production yield of YCH4 of 0.30 ± L/g COD removed.However, the methane yield dropped dramatically during Stage6, when the ETS dosing was altered from 17 to 18 mg/L and thenremained nearly stable until Stage 10. Following Stage 10, withan ETS dose of 45 mg/L, the significant adverse effect of the reactorperformance was observed for methane yield, which dropped from44% to 24% between days 300 and 322. Methane yield was at analmost stable level in the ST reactor until Stage 4, and thendecreased dramatically during Stage 9. Consequently, the methaneyield decreased substantially after Stage 9 with the addition of ST

in the ST and ETS reactors.

Fig. 2. Biogas production in the ETS and ST reactors.

Fig. 3. Methane yield in the ETS and ST reactors.

S. Aydin et al. / Bioresource Technology 186 (2015) 207–214 211

mixtures (22 mg/L) on day 266 (25%). Methane production in allreactors showed a strong correlation (p < 0.05) with biogas produc-tion during the operational time. This also indicates that the effectthat antibiotic concentrations have on different microbial groups ofthe anaerobic process (Aydin et al., 2015b). The average amount ofmethane produced in the control reactor was 879 ± 4 mL/day.Therefore, the methane percentage of biogas production in thecontrol reactor was calculated to be about 66%, indicating amethane production yield of YCH4 per 0.32 L/g COD removed.

A limited study on the combined effect of antibiotics on biogasproduction in existing literature was found. Massé et al. (2000)reported that tetracycline and penicillin reduced the methane pro-duction of psychrophilic anaerobic digestion (20 �C) in swine man-ure slurry in sequencing batch reactor by 25% and 35% respectively.Similarly, Álvarez et al. (2010) reported the significant inhibition ofanaerobic digestion in swine manure containing a combination ofchlortetracycline (CTC) and oxytetracycline (OTC) at concentra-tions of 10, 50 and 100 mg/L at 35 �C, where maximum methaneproduction decreased by 64% in manure containing 100 mg/Lof both CTC and OTC. Christensen et al. (2006) observed thesignificant synergistic effects of antibiotic mixtures including

erythromycin and oxytetracycline on activated sludge sample.The results indicated that, antibiotics have synergistic effects onbiological treatment processes. The impacts of antibiotic mixtureson a mixed culture may be different than their impact on pure cul-tures, as each group of bacteria may respond in dissimilar ways tothe different groups of antibiotics (Aydin et al., 2015b).

3.1.4. Effluent VFA compositionThe presence and composition of the volatile fatty acids concen-

tration during all stages of the operation of the ST and ETS reactorsare shown in Fig. 4. VFA could not be detected in the effluent of thecontrol reactor during the entire operation period (Aydin et al.,2015b). Furthermore, VFAs were not detected in the ST reactor’seffluent until the 110th day (Stage 4), at which point acetic acidand propionic acid accumulation started at 47 and 59 mg/L respec-tively. While butyric acid concentration varied between 7 and49 mg/L and valeric acid concentration slowly increased from17 mg/L to 105 mg/L until the end of operation. At Stage 10, aceticacid and propionic acid were detected at 1000 and 691 mg/L,respectively. The results indicated that ST antibiotic combinationsat higher concentrations affected the propionic and acetic acid

Fig. 4. VFA profile in the (a) ST reactors, and (b) ETS reactor.

212 S. Aydin et al. / Bioresource Technology 186 (2015) 207–214

utilization pathways. The degradation of propionate is most oftenutilized by gram-negative bacteria, and combinations of STantibiotics would be expected to inhibit sensitive strains of thismicrobial group.

The first VFA in the effluent of the ETS reactor was measured onthe 180th day, the first day of Stage 7, at 50 mg/L of acetic acid(Fig. 4b). On the 240th day, while the acetic acid concentrationincreased to 210 mg/L; valeric, butyric and isobutyric acids weredetermined for the first time at 20, 140 and 70 mg/L, respectively.Butyric acid concentration increased on daily basis and reached710 mg/L at the end of operation. Similar findings were reportedby Beneragama et al. (2013) who also reported an accumulationof VFAs resulting in decreased methane production in the presenceof cefazolin and oxytetracycline antibiotics.

