Bioresource Technology 143 (2013) 431–438

Contents lists available at SciVerse ScienceDirect

Bioresource Technology

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

Evaluation of system performances and microbial communities of twotemperature-phased anaerobic digestion systems treating dairy manure

0960-8524/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.biortech.2013.06.013

⇑ Corresponding author. Address: Department of Animal Sciences, The Ohio StateUniversity, 2029 Fyffe Road, Columbus, OH 43210-1095, United States. Tel.: +1 614292 3057; fax: +1 614 292 2929.

E-mail address: yu.226@osu.edu (Z. Yu).

Wen Lv a, Wenfei Zhang c, Zhongtang Yu a,b,⇑a Department of Animal Sciences, The Ohio State University, Columbus, OH 43210, United Statesb Environmental Science Graduate Program, The Ohio State University, Columbus, OH 43210, United Statesc Department of Biostatistics, Columbia University, New York, NY 10032, United States

h i g h l i g h t s

� The NT-TPAD system achieved better overall performance than the AT-TPAD system.� Both digesters of each TPAD system had different roles between two TPAD systems.� Each digester harbored distinctive microbial populations.� Certain microbial groups were significantly correlated with system performance.� Methanosarcina was important in both systems but Methanosaeta only in NT-TPAD.

a r t i c l e i n f o

Article history:Received 9 April 2013Received in revised form 3 June 2013Accepted 5 June 2013Available online 13 June 2013

Keywords:BiogasDGGEMethaneMethanogensqPCR

a b s t r a c t

Two temperature-phased anaerobic digestion (TPAD) systems, with the thermophilic digesters acidifiedby acidogenesis products (AT-TPAD) or operated at neutral pH and balanced hydrolysis/acidogenesis andmethanogenesis (NT-TPAD), were evaluated to treat high-strength dairy cattle manure. Despite similarmethane productions (about 0.22 L/g VS fed), the NT-TPAD system removed significantly more VS(36%) than the AT-TPAD system (31%) and needed no pH adjustments. The thermophilic digester ofthe NT-TPAD system dominated the system performance and performed significantly better than thatof the AT-TPAD system. The opposite held true for the mesophilic digesters. Differences of the thermo-philic digesters between two TPAD systems affected the microbial communities of both local and down-stream digesters. Each digester harbored distinctive microbial populations, some of which weresignificantly correlated with system performance. Methanosarcina was the most important methanogenicgenus in both TPAD systems, while Methanosaeta only in the NT-TPAD system. Their populations wereinversely related to VFA concentrations.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Compared to mesophilic single-stage digesters that are com-monly used in full scale (De Baere, 2000), temperature-phasedanaerobic digestion (TPAD) systems are considered one of the mostpromising approaches that improve both efficiency and reliabilityof the anaerobic digestion (AD) process in renewable energyproduction and biomass waste management (Lv et al., 2010;Mata-Alvarez et al., 2000). Previous studies suggested that TPADsystems could achieve improved pathogen control, solid removal,and methane production compared to mesophilic single-stage

digesters (Riau et al., 2010; Santha et al., 2006), and other ADsystems (Kim et al., 2004; Riau et al., 2010).

A TPAD system consists of a first stage thermophilic digesterand a second stage mesophilic digester. The thermophilic digestercan enhance hydrolysis and acidogenesis due to reduced feedstockrecalcitrance and increase microbial metabolism at an elevatedtemperature, while the mesophilic digester provides lenient condi-tions supporting efficient and stable syntrophic acetogenesis andmethanogenesis due to reduced inhibitor toxicity at a lower tem-perature (Lv et al., 2010). Depending on the pH of the thermophilicdigester, a TPAD system can be categorized as either an AT-TPADsystem, where the thermophilic digester is operated at acidic pH(Youn and Shin, 2005), or a NT-TPAD system, where the thermo-philic digester is operated at neutral pH (Sung and Santha, 2003).The acidic pH of the thermophilic digester in an AT-TPAD systemhas been shown to favor hydrolysis and acidogenesis at the ex-pense of syntrophic acetogenesis and methanogenesis (Chyi and

432 W. Lv et al. / Bioresource Technology 143 (2013) 431–438

Dague, 1994; Angelidaki et al., 2002), while the neutral pH of thethermophilic digester in a NT-TPAD system is aimed at reachinga dynamic balance between hydrolysis/acidogenesis and methano-genesis. Schmit and Ellis (2001) reported that a NT-TPAD systemhad either better or similar performances in solid removal andmethane production compared to an AT-TPAD system, dependingon the feedstock composition.

Feedstock composition can significantly affect the performancesof AT-TPAD and NT-TPAD systems (Lv et al., 2010; Schmit and Ellis,2001). However, these two systems have not been compared intreating dairy manure, one of the most important biomass wastesfor renewable energy production (Yu et al., 2009). A comparativestudy between these two types of TPAD systems in digestinghigh-strength dairy manure slurry is necessary. In addition, littleis known about microbial population dynamics and interactionsimportant to the AD process (Hofman-Bang et al., 2003; O’Flahertyet al., 2006), and the understanding of how the conditions within athermophilic digester affect the system performance and themicrobial community of the downstream mesophilic digester andthe entire TPAD system is limited. Thus, the objective of this studywas to evaluate and compare both system performances andmicrobial communities between an AT-TPAD system and aNT-TPAD system in order to understand the microbial populationsunderpinning system performance.

