Journal of Bioscience and BioengineeringVOL. 107 No. 1, 54–60, 2009

www.elsevier.com/locate/jbiosc

Thermophilic anaerobic co-digestion of garbage, screened swine anddairy cattle manure

Kai Liu,1 Yue-Qin Tang,1,§ Toru Matsui,2 Shigeru Morimura,1 Xiao-Lei Wu,3 and Kenji Kida1,⁎

⁎ CorrespondE-mail add

§ Present adEngineering, Pe

1389-1723/$doi:10.1016/j.

Graduate School of Science and Technology, Kumamoto University, 2-39-1 Kurokami, Kumamoto-City, Kumamoto 860-8555, Japan1 Fundamental ResearchInstitute, Tokyo Gas Co., Ltd., 1-7-7 Suehiro-Cho, Tsurumi-Ku, Yokohama-City, Kanagawa 230-0045, Japan2 and Department of Energy and Resources

Engineering, College of Engineering, Peking University, Beijing 100871, China3

Received 29 February 2008; accepted 11 September 2008

Methane fermentation characteristics of garbage, swine manure (SM), dairy cattle manure (DCM) and mixtures of thesewastes were studied. SM and DCM showed much lower volatile total solid (VTS) digestion efficiencies and methane yield thanthose of garbage. VTS digestion efficiency of SM was significantly increased when it was co-digested with garbage (Garbage:SM=1:1). Co-digestion of garbage, SM and DCM with respect to the relative quantity of each waste discharged in the Kikuchi (1:16: 27) and Aso (1: 19: 12) areas indicated that co-digestion with garbage would improve the digestion characteristic of SM andDCM as far as the ratio of DCM in the wastes was maintained below a certain level. When the mixed waste (Garbage: SM:DCM=1:19:12) was treated using a thermophilic UAF reactor, methanogens responsible for the methane production wereMethanoculleus and Methanosarcina species. Bacterial species in the phylum Firmicutes were dominant bacteria responsiblefor the digestion of these wastes. As the percentage of garbage in the mixed wastes used in this study was low (2–3%) and thedigestion efficiency of DCM was obviously improved, the co-digestion of SM and DCM with limited garbage was a prospectivemethod to treat the livestock waste effectively and was an attractive alternative technology for the construction of asustainable environment and society in stock raising area.

© 2008, The Society for Biotechnology, Japan. All rights reserved.

[Key words: Anaerobic digestion; Methane fermentation; Swine manure; Dairy cattle manure; Microbial community]

In Japan, approximately 300 million tones of biomass wastes aregenerated annually. Livestock wastes and garbage account for about20% and 5%, respectively. Severalmethods are used for the treatment oflivestockwaste, including field dispersal, composting, activated sludgeand methane fermentation. Most of garbage is treated by incinerationand composting, although methane fermentation is also employed insome areas. The use of methane fermentation for the treatment ofthese organic wastes has gained importance and increased in recentyears (1,2), due to some new regulations for the disposal of thesewastes and production of electricity from renewable energy sources.

Compared to the quantity of garbage generated, livestock waste isthe dominant biomass generated in stock raising area such as theKyusyu area in Japan. A proper treatment method of these wastes,particularly suited to stock raising areas, is important for sustainabledevelopment and environment protection. Methane fermentationseems to be a more suitable method compared to composting, sincethe demand on the amount of compost is very limited in these areasdue to the huge quantity of manure generated. Garbage (3–5) andmanure from livestock (6–9) for methane fermentation have beenextensively studied. Due to the very large biodegradable organic

ing author. Tel.: +81 96 342 3668; fax: +81 96 342 3669.ress: kida@gpo.kumamoto-u.ac.jp (K. Kida).dress: Department of Energy and Resources Engineering, College ofking University, Beijing 100871, China.

- see front matter © 2008, The Society for Biotechnology, Japan. Alljbiosc.2008.09.007

content of garbage, anaerobic digestion of this waste is very rapid andusually high loading rates with high VTS digestion efficiencies areeasily obtained even though treated in full scale (3,10). However,manure is known to have a poor methane yield and it is not commonto use manure as a sole substrate for biogas production on a largescale. The biodegradability of cattle manure is low, typically in therange of 30–43% yielding 150–240 l CH4/kg-volatile solid (VS) (6,7),while swine manure has a slightly higher biodegradability, in therange of 40–69%, potentially giving 280–360 l CH4/kg-VS (6,9). One ofthe possibilities to increase the biogas production from manure is toco-digest it with other organic wastes such as garbage from whichnecessary or inadequate nutrients are provided. Co-digestion is atechnology that is increasingly applied for simultaneous treatment ofseveral solid and liquid wastes. The main advantages of thistechnology are improved methane yield because of the supply ofadditional nutrients from codigestates and more efficient use ofequipment and cost-sharing by processingmultiplewaste streams in asingle facility (11–13). However, the percentage of garbage added forco-digestion reported till now is relatively high, from 10% to nearly100%. It is not possible and not realizable for most of stock raisingareas since the actual quantity of garbage generated in those areas isvery limited. The percentage of garbage generated in Kikuchi and Asoarea, famous stock raising areas in Kyushu, is only about 2–3%. Theeffect of garbage on co-digestion should be studied in a relatively lowpercentage range.

rights reserved.

