Waste Management 34 (2014) 2080–2089

Contents lists available at ScienceDirect

Waste Management

journal homepage: www.elsevier .com/ locate/wasman

Investigation into the effect of high concentrations of volatile fatty acidsin anaerobic digestion on methanogenic communities

http://dx.doi.org/10.1016/j.wasman.2014.07.0200956-053X/� 2014 The Authors. Published by Elsevier Ltd.This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/3.0/).

⇑ Corresponding author.E-mail address: ingrid.whittle@uibk.ac.at (I.H. Franke-Whittle).

Ingrid H. Franke-Whittle a,⇑, Andreas Walter a, Christian Ebner b, Heribert Insam a

a Institut für Mikrobiologie, Universität Innsbruck, Technikerstraße 25, 6020 Innsbruck, Austriab Abwasserverband Zirl und Umgebung, Meilbrunnen 5, 6170 Zirl, Austria

a r t i c l e i n f o a b s t r a c t

Article history:Received 25 February 2014Accepted 29 July 2014Available online 24 August 2014

Keywords:Volatile fatty acids (VFAs)ANAEROCHIPReal-time PCRMethanogensAnaerobic digestion (AD)

A study was conducted to determine whether differences in the levels of volatile fatty acids (VFAs) inanaerobic digester plants could result in variations in the indigenous methanogenic communities. Twodigesters (one operated under mesophilic conditions, the other under thermophilic conditions) weremonitored, and sampled at points where VFA levels were high, as well as when VFA levels were low.Physical and chemical parameters were measured, and the methanogenic diversity was screened usingthe phylogenetic microarray ANAEROCHIP. In addition, real-time PCR was used to quantify the presenceof the different methanogenic genera in the sludge samples. Array results indicated that the archaealcommunities in the different reactors were stable, and that changes in the VFA levels of the anaerobicdigesters did not greatly alter the dominating methanogenic organisms. In contrast, the two digesterswere found to harbour different dominating methanogenic communities, which appeared to remain sta-ble over time. Real-time PCR results were inline with those of microarray analysis indicating only mini-mal changes in methanogen numbers during periods of high VFAs, however, revealed a greater diversityin methanogens than found with the array.

� 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license(http://creativecommons.org/licenses/by/3.0/).

1. Introduction

The anaerobic digestion (AD) of organic wastes is a sustainablewaste management strategy that is gaining significance due to theincreasing costs of fossil fuels and the need to mitigate anthropo-genic global warming. Biogas production from various types ofraw materials (e.g. manure, sewage sludge, food waste) has beenshown to be a source of renewable energy that can occur sustainablyin many different countries around the world (Bond and Templeton,2011). The process produces a sludge of agricultural value, as well asbiogas, which can be used to generate electricity and heat (Lastellaet al., 2002; Insam and Wett, 2008). The efficient conversion oforganic matter to methane in an anaerobic digester is dependenton the mutual and syntrophic interactions of functionally distinctmicroorganisms (Akuzawa et al., 2011). However, despite a contin-uously increasing interest and popularity of AD, the process remainsinefficient and enigmatic. This is because of a lack of knowledgelinking microbial community content, dynamics, and activity withreactor performance (Lee et al., 2009; Pycke et al., 2011).

The initial step in the AD process is the hydrolysis of the organicmaterials. During this step, macromolecules, such as proteins,

carbohydrates and fats, are broken down to amino acids, sugarsand fatty acids, respectively, by bacteria (Nielsen et al., 2007) andfungi (Leis et al., 2014). In the second stage of the process, acido-genic bacteria convert sugars, amino acids, and fatty acids toorganic acids, alcohols and ketones, acetate, CO2, and H2. Acetogen-ic bacteria then convert fatty acids and alcohols into acetate, H2

and CO2, products used by methanogenic archaea to form biogas(typically 60% methane, 38% carbon dioxide and 2% trace gases).Archaea thus hold the key position in the methanisation. Methanecan be produced by both acetotrophic and hydrogenotrophicmethanogens, and differences in environmental conditions as wellas reactor operating conditions (pH, temperature, hydraulic reten-tion time, input material) have been reported to affect the compo-sition of these communities (Demirel and Scherer, 2008; Akuzawaet al., 2011; Kim et al., 2013). Because the degradation phases areall closely connected with each other, an imbalance between thebacterial and archaeal communities can cause a deterioration inreactor performance, and thus changes in the amount of methaneproduced (Demirel and Yenigün, 2002; Rastogi et al., 2008;Akuzawa et al., 2011).

Reactor acidification through reactor overload is one of the mostcommon reasons for process deterioration in anaerobic digesters(Akuzawa et al., 2011). This occurs because of a build-up of volatilefatty acids (VFAs) which are produced by acidogenic and acetogenic

I.H. Franke-Whittle et al. / Waste Management 34 (2014) 2080–2089 2081

bacteria, and reflects a kinetic uncoupling between the acid produc-ers and consumers (Ahring, 1995). High VFA concentrations causepH values to decrease, and result in toxic conditions in the reactor.In anaerobic digesters with low buffering capacity, pH, partial alka-linity and VFAs are reliable indicators for process imbalance, how-ever, in highly buffered systems, pH changes can be small, evenwhen the process is extremely stressed, and only VFAs can be con-sidered reliable for process monitoring (Murto et al., 2004). VariousVFAs exist in ADs, and they have different and co-operative effectson bacteria and archaea. Wang et al. (2009) reported that acetic acidand butyric acid concentrations of 2400 and 1800 mg L�1, respec-tively, resulted in no significant inhibition of the activity of metha-nogens, while a propionic acid concentration of 900 mg L�1

resulted in significant inhibition of the methanogens. Opinions varyregarding which VFA is the best indicator for impending reactor fail-ure, with different authors suggesting i-butyric, i-valeric, propionicacid, or the ratio of propionic:acetic acid as the most appropriateindicator (Boe, 2006). Nonetheless, it does not appear to be possibleto define VFA levels to indicate the state of an anaerobic process, asdifferent systems have their own levels of VFAs that can be consid-ered ‘normal’ for the reactor, and conditions that cause instabilityin one reactor do not cause problems in another reactor(Angelidaki et al., 1993).

Other chemical compounds which are known to cause toxiceffects in biogas reactors and can lead to a complete failure ofmethanogenesis are ammonia and hydrogen sulfide (Kayhanian,1994; Chen et al., 2008), as well as accumulations of hydrogenand acetate, excess tannins, salts and heavy metals (Karri et al.,2006; Panyadee et al., 2013).

The aim of this study was to investigate the hypothesis thatchanges in the VFA levels of anaerobic digester plants can influenceindigenous methanogenic communities. Archaeal communitiespresent in two different anaerobic digesters were monitored usingthe ANAEROCHIP microarray. This microarray, which targets the16S rRNA gene, offers the possibility to analyse an entire array ofmethanogens concerning their presence or absence in a particularsludge sample in a single experiment (Franke-Whittle et al.,2009a). Methanogens detected using the microarray were quanti-fied using real-time PCR to determine exact numbers.