2. Methods

2.1. Experimental setup

Both the thermophilic and the mesophilic digesters were madefrom two Nalgene polypropylene wide-mouth bottles of 4.3 Lcapacity with 100 mm screw caps (Fisher Scientific, PA). Each di-gester had a feeding port, a sampling port, and a gas outlet. Thefeeding port and the sampling port were sealed by rubber stoppersto keep the digesters airtight. The biogas produced was collectedand measured by water displacement. The first stage thermophilicdigester and the second stage mesophilic digester had working vol-umes of 1 L and 2 L, respectively. The two digesters were placed intwo water baths to maintain their temperatures (50 �C for the ther-mophilic digester and 35 �C for the mesophilic digester).

2.2. Seed sludge

A thermophilic (55 �C) and a mesophilic (37 �C) bench-scalereactors had been maintained separately for 6 months digestingdairy cattle manure. The contents from these two reactors wereused as seed sludge for the thermophilic and the mesophilic digest-ers of the AT-TPAD system in a respective manner. The contents ofthe thermophilic and the mesophilic reactors contained 8.80% and9.00% total solid (TS), respectively, and about 6.40% volatile solid(VS).

After the AT-TPAD system was finished in the study, its thermo-philic and mesophilic digester contents were then used as seedsludge for the NT-TPAD system. The thermophilic and the meso-philic seed sludge contents contained 11.52% and 9.33% TS, and9.67% and 7.44% VS, respectively.

2.3. Feedstock

Dairy manure (including feces and urine) from cattle was col-lected on a daily basis from the Waterman Dairy Center, The OhioState University. Based on dry matter (DM), the ration fed to thecattle during this study was primarily composed of 50.00% corn si-lage, 4.50% alfalfa hay, 21.00% co-product of corn wet milling,9.05% ground corn, 4.64% soybean meal, 1.30% Aminoplus�, 1.30%

soyhulls, 0.38% fat, 2.01% vitamin and minerals. The average TSand VS contents of the collected manure were 14.61% (w/v) and12.81% (w/v), respectively. Prior to use, the manure was dilutedto desired TS and VS contents using tap water and mixed thor-oughly into slurry. This was done to reduce potential clogging indigesters (Sung and Santha, 2003) and improve the substrateaccessibility to the microbial community (Angelidaki and Ahring,2000).

2.4. Start-up, operation, and sampling

In the AT-TPAD system, the thermophilic and the mesophilicdigesters were seeded with 100 mL and 200 mL of the correspond-ing seed sludge, respectively, and then filled to their working vol-umes with pre-warmed manure slurry. After maintained at desiredtemperatures for 4 days to allow temperature equilibration andmicrobial activity recovery, the AT-TPAD system was then fed withthe same manure slurry of 11.25% TS and 9.86% VS. In the NT-TPADsystem, both digesters were filled to their working volumes withseed sludge: the thermophilic digester was filled with a mixtureof the thermophilic and the mesophilic seed sludge (0.5 L each),while the mesophilic digester was filled with only the mesophilicseed sludge (2 L). The NT-TPAD system was then fed with manureslurry of 13.56% TS and 11.90% VS.

Both TPAD systems were operated in a fed-batch mode on a dai-ly basis: the effluent was discharged from the mesophilic digester,then the same volume of the content from the thermophilic diges-ter was transferred to the mesophilic digester, and finally the samevolume of manure slurry was fed to the thermophilic digester. Dai-ly biogas production and effluent pH from each digester were re-corded daily before feeding. The content of each digester wasmanually mixed before and after feeding, and the pH of the ther-mophilic digester of the AT-TPAD system was adjusted to 6.5 using10 M NaOH solution when necessary. During the start-up of theNT-TPAD system, an extra volume of sludge (the same volume asthe feeding slurry) was recycled once daily between the twodigesters of the NT-TPAD system in case of acidification (indicatedby pH decreases) in the thermophilic digester until the thermo-philic digester reached balance between hydrolysis/acidogenesisand methanogenesis (indicated by stable and neutral pH).

The operation of each TPAD system lasted 144 days and wassummarized in Table 1. Both systems were operated until theyreached their steady states when the variation of daily biogas pro-duction from each digester was less than 10% for five consecutivedays without any upward or downward trend (Wen et al., 2007).Biogas and sludge samples from each TPAD system were collectedmultiple times during steady-state operation. The sludge sampleswere aliquoted and stored at �80 �C until being analyzed.

2.5. Analysis

2.5.1. System performance analysisContents of CH4 in biogas samples were analyzed by gas chro-

matography (GC) as described previously (Patra and Yu, 2012),and methane yields were calculated from corresponding biogasproductions and methane contents. Sludge samples were centri-fuged to collect their supernatants, prepared per the same protocolused by Oelker et al. (2009), and subjected to VFA analysis usingGC. Concentrations of TS and VS of each sludge sample were deter-mined following the standard methods (American Public HealthAssociation et al., 2005), and solid removal was calculated for eachsludge sample.

2.5.2. Microbial community analysisCommunity DNA was extracted from each sludge sample using

the RBB + C method, which allows efficient extraction of microbial

Table 1The operational parameters of the two TPAD systems used in this study.