Acetic acid accumulation displayed a similar trend; however, itwas at a higher concentration than butyric acid and propionic acid.Butyric acid utilization pathways were inhibited in the ETS reactorbut not in the ST reactor. The results indicate that ETS combina-tions have a more dramatic effect on gram-positive bacteria thangram-negative bacteria (Aydin et al., 2015a). The degradation ofbutyric acid is most often utilized by gram-positive bacteria andETS combinations would be expected to inhibit sensitive strainsof this microbial group.

The VFA results revealed that both the ST and ETS reactors had aclear effect on the acetoclastic methanogens, which utilize acetateto produce methane. This was indicated that homoacetogen affectduring antibiotic biodegradation. However, no hydrogen was mea-sured in the biogas evaluation, and this corresponded with theresults of a study by Shimada et al. (2008). This is a clear indication

that hydrogenotrophic methanogens can instantly changehydrogen to CH4. It should be stated that it is not anticipated thathomoacetogenic bacteria perform any vital function in the degra-dation of acetate. The findings of the current study support theinformation and results of previous studies. Similarly, Stone et al.(2010) speculated that chlortetracycline, as a tetracycline antibi-otic, might have contributed to the inhibition of acetoclasticmethanogens and that the concentration of VFA increased duringthe operation of an anaerobic digester.

The VFA results indicated that the high dosage of antibiotics hadan impact on the acetoclastic methanogens in the ETS and STreactors. This could be attributed to the accumulation of organicacids due to the failure of the methanogens to utilize acetate toproduce methane. Also, ETS combinations have a more dramaticeffect on Gram-positive bacteria than Gram-negative bacteria.However, unlike in the ETS reactor, the propionic acid utilizationpathway was inhibited in the ST reactor. The degradation ofpropionate is most often utilized by gram-negative bacteria (e.g.Syntrophobacter species, Pelotomaculum species), and combinationsof ST antibiotics would be expected to inhibit sensitive strains ofthis microbial group (Aydin et al., 2015a).

3.1.5. Erythromycin, tetracycline and sulfamethoxazole eliminationThe degradation efficiency of erythromycin (ERY), sulfamethoxa-

zole (SMX) and tetracycline (TET) antibiotics is illustrated inFig. S2a–c. This indicates a pattern of reduction in SMX and ERYthat started with a more than 60% reduction, increased to morethan 80% during Stage 6, and sustained around 10% at the end ofStage 12, when the COD removal efficiency and biogas production

S. Aydin et al. / Bioresource Technology 186 (2015) 207–214 213

were practically stopped. TET removal efficiency started with 60%,then throughout Stages 2–3–4 the TET removal efficiency of theETS reactor remained constant (95%) before reaching 3% duringStage 10. The results indicated that TET was removed from the sys-tem during Stages 1–9, even though TET is not biodegradableunder anaerobic conditions. Cetecioglu et al. (2013) also examinedthe efficiency of tetracycline removal under anaerobic conditions.According to their results, an average of 80% tetracycline reductionwas achieved in the SBR, indicating that this antibiotic could bedegraded efficiently in the anaerobic reactor system. Fig. S3 indi-cates SMX reduction patterns that started at 62%, increased to80% during Stage 5, and sustained at around 20% at the end ofStage 10; TET reduction started at 50% then increased to 60% dur-ing Stages 2–3–4, and continued to 12% at the end of Stage 12,where substrate/COD utilization and biogas production was practi-cally stopped. Also, a comparison of the removal behavior of SMXand TET in the ST reactor demonstrated that SMX had a higherremoval efficiency than TET, and it was clear that anaerobic treat-ment is a suitable for this compound to remove from wastewaterat lower concentration (10.5 mg/L and 16.5 mg/L for the ST andETS reactor respectively).

It is clear from Figs. S2 and S3 that despite the collapse of thereactors during Stage 12 (46 mg/L) of the ETS reactor and, stage10 (27.5 mg/L) of the ST reactor, antibiotic removal efficiency inboth reactors was not zero. One possible explanation for this couldbe attributed to the amount of antibiotics and their metabolitesthat were physically removed by means of sorption onto biomass.In fact, these results match those observations on earlier studiesthat shown sorption as the main mechanism for the removal ofantibiotics (Le-Minh et al., 2010; Cetecioglu et al., 2013; Aydinet al., 2014).