TABLE 1. Characteristics of wastes used in the study

Parameter Garbage SM DCM Garbage:SM Garbage:SM:DCM

1:1 1:16:27 1:19:12

TS (g/l) 202.2 62.7 69.8 121.4 68.8 65.5VTS (g/l) 195.2 49.4 52.7 110.1 54.4 50.0SS (g/l) 158.4 38.6 47.4 74.2 46.5 37.1VSS (g/l) 154.4 33.6 36.9 69.9 38.2 34.9T-CODCr (mg/l) N.M. 89,000 71,000 162,000 N.M. 81,000S-TOC (mg/l) N.M. 19,451 9726 7825 4505 16,910S-IC (mg/l) N.M. 849 1926 158 1700 977T-VFA (mg/l) 1267 37,145 11,822 9560 8053 31,064Succinic acid N.D. 348 N.D. 237 N.D. 256Lactic acid N.D. 19,770 N.D. 3202 2087 16,980Acetic acid 452 7050 9261 3000 3827 7830Propionic acid 815 3525 1480 1257 1077 2282Butyric acid N.D. 3960 418 1869 490 2479I-valeric acid N.D. 1046 N.D. N.D. 275 571N-valeric acid N.D. 1446 663 494 N.D. 666

PO43- (mg/l) 120 625 50 N.M. N.M. 463

NH4+ (mg/l) 245 1382 1412 N.M. 806 1432

pH (−) 5.3 5.5 7.1 6.1 7.1 6.1Viscosity (cP) N.M. 200 4900 N.M. N.M. 280

SM: screened swine manure; DCM: screened dairy cattle manure; TS: total solid; VTS:volatile total solid; SS, suspended solid; VSS, volatile suspended solid: T-COD, totalchemical oxygen demand; S-TOC, soluble total organic carbon; S-IC, soluble inorganiccarbon; T-VFA, total volatile fatty acid; N.M., not measured; N.D., not detected. Ni2+ andCo2+ were added to each waste at final concentrations of 74 μg and 21 μg per g VTS,respectively.

CO-DIGESTION OF WASTE BIOMASS AND MICROBIAL DIVERSITY 55VOL. 107, 2009

In this study, the methane fermentation of garbage, swine manure(SM), dairy cattle manure (DCM), and mixed wastes of these threekinds of wastes, especially those reflecting the real relative quantity ofwastes generated in two representation areas, Kikuchi and Aso, arestudied. In addition, the microbial community responsible for thedigestion of thesemixedwastes is analyzed using 16S rRNA gene cloneanalysis.

Materials and methods

Garbage, screened swine manure (SM) and screened dairy cattle manure(DCM) Swine and dairy cattle manure were kindly provided by the Mashiki andKojima farm (Kumamoto, Japan). Synthetic garbage, screened swine and dairy cattlemanure were prepared as described (3,9). Synthetic garbage contained fruits,vegetables, meat and fish, and staple foods. These materials were mashed using amixer and tap water was added to obtain a final total solid (TS) concentration of 20%.

FIG. 1. Schematic diagram of an anaerobic digestion pro

Swine manure and dairy cattle manure were screed using a vibrating sieve (opening2.8 mm) liquid–solid separator. Table 1 shows the representative characteristics of theindividual wastes and mixture of them. Ni2+ and Co2+ were added to the waste to givefinal concentrations of 74 μg and 21 μg per gram VTS to enhance methane fermentationrate (14,15).

Anaerobic digestion of screened dairy cattle manure (DCM) As the viscosityof DCM was high (4900 cP), we used a continuous stirred tank reactor (CSTR) with aworking volume of 1.5 l to study the anaerobic digestion of DCM. Thermophilicanaerobic digestion sludge provided by Nishihara-Syouten Co. Ltd. (Kumamoto) wasused as seeds for starting up the reactor. DCM was supplied once a day using thedraw-and-fill method. The reactor was maintained at the temperature of 53 °C bycirculation of thermostated water through a water jacket. pH in the reactor was keptat 7.5–7.8 by adding 1N HCl solution using a peristaltic pump with a pH controller.The effect of VTS volumetric loading rate was studied by changing the volume of DCMfed into the tank.

Anaerobic digestion of the mixed wastes The digestion characteristics ofdifferent mixed wastes were studied under different VTS loading rates using an upflowanaerobic filter (UAF) reactor (working volume, 4 l) as described (16) (Fig. 1). Mixedwastes were fed once a day by the draw-and-fill method. The temperature and pHweremaintained at 53 °C and 7.5–7.8, respectively. Four sheets of nonwoven fabric material(each was 5 mm in thickness) were fixed in the reactor and served as the support formicroorganisms. Microorganisms accumulated in the support, when the mixed wastewith 1:19:12 (wet weight) of Garbage: SM: DCM was treated at a VTS loading rate of6 g/l·d, were analyzed by 16S rRNA gene clone analysis.