System OLR (g VS/L/d) Thermophilic digester Mesophilic digester

HRT/SRT (d) Working volume (L) Temperature (�C) HRT/SRT (d Working volume (L) Temperature (�C)

AT-TPAD 6.58 5 1 50 10 2 35NT-TPAD 7.93 5 1 50 10 2 35

W. Lv et al. / Bioresource Technology 143 (2013) 431–438 433

DNA of high purity (Yu and Morrison, 2004a). The DNA quality wasevaluated using agarose gel (0.8%) electrophoresis. Each DNA sam-ple recovered from the extraction was quantified by Nanodrop(Thermo Scientific, MA) and diluted to a final concentration ofabout 50 ng/lL with TE buffer.

The archaeal and the bacterial communities in each sample ofthe two TPAD systems were analyzed by PCR-denaturing gradientgel electrophoresis (DGGE) as previously described (Yu and Morri-son, 2004b; Yu et al., 2008). Briefly, the V3 hyper-variable region of16S rRNA genes was amplified from each DNA samples by both do-mains archaea- and bacteria-specific primers (Table 2) and re-solved by DGGE with a 40–60% denaturing gradient. The DGGEgel images of both TPAD systems were analyzed by BioNumericsversion 5.1 (Applied Maths, Inc., TX) to determine banding profilesof each DNA sample as described by Cressman et al. (2010).Dendrograms based on the similarity matrix of DGGE profiles wereconstructed using the Jaccard similarity coefficient and the UPGMAclustering method. The determined DGGE profiles were also sub-jected to principal component analysis (PCA).

Four methanogen genera, including Methanobacterium,Methanoculleus, Methanosaeta, Methanosarcina, the WSA2/ArcIgroup, and total archaea in each DNA sample were quantified usingrespective primer/TaqMan probe sets (Table 2). Briefly, a sample-derived qPCR standard was prepared for each target genus or groupfrom a pooled DNA sample representing the DNA samples to beanalyzed as previously described (Yu et al., 2005a,b). Each standardwas subsequently quantified and serially diluted with TE buffer toachieve a nine log range of concentrations (i.e. from 100 ampli-con copies/lL to 108 amplicon copies/lL). Each targeted grouppresent in each DNA sample was quantified in three technicalreplicates with its corresponding qPCR standard (also in threetechnical replicates) included on the same qPCR plate. After an

Table 2Primers and primer/probe sets used in this study for DGGE and qPCR analyzes.

Analysis Target Primer Sequence (50–30)

DGGE Archaea GC-Arc344F ACGGGYGCAGCAGGCGCGAa

Univ519R ATTACCGCGGCKGCTGBacteria GC-Eub357F CCTACGGGAGGCAGCAGa

Univ519R ATTACCGCGGCKGCTG

qPCR WSA2/ArcI ArcI-F GCTCATGCATTGCATGGArcI-Taq Cy5-GTAATACCGGCAGCTCGAGTGArcI-R TATCCGGCTACGAACGTT

Methanobacterium Mbt-210F CCAAGCCWKTRATCTGTACGMbt-341Taq FAM-CGCGAAACCTCCGCAATGC-BMbt-359R CGTTAAGAGTGGCACTTGGG

Methanoculleus Mcu-298F GGAGCAAGAGCCCGGAGTMcu-Taq HEX-CGGTCTTGCCCGGCCCTTTCTMcu-586R ATTTCCCAAGAGACTTAACAACCCA

Methanosarcina MB1b CGGTTTGGTCAGTCCTCCGGSAR761Taq HEX-ACCAGAACGGGTTCGACGGTSAR835R AGACACGGTCGCGCCATGCCT

Methanosaeta MS1b CCGGCCGGATAAGTCTCTTGASAE761Taq FAM-ACCAGAACGGACCTGACGGCSAE835R GACAACGGTCGCACCGTGGCC

Total archaea TArc-787F ATTAGATACCCSBGTAGTCCTArc-915Taq HEXAGGAATTGGCGGGGGAGCACTArc-1059R GCCATGCACCWCCTCT

a GC-Arc344F and GC-Eub357F both had a 40nt GC-clamp attached at their 5’ ends.

initial denaturation at 95 �C for 5 min, all qPCR assays were carriedout in 40 PCR cycles of 95 �C for 15 s and 60 �C for 1 min, withfluorescent signals being acquired at the end of the annealing/extension step at 60 �C. Abundances of methanogen genera andthe total archaea in each DNA sample were quantified against theirrespective sample-derived standards as copies, and converted intocopies/mL sludge sample according to individual dilution rates ofeach DNA sample.

2.5.3. Statistical analysisSince the overall organic loading rate (OLR) was different be-

tween the two TPAD systems, data of system performances andmicrobial communities from the NT-TPAD system were both nor-malized against the OLR of the AT-TPAD system before they werepresented and subjected to statistical analysis.

To examine the correlation between microbial communitiesand system performances of both TPAD systems, Pearson correla-tion coefficients and their p-values were calculated between DGGEbanding profiles and corresponding system performances(methane production and VS removal). Significant correlationwas declared at p < 0.05. Correlations between populations ofmethanogen genera/total archaea and methane productions/VFAconcentrations were assessed in the same way. Correlations weredetermined from the data for the thermophilic digesters and themesophilic digesters separately and also from the data for eachTPAD system.

T-test was used to make pairwise comparisons between meansof system performances (biogas/methane production and VS re-moval), VFA concentrations, and population abundances of meth-anogen genera/total archaea. Statistical significance was declaredat p < 0.05. Standard deviations (SD) were also calculated for allthe data.