Comparison of the TET removal behavior in the ETS and ST reac-tors demonstrated that the ETS reactor removed TET more effi-ciency (44%) than the ST reactor (35% TET removal efficiency).This suggests that TET and ERY antibiotic combinations have sig-nificant (p < 0.05) synergistic effects on the anaerobic biodegrada-tion pathway.

The results obtained from the ETS and ST reactors showed thatthe increase in the dosage of antibiotic combinations caused adecrease in antibiotic removal efficiency; however when repeatinga constant dosage during the next stage, the reactor exhibited anon-significant removal efficiency because of the microorganisms’development resistance to antibiotics. Antibiotic resistance genescan be transferred between bacteria that are found in the environ-ment through plasmids, integrons, and transposons (Resende et al.,2014). It is largely accepted that an aqueous environment providesthe ideal conditions through which resistant genes can be trans-ferred between bacteria (Baquero et al., 2008). Fan and He (2011)established that the existence of erythromycin at concentrationsas low as those found in the natural environment can significantlyincrease antibiotic resistance. Furthermore, Gao et al. (2012) foundthat there is a positive relationship between antibiotics and thenumbers of antibiotic resistance genes and bacteria found in con-ventional wastewater treatment plants.

3.2. Effect of elevated erythromycin, tetracycline and sulfamethoxazoleconcentrations

The COD removal efficiency compared to the ERY, SMX and TETremoval efficiency in the ETS and ST reactors can be observed inFigs. S4 and S5. As shown, COD reduction was higher than theantibiotic removal efficiency during all stages. This may be dueto the inadequate degradation efficiency of the anaerobic system.Furthermore, the COD removal efficiency of the ETS and ST reactorsdecreased with respect to time and concentration of antibiotics. Asseen in Fig. S4, antibiotics degradation significantly (p < 0.05)

decreased during Stage 7. At the same time, the removal of antibi-otics gradually impaired the COD removal efficiency. There was ahigh degree of SMX degradation during Stages 4–5–6–7 of theETS reactor. COD reduction was gradually decreased and Fig. S5shows that COD removal efficiency was nearly the same as SMXdegradation efficiency during Stages 4–5. Also, COD removal effi-ciency and SMX degradation decreased dramatically during Stage6. It is indicated that the long-term antibiotic combinations feedingto the reactors caused an increase in the accumulation of theantibiotics and this directly impacted the efficiency of the reactors.It would be useful for future studies to compare the fullchromatograms of the influent, effluent and sludge to ascertainwhether the intermediates metabolites were formed during thetreatment process.

4. Conclusions

This study indicated that the anaerobic process tolerated antibi-otic combinations until Stage 3 and Stage 6 of the operation of theST and ETS reactors. Following these stages, there were negativeimpacts of increasing antibiotic concentrations on COD removaland biogas production in the SBRs. The antibiotic removal behaviorin the SBRs demonstrated that the ETS reactor removed antibioticsmore efficiency than the ST reactor. The 8th and 10th Stages werecritical for the ST and ETS reactors respectively. After these stages,the accumulation of antibiotics and VFA increased rapidly anddirectly impacted the performance of the reactors.

Acknowledgement

Authors thank TUBITAK (Turkish Association of Science andTechnology) for support of this project (No: 110Y310).

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.biortech.2015.03.043.

References

Amin, M.M., Zilles, J.L., Greiner, J., Charbonneau, S., Raskin, L., Morgenroth, E., 2006.Influence of the antibiotic erythromycin on anaerobic treatment of apharmaceutical wastewater. Environ. Sci. Tech. 40 (12), 3971–3977.

Ahring, B.K., Sandberg, M., Angelidaki, I., 1995. Volatile fatty acids as indicators ofprocess imbalance in anaerobic digestors. Appl. Microbiol. Biotechnol. 43 (3),559–565.

Álvarez, J.A., Otero, L., Lema, J.M., Omil, F., 2010. The effect and fate of antibioticsduring the anaerobic digestion of pig manure. Bioresour. Technol. 101 (22),8581–8586.