2. Materials and methods

2.1. Sampling of sludge from biogas producing reactors

Anaerobic sludges were collected from two AD plants (Inzing andNeustift) in Tirol, Austria, at various times (May 18, 2009; August 5,2009; September 16, 2009 and October 27, 2009; for Neustift reac-tor, no samples for the August sampling were available). The Inzingreactor (I) was run under mesophilic conditions and had a reactorvolume of 173 m3, an organic loading rate (OLR) of 2.8 kg VS m�3

d�1, a hydraulic retention time (HRT) of 57 d and a combined heatand power plant (CHP) with 20 kW electrical performance. TheNeustift reactor (N) was run under thermophilic conditions, had areactor volume of 110 m3, an OLR of 5.2 kg VS m�3 d�1, a HRT of26 d and a CHP with 25 kW electrical performance. Both plantsproduced sludges that were used for agricultural purposes.

Information on the AD plants is listed in Table 1. Three bulkedsamples (each about 0.5 L) were collected from the reactorsthrough the sampling ports.

2.2. Physical–chemical parameters

Temperature and biogas production were measured online inthe reactors, and gas quality (CH4 [%], CO2 [%]) of the reactor wasanalysed with a portable Biogas Check BM 2000 instrument

(Geotechnical Instruments, Warwickshire, UK). pH and electricalconductivity (EC) were measured in sludge samples at the time ofcollection using a portable multi-parameter meter Multi 340i(WTW, Weilheim, Germany). Total solids (TS) were calculated asthe amount of solids remaining after oven-drying the samples over-night (105 �C). Volatile solids (VS) were calculated as the loss ofweight after igniting the oven-dried residue at 550 �C for 5 h. Sam-ple preparation for HPLC analysis was performed using dialysis andthe method of Wagner et al. (2012). Following sample collection, adialysis tube filled with 10 ml of distilled water was submerged intothe liquid sample. The bottle was shaken three times and stored at4 �C overnight in order to reach a total equilibrium in the dialysate.The tubing was removed, washed with a minimum volume of dis-tilled water and opened. Dialysate (0.5 ml) was subjected to highperformance liquid chromatography (HPLC) analysis on an AminexHPX-87H column (Bio-Rad, Hercules, USA). A 5 mM H2SO4 mobilephase run at 0.7 ml min�1 and a detection wavelength of 210 nmwere used. The detection limit for VFAs was >1 mmol l�1. Ammo-nium nitrogen (NH4–N) was measured photometrically after appro-priate dilution with distilled water using the colorimetric tube testfrom Macherey-Nagel (Düren, Germany). NH3–N was calculatedfrom NH4–N concentrations according to the formula of Calli et al.(2005). The PASW-SPSS 17.0 software was used to determine corre-lations between physical–chemical parameters and methanogenicgenera present in sludges.

2.3. DNA extraction

The PowerSoil DNA Isolation Kit (MO BIO Laboratories, Carlsbad,California) was used to extract genomic DNA from sludge samplesaccording to the instructions of the manufacturer, with the excep-tion that the sample after being subjected to lysis buffers and mix-ing was exposed to three freeze–thaw cycles (30 min at �80 �Cfollowed by 5 min at 65 �C). This was done in order to improvethe efficiency of cell lysis. DNA extractions were conducted in trip-licate from the well mixed bulked sludge samples. Extracted DNAwas subjected to electrophoresis in a 1% agarose gel in 1 X TAE buf-fer, and DNA concentration was determined by fluorescence using aPicoGreen� dsDNA quantitation kit (Molecular Probes Inc., Oregon,USA) and a fmax Fluorescence Microplate Reader (MolecularDevices, CA, USA), as described by the manufacturer.

2.4. ANAEROCHIP microarray analysis

The 109F and 934R primers (Grosskopf et al., 1998) were usedto amplify the 16S rRNA gene of methanogens in the sludge sam-ples by PCR, as described by Franke-Whittle et al. (2009b). Sin-gle-stranded Cy5-labeled PCR product was generated usingLambda exonuclease and 500 ng of single-stranded DNA washybridised on an ANAEROCHIP microarray at 55 �C for 4 h(Franke-Whittle et al., 2009b). Arrays were washed after hybridisa-tion, and a ScanArray Gx microarray scanner (Perkin Elmer, MA,USA) was used to scan hybridised microarray slides. The ScanArrayGx software (Perkin Elmer, MA, USA) was used to analyse fluores-cent images, as described by Franke-Whittle et al. (2009b). For allspots, the median foreground and background signals were deter-mined. The signal-to-noise ratio (SNR) for all spots was calculatedusing the following calculation, as described by Loy et al. (2002):

SNR = [Ip-(Inp–Ibnp)]/Ibp where Ip is median intensity of fluores-cence of the probe, Inp is the median intensity of fluorescence ofthe nonbinding control probe, Ibnp is the median intensity of fluo-rescence of the background area around the nonbinding controlprobe, and Ibp is the median intensity of fluorescence of the back-ground area around the probe. Signals were treated as positive if aSNR value of P2 was obtained (Loy et al., 2002).

Table 1Reactor sampling dates, operational details and input materials.

Plant Sampling date Sample name Operation temperature Input material

Inzing 18.05.2009 I1 37–38 �C Cow manure (46%), corn silage (36%), vegetable waste (9%), potato (9%)05.08.2009 I2 37–38 �C Cow manure (46%), corn silage (36%), vegetable waste (9%), potato (9%)16.09.2009 I3� 37–38 �C Cow manure (46%), corn silage (36%), vegetable waste (9%), potato (9%)27.10.2009 I4 37–38 �C Cow manure (46%), corn silage (36%), vegetable waste (9%), potato (9%)

Neustift 18.05.2009 N1� 55 �C Cow manure (52%), food waste (48%)16.09.2009 N3 55 �C Cow manure (52%), food waste (48%)27.10.2009 N4� 55 �C Cow manure (52%), food waste (48%)

Note: Vegetable waste refer to wastes obtained from the field after harvesting vegetables. Food waste refers to household kitchen waste. Low quality potatoes not fit forconsumption included in Inzing reactor.� High VFA samples.

2082 I.H. Franke-Whittle et al. / Waste Management 34 (2014) 2080–2089

Principal component analysis (PCA) of the total SNR microarraydata of the study was conducted, using CANOCO for windows 4.5(Ter Braak and Šmilauer, 2002).