Amplicon length (bp) Reference

208 Yu et al., 2008

195 Yu and Morrison, 2004b

143 Nelson, 2011G-BHQ

148 Nelson, 2011HQ

314 Franke-Whittle et al., 2009, this study-BHQ

271 Shigematsu et al., 2003GAGG-BHQ

272 Shigematsu et al., 2003AAGG-BHQ

271 Yu et al., 2005a,b-BHQ

Table 4VS removal by different TPAD systems.

System VS removal (%)

Thermophilic digester Mesophilic digester Overall

AT-TPAD 7.92 ± 0.47a 22.60 ± 1.33a 30.52 ± 0.88a

NT-TPAD 28.11 ± 1.48b 7.93 ± 2.37b 36.05 ± 1.40b

Note: Data were presented as the means ± SD (AT-TPAD, n = 3; NT-TPAD, n = 6).Values within each column with different superscripts were significantly different(p < 0.05).

434 W. Lv et al. / Bioresource Technology 143 (2013) 431–438

3. Results and discussion

3.1. System performance

As suggested by a previous study (Sung and Santha, 2003), thethermophilic and the mesophilic digesters of both TPAD systemswere operated at a hydraulic retention time (HRT)/solid retentiontime (SRT) of 5 days and 10 days, respectively. Preliminary resultsshowed that the TPAD system was optimized when the thermo-philic digester was operated at 50 �C (unpublished results), sothe thermophilic and the mesophilic digesters of both TPAD sys-tems were maintained at 50 �C and 35 �C in this study. In the AT-TPAD system, the thermophilic digester was acidified (pH around6.0) by acidogenesis and needed daily pH adjustments, while themesophilic digester was operated at stable pH around 7.5 withoutany need for pH adjustment. In the NT-TPAD system, both digest-ers were operated at stable pH around 7.2 without the need to ad-just the pH. This indicated that proper inoculation and sludgerecycling during the start-up of a TPAD system could prevent theacidification in the thermophilic digester and help establish aNT-TPAD system. This finding corroborates the findings of a previ-ous study in that the start-up procedure was crucial for perfor-mances of the thermophilic AD systems (Ahring, 1994).

The thermophilic digester of the NT-TPAD system achieved sig-nificantly higher VS removal and biogas/methane productions thanthat of the AT-TPAD system, while the opposite holds true for themesophilic digesters of these two TPAD systems. The overall VS re-moval and biogas production from the NT-TPAD system were sig-nificantly higher than those from the AT-TPAD system, but theoverall methane productions were similar between these twoTPAD systems (Tables 3 and 4). These results indicated that thepH of the thermophilic digester not only had local effects, but alsohad extended influences on the downstream mesophilic digester,affecting the entire TPAD system. Performances of the thermophilicdigesters differed between the two TPAD systems, possibly due todifferent microbial communities formed under different conditionsincluding different pH values. The mesophilic digesters of the twoTPAD systems might also have different microbial communitiesthat caused different performances, likely attributed to the influxesfrom the thermophilic digesters that contained significantly differ-ent concentrations of VFA between the two TPAD systems (Fig. 1A).In addition, the NT-TPAD system converted more degraded VS tonon-methane substances, which might include CO2, intermediates,and microbial cell biomass.

In the AT-TPAD system, the mesophilic digester accounted for74.0% of the VS removal and 84.8% of the methane production,while the thermophilic digester only accounted for 26.0% of theVS removal and 15.2% of the methane production. On the contrary,in the NT-TPAD system, the thermophilic digester accounted formost of the VS degradation (78.0%) and methane production(75.3%), while the mesophilic digester only accounted for 22.0%of the VS removal and 24.7% of the methane production. Theseopposite patterns in performance suggested that the thermophilicand the mesophilic digesters had different roles in each TPAD sys-tem. In the AT-TPAD system, the thermophilic digester was acidi-fied by and mainly used for hydrolysis and acidogenesis,

Table 3Biogas and methane productions of different TPAD systems.

System Thermophilic digester Mesophilic diges

Biogas production* Methane production CH4 (%) Biogas productio

AT-TPAD 1.18 ± 0.19a 0.65 ± 0.10a 55.00 2.58 ± 0.11a

NT-TPAD 5.63 ± 0.10b 3.30 ± 0.12b 58.61 0.89 ± 0.02b

Note: Data were presented as the means ± SD (AT-TPAD, n = 5; NT-TPAD, n = 6). Values w* liter/liter working volume/day.

facilitating performance of the mesophilic digester by providinghydrolyzed substrates and high concentrations of VFA. The highconcentrations of VFA might have impaired both VS removal andmethane production of the thermophilic digester, possibly by lim-iting microbial populations (particularly methanogens). The meso-philic digester made the primary contribution to methanogenesisthat accounted for most VS removal of the AT-TPAD system. Inthe NT-TPAD system, although both digesters were operated withbalanced hydrolysis/acidogenesis and methanogenesis, the ther-mophilic digester made the primary contribution to the overallsystem performance, with the mesophilic digester only scavengingresidual hydrolysis and acidogenesis products received from thethermophilic digester. This indicates that the balance betweenhydrolysis/acidogenesis and methanogenesis resulted in efficientVS removal and methane production from both digesters of theNT-TPAD system, which was consistent with the findings of a pre-vious study in that a balanced loading and operation were impor-tant to performances of the thermophilic AD systems (Ahring,1994).

