Bioresource Technology 101 (2010) S2–S6

Contents lists available at ScienceDirect

Bioresource Technology

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

Effects of prolonged starvation on methanogenic population dynamicsin anaerobic digestion of swine wastewater

Kwanghyun Hwang, Minkyung Song, Woong Kim, Nakyung Kim, Seokhwan Hwang *

School of Environmental Science and Engineering, Pohang University of Science and Technology, San 31, Hyoja-dong, Nam-gu, Pohang, Kyungbuk 790-784, Republic of Korea

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

Article history:Received 1 November 2008Received in revised form 25 March 2009Accepted 27 March 2009Available online 9 May 2009

Keywords:Anaerobic batch reactorsDenaturing gradient gel electrophoresisMethanogensReal-time polymerase chain reactionStarvation

0960-8524/$ - see front matter � 2009 Elsevier Ltd. Adoi:10.1016/j.biortech.2009.03.070

* Corresponding author. Tel.: +82 54 279 2282; faxE-mail address: shwang@postech.ac.kr (S. Hwang)

This study investigated the relationship between the processes and microbial populations induced bylong-term starvation. To demonstrate the effects of starvation, a laboratory-scale anaerobic reactorwas operated in three phases (first reaction, starvation, second reaction) for 316 days. During the firstreaction, the chemical oxygen demand (COD) concentration decreased by about 70% of the input swinewastewater and 64 L of methane gas was produced; during the second reaction, there was a 63% CODreduction and 36 L of methane was produced. The methanogenic diversity, qualitatively monitored withdenaturing gradient gel electrophoresis using archaeal 16S rRNA gene primers, was not different betweentwo reactions. However, DNA copy numbers of Methanosarcinales, quantitatively monitored with quan-titative real-time polymerase chain reaction using order-level 16S rRNA gene primers, showed the chan-ged results. Cell numbers of Methanosarcinales and methanogenic activity were important factorsdetermining the different efficiencies of the process.

� 2009 Elsevier Ltd. All rights reserved.

1. Introduction

Anaerobic digestion of wastes is a three-stage process in whichthe complex organic components of the waste are hydrolyzed, bro-ken down, and converted to a variety of intermediates that are sub-sequently reduced into methane and carbon dioxide. The stages arereferred to as hydrolysis or liquefaction, acidogenesis, and metha-nogenesis. In the first phase, microorganisms release a variety ofextracellular enzymes hydrolyzing complex organics such as car-bohydrates, proteins and fats into soluble products such as glucose,alcohols, amino acids and long chain organic acids. A group ofmicroorganisms, referred to as acidogenic microorganisms or acid-ogens, then metabolize the reduced end products of the hydrolysisor liquefaction phase to produce various short chain volatile fattyacids (VFAs), H2 and CO2. Finally, methanogens, methane (CH4)forming archaea, produce mostly CH4 and CO2 from the end prod-ucts of the acidogenic phase (Jiang et al., 2005; Hori et al., 2006;Sousa et al., 2007; Ward et al., 2008). Anaerobic digestion offersa cost-effective solution for the treatment of high strength organicwaste such as swine wastewater because of methane formationand low sludge production. Thus, many different types of anaero-bic digesters have been installed in swine farms worldwide as ameans to solve environmental problems and/or a source of addi-tional revenue (Fernandes, 1994; Kim et al., 2004; Ndegwa et al.,2008).

ll rights reserved.

: +82 54 279 8299..

Swine diseases such as porcine reproductive and respiratorysyndrome (PRRS), classical swine fever (CSF), and foot-and-mouthdisease (FMD) are detrimental to the swine farms and often lead tohuge economic losses as well as massive slaughtering the infectedpigs. In an FMD outbreak that occurred in Taiwan in 1997, forexample, the disease spread from 3 farms in 2 prefectures to morethan 57 hundred farms in 15 prefectures and 5 cities within2 weeks, and ultimately more than 300 thousands swine had tobe killed. It is usually required longer than 2 months to reopenthe swine farm when these outbreaks occur (Wong et al., 2000;Meuwissen et al., 2003). Consequently, the anaerobic digesters in-stalled in the farms are likely to undergo a long period of starvationwithout receiving swine wastewater.

