Microbial community analysis of a biogas-producing completely stirred tank reactor fed continuously with fodder beet silage as mono-substrate

ARTICLE IN PRESS

Systematic and Applied Microbiology 30 (2007) 139–151 www.elsevier.de/syapm

Microbial community analysis of a biogas-producing completely stirred tank reactor fed continuously with fodder beet silage as mono-substrate Michael Klockea,, Pia Ma¨hnerta,b, Kerstin Mundta, Khadidja Souidia,b, Bernd Linkea,b a

Leibniz-Institut fu¨r Agrartechnik Potsdam-Bornim e.V., Abteilung Bioverfahrenstechnik, Max-Eyth-Allee 100, D-14469 Potsdam-Bornim, Germany b Humboldt-Universita¨t zu Berlin, Landwirtschaftlich-Ga¨rtnerische Fakulta¨t, Institut fu¨r Pflanzenbauwissenschaften, Fachgebiet Agrartechnik, Philippstrasse 13, D-10115 Berlin, Germany

Abstract The bioconversion of renewable raw material to biogas by anaerobic microbial fermentation processes in completely stirred tank reactors (CSTR) is a valuable alternative resource of energy especially for rural areas. However, knowledge about the microorganisms involved in the degradation of plant biomass is still poor. In this study, a first analysis of the biogas-forming process within a CSTR fed continuously with fodder beet silage as mono-substrate is presented in the context of molecular data on the microbial community composition. As indicated by the conventional process parameters like pH value, content of volatile fatty acids, N:P ratio and the biogas yield, the biogas-forming process within the CSTR occurred with a stable and efficient performance. The average biogas yield based on volatile solids was 0.87 m3 kg1 at an organic loading rate of 1.2–2.3 kg m3 d1. This amounts to 94% of the theoretical maximum. In order to identify microorganisms within the CSTR, a 16S rDNA clone library was constructed by PCR amplification applying a prokaryote-specific primer set. One hundred and forty seven clones were obtained and subsequently characterized by amplified rDNA restriction analysis (ARDRA). The sequences of 60 unique ARDRA patterns were estimated in a length of approximately 800–900 bp each. Four of them were assigned to the domain Archaea and 56 to the domain Bacteria. Within the domain Archaea, all clones showed a close relationship to methanogenic species. Major bacterial groups represented in the clone library were the class Clostridia of the phylum Firmicutes (22% of all 16S rDNA clones), the class Deltaproteobacteria of the phylum Proteobacteria (24%), the class Bacilli of the phylum Firmicutes (22%) and members of the phylum Bacteroidetes (21%). Within these major groups, the highest biodiversity was found within the class Clostridia (35% of all operational taxonomic units). Members of the phyla Actinobacteria and Spirochaetes were represented only by 5 and 2 clonal sequences, respectively. r 2006 Elsevier GmbH. All rights reserved. Keywords: 16S rDNA clone library; Amplified 16S rDNA restriction analysis; ARDRA; Bioconversion; Biogas; CSTR; Energy crops

Introduction Corresponding author.

E-mail address: [email protected] (M. Klocke). 0723-2020/$ - see front matter r 2006 Elsevier GmbH. All rights reserved. doi:10.1016/j.syapm.2006.03.007

In Germany as well as in many other countries, research on biogas production as a renewable source of energy has increased in recent times due to strong public

ARTICLE IN PRESS 140

M. Klocke et al. / Systematic and Applied Microbiology 30 (2007) 139–151

interest in protecting climate and environment. Besides the digestion of slurry, the methane production from crops as co- and mono-substrate is also becoming more important. Major ‘‘energy crops’’ are predominantly grain crops like maize, green forage plants like fodder grass and root crops like forage beet. In order to be independent of harvesting, these crops are mostly applied as silage either made from entire crops or parts of crops. The use of forage beet for biogas production is of particular interest because of its high biomass yield per hectare and the easy handling of the ensiled beet [46]. Furthermore, fodder beet possesses the highest energy concentration of all fodder plants. Thus, the application of fodder beet for biogas production could decrease the dependence of agronomists on costly organic waste as substrate [39]. However, until now biogas production based on the fermentation of fodder beet silage has not been investigated sufficiently. Moreover, the important role of process balance and the activity of anaerobic digestion is mostly described by means of volatile fatty acids [1,32], pH value [45] and ammonia concentration [24]. Beside these chemical parameters, knowledge about the composition of the microbial community participating in the degradation process of plant biomass to organic acid and finally to methane, hydrogen and carbondioxide, will also be crucial to understanding the biogas-forming process. In most cases, classical microbiological techniques for studying microbial biodiversity are of limited use due to the specialized environmental requirements of certain Archaea and Bacteria species especially if syntrophic organisms can be expected [3,12]. Thus, in addition to the traditional isolation, several cultivation-independent approaches based on the detection of differences within the microbial genome were established. Fingerprinting techniques like denaturating gradient gel electrophoresis (DGGE) or terminal restriction fragment length polymorphisms (T-RFLP) are commonly applied in various micro-ecological studies. Additionally, techniques based on the cloning and subsequent sequencing of selected DNA regions like the gene for the small subunit ribosomal RNA (16S rDNA) [3] or, alternatively, metabolism-specific genes like the mrcA gene for methanogenic organisms were described [28]. Such DNA libraries are often screened with fingerprint techniques like amplified 16S rDNA analysis (ARDRA) [23], which enables a reduction of sequencing efforts. Molecular approaches facilitate the characterization of microbial consortia as well as the monitoring of changes in microbial diversity [e.g. 11]. They are also well suited to uncovering new groups of microorganisms [eg. 35]. In contrast to granular sludge reactors like UASB reactors, relatively few studies are available regarding the microbial diversity in biogas-producing anaerobic completely stirred tank reactors (CSTR) (for review:

[19]). Within UASB reactors, the degradation process is performed in visible units of certain microbial species, the granules. They provide a close spatial orientation of the biogas-forming syntrophic microbial community [26]. Within CSTR, such granules do not appear, especially if the CSTR are loaded with high amounts of organic biomass like herbal raw material. Recent studies on the microbiology within CSTR analysed reactors processing manure [17] or sewage sludge [33]. Reports on the microbial diversity in reactors fed with defined substrates were also published [e.g. 41,42]. However, to our knowledge the microbiology within an anaerobic plant-biomass utilizing CSTR has not been described up to now. This study presents results of semi-continuous experiments to determine the biogas yield from fodder beet silage in a CSTR. To study the diversity of microorganisms involved in the biogas process, a 16S rDNA library was constructed. Subsequently, the 16S rDNA library was analysed by ARDRA combined with sequencing individual clones representative for particular ARDRA patterns. A phylogenetical analysis was performed to reveal similarities to already cultured microorganisms.

