Sie sind auf Seite 1von 10

New Biotechnology  Volume 33, Number 1  January 2016 RESEARCH PAPER

Volume ratios between the thermophilic and

the mesophilic digesters of a temperature-

Research Paper
phased anaerobic digestion system affect their
performance and microbial communities
Wen Lv1, Wenfei Zhang3 and Zhongtang Yu1,2
Department of Animal Sciences, The Ohio State University, Columbus, OH 43210, USA
Environmental Science Graduate Program, The Ohio State University, Columbus, OH 43210, USA
Department of Biostatistics, Columbia University, New York, NY 10032, USA

An experimental temperature-phased anaerobic digestion (TPAD) system, with the thermophilic

digester operated at neutral pH and with a balanced acidogenesis and methanogenesis (referred to as
NT-TPAD), was evaluated with respect to the microbial communities and population dynamics of
methanogens when digesting dairy cattle manure at 15-day overall system hydraulic retention time
(HRT). When fed a manure slurry of 10% total solid (TS), similar system performance, 36–38% volatile
solid (VS) removal and 0.21–0.22 L methane g1 VS fed, was achieved between a 5-day and 7.5-day HRT
for the thermophilic digester. However, the thermophilic digester achieved a greater volumetric biogas
yield when operated at a 5-day RT than at a 7.5-day HRT (6.3 vs. 4.7 L/L/d), while the mesophilic digester
had a stable volumetric biogas yield (about 1.0 L/L/d). Each of the digesters harbored distinct yet
dynamic microbial populations, and some of the methanogens were significantly correlated with
methane productions. Methanosarcina and Methanosaeta were the most important methanogenic
genera in the thermophilic and the mesophilic digesters, respectively. The microbiological findings may
help understand the metabolism that underpins the anaerobic processes within each of the two digesters
of TPAD systems when fed dairy manure.

Introduction dynamic and their composition and structure can be shaped by

Known and utilized for many decades, anaerobic digestion (AD) many factors including feedstock, design, and operation [5]. The
has received renewed interests recently in the pursuit of renewable complex and dynamic features of the microbial communities
energy and mitigation of greenhouse gas emissions. Indeed, AD is create many challenges to increase biogas yield and enhance
one of the few dual-purpose technologies that can simultaneously process stability [4,6]. For example, the four major guilds of
produce biogas from biomass wastes and reduce/prevent environ- microbes, hydrolytic bacteria, acidogens, syntrophic acetogens,
mental pollutions [1,2]. While the main goal of AD was previously and methanogens, differ in several aspects relevant to AD, includ-
waste management, its current emphasis is on cost-effective pro- ing growth rate and ability to survive and function at low pH and
duction of biogas as renewable energy. New designs and opera- high ammonia concentration that are often observed during AD
tional strategies have been sought after to increase biogas yield operation [7,8]. As a result, conditions and operations that meet
while maintaining process stability [1,3]. A complex microbial the requirements of hydrolytic bacteria and acidogens often over-
community is responsible for the sequential steps (hydrolysis, whelm the fastidious and exigent methanogens [9–11]. Thus,
acidogenesis, syntrophic acetogenesis, and methanogenesis) of single-stage digesters often suffer from poor system performance
AD process [4]. The microbial communities in digesters are also [12] and susceptibility to high organic loading rates (OLR) [11],
due to their inabilities to provide the conditions that are needed
Corresponding author: Yu, Z. ( for optimal growth and function of each microbial guild.
1871-6784/ß 2015 Elsevier B.V. All rights reserved. 245
RESEARCH PAPER New Biotechnology  Volume 33, Number 1  January 2016

Two-stage anaerobic digester systems provide opportunities to manure slurry for 144 days in a previous study [22]. Briefly, this AT-
create and maintain two separate intra-system environments in the TPAD system had a thermophilic digester operated at 508C and
two digesters. The advantages of two-stage digesters over single- acidic pH mainly for hydrolysis/acidogenesis and a mesophilic
stage digesters have been demonstrated in a numerous studies digester operated at 358C for balanced hydrolysis/acidogenesis
where two-stage anaerobic digester systems outperformed single- and methanogenesis. The content from these two digesters was
stage digesters when readily digestible feedstock were fed [9,13–17]. used as the seed sludge for the two digesters of the NT-TPAD system.
For livestock manure, including its co-digestion with a readily The thermophilic and the mesophilic seed sludge contained 11.52%
digestible feedstock, a few studies also showed that two-stage diges- and 9.33% total solid (TS), and 9.67% and 7.44% volatile solid (VS),
ters achieved greater biogas yield and VS removal than single-stage respectively.
digesters [17–19]. In most studies on two-stage digesters, both stages The feedstock was dairy manure (including both feces and urine)
were operated at similar temperatures, about 358C for mesophilic or slurry that was prepared as described previously [22]. Briefly, fresh
Research Paper