16S rRNA gene clone analysis The nonwoven fabric support from the UFAreactor treating mixed waste GSD (Garbage:SM:DCM=1:19:12) at the VTS loading rateof 6 g/l·d was sampled and the microbial community attached in the support wasstudied by 16S rRNA gene clone analysis (17,18). One arachaeal-16S rRNA gene library(GSDA library) and one bacterial-16S rRNA gene library (GSDB library) were constructedusing extracted community DNA and 19 clones from GSDA library and 21 clones fromGSDB library were sequenced with a CEQ8000 genetic analysis system (BeckmanCoulter, Fullerton, CA, USA). The phylogenetic analyses of these clones were carried outas described (16). The operational taxonomic unit (OTU) were designated GSDA01 toGSDA08 for clones of the GSDA library, and GSDB01 to GSDB16 for clones of the GSDBlibrary. The DDBJ/EMBL/GenBank accession numbers for the sequences of OTUs GSDA01to GSDB16 are AB425245 to AB425268.

Fluorescence in situ hybridization (FISH) A small cube (approximately 5 by 5by 5 mm) of the support on which microorganisms were attached, was cut from thesurface of the support fixed in the UFA reactor operated at the total organic carbon(TOC) loading rate of 6 g/l·d when mixed waste GSD (Garbage:SM:DCM=1:19:12) weretreated. FISH was performed as described (16).

Other analytical methods All of the following parameters of the culturesolution in reactors, except for TS, VTS, suspended solids (SS), and volatile suspendedsolids (VSS), were analyzed in supernatants obtained after centrifugation at 8000 ×g for10 min. TS, VTS, SS, and VSS were analyzed in accordance with standard methods (19).Soluble TOC and inorganic carbon (IC) were analyzed using a TOC auto analyzer (TOC-500; Shimadzu, Kyoto), according to the testing methods for industrial wastewater,JISK0102-1986 (19). Volatile fatty acids (VFAs) were analyzed as previously described(20). PO4

3−, NH4+ and total chemical oxygen demand (T-CODCr) were measured using the

HACH method (HACH Co., Loveland, CO, USA). The methane content of the biogas wasmeasured by gas chromatography using a thermal conductivity detector (TCD) (KOR-

cess using an upflow anaerobic filter (UAF) reactor.

TABLE 2. Performance parameters of anaerobic digestion when different wasteswere used

Parameter Garbage SM DCM Garbage:SW Garbage:SM:DCM

1:1 1:16:27 1:19:12

Reactor (°C) CSTR(53 oC)

UAFR(37 oC)

CSTR(53 oC)

UAFR(53 °C)

Maximal VTSloading rate(g/l·d)

8.0 15.0 8.0 12.0 8.0 10.0

Hydraulicretentiontime (d)

12.5 3.2 10.4 9.2 6.8 5.0

VTS digestionefficiency (%)

82 42 28 78 40 47

Methane contentin biogas (%)

50 58 59 50 58 58

Biogas yield(ml/g-VTS fed)

875 460 250 730 250 464

Methane yield(ml/g-VTS fed)

440 267 148 365 145 270

CSTR, continuous stirred tank reactor; UAFR, upflow anaerobic filter reactor.

56 LIU ET AL. J. BIOSCI. BIOENG.,

2G; GL Science, Tokyo) with a packed column (Porapack Q; GL Science). Hydrogensulfide was measured using Kitagawa precision gas detector tubes (Komyo Kitagawa,Tokyo).

RESULTS AND DISCUSSION

Anaerobic digestion of garbage, SM and DCM Previously, westudied the anaerobic digestion of synthetic garbage and SM (3,9). Asshown in Table 2, the maximal VTS loading rate of 8 g/l·d wasobtained when two-time diluted synthetic garbage was used as feedfor a CSTR operated at 53 °C. VTS digestion efficiency at the VTSloading rate was approximately 82% and methane yield was about440 ml/g-VTS fed. SM was mesophilic treated using a UAF reactor(Fig. 1), as described in Materials and methods. Even though a highermaximal VTS loading rate of 15 g/l·d was achieved, the VTS digestionefficiency was only half that of garbage. Methane yield at this VTSloading rate was approximately 267 ml/g-VTS fed (Table 2). Anexplanation for the different digestion efficiencies was the NH3

inhibition. However, the concentration of NH4+ was approximately

2300 mg/l at the VTS loading rate of 15 g/l·d when SM was treated,which was not high enough to result in strong repression effect onmethanogenic microorganisms as the pH in the reactor wascontrolled at 7.5 (21). Therefore, different biodegradability of thesetwo kinds of wastes in might be the reason for the different digestion

FIG. 2. Effect of VTS volumetric loading rate on treatment efficiency of screened dairy cattleclosed squares, VTS digestion efficiency; open triangles, TOC concentration; closed trianglescontent in biogas.

results, since garbage was considered to contain nutrient-balancedcomponents suitable for methane fermentation.