Material and methods Operation of CSTR The lab-scale experiment regarding the anaerobic digestion of fodder beet silage was carried out in a completely stirred tank reactor (CSTR) with a total volume of 8 (Fig. 1). Over a period of 19 weeks the reactor was operated in a daily fill and draw mode and mixed slowly by a stirrer. A constant temperature of 35 1C was maintained by a water jacket heated from a thermostat. The biogas was stored in a gas bag and analysed automatically once a day by a gas meter (Ritter, Bochum, Germany) and a biogas analyser (Pronova, Berlin, Germany).

Fig. 1. Scheme of the CSTR: (1) thermostat for water jacket, (2) effluent, (3) reactor content, (4) feed, (5) stirrer, (6) gas bag, (7) magnetic valve, (8) gas meter, and (9) biogas analyzser.

ARTICLE IN PRESS M. Klocke et al. / Systematic and Applied Microbiology 30 (2007) 139–151

141

Table 1. Chemical composition of fodder beet silage and CSTR effluent: pH, total solids TS, volatile solids VS and volatile fatty acids VFA Substrate

PH (dimension less)

TS105 (%)

VS (% TS)

VFA (g kg1)

Fodder beet silage Effluent (average)

3.670.1 7.870.2

13.870.3 3.270.1

92.770.6 60.373.8

10.2 1.870.4

At the start of the experiment the reactor was filled once with anaerobic digested manure from previous experiments with fodder beet silage and cattle slurry. Afterwards the fodder beet silage was applied as monosubstrate without slurry. Beginning with an organic loading rate (OLR) of 0.3 and 1.0 kg VS m3 d1 in the first and second week, respectively, the daily input was increased until an OLR of 2.3 kg VS m3 d1 was reached. The fodder beet (Beta vulgaris L., var. Kyros) was ensiled at laboratory scale. Subsequently, the silage was analysed 10 times by total solids (TS) at 105 1C (13.8%) and volatile solids (VS) at 550 1C (92.7% TS), once by volatile fatty acids VFA (10.2 g kg1) and five times by pH (3.6), nitrogen N analysed according to Kjeldahl (1.7 g kg1), crude protein XP (7.5% TS), crude fibre XF (7.1% TS), crude fat XL (0.4% TS) and saccharide (31.2% TS) according to standard methods (Table 1). The reactor effluent was measured weekly for pH, TS, VS, VFA, N and P.

0.7 M NaCl and 1% (w/v) CTAB [30]. The samples were incubated at 65 1C for 10 min. Depending on the grade of brownish colouring of the samples, one or two extraction steps were performed by adding equal volumes of a chloroform–isopentylalcohol mixture (24:1 v/v). Subsequently, the genomic DNA was precipitated by adding 0.25 volumes 3 M sodium–acetate and one volume isopropanol. To ensure a complete precipitation, the mixture was stored at 20 1C over night. The precipitated DNA was recovered by centrifugation at 20,800 g1 for 10 min, washed twice with 70% (v/v) ethanol, resuspended in 10 mM tris/HClbuffer (pH 8.0) and stored at 4 1C. All chemicals were provided by AppliChem, Darmstadt, Germany. DNA preparations were separated by electrophoresis in a 0.8% agarose gel in tris–acetate–EDTA buffer and quantified visually under UV light through staining with ethidium bromide and compared with standards of known length [37].

PCR amplification

DNA extraction All microbiological analyses were performed with samples taken from the effluent of the CSTR after 19 weeks of operation. For the extraction and subsequent purification of prokaryotic DNA, a modified protocol based on the protocol published by Rheims and Stackebrandt [35] was applied. Two millilitre samples were centrifuged for a short time at 1000 g1 to separate the undigested plant material from the aqueous supernatant containing the microorganisms. To collect the microorganisms, the supernatants were centrifuged at 20,800 g1 for 10 min. The microorganisms were resuspended in 1300 ml of saline-EDTA buffer (0.1 mol l1 EDTA, 0.15 mol l1 NaCl) containing approximately 30–40 mg of acid-washed PVPP. For enzymatic lysis of cells, 20 ml of a lysozyme solution (10 mg ml1) were added. After incubation at 37 1C for 30 min, 20 ml of a 1% (w/v) proteinase K solution and 100 ml of a 10% (w/ v) SDS were added. Subsequently, the samples were incubated at 65 1C for 45 min. To ensure the complete lysis of cells, three freeze–thaw steps were performed through freezing the probes in liquid nitrogen and thawing at 65 1C in a water bath. After centrifugation at 7500 g1 for 10 min, the supernatant was adjusted to