Starved conditions may induce changes in microbial dynamicssuch as physiological status and community composition affectingprocess performance (Cox and Deshusses, 2002; Koutinas et al.,2006). It can also cause changes in microbial growth, activity (Car-rero-Colón et al., 2006), and the cell surface properties (Castellanoset al., 2000; Sanin et al., 2003; Li et al., 2006). Konopka et al. (2002)reported that substrate starvation for less than 1 day results inmodest changes in adenosine 50-triphosphate (ATP) content andbiomass levels, whereas a starvation period of 32 days results insubstantial changes in microbial community structure. It is also re-ported that a long starvation period lowers the wastewater treat-ment efficiency (Lopez et al., 2006; Liu and Tay, 2008).

Acidogens and methanogens differ widely with respect to theirgrowth and substrate utilization rate, and sensitivity to environ-mental changes; for example, methanogens are relatively slower

K. Hwang et al. / Bioresource Technology 101 (2010) S2–S6 S3

growing and less stable against environmental variation than acido-gens. Unbalance between the rate of production and consumption ofVFAs usually results in the accumulation of acids, low pH, loss ofmethanogenic activities and digester failure in that order. It mustbe addressed that the major waste stabilization in terms of chemicaloxygen demand (COD) or biological oxygen demand (BOD) takesplace in the methanation phase by releasing CH4 and CO2 from themedium. Consequently, proper control of the methanogenic phaseas well as monitoring its diversity has been a key factor in the suc-cessful operation of most anaerobic processes (Raskin et al., 1994;Griffin et al., 1998; Angenent et al., 2002; Yu et al., 2006). However,the information on the changes in methanogenic diversity and pop-ulation undergoing a long period of starvation is lacking in literature.

Therefore, the aim of this study was to evaluate experimentallythe significant aspects of methanogenic population dynamics be-fore and after substrate-limited conditions in a batch reactor. Toidentify changes in the microbial profiles, methanogenic diversitybefore and after starvation were analyzed by denaturing gradientgel electrophoresis (DGGE). Quantitative real-time polymerasechain reaction (Q-PCR) was used to quantify the methanogen pop-ulations. The operational conditions were monitored by measuringthe degradation of VFAs, the production of methane gas, and thereduction in COD concentration.

2. Methods

2.1. Reactor operation

An anaerobic batch reactor with a working volume of 4 L wasoperated at the mesophilic temperature (35 �C), and the pH wasmaintained above 7.0 for the period of operation. Swine wastewa-ter, screened with a 600 lm sieve, was obtained from a local pigfarm. The wastewater was homogeneously mixed and frozen at�25 �C until analysis. Swine wastewater was used as the substrate.Physical and chemical characteristics of substrate are shown in Ta-ble 1. For the first reaction, the reactor was inoculated with 30% (v/v) anaerobic sludge with 24.7 g COD/L and 13.7 g of volatile sus-pended solids (VSS)/L obtained from a local domestic wastewatertreatment plant in Pohang, Korea. The first reaction was performed90 days until no more methane gas was produced. This anaerobicreactor was maintained under starved condition for 4 months.There was no external substrate supplied into the bioreactor. Toinvestigate the population dynamics of long time starved metha-nogens, the material remaining on day 205 after the starvation per-iod was used as the seed for the second reaction of the anaerobicreactor. The concentrations of COD and VSS in the material remain-ing were 21.2 g/L and 9.1 g/L, respectively. In the second reaction,the reactor was filled with 50% (v/v) swine wastewater as substrateand 50% (v/v) of the remaining material from starved conditionafter the first reaction. The bioreactor was operated for the tworeactions and one starvation periods over 10 months: the first

Table 1Characteristics of the swine wastewater used in this research.