2.5. Real-time quantitative PCR

Real-time quantitative PCR (RT-PCR) was conducted on sludgeDNA samples with specific primers targeting the 16S rRNA geneof five different genera of methanogens, determined according tothe analysis of the microarray results. The genera investigatedwere Methanosaeta, Methanosarcina, Methanothermobacter, Met-hanoculleus and Methanobacterium. RT-PCR amplifications wereconducted using the Sensimix SYBR No-ROX kit (Biotools, Spain)and performed in a Rotor-Gene™ 6000 (Corbett Life Sciences, Syd-ney, Australia) in 20 ll volumes. Each standard reaction mix con-tained a final concentration of 1 X Sensimix reaction premix,100 nM each primer (150 nM for Methanoculleus and Methanobac-terium), 0.04 mg ml�1 BSA and distilled water (Franke-Whittleet al., 2009a; Goberna et al., 2010). Sequences of the Methanosaetaprimers (MS1b and SAE835R), Methanosarcina primers (240F and589R), Methanothermobacter primers (410F and 667R), Methanocul-leus primers (298F and 586R) and Methanobacterium primers(fMbium and 748R) are listed in Goberna et al. (2010). One llsludge DNA (undiluted) was used as the template in each reaction.After an initial denaturation at 95 �C for 5 min, thermal cyclingcomprised 40 cycles of 20 s at 95 �C, 20 s at 58–65 �C (annealingtemperature) and 20 s at 72 �C. The annealing temperatures forthe various PCR programs were as follows: 58 �C for Methanobacte-rium, 60 �C for Methanosaeta, 61 �C for Methanothermobacter, 64 �Cfor Methanosarcina and 65 �C for Methanoculleus. Thermal cyclingwas completed with a melting analysis (65–95 �C, ramp 0.5 �C/min) to check for primer dimer formation and product specificity.Standard curves were constructed with PCR amplified 16S rDNAfrom pure cultures (Methanobacterium formicicum DSMZ 1535,Methanosaeta concilii DSMZ 2139, Methanothermobacter wolfeiiDSMZ 2970, Methanosarcina barkeri DSMZ 800 and Methanoculleus

Table 2Physical–chemical parameters of the biogas reactor sludges sampled at different times.

Plant pH EC(mS cm�1)

TS (%[wt/vol])

VS (%[wt/wt]of TS)

NH3–N(mg L�1)

NH4–N(mg L�1)

CH4

(%)

I1 7.3 9.97 3.03 71 15 640 57.2I2 7.5 11.90 3.58 64 26 720 58.8I3� 7.4 11.81 3.53 72 19 720 56.8I4 7.4 12.33 2.91 66 21 800 52.3N1� 7.7 28.20 3.67 62 539 3840 64.5N3 7.9 24.90 3.50 65 636 2880 63.4N4� 8.0 25.10 5.13 65 721 2960 65.0

Note: nd- not detected.� High VFA samples.

thermophilus DSMZ 2640) as described in Franke-Whittle et al.(2009a). All standards and samples were run in duplicate.

3. Results and discussion

3.1. Physical–chemical parameters

Table 2 shows the results of physical–chemical parameter anal-ysis for the two biogas reactors. Neutral, or close to neutral pH val-ues (7.3–7.5) were observed for the duration of the experiment inthe Inzing reactor, indicating stable digester conditions. Likewise,the Neustift reactor revealed stable values (7.7–8.0), despite beinghigher. The values did not vary significantly, despite changes inVFA levels. According to Sandberg and Ahring (1992) and Wardet al. (2008), AD occurs optimally at pH values of 6.8–7.2, and anexcessively alkaline pH can potentially result in disintegration ofmicrobial granules and the subsequent failure of the process. Thetwo reactors under investigation nonetheless produced acceptablelevels of methane (57.2–65.0%) and were operating stably, despitetheir higher pH values.

Conductivity in the two digestion plants varied, with values of9.97–12.33 mS cm�1 obtained for reactor I, and values of 24.90–28.20 mS cm�1 for reactor N. Similarly, NH4–N and NH3–N concen-trations varied considerably in the two reactors, as shown inTable 1. Reactor I had significantly lower values (640–800 mg L�1

NH4–N; 15–26 mg L�1 NH3–N) than those found in the thermo-philic reactor N (2880–3840 mg L�1 NH4–N; 539–721 mg L�1

NH3–N). Kayhanian (1999) also showed that the free ammonianitrogen concentration in a thermophilic reactor can be expectedto be six times higher than when compared to a mesophilic reactorat the same pH. Total ammonia nitrogen concentrations of 1.7 g L�1

or higher are known to inhibit methanogenesis (McCarty andMcKinney, 1961; Koster and Lettinga, 1984; Rajagopal et al.,2013), and archaeal diversity was reported to be affected by con-centrations of ammonia exceeding 2000 mg L�1 (Garcia et al.,2000; McMahon et al., 2001; Demirel and Scherer, 2008). Total

CO2

(%)Acetate(mg L�1)

Propionate(mg L�1)

Isobutyrate(mg L�1)

Butyrate(mg L�1)

Valerate(mg L�1)

Isovalerate(mg L�1)

41.6 25.3 nd nd nd nd 54.040.7 20.8 2.7 159.4 nd 111.7 227.942.1 1249.6 3910.9 510.0 104.7 647.8 797.947.0 131.2 27.6 nd nd nd 24.435.4 2281.9 8741.3 650.5 1329.8 1237.2 2623.035.7 15.7 4.2 nd nd nd nd34.6 1184.9 4977.5 78.3 712.3 766.8 1189.2

I.H. Franke-Whittle et al. / Waste Management 34 (2014) 2080–2089 2083

ammonia nitrogen levels in the Neustift reactor sludges werefound to be rather high, and thus some inhibition of methanogen-esis could have been expected, although it did not occur.

Higher methane concentrations were found in the thermophilicreactor N (63.4–65%). This finding is thus inline with otherresearch, whereby thermophilic processes have been reported tobe more efficient than mesophilic processes, and have higher ratesof methane production (Cecchi et al., 1991; Griffin et al., 1998; Hoet al., 2013).

VFA levels were found to vary significantly in the two reactorsat the different sampling times. Of the four samples collected atdifferent times from the Inzing reactor, only the I3 sludge wasfound to contain high levels of many VFAs. An acetate level of1249.6 mg L�1 was measured, an amount that according to Hillet al. (1987) would indicate process instability (>13 mM). Elevatedpropionate levels (3910.9 mg L�1) were also detected, resulting in apropionate:acetate ratio far exceeding 1.4, a level considered toindicate process instability (Hill et al., 1987). Butyrate, isobutyrate,isovalerate and valerate levels were, in addition, higher in the I3sludge sample than in all other Inzing reactor sludges. It is possiblethat an overloading of the reactor prior to the sampling occurred,however, no information from the plant operators regarding thissampling time is available.

According to VFA analysis, two of the three Neustift reactorsludges analysed were found to contain high VFA levels (N1 andN4). The N1 reactor sludge was found to have the highest level ofall VFAs, including acetate and propionate levels (2281.9 mg L�1

and 8741.3 mg L�1, respectively). VFAs measured in samples col-lected from the Neustift reactor at other times also contained highVFA levels (data not shown). This would indicate that the Neustiftreactor was operating under higher loading rates. Possibly, therewas a difference in the OLR of the reactor prior to the collectionof the N3 sample, where much lower VFA levels were found. Differ-ent researchers have proposed various VFAs as indicators of reactorimbalance, e.g. Kaspar and Wuhrmann (1978) and Nielsen et al.(2007), who investigated reactors treating municipal wastes and

Fig. 1. Principal component analysis loading plot depicting the organisms responsible fNote: The lengths of the arrows indicate the significance for sludge sample differentiatReactor I samples are from a mesophilic anaerobic digestion plant in Tirol (I1- 18.05.2009a thermophilic anaerobic digestion plant in Tirol (N1� -18.05.2009, N3 -16.09.2009 and

manure/industrial waste, respectively, and proposed propionateto be a more appropriate indicator for process instability than ace-tate. However, according to Hill and Bolte (1989), who investigatedreactors treating swine wastes, isoforms of butyrate and valerateare the best indicators of imbalance. In any case, the elevated levelsof VFAs in the N1 and N4 sludge samples would suggest processinstability and impending digester failure (Nielsen et al., 2007).Methane production rates from the Neustift reactor did not how-ever indicate any problems, and thus it would appear that themicrobial communities in the reactor were adapted to a high VFAenvironment.