3.2. Microbial community

3.2.1. DGGEDGGE profiling revealed 20 band positions (A1–20) for the

archaeal community and 50 band positions (B1–50) for the bacte-rial community (Suppl. Figs. S1–S3). Clustering analysis of theDGGE profiles showed that both the archaeal and the bacterialcommunities present in both the thermophilic and the mesophilicdigesters generally differed between the two TPAD systems andthat the thermophilic and the mesophilic digesters harbored differ-ent archaeal and bacterial communities in each TPAD system(Fig. 2). The PCA analysis results also showed that both the archa-eal and the bacterial communities of the thermophilic and themesophilic digesters were different between the two TPAD sys-tems (Fig. 3). This indicates that different operation conditions(i.e., pH values and VFA concentrations) of the thermophilic digest-ers led to the development of different microbial communities inboth the thermophilic and the mesophilic digesters between thetwo TPAD systems. In addition, the thermophilic and the meso-philic digesters of each TPAD system generally had distinctivemicrobial communities that were selected by the conditions there-in and fulfilled different roles in their system, which was consistentwith the findings of a previous study in that different operationtemperatures led to the development of different microbial com-munities in a hybrid anaerobic reactor (Kundu et al., 2013).

ter Overall

n Methane production CH4 (%) Biogas production Methane production

1.82 ± 0.08a 70.65 2.11 ± 0.06a 1.43 ± 0.04a

0.54 ± 0.03b 60.32 2.47 ± 0.04b 1.46 ± 0.05a

ithin each column with different superscripts were significantly different (p < 0.05).

A B

0

50

100

150

200

250

Ace�c Acid Propionic Acid

Isobutyric Acid

Butyric Acid

Isovaleric Acid

Valeric Acid

Total VFA0

5

10

15

20

25

30

35

40

45

50

Ace�c Acid Propionic Acid

Isobutyric Acid

Butyric Acid

Isovaleric Acid

Valeric Acid

Total VFA

a

b

a

b b b b b baa

a a

a

a

b b b aa a a a aaa

a

b

*

**

** * * * **

* *

*

*

* *

Fig. 1. VFA concentrations (mM) in the thermophilic (A) and the mesophilic (B) digesters of the AT-TPAD (open) and the NT-TPAD (filled black) systems. Error bars representSD (AT-TPAD, n = 3; NT-TPAD, n = 6). Columns with different letters in the same figure indicate significant differences (p < 0.05) in VFA concentrations between the two TPADsystems. Columns of the same type with asterisks indicate significant differences (p < 0.05) in VFA concentrations between the thermophilic and the mesophilic digesters of aTPAD system.

100

80

60

100

80

60

NT-MNT-TAT-TAT-M NT-MAT-MNT-TAT-T

A B

Fig. 2. Archaeal (A) and bacterial (B) DGGE banding patterns of the thermophilic (T) and the mesophilic (M) digesters of the AT-TPAD (AT) and the NT-TPAD (NT) systems.

W. Lv et al. / Bioresource Technology 143 (2013) 431–438 435

3.2.2. qPCRIn the thermophilic digester of the AT-TPAD system, Methano-

bacterium and Methanoculleus were the two most predominantmethanogen genera (Fig. 4A), which suggests that the methaneconversion in this digester was dominated by the hydrogenotroph-ic methanogenesis pathway. The facultatively acetoclastic genus,Methanosarcina, was only detected at a low abundance. In the ther-mophilic digester of the NT-TPAD system, Methanosarcina and Met-hanobacterium were the two most predominant methanogengenera (Fig. 4A), suggesting that the methane conversion wasconducted by both the acetoclastic and the hydrogenotrophicmethanogenesis pathways in this digester. In the thermophilicdigesters of TPAD systems, especially when the pH was low as inthe case of the thermophilic digester of the AT-TPAD system,methanogenesis carried out by hydrogenotrophic methanogensmight have been facilitated by enhanced hydrogen production

(O-Thong et al., 2011). Considering the significantly lower VFAconcentrations and the significantly higher predominance of Met-hanosarcina in the thermophilic digester of the NT-TPAD systemthan in the thermophilic digester of the AT-TPAD system, aceto-clastic methanogenesis carried out by members of Methanosarcinamight be a key for efficient and stable operation of the thermo-philic digester of the NT-TPAD system. Similar findings were re-ported by a previous study on a two-stage AD system in that theenriched population of Methanosarcina likely contributed to thesuccessful and stable operations (Coats et al., 2012). The popula-tions of the three methanogen genera and total archaea detectedin the thermophilic digesters were significantly larger in the NT-TPAD system than in the AT-TPAD system, which corroboratesthe inverse correlation between populations of methanogens andVFA concentrations in digesters. The other acetoclastic genus,Methanosaeta, was not detected in the thermophilic digesters of

NT-T NT-M AT-T AT-M

A BFig. 3. PCA analysis results for archaeal (A) and bacterial (B) DGGE profiles of the AT-TPAD and the NT-TPAD systems. Values in parentheses of each axis are percentages ofvariances explained by each principal component (PC).

A B

1.E+001.E+011.E+021.E+031.E+041.E+051.E+061.E+071.E+081.E+091.E+10

1.E+001.E+011.E+021.E+031.E+041.E+051.E+061.E+071.E+081.E+091.E+10

aa

a a

a

ab

b

b b a b

a a

a

b a b a a*

*

*

*

*

*

*

*

*

*

Fig. 4. Population abundances (16S rRNA gene copies/g sludge sample) of four methanogen genera and total archaea in the thermophilic (A) and the mesophilic (B) digestersof the AT-TPAD (open) and the NT-TPAD (filled black) systems. Error bars represent SD (AT-TPAD, n = 3; NT-TPAD, n = 6). Columns with different letters in the same figureindicate significant differences (p < 0.05) in the population abundance of a specific group between two TPAD systems. Columns of the same color with asterisks indicatesignificant differences (p < 0.05) in the population abundance of a specific group between the thermophilic and the mesophilic digesters of a TPAD system.