558C for thermophilic operations. However, temperature-phased dairy manure was collected on a daily basis from the Waterman
AD (TPAD) with the first stage operated at thermophilic temperature Dairy Center, The Ohio State University, where Jersey cattle were fed
(typically about 558C) and the second stage at mesophilic tempera- the same total mixed ration (TMR, based on dry matter, 50.00% corn
ture (mostly about 358C) is an emerging technology attracting much silage, 4.50% alfalfa hay, 21.00% co-product of corn wet milling,
recent research interest. This is because the thermophilic digester of 9.05% ground corn, 4.64% soybean meal, 1.30% Aminoplus1,
a TPAD system can enhance destruction of solid and pathogen 1.30% soy hulls, 0.38% fat, 2.01% vitamin and minerals). The
[20,21], while the mesophilic digester facilitates stable biogas pro- average TS and VS contents of the collected manure were 14.61%
duction [11,21]. (w/v) and 12.81% (w/v), with variations less than 1.01% and 0.74%,
A few studies have been reported on TPAD systems that digest respectively. Prior to use, manure was diluted to desired TS and VS
cattle manure as the solely feedstock [22–24]. In these studies, contents using tap water and mixed thoroughly into slurry with a
TPAD systems achieved 36–40% VS removal and 0.21–0.22 L 10% TS to reduce potential clogging in digesters [24] and improve
methane g1 VS fed when operated with a short HRT (14–15 days) the substrate accessibility to the microbial community [25].
and when fed a dairy cattle manure slurry containing about 10%
TS. Sung and Santha [24] also showed that the final digestate met Start-up, operation, and sampling
the criteria specified for Class A biosolids. In a recent study [23], Both digesters of the NT-TPAD system were filled to their working
comparing the performance of two TPAD systems with the ther- volumes with the seed sludge as done previously [22]. Briefly, the
mophilic digester operated at either acidic or neutral pH (referred thermophilic digester received a mixture of the thermophilic and
to as AT-TPAD and NT-TPAD, respectively), a NT-TPAD system was the mesophilic seed sludge (0.5 L each), while the mesophilic di-
shown to outperform an AT-TPAD system. The objectives of the gester received only the mesophilic seed sludge (2 L). The thermo-
present study were to further evaluate NT-TPAD systems in terms philic digester was also inoculated with the mesophilic seed sludge
of performance when operated at two different HRT by varying the to augment the methanogenesis activity. The NT-TPAD system was
volume ratio between the thermophilic and the mesophilic diges- then operated in a feed-batch mode on a daily basis [23]. Contents of
ters and to investigate the microbial (bacterial and archaeal) both digesters were manually mixed before and after feeding. Biogas
communities and population dynamics of methanogens in each production and effluent pH from each digester were recorded daily
digester. Correlation between system performance and methano- before feeding. During the startup, about 150 ml of sludge was
gen populations was also examined. recycled between the two digesters when acidification (as indicated
by pH decrease below 6.5) was observed in the thermophilic digest-
Materials and methods er. The sludge recycling ended when the thermophilic digester
Experimental setup, seed sludge, and feedstock reached and maintained neutral pH without any recycling.
Two bench-scale digesters were made from two Nalgene polypro- The operation of the NT-TPAD system was essentially the same
pylene wide-mouth bottles of 4.3-L capacity (Fisher Scientific, PA) as described in the previous study [23] and was separated into three
and used as the thermophilic and the mesophilic digesters of the sequential periods: startup period, period 1, and period 2 (Fig. 1).
NT-TPAD system [22]. Each of the digesters had a feeding port, a As defined in a previous study [26], the NT-TPAD system was
sampling port, and a biogas outlet. The feeding port and the considered to have reached steady state in each period when
sampling port were each sealed by a rubber stopper to keep both the variation of daily biogas production by both digester was less
digesters airtight except during feeding and sampling. The biogas than 10% for five consecutive days without any upward or down-
produced by each digester was collected and measured by water ward trend. During the startup period, the HRT (also solid reten-
displacement using an inverted graduated cylinder, which was tion time SRT because of mixing of digester content) was 33%
maintained at room temperature. The first stage thermophilic longer than that applied during periods 1 and 2, while the OLR was
digester and the second stage mesophilic digester were started 75% of that applied during periods 1 and 2. The overall HRT/SRT
with working volumes of 1 and 2 L, respectively. The two digesters was maintained at 15 days for all the three operation periods. Two
were placed in two adjacent water baths to maintain their respec- volume ratios of thermophilic over mesophilic digesters were used
tive temperatures (508C for the thermophilic digester and 358C for during the study: 1:2 during the startup period and period 1, and
the mesophilic digester). 1:1 during period 2. Six biogas samples and six sludge samples were
The seed sludge for the thermophilic digester (pH about 6.0) and collected from each digester at its stable state during each of these
mesophilic digester (pH about 7.5) was the content of two digesters three periods of operation: at days 24, 25, 28, 29, 32, and 33 during
of an AT-TPAD system that had been operated using dairy cattle the startup; at days 77, 78, 82, 86, 87, and 92 of period 1; at days

New Biotechnology  Volume 33, Number 1  January 2016 RESEARCH PAPER

Thermophilic digester Mesophilic digester

Startup Period 1 Period 2 Startup Period 1 Period 2

Working volume (L) 1 1 1.5 2 2 1.5

Operation duration (d) 1-34 35-93 94-144 1-34 35-93 94-144
Operation temp (°C) 50 50 50 35 35 35
OLR (g VS/L/d) 13.09 17.54 11.69 6.57 8.55 11.58
HRT/SRT (d) 6.7 5 7.5 13.3 10 7.5

Research Paper
The operational parameters of the NT-TPAD system used in this study. The organic loading rate (OLR) for the mesophilic digester was calculated from the volatile
solid (VS) content of the effluent of the thermophilic digester. HRT, hydraulic retention time; SRT, solid retention time.

135, 136, 139, 140, 143, and 144 of period 2 (see Fig. 1 for the by repeated bead beating, and the released DNA was purified by
duration of each period). The sludge samples were aliquoted and enzymatic digestion of protein and RNA and columns. The DNA
stored at 808C until being analyzed. quality was evaluated using agarose gel (0.8%) electrophoresis.
Concentration of each DNA sample was quantified using Nano-
System performance analysis Drop spectrophotometer (Thermo Scientific, MA). The average
Contents of CH4 and CO2 in the biogas samples were analyzed DNA yields (mg/ml sample) were 44.23, 44.79, and 60.74 for the
using gas chromatography (GC) [27], and methane yields were startup, period 1, and period 2, respectively, of the thermophilic
calculated from corresponding biogas productions and methane digester samples, and 42.12, 41.14, and 22.58 for the startup,
contents. To compare the performance of the two digesters, only period 1, and period 2, respectively, of the mesophilic digester
volumetric biogas and methane yields (L/L/d) were used. Sludge samples diluted to a final concentration of about 50 ng/ml with
samples were centrifuged and their supernatants were analyzed for Tris–EDTA buffer. The archaeal and the bacterial communities
volatile fatty acids (VFAs) using GC [28]. Contents of TS and VS of represented in each DNA sample were analyzed using domain-
each sludge sample were determined following the standard meth- specific PCR and denaturing gradient gel electrophoresis (DGGE)
ods [29], and solid removal was calculated for each sludge sample. [31,32]. Briefly, the V3 hyper-variable region of 16S rRNA genes of
bacteria and archaea was separately amplified by PCR using both
Analysis of microbial community and methanogen populations Archaea- or Bacteria-specific primers (Table 1). The amplicons were
Community DNA was extracted from each sludge sample as verified using agarose gel (1.5%) electrophoresis and then resolved
described previously [30]. Briefly, the microbial cells were lysed using a 7.5% polyacrylamide gels (37.5:1) containing a 40–60%