As the viscosity of DCM was higher than 4900 cP, UAF reactorwas not suitable for its treatment. We treated DCM at 53 °C usingthe same CSTR as that used for the treatment of garbage. Theoperational period at each VTS loading rate was maintained for twoto three weeks to obtain reliable data. Fig. 2 shows the effect of VTSloading rate on the treatment performance. Though the gasproduction was low, methane production happened at the VTSloading rate from 2 to 8 g/l·d and almost ceased at the VTS loadingrate at 10 g/l·d. VTS digestion efficiency was near 30% at 2–6 g/l d ofVTS loading rates and decreased to approximately 28% at the VTSloading rate of 8 g/l·d. Methane yield at the VTS loading rate of 8 g/ld was only 148 ml/g-VTS fed (Table 2). Though NH4

+ concentration inthe reactor increased to approximately 3300 mg/l when the VTSloading rate was increased to 6 and 8 g/l·d, it was not high enoughto result in serious NH3 inhibition (21–23). DCM seemed much moredifficult for methane fermentation compared to SM, not only for itsrelatively low VTS digestion efficiency but also for its low methaneproduction.

From the results of these wastes digested separately, garbage gavethe most satisfied methane fermentation while SM and DCM showedmuch lower fermentation efficiency, probably due to their lowerbiodegradability. Similar results for the treatment of SM and DCMwere reported by Møller et al. (6). They tried to improve the methaneyield of these wastes by thermal pretreatments and combination ofreactors operated at different temperatures (8,24).

As the biogas yield is very dependent on waste composition, co-digestion of two or more organic wastes proves to be an effectivetechnology for improving methane fermentation of wastes. One of themain advantages of this technology is the improved methane yielddue to the supply of additional nutrients from codigestates. This isespecially effective for the treatment of SM and DSM, whosebiodegradability need improvement. The degradation of SM andDSM can be supposedly increased by co-digestion with garbage.

Anaerobic digestion of different mixed wastesGarbage: SM=1:1 The co-digestion effect of garbage and SM

was investigated using the mixed waste with the same wet quantityof garbage and SM. As shown in Fig. 3A, VTS loading rate of 12 g/l·dseemed to be the maximal loading rate and a VTS loading rate of14 g/l·d lead to abrupt increases in TOC and VFA concentrations anda decrease in H2S concentration. The loading rate of 12 g/l·d was justbetween the maximal loading rates, 8 and 15 g/l·d, when garbageand SM was treated separately. However, the VTS digestionefficiency was kept high at all VTS loading rates studied, though a

manure during thermophilic anaerobic digestion. Open squares, CH4 content in biogas;, VFA concentration; star, NH4

+ concentration; open circles, gas yield; closed circles, H2S

CO-DIGESTION OF WASTE BIOMASS AND MICROBIAL DIVERSITY 57VOL. 107, 2009

lower VTS loading rate resulted in a higher digestion efficiency. TheVTS digestion efficiency at the VTS loading rate of 1–6 g/l·d keptabove 85%, which was higher than that when garbage was treatedseparately. Although it decreased to 82% and 78% when the VTSloading rate was increased to 10 and 12 g/l·d, respectively, it was stillnearing the values when garbage was treated alone and was almosttwo time that when SW was treated separately. NH4

+ concentrationin the reactor was approximately 3800 mg/l when the VTS loadingrate was between 4 to 12 g/l·d, which was higher that that when SMwas treated separately, also suggesting an enhanced digestionefficiency of SM when it was co-treated with garbage. The methaneyield at the VTS loading rate at 12 g/l·d was about 365 ml/g-VTS fed,somewhat higher than could be expected from the methane yield ofthe two types of wastes digested separately (Table 2). Feng et al.reported a successful pilot-plant treatment result for the co-digestion of garbage and swine wastes, where a higher loadingrate was obtained comparing that to when each waste was treatedseparately (25). Garbage is, therefore, an ideal codigestate for wasteswhich are difficult to be digested.