The accumulation of VFAs in most situations reflects an imbal-ance between acid producers (mostly bacteria) and consumers, andis usually associated with a drop in pH and a breakdown of the buf-fering capacity of the reactor sludge (Ahring, 1995; Akuzawa et al.,2011). The reduction in pH can inhibit the growth of methanogens(Bouallagui et al., 2005; Ye et al., 2013). However, reactor failure,which would be predicted according to the VFA levels in I3, N1and N4, did not result in the Inzing and Neustift reactors, pH valuesdid not change significantly, and a stable methane production con-tinued. Cow manure was included in the input materials of the tworeactors under investigation, and the high bicarbonate and ammo-nia contents most likely contributed to the buffering of the system(Pind et al., 2003) and prevention of reactor failure. According toAhring (1995), acetate and propionate concentrations of up to6000 and 3000 mg L�1, respectively, have been shown to causeno inhibition of the AD process in reactors treating manure. Thus,the high acetate concentrations seen in the reactors under investi-gation may not have been problematic, although the high propio-nate levels should alleviate concern.

It should also be taken into consideration that all the abovementioned VFA indicator system proposals are based on laboratoryscale reactors and research. Our data comes from full scale ADplants. Angelidaki et al. (1993) suggested that various AD plantshave their own ‘normal’ levels of VFAs, determined by the compo-sition of the input material entering the reactor, and thus, that it is

or community differences amongst the Inzing and Neustift reactor sludge samples.ion. Arrows of probes point in the direction of samples with above average signal., I2- 05.08.2009, I3�- 16.09.2009 and I4 -27.10.2009) and reactor N samples are fromN4� -27.10.2009). High VFA samples designated by �.

Table 3Signal to noise ratios (SNR) of ANAEROCHIP hybridisations with Inzing and Neustift reactor sludges.

Note: Probes with SNR values � 2 are highlighted in dark green. Probes with SNR � 1.5 are highlighted in light green.Probes for Methanocalculus, Methanocaldococcus, Methanococcoides, Methanomicrobium, Methanobrevibacter smithii,Methanosphaera stadtmanae and Methanothermobacter thermoautotrophicus excluded from table as no hybridisationsfor any of the probes were detected.Abbreviations: Mbact/Mthermbac = Methanobacterium/Methanothermobacter; Mlobus/Mmethylovorans = Methanolo-bus/Methanomethylovorans; Mmicrobium/Mgenium/Mplanus = Methanomicrobium/Methanogenium/Methanoplanus;Mthermbac/Mbact/Msph/Mbrev = Methanothermobacter/Methanobacterium/Methanosphaera/Methanobrevibacter.High VFA samples designated by ⁄.

2084 I.H. Franke-Whittle et al. / Waste Management 34 (2014) 2080–2089

I.H. Franke-Whittle et al. / Waste Management 34 (2014) 2080–2089 2085

not feasible to define specific VFA levels which indicate reactorfailure. It would appear that these two AD systems had a good spe-cific buffering capacity, and were able to operate stably despite thehigh VFA levels.

3.2. ANAEROCHIP microarray analysis

The ANAEROCHIP microarray, an oligonucleotide array targetingmethanogenic genera and species within the orders Methanobacte-riales, Methanococcales, Methanomicrobiales and Methanosarcinaleswas used to investigate the diversity of the methanogenic commu-nity structure present in the sludges collected from the two reactorsat different times. Organisms belonging to these orders are knownto dominate and thrive in the anaerobic digester environment (Horiet al., 2006; Demirel and Scherer, 2008; Rastogi et al., 2008). TheSNR was calculated for all probes, and those with an SNR P 2 inone or more samples were used for subsequent analyses.

Fig. 1 shows a principal component analysis (PCA) loadingplot, whereby the two first axes explain 87.8% of the variance,the first axis representing 79.5% of the variance, the second axisrepresenting 8.3%. Certain probes (indicated by the arrows) canbe seen to be more influential in discriminating the samples.The lengths of the arrows indicate the significance for sludgesample differentiation, and arrows point in the direction ofsamples with above average signal. Probes with similar arrowdirections have high covariance, meaning they tend to occurjointly on the microarrays.

Canonical analysis clearly shows the separate grouping of themesophilic and thermophilic reactor samples. The Inzing reactorsludge samples were found to be dominated by the generaMethanosaeta and Methanosarcina. Both genera belong to theorder Methanosarcinales, acetoclastic methanogens which havebeen reported to be responsible for approximately 70% of themethane produced in biogas reactors (Jetten et al., 1992;Ahring, 1995). Members of the genus Methanosaeta have beenreported to dominate in reactors with low levels of NH3 and VFAs(Karakashev et al., 2005; Walter et al., 2012), conditions thatexisted in the I1, I2, and I4 sludges, but not in I3,which containedhigh VFA levels. Nonetheless, Methanosaeta was found in the I3sample, indicating that the period of high VFAs did not signifi-cantly affect the community. Methanosarcina, a generalist knownto have a high metabolic versatility and able to use acetate,hydrogen, formate, secondary alcohols and methyl compoundsas energy sources (Kendall and Boone, 2006), as well as beingcapable of quickly adapting to changing conditions throughhigher growth rates (Goberna et al., submitted), was also detectedin Inzing reactor samples.

Table 4Gene copy number g�1 sample and standard deviation of Methanosarcina, Methanocullequantitative real-time PCR of the 16S rRNA gene and standard curve parameters. Numbemicroarray, while numbers in bold represent genera not detected using the array.

Methanosarcina Methanoculleus Met

I1 6.31E+04 ± 7.01E+02 1.07E+04 ± 4.17E+03 3.65I2 9.00E+04 ± 7.23E+04 6.67E+04 ± 1.05E+05 1.53I3� 1.57E+05 ± 1.33E+05 3.68E+04 ± 2.64E+04 2.75I4 7.33E+04 ± 7.79E+04 2.92E+04 ± 2.75E+04 4.15N1� 2.56E+04 ± 4.74E+03 3.03E+04 ± 8.90E+03 5.82N3 1.34E+03 ± 7.21E+02 1.27E+05 ± 1.30E+05 9.07N4� 1.32E+04 ± 4.01E+03 6.18E+03 ± 4.72E+03 7.53Rangea 102–107 102–106 102

R2 0.99958 0.99956 0.99Slope �4.268 �4.303 �3.Intercept 44.600 39.013 41.9Efficiency 72% 71% 80%

a Range of standards (gene copies ll�1) used for quantifying each genus.� High VFA samples.

Methanobacterium, a hydrogenotrophic methanogen, wasdetected in reactor I sludges using the microarray, with all fourprobes yielding signals in the sludges. As expected, the two Meth-anobacteriaceae family probes (313 and 855) and the group probes405 (Methanobacterium/Methanothermobacter) and 817 (Methan-othermobacter/Methanobacterium/Methanosphaera/Methanobrevib-acter) also yielded strong signals. Hybridisations for the probestargeting Methanoculleus and Methanosphaera were obtained,revealing the digester to host a metabolically diverse methanogen-ic community, comprised of organisms capable of methane pro-duction via various biochemical pathways.