Primers and primer/probe sets used in this study for DGGE and qPCR analyses
Analysis Target Primer Sequence (50 –30 ) Amplicon length (bp) Refs
Bacteria GC-Eub357F CCTACGGGAGGCAGCAGa 195 [31]
Methanobacterium Mbt-210F CCAAGCCWKTRATCTGTACG 148 [61]
Methanoculleus Mcu-298F GGAGCAAGAGCCCGGAGT 314 [62]
Methanosarcina MB1b CGGTTTGGTCAGTCCTCCGG 271 [63]
Methanosaeta MS1b CCGGCCGGATAAGTCTCTTGA 272 [63]
Total archaea TArc-787F ATTAGATACCCSBGTAGTCC 271 [64]
GC-Arc344F and GC-Eub357F both had a 40nt GC-clamp attached at their 50 end. 247
RESEARCH PAPER New Biotechnology  Volume 33, Number 1  January 2016

gradient of denaturants (100% denaturants consisting of 40% [vol/ digesting dairy manure [22–24,40]. This study investigated the
vol] formamide and 7 M urea). The DGGE gels were stained with effect of two different volume (or HRT) ratios between the ther-
GelStar (BioWhittaker Molecular Applications, Rockland, Maine) mophilic and the mesophilic digester on TPAD performance and
as per the manufacturer’s instruction, and the images were cap- microbial communities/populations. The whole study lasted for
tured using a FluorChem Imager (Alpha Innotech Corp., San more 140 days, with each ratio being evaluated at stable stage as
Leandro, Calif.). The DGGE images were analyzed using BioNu- indicated by stable biogas production. The results showed that
merics version 5.1 (Applied Maths, Inc., TX) to determine the variation of HRT in each digester of A TPAD system can affect
banding profiles of each sample [33]. The DGGE banding profiles TPAD performance and microbial communities, and that certain
were clustered using the Jaccard similarity coefficient and the groups of microbes are particularly associated with the perfor-
Unweighted Pair Group Method with Arithmetic Mean (UPGMA) mance of the digesters of a TPAD system when digesting diary
clustering method and also subjected to principal component manure.
Research Paper

analysis (PCA) as described in previous studies [22,33].

The population sizes of four methanogen genera, including System performance
Methanobacterium, Methanoculleus, Methanosaeta, and Methanosar- In a previous study, a TPAD system with the thermophilic digester
cina, the WSA2/ArcI group, and total archaea in each DNA sample operated at neutral pH (NT-TPAD) was found to outperform an AT-
were quantified using respective specific primer/TaqMan probe TPAD system when fed dairy manure and operated with a HRT/SRT
sets (Table 1) [22,23]. Briefly, a sample-derived qPCR standard was of 15 days (7.5 day for each digester). In this study two volume
prepared for each target genus or group from a pooled DNA sample ratios between the thermophilic and the mesophilic digesters, 1:2
representing all the DNA samples to be analyzed [34]. Each stan- (in period 1) and 1:1 (in period 2), were evaluated with respect to
dard was subsequently quantified and serially diluted with Tris– digester performance and effect on microbiome. Besides, because
EDTA buffer to achieve a 9-log range of concentrations (i.e. 100– 50, 55, and 608C in the thermophilic digester resulted in similar
108 amplicon copies/ml). The population size of each targeted biogas yields from dairy manure [22], the thermophilic digester of
genus or group present in each DNA sample was quantified in the NT-TPAD system was maintained at 508C to save energy input
three technical replicates against its corresponding qPCR standard while the mesophilic digester at 358C throughout this study. The
included on the same qPCR plate. Abundances of methanogen VFA concentrations were consistently low (<5 mM) in both diges-
genera and the total archaea in each DNA sample were expressed as ters throughout the study (Fig. 2), rendering stable neutral pH
copies/ml sludge sample. around 7.2 during the steady state of all the three experimental
Statistical analysis Volumetric yields (L/L/d) of both biogas and methane were
Correlations between populations of methanogen genera/total significantly higher after the startup period (Table 2), correspond-
archaea and methane production were also assessed using Pearson ing to the higher OLR during periods 1 and 2. Consistent with the
correlation coefficients. Correlations were determined from the previous study [23], in all the three operation periods, the ther-
data for each digester and from the data for the entire NT-TPAD mophilic digester achieved much greater (>4 folds) volumetric
system. Significant correlation was declared at p < 0.05. T-test was yields of biogas and methane productions than the mesophilic
used to make pairwise comparisons among means of system digester. Although the total biogas and methane productions were
performances (biogas/methane production and VS removal) and similar between the two volume ratios in periods 1 and 2, the
population abundances of methanogen genera/total archaea. p- thermophilic digester had a greater volumetric biogas and meth-
Values less than 0.05 were defined as statistically significant. ane yield in period 1 (1:2 volume ratio and 5 days of HRT/SRT) than
Standard deviations (SD) were also calculated for all the data. in period 2 (1:1 volume ratio and 7.5 days of HRT/SRT). In all the
three periods, the thermophilic digester also yielded greater meth-
Results and discussion ane content in the biogas than the mesophilic digester. The
Hydrolysis is a key rate-limiting step of AD as the hydrolysis thermophilic digester also achieved a significantly higher daily
products can be rapidly fermented by acidogenic bacteria and VS removal rate (g/L/d) than the mesophilic digester in all the
then converted to methane by methanogens [35]. This is particu- three periods (Table 3), suggesting more rapid conversion of VS to
larly true for AD of feedstocks containing cellulosic biomass. Dairy biogas in the thermophilic digester. This is probably attributable to
cattle manure belongs to this category of feedstock because it the higher organic loading rate in period 1 than in period 2. The
contains no or little readily fermentable carbohydrates, but large daily VS removal rate of the entire NT-TPAD system increased
amounts of fibers that escaped digestion by dairy cattle and fecal significantly from the startup to period 2. However, when the
bacterial cells. Activated sludge is similar with respect to its recal- HRT/SRT of the thermophilic digester was increased from 5 days in
citrance to AD despite its difference in chemical composition. period 1 to 7.5 in period 2, the daily VS removal rate decreased in
Several studies have shown that TPAD systems can enhance hy- the thermophilic digester but increased in the mesophilic digester.
drolysis of organic solids of activated sludge and increase methane This change in daily VS removal rate is consistent with the change
production [36,37]. The TPAD systems used in AD of activated in biogas production from both digesters and can be explained by
sludge typically consist of a thermophilic digester maintained at the decreased organic loading rate in the thermophilic digester
50–658C with a short (2–4 days) HRT and a mesophilic digesters and concomitant increased organic loading rate in the mesophilic
operated at about 358C with longer (10–15 day) HRT [38,39]. No digester. Such a shift in daily VS removal rate between the two
full-scale TPAD system has been used in digesting dairy manure, digesters is also expected given the low proportion of readily
but laboratory-scale TPAD systems were shown to be effective in digestible substrate present in dairy manure (readily digestible