All PCR amplifications were performed with a Biometra T gradient 96 (Whatman Biometra, Go¨ttingen, Germany). The reaction mixture was set up on ice as follows: 10 ng template DNA, 1* PCR buffer, 1.5 mmol l1 MgCl2, 0.2 mmol l1 forward primer, 0.2 mmol l1 reverse primer, 0.2 mmol l1 of each dNTP, 0.8 U Taq-DNA-Polymerase (recombinant), ad 20 ml final volume bi-distilled H2O. All chemicals and enzymes were provided by Fermentas (St. Leon-Rot, Germany). The following primer pair specific for the bacterial and archaeal 16S rDNA gene was used: as forward primer, the primer U341F 50 -CCTACGGGRSGCAGCAG-30 (341–358 bp Escherichia coli position) [16] was applied. The primer 16Srev 50 -TACGGYTACCTTGTTACGACTT-30 (1,488–1,509 bp E. coli position) modified after Sekiguchi et al. [41] was used as reverse primer. The PCR conditions were: (1) initial denaturation at 94 1C for 120 s, (2) cycle denaturation at 94 1C for 30 s, (3) cycle annealing at 53 1C for 30 s, (4) elongation at 70 1C for 60 s, (5) final extension at 70 1C for 900 s and (6) hold at 4 1C. Steps (2)–(4) were repeated 25 times. The PCR product was separated by gel electrophoresis on a 0.8% agarose gel in tris/acetate-buffer and

ARTICLE IN PRESS 142

M. Klocke et al. / Systematic and Applied Microbiology 30 (2007) 139–151

analysed by staining with ethidium bromide under UV light [37]. The band of the expected size (approximately 1100 bp) was cut off and purified with a commercial kit (Qiaquick Gel Extraction Kit, Qiagen, Hilden, Germany). To reduce the bias, the amplification product of six single PCR reactions were pooled as recommended by several authors [e.g. 6].

eSnap version 6.00.20 and Syngene GeneTools version 3.04.00 (Synoptics). The diversity of the 16S rDNA clone library was analysed by a rarefaction analysis with the software Analytic Rarefaction 1.3 [20] as described by Ravenschlag et al. [34].

Sequencing and phylogenetic analysis Cloning and amplified rDNA restriction analysis (ARDRA) The purified amplicons were cloned into the pGEM-T plasmid (Promega, Mannheim, Germany) via TA cloning and subsequently transformed into JM109 competent cells (Promega) according to manufacturers guidelines. Clones were denominated as ATB-KM followed by a consecutive number. Recombinant plasmids were purified applying a commercial kit (E.N.Z.A. Plasmid Miniprep Kit I, Peqlab Biotechnologie, Erlangen, Germany) and tested with restriction digest (NcoI/SalI double digest according to manufacturers guidelines) (Fermentas). From plasmids containing an insert of expected length, a PCR was performed with a Biometra T gradient 96 (Whatman Biometra). The reaction mixture was set up on ice as follows: 1 ml plasmid DNA, 1  PCR buffer, 1.5 mmol l1 MgCl2, 0.2 mmol l1 forward primer, 0.2 mmol l1 reverse primer, 0.2 mmol l1 of each dNTP, 0.8 U Taq-DNA-Polymerase (native), ad 20 ml final volume bi-distilled H2O. All chemicals and enzymes were provided by Fermentas. As primers, standard primers targeting the T7 and SP6 promoter region located on the pGEM-T plasmid were used (T7 promoter primer 50 -TAATACGACTCACTATAGGG-30 , SP6 promoter primer 50 -ATTTAGGTGACACTATAG-30 ). The PCR conditions were: (1) initial denaturation at 94 1C for 120 s, (2) cycle denaturation at 94 1C for 30 s, (3) cycle annealing at 47 1C for 60 s, (4) elongation at 70 1C for 120 s, (5) final extension at 70 1C for 600 s and (6) hold at 4 1C. Steps (2)–(4) were repeated 30 times. Two microlitre of the PCR product were tested by gel electrophoresis on a 1.2% agarose gel in tris–acetate-buffer and analysed by staining with ethidium bromide under UV light [37]. Only PCR products of desired length were included into subsequent ARDRA. Eighteen microlitre of the PCR product were digested by addition of 1.5 U BsuRI and 1.5 U Hin6I (Fermentas) and incubation at 37 1C for 12 h according to manufacturers guidelines. Subsequently, the reaction mixture was analysed by gel electrophoresis on a 3.5% metaphor agarose gel (Cambrex Bioscience, Rockland, USA) in tris/acetate-buffer and subsequent ethidium bromide staining. The ARDRA patterns were documented with a Syngene GeneGenius Bio Imaging System (Synoptics, Cambridge, UK) applying the software Syngene Gen-

After a visual comparison of ARDRA patterns, clones representative for individual ARDRA patterns were selected and sequenced using the T7 promoter region as starting point. The sequencing was performed by MWG Biotech (Ebersberg, Germany). The length of resulting sequences was about 800 up to 900 bp (of approximately 1100 bp cloned). Sequences were used as operational taxonomic units (OTU) and denominated accordingly to the clone. All sequences were checked for chimeric artifacts by the Chimera Check software tool [9]. Sequences were compared with NCBI GenBank [5] entries using the nucleotide–nucleotide BLAST [2]. This software was also used to calculate the identity value. Alignment of clonal sequences and sequences from selected reference species or uncultured clones from the NCBI GenBank were performed with the software ClustalW [8] using standard settings. Phylogenetic trees were constructed by the neighbour-joining method [36]. Bootstrap resampling analysis [10] for 100 replicates was performed to estimate the confidence of tree topologies. Except for alignment, all phylogenetic and molecular evolutionary analyses were conducted using MEGA version 3.0 [22]. All estimated sequences were deposited in the NCBI GenBank database with following accession numbers: uncultured archaeon clones DQ390259 to DQ390262, uncultured bacterium clones DQ390263 to DQ390318. The taxonomic system of Bergey’s Manual was used as base for all phylogenetic analyses [13].

Results CSTR performance The weekly reactor performance data for VS-biogas yield yB and organic loading rate OLR is shown in Fig. 2 with the N and P content during 19 weeks of the CSTR’s semi-continuous operation. The biogas yield was obtained between 0.80 and 0.93 m3 kg1 VS with an average of 0.87 m3 kg1 VS. The production of biogas during the experiment was nearly constant and independent of the chosen OLR, which increases from 1.2 to 2.3 kg m3 d1. During the experiment the contents of N and P decreased from 2.9 to 2.1 g kg1 and 0.49 to 0.38 g kg1, respectively.