In contrast, samples from the thermophilic reactor were notfound to be as diverse, a finding supported by Karakashev et al.(2005). Reactor N sludges were dominated by a community ofhydrogenotrophic Methanothermobacter. The signals obtained forthe Methanothermobacter probes were the highest of all genusprobes detected (Table 3), indicating both its dominance in thesample as well as a high hybridisation efficiency of these probes.Low signals for 2 of the 5 Methanoculleus probes were additionallydetected. These findings support the findings of Lee and Zinder(1988) and Petersen and Ahring (1991), whereby high levels ofhydrogenotrophic methanogens were found in thermophilicdigesters with low acetate concentrations. The hydrogenotrophicdominating methanogen community would indicate that syn-trophic relationships between acetate oxidisers and hydrogeno-trophic methanogens could be the main pathway for acetatedegradation and methanogenesis in the N reactor.

Hybridisation signals were also detected for the group probestargeting Methanobacteriaceae (313 and 855), Methanomicrobium/Methanogenium/Methanoplanus (309) and Methanothermobacter/Methanobacterium/Methanosphaera/Methanobrevibacter (817). LowSNR values were obtained for two of the Methanogenium probesin reactor N samples, explaining the signal for the 309 probe.

Interestingly, Methanoculleus was the only genus detected by theANAEROCHIP microarray in all Inzing and Neustift reactor samples.Schnürer et al. (1999) suggested that its dominance over otherhydrogenotrophic methanogens might be related to its toleranceof high salt concentrations. Ammonium levels in reactor N can beconsidered high (2.9–3.8 g L�1), and, according to McCarty andMcKinney (1961) and Koster and Lettinga (1984), concentrationsabove 1.7 g L�1 can be inhibitory to some organisms when biogasdigesters are not adapted to high ammonium levels. Methanoculleushas been reported to dominate in cattle manure fed reactors(Chachkiani et al., 2004; Goberna et al., 2009), despite significantchanges in pH and VFAs, supporting the findings of this study.

Table 3 shows the SNRs obtained after hybridisation of theANAEROCHIP microarray with DNA extracted from AD plant

us, Methanobacterium, Methanothermobacter, and Methanosaeta as revealed throughrs written in italics represent methanogenic genera detected using the ANAEROCHIP

hanobacterium Methanothermobacter Methanosaeta

E+05 ± 9.16E+04 1.05E+04 ± 1.16E+02 5.37E+06 ± 4.88E+06E+05 ± 5.57E+04 3.83E+03 ± 2.40E+03 8.12E+05 ± 8.17E+05E+05 ± 1.86E+05 1.09E+04 ± 1.15E+04 1.90E+06 ± 1.36E+06E+05 ± 4.25E+05 1.57E+04 ± 1.38E+04 1.83E+06 ± 2.25E+06E+03 ± 2.63E+02 4.94E+06 ± 3.18E+05 3.45E+04 ± 4.70E+04E+03 ± 4.43E+03 9.40E+06 ± 4.62E+06 2.68E+03 ± 1.85E+03E+03 ± 2.78E+03 9.73E+06 ± 4.68E+06 1.23E+04 ± 1.62E+04

–107 102–106 102–107

969 0.99927 0.99983926 �3.518 �3.65395 32.192 34.643

92% 88%

Table 5Correlation analysis of physical–chemical parameters and methanogenic genera present in Inzing reactor sludges.

Note: Significant positive correlations (p < 0.01) are highlighted in dark green, those with a p < 0.05 in light green.Abbreviations: Msarcina- Methanosarcina; Mculleus- Methanoculleus; Mbacter- Methanobacterium; Mtherbact- Methanothermobacter; Msaeta-Methanosaeta.cNot able to be determined.

2086 I.H. Franke-Whittle et al. / Waste Management 34 (2014) 2080–2089

sludges. There did not appear to be any significant differences inthe hybridisation signals obtained, or in the fluorescence intensityof signals between the sludges sampled at different times from thesame AD plant. This would indicate that the methanogens were notinfluenced by such fluctuations in VFA levels and reactor condi-tions, and that the established microbial community was able towithstand some variations in reactor conditions without majorchanges in methanogen profiles, a finding also found by Wagneret al. (2011). It is also possible, however, that had RNA and notDNA been used in these studies, variations in methanogen num-bers may have been detected.

The ANAEROCHIP microarray approach has to be considered‘semi-quantitative’, and the relative abundance of different micro-organisms derived from the microarrays needs to be interpretedwith some caution. This is because different probes have differentaffinities for their targets, and because of the inherent biasinvolved with PCR amplification. However, a linear correlationbetween signal intensity of probes and target concentration hasbeen reported by others (Taroncher-Oldenburg et al., 2003;Tiquia et al., 2004). Thus, in order to more accurately investigatemethanogen numbers, real-time PCR was applied to generadetected using the microarray.

3.3. Real-time PCR

Dominant genera in the sludge sample DNAs according to theresults of ANAEROCHIP microarray analysis were chosen as targetsfor quantification using real-time PCR. Quantitative real-time PCRresults largely supported the findings of the microarrays, althoughwere more sensitive, and all genera were detected in all Inzing andNeustift reactor samples. The results and standard curve parame-ters (slope, intercept, range, R2 and efficiency) for quantificationassays are included in Table 4. Numbers written in italics representmethanogenic genera detected using the ANAEROCHIP microarray,while numbers in bold represent genera not detected using thearray.

Methanosaeta dominated the sludges of reactor I, with genecopy numbers of 8.12 � 105–5.37 � 106 g�1 reactor sampledetected at different times, despite differing VFA levels. This evi-dence supports the results from ANAEROCHIP microarray analysis,

and indicates that the accumulation of VFAs did not significantlyaffect the numbers of Methanosaeta in the mesophilic reactor, amethanogen known to be sensitive to higher concentrations ofVFAs (Griffin et al., 1998).

Methanosarcina, as well as the hydrogenotrophic genera Met-hanobacterium and Methanoculleus, were also detected in reactorI samples (Methanosarcina: 6.31 � 104–1.57 � 105 copies g�1 reac-tor sample, Methanobacterium: 1.53 � 105–4.15 � 105 copies g�1

reactor sample, and Methanoculleus: 1.07 � 104–6.67 � 104 copiesg�1 reactor sample). Methanobacterium and Methanoculleus num-bers were found to remain stable when VFA levels increased (I3),while numbers of Methanosarcina were found to increase in dom-inance when higher levels of VFAs were detected. Growth of mem-bers of the family Methanosarcinaceae are favoured underconditions where acetate has accumulated (Demirel and Scherer,2008), and our findings were thus inline with the findings ofothers.

Despite not being detected in Inzing reactor sludges using themicroarray, Methanothermobacter was detected at lower levelsthan other methanogens (3.83 � 103–1.57 � 104 copies g�1 reactorsample) in all four samples investigated. The levels of VFAs in thereactor did not appear to be correlated with Methanothermobacternumbers. Methanothermobacter has not been frequently reportedin mesophilic reactors in the literature (Ahring, 1995).