New Biotechnology  Volume 33, Number 1  January 2016 RESEARCH PAPER

9 9
a b
8 8
7 7
6 6
5 5
4 4
3 3
2 2
1 1
0 0

Research Paper
Average VFA concentrations (mM) in the thermophilic (a) and the mesophilic (b) digesters of the NT-TPAD system during the startup period (open), period 1 (filled
gray), and period 2 (filled black). Error bars present SD (n = 6). Total VFA includes all the six individual VFAs shown in this figure.

Volumetric biogas and methane productions (L/L/d) of the NT-TPAD system during different periods
Period Thermophilic digester Mesophilic digester NT-TPAD system
Biogas production CH4 production CH4 (%) Biogas production CH4 production CH4 (%) Biogas production CH4 production
a a a a a
Startup 5.60  0.17 3.53  0.17 62.99 0.84  0.09 0.57  0.07 68.08 2.43  0.06 1.56  0.08a
b b b b b
1 6.78  0.12 3.98  0.14 58.61 1.08  0.03 0.65  0.04 60.32 2.98  0.05 1.76  0.06b
c c b b b
2 4.77  0.18 2.84  0.08 59.60 1.01  0.07 0.69  0.05 68.00 2.89  0.12 1.76  0.06b
Data were presented as the means  SD (n = 6). Values within each column with different superscripts were significantly different (p < 0.05).

VS removal of the NT-TPAD system during different periods
Period VS removal (g VS L1 working volume d1)
Thermophilic digester Mesophilic digester NT-TPAD system
Startup 4.70  0.43 1.19  0.22 2.36  0.09
a a
1 6.69  0.53 0.94  0.28 2.86  0.11a
b b
2 4.28  0.55 1.76  0.70 3.02  0.11b
Note: Data were presented as the means  SD (n = 6). Values in each column with different superscripts were significantly different (p < 0.05). Only periods with the same overall OLR
(periods 1 and 2) were compared with each other.

polysaccharides have been digested in the rumen and colon of thermophilic digester. Compared to the AT-TPAD system [22], the
cows). NT-TPAD system greatly improved production of methane, espe-
Overall, the thermophilic digester accounted for 66.5–78.0% of cially in the thermophilic digester, probably owing to the balanced
the total VS removal and 75.4–80.7% of the total methane pro- hydrolysis/acidogenesis and methanogenesis therein. In the pres-
duction of the NT-TPAD system, while the mesophilic digester ent study, the thermophilic digester was seeded with seed sludge
only accounted for 22.0–33.5% of the total VS removal and 19.3– from both an acidic thermophilic digester and a mesophilic di-
24.6% of the total methane production. These results are consis- gester, and the effluent of the mesophilic digester content was
tent with most previous studies on TPAD treating activated sludge recycled back to the thermophilic digester during the startup
and dairy manure [38,41,42], and the results also collectively period. Such a seeding regimen and recycling might have helped
indicate that the thermophilic digester reached a dynamic balance establishing a microbial community that can carry out balanced
between hydrolysis/acidogenesis and methanogenesis, resulting hydrolysis/acidogenesis and methanogenesis. The low concentra-
in efficient VS removal and methane production and both hydro- tions of hydrolyzed substrates and VFA in the effluent of the
lysis/acidogenesis and methanogenesis were enhanced at the ele- thermophilic digester might lead to consistent biogas and meth-
vated temperature (508C). The mesophilic digester was also ane productions from the mesophilic digester during these two
operated with balanced hydrolysis/acidogenesis and methanogen- periods.
esis, but it probably only scavenged low concentrations of hydro- The greater VS removal in the thermophilic digester than in the
lyzed substrates and the VFA carried over by the influx from the mesophilic digester is consistent with the results of TPAD systems 249
RESEARCH PAPER New Biotechnology  Volume 33, Number 1  January 2016

fed activated sludge [38,41,42], but these studies showed that the bacterial and archaeal communities between the thermophilic and
mesophilic digester produced more methane than the thermo- the mesophilic digesters is expected and consistent with the previ-
philic digester. A previous study on AD of dairy manure using a ous reports [22,23,46,47]. Both the archaeal and the bacterial com-
TPAD system also showed that the thermophilic digester produced munities present in both the thermophilic and the mesophilic
more methane than the mesophilic digester [24]. This discrepancy digesters generally differed between periods 1 and 2 (Figs 3 and
is probably attributed to the difference in the HRT in the thermo- 4). This indicates that the two different HRT/SRTs applied to both
philic digester: 2–3 days in the aforementioned studies vs. 5–7.5 the thermophilic and the mesophilic digesters resulted in two
days in the present study. The longer HRT allowed for the ther- different microbial communities in both digesters even though
mophilic digester in the present study probably have allowed the overall OLR remained unchanged.
methanogens to establish therein. This explanation is consistent Operational and chemical factors affect the anaerobes in certain
with the slow growth rate of methanogens and the finding that ways. It is believed that there are many ambiguous points in AD
Research Paper