Parameters Concentrations (g/L)

Total COD 124.1 (3.9)Soluble COD 64.4 (0.8)Total carbohydrate 7.7 (0.7)Soluble carbohydrate 2.4 (0.3)TKNa 7.5 (0.1)Total solids 65.6 (1.1)Volatile solids 47.8 (0.5)Total suspended solids 40.6 (0.3)Volatile suspended solids 34.6 (0.4)

Standard deviations are in parentheses.a Total Kjeldahl nitrogen.

reaction for 3 months, from day 0 to day 90; starvation for4 months, from day 91 to day 205; and the reaction after starvationfor 3 months, from day 206 to day 316.

2.2. Process data analysis

A gas chromatograph model 6890 Plus (Agilent, Palo Alto, CA)equipped with an Innowax capillary column and a flame ionizationdetector was used to determine the concentrations of C2–C6 VFAs.Another 6890 Plus gas chromatograph (Agilent), with an HP-5 cap-illary column and a thermal conductivity detector, was used toanalyze the gas composition of the biogas. Total Kjeldahl nitrogenwas measured according to the Kjeldahl method and the amount ofcarbohydrate was determined by phenol–sulfuric acid assay. TheCOD and solid concentrations were measured according to the pro-cedures in Standard Methods (APHA-AWWA-WEF, 2005). All anal-yses were performed in duplicate, and the results are given asmean values.

2.3. DNA extraction and amplification

DNA was extracted from the anaerobic reactor at each samplingtime using an automated nucleic acid extractor, Magtration System6GC (PSS, Chiba, Japan). To construct the DGGE profiles, archaeal16S rRNA gene-specific primers, GC-787F (5-CGCCCGCCGCGCC-CCGCGCCCGTCCCGCCGCCCCCGCCCGATTAGATACCCSBGTAGTCC-3)and 1059R (5-GCCATGCACCWCCTCT-3) were used to amplify a273-base pair (bp) fragment from the extracted DNA for the detec-tion of specific methanogenic groups (Yu et al., 2005). A 40-bp GCclamp was added to the 5 end of the forward primer and a touch-down PCR method was used with the following conditions: initialdenaturation at 94 �C for 10 min, followed by 20 cycles of denatur-ation at 94 �C for 1 min; annealing at a temperature that decreasedby 0.5 �C every cycle from 65 �C to the ‘touchdown’ at 55 �C,remaining at each temperature for 1 min; and chain extension at72 �C for 1 min. This was followed by 20 cycles of denaturationat 94 �C for 1 min, annealing at 55 �C for 1 min, and extension at72 �C for 1 min. A final extension was performed at 72 �C for3 min (Hwang et al., 2008).

2.4. DGGE analysis and DNA sequencing

The PCR products were loaded onto 8% polyacrylamide gels con-taining a range of 30–60% denaturant concentrations (100% denatur-ant was a mixture of 7 M urea and 40% [v/v] formamide). Each DGGEgel was run for 7 h at 150 V in 1 � TAE electrophoresis buffer withthe D-Code system (BioRad, Hercules, CA). Following electrophore-sis, the gel was stained with ethidium bromide solution for 20 min,rinsed for 20 min in deionized water (DW), and photographed underUV transillumination. The visible DGGE bands in each DGGE profilewere excised directly from the gels with a sterile blade, mixed with40 lL of DW, and incubated overnight at 4 �C. Each band was thuseluted into solution and 5 lL was used as the template in a reampli-fication reaction using the target primers. The PCR products werepurified on 1% agarose gel. The final products, the partial 16S rRNAsequences amplified with archaeal-specific primers were clonedinto the pGEM-T Easy vector (Promega, Mannheim, Germany). The16S rRNA gene inserts were sequenced with a 3730XL DNA Analyzer(Applied Biosystems, Foster City, CA). Database homology searchesfor these sequences were performed using the BLAST program inthe National Center for Biotechnology Information (NCBI) database.