The thermophilic Neustift reactor was found to be dominatedby Methanothermobacter, with numbers of 4.94 � 106–9.73 � 106

copies g�1 reactor sample. Hori et al. (2006) found that species ofMethanothermobacter began to dominate in a thermophilic reactorwhen VFA levels increased, and in particular, propionate. As wasfound in reactor I, the levels of VFAs in the reactor did not appearto be correlated with Methanothermobacter numbers, and ourresults did not indicate an increasing dominance of the genus inthe N1 and N4 samples which contained high VFAs.

Methanosaeta was detected by real-time PCR in reactor N sam-ples (2.68 � 103–3.45 � 104 copies g�1 reactor sample), despite notbeing detected using the ANAEROCHIP array. Methanosarcina andMethanobacterium were also detected at lower levels in reactor Nsamples (up to 2.56 � 104 and 9.07 � 103copies g�1 reactor sample,respectively), despite not having been detected using the micro-array. This difference may be attributable to the fact that general

Table 6Correlation analysis of physical–chemical parameters and methanogenic genera present in Neustift reactor sludges.

Note: Positive correlations with a p < 0.05 are highlighted in light green, and negative correlations with a p < 0.05 are highlighted in yellow.Abbreviations: Msarcina- Methanosarcina; Mculleus- Methanoculleus; Mbacter- Methanobacterium; Mtherbact- Methanothermobacter; Msaeta-Methanosaeta.

I.H. Franke-Whittle et al. / Waste Management 34 (2014) 2080–2089 2087

archaeal primers were used to amplify DNA from the sludge sam-ples for microarray analysis, while genus specific primers wereused in real-time PCR analysis. It is possible that a bias existed withthe primers towards the more dominant methanogens, such thatthe less dominant organisms were not amplified enough to bedetected using the microrarray. It is also possible that the cellnumbers of Methanosarcina and Methanobacterium were belowthe detection limit of the array (�0.4 pg DNA, Franke-Whittleet al., 2009b). Methanosarcina was found to increase in dominancewhen higher levels of VFAs were present, as was found in reactor I.The numbers of Methanobacterium did not appear to be affected byVFA concentration. Methanobacterium is a genus containing hydro-genotrophic methanogens. Studies conducted by Griffin et al.(1998) concluded that species of the genus Methanobacterium werethe syntrophic partners of acetate oxidisers in a thermophilicdigester treating municipal solid waste and biosolids, and domi-nated by Methanobacteriaceae. Although Methanobacterium didnot dominate in samples from either reactor, species of the genushave been reported by others to be dominant in anaerobic digest-ers (Godon et al., 1997; Leclerc et al., 2001).

Similar copy numbers of Methanoculleus were also detected in allNeustift reactor samples (3.03 � 104–1.27 � 105 copies g�1 reactorsample). According to Goberna et al. (2009), Methanoculleusremained the dominant methanogen in a thermophilic reactor,despite significant changes in pH and VFAs. In contrast, Hori et al.(2006) showed that Methanoculleus numbers in a thermophilicanaerobic digester declined during the accumulation of VFAs.

3.4. Correlation analyses

To detect correlations between the various physical–chemicalparameters and methanogenic genera in the two reactors, datawere subjected to analysis using the PASW-SPSS 17.0 software.Results are presented in Tables 5 and 6. Significant positive corre-lations (p < 0.01) are shaded in dark green, those with a p < 0.05 inlight green. Negative correlations (p < 0.05) are shaded in yellow.Correlation analysis for the reactor I reactor indicated significantcorrelations between Methanosarcina and all VFAs (Table 5). Thehigher growth rates at high acetate levels of species of Methanosar-cina when compared to Methanosaeta, have been reported by oth-ers (Zinder, 1993; Conrad and Klose, 2006; Demirel and Scherer,2008; Walter et al., 2012), and thus the correlation was not surpris-

ing. Similarly, the majority of VFAs were positively correlated witheach other, as expected. Interestingly, apart from significant corre-lations (p < 0.05) between NH3–N with pH and NH3–N with Met-hanoculleus, there were no other significant correlations ofparameters in the reactor with each other, or with the methano-gens. These correlations are in contrast with the findings of Niuet al. (2014), who reported a distinct rise in VFA accumulation athigh NH3–N levels, as well as the dominance of Methanoculleus inan AD reactor prior to the addition of high amounts of NH3–N,and the absence of the genus after inhibiting ammonia conditionshad been created.

Different significant correlations were detected in the reactor Nsludges (Table 6). Certain VFAs were positively correlated witheach other, as seen in reactor I. However, in contrast to reactor I,Methanosarcina was not significantly positively correlated withall VFAs, rather, only with acetate, butyrate and isovalerate. Met-hanobacterium, a hydrogenotrophic methanogen, was significantlynegatively correlated with the same VFA, as well as with Methano-sarcina. Methanothermobacter, which was dominant in the thermo-philic reactor, only correlated significantly positively with the VS.

4. Conclusions

In this study, the hypothesis that changes in VFA levels ofanaerobic digester plants affect the indigenenous methanogeniccommunities was investigated. Two different anaerobic digesters(one operating under mesophilic conditions and the other underthermophilic conditions) which experienced variation in VFA lev-els were monitored in terms of physical and chemical analyses.In addition, methanogenic communities were investigated usingDNA based tools.

The methanogenic community composition was similar in themesophilic and thermophilic reactors, however, the structurewas different. According to microarray analysis, the mesophilicreactor was shown to be dominated by a more diverse community,comprising Methanosarcina, Methanoculleus, Methanobacterium andMethanosaeta, while the thermophilic community was dominatedby Methanothermobacter. Real-time PCR results confirmed the find-ings of the microarrays, but were quantitative and more sensitive,indicating a greater diversity.

Our findings support those obtained not only in other biogasreactors, but also in other environments such as upland pasture rhi-

2088 I.H. Franke-Whittle et al. / Waste Management 34 (2014) 2080–2089

zosphere soil (Nicol et al., 2004), where acidification appeared tohave no significant effects on the indigenous community ofmethanogenic archaea. It would appear that the two AD systemsinvestigated had a good buffering capacity, and that the microbialcommunities were able to withstand changes in VFA concentrations.

Acknowledgments

We are grateful to Theresa Farbmacher for her collection of thesamples, and determination of physical and chemical parameters.Similarly, we acknowledge Jennifer Stoifl for the measurement ofvarious other parameters. The project was funded by the Fondszur Förderung der Wissenschaftlichen Forschung (FWF) Austria(Project FP200010).

References

Ahring, B.K., 1995. Methanogenesis in thermophilic biogas reactors. A. Van Leeuw.67, 91–102.

Akuzawa, M., Hori, T., Haruta, S., Ueno, Y., Ishii, M., Igarashi, Y., 2011. Distinctiveresponses of metabolically active microbiota to acidification in a thermophilicanaerobic digester. Microbial Ecol. 61, 595–605.

Angelidaki, I., Ellegaard, L., Ahring, B.K., 1993. A mathematical model for dynamicsimulation of anaerobic digestion of complex substrates: focusing on ammoniainhibition. Biotechnol. Bioeng. 42, 159–166.

Boe, K., 2006. Online monitoring and control of the biogas process. PhD thesis,Institute of Environment and Resources, Technical University of Denmark.