increased HRT in the thermophilic digester increases methane that are not yet known. Therefore, by controlling and monitoring
yield therein [43]. It should be noted that dairy manure contains each of the microbiological, operational, and chemical parame-
pathogens, including pathogens causing mastitis of dairy cattle ters, AD performance may be enhanced [7]. In the present study,
[44]. A previous study using bench-scale TPAD systems showed the chemical parameters, digester performance, and microbial
that Class A biosolid can be produced from digest dairy manure communities were monitored to help understand the AD process
[24] and activated sludge [42,45]. A longer HRT in the thermo- dynamics. Over the course of the operation of each period, the
philic digester will require more energy input, but it can enhance methanogenic and the bacterial communities were rather stable
inactivation of pathogens. within each digester even though the HRT and OLR varied. The
stable microbial communities might be attributable to the stable
Microbial community DGGE profiles and low VFA concentrations and neutral high pH. The premise is
The succession of both bacterial and archaeal (including methano- consistent with the previous finding that pH is among the most
gens) communities in each of the two digesters was evaluated using influential factors that affect the bacterial communities in anaer-
DGGE. Clustering analysis of the DGGE profiles showed that the obic digesters [48,49]. These results also suggest that although the
thermophilic and the mesophilic digesters harbored different ar- overall HRT/SRT remained unchanged, changes in HRT/SRT of
chaeal and bacterial communities during all the three periods (Figs 3 each digester shifted the populations and metabolic activities of
and 4). The results of PCA analysis of the DGGE profiles were important microbial groups in both the thermophilic and the
consistent with those of the clustering analysis of DGGE profiles mesophilic digesters.
(Fig. 5). Except for the samples from period 1 that had relatively
similar (with a similarity about 90%) archaeal profiles between the Effect of HRT on abundance of methanogens
two digesters (Fig. 5e), both the archaeal and the bacterial commu- Methanogens are the direct producers of methane and they are
nities were different between the two digesters during all the more susceptible to variations of AD operation than bacteria
three operation periods. The temperature-dependent difference in [6,10]. The population dynamics of common methanogen genera

Archaeal DGGE banding patterns of the thermophilic (T) and the mesophilic (M) digesters of the NT-TPAD system during the startup period (S), period 1 (1), and
period 2 (2). The same amount of DNA template (50 ng each) was used in the PCR, and the same volume of PCR product was resolved on the DGGE gel. The
similarity scale indicates the Jaccard similarity coefficient.

New Biotechnology  Volume 33, Number 1  January 2016 RESEARCH PAPER

Research Paper
Bacterial DGGE banding patterns of the NT-TPAD system. The same amount of DNA template (50 ng each) was used in the PCR, and the same volume of PCR
product was resolved on the DGGE gel. The similarity scale indicates the Jaccard similarity coefficient.

was examined by deterring their abundance using qPCR. In the might also enable Methanosarcina to be more competitive in the
thermophilic digester, Methanosarcina and Methanobacterium were thermophilic digester [56]. The abundances of Methanosarcina and
the two most predominant methanogen genera during all the Methanobacterium were significantly higher during periods 1 than
three periods (Fig. 6a). These two genera had similar abundance period 2. It remains to be determined if the change in abundance
at the beginning of the operation, but the population of Metha- of these two genera was caused sole by changes in HRT/SRT.
nosarcina became significantly larger than that of Methanobacter- In the mesophilic digester, Methanosaeta was the most predom-
ium after the startup period (p < 0.05). This indicates that methane inant methanogen genus with a population significantly larger
was produced through both the acetoclastic and the hydrogeno- than that of Methanosarcina during the startup period and period 1
trophic methanogenesis pathways during the startup period, but (p < 0.05). However, Methanosaeta could not be detected during
thereafter acetoclastic methanogenesis became dominating as period 2, and Methanosarcina became the most predominant
hydrolysis/acidogenesis intensified after the overall OLR was in- methanogen genus of this period (Fig. 6b). Again, the shortening
creased. This premise agrees with the finding that Methanosarcina of the HRT/SRT from 10 days in period 1 to 7.5 days in period 2
is generally responsible for methane production in anaerobic might be the main reason that caused washout of Methanosaeta.
digestion systems when acetate concentrations are high [50–52]. Additionally, Methanosaeta spp. generally have lower Ks values
Despite decreased predominance after the startup period, the than Methanosarcina spp., which means that Methanosaeta spp.
population of Methanobacterium might have played an important generally have higher affinities for substrate [56]. This might have
role in methanogenesis, as indicated by its population size enabled Methanosaeta to be more competitive during the startup
throughout the study, possibly due to enhanced hydrogen pro- and period 1 when the VFA concentrations (primarily acetate)
duction at thermophilic temperature [53]. The obligate acetoclas- were consistently low but not when the VFA concentrations were
tic genus, Methanosaeta, was not detected during any period in the high in period 2 (Fig. 2b). This premise is consistent with the
thermophilic digester. This is consistent with the finding of previ- observation of Conklin et al. [57]. The abundance of Methanobac-
ous studies [22,23,54,55] and agrees with most species of Metha- terium was similar to that of Methanosaeta at the beginning of the
nosaeta being mesophilic organisms. In addition, its slow growth startup but became significantly lower after the startup period,
and long generation time (3.5–9 days) might be another reason while the population of Methanosarcina was consistent throughout
[56]. Furthermore, high growth rate and the ability to utilize both the study. Methanobacterium spp. generally has longer generation
acetoclastic and hydrogenotrophic methanogenesis pathways times than Methanosarcina. Thus, the decreased Methanobacterium 251
RESEARCH PAPER New Biotechnology  Volume 33, Number 1  January 2016
Research Paper

PCA analysis results of the profiles of six samples from each digester in each periods. (a–d) Triangles = startup; inverted triangles = period 1; circles = period 2.
(e,f) triangles = thermophilic digester; inverted triangles = mesophilic digester. Values in parentheses of each axis are percentages of variances contributed by
each principal component (PC). Some of the samples had identical profiles so they were superposed.