American Public Health Association (APHA), 2005. Standard Methods for theExamination of Water and Wastewater, 21st ed. American Public HealthAssociation, Washington, DC.

Auerbach, E.A., Seyfried, E.E., McMahon, K.D., 2007. Tetracycline resistance genes inactivated sludge wastewater treatment plants. Water Res. 41 (5), 1143–1151.

Aydin, S., Ince, B., Cetecioglu, Z., Ozbayram, E.G., Shahi, A., Okay, O., Ince, O., 2014.Performance of anaerobic sequencing batch reactor in the treatment ofpharmaceutical wastewater containing erythromycin and sulfamethoxazolemixture. Water Sci. Technol. 70 (10), 1625–1632.

Aydin, S., Cetecioglu, Z., Arikan, O., Ince, B., Ozbayram, E.G., Ince, O., 2015a.Inhibitory effects of antibiotic combinations on syntrophic bacteria,homoacetogens and methanogens. Chemosphere 120, 515–520.

Aydin, S., Ince, B., Ince, O., 2015b. Application of real-time PCR to determination ofcombined effect of antibiotics on bacteria, methanogenic archaea, archaea inanaerobic sequencing batch reactors. Water Res. 76, 88–98.

Baquero, F., Martínez, J.-L., Cantón, R., 2008. Antibiotics and antibiotic resistance inwater environments. Curr. Opin. Biotechnol. 19 (3), 260–265.

Beneragama, N., Lateef, S.A., Iwasaki, M., Yamashiro, T., Umetsu, K., 2013. Thecombined effect of cefazolin and oxytertracycline on biogas production fromthermophilic anaerobic digestion of dairy manure. Bioresour. Technol. 133, 23–30.

214 S. Aydin et al. / Bioresource Technology 186 (2015) 207–214

Cetecioglu, Z., Ince, B., Gros, M., Rodriguez-Mozaz, S., Barceló, D., Orhon, D., Ince, O.,2013. Chronic impact of tetracycline on the biodegradation of an organicsubstrate mixture under anaerobic conditions. Water Res. 47 (9), 2959–2969.

Chelliapan, S., Wilby, T., Sallis, P.J., 2006. Performance of an up-flow anaerobic stagereactor (UASR) in the treatment of pharmaceutical wastewater containingmacrolide antibiotics. Water Res. 40 (3), 507–516.

Christensen, A.M., Ingerslev, F., Baun, A., 2006. Ecotoxicity of mixtures of antibioticsused in aquacultures. Environ. Toxicol. Chem. 25 (8), 2208–2215.

Deubel, A., Holzgrabe, U., 2007. Development of an enhanced separation oferythromycin and its related substances by liquid chromatography. J. Pharm.Biomed. Anal. 43 (2), 493–498.

Fan, C.A., He, J.Z., 2011. Proliferation of antibiotic resistance genes in microbialconsortia of sequencing batch reactors (SBRs) upon exposure to traceerythromycin or erythromycin-H2O. Water Res. 45 (10), 3098–3106.

Gao, P., Munir, M., Xagoraraki, I., 2012. Correlation of tetracycline and sulfonamideantibiotics with corresponding resistance genes and resistant bacteria in aconventional municipal wastewater treatment plant. Sci. Total Environ. 421,173–183.

Hu, Z., Liu, Y., Chen, G., Gui, X., Chen, T., Zhan, X., 2011. Characterization of organicmatter degradation during composting of manure-straw mixtures spiked withtetracyclines. Bioresour. Technol. 102 (15), 7329–7334.

Karcı, A., Balcıoglu, I.A., 2009. Investigation of the tetracycline, sulfonamide, andfluoroquinolone antimicrobial compounds in animal manure and agriculturalsoils in Turkey. Sci. Total Environ. 407 (16), 4652–4664.

Kümmerer, K., 2009. The presence of pharmaceuticals in the environment due tohuman use – present knowledge and future challenges. J. Environ. Manage. 90(8), 2354–2366.

Le-Minh, N., Khan, S., Drewes, J., Stuetz, R., 2010. Fate of antibiotics duringmunicipal water recycling treatment. Water Res. 44 (15), 4295–4323.

