International Journal of Coal Geology 154155 (2016) 205212

Contents lists available at ScienceDirect

International Journal of Coal Geology

journal homepage: www.elsevier.com/locate/ijcoalgeo

The effect of coal rank on biogenic methane potential and

microbial composition
Steven J. Robbins a,, Paul N. Evans a, Joan S. Esterle b, Suzanne D. Golding b, Gene W. Tyson a,
a
b

Australian Centre for Ecogenomics, School of Chemistry and Molecular Biosciences, The University of Queensland, St. Lucia, Queensland 4072, Australia
School of Earth Sciences, The University of Queensland, St. Lucia, Queensland 4072, Australia

a r t i c l e

i n f o

Article history:
Received 17 July 2015
Received in revised form 4 January 2016
Accepted 4 January 2016
Available online 5 January 2016
Keywords:
Coal bed methane
Coal rank
Acetate
16S rRNA gene
Microbial activity

a b s t r a c t
Demand for natural gas is expected to increase faster than any other fossil fuel over the coming decades.
Australian coal bed methane (CBM) resources are among the largest in the world and are already being utilized
to meet increasing demand. The majority of methane contained within CBM producing coal beds is microbially
generated and low rank coals are often associated with higher bioavailability. However, the results of previous
studies are conicting, and it is unclear how or if coal rank has an effect on microbial community structure.
Here, enrichment cultures grown on coals spanning 13 different ranks (lignite to bituminous) were characterized
with 16S rRNA gene amplicon sequencing and combined with volatile fatty acid (VFA) and headspace methane
measurements to understand the effect of coal rank on CBM microbial communities. Our results show that there
is a signicant negative correlation between nal methane yield and rank, suggesting that the bioavailability of
the coal organic material decreases with increasing thermal maturity. The concentration of VFAs generally increased with decreasing rank and revealed that community composition was signicantly correlated with the
concentration of specic VFAs (e.g. acetate). These data suggest that the observed higher methane production
from lower rank coals is linked to increased concentrations of low molecular weight acids desorbing from the
coal. The increase in low molecular weight VFAs was associated with the enrichment of specic taxa known to
specialize in the degradation of low molecular weight acids and alcohols.
2016 Elsevier B.V. All rights reserved.

1. Introduction
Coal is a complex sedimentary rock composed of plant matter that
has been ooded, buried, heated, and compacted over the course of
millions of years. In general, coal can be classied by several dening
characteristics, which reect its burial history and the geological forces
that have acted on it over time (Diessel, 1992; O'Keefe et al., 2013;
Surez-Ruiz et al., 2012; Taylor, 1998; Ward, 1984). These characteristics include the original plant matter (peat) from which the coal is derived, the minerals incorporated into the coal, and coal rank, the
extent of diagenetic transformation experienced by the coal, primarily
as a result of compaction and heating (O'Keefe et al., 2013). Coals are
classied into a rank system reecting their thermal maturity, with
rank designations increasing from lignite, sub-bituminous, bituminous,
to anthracite. As the original peat is coalied, individual plant components of similar chemical composition are compacted into semidiscrete units within the coal called macerals. Over time, geothermal
heating of the coal results in predictable geochemical changes in the organic chemistry of each maceral. For example, vitrinite, derived from
Corresponding authors.
E-mail addresses: steven.robbins@uqconnect.edu.au (S.J. Robbins), g.tyson@uq.edu.au
(G.W. Tyson).

http://dx.doi.org/10.1016/j.coal.2016.01.001
0166-5162/ 2016 Elsevier B.V. All rights reserved.

woody, lignin-rich tissue of plants (Taylor, 1998), undergoes a high degree of condensation of its aromatic structure with increasing rank,
resulting in an increase in shine/reectance (O'Keefe et al., 2013;
Strpo et al., 2011). For this reason, vitrinite reectance, the percentage
incident light reected from the surface of the coal, is commonly used as
a proxy for rank and maturity.
As coal increases in rank, heteroatoms such as oxygen, sulfur, and nitrogen are lost. At the same time, the aromatic lignin-derived structure of
the coal condenses to form higher order polyaromatic compounds and
eventually aromatic sheets (Strpo et al., 2011). The increase in aromaticity, coupled with a loss of heteroatom moieties amenable to microbial
attack, is associated with a loss in bioavailability. As a result, it has been
assumed that coals of higher rank are less bioavailable and will produce
less methane than coals of lower rank (Fallgren et al., 2013; Strpo
et al., 2011). To date, very few studies have been conducted to validate
this assumption. While a meta-analysis of published methanogenesis
rates showed support for the proposed inverse correlation between
rank and methane biogenesis potential (Strpo et al., 2011), other studies have shown that higher rank can be associated with increased methane production (Fallgren et al., 2013) or found no correlation (Wawrik
et al., 2012). In the latter cases, the authors suggested that low molecular
weight hydrocarbons trapped within the coal matrix were used, rather
than the coal matrix itself. The nature of the relationship between rank