ARTICLE IN PRESS M. Klocke et al. / Systematic and Applied Microbiology 30 (2007) 139–151

Table 2. CSTR

143

Distribution of 16S rDNA clones detected in the No. of OUT

Fig. 2. Performance data for semi-continuous biogas production from fodder beet silage: biogas yield (-~-), OLR (&) and content of N (-m-) and P (-K-).

Domain Archaea Phylum Euryarchaeota Domain Bacteria Phylum Proteobacteria Phylum Firmicutes Class Clostridia Class Bacilli Phylum Actinobacteria Phylum Spirochaetes Phylum Bacteroidetes

No. of 16S rDNA clones

4 (6.7%)

9 (6.1%)

8 (13.3%)

35 (23.8%)

21 11 1 2 13

(35.0%) (18.3%) (1.7%) (3.3%) (21.7%)

33 32 5 2 31

(22.4%) (21.8%) (3.4%) (1.4%) (21.1%)

Corresponding to the VS-content of the fodder beet silage, the biogas yield based on fresh matter resulted in 0.11 m3 kg1. On average, the produced biogas contained 55.2% methane, 42.9% carbon dioxide and 12 ppm hydrogen sulphide. Consequently, the methane yield was about 0.47 m3 kg1 VS. Table 1 presents the differences between the chemical composition of the fodder beet silage and the effluent of the biogas process. The amounts were calculated as averages and standard deviations from the analysis of the silage and the 19 samples of the weekly effluent.

Overall phylogenetic analysis In total 147 clones of the 16S rDNA clone library were analysed by ARDRA. Among these clones, 60 different ARDRA patterns were detected and taken as operational taxonomic units (OTU). After estimating the DNA sequence of clones representative for unique fingerprints, a large-scale phylogenetic analysis with these operational taxonomic units (OTU) was performed to affiliate the estimated clonal sequences to hitherto-determined groups (Table 2). From this analysis, 9 (i.e. 6%) of total 16S rDNA clones were affiliated with the domain Archaea and 137 clones (i.e. 94%) were assigned to the domain Bacteria. The major groups in the 16S rDNA clone library were the Firmicutes (44% of the total clones), the Bacteroidetes (21%) and the Proteobacteria (24%). Beside these major groups members of the Actinobacteria (3%) and Spirochaetes (1%) were also detected. No clonal sequences were detected as chimeric artifact. Within the Firmicutes, 22% of the clones were assigned to the class Clostridia and 22% to the class Bacilli. Within the Proteobacteria, all clones were affiliated to the class Deltaproteobacteria, no clonal sequences belonging to the other classes of the Proteobacteria were detected.

Fig. 3. Rarefaction curve for the different ARDRA patters of 16S rDNA clones. The dashed lines represent 95% confidence intervals [20].

A rarefaction analysis was carried out to evaluate the adequacy of the sample size for the determination of diversity within the 16S rDNA clone library. As shown in Fig. 3, the calculated rarefaction curves did not reach a clear saturation. This indicates that the analysis of 147 clones in total had only partially covered the diversity.

Domain archaea Four OTU of the 16S rDNA library were assigned to the domain Archaea (Fig. 4). All of these clones were affiliated to the phylum Euryarchaeota, and all clones were closely related to the methanogenic archaea. Two clonal sequences, ATB-KM1219 and ATB-KM1253, were assigned to the order Methanosarcinales with a close relationship (identity values between 97% and 98%) to Methanosarcina acetivorans C2A, Methanosarcina mazei DSM 9195 and to Methanosarcina barkeri

ARTICLE IN PRESS 144

M. Klocke et al. / Systematic and Applied Microbiology 30 (2007) 139–151

Fig. 4. Phylogenetic tree of the clones among the phylum Euryarchaeota of the domain Archaea based on neighbour-joining analysis [65] of partial 16S rDNA sequences. Numbers at nodes represent bootstrap values [21] for the nodes in percent (100 times resampling analysis). Only bootstrap values above 50% were displayed. Numbers in brackets indicate the NCBI genbank accession number of 16S rDNA sequence of reference strains. If obtainable, type strains were used as reference. Sulfolobolus acidocaldarius DSM 639 was used as the outgroup (marked by an asterisk).

DSM 800. High similarities were also obtained to a range of clones derived from environmental samples like landfill leachate (clone GW70-10-1) [44] or biodegraded oil reservoirs (clone PL-36A11) [AY570657] as examples. Beside these, similar OTU were also found in methanogenic enrichment cultures (clone ALU04) [41]. A third OTU, ATB-KM1362, was also assigned to the order Methanosarcinales, but within the genus Methanosaeta. ATB-KM1362 revealed an identity value of 93% to Methanosaeta concilii. Highest identity values between 99% and 100% were obtained in comparison with uncultured archaeon clones derived from environmental samples like landfill leachate (clone GW70-20-23 [44] as example) or in sludge-fed methanogenic USAB bioreactors (clone CM21 [7]). One further clonal sequence, ATB-KM1251, was affiliated to the order Methanobacteriales. ATBKM1251 was assigned to a comparatively distinct group together with other uncultured archaeon clones like clone GZK72, clone KuA13 and clone 69-1 with identity values of 99–100%. These clones were also derived from anaerobic environments like landfill leachate and oil or water reservoirs, respectively. The closest related cultured strain was Methanobacterium formicium DSM 1535 [AF169245] with an identity value of 86%.