2.5. Real-time quantitative PCR (Q-PCR)

Q-PCR was performed using the LightCycler 480 (Roche Diag-nostics, Mannheim, Germany) with primer and probe sets target-

S4 K. Hwang et al. / Bioresource Technology 101 (2010) S2–S6

ing the orders Methanomicrobiales and Methanosarcinales. The20 lL Q-PCR mixtures were prepared using the LightCycler 480Master Kit (Roche Diagnostics). The two-step amplification proto-col was as follows: initial denaturation for 10 min at 94 �C followedby 45 cycles of 10 s at 94 �C and combined annealing and extensionfor 30 s at 60 �C (Yu et al., 2005; Shin et al., 2008). The fluorescentsignal was measured at the end of each annealing and extensionstep.

3. Results

3.1. Performance of the anaerobic batch reactor

VFAs were not present and methane gas was not produced any-more at day 90. After 90 days, the production of methane gas wasobserved intermittently to confirm the methanogenic activity andthere was not detected further produced gas. This result indicatedthat the methanogenic activity for methane production was ar-rested due to substrate limitation. To maintain the state of metha-nogens influenced by long time starvation, the reactor was notprovided swine wastewater and other nutrients during about4 months. And then, the bioreactor started on the second reactionby providing new substrate from day 206.

The VFA and methane profiles in Fig. 1 show that the opera-tional efficiency of the first reaction was different from that ofthe second reaction. During the first reaction, the COD concentra-tion decreased by about 70% of that in the input swine wastewater,from 91 to 27 g/L, whereas that of the second reaction showed a63% reduction, from 70 to 25.8 g/L. Among the VFAs, acetate, themajor fermentation product, reached a concentration of 10,293

A

0

2

4

6

8

10

12

14

Acc

umul

ated

gas

pro

duct

ion

(L/L

)

0

10

20

30

40

50

60

70

starvation

starvation

Time (days)

0

2

4

6

8

10

12

14

Acc

umul

ated

gas

pro

duct

ion

(L/L

)

0

10

20

30

40

50

60

70

0 50 25 300 150 200 250 3000 100

Time (days)0 50 30150 200 250 300100

Vola

tile

fatty

aci

ds (g

/L)

0

2

4

6

8

10

12

14

0

10

20

30

40

50

60

70

B

Vola

tile

fatty

aci

ds (g

/L)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

Acc

umul

ated

gas

pro

duct

ion

(L/L

)

0

10

20

30

40

50

60

70

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

0

10

20

30

40

50

60

70

Fig. 1. Methane production according to the formation and degradation of volatilefatty acids for overall process period (about 10 months) including two reactions andstarvation periods in an anaerobic batch reactor treating swine wastewater. (A)Profiles of acetate, propionate, and n-butyrate, three major VFAs, and methaneproduction. (B) Profiles of i-butyrate, i-valerate, n-valerate, and n-carproate, minorVFAs, and methane production.

mg/L on day 14 and then decreased rapidly. In contrast, in the sec-ond reaction, the maximum acetate concentration was 11,327 mg/L on day 231 and then decreased slowly. It took a relatively longtime, about 81 days, for propionate, another major VFA, to beginto decrease in the second reaction, compared with 55 days in thefirst reaction. During the first reaction, 64 L of methane gas wasproduced, whereas 36 L of methane gas was produced during thesecond reaction.