206

S.J. Robbins et al. / International Journal of Coal Geology 154155 (2016) 205212

and methane production remains unclear. While it has been shown that
coal properties can affect microbial community composition (Fry et al.,
2009; Susilawati et al., 2014), the effect of coal rank on microbial community composition and function has not been fully addressed.
Here, 14 enrichment cultures grown using coals of different rank
(peat to bituminous) as the sole carbon and energy source were characterized using 16S rRNA gene amplicons. Methane production, as well as
volatile fatty acid composition, was coupled with microbial community
composition analysis to provide insight into how CBM producing microorganisms and gas production are inuenced by coal rank.
2. Materials and methods
2.1. Core collection and processing
Thirteen coal samples spanning a range of ranks from lignite to high
volatile bituminous were collected from sites located in the Bowen
Basin (Australia), Kutai Basin (Indonesia), Ohai Basin (New Zealand),
and Greymouth Basin (New Zealand) as described in Table 1. All
samples were collected and characterized as part of previous studies
(Anggara et al., 2014; Dmyterko, 2014; Rahmat, in progress). Coals
were obtained for processing as small blocks ( 2 cm3) and stored in
air for several months prior to use in these experiments. Due to the
small size of each coal block, the outer core could not be removed. Preparation of core material for use as an enrichment culture substrate was
carried out in an anaerobic chamber (COY Laboratory Products, MI,
USA) under an atmosphere containing 95% nitrogen and 5% hydrogen.
To ensure that the coal did not become contaminated with volatile substances during storage or preparation, the chamber atmosphere had
been fully purged several months before the commencement of this
experiment, and no volatile substances (e.g. ethanol) were used in the
chamber. The core material was ground using a bur-style coffee grinder
and then a mortar and pestle. The resulting granules were sieved to between 150 and 500 m in size, and maceral analysis and vitrinite reectance measurements were performed at the School of Earth Sciences
(University of Queensland) in accordance with Australian Standards
AS 2856.2-1998 and AS 2856.3-2000, respectively (Table 2). A wood
sample of Huon Pine (HP) was included in the analysis as it has a rank
of approximately 0.1.
2.2. Establishment of enrichment cultures
Enrichment cultures were prepared using each coal as the dominant
carbon and energy source as previously described (Green et al., 2008;
Papendick et al., 2011). In total, 3 g/L 2-((1,3-dihydroxy-2(hydroxymethyl)propan-2-yl)amino)ethanesulfonic acid (TES) and
1 g/L sodium bicarbonate were added to the growth medium (Tanner,

2007) and the pH was adjusted to 7.4. Resazurin (1 g/L stock) was
added as a redox indicator to 50 g/L nal concentration. Media was
boiled for 2 min and cooled to room temperature under a stream
of oxygen-free nitrogen with sodium sulde nonahydrate reductant
added to a nal concentration of 0.3 mM. The cooled anaerobic medium
(9.8 ml) was dispensed into 36 ml serum bottles (Wheaton, USA) containing 0.25 g of coal granules. Serum bottles were sealed with butyl
rubber stoppers and aluminum crimps. The headspace was exchanged
with oxygen-free nitrogen gas and serum bottles were autoclaved for
20 min at 120 C and 1 atm of pressure. Prior to inoculation, the volatile
fatty acid composition (VFA) of the supernatants of three un-inoculated
enrichments containing each coal were measured using standard
methods (Eaton et al., 2005) in order to determine the extent of labile
carbon desorbing from the coal, as well as to ensure that ethanol was
not present in these enrichments. VFAs measured include ethanol,
propanol, and butanol, as well as acetic, propionic, iso-butyric, butyric,
iso-valeric, valeric, and hexanoic acids.
To create a diverse inoculum capable of degrading a wide range of
organic substrates, biomass from termite gut contents (degutted and
homogenized using sterile instruments), digester uid, koala feces,
sediment from the University of Queensland Lake, and the PK18 enrichment culture were combined. The PK18 enrichment community is derived from formation waters obtained from a CBM production well
located in the Surat Basin, Australia that was previously shown to produce methane using coal as the dominant carbon and energy source
(unpublished data). To ensure that equal cell densities from each sample were present in the inoculum, the samples were diluted to remove
debris or decrease the cell density, and estimates of cell density were
made using Thoma slide count. After sample mixing, approximately
1 107 cells in total from each source were combined into a master inoculum. One milliliter of the master inoculum was stored at 80 C for
subsequent 16S rRNA gene proling. Enrichment cultures containing
each of the 13 coals or Huon Pine sample were inoculated in triplicate
with 0.2 ml of the master inoculum and incubated without shaking
at 37 C for 50 days. Cumulative headspace methane concentration
was measured for each enrichment culture at regular intervals until
methane production had ceased, at which point microbial biomass
was harvested for 16S rRNA gene sequencing. Methane generated
from the degradation of organic matter from the inoculum was determined using controls containing only the inoculum without coal. Uninoculated controls containing each coal served as negative control for
methane desorbing from the coal.
2.3. Methane measurement
Cumulative headspace methane concentrations were measured
using a Shimadzu GC-2014 gas chromatograph (Shimadzu, Japan) tted

Table 1
Provenance, depth, and vitrinite reectance of one wood (Huon Pine, HP) and 13 coal samples. Samples listed as having No depth were taken from open cut mines where the coal was
positioned 4060 m below surface. Basin, formation, and depth measurement are not applicable (NA) for Huon Pine. Vitrinite reectance was measured as maximum telovitrinite
reectance and is here reported as Rmax for all coal regardless of rank for consistency.
Sample ID

Country

Basin

Coal Measures/formation

Depth (meters)

Vitrinite reectance (Rmax)

HP
JE2
KSA
ST7
JE4
ST3
JE9
IP
JE10
BM
CD
JE11
MVL
CP

Australia
New Zealand
Indonesia
Indonesia
New Zealand
Indonesia
New Zealand
Australia
New Zealand
Australia
Australia
New Zealand
Australia
Australia