436 W. Lv et al. / Bioresource Technology 143 (2013) 431–438

either TPAD system, which might be attributed to its generationtime (Yu et al., 2009) being longer than the HRT/SRT of the thermo-philic digesters.

In the mesophilic digester of the AT-TPAD system, the most pre-dominant methanogen genus was Methanosarcina, while Methan-osaeta was not detected (Fig. 4B). Given the high total VFAconcentration (primarily acetic acid) in the influx from the thermo-philic digester of the AT-TPAD system, the methane produced inthis mesophilic digester might be mainly from the acetoclasticmethanogenesis mediated by Methanosarcina spp. In the meso-philic digester of the NT-TPAD system, however, the most predom-inant methanogen genus was Methanosaeta, with its populationsignificantly larger than that of Methanosarcina (Fig. 4B). This indi-cates that the methane might also be primarily produced via theacetoclastic methanogenesis. Methanosarcina spp. generally havehigher growth rates and tolerances against high concentrationsof VFA than Methanosaeta spp. (Yu et al., 2009), which might makeMethanosarcina more competitive than Methanosaeta in the

mesophilic digester of the AT-TPAD system where the influx fromthe thermophilic digester contained high concentrations of VFA(primarily acetate). On the other hand, Methanosaeta might bemore competitive in the mesophilic digester of the NT-TPAD sys-tem where the influx from the thermophilic digester containedlow concentrations of VFA because Methanosaeta spp. generallyhave lower Ks values for acetate and thus higher affinities for ace-tate than Methanosarcina spp. (Yu et al., 2009). The presence ofMethanosaeta in the mesophilic digester of the NT-TPAD systemalso suggests that the HRT/SRT (5 days) of the thermophilic digest-ers might be too short for the population of Methanosaeta to estab-lish therein.

The WSA2/ArcI group, first discovered in several AD systemstreating municipal wastewater and proposed to have a hydrogeno-trophic metabolism (Chouari et al., 2005; Rivière et al., 2009), wasnot detected from the pooled community DNA sample that wasused to prepare the sample-derived qPCR standards, so this groupof methanogens was not analyzed in this study. It is possible that

W. Lv et al. / Bioresource Technology 143 (2013) 431–438 437

the WSA2/ArcI group is not essential for methanogenesis fromdairy cattle manure.

3.3. Correlation between system performance and microbialcommunity

3.3.1. DGGEIn the thermophilic digesters of both TPAD systems, a number of

archaeal DGGE bands (i.e., A4, 5, 7, 8, 9, 10, 15, 16, 18, 19) and bac-terial DGGE bands (i.e., B4, 8, 9, 12, 14, 15, 16, 20, 21, 28, 30, 33, 38,39, 40, 42, 43, 44, 45, 47, 48, 49) appeared to be significantly corre-lated with system performances (i.e., methane production and VSremoval). These band profiles differed between the thermophilicdigesters of the two TPAD systems, mirroring the significantly dif-ferent system performances between these two digesters. In themesophilic digesters of both TPAD systems, some archaeal bands(i.e., A3, 4, 5, 6, 7, 10, 16, 17) and bacterial bands (i.e., B1, 2, 3, 7,9, 16, 18, 19, 20, 28, 31, 32, 34, 36, 38, 39, 40, 42, 43) were also sig-nificantly correlated with system performances. These band pro-files differed between the mesophilic digesters of the two TPADsystems, in accordance with the significantly different system per-formances between these two mesophilic digesters. This was con-sistent with the findings of a previous study in that a strongcorrelation existed between the reactor performance and themicrobial community in anaerobic digesters (Kundu et al., 2013).Given that system performances and microbial communities ofboth the thermophilic and the mesophilic digesters were differentbetween the two TPAD systems, it was hypothesized that the differ-ent conditions, including pH values, of the thermophilic digestersnot only resulted in different microbial communities and systemperformances therein, but also had extended influences on themicrobial communities and system performances of the down-stream mesophilic digesters, due to the sequential operations ofboth TPAD systems. This premise was supported by the differentDGGE band profiles of both the thermophilic and the mesophilicdigesters between these two TPAD systems, and the correlationsbetween these band profiles and system performances.

3.3.2. qPCRIn the thermophilic digesters of both TPAD systems, methane

productions and populations of all the three detected methanogengenera and total archaea were significantly correlated. Populationsof these groups in the thermophilic digester were significantly lar-ger of the NT-TPAD system than of the AT-TPAD system, whichmirrored the significantly higher methane production from thethermophilic digester of the NT-TPAD system than of the AT-TPADsystem. Since the populations of Methanosarcina had the greatestdifference among all the three detected methanogen genera be-tween the thermophilic digesters of the two TPAD systems, itwas concluded that the population of Methanosarcina might bethe key for the thermophilic digester to be operated with balancedhydrolysis/acidogenesis and methanogenesis leading to highmethane production. The population data of the other methanogengenera and total archaea also corroborated the higher methaneproduction from the thermophilic digester of the NT-TPAD system.