3.2. PCR-based DGGE analysis of the domain Archaea

To investigate the microbial structures of the processes beforeand after starvation, we tried DGGE analysis about methanogenicgroups in each reaction period. DGGE banding profiles in Figs. 2and 3 show methanogenic compositions during the first and sec-ond reaction periods. To compare the distinct DGGE bands pattern,the 16S rRNA genes amplified on day 100 and day 130 were loadedonto both gels according to the time. Although there were manydiscernible and weak DGGE bands in the two DGGE profiles, themethanogens correlated with those bands that were consistentlyrepresented on the two DGGE gels (Figs. 2 and 3 and Table 2).The DGGE profiles show that bands 5 and 13 appeared with a dis-tinct pattern from day 11, when methane gas began to be pro-duced. Bands 5 and 13 are closely related to the methanogenicspecies Methanoculleus bourgensis, which uses H2/CO2 and formateas methanogenic substrates, and requires acetate for growth, withan optimal temperature of 40 �C and an optimal pH of 6.7–8.0(Asakawat and Nagaoka, 2003.). Band 16 was distinctly intensefrom day 25, when acetate, the major substrate of the order Meth-anosarcinales, decreased rapidly. This band is closely related toMethanosarcina acetivorans, an acetate-utilizing methanogen. Theoptimal conditions for its growth are a temperature of 40 �C andpH 5.5–8.0 (Sowers et al., 1984). DGGE bands 4 and 7 were ob-served in the swine wastewater. These bands are closely relatedto the methanogenic species Methanocorpusculum bavaricum. Theoptimal growth of this strain is in the mesophilic temperaturerange and at a pH of around 7, and it produces methane from H2/CO2, formate, and 2-propanol/CO2 (Zellner et al., 1989). In theDGGE profiles of Fig. 3, bands 20 and 22 were observed from day212, the starting point of methane gas production in the secondreaction. Associated with these two bands, band 21 detected inthe period of propionate reduction in the second reaction corre-sponds to M. bourgensis. Band 16 which was observed in the first

Fig. 2. Archaea DGGE banding profiles from an anaerobic bioreactor during firstreaction (S, swine wastewater; lane labels at the top show the sampling time [days]from start-up of the bioreactor; numbers on the DGGE gel indicate the bandsexcised for sequencing).

Fig. 3. DGGE banding profiles during second reaction (S, swine wastewater;previous 205 in gel means remained material before putting substrate; lane labelsat the top show the sampling time [days]; numbers on the DGGE gel indicate thebands excised for sequencing).

Table 2Identification of amplified 16S rRNA gene sequences excised from the DGGE gels.

DGGEband/s

Highest similarity GenBank accessionnumber

Homology in basepair (%)

4,7 Methanocorpusculumbavaricum

AY196676 266/271 (98%)

16 Methanosarcinaacetivorans

U89773 269/271 (99%)

Others Methanoculleusbourgensis

AB065298 270/271 (99%)

K. Hwang et al. / Bioresource Technology 101 (2010) S2–S6 S5

reaction appeared with the same pattern in the second reaction.From these DGGE profiles, we can infer that the methane gas ob-served in the overall process period was produced by the activityof M. bourgensis and M. acetivorans, belonging to the two differentorders Methanomicrobiales and Methanosarcinales, respectively.

3.3. Q-PCR analysis

The Q-PCR results show that the 16S rRNA gene copy number ofthe order Methanomicrobiales was relatively constant; in a rangefrom 3.4 � 107 copies/mL to 4.6 � 107 copies/mL during overallprocess period in both the first and second reaction. However,the 16S rRNA gene copy number of the order Methanosarcinalesincreased from 2.1 � 106 copies/mL on day 0 to 1.4 � 107 copies/mL on day 25, when acetate decreased sharply during the firstreaction. In the second reaction, the gene copy number increasedslightly from 5.1 � 106 copies/mL on day 206 to 5.7 � 106 copies/mL on day 259, the acetate-decreasing period. Therefore, in thefirst reaction, the increase in the gene copy number of Methanos-arcinales could explain the distinct increase in methane productionwith acetate utilization from day 21 to day 28, unlike the slow in-crease in methane production in the second reaction. Based on theQ-PCR results, it is inferred that Methanosarcinales affected theprocess results, including methane production and acetate degra-dation, as its gene copy number increased during the first and sec-ond reaction periods. Although the microbial diversity did notchange, as revealed by the DGGE profiles, the microbial concentra-tions differed in each reaction period.