NA
Ohai
Kutai
Kutai
Ohai
Kutai
Greymouth
Bowen Basin
Greymouth
Bowen Basin
Bowen Basin
Greymouth
Bowen Basin
Bowen Basin

NA
Morley coal measures
Balikpapan formation
Balikpapan formation
Morley coal measures
Balikpapan formation
Rewanui coal measures
Rangal coal measures
Rewanui coal measures
Moranbah coal measures
Rangal coal measures
Rewanui coal measures
Rangal coal measures
Rangal coal measures

NA
34 m
No depth
5860 m
804 m
134136 m
Outcrop
No depth
outcrop
No depth
No depth
Outcrop
No depth
No depth

0.1
0.5
0.5
0.5
0.5
0.5
0.7
0.9
0.9
1.1
1.3
1.5
1.5
1.9

S.J. Robbins et al. / International Journal of Coal Geology 154155 (2016) 205212

207

Table 2
Petrological measurements of the coals used as enrichment culture substrates.
Sample ID

Country

JE2
KSA
ST7
JE4
ST3
JE9
IP
JE10
BM
CD
JE11
MVL
CP

New Zealand
Indonesia
Indonesia
New Zealand
Indonesia
New Zealand
Australia
New Zealand
Australia
Australia
New Zealand
Australia
Australia

Moisture
(%ad)

Ash
(%ad)

Volatile
(%ad)

Vitrinite
%

Liptinite
%

Inertinite
%

Mineral
%

26.2
5.4
13.3
16.5
10.1
10.6
13.6
0.9
9.3
11.7
7.1
18
12.3

7.3
9.2
0.8
3.1
0.6
4.1
11.8
4.7
7.6
10.5
6.9
16.2
10.7

31.1
46
39.3
34.7
42
38
25.3
35.2
27.1
20.9
19.3
16.4
12

79
65
83.4
94
79.8
100
35.2
90
72.9
53.4
96
48.5
48.2

8
14
8
3
6.8
0
5.1
9
1.2
1.8
0
0
0

6
4
6.6
1
12.2
0
53.8
0
21.5
42.1
0
36.9
45.4

7
17
2
2
1.2
0
6
1
4.4
2.7
4
14.6
6.4

with a Shincarbon micropack column (Restek catalog number 19,808)

and ame ionization detector (FID) with helium as the carrier gas
(BOC, Australia). Gas standards of 0%, 1.03% and 15.1% methane were
injected into the GC at atmospheric pressure to produce calibration
curves. Total headspace methane was calculated by multiplying the
total quantity of methane (mol) measured in the injection volume
(100 L) contained in a gas-tight syringe by total headspace volume
(26 ml). A conservative estimate for the theoretical yield of methane
that could be produced from the quantity of acetate measured in each
enrichment was calculated assuming that both acetate carbons were
converted to methane (i.e. 1 CH3C2OOH = 2 CH4).
2.4. DNA extraction, sequencing, and community proling
Microbial biomass from each enrichment culture was harvested for
DNA extraction by centrifuging 5 ml of the culture at 3270 g for
30 min. DNA extraction was performed on pelleted microbial biomass
and coal granules using the MoBio Powersoil DNA isolation kit as per
the manufacturer's instructions. Extracted DNA was used as a target
for polymerase chain reaction (PCR) using the universal primer sequences 926F and 1392R (Kunin et al., 2010) modied at the end to contain Illumina specic adapter sequence (926F: 5-TCGTCGGCAGCGTCAG
ATGTGTATAAGAGACAGAAACTYAAAKG AATTGRCGG-3 and 1392wR:
5-GTCTCGTGGGCTCGGGTCTCGTGGGCTCGGAGATGTGTATAAGAGACA
GA CGGGCGGTGWGTRC-3). The 16S rRNA gene library preparation
workow from Illumina (#15044223 Rev.B) was performed as described. In the initial stage of this workow, PCR products of ~ 500 bp
were amplied according to the specied workow with an alteration
in PCR reactions used to substitute Q5 HotStart High-Fidelity 2X
Master Mix (NEB) in standard PCR conditions. Resulting PCR amplicons
were puried using Agencourt AMPure XP beads (Beckman Coulter).
Puried DNA was indexed with unique 8 bp barcodes using the Illumina
Nextera XT V2 Index Kit Set AD (Illumina FC-131-1002) in standard
PCR conditions with 2X Q5 HotStart HiFidelity MasterMix (NEB).
Indexed amplicons were pooled together in equimolar concentrations
and sequenced on MiSeq Sequencing System (Illumina) using paired
end sequencing with V3 300 bp chemistry at the Australian Centre for
Ecogenomics according to manufacturer's protocol.
Only the forward read was used to generate community composition proles. The rst 20 bases of each read were trimmed to remove
primer sequences, and quality trimmed to remove poor quality
sequence using a sliding window of 4 bases with an average base
quality above 15 using Trimmomatic (Bolger et al., 2014). All reads
were hard trimmed to 250 bases, and any with less than 250
bases was excluded. To assign operational taxonomic units (OTUs),
clustering was performed at 97% identity using QIIME's pick_open_
reference_otus.py workow using default parameters with taxonomy assignment and alignment features suppressed. The resulting
OTU table was ltered to remove any OTU with an abundance of