A. B.
1.E+10 a a a 1.2E+10
a a a,b a b a
1.E+08 a b b a,b
1.E+06 b
1.E+04 a a
1.E+02 2.0E+09 a a a
a a a b a a a b
1.E+00 0.0E+00

Population abundances (16S rRNA gene copies/ml sludge sample) of four methanogen genera and total archaea in the thermophilic (a) and the mesophilic (b)
digesters of the NT-TPAD system during the startup period (open), period 1 (filled gray), and period 2 (filled black). The R2 values for the qPCR standard curves
ranged from 0.995 to 0.998. The PCR efficiencies were 91.6%, 87.8%, 81.2, 71.9, and 86.8 for the qPCR of Methanobacterium, Methanoculleus, Methanosarcina,
Methanosaeta, and total archaea, respectively. Error bars presented SD (n = 6). Columns with different letters in the same figure indicate significant differences
(p < 0.05) in the population abundance of a specific group among the three periods.

population in period 2 might be attributed, at least partially, to the The WSA2/ArcI group, first discovered in several AD systems
shortened HRT/SRT. Given their high abundance in the mesophi- treating municipal sludge and proposed to have a hydrogenotrophic
lic digester, acetoclastic methanogens (Methanosaeta and Metha- metabolism [58], was found in AD reactors fed municipal sludge
nosarcina) were probably the major biogas producer therein. [5,59,60], but it was not detected from the pooled community DNA

New Biotechnology  Volume 33, Number 1  January 2016 RESEARCH PAPER

sample that was used to prepare the sample-derived qPCR standards. decrease of the Methanosaeta population from the startup to period
Thus, this group of methanogens was not analyzed in the samples. 2 coincided with significant increase in methane production
Because PCR is a very sensitive detection method, the lack of during these two periods. In the whole NT-TPAD system, the
detection of this methanogen group suggests that the WSA2/ArcI abundance of Methanosaeta was also negatively correlated (Pearson
group might be a minor group at most, contributing little to correlation efficient = 0.6363, p < 0.0001) with the methane
methanogenesis from dairy cattle manure. production. This may be explained by the ecophysiological fea-
tures of this genus: slow growth, requirement of long HRT/SRT,
Correlation between abundance of methanogens and methane and preference for low concentration of VFA, particularly acetate.
Theoretically, correlation exists between AD performance and Conclusions
individual groups of microbes. However, few studies correlated Balanced hydrolysis/acidogenesis and methanogenesis were estab-

Research Paper
AD performance to specific groups of bacteria or methanogens. In lished in the thermophilic digester of a NT-TPAD system fed dairy
one recent study, some bacterial phylotypes, but not specific manure when it was operated at neutral pH. The thermophilic
bacterial taxa (species, genus or high taxonomic units), were found digester functioned as the major reactor for methane production
to be weakly linked with some process performance parameters rather than as a pretreatment reactor. The NT-TPAD system had a
[46]. In the present study, correlation between the system perfor- greater volumetric biogas production rate than that reported for an
mance and the abundance of known methanogen genera was AT-TPAD system. At the same overall HRT/SRT, 1:2 and 1:1 volume
examined for each of the two digesters and for the whole TPAD ratios between the thermophilic and the mesophilic digesters led
system. In the thermophilic digester, abundances of both Metha- to similar biogas production and VS removal. A volume ratio of
nosarcina (Pearson correlation coefficient = 0.6110, p = 0.0071) 1:2, corresponding to 5 and 10 days of HRT/SRT for the thermo-
and Methanobacterium (Pearson correlation coefficient = 0.4748, philic and the mesophilic digester, respectively, will be preferred
p = 0.0465) were positively correlated with methane productions. because a smaller thermophilic digester and thus less energy input
The abundances of both Methanosarcina and Methanobacterium is needed to maintain the thermophilic temperature. Each digester
were significantly higher during period 1 than during period 2 of the NT-TPAD system harbored distinct yet dynamic microbial
(Fig. 6), which mirrors the significantly higher methane produc- populations. Methanobacterium and Methanosarcina were the most
tion during period 1 than period 2 (Table 2). Given that Metha- important methanogenic genera in the thermophilic digester,
nosarcina and Methanobacterium were the two most predominant while Methanobacterium, Methanosarcina, and Methanosaeta were
methanogen genera in the thermophilic digester, it was hypothe- the most important in the mesophilic digesters.
sized that methane was primarily produced by these two genera
through both acetoclastic and hydrogenotrophic methanogenesis Acknowledgements
pathways. In the mesophilic digester, the abundance of Methano- This work was partially supported by a DOE grant (award number:
saeta (Pearson correlation efficient = 0.6014, p = 0.0083) was neg- DE-FG36-05GO85010) and a North East SUN grant (award
atively correlated with the methane production. The significant number: 52110-8512).