Garbage:SM:DCM=1:16:27 The co-digestion effect of garbage,SM and DCM was firstly investigated using the mixed waste withthe relative ratio of 1:16:27 (wet weight), which was in order toapproach the real mixture of these wastes discharged in the Kikuchi

FIG. 3. Effect of VTS volumetric loading rate on treatment efficiency ofmixedwastes during th(C), Garbage: SM: DCM=1:19:12. All symbols represent the same parameters as in Fig. 2.

zone, Kumamoto. Due to the low population in the Kikuchi zone, theratio of garbage in these wastes was relatively low, only 2.3%. Inaddition, as the number of cattle bred was much higher than that ofswine bred, the quantity of DCM was almost double that of SM andthe relative ratio of DCM in the mixed waste was about 61.0%. Theeffect of VTS loading rate on the digestion result is shown in Fig. 3B.The maximal VTS loading rate was 8 g/l·d and a VTS of 10 g/l dresulted in obvious increases in TOC and VFA concentrations. TheVTS digestion efficiency decreased with the increasing VTS loadingrate and it was 40% at the VTS loading rate of 8 g/l·d. Though 40%was higher than when DCM was treated separately and almost thesame when SM was treated separately, the methane yield of about145 ml/g-VTS fed was low and similar to the methane yield whenDCM was digested separately (Table 2). NH4

+ concentration in thereactor was approximately 2800 mg/l when the VTS loading ratewas from 4 to 8 g/l·d, which was lower than when the mixed waste(Garbage: SW=1:1) was treated, also suggesting a low digestionefficiency. The results indicated that too high a percentage of DCMin the mixed waste would make the methane fermentation difficult,probably due to its unbalanced components. Under this condition,the percentage of garbage should be increased or/and other organicwastes should be added to the mixture to make the ratio of DCMdecrease, thus suitable for methane fermentation.

ermophilic anaerobic digestion. (A), Garbage: SM=1:1; (B), Garbage: SM: DCM=1:16:27;

58 LIU ET AL. J. BIOSCI. BIOENG.,

Garbage: SM: DCM=1: 19: 12 The mixed waste with a relativeratio of 1:19:12 (wet weight), which was formulated to approximatethe real mixture of these wastes discharged in the Aso zone,Kumamoto, was studied. The quantity of SM discharged in this areawas much higher than that of DCM and quite different from that inthe Kikuchi zone where the quantity of DCM was much higher. Therelative ratio of DCM in the mixed waste was about 38.0%. Inaddition, the relative ratio between garbage and manure was 3.1%,which was almost similar to that in the Kikuchi area. As shown inFig. 3C, methane fermentation occurred normally under all VTSloading rates studied. A VTS digestion efficiency of 47% was achievedat the VTS loading rate of 10 g/l·d and was higher than those whenSM and DCM were treated separately (Table 2). NH4

+ concentration inthe reactor kept at 3000 mg/l when the VTS loading rate wasbetween 4 to 10 g/l·d, a little higher than that when SW was treatedseparately. TOC and VFA concentrations maintained stable in all VTSloading rate studied. Methane yield did not change obviously withthe VTS loading rate and was about 270 ml/g-VTS fed at the VTSloading rate of 10 g/l·d. This methane yield was almost the same aswhen SM was digested separately, though the mixed wastecontained up to 38.0% of DCM. By co-digestion with garbage andSM, the digestion efficiency of DCM was enhanced. The percentage ofDCM in the wastes was proved to be important for a favorableoutcome. However, in this study we did not investigate the suitableratio of DCM in mixed waste.

As a good performance was achieved when using mixed wastewith Garbage: SM: DCM=1:19:12, the microorganism communityresponsible for the digestion of these wastes was thereforeanalyzed when the reactor was operated at the VTS loading rateof 6 g/l·d.

16S rRNA gene analysis of microorganisms collected from thesupport All clones sequenced in the archaeal library (GSDA) wereaffiliated with the phylum Euryarchaeota. Among the clones analyzed,8 different sequences (OTUs) were found. OTU GSDA01 to 04 of 12clones had 97–100% sequences similarities to Methanoculleus thermo-philicus. OTU GSDA05 and GSDA06 were related to Methanoculleusbourgensis with a 97% sequence similarity. OTU GSDA07 and GSDA08of 3 clones were closely related toMethanosarcina thermophiliawith a98% sequence similarity. Methanogens, closely related to generaMethanoculleus and Methanosarcina, were predominant hydrogeno-trophic and aceticlastic methanogens. Methanogen closely related tothese two species had been previously detected as dominantmethanogens in other thermophilic reactors including one treatinggarbage (16,26,27), indicating they were common species in thermo-philic methane fermentation reactors, no matter what type of wasteswere treated.

Sixteen OTUs were found in the bacterial library (GSDB). FifteenOTUs (GSDB01 to GSDB15, 20 clones) were affiliated with the phylumFirmicutes and OTU GSDB16 (1 clones) were classified to the Candidatedivision OP9. As illustrated in Fig. 4, OTUs in the phylum Firmicutesshowed high diversities, and most of the clones (GSDB01 to GSDB10,15 clones) formed a distinct cluster (hereafter named GSD-dominantcluster) and did not closely relate to any pure-cultured species in thisphylum. Nine OTUs (GSDB01 to GSDB08, GSDB10) in GSD-dominantcluster were closely related to uncultured clones, LKB177, LKB36,LKB45 and LKB89, obtained from the leachate of a full-scalerecirculating landfill (28), with sequence similarities above 96%. OTUGSDB06 and GSDB10 were also closely related to uncultured clonesA55_D21_H_B_H04 (EF559060) and A55_D21_H_B_B02 (EF559067)obtained from a thermophilic anaerobic solid digestor with 99% and100% sequence similarities. By the result of BLAST, the closest pure-cultured species of all OTUs in GSD-dominant cluster were thermo-philic acetogens, Moorella glycerini and Moorella thermoacetica,though the sequence similarities were only 86–87%. It was likelythat the microorganisms represented as these OTUs were responsible