less than 0.05%. BLASTn (Altschul et al., 1990) was used to assign a
GreenGenes taxonomy (DeSantis et al., 2006) to the representative
sequences from each OTU.
2.5. Statistical analyses
All statistical analyses were performed in R v3.1.2 (R_Core_team,
2013). A linear model was used to determine the coefcient of determination (r2) and p-value for each regression. A heatmap showing the
relative abundances of all OTUs present at a minimum of 2.5% in at
least one sample was generated using the CRAN packages RColorBrewer
(Neuwirth, 2011) and gplots (Warnes et al., 2014). Differences in
community composition were further explored through principal
components analysis of Hellinger transformed OTU relative abundances
(Legendre and Gallagher, 2001) using the CRAN package vegan (Dixon,
2003), and the environmental tting function was used to correlate
community composition with metadata parameters.
3. Results
3.1. Establishment of enrichment cultures and methane production
In total, one wood (HP) and 13 coal samples of different rank from
Australia, New Zealand, and Indonesia (Tables 1; 2) were assessed for
their ability to produce methane in microbial enrichment cultures seeded with a diverse inoculum derived from termite gut, anaerobic digester
uid, koala feces, lake sediment, and a CBM enrichment culture. Both

Fig. 1. Average total headspace methane concentrations for replicate enrichment cultures.
Samples for community composition analysis were taken at Day 50 for all enrichments.
Error bars represent the standard error for each set of triplicate enrichment cultures and
the vitrinite reectance of each coal shown in parentheses. Un-inoculated controls
containing each coal did not produce detectable quantities of methane.

208

S.J. Robbins et al. / International Journal of Coal Geology 154155 (2016) 205212

the New Zealand and Indonesian coals showed high vitrinite contents
(65100%), while the Australian coals were higher in inertinite (21
53%). Methane production above the controls was observed in all coal
enrichments except IP and JE11 (Fig. 1), which had vitrinite reectance
values of 0.9 and 1.5, respectively (Table 1). Interestingly, JE11 appeared
under the microscope to be highly tectonized compared to the rest of
the rank suite. The majority of methane was produced after Day 10
and the total yield was between 1 and 25 mol of methane per gram
coal above controls by Day 50. Lower rank cultures tended to show
higher acetate concentrations in the media supernatant before inoculation (Table 3) and yield more methane at the end of enrichment than
high rank enrichments (Fig. 1). No other VFA was detected in any
enrichment. The three highest producing enrichments (HP, JE2 and JE
4) yielded an average of 20.1 6.56 mol of methane per gram coal
and showed an average acetate concentration of 24.5 6.55 mg/L.
Comparison of the nal methane yield to the amount of methane that
could theoretically be produced if acetate was completely converted
to methane showed that only the HP, KSA, ST3, BM, and MVL enrichments exceeded the theoretical yield (Table 3).
Examination of the vitrinite reectance (a proxy for rank) of each
coal versus the average nal methane yield from its respective enrichments (Table 3) revealed a signicant inverse correlation between the
two variables (r2 = 0.72, p = 7.2 105; Fig. 2A). The best t of this
line was logarithmic, reaching an asymptote at a reectance of ~ 1. In
contrast, linear tting showed a signicant, but weaker correlation
(r2 = 0.0.52, p = 2.2 103; Fig. 2B) when all ranks were considered.
Interestingly, this relationship improved if only samples below a rank of
one were considered (r2 = 0.82, p = 6.8 104; Fig. 2B dotted line).
Signicant but somewhat weaker linear correlations were also observed
between vitrinite reectance and acetate concentration (r2 = 0.34, p =
0.017; Fig. 3A), and acetate concentration and nal methane yield (r2 =
0.39, p = 0.017; Fig. 3B). Logarithmic tting of vitrinite reectance and
nal acetate concentration produced approximately the same result as
linear tting, raising the coefcient of determination from 0.34 to 0.38.
3.2. Characterization of enrichment culture community composition
In order to examine the relationship between community composition, rank, and methane production, enrichment cultures were characterized using 16S rRNA gene amplicon sequencing at the end of gas
production (Day 50; Fig. 4). Although no single operational taxonomic unit (OTU) was found in all enrichments, members of the genera Tissierella and Pelobacteraceae, order Bacteroidales, and family
Anaerolinaceae were identied in all enrichments grown on coal, but
Table 3
Measurements of vitrinite reectance, acetate concentration prior to inoculation, and nal
methane yield. Acetate concentrations represent the concentration of acetate in the medium supernatant prior to inoculation. Theoretical methane yield is calculated by assuming
that both acetate carbons will be converted to methane. Yield difference is calculated as
the difference between the theoretical methane yield from acetate and the observed yield.
Observed
Initial
Sample Vitrinite
Theoretical
Yield
ID
reectance acetate methane yield
methane yield
difference
(mg/L) (molmethane/gcoal) (molmethane/gcoal)
(Rmax)
HP
JE2
KSA
ST7
JE4
ST3
JE9
IP
JE10
BM
CD
JE11
MVL
CP