In the mesophilic digesters of both TPAD systems, except forMethanoculleus (p = 0.08), methane productions and populationsof the other three methanogen genera were significantly corre-lated. Populations of Methanosarcina and Methanobacterium inthe mesophilic digester were significantly larger of the AT-TPADsystem than of the NT-TPAD system, which mirrored the signifi-cantly higher methane production from the mesophilic digesterof the AT-TPAD system than from that of the NT-TPAD system.The population of Methanosaeta was significantly larger in themesophilic digester of the NT-TPAD system than in the mesophilicdigester of the AT-TPAD system, which was inversely correlated

with methane productions of these two mesophilic digesters. Be-cause both the VFA concentrations (primarily acetate) in the ther-mophilic influxes and the predominance of each acetoclasticmethanogen genus in the mesophilic digesters were different be-tween the two TPAD systems, the major methane producers inthe mesophilic digesters were probably different between thetwo TPAD systems: Methanosarcina spp. for the mesophilic digesterof the AT-TPAD system, whereas Methanosaeta spp. for the meso-philic digester of the NT-TPAD system. In both the thermophilicand the mesophilic digesters of the TPAD systems, the populationsof Methanosarcina were significantly correlated with the methaneproductions from these digesters, which was similar to the findingsof a previous study on a two-stage AD system in that the popula-tion of Methanosarcina appeared linked with the biogas production(Coats et al., 2012).

Except for Methanosaeta, population abundances of the otherthree methanogen genera and total archaea were also significantlycorrelated with the concentration of each VFA. Population abun-dances of these groups were significantly higher in the digesterswith significantly lower VFA concentrations. This confirmed thathigh VFA concentrations were inversely correlated with populationabundances of methanogens. Because populations of methanogenswere associated with methane productions, high VFA concentra-tions and methane productions were thus also in an inverse rela-tionship, which was consistent with the findings reported byChakraborty et al. (2011) in that excessive VFA accumulation re-sulted in both decrease in biogas generation and loss of reactor sta-bility. In addition, Methanosaeta was only detected in themesophilic digester of the NT-TPAD system, where both the influxfrom the thermophilic digester and the mesophilic digester contentcontained low concentrations of VFA. This indicated that Methan-osaeta might also be highly sensitive to VFA concentrations.

4. Conclusion

The NT-TPAD system achieved better overall performance thanthe AT-TPAD system. The operation conditions of the thermophilicdigesters affected performances and microbial communities ofboth the thermophilic and the mesophilic digesters. A balance be-tween hydrolysis/acidogenesis and methanogenesis appeared tobe important for efficient and stable operations of thermophilicdigesters. Each digester harbored distinctive microbial popula-tions, some of which were significantly correlated with the systemperformance. Methanosarcina was the most important methano-genic genus in both TPAD systems, while Methanosaeta was onlythe most important in the mesophilic digester of the NT-TPAD sys-tem. Their populations were inversely correlated to high VFAconcentrations.

Acknowledgement

This work was partially supported by a DOE grant awarded toZ.Y. (award number DE-FG36-05GO85010).

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.biortech.2013.06.013.

References

Ahring, B.K., 1994. Status on science and application of thermophilic anaerobicdigestion. Water Sci. Technol. 30, 241–249.

438 W. Lv et al. / Bioresource Technology 143 (2013) 431–438

American Public Health Association, American Water Works Association, WaterEnvironment Federation, 2005. Standard Methods for the Examination of Waterand Wastewater, 21st ed. APHA, Washington, DC.

Angelidaki, I., Ahring, B.K., 2000. Methods for increasing the biogas potential fromthe recalcitrant organic matter contained in manure. Water Sci. Technol. 41,189–194.

Angelidaki, I., Ellegaard, L., Ahring, B.K., 2002. Applications of the anaerobicdigestion process. In: Ahring, B.K. (Ed.), Biomethanation II. Springer, Heidelberg,Berlin, pp. 1–33.

Chakraborty, C., Chowdhury, R., Bhattacharya, P., 2011. Experimental studies andmathematical modeling of an up-flow biofilm reactor treating mustard oil richwastewater. Bioresour. Technol. 102, 5596–5601.

Chouari, R., Le Paslier, D., Daegelen, P., Ginestet, P., Weissenbach, J., Sghir, A., 2005.Novel predominant archaeal and bacterial groups revealed by molecularanalysis of an anaerobic sludge digester. Environ. Microbiol. 7, 1104–1115.

Chyi, Y.T., Dague, R.R., 1994. Effects of particulate size in anaerobic acidogenesisusing cellulose as a sole carbon source. Water Environ. Res. 66, 670–678.

Coats, E.R., Ibrahim, I., Briones, A., Brinkman, C.K., 2012. Methane production onthickened, pre-fermented manure. Bioresour. Technol. 107, 205–212.

Cressman, M.D., Yu, Z., Nelson, M.C., Moeller, S.J., Lilburn, M.S., Zerby, H.N., 2010.Interrelations between the microbiotas in the litter and in the intestines ofcommercial broiler chickens. Appl. Environ. Microbiol. 76, 6572–6582.

De Baere, L., 2000. Anaerobic digestion of solid waste: state-of-the-art. Water Sci.Technol. 41, 283–290.

Franke-Whittle, I.H., Goberna, M., Insam, H., 2009. Design and testing of real-timePCR primers for the quantification of Methanoculleus, Methanosarcina,Methanothermobacter, and a group of uncultured methanogens. Can. J.Microbiol. 55, 611–616.