for the acetogenesis in the reactor. OTU GSDB12 to GSDB15 wereclassified in the genus Clostridium. OTU GSDB12 was clustered withClostridium species responsible for the degradation of cellulose andwas closely related to uncultured clone GZKB92 obtained from theleachate of a closedmunicipal solid waste landfill with a 96% sequencesimilarity (29). OTU GSDB15 was clustered with Clostridium glycoli-cum, a bacterium capable of fermenting various carbohydrates (30),and was closely related to uncultured clone A-3B (AY953232)obtained from a swine lagoon with a 99% sequence similarity. OTUGSDB13 was closely related to Clostridium butyricum (31), a butyrateproducing bacterium, with a 98% sequence similarity. OTU GSDB14was closely related to Clostridium ultunense (32), an acetate-oxidizingbacterium, with a 94% sequence similarity. OTU GSDB11 was closelyrelated to Lactobacillus amylophilus (33), a lactate producing bacter-ium, with a 92% sequence similarity. Microorganisms represented asOUT GSDB11–13 and GSDB15 could be considered responsible for theacidogenesis and OUT GSDB14 might be connected with the acetateoxidizing in the reactor.

Clones classified to the phylum Firmicutes were dominant, whichwas consistent with the results for other thermophilic reactors(16,26,34,35) in which bacteria affiliated to this phylum were alwaysfound to be predominant. Pure-cultivated species in this phylum arehighly diverse, including many bacteria responsible for anaerobichydrolysis, acidogenesis and acetogenesis, which would play animportant role in anaerobic digestion of solid organic wastes such asthe wastes used in this study.

By the result of clone analysis, in our reactor bacteria responsiblefor anaerobic hydrolysis, acidogenesis and acetogenesis, and hydro-genotrophic and aceticlastic methanogens were detected. We coulddeduced that high molecules contained in waste biomass, such ascellulose, were first degraded to low chain fatty acids and hydrogen,carbon dioxidate and acetate produced in the degradation processwere therefore transformed to methane by methanogens.

We also observed the inside of nonwoven supports packed in theUAF reactor by the FISH. Though the pictures were not shown,bacterial cells were found throughout the whole 5-mm-depthsupport and archaeal cells existed mainly in the upper 3 mm.Deducing from the shape of cells, most of the archaeal cells wereMethanosarcia-like. The spatial distribution of microorganisms wasquite similar to that when awamori distillery wastewater wastreated using a similar UAF reactor installed with the same supportmaterials (16). As the support was highly porous and had highmicroorganism-adsorption ability, the bacterial microorganismswere immobilized throughout the total 5-mm-depth regionobserved. However, Methanosarcina-like methanogens were favoredto stay in the relative outside region of the support, from whereproduced biogas could be easily released.

The anaerobic treatment of the mixed waste could be carried outstably even in high organic loading rates because of the high quantityof microorganism accumulated on the support.

CONCLUSION

Anaerobic digestion of garbage, swine manure (SM), dairy cattlemanure (DSM) and the thermophilic co-digestion of these wasteswas investigated under methane fermentation conditions. SM andDCM showed much lower VTS digestion efficiency and methane yieldthan that of garbage, while DCM showed the worst performance. VTSdigestion efficiency of SM was significantly enhanced when SM wasco-digested with garbage (Garbage: SM=1:1), though the methaneyield only increased a little. The addition of garbage in anaerobictreatment would make the methane fermentation of manure mucheasier and improve the performance. Co-digestion of garbage, SMand DCM with respect to the relative quantity of each wastedischarged in the Kikuchi and Aso areas indicated that even though

FIG. 4. Phylogenetic tree showing the genetic relationships among clones affiliated with the phylum Firmicutes (low G+C gram-positive bacteria). The tree was constructed by theneighbor-joining method, using partial sequences of the 16S rRNA gene. Numbers of clones with identical sequences are shown in parentheses. The bar represents two substitutionsper 100 nucleotide positions. Bootstrap probabilities are indicated at the branch nodes. The DDBJ/EMBL/GenBank accession numbers for reference strains and clones obtained in thisstudy are shown in parentheses.