0.1
0.5
0.5
0.5
0.5
0.5
0.7
0.9
0.9
1.1
1.3
1.5
1.5
1.9

21.3
20.1
3.4
10.6
32
6.5
20.8
4.1
5.6
3.2
3.4
3.8
3.5
2.7

26.4
20.5
9.3
10.9
13.3
11.8
4.5
0.3
4.1
9.7
2.2
0.2
5.4
2.5

23.1
21.8
3.7
11.5
34.7
7.1
22.6
4.4
6.1
3.5
3.7
4.1
3.8
2.9

3.3
1.3
5.6
0.6
21.4
4.7
18.1
4.1
2.0
6.2
1.5
3.9
1.6
0.4

were not observed in the Huon Pine (HP) sample. Putative hydrolyzers
of coal, those potentially capable of aromatic hydrocarbon degradation,
were present in several enrichments and included members of the genera Pseudomonas (03%), Thauera (03%), Dechloromonas (017%), and
families Peptococcaceae (09%) and Aeromonadaceae (012%). The
original inoculum and HP enrichments showed very little overlap in
terms of community composition with enrichments grown using coal
as the dominant carbon and energy source. Most of the OTUs present
in coal enrichments were rare members of the inoculum, with only
one OTU, a member of candidate phylum WWE1, detected in both
the inoculum and enrichments. Both the inoculum and HP enrichments
were dominated by hydrogenotrophic methanogens from the
order Methanobacteriales, while the coal enrichments (reectance
N0.1) were dominated by acetoclastic methanogens from the order
Methanosarcinales.
3.3. The effect of coal rank on community composition
Principal components analysis (PCA) was used to identify key differences in community composition between enrichments grown on coals
of different rank (Fig. 5). All coal enrichments clustered away from the
Huon Pine (HP) sample, further illustrating the unique community
composition of the coal enrichments. It is likely that HP contains
compounds such as simple sugars that would not be found in coal
(Stout et al., 1988), accounting for the disparity between coal and
wood enrichments. In addition, JE2 and JE4 communities clustered
away from the other enrichments, which appear to be correlated with
higher acetate concentrations in these samples. Environmental tting
analysis revealed signicant correlations between community composition and rank (r2 = 0.49, p = 0.009), acetate concentration prior to inoculation (r2 = 0.71, p = 0.001), and nal methane production (r2 =
0.70, p = 0.003).
In order to examine community composition within enrichments
grown using coal as a substrate, a separate PCA plot was constructed
without HP or the original inoculum (Fig. 6). Samples appeared to
cluster mainly by country of origin, and environmental tting analysis
supported this correlation (r2 = 0.41, p = 0.02). However, stronger
correlations with community composition were identied between
rank (r2 = 0.56, p = 0.022), acetate concentration prior to inoculation
(r2 = 0.67, p = 0.006), average nal methane yield (r2 = 0.73, p =
0.003). Operational taxonomic units (OTUs) related to members of
the class Gammaproteobacteria, family Aeromonadaceae, and family
ML635J-40 appeared to be positively correlated with increasing rank,
but negatively associated with nal methane production and acetate
concentrations. In contrast, OTUs belonging to the bacterial family
Bacteroidales, and the genus Pelobacter, and the archaeal genus
Methanosaeta appeared to be positively correlated both with higher
methane production and acetate concentrations, but negatively
correlated with rank (Fig. 6).
In order to further investigate the effect of acetate in the coal on the
microbial community, environmental tting of acetate concentrations
prior to inoculation was applied to the bacterial community and
archaeal community separately. While no correlation between bacterial
community composition and rank, acetate concentration, or methane
production could be found, all of these variables were correlated with
archaeal community composition. Additionally, acetate concentrations
were strongly correlated with the relative abundance of methanogens
(r = 0.61, p = 5.6 104). Conversely, correlations between acetate
concentration and acetoclastic methanogens (r = 0.30, p = 0.025)
and between vitrinite reectance and total methanogen abundance
(r = 0.26, p = 0.034) were signicant but weak.
4. Discussion
The identication of key physical features affecting biogenic methane
formation from coal will be critical in assessing the economic viability of

S.J. Robbins et al. / International Journal of Coal Geology 154155 (2016) 205212

209

Fig. 2. Comparison of vitrinite reectance of a Huon Pine wood sample and 13 coals with the nal methane yield from the enrichment culture. Coal samples are shown as circles and the
Huon Pine as a star.

sources of coal bed methane (CBM). A detailed characterization of how

these features inuence coal bioavailability and microbial community
composition will aid in the development of strategies for stimulating additional methane production from active coal-associated microbial communities (Hamilton et al., 2015). However, our understanding of the
factors that govern biogenic methane formation from coal is currently
incomplete. Here, the microbial communities actively producing
methane when grown using coals of different rank as the sole carbon
and energy source were characterized. Coal rank was shown to be inversely correlated with nal methane yields and suggests that acetate
is an important factor in determining microbial community composition. This nding adds to a growing body of evidence that indicates
that coal rank and acetate are strong determinants of CBM microbial
community composition and biogenic methane potential from coal
(Orem et al., 2014; Ulrich and Bower, 2008).
All enrichment cultures, other than JE11 and IP, produced between 1
and 25 mol of methane per gram coal above controls (Fig. 1), similar to
previously published yields from coal enrichment cultures (Green et al.,
2008; Jones et al., 2008; Papendick et al., 2011; Penner et al., 2010).
Although bicarbonate may have served as a carbon source, it was
added equally to each culture and any differences in methane production can be attributed to the utilization of organic matter from coal.
The JE11 and IP coals are comparatively high rank, showing a vitrinite
reectance of 1.5 and 0.9 respectively, which may account for the
observed lack of methane production. In addition, the high inertinite
content of IP (53.8%) may have resulted in a decrease in bioavailability,
as inertinite shows a high degree of ring condensation and loss of
heteroatoms compared to other macerals and is therefore expected to
be recalcitrant (Strpo et al., 2011). In contrast, the high rate of tectonic
shearing observed in the JE11 coal may have reduced its porosity and
restricted absorption capacity and microbial access to the poor space.