Hofman-Bang, J., Zheng, D., Westermann, P., Ahring, B.K., Raskin, L., 2003. Molecularecology of anaerobic reactor systems. Adv. Biochem. Eng.: Biotechnol. 81, 151–203.

Kim, H.W., Han, S.K., Shin, H.S., 2004. Anaerobic co-digestion of sewage sludge andfood waste using temperature-phased anaerobic digestion process. Water Sci.Technol. 50, 107–114.

Kundu, K., Sharma, S., Sreekrishnan, T.R., 2013. Changes in microbial communitiesin a hybrid anaerobic reactor with organic loading rate and temperature.Bioresour. Technol. 129, 538–547.

Lv, W., Schanbacher, F.L., Yu, Z., 2010. Putting microbes to work in sequence: recentadvances in temperature-phased anaerobic digestion processes. Bioresour.Technol. 101, 9409–9414.

Mata-Alvarez, J., Mac, S., Llabrs, P., 2000. Anaerobic digestion of organic solidwastes, an overview of research achievements and perspectives. Bioresour.Technol. 74, 3–16.

Nelson, M., 2011. An integrated investigation of the microbial communitiesunderpinning biogas production in anaerobic digestion systems. Ph.D.Dissertation, Ohio State University.

Oelker, E.R., Reveneau, C., Firkins, J.L., 2009. Interaction of molasses and monensinin alfalfa hay- or corn silage-based diets on rumen fermentation, total tractdigestibility, and milk production by Holstein cows. J. Dairy Sci. 92, 270–285.

O’Flaherty, V., Collins, G., Mahony, T., 2006. The microbiology and biochemistry ofanaerobic bioreactors with relevance to domestic sewage treatment. Rev.Environ. Sci. BioTechnol. 5, 39–55.

O-Thong, S., Mamimin, C., Prasertsan, P., 2011. Effect of temperature and initial pHon biohydrogen production from palm oil mill effluent: long-term evaluationand microbial community analysis. Electron. J. Biotechnol. 14, 1–9.

Patra, A.K., Yu, Z., 2012. Effects of essential oils on methane production andfermentation by, and abundance and diversity of, rumen microbial populations.Appl. Environ. Microbiol. 78, 4271–4280.

Riau, V., De la Rubia, M.A., Pérez, M., 2010. Temperature-phased anaerobic digestion(TPAD) to obtain class A biosolids: a semi-continuous study. Bioresour. Technol.101, 2706–2712.

Rivière, D., Desvignes, V., Pelletier, E., Chaussonnerie, S., Guermazi, S., Weissenbach,J., Li, T., Camacho, P., Sghir, A., 2009. Towards the definition of a core ofmicroorganisms involved in anaerobic digestion of sludge. ISME J. 3, 700–714.

Santha, H., Sandino, J., Shimp, G.F., Sung, S., 2006. Performance evaluation of a‘sequential-batch’ temperature-phased anaerobic digestion (TPAD) scheme forproducing Class A biosolids. Water Environ. Res. 78, 221–226.

Schmit, K.H., Ellis, T.G., 2001. Comparison of temperature-phased and two-phaseanaerobic co-digestion of primary sludge and municipal solid waste. WaterEnviron. Res. 73, 314–321.

Shigematsu, T., Tang, Y., Kawaguchi, H., Ninomiya, K., Kijima, J., Kobayashi, T., et al.,2003. Effect of dilution rate on structure of a mesophilic acetate-degradingmethanogenic community during continuous cultivation. J. Biosci. Bioeng. 96,547–558.

Sung, S., Santha, H., 2003. Performance of temperature-phased anaerobic digestion(TPAD) system treating dairy cattle wastes. Water Res. 37, 1628–1636.

Wen, Z., Frear, C., Chen, S., 2007. Anaerobic digestion of liquid dairy manure using asequential continuous-stirred tank reactor system. J. Chem. Technol. Biotechnol.82, 758–766.

Youn, J.H., Shin, H.S., 2005. Comparative performance between temperature-phasedand conventional mesophilic two-phased processes in terms of anaerobicallyproduced bioenergy from food waste. Waste Manage. Res. 23, 32–38.

Yu, Z., Morrison, M., 2004a. Improved extraction of PCR-quality community DNAfrom digesta and fecal samples. Biotechniques 36, 808–812.

Yu, Z., Morrison, M., 2004b. Comparisons of different hypervariable regions of rrsgenes for use in fingerprinting of microbial communities by PCR-denaturinggradient gel electrophoresis. Appl. Environ. Microbiol. 70, 4800–4806.

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

Yu, Z., Michel, F.C., Hansen, G., Wittum, T., Morrison, M., 2005b. Development andapplication of real-time PCR assays for quantification of genes encodingtetracycline resistance. Appl. Environ. Microbiol. 71, 6926–6933.

Yu, Z., Garcia-Gonzalez, R., Schanbacher, F.L., Morrison, M., 2008. Evaluations ofdifferent hypervariable regions of archaeal 16S rRNA genes in profiling ofmethanogens by archaea-specific PCR and denaturing gradient gelelectrophoresis. Appl. Environ. Microbiol. 74, 889–893.

Yu, Z., Morrison, M., Schanbacher, F.L., 2009. Production and utilization of methanebiogas as renewable fuel. In: Vertès, A.A., Blaschek, H.P., Yukawa, H., Qureshi, N.(Eds.), Biomass to Biofuels: Strategies for Global Industries. John Wiley & Sons,Inc., Hoboken, NJ, pp. 403–433.