Bond, T., Templeton, M.R., 2011. History and future of domestic biogas plants in thedeveloping world. Energy Sustain. Dev. 15, 347–354.

Bouallagui, H., Touhami, Y., Ben Cheikh, R., Hamdi, M., 2005. Bioreactor performancein anaerobic digestion of fruit and vegetable wastes. Process Biochem. 40, 989–995.

Calli, B., Mertoglu, B., Inanc, B., Yenigun, O., 2005. Effects of high free ammoniaconcentrations on the performances of anaerobic bioreactors. Process Biochem.40, 1285–1292.

Cecchi, F., Pavan, P., Mata-Alvarez, J., Bassetti, A., Cozzolino, C., 1991. Anaerobicdigestion of municipal solid waste: thermophilic vs. mesophilic performance athigh solids. Waste Manage. Res. 9, 305–315.

Chachkiani, M., Dabert, P., Abzianidze, T., Partskhaladze, G., Tsiklauri, L., Dudauri, T.,Godon, J.J., 2004. 16S rDNA characterization of bacterial and archaealcommunities during start-up of anaerobic thermophilic digestion of cattlemanure. Bioresource Technol. 93, 227–232.

Chen, Y., Cheng, J.J., Creamer, K.S., 2008. Inhibition of anaerobic digestion process: areview. Bioresource Technol. 99, 4044–4064.

Conrad, R., Klose, M., 2006. Dynamics of the methanogenic archaeal community inanoxic rice soil upon addition of straw. Eur. J. Soil Sci. 57, 476–484.

Demirel, B., Scherer, P., 2008. The roles of acetotrophic and hydrogenotrophicmethanogens during anaerobic conversion of biomass to methane: a review.Rev. Environ. Sci. Biotechnol. 7, 173–190.

Demirel, B., Yenigün, O., 2002. Two-phase anaerobic digestion processes: a review. J.Chem. Technol. Biotechnol. 77, 743–755.

Franke-Whittle, I.H., Goberna, M., Insam, H., 2009a. Design and testing of real-timePCR primers for the quantification of Methanoculleus, Methanosarcina,Methanothermobacter and a group of uncultured methanogens. Can. J.Microbiol. 55, 11–616.

Franke-Whittle, I.H., Goberna, M., Pfister, V., Insam, H., 2009b. Design anddevelopment of the ANAEROCHIP microarray for investigation ofmethanogenic communities. J. Microbiol. Meth. 79, 279–288.

Garcia, J.L., Patel, B.K.C., Ollivier, B., 2000. Taxonomic, phylogenetic and ecologicaldiversity of methanogenic Archaea. Anaerobe 6, 205–226.

Goberna, M., Gadermaier, M., García, C., Wett, B., Insam, H., 2010. Adaptation ofmethanogenic communities to the cofermentation of cattle excreta and olivemill wastes at 37 �C and 55 �C. Appl. Environ. Microbiol. 76, 6564–6571.

Goberna, M., Gadermaier, M., Franke-Whittle, I.H., García, C., Wett, B., Insam, H. Fourstart-up strategies in biogas reactors treating cattle manure: processparameters and methanogenic communities. Biomass & Bioenergy (submitted).

Goberna, M., Insam, H., Franke-Whittle, I.H., 2009. Effect of biowaste sludgematuration on the diversity of thermophilic bacteria and archaea in ananaerobic reactor. Appl. Environ. Microbiol. 75, 2566–2572.

Godon, J.J., Zumstein, E., Dabert, P., Habouzit, F., Moletta, R., 1997. Molecularmicrobial diversity of an anaerobic digester as determined by small-subunitrDNA sequence analysis. Appl. Environ. Microbiol. 63, 2802–2813.

Griffin, M.E., McMahon, K.D., Mackie, R.I., Raskin, L., 1998. Methanogenic populationdynamics during start-up of anaerobic digesters treating municipal solid wasteand biosolids. Biotechnol. Bioeng. 57, 342–355.

Grosskopf, R., Janssen, P.H., Liesack, W., 1998. Diversity and structure of themethanogenic community in anoxic rice paddy soil microcosms as examined bycultivation and direct 16S rRNA gene sequence retrieval. Appl. Environ.Microbiol. 64, 960–969.

Hill, D.T., Bolte, J.P., 1989. Digester stress as related to iso-butyric and iso-valericacids. Biol. Wastes 28, 33–37.

Hill, D.T., Cobb, S.A., Bolte, J.P., 1987. Using volatile fatty acid relationships to predictanaerobic digester failure. T. Asabe 30, 496–501.

Ho, D.P., Jensen, P.D., Batstone, D.J., 2013. Methanosarcinaceae and acetate oxidisingpathways dominate in high-rate thermophilic anaerobic digestion of wasteactivated sludge. Appl. Environ. Microbiol.. http://dx.doi.org/10.1128/AEM.01730-13.

Hori, T., Haruta, S., Ueno, Y., Ishii, M., Igarashi, Y., 2006. Dynamic transition of amethanogenic population in response to the concentration of volatile fatty acidsin a thermophilic anaerobic digester. Appl. Environ. Microbiol. 72, 1623–1630.

Insam, H., Wett, B., 2008. Control of GHG emission at the microbial communitylevel. Waste Manage. 28, 699–706.

Jetten, M.S.M., Stams, A.J.M., Zehnder, A.J.B., 1992. Methanogenesis from acetate: acomparison of the acetate metabolism in Methanothrix soehngenii andMethanosarcina spp. FEMS Microbiol. Rev. 88, 181–198.

Karakashev, D., Batstone, D.J., Angelidaki, I., 2005. Influence of environmentalconditions on methanogenic compositions in anaerobic biogas reactors. Appl.Environ. Microbiol. 71, 331–338.

Karri, S., Sierra-Alvarez, R., Field, J.A., 2006. Toxicity of copper to acetoclastic andhydrogenotrophic activities of methanogens and sulfate reducers in anaerobicsludge. Chemosphere 62, 121–127.

Kaspar, H.F., Wuhrmann, K., 1978. Kinetic parameters and relative turnovers ofsome important catabolic reactions in digesting sludge. Appl. Environ.Microbiol. 36, 1–7.

Kayhanian, M., 1994. Performance of a high-solids anaerobic digestion processunder various ammonia concentrations. J. Chem. Technol. Biotechnol. 59, 349–352.

Kayhanian, M., 1999. Ammonia inhibition in high solids biogasification: anoverview and practical solutions. Environ. Technol. 20, 355–365.

Kendall, M.M., Boone, D.R., 2006. The order methanosarcinales. In: Dworkin, M.,Falkow, S., Rosenberg, E., Schleifer, K.-H., Stackebrandt, E. (Eds.), TheProkaryotes. A Handbook on the Biology of Bacteria, Archaea. Bacteria:Firmicutes, Actinomycetes, vol. 3. Springer-Verlag, New York, pp. 244–256.

Kim, J., Kim, W., Lee, C., 2013. Absolute dominance of hydrogenotrophicmethanogens in full-scale anaerobic sewage sludge digesters. J. Environ. Sci.25, 2272–2280.

Koster, I.W., Lettinga, G., 1984. The influence of ammonium-nitrogen on the specificactivity of pelletized methanogenic sludge. Agric. Waste 9, 205–216.