Vitrinite reectance was strongly negatively correlated with nal

methane yield, supporting the hypothesis that lower rank coals should
allow higher biogenic methane production than coals of higher rank. Interestingly, we show that coals in the diagenetic and early catagenetic
phases produce the highest methane yields, potentially reecting a
loss of bioavailable carbon in late catagenesis. This trend has been
shown previously for New Zealand coals, which became depleted in hydrolysable fatty acids toward late catagenesis (Glombitza et al., 2009a).
This explanation is consistent with the observation that lower rank
coals from this study are associated with higher acetate concentrations
and methane yields (Figs. 2; 3). These data strongly suggest that the increase in nal methane yield is directly related to higher acetate concentrations in low rank coal.
Characterization of the community composition in each enrichment at the end of gas production also provided insight into the inuence of coal rank on community structure (Figs. 4; 6). Members
of the bacterial class Gammaproteobacteria, family Aeromonadaceae,
and family ML635J-40 were found to be associated with an increase
in coal rank, while members of family Bacteroidales and genus
Pelobacter were associated with a decrease in rank. These changes
in community composition with coal rank may reect a difference
in organic substrate range between the microbial groups, which is
expected to change with increasing rank of the coal (Strpo et al.,
2011). Specically, the abundance of Pelobacter and Bacteroidales in
lower rank coals is likely due to their ability to use low molecular
weight compounds from the coal, including acetate. For example, acetate is a required co-substrate of Pelobacter spp. when growing on
several linear alcohols (Schink, 2006). Similarly, the observed clustering of samples by their country of origin may be a result of organic
facies differences in the coal between distant, geologically distinct
locations (e.g. source plant material and inorganic content). In

Fig. 3. Correlation between vitrinite reectance, acetate concentration prior to inoculation, and nal methane production in a Huon Pine wood sample and 13 coals. Coal samples are
shown as circles and the Huon Pine as a star.

210

S.J. Robbins et al. / International Journal of Coal Geology 154155 (2016) 205212

Fig. 4. Heatmap of the relative abundance of community members from each enrichment culture. Only operational taxonomic units present at 2.5% relative abundance in a least one
sample are shown. The color intensity of each box indicates the relative abundance. Samples were collected from each enrichment on Day 50. Coal rank and observed methane yield
per gram of coal are shown in parentheses.

contrast, environmental tting analysis showed that the bacterial

community was not signicantly correlated with changes in coal
rank, acetate concentrations, or methane production. Although

bacterial community composition as a whole was not correlated

with these variables, it is likely that specic bacterial community
members were inuenced by rank and acetate concentration.

Fig. 5. Principal components analysis of Hellinger transformed relative abundances for 13 coal and one Huon Pine wood enrichment cultures on Day 50.

S.J. Robbins et al. / International Journal of Coal Geology 154155 (2016) 205212

211

Fig. 6. PCA of Hellinger transformed relative abundances for all enrichment cultures, excluding the Huon Pine wood sample, HP, and the original inoculum.

The archaeal community, composed almost entirely of acetoclastic

methanogens in all coal enrichments, was found to correlate with
rank, acetate concentration prior to inoculation, and nal methane
yield (Fig. 6). These data suggest that small organic molecules absorbed
to the coal, such as acetate, may have a marked effect on community
composition and the biogenic methane potential. However, the origin
of the acetate in the coal is difcult to determine. It has been suggested
that acetate found in CBM formation waters may result from microbial
breakdown of coal organic matter (Ulrich and Bower, 2008), or biotic
or abiotic ester hydrolysis in coals of lower rank (Glombitza et al.,
2009b; Vieth et al., 2008). Although the origin of the low molecular
weight hydrocarbons present in coals prior to use in microbial enrichment is difcult to determine, they represent an important carbon and
energy source for CBM microbial communities.
While proposed models of CBM production focus on the degradation
of aromatic and alkane hydrocarbons (Strpo et al., 2011), it has been
suggested that alternative substrates absorbed to the coal surface may
contribute to, or even enhance, CBM production (Fallgren et al., 2013;
Fry et al., 2009; Pfeiffer and Ulrich, 2014; Ulrich and Bower, 2008). For
example, Fallgren et al. (2013) found that subbituminous coals produced higher quantities of methane than lignite, but suggests that this
is likely a result of non-coal organic compounds trapped within the
coal matrix. Similarly, Furmann et al. (2013) have shown that both aromatic and alkane hydrocarbons extracted from the coal using various
organic solvents are used by CBM microbial communities. However,
these compounds are not necessarily hydrolyzed from the coal structure
and may represent desorbing organic matter. Consistent with prior
work on methane production from coal enrichments (Furmann et al.,
2013; Green et al., 2008; Papendick et al., 2011; Strpo et al., 2011),
the total methane yield in our enrichments accounted for less than
0.7% of total coal carbon based on a conservative estimate using lignite
(70% carbon) (Strpo et al., 2011). This suggests that the matured,
lignin-like coal structure itself may not always be the primary substrate
targeted by CBM microbial communities, but instead methane is produced from desorbing organics, such as acetate, trapped within the
coal matrix or easily hydrolysable esters.
It is likely that some of the acetate in the coal is directly converted
to methane by acetoclastic methanogens, consistent with the observed
shift from hydrogenotrophic methanogens in the inoculum to acetoclastic
methanogens in the coal enrichments (Fig. 4). However, the weak
correlation between acetate concentrations and nal methane yields,
indicates that the community is also using another organic substrate