[1] Appels L, Lauwers J, Degrève J, Helsen L, Lievens B, Willems K, et al. Anaerobic [12] Angelidaki I, Ellegaard L, Ahring BK. Applications of the anaerobic digestion
digestion in global bio-energy production: potential and research challenges. process. In: Ahring BK, editor. Biomethanation II. Berlin/Heidelberg: Springer;
Renew Sustain Energy Rev 2011;15:4295–301. 2002. p. 1–33.
[2] Yasui H, Komatsu K, Goel R, Matsuhashi R, Ohashi A, Harada H. Minimization of [13] Demirel B, Yenigun O. Two-phase anaerobic digestion processes: a review. J
greenhouse gas emission by application of anaerobic digestion process with Chem Technol Biotechnol 2002;77:743–55.
biogas utilization. Water Sci Technol 2005;52:545–52. [14] Ke S, Shi Z, Fang HHP. Applications of two-phase anaerobic degradation in
[3] Weiland P. Biogas production: current state and perspectives. Appl Microbiol industrial wastewater treatment. Int J Environ Pollut 2005;23:65–80.
Biotechnol 2010;85:849–60. [15] Parawira W, Reed JS, Mattiasson B, Björnsson L. Energy production from
[4] Lv W, Schanbacher FL, Yu Z. Putting microbes to work in sequence: recent agricultural residues: high methane yields in pilot-scale two-stage anaerobic
advances in temperature-phased anaerobic digestion processes. Bioresour Tech- digestion. Biomass Bioenergy 2008;32:44–50.
nol 2010;101:9409–14. [16] Zahller JD, Bucher RH, Ferguson JF, Stensel HD. Performance and stability of
[5] Nelson MC, Morrison M, Yu Z. A meta-analysis of the microbial diversity two-stage anaerobic digestion. Water Environ Res 2007;79:488–97.
observed in anaerobic digesters. Bioresour Technol 2011;102:3730–9. [17] Nielsen HB, Mladenovska Z, Westermann P, Ahring BK. Comparison of two-
[6] Chen Y, Cheng JJ, Creamer KS. Inhibition of anaerobic digestion process: a stage thermophilic (688C/558C) anaerobic digestion with one-stage thermophil-
review. Bioresour Technol 2008;99:4044–64. ic (558C) digestion of cattle manure. Biotechnol Bioeng 2004;86:291–300.
[7] Amani T, Nosrati M, Sreekrishnan TR. Anaerobic digestion from the viewpoint [18] Lateef SA, Beneragama N, Yamashiro T, Iwasaki M, Umetsu K. Batch anaerobic
of microbiological, chemical, and operational aspects – a review. Environ Rev co-digestion of cow manure and waste milk in two-stage process for hydrogen
2010;18:255–78. and methane productions. Bioprocess Biosyst Eng 2013;1–9.
[8] O’Flaherty V, Collins G, Mahony T. The microbiology and biochemistry of [19] Bertin L, Grilli S, Spagni A, Fava F. Innovative two-stage anaerobic process for
anaerobic bioreactors with relevance to domestic sewage treatment. Rev Environ effective codigestion of cheese whey and cattle manure. Bioresour Technol
Sci Bio/Technol 2006;5:39–55. 2013;128:779–83.
[9] Zuo Z, Wu S, Zhang W, Dong R. Effects of organic loading rate and effluent [20] Yilmaz T, Yuceer A, Basibuyuk M. A comparison of the performance of meso-
recirculation on the performance of two-stage anaerobic digestion of vegetable philic and thermophilic anaerobic filters treating papermill wastewater. Bior-
waste. Bioresour Technol 2013;146:556–61. esour Technol 2008;99:156–63.
[10] Chen S, Zamudio Canas EM, Zhang Y, Zhu Z, He Q. Impact of substrate over- [21] Suryawanshi PC, Chaudhari AB, Kothari RM. Thermophilic anaerobic digestion:
loading on archaeal populations in anaerobic digestion of animal waste. J Appl the best option for waste treatment. Crit Rev Biotechnol 2010;30:31–40.
Microbiol 2012;113:1371–9. [22] Lv W, Zhang W, Yu Z. Evaluation of system performance and microbial
[11] Speece RE, Boonyakitsombut S, Kim M, Azbar N, Ursillo P. Overview of anaero- communities of a temperature-phased anaerobic digestion system treating dairy
bic treatment: thermophilic and propionate implications. Water Environ Res manure: thermophilic digester operated at acidic pH. Bioresour Technol 2013;
2006;78:460–73. 142:625–32. 253
RESEARCH PAPER New Biotechnology  Volume 33, Number 1  January 2016

[23] Lv W, Zhang W, Yu Z. Evaluation of system performances and microbial [45] Santha H, Sandino J, Shimp GF, Sung S. Performance evaluation of a ‘sequential-
communities of two temperature-phased anaerobic digestion systems treating batch’ temperature-phased anaerobic digestion (TPAD) scheme for producing
dairy manure. Bioresour Technol 2013;143:431–8. class A biosolids. Water Environ Res 2006;78:221–6.
[24] Sung S, Santha H. Performance of temperature-phased anaerobic digestion [46] Lee S-H, Kang H-J, Lee YH, Lee TJ, Han K, Choi Y, et al. Monitoring bacterial
(TPAD) system treating dairy cattle wastes. Water Res 2003;37:1628–36. community structure and variability in time scale in full-scale anaerobic diges-
[25] Angelidaki I, Ahring BK. Methods for increasing the biogas potential from the ters. J Environ Monit 2012;14:1893–905.
recalcitrant organic matter contained in manure. Water Sci Technol 2000;41: [47] Pervin HM, Dennis PG, Lim HJ, Tyson GW, Batstone DJ, Bond PL. Drivers of
189–94. microbial community composition in mesophilic and thermophilic tempera-
[26] Wen Z, Frear C, Chen S. Anaerobic digestion of liquid dairy manure using a ture-phased anaerobic digestion pre-treatment reactors. Water Res 2013;47:
sequential continuous-stirred tank reactor system. J Chem Technol Biotechnol 7098–108.
2007;82:758–66. [48] Hori T, Haruta S, Ueno Y, Ishii M, Igarashi Y. Dynamic transition of a metha-
[27] Patra AK, Yu Z. Effects of essential oils on methane production and fermentation nogenic population in response to the concentration of volatile fatty acids in a
by, and abundance and diversity of, rumen microbial populations. Appl Environ thermophilic anaerobic digester. Appl Environ Microbiol 2006;72:1623–30.
Microbiol 2012;78:4271–80. [49] Supaphol S, Jenkins SN, Intomo P, Waite IS, O’Donnell AG. Microbial commu-
[28] Oelker ER, Reveneau C, Firkins JL. Interaction of molasses and monensin in nity dynamics in mesophilic anaerobic co-digestion of mixed waste. Bioresour
alfalfa hay- or corn silage-based diets on rumen fermentation, total tract Technol 2011;102:4021–7.
Research Paper