Lastella, G., Testa, C., Cornacchia, G., Notornicola, M., Voltasio, F., Sharma, V.K., 2002.Anaerobic digestion of semi-solid organic waste: biogas production and itspurification. Energy Convers. Manage. 43, 63–75.

Leclerc, M., Delbes, C., Moletta, R., Godon, J., 2001. Single stand conformationpolymorphism monitoring of 16S rDNA Archaea during start-up of an anaerobicdigester. FEMS Microbiol. Ecol. 34, 213–220.

Lee, C., Kim, J., Hwang, K., O’Flaherty, V., Hwang, S., 2009. Quantitative analysis ofmethanogenic community dynamics in three anaerobic batch digesters treatingdifferent wastewaters. Water Res. 43, 157–165.

Lee, M.J., Zinder, S.H., 1988. Hydrogen partial pressures in a thermophilic acetate-oxidizing methanogenic co-culture. Appl. Environ. Microbiol. 54, 1457–1461.

Leis, S., Dresch, P., Peintner, U., Fliegerova, K., Sandbichler, A.M., Insam, H.,Podmirseg, S.M., 2014. Finding a robust strain: anaerobic fungi(Neocallimastigomycota) and their associated methanogens from capricorn(Capra ibex). Anaerobe 29, 34–43.

Loy, A., Lehner, A., Lee, N., Adamczyk, J., Meier, H., Ernst, J., Schleifer, K., Wagner, M.,2002. Oligonucleotide microarray for 16S rRNA gene-based detection of allrecognized lineages of sulfate-reducing prokaryotes in the environment. Appl.Environ. Microbiol. 68, 5064–5081.

McCarty, P.L., McKinney, R.E., 1961. Volatile acid toxicity in anaerobic digestion. J.Water Pollut. Con. F 33, 223–232.

McMahon, K.D., Stroot, P.G., Mackie, R.I., Raskin, L., 2001. Anaerobic codigestion ofmunicipal solid waste and biosolids under various mixing conditions – II:microbial population dynamics. Water Res. 35, 1817–1827.

Murto, M., Björnsson, L., Mattiasson, B., 2004. Impact of food industrial waste onanaerobic co-digestion of sewage sludge and pig manure. J. Environ. Manage.70, 101–107.

Nicol, G.W., Webster, G., Glover, L.A., Prosser, J.I., 2004. Differential response ofarchaeal and bacterial communities to nitrogen inputs and pH changes inupland pasture rhizosphere soil. Environ. Microbiol. 6, 861–867.

Nielsen, H.B., Uellendahl, H., Ahring, B.K., 2007. Regulation and optimization of thebiogas process: propionate as a key parameter. Biomass Bioenergy 31, 820–830.

Niu, Q., Hojo, T., Qiao, W., Qiang, H., Li, Y.-Y., 2014. Characterization ofmethanogenesis, acidogenesis and hydrolysis in thermophilic methanefermentation of chicken manure. Chem. Eng. J 244, 587–596.

Panyadee, S., Petiraksakul, A., Phalakornkule, C., 2013. Biogas production from co-digestion of Phyllanthus emblica residues and food waste. Energy Sustain. Dev.17, 515–520.

Petersen, P.S., Ahring, B.K., 1991. Acetate oxidation in a thermophilic anaerobicsewage-sludge digester: the importance of non-aceticlastic methanogenesisfrom acetate. FEMS Microbiol. Lett. 86, 149–152.

Pind, P., Angelidaki, I., Ahring, B.S.K., Lyberatos, G., 2003. Monitoring and control ofanaerobic reactors. In: Ahring, B.K. (Ed.), Biomethanation II. Springer-Verlag,Berlin, pp. 135–182.

Pycke, B.F., Etchebehere, C., Van de Caveye, P., Negroni, A., Verstraete, W., Boon, N.,2011. A time-course analysis of four full-scale anaerobic digesters in relation tothe dynamics of change of their microbial communities. Water Sci. Technol. 63,769–775.

I.H. Franke-Whittle et al. / Waste Management 34 (2014) 2080–2089 2089

Rajagopal, R., Masse, D.I., Singh, G., 2013. A critical review on inhibition of anaerobicdigestion process by excess ammonia. Bioresource Technol. 143, 632–641.

Rastogi, G., Ranade, D.R., Yeole, T.Y., Patole, M.S., Shouche, Y.S., 2008. Investigationof methanogen population structure in biogas reactor by molecularcharacterization of methyl-coenzyme M reductase A (mcrA) genes.Bioresource Technol. 99, 5317–5326.

Sandberg, M., Ahring, B.K., 1992. Anaerobic treatment of fish-meal processwastewater in a UASB reactor at high pH. Appl. Microbiol. Biot. 36, 800–804.

Schnürer, A., Zellner, G., Svensson, B.H., 1999. Mesophilic syntrophic acetateoxidation during methane formation in biogas reactors. FEMS Microbiol. Ecol.29, 249–261.

Taroncher-Oldenburg, G., Griner, E.M., Francis, C.A., Ward, B.B., 2003.Oligonucleotide microarray for the study of functional gene diversity in thenitrogen cycle in the environment. Appl. Environ. Microbiol. 69, 1159–1171.

ter Braak, C.J.F., Šmilauer, P., 2002. Microcomputer Power, Ithaca. CANOCOReference Manual and CanoDraw for Windows User’s Guide: Software forCanonical Community Ordination (version 4.5) Microcomputer Power: Ithaca.

Tiquia, S.M., Wu, L., Chong, S.C., Passovets, S., Xu, D., Xu, Y., Zhou, J., 2004. Evaluationof 50-mer oligonucleotide arrays for detecting microbial populations inenvironmental samples. Biotechniques 36, 664–675.

Wagner, A., Malin, C., Lins, P., Illmer, P., 2011. Effects of various fatty acidamendments on a microbial digester community in batch culture. WasteManage. 31, 431–437.

Wagner, A.O., Hohlbrugger, P., Lins, P., Illmer, P., 2012. Effects of different nitrogensources on the biogas production – a lab-scale investigation. Microbiolog. Res.167, 630–636.

Walter, A., Knapp, B.A., Farbmacher, T., Ebner, C., Insam, H., Franke-Whittle, I.H.,2012. Searching for links in the biotic characteristics and abiotic parameters ofnine different biogas plants. Microbial Biotechnol. 5, 717–730.

Wang, Y., Zhang, Y., Wang, J., Meng, L., 2009. Effects of volatile fatty acidconcentrations on methane yield and methanogenic bacteria. BiomassBioenergy. 33, 848–853.

Ward, A.J., Hobbs, P.J., Holliman, P.J., Jones, D.L., 2008. Optimisation of the anaerobicdigestion of agricultural resources. Bioresource Technol. 99, 7928–7940.

Ye, J., Li, D., Sun, Y., Wang, G., Yuan, Z., Zhen, F., Wang, Y., 2013. Improved biogasproduction from rice straw by co-digestion with kitchen waste and pig manure.Waste Manage. 33, 2653–2658.

Zinder, S.H., 1993. Physiological ecology of methanogens. In: Ferry, J.G. (Ed.),Methanogenesis: Ecology, Physiology, Biochemistry, and Genetics. Chapman &Hall, New York, pp. 128–206.