derived from the coal. Five of the coals in this study, which span a range
of ranks (lignite to bituminous), produced more methane than could be
accounted for by full stoichiometric conversion of both acetate carbons
to methane (Table 3). However, this estimate is very conservative, as it
does not take into account carbon incorporated into biomass. A similar
nding was reported by Ulrich and Bower (2008) in full scale eld trials,
where acetate amendment increased methane recovery above the yield
expected from conversion of the acetate to methane. Taken together
with our ndings, we hypothesize that acetate can be used to cometabolize other substrates in coal. This is supported by the metabolic
properties of some Pelobacter spp., found at high abundance in our low
rank enrichments, that have previously been found to only metabolize
propanol or butanol in the presence of acetate (Schink, 2006).
5. Conclusions
In contrast to studies conducted by Fallgren et al. (2013) and
Wawrik et al. (2012), but consistent with the results of Strpo et al.
(2011), we demonstrate that a decrease in coal rank was associated
with an increase in biogenic methane potential. The decrease in rank
was correlated with an increase in acetate concentration, which suggests that the increase in methane production observed in low rank
coals is not only a function of the bioavailability of the predominantly
aromatic coal matrix, but also the availability of labile organic matter
absorbed to the coal. Further, we show that community composition
is inuenced by geographic location, coal rank, and the amount of acetate present within the coal. This nding suggests that different sets of
microoganisms may be coal specic. However, our understanding of
the metabolic interactions between CBM microorganisms and the coal
substrate is far from complete. A more detailed prole of the organic
composition of coal involving high-resolution mass spectrometry,
coupled to metagenomic and metatransciptomic analysis of active microbial communities will be critical in elucidating these interactions.
Acknowledgments
We thank S. Low for her help with sequencing library preparation,
the staff at Santos and Queensland Gas Company (QGC) for providing
access to CBM formation waters, and Dr. Ana Paula Oliveira from
TOTAL for critical review of the manuscript. We also thank Dr. Jane
Newman and Soleh Basuki Rahmat for providing us with the New
Zealand coals, Tasmanian Special Timbers for the Huon Pine, Ferian

212

S.J. Robbins et al. / International Journal of Coal Geology 154155 (2016) 205212

Angara for the Indonesian coals, and Anastasia Dmyterko for the
Australian coals. This study is supported by the Australian Research
Council (ARC) through a Linkage Project (LP100200730). S.R. is supported by an Australian Postgraduate Award Industry (APAI) scholarship. G.W.T. is supported by an ARC Queen Elizabeth II Fellowship
(DP1093175).
References
Altschul, S.F., Gish, W., Miller, W., Myers, E.W., Lipman, D.J., 1990. Basic local alignment
search tool. J. Mol. Biol. 215, 403410.
Anggara, F., Sasaki, K., Rodrigues, S., Sugai, Y., 2014. The effect of megascopic texture on
swelling of a low rank coal in supercritical carbon dioxide. Int. J. Coal Geol. 125,
4556.
Bolger, A.M., Lohse, M., Usadel, B., 2014. Trimmomatic: a exible trimmer for Illumina
sequence data. Bioinformatics btu170.
DeSantis, T.Z., Hugenholtz, P., Larsen, N., Rojas, M., Brodie, E.L., Keller, K., Huber, T., Dalevi,
D., Hu, P., Andersen, G.L., 2006. Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl. Environ. Microbiol. 72, 50695072.
Diessel, C.F., 1992. Coal Formation and Sequence Stratigraphy. Springer.
Dixon, P., 2003. VEGAN, a package of R functions for community ecology. J. Veg. Sci. 14,
927930.
Dmyterko, A., 2014. Variability of Coal Seam gas Reservoir Parameters in the Leichhardt
Seam, Rangal Coal Measures, Bowen Basin, Central Queensland, School of Earth
Science. University of Queensland, p. 72.
Eaton, A.D., Clesceri, L.S., Greenberg, A.E., 2005. Standard Methods for the Examination of
Water and Wastewater. American Public Health Association, Washington, DC,
pp. 2000123710.
Fallgren, P.H., Jin, S., Zeng, C., Ren, Z., Lu, A., Colberg, P.J.S., 2013. Comparison of coal rank
for enhanced biogenic natural gas production. Int. J. Coal Geol. 115, 9296.
Fry, J.C., Horseld, B., Sykes, R., Cragg, B.A., Heywood, C., Kim, G.T., Mangelsdorf, K.,
Mildenhall, D.C., Rinna, J., Vieth, A., Zink, K.-G., Sass, H., Weightman, A.J., Parkes, R.J.,
2009. Prokaryotic populations and activities in an interbedded coal deposit, including
a previously deeply buried section (1.62.3 km) above b150 Ma basement rock.
Geomicrobiol J. 26, 163178.
Furmann, A., Schimmelmann, A., Brassell, S.C., Mastalerz, M., Picardal, F., 2013. Chemical
compound classes supporting microbial methanogenesis in coal. Chem. Geol. 339,
226241.
Glombitza, C., Mangelsdorf, K., Horseld, B., 2009a. Maturation related changes in the distribution of ester bound fatty acids and alcohols in a coal series from the New Zealand Coal
Band covering diagenetic to catagenetic coalication levels. Org. Geochem. 40, 10631073.
Glombitza, C., Mangelsdorf, K., Horseld, B., 2009b. A novel procedure to detect low molecular
weight compounds released by alkaline ester cleavage from low maturity coals to
assess its feedstock potential for deep microbial life. Org. Geochem. 40, 175183.
Green, M.S., Flanegan, K.C., Gilcrease, P.C., 2008. Characterization of a methanogenic
consortium enriched from a coalbed methane well in the Powder River Basin,
U.S.A. Int. J. Coal Geol. 76, 3445.
Hamilton, S., Golding, S., Baublys, K., Esterle, J., 2015. Conceptual exploration targeting for
microbially enhanced coal bed methane (MECoM) in the Walloon Subgroup, eastern
Surat Basin, Australia. Int. J. Coal Geol. 138, 6882.
Jones, E.J.P., Voytek, M.A., Warwick, P.D., Corum, M.D., Cohn, A., Bunnell, J.E., Clark, A.C.,
Orem, W.H., 2008. Bioassay for estimating the biogenic methane-generating potential
of coal samples. Int. J. Coal Geol. 76, 138150.