Domain bacteria (I) Deltaproteobacteria A range of clonal sequences was assigned to the class Deltaproteobacteria (Fig. 5). From 8 clonal sequences, most of them (5) were affiliated as unclassified Deltaproteobacteria, together with other uncultured environmental bacterium clones. These clonal sequences were closely related to the clones BSA2B-20, obtained from the analysis of a protein-degrading bioreactor [43] and GZKB79, derived from landfill leachate and showed an identity value between 99% and 100%. Further uncultured bacteria showed lower similarities, like clone TANB52a prepared out of a sample of a dechlorinating microbial community [29], clone PL-21B5 from a sample of an oil-degrading consortium [AY570604] or clone W22 isolated from a water injection dwell of an oilfield [AY770692]. This group was related to the order Desulfovibrionales. These five OTU were found in 32 a total of 150 clones of the 16S rDNA library (21.8%) as revealed by ARDRA analysis. Two further clonal sequences, ATB-KM1325 and ATB-KM1289, were also highly similar to yet

ARTICLE IN PRESS M. Klocke et al. / Systematic and Applied Microbiology 30 (2007) 139–151

145

Fig. 5. Phylogenetic tree of the clones among the class Deltaproteobacteria of the domain Bacteria based on neighbour-joining analysis of partial 16S rDNA sequences. Lactobacillus delbrueckii DSM 20074 was used as the outgroup (marked by an asterisk). The abbreviation ‘‘Uncult. bact. clone’’ means ‘‘Uncultured bacterium clone’’. Further details are as given in Fig. 4.

uncultured bacteria. As an example, the clone cs25 [DQ088249], derived from an analysis of the microbial community in a moat sediment, was included in the phylogenetical analysis (identity value 95%). These clones seemed to be more closely related to strains of the orders Syntrophobacterales, Desulfobacterales, Desulfomonadales, Myxococcales and Bedellovibrionales than to members of the orders Desulfovibrionales and Desulforellales. One clonal sequence was found to be closely related to the genus Pelobacter, order Desulfomonadales. Highest identity values of between 97% and 98% were obtained in comparision with Pelobacter acetylenicus DSM 2348, Pelobacter venetianus, Pelobacter carbinolicus DSM 2380 and the environmental clone EtOHpelo isolated out of an intertidial mud flat [31].

(II) Firmicutes Many OTUs (21 representing 33 clones) revealed a relationship to members of the order Clostridiales (Fig. 6). 21 OTU representing 33 clones were assigned to this order. Two OTU, ATB-KM1344 and ATBKM1254, were found to be related to the genus Sedimentibacter within the family Peptostreptococca-

ceae. Further clonal sequences were assigned to the family Peptococcaceae, to the genus Desulfotomaculum and the genus Peptococcus, respectively. Several OTU were also found to be affiliated with the family Syntrophomonadaceae and the genus Aminomonas, respectively. Three OTU were assigned to Acetovibrio cellulolyticus DSM 1870 [L35516], which is a member of the family Clostridiaceae. Interestingly,some clonal sequences (ATB-KM1249 and ATB-KM1239) were also found with similarities to the endosymbiontic bacterium Sporobacter termitidis DSM 10068 [14]. Eleven OTUs representative for a total of 32 clones were assigned to the class Bacilli of the phylum Firmicutes (Fig. 7). Most of them (7 OTU) are associated with other uncultured clones derived from environmental samples in a group distantly related to the genus Paenibacillus. Closely related, but uncultured clones were derived predominantly from anaerobic environments like landfill leachate (clone GZKB26), feedlot manure (clone B4) [AF317372], sludge digesters (clone 061D08 and 008C09) [CR933290] and the gastrointestinal tract of various animals (e.g. clone p-1528-b5 [25]) as examples. Furthermore, one OTU (ATB-KM1376) was assigned to the genus Lactobacillus together with the uncultured

ARTICLE IN PRESS 146

M. Klocke et al. / Systematic and Applied Microbiology 30 (2007) 139–151

Fig. 6. Phylogenetic tree of the clones among the order Clostridiales of the domain Bacteria based on neighbour-joining analysis of partial 16S rDNA sequences. E. coli ATCC 1175 was used as the outgroup (marked by an asterisk). Further details are as given in Fig. 4.

clone U148 isolated from a phenanthrene-degrading microbial community [AY871193]. Another clonal sequence, ATB-KM1384, was affiliated with the genus Syntrophococcus. This OTU revealed the highest identity values (99%) to the uncultured clone 012F10 obtained from the analysis of an anaerobic sludge digester [CR933135].

(III) Actinobacteria Only one OTU (ATB-KM1232) representative for five clones of the 16S rDNA library was assigned to the genus Atopobium, a member of the phylum Actinobacteria. The closest relationship was obtained with Atopobium parvulum ATCC 22793 with an

ARTICLE IN PRESS M. Klocke et al. / Systematic and Applied Microbiology 30 (2007) 139–151

147

Fig. 7. Phylogenetic tree of the clones among the class Bacilli of the domain Bacteria based on neighbour-joining analysis of partial 16S rDNA sequences. E. coli ATCC 1175 was used as the outgroup (marked by an asterisk). Further details are as given in Fig. 4.

identity value of 95%. Higher identity values were also observed in comparison with the uncultured clones rRNA160 [21], PrebhufecO3 and T15FO04 [DQ093271] of predominately human origin (gastrointestinal tract).

(IV) Spirochaetes Two clonal sequences, ATB-KM1283 and ATBKM1331, were affiliated in the phylum Spirochaetes and, both of them, in the family Spirochaetaceae. No Leptospiraceae or Serpulinaceae were detected. Both clonal sequences were assigned to yet uncultured Spirochaeta clones. ATB-KM1283 were closely related with the uncultured clone TANB18 [29] and clone DCE33 [15] with identity values of 99%. Both clones were obtained from the analysis of biodegrading microbial consortia. ATB-KM1331 was assigned with an identity value of 90% to the uncultured clone AUSPI67 [AY648566] isolated from an anaerobic bioreactor processing sulfate-rich waste streams.