4. Discussion

Konopka et al. (2002) also found a longer lag phase with in-creased starvation period of methanogenic growth. This finding

indicates that methanogenic activity is influenced by starvationperiod. The acetate profile in Fig. 1 implies that the acidogens pro-duced acetate during the early stage of the second reaction, whilethe methanogens experienced a lag phase, after the long period ofstarvation. These microbial and chemical data indicate that duringthe process, the acetate concentration was higher and the methaneproduction was lower in the second reaction than in the first reac-tion, attributable to differences in methanogenic activity andmicrobial concentrations. As a result of this study, microbial diver-sities were not shifted in each reaction period before and after star-vation. However, microbial quantification showed some changedaspects according to the reaction period generating formationand degradation performances of methane gas and VFAs duringthe first and second reactions. Therefore, important factors tocause the different efficiency of process results were methanogenicconcentration and activity not diversity.

These data provide insight into the diversity and recovery offunctionally important methanogens and processing performanceunder substrate-limited conditions in an anaerobic batch system.

5. Conclusions

This study has demonstrated the relationship between thetreatment processes and microbial dynamics before and after star-vation in an anaerobic reactor with a long starvation period. Themicrobial population diversity examined with DGGE was not themain factor affecting the different operational results. Cell num-bers of the order Methanosarcinales and methanogenic activitywere important factors determining the different efficiencies ofthe process results.

Acknowledgements

This research was supported in part by the BK-21, the Korea En-ergy Management Corporation (2006-N-BI-02-P-09-0-000-2007)programs, and the Korea Ministry of Environment (MOE) as HumanResource Development Project for Waste to Energy, ManpowerDevelopment Program for Energy and Resources by the Ministryof Knowledge and Economy (MKE).

References

Angenent, L.T., Sung, S., Raskin, L., 2002. Methanogenic population dynamics duringstartup of a full-scale anaerobic sequencing batch reactor treating swine waste.Water Res. 36, 4648–4654.

APHA-AWWA-WEF, 2005. Standard Methods for the Examination of Water andWastewater, 21st ed. American Public Health Association, Washington, DC.

Asakawat, S., Nagaoka, K., 2003. Methanoculleus bourgensis, Methanoculleusolentangyi and Methanoculleus oldenburgensis are subjective synonyms. Int. J.Syst. Evol. Microbiol. 53, 1551–1552.

Carrero-Colón, M., Nakatsu, C.H., Konopka, A., 2006. Effect of nutrient periodicity onmicrobial community dynamics. Appl. Environ. Microbiol. 72, 3175–3183.

Castellanos, T., Ascencio, F., Bashan, Y., 2000. Starvation-induced changes in the cellsurface of Azospirillum lipoferum. FEMS Microbiol. Ecol. 33, 1–9.

Cox, H.H.J., Deshusses, M.A., 2002. Effect of starvation on the performance and re-acclimation of biotrickling filters for air pollution control. Environ. Sci. Technol.36, 3069–3073.

Fernandes, L., 1994. Effect of temperature on the performance of an SBR treatingliquid swine-manure. Bioresour. Technol. 47, 219–227.

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.

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.

Hwang, K., Shin, S.G., Kim, J., Hwang, S., 2008. Methanogenic profiles by denaturinggradient gel electrophoresis using order-specific primers in anaerobic sludgedigestion. Appl. Microbiol. Biotechnol. 80, 269–276.

Jiang, W.Z., Kitamura, Y., Li, B., 2005. Improving acidogenic performance inanaerobic degradation of solid organic waste using a rotational drumfermentation system. Bioresour. Technol. 96, 1537–1543.

Kim, B.U., Won, C.H., Rim, J.M., 2004. Combined UASB reactor and DAF/BF/anoxic/aerobic process for the removal of high-concentration organic matter andnutrients from slurry-type swine waste. Water Sci. Technol. 49, 199–205.

S6 K. Hwang et al. / Bioresource Technology 101 (2010) S2–S6

Konopka, A., Zakharova, T., Nakatsu, C., 2002. Effect of starvation length uponmicrobial activity in a biomass recycle reactor. J. Ind. Microbiol. Biotechnol. 29,286–291.