Kunin, V., Engelbrektson, A., Ochman, H., Hugenholtz, P., 2010. Wrinkles in the rare
biosphere: pyrosequencing errors can lead to articial ination of diversity estimates.
Environ. Microbiol. 12, 118123.
Legendre, P., Gallagher, E., 2001. Ecologically meaningful transformations for ordination
of species data. Oecologia 129, 271280.
Neuwirth, E., 2011. RColorBrewer: ColorBrewer Palettes.
O'Keefe, J.M.K., Bechtel, A., Christanis, K., Dai, S., DiMichele, W.A., Eble, C.F., Esterle, J.S.,
Mastalerz, M., Raymond, A.L., Valentim, B.V., Wagner, N.J., Ward, C.R., Hower, J.C.,
2013. On the fundamental difference between coal rank and coal type. Int. J. Coal
Geol. 118, 5887.
Orem, W., Tatu, C., Varonka, M., Lerch, H., Bates, A., Engle, M., Crosby, L., McIntosh, J., 2014.
Organic substances in produced and formation water from unconventional natural
gas extraction in coal and shale. Int. J. Coal Geol. 126, 2031.
Papendick, S.L., Downs, K.R., Vo, K.D., Hamilton, S.K., Dawson, G.K.W., Golding, S.D.,
Gilcrease, P.C., 2011. Biogenic methane potential for Surat Basin, Queensland coal
seams. Int. J. Coal Geol. 88, 123134.
Penner, T.J., Foght, J.M., Budwill, K., 2010. Microbial diversity of western Canadian subsurface coal beds and methanogenic coal enrichment cultures. Int. J. Coal Geol. 82,
8193.
Pfeiffer, R.S., Ulrich, G.A., 2014. Chemical Amendments for the Stimulation of Biogenic gas
Generation in Deposits of Carbonaceous Material. Google Patents.
R_Core_team, 2013. A Language and Environment for Statistical Computing. 3.1.2 ed.
Rahmat, S.B.,. School of Earth Science. University of Queensland (in progress).
Schink, B., 2006. The genus pelobacter. In: Dworkin, M., Falkow, S., Rosenberg, E.,
Schleifer, K.-H., Stackebrandt, E. (Eds.), The Prokaryotes. Springer, New York,
pp. 511.
Stout, S.A., Boon, J.J., Spackman, W., 1988. Molecular aspects of the peatication and early
coalication of angiosperm and gymnosperm woods. Geochim. Cosmochim. Acta 52,
405414.
Strpo, D., Mastalerz, M., Dawson, K., Macalady, J., Callaghan, A.V., Wawrik, B., Turich, C.,
Ashby, M., 2011. Biogeochemistry of microbial coal-bed methane. Annu. Rev. Earth
Planet. Sci. 39, 617656.
Surez-Ruiz, I., Flores, D., Mendona Filho, J.G., Hackley, P.C., 2012. Review and update of
the applications of organic petrology: part 1, geological applications. Int. J. Coal Geol.
99, 54112.
Susilawati, R., Evans, P.N., Esterle, J.S., Robbins, S.J., Tyson, G.W., Golding, S.D., Mares, T.E.,
2014. Temporal changes in microbial community composition during culture enrichment experiments with Indonesian Coals. Int. J. Coal Geol.
Tanner, R.S., 2007. Cultivation of bacteria and fungi. In: Hurst, C., Crawford, R., Garland, J.,
Lipson, D., Mills, A., Stetzenbach, L. (Eds.), Manual of Environmental Microbiology,
pp. 6978.
Taylor, G., 1998. Organic Petrology: A New Handbook Incorporating Some Revised Parts
of Stach's Textbook of Coal Petrology. Gebruder Borntraeger Verlagsbuchhandlung.
Ulrich, G., Bower, S., 2008. Active methanogenesis and acetate utilization in Powder River
Basin coals, United States. Int. J. Coal Geol. 76, 2533.
Vieth, A., Mangelsdorf, K., Sykes, R., Horseld, B., 2008. Water extraction of coalspotential
for estimating low molecular weight organic acids as carbon feedstock for the deep
terrestrial biosphere. Org. Geochem. 39, 985991.
Ward, C.R., 1984. Coal Geology and Coal Technology.
Warnes, G.R., Bolker, B., Bonebakker, L., Gentleman, R., Huber, W., Liaw, A., Lumley, T.,
Maechler, M., Magnusson, A., Moeller, S., Schwartz, M., Venables, B., 2014. gplots:
Various R Programming Tools for Plotting Data.
Wawrik, B., Mendivelso, M., Parisi, V.A., Suita, J.M., Davidova, I.A., Marks, C.R., Van
Nostrand, J.D., Liang, Y., Zhou, J., Huizinga, B.J., Strapoc, D., Callaghan, A.V., 2012.
Field and laboratory studies on the bioconversion of coal to methane in the San
Juan Basin. FEMS Microbiol. Ecol. 81, 2642.