(V) Bacteroidetes Within the phylum Bacteroidetes, 13 OTU representatives for 31 clones were found (Fig. 8). All detected OTU were assigned to the class Bacteroides, no OTU was affiliated with the class Sphingobacteria or Flavobacteria. Most clonal sequences showed higher identity values with certain species of the genus Bacteroides, the closest relationship was obtained with the clone ATB-KM1330. Three OTU, ATB-KM1392, ATB KM1240 and ATB-KM1328, were assigned to Proteophilum aceticum TB107 [AY742226] with an identity value of 97%. Furthermore, one OTU (ATBKM1382) was found to be related with Rikenella microfusus DSM 15922. However, most OTU were grouped together with uncultured clonal sequences forming comparatively distinct clusters. As an example, the OTU ATB-KM1272 was assigned to the class Bacteroides, but the clones EUB53-2 [AY69830] and 054E07-B-DI-P58 [CR933135] were found as closest neighbours. These clones were obtained from bioreactors treating wastewater and sludge, respectively. A further relationship was given to the clones IA-16

ARTICLE IN PRESS 148

M. Klocke et al. / Systematic and Applied Microbiology 30 (2007) 139–151

Fig. 8. Phylogenetic tree of the clones among the phylum Bacteroidetes of the domain Bacteria based on neighbour-joining analysis of partial 16S rDNA sequences. E. coli ATCC 1175 was used as the outgroup (marked by an asterisk). Further details are as given in Fig. 4.

[AJ488070] and TSAI15 from a degrading microbial consortium [47].

Discussion The fermentation of fodder beet silage as monosubstrate revealed a high potential for biogas production. The applied OLR of 1.23 kg m3 d1 in the third week and up to 2.32 kg m3 d1 in the nineteenth week is equal to a hydraulic retention time of 106 and 56 days, respectively. This provides sufficient time for degradation of plant biomass and methanogenesis. According to Linke and Ma¨hnert [27], the maximum biogas yield from fodder beet silage amounted to 0.93 m3 kg1 VS. Thus, the biogas yield of 0.87 m3 kg1 VS obtained in this study equals 94% of the maximum. The methane content of the biogas of about 55.2% also corresponds well with previously published values (e.g. 54%) [46]. It can therefore be concluded that the performance of the CSTR was almost optimal. This indicates the existence of a functional biogas-producing microbial consortium.

The content of N and P in the effluent decreased during the experiment caused by a wash-out due to the daily input of the silage with a lower content of 1.7 g kg1 N and 0.22 g kg1 P. According to the literature [40], the N:P ratio ranging from 5.8:1 to 5.5:1 and 5:1, respectively, can be described as wellbalanced. The pH value of the applied substrate and also of the effluent with values of 3.6 and 7.8 correlates closely with the literature. Scherer et al., for example, obtained pH values of fodder beet silage and digested fodder beet of 3.4–4.1 and 7.5, respectively [39]. Normally, a neutral milieu with a pH of 6.8 and 7.5 is assumed to be sufficient for a growth of methanogenic Archaea [38]. Furthermore, the content of volatile fatty acids in the effluent (1.8 g kg1) should be low enough to exclude any inhibitory effects. In conclusion, the process conditions are suited to allow the establishment of methanogenic archaea. Molecular analysis revealed the presence of H2/CO2/ formate-oxidizing Methanobacteriales, the H2/CO2-oxidizing Methanosarcinaceae or the acetate-splitting Methanosaetaceae [13]. These findings are in agreement with the common hypothesis that biogas is normally

ARTICLE IN PRESS M. Klocke et al. / Systematic and Applied Microbiology 30 (2007) 139–151

produced by both hydrogenotrophic and acetoclastic Archaea [41]. As producers of the substrates for methanogenic Archaea, a broad range of chemoorganoheterotrophic Bacteria were found. Sequence similarities suggest that most of them were anaerobic mesophiles. Several OTU were affiliated with carbohydrate utilizing Bacteria like the Bacteroides, Spirochaetaceae or members of the Clostridiales. OTU related to species using aminoacids (e.g. Peptostreptococcus anaerobicus) or short branched fatty acids (e.g. Peptococcus niger) were detected. Beside this, some OTU were assigned to potential cellulose utilizing Clostridiales like Acetivibrio cellulyticus. End product of the dissimilation metabolism is in most cases acetate or other VFA, ethanol, H2 and CO2. Interestingly, no clearly syntrophic bacterial species from the order Syntrophobacteriales, which are normally part of granules in UASB reactors, were detected [18]. Beside the utilization of carbohydrates and aminoacids, Bacteria performing a dissimilatory sulphur reduction (end product H2S) for energy production were also found. The H2S-content in the biogas, with an average of 12 ppm, is also negligible. Several OTU were assigned in close relationship to Desulfuromonas spp. (class Deltaproteobacteria) and Desulfotomaculum (class Firmicutes). Bacterial species (or cloned 16S rDNA sequences representative for such species) with a high similarity to the OTU prepared from the fodder beet CSTR were described for a broad range of natural environments and artificial systems, respectively, such as feedlot manure, sludge digesters, wastewater treating bioreactors, but also for the gastrointestinal tract of human and animals. Thus, the microbial community of the CSTR seemed to be typical for a biomass degrading anaerobic microbial consortium. It is well known that the molecular approach for studying microbial ecology includes a broad range of potential pitfalls. Non-suitable primer sets can lead to a discrimination of certain microbial species [4,12]. The estimation of a quantitative proportion of a single species is also critical[6]., As a result of the influence of primer design and the amplification process during PCR, single species or groups can be overestimated or, in contrast, discriminated. Thus, the dominance of a certain species revealed in a molecular analysis can be no more than initial evidence. In accordance with other studies [e.g. 41], a comparatively low number of methanogenic Archaea species was found in this study. However, the primer set applied in this study yielded a 16S rDNA library containing methanogenic Archaea and Bacteria. This study gives first insights into the microbial conversion process from plant biomass to biogas in technical systems. Further studies analyzing microbial diversity in other types of biogas-reactor processing different herbal substrates will deepen the knowledge

149

and understanding of this important bioconversion process. Quantitative data on population size or specific bioconversion capacities are required for an empiric model which could facilitate the development of better process regulation mechanisms and, finally, the engineering of more efficient and, thus, more economic biogas reactors.

Acknowledgements M. Klocke and P. Ma¨hnert were supported by research grants from the German Federal Ministry of Food, Agriculture and Consumer Protection (grant 22011804 and 22011402), K. Souidi was supported by a research grant from the German Federal Ministry of Education and Research (grant 03SF0317M).