digestibility, and milk production by Holstein cows. J Dairy Sci 2009;92:270–85. [50] Jetten MSM, Stams AJM, Zehnder AJB. Acetate threshold values and acetate
[29] American Public Health Association, American Water Works Association, Water activating enzymes in methanogenic bacteria. FEMS Microbiol Lett 1990;6:
Environment Federation, editors. Standard methods for the examination of 339–44.
water and wastewater. 21st ed., Washington, DC: American Public Health [51] McMahon KD, Stroot PG, Mackie RI, Raskin L. Anaerobic codigestion of mu-
Association; 2005. nicipal solid waste and biosolids under various mixing conditions – II: microbial
[30] Yu Z, Morrison M. Improved extraction of PCR-quality community DNA from population dynamics. Water Res 2001;35:1817–27.
digesta and fecal samples. Biotechniques 2004;36:808–12. [52] Vavilin VA, Qu X, Mazeas L, Lemunier M, Duquennoi C, He P, et al. Metha-
[31] Yu Z, Morrison M. Comparisons of different hypervariable regions of rrs genes nosarcina as the dominant aceticlastic methanogens during mesophilic anaer-
for use in fingerprinting of microbial communities by PCR-denaturing gradient obic digestion of putrescible waste. Antonie Van Leeuwenhoek 2008;94:
gel electrophoresis. Appl Environ Microbiol 2004;70:4800–6. 593–605.
[32] Yu Z, Garcia-Gonzalez R, Schanbacher FL, Morrison M. Evaluations of different [53] O-Thong S, Mamimin C, Prasertsan P. Effect of temperature and initial pH on
hypervariable regions of archaeal 16S rRNA genes in profiling of methanogens biohydrogen production from palm oil mill effluent: long-term evaluation and
by archaea-specific PCR and denaturing gradient gel electrophoresis. Appl microbial community analysis. Electron J Biotechnol 2011;14:1–9.
Environ Microbiol 2008;74:889–93. [54] Demirel B, Scherer P. The roles of acetotrophic and hydrogenotrophic metha-
[33] Cressman MD, Yu Z, Nelson MC, Moeller SJ, Lilburn MS, Zerby HN. Interrela- nogens during anaerobic conversion of biomass to methane: a review. Rev
tions between the microbiotas in the litter and in the intestines of commercial Environ Sci Bio/Technol 2008;7:173–90.
broiler chickens. Appl Environ Microbiol 2010;76:6572–82. [55] Petersen SP, Ahring BK. Acetate oxidation in a thermophilic anaerobic sewage-
[34] Yu Z, Michel Jr FC, Hansen G, Wittum T, Morrison M. Development and sludge digestor: the importance of non-aceticlastic methanogenesis from ace-
application of real-time PCR assays for quantification of genes encoding tetra- tate. FEMS Microbiol Lett 1991;86:149–58.
cycline resistance. Appl Environ Microbiol 2005;71:6926–33. [56] Yu Z, Morrison M, Schanbacher FL. Production and utilization of methane
[35] Rittmann BE. Opportunities for renewable bioenergy using microorganisms. biogas as renewable fuel. In: Vertès AA, Qureshi N, Blaschek HP, Yukawa H,
Biotechnol Bioeng 2008;100:203–12. editors. Biomass to biofuels: strategies for global industries. Hoboken, NJ: John
[36] Paul E, Carrère H, Batstone DJ. Thermal methods to enhance biological treat- Wiley & Sons, Inc; 2010. p. 403–33.
ment processes. Biological sludge minimization and biomaterials/bioenergy [57] Conklin A, Stensel HD, Ferguson J. Growth kinetics and competition between
recovery technologies. John Wiley & Sons, Inc; 2012: 373–404. Methanosarcina and Methanosaeta in mesophilic anaerobic digestion. Water
[37] Ge H, Jensen PD, Batstone DJ. Increased temperature in the thermophilic stage Environ Res 2006;78:486–96.
in temperature phased anaerobic digestion (TPAD) improves degradability of [58] Chouari R, Le Paslier D, Daegelen P, Ginestet P, Weissenbach J, Sghir A. Novel
waste activated sludge. J Hazard Mater 2011;187:355–61. predominant archaeal and bacterial groups revealed by molecular analysis of an
[38] Ge H, Jensen PD, Batstone DJ. Temperature phased anaerobic digestion anaerobic sludge digester. Environ Microbiol 2005;7:1104–15.
increases apparent hydrolysis rate for waste activated sludge. Water Res 2011; [59] Wilkins D, Lu XY, Shen Z, Chen J, Lee PK. Pyrosequencing of mcrA and archaeal
45:1597–606. 16S rRNA genes reveals diversity and substrate preferences of methanogen
[39] Pervin HM, Batstone DJ, Bond PL. Previously unclassified bacteria dominate communities in anaerobic digesters. Appl Environ Microbiol 2015;81:
during thermophilic and mesophilic anaerobic pre-treatment of primary sludge. 604–13.
Syst Appl Microbiol 2013;36:281–90. [60] Kuroda K, Hatamoto M, Nakahara N, Abe K, Takahashi M, Araki N, et al.
[40] Dugba PN, Zhang R. Treatment of dairy wastewater with two-stage anaerobic Community composition of known and uncultured archaeal lineages in anaer-
sequencing batch reactor systems – thermophilic versus mesophilic operations. obic or anoxic wastewater treatment sludge. Microb Ecol 2015;69:586–96.
Bioresour Technol 1999;68:225–33. [61] Nelson MC. An integrated investigation of the microbial communities under-
[41] Ge H, Jensen PD, Batstone DJ. Pre-treatment mechanisms during thermophilic– pinning biogas production in anaerobic digestion systems. Environmental
mesophilic temperature phased anaerobic digestion of primary sludge. Water Sciences Graduate Program. Columbus: The Ohio State University; 2011.
Res 2010;44:123–30. [62] Franke-Whittle IH, Goberna M, Insam H. Design and testing of real-time PCR
[42] Riau V, De la Rubia MA, Perez M. Temperature-phased anaerobic digestion primers for the quantification of Methanoculleus, Methanosarcina, Methanother-
(TPAD) to obtain class A biosolids: a semi-continuous study. Bioresour Technol mobacter, and a group of uncultured methanogens. Can J Microbiol 2009;55:
2010;101:2706–12. 611–6.
[43] Riau V, De la Rubia MA, Perez M, Martin A, Borja R. Modelling of the tempera- [63] Shigematsu T, Tang Y, Kawaguchi H, Ninomiya K, Kijima J, Kobayashi T, et al.
ture-phased batch anaerobic digestion of raw sludge from an urban wastewater Effect of dilution rate on structure of a mesophilic acetate-degrading methano-
treatment plant. J Environ Sci Health A Tox Hazard Subst Environ Eng genic community during continuous cultivation. J Biosci Bioeng 2003;96:
2012;47:221–7. 547–58.
[44] Toth JD, Aceto HW, Rankin SC, Dou Z. Short communication: survey of animal- [64] Yu Y, Lee C, Kim J, Hwang S. Group-specific primer and probe sets to detect
borne pathogens in the farm environment of 13 dairy operations. J Dairy Sci methanogenic communities using quantitative real-time polymerase chain
2013;96:5756–61. reaction. Biotechnol Bioeng 2005;89:670–9.