Loftin, K.A., Henny, C., Adams, C.D., Surampali, R., Mormile, M.R., 2005. Inhibition ofmicrobial metabolism in anaerobic lagoons by selected sulfonamides,tetracyclines, lincomycin, and tylosin tartrate. Environ. Toxicol. Chem. 24 (4),782–788.

Ma, J., Frear, C., Wang, Z.-W., Yu, L., Zhao, Q., Li, X., Chen, S., 2013. A simplemethodology for rate-limiting step determination for anaerobic digestion ofcomplex substrates and effect of microbial community ratio. Bioresour.Technol. 134, 391–395.

Martín, J., Camacho-Munoz, D., Santos, J.L., Aparicio, I., Alonso, E., 2012. Occurrenceof pharmaceutical compounds in wastewater and sludge from wastewater

treatment plants: removal and ecotoxicological impact of wastewaterdischarges and sludge disposal. J. Hazard. Mater. 239, 40–47.

Massé, D.I., Lu, D., Masse, L., Droste, R.L., 2000. Effect of antibiotics on psychrophilicanaerobic digestion of swine manure slurry in sequencing batch reactors.Bioresour. Technol. 75 (3), 205–211.

McDermott, P.F., Walker, R.D., White, D.G., 2003. Antimicrobials: modes of actionand mechanisms of resistance. Int. J. Toxicol. 22 (2), 135–143.

Oktem, Y.A., Ince, O., Sallis, P., Donnelly, T., Ince, B.K., 2008. Anaerobic treatment of achemical synthesis-based pharmaceutical wastewater in a hybrid upflowanaerobic sludge blanket reactor. Bioresour. Technol. 99 (5), 1089–1096

Ozbayram, E.G., Arikan, O., Ince, B., Cetecioglu, Z., Aydin, S., Ince, O., 2014. Acuteeffects of various antibiotic combinations on acetoclastic methanogenicactivity. Environ. Sci. Pollut. Res., 1–6

Resende, J.A., Silva, V.L., de Oliveira, T.L.R., de Oliveira Fortunato, S., da CostaCarneiro, J., Otenio, M.H., Diniz, C.G., 2014. Prevalence and persistence ofpotentially pathogenic and antibiotic resistant bacteria during anaerobicdigestion treatment of cattle manure. Bioresour. Technol. 153, 284–291.

Shi, X., Lefebvre, O., Ng, K.K., Ng, H.Y., 2014. Sequential anaerobic–aerobic treatmentof pharmaceutical wastewater with high salinity. Bioresour. Technol. 153, 79–86.

Shimada, T., Zilles, J.L., Morgenroth, E., Raskin, L., 2008. Inhibitory effects of themacrolide antimicrobial tylosin on anaerobic treatment. Biotechnol. Bioeng.101 (1), 73–82.

Sponza, D.T., Demirden, P., 2007. Treatability of sulfamerazine in sequential upflowanaerobic sludge blanket reactor (UASB) completely stirred tank reactor (CSTR)processes. Sep. Purif. Technol. 56 (1), 108–117.

Sreekanth, D., Sivaramakrishna, D., Himabindu, V., Anjaneyulu, Y., 2009.Thermophilic treatment of bulk drug pharmaceutical industrial wastewatersby using hybrid up flow anaerobic sludge blanket reactor. Bioresour. Technol.100 (9), 2534–2539.

Stone, J.J., Clay, S.A., Spellman, G.M., 2010. Tylosin and chlortetracycline effectsduring swine manure digestion: influence of sodium azide. Bioresour. Technol.101 (24), 9515–9520.

Tenson, T., Lovmar, M., Ehrenberg, M., 2003. The mechanism of action of macrolides,lincosamides and streptogramin B reveals the nascent peptide exit path in theribosome. J. Mol. Biol. 330 (5), 1005–1014.

Yuan, S., Wang, Q., Yates, S.R., Peterson, N.G., 2010. Development of an efficientextraction method for oxytetracycline in animal manure for high performanceliquid chromatography analysis. J. Environ. Sci. Health B 45 (7), 612–620.