References [1] B.K. Ahring, M. Sandberg, I. Angelidaki, Volatile fatty acids as indicators of process imbalance in anaerobic digestors, Appl. Microbiol. Biotechnol. 43 (1995) 559–565. [2] S.F. Altschul, W. Gish, W. Miller, E.W. Myersand, D.J. Lipman, Basic local alignment search tool, J. Mol. Biol. 215 (1990) 403–410. [3] R.I. Amann, W. Ludwig, K.H. Schleifer, Phylogenetic identification and in-situ detection of individual microbial-cells without cultivation, Microbiol. Rev. 59 (1995) 143–169. [4] G.C. Baker, J.J. Smith, D.A. Cowan, review and reanalysis of domain-specific 16S primers, J. Microbiol. Methods 55 (2003) 541–555. [5] D.A. Benson, I. Karsch-Mizrachi, D.J. Lipman, J. Ostell, D.L. Wheeler, GenBank, Nucleic Acids Res. 31 (2003) 23–27. [6] R. Bonnet, A. Suau, J. Dore, G.R. Gibson, M.D. Collins, Differences in rDNA libraries of faecal bacteria derived from 10-and 25-cycle PCRs, Int. J. Syst. Evol. Microbiol. 52 (2002) 757–763. [7] B. Calli, B. Mertoglu, B. Inanc, O. Yenigun, Methanogenic diversity in anaerobic bioreactors under extremely high ammonia levels, Enzyme Microb. Technol. 37 (2005) 448–455. [8] R. Chenna, H. Sugawara, T. Koike, R. Lopez, T.J. Gibson, D.G. Higgins, J.D. Thompson, Multiple sequence alignment with the Clustal series of programs, Nucleic Acids Res. 31 (2003) 3497–3500. [9] J.R. Cole, B. Chai, T.L. Marsh, R.J. Farris, Q. Wang, S.A. Kulam, S. Chandra, D.M. McGarrell, T.M. Schmidt, G.M. Garrity, J.M. Tiedje, The Ribosomal Database Project (RDP-II): previewing a new autoaligner that allows regular updates and the new prokaryotic taxonomy, Nucleic Acids Res. 31 (2003) 442–443. [10] J. Felsenstein, Confidence limits of phylogenies: an approach using the bootstrap, Evolution 39 (1985) 783–791.

ARTICLE IN PRESS 150

M. Klocke et al. / Systematic and Applied Microbiology 30 (2007) 139–151

[11] A. Fernandez, S.Y. Huang, S. Seston, J. Xing, R. Hickey, C. Criddle, J. Tiedje, How stable is stable? Function versus community composition, Appl. Environ. Microbiol. 65 (1999) 3697–3704. [12] L.J. Forney, X. Zhou, C.J. Brown, Molecular microbial ecology: land of the one-eyed king, Curr. Opin. Microbiol. 7 (2004) 210–220. [13] G.M. Garrity, (Editor-in-Chief), Bergey’s Manual of Systematic Bacteriology, second ed., 2001, Springer, Berlin, Germany, pp. 155–166. [14] I. Grech-Mora, M.-L. Fardeau, B.K.C. Patel, B. Ollivier, A. Rimbault, G. Prensier, J.-L. Garcia, E. Garnier-Sillam, Isolation and characterization of Sporobacter termitidis gen. nov. sp. nov., from the digestive tract of the woodfeeding termite Nasutitermes lujae, Int. J. Syst. Bacteriol. 46 (1996) 512–518. [15] A.Z. Gu, B.P. Hedlund, J.T. Staley, S.E. Strand, H.D. Stensel, Analysis and comparison of the microbial community structures of two enrichment cultures capable of reductively dechlorinating TCE and cis-DCE, Environ. Microbiol. 6 (2004) 45–54. [16] K.H. Hansen, B.K. Ahring, L. Raskin, Quantification of syntrophic fatty acid-b-oxidizing bacteria in a mesophilic biogas reactor by oligonucleotide probe hybridization, Appl. Environ. Microbiol. 65 (1999) 4767–4774. [17] M.C. Hansen, T. Tolker-Nielsen, M. Givskov, S. Molin, Biased 16S rDNA PCR amplification caused by interference from DNA flanking the template region, FEMS Microbiol. Ecol. 26 (1998) 141–149. [18] H.J.M. Harmsen, H.M.P. Kengen, A.D.L. Akkermans, A.J.M. Stams, W.M. de Vos, Detection and localization of syntrophic propionate-oxidizing bacteria in granular sludge by in situ hybridization using 16S rRNA-based oligonucleotide probes, Appl. Environ. Microbiol. 62 (1996) 1656–1663. [19] J. Hofman-Bang, D. Zheng, P. Westermann, B.K. Ahring, L. Raskin, Molecular ecology of anaerobic reactor systems, Adv. Biochem. Eng./Biotechnol. 81 (2003) 151–203. [20] S. Holland, Analytic Rarefaction 1.3. URL ftp:// 3rdrock.gly.uga.edu/pub/stratum/aRarefactWin.exe (posting date October 2003, last date accessed 7 December 2005). [Neu1] [21] R.W. Hyman, M. Fukushima, L. Diamond, J. Kumm, L.C. Giudice, R.W. Davis, Microbes on the human vaginal epithelium, Proc. Natl. Acad. Sci. USA 102 (2005) 7952–7957. [22] S. Kumar, K. Tamura, M. Nei, MEGA3: Integrated software for molecular evolutionary genetics analysis and sequence alignment, Briefings Bioinform. 5 (2004) 150–163. [23] G. Laguerre, M.R. Allard, F. Revoy, N. Amarger, Rapid identification of Rhizobia by restriction-fragment-length polymorphism analysis of PCR-amplified 16S ribosomalRNA genes, Appl. Environ. Microbiol. 60 (1994) 56–63. [24] J.J. Lay, Y.Y. Li, T. Noike, The influence of pH and ammonia concentration on the methane production in high-solids digestion processes, Water Environ. Res. 70 (1998) 1075–1082. [25] T.D. Leser, J.Z. Amenuvor, T.K. Jensen, R.H. Lindecrona, M. Boye, K. Moller, Culture-independent analysis of