CO-DIGESTION OF WASTE BIOMASS AND MICROBIAL DIVERSITY 59VOL. 107, 2009

the percent of garbage was low, about 2–3%, the co-digestion withgarbage improved the digestion characteristic of SM and DCM as faras the ratio of DCM was controlled under a certain level. When themixed waste (Garbage: SM: DCM=1:19:12) was treated using a UAFreactor, methanogens responsible for the methane production fromthese wastes was Methanoculleus and Methanosarcina species.Bacterial species in the phylum Firmicutes were dominant bacteriaresponsible for the digestion of these wastes. Porous support used inthis study successfully provided a good growth condition formicroorganism and therefore resulted in a good and stable wastestreatment performance. In conclusion, co-digestion of swine manureand dairy cattle manure with garbage of limited quantity was a

prospective method to treat them simultaneously and effectively andis therefore an attractive technology for the construction of asustainable society in stock raising areas.

References

1. Nishio, N. and Nakashimada, Y.: Recent development of anaerobic digestionprocesses for energy recovery from wastes, J. Biosci. Bioeng., 103, 105–112 (2007).

2. Kim, J. K., Oh, B. R., Chun, Y. N., and Kim, S. W.: Effects of temperature andhydraulic retention time on anaerobic digestion of food waste, J. Biosci. Bioeng.,102, 328–332 (2006).

3. Ikbal, Y. Q., Tang, T., Shigematsu, S., Morimura, K., and Kida: Methanogenicactivity and repression of hydrogen sulfide evolved during high rate thermophilic

60 LIU ET AL. J. BIOSCI. BIOENG.,

methane fermentation of municipal solid waste, Jpn. J. Water Treatment Biol., 39,17–24 (2003).

4. Hartmann, H. and Ahring, B. K.: Strategies for the anaerobic digestion of theorganic fraction of municipal solid waste: an overview, Water Sci. Technol., 53,7–22 (2006).

5. Park, Y., Hong, F., Cheon, J., Hidaka, T., and Tsuno, H.: Comparison of thermophilicanaerobic digestion characteristics between single-phase and two-phase systemsfor kitchen garbage treatment, J. Biosci. Bioeng., 105, 48–54 (2008).

6. Møller, H. B., Sommer, S. G., and Ahring, B. K.: Methane productivity of manure,straw and solid fractions of manure, Biomass and Bioenergy, 26, 485–495 (2004).

7. Ahring, B. K., Ibrahim, A. A., and Mladenovska, Z.: Effect of temperature increasefrom 55 to 65 °C on performance and microbial population dynamics of ananaerobic reactor treating cattle manure, Water Res., 35, 2446–2452 (2001).

8. Mladenovska, Z., Hartmann, H., Kvist, T., Sales-Cruz, M., Gani, R., and Ahring, B. K.:Thermal pretreatment of the solid fraction of manure: impact on the biogas reactorperformance and microbial community, Water Sci. Technol., 53, 59–67 (2004).

9. Liu, K., Tang, Y. Q., Fujimura, Y., Shigematsu, T., Morimura, S., and Kida, K.:Anaerobic digestion of screened swine waste with high concentrations of organicmatter and suspended solids using an upflow anaerobic filter reactor, Jpn. J. WaterTreatment Biol., 42, 93–98 (2006).

10. Kübler, H., Hoppenheidt, K., Hirsch, P., Kottmair, A., Nimmrichter, R., Nordsieck,H., Mücke, W., and Swerev, M.: Full-scale codigestion of organic waste, Water Sci.Technol., 41, 195–202 (2000).

11. Alatriste-Mondragón, F., Samar, P., Cox, H. H. J., Ahring, B. K., and Iranpour, R.:Anaerobic codigestion of municipal, farm, and industrial organic wastes: a surveyof recent literature, Water Environ. Res., 78, 607–636 (2006).

12. Sosnowski, P., Wieczorek, A., and Ledakowicz, S.: Anaerobic co-digestion ofsewage sludge and organic fraction of municipal solid wastes, Adv. Environ. Res., 7,609–616 (2003).

13. Dohányos, M., Zábranská, J., Kutil, J., and Jenícek, P.: Improvement of anaerobicdigestion of sludge, Water Sci. Technol., 49, 97–104 (2004).

14. Kida, K., Shigematsu, T., Kijima, J., Numaguchi, M., Mochinaga, Y., Abe, N., andMorimura, S.: Influence of Ni2+ and Co2+ onmethanogenic activity and the amountsof coenzymes involved in methanogenesis, J. Biosci. Bioeng., 91, 590–595 (2001).

15. Kida, K. and Sonoda, Y.: Influence of mineral nutrients on high performanceduring anaerobic treatment of distillery wastewater from barley shochu making,J. Ferment. Bioeng., 75, 235–237 (1993).

16. Tang, Y. Q., Fujimura, Y., Shigematsu, T., Morimura, S., and Kida, K.: Anaerobictreatmentperformanceandmicrobial populationof thermophilic upflowanaerobic filterreactor treating awamori distillery wastewater, J. Biosci. Bioeng.,104, 281–287 (2007).