Koutinas, M., Baptista, I.I.R., Peeva, L.G., Jorge, R.M.F., Livingston, A.G., 2006. The useof an oil absorber as a strategy to overcome starvation periods in degrading 1,2-dichloroethane in waste gas. Biotechnol. Bioeng. 96, 673–686.

Li, Z.H., Kuba, T.K., Kusuda, T., 2006. The influence of starvation phase on theproperties and the development of aerobic granules. Enzyme Microb. Technol.38, 670–674.

Liu, Y.Q., Tay, J.H., 2008. Influence of starvation time on formation and stabilityof aerobic granules in sequencing batch reactors. Bioresour. Technol. 99, 980–985.

Lopez, C., Pons, M.N., Morgenroth, E., 2006. Endogenous processes during long-termstarvation in activated sludge performing enhanced biological phosphorusremoval. Water Res. 40, 1519–1530.

Meuwissen, M.P.M., Van Asseldonk, M.A.P.M., Huirne, R.B.M., 2003. Alternative riskfinancing instruments for swine epidemics. Agric Syst. 75, 305–322.

Ndegwa, P.M., Hamilton, D.W., Lalman, J.A., Cumba, H.J., 2008. Effects of cycle-frequency and temperature on the performance of anaerobic sequencing batchreactors (ASBRs) treating swine waste. Bioresour. Technol. 99, 1972–1980.

Raskin, L., Poulsen, L.K., Noguera, D.R., Rittmann, B.E., Stahl, D.A., 1994.Quantification of methanogenic groups in anaerobic biological reactors byoligonucleotide probe hybridization. Appl. Environ. Microbiol. 60, 1241–1248.

Sanin, S.L., Sanin, F.D., Bryers, J.D., 2003. Effect of starvation on the adhesiveproperties of xenobiotic degrading bacteria. Process Biochem. 38, 909–914.

Shin, S.G., Lee, C., Hwang, K., Ahn, J., Hwang, S., 2008. Use of order-specific primersto investigate the methanogenic diversity in acetate enrichment system. J. Ind.Microbiol. Biotechnol. 35, 1345–1352.

Sousa, D.Z., Pereira, M.A., Smidt, H., Stams, A.J.M., Alves, M.M., 2007. Molecularassessment of complex microbial communities degrading long chain fatty acidsin methanogenic bioreactors. FEMS Microbiol. Ecol. 60, 252–265.

Sowers, K.R., Baron, S.F., Ferry, J.G., 1984. Methanosarcina acetivorans sp. nov., anacetotrophic methane-producing bacterium isolated from marine sediments.Appl. Environ. Microbiol. 47 (5), 971–978.

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

Wong, H.T., Cheng, S.C.S., Chan, E.W.C., Sheng, Z.T., Yan, W.Y., Zheng, Z.X., Xie, Y.,2000. Plasmids encoding foot-and-mouth disease virus VP1 epitopes elicitedimmune responses in mice and swine and protected swine against viralinfection. Virology 278, 27–35.

Yu, Y., Lee, C., Kim, J., Hwang, S., 2005. Group-specific primer and probe sets todetect methanogenic communities using quantitative real-time polymerasechain reaction. Biotechnol. Bioeng. 89, 670–679.

Yu, Y., Kim, J., Hwang, S., 2006. Use of real-time PCR for group-specificquantification of aceticlastic methanogens in anaerobic processes: populationdynamics and community structures. Biotechnol. Bioeng. 93, 424–433.

Zellner, G., Stackebrandt, E., Messner, P., Tindall, B.J., Macario, E.C., Kneifel, H., Sleytr,U.B., Winter, J., 1989. Methanocorpusculaceae fam. nov., represented byMethanocorpusculum parvum, Methanocorpusculum sinense spec. nov. andMethanocorpusculum bavaricum spec. nov. Arch. Microbiol. 151, 381–390.