[26] [27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

gut bacteria: the pig gastrointestinal tract microbiota revisited, Appl. Environ. Microbiol. 68 (2002) 673–690. G. Lettinga, Anaerobic digestion and wastewater treatment systems, Antonie van Leeuwenhoek 67 (1995) 3–28. B. Linke, P. Ma¨hnert, Biogasgewinnung aus Rindergu¨lle und nachwachsenden Rohstoffen—Einfluß der Belastung des Fermenters, Agrartechn. Forschung. 11 (2005) 125–132. P.E. Luton, J.M. Wayne, R.J. Sharp, P.W. Riley, The mcrA gene as an alternative to 16S rRNA in the phylogenetic analysis of methanogen populations in landfill, Microbiology 148 (2002) 3521–3530. T.W. Macbeth, D.E. Cummings, S. Spring, L.M. Petzke, K.S. Sorenson Jr., Molecular characterization of a dechlorinating community resulting from in situ biostimulation in a trichloroethene-contaminated deep, fractured basalt aquifer and comparison to a derivative laboratory culture, Appl. Environ. Microbiol. 70 (2004) 7329–7341. M.G. Murray, W.F. Thompson, Rapid isolation of high molecular weight plant DNA, Nucleic Acids Res. 8 (1980) 4321–4325. M. Mussmann, K. Ishii, R. Rabus, R. Amann, Diversity and vertical distribution of cultured and uncultured Deltaproteobacteria in an intertidal mud flat of the Wadden Sea, Environ. Microbiol. 7 (2005) 405–418. P.F. Pind, I. Angelidaki, B.K. Ahring, Dynamics of the anaerobic process: effects of volatile fatty acids, Biotechnol. Bioeng. 82 (2003) 791–801. L. Raskin, D. Zheng, M.E. Griffin, P.G. Stroot, P. Misra, Characterization of microbial communities in anaerobic bioreactors using molecular probes, Antonie van Leeuwenhoek 68 (1995) 297–308. K. Ravenschlag, K. Sahm, J. Pernthaler, R. Amann, High bacterial diversity in permanently cold marine sediments, Appl. Environ. Microbiol. 65 (1999) 3982–3989 [Neu2]. H. Rheims, E. Stackebrandt, Application of nested polymerase chain reaction for the detection of as yet uncultured organisms of the class Actinobacteria in environmental samples, Environ. Microbiol. 1 (1999) 137–143. N. Saito, M. Nei, The neighbor-joining method: a new method for constructing phylogenetic trees, Mol. Biol. Evol. 4 (1987) 406–425. J. Sambrook, D.W. Russell, Gel electrophoresis of DNA and pulsed field agarose gel electrophoresis, in: J. Sambrook, D.W. Russell (Eds.), Molecular Cloning, A Laboratory Manual, third ed, Cold Spring Harbour Laboratory Press, New York, NY, USA, 2001 (pp. 5.4–5.17). A. Schattauer, P. Weiland, Grundlagen der anaeroben Fermentation, in: Bundesministerium fu¨r Verbraucherschutz, Erna¨hrung und Landwirtschaft, Fachagentur Nachwachsende Rohstoffe (Eds.), Handreichung Biogasgewinnung und –nutzung, Leipzig, Germany, 2004, pp. 25–35. P.A. Scherer, S. Dobler, S. Rohardt, R. Loock, B. Bu¨ttner, P. No¨ldeke, A. Brettschuh, Continuous biogas production from fodder beet silage as sole substrate, Water Sci. Technol. 48 (2003) 229–233.

ARTICLE IN PRESS M. Klocke et al. / Systematic and Applied Microbiology 30 (2007) 139–151

[40] D. Schmack, Na¨hrstoffe im Fertmenter – Limitierender Faktor fu¨r die Verga¨rung von Nachwachsenden Rohstoffen?, in: Fachverbandes Biogas e.V. (Ed.), Proceedings Biogas–Nachwachsende Rohstoffe, Neue Wege fu¨r die Landwirtschaft, 14, Jahrestagung des Fachverbandes Biogas e.V., 11–14 January 2005, Nu¨rnberg, pp. 121–124. [41] Y. Sekiguchi, Y. Kamagata, K. Syutsubo, A. Ohashi, H. Harada, K. Nakamura, Phylogenetic diversity of mesophilic and thermophilic granular sludges determined by 16S rRNA gene analysis, Microbiology 144 (1998) 2655–2665. [42] T. Shigematsu, Y. Tang, H. Kawaguchi, K. Ninomiya, J. Kijima, T. Kobayashi, S. Morimura, K. Kida, Effect of dilution rate on structure of a mesophilic acetatedegrading methanogenic community during continuous cultivation, J. Biosci. Bioeng. 96 (2003) 547–558. [43] Y. Tang, T. Shigematsu, S. Morimura, K. Kida, Microbial community analysis of mesophilic anaerobic

[44]

[45]

[46]

[47]

151

protein degradation process using bovine serum albumin (BSA)-fed continuous cultivation, J. Biosci. Bioeng. 99 (2005) 150–164. I. Uz, M.E. Rasche, T. Townsend, A.V. Ogram, A.S. Lindner, Characterization of methanogenic and methanotrophic assemblages in landfill samples, Proc. R. Soc. London B—Biol. Sci. 270 (2003) S202–S205. A. Veeken, S. Kalyuzhnyi, H. Scharff, B. Hamelers, Effect of pH and VFA on hydrolysis of organic solid waste, J. Environ. Eng.—ASCE 126 (2000) 1076–1081. P. Weiland, Production and energetic use of biogas from energy crops and wastes in Germany, Appl. Biochem. Biotechnol. 109 (2003) 263–274. N. Yoshida, N. Takahashi, A. Hiraishi, Phylogenetic characterization of a polychlorinated-dioxin-dechlorinating microbial community by use of microcosm studies, Appl. Environ. Microbiol. 71 (2005) 4325–4334.