17. Shigematsu, T., Tang, Y. Q., Kawaguchi, H., Ninomiya, K., Kijima, J., Kobayashi, T.,Morimura, S., andKida,K.:Effect of dilution rate on structure of a hemophilic acetate-degrading methanogenic community during continuous cultivation, J. Biosci. Bioeng.,96, 547–558 (2003).

18. Shigematsu, T., Era, S., Mizuno, Y., Ninomiya, K., Kamegawa, Y., Morimura, S.,and Kida, K.: Microbial community of a hemophilic propionate-degradingmethanogenic consortium in chemostat cultivation analyzed based on 16S rRNAand acetate kinase genes, Appl. Microbiol. Biotechnol., 72, 401–415 (2006).

19. Namiki, H.: Testing methods for industrial wastewater, JIS K0102–1986, JapaneseStandards Association, Tokyo, 1986.

20. Kida, K., Morimura, S., and Sonoda, Y.: Accumulation of propionic acidduring anaerobic treatment of distillery wastewater from barley-shochumaking, J. Ferment. Bioeng., 75, 213–216 (1993).

21. Ikbal, Tang, Y. Q., Fujimura, Y., Shigematsu, T., Morimura, S., and Kida, K.: Theaffecting factors for optimization of hemophilic aceticlastic methanogenesis, Jpn. J.Water Treatment Biol., 39, 189–197 (2003).

22. Angelidaki, I. and Ahring, B. K.: Thermophilic anaerobic digestion of livestockwaste: the effect of ammonia, Appl. Microbiol. Biotechnol., 38, 560–564 (1993).

23. Angelidaki, I. and Ahring, B. K.: Anaerobic thermophilic digestion of manure atdifferent ammonia loads: Effect of temperature, Water Res., 28, 727–739 (1994).

24. Nielsen, H. B., Mladenovska, Z., Westermann, P., and Ahring, B. K.: Comparisonof two-stage thermophilic (68 °C/55 °C) anaerobic digestion with one-stagethermophilic (55 °C) digestion of cattle manure, Biotechnol. Bioeng., 86, 291–300(2004).

25. Feng, C., Shimada, S., Zhang, Z., and Zhang, T.: A pilot plant two-phase anaerobicdigestion system for bioenergy recovery from swine wastes and garbage, WasteManage., 28, 1827–1834 (2008).

26. Tang, Y. Q., Shigematsu, T., Ikbal, Shigematsu, S., and Kida, K.: The effect of micro-aeration on the phylogenetic diversity of microorganisms in a thermophilicanaerobic municipal solid-waste digester, Water Res., 38, 2537–2550 (2004).

27. Sasaki, K., Haruta, S., Tatara, M., Yamazawa, A., Ueno, Y., Ishii, M., and Igarashi,Y.: Microbial community in methanogenic packed-bed reactor successfullyoperating at short hydraulic retention time, J. Bosci. Bioeng., 101, 271–273 (2006).

28. Huang, L.-N., Zhou, H., Zhu, S., and Qu, L.-H.: Phylogenetic diversity of bacteria inthe leachate of a full-scale recirculating landfill, FEMS Microbiol. Ecol., 50, 175–183(2004).

29. Huang, L.-N., Zhu, S., Zhou, H., and Qu, L.-H.: Molecular phylogenetic diversity ofbacteria associated with leachate of a closed municipal solid waste landfill, FEMSMicrobiol. Lett., 242, 297–303 (2005).

30. Gaston, L. W. and Stadtman, E. R.: Fermentation of ethylene glycol by Clostridiumglycolicum, sp, N. J. Bacteriol., 85, 356–362 (1963).

31. Saint-Amans, S., Girbal, L., Andrade, J., Ahrens, K., and Soucaille, P.: Regulation ofcarbon and electron flow in Clostridium butyricum VPI 3266 grown on glucose-glycerol mixtures, J. Bacteriol., 183, 1748–1754 (2001).

32. Schnürer, A., Schink, B., and Svensson, B. H.: Clostridium ultunense sp. nov., ahemophilic bacterium oxidizing acetate in syntrophic association with a hydro-genotrophic methanogenic bacterium, Int. J. Syst. Bacteriol., 46, 1145–1152 (1996).

33. Nakamura, L. K. and Crowell, C. D.: Lactobacillus amylophilus, a new starch-hydrolyzing species from swine waste-corn fermentation, Dev. Ind. Microbiol., 20,532–540 (1979).

34. Crowell, l., Eriksson, A. R. B., and Schnurer, A.: Effect of process temperature onbacterial and archaeal community in two methanogenic bioreactors treatingorganic household waste, FEMS Microbiol. Ecol., 59, 683–693 (2007).

35. Schnurer, K., Haruta, S., Ueno, Y., Ishii, M., and Igarashi, Y.: Microbial populationin the biomass adhering to supporting material in a packed-bed reactor degradingorganic solid waste, Appl. Microbiol. Biotechnol., 75, 941